Treatise On Geomorphology - John F. Shroder (AP, 2013)

2.
1 Quantitative Modeling of Geomorphology

ACW Baas, Kings College London, London, UK
r 2013 Elsevier Inc. All rights reserved.
2.1.1
Introduction
2.1.2
Structure of this Volume
Acknowledgments
References
1
3
4
5
Abstract
This introductory chapter of the volume on Quantitative Modeling of Geomorphology considers different types of models
in geomorphology as recognized in past research and review literature and the various ways of classifying and characterizing
different kinds of modeling approaches and philosophies. This is followed by reflections on one particular aspect that is not
generally considered in the context of geomorphic models: questions of beauty and aesthetics, what these might be, and
whether they may be fruitful to consider in this realm. The remainder of the introduction explains the general development
approach and objectives of the volume, describes the volume structure and logic of the sequence of chapters, and concludes
with important acknowledgements.
See from the first your proportions, and do not lose them. But
proportions according to correct measurement are after all but very
little unless confirmed by sentiment, and expressive of the particular physical character of the model.
Henri Matisse, First Instruction for Sculpture
2.1.1
Introduction
A model can be a many great things: a concept, a hypothesis,

a theory, a law, a paradigm, a description, an analogy, an
idealization, an equation, a calculation, an algorithm, a
dataset, or a map, among many other things. In its broadest
sense a model can be any representation of external reality,
and thus arguably all of scientific understanding is nothing
but a model. An abundance of argument and reflection exists
on the philosophy and practice of modeling by wiser men and
women, and the brief introduction here to this volume on
Quantitative Modeling of Geomorphology will not attempt
to review this in any comprehensive way. It can, however, be
fruitful for practicing modelers to reflect on the variety of
attributes, styles, types, and characteristics of modeling that
have been recognized in the geomorphological tradition.
Commonly a primary distinction is made between: (1) conceptual models, (2) scale models or analogs, and (3) mathematical models (e.g., Huggett, 1985).
Conceptual models are abstract representations rendered
in a verbal or visual form, ranging from a sequence of relationship statements or story of a geomorphic system, to flow
diagrams and schematics, commonly in a graphical or cartoon
form. Many historically (in)famous geomorphic models, and
Baas, A.C.W., 2013. Quantitative modeling of geomorphology. In: Shroder,
J. (Editor in Chief), Baas, A.C.W. (Ed.), Treatise on Geomorphology.
Academic Press, San Diego, CA, vol. 2, Quantitative Modeling of
Geomorphology, pp. 15.
Treatise on Geomorphology, Volume 2
indeed paradigms, are of this kind, such as Daviss (1899)

Geographical Cycle, characteristically represented as a sequence of landscape cartoons, or the general systems framework championed by Chorley (1962) and Strahler (1980),
generally drawn as detailed and elaborate flow schemes
resembling electronic circuitry diagrams.
Scale models and analogs aim to represent a geomorphic
system on a (generally reduced) physical or temporal scale
so as to replicate or mimic the processes and forms in an
observable and reproducible form. This includes hardware
models or physical analogs such as water flumes (e.g., Peakall
et al., 1996), wind tunnels (e.g., White, 1996), and wave tanks,
where processes or properties are scaled in terms of their
geometry (sizes), kinematics (velocities), or dynamics (forces),
relative to the real-world system. Natural analogs may also
be identified as models: observable real-world systems that
exhibit a behavior or characteristic that is similar in kind to
(and thereby an explanation of) another geomorphic system.
Chorley (1967) in his taxonomy of geomorphic models
recognized in this category both historical analogs (history
repeats itself, as in the case of past landscaping events) as well
as spatial analogs (like goes with like, as in disparate terrain
units that yet share a common aspect).
Mathematical models, lastly, quantify features, objects, and
processes of a geomorphic system numerically as variables,
parameters, and coefficients, representing the relationships
between them as a collection of mathematical expressions.
Such models range from single equations, like the Universal
Soil Loss Equation (USLE), to coded algorithms of varying
levels of complexity running on computers. Mathematical
models are generally developed out of initially conceptual or
scale models. It is these types of geomorphic models that are
the primary focus of this volume here.
The distinction of the three modeling types above is tied to
their methodology or the means by which the representation
of a geomorphology is achieved. Chorleys (1967) model
http://dx.doi.org/10.1016/B978-0-12-374739-6.00023-3
Quantitative Modeling of Geomorphology
taxonomy, however, was instead developed within a general

systems theory framework, recognizing the act of modeling
as a function of three distinct purposes of scientific activity:
(1) translation, leading to natural analogs mentioned previously; (2) dissection, including hardware models, mathematical models, and experiments; and (3) conception, as
synthetic systems, partial systems, or black box modeling.
Of course there is no one single authoritative model
typology and many different versions of classification have
developed over the years, both past (e.g., those by Chorley,
1967; Huggett, 1985, mentioned above) as well as more recent
(Hooke, 2003; Mulligan and Wainwright, 2004). The main
interest of considering taxonomies like these is, rather, for the
practicing modeler to reflect on their own model, to compare
and contrast with other types and to situate it within a
classification scheme as a means for discovering potential
cross-links, limitations, and possibilities. Another way of
exploring and confronting geomorphic models is to consider
their characteristics along dimensions of dichotomy (e.g.,
Haggett and Chorley, 1967; Slingerland and Kump, 2011),
including:
static
versus dynamic: models that represent equilibrium

features or structures as opposed to models that focus on
processes and changes over time,
descriptive versus normative: models that are concerned
with a stylistic or simplified description of reality as opposed
to models that attempt to predict outcomes under certain
conditions,
stochastic versus deterministic: models that represent aspects
of a geomorphic system in terms of probabilities and statistics as opposed to models that quantify specific precise
values,
process-based versus form-based: models that simulate
physical processes that are thought to operate in reality as
opposed to models that represent shapes and forms occurring in the geomorphic landscape,
forward versus inverse: models that attempt to predict a final
or future state of a geomorphic system as opposed to models
that aim to determine past initial or boundary conditions
based on a current state,
black box versus white box: models that use invisible or
untraceable internal procedures (commonly involving advanced statistical methods) as opposed to models where all
processes and relationships are transparent and precisely
specified,
inductive versus deductive: models that represent a system
by generalizing and categorizing from empirical findings as
opposed to models that assume a theory or framework to
simulate consequent forms and processes.
All of the terms above are clearly subject to semantics and
philosophical debate, but the principal purpose of reflecting
on these dichotomies is to elicit alternative and novel ways of
thinking about ones own modeling approach and context
that can help spur new ideas, opportunities, and extensions.
Certain combinations of dichotomies may occur more naturally, for example, deductive deterministic versus inductive
stochastic, or the similarity between static and form-based
versus dynamic and process-based, but contemplating other
combinations may identify modeling approaches that are
relatively rare: for example, stochastic, form-based dynamic

models (akin perhaps to allometry: Bull, 1975; Maritan et al.,
2002).
A different means for reflecting on geomorphic modeling is
the question of what makes for a good model, a subject more
extensively discussed in Chapter 2.5 of this volume. Kirkby
(1996), for example, argues very precise criteria for a good
model, stipulating four requisite attributes: (1) a physical
basis; (2) simplicity; (3) generality and richness; and (4) potential for scaling up and down. The question, however, relies
on the meaning of good, which in a modeling context may
involve various aspects, ranging from practical utility, or the
delivery of results that are useful and relevant to end-users,
through representation, or the synthesis and integration of
concepts and understanding that help to consolidate knowledge, to inspiration, or the discovery of links and insights that
lead to new research questions. A stochastic, black-box, river
flow discharge model, for example, may not offer much inspiration for novel research questions, nor a good representation of scientific understanding (and it would not meet
Kirkbys criteria either), but if it can predict a 50-year flood
very accurately it is undoubtedly of good value to a water
resource manager. Reflections on utility and representation
are generally extensively considered in the context of model
validity (Oreskes et al., 1994) and prediction (Wilcock and
Iverson, 2003), but the opportunities for inspiration are generally only implicitly articulated in the literature. Some geomorphic models, such as reduced complexity models or
cellular automata, however, can have a great capacity for inspiring new insights and research questions, even though they
may not be able to produce very accurate predictions for
practical end-users, nor represent particularly comprehensive
integrations of detailed process-based understanding.
Just as merit and worth are ultimately a subjective judgment, we may ask whether we can consider the beauty or
aesthetics of a geomorphological model, something that the
opening quote of this introduction tries to allude to. Although
scientists generally shy away from artistic reflection, in the
field of mathematics and physics there are some equations, or
models, that are widely regarded as beautiful (Rota, 1997),
such as the Maxwell equations, Eulers identity, eip 1,
admired by the likes of Richard Feynman (Feynman et al.,
1964), or indeed the quantum theory diagrams (or graphs)
that bear his name. These models may be considered beautiful
because they combine, synthesize, and abstract a large and
complex body of knowledge into a pinnacle of simplification,
built on extensive foundations from various (commonly disconnected) areas. Eulers formula, for instance, unfolds into an
understanding of exponential functions, trigonometry, and
complex numbers, and garners insights and connections
between these domains. The idea of elegance in simplicity
is of course found in Kirkbys second criterion, mentioned
previously, and is equivalent to the general principle of
Ockhams Razor, a philosophical underpinning that recurs in
Chapter 2.2. It is, however, when simplicity is found in the
integration and combination of a large number of concepts,
theories, and observations, a model becomes most elegant and
beautiful, or equivalently: efficient, powerful, or rich in content. The conception of beauty and elegance in terms of information content and richness finds similarities in practices
like computer programming contests, where the goal is to

write a piece of code that achieves some application while
remaining below a very small size limit (e.g., less than 1024
bytes), and has also been explored on a quantitative basis by
Schmidhuber (1997) using algorithmic information theory. In
geomorphic modeling there may be less scope for beauty,
because it rarely involves purely mathematical equations that
resemble the power and simplicity of an Eulers formula.
Perhaps the Exner equation (Chapter 2.3, eqn [1]) comes
close, as it efficiently and powerfully relates surface expression
and sediment transport in time and space. It does not, however, particularly integrate or connect disparate areas of research, but rather simply provides a foundation of massconservation to the numerical modeling of Earth surface
processes and landforms. Concepts of self-organization and
associated cellular-automaton models may also qualify as
beautiful, as they can transcend traditional subdisciplinary
boundaries and represent a variety of geomorphic systems in a
simplified and elegant way. For the kind of extensive, complicated (rather than complex), grid-based landscape evolution models like those reviewed in Chapter 2.13 perhaps the
quest for beauty will lie in the formulation of the shortest,
most efficient, algorithmic description (or a low Kolmogorov
complexity), and in particular in balancing the representations and scales of detail of the various modeled processes
and components. Some choices and decisions involved in
geomorphic modeling can be subjective, and so a general
striving for beauty and elegance may serve an inherent
purpose.
2.1.2
Structure of this Volume
The theme of this volume is on quantitative modeling of

geomorphology. The term quantitative, rather than mathematical or numerical, has been used because it encompasses
any modeling approach that involves the quantification of
processes or forms, as in assigning magnitudes that may be
variable over time or space. The term mathematical may seem
too much confined to models that involve only explicit (or
even analytical) equations of some kind to represent geomorphology, thereby somewhat excluding cellular automata
or purely algorithmic approaches, whereas the term numerical has too strong a connotation nowadays with digital
computers and computational methods for calculation and
solution (e.g., finite element methods for fluid flow problems), which often form part of a geomorphic modeling
approach but do not necessarily represent their defining
character.
Instead of collating individual reviews of modeling
literature in each of the traditional subdisciplines of geomorphology for example, modeling in glacial, coastal,
aeolian, fluvial geomorphology, and so forth this volume
of the Major Reference Work takes a more eclectic approach to
bring together a perhaps more inventive combination of reviews and discussions that are relevant across the whole discipline as a complementary resource for the reader that
may not be easily found elsewhere, particularly as many of
the companion volumes in this Major Reference Work also
include their own separate chapters on modeling in their respective field (e.g., Chapters 3.6, 3.9, 7.10, 9.34, and 11.16).
This volume is structured along three sections: I
Fundamental Aspects, II Innovative Methods, and III
Geomorphic Modeling from Soil to Landscape. The first four
chapters review philosophical issues, fundamental principles,
and technical aspects of quantitative modeling in geomorphology. The first chapter, by Malamud and Baas on Nine
Considerations for Constructing and Running Geomorphological Models, reviews practical as well as conceptual challenges and questions that commonly arise when developing
or applying quantitative geomorphic models, including
such issues as model parsimony, benchmarking, calibration,
and sensitivity analysis. Some of these topics have been discussed at length in the broader geosciences literature, for example, the issue of calibration and validation, whereas others
are only found in the margins of physical sciences textbooks
or in highly specialized papers, for example, dimensional
analysis and model data exploration. The second chapter, by
Pelletier on Fundamental Principles and Techniques of
Landscape Evolution Modeling, provides a thorough overview
of the basic equations and mathematical techniques that are
used in quantifying and simulating the redistribution of
sediment around a landscape and the resultant surface evolution. It includes a detailed explanation to the key attributes
of diffusive and advective equations and the methods and
assumptions used for simplifying and solving them. The third
chapter, by Slingerland and Syvitski on A Community Approach to Modeling Earth- and Seascapes, introduces the innovative and cutting-edge development of community
modeling, harnessing and integrating the collective efforts of
geomorphologists across many subdisciplines toward a comprehensive and transparent simulation framework. This new
approach takes advantage of the latest IT solutions and
infrastructure to develop new model platforms that would
be unattainable by individual researchers alone. The fourth
chapter in this section, by Murray on Which Models Are Good
(Enough), and When?, contemplates some of the philosophical and theoretical concerns that are peculiar to geomorphological modeling, grappling with the question of
how to judge a model, as well as with the issues of scale,
reductionism, and emergence.
The four chapters in the second section review specific
methodological approaches and techniques that are currently
at the forefront of innovative modeling in geomorphology
and are seldom part of the standard repertoire, including such
topics as artificial neural networks, evolutionary modeling,
fractional derivatives, and cellular automata. The first chapter
in this section, by Hjort and Luoto on Statistical Methods
for Geomorphic Distribution Modeling provides an entry into
the world of geomorphic distribution modeling and the statistical methods that are used to relate spatially distributed
data, for example, a geomorphic surface, to multiple and
commonly complex environmental foring variables. This
chapter includes detailed explanations of generalized linear
models and artificial neural network methods that have
recently gained significant interest. The second chapter, by
DAmbrosio, Spataro, Rongo, and Iovine on Genetic Algorithms, Optimization, and Evolutionary Modeling, is closely
related to the previous one and focuses specifically on
automated computational methods that mimic Darwinian

natural selection to evolve toward an optimized model of a
geomorphic system. Such genetic algorithms can also be
applied to find a best combination of parameter values
for calibration purposes. The third chapter, by FoufoulaGeorgiou and Passalacqua on Nonlocal Transport Theories in
Geomorphology: Mathematical Modeling of Broad Scales of
Motion, builds on the basic principles of diffusive equations
that describe sediment transport and surface evolution (reviewed by Pelletier in the first section of this volume), by
expanding the fundamental mathematical approach to include fractional derivatives in the equations, that is, derivatives
that are not integer-order (first, second, etc.) but fractions.
Such fractional derivatives broaden the scope of diffusive
sediment transport modeling from purely local quantifications at individual points to a more integrated inclusion of
the surrounding surface conditions, now explicitly contained
within the mathematical approach. The fourth chapter, by
Fonstad on Cellular Automata in Geomorphology, provides
an extensive overview of perhaps a more widely recognized
alternative modeling approach in geomorphology. Cellular
automata have been developed across a wide range of geomorphic applications and environments and have proved the
most accessible testing ground for exploring emergent phenomena in Earth surface systems, raising some of the intriguing philosophical questions that Murray grappled with in
his chapter.
The third section of this volume presents a more topical
collection of chapters that focus on the traditional core of
geomorphological modeling, following a sequence of scale
and sediment movement, starting with the erosion of soil on
hill slopes, followed by the transport of the sediment in water
and air, to the consequences of this redistribution in terms of
floodplain development on a valley-scale to landscape evolution on catchment and continental scale, and concluding
with the important influence of biology at all these levels of
geomorphic systems. The first chapter, by Brazier on Hillslope
Soil Erosion Modeling, presents a comprehensive overview of
the many models and algorithms that have been developed
over the past decades for quantifying soil erosion on slopes, of
varying degrees of complexity and detail. The review furthermore typifies and compares the temporal and spatial scales of
operation of a host of erosion models. The second chapter, by
Hardy on Process-Based Sediment Transport Modeling, discusses the numerical simulation of the transport of eroded
sediments by fluids, principally water in open channel flow
rivers and wind. The principles of mass and momentum
equations for sediment particles are reviewed together with
different methods of their representation in 1-D, 2-D, and 3-D
modeling approaches. The third chapter, by Nicholas on
Morphodynamic Modeling of Rivers and Floodplains, looks
at modeling sediment transport processes and their geomorphic consequences at the larger scale of the river reach and
flood plain. It covers such topics as simulating the evolution
and migration of river meander bends, braided river channel
patterns, and the modeling of longer term longitudinal profile
development and floodplain sedimentation. The fourth
chapter, by Temme, Schoorl, Claessens, and Veldkamp on
Quantitative Modeling of Landscape Evolution, takes the review to the largest spatial and temporal scales, that of
modeling topographic development of whole river basins and

mountain ranges, as well as considering glaciers, river deltas,
and aeolian dune landscapes. The chapter also grapples
with the fundamental issues of uncertainty analysis and validation attempts for geomorphic modeling on these scales.
The fifth chapter, by Saco and Rodriguez on Modeling
Ecogeomorphic Systems, concludes this section (and the volume) on a very important topic that is increasingly recognized
as such across the whole of the geomorphological research
community. It is becoming clear that biological components,
flora and fauna and their combined ecology, can have a far
greater impact on the dynamics of Earth surface systems than
previously thought, from their impact on river channel patterns and catchment evolution, to tidal wetland development
and aeolian dune fields. This chapter provides an extensive
and original review of the recent progress and strategies in
modeling the great variety of ecogeomorphic systems.
The above collection of chapters hopefully offers the reader
a novel approach to considering the principles and practices of
quantitative modeling of geomorphology. More traditional
coverage of modeling in geomorphology can be found in a
number of established books and edited volumes: Chorley
and Haggetts (1967) Models in Geography is a somewhat
older text, but it contains still highly relevant discussions on
fundamental principles and philosophical issues. Kirkby et al.
(1992) in their Computer Simulation in Physical Geography
presented an extensive overview of all the technical and
practical aspects of numerical modeling. Rhoads and Thorns
(1996) Scientific Nature of Geomorphology, proceedings of a
Binghamton symposium (also available online), includes an
important section on modeling prospects and problems.
Anderson and Bates (2001) Model Validation: Perspectives in
Hydrological Science provides a thorough overview on the
issues and practices of testing and evaluating models, in this
case specifically related to hydrology. Wilcock and Iversons
(2003) edited volume on Prediction in Geomorphology is a
rich collection of chapters that discuss the application, limits,
and testing of various types of models in Earth sciences.
Wainwright and Mulligans (2004) Environmental Modeling
presents an extensive collection of writings on state of the art
modeling of very specific environmental systems and phenomena, including plants and animals, as well as modeling
for end-users. Finally, two recent textbooks focus specifically
on mathematical aspects and their technical implementation
in geomorphic modeling: Pelletiers (2008) Quantitative
Modeling of Earth Surface Processes, and Slingerland and
Kumps (2011) Mathematical Modeling of Earths Dynamical
Systems.
Although the literature cited above presents extensive
background and further reading, this volume on Quantitative
Modeling of Geomorphology may contribute a valuable and
complementary resource for anyone interested in this type of
research.
Acknowledgments
The quality of this volume as a comprehensive resource for
students, researchers, and other users who want to learn more
about quantitative modeling of geomorphology has been
greatly improved by the valuable contributions of a large

number of peer-reviewers providing many constructive comments and suggestions to the authors. The authors would like
to thank the following colleagues for their generous help:
Alexander Brenning
Andrea DAlpaos
Sergio Fagherazzi
Dennis Flanagan
Stefan Hergarten
Erkan Istanbulluoglu
James Millington
Bruce Rhoads
Rina Schumer
Greg Tucker
Alexey Voinov
Marco van der Wiel
Emily Zechman
References
Anderson, M.G., Bates, P.D. (Eds.), 2001. Model Validation: Perspectives in
Hydrological Science. Wiley, Chichester, 500 pp.
Bull, W.B., 1975. Allometric change of landforms. Geological Society of America
Bulletin 86(11), 14891498.
Chorley, R.J., 1962. Geomorphology and general systems theory. US Geological
Survey Professional Paper, 500-B: B1-B10.
Chorley, R.J., 1967. Models in geomorphology. In: Chorley, R.J., Haggett, P. (Eds.),
Models in Geography. Methuen, London, pp. 5996.
Chorley, R.J., Haggett, P. (Eds.), 1967. Models in Geography. Methuen & Co Ltd,
Worcester, 801 pp.
Davis, W.M., 1899. The geographical cycle. Geographical Journal 14(5), 481504.
Feynman, R.P., Leighton, R.B., Sands, M., 1964. The Feynman Lectures on Physics,
vol. 1. Addison-Wesley Publishing Company, Reading, MA.
Haggett, P., Chorley, R.J., 1967. Models, paradigms, and the new geography.
In: Chorley, R.J., Haggett, P. (Eds.), Models in Geography. Methuen, London,
pp. 1941.
Hooke, R.L., 2003. Predictive modelling in geomorphology: an oxymoron?

In: Wilcock, P.R., Iverson, R.M. (Eds.), Prediction in Geomorphology. American
Geophysical Union, Washington, pp. 5161.
Huggett, R.J., 1985. Earth Surface Systems. Springer-Verlag, Heidelberg, 270 pp.
Kirkby, M.J., 1996. A role for theoretical models in geomorphology? In: Rhoads,
B.L., Thorn, C.E. (Eds.), The Scientific Nature of Geomorphology: Proceedings
of the 27th Binghamton Symposium in Geomorphology. Wiley, Chichester,
pp. 257272.
Kirkby, M.J., Naden, P.S., Burt, T.P., Butcher, D.P., 1992. Computer Simulation in
Physical Geography. Wiley, Chichester, 180 pp.
Maritan, A., Rigon, R., Banavar, J.R., Rinaldo, A., 2002. Network allometry.
Geophysical Research Letters 29, 11.
Mulligan, M., Wainwright, J., 2004. Modelling and model building. In: Wainwright,
J., Mulligan, M. (Eds.), Environmental Modelling. Wiley, Chichester, pp. 773.
Oreskes, N., Shraderfrechette, K., Belitz, K., 1994. Verification, validation, and
confirmation of numerical models in the earth-sciences. Science 263(5147),
641646.
Peakall, J., Ashworth, P.J., Best, J., 1996. Physical modelling in fluvial
geomorphology: principles, applications, and unresolved issues. In: Rhoads,
B.L., Thorn, C.E. (Eds.), The Scientific Nature of Geomorphology. Wiley,
Chichester, pp. 221253.
Pelletier, J., 2008. Quantitative Modeling of Earth Surface Processes. Cambridge
University Press, Cambridge, 295 pp.
Rhoads, B.L., Thorn, C.E. (Eds.), 1996. The Scientific Nature of Geomorphology.
Wiley, Chichester, 481 pp.
Rota, G.C., 1997. The phenomenology of mathematical beauty. Synthese 111(2),
171182.
Schmidhuber, J., 1997. Low-complexity art. Leonardo 30(2), 97103.
Slingerland, R.L., Kump, L., 2011. Mathematical Modelling of Earths Dynamical
Systems. Princeton University Press, Princeton, 231 pp.
Strahler, A.N., 1980. Systems theory in physical geography. Physical Geography
1(1), 127.
Wainwright, J., Mulligan, M. (Eds.), 2004. Environmental Modelling. Wiley,
Chichester, 408 pp.
White, B.R., 1996. Laboratory simulation of aeolian sand transport and physical
modeling of flow around dunes. Annals of the Arid Zone 35(3), 187213.
Wilcock, P.R., Iverson, R.M. (Eds.), 2003. Prediction in Geomorphology. American
Geophysical Union, Washington, 256 pp.
Biographical Sketch
Andreas Baas is a Senior Lecturer at the Department of Geography at Kings College London. He is an editor for
Nonlinear Processes in Geophysics, an editorial board member for Geomorphology, and an active session convener for the Geomorphology Division of the European Geosciences Union (EGU). Andreas research is supported by The Leverhulme Trust, The Nuffield Foundation, and the UK Natural Environment Research Council,
and spans all aspects of aeolian geomorphology across a range of spatiotemporal scales, including the quantitative
modeling of dunes and dune fields, particularly simulating the effects of vegetation on the development of dune
landscapes in coastal environments.
3.1 Remote Sensing and GIScience in Geomorphology: Introduction

and Overview
MP Bishop, Texas A&M University, College Station, TX, USA
3.1.1
3.1.2
3.1.3
3.1.3.1
3.1.3.2
3.1.3.2.1
3.1.3.2.2
3.1.3.3
3.1.3.4
3.1.3.5
3.1.3.6
3.1.4
3.1.4.1
3.1.4.2
3.1.4.3
3.1.4.4
3.1.4.5
3.1.4.6
3.1.4.7
3.1.5
References
Introduction
Geospatial Technology and Fieldwork
Remote Sensing and Geomorphology
Photography and Videography
Imaging Spectroscopy
Sensor parameters
Reflectance properties and applications
Microwave Remote Sensing
The Atmosphere and Climate Forcing
Land-Cover Assessment and Mapping
Near-Surface Geophysics
GIS and Geomorphology
Digital Terrain Modeling (DTM)
Terrain Analysis
Landform Mapping
Spatial Hydrology
Erosion Modeling
Natural Hazards
Visualization
Conclusions
Glossary
Digital elevation model A digital elevation model (DEM)
is generally a land-surface model that attempts to accurately
portray the altitude field of the topography. In
geomorphology, it commonly takes the form of a
raster data layer representing a field of square
tessellations. The grid cell resolution is based upon
the source data and the desired scale for representing the
topography.
Digital terrain modeling Digital terrain modeling (DTM)
refers to a workflow process of acquiring data that samples
the altitude field, preprocessing the data to generate a digital
elevation model, and error and uncertainty analysis to
identify and remove systematic and random errors.
Geographic information science Geographic
information science (GIScience) is an emerging
multidisciplinary field that attempts to understand the
nature of spatio-temporal information, with a focus on
geographic representation, spatial analysis and modeling,
and addressing scientific problems. It represents a body of
geographical and technical knowledge concerned with
3
3
5
5
6
6
7
10
10
12
12
12
13
13
14
15
17
18
19
20
21
philosophical, cognitive, and scientific treatments of spatiotemporal theory and concepts. GIScience is often used as an
umbrella term to refer to developments in a variety of fields
that have contributed theory, concepts, and new
information technology. It also refers to those disciplines
involved in the investigation of the utility of information
technology for applied information production and
problem solving.
Geographic information systems A geographic
information system (GIS) is a software system that can be
used to store, manage, manipulate, analyze, and display
spatially referenced data. It can also be thought of as a
decision support system, as spatial data are analyzed to
produce information that is used to support management
and planning decisions.
Geomorphological mapping Geomorphological
mapping is a general term that refers to mapping various
aspects of the geomorphological system. This can include
numerous themes such as land systems, hydrology, surface
material and structure, morphometry, sediment transfer,
surface-process regimes, and chronology.
Bishop, M.P., 2013. Remote sensing and GIScience in geomorphology:

introduction and overview. In: Shroder, J. (Editor in Chief), Bishop, M.P.
(Ed.), Treatise on Geomorphology. Academic Press, San Diego, CA, vol. 3,
Remote Sensing and GIScience in Geomorphology, pp. 124.
http://dx.doi.org/10.1016/B978-0-12-374739-6.00040-3
Remote Sensing and GIScience in Geomorphology: Introduction and Overview
Geomorphometry The discipline that is concerned with

the science of quantitative land-surface characterization.
Imaging spectroscopy Imaging spectroscopy refers to the
collection of surface spectra from a spectrally contiguous
region of the electromagnetic spectrum that includes the
visible, near-infrared, and short-wave infrared regions of the
electromagnetic spectrum. The spectra are collected using a
very narrow spectral interval (812 nm) and hundreds of
images are usually acquired. This allows detailed
biophysical assessment and mapping of the Earth surface.
Imaging spectroscopy is also commonly called imaging
spectrometry or hyperspectral remote sensing, as the
dimensionality of spectral data (images) is very high. The
accuracy of the surface spectra for each pixel is dependent
on the spatial, spectral, and radiometric characteristics of
the sensor.
Land-surface objects Land-surface objects are spatial
entities that represent a meaningful segmentation of the
Earths surface. They are generated from land-surface
parameters using a variety of algorithms and methods, and
generally relate to the morphology of the topography in
terms of landform elements, features, or functional units.
Terrain segmentation can also relate to topographic
position and structure, surface material, or process
domains, if properly defined and delineated.
Land-surface parameters Land-surface parameters are
also called geomorphometric parameters, and they attempt
to quantitatively characterize various aspects of the
topography. They can be defined and classified based on
geometry, scale, and by surface-process characterization.
They are used to generate land-surface objects and
characterize process mechanics in surface-process modeling.
A variety of parameters such as slope, slope azimuth,
curvature, surface roughness, and relief are used for
studying geomorphological systems and for
geomorphological mapping.
Light detection and ranging Light detection and ranging
(LiDAR) is a form of active remote sensing that is based on
laser-light technology to obtain information about the
Earth surface. LiDAR data can be used to produce highresolution digital elevation models that attempt to
characterize a surface or a bare-earth representation of the
topography.
Near infrared The near-infrared (NIR) represents a region
of the electromagnetic spectrum that ranges from 0.7 to
1.3 mm. Optical sensors record NIR light reflected from the
surface. NIR imagery can be used to assess and map
vegetation, water, and other environmental characteristics.
Object-oriented analysis Object-oriented analysis refers

to the analysis of surface objects to determine their inherent
properties. Terrain or image segmentation is first required to
group tessellations into meaningful spatial entities that
represent land-cover or terrain features. Spatial analysis is
required to isolate surface objects at a variety of scales.
Object-oriented analysis is then used to characterize object
properties such as size, shape, and topographic conditions.
Spatial analysis of objects can then provide detailed
information about the context and topological
relationships among surface objects. This analysis has been
found to be superior to pixel-based methods for mapping a
variety of Earth-surface features.
Short-wave infrared The short-wave infrared (SWIR)
represents a region of the electromagnetic spectrum that
ranges from 1.3 to 3.0 mm. Optical sensors record SWIR
light reflected from the surface. SWIR imagery is routinely
used for snow and ice, mineralogical, and lithological
mapping.
Spatial hydrology Spatial hydrology refers to the linking
of remote sensing and GIS with hydrological modeling
efforts. Numerous levels of coupling of GIScience and
hydrology exist, although a major focus has been on using
various forms of data and methods to estimate key
hydrological parameters that drive hydrological models.
Spectral absorption feature A spectral absorption feature
is a wavelength-dependent feature of a spectral reflectance
curve. It is characterized by a localized decrease in reflectance
due to the absorption of energy by surface matter, given its
composition and chemistry. The prominence of the
absorption feature is determined by its depth, width, and
shape. A spectral curve can exhibit numerous spectral
absorption features depending on sensor-system
characteristics and surface compositional variations. Many
surface materials such as biochemicals or minerals have
diagnostic absorption features that can be used to assess
surface biophysical characteristics. Imaging spectroscopy
makes use of spectral absorption features to assess and map
biophysical conditions on the Earths surface.
Thermal infrared The thermal infrared (TIR) represents a
region of the electromagnetic spectrum that effectively ranges
from 3.0 to 15 mm. This region of the spectrum is used to
assess the thermal properties of the landscape and emission is
related to surface temperature. Thermal imagery is used in a
variety of disciplines including geomorphology, volcanology,
glaciology, and for mineralogical/lithological mapping.
Abstract
Geospatial technologies are having a profound effect on geomorphology. Remote sensing and geographic information
system studies are now commonplace in Earth science investigations. Significant advances have occurred in sensors,
geodesy, photogrammetry, geophysics, computer science, statistics, and pattern recognition. Consequently, it is now possible to quantify landscape morphology, investigate climate forcing, link process with patterns and form, and enhance our
understanding of scale dependence and the polygenetic nature of landscape evolution. This chapter introduces current
capabilities and new developments that are relevant to geomorphological investigations. The emphasis is on using data and
new analysis approaches to better understand geomorphological systems and landforms.
3.1.1
Introduction
Earth science investigations are increasingly utilizing geospatial

technologies (Bishop and Shroder, 2004b; Hengl and Reuter,
2009; Bishop et al., 2012). The rapid proliferation of geospatial
technologies is related to advances in geodesy, photogrammetry, geophysics, computer science, statistics, remote sensing,
geographic information technology (GIT), and numerical
modeling, which have collectively revolutionized the field of
geomorphology (Bishop et al., 2001; Shroder and Bishop,
2003; Bishop and Shroder, 2004a;). Scientists are routinely
utilizing new spatio-temporal data, geocomputational algorithms, and processing approaches and models, which now
allow assessments far beyond traditional geomorphological
mapping. It is now possible to quantify landscape morphology
(Pike, 2000; Hengl and Reuter, 2009), assess surface biophysical
conditions (Florinsky, 1998; Liang, 2007; Smith and Pain,
2009; Tarolli et al., 2009), assess near-surface conditions, link
process with form and patterns (Allen and Walsh, 1993;
Montgomery et al., 2004), and improve our understanding of
scale dependence and the polygenetic nature of landscape
evolution (Walsh et al., 1997; Tate and Wood, 2001; Bishop
et al., 2003). Nevertheless, such rapid utilization of information
technologies must be carefully examined, given the empirical
nature of utilizing geographic information systems (GIS), the
need for formalization to address numerous issues, and requirements of accuracy and repeatability (Bishop et al., 2012).
Numerous conceptual/theoretical and methodological
issues are at the heart of effectively utilizing spatio-temporal
data and GIT to study geomorphological systems (Bishop
et al., 2012). Therefore, Earth scientists need to be fully aware
of current capabilities as well as the issues and challenges related to geomorphology and geographic information science
(GIScience) (Bishop and Shroder, 2004b; Bishop et al., 2012).
Geospatial technologies can be used to address various conceptual and practical issues such as heterogeneous surface
composition with fuzzy-classification membership (Warner
and Shank, 1997), indeterminant boundaries and features
(Burrough, 1989; Usery, 1996; Burrough et al., 2000; Smith
et al., 2000; Deng and Wilson, 2008), hierarchical organization and spatial analysis using object-oriented technology
(Ralston, 1994; Brandli, 1996; Schmidt and Dikau, 1999),
scale dependence of properties and patterns using geostatistics
(Tate and Wood, 2001), and objective mapping using different
analytical approaches (e.g., descriptive statistics, inferential
statistics, artificial intelligence, and various analytical reasoning technologies).
Nevertheless, numerous limitations are associated with the
use of existing cartographic representations of environmental
and geomorphological information, as parameterization
schemes that uniquely characterize the formal structure of
natural systems and information associations (i.e., attributes,
space, and time), and the formal linkage of multiple processes
and forms in spacetime is not readily available. Consequently, cartographic representational schemes are still effectively being used, although more complex representational
schemes typically occur outside of GIS environments. Addressing numerous science issues will most likely require
multidisciplinary collaboration between Earth and information scientists (Bishop and Shroder, 2004b).
From a more methodological perspective, advances in

the development and deployment of new sensors and mobile
platforms will provide for new forms of data across a multitude of scales. This has enabled researchers to address numerous issues at the requisite scale. Active research areas
include the assessment of surface processes and their impact
on erosion and landscape evolution (e.g., Finlayson et al.,
2002; Finlayson and Montgomery, 2003) and the evaluation
of data and methods for geomorphological mapping (e.g.,
Saadat et al., 2008; Schneevoigt et al., 2008). The role of remote sensing in generating high-quality digital elevation
model (DEMs) is essential (Wilson and Gallant, 2000), as the
topography inherently incorporates a multitude of morphologies that are generated due to the interaction of climate,
tectonic, and surface processes. The development and evaluation of new techniques and analytical approaches for information extraction from multispectral, multitemporal, and
DEMs is an active research area (Bishop and Shroder, 2004b;
Bishop et al., 2012). Investigators have focused on the technical aspects of developing GIS databases (Gustavsson et al.,
2008), developing geomorphometric mapping software
(Klingseisen et al., 2008), mapping specific landform features,
and developing new ways to visualize geomorphological
information (Vitek et al., 2008). A plethora of quantitative
metrics and approaches exist. The advantages and limitations,
however, of numerous algorithms and multistage processing
approaches have not been rigorously evaluated and compared
for specific applications. Furthermore, standardized formalizations for specific applications are urgently needed.
Consequently, numerous issues related to taxonomy, scale,
process mechanics, feedback mechanisms, system dynamics
and states, representational schemes, algorithms and processing protocols, visualization of complex information, and
effective information distribution need to be accounted for.
The objective of this chapter is to place the current volume
into perspective and provide an overview of remote sensing
and GIScience contributions to geomorphology and the Earth
sciences in general. It is important that significant capabilities
be highlighted so that scientific inquiry may be facilitated by
the use of spatio-temporal data and GIT. It is also necessary to
highlight the challenges and issues associated with the use and
evolution of GIT, so that rapidly evolving capabilities effectively address scientific and practical issues. Given the exceptionally diverse fields of study that encompass Earth
science and GIScience, it is not possible to cover all developments in a comprehensive fashion. Rather, the topics that have
the most direct significance for understanding geomorphological systems and landforms have been selected.
3.1.2
Geospatial Technology and Fieldwork
Fieldwork has been and will continue to be an important aspect of scientific inquiry and mapping. The nature of field
equipment has radically changed from analog to digital devices, and automated sensors and systems ensure specific
levels of accuracy (e.g., location, spectra-based mineral identification, microclimate). Field sensors can be used for recording many different landscape parameters related to
microclimate, ecological, hydrological, sedimentological, and
lithological systems. Telecommunications and computers

allow the rapid collection and analysis, whereas mobile sensor
platforms allow expanded coverage.
Relatively new quantitative spatio-temporal data are related
to advances in geodesy. A significant milestone was the development of the global positioning system (GPS). A constellation of GPS satellites communicate their position with
each other and ground receivers, such that attribute information collected in the field is tagged with precise spatiotemporal coordinates. Spatial and temporal information can
be used to compute distances, directions, and rates of change.
Consequently, GPS technology is routinely used in fieldwork,
and GPS data have significantly contributed to the understanding of environmental change, glaciation and glacierization, tectonics, and facilitate soils and landform mapping
(e.g., Bilham et al., 1997; Gao and Liu, 2001; Banerjee and
Burgmann, 2002; Sella et al., 2007; Bishop et al., 2010; Flowers
et al., 2011).
Similarly, advances in laser, computer, and communication
technology have resulted in new surveying equipment.
Field devices include laser-based total stations, laser rangefinders, hand-held spectral radiometers, radiation sensors, and
many other portable sensors that transfer data to loggers or
satellites for subsequent downloading. Interactive data loggers
can be used to visually examine digital aerial photography or
satellite imagery in the field, and GIS data layers can also be
viewed to facilitate fieldwork. This allows improved data collection for geomorphological mapping, sampling, and targeted investigations.
Terrestrial laser scanners can now be used to study surface
processes in great detail. These laser-scanning devices can
be mounted on tripods, jeeps, or terrain rovers to collect
high-resolution 3-D point clouds (Figure 1). High-resolution
DEMs can be generated over relatively short time intervals.
Rates of erosion, deposition, and ablation can be estimated by
comparing altitude values over time. Furthermore, process
mechanics can be studied, as the relationships between process, form and topographic evolution can be quantitatively
characterized.
New developments in field-based radiation sensors also
facilitate field and remote sensing studies of geomorphological systems. Investigators can use field-based spectral
radiometers and spectra-based mineralogical identification
systems to collect spectra in the field and identify the mineralogical composition of rocks, sediments, and soils. Specific
biophysical characteristics of surface materials include primary
silicate minerals, secondary silicate minerals, water content,
and organic matter. Furthermore, samples of material collected in the field can be brought back to the laboratory for
subsequent spectral analysis.
Other radiation sensors are used to measure the magnitude
of incoming and outgoing radiation that governs many surface
processes and the surface-energy budget. Consequently,
pyranometers, pyrheliometers, albedometers, and shortwave
and longwave net radiometers are used to measure various
components of the radiation-transfer cascade. Microclimatic
conditions influence weathering, ablation and melt water
production, glacier erosion, avalanching, and other mass
movement processes. Such field-based investigations are critical for evaluating the information content in multispectral
Figure 1 Terrestrial laser scanning at the Chalk Cliffs Natural Debris

Flow Laboratory (Summer 2010). A Leica HDS C10 laser scanner was
used for repeat surveying. Pictured in the image are Thad Wasklewicz
(yellow helmet) and Dennis Staley (red helmet). Photograph courtesy
of Thad Wasklewicz, East Carolina University.
satellite imagery and validating the accuracy of thematic information generated from satellite-derived data and numerical
models. Such validations (e.g., surface temperature and albedo
estimates from imagery) provide new opportunities for geomorphological assessment via remote sensing.
Given the increasing miniaturization of sensors and
near-field communication capabilities, investigators are
recognizing the potential of establishing field-site sensor networks to characterize and monitor changes on the landscape.
The parameter list for such networks is ever increasing
and potentially includes atmospheric variables, photography,
and surface parameters. Depending on the spatial density of
station sensors, data can be transmitted to satellites and
downloaded to produce spatio-temporal information layers
via spatial interpolation. Future developments in sensor networks will most likely include robotic rovers, where information collected from the network governs the location of
sensors, such that the network adapts to environmental conditions in an attempt to better characterize the spatio-temporal variability of phenomena being collected by the sensor
network.
Communication technology now allows ground control of
unmanned ground and aerial vehicles, which provides new
opportunities for collecting data in the field. The University of
Arizona, in a project led by Jeffrey Kargel, and assisted by the
Kathmandu-based International Centre for Integrated Mountain Development (ICMOD), has conducted feasibility studies
for the use of remotely controlled boats equipped with side
scan sonar and other instruments to study glacial lakes
(Figure 2). Featured here is a boat deployment on Imja Lake
(Imja Glacier in the background), near Mount Everest, Nepal.
These studies have investigated lake bathymetry and suspended sediments, and density stratification of the water column. Routine deployments in hazardous or dangerous lakes
are the goal of these studies. The potential advantage of boat
deployments, as opposed to conventional human-piloted boat
deployments, is that the small watercrafts are more readily
transported to remote locations and their deployment leaves
the scientific crew safely onshore.
Figure 2 Remote-controlled boat on Imja Lake (Imja Glacier in the background), near Mount Everest, Nepal. The boat has a sensor payload that
allows the acquisition of bathymetric information, video, and water temperature. Such mobile platforms and sensors are useful for studying
supraglacial lake development and monitoring flood hazards associated with proglacial lakes and landslide-valley impoundments that cause
catastrophic flooding. Photography courtesy of Jeffrey Kargel, University of Arizona.
Deployment of terrain rovers with sensors is also being

tested at the University of Arizona by Wolfgang Fink. Research
is aimed toward the development and deployment of truly
autonomous observing systems. These platforms and sensors
involve semi-autonomous navigation, where input from
scientists would be greatly reduced. Such technology and
equipment involves advancements in semi-automated data
analysis and autonomous scientific interpretation and decision making using various forms of artificial intelligence.
Other possibilities also include aircraft and gliders with various sensor arrays.
Supplementary material related to this article can be found
online at doi:10.1016/B978-0-08-088523-0.00040-X.
A host of different geoscience fieldwork applications are
also becoming available for field-portable laptop computers,
wherein checklists for data collection, section measuring, and
other typical field data can be collected electronically. Given
many new possibilities, fieldwork is becoming more complex
and automated. Special attention must be paid to the choice
and suitability of sensors and technologies, sensor-system
deployment and sampling, key data parameters, and storage
issues. Collectively, geospatial technologies have greatly enhanced fieldwork and data collection.
3.1.3
Remote Sensing and Geomorphology
Advances in optical, microwave, and light detection and ranging (LiDAR) remote sensing now allow a multitude of
landscape information to be acquired. Data and information
characteristics are closely tied to sensor characteristics that
must be carefully considered. Improvements in spatial resolution enable better recognition of small-scale objects. A recent trend is high-resolution imagery which is now available
via commercial systems, although cost is an issue for many.
Similarly, the evolution of sensors has brought about an increase in spectral resolution that includes measuring more
regions of the electromagnetic spectrum with improved spectral bandwidths. Hyperspectral sensors, or imaging spectrometers, record hundreds of spectral images using 812 nm
bandwidth ranges. Better radiometric resolution (1012 bits)
also allows sensors to be more sensitive to reflectance and
emission variations. Microwave remote sensing has also seen
significant advancements, and LiDAR remote sensing provides
numerous possibilities for the study of the atmosphere and
landscape. Consequently, a plethora of spatio-temporal data
can be used by Earth scientists.
3.1.3.1
Photography and Videography
Ground photography has historically played a significant role

in conducting fieldwork. Pictures of landforms, topography,
surface cover, and environmental conditions have been used
to document landscape features, and provide the basis for
qualitative interpretations of landforms and surface processes.
Photography, based on the chemical processing of film, was
mainstream for a considerable amount of time. Technological
advances have recently resulted in the use of digital photography and fieldwork now involves the use of digital cameras
that generate megapixel images (Figure 3). Nevertheless, the
ability to acquire repeat ground photography is important in
documenting the Earths surface and for change detection.
Classic examples involve monitoring lake water levels, glacier
advancement and retreat, soil erosion, vegetation succession,
mass movements, and in general, landscape evolution (given
sufficient time).
Digital camera technology is rapidly advancing and it will
soon be possible to obtain giga-pixel digital images. Research
is underway to produce a digital camera that can take panoramic images. A single-shot image of a landscape would
allow a detailed examination of the entire landscape, as such
high-resolution systems generate detailed data for subareas
within the panoramic scene. This technology, coupled with
photogrammetry and GPS technology, will markedly improve

3.1.3.2
Figure 3 Ground photograph of the north-facing Braldu Valley in

the central Karakoram Himalaya of Pakistan. Glaciation has had a
profound influence on the topographic evolution of the Karakoram,
and such data document glacially polished valley walls, high-altitude
erosion surfaces, and a variety of depositional features associated
with mass movement and catastrophic flooding. Fieldwork and
ground photography allow geomorphological mapping and
reconstruction of significant erosion events.
our ability to comprehensively document the landscape as

never before. Critical issues involving data acquisition include
data volume and storage, computer memory, and processor
speeds.
Such technology will inevitably be placed on mobile
platforms, thereby providing detailed spatial coverage. One
may also envision giga-pixel multispectral images. Consequently, these systems would represent state-of-the-art sensors for aerial and space photography, and such systems
would provide for a tremendous volume of multitemporal
data for change-detection studies.
It should be kept in mind, however, that a tremendous
amount of historical aerial and space photography can be
used to study geomorphological systems. Classic examples
include the use of multitemporal aerial photography for
studying coastline changes and sediment transport, aerial and
space photography for monitoring glacier fluctuations, space
photography and imagery for studying regional geomorphological conditions, and ground and aerial photography for
inventory and assessment of mass movements. It is critical to
understand that multitemporal photography serves as baseline information on which existing data for quantifying environmental change can be compared.
Developments in videography can also considerably improve our understanding of surface processes and landforms.
Ground and aerial video of catastrophic events document such
events and provide insights into the nature of process mechanics. Examples include debris flows and landslide events,
earthquakes, and associated landscape changes, tsunamies,
and flooding events. Thermal videography is also being increasingly utilized in volcanology (e.g., Vaughan et al., 2005),
and in many surface applications to study energy budget
and depositional processes (e.g., Hardgrove et al., 2009).
Many examples are presented in Chapter 3.2 by Shroder, as he
specifically addresses the use of photography and videography
in geomorphology.
Imaging Spectroscopy
Optical imaging sensors record the magnitude of reflected

and emitted radiation from planetary surfaces. Sensors can be
mounted on aircraft and satellite platforms to obtain
multispectral and multitemporal information about surface
characteristics. The data are qualitatively and quantitatively
analyzed to generate thematic and quantitative biophysical
information. The nature and accuracy of remotely derived
information is highly dependent on sensor characteristics regarding spatial, spectral, and radiometric sensitivity. Consequently, not all sensors can provide similar information, as
they have been specifically designed to address issues related
to sensor-system evolution and application objectives.
With the advent of resource satellites, starting in the 1970s
and 1980s, Earth scientists were able to assess and map the
regional geomorphological conditions on the Earth (e.g.,
Short and Blair, 1986). Numerous governments subsequently
launched their own satellites with different sensor payloads,
such that currently, a multitude of information products and
satellite data are routinely used in Earth science applications.
The evolution of imaging sensors has been one of improving
spatial, spectral, and radiometric resolution and signal-tonoise ratio. This evolution has already had a profound influence on Earth science, and high spatial-resolution sensors
(15 m) allow improved thematic mapping capabilities.
Nevertheless, many sensors do not have a comparable spectral
resolution, exhibiting a limited number of spectral bands that
may or may not cover key regions of the electromagnetic
spectrum.
The primary sensor characteristic that governs image information content is spectral resolution. Therefore, more
spectral bands theoretically facilitate obtaining unique information. In practice, this is not the case, as the amount of
generalization associated with the spectral bandwidth determines the degree of multicollinearity in the data. Imaging
spectroscopy represents the collection of many narrow, spectrally contiguous bands, such that each pixel contains a
spectrum. In the literature, imaging spectroscopy is also
commonly referred to as imaging spectrometry or hyperspectral remote sensing. Imaging spectroscopy has existed
for the past three decades, and has witnessed exponential
growth recently, as new sensors are being developed and
evaluated, and hyperspectral data are being increasingly utilized in a large variety of domains including ecology, hydrology, soil science, geology, and geomorphology (Schaepman
et al., 2009).
3.1.3.2.1
Sensor parameters
Electro-optical sensors consist of detectors that record the incident at-sensor radiance as an electronic signal. The signal
varies over time, space, wavelength, and amplitude depending
on sensor responsivity, which in general is called resolution.
The sensor characteristics can significantly alter the signal
and ultimately determine the level of generalization and the
nature of the information represented in the imagery. Furthermore, knowledge of sensor characteristics is required for
interpretation and analysis, given sensor influence on spectral
variability.
The spatial resolution of a sensor is usually referred to as

the ground instantaneous field-of-view (GIFOV). It is the
geometric projection of the detector width, w, and is defined as

IFOV
H
w ;
1
GIFOV 2:0H tan
2:0
f
where IFOV is the instantaneous field-of-view, H is the height
of the sensor above the Earth surface, and f is the focal length
of the sensor. The GIFOV determines whether an object will be
resolved by the system, as GIFOV and object dimensions determine the sensors object-discrimination capabilities. The
sensor-system averages the recorded signal over the spatial
extent of a pixel. Specifically, the spatial response of a sensor is
characterized by the optical point-spread function (PSF). It is a
weighting function that spatially distorts the inherent spatial
geometry at the surface of the Earth. A common model for the
PSF is the 2-D Gaussian function such that
PSFx,y
1:0
expx2 =2a2 expy2 =2b2 ;
2:0pab
DNl intgles ol;
where lc is the wavelength corresponding to the peak

weighting and s FWHM(8 log(2))0.5. The full-width at halfmaximum (FWHM) effectively represents the spectral bandwidth, although technically this represents the width where W
is 0.5. A wider spectral bandwidth results in more spectral
generalization. The central wavelength peak and bandwidth
determine whether spectral absorption features will be characterized in recorded spectra. The spectral averaging can be
represented as
Pl2
l1 WlLl
;
Pl2
l1 Wl
In practice, sensor selection is an important aspect of remote sensing, as resolution characteristics determine applicability for problem solving. For example, some panchromatic
sensors may have a higher spatial resolution, although the
data are not useful for biophysical applications, given the SRF
(Figure 4). Other sensors provide more detailed spectral information with less spectral generalization (Figure 5). Finally,
imaging spectroscopy depends on a very fine spectral response,
such that there are a large number of spectral bands and
narrow spectral bandwidths (Figure 6).
The electronic signal must also be amplified to provide for
a sufficient signal for quantification. The electronic gain and
offset values are set based on an expectation of the range in the
magnitude of incident radiance from different environments.
The amplified signal is then sampled and quantified into
digital number (DN) values. Therefore, a DN value for each
pixel is represented as
5
2
where a and b determine the width of the PSF in the cross- and
in-track directions. For well-designed optics, a b. Consequently, moderate- to coarse-resolution sensors produce
spectral data that are not representative of a single type of
matter or biophysical property, and the system produces
composite spectra.
The spectral resolution of the sensor is also an important
consideration, as sensors exhibit spectral bands that measure
energy in different regions of the spectrum. In theory, different
regions of the spectrum can be used to obtain different information, as matter/energy interactions can be fundamentally
different. Consequently, it is advantageous to sample the visible
(0.40.7 mm), near-infrared (0.71.3 mm), shortwave infrared
(SWIR) (1.33.0 mm), and thermal (3.0100 mm) regions of the
spectrum. The total amount of energy measured in each spectral
band must be averaged over a spectral bandwidth, and each
detector has a spectral-response function (SRF) that characterizes the wavelength weighting. An ideal SRF is modeled as a
Gaussian function such that
"
#
l lc 2
;
3
Wl exp
2s2
Ll
4
where L represents the average radiance value and L represents

the radiance recorded by the sensor.
where es is the electronic signal, g is the electronic gain, and o

is the electronic offset. The number of discrete DN values that
represent the magnitude of radiance defines the radiometric
resolution of the sensor. This represents 2n bits over the dynamic range in radiance. This characterizes the sensitivity of
the sensor to record variations in the magnitude of energy
reflected from the landscape. Most modern sensor have an
812-bit radiometric resolution.
3.1.3.2.2
Reflectance properties and applications
The reflectance properties of surface features are based on their

composition and molecular processes. Incident radiation is
preferentially absorbed by molecules. The total energy of a
molecule is the sum of electronic, vibrational, and rotational
energy. Changes in the energy state are governed by electronic
and vibrational processes. The absorption of energy results in
absorption features in reflectance spectra. Emission of photons
results from a transition to a lower energy state. Consequently,
reflectance spectra can be used to diagnostically detect the
presence of a variety of materials, as many exhibit absorption
features and unique spectral reflectance patterns.
Imaging spectroscopy can be used for a variety of geological and soil-science applications. Mineral detection is
critical for many applications. Imaging spectroscopy has long
been used to explore for mineral deposits and for lithological
mapping (Goetz and Rowan, 1981; Kruse et al., 1993). This is
possible as electronic transition and charge transfer processes
associated with transition metal ions determine the position
of diagnostic absorption features in the spectra of minerals
(Burns, 1970). Vibrational processes in H2O and OH also
produce overtone absorptions. Consequently, the position,
shape, depth, width, and asymmetry of absorption features are
controlled by the crystal and chemical structure of the mineral
(van der Meer and de Jong, 2006). Absorption bands can be
found in the visible, NIR, and SWIR regions of the spectrum.
For a more detailed characterization of mineral reflectance
spectra, see Grove et al. (1992).
The reflectance spectra of minerals are well known, and
numerous spectral libraries can be used to facilitate remotesensing studies. Specifically, the United States Geological
1.0
0.9
Spectral response weight (w)
0.8
0.7
0.6
0.5
0.4
0.3
0.2
0.1
0.0
0.45
0.50
0.55
0.60
0.65
0.70
0.75
0.80
085
0.90
0.95
1.00
Wavelength (m)
Figure 4 Landsat-7 Panchromatic sensor spectral response function. The sensor is more sensitive to the near-infrared region of the spectrum,
and this reduces the atmospheric effects in the imagery. The wide spectral bandwidth does not allow the imagery to be utilized to assess specific
biophysical conditions because extensive spectral averaging does not accurately characterize narrow absorption features.
0.80
0.75
0.70
0.65
Reflectance
0.60
0.55
0.50
ASTER illite spectral curve
Landsat 7 TM Illite spectral curve
0.45
0.40
0.35
0.30
0.25
0.4
0.6
0.8
1.0
1.2
1.4
1.6
1.8
2.0
2.2
2.4
Wavelength (m)
Figure 5 Simulated ASTER and Landsat-7 Thematic Mapper spectral-reflectance curves for the secondary-silicate mineral illite. Sensor spectralresponse functions determine the degree of generalization associated with the spectral curve. This example demonstrates that ASTER data depict
the illite absorption feature in the SWIR, whereas the TM spectral curve does not. Consequently, spectral resolution must be carefully considered
for different applications.
Survey Spectral Library and the ASTER spectral library can be

used to simulate composite spectra using spectral mixing
models. In addition, spectral matching algorithms can be used
to assess the degree of similarity between pixel spectra and
laboratory spectra. These types of analyses allow assessment of
the potential presence of specific minerals at the surface. Rock
spectra can be simulated based on standard or assumed
mineralogical composition. Linear and nonlinear spectral
mixing models and model inversion can be used for
lithological and soils mapping and finding mineral endmember distributions, respectively.
Imaging spectroscopy provides many new capabilities for
studying soil properties (Ben-Dor et al., 2009). Soil degradation
due to salinity, and caused by rising water tables or irrigation,
can be assessed via hyperspectral remote sensing (Taylor et al.,
1994; Metternicht and Zinck, 2003). Specifically, Taylor et al.
(1994) showed that soil salinity could be mapped using airborne spectrometer data. More specifically, Taylor (2004)
12.5
12.0
AVIRIS FWHM (nM)
11.5
11.0
10.5
10.0
9.5
9.0
0.4
0.6
0.8
1.0
1.2
1.4
1.6
1.8
2.0
2.2
2.4
AVIRIS wavelength center (m)

Figure 6 Airborne visible/infrared imaging spectrometer (AVIRIS) spectral response-function characteristics. The sensor exhibits 224 spectral
bands with very narrow spectral bandwidths. Notice that the wavelength interval and FWHM values vary across the spectrum. The very high
spectral resolution allows improved characterization of measured spectra.
demonstrated that the depth and width of the hydroxyl absorption feature at 220 nm changes with increased soil salinity.
Others have found that surface gypsum is highly correlated with
the NaCl content in some soils and can be spectrally identified.
Other soil processes can be examined and include soil
erosion and deposition (Ben-Dor et al., 2009). Using airborne
visible/infrared imaging spectrometer (AVIRIS) data, Hill et al.
(1995) used a spectral mixing model with end-member
spectra to describe the status of soil erosion. They estimated
the relative abundance of parent material and soil particles on
the surface. This represents the mapping of erosion state, as it
is related to the mixing ratio between developed substrates
and components of the parent material.
Another approach to assessing soil erosion involves
quantitative assessment of soil chemical properties (Ben-Dor
et al., 2009). For example, Hill and Schutt (2000) suggest that
organic carbon can be a tracer substance for identifying accumulation areas and relatively stable soil conditions.
Stable conditions correspond to higher infiltration and water
retention capacity; therefore, erosion and depositional areas
can be mapped based on detecting organic carbon. The
curvature of the spectral continuum from 0.4 to 1.6 m was
used to derive the organic carbon content of soils (Hill and
Schutt, 2000). Research has indicated that spectral differences
are associated with variations in weathering and natural versus
eroded soils (Dematte, 2002).
Other capabilities are also feasible and include soil mapping
and classification, extraction of information regarding soil genesis and formation, and assessment of soil contamination and
swelling (Ben-Dor et al., 2009). These capabilities are related to
assessment of specific soil properties including: (1) presence and
absence of primary and secondary silicate minerals; (2) quantification of iron oxides; (3) content, composition, and maturity
of organic matter; (4) quantitative estimation of heavy metals;
and (5) assessment of soil moisture content.
Imaging spectroscopy is also used in vegetation science and

ecology. Its utility is based on assessment of specific physical
and chemical characteristics of vegetation. In the visible region
of the spectrum, plant pigments absorb radiation and absorption features can be used to assess a variety of plant pigments. In the NIR region, plants exhibit relatively high
reflectance and transmission. The distribution of air spaces
and the size, shape, and arrangement of cells determine the
passage of light through plant leaves and the amount of scatter
in the mesophyll layer of leaves. In the SWIR region, leaf
moisture and foliar biochemicals influence reflectance, and
this region contains strong water absorption features and
minor biochemical absorption features.
Variations in environmental conditions can cause variations
in chlorophyll production, leaf cellular structure, and leaf
moisture conditions. Consequently, variations in these and
other canopy characteristics (e.g., leaf area index) can be detected via imaging spectrometer data, and provide an insight
into surface/subsurface conditions, as plants respond to soil
geochemical, lithological, and structural conditions. Numerous
studies have evaluated the use of spectra and the shifting of the
red edge to detect plant stress and their response to mineral
deposits, lithological changes, and other environmental characteristics (e.g., Collins et al., 1983; Boochs et al., 1990).
Finally, imaging spectroscopy can be used for assessing
water in various phases within the Earths natural systems, and
this supports new strategies for hydrological research and assessment of quality and distribution (Green et al., 2006). Remote sensing has commonly been used for the assessment
and mapping of inland and coastal water quality conditions.
With the advent of spectrometers, atmospheric water-vapor
conditions and ice-crystal size variations in snow can be assessed quantitatively (Dozier et al., 2009). This allows distributed water-budget assessment in high-altitude basins that
facilitates water management and planning activities.
10
Each phase of water exhibits absorption features between

400 and 2500 nm. Water vapor exhibits an extremely fine
spectral absorption structure, and liquid water and ice exhibit
broad molecular absorption bands (Green et al., 2006). This
allows spectroscopic separation of the three phases. Consequently, imaging spectroscopy will play an ever-increasing
role in hydrological research, as detailed information regarding surface-energy conditions (i.e., albedo) can be used
for energy-budget modeling, the update and validation of
distributed snowmelt and runoff models, and in assessing
water-quality condition and bathymetry in relatively shallow
inland and coastal areas. Unfortunately, a chapter on Imaging
Spectroscopy was not completed in time for inclusion in this
Treatise volume.
3.1.3.3
Microwave Remote Sensing
Unlike optical-based sensors that depend on reflected and

emitted radiation from objects and surfaces, microwave sensors collect information from the microwave region of the
electromagnetic spectrum with frequencies ranging from 0.3
to 40 GHz. Passive and active sensors exist; however, active
sensors play a dominant role.
Numerous sensor types provide a wealth of information
obtained at planetary, region, and local scales. Consequently,
it is essential to be familiar with the advantages and disadvantages associated with the use of specific microwave
sensors. These include radio detection and ranging (radar)
altimeters, synthetic aperture radar (SAR), polarimetric
SAR, stereo SAR, and interferometric SAR. The choice of a
particular sensor and analysis for information extraction is
slightly more complicated compared with optical-based sensors, as the backscatter of microwave radiation is dependent
on frequency selection, sensor-system imaging geometry, polarization, surface composition and roughness, near-surface
structure, and the electrical composition of the surface that
determines the amount of the energy reflected, absorbed, and
transmitted.
Radar imagery can be qualitatively evaluated using human
interpretation techniques. Quantitative analysis is used to
produce topographic information (i.e., DEMs) that is routinely utilized by Earth scientists. In addition, the ability to
assess deformation patterns resulting from natural (e.g.,
earthquakes) and anthropogenic events (e.g., oil and ground
water extraction) allows new insights into the nature of numerous processes related to landscape evolution and natural
hazards. Numerous applications involving geodesy, land
cover, ecology, hydrology, geology, geomorphology, and glaciology are possible.
Hensley and Farr, in Chapter 3.3, provide an authoritative
treatment of microwave remote sensing in geomorphology.
They specifically address the different types of active microwave sensors and relate scale to various geomorphological
applications. Microwave remote-sensing principles are also
addressed so that users take into consideration the multitude
of factors that influence backscatter variations in imagery such
as frequency, resolution, polarization, scattering, and penetration. For each sensor, the theory of operation is examined
and specific geomorphological applications are presented.
This treatment of the topic should provide readers with an

insight into the complex nature of matterenergy interactions
in the microwave region of the spectrum, while highlighting
the wealth of landscape information that can be obtained and
utilized by the Earth science community.
3.1.3.4
The Atmosphere and Climate Forcing
The linkages between climate and surface processes have long

been established by the geomorphological community. Atmospheric conditions regulate surface energy, temperature,
and precipitation. Climate forcing is an active research area
and is known to govern process domains, rates of erosion, and
the presence/absence of specific landforms. A relatively recent
development has been the recognition of the complex interrelationships between climate, surface processes, and tectonics
(Molnar and England, 1990; Shroder and Bishop, 2000). This
has prompted vigorous debate regarding the magnitude
of erosion, role of surface processes, and issues of climate
versus tectonic forcing in orogenesis. Investigations into
climategeomorphology linkages require quantitative estimates of atmospheric variables and maps depicting the spatial
variability of atmospheric conditions. Such detailed information has not been historically available to geomorphologists. Recent advances now allow a detailed examination of
the atmospheric conditions that facilitate climate forcing
studies, and remote sensing provides new avenues for geomorphological research.
The moderate resolution imaging spectro-radiometer
(MODIS) on the Terra and Aqua satellites can be used to
collect a variety of atmospheric parameters. It has a swath
width of 2330 km and can nearly provide global coverage of
atmospheric conditions using 36 spectral bands ranging from
the visible to the thermal region of the electromagnetic spectrum. Specifically, it allows global monitoring of atmospheric
profiles, precipitable water-vapor amount, aerosol particles,
and cloud characteristics. An atmospheric profile algorithm is
used to estimate the atmospheric temperature and moisture
conditions. Consequently, the spatial pattern of the atmospheric temperature and moisture variations at a particular
level in the atmosphere can be examined. See King et al.
(2003) for technical details and a more complete description
of data products generated from the sensor.
Other imaging spectrometers can also be used to assess
atmospheric conditions. For example, Green et al. (2006) used
AVIRIS over Mount Rainer in Washington to assess water vapor
content. At the water vapor absorption bands (near 940, 1150,
1380, and 1900 nm), the upwelling radiance at the sensor
varies in strength as a function of the column water vapor
content. Numerous investigators have found a strong relationship between water vapor and elevation (Green et al.,
2006). Consequently, AVIRIS data can be used to generate
high-resolution atmospheric moisture maps.
The magnitude of precipitation is another key atmospheric
parameter. Satellite remote sensing is an indispensable means
of measuring and monitoring precipitation on a global scale
(Masunaga et al., 2002). The tropical rainfall measuring mission (TRMM) has made significant contributions to the Earth
sciences, as scientists have access to a variety of precipitation
information products. This information is critical for waterbudget studies involving ground water recharge and extraction, surface runoff and human impact on basin hydrology, water storage and resources in mountain environments,
and in the assessment of flooding and flood inundation.
Furthermore, this parameter is critical for studying the influence of climate forcing on mountain topographic evolution,
studying specific climate systems (e.g., Houze et al., 2007),
and validating climate simulations.
Numerous researchers have examined spatial patterns in
precipitation in an attempt to quantitatively relate climate to
topography, erosion, and specific process domains (e.g., Burbank et al., 2003; Wulf et al., 2010; Bishop et al., 2010). Research indicates a coupling between precipitation and
topography (Nesbitt and Anders, 2009) and magnitude variation can characteristically reveal the presence of orographic
precipitation and erosion zones that may be related to uplift
patterns. For example, Wulf et al. (2010) examined seasonal
precipitation gradients using TRMM data to determine their
impact on fluvial sediment flux in the Himalaya. They found
that the Indian summer monsoon is the main driving force for
erosional processes, despite more precipitation falling in the
winter season. Similarly, Bishop et al. (2010) used TRMM data
to find that the highest amounts of precipitation in
the Karakoram occur during the spring by the westerlies
(Figure 7). They also noted a spatial coincidence between a
precipitation anomaly in the Hunza region and increased
landscape dissection and steep slopes, suggesting the presence
of a high-magnitude erosion zone.
Remotely derived atmospheric information has not been

thoroughly utilized in geomorphology and numerous research
opportunities exist. For example, an important part of landscape evolution is related to physical and chemical weathering
and regolith production. Research has demonstrated links
between weathering and surface and atmospheric conditions
(Curtis, 1976; Trudgill, 1976). Spatial distribution patterns of
air and surface temperature, air and surface moisture, surface
mineralogy, and vegetation characteristics should provide insights into the magnitude and distribution of weathering and
regolith production. Perhaps modelers might be able to develop new parameterization schemes that incorporate satellitederived atmospheric and surface parameters. This would significantly improve the ability to account for the spatial variability in weathering and sediment transport. Nevertheless,
there would be other difficult issues to address that include
accounting for biological processes, acid production, and the
integrated influence of climate change that extends beyond the
temporal availability of the data.
Atmospheric information is also critical for understanding
and predicting the surface-energy budget that governs many
surface processes including weathering, moisture availability,
erosion and sediment transport. Numerous field studies typically use point station data to obtain measurements of atmospheric conditions, not knowing the degree of variability
associated with the shortwave and longwave net radiation flux.
Atmospheric conditions, surface albedo and temperature, and
topographic information via remote sensing allow more sophisticated modeling that can be used for hydrological and
12
11
10
9
8
Month
11
0
20
40
60
80
50
60
70
80
90
7
6
5
4
3
2
1
1998 1999 2000 2001 2002 2003 2004 2005 2006 2007 2008
Year
Figure 7 Tropical rainfall mapping mission (TRMM) precipitation data (3B43V6) for the Baltoro-Mustagh region in the Karakoram Himalaya,
Pakistan. The temporal variation in cumulative precipitation (mm per month) depicts the influence of the westerlies during the spring and the
combined influence of the southwestern monsoon and westerlies during the summer months. TRMM data can also be used to examine spatial
patterns in precipitation. Such satellite data allow a spatio-temporal evaluation of precipitation conditions.
12
glaciological investigations. Consequently, satellite-derived

atmospheric information can greatly improve the understanding of many aspects of geomorphological systems. Unfortunately, a chapter on this topic was not completed in time
for inclusion in this Treatise volume.
3.1.3.5
Land-Cover Assessment and Mapping
Aerial photography and satellite imagery have been historically utilized to generate land-cover and land-use information.
A multitude of new high-resolution and multispectral sensors
acquire data that can be used to characterize a wide range
of biophysical landscape properties. Collectively, this information can be used for thematic mapping of land-cover
characteristics. Research has focused on evaluating the information content within multispectral datasets, evaluation and
comparison of pattern-recognition techniques for improved
classification of thematic content, and the development
of new methodological approaches for data fusion, spectralfeature extraction, spatial-feature extraction, and multitemporal analysis. Although challenges remain with respect
to addressing increased data volumes and increased spectral
and temporal variability, existing software systems allow
routine mapping of fundamental land-cover classes. Thematic
mapping and study of land-cover dynamics represent an
important component of geomorphological assessment and
mapping. Furthermore, land-cover and land-use patterns
provide an insight into the magnitude of anthropogenic
forcing.
New capabilities include developments in a variety of subdisciplines. For example, data fusion is an approach to mapping
and analysis that exploits the power of multiple representations
of the landscape. This involves integrating data with different
spatial, spectral, and radiometric resolutions. A classic example
is merging multispectral satellite data with higher-resolution
panchromatic data. In a GIS, multiresolution airborne and
satellite data can be fused with a DEM, terrestrial photography,
maps, and graphics. Digital mapping can be accomplished by
utilizing various feature sets that represent multiple landscape
dimensions and perspectives.
Object-oriented land-cover mapping also represents a
relatively new development. It first requires meaningful segmentation based on specific criteria to generate spatial entities
called objects. Initial segmentation is typically based on information in imagery and DEMs. Numerous approaches to
segmentation can be used including homogeneity and shape
analysis, region growing, pattern recognition, and rule-based
segmentation. Segmentation results are then analyzed via
spatial clumping to identify individual homogeneous spatial
entities. These objects then serve as a spatial constraint for
subsequent analysis. Object-oriented analysis involves computing the attributes of individual objects such as object location, size, shape, and its topological relationships with other
objects on the landscape.
Mapping can be facilitated by spatial aggregation and
spatial intersection of objects and by identifying unique patterns of object attributes in an n-dimensional feature space.
This approach is widely recognized as superior to purely pixelbased classification procedures, as it allows the integration of
image elements (i.e., tone, texture, size, shape, pattern, site)

and the linkage of spatial objects across multiple scales.
This approach is also valuable for mapping specific
landforms based on the segmentation of terrain units using
geomorphometry.
Allen and others in Chapter 3.4 specifically address the
issues of land-cover and land-use assessment via optical remote sensing. They highlight the data sources that are valuable
for land-cover mapping and review some of the methodological approaches that are routinely utilized. Finally, they
provide classic application examples of how land-cover information is used in geomorphology and provide a treatment
on land-cover change detection.
3.1.3.6
Near-Surface Geophysics
Remote sensing of the subsurface is commonly required to

assess geomorphological systems and to accurately map the
three-dimensional extension of landforms and structural features. Subsurface compositional variations and the occurrence
of subsurface structures can have a significant influence on
surface processes and landscape evolution.
Information on subsurface materials and characteristics
can be obtained by passive gamma-ray spectrometry and
geophysical techniques such as seismic, gravity, aeromagnetics,
electromagnetics, and ground penetrating radar (GPR).
Gamma-ray spectrometry may indicate the composition of
materials in the upper 50 cm of the surface (Smith and Pain,
2009), whereas gravity, aeromagnetics, electromagnetics, and
GPR can be used to assess density, subsurface features, conductivity variations, and depths, respectively (Lane, 2002;
Wilford, 2002). Consequently, subsurface lithological variation can be compared with surface morphometry and other
biophysical properties to characterize the 3-D nature of
landforms. The cost and availability of such subsurface information is currently a serious limitation, as expensive airborne or field surveys are required.
Nevertheless, the use of geophysical techniques in geomorphology crosses many subdisciplines, as revealed by Kruse
in Chapter 3.5. She summarizes the relative significance of
different geophysical methods in various subdisciplines of
geomorphology. She also provides practical advice for prospective users of near-surface geophysics and highlight the
importance of reference data collected in the field, in addition
to geophysical data. There is a need for the comparison of
methods and caution in the interpretation based on forward
or inverse modeling. Kruse notes that the uses of near-surface
geophysics in geomorphology are not fundamentally different
from the uses in other geoscience disciplines, and that geomorphologists would be well served by examining methods
and results from tectonic, hydrogeophysical, applied geophysical, and engineering studies.
3.1.4
GIS and Geomorphology
The rapid proliferation of GIT allows improved data management, manipulation, analysis, modeling, and visualization
capabilities. Various forms of spatio-temporal data can be
stored in GIS databases, and a plethora of software tools

allows scientists to effectively study spatio-temporal patterns
and relationships.
Such new capabilities represent a substantial evolution in
geomorphological assessment and mapping compared with
traditional approaches (Bishop et al., 2012). Yet, the traditional approaches of information integration via analytical
reasoning, which is the pillar of qualitative interpretation, are
poorly represented by statistical metrics and mathematical
operators that are commonly used in GIS analysis. Furthermore, the results of GIS-based quantitative analysis and numerical modeling are dependent on numerous factors and
simplifying assumptions, and may not be representative of
objective measurements obtained in the field. Consequently,
conceptual and practical issues need to be recognized,
and Bishop et al. (2012) have identified a number of issues
that should be considered. These include representation,
the predominantly empirical nature of using modern-based
GISs, scale, and mapping perspectives. In general, advances
in geomorphology have resulted from addressing these
issues and utilizing geospatial technologies to address specific
problems.
3.1.4.1
Digital Terrain Modeling (DTM)
Quantitative land-surface information is required in geomorphology. A major contribution of remote sensing has been
the development and use of passive and active sensors to
generate DEMs. A variety of techniques can be utilized for
DTM including image photogrammetry, radar or laser altimetry, and interferometric SAR. Photogrammetric applications
utilizing Satellite Pour lObservation de la Terre (SPOT) and
Advanced Spaceborne Thermal Emission and Reflection
Radiometer (ASTER) data are commonly used by scientists. In
the case of SPOT imagery, alternate view perspectives from
multiple satellite passes enable stereoscopic representations,
whereas the ASTER system relies on forward- and back-looking
telescopes to characterize topography through a merged
characterization. Similarly, radar imagery and specifically
shuttle radar topographic mapping mission (SRTM) data are
widely used. The SRTM and ASTER mission objectives were
specifically designed to produce a global DEM data product to
facilitate Earth science applications. These DEMs have resulted
in many new developments to characterize surface morphology and better assess and map the landscape.
More recently, airborne high-resolution LiDAR systems and
terrestrial-laser-scanning systems now generate millions of 3-D
point measurements. These point clouds must be analyzed
and manipulated to ensure accurate interpolation to generate
a bare-Earth altitude field. LiDAR high-resolution DEMs allow
detailed geomorphometric characterization of the surface and
greater mapping accuracy (Figure 8). Such data allow developments in geomorphometry to be exploited, whereas the
same techniques may not be as useful, given a coarser DEM
measurement scale. For example, DEM differencing is an important aspect of change detection suitable for examining
spatial patterns of surface dynamics and volumetric analysis,
but the availability of high-resolution, geo-referenced elevation grids is critical. Numerous studies have revealed the
13
Figure 8 One meter digital elevation model generated from LiDAR

data with an average point spacing of approximately 5 m. ESRI
terrain dataset generated from original bare-earth LiDAR points from
the North Carolina Floodplain Mapping Program. The 1 m DEM was
created to enhance hydrographic modeling operations within the
watershed. DEM courtesy of Jeffrey Colby, Appalachian State
University.
significance of using an optimum resolution to appropriately

characterize geomorphometric parameters for erosion and
mapping investigations (Zhang and Montgomery, 1994;
Napieralski and Nalepa, 2010).
DTM involves many issues related to the nature of data
acquisition and sampling, preprocessing, spatial interpolation,
quantitative characterization of error and uncertainty, and
postprocessing. Evaluation of DTM for geomorphological
applications is an active research theme. Wasklewicz and
others in Chapter 3.6 discuss the generation and utility of
DEMs in geomorphology. They address the technical advances
in measuring the topography including laser-scanning, shuttle-based radar, and terrestrial photogrammetry techniques.
They also address numerous technical issues that must be
accounted for via preprocessing and postprocessing. They
correctly acknowledge the need to understand a variety of
DTM issues that are critical in analysis and communication of
information generated from a DEM.
3.1.4.2
Terrain Analysis
The quantification and analysis of the land surface is

called gemorphometry (Pike, 1995, 2000). Geomorphometry
plays a central role in studying surface processes and
for geomorphological mapping. Consequently, it has been
characterized as general and specific geomorphometry,
respectively.
Geomorphometry addresses issues of: (1) sampling attributes of land surfaces; (2) geodesy and DTM; (3) DEM preprocessing and error assessment; (4) generation of landsurface parameters, indices, and objects; and (5) geomorphic
information production and problem-solving using parameters and objects. Each aspect of geomorphometry represents a research subdiscipline and contributes significantly
toward the development of software tools and geospatial
technology. Its significance in geomorphology is expected to
increase, as it can be used for assessing and mapping geology
and tectonics, landform elements and landform, functional
units related to water resources and hydrology, process domains, erosion patterns, as well as climate and meteorological
conditions (Bishop et al., 2012).
14
Progress has focused on:
The development and use of geomorphometric algorithms.

New and modified forms of parameters and indices are
being developed and evaluated for assessment and mapping.
Spatial analysis involves neighborhood operations, subgrid
operations, and multiscale analysis. The primary mathematical approach has been statistical analysis and probability theory; however, geostatistics, artificial intelligence,
and fuzzy-set theory are increasingly being utilized.
New software tools and systems for geomorphometric analysis and mapping. A number of programs are specifically
designed to compute numerous geomorphometric parameters, although many GISs are limited to basic parameters.
Consequently, geomorphometry and mapping can be carried
out using ESRI software (Reuter and Nelson, 2009), SAGA
(Olaya and Conrad, 2009), ILWIS (Maathuis and Wang,
2009), LandSerf (Wood, 2009), MicroDEM (Guth, 2009), TAS
GIS (Lindsay, 2009), GRASS GIS (Hofierka et al., 2009), and
River-Tools (Peckham, 2009), just to name a few.
Existing and new applications. Numerous algorithms and
approaches for characterizing spatial variation, scale, landscape position, fuzzy boundaries, and complexity exist, and
many landforms and features such as drainage basins and
networks, ridges, and peaks can be mapped to various degrees. Nevertheless, researchers have a daunting task of determining which metrics and approaches are best for specific
objectives. Geomorphometry has significantly contributed
to geological, soil, vegetation, landform, ecological, hydrological, mass movements, hazards, meteorological, and
agricultural mapping applications, and new applications are
likely to evolve (Gessler et al., 2009).
Mapping other aspects of the geomorphic system related to
climate and tectonic forcing, process domains, and erosion,
however, is more complex and may require very different
morphometric approaches. For example, quantifying the
extent to which geomorphic parameters or landforms and
landform elements can be used to assess and characterize
tectonic signals, or the influence of tectonics on the landscape, remains a key challenge in the Earth Sciences (Boulton
and Whittaker, 2009; Whipple, 2009). A typical approach
includes the analysis of drainage basins and patterns, and an
evaluation of the longitudinal profiles of bedrock rivers.
Asymmetric drainage patterns, elongated drainage basins,
and convexities and the presence of knick points are thought
to reflect the system response to ongoing tectonic uplift
(Jamieson et al., 2004; Boulton and Whittaker, 2009). Other
applications, such as the sampling and estimation of surface
cosmogenic nuclides, allow estimates of catchment erosion
rates using GIS. This requires knowledge of the production
rate of various isotopes related to the incoming cosmic-ray
flux, which is governed by latitude, altitude, slope, azimuth,
and topographic shielding (Figure 9).
It is essential that geomorphologists be familiar with
geomorphometric parameters and know how to use them
for various applications. Wilson and Bishop in Chapter 3.7
provide a treatment of geomorphometry that highlights various parameters and how they can be used for studying geomorphological systems and for geomorphological
mapping.
3.1.4.3
Landform Mapping
Historically, landform mapping was primarily a field-based

activity, although it has rapidly evolved, given the advent of
geospatial technologies and the ever-increasing availability of
spatio-temporal data (Bishop et al., 2012). New imagery and
DEMs, along with new algorithms and processing approaches,
now allow assessment of surface materials, assessment of nearsurface conditions using geophysical techniques, and assessment of terrain morphology. These new mapping capabilities,
however, bring about new challenges involving theory, concepts, and technology that need to be carefully addressed. This
will require new developments in data, representation, analysis, and visualization.
A fundamental issue in mapping is that of spatial complexity and taxonomy, as robust and reliable classification is
required for landform mapping. Bishop (2009) addressed this
issue and discussed the need for taxonomic criteria that facilitate the classification of volcanic landforms. Spatial complexity was addressed using a generic approach that described
the complexity of landform regardless of scale, climate regime,
or geological and tectonic setting. This concept is critical for
GIS-based landform mapping as formalized taxonomies seem
warranted to address the indeterminant nature of landform
terminology.
Similarly, the indeterminant nature of landform boundaries is an issue, as many landforms do not exhibit a discrete
demarcation in material and form (Burrough, 1996; Lagacherie et al., 1996; Bishop et al., 2012). For example, Deng and
Wilson (2008) used fuzzy theory and fuzzy spatial entities to
map mountain peaks. Peaks were mapped as multiscale entities with modifiable boundaries. Their approach allowed
vague peak entities to be assessed. The approach addresses
landform delineation, spatial continuity, and multicriteria
definition, all important concepts in landform mapping.
Assessment of the spatial uncertainty associated with process domains and landform mapping is another critical area of
investigation. Increasingly, landform mapping is being accomplished by the development and assessment of complex
spatial models that inherently incorporate the error that is
associated with data, algorithm selection, processing procedures, and model assumptions. There is a need to quantitatively characterize the degree of uncertainty in such mapping
models, and selected techniques including artificial intelligence (i.e., neural networks and fuzzy uncertainties) and fuzzy
theory can be used to facilitate such investigations. For example, Luoto et al. (2010) produced spatial uncertainty maps
based on the agreement between different modeling techniques. The uncertainty maps reflected the reliability of assessing
periglacial landforms.
A multitude of studies have focused on improving mapping capabilities for assessing fluvial and hydrological systems
(e.g., Schaper et al., 1999; Marcus et al., 2003), glacial and
periglacial systems (e.g., Schneevoigt et al., 2008; Kaab and
Vollmer, 2000; Frankl et al., 2010), and various aspects of the
sediment transfer cascade including mass movements (e.g.,
Wichmann et al., 2009). Napieralski and others in Chapter 3.8
address the ways in which remote sensing and GIS can be used
for mapping a variety of landforms. Specifically, they highlight
the numerous forms of data and methods that can be used.
15
Figure 9 Skyview-factor image over the Mount Everest region in Nepal. The skyview factor represents the degree to which the atmosphere can
be viewed, given hemispherical topographic variation. Flat terrain exhibits a skyview factor of 1 (white) and lower values (darker greytones)
depict valley bottoms exhibiting more relief. The algorithm examines the relief characteristics around each pixel using an azimuth direction
interval out to a specified distance. The magnitude of the parameter is directly related to topographic shielding of incoming radiation.
Consequently, this parameter is valuable in depicting variation in the cosmic-ray flux required for cosmogenic-based erosion modeling and in
assessing surface irradiance conditions for surface energy-budget modeling, as it governs the diffuse-skylight irradiance. Its evaluation provides
new research opportunities in geomorphology, hydrology, and glaciology.
They also cover landslide mapping and relate this to hazard

studies. This treatment represents many of the traditional and
new forms of analysis that have been used in geomorphological mapping.
3.1.4.4
Spatial Hydrology
Understanding the spatio-temporal complexities of the

hydrological system is essential for characterizing geomorphological systems and for the management of water resources. Detailed information is required about the spatial
variability of functional units and processes that govern the
collection, flow, and storage of water (Beven and Moore,
1993). Consequently, information regarding climate, geology,
topography, land cover, soils, and regolith is needed to characterize various components of the system.
Meso-scale information regarding atmospheric moisture
and precipitation rates can be obtained with passive and active
sensors (e.g., MODIS and TRMM). Spatio-temporal variations
in surface-water conditions are assessed by mapping water,
snow, and ice/glacier distributions (Bishop et al., 2004; Green
et al., 2006; Dozier et al., 2009). Active sonar systems can
produce detailed bathymetric data to facilitate volume estimates for rivers and lakes. Remotely sensed data and analysis
can also be used to estimate the grain size of the snow, the
fraction of each pixel covered by snow, and the amount of
radiative forcing caused by absorbing impurities (Dozier et al.,

2009).
Satellite gravimetry allows assessment of water mass variations in the cryosphere and subsurface (Rodell et al., 2009;
Matsuo and Heki, 2010). Data from the gravity recovery and
climate experiment (GRACE) satellite mission can provide
monthly estimates of the Earths gravitational field (Tapley
et al., 2004). Time variations in the gravitational field can be
used to determine changes in the Earths mass distribution.
Water mass variations represent a dominate signal that can
influence the gravitational field (Wahr et al., 1998). Consequently, if the data are preprocessed appropriately, and mass
variations due to the atmospheric and tectonic conditions are
accounted for, GRACE data can be used to evaluate water mass
variations in the cryosphere (e.g., Chen et al., 2007; Matsuo
and Heki, 2010) and detect large regional anomalies in ground
water fluctuations including the Amazon and ground water
depletion in India (Syed et al., 2005; Rodell et al., 2009).
Consequently, remote-sensing investigations have already
provided new insights into hydrological systems at local, regional, and global scales.
Remote sensing and GIS have also contributed significantly
to hydrological studies, given the advent of DEMs and terrain
analysis capabilities (Beven and Moore, 1993; Wilson and
Gallant, 2000). The production of high- and moderate-resolution DEMs using active sensor systems has markedly
16
influenced hydrological modeling. Numerous topographic

parameters directly and indirectly control hydrological and
geomorphological processes. In general, these parameters have
been classified as primary and secondary (compound) parameters that play a significant role in the spatial complexity of
the hydrological system (Beven and Moore, 1993; Wilson and
Gallant, 2000). Topographic information is now routinely
used for assessing and delineating drainage basins and
drainage networks, and DEMs are required for assessing water
flow direction and regimes.
Topographic parameters govern a variety of climate and
lithological processes including precipitation, surface-energy
balance, erosion, deposition, and rock stress fields and
strength. These in turn influence specific hydrologic processes
and storage including recharge, evaporation, infiltration, soil
moisture content, and surface saturation zones. One of the
most commonly used hydrologically based topographic parameters or indexes is the wetness index, which is used to determine the effects of the topography on the location and size
of saturated source areas (Wilson and Gallant, 2000). Landcover conditions also govern the magnitude of erosion and
infiltration capacity, and remote sensing provides detailed
information required for watershed analysis and the prediction of discharge. Detailed mapping of impermeable surfaces
and GIS-based modeling of urban expansion represent other
aspects of watershed modeling and planning.
Terrain analysis is also critical for assessing precipitation
patterns, and topographic parameters have been found to influence precipitation rates. The orographic precipitation
mechanism is well understood and results in spatial anomalies and gradients in precipitation magnitude in mountainous
terrain. However, valley structure and orientation can also
influence precipitation, as atmospheric flow can be topographically directed into basins or regions. This is the case in
Asia, where monsoon conditions move further inland due to
large glacial valleys that do not restrict air flow inland. Similarly, slope angles and relief also govern precipitation, as these
parameters regulate the rate at which the air will be forced
upward. In orogens such as the Himalaya, precipitation can
vary significantly over relatively short distances. Even a
microtopographical variation has been shown to be related to
precipitation variations (Sharon et al., 1988).
The surface-energy budget is a critical component of the
hydrological system. It regulates ablation, snow melt, evaporation, and transpiration. Energy-budget modeling represents
the most rigorous approach for assessing surface-water runoff
and basin discharge. This type of modeling is critical for the
assessment of water resource potential and can be used to
address issues of sustainability. Currently, governments
around the world are monitoring drought conditions, snowpack variations, and glacier fluctuations to assess future water
supplies. Information related to the key energy-balance parameters can be estimated via remote sensing and numerical
modeling including surface irradiance, albedo, and surface
temperature. Specifically, the net shortwave radiation component can be assessed as atmospheric, topographic, and
surface albedo information is required to predict the direct,
diffuse-skylight, and adjacent-terrain irradiance. GIS-based irradiance modeling can be used, and it is feasible to account
for multiscale topographic effects. Spectral and surface albedo
can be estimated using satellite imagery and bidirectional reflectance distribution models. Similarly, the net longwave radiation component can also be evaluated.
The aforementioned discussion summarizes the many ways
in which remote sensing and GIS play a fundamental role in
hydrological modeling. Data analysis and spatial modeling
provide spatio-temporal information for characterizing the
spatial structure of hydrological systems and key parameters.
Hydrological modeling, however, is mostly concerned with the
flow of water over the surface and in the subsurface (Maidment,
1993). Historically, hydrological modeling has focused on the
temporal evolution of systems, characteristically assuming
uniform spatial properties for various system components.
Spatial hydrology represents the linkage of remote sensing and
GIS with hydrological models to more accurately account for
the spatio-temporal complexities in the hydrological system.
Maidment (1993) reviewed hydrological modeling independent of geospatial technologies and characterized the
nature of such models. They typically focus on surface water
hydrology, surface water quality, groundwater flow, and
groundwater transport. The spatial components associated
with the GIS-hydrological model linkage include watersheds,
pipes and stream channels, aquifers, lakes, and estuaries. A
treatment of the process mechanics of the fundamental flow
systems is beyond the scope of this chapter, although there are
several levels of GIS-based coupling that are itemized below.
Hydrological
assessment involves the mapping of hydrological factors that relate to a situation. A classic example is
the use of the DRASTIC model for mapping groundwatercontamination potential, which characterizes the likelihood
that the groundwater will be contaminated based on point
and areal sources of contamination, and topographic and
subsurface conditions. It represents the utility of GIS-based
spatial modeling and does not utilize any explicit physical
laws. Such spatial modeling is highly empirical and the results are also a function of ranking and weighting of information layers.
Hydrologic parameter determination is a very active research
area in hydrology. The objective is to accurately estimate the
parameters that go into hydrological models, based on atmospheric, terrain, and land-cover analysis. Many examples
have been previously presented.
GIS-based hydrologic modeling represents a more detailed
level of coupling. This is possible, provided that the modeling does not require significant temporal simulations.
Maidment (1993) indicates that one- and two-dimensional
steady-flow computations may be carried out using GISbased modeling, although often times, numerous assumptions are used to eliminate or reduce temporal variability.
Hydrologic modeling of flood inundation is an example of a
2-D GIS-based hydrological model where the groundwater
conditions are not taken into account.
Hydrological modeling that links GIS-based modeling with
hydrological subsystem models accounts for the connection
between the surface and subsurface systems to examine
piezometric head surface and contamination plumes.
The contributions of remote sensing and GIS to hydrology
have the potential to open up new fields of study, as new
information and methodologies allow for the exploration of

new ways to characterize hydrological systems and processes.
The limiting factor is not the ability to characterize hydrological processes mathematically or to solve the resulting
equations, but to accurately characterize the model parameters, given their scale dependencies (Maidment, 1993).
Consequently, recent developments in geostatistics, objectoriented analysis, geomorphometry, and spatial analysis and
modeling have considerably strengthened hydrology. New
developments in spacetime representation and temporal
analysis are also required to facilitate GIS-based hydrological
modeling. Unfortunately, a chapter on this topic was not
completed in time for inclusion in this Treatise volume.
3.1.4.5
Erosion Modeling
Landscape evolution theories indicate that the topography inherently records the interaction between climatic, tectonic, and
surface processes. Consequently, geological and topographic
information can be used to assess erosion and topographic
evolution. Rather than focusing on empirical relationships,
landscape-evolution modeling attempts to formalize the
understanding of process mechanics and systems coupling.
Clearly, a better understanding of processform relationships is needed for developing improved erosion models.
Depending on the climatic and tectonic setting, variations in
weathering, fluvial, mass movement, and glacial processes
make it difficult to accurately assess the magnitude and spatial
distribution of erosion and denudation.
Erosion modeling facilitates the understanding of such
complications, as models account for the conservation of mass
and energy, and are based on a series of mass continuity
equations to address the erosion and deposition of rock and
sediment. Continuity assumes that the rate of change of altitude is proportional to the volumetric sediment flux.
Uplift should account for isostatic and tectonic forcing
components. Tectonic forcing includes the advection of rock
mass, given structural controls, and the alteration of rock
strength, given topographic stress fields and far-field velocities.
The tectonics component requires the integration of mechanical models, as feedbacks exist between the topographic
stress field, rock strength, and erosion and uplift (Koons,
1995; Koons et al., 2002). The magnitude of denudation also
influences the isostatic compensation, which is a function of
the flexural rigidity of the crust and the wavelength of the
topography (Gilchrist et al., 1994).
Most models use a flexible parameterization scheme that
accounts for the depth of regolith production from weathering
(Tucker and Hancock, 2010). Parameterization schemes
should account for variations in lithology, precipitation, and
surface temperature and moisture conditions. Remote sensing
and terrain analysis of surface and atmospheric conditions can
be used to generate maps that may be associated with weathering patterns. Key variables include surface irradiance,
temperature variation, atmospheric water vapor content, and
precipitation patterns. In mountain environments, temperature and precipitation variations may be considerable, given
highly variable topography and forcing factors (Barros et al.,
2006).
17
Hillslope sediment flux can be accounted for using a linear

or a nonlinear relation with the hillslope gradient. Diffusivity
coefficients are commonly used, with different values for different environments.
Bedrock river incision is important in many regions and
investigators have utilized the stream-power bedrock river incision law to account for fluvial erosion. In general, the change
in elevation is modeled as:
qz
KAm Sn
qt
6
where K is bedrock erodability and A is the upstream catchment area that is used as a proxy for discharge. The exponents
m and n are constants used to differentiate between the stream
power and shear stress-based rules.
Simulations of glacier erosion have also been conducted
where erosion is based on basal-sliding velocity and ice
thickness (MacGregor et al., 2000; Tomkin and Braun, 2002;
Pelletier et al., 2010). An abrasion model (Hallet, 1979) can be
used such that the rate of erosion is
qz
aubs
qt
7
where us is the basal sliding speed, and a and b are empirical

coefficients usually set to 1 or 2. Basal sliding is primarily
dependent on the basal shear stress, tb, and a bed-friction
parameter. Simulations of glacier erosion demonstrate the
complexity associated with relating process to form as glacier
erosion can enhance or reduce relief and controls valley spacing and slope variability (Harbor, 1992; Bishop et al., 2003;
Tomkin and Braun, 2002; Pelletier et al., 2010).
Although such landscape-evolution modeling could be
implemented in a GIS (numerous issues do exist), most
models for orogen evolution simulations exist outside of GIS
environments. Common GIS-based erosion models focus on
soil erosion modeling, given relatively mild to moderate
topographic conditions. Soil erosion and sediment transport
studies are important in agriculture, water quality, and sediment budget modeling, and in determining the magnitude
of anthropogenic forcing. Geospatial technologies are ideally
suited for soil erosion modeling and numerous models
exist that use different sediment transport equations. Mitasova
and others in Chapter 3.9 describe different types of erosion
models and provide a physical and mathematical foundation
for understanding their ability to predict the magnitude and
spatial distribution of soil erosion over the landscape. Special
emphasis is on using existing GIS-based erosion models and
the visualization of erosion patterns.
Ultimately, remote sensing and GIS can provide new
information that can be used to develop more rigorous
parameterization schemes for erosion modeling. Geomorphometric characterization is required at each time interval to
drive process mechanics and process domain states, such that
modeling allows parameters, processes, and system characteristics to be mapped. Further advances in GIS spacetime
representations and formal processform linkages are urgently
needed to facilitate improved GIS-based erosion modeling.
18
3.1.4.6
Natural Hazards
Advances in remote sensing, GIT, and numerical modeling

have greatly improved the ability to assess a variety of natural
hazards. To date, remote sensing and GIS are routinely used
for assessing volcanic, earthquake, flood, slope stability, meteorological, and other environmental hazards. Given population growth and rapid environmental change, extreme
hydro-meteorologic events are expected to increase, thereby
causing a higher frequency of hazards.
The literature is replete with research investigating the use
of remote sensing and GIS technology for landslide inventory,
slope-failure susceptibility mapping, and landslide hazard
assessment. Researchers are actively involved with developing,
testing, and validating new GIS-based spatial models to predict slope failures in order to keep landslide hazards from
becoming disasters in highly populated areas. The methodological approach is typically based on empirical relationships
or heuristics, statistical analysis, and the use of deterministic
physical-based models. Carrara and Pike (2008) indicated that
despite the variety of approaches, our abilities to spatially
predict slope failure and hazards are based on unsuitable data,
lack of numerical modeling, or improper characterization of
processes in slope-failure modeling.
A classic GIS example involves the common approach of
utilizing a criterion-weighted scheme based on controlling/
triggering factors that include topographic parameters, lithology, land cover, and other environmental variables. These
factors are assumed to be causative and are subjectively ranked
based on knowledge of the location and existing inventory
information. Spatial overlay and weighting of GIS layers are
used to produce a composite index that is assumed to be related to slope-failure susceptibility. Such GIS-based empirical
approaches have significant limitations as key causative factors/variables involving earthquake activity, rock strength and
structure, topographic stress fields, soil depth, cohesion and
moisture, and climate forcing factors (precipitation and wind
direction) are not usually accounted for. It is important to
realize that many of these factors are dynamic and not accurately characterized by static cartographic representations
using an ordinal measurement scale. The weighting schemes
also do not necessarily portray the relative importance of selected processes and the dominance of site-specific feedback
mechanisms.
Other more sophisticated spatial analysis and pattern recognition approaches involving geostatistics and artificial intelligence techniques have also been evaluated, although it is
essential to recognize the difference between characterizing
spatial and nonlinear patterns and process-based modeling
based on fundamental physics and rock geomechanical
properties. Given the numerous approaches, both qualitative
and quantitative, remote sensing and GIS technology have
promoted more quantitative slope-failure studies. Research
indicates that there is a high degree of uncertainty with respect
to a dominant GIS-based approach to landslide-hazard prediction (Carrara and Pike, 2008), as research tends to focus
more on investigating the method or technology rather than
focusing on inventories and causative factors. GIS data
manipulation cannot be expected to accurately characterize
key parameters and processes. GIS spatial analysis may
potentially produce new information that can help provide an

insight into better understanding mass-movement processes
and generate improved predictive spatial patterns. This most
likely will require an emphasis on the generation of new
causal factors and their integration into GIS-based numerical
models.
Another key example of the role of remote sensing and GIS
in hazards assessment is in characterizing flood magnitude
and inundation. Numerous types of flooding around the
world pose various risks to populations and infrastructure.
Flash, management-induced, and catastrophic flooding
(caused by landslide break-out floods and glacier hazards) can
all drastically alter the landscape and create flood disasters in
more populated areas. For example, in the spring of 2010 in
the Hunza region of Pakistan, a large landslide blocked the
Hunza River. Given the relatively large river discharge and the
increasing meltwater contributions by glaciers in the region,
the water impoundment grew at a significant rate. Multitemporal satellite imagery was used to monitor the progress of
the rapidly growing Lake Gojal (Figure 10). When combined
with topographic data and discharge information, the water
volume and spill-over time can be accurately predicted. It is
still slightly unclear as to the reasons why a catastrophic
breakout flood did not occur, although landslide experts
speculate that it is probably due to remobilized lake clays
imbedded with boulders greater than 1020 m, making an
ideal lake dam.
Given the complexity of assessing various types of natural
hazards, it is reasonable to assume that an integrated approach
involving the use of numerical modeling, field data and
mapping, and remote sensing may be required. Pelletier et al.
(2005) used this approach for flood-hazard assessment on
alluvial fans. Raster-based hydraulic modeling, satellite-image
change detection, field mapping of recent flood inundation,
and surficial geological mapping were used to characterize
specific spatial details that are lacking in standard GIS approaches. Model predictions of flood inundation and flow
depths were tested against field and satellite-based flood maps
for two extreme events. They were able to predict spatially
complex flood hazards that strongly reflect small-scale topographic and geologic conditions.
Geomorphologists typically address the issue of using remote sensing and GIS from a specific hazard susceptibility
perspective, although this is only the first phase or component
in a series of steps and analyses that are required for hazards
assessment and disaster risk management. Chapter 3.10 by van
Westin provides a comprehensive treatment of this topic,
highlighting the significance and difficulty of assessing specific
natural hazards, but also incorporating multihazard assessment as part of a system to understand and manage complex
cascading hazard influences. Specifically, van Westin describes
the importance of the integration of approaches for hazards
assessment and the need to conduct hazard and risk assessments at different scales, from global to community levels. van
Westin reveals that each scale or level of analysis has its own
objectives and spatial data requirements for hazard inventories, causal factors, and elements at-risk mapping and
database development. van Westin also addresses vulnerability
assessment approaches. Collectively, his treatment reveals that
very little research has been conducted on establishing
19
Lake Gojal Growth

12 Jan, 2010: SPOT
16 Mar, 2010: ALI
02 May, 2010: ASTER
25 May, 2010: ASTER
07 July, 2010: ALI
(a)
(b)
(c)
Figure 10 Development of Lake Gojal. (a) Prerockslide ASTER FCC image mosaic of the Hunza Valley. Red represents vegetation, mainly
agricultural fields associated with villages. (b) ASTER false-color image 4 day before spillover. Note extensive late spring snowfields and glaciers
feeding Lake Gojal. (c) Advanced Land Imager (ALI) near-true-color base image, 7 July 2010, showing the growth of Lake Gojal based on SPOT,
ALI, and ASTER. Reproduced from Kargel, J.S., Leonard, G.J., Crippen, R.E., Delaney, K.B., Evans, S.G., Schneider, J.S., 2010. Satellite
monitoring of Pakistans rockslide-dammed Lake Gojal. Eos 91(43), 394395, with permission from AGU.
complete GIS-based multihazard assessments, and that geomorphologists typically focus on evaluating individual hazard
types. Although geospatial technologies serve as a basis for
hazard and risk assessment, the scientific advances in hazards
and risk assessment have remained in the scientific community and have not effectively reached end-users such as
planner, managers, and policy-decision makers.
techniques. In the geosciences, visualization is used in a variety of ways, and Earth scientists need to be familiar with
visualization techniques suitable for working with specific
data sets and applications. The major functional uses can be
categorized as follows, although these are not necessarily
mutually exclusive.
Exploratory visual analysis (EVA). The goal is to explore the

3.1.4.7
Visualization
Visualization is an emerging science that represents research

developments in cartography, computer graphics, computational geometry, cognitive science, and psychology (Gahegan,
2000). It is used in a variety of ways to view imagery, integrate
information, view spatial and temporal patterns, and interactively examine abstract-data-spaces and numerical simulations. In the Earth sciences, many have recognized the
significance of visualizing and perceiving patterns of structure,
processes, features, and relationships, in an attempt to validate
analysis and modeling approaches and understand many
concepts in geomorphology regarding processes, landforms,
and system dynamics. With the advent of rapidly improving
computer and visualization technologies, a trend has emerged
involving more humancomputer interaction that exploits the
significance of the human visualization system.
In general, visualization is the interaction between computers and humans based on a graphical-user-interface environment that makes use of a variety of visualization
nature of the data such that spatial and temporal relationships and structure become apparent. This is analogous to
data mining or knowledge discovery and is closely related to
finding statistical anomalies in the data. Classic examples
include visualizations of scatterplots and spectral-feature
spaces in remote sensing. Such visualizations allow the exploration of spectral end-members that can be used in linear
and nonlinear spectral mixing to facilitate accurate mapping
of surface composition and assessing the composition mixture of materials that influence surface processes (e.g., surface energy budget, ablation).
A critical EVA example in geomorphology is visualizing
data validity and uncertainty. DTM and the accuracy of
DEMs is a critical first step toward meaningful terrain analysis. Three-dimensional perspective viewing of the landscape using anaglyphs or fly-by simulations can be
effectively utilized to detect high-frequency errors in DEMs
and low-frequency patterns caused by spatial interpolation
algorithms. Similarly, the use of pattern-recognition algorithms such as neural networks and fuzzy classifiers can be
used to generate fuzzy uncertainties or likelihoods of
20
accurate land-cover or surface matter classifications. These

approaches and quantitative measures communicate the
concepts of uncertainty, error, validity, or probability.
A geomorphometry example of EVA involves the exploration of scale dependence and prediction of spatial patterns
in understanding complex landscapes. Surface parameter
magnitudes such as relief, slope, and other important multiscale topographic parameters are dependent on the computational scale of analysis (Bishop et al., 2012).
Consequently, visualization of magnitude-scale relations
and distribution must be explored to determine the appropriate scale. Spatial scale dependence involves analysis and
display to determine the scale of spatial autocorrelation and
the directional fabric of the topography. Semivariograms can
be used to characterize scale dependence and the anisotropic
nature of the topography, although visualization is required
to appropriately assign semivariogram models to experimental variograms in order to view simulated spatial patterns (Webster and Oliver, 2007). Furthermore, visualization
is critical in an attempt to understand the anisotropic nature
of topography caused by various landscape evolution components, as erosion, deformation, rock strength, and faulting
govern the directional dependence. To date, this aspect of
EVA in geomorphometry is yet to be fully exploited, although the potential for new knowledge generation and new
theory exists.
Visual analysis and modeling. A strong empirical basis is
associated with the use of geospatial technologies for analysis and modeling (Bishop and Shroder, 2004b; Deng,
2007; Bishop et al., 2012). Whereas this provides for flexibility in developing metrics, software tools, and new algorithms and analysis approaches, it also raises important
questions concerning the validity of analysis and the use of
geomorphological information in integrative science.
Spatial analysis commonly relies on an index approach to
characterize various types of landscape information. These
indices or metrics are generally based on a association
with spatial position or a topographic parameter, although
they do not adequately characterize process mechanics,
scale dependencies, or temporal dynamics. Furthermore,
manipulations of these indices are based on concepts of
ranking, weighting, membership, prototypicality, scaling,
thresholding, heuristic rules, and additional empirical coefficients that can be used to alter the spatial extent of analysis results. Whereas these metrics have value in terms of
flexibility to address issues associated with semantic meanings, spatial uncertainty, subjective interpretations, data integration, classification, and variable definitions, the results
can be highly variable. Visualization is required to evaluate
the magnitude of index/metric values and determine whether or not such patterns actually represent reality (i.e.,
morphology, physical properties, genetics, dynamics, landforms). Although new spatio-temporal information can be
potentially generated in this way, spatial patterns and the
delineation of boundaries and zones must be carefully
examined to determine the sensitivity of combinations of
parameters. Consequently, visualization is required for adequate and accurate analysis and information production.
Finally, numerical modeling of climate, surface processes,
and tectonics can provide valuable insights into erosion,
relief production, feedback mechanisms, and the role of

surface processes in landscape evolution. Animations of
simulations can provide three-dimensional perspectives on
the nature of polygenetic evolution and the genesis of
landforms. Visualizations of the temporal dynamics of system-critical parameters such as precipitation, surface-sediment flux, influx of mass due to uplift, and topographic
constraints can be very useful for evaluating existing representations of process mechanics and dominant forcing factors. Consequently, visualization will play an ever-increasing
role in fundamental analysis and modeling efforts.
Visual Synthesis. Visualization approaches that allow an
evaluation of various types of thematic information can facilitate geomorphological studies. Classic examples include
examination of false-color composite images to facilitate
geological and land-cover mapping. Spatial overlay and the
use of symbols, color, transparency, and material properties
can be used to depict a variety of landscape conditions. In
essence, geomorphological mapping represents a visual synthesis of landscape conditions, as traditional and modern-day
geomorphological maps are commonly generated based on
the integration of information reflecting climatic, geological,
morphometric, pedogenic, land cover, and geochronological
variation. Bishop et al. (2012) provided a treatment of the use
of geospatial technologies for geomorphological mapping,
and indicate the role of visualizing and integrating a multitude of data types for assessing geomorphological conditions.
Given the complexity of polygenetic evolution and the general lack of standard protocols for information integration
and mapping, new techniques and cartographic approaches
seem warranted, as there are different integration requirements based on a variety of mapping perspectives.
Presentation. Numerous techniques and approaches can be
used for presenting information. These techniques and
products include traditional graphic products such as cross
sections, profiles, images, maps, animations, virtual globes,
and virtual realities. Effective communication and dissemination of geomorphological information to planners,
managers, and policy makers is a central theme. This aspect
of visualization is essential to promote and facilitate the use
of scientific information in the decision-making process.
Geomorphological research has been greatly facilitated by
rapid advances in geospatial technologies and augmented
realities. Chapter 3.11 by Smith and others provides a more
detailed examination of some fundamental aspects of visualization in the geosciences. Specifically, they provide examples
of techniques and approaches that allow various aspects of
geomorphology to be effectively studied. Important examples
involving remote sensing and terrain analysis are included.
The treatment clearly demonstrates that visualization techniques and approaches are valuable in Earth science
investigations.
3.1.5
Conclusions
The rapid advancement of geospatial technologies has had a

profound effect on the discipline of geomorphology. Remote
sensing and GIS studies are now commonplace in geomorphological investigations, as a better understanding of
surface processes and landscape evolution is sought. Earth
scientists are increasingly using new spatio-temporal datasets
and GIS technology for analyzing and modeling various aspects of geomorphological systems and for addressing conceptual and practical issues (i.e., scale dependency, processform and process-pattern relationships, digital geomorphological mapping, landscape evolution modeling, natural hazard modeling).
The 41st Binghamton symposium highlighted many of the
aforementioned developments in geomorphology related to
geospatial technologies. The delegates were amazed at the
rapid evolution of geospatial technologies and were interested
in the many new ways to study geomorphological systems and
facilitate practical-problem solving. Nevertheless, the delegates
realized the difficulty of effectively utilizing new data and
analysis/modeling approaches, because the effective use of
remote and GIS in geomorphology requires multidisciplinary
domain knowledge including radiation transfer, matter/energy
interactions, sensor-system characteristics, mathematical underpinnings of algorithms, analysis approaches, models, and
Earth science. It is not possible to push a button and obtain
the desired results, as expertise and time is required to effectively produce accurate information. It is also essential to recognize the inherently empirical nature associated with using
GIS in geomorphology (Bishop et al., 2012). Consequently, it
is now important for geomorphologists to take remote sensing
and GIS coursework.
Clearly, the incorporation of remote sensing and GIScience
investigations into geomorphology has greatly contributed to
the quantitative evolution in geomorphology, compared with
its more qualitative beginnings. Geospatial technologies have
revolutionized the way in which scientists study the Earth, and
quantitative analysis and modeling of the landscape offer
many new research opportunities. It should be kept in mind,
however, that qualitative information derived from the field
and via human interpretation of data can contribute greatly to
the interpretation of quantitative results. The inherent digital
and quantitative nature of modern-day analysis should be
carefully examined with respect to GIS-based empiricism and
the ability to produce repeatable results. Only through quantitative formalization of geomorphological concepts and theories can such repeatable results be expected, which will go a
long way toward establishing geomorphology as a required
component in integrated science. Advances in remote sensing
and GIScience contribute to this important goal.
References
Allen, T.R., Walsh, S.J., 1993. Characterizing multitemporal alpine snowmelt
patterns for ecological inferences. Photogrammetric Engineering and Remote
Sensing 59(10), 15211529.
Banerjee, P., Burgmann, R., 2002. Convergence across the northwest Himalaya from
GPS measurements. Geophysical Research Letters 29(13), 14.
Barros, A.P., Chiao, S., Lang, T.J., Burbank, D., Putkonen, J., 2006. From weather
to climate-seasonal and interannual variability of storms and implications for
erosion processes in the Himalaya. In: Willett, S.D., Hovius, N., Brandon, M.T.,
Fisher, D.M. (Eds.), Tectonics, Climate, and Landscape Evolution. No. 398 in
Special paper. Geological Society of America, Boulder, Colorado, pp. 1738.
21
Ben-Dor, E., Chabrillat, S., Dematte, J.A.M., Taylor, G.R., Hill, J., Whiting, M.L.,
Sommer, S., 2009. Using imaging sepctroscopy to study soil properties. Remote
Sensing of Environment 113, S38S55.
Beven, K.J., Moore, I.D. (Eds.), 1993. Terrain Analysis and Distributed Modelling in
Hydrology. Advances in Hydrological Processes. John Wiley and Sons,
Chichester.
Bilham, R., Larson, K., Freymueller, J.Project Idylhim Members, 1997. GPS
measurements of present-day convergence across the Nepal Himalaya. Nature
386, 6164.
Bishop, M.A., 2009. A generic classification for the morphological and
spatial complexity of volcanic (and other) landforms. Geomorphology 111,
104109.
Bishop, M.P., Barry, R.G., Bush, A.B.G., et al., 2004. Global land-ice measurements
from space (GLIMS): remote sensing and GIS investigations of the Earths
cryosphere. Geocarto International 19(2), 5784.
Bishop, M.P., Bonk, R., Kamp, Jr. U., Shroder, Jr. J.F., 2001. Terrain analysis
and data modeling for alpine glacier mapping. Polar Geography 24(4),
257276.
Bishop, M.P., Bush, A.B.G., Copland, L., Kamp, U., Owen, L.A., Seong, Y.B.,
Shroder, J.F., 2010. Climate change and mountain topographic evolution in the
Central Karakoram, Pakistan. Annals of the Association of American Geographers
100(4), 772793.
Bishop, M.P., James, L.A., Shroder, J.F., Walsh, S.J., 2012. Geospatial technologies
and digital geomorphological mapping: concepts, issues and research.
Geomorphology 136, 526.
Bishop, M.P., Shroder, Jr. J.F. (Eds.), 2004a. Geographic Information Science and
Mountain Geomorphology. Springer-Praxis, Chichester.
Bishop, M.P., Shroder, Jr. J.F. (Eds.), 2004b. GIScience and mountain
geomorphology: overview, feedbacks, and research directions. Springer-Praxis,
Bishop, M.P., Shroder, Jr. J.F., Colby, J.D., 2003. Remote sensing and
geomorphometry for studying relief production in high mountains.
Boochs, F., Kupfer, G., Dockter, K., Kuhbauch, W., 1990. Shape of the red edge as
vitality indicator for plants. International Journal of Remote Sensing 11,
17411753.
Boulton, S.J., Whittaker, A.C., 2009. Quantifying the slip rates, spatial distribution
and evolution of active normal faults from geomorphic analysis: field examples
from oblique-extensional graben, southern Turkey. Geomorphology 104,
299316.
Brandli, M., 1996. Hierarchical models for the definition and extraction of terrain
features. In: Burrough, P., Frank, A.U. (Eds.), Geographic Objects with
Indeterminate Boundaries. Taylor and Francis, London, pp. 257270.
Burbank, D.W., Blythe, A.E., Putkonen, J., et al., 2003. Decoupling of erosion and
precipitation in the Himalayas. Nature 426, 652655.
Burns, R.G., 1970. Mineralogical Application to Crystal Field Theory. Cambridge
University Press, Cambridge, UK.
Burrough, P., 1989. Fuzzy mathematical methods for soil survey and land
evaluation. Journal of Soil Science 40, 477492.
Burrough, P., 1996. Natural objects with indeterminate boundaries. In: Burrough, P.,
Frank, A.U. (Eds.), Geographic Objects with Indeterminate Boundaries. Taylor
and Francis, London, pp. 328.
Burrough, P., van Gaans, P., MacMillan, R., 2000. High-resolution landform
classification using fuzzy k-means. Fuzzy Sets and Systems 113(1), 3752.
Carrara, A., Pike, R.J., 2008. GIS technology and models for assessing landslide
hazard and risk. Geomorphology 94, 257260.
Chen, J.L., Wilson, C.R., Tapley, B.D., Blankenship, D.D., Ivins, E.R., 2007.
Pategonia Icefield melting observed by Gravity Recovery and Climate Experiment
(GRACE). Geophysical Research Letters 34, L22501.
Collins, W., Chang, S.H., Raines, G., Canney, F., Ashley, R., 1983. Airborne
biogeochemical mapping of hidden mineral deposits. Economic Geology 78,
737743.
Curtis, C.D., 1976. Chemistry of rock weathering: fundamental reactions and
controls. In: Derbyshire, E. (Ed.), Geomorphology and Climate. John Wiley and
Sons, New York, Chapter 2, pp. 2557.
Dematte, J.A.M., 2002. Characterization and discrimination of soils by their reflected
electromagnetic energy. Brazilian Journal of Agricultural Research 37,
14451458.
Deng, Y., 2007. New trends in digital terrain analysis: landform definition,
representation, and classification. Progress in Physical Geography 31, 405419.
Deng, Y., Wilson, J.P., 2008. Multi-scale and multi-criteria mapping of mountain
peaks as fuzzy entities. International Journal of Geographical Information Science
22(2), 205218.
22
Dozier, J., Green, R.O., Noline, A.W., Painter, T.H., 2009. Interpretation of snow
properties from imaging spectrometry. Remote Sensing of Environment 113,
S25S37.
Finlayson, D.P., Montgomery, D.R., 2003. Modeling large-scale fluvial erosion in
GIS. Geomorphology 53, 147164.
Finlayson, D.R., Montgomery, D.R., Hallet, B., 2002. Spatial coincidence of rapid
inferred erosion with young metamorphic massifs in the Himalayas. Geology 30,
219222.
Florinsky, I.V., 1998. Combined analysis of digital terrain models and remotely
sensed data in landscape investigations. Progress in Physical Geography 22(1),
3360.
Flowers, G.E., Roux, N., Pimentel, S., Schoof, C.G., 2011. Present dynamics and
future prognosis of a slowly surging glacier. The Cryosphere 5, 299313.
Frankl, A., Nyssen, J., Calvet, M., Heyse, I., 2010. Use of Digital Elevation Models
to understand and map glacial landforms-The case of the Canigou Massif
(Eastern Pyrenees, France). Geomorphology 115, 7889.
Gahegan, M., 2000. Visualization as a tool for geocomputation. In: Openshaw, S.,
Abrahart, R.J. (Eds.), GeoComputation. Taylor and Francis, London, pp.
253274.
Gao, J., Liu, Y., 2001. Applications of remote sensing, GIS and GPS in glaciology:
a review. Progress in Physical Geography 25(4), 520540.
Gessler, P., Pike, R., MacMillan, R.A., Hengl, T., Reuter, H.I., 2009. The future of
geomorphometry. In: Hengl, T., Reuter, H.I. (Eds.), Geomorphometry: Concepts,
Software, and Applications. Developments in Soil Science. Elsevier, Amsterdam,
The Netherlands, pp. 637651.
Gilchrist, A.R., Summerfield, M.A., Cockburn, H.A.P., 1994. Landscape dissection,
isostatic uplift, and the morphologic development of orogens. Geology 22,
963966.
Goetz, A.F.H., Rowan, L.C., 1981. Geologic remote sensing. Science 211,
781791.
Green, R.O., Painter, T.H., Roberts, D.A., Dozier, J., 2006. Measuring the expressed
abundance of the three phases of water with an imaging spectrometer over
melting snow. Water Resources Research 42, W10402.
Grove, C.I., Hook, S.L., Paylor, E.D., 1992. Laboratory reflectance spectra of 160
minerals, 0.4 to 2.5 micrometers. Technical Report 92-2, NASA-Jet Propulsion
Laboratory, Pasadena.
Gustavsson, M., Seijmonsbergen, A.C., Kolstrup, E., 2008. Structure and contents
of a new geomorphological GIS database linked to a geomorphological map
with an example from Liden, central Sweden. Geomorphology 95, 335349.
Guth, P., 2009. Geomorphometry in MicroDEM. In: Hengl, T., Reuter, H.I. (Eds.),
Geomorphometry: Concepts, Software, and Applications. Developments in Soil
Science. Elsevier, Amsterdam, The Netherlands, pp. 351366.
Hallet, B., 1979. A theoretical model of glacial abrasion. Journal of Glaciology 23,
3950.
Harbor, J., 1992. Numerical modeling of the development of U-shaped valleys by
glacial erosion. Geological Society of America Bulletin 104, 13641375.
Hardgrove, C., Moersch, J., Whisner, S., 2009. Thermal imaging of alluvial fans: a
new technique for remote classification of sedimentary features. Earth and
Planetary Science Letters 285, 124130.
Hengl, T., Reuter, H.I. (Eds.), 2009. Geomorphometry: Concepts, Software, and
Applications. Developments in Soil Science. Elsevier, Amsterdam, The
Netherlands, 33, 765 pp.
Hill, J., Megier, J., Mehl, W., 1995. Land degradation, soil erosion and
desertification monitoring in Mediterranean ecosystems. Remote Sensing
Reviews 12, 107130.
Hill, J., Schutt, B., 2000. Mapping complex patterns of erosion and stability in dry
Mediterranean ecosystems. Remote Sensing of Environment 74, 557569.
Hofierka, J., Mitasova, H., Neteler, M., 2009. Geomorphometry in GRASS GIS.
In: Hengl, T., Reuter, H.I. (Eds.), Geomorphometry: Concepts, Software, and
Applications. Developments in Soil Science. Elsevier, Amsterdam, The
Netherlands, pp. 387410.
Houze, R.A., Wilton, D.C., Smull, B.F., 2007. Monsoon convection in the Himalayan
region a seen by the TRMM Precipitation Radar. Quarterly Journal of Royal
Meteorological Society 133, 13891411.
Jamieson, S.S.R., Sinclair, H.D., Kirstein, L.A., Purves, R.S., 2004. Tectonic forcing
of longitudinal valleys in the Himalaya: morphological analysis of the Ladakh
Batholith, North India. Geomorphology 58, 4965.
Kaab, A., Vollmer, M., 2000. Surface geometry, thickness changes and flow fields
on creeping mountain permafrost: automatic extraction by digital image analysis.
Permafrost and Periglacial Processes 11, 315326.
Kargel, J.S., Leonard, G.J., Crippen, R.E., Delaney, K.B., Evans, S.G., Schneider,
J.S., 2010. Satellite monitoring of Pakistans rockslide-dammed Lake Gojal. Eos
91(43), 394395.
King, M.D., Menzel, P., Kaufman, Y.J., et al., 2003. Cloud and aerosol properties,
precipitable water, and profiles of temperature and water vapor from MODIS.
IEEE Transactions on Geoscience and Remote Sensing 41(2), 442458.
Klingseisen, B., Metternicht, G., Paulus, G., 2008. Geomorphometric landscape
analysis using a semi-automated GIS-approach. Geomorphology 23, 109121.
Koons, P.O., 1995. Modeling the topographic evolution of collisional belts. Annual
Review Earth and Planetary Science 23, 375408.
Koons, P.O., Zeitler, P.K., Chamberlain, C.P., Craw, D., Meltzer, A.S., 2002.
Mechanical links between erosion and metamorphism in Nanga Parbat, Pakistan
Himalaya. American Journal of Science 302, 749773.
Kruse, F.A., Lefkoff, A.B., Dietz, J.B., 1993. Expert system-based mineral mapping in
Northern Death Valley, California/Nevada, using the Airborne Visible/Infrared
Imaging Spectrometer (AVIRIS). Remote Sensing of Environment 44, 309336.
Lagacherie, P., Andrieux, P., Bouzigues, R., 1996. Fuzziness and uncertainty of soil
boundaries: from reality to coding in GIS. In: Burrough, P., Frank, A.U. (Eds.),
Geographic Objects with Indeterminate Boundaries. Taylor and Francis, London,
pp. 275286.
Lane, R., 2002. Ground and airborne electromagnetic methods. In: Papp E. (Ed.).
Geophysical and Remote Sensing Methods for Regolith Exploration. CRC LEME.
Open File Report, http://crcleme.org.au/Pubs/OPEN20FILE20REPORTS/
OFR20144/OFR144.html.
Liang, S., 2007. Recent developments in estimating land surface biophysical
variables from optical remote sensing. Progress in Physical Geography 31(5),
501516.
Lindsay, J.B., 2009. Geomorphometry in TAS GIS. In: Hengl, T., Reuter, H.I. (Eds.),
Luoto, M., Marmion, M., Hjort, J., 2010. Assessing spatial uncertainty in predictive
geomorphological mapping: a multi-modeling approach. Computers and
Geosciences 36, 355361.
Maathuis, T.H.B.H.P., Wang, L., 2009. Geomorphometry in ILWIS. In: Hengl, T.,
Reuter, H.I. (Eds.), Geomorphometry: Concepts, Software, and Applications.
Developments in Soil Science. Elsevier, Amsterdam, The Netherlands, pp.
309331.
MacGregor, K.R., Anderson, R.S., Anderson, S.P., Waddington, E.D., 2000.
Numerical simulations of glacial-valley longitudinal profile evolution. Geology
28, 10311034.
Maidment, D.R., 1993. GIS and hydrological modeling. In: Goodchild, M.F., Parks,
B.O., Steyaert, L.T. (Eds.), Environmental Modeling with GIS. Oxford University
Press, New York, pp. 147167.
Marcus, W.A., Legleiter, C.J., Aspinall, R.J., Boardman, J.W., Crabtree, R.L., 2003.
High spatial resolution, hyperspectral (HSRH) mapping of instream habitats,
depths, and woody debris in mountain streams. Geomorphology 55, 363380.
Masunaga, H., Iguchi, T., Oki, R., Kachi, M., 2002. Comparison of rainfall products
derived from TRMM Microwave Imager and Precipitation Radar. Journal of
Applied Meteorology 41, 849862.
Matsuo, K., Heki, K., 2010. Time-variable ice loss in Asia high mountains from
satellite gravimetry. Earth and Planetary Science Letters 290, 3036.
Metternicht, G.I., Zinck, J.A., 2003. Remote sensing of soil salinity: potentials and
constraints. Remote Sensing of Environment 85, 120.
Molnar, P., England, P., 1990. Late Cenozoic uplift of mountain ranges and global
climate change: chicken or egg? Nature 346, 2934.
Montgomery, D.R., Hallet, B., Yuping, L., Finnegan, N., Anders, A., Gillespie, A.,
Greenberg, H.M., 2004. Evidence for Holocene megafloods down the Tsangpo
River gorge, Southeastern Tibet. Quaternary Research 62, 201207.
Napieralski, J., Nalepa, N., 2010. The application of control charts to determine the
effect of grid cell size on landform morphometry. Computers and Geosciences
36, 222230.
Nesbitt, S.W., Anders, A.M., 2009. Very high resolution precipitation climotologies
from the Tropical Rainfall Measureing Mission precipitation radar. Geophysical
Research Letters 36, L15815.
Olaya, V., Conrad, O., 2009. Geomorphometry in SAGA. In: Hengl, T., Reuter, H.I.
(Eds.), Geomorphometry: Concepts, Software, and Applications. Developments in
Soil Science. Elsevier, Amsterdam, The Netherlands, pp. 293308.
Peckham, S.D., 2009. Geomorphometry in RiverTools. In: Hengl, T., Reuter, H.I.
(Eds.), Geomorphometry: Concepts, Software, and Applications. Developments in
Soil Science. Elsevier, Amsterdam, The Netherlands, pp. 411430.
Pelletier, J.D., Comeau, D., Kargel, J., 2010. Controls of glacial valley spacing on
Earth and Mars. Geomorphology 116, 189201.
Pelletier, J.D., Mayer, L., Pearthree, P.A., House, P.K., Demsey, K.A., Klawon, J.E.,
Vincent, K.R., 2005. An integrated approach to flood hazard assessment on
alluvial fans using numerical modeling, field mapping, and remote sensing.
Geological Society of America Bulletin 117(9/10), 11671180.
Pike, R.J., 1995. Geomorphometry progress, practise and prospect. Zeitschrift fur
Geomorphologie 101(7), 221238.
Pike, R.J., 2000. Geomorphometry Diversity in quantitative surface analysis.
Progress in Physical Geography 24(1), 120.
Ralston, B.A., 1994. Object oriented spatial analysis. In: Fotheringham, S.,
Rogerson, P. (Eds.), Spatial Analysis and GIS. Taylor and Francis, London, pp.
165185.
Reuter, H.I., Nelson, A., 2009. Geomorphometry in ESRI packages. In: Hengl, T.,
Reuter, H.I. (Eds.), Geomorphometry: Concepts, Software, and Applications.
Developments in Soil Science. Elsevier Amsterdam, The Netherlands, pp.
269291.
Rodell, M., Velicogna, I., Famiglietti, J.S., 2009. Satellite-based estimates of
groundwater depletion in India. Nature 460, 9991002.
Saadat, H., Bonnell, R., Sharifi, F., Mehuys, G., Namdar, M., Ale-Ebrahim, S., 2008.
Landform classification from a digital elevation model and satellite imagery.
Schaepman, M.E., Ustin, S.L., Plaza, A.J., Painter, T.H., 2009. Earth system science
related imaging spectroscopy- An assessment. Remote Sensing of Environment
113, S123S137.
Schaper, J., Martinec, J., Seidel, K., 1999. Distributed mapping of snow and
glaciers for improved runoff modelling. Hydrological Processes 13(12-13),
20232031.
Schmidt, J., Dikau, R., 1999. Extracting geomorphic attributes and objects from
digital elevation models - Semantics, methods, future needs. In: Dikau, R.,
Saurer, H. (Eds.), GIS for Earth Surface Systems: Analysis and Modelling of the
Natural Environment. Gebrder Borntraeger, Berlin, Stuttgart, pp. 153173.
Schneevoigt, N.J., van der Linden, S., Thamm, H., Schrott, L., 2008. Detecting
alpine landforms from remotely sensed imagery. A pilot study in the Bavarian
Alps. Geomorphology 93, 104119.
Sella, G.F., Stein, S., Dixon, T.H., Craymer, M., James, T.S., Mazzotti, S., Dokka,
R.K., 2007. Observation of glacial isostatic adjustment in stable North America
with GPS. Geophysical Research Letters 34, L02306.
Sharon, D., Morin, J., Moshe, Y., 1988. Micro-topographical variations of rainfall
incident on ridges of a cultivated field. Transactions of the American Society of
Agricultural Engineering 31, 17151722.
Short, N.M., Blair, R.W. (Eds.), 1986. Geomorphology From Space: A Global
Overview of Regional Landforms. NASA, Washington, DC.
Shroder, Jr. J.F., Bishop, M.P., 2000. Unroofing of the Nanga Parbat Himalaya.
In: Khan, M.A., Treloar, P.J., Searle, M.P., Jan, M.Q. (Eds.), Tectonics of the
Nanga Parbat Syntaxis and the Western Himalaya. No. 170 in Special
Publication. Geological Society, London, pp. 163179.
Shroder, Jr. J.F., Bishop, M.P., 2003. A perspective on computer modeling and
fieldwork. Geomorphology 53, 19.
Smith, G.R., Woodard, J.C., Heywood, D.I., Gibbard, P.L., 2000. Interpreting
Pleistocene glacial features from SPOT HRV data using fuzzy techniques.
Computers and Geosciences 26, 479490.
Smith, M.J., Pain, C.F., 2009. Applications of remote sensing in geomorphology.
Syed, T.H., Famiglietti, J.S., Chen, J., Rodell, M., Seneviratne, S.I., Viterbo, P.,
Wilson, C.R., 2005. Total basin discharge for the Amazon and Mississippi river
basins from GRACE and a land-atmosphere water balance. Geophysical Research
Letters 32, L24404.
Tapley, B., Bettapur, S., Ries, J., Thompson, P.F., Watkins, M., 2004. GRACE
measurements of mass variability in the Earth system. Science 305, 503505.
Tarolli, P., Arrowsmith, J.R., Vivoni, E.R., 2009. Understanding Earth surface
processes from remotely sensed digital terrain models. Geomorphology 113,
13.
Tate, J.N., Wood, J., 2001. Fractals and scale dependencies in topography. In: Tate,
J.N., Atkinson, P.M. (Eds.), Modelling Scale in Geographical Information
Science. John Wiley and Sons, Chichester, pp. 3552.
23
Taylor, G.R., 2004. Field and image spectrometry for soil mapping. Proceedings
of the 12th Australian Remote Sensing Conference. Fremantle, WA,
Australia.
Taylor, G.R., Bennett, B.A., Mah, A.H., Hewson, R.D., 1994. Spectral properties of
salinised land and implications for interpretation of 24 channel imaging
spectrometry. Proceedings of the First International Remote Sensing Conference
ane Exhibition. Vol. 3. Strasbourg, France.
Tomkin, J.H., Braun, J., 2002. The effect glaciation sas on the relief of a fast
growing orogen: a numerical modelling study. American Journal of Science 302,
169190.
Trudgill, S.T., 1976. Rock weathing and climate: quantitative and experimental
aspects. In: Derbyshire, E. (Ed.), Geomorphology and Climate. John Wiley and
Sons, New York, Chapter. 3, pp. 5999.
Tucker, G.E., Hancock, G.R., 2010. Modeling landscape evolution. Earth Surface
Processes and Landforms 35, 2850.
Usery, E.L., 1996. A conceptual framework and fuzzy set implementation for
geographic features. In: Burrough, P., Frank, A.U. (Eds.), Geographic Objects
with Indeterminate Boundaries. Taylor and Francis, London, pp. 7185.
van der Meer, F.D., de Jong, S.M. (Eds.), 2006. Imaging Spectrometry: Basic
Principles and Prospective Applications. Remote Sensing and Digital Image
Processing. Springer, Dordrecht.
Vaughan, R.G., Hook, S.J., Ramsey, M.S., Realmuto, V.J., Schneider, D.J., 2005.
Monitoring eruptive activity at Mount St. Helens with TIR image data.
Geophysical Research Letters 32, L19305.
Vitek, J.D., Giardino, J.R., Fitzgerald, J.W., 2008. Mapping geomorphology: a
journey from paper maps through computer mapping to GIS and virtual reality.
Wahr, J., Molenaar, M., Bryan, F., 1998. Time variability in the Earths gravity field:
Hydrological and oceanic effects and their possible detection using GRACE.
Journal of Geophysical Research 103, 30,20530,229.
Walsh, S.J., Moody, A., Allen, T.R., Brown, D.G., 1997. Scale dependence of NDVI
and its relationship to mountainous terrain. In: Quattrochi, D.A., Goodchild, M.F.
(Eds.), Scale in Remote Sensing and GIS. CRC/Lewis Publishers, Boca Raton,
pp. 2755.
Warner, T., Shank, M., 1997. An evaluation of the potential for fuzzy classification
of multispectral data using artificial neural networks. Photogrammetric
Engineering and Remote Sensing 63(11), 12851294.
Webster, R., Oliver, M.A., 2007. Geostatistics for Environmental Scientists. Statistics
in Practice. John Wiley and Sons Ltd., Chichester.
Whipple, K.X., 2009. The influence of climate on the tectonic evolution of mountain
belts. Nature Geoscience 2, 97104.
Wichmann, V., Heckmann, T., Haas, F., Becht, M., 2009. A new modeling approach
to delineate the spatial extent of alpine sediment cascades. Geomorphology 111,
7078.
Wilford, J., 2002. Airborne gammay-ray spectrometry. In: Papp, E. (Ed.),
Geophysical and Remote Sensing Methods for Regolith Exploration. CRC LEME.
Open File Report, http://crcleme.org.au/Pubs/OPEN20FILE20REPORTS/
OFR20144/OFR144.html.
Wilson, J.P., Gallant, J.C. (Eds.), 2000. Terrain Analysis: Principles and
Applications. John Wiley and Sons, New York.
Wood, J., 2009. Geomorphometry in LandSerf. In: Hengl, T., Reuter, H.I. (Eds.),
Wulf, H., Bookhagen, B., Scherler, D., 2010. Seasonal precipitation gradients and
their impact on fluvial sediment flux in the Northwest Himalaya. Geomorphology
118, 1321.
Zhang, W., Montgomery, D.R., 1994. Digital elevation model grid size, landscape
representation, and hydrological simulations. Water Resources Research 30,
10191028.
24
Biographical Sketch
Dr Michael P Bishop is a professor and Haynes Chair in Geosciences in the Department of Geography at Texas
A&M University. He received his PhD from the Indiana State University (1987) in physical geography with a focus
on geographic information science (GIScience). His areas of expertise are in remote sensing, geographic information systems (GIS), geomorphometry, numerical modeling, and mountain geomorphology. He has published
more than 30 articles in scientific journals, three books, and numerous book chapters on topics including
radiation transfer, image and terrain spatial analysis, surface processes and landforms, climate and glacier change,
and landscape evolution modeling. Furthermore, he has presented more than 200 national and international
professional papers on various remote sensing, GIScience, and mountain geomorphology-related topics. Financial
support for his research has been obtained through the National Geographic Society, National Science Foundation, NASA, DOE, USGS, and numerous foundations. His current research is focused on the use of satellite
imagery and terrain analysis for characterizing surface processes and features in complex mountain environments.
3.2 Ground, Aerial, and Satellite Photography for Geomorphology and

Geomorphic Change
JF Shroder, University of Nebraska at Omaha, Omaha, NE, USA
3.2.1
3.2.2
3.2.2.1
3.2.2.2
3.2.2.3
3.2.2.3.1
3.2.2.4
3.2.3
3.2.3.1
3.2.4
References
Introduction
Data Acquisition
Photographic Scale
Temporal Coverage
Digital Cameras and Videography
Gigapan technology
Thermal Imaging Technology and Geomorphology
Image Interpretation
Change Detection
Conclusions
Glossary
Change detection The use of repeat photography or
images acquired at different time periods to detect and map
changes on the landscape.
Colorimetry Science and technology used to describe
physically, the human perception of color.
Film-return satellite An early generation of orbital
satellites with onboard cameras whose film was ejected
periodically by parachute return vehicle to be collected and
developed.
Forward-looking, infrared radiometer (FLIR) A
portable device to obtain thermal or temperature-based
images of phenomena.
Gigapan technology Gigapixel-panorama technology
using digital cameras for an aggregated single image that
can be browsed and zoomed at multiple scales from
macroscale to microscale.
Megageomorphology Geomorphology of large regions
(regional geomorphology) enabled by space photography
or satellite images that cover broad areas.
26
26
29
30
31
32
33
33
34
38
40
Photogrammetry Practice of determining geometric

properties of objects from photographic images, in some
cases with dimensionally accurate and precisely rectified
aerial photography.
Spectrophotometry Quantitative measurement of the
spectral reflectance or transmittance of an object as a
function of wavelength.
Stereoscopy A technique for creating an illusion of depth
by presenting two offset photographs or images to the left
and right eye of the viewer.
Thermal imaging (thermography) Thermal imaging
cameras detect radiation in the thermal infrared region of
the electromagnetic spectrum and produce images of the
emitted radiation as thermograms.
Time-lapse photography Individual picture frames of
slowly moving or changing objects taken repeatedly from a
fixed position over some intermittent time period, and then
sequentially viewed to depict environmental change.
Videography The process of capturing sequential images
that depict object or landscape change.
Abstract
Historically, repeat coverage enabled change detection and study of landscape evolution. Today, digital cameras and
videography permit new capabilities in terms of producing ground, aerial, and space photography, as geomorphologists
still rely on spatial data collected from camera sensors that enable them to collect field data and document geomorphic
events and environmental change. Active geomorphological research now involves the acquisition of digital photography/
imagery, videography, and the use of historical photography, as historical data must serve as baseline data that is compared
to recently collected data for detecting and measuring geomorphic change. Digital camera technology has rapidly evolved,
Shroder, J.F., 2013. Ground, aerial, and satellite photography for

geomorphology and geomorphic change. In: Shroder, J. (Editor in Chief),
Bishop, M.P. (Ed.), Treatise on Geomorphology. Academic Press, San Diego,
CA, vol. 3, Remote Sensing and GIScience in Geomorphology, pp. 2542.
http://dx.doi.org/10.1016/B978-0-12-374739-6.00041-5
25
26
Ground, Aerial, and Satellite Photography for Geomorphology and Geomorphic Change
and scientists now use these sensors to record flooding, volcanic, mass movement, and other events. Movies and videos of
geomorphic phenomena, especially transient processes such as landslides, debris flows, or time-lapse glacial activities are an
essential research and teaching element. Finally, new gigapixel-panorama (gigapan) technology will provide new capabilities to assess multiscale geomorphic phenomena.
3.2.1
Introduction
Understanding surface processes and processform relationships, and conveying that understanding requires conceptual,
qualitative description, quantitative characterization, and
graphic depiction. Several centuries ago, graphic depiction
consisted solely of scientific illustrations, primarily artistic
field sketches of greater or lesser capability and complexity
(Merriam, 2009). Line drawings, sketches, or watercolors were
commonly used in the field, and a well-trained young scientist, particularly in Europe, normally had some formal education in drawing to assist their field work. This useful manner
of touristic, military, or scientific illustration was largely supplanted by widespread use of cameras by the early twentieth
century, and the ability to make skillful field sketches and
hand-drawn illustrations has been replaced with newer technologies, involving new platforms and sensor technologies for
acquisition of various forms of photography.
The idea of using a camera for field documentation,
however, is almost as old as the first cameras themselves. For
example, in 183842, F. Catherwood used camera-lucida
projections to make accurate drawings of Mayan ruins, while
also documenting their weathered condition with daguerreotypes only shortly after the invention of that camera-film type
(von Hagen, 1947). Malde (1973) noted that the camera sees
all of the intricate qualities of the terrain that might otherwise
escape notice and never make it into the field notebook. Such
things include weathering stains on outcrops, soil cracks and
small rills on bare ground, plant species and growth patterns,
and countless other features and environmental characteristics
that are impossible to specifically map and describe by any
other reasonable means.
The earliest known use of what is referred to by the military
as an overhead platform to observe the ground during the
French Revolution in the late eighteenth century was when
aerostiers, or balloonists took to the skies. In the nineteenth
century, military photographs from balloons in the American
Civil War were used to document enemy positions in landscapes, as well as to obtain high-angle views of the terrain. This
capability from the air has undergone various stages of development over time that have proven quite useful to geomorphologists. For example over a century ago, G. Lawrence devised
cameras weighing 4453 kg (1000 lb) that took panoramic
pictures as large as 1.4 m 2.4 m (4.5 ft 8 ft) from heights of
4600 m from balloons and kites. Of relevance to geomorphology were his panoramas aloft of the results of the great San
Francisco earthquake (Rosenberg et al., 1966). In the early
twentieth century, shortly after the invention of the airplane,
aerial cameras were beginning to be used for mapping, and
between World War I (WWI) and World War II (WWII),
aerial-survey techniques were extensively developed. Cameras in
airplanes were flown for reconnaissance before WWI, but geomorphologic studies with them were not done at that time. This
is thought in part to be because early conceptions of idealized
and hypothetical terrain conditions and preconceived notions

of geomorphic genesis through time were presented primarily
as artistic block diagrams (Lobeck, 1958; Hayden, 1986).
Between the two world wars of the twentieth century, an
enhanced development of aerial photography was undertaken
and the photogrammetric principles of aerial photography for
mapping were elucidated (Bagley, 1922; Lee, 1922; Reeves,
1927; Ashworth, 1937). Then with the advent of new aerial
photography for WWI, extensive development of the use of
aerial photographs occurred. Where direct overflights were
possible, vertical aerial photographs prevailed, in which
overlapping photographs along flight paths allowed stereography, wherein pseudo-three-dimensional (3-D) relief was
possible to visualize. Commonly however, low-oblique aerial
photographs (without the horizon), or high-oblique photos
(with the horizon) had to be taken from a distance, and then
projected into the vertical using the principles of photogrammetry in order to make a map. After the war, the systematic use of such postwar photography permitted detailed
topographic mapping (Smith, 1943), thereby allowing this
geomorphic research medium to achieve its apogee of technical formulation.
In the latter half of the twentieth century, the Cold-War
drive to command the high ground led to the race into space
by the USA and the Soviet Union. This endeavor resulted in
the development of a host of new technologies, including
photographic and imaging sensors aboard a multitude of
satellites.
Numerous sensing devices have moved from experimental
to effectively operational, and from aerial to space-based, or
from space-based to ground-based, such that new sensors for
producing digital photography and imagery are commonplace. Many of these technologies overlap considerably in
terms of data acquisition, manipulation, analysis, and geomorphic application. Therefore, the objective of this chapter is
provide examples of the use of historical photography, as well
as some newer technologies that are not be covered elsewhere
in the Treatise on Geomorphology.
3.2.2
Data Acquisition
The first known photographs were daguerreotypes produced

in 1839, and shortly thereafter the science of photogrammetry
started, although the term photogrammetry did not come
into common usage until the mid-twentieth century (Whitmore and Thompson, 1966). In the mid-nineteenth century,
photography began to be used to make topographic maps by
combining surveying theodolites with cameras. In the latter
half of the nineteenth century many early photographers of
the new (to them) lands of the western USA, such as W.H.
Jackson, A.J. Russell, and others, were led to document the
exciting new landscapes that were being discovered. These
professional photographs, and any number of more amateur
attempts up to the present day, constitute a collection of

photographs that are valuable for landform analysis as well as
constituting historical documents. Consequently, much effort
is being focused nowadays on electronic storage and computer
analysis to generate digital presentations and animations of
geomorphic change.
Cameras, of course, can be used in a variety of ways to
permit accurate delineation and measurement of geomorphologic phenomena. It is not the purpose of this chapter,
however, to relate the geometric or algebraic details of
photogrammetry to enable quantitative geomorphic measurements because other sources do that in detail (e.g.,
Thompson, 1966; Malde, 1973; Graf, 1985; Lane et al., 1993).
Nonetheless, it should be noted that major objectives of
ground-based geomorphologic research are characterizing,
measuring, and explaining surficial processes as they vary
spatially and temporally across landscapes. Measuring dimensions of landforms is a standard method to make such
observations, and the methods of photogrammetry using
ground-based and aerial cameras enable many useful measures to be obtained.
In an interesting new usage of photography in documentation for later geomorphologic interpretation, J.P. MacCalpin
has described what he terms as objective, 2-D photomosaic,
and 3-D photogrammetric logging of exposed trench faces (see
Chapter 14.12). These techniques (Coe et al., 1991; Fairer
et al., 1989) are designed to capture all the essential details of
sediment clast size, grain orientation, color, structure, and all
other fine information to document exactly what a subsurface
trench wall looks like so as to be best able to interpret past
events in the stratigraphy relating to the formation of the
landforms. Processes of ground rupture and sedimentation
attendant to past earthquakes, slope failures, or karst subsidence may be much better understood and explained in this
fashion in the investigation of the subsurface of fault scarps,
landslides, or sinkholes.
Beginning in the late 1940s, the desire to monitor developments inside the Soviet Union and its satellite countries to
prevent another Pearl Harbor surprise attack led to a plethora
of classified airborne and spaceborne platforms for photography of the Earths surface. The reconnaissance versions of
high-altitude bombers, especially the RB-26, RB-29, RB-50,
RB-47, RB-57, RB-58, and the later purpose-built reconnaissance aircraft the U-2 and the SR-71 allowed photography
during the Cold War, but little useful geomorphology is
known to have ever been obtained from them.
Before the declassification in 1995 of the historic photographs taken by the generations of spy satellites, the few
available photographs of the Earths terrain from space in the
early days were those taken by orbiting US astronauts. At first
in the manned space programs in the 1960s, handheld camera
photographs by the astronauts through porthole windows in
space capsules were only barely possible, and their use in
geomorphology has been rather limited. The four Mercury
orbital missions primarily used the Swedish-made precision
Hassleblad cameras, but it was only on the last two missions,
MA-8 and MA-9, that systematic terrain photography was a
formal experiment (Lowman, 1996). Astronauts W.M. Schirra
and L.G. Cooper took photographs that received widespread
publicity.
27
The follow-after Gemini program (Underwood, 1967;

Lowman and Tiedemann, 1971) had a Synoptic Terrain Photography Experiment (S005) that produced some 1300 usable
70 mm color pictures of Earths terrain (Figure 1). Some resulting Gemini photographs were published in Life, National
Geographic, and many newspapers to engage the publics
interest. It was only later that the progressive diminutions of
Lake Chad in Africa and the Aral Sea in Central Asia were
documented through repeat photography that the values of
such data were realized for change detection. Photography
obtained from the Gemini astronaut program was the stimulus to what ultimately became the Landsat series of fairly
standardized acquisition of satellite multispectral imagery. In
1966, the US Geological Survey had proposed an Earth Resources Observation Satellite (EROS), whose name was later
changed to Earth Resources Technology Satellite (ERTS), and
finally, Landsat. Early satellite imagery was low resolution that
showed only broad areas at small scale, but even so this enabled views of megageomorphology (wherein huge areas of
the Earths crust could be viewed collectively to gain understandings of regional geomorphology).
Well before astronaut photographs entered the public domain, several series of satellites were acquiring greater than
860 000 detailed photographs of many parts of the Earths
surface, but the photos were unavailable for general scientific or
geomorphologic use. After declassification by the ClintonGore
executive branch of government in 1995, geomorphologists
were able to access the treasure trove of available espionagesatellite data for the first time (Figure 2). In spite of the unavailability of some critical photogrammetric documentation
and deficient metadata that impedes some usages, the historical
data from widely dispersed parts of the world are an unusual
source of information that can be used in any number of
change detection or other studies in geomorphology.
Figure 1 Astronaut photograph from the Gemini 4 mission of 1966

showing the delta of the Nile River in the foreground and center, with
the Suez Canal area in the middle ground and the Red Sea, Gulf of
Aqaba, and Dead Sea in the background. Courtesy of US Government.
28
1,000
500
1.000 Meters
(a)
150 75
(e)
200
400 Meters
100
200 Meters
100
200 Meters
(b)
150 Meters
(c)
500
400
200
(d)
250
500 Meters
200
(f)
Figure 2 Periglacial geomorphological features on the Bykovsky Penninsula of the northeast Siberian coast identified from Corona photographs
of July 1969. Courtesy of US Government.
Table 1 Early generations of espionage satellites with film return by

ejected canister in the Keyhole program whose large number of
declassified photographs can be accessed from the US Geologic
Surveys EROS Data Center for change detection and other studies in
geomorphology. Each satellite number may represent multiple satellite
systems as 144 satellites alone were launched in the Corona program
Satellite number
Satellite name
Resolution (m)
System operation
KH-1
KH-3
KH-4
KH-4A
KH-4B
KH-7
KH-8
KH-9
Corona
Corona
Corona
Corona
Corona
Gambit
Gambit
Octagon
B7.5 m
B7.5 m
B7.5 m
B2.75 m
B1.8 m
B48 cm
B10 cm
B15 cm
196072
196367
196684
197186
The US National Reconnaissance Office (NRO) has operated satellites for the national intelligence community for
some time. Beginning in June of 1959 the NRO launched and
operated a number of Earth-observation satellites. Only the
early film-return satellites are discussed here (Table 1); imagery
from later satellites has not been declassified. In general the
film-return CORONA, ARGON, and LANYARD satellites were
in elliptical orbits ranging from a low of B280 km to a high of
B1006 km above the Earths surface. Each satellite passed over
its assigned observation target on the ground approximately
twice a day. At appropriate times, canisters of exposed films
were ejected over oceans, made reentry and deployed parachutes at lower altitudes, where most were successfully caught
by US aircraft, although some floated for a time until the US
Navy could intercept them. Others that were not recovered
quickly were designed to sink to deny them to other espionage
agencies. A large number (144) of satellites were launched
under the CORONA program and 102 returned usable data.
Many of the now-declassified scenes were acquired in stereographic mode from forward and aft-pointed cameras
(relative to the direction of motion of the satellites). Some
stereo viewing of the ground surface is possible, although
spatial distortions caused by forward camera motion, scan
time, and imaging of a bowtie shaped area of the ground
that was compressed into a rectangular image frame in the
Corona system makes such viewing rather difficult (Casana
and Cothren, 2008). Distortions in the Corona images can be
corrected using a rigorous model (e.g., Schenk et al., 2003;
Sohn et al., 2004), or a simple photogrammetric frame model
(Altmaier and Kany, 2002). The distortions can be further removed using ground-control points for geometrically correcting the data.
In practice, one might have expected such declassified data
to have been used much more in geomorphology. For example, in the Global Land Ice Measurements from Space
(GLIMS) project begun in the late 1990s that was designed to
monitor the world ice masses, researchers have used declassified satellite data to identify the terminus positions of
glaciers in the Hindu Kush and the Himalaya (Bishop et al.,
2004; Kargel et al., 2005), and stereoscopic mapping of
rockslide denudation there as well (Shroder et al., 2010). Rao
(2009) used declassified satellite data to map beach ridges on
29
the Godavari delta in India, and Grosse et al. (2005) were able
to map a wide variety of periglacial geomorphology (thermokarst depressions, lakes, and lagoons; thermoerosional
cirques and valleys; pingos) in northeast Siberia with the imagery (Figure 2).
The US Geological Survey maintains these images at its
EROS Data Center, in Sioux Falls, SD, and beginning in 2005
began distributing images scanned directly from the original
film strips at resolutions of up to 7 mm, which produces much
crisper images than the previous contact prints provided
(Casana and Cothren, 2008). Inasmuch as over 800 000 of
these images were acquired and the collection of Corona
material alone included 2.1 million ft of film in 39 000 cans,
the terrain information contained in these sources is enormous and can be used to facilitate geomorphological research.
Other official NASA campaigns of terrain photography
were also undertaken (Robinson et al., 2002). These include
data acquisition during the Earth-orbiting Apollo missions
(Colwell, 1971), the ApolloSoyuz mission (El-Baz, 1977;
El-Baz and Warner, 1979), Skylab (NASA, 1974; Wilmarth
et al., 1977), some Shuttle missions (Jones et al., 1996), and
the ShuttleMir missions (Evans et al., 2000). As of 30
September 1999, some 378 461 photo frames had been
included in the database (Office of Earth Sciences, 2000), but
approximately 50% have been deemed not useful for remote
sensing applications, which leaves some 190 911 possibly
suitable for Earth science related studies (Robinson et al.,
2002).
With the large number available, astronaut photographs
can be an excellent source of data for different studies, and the
best case resolutions are similar to Landsat or SPOT imagery
with pixel resolutions down to o10 m (Robinson et al., 2002).
As public domain information, costs are minimal to nothing,
and these images can be quite useful in filling in time-series
gaps where other imagery are not available. Access to the
complete and available database of astronaut photographs,
including low-spatial-resolution browse images, is available via
the web (Office of Earth Sciences, 2000). Such images posted
on the web are of low quality, but high-quality images can be
obtained on request for no cost (Robinson et al., 2002).
In recognition of the great advances made in visualizing
the geomorphology of the Earths surface from space, NASA
produced a large volume with copious pictures that represent
an atlas of regional landforms (Short and Blair, 1986). This
atlas has as its core, 237 color or black and white plates, each
of which consists of a space image with accompanying commentaries, explanations, together with three or four informative and enhancing aerial and/or ground photographs
(Lattman, 1987). Astronaut photographs, Landsat images,
radar, and thermal images are included as well. Twelve extensively illustrated chapters on regional, tectonic, and global
geomorphology are included, as well as detailed information
regarding process types and geomorphological mapping.
3.2.2.1
Photographic Scale
The mere existence of historical photography does not mean

that the data are suitable for specific geomorphological studies. In addition to the typical problems of cloud cover,
30
vegetation cover, and differential illumination causing

extensive shadowing, the issue of photographic scale must be
accounted for. This is especially the case for process-based
versus mapping-based studies, as photographic scale in relation to the spatial complexity of the landscape, and the
phenomena of interest, dictates photographic-scale requirements. In general, large-scale photography can be used for
detailed geomorphological mapping, although aerial coverage
is limited and requires the mosaicking of photographs. Medium scale photography addresses this issue as long as the
landforms or features can be detected and differentiated from
other objects. Regional geomorphological mapping can be
facilitated by medium- and small-scale photography although
the level of detail progressively worsens. These issues have
been extensively addressed in the past and readers are directed
to manuals and textbooks, especially those published by the
American Society for Photogrammetry and Remote Sensing.
3.2.2.2
Temporal Coverage
Two types of repeat photography or repeat imagery can be

recognized: (1) obtaining historical ground or aerial photographs for modern replication and (2) establishment of
photographic or videographic monitoring stations that are
meant to be replicated in the future for various purposes.
Sources of historical photography are varied and commonly
obscure enough that Graf (1985) recognized a convergence of
interest between the geomorphologist, historian, and even the
detective to ferret out locations of known historical photographs, as well as to find out whether photographs exist at all
for certain important areas. Government agencies are obvious
sources of a plethora of photographs, and local historical societies are another major source. Postcard collections in antique shops and the popular parlor stereopticon pictures of the
nineteenth century provide useful historical photographs as
well. A major problem in all searches for useful photographs to
use in rephotography studies is that indexes are generally not at
all relevant, and the search for photographs to use
in repeat photography forces the researcher to use visual
inspection of historical photographs, one photo at a time. In
some cases somewhat fortuitous photographs have given geomorphologists the ability to see long-term processes in action.
For example, Stephenson et al. (Chapter 10.11 in this volume)
and Shepard and Kuhn (1983) used historical photographs to
document the evolutionary development of coastal sea caves
on peninsulas into arches by wave erosion.
Historical photographs of any area in the world may or
may not be difficult to acquire, although by far the most difficult job once the photograph has been obtained, is to reestablish the original standpoint in the target photograph. In
some cases, only a reasonable approximation is possible,
whether because the original standpoint was on a boat on a
body of water, or was inundated subsequently, or is covered
with new or growing vegetation, or might be too hard to find
or to get to (A. Byers, oral communication, 2007; B. Molnia,
oral communication, 2009). In order to facilitate finding
difficult locations, Hanks (2006) for example, has engaged
in virtual repeat photography wherein digital aerial photographic data are draped over a digital elevation model (DEM)
to locate past camera stations that would have been difficult or

impossible to find otherwise.
In terms of temporal coverage where historical photographs were not available, R. McInnes of the Isle of Wight
consultancy, Coastal and Geotechnical Services in the UK,
compiled an array of more than 1000 sketches, prints, watercolors, and oil paintings of the Isle of Wight off the southern
coast of England and the adjacent Hampshire coast that
record the changes and landform evolution there since 1770
(Johnson, 2009a). He and his colleagues developed a ranking
system of the art based on up to five factors for accuracy determinations that enabled useful comparison for delineation
of change over time. In general, the watercolors of the nineteenth century were the most accurate because they were designed to record the actual landscape, especially as tourists or
military artists wanted to record information to explain their
travels and observations (Figure 3).
The capture of photographs of geomorphic processes in
action have been obtained over the years by people with
movie or video cameras at the right place and time in the
field, or in wind tunnel or flume experiments, and other such
laboratory manipulations (Table 2). For example, videos of
landslides have been taken, such as the famous quick-clay
failure in Rissa, Norway, in April 1978, as well as a wide
variety of other slope failures, or rapid wet debris flows
(Video 1) in various parts of the world (California, Japan,
China, Afghanistan, and Pakistan). Many of these videos are
available online through You Tube, Flickr, and other media
outlets. Other processes that have been captured using
videography include saltating sand grains, river meandering,
flow and flooding (Gough, 2007), action of waves in eroding
landforms and transporting coastal sediment, as well as a
wide variety of volcanic processes including catastrophic
eruptions, lava flows, formation of pillow lavas underwater,
and so forth.
Figure 3 Wash sketch of rocks displaced by mass movement and

an erosional residual pillar in Sinjao Nullah (gulley) a few km west of
Herat Afghanistan, painted in 1885 by Edward Law Durand of the
Afghan Frontier Commission during the border delineations of the
country. Such sketches could be quite accurate delineations of
landscapes in the 18th and 19th centuries when cameras were
unavailable or quite rare. Print purchased from India Office Library,
1979; identification India Office, British Library, Prints and Drawings,
shelfmark WD427. Copyright r The British Library Board.
31
Table 2 List of geomorphology videos that are available online at http://serc.carleton.edu/NAGTWorkshops/geomorph/visualizations.html

(accessed 10 February 2011). Each of these topics listed in the original website opens onto a plethora of animations, images, and videos that
show geomorphic processes at work in real field situations, or contrived laboratory ones
River Geomorphology Videos Videos of geomorphic process at work in rivers and river beds.
Chemical Weathering Animations and images dealing with chemical weathering.
Coastal Wave Mechanics (in ocean systems) Animations and movies depicting how waves and water molecules act along the shore and in deep
water settings.
Cryosphere Visualizations dealing with the cryosphere as a whole.
Deltas and Plumes Images and animations that illustrate both modern and ancient delta systems. Photos and 3-D animations depict the
development of deltas at various spatial and temporal scales.
Dunes: Process and Form Animations describing conditions necessary for producing form variations in sand dunes.
Examples of Deglaciation Animations of the deglaciation of North America and the removal of ice from the Bering Strait.
Glacial Landforms Resulting from Erosion and Deposition Animations presenting how depositional landforms like moraines and outwash
plains form.
Glacier Physics Animations, images, and movies revealing how a glacier forms, moves, retreats, and in the case of tidewater glaciers, calving.
Longshore Drift and Depositional Landforms Animations and images showing a variety of depositional landforms resulting from longshore drift
and the impact of river and coastal engineering projects on sediment transport processes.
Mass Wasting/Landslides Animations showing different types of landslides like slumps, slides, and falls in a variety of environments.
Mountain Uplift and Erosion (in structural geology) Visualizations illustrating the physical processes interacting to create mountain uplift and
erosion.
Physical Weathering Animations and images dealing with physical weathering.
Processes of River Erosion, Transport and Deposition Animations showing processes of river erosion, transport and deposition.
River Systems: Process and Form (in sedimentary geology) Visualizations and supporting material that can be used effectively to teach students
about physical processes acting in rivers and their floodplains.
Rocky Coastlines and Erosional Landforms Animations and images related to erosional landforms such as cliffs, arches, sea stacks and other
morphological features.
Soil Erosion Animations of soil erosion, images of soils, and audio files about the Dust Bowl.
Soil Horizons Animations illustrating the development of soil horizons and their characteristics.
Soil Orders Animations showing distinct soil orders.
Soils Physical Properties Animations about the physical properties of soil.
Waterfall Formation/Nick Point Migration Animations showing how waterfalls are created; examples of spectacular waterfalls from around the
world, and how falling water is used in hydroelectric power.
Figure 4 Extreme panoramic Gigpan picture of the north wall of Yosemite National Park. On the website, 14 snapshots can be selected and
then zoomed in and out from one to another, showing fine details, such as climbers on walls, details of the rock and water, individual people,
and other features. Interactive version available on http://www.xrez.com/, with permission from xRez Studio.
The online version of this chapter contains a Video with an

Animation 1. The online version can be found at doi:10.1016/
B978-0-12-374739-6.00041-5
Some largely imperceptible processes such as the slow ice
flow of glaciers, or terminus retreat and calving are best captured by time-lapse photography in which individual picture
frames are taken automatically once an hour or day, or some
other time period. These data can then be used together to
depict the movement of the phenomena that provides insights
into process and mechanism. Where this is done by film
production companies, they produce a product that provides
for a relatively seamless transition, and maintain lighting
conditions. In the real world of science given logistics, such
conditions are not usually obtained. Instead, cameras are set
to run automatically with batteries or solar power regardless of
lighting conditions or processes velocities, which generally
produces a jumpy, light-flickering, time-lapse series of images.

Even so, and depending on the results captured, this can
produce quite dramatic footage. For instance, the Extreme Ice
Survey (EIS) is an example of recent (B2008) expedition work
that Balog (2009) has done for glacier advance, retreat, and
extreme calving for glaciers in Alaska, Greenland, Iceland, the
Alps, and Andes, as part of the attempt to capture major
cryospheric-change visuals for public presentation.
3.2.2.3
Digital Cameras and Videography
The transition from film to digital photography is a collision

of interests and debate that goes deeply into arguments of
feature resolution with random silver halide crystals of various
sizes in film, versus pixel resolution and radiometric sensitivity
(Galer and Horvat, 2005; Rand et al., 2005). In addition,
32
digital sensor geometry is arranged in a rectangular grid pattern, which makes images susceptible to moire interferencepattern artifacts, whereas the random orientation of film
grains precludes this problem. In general, to approach 35 mm
film quality with digital cameras requires at least 6 million
pixels, but such an oversimplified statement needs to be
amended very much on what one wishes to do subsequently
with the acquired image (enlargement, publication, etc.). Part
of the issue is also the dynamic range one is interested in
capturing in the image, which is the amount of detail that can
be seen in the shadows before the bright areas start to lose
texture. In general one needs uncompressed file sizes 430 MB
to come close to film when making enlargements. At the
outset, many digital cameras were not as capable as film for
capturing the level of detail that geomorphologists generally
needed, although that is no longer the case. A number of
digital cameras now exhibit a huge dynamic range, compared
to either print or slide film, and film production companies
are gradually ceasing to produce film for use anymore, as it
becomes progressively more obsolete.
In addition, the line between still and video digital cameras has become blurred, if not quite nonexistent, and digital
camera technology is one of explosive growth and sudden
collapse. For example, the once quite popular Flip video
cameras, introduced in the spring of 2006, ended production
in spring of 2011, as still cameras increased onboard storage
to the point that significant lengths of video imagery could
be recorded. Cameras suitable for ground-based acquisition
of instructional or simple photographs are in many smart
phones and tablet computers. A profusion of more complex
viewfinder or digital single-lens reflex cameras suitable for
very detailed studies are available for less than $1500, and
prices are continuing to fall. Photogrammetric quality cameras, being more complex and less popular, are rarer and more
expensive. Because of the variety of new cameras available in
this growing field, expert advice should be sought before
investing in a new digital camera for collecting field data.
3.2.2.3.1
(a)
Gigapan technology
Gigapan or gigapixel-panorama technology is a new system

using digital cameras for acquiring composite pictures composed of billions of picture elements (pixels). The system
produces extremely high-resolution panoramic images that are
capable of resolutions from macropanoramas from a considerable distance away, zooming down without leaving the
scene to microresolutions of rock thin sections (Johnson,
2009b). The goal of gigapan technology is to facilitate acquisition and presentation of data as a single image that can be
stored efficiently on the web, and that can be examined and
displayed at multiple scales. The system is established as single
or multiple robotic cameras set to full zoom with automated
image acquisition of multiple individual pictures arranged
across a grid. The motorized camera system then automatically
moves, taking hundreds to thousands of slightly overlapping
scenes until an entire scene or viewscape is captured. Such
images are then downloaded to a personal computer and
software stitches the individual pictures into one large
explorable, gigapixel-sized, super-image, or gigapan.
Numerous examples of gigapan use in geomorphology are
available, from full-gigapixel anaglyph (red/blue 3-D images)
(b)
Figure 5 Photographs of the following: (a) Russias Klyuchevskoy

volcano showing a thermal image (upper), and a photograph of the
volcano (lower image. Reproduced from http://
news.nationalgeographic.com/news/bigphotos/3626065.html). (b)
Italys Stomboli Island volcano showing a thermal image of openings
of a new fracture vent in the middle of a flowfield along the Sciara
del Fuoco, which is the hillside from which the lava flows to the sea.
A curtain of smoke and gas emissions obscured the view in
visible-light but not this infrared.
33
Table 3 List of geomorphology picture libraries that are available online with numerous photographs available of landforms and geomorphic
processes. Many of these photographs are free (F) but some require payment ($)
Lisa Wells Geomorphology Images http://geoimages.berkeley.edu/GeoImages/Wells/wells.html
Geomorphology from Space (NASA) http://disc.sci.gsfc.nasa.gov/geomorphology
Gallery of Landform Images http://www.geomorph.org/gal/mslattery/world.html
Ediafocus Stock Photos (Geomorphology) http://www.mediafocus.com/image-search/geomorphology-stock-photos.html
About.Com: Geology (Landform Picture Gallery) http://geology.about.com/library/bl/images/bllandformindex.htm
Overview of Geomorphology Sites http://www.falw.vu/Bbalr/geomorphology.htm
of sea stacks and uplifted wave-cut terraces in Sonoma County,

CA (Johnson, 2009b) that can be viewed stereoscopically, to
the Extreme Resolution Panoramic Imaging Project (ERPIP) in
Yosemite National Park, where comprehensive assessment and
mapping of rockfall hazards was required (Figure 4). Because
the Yosemite Valley undergoes numerous large rock falls every
year, with 4600 recorded since 1850, evaluating, mapping,
and quantifying the geomorphic hazard was deemed a major
goal of the ERPIP activities. Thus high-resolution imagery of
the precipitous valley walls was needed to establish a baseline
datum for before-and-after comparisons. With establishment
of a collaboration between the National Park Service and xRez
Studios of Los Angeles, 410 000 concurrent images of the
425 km of granite walls in Yosemite were obtained at the
same time on 9 May 2010 to ensure the same lighting and
atmospheric conditions.
The basic technology included ground-based, gigapixel
panoramic photography, light detecting and ranging (LiDAR)
data acquisition from the air (plane and hang glider), and 3-D
computer rendering. Twenty photo-shooting teams totaling 70
photographers between them ascended a total of B11 000
vertical meters of trail. At each vantage point 4500 overlapping shots were obtained, which when merged and printed at
magazine-quality, 300 dpi resolution, the photos extended
uninterrupted for 412 m (Madrigal, 2008). All 20 gigapixel
panoramas were projected onto a 1-m resolution DEM using
Maya 3-D animation software, unifying the 25 km of Yosemite
walls into two single vertical orthographic (undistorted) views,
which yielded a unique, nonperspective elevational view of
the valley walls, which is a first in landscape photography.
In other future studies, obtaining such high-resolution
imagery can be used to establish baseline conditions for before-and-after comparisons of geomorphic phenomena.
Evaluation, mapping, and quantification of process rates,
geomorphic hazard, and other useful measures of geomorphic
change could be obtained in this way.
3.2.2.4
Thermal Imaging Technology and Geomorphology
A thermal imaging camera (TIC) records emitted thermal radiation and can be used to assess thermal properties and
surface temperature, which is a useful for monitoring volcanic
activity, especially in remote areas from the air, or where
dangerous eruptions threaten and some distance needs to be
maintained. Low resolution (1 km pixel 1) satellites such as
the Advanced Spaceborne Thermal Emission and reflection
Radiometer (ASTER) on board the Terra satellite produce
thermal imagery that can be compared to portable thermal
(F)
(F)
(F)
($)
(F)
(F)
camera systems. An example of a portable system is the forward-looking infrared radiometer (FLIR) that can then be used
for generating high-resolution thermal images, such as from a
helicopter hovering near a volcano (Figure 5), or a fixed
camera position to record thermal change through time as
lavas are emplaced (Video 2).
B978-0-12-374739-6.00041-5
Such instruments enable scientists to assess volcanic hazards more accurately because they can not only be used to
assess surface temperatures, but they can also measure the
presence of phenomena in the sky above the ground, such as
ash or steam. From such data, calculations of emitted volumes
of gas, rock, ice, water, ash, sulfur dioxide, and lava are possible. For example, in this sort of work on Erebus volcano in
Antarctica in 2004, radiative heat outputs calculated for the
Ray (B1400 m2) and Werner (B10001200 m2) lava lakes
were 3035 and 20 MW, respectively (Calkins et al., 2008).
The estimated magma flux necessary to sustain the combined
heat loss was B250710 kg s 1, the minimum volume of the
magma reservoir underground was B2 km3, and the radius of
the conduit feeding the Ray Lava Lake was B2 m.
3.2.3
Image Interpretation
Human interpretation of photography and imagery is a

standard approach for information extraction that has been
used for teaching and research for almost as long as the discipline of geomorphology has been practiced. Photo interpretation is based on the fundamental photographic or image
information elements that include the following: (1) tone, or
the relative brightness and color of landforms, soils and other
surface features that make up the landscape; (2) texture, the
local spatial variability in tone caused by variations in the
surface structure or morphology of the landscape or feature;
(3) size, both absolute and relative, that may be important
with respect to classification; (4) shape, which is related to
morphology and may be distinctive for specific landform
features; (5) shadow, that reflects the shape and the nature of
the object and topographic relief; (6) pattern, in which the
spatial arrangements of landforms may be significant with
respect to repetition, layout, and ordering; and (7) site, the
topographic position of landforms on the landscape (Lillesand and Kiefer, 1994).
Collectively these elements constitute the basic information that can be extracted and synthesized via human
34
visualization and analytical reasoning. When combined with

a-priori experience, geographic-, and discipline-domain
knowledge, effective analysis and information can be produced. Computer-assisted analysis that attempts to mimic the
human interpretation process and information production
can be utilized with digital photography, although quantitative formalization of human interpretation still remains an
active research area.
Human interpretation of the landscape and landforms is
relatively more accurate under complex landscape scenarios,
although great progress is being made in quantitative landform analysis and mapping from digital data (Bishop et al.,
2011).
A number of web sites provide pictures of landscape and
landforms that can be used most effectively in education
(Table 3). Commonly ignored as an important means of
doing research or making effective presentations, the ability to
obtain field photography and images, and interpreting field
data is an important skill that needs to be thought about
carefully involving a fair degree of competence. This skill is not
generally taught in graduate school or field camps, although it
is an essential part of any project in geomorphology. Practical
exercises in such landform analysis occur in some older texts,
such as Miller (1961), Wanless (1986), and Way (1973), and
the aforementioned manuals from the American Society of
Photogrammetry and Remote Sensing. The Remote Sensing
Tutorial, by Dr. Nicholas Short, contains several sections of
interest to geomorphologists desiring further instruction in
image interpretation, and is available at multiple web sites.
3.2.3.1
Change Detection
The elementary, although significant aspect of acquiring repeat

photographs and images from marked locations seems to have
been first used in Europe in the 1880s to monitor glacier
changes, and then in 1896 in the USA by Israel Russell (1898),
who used the technique to monitor and measure the terminus
of the Nisqually Glacier on Mount Rainier. After a hiatus of
approximately five decades, the technique of change detection
using photography was finally taken up again with more attention and utilized by a number of people looking for other
geologic change (Shepard and Grant, 1947; Phillips, 1963;
Veatch, 1969; Baars and Molenaar, 1971). This flurry of
interest by a number of people led to more general recognition
of the potential of repeat photography to provide important
new information, with the result that the United States Geological Survey (USGS) decided to utilize its vast historical
collections of photographs to accomplish major efforts in
change detection by trying to reoccupy many remote locations. The early expeditions by the USGS into the western
USA in the 1860s and 1870s had resulted in thousands of
photographs by such renowned photographers as W.H. Jackson, T. OSullivan, J.K. Hillers, and A.J. Russell that were envisioned in the USGS as ripe for rephotography. In the late
1970s then, the USGS began a second phase of taking new
photographs at the same date and time of day in the same
places as an experiment in repeat photography. A third phase
of this repeat photography began in 1997 (Klett, 2004)
(Figures 6 and 7).
(a)
(b)
(c)
Figure 6 Tufa knobs, Pyramid Lake, NV: (a) Photograph by T.

OSullivan for the USGS, 1867. (b) Photograph by M. Klett for the
Rephotographic Survey Project, 1979. (c) Photograph by M. Klett and B.
Wolfe for the Third View Project, 2000. Such tufa deposits are a
characteristic of many playa lakes that were active in the late Pleistocene
and some of the Holocene in the western USA but the progressive
abstraction of ever more water from these sites for irrigation and other
human uses indicates their progressive degradation thereafter. Such
photographic documentation enables measurement of former water
depths and potential erosional and depositional effects above and below
the former water line through time. Courtesy of US Government.
(a)
(b)
35
(c)
Figure 7 Eroded sandstones in Monument Park area of Colorado Springs, CO: (a) Photograph by W.H. Jackson for the USGS, 1873. (b) Photograph
by J. Verburg for the Rephotographic Survey Project, 1977. (c) Photograph by M. Klett, B. Wolfe, T. Ueshina, K. Bajakian for the Third View Project,
1997. The prominent changes due to rock erosion between scenes allow some calculation of episodic erosion rate. Courtesy of US Government.
In other cases, commercial interests in developing the

American West also eventually provided opportunity for repeat photography, as when in 188990 R.B. Stanton took
photographs every few km along the bottom of the Grand
Canyon in connection with a railroad that was planned close
to the water level but that was never constructed. A century
later, Webb (1996), a USGS hydrologist investigating debris
flows in the Canyon, was able to replicate all 445 of Stantons
photographs. As with so many rephotography projects, the
changes in vegetation were most obvious, but also apparent
were changes exerted by the newer Glen Canyon Dam
upstream so that eroded sand bars, aggraded debris fans, and
new marshes have resulted from this anthropogenic
geomorphic interference with the natural river flows. The experience gained in this project enabled Webb et al. (2010) to
write a text on the history, techniques, and applications of
repeat photography that other geomorphologists may find
quite useful.
Similarly, P. Bierman set about to replicate as many old
photographs in the state of Vermont as he could find, with the
result that he was able to document major landscape change
in the state, especially as the clear-cutting of old growth forest
in the nineteenth century led to extensive soil erosion, mass
movement, debris-fan growth, and flooding exacerbated by
denuded watersheds. Subsequently in the twentieth century,
as farm fields were abandoned wholesale back to secondgrowth forests, dirt roads were paved, and major new interstate
routes were constructed, the anthropogenic forcings of geomorphic change resulted in major changes in sedimentation
and slope and river-valley changes. White and Hart (2007) did
a similar rephotography project for the Canadian Rockies,
which was dominantly about vegetation change, but also included change in some glaciers and other landform features.
Some landscapes in the Himalaya, such as those documented by R. Finsterwalder in the 1930s on Nanga Parbat,
were so well photographed as major parts of their large-scale

photogrammetry-based mapping, that it became desirable to
see what changes have occurred there in the intervening decades. With the exception of the later German (Schmidt and
Nusser, 2009) and other rephotographic work on Nanga
Parbat (Figure 8) in the western Himalaya, however, rephotography has been predominantly confined to the eastern
Himalaya where so many iconic mountains, valleys, and glaciers occur. For example in the 1950s, several Austrian and
Swiss scientists traversed the terrain in Nepal thoroughly and
made extensive photographic records that A. Byers rephotographed in 2007 (Figure 9). In the western Himalaya, however, many of the superb high-altitude panoramas of the
Karakoram Himalaya taken by Sella (De Filippi, 1912) on the
expedition with the Duke of Abruzzi in 1909, remain to be
rephotographed, although Diolaiuti et al. (2003) have
reproduced some.
In other areas multitemporal photographs of glaciers have
proven useful in geomorphologic change detection. This includes the glaciers of Alaska by B. Molnia after many other
original photographers (Figure 10), or in a more restricted area
such as Glacier National Park those by D. Fagre after many
other photographers (Figure 11). Many other areas of interest
make ground-based, repeat photography very useful to understand geomorphic (and ecologic) change through time, especially with the rapidly changing and labile ice masses of the
world. Although commonly hard to do, ancillary quantitative
measurements from the photographs are possible as well.
Repeat aerial photography has also allowed a number of
special geomorphologic applications. For instance, geomorphic
change can be captured and better understood by limited
extrapolation between successive scenes, as for example,
where a sequence of migrating river meanders can reasonably
be constructed between scenes and linked into an
animation Video 3 (Gough, 2007). In another example, an
36
(a)
1934
(b)
1985
(c)
1991
(d)
1993
(e)
1994
(f)
1995
(g)
1997
(h)
2006
Figure 8 Pictures of the terminus of Raikot (Rakhiot) Glacier on the north face of Nanga Parbat, western Himalaya, Pakistan that were taken
from the same standpoint at the northeast end of Fairy Meadows. In part after Schmidt, S., Nusser, M., 2009. Fluctuations of Raikot Glacier
during the last 70 years a case study from the Nanga Parbat massif, northern Pakistan. Journal of Glaciology 55(194), 949959. (a)
Photograph by R. Finsterwalder, 1934. (b) Photograph by J. Gardner, 1985. (c) Photogaph by J. Shroder in 1992. (d) Photograph by M. Nusser
in 1994. (e) Photogaph by J. Shroder in 1996. (f) Photograph by M. Nusser in 2006. Such variation allows determination of the highly labile
nature of glacier geomorphology.
(a)
Late 1950s
(b)
2007
37
Figure 9 Taboche (6367 m) in center, Jobo Lapstan, the peak leaning to the right, and Khumbu Valley from the east near Mount Everest
(Sagamatha) in the eastern Himalaya of Nepal: (a) ca. 1955 photograph by Erwin Schneider, courtesy of the Association for Comparative Alpine
Research, Munich. Archives of Alton C. Byers, The Mountain Institute. (b) Photograph by A. Byers, The Mountain Institute, 2007. The clean,
debris-free ice below the summit of Taboche has been reduced considerably, as have the size of many small glaciers. Tsholo Tso is the morainedammed lake at the base of Jobo Lapstan and the moraine dam can be seen to have been more active half a century ago, whereas the lake was
higher in 2007.
intermittently active landslide in the Slovakian Carpathian

mountains was photographed at large scale from the air four
times at Bdecadal intervals and used to construct a detailed
DEM that enabled calculation of the net vertical and horizontal, mass-flux volumes through the system Figure 12
(Prokesova et al., 2010). Similarly Chandler (1989) and
Chandler and Brunsden (2006) applied analytical photogrammetry to archival photographs of five epochs of movement between 1946 and 1988 of the Black Venn landslide in
Dorset, England (Figure 13). Digital terrain models were produced to assess slope changes through time.
B978-0-12-374739-6.00041-5
In an unusual application of repeat photography to geomorphic process, albeit of an anthropogenic nature, the
Limestone Project in Oxford, England, was designed to assess
the catastrophic decay of building limestone with a wide
variety of tools (Meneely et al., 2008). Chief among these
techniques is rephotography (Thornbush and Viles, 2005,
2007; Thornbush, 2010) wherein Adobe PhotoshopTM was
used as image-processing software to obtain histogram-based
measurements of soiling and weathering of the building stone.
Results showed that photographs could be used to measure
and quantify change using integrated digital photos and
image-processing techniques calibrated by photographic incorporation of a grayscale and spectrophotometry. In addition

to the digital photography used in these analyses, comprehensive studies were also done using colorimetry, permeability, ground penetrating radar, thermography, X-ray
fluorescence, and stone-condition surveys. Such detailed analyses devoted to buildings can also be applied to landforms in
the natural rather than anthropogenic world.
Photomonitoring of sites is thus increasingly recognized as
scientifically advantageous, especially in vegetation studies,
range management, or ecological analysis (Rasmusson and
Voth, 2001), and assessment of geomorphological conditions
could be done in more locations. Equipment required to get
started is minimal, and includes a camera, film or camera
computer chips, a photo-information board, a reference-scale
pole, evaluation forms, and a notebook or field computer in
which to record the relevant information. The reference pole,
which is particularly important in the background of the
photograph for range managers to assess long-term health of
vegetation, would be less essential for many geomorphologic
situations, but might be relevant in gauging such things as
ripple and dune size and character, badland topography,
erosional scarps, depositional moraines, and such.
Most essential in setting up the picture for long-term
monitoring is to make absolutely sure that a distinctive and
38
(a)
(b)
1941
1950
(a)
1913
(b)
2005
Figure 11 Shepard Glacier in Glacier National Park, MT: (a)

Photograph in 1913 by W. Alden. (b) Photograph by B. Reardon. The
highly labile nature of glacier ice is obvious.
(c)
2004
Figure 10 Muir and Riggs glaciers, Glacier Bay National Monument,

AL: (a) Photograph in August 1941 after nearly two centuries of
retreat. (b) Photograph in August 1950. (c) Photograph in August
2004 by B. Molnia. Muir Glacier has retreated out of the field of view
and is now nearly 8 km to the northwest. Riggs Glacier is still in view
but has retreated as much as 610 m and thinned or downwasted as
much as 250 m.
permanent landmark occurs in the background or on the

skyline. One must make sure that the camera frame includes
the skyline, and that there are distinctive rock outcrops,
mountain slopes, or other geologic or geomorphic features
that will remain over long periods of human history so that
the site can be found again and again. A secondary, but also
important feature is the photoboard set up in the foreground
that has the data and location of the monitoring site. Both the
photoboard and the skyline or landmarks need to be quite
visible in the photograph. Finally, a plan for preservation and

long-term storage of the photographic images and written
records is critical if they are to ever be of real scientific use to
anyone in the future. This can be a problem if a geomorphologist has not had time to put things in order before retirement, or who has left inadequate documentation at death.
Special efforts should be exerted by older geomorphologists
who have had well published careers to see that their field
photographs and mapped materials are not lost to science
after their retirement. Film slides and photographs should
be digitized at high resolution and stored in well indexed
electronic media.
3.2.4
Conclusions
The uses of photography as a primary data-collection and

information-recording tool, as well as a device for obtaining
some quantitative information about landforms and processes, has been established in geomorphology for some time.
These qualitative as well as quantitative procedures can be
used to aid geomorphologists in studying landscape dynamics
and for geomorphological mapping. Most geomorphology
papers use photographs of surface processes or landforms as a
means to illustrate important aspects of the landscape or vital
concepts.
19771969
39
7.0 5.0 3.0 1.7 +1.7 +3.0 +5.0 +7.0 m
[m a.s.l]
650
445
0
200
400
600
800
1000
Landslide boundary
Obscured areas
1200
1400
[m]
Figure 12 Differential image of the Lubietova landslide in the western Carpathian Mountains of Slovakia that is draped over a 1977 DEM that
shows the removal (blue) and accumulation of transported slope debris (red) measured from four sets of stereographic aerial photographs
during the 196977 period. Reproduced from Prokesova, R., Kardos, M., Medvedova, A., 2010. Landslide dynamics from high-resolution aerial
photographs: a case study from the Western Carpathians, Slovakia. Geomorphology 115, 90101.
25
1988
1976
1969
1958
1946
Percentage
20
15
10
0
0
10
20
30
40
50
60
70
Interval 5 degrees
Figure 13 Percentages of changing slope angles of the Black Venn landslide measured from aerial photographs acquired in 1946, 1958, 1969,
1976, and 1988. Reproduced from Chandler, J.H., 1989. The acquisition of spatial data from archival photographs and their application to
geomorphology. Ph.D. thesis, The City University, London, England, 300 pp., unpublished, and Lane, S.N., Richards, K.S., Chandler, J.H., 1993.
Developments in photogrammetry; the geomorphological potential. Progress in Physical Geography 17(3), 306328, with permission from Sage.
40
The great diversity of general photographs of the surface of

Earth is a vast collection scattered in various places that has
only been accessed and used in a limited way. Thus the discipline of geomorphology over the past century has produced
large collections of still photographs of landforms, as well as
hosts of movies and videos that show geomorphic processes
in action. On top of the plethora of ground photographs, the
huge amounts of aerial and space-based photos taken
worldwide in the past 40 years are also tremendous sources
that could be processed and stored in databases so that they
could be used better than they have been. Stereographic aerial
photographs are a long-term staple of geomorphologic research, and with the addition of photos from astronauts and
film-return satellites, the potential data sources have grown
ever larger over time. A problem is that with the millions of
potentially rewarding and available historical photographic
materials, a certain amount of time and energy is necessary to
access, review, and store the scattered sources. This effort can
put off busy people so that they do not make effective use of
the materials of which they are unaware, or for which they
lack the requisite skills to access effectively. The addition of
new LiDAR and gigapan technologies and thermal sensor
technologies add to the photographic disciplinary diversity
as well.
New research using photographic techniques could certainly include greater use of the enormous store of Corona
film returned from space, as well as the many historical
photographs preserved unused in the many historical societies
in towns and cities all across the world where such photo data
are kept. Any number of quantitative or qualitative changedetection studies remain to be done with permafrost, glaciers,
mass movement, fluvial, aeolian or coastal environments. The
common availability of video cameras of various types, even
in some cases only low resolution cell phones, sometimes
have captured rare geomorphic phenomena in action, which
can be useful in analysis. Different velocities, viscosities, and
grain sizes of such unusual complex debris-flow continua are
being caught increasingly. Eventually, such features as the high
speed and enigmatic mass movements, such as long runout
zone landslides or slow debris flows may be captured in motion photographically and assessed for better understanding of
mechanics of their motion. Many other somewhat enigmatic
processes may be similarly assessed in the future with these
techniques.
In sum, the use of photographs is an essential and
central part of almost all geomorphologic research, and the
ubiquity and availability of so many old and new cameras
and photographic techniques enhances geomorphological
investigations. Attention to the details of making or obtaining
good digital photographic-image coverage of any particular
geomorphologic problem is viewed as an important, even
essential part of the modern discipline that should be taught
in some fashion in the more advanced graduate programs.
References
Altmaier, A., Kany, C., 2002. Digital surface model generation from CORONA
satellite images. ISPRS Journal of Photogrammetry and Remote Sensing 56(4),
221235.
Ashworth, F.L., 1937. Aerial mapping. Leicia Photography 6(9), 7 and 16.
Baars, D.L., Molenaar, C.M., 1971. Geology of canyonlands and cataract canyon.
Four Corners Geological Society, Sixth Field Conference, Cataract Canyon River
Expedition, 1971, 99 pp.
Bagley, J.W., 1922. Concerning aerial photographic mapping: a review.
Geographical Review 12, 628635.
Balog, J., 2009. Extreme Ice Now: Vanishing Glaciers and Changing Climate; A
Progress Report. National Geographic Society, Washington, DC, 118 pp.
Bishop, M.P., James, L.A., Shroder, Jr. J.F., Walsh, S.J., 2012. Geospatial
technologies and digital geomorphological mapping: concepts, issues and
research. Geomorphology, Binghamton Symposium Special Issue 137(1), 526.
http://dx.doi.org/10.1016/j.geomorph.2011.06.027.
Bishop, M.P., Olsenholler, J.A., Shroder, J.F., et al., 2004. Global land ice
measurements from space (GLIMS): remote sensing and GIS investigations of
the Earths cryosphere. Geocarto International 19(2), 5784.
Calkins, J., Oppenheimer, C., Kyle, P.R., 2008. Ground-based thermal imaging of
lava lakes at Erebus volcano, Antarctica. Journal of Volcanology and Geothermal
Research 177(3), 695704.
Casana, J., Cothren, J., 2008. Stereo analysis, DEM extraction and orthorectification
of CORONA satellite imagery: archaeological applications from the Near East.
Antiquity 82(317), 732749.
Chandler, J.H., 1989. The acquisition of spatial data from archival photographs and
their application to geomorphology. Ph.D. thesis, The City University, London,
England, 300 pp., unpublished.
Chandler, J.H., Brunsden, D., 2006. Steady state behaviour of the Black Ven
mudslide: the application of archival analytical photogrammetry to studies
of landform change. Earth Surface Processes and Landforms 20(3),
255275.
Coe, J.A., Taylor, E.M., Schilling, S.P., 1991. Close-range geophotogrammetric
mapping of trench walls using multi-model stereo restitution software. ACSMASPRS annual convention. Primary Data Acquisition 5, 3043.
Colwell, R.N., 1971. Monitoring Earth Resources from Aircraft and Spacecraft,
NASA SP-275. National Aeronautics and Space Administration, Washington, DC.
De Filippi, F., 1912. Karakoram and western Himalaya 1909: an account of the
expedition of H.R.H. Prince Luigi Amedeo of Savoy Duke of the Abruzzi. E.P.
Dutton and Company, NY, 496 pp., 18 panoramas.
Diolaiuti, G., Pecci, M., Smiraglia, C., 2003. Liligo Glacier, Karakoram, Pakistan: a
reconstruction of the recent history of a surge-type glacier. Annals of Glaciology
36, 168172.
El-Baz, F., 1977. Astronaut Observations from the ApolloSoyuz Mission.
Smithsonian Studies in Air and Space. Smithsonian Institution Press,
Washington, DC, vol. 1, 400 pp.
El-Baz, F., Warner, D.M., 1979. ApolloSoyuz Test Project, vol. II, Earth
Observations and Photography, NASA SP-412; Houston, Texas; National
Aeronautics and Space Administration, Lyndon B. Johnson Space Center,
692 pp.
Evans, C.A., Lulla, K.P., Dessinov, L.V., Glazovskiy, N.F., Kasimov, N.S.,
Knizhnikov, Yu.F., 2000. ShuttleMir Earth science investigations: studying
dynamic Earth environments from the Mir space station. In: Lulla, K.P.,
Dessinov, L.V. (Eds.), Dynamic Earth Environments: Remote Sensing
Observations from ShuttleMir Missions. John Wiley & Sons, New York, pp.
114.
Fairer, G.M., Whitney, J.W., Coe, J.A., 1989. A close-range photogrammetric
technique for mapping neotectonic features in trenches. Bulletin of the
Association of Engineering Geologists 26, 521530.
Galer, M., Horvat, L., 2005. Digital Imaging. Elsevier, Amsterdam.
Gough, S., 2007. River Geomorphology Videos. DVD. Little River Research &
Design, Carbondale, IL.www.emriver.com
Graf, W.L., 1985. Geomorphic measurements from ground-based photographs.
In: Pitty, A. (Ed.), Geomorphology Themes and Trends, pp. 211225.
Grosse, G., Schirrmeister, L., Kunitsky, V.V., Hubberten, H.-W., 2005. The use
of CORONA images in remote sensing of periglacial geomorphology: an
illustration from the NE Siberian coast. Permafrost and Periglacial Processes 16,
163172.
von Hagen, V.W., 1947. Maya explorer, John Lloyd Stephens and the Last Cities
of Central America and Yucatan. University of Oklahoma Press, Norman, OK,
324 pp.
Hanks, T.C., 2006. Repeat photography, virtual repeat photography, and Earthsurface change in the photographic era. Geological Society of America 38(7),
451(Philadelphia Annual Meeting, Abstracts with Programs).
Hayden, R.S., 1986. Geomorphological mapping. In: Short, N.M., Blair, Jr. R.W.
(Eds.), Geomorphology from Space: A Global Overview of Regional Landforms,
pp. 637656.
Johnson, B.F., 2009a. Art: old paintings give new insight into coastal change. Earth
54(8), 6667.
Johnson, B.F., 2009b. GigaPan imaging. Earth 54(11), 6667.
Jones, T.D., Godwin, L.M., Wisoff, P.J., Amsbury, D.L., Evans, C.A., 1996. Astronaut
Earth observations during the Space Radar Laboratory missions. Proceedings of
the IEEE Aerospace Applications Conference, Snowmass at Aspen, Colorado,
Institute of Electrical and Electronics Engineers, vol. 2, 310 February 1996,
pp. 2946.
Kargel, J.S., Abrahams, M.J., Bishop, M.P., et al., 2005. Multispectral imaging
contributions to global land ice measurements from space. Remote Sensing of
Environment 99, 187219.
Klett, M. (Ed.), 2004. Third Views, Second Sights: A Rephotographic Survey of the
American West. Museum of New Mexico Press, Albuquerque, 256 pp.
Lane, S.N., Richards, K.S., Chandler, J.H., 1993. Developments in photogrammetry;
the geomorphological potential. Progress in Physical Geography 17(3), 306328.
Lattman, L.H., 1987. Landform studies from space. Science 238, 1298.
Lee, W.T., 1922. The Face of the Earth as Seen from the Air: A Study in the
Application of Airplane Photography to Geography. American Geographical
Society Special Publication No. 4, New York.
Lillesand, T.M., Kiefer, R.W., 1994. Remote Sensing and Image Interpretation. John
Wiley and Sons, NY, 750 pp.
Lobeck, A.K., 1958. Block Diagrams: And Other Graphic Methods used in Geology
and Geography. Emerson-Trussell Book Co, Amherst, MA, 212 pp.
Lowman Jr. P.D., 1996. The human remote sensor in space: astronaut photography. 6
pp. http://fas.org/irp/imint/docs/rst/Sect12/Sect12_1.html (accessed 9 June 2010).
Lowman Jr. P.D., Tiedemann, H.A., 1971. Terrain Photography from Gemini
Spacecraft. Final Geologic Report, Report X-644-71-15. National Aeronautics
and Space Administration, Goddard Space Flight Center, Greenbelt, MD, 15 pp.
Madrigal, A., 2008. Extreme resolution photos and lasers create virtual Yosemite.
Wired Science 31. October 2008.
Malde, H.E., 1973. Geologic bench marks by terrestrial photography. US Geological
Survey Journal of Research 1, 193206.
Meneely, J.D., Smith, B.J., Viles, H.A., Gomez-Heras, M., 2008. In-situ monitoring
of limestone buildings in Oxford. In: Tiano, P., Pardini, C. (Eds.), In-Situ
Monitoring of Monumental Surfaces. CNR, Florence, pp. 327334.
Merriam, D.F., 2009. A lost art: geological illustrations. Special Commentary, GSA
Today 19(11), 3034.
Miller, V.C., 1961. Photogeology. McGraw-Hill, New York, NY.
National Aeronautics and Space Administration, 1974. Skylab Earth Resources Data
Catalog. Lyndon B. Johnson Space Center, Houston, TX, JSC-09016.
Office of Earth Sciences, NASA-Johnson Space Center, 14 March 2000. The
Gateway to Astronaut Photography of Earth. http://eol.jsc.nasa.gov/sseop/
default.htm.
Phillips, W.S., 1963. Photographic documentation, vegetational changes in northern
Great Plains. University of Arizona, Agricultural Experiment Station Report 214,
185 pp.
Prokesova, R., Kardos, M., Medvedova, A., 2010. Landslide dynamics from highresolution aerial photographs: a case study from the Western Carpathians,
Slovakia. Geomorphology 115, 90101.
Rand, G., Litschel, D., Davis, R., 2005. Digital Photographic Capture. Elsevier,
Amsterdam.
Rao, K.N., 2009. CORONA satellite photographs A new (Old) tool for earth
scientists. Journal of the Geological Society of India 73(3), 425429
http://dx.doi.org/10.1007/s12594-009-0022-z.
Rasmusson, G.A., Voth, K., 2001. Repeat photography monitoring made easy. Utah
State University Cooperative Extension, USEPA NR504, 11 pp. (obtained in-line
6 July 2001).
Reeves, D.M., 1927. Aerial Photographs, Characteristics and Military Applications.
The Ronald Press, NY, 307 pp.
Robinson, J.A., Amsbury, D.L., Liddle, D.A., Evans, C.A., 2002. Astronaut orbital
photographs as digital data for remote sensing: spatial resolution. International
Journal of Remote Sensing 23(20), 44034438.
Rosenberg, P., Katz, A.H., Rosenfield, G.H., 1966. Photogrammetry in the space age.
In: Thompson, M.M. (Ed.), Manual of Photogrammetry, vol. II, pp. 10491089.
Schenk, T., Csatho, B., Shin, S.W., 2003. Rigorous panoramic camera model for
DISP imagery. Proceedings of the Joint ISPRS/EARSeL Workshop: High
41
Resolution Mapping from Space 2003, Hannover, Germany, October 68,

www.ipi.uni-hannover.de/fileadmin/institut/pdf/schenk.pdf.
Schmidt, S., Nusser, M., 2009. Fluctuations of Raikot Glacier during the last 70
years a case study from the Nanga Parbat massif, northern Pakistan. Journal
of Glaciology 55(194), 949959.
Shepard, F.P., Grant, IV U.S., 1947. Wave Erosion along the Southern California
Coast. Bulletin of the Geological Society of America 58, 919926.
Shepard, F.P., Kuhn, G.G., 1983. History of sea arches and remnant stacks of La
Jolla, California, and their bearing on similar features elsewhere. Marine
Geology 51, 139161.
Short, Sr. N.M., Blair, Jr. R.W. (Eds.), 1986. Geomorphology from Space: A Global
Overview of Regional Landforms. National Aeronautics and Space Administration
(NASA), Washington, DC, 717 pp.
Shroder, Jr. J.F., Owen, L.A., Seong, Y.B., et al., 2010. The role of mass movements
on landscape evolution in the Central Karakoram: discussion and speculation.
Quaternary International, http://dx.doi.org/10.1016/j.quaint.2010.05.024.
Smith, H.T.U., 1943. Aerial Photographs and Their Applications. The Century Earth
Science Series. Appleton-Century Crofts, Inc., New York, 372 pp.
Sohn, H.G., Kim, G.H., Yom, J.H., 2004. Mathematical modeling of historical
reconnaissance Corona KH-4B imagery. The Photogrammetric Record 19(105),
5166.
Thompson, M.M. (Editor-in-Chief), 1966. Manual of Photogrammetry. Third ed.,
vols. 1 and 2. American Society of Photogrammetry, Falls Church, VA, 1199 pp.
Thornbush, M.J., 2010. Measurements of soiling and colour change using outdoor
rephotography and image processing in Adobe Photoshop along the southern
fac- ade of the Ashmolean Museum, Oxford. Geological Society, London, Special
Publications 331, 231236.
Thornbush, M.J., Viles, H.A., 2005. The changing fac- ade of Magdalen College,
Oxford: reconstructing long-term soiling patterns from archival photographs and
traffic records. Journal of Architectural Conservation 11(2), 4057.
Thornbush, M.J., Viles, H.A., 2007. Photo-based decay mapping of replaced stone
blocks on the boundary wall of Worcester College, Oxford. In: Prikryl, R., Smith,
B.J. (Eds.), Building Stone Decay: From Diagnosis to Conservation. Geological
Society Special Publication, vol. 271, pp. 6975.
Underwood, R.W., 1967. Space photography. Proceedings Gemini Summary
Conference; NASA-SP-138, 231290.
Veatch, F.M., 1969. Analysis of a 24-Year Photographic Record of Nisqually Glacier,
Mount Rainier National Park, Washington. U.S. Geological Survey Professional
Paper 631.
Wanless, Harold, 1986. Aerial Stereo Photographs. Hubbard Scientific, Fort Collins, CO.
Way, D.S., 1973. Terrain Analysis. Dowden, Hutchinson, and Ross, Stroudsburg, PA.
Webb, R.H., 1996. Grand Canyon: A Century of Change. Rephotography of the
18891890 Stanton Expedition. University of Arizona Press, Tucson, AZ, 290 pp.
Webb, R.H., Boyer, D.E., Turner, R.M., 2010. Repeat Photography: Methods and
Applications in the Natural Sciences. Island Press, Washington, DC, 530 pp.
White, C., Hart, E.J., 2007. The Lens of Time: A Repeat Photography of Landscape
Change in the Canadian Rockies. University of Calgary Press, Calgary, Alberta,
312 pp.
Wilmarth, V.R., Keltenbach, J.L., Lenoir, W.B., 1977. Skylab Explores the Earth.
NASASP-380. National Aeronautics and Space Administration, Washington, DC,
517 pp.
Relevant Websites
www.ExtremeIceSurvey.com
Extreme Ice Survey.
http://www.fas.org/irp/imint/docs/rst/Front/tofc.html
Federation of American Scientists.
www.lib.utulsa.edu/Speccoll/collections/rephotosurvey/index.htm
McFarlin Library, University of Tulsa.
http://rst.gsfc.nasa.gov/
NASA Remote Sensing Tutorial.
www.nro.gov/corona/facts.html
National Reconnaissance Office.
42
Biographical Sketch
Jack Shroder has had a strong interest in geomorphology ever since working in the Mojave Desert as a field
assistant on the US Air Force-funded Playa Project in 1963 for his professors, HTU Smith and Ward Motts. As a
result of this early experience in geomorphology, rather than becoming an oil company geologist, that he was
expected to do, he instead elected to study the landslides of Utah for his dissertation at the University of Utah, UT,
USA, in 1967. His eventual pursuit of landslides in the Hindu Kush and Himalaya (HKH) ultimately led him to
glacier studies in the region, and especially the use of repeat satellite imagery and photography to discern longterm changes in the highly labile geomorphology of glaciers. Dr. Shroder is planning on maintaining his activefunded research in the HKH for the foreseeable future.
Dr. Shroder has published more than 150 scientific papers in journals and books, 30 edited books, and several
hundred published abstracts, opinion editorials, and other writings. He is a Trustee of the Foundation of the
Geological Society of America (GSA), as well as a GSA Scientific Fellow, and also an elected Fellow of the American
Association for the Advancement of Science. Dr. Shroder won Distinguished Career awards from both the Geomorphology Specialty Group of the Association of American Geographers in 2010, as well as their Mountain
Specialty Group in 2001.
3.3 Microwave Remote Sensing and Surface Characterization

S Hensley and T Farr, Jet Propulsion Laboratory, California Institute of Technology, Pasadena, CA, USA
3.3.1
3.3.2
3.3.2.1
3.3.2.2
3.3.2.2.1
3.3.2.2.2
3.3.2.3
3.3.2.4
3.3.3
3.3.3.1
3.3.3.2
3.3.3.3
3.3.4
3.3.4.1
3.3.4.2
3.3.5
3.3.5.1
3.3.5.2
3.3.6
3.3.6.1
3.3.6.2
3.3.7
References
Types of Microwave Sensors

Microwave Remote-Sensing Principles
Frequency or Wavelength Considerations
Resolution
Beam-limited resolution
Pulse-limited resolution
Polarization
Scattering, Surface Compositions, and Surface Roughness
Altimeters
Theory of Operation
Example Echo Profile
Geomorphological Applications
Synthetic-Aperture Radars
Theory of Operation
Stereo SAR
Theory of Operation
Interferometric SAR
Theory of Operation
Summary
Glossary
Active sensor An active sensor is a sensor that generates its
own illumination such as a radar or light detection and
ranging (LiDAR).
Airborne Synthetic Aperture Radar (AIRSAR) A NASA/
JPL airborne synthetic aperture radar that operated from
1986 to 2004. It was a multi-frequency (C-, L-, and
P-bands) fully polarimetric radar that also had both alongtrack and cross-track interferometric observing capabilities
and C- and L-bands.
Altimeter A radar altimeter is a device for determining the
height or elevation above a surface by recording the roundtrip transit time for a pulse to propagate to the surface and
return.
Antenna A device used to radiate and receive
electromagnetic radiation.
Azimuth compression The signal processing needed to
combine data from multiple pulses to achieve fine resolution
in the along-track direction for microwave sensors.
45
47
47
48
48
48
49
50
51
51
53
56
57
58
62
65
65
68
69
69
74
75
77
Azimuth resolution Resolution in the along-track

direction (direction of travel of the sensor).
Backscatter The amount of electromagnetic signal
reflected back toward the direction of incidence.
Bandwidth The difference between the highest and lowest
frequencies in a contiguous set of frequencies in a radar
waveform. Bandwidth is related to the achievable resolution
of a radar sensor.
Beam-limited Sensor whose ground resolution is
determined from the size of the antenna footprint on the
ground.
Beamwidth The half-power angular width of the
mainlobe of an antenna.
Chirp A type of radar waveform that varies the frequency
linearly over the transmitted pulse.
Correlation See interferometric correlation.
Decorrelation The loss of signal coherence between the
pair of signals in a radar interferometer. Mathematically, it is
one minus the correlation.
Hensley, S., Farr, T., 2013. Microwave remote sensing and surface
characterization. In: Shroder, J. (Editor in Chief) Bishop, M.P. (Ed.), Treatise
on Geomorphology. Academic Press, San Diego, CA, vol. 3, Remote Sensing
and GIScience in Geomorphology, pp. 4379.
http://dx.doi.org/10.1016/B978-0-12-374739-6.00044-0
43
44
Microwave Remote Sensing and Surface Characterization
Dielectric constant A physical property of a material that

determines how much electric energy is reflected and
absorbed.
Digital elevation model (DEM) A digital representation
of the topography of a surface.
Echo profile The sequence of received signals for
multiple-range samples within the antenna footprint of a
radar altimeter.
Ellipticity One of two parameters (the other being tilt)
necessary to fully describe the polarization state of an
electromagnetic wave. Ellipticity measures the elongation of
the ellipse formed in the plane perpendicular to
the direction of propagation for an elliptically
polarized wave.
Foreshortening A form of geometric distortion occurring
in radar imagery of sloped surfaces that results in the
apparent contraction or expansion of these features relative
to their true dimensions.
Gain The relative increase in radiation at the maximum
point compared to an idealized isotropic radiator.
Geometric decorrelation The loss of correlation in an
interferometer resulting from spatial separation of the
observing vantages.
H Electromagnetic waves whose electric field orientation
is fixed and parallel to the ground surface are referred to as
being horizontally polarized.
Image parallax The apparent shift or displacement of an
object between images taken at different vantages whose
elevation is above or below a reference surface.
Incidence angle The angle between the direction of
incident radiation and a vector normal or perpendicular to
the surface.
Interferometric correlation A measure of the similarity of
the pair of signals received by a radar interferometer.
Correlation equals zero if there is no similarity and equals
one if they are identical.
Interferometry The science and techniques used to study
the interference of superimposed electromagnetic waves
transmitted from spatially and/or temporally separated
vantages.
Jet Propulsion Laboratory (JPL) One of the major NASA
centers primarily responsible for robotic exploration of the
solar system.
Layover This occurs when the radar beam reaches the top
of a tall feature before it reaches the base. In this case, the
top of the object appears displaced closer to the sensor from
its base and hence appears to lay over its base.
Mainlobe The angular region or lobe containing the
maximum power of an antenna.
Microwave sensor Any sensor transmitting or receiving
microwaves.
Microwaves A form of electromagnetic radiation with
wavelengths from 1 mm to 1 m. Microwaves are about
100 000 times longer than visible light waves and are
employed in a variety of Earth science applications.
National Aeronautics and Space Administration
(NASA) The United States space agency.
Noise decorrelation The loss in interferometric
correlation due to thermal noise.
Passive sensor One that senses the emitted or reflected

radiation from object.
Polarimetry The science and techniques used to study the
polarimetric properties of electromagnetic waves and the
scattering properties of these waves.
Polarization signature A plot of the backscattered power
received from specific a scatterer as a function of the
polarizations of the incident and scattered
electromagnetic waves.
Pulse repetition frequency (PRF) The frequency at which
a radar emits pulses and is measured in units of hertz (s1).
Pulse repetition interval (PRI) The time between pulses
emitted by a radar.
Pulse length The duration of a transmitted radar pulse.
Pulse-limited Sensor whose ground resolution is
determined from the range resolution of the transmitted
waveform.
Radar Any sensing device that transmits and then receives
reflected microwaves to measure the properties of the object
or scene under study.
Range bandwidth See bandwidth.
Range compression The signal processing needed to
achieve fine resolution in the range or line-of-sight direction
for microwave sensors.
Range resolution Resolution in the line-of-sight or
direction of propagation.
Resolution A measure of the ability to spatially separate
signals obtained from a sensor. For radar sensors, the
resolution is a function of the bandwidth.
Root mean square (RMS) A statistical measure of the
magnitude of a varying quantity.
Shadow Those regions in a radar image that are occluded
from view resulting either from obstruction by another
object or a portion of the terrain.
Sidelobes The angular region outside the antenna
mainlobe.
Signal-to-noise ratio (SNR) The ratio of the signal
strength to the background thermal noise.
Small perturbation model (SPM) An electromagnetic
scattering model that predicts the amount of radar
backscatter from a surface subject to conditions on the
surface geometry.
Speckle Speckle noise is a granular noise that inherently
exists because of the coherent form of the radiation that
degrades the quality of the active radar and SAR images.
Shuttle Radar Topography Mission (SRTM) A single pass
radar interferometer that flew aboard the space shuttle in
February 2000 and generated fine-resolution topographic
maps of the Earths surface between 7601 latitude.
Synthetic aperture radar (SAR) A type of radar that uses
data from multiple pulses to synthesize a long antenna,
thereby achieving fine resolution in the along-track or
azimuth direction.
Temporal decorrelation The loss of interferometric
correlation as a result of changes within a resolution
element between observations.
Tilt One of two parameters (the other being ellipticity)
necessary to fully describe the polarization state of an
electromagnetic wave. Tilt measures the orientation of the
ellipse formed in the plane perpendicular to the direction of

propagation for an elliptically polarized wave.
Uninhabited Aerial Vehicle Synthetic Aperture Radar
(UAVSAR) UAVSAR is a NASA/JPL L-band actively
scanned synthetic aperture radar operational since
2008.
V Electromagnetic waves whose electric-field orientation is

fixed and parallel to the local vertical are referred to as being
vorizontally polarized.
Volumetric decorrelation The loss of interferometric
correlation as a result of a vertical distribution of scatterers
within a resolution element.
Nomenclature
c^
fd
~f
d
NL
/
c
dR
rh
r/
Dqext
Radar wavelength (m)

Speed of light (m s1)
Frequency (Hz)
Antenna length (m)
Along-track spacing (m)
Platform velocity (m s1)
Antenna half-power beamwidth (radians)
Gain (dB)
Antenna area (m2)
Beam-limited spatial resolution (m)
Range (m)
Propagation time (s)
Pulse length (s)
Bandwidth (Hz)
Ground spacing (m)
Range resolution or spacing (m)
Look angle (radians)
Incidence angle (radians)
Cross-track slope (radians)
Pulse repetition frequency (Hz)
Radar backscatter (m m1)
Relative complex dielectric constant ( )
k
C
F
L
Ds
vp
h3dB
G
A
Db
q
tp
s
B
Dg
Dq
yc
hi
sq
PRF
ro
e
ht
Laz
!
b
Lsyn
Dp, Dpr
Dssyn
b
a
!
d
45
Unit look vector ( )

Beam-limited ground resolution (m)
Pulse-limited ground resolution (m)
Number of looks ( )
Interferometric phase (radians)
Interferometric correlation ( )
Ground area element (m2)
Elevation precision (m)
Phase precision (radians)
Range extent of antenna footprint on
ground (m)
Terrain elevation (m)
Azimuth antenna length (m)
Interferometric baseline vector (m)
Length of synthetic aperture (m)
Optical and radar parallax (m)
Azimuth resolution from synthetic
aperture (m)
Baseline length (m)
Baseline orientation angle (radians)
Surface deformation vector (m)
Abstract
The scientific study of landforms and the processes that shape them have benefited tremendously from the rapid progress in
remote sensing and computer technologies. Microwave sensors have the ability to inform geomorphological studies at
length scales that range from planetary scales of thousands of kilometers to micro-topography of surfaces at the centimeter
scale. This tremendous range of applicability of microwave sensors is a result of the variety of sensor types and the large
frequency range over which these sensors operate. Microwave sensors, operating from both airborne and space-borne
platforms, have had a profound influence on our understanding of landforms and the processes that shaped them. In
addition to their standalone intrinsic value to geomorphological studies, these sensors can be combined synergistically with
other sensor data (e.g., light detection and ranging (LiDAR), optical, or hyperspectral data). This chapter focuses on the
basic theory of sensor operation and selected applications in geomorphology.
3.3.1
Types of Microwave Sensors
Microwaves are a form of electromagnetic radiation. Electromagnetic waves are comprised of oscillating electric
and magnetic fields that propagate at the speed of light or
299 792 458 m s1. They are characterized by several key
properties that include their frequency or wavelength, their
intensity, or strength and polarization. Frequency, f, is the rate
at which the amplitude of an electromagnetic-wave changes
and is related to wavelength via the relationship
c
1
l
f
where c is the speed of light. The intensity is a measure of

the strength or amplitude of the electromagnetic radiation and
is related to the amount of energy carried by the wave.
The electric and magnetic fields comprising an electromagnetic wave are vector quantities, which means they have
both a magnitude and direction. The direction of the electric
field (and the magnetic field) lies in the plane perpendicular
to the direction of propagation and defines the polarization
of the wave as shown in Figure 1. The electric field can point
in a fixed direction as the wave propagates, called linear polarization, or it can trace out a circular (circular polarization)
or more generally an elliptical path in the plane perpendicular
46
to the direction of propagation. The vector nature of electromagnetic radiation gives rise to sensitivity to the orientation of
an object from which it scatters, and, hence, provides additional information about surface morphology.
Sensors that employ electromagnetic waves with frequencies in the range of 0.340 GHz (wavelengths from 1 mm to
1 m) are referred to as microwave sensors. These sensors can
operate either passively, that is, they sense natural microwave
radiation emitted by a surface, or they may be active sensors,
that is, sensors that emit microwaves and then sense reflected
microwaves from the object or scene under observation.
Microwave sensor systems are currently operating over a wide
range of frequencies depending on their intended applications.
Operating frequencies vary from as low as 300 MHz to as high
as 40 GHz. The choice of frequency is dictated by a number of

factors including intended application, platform and power
constraints, and availability of the spectrum at the desired
frequency range. Table 1 shows the correspondence between
frequency, wavelength, and the band designation letter code
(assigned in World War II for security reasons) that are often
used to specify the operating frequency of microwave sensors.
Microwave sensors capable of making measurements
suitable for geomorphological applications can be grouped
into five basic types (Table 2). All are active sensors (that is,
they emit and then record reflected microwaves, to make
measurements of a surface. (Passive microwave sensors, those
that receive microwaves naturally emitted from a surface, can
also provide useful information about a surface; e.g., passive
microwave sensors are used to measure soil moisture and seasurface salinity. This chapter restricts its coverage to active
sensors.) Sensor types have distinctive measurement paradigms
and provide information at length scales from thousands of
kilometers down to centimeters. They include the following:
Altimeters.
Direc
t
prop ion of
agat
ion
Pola
rizat
plan ion
e
Figure 1 The polarization plane is defined by the electric field and

the direction of propagation vectors.
Table 1 Frequency and wavelength relationship table

Frequency band
(MHz)
Wavelength range
(cm)
Band identification
26 50040 000
18 00026 500
12 50018 000
800012 500
40008000
20004000
10002000
300900
1.130.75
1.661.13
2.41.66
3.752.4
7.53.75
157.5
3015
10033
Ka
K
Ku
X
C
S
L
P or UHF
These determine the height of a surface by

measuring the round-trip time it takes for a microwave to
reflect from the surface and converting it to a distance that is
subtracted from the platform altitude.
Synthetic aperture radar (SAR). These radars generate fineresolution imagery in the along-track or azimuth direction
using the motion of the platform to synthesize a long antenna. The magnitude of the reflected energy from the surface, called the backscatter, in each pixel is related to both
the surface composition, through the dielectric constant,
and the surface roughness at the wavelength scale. SARs
operate during both day and night and can operate during
inclement weather as well (e.g., clouds or rain, thus providing remote-sensing observations in regimes where other
sensors cannot operate).
Polarimetric SAR. Polarimetric SARs are SAR systems that
generate imagery from multiple polarizations. Polarimetric
information helps resolve surface roughness information
from surface composition effects on the backscatter. Sensitivity to orientation and increased information on surface
scattering permits more detailed surface classification and
more accurate quantitative parameter estimation.
Stereo SAR. Stereo uses SAR imagery collected from different
vantages to determine topographic information. Like optical
image stereo pairs, objects at different elevations in SAR
images give rise to parallax, or image displacement, that is
proportional to the height above a reference surface.
Table 2 Types of microwave sensors

Sensor type
Length scale
Geomorphological applications
Altimeter
SAR
Polarimetric SAR
101000 s km
1 cm100 s km
1 cm100 s km
Stereo SAR
10 m100 s km
Interferometric SAR
1 m00 s km
Geodetic studies of earth shape, ice profiling.

Surface roughness, fluvial and aeolian processes, land-cover types.
Surface roughness, fluvial and aeolian processes, land-cover types,
paleohydrology.
Topographic mapping, hydrological processes, fault mapping, hypsometry,
glaciology.
Topographic mapping, subcanopy topographic mapping, hydrological
processes, fault mapping, hypsometry, glaciology.
Interferometric SAR. Like stereo SAR, interferometric SARs use

data collected from different vantages to determine topographic or surface displacement information. In interferometric systems, the parallax is typically much less than a
pixel; so, the topographic information is obtained from a
phase measurement that makes extremely accurate parallax,
or range difference, measurements possible. These phase
measurements are then converted into elevation measurements or surface displacements in the case of differential
radar interferometry. Differential radar interferometry can
measure surface displacements at the millimeter level at fine
resolution and over wide spatial scales, thus providing detailed surface morphological change information for both
natural (e.g., earthquakes and volcanoes) and anthropogenic
causes (e.g., pumping of oil and groundwater).
3.3.2
Microwave Remote-Sensing Principles
Although microwaves are a form of electromagnetic radiation,

they interact differently with surfaces compared to shorterwavelength electromagnetic waves (e.g., visible light). Moreover, key sensor parameters of interest (e.g., resolution) are
often specified in a manner not usual for optical sensors. This
section provides a brief overview of key sensor parameters of
interest and a very elementary introduction to the scattering of
microwaves from surfaces (Raney, 1999).
3.3.2.1
reflected from the surface related to the quantity or quantities

of interest. For example, C-band or 5-cm wavelength systems
are often desired for ice imaging applications, whereas L-band
(23-cm wavelength) or P-band (80-cm wavelength) is used for
vegetation and land-cover applications due to their increased
penetration capabilities.
Antennas are used to both transmit and receive microwave
radiation. The size of an antenna very much depends on the
wavelength. To see why this is the case we consider how the
beamwidth and gain depend on wavelength and antenna size.
Beamwidth is an angular measure of how much the microwave
radiation spreads as it propagates from an antenna. Gain relates
to the intensity of the radiation in a given direction relative to an
idealized antenna that radiates equally in all directions. The intensity or gain pattern of a typical antenna can be separated into
two regions. First, is the mainlobe where most of the energy is
directed and the second is the sidelobe region which is the area
outside of the mainlobe. The boresight direction is the direction
of peak gain of the antenna. Quantitatively, the beamwidth is
measuring the angular width of the mainlobe where the intensity drops by a factor of 2. (Intensity or gain is often measured in
decibels. The gain in dB, GdB, is equal to 10 log10(G), where G is
intensity. In decibels a factor of 2 reduction in intensity is a loss
of 3 dB since 10 log(0.5)E 3.0103.) Figure 2 shows the intensity pattern of a typical antenna with the mainlobe and
sidelobe regions identified.
The relationship between beamwidth, y3dB, and antenna
size, L, is given by
y3dB k
Frequency or Wavelength Considerations
Frequency and wavelength are related as shown by eqn [1].

Choice of frequency or wavelength is governed by the intended application and platform constraints. Microwaves tend
to interact most strongly with objects that are 1/20th of a
wavelength or larger, and penetrate more deeply into soil,
vegetation, and ice for longer wavelengths. Thus, depending
on the desired application, the wavelength is chosen to
maximize the information content contained in the radiation
47
l
L
2
where l is the wavelength and k is constant that depends on

the design of the antenna. Typical values of k are between 0.3
and 1.0. (For simplicity k will be assumed equal to 1
throughout this chapter.) Notice, that as the antenna size becomes larger, the beamwidth becomes smaller, and, hence, the
microwaves are directed into a narrower angular region. As the
antenna dimensions can be different in different directions,
the beamwidth will also vary accordingly.
0
Beamwidth
Intensity or gain (dB)
10
20
30
Sidelobes
40
Mainlobe
Sidelobes
50
60
60
50
40
30
0
10
20
20 10
Angle off-boresight (deg)
30
40
50
60
Figure 2 Intensity pattern for a typical antenna is divided into two primary regions. The mainlobe, shaded in light green, is the primary region
where a microwave sensor ideally would supply and receive illumination. The sidelobe region, indicated by the red arrows, is a region of reduced
gain where undesired energy is both emitted and received. Techniques exist for reducing sidelobes in particular applications; however, the laws
of physics preclude their complete elimination.
48
The peak intensity or gain, G, of an antenna at the center of

the mainlobe is given by
G
4pA
l2
3
3.3.2.2.2
where A is the effective area of the antenna. In order to obtain

useful microwave signals reflected from a surface, a microwave
sensor must employ an antenna of appropriate size. Practical
limits exist for the size of an antenna that can be accommodated
on either space-borne or air-borne platforms, and, hence, places
constraints on frequency selection for microwave sensors.
3.3.2.2
Resolution
One of the most important parameters in selecting a sensor for

any remote-sensing application is the resolution, or the ability
to spatially discriminate signals from that sensor. Microwave
sensor resolution, depending on the type and operation of the
sensor, can be limited either by the antenna beamwidth, called
beam-limited resolution, or by the form of the transmitted
signal called the waveform, called pulse-limited resolution.
3.3.2.2.1
where r is the range. The important fact to note about beamlimited resolution is that it depends on the distance, with
greater distances resulting in coarser resolution.
Beam-limited resolution
Beam-limited systems are ones for which the range variation

within a beam footprint is less than the range resolution of the
system. The size of the antenna footprint on the ground is
approximately the range, or distance between the antenna and
ground, times the beamwidth measured in radians. Figure 3
shows the region on the ground illuminated by an antenna at
a range, r.
Beam-limited spatial resolution, Db, derived from eqn [2]
is given by
l
Db r
L
4
Laz
Lel
Pulse-limited resolution
Before proceeding directly to the definition and quantification

of pulse-limited resolution, it is useful to understand a little
about the operational aspects of microwave sensors. As mentioned earlier, active sensors supply their own illumination by
transmitting microwaves. Microwaves may be transmitted
continuously, called continuous wave (CW) systems, or be
transmitted in segments called pulses of limited duration that
are interrupted by periods when the sensor is not transmitting,
called pulsed operation. As nearly all the systems used for
geomorphological applications are of a pulsed nature, we will
assume pulsed operation of microwave sensors for the remainder of this chapter. For pulsed systems, the time duration
between pulses is called the pulse repetition interval (PRI)
and the associated pulse frequency is called the pulse repetition frequency (PRF 1/PRI). The time duration of the
pulse is called the pulse length.
Locating where the microwaves are reflected from a surface
depends on the sensor type; however, all the microwave sensor
types under consideration use range or the distance from the
antenna to an object for location determination. Range is
measured by recording the time it takes from the generation of
a pulse till it propagates to an object and then returns to the
radar. The one-way distance or range to an object, r, is then
given by
r
ctp
2
where tp is the round-trip propagation time. Nominally, range

resolution (i.e., ability to discriminate objects in the direction
of propagation of the microwaves) is limited by the width of
the transmitted pulse, since energy returned from any point of
a pulse cannot be distinguished from another point within the
same pulse. In equation form, the range resolution, Dr, is
given as
Dr

Laz

Lel
Figure 3 An antenna of dimensions Laz in the along-track or azimuth

direction and Lel in the cross-track or range direction illuminate the
ground. The shaded region shows the area on the ground illuminated
by the mainlobe of an antenna that has approximate dimensions given
by the range, r, multiplied by the beamwidth of the antenna. The red
arrow shows the range or elevation antenna footprint dimension that is
proportional to the elevation beamwidth, and the blue arrow shows the
along-track antenna footprint dimension that is proportional to the
azimuth beamwidth.
5
ct
2
6
where t is the pulse length. Many operational systems find it

impractical to transmit pulses narrow enough to achieve their
desired resolution due to peak power limitations or other
hardware considerations. In order to reduce the peak power in
a transmitted pulse yet maintain the same average power, it is
desirable to have longer pulses without somehow sacrificing
range resolution. This is achieved by encoding the transmitted
pulse in such a way as to be able to distinguish where within a
pulse the returned energy originated. Several schemes exist for
achieving this discrimination; however, the method used by
most microwave sensors is chirp encoding, where the frequency
is linearly changed across the pulse as illustrated in Figure 4.
The amount of frequency variation across a pulse is called
the range bandwidth. Range resolution is inversely proportional to the bandwidth, B, and is given by
Dr
c
2B
7
49
Frequency
PRI
i
Bandwidth

0
Time
Figure 4 A typical microwave sensor emits a sequence of pulses
separated in time. The time duration between pulses is called the
pulse repetition interval (PRI) and the associated pulse frequency is
called the pulse repetition frequency (PRF 1/PRI). The pulse
duration, t, is called the pulse length. For many microwave sensors,
each pulse is frequency encoded with a linear-frequency ramp across
the pulse known as a chirp.
Table 3 Bandwidth to resolution table

Bandwidth (MHz)
Resolution (m)
400
300
160
80
40
20
10
0.37
0.50
0.95
1.90
3.70
7.50
15.0
g
i
sight and nadir, and tr is the terrain slope in the range direction as illustrated in Figure 5.
When the ground-projected range resolution is less than
beam-limited resolution, the system is said to be pulselimited.
From eqns [6] and [7], the bandwidth associated with a

nonchirped pulse of duration, t, is simply 1/t. For coded
pulses, the desired range resolution is achieved only after a
signal-processing step called range compression. Note that in
this case range resolution is independent of range! In fact, 4-m
resolution imagery of the moon has been obtained from the
Earth at a distance of 373 000 km using this technique
(Hensley et al., 2009). Table 3 shows the conversion between
range resolution and bandwidth for a wide range of currently
operational systems.
Resolution in a direction different from the range direction
is needed to generate two- (or three-)dimensional image
products. Direction along the platform flight path is called the
along-track or azimuth direction and is generally the second
spatial coordinate that is not aligned with the range direction
for Earth (or planetary) observing sensors. How resolution is
obtained is a function of the type of sensor and the sensor
imaging geometry. It is possible to express azimuth resolution
in terms of bandwidth, but this will not be pursued here.
Specific details concerning azimuth resolution are discussed in
the sections covering the sensor types.
Range resolution, which is measured along the line of
sight, can be converted into resolution along the ground, Dg,
and is given by
Dr
Dr
sinyi sinyl tr
Figure 5 The ground resolution, or resolution along the surface

being imaged, is a function of the intrinsic range resolution of the
sensor and the imaging geometry. Variations in the local incidence
angle, yi, which is a function of the look angle, yl, and the crosstrack slope, tr, results in changes to the ground resolution, Dg. The
sensor resolution, Dr, is shown as bounded by the two green
dashed lines and the ground resolution, Dg, by the thick brown line.
Larger incidence angles result in finer resolution, whereas small
incidence angles result in coarser resolution.
3.3.2.3
Dg
8
where yi is the incidence angle, which is the angle between the

line-of-sight vector and the local normal vector to the surface,
^, yl is the look angle, which is the angle between the line of
n
Polarization
A microwave signal impinging on a resolution element (area

of the surface contained within a single range and azimuth
resolution cell) will in general scatter energy in all directions.
The signal reflected back toward the microwave sensor or radar
is referred to as the backscatter. The amount of backscattered
energy is a function of a number of parameters including the
polarization that was introduced in Section 3.3.1. Polarization
diversity can be a powerful tool for characterizing the type of
surface and for separating the various components contributing to the backscatter magnitude. Although even a cursory
treatment of radar polarization theory is beyond the scope of
this chapter, a few salient points and terminology will prove
useful to subsequent sections. See Mott (2007), Lee and
Pottier (2009), and Cloude (2010) for more details.
Microwave polarization is characterized by the shape of the
path that the electric-field vector traces out in the plane perpendicular to the direction of propagation. The most general
path is an ellipse that can be characterized by two parameters.
Tilt measures the orientation of the ellipse with respect to
horizontal direction and ellipticity is a measure of ellipse
elongation, as illustrated in Figure 6.
Now although there are an infinite number of polarization
states, it turns out that by transmitting and receiving data on
two suitably chosen polarizations (a sufficient requirement is
that they be orthogonal), it is possible to synthesize all other
polarizations from a combination of these representative polarizations. The most common representative set used by most
radars are horizontal and vertical polarization denoted by H
and V, respectively. Polarimetric radars operate by transmitting
H and V polarization on alternate pulses and receiving both H
and V polarization on every pulse. Radars can also operate in
50
|E |y0
|E |x0
Figure 6 Microwaves are a form of electromagnetic radiation

whose polarization state affects the returned signal from a
microwave sensor. Polarization is characterized by the shape the
electric-field vector traces out in the plane perpendicular to the
direction of propagation. The most general shape traced out by the
electric-field vector is an ellipse that is typically specified by two
numbers. Illustrated above are the tilt, f, a measure of the
orientation of the ellipse and the ellipticity, a measure of the
oblateness of the ellipse.
restricted polarization combinations, that is, transmitting only

H and receiving H and V, or transmitting H and receiving H
and transmitting V and receiving V. When the sensor is operated in a restricted polarization combination, it is not possible
to synthesize all possible polarizations any longer.
Various polarization combinations or quantities that
can be computed from the various polarizations (e.g., polarization signatures, eigenvalue and eigenvector decompositions,
and entropy) are useful both for land classification and for
determining quantitative information about the surface. For
example, the pedestal height in the polarization signature can
be related to surface roughness. Polarization diversity can
also be used to identify dominant scattering mechanisms
within a resolution element, as well as determine quantitative
information about the orientation of objects. A nice example
of the sensitivity to surface orientation is the estimation of
surface slope in the along-track direction from shifts in the peak
of the polarization signature. See Lee et al. (2000) for more
details.
3.3.2.4
Scattering, Surface Compositions, and Surface

Roughness
In order to understand how wavelength-scale roughness, that

is, commonly much finer than the resolution of the radar, can
be inferred from microwave-sensor observations, it is necessary to understand a little about how microwaves interact with
surfaces. As this subject is very rich and can be quite technical,
we will only touch upon this topic to illustrate how one might
go about obtaining centimeter roughness data from microwave measurements.
Backscatter strength is a function of the composition of the
surface, its structure, polarization, and imaging geometry.
Electrical composition of a surface is characterized by its
dielectric constant. The dielectric constant of a material
determines how much energy is absorbed or reflected from the

surface and depends weakly on the radar frequency. The dielectric constant is a complex quantity, that it is has a real and
imaginary part, with the real part determining the amount of
energy reflected and the imaginary part determining the
amount of penetration and absorption. Surface moisture affects the dielectric and will affect the level of backscatter.
Surface structure is generally characterized by roughness, a
measure of how much the surface morphology varies in a
resolution element. Roughness is measured in terms of the
incident radiations wavelength, so surfaces that are smooth at
one wavelength may appear rough at another wavelength. As a
general rule of thumb, the rougher the surface, the greater the
backscatter. For example, a road that is relatively flat and free
of potholes or other major imperfections may appear very
smooth when imaged using an L-band (23-cm wavelength)
radar, since a typical roads micro-topography may be less than
5 mm. The same road imaged with a Ka-band (7.5-mm
wavelength) radar, however, may seem quite rough and appear
relatively bright in a radar image. Figure 7 shows the same
area imaged at three different wavelengths.
Imaging geometry is generally characterized in terms of the
incidence angle, which is the angle between the incident wave
and the local surface normal, as illustrated in Figure 5. In
general, the backscatter decreases as the incidence angle becomes larger. Thus, the largest backscatter is achieved when
the incoming wave is perpendicular to the surface (i.e., zero
incidence angle) and the smallest backscatter occurs for incoming waves parallel to the surface (i.e., grazing incidence or
901 incidence angle).
One of the simplest mircrowave-backscatter models is the
small-perturbation model (SPM) that we will use to illustrate
the several ways surface-roughness information can be obtained
from radar data. This model is valid for surfaces that are not too
rough at the wavelength scale. The horizontal and vertical polarized backscatter, sxx, where (xx hh or vv) is given by
sxx 4pk4 h2 cos4 yi 9axx 9W2sinyi
9
where k 2p/l, h is the root-mean-square surface roughness, yi

is the incidence angle, W is the two-dimensional normalized
surface-roughness spectrum, and axx are given by
ahh
avv
e1
p 2
cosyi e sin2 yi
e 1e 1sin2 yi e
p
ecosyi e sin2 yi 2
10
where e is the complex dielectric constant of the surface. Note

from eqns [9] and [10] that the backscatter is a function of the
surface roughness and the dielectric constant, as well as the
imaging geometry. Surface roughness can be separated from
these other factors in several ways, provided enough measurements are available.
One method of separating surface-roughness information
uses polarization diversity. In cases where the SPM model is
valid, this can be seen in principle by noting that the ratio
of shh/svv ahh/avv depends only on the dielectric constant
and incidence angle. Using this relationship to solve for the
51
Figure 7 GeoSAR X-band (left), AISRAR L-band (middle), and GeoSAR P-band (right) orthorectified SAR images. Notice how the vegetated
areas in the center portion of the images have much greater contrast at L-band (23-cm wavelength) and P-band (85 cm wavelength) than X-band
(3-cm wavelength). This contrast differential results from open areas appearing smoother at P-band than X-band, whereas the vegetated areas
appear rough at both wavelengths.
dielectric constant, eqn [8] can then be inverted for the surface
roughness. More complicated models, both empirical and
theoretical, of the polarization dependence on surface
roughness are used in cases where the simple SPM model does
not apply.
Another means of trying to separate out surface roughness
from the other scattering factors uses incidence-angle diversity.
Measurements of the backscatter over a range of incidence
angles, preferably covering as large an angle range as possible
and with measurements close to normal incidence where the
backscatter variation is typically the largest, can be inverted to
give surface-roughness information under appropriate conditions. In some cases, quantitative surface-roughness information does not have the desired precision but may allow
classification of the surface into smooth, moderately rough,
and very-rough classes. Again, we note that surface roughness
means surface roughness at the wavelength scale.
3.3.3
Altimeters
We begin the discussion of microwave sensors with altimeters

which have the simplest implementation of the sensors we
will describe and use it to introduce several other key concepts
of microwave remote sensing.
3.3.3.1
Theory of Operation
Microwave or radar altimeters are downward- or nadir-pointing

sensors that measure terrain elevation. Altimeters measure terrain elevation by taking the difference of the platform height
and the range to the surface as shown in Figure 8.
Figure 8 Altimeters are nadir-pointing microwave sensors that

determine terrain height by measuring the time delay for a pulse to
transit to and from the surface. Measured elevation is a measure
average elevation in the antenna footprint shaded in light green.
Elevation profiles are obtained by taking measurements from multiple
pulses while the sensor moves in the along-track direction (i.e., in
the direction of the velocity vector shown in black above). Additional
elevation measurement footprints are shown shaded in yellow.
Although the basic concept of altimeter operation is very

simple, in practice understanding the measurement is complex
due to the fact that the terrain elevation is not constant within
the footprint of the antenna beam on the ground and the
manner in which microwaves backscatter from the terrain.
Altimeters are profiling instruments that measure terrain
elevation beneath the aircraft or spacecraft by emitting a sequence of pulses and measuring the distance to the surface for
each pulse. Spatial resolution in the cross-track or range direction depends on whether the altimeter is a beam-limited or
pulsed-limited sensor. Beam-limited systems are those for
which the beam size on the ground, fd, given by
l
L
where Dr is the range resolution of the sensor and r is the range

to the nadir point (point directly beneath the sensor). Figure 9
shows the pulse-limited and beam-limited dimensions.
From eqns [11] and [12], an altimeter is pulsed-limited if
2
l
Dr
13
8
L
r
A pulse-limited altimeter can have multiple range-resolution
elements within an antenna footprint and thus provide finerresolution topography measurements within the antenna
footprint. Since the range varies for range-resolution elements
within the antenna footprint as a result of topographic variation within the footprint, it means there are multiple possible
elevation values that can be assigned to a given pulse depending on the algorithm chosen. Possible algorithms for
choosing the reported elevation are the highest, lowest, or
average elevation within the antenna footprint as illustrated in
Figure 10. Understanding of what a particular altimeter is
measuring is critical to proper interpretation of the data. Using
the echo profile, which is the echo intensity as a function of
range, can provide information on the topographic variability
and surface composition at the scale of the footprint size.
What can be estimated and to what level of accuracy depends
on the applicability of the particular scattering model used to
invert the echo-profile data for the geophysical parameters of
interest.
Spatial resolution in the along-track direction depends on
several factors, including how the data is processed. Without
sophisticated processing to sharpen the along-track resolution
similar to synthetic aperture radars (discussed in the next
section), the along-track resolution is the greater of the beam
footprint size in the along-track direction and the spacing or
distance between pulses. If the distance between pulses in the
along-track direction, Ds, given by
vp
PRF
14
where vp is the platform velocity and PRI is the inter-pulse

period, is greater than the antenna footprint size, then the
along-track resolution equals Ds. However, if the along-track
spacing between pulses in less than the footprint size, then the
along-track resolution will be less than or equal to the footprint size. It is possible to refine the resolution in the alongtrack direction to better than the antenna footprint size if the
PRF can be commanded high enough, such that the pulse
spacing is less than half the antenna length in the along-track
direction. By exploiting the fact that points within the antenna
+
is less than the size of a ground-projected range resolution

element at nadir, ~fd , given by
q
pp
~f 2 r Dr2 r2 E2 2 rDr
12
d
Ds vp PRI
Pulse-limited
11
fd r
Beam-limited
+
52
Figure 9 In beam-limited altimeters (left) the antenna beam

footprint on the ground is less than the distance to the point where a
resolution element centered at nadir intersects the terrain (illustrated
by the dotted line). For pulse-limited altimeters (right), however, the
antenna footprint (green-shaded region) on the ground is wider than
the dimension of a range-resolution cell centered at nadir (area on
ground inside the dotted line).
o
Figure 10 A pulse-limited altimeter will make multiple range

measurements within an antenna footprint that is related to the
amount of topographic variation. Depending on the algorithm used to
process the data, the reported elevation may correspond to the
highest (blue line), lowest (red line), or some average (yellow line)
elevation within the echo profile.
footprint have differential Doppler shifts, sophisticated signalprocessing algorithms can be used to improve resolution.
The Doppler shift is the change in frequency of the radar
signal when an object is moving relative to the radar antenna.
(For the radar systems of interest in this chapter, it is the
platform that is moving and generating the relative motion
between the ground and platform). Objects moving toward
the radar have compressed or smaller wavelengths, hence a
higher frequency, whereas objects moving away from the radar
have longer wavelengths, hence lower frequency, as illustrated
in Figure 11.
The magnitude of the frequency shift is proportional to
the projection of the velocity vector of the radar onto the lineof-sight vector. Thus, objects located perpendicular to the
velocity-vector have zero Doppler shift. Explicitly, the Doppler
53
Table 4 Cassini radar-altimeter parameters
Figure 11 Objects moving relative to an antenna experience a

frequency shift called the Doppler shift. Objects moving toward the
radar have higher frequencies, since the reflected waves are
compressed relative to the transmitted wave (reflected wave shown
in blue), whereas objects moving away from the radar have lower
frequencies, since the received wavelengths are elongated compared
to the transmitted waves (reflected wave shown in red).
Parameter
Value
Symbol
Radar parameters
Bandwidth
Wavelength
PRF
Range resolution
Antenna beamwidth
Number of pulses
4.24886 MHz
2.1 cm
5000 Hz
15.0 m
0.351
8
B
l
PRF
Dr
y3dB
Np
Platform parameters
Platform altitude
Along-track velocity
Radial velocity
5000 km
2.0 km/s1
4.5 km/ss1
hp
v
r_
Backscatter function parameters

Backscatter function
RMS surface slope
Dielectric constant
Gaussian
101
3
k
e
shift, f, is given by
2
2
v p , c^ S vp cosysq
f /!
l
l
15
where !
v p is the velocity vector, vp is the magnitude of
x ,!
yS
the velocity vector, c^ is the unit line-of-sight vector, /!
!
!
represents the dot product between two vectors x and y , and
ysq is the angle between the velocity vector and line-of-sight
vector called the squint angle. Points in the beam forward of
the radar have positive Doppler shifts since the radar is
moving toward them, whereas points in the beam aft of the
radar have negative Doppler shifts, as the radar is moving
away from them. By combining data from multiple pulses and
using signal-processing techniques similar to those employed
in SAR, it is possible to refine the along-track resolution by
separating points by their Doppler shifts. This process is called
Doppler-beam sharpening. The increase in along-track
resolution depends on how many pulses are combined, as is
discussed in greater detail in Section 3.3.4.
The precision to which an altimeter can measure elevation
is roughly equal to the range resolution of the sensor. Somewhat better performance is possible in cases where the scattering from the surface can be well modeled (e.g., ocean
surfaces), where more sophisticated algorithms can be employed to resolve the range to a fraction of range-resolution
element or range bin. Other factors contributing to the elevation measurement accuracy are the signal-to-noise ratio
(SNR, i.e., how strong the radar signal is in comparison to
background thermal noise and the accuracy of the platform
position). With the advent of global position system (GPS)aided navigation systems, the position accuracy is normally
not a dominating factor in the overall altimeter accuracy for
Earth-based systems, but represents a larger component for
planetary altimeters.
3.3.3.2
Example Echo Profile
To better understand the complexity in an altimeter-echo

profile, it is useful to see more explicitly how various terms
combine to form the returned signal. The signal returned from
each resolution element represents the backscatter-weighted
average over that element. This means if a resolution element

on the ground is broken up into small subareas, the subareas
with the largest backscatter will contribute most to an echoprofile measurement from that range. Since the returned
power level is a function of the antenna gain, backscatter
function, incidence angle, and distance to the antenna, all
these terms factor into the returned signal. Here, we exhibit a
couple of simulated echo profiles from an altimeter with
characteristics similar to the Cassini-radar altimeter used to
map the surface of Titan to illustrate these points. Figure 9
shows the notional imaging geometry and Table 4 provides a
list of the radar-altimeter parameters.
Using eqn [13] and values for the altitude and antenna
beamwidth (recall it is equal to l/L) in Table 4, it follows that
the Cassini-radar altimeter is a pulse-limited system. Therefore, it is possible to generate an echo profile with multiple
range samples within an antenna footprint. To generate
simulated altimeter returns, the area within the antenna
footprint is divided into many small subareas that are much
smaller than either the pulse-limited dimension or the alongtrack resolution of a resolution cell. Subareas within a resolution cell will be called a scattering element, or scatterer for
short. The amount of backscattered power from each scatterer
is computed from the radar equation below.
The range from the platform to a scatterer denoted by, r, is
simply the distance from scatterer to the platform. The intensity or power, P, contributing to the returned signal for each
scatterer is computed using a simplified version of the radar
equation given by
P ks
so yi G2 ZdS
r4
16
where so is the surface backscatter function, yi is the incidence

angle, G is the antenna gain, Z is the angle off boresight, dS is
the area on surface within the scattering element, and ks is an
overall scaling factor that is function of the transmitted power
and various gains and losses in the system.
Each of the terms in the radar equation has a simple
physical basis. To understand the r4 term in the denominator,
consider a microwave that is radiating isotropically. At a distance r from the point of origin of the wave, the energy is
distributed evenly on a sphere of that radius. As the radiated
energy is evenly distributed over a sphere of radius equal to the
range, and since the area of a sphere of radius R is 4pR2,
the power drops off, as 1/r2. Since this effect occurs for both
the transmitted and reflected waves, the resulting overall reduction in power is 1/r4. In Section 3.3.2, we learned that
antennas angularly direct the energy that results in nonisotropic radiation. The gain factor, G, accounts for the directivity of the antenna (i.e., how much the radiation is focused in
a particular direction), and it is squared in the radar equation
to account for two-way transmission to and from the scatterer.
The final two terms determine the strength of the reflection
from the surface. Somewhat obviously, the larger the amount
of surface area reflecting the energy, the larger the return
should be. The surface area over the scattering element appears
as a simple multiplicative factor because of the way the backscatter function, so, is defined. The backscatter function is a
measure of how much energy is reflected from the surface back
toward the radar. It is a function primarily of the surface reflectivity, which is in turn a function of the complex dielectric
constant, and the local surface roughness or slope distribution
at wavelength scales. Backscatter from a surface depends on the
angle between the incident wave and the surface that is parametrized in terms of the incidence angle. Backscatter functions vary quite dramatically depending on the surface
composition, and many models, both theoretical and empirical, have been developed for various types of surfaces.
Finally, we specify a simple backscatter function to illustrate how it affects the reflected signal. We chose a standard
Gaussian backscatter model where the backscatter, as a function of incidence angle, is given by
2
so yi gm
egtan yi
cos4 yi
Backscatter (dB)
54
0
4
8
12
16
0
1
,
tan2 k
p
2
e1
p
e1
18
where k is the RMS surface slope within a scattering element

and e is the dielectric constant. Values for the RMS surface
slope and dielectric constant used in the echo-profile simulations are given in Table 4, and Figure 12 shows a plot of the
backscatter function as a function of incidence angle. Note the
reduction in backscatter as the incidence angle increases (this
is typical of the backscatter of many surfaces).
For the first example, we consider the echo profile for a
perfectly flat surface. The left side of Figure 13 shows an image
of the amplitude, including antenna gain, backscatter, and
range effects, for each scattering element within the antenna
footprint and extending into the sidelobe region. On the right
is a plot of the range difference for the scattering elements
within the footprint. From the graphic, we can see there is
approximately 300 m of range variation within the antenna
footprint. With the 15-m range-resolution size, this corresponds to approximately 20 range samples (300/15) in the
echo profile. Eight pulses were simulated including both
20
thermal noise added to each pulse. The lower part of Figure 13

shows the echo power for the eight pulses and the average
profile obtained from these pulses.
Several features of the echo profile should be noted. First,
the leading edge of the echo profile (i.e., location in the profile
where intensity rises out of the background noise) corresponds to the range closest to the radar or the highest elevation
within the antenna footprint. Using the leading edge to determine the resulting elevation value gives the unambiguous
value expected for this example. Determining the location of
the leading edge is simple for this terrain; however, finding the
leading edge can be quite complicated for more complex terrain. Moreover, it is not always clear that this the best method
for assigning an altitude, as the next example illustrates. The
width of the average echo profile is fairly narrow in this example and roughly corresponds to the range extent of the
beamwidth of the antenna, Drext, which is given by
Drext
16
Figure 12 Plot of the Gaussian backscatter function as a function of

incidence angle. Note the reduction in backscatter as the incidence
angle increases.
17
where
8
12
Incidence angle (deg)
s

hp y3dB 2
hp y23dB
hp E
h2p
2
8
19
This equates to 23 m using the parameters in Table 4. Thus, a

wider beam or increased altitude results in a wider echo profile
and increased difficulty in extracting the location of the
leading edge.
For more complex topography, we consider terrain consisting of sinusoidally rolling hills or dunes in the along-track
direction using a height of 200 m and a spatial wavelength
of 2 km. This terrain has a maximal along-track slope of approximately 321. Figure 14 shows the intensity and range
variation images for the sinusoidal terrain. Observe in the
intensity image the horizontally stripped banding of the
scattering element intensities. This banding results from
changes in the backscatter function, as the incidence angle
varies on the terrain. The points of maximal return occur when
the tangent plane to the terrain is perpendicular to the nadir
direction (i.e., points where the incidence angle is zero, which
occurs at the peaks and troughs of the terrain). Thus, the
highest and lowest points in the terrain contribute the most to
echo intensity in this example. The delta-range image also
exhibits the same horizontal banding corresponding to the
200-m variation in height of the terrain. Echo profiles for eight
Intensity image
55
Delta range image
range (m)
600
500
400
300
200
100
0
Compressed echo profile
1.0
Cassini altimeter simulation

flat terrian
P1
P2
P3
P4
P5
P6
P7
P3
AVG
Normalization power
0.8
0.6
0.4
0.2
0
4999.5
5000
Range (km)
5000.5
Figure 13 Echo profiles and the average profile for eight simulated pulses over flat terrain. The top-left image in this figure shows the amount
of signal reflected back toward the radar for scattering elements within the antenna footprint. Lighter-colored areas correspond to stronger
returns. Since the range is smallest and the antenna gain is largest, the greatest reflected signal occurs at the center of the beam. The top-left
image shows the variation in range within the antenna footprint where distance is color-coded via the scale at right. The bottom graph shows the
echo profile, signal strength as a function of range, for the simulated flat terrain. Note, the echo power is centered and sharply at the distance or
range of the platform to the surface. In this case, we get a nice unambiguous measure of surface elevation.
pulses and their average are shown at the bottom of Figure 14.
Observe that the width of the profile is now approximately
300 m, a factor of 10 larger than in the previous example.
Here, the leading edge of the echo profile is fairly well defined
since it originates from the high-intensity portion of the echo
coming from the peaks of the hills. The profile is double
peaked with the second peak corresponding to the troughs,
and is considerably more complex than the flat terrain example. Obviously, more complex terrain will result in more
complex profiles.
Generically, given an echo profile, it is not possible to fully
reconstruct the terrain elevations for all points within the
antenna footprint. Note that the same underlying terrain can
yield many different echo profiles simply by spatially varying
the backscatter function. Different backscatter functions arise
naturally from different terrain-cover types (e.g., vegetated, wet
vs. dry, and lava or sand). Generally, one must be content with
providing an elevation value and some metric(s) of the terrain

variability with the antenna footprint. Deciding which elevation value to pick and what metrics to use depends on the
type of surface being mapped with the altimeter.
In cases where the terrain being mapped is fairly homogeneous and a well-defined backscatter function can be identified, it is possible to generate a library of template-echo
profiles that can then be compared with actual ones to determine the best match in a least-square or maximal-likelihood sense. Terrain characteristics are then assigned based on
the parameters used to simulate the echo profile that best
matches the actual echo profile. An example of this paradigm
was Magellan, a National Aeronautics and Space Administration (NASA) mission that used radar to map the surface of
Venus. In that case, the echo profiles were generated based on
RMS surface slope and radar reflectivity using a Hagfors
backscatter model. Echo profiles were inverted to provide an
56
Intensity image
Delta range image
range (m)
600
500
400
300
200
100
0
Compressed echo profile
1.0
Cassini altimeter simulation

Dunes: 2000 m spacing - 200 m height
P1
P2
P3
P4
P5
P6
P7
P8
AVG
Normalizd power
0.8
0.6
0.4
0.2
0
4999.5
5000
Range (km)
5000.5
Figure 14 Simulated altimeterecho profile for sinusoidal hills of 200 m in height. Echo power shown in the upper right shows that in addition
to the range and antenna-gain variation we had in the first example, the backscatter variation due to changes in local slope leads to a horizontal
banding of the returned power. Similar banding is seen in the range-variation image shown in the top left. Echo profiles for eight pulses and their
average are plotted in the bottom figure. This echo profile is more complex than the flat terrain example. The shape of the profile has a doublepeaked appearance, where the peaks correspond to the higher-powered returns from the peaks and valleys in the terrain. More complex terrain
variation exhibits more complex echo profiles, requiring more sophisticated signal-processing algorithms to assign an elevation value.
elevation measurement with 20-km resolution as well as terrain reflectivity and RMS surface slope at that scale (Ford and
Pettengill, 1992).
3.3.3.3
Radar altimeters have been used to map the Earth as well as

several planetary bodies including Venus and Titan. Most
radar altimeters used by remote-sensing scientists are flown on
orbiting platforms. Due to the nadir-pointing geometry and
limited swath and orbital constraints, radar-altimeter tracks
rarely consist of contiguous tracks providing complete coverage of the planet. A notable exception is the Magellan radar
altimeter that generated topographic data for nearly the entire
surface of Venus at 2040-km resolution. Nevertheless, radar-
altimeter data can be used in many geomorphological studies.

Earth-orbiting radar altimeters, for example, Topex/Poseidon
and Jason I altimeters have been used primarily to measure
sea-surface heights (e.g., Nerem et al., 2006) and in determining Earth geoid parameters (e.g., Nerem et al., 1994).
As an example of applying radar-altimeter data to geomorphology, we consider the Cassini-radar altimeter. Cassiniradar data have been applied to understand the surface
morphology of Titan from planetary length scales down to
wavelength scales. This shows the versatility of microwave remote sensing to obtain multi-scale topographic information
that characterizes the morphology of the terrain over very large
length scales.
Large-scale planetary shape parameters can be computed
from radar-altimeter data. Given altimeter tracks with sufficient geographic distribution on a planetary body, it is
57
Elevation (m)
800
Altimeter
Monopulse
400
0
400
800
90
North latitude (deg)
45
45
90
360
270
180
West longitude (deg)
90
Figure 15 Altimeter and radar monopulse-derived topographic measurements of Titan (elevations relative to a reference sphere of 2575 km)
used to estimate the ellipsoidal and higher-order shape parameters of Titans surface. Reproduced from Zebker, H.A., Stiles, B., Hensley, S.,
Lorenz, R., Kirk, R.L., Lunine J., 2009. Size and shape of Saturns moon Titan. Science 324(5929), 921923, with permission from AAAS.
possible to solve for low-order shape parameters such as triaxial ellipsoid axes and their orientation relative to the spin
axis (Zebker et al., 2008, 2009). Using a combination of radar
altimetry tracks and surface topography derived from monopulse measurements (Stiles et al., 2009) as shown in Figure 15, spherical, bi-axial, and tri-axial ellipsoidal fits to
Titans surface were made. The best-fit tri-axial ellipsoid to
Titan has axial parameters of 2575.15 and 2574.78 km in the
equatorial plane, and 2574.47 km in the direction of the spin
axis. These parameters describe the shape of Titan at the
planetary scale.
Depending on the density of tracks, higher-order fits are
possible (e.g., Legrendre polynomials), providing greater details on the global-shape distribution. Shape parameters can in
turn be used to estimate geophysical quantities relating to the
interior density distribution. Indeed Legendre polynomial fits
to seventh order were generated (Zebker et al., 2009) and used
to characterize higher-order facets to Titans shape. These
global shape parameters were compared with a planetary
model for Titans shape, assuming it were in hydrostatic
equilibrium for a body at its present orbital location and rotation state. Its observed larger oblateness suggested that the
icy crust has appreciable rigidity.
Long tracks of altimeter data are ideally suited to characterizing the power- spectral-density (PSD) function of the
topography. From the PSD, other quantities such as structure
functions and fractal dimension can be computed. PSDs can
be computed for spatial scales ranging from the length of an
altimeter track, hundreds-to-thousands of kilometers down to
the resolution of the altimeter, and several-to-tenths of
kilometers depending on the altitude. Along-track slope information, also at these scales, can be estimated providing
constraints on surface composition and geophysical processes.
Comparisons of simulated echo profiles to actual terrain can
be used to further constrain morphology at the subresolution
length scales, albeit these inferences may in general impose
broad constraints.
Information at the wavelength and sub-wavelength scales
can be derived from altimeter-backscatter data. These estimates
can be obtained using altimeter data alone via backscatter
models or in conjunction with backscatter and radiometer data
at other incidence angles to obtain better estimates. Titans
surface parameters, dieletric constant, and surface roughness
were estimated using a combination altimeter, scatterometer
(backscatter measured at a variety of incidence angles), and
radiometer data (Zebker et al., 2009). The estimated surface
dielectric constants varied from 1.75 to 2.5, consistent with a
surface composed of solid hydrocarbons or CO2 with considerable small-scale structure. Inferred RMS surface slopes at
the 10100-wavelength scale (Cassini has a 2.1-cm wavelength)
were in the range of 10161. Additional modeling of the
scattering further indicated volume scattering from millimetersized bubbles embedded in the ice or structures similar to ice
pipes and lenses occurring in terrestrial ice sheets.
3.3.4
Synthetic-Aperture Radars
In the early 1950s a technique that enabled the generation of

fine-resolution imagery using radar was developed called SAR.
58
irectio
d
track
Along
Ra
ng
di
re
ct
io
Figure 16 A typical SAR-imaging geometry has a platform containing a radar instrument moving in the along-track direction and imaging the
terrain to one side of the flight path. The SAR transmits a series of pulses at regular intervals along track that simultaneously illuminates an area
in the along-track direction, much greater than the desired azimuth resolution. By recording the returned echo from each pulse, and using signalprocessing techniques to synthesize a larger antenna, fine resolution in azimuth is achieved. The blue square in the center of beam shows the
size of a resolution element compared with the illuminated area from a single pulse indicated in green.
Since then many applications both civilian and military have

been developed that make use of SAR imagery.
3.3.4.1

Lsyn = 2
Laz
Theory of Operation
SAR can be used to produce fine-resolution imagery from either air-borne or space-borne platforms (Raney, 1999; Kovaly,
1976; Curlander and McDonough, 1991). Unlike optical
sensors operating at wavelengths between 3 nm and 30 mm,
such as photogrammetric or hyperspectral systems that form
images from reflected solar radiation, SAR systems transmit
their own radiation and record the signals reflected from the
terrain. With optical systems, images are generally formed
instantaneously, whereas for SAR, data collected from multiple
pulses along the flight path are required in order to achieve
useful resolution in the along-track, or azimuth, direction.
(Exceptions in optical systems include scanning optical systems such as Landsat where the optics are scanned and the
image is generated one line (or pixel) at a time. This is not a
fundamental sensor constraint, that is, if adequate lenses and
optical-recording technology are available, then a full twodimensional image could be made instantaneously.) Rather
sophisticated image processing is required to form recognizable images from the raw data. The resolution and quality of
the imagery depend on a number of system parameters, as
well as how the data are collected and processed.
SAR takes advantage of the motion of the platform to
synthesize a large antenna that may be many hundreds of
meters in length to achieve fine along-track resolution. Figure 16 shows the typical SAR-imaging geometry with the SARplatform moving along in flight. The radar antenna points in a
direction perpendicular to the flight path called the range or
cross-track direction imaging the terrain below. At approximately regular intervals along the flight path, the radar transmits a signal called a pulse, and then records the returned echo.

Lfp =
Laz
Figure 17 An object (yellow triangle) is being illuminated by an
antenna of azimuth dimension Laz at multiple points along the flight
trajectory. The synthetic-aperture length is the distance over which
the object is illuminated and is equal to twice the azimuth-footprint
size on the ground.
Azimuth resolution is achieved by synthesizing a large

antenna from the echoes received from a sequence of pulses
illuminating a target. Without signal processing, the intrinsic
azimuth resolution from a single transmitted pulse would be
the azimuth-angular width of the antenna beam times the
range. By combining the echoes using appropriate signalprocessing algorithms from all the pulses imaging a point, the
azimuth resolution is dramatically improved. Azimuth resolution after processing is determined by the size of the synthetic aperture (or antenna), which is the length of flight track
over which a fixed point stays within the azimuth-antenna
beamwidth. Figure 17 shows an object (yellow triangle) at
range, r, being illuminated by an antenna of length, Laz, from

multiple points along its trajectory.
The distance that the object is illuminated, called the
synthetic-aperture length, Lsyn, is twice the azimuth antennafootprint size and is given by
Lsyn 2Lf p 2r
l
Laz
20
The corresponding azimuth resolution, Dssyn, that is achieved

from a synthesized antenna of length, Lsyn, is given by
Dssyn r
l
Lsyn
l
Laz
2rl=Laz
2
21
Note that the resolution is simply half the azimuth-antenna

length, independent of range! This remarkable feature of SAR
systems enables them to make very fine resolution imagery at
large distances with moderately sized antennas. If data are not
combined over the full synthetic-aperture length, then there is
a corresponding reduction in the along-track or azimuth
resolution. The process of forming the synthetic aperture to
achieve the increased azimuth resolution is called azimuth
compression.
A natural question that arises is if there is a way to achieve
an even finer azimuth resolution than half the azimuth-antenna length. An alternate question is if there is some resource
savings that the system could incur if it did not need the full
azimuth resolution. The answer is yes to both questions, and
depends on the operational mode of the SAR that may be one
of the following:
59
data-rate reduction are achieved at the expense of azimuth

resolution and a reduction in SNR.
Spotlight mode: In spotlight mode, the radar transmits a
sequence of pulses while directing the antenna to a specified
point on the ground. This increases the amount of time an
object is within the antenna beam and consequently the
length of the synthetic aperture, and thereby the azimuth
resolution. A consequence of this mode of operation is that
areal coverage is limited compared with the long-continuous
imagery obtained in strip mode.
Figure 18 illustrates the basic modes of SAR operation.
During the time a target is in the beam, the range to the
target is changing from pulse to pulse. After generating an SAR
image, we identify a targets location in the image by its azimuth or along-track coordinate and range position as shown
in Figure 19. To select a unique range from the family of
ranges that are changing from pulse to pulse during the synthetic aperture, the angle from the velocity vector to the target
(or equivalently the Doppler frequency see eqn [14]) is
specified for processing, as shown in Figure 20.
The bold-dashed line from pulse N-2 to the target indicates
the desired Doppler (or equivalently angle) at which the target
will be imaged. Observe that the selection of the Doppler
frequency affects not only the range at which a target is
Range
sequence of pulses over the area to be imaged. Achievable

azimuth resolution is half the along-track antenna length.
ScanSAR or burst mode: In ScanSAR or burst-mode systems,
the radar transmits a sequence of pulses at the pulse-repetition frequency (PRF) followed by a period of no transmission. This allows the radar to either scan to another
elevation angle during the no-transmission period, thereby
increasing swath width (ScanSAR mode), or remain silent,
thereby reducing the data rate. The increased swath-width or
Strip mode
Azimuth
Strip mode: In strip mode, the radar transmits a continuous
o
= o + j
Figure 19 Shown above is a target imaged in the j th range bin.
Burst mode
Spotlight mode
Figure 18 Figure illustrating three basic modes of SAR operation which are strip, burst, and spotlight modes. In strip mode (left image), the
radar emits a continuous stream of pulses indicated by the red rectangles that allow fine resolution to be achieved in long-imaging strips. Burst
mode (middle images) transmits a short sequence of pulses (typically tens to several hundred) called a burst, followed by a suspension of
transmission usually of similar duration. This enables wider swaths by scanning to another location during the interburst period or a reduction in
the amount of recorded data at the expense of azimuth resolution. The finest resolutions are achieved by spotlight-mode systems that direct the
antenna at a point on the ground to keep in the beam for a longer period of time, and, hence, improve azimuth resolution.
60
N+3
N+2
N+1
N
N-1
N-2
N-3
Target
Figure 20 This figure shows the space shuttle imaging a fixed point on the ground from a number of pulses in a synthetic aperture. The range
at which a target appears in a synthetic-aperture image depends on the processing parameters and algorithm used to generate the image. For
standard range/Doppler processing, the range is fixed by choosing the pulse that has a user-defined fixed angle between the velocity vector and
the line-of-sight vector to the target. This is equivalent to selecting the Doppler frequency.
imaged, but also the corresponding position of the platform

(and hence azimuth location) when the target is imaged.
It is useful for our subsequent discussions to distill the
above information on SAR-image coordinates to the simple
geometry of the intersection of two surfaces. As discussed
earlier, range information is obtained by measuring the time it
takes a radar pulse to propagate from the antenna to a target
and return. Azimuth location is determined from the Dopplerfrequency shift that is related to the angle from the velocity
vector when a target is imaged. Viewing SAR-target location
geometrically, the range/azimuth location locus is the intersection of a sphere centered at the antenna, with radius equal
to the radar range and a cone with generating axis along the
velocity vector, with cone-angle proportional to the Doppler
frequency, as shown in Figure 21. A target in the radar image
could be located anywhere on the intersection locus, which is
a circle in the plane formed by the radar line of sight to the
target and vector pointing from the aircraft to nadir. Since the
intersection is a curve in three-dimensional space, further information is required in order to locate a target uniquely.
Since the range direction is not parallel to ground coordinates as shown in Figure 22, SAR images are distorted
relative to a planimetric view. In many applications, this distortion can adversely affect data interpretation, particularly
when one is not well acquainted with SAR imagery (Leberl,
1990). This distortion can be removed as part of the geocoding process if a digital elevation model (DEM) is available, or
from three-dimensional position data generated directly from
SAR-stereo or interferometric systems, at points where these
techniques produce useful data. These corrected images are
often easier both to interpret and to register with other data
layers that might be required for analysis. Three common
features observed in SAR imagery that warrants particular

mention include foreshortening, layover, and shadow.
Foreshortening in radar imagery results from the fact that
relief displacement is toward the direction of the radar. Because the range increases more slowly than ground coordinates on slopes facing toward the radar (higher elevations
contend with increasing ground-distance slowing the range
increase), they tend to appear bunched relative to a planimetric view. The opposite occurs on slopes facing away from
the radar (lower elevations coupled with increasing grounddistance speeds the range increase) where they tend to expand
out when compared to a planimetric view. Both situations are
illustrated in Figure 22. Equation [8] can be used to determine
the amount of foreshortening from the imaging geometry and
terrain slope in the range direction. Note that foreshortening
in radar images is opposite to that of optical imagery where
relief displacement is away from the direction of the camera.
Layover is a limiting case of foreshortening where points
arranged with increasing ground coordinates appear reversed
in the radar imagery. Layover occurs because the range to
objects with larger ground coordinates is less than the range to
other objects with smaller ground coordinates. Geometrically
this happens when the slope of the terrain is greater than the
angle the incident radiation makes with respect to vertical.
Also, we note that for interferometric or stereometric radar
systems, layover causes a loss of useful signal and therefore
precludes the determination of elevation in layover regions.
Shadow occurs when the radar beam cannot reach a portion of the terrain being imaged, because it is occulted by
other parts of the terrain or other objects in the scene. Where
the terrain is shadowed, the radar image will appear dark and
the signal in these range cells is only due to thermal noise.
61
Range sphere
Doppler cone
Aircraft
position
Velocity
vector
Assumed reference
plane
Scatter is assumed at the intersection of range
sphere, doppler cone and reference plane
Figure 21 Geometric perspective of target location in SAR imagery is illustrated above. Range information locates a target on a sphere centered
about the SAR platform and the Doppler frequency locates a target on a cone centered about the velocity vector. The intersection locus of these
two surfaces is a circle, and thus the three-dimensional location of a target cannot be uniquely determined from a single image.
Ground coordinates
Figure 22 The three-dimensional world is collapsed to two dimensions in conventional SAR imaging. After image formation, the radar return is
resolved into an image in rangeazimuth coordinates. This figure shows a profile of the terrain at constant azimuth, with the radar-flight track
into the page. Notional radar-backscatter values are shown in gray scale above the radar pixels that are delineated by dotted lines. Radar
foreshortening on slopes is indicated by the green and blue arrows above. On the slope facing toward the radar it is imaged in three radar pixels,
whereas the slope facing away from the radar is imaged in five radar pixels. Notional layover and shadow regions are indicated by the
semitransparent green and blue ellipses, respectively. Note the bright return indicated for the layover region and the dark return for the
shadowed region. Finally, the Leaning Tower of Pisa illustrates a classic layovered object where the top of the image appears nearer in range
than the base.
62
Minor PRF
2
2
Major PRF
Tx LH1
Rx LH1
Tx LV1
Rx LH1
Rx LV1
Rx LV1
Channel 1 data
LH1LH1
LV1LH1
Channel 2 data
LH1LV1
LV1LV1
Figure 23 Polarization diversity is achieved by transmitting H

polarization on alternate pulses (designated by a 1 above) and V
polarization on the adjacent pulses (designated by a 2 above). Four
images are formed from the data, an HH (H transmit, H receive) HV,
VH, and VV. From these four basic images, it is possible to
synthesize any other polarization combination.
As with layover regions, shadowed regions have no useful

signal and consequently no elevation values or other geophysical parameters for the surface can be estimated.
Because radar primarily interacts with structures that have
lengths comparable to the wavelength or larger, longer wavelength (lower frequency) radars tend to penetrate deeper into
the vegetation canopy or ground surface. The amount of
penetration in a vegetation canopy depends on the structure
and density of the vegetation. Radar wavelengths less than
roughly 10 cm mostly scatter off the upper portions of canopies, whereas wavelengths longer than 20 cm penetrate
deeper into a canopy. This differential penetration effect for
lower-frequency radars has led to the development of radar
systems designed to exploit this phenomenon. Ground-surface
penetration depends on the type and composition of the
ground layers, ground cover, and soil moisture. Longer-wavelength radars have been known to penetrate several meters or
more in dry sandy soil, and even deeper into certain types
of ice.
The vector nature of electromagnetic radiation can be exploited to gain further information about a surface. SAR systems that transmit and/or receive multiple polarizations of an
electromagnetic wave are called polarimetric SAR systems.
Systems that transmit and receive on two orthogonal polarizations (usually H and V) are said to be fully or quadpolarimetric. Most polarimetric-airborne systems are fully
polarimetric, whereas a number of polarimetric-space-borne
systems operate primarily using a restricted set of polarization
combinations (e.g., HH and HV or HH and VV) due to performance constraints. Polarization diversity is achieved by
transmitting H and V polarized waves on alternate pulses and
then receiving H and V data for every pulse as illustrated in
Figure 23.
Imagery from multiple polarizations is often combined to
form a false-color composite image (Figure 24). For example,
from the NASA/JPL L-band fully polarimetric radar over Death
Valley, CA, the color-composite image shown in Figure 24 was
formed with the red layer from an HH image, the blue layer
from the VV image, and the green layer from the HV image.
Color-coding of the image components was chosen to be
Figure 24 NASA/JPL uninhabited aerial vehicle synthetic aperture

radar (UAVSAR) color composite (red HH, blue VV, green HV)
image of polarimetric data collected over Death Valley, CA. Color
scheme is chosen to highlight various scattering mechanisms useful
in the delineation and classification of various surface types.
suggestive of particular scattering mechanisms that aid in

classification of the surface types, such as red for double
bounce, green for volume scattering, and blue for surface
scattering. In this image, we see that green represents the alluvium and the vegetated regions where volume scattering
dominates. Blue regions correspond to surfaces where the
radar is not penetrating and surface scattering dominates.
3.3.4.2
Radar images can be used for a variety of geomorphic applications that make use of both qualitative, photointerpretation
techniques and quantitative inversions of the data. Photointerpretation relies on the basic image elements including
tone and texture, which in radar images, are dominated both
by the imaging geometry and by the surface roughness and
dielectric constant. Because radar images are obtained by imaging to the side of the flight line, geometry is more important
in radar image interpretation than for interpreting opticalsensor data such as satellite imagery or aerial photography. As
described above, radar-image layover and shadowing occur for
extreme combinations of radar look angle and topography.
Small radar look angles combined with steep topography can
lead to layover, and large look angles combined with high
relief can cause shadows. Both cases lead to loss of data.
Outside of those extremes, radar images at small look angles
emphasize low relief, much like airphotos at low sun angle.
This is at the cost of distorting higher relief through foreshortening or layover. By contrast, radar images obtained at
larger look angles are more map like, and tone in these images
is more directly related to surface characteristics such as
roughness near the scale of the radar wavelength and dielectric
constant.
We illustrate image interpretation aspects of using SAR data
for geomorphological and geological application using data
collected by the NASA/JPL AIRSAR in Australia (Tapley and
Craig, 1995; Tapley, 1996, 1998, 2000). Here, polarimetric
multi-frequency, C-, L-, and P-bands (wavelengths 5.6, 24, and
70 cm, respectively) are combined in color-composite imagery
and filtered to highlight various geologic units. Figure 25

shows a three-frequency false-color composite image of the
Ophthalmia Range region of the Hamersley Basin in Western
Australia. Interpretation of the radar imagery relies on the
ability to distinguish the interplay between surface morphology and composition of landform units. The labels 1, 2, and
3 indicate several morphological elements that are discernable
in this imagery. Outcrops of the Ophthalmia Range have been
and continue to be a supply of an erosional material forming
both colluvial (location 1) and alluvial fans (location 2), as
well as sheetwash plains (location 3).
Figure 25 AISRAR three frequency false-color composite image

consisting of red from C-VV, blue from L-VV, and green from P-VV in
the vicinity of the Ophthalmia Range, Hamersley Basin in Western
Australia. Image dimensions are approximately 16 8 km.
HH
HV
VV
63
Another interesting application of the sensitivity of radar

images to slope angle occurs for sand dunes. As the angle of
repose for loose sand is about 331, it has been observed that
radar images of sand dunes at look angles smaller than that
value contain many bright spots caused by direct reflection
from the dune slip faces and other lower-angle surfaces (Blom
and Elachi, 1981). At larger look angles, the smooth sand
yields little scattering of the radar waves, as described in the
following.
Surface roughness, although not a quantity that most Earth
scientists consider routinely, is related to a number of geologic
characteristics. Roughness of lava flows, particularly the differences between pahoehoe and aa, reflects differences in
eruption and composition. Weathering of different lithologies
typically yields varying sizes of debris that can be detected
easily in radar images. Other surficial processes such as
mantling by aeolian dust, desert pavement formation, and salt
weathering, all serve to modify surface roughness (Farr, 1992;
Farr and Chadwick, 1996).
Radar imagery collected over the Big Island of Hawaii by
the NASA/JPL uninhabited aerial vehicle synthetic aperture
radar (UAVSAR) fully polarimetric L-band (23-cm wavelength)
radar is used to illustrate how both qualitative and quantitative aspects of surface roughness are reflected in polarimetric
imagery. Figure 26 shows HH, HV, and VV individual polarization images and a color-composite image from a 23-kmwide strip of data imaging the Kilauea volcano. Lava
flows, vegetation, volcanic craters, fault zones, and ocean
Composite
Figure 26 Individual polarization (HH, HV, and VV) and false-color composite imagery from the NASA/JPL L-band polarimetric radar collected
over the Kilauea volcano on the Big Island of Hawaii. To the right are expanded views of the three orange-boxed areas. Lava flows, vegetation,
volcanic craters, fault zones, and ocean are all easily differentiated using typical photo-interpretation methodologies.
64
Co-polarization signature
1
Normalized power
Normalized power
1
0.8
0.6
0.4
0.2
0
150
100

50
0
40
20
20
0.8
0.6
0.4
0.2
0
40
150
100

40
20
20
40
1
Normalized power
1
Normalized power
50
0.8
0.6
0.4
0.2
0
150
100

50
0
40
20
20
40
0.8
0.6
0.4
0.2
0
150
100

50
0
40
20
20
40
Figure 27 Co-polarized radar signatures generated for four surface types using the NASA/JPL UAVSAR L-band polarimetric image data shown
in Figure 26. The top-left signature is of the lava near the Puu Oo cone; top right is the signature of a forested area; bottom left is a signature of
ocean data; and, finally, the bottom right is a signature of a building. Note how the shape and pedestal height of the signature vary as a function
of the surface type. Pedestal height can often be linked to the amount of surface roughness, and the shape and number of peaks provide insight
into the scattering mechanism involved.
are all easily discerned using standard photo-interpretation

strategies. On the left are expanded views of the regions enclosed by the orange boxes. Lava flows (reddish-purple) and
volcanic craters and forested regions (green) are clearly visible
in the top two images. Wind-roughed ocean surfaces show up
as magenta-like color in the bottom image. Fault zones and
other geologic units are also clearly distinguishable in the
imagery.
More quantitative interpretations using polarization
signatures or polarimetric decompositions can also be used
to distinguish surface types and morphologies. Figure 27
shows the polarization signatures for lava near the Puu Oo
cone, a forested region, ocean, and building in the Kilauea
radar image shown in Figure 26. Polarization signatures
depict the normalized power return from the surface, as the
polarization is varied (i.e., as the tilt and ellipticity are
varied). Pedestal height of the polarimetric signatures can
generally be related to surface roughness and the shape of
the polarization signature can provide information concerning the dominant scattering mechanism (e.g., surface or
volume scattering), providing additional information about
the surface.
Vegetation appears rough to radar sensors and radar images

have been used extensively for vegetation type and density
mapping. In fact, quantitative relationships between biomass
and radar-image brightness have been developed and used for
mapping large regions (Saatchi et al., 2007).
Smooth surfaces (at the scale of the wavelength) show up
as dark surfaces in radar images because they scatter the incident waves away from the receiving antenna. Common examples of these types of surfaces include calm-water bodies
and smooth-sand sheets. An interesting effect occurs, however
for flooded-vegetated areas. Even if the flooded surface is not
visible due to vegetation cover, enough of the radar signal
penetrates the vegetation to reflect off of the smooth water
surface only to then scatter among the vegetation. This has
been observed to enhance the radar brightness of a flooded
forest (Hess et al., 1990).
Generally, radar waves do not penetrate significantly into
rock, soil, or ice surfaces. If the soil is very dry and uniform
(no large rocks or horizons) or the ice very cold, however,
radar images may be obtained of the subsurface. The amount
of penetration is also proportional to the wavelength and has
been documented most for L-band (25 cm) (Schaber et al.,
1997) and C-band (5 cm) (Robinson, 2002). Nonimaging

ground-penetrating radars, which operate at the surface at
wavelengths ranging from 25 cm to several meters, have been
used in the field for some time to characterize geologic and ice
volumes (Bristow and Jol, 2003). Radar-wave penetration has
been found to be important at Saturns moon Titan, where the
composition and temperature of the surface promotes penetration of the short (2.2 cm)-wavelength Cassini SAR up to
several meters.
Another, secondary effect on radar-image brightness is the
surface dielectric constant. On Earth, most geologic materials
have a range of dielectric constants from about 3 to 10. However, water has a dielectric constant of about 80 and so dominates dielectric constant differences. Thus, wet soils can
commonly be differentiated from drier soils and quantitative
models of radar scattering can be used to estimate soil moisture.
Scattering models allow more quantitative interpretations of
radar images in terms of surface roughness, structure, and
moisture content. Generally, additional radar-image parameters
are utilized in modeling including multiple wavelengths and
polarizations. Multiple wavelengths provide information about
multiple scales of roughness and the polarization data allow
inferences about surface or vegetation structure.
3.3.5
b
1
2
+
h
x1
x2
p
Figure 28 A typical optical-stereo viewing geometry is illustrated.
Two optical cameras view an object of height, h, from two distinct
vantages from which relative image displacements in the resulting
imagery are converted to elevation measurements.
b
1
2
Stereo SAR
We have seen that surface morphology can be obtained directly by examination of SAR imagery, or, more quantitatively
in some cases, from the backscatter measurements. In these
cases, however, a direct measurement of topography was not
obtained. This naturally leads to the question of whether
topographic information can be obtained directly from SAR
data. The answer is affirmative, provided imagery is acquired
from two spatially separated vantages having the appropriate
geometry.
pr
Theory of Operation
x1
x2
Figure 29 SAR stereo-imaging geometry where a scene is imaged
from two vantages with incidence angles y1 and y2.
that is related to height by

h
3.3.5.1
65
Dp
tany2 tany1
23
Measurement of topography using stereo techniques uses

relative displacement in imagery collected from two distinct
vantages to extract elevation information. This technique in
the SAR area is a natural extension of the classical stereo
techniques employed very successfully by the photogrammetry
community. Sensor geometry and differences in the nature of
optical versus SAR imagery result in some substantial differences in the processing required to generate topographic
maps. We start with a brief overview of optical-stereo geometry. Figure 28 shows a typical optical-stereo imaging geometry. Two cameras are positioned to view an area to be mapped
from with off-nadir angles of y1 and y2 (angles to left of nadir
in the figure are negative and angle to the right are positive)
typically in the along-track direction. An object of height, h,
above a reference image plane is displaced by amount Dxi
(i 1,2) as shown in Figure 28. Image parallax, defined as the
relative displacement of object as seen in the two images, Dp,
is equal to
Note that since y1 and y2 have opposite signs, eqn [23] is

well defined and reduces to Dp=2tany1 when the two vantages have viewing angles that are of equal magnitude.
Although a couple of distinct imaging geometries are
possible for SAR-stereo topography generation, we will concentrate on the cross-track stereo geometry that is most commonly employed in this section. Figure 29 shows the SAR
cross-track stereo imaging geometry.
Unlike optical images, where objects appear in imagery
based on the viewing angle, objects in SAR images are positioned based on range. An object of unknown height is located in an SAR image where the sphere of that range
intersects the image reference plane. Figure 29 shows an object of height, h, being imaged from two vantages. The range
sphere can be well approximated by a line perpendicular to
the line of sight, and the location of an object in an SAR image
is the intersection of that line with the reference surface as
shown in Figure 29. The radar parallax, Dpr, is the relative
position of an object in the two images which is given by
Dp Dx2 Dx1
Dpr Dx2 Dx1 hcoty2 coty1
22
24
66
that can be inverted for height to give

h
Dpr
coty2 coty1
25
Given eqn [25], several important observations about SAR

stereo can be made. The sensitivity of height to the parallax
measurement is given by
qh
1
q Dpr coty2 coty1
26
Thus, the larger the difference in incidence angles, the greater

is the height sensitivity. This relationship would seem to imply
that to maximize stereo-mapping accuracy, large incidenceangle differences are desired. However, another aspect of stereo processing dominates radar-stereo-derived elevation
measurement accuracy that constrains the incidence angle
difference selection. If the difference in incidence angles is not
too large (which is usually the case see discussion below on
automated scene matching), then eqn [26] can be approximated as
qh
sin2 y
Dy
q Dpr
27
generated via SAR stereo. When no a priori elevation data are

available, geocoding to a flat reference surface is done.)
Figure 31 shows data collected over Mt. Everest in Nepal by
shuttle radar SIR-C in 1994 at two incidence angles of 541 and
581. By collecting data with relatively large incidence angles,
layover problems were reduced at the expense of greater
shadowing in the imagery.
After geocoding, automated scene matching is used to
identify corresponding points in the two images. Careful
filtering of the match data is required to eliminate bad matches obtained from automated-matching algorithms. Often
pyramid- or hierarchical-matching schemes are employed that
start at low resolution and then go to progressively higher
levels of resolution until the desired resolution level is
reached. Filtering algorithms can then take advantage of the
various matching levels to eliminate bad matches. From the
matched points, the parallax can be computed and converted
into a height measurement. In practice, height determination
is done in a more rigorous fashion than simply using eqn [30].
An SAR sensor-imaging model that incorporates all the
measurements is used and solves for the three-dimensional
location of a point, which is discussed later in this section.
If a rasterized-elevation product is desired, the irregularly
where y is the average of y1 and y2, Dy y2 y1 and, therefore,

the elevation error, sh, in terms of the matching accuracy,
sDpr , is
sh
sDpr sin2 y
Dy
28
with typical automated stereo-matching algorithms having an

accuracy of 0.3 a pixel or larger and y between 301 and 401,
and Dy about 101, the height accuracy is about the pixel
size. This simple rule of thumb is often heard with regard to
SAR-stereo height accuracy, but it is predicated on several assumptions about the stereo geometry and matching accuracy.
Figure 30 shows the generic processing flow for SAR-stereo
topographic map generation.
Radar-stereo processing consists of four main steps. The
process begins by processing SAR imagery for each of the two
channels and then geocoding to some reference surface. The
reference surface may be a simple plane or some higherfidelity representation of the surface (e.g., a low-resolution
digital elevation model. (Usually, the reference surface has
much lower resolution than the product that is being
Raw
data
image 1
SAR image
formation
Raw
data
image 2
SAR image
formation
Incidence angles
5660
Incidence angles
5257
Figure 31 Stereo pair of Mt. Everest collected by the shuttle radar
SIR-C in 1994.
Ephemeris
data
Geocode
Low
resolution
dem
Scene
matching
Match
filtering
Height
determination
Geocode
Rasterize to
desired grid
Figure 30 The four main steps in the generic processing flow used to generate SAR-stereo topographic maps.
spaced points obtained in the height-determination

process must be resampled or interpolated onto a chosen
map grid.
Scene matching of SAR imagery is the limiting factor affecting stereo elevation-estimation accuracy. There are two
primary mechanisms that limit the matching accuracy. First,
the two vantages with different incidence angles result in
image distortions related to SAR image formation. For example, foreshortened and shadow regions have different dimensions due to the difference in incidence angles. Moreover,
the location of the brightest point on a slope facing toward the
radar may come from different physical locations that can be
misidentified in the matching process, as illustrated in
Figure 32.
The larger the incidence-angle difference, the greater the
relative distortion between the images and the larger the
matching error. However, reducing the incidence-angle difference results in decreased height acuity and, hence, the optimal incidence-angle difference is a balance between better
matching accuracy and height acuity. Studies have shown incidence-angles differences of about 101 result in the best SAR
topographic maps.
The second factor limiting matching accuracy is SAR
speckle. Speckle is point-to-point random fluctuations in
the image magnitude that is a result of using a monochromatic imaging source. This type of noise is not present in
optical imagery and results in reduced matching accuracy of
67
SAR data relative to optical data. It can be shown that the

magnitude of single-look SAR imagery is exponentially distributed, meaning that the mean and standard deviation are
equal. The random fluctuations are uncorrelated between
images when the incidence-angle difference exceeds a critical
threshold that is typically the case for SAR stereo-image pairs.
Speckle noise is reduced via spatial averaging of pixels that
results in reduced spatial resolution. The amount of spatial
averaging to reduce speckle noise needs to be balanced against
having higher-resolution data for the matching process.
Nominally, the minimum amount of spatial averaging commensurate with a significant reduction in speckle noise
is done.
Low-contrast, shadowed, and layover regions generally are
not mapped by SAR stereo. Selecting the incidence angles used
in mapping can minimize the area affected by shadow and
layover. SAR stereo collections in mountainous areas from
opposite sides can also reduce dropout area from shadow and
layover. Low-contrast featureless terrain precludes good stereo
matches and these areas are better mapped via other techniques such as SAR interferometry. Figure 33 shows the output
of an automated hierarchical matching algorithm for the SIRC Mt. Everest data. The matching algorithm also outputs a
match quality metric (estimated covariance error in pixels)
that is used in the filtering algorithm and to generate a heighterror map for stereo-derived heights. (The height error (more
accurately height precision) map is the estimated statistical
2
1
Images from 1
Images from 2
Mismatched
object
Figure 32 Illustration of the matching ambiguity that can occur with the radar stereo-imaging geometry. In this example, a hill is illuminated
from two different incidence angles with the brightest returns in their associated images corresponding to the points on the hill near normal
incidence. An automated matcher can misidentify the bright streaks as the same points and thereby lead to error in the SAR-stereo elevation.
75 pixel contour wrap
68
Radar stereo disparity
Matching error
Figure 33 Disparity or parallax maps generated from the output of

an automated matching and filtering algorithm applied to the SIR-C
Mt. Everest data.
Mt Everest
N
6000 m per color cycle
Figure 35 UTM projection of the SIR-C stereo-derived DEM of Mt.
Everest.
Elevation error
6000 m contour wrap
25
Elevation
Elevation error
Figure 34 Elevations and elevation precision (height-error map)

derived from SIR-C radar-stereo data over Mt. Everest. Color
contours of elevations with a 6000-m color wrap and height errors
with a 25-m color wrap are overlain on radar-backscatter data.
height accuracy of a point, and not the actual height error

relative to the reference map. This layer is very valuable when
using elevation data for quantitative analysis, or filtering data
that contain points not meeting a desired map accuracy for a
particular application.)
Equation [30] relates the height of a point in terms of its
observed parallax; however, for precision mapping applications, a set of equations based on direct SAR imaging
measurements is used. SAR imaging from two vantages gives
rise to a set of four nonlinear equations for the target-position
vector that will be referred to as the nonlinear stereo equations. These equations geometrically represent the intersection
of two range spheres centered at the platform positions when
the target was imaged, and the corresponding two Doppler

cones. Exact closed-form solutions are available to these
equations; however, the exact solution requires the use of only
three of the four equations (Hensley, 2008). Use of all the
measurement data from the four stereo equations is possible
using least-square techniques. There are a couple advantages
to using least squares to obtain the stereo-solution vector.
First, by using all the measurement data and by providing
input-error covariance information on the matching accuracy,
platform ephemeris, and range data, a formal position-error
estimate can be obtained for every point in the map. This can
be useful when deciding the appropriateness of radar stereoderived elevation measurements to particular geomorphological applications. A second benefit is that corrections to the
range for atmospheric delay and bending can be done as part
of the least-square iterative procedure. Figure 34 shows elevations and estimated elevation precisions derived from the
SIR-C stereo data collected over Mt. Everest. Elevations in the
scene varied from 3500 m to 8800 m. Formal height-precision
estimates that varied from 15 to 50 m were verified using
higher-accuracy radar interferometric data. Average estimated
elevation precision was 25 m with a standard deviation 5.3 m.
Finally, Figure 35 shows the radar-derived DEM in a
standard map projection (Universal Transverse Mercator
(UTM)).
3.3.5.2
Radar-stereo data have not been used extensively for Earthbased geomorphological investigations primarily due to the
ready availability of more accurate topographic data obtained
from optical or radar-interferometric techniques. However,

planetary investigations, particularly Venus and Titan, have
benefited from radar-stereo data, which to this point is the
only available means of obtaining high-resolution topographic data. Because the applications and interpretation
techniques for stereo SAR are similar to optical stereo, the
extensive literature on the use of optical-stereo images will not
be repeated here. Main differences between optical- and radarstereo-derived topographic data are related to the geometric
differences between side-looking radar images and nadirlooking optical images. The presence of radar image speckle
and geometric distortions induced by the incidence-angle
difference discussed earlier are a major factor limiting stereo
topographic mapping accuracy. This contributes to high spatial frequency noise in the resulting topographic maps that can
cause difficulty interpreting small features and can yield
anomalously large values for small-scale slopes.
Radar stereo has been used to constrain impact crater
morphology on Venus (Cochrane and Ghail, 2006). The
Magellan radar systematically mapped Venus in the early
1990s owing to its optically opaque atmosphere, precluding
other means of obtaining high-resolution imagery of its surface. Mapping was conducted in three cycles, each in one
Venus day or 243 earth days in duration. On the first and third
imaging cycles, left-looking radar-stereo data were obtained
for approximately 30% of the planets surface before recorder
problems forced the suspension of further mapping operations. Stereo data from approximately 421 N to 441 S latitude
and 511E to 901 latitude containing the Aphrodite region were
used to study the morphology of impact craters. By comparing
depth-to-diameter trends, the authors found that the data were
not in accordance with that predicted from theory. They found
that craters smaller than 18 km in diameter were unexpectedly
shallow, leading them to conclude that midsize bolides
(2001000 m in diameter) fragment in the dense Venusian
atmosphere and then disperse over an area 1020 times the
diameter of the bolide diameter.
3.3.6
Interferometric SAR
3.3.6.1
Theory of Operation
By augmenting a conventional SAR system with another spatially

separated receiving antenna, as illustrated in Figure 36, it is possible to extract topographic information. More details on interferometric synthetic-aperture radar (IFSAR) systems and
processing can be found in Bamler and Hartl (1998), Franceschetti and Lanari (1999), Madsen and Zebker (1999), Rosen
et al. (2000), and Hensley et al. (2001a). By receiving data on
both antennas, it is possible to extract fine-resolution topographic
information from either space-borne or air-borne platforms.
Up to this point, we have considered SAR images to consist
of only a brightness or magnitude value at each image pixel. In
actuality, SAR-pixel values are complex numbers consisting of
both a magnitude and phase. (Although glossed over when
polarimetric-imaging radars were introduced, phase information is also critical to the understanding and use of polarimetric data. We suppressed this complication until the
introduction of interferometry where it is essential to
69
H
Swath width
Figure 36 Interferometric SAR for topographic mapping uses two

apertures separated by a baseline to image the surface. The phase
difference between the apertures for each image point, along with the
range and knowledge of the baseline, can be used to infer the precise
shape of the imaging triangle (in red) to determine the topographic
height of an object.
Phase =
2
4
1
2
2
D
Radians per wavelength
Number of Wavelengths
Figure 37 Distance in a radar system is converted to a phase

measurement by dividing the distance by the radar wavelength to get
the distance in wavelengths and then multiplying by 2p, the number
of radians per wavelength to get to phase. Measured-phase values in
a radar image do not contain the total number of wavelengths but
only retain the fractional portion of a wavelength. Typically, phase in
a radar interferometric image is color coded as shown above, and
one wavelengths worth of phase change is called a fringe.
obtaining a basic understanding of the measurement concept.) Phase information in a single SAR image encodes subwavelength range information that normally is not accessible.
Extraction of this sub-wavelength range information is only
possible by using image pairs satisfying geometric imaging
constraints. There are two primary components to the phase
measurement consisting of a systematic and a random part.
The systematic part is the range to the resolution element that
is converted to a phase value by dividing by the wavelength
(gives distance in unit of wavelengths) and then multiplying
by 2p, which is the number of radians in one wavelength as
illustrated in Figure 37. Even though there may be many
millions of wavelengths (hence many multiples of 2p) from
70
Range from antenna

Phase =
+
to center of cell
Systematic component desired
by interferometric measurement
Coherent sum of elemental

+
scatters arranged randomly in cell
Random component that if look direction is nearly the same
and scatterers within cell do not move relative to each other
this component cancels in the interferogram formation process.
Thermal noise
Random component that does not cancel
and results in interferometric phase noise
Figure 38 Phase is composed of a systematic component that encodes the range difference between the two antennas and point, and a
random part that consists of both random thermal noise and random-arranged elemental scatterers within a resolution element.
the antenna to a resolution element, only the principal value

(a number between p and p) can be extracted from a
complex-valued resolution element.
The random component may also be thought of as consisting of two components summarized in Figure 38. The first
component, thermal noise, corrupts all electronic measurements and is present in every pixel value. Its impact on the
phase depends on the relative magnitude of the reflected signal
magnitude to the noise level. The second component is geometric in nature and depends on the arrangement of so-called
elemental scatterers, points within a resolution element that
dominate the signal value, and their contribution depends only
on the viewing geometry and the composition of the scatterers.
In general, because the distribution of elemental scatterers
within a resolution element changes from element to element
and the range converted to phase modulo 2p (subtracting off
the integer multiples of 2p) to a resolution element is randomly
distributed, phase values in SAR images are also randomly
distributed. It is important for the interferometric SAR discussion to follow to note that if the viewing geometry is nearly
unchanged and the elemental scatterers within a resolution cell
are undisturbed, then this portion of the random phase remains the same. Thus, the thermal noise is random in time and
the elemental scatterer noise is random in space.
As mentioned previously by using two images sufficiently
similar in imaging geometry and combining the radar data
coherently, the elemental-scatterer portion of the interferometric phase cancels out, leaving only the difference of the
systematic portions of the phase. Thus, interferometry recovers
the sub-wavelength range information encoded in the phase
from two images to extract differential-range information with
an accuracy of a fraction of a wavelength. The interferometricphase difference is essentially related to the geometric pathlength difference to the image point, which depends on the
topography as is shown in Figure 39. With knowledge of the
interferometer geometry, the phase difference can be converted into an elevation measurement for each image point.
Differential-radar interferometry uses observations separated
in time to measure sub-wavelength changes in the surface
topography. Observations may be separated by time intervals
ranging from fractions of a second to many years.
Previously, it was shown that knowing the SAR coordinates
of a point constrained its location to be on the intersection
locus of a sphere and cone that from Figure 21 was seen to be
a circle. Parametrizing the location on this circle by an angle,
referred to as the elevation angle, y, reduces the three-dimensional location problem to determining this angle. For
this, we need the interferometric measurement. Given two
identical antennas, A1 and A2, are receiving radar-echo signals
from a single point as shown in Figure 40.
Baseline
D2
D1
L4
L3
L1
L2
Differential ranges
D2 = L4 L3
D1 = L2 L1
Figure 39 Topographic information can be extracted from the

differential-distance encoded in the interferometric phase obtained
from two antenna separated in distance in the cross-track direction
called the baseline. D1 and D2 are the differential distances measured
by the interferometric phase for two points at different elevations.
Since the elevation affects the differential distance and hence the
phase, it contains topographic information.
A2
B

A1
2

h
1
h
Figure 40 Simplified broadside looking (zero Doppler) radarinterferometry geometry. The difference in range from the two
observing antennas to the target is approximately equal to the
projection of the baseline vector onto the line-of-sight vector shown
in blue. This range difference can be related to a phase measurement
and forms the primary interferometric observable.
The path-length difference, Dr, of the signals received by

the two antennas is approximately given by
Dr r2 r1 Ebsiny a
29
where ri indicates the range from antenna i to the target, b is

the length of the baseline vector which is the vector pointing
from antenna 1 to antenna 2, y is the desired elevation (or
look) angle, and the baseline-orientation angle, a, is the angle
the baseline vector makes with respect to the horizontal.
Observe that the range difference to a good approximation for
most systems is simply the length of the projection of the
baseline vector onto the line of sight. The range difference, Dr,
may be obtained by measuring, f, the phase between the two
interferometer signals, using the relation
f
2ppDr
,
l
p 1,2
30
Range sphere
where l is the radar wavelength, and p equals 1, when the

path-length difference is associated with the one-way path
difference, or 2 for the two-way path difference, as is the case
for Ping-Pong or repeat-pass systems described below. Geometrically, the phase measurement represents a second cone
with cone axis aligned with the interferometric baseline.
Intersecting the phase cone with the range sphere and Doppler
cone determines the elevation angle to the target, and therefore the targets full three-dimensional location, as shown in
Figure 41.
Using the simplified geometry of Figure 39, the height of a
target, ht, is given by
ht h r cosy,
Baseline
vector
SAR
position
Velocity
vector
Phase cone
Target location

lf
a:
2ppb
32
It is given from eqns [31] and [32] that determining the

height of a target requires knowledge of the platform position,
the range, the interferometric baseline length, the baseline
angle, and the interferometric phase. Generation of accurate
topographic maps using radar interferometry places stringent
requirements on the knowledge of the platform and baseline
vectors. Figure 42 shows the interferometric-phase measurements and amplitude image for data collected by the SIR-C
imaging radar over the Mojave Desert in California.
It is important to appreciate the consequences of the fact
the interferometric-phase measurement is made modulo 2p.
The total-range difference between the two observations
that the phase represents in general can be many multiples of
the radar wavelength, or expressed in terms of phase, many
Azimuth
Figure 41 Target location in an SAR image could be anywhere on

the intersection of a range sphere and Doppler cone, thereby
providing no information on the targets elevation. 3-D information is
obtained by the intersection of the phase cone with range sphere and
Doppler Cone.
31
where h is the altitude of the radar antenna and r is the slant

range from the antenna to the target. Since the signal phase is
sensitive to displacements between images of a fraction of a
wavelength, the interferometric technique provides a very
accurate means of determining topographic heights. Using
eqns [29][31] the elevation angle can be determined to be
y sin1
Doppler cone
71
Range
Figure 42 Interferometric phase (right) and associated amplitude image (left) of the Mojave Desert, CA, generated from repeat-pass
observations using the SIR-C radar. Observe that the phase values soon on the right mimic topographic contours and hence can be converted
into topographic measurements.
72
multiples of 2p. It is this value that is required in order to

make height measurements. The standard approach for determining the unique phase that is directly proportional to the
range difference is to first determine the relative phase between pixels via the so-called phase unwrapping process.
Unwrapping of IFSAR imagery is a nontrivial process for
which a number of algorithms have been developed. Complications arise in avoiding unwrapping errors in regions of
shadow, layover, and low signal return. The connected phase
field after unwrapping may still need to be adjusted by an
overall constant of 2p. The step that determines the overall
constant of 2p is referred to as absolute phase determination.
Interferometric correlation, a measure of the similarity of
the signal received at the two antennas, can be estimated
directly from the image data of the two interferometric
channels (Zebker and Villasenor, 1992). Correlation measurements have values between 0 and 1, with 1 designating
perfect correlation between the channels. Sometimes it is more
convenient to refer to the amount of interferometric decorrelation, which is defined as one minus the correlation. The
amount of decorrelation due to the slightly different viewing
geometry is called geometric decorrelation. Thermal noiseinduced signal decorrelation is called noise decorrelation.
Shadowed regions suffer from noise decorrelation and areas
on steep slopes exhibit geometric decorrelation that increases
phase noise and can preclude useful phase measurements
altogether. Another form of decorrelation occurs when there is
a vertical distribution of scattering elements within a resolution element as shown in Figure 43. Not only is the signal
decorrelated, but also the point within the resolution cell
corresponding to the interferometric-phase measurement depends on the wavelength and the scatterer distribution in the
cell. This form of decorrelation is called volumetric decorrelation and can be used to infer information about the vertical structure of the volume. Recently, there has been a great
deal of activity using volumetric correlation to estimate tree
and canopy structure within the interferometric SAR community (Cloude and Papathanassiou, 1997; Hensley et al.,
2001a; Hensley et al., 1995; Hokeman and Varekamp, 2001;
Hagberg et al., 1995).
There is an upper value on the useful baseline length
known as the critical baseline (Rosen et al., 2000). The critical
baseline is reached when the amount of phase change per
resolution element exceeds 2p radians. This limitation is a
result of the fact that the interferometric-phase measurement
is made only modulo 2p. As the baseline approaches the
critical baseline, the phase values from the two antennas become completely decorrelated. However, as the baseline increases, the sensitivity of phase to height increases improving
the accuracy of interferometric SAR systems. Practical mapping
systems must select baselines with a balance between adequate
phase to height sensitivity to meet mapping requirements
and excessive decorrelation with corresponding processing
difficulties.
Combining the SAR-image formation process, interferometric-phase measurement, unwrapping, and height determination into an automated processing algorithm require a
process flow that is shown in Figure 44 (Madsen et al.,
1993b). The first processing step is decoding the byte data,
followed by range compression for each of the two interferometric channels. Using the platform-motion information obtained from inertial navigation unit (INU) and GPS
measurements, as well as any other baseline metrology devices, the data are compensated for perturbations in aircraft
motion from a reference path and then azimuth compressed.
This generates two single-look complex images. (The process
of correction for motion away from an ideal path is referred to
as motion compensation. Motion compensation corrects for
motion on the order of fractions of a wavelength in order to
generate well-focused SAR imagery. The requirements for
motion compensation are even more demanding for IFSAR
systems and place stringent requirements on the platform and
baseline metrology systems.)
One of the single-look complex image pairs is resampled to
overlay the other. This registration must be done to a small
fraction of a pixel (typically o0.1 of a pixel) in order to avoid
phase decorrelation. Multiplying the complex pixel value in
one image by the complex conjugate of the corresponding
pixel value in the second image forms an interferogram. The
resulting interferogram is multi-looked, by spatially averaging
the complex pixels in a box about a given pixel to reduce the
amount of phase noise.
After the multi-looked interferogram has been generated,
the phase for each complex sample is computed. To generate a
continuous height map, the two-dimensional phase field must
be unwrapped. After the unwrapping process, an overall
multiple of 2p is estimated and added to the unwrapped phase
(the estimated value may be 0).
Subsequent to determining the absolute phase for each pixel
in the interferogram and possibly taking additional looks, the
three-dimensional target position can be determined. Phase
corrections are applied to the interferometric phase to account
Look direction
Figure 43 Vertical structure of scatterers within a resolution element due to vegetation or other structures present in the cell causes increased
decorrelation. This form of decorrelation can be used to infer information about vertical structure within a resolution element. The increased
decorrelation results from the increased size of the range cell projected back toward the direction of the radar (shown in magenta) when
compared to a flat surface.
Decode and range

compression
Raw signal
data antenna 1
Raw signal
data antenna 2
73
Azimuth
compression
Decode and range

compression
Azimuth
compression
Image resample
and interferogram
formation
Looks
NAV data
Calibration
data
Baseline
measurement
Coarse DTE
Absolute phase
determination
Phase unwrapping
Phenomenological
studies
Troposphere
model
Looks
SAR
signal processing
Post
processing
3 dimensional
location and gridding
Module for which
exact position in
processing is
algorithm dependent
Primary output
product generation
Ancillary
data
Mosaicking
Used if needed
by selected
algorithm
Figure 44 Interferometric processing block diagram.
for tropospheric effects, and the range is corrected to account

for changes in the speed of light through the atmosphere. Using
accurate baseline and platform position information, the phase
and range information for the target position is computed.
A relief map is generated by gridding the unevenly sampled
three-dimensional target locations into a desired coordinate
system. The gridded products include the target elevations, the
SAR image (which has been orthorectified), a correlation map,
and a height-error map described below. The resulting radarrelief map may be measuring the heights above the ground,
within the vegetation canopy, or beneath the surface in arid
regions. To convert this map into a true ground-surface DEM,
corrections based on phenomenological studies (e.g., using
scattering or semi-empirical curves to correct elevation measurements based on the amount of decorrelation in the canopy)
must be incorporated either into the three-dimensional
location algorithms or into a post-processing step.
One of the unique aspects of interferometric-SAR systems is
the ability to determine the statistical-height precision, that is
the degree of height noise from pixel to pixel, estimated from
knowledge of the correlation, g (Hensley and Webb, 1994). The
amount of phase noise between the two channels is simply and
directly related to the correlation and number of looks used to
reduce phase noise. (The CramerRao bound (expresses a lower
bound on the variance of estimators of a deterministic parameter) used to relate the phase noise to correlation and
number of looks is only valid when the number of looks exceeds four or five. The number of looks in most interferometric
systems used to generate topographic maps usually is much
larger than four.) The CramerRao bound relating the phase

variance, sf, to the correlation coefficient, g, is given by
s
1
1 g2
33
sf p
g2
2NL
where NL is the number of looks. From eqns [32] and [33] the
height error, sh, as a function of the phase noise is found to be
sh
lrsiny
sf
2ppbcosy a
34
Equations [33] and [34] allow the generation of an elevation-precision map showing the local height precision for
each post in an interferometrically derived DEM (Zebker et al.,
1994; Madsen et al., 1993a).
Differential-radar interferometry uses repeat-pass observations separated in time by seconds to years to measure
change in the surface under observation. This technique can
measure surface displacements at the millimeter level over
large areas, making it ideal for a variety of geophysical investigations. Differential interferometry has become an accepted tool in the geophysical community for mapping surface
deformation from natural sources such as earthquakes, volcanoes, and glaciers, and from anthropogenic sources such as
oil or groundwater pumping. Differential interferometry
works because the radar is far more sensitive to change than it
is to topography. Ideally, for surface deformation measurements, one would like to have repeat passes with zero
74
baseline, and, therefore, have no topography sensitivity. In

that case, the only differential-range change observed would
be due to surface movement directed along the radar line of
sight, as illustrated in Figure 45. Thus, the phase measurement
due to surface change is simply given by
f
4p
4p ! ^
Dr
/ d ,c S
l
l
35
!
where Dr / d , c^ S is the differential range shift of the surface along the line-of-sight vector, c^ , which is simply the
!
projection of the surface-displacement vector, d, onto c^ :
In general, it is not possible to make a repeat-pass measurement with a zero baseline, so the interferometric measurement is in general a combination of both surface change
and a topographic measurement. The topographic-phase

component is computed and subtracted from the measured
phase using a DEM and knowledge of the interferometric
baseline. The smaller the baseline the less accurate the
DEM needs to be to remove the topographic component. Atmosphere changes, particularly those due to tropospheric
water vapor, change the propagation speed in the atmosphere
and represents a major-error component to differential-interferometric surface-displacement measurements (Goldstein,
1995; Tarayre and Massonnet, 1996). Changes in the surface
at the wavelength scale within a resolution element lead to
temporal decorrelation, which can either degrade the quality
of the phase measurements or lead to the complete loss of
signal. Large disruptive events, such as earthquakes, generally
have severe decorrelation on areas of the fault where there is
surface rupture or buildings have been damaged. Most differential-radar-interferometric measurements have been made
with satellite systems because obtaining knowledge regarding
irregular aircraft trajectories at the required accuracy (millimeter level) is considerably more challenging.
3.3.6.2
Figure 45 Surface deformation at the millimeter scale can be

measured over large spatial scales and fine resolution using the
technique of differential-radar interferometry. In the above figure,
SAR imagery is acquired a first time over a volcano. A subsequent
collection, which may be minutes to years later, is collected at a
second time spanning a deformation event to be monitored. In this
illustration, deformation is surface inflation on the flank of the
volcano. If the plane flew exactly the same trajectory, then the
change in range to a point on the deforming flank would be solely
due to surface motion. This surface motion can be extracted from the
interferometric phase using differential-radar interferometry. Repeat
passes with nonzero baselines need to remove the topography term
in the phase using an external DEM and knowledge of the baseline.
Perhaps the single biggest contribution of single-pass radar

interferometry for topographic mapping was the Shuttle Radar
Topography Mission (SRTM) that flew on the Space Shuttle in
2000 (Farr et al., 2007). This mission produced the first
globally consistent fine-resolution (30 and 90 m products)
topographic map of the Earths surface between 7601 latitude.
The shuttle radar consisted of a C-band (and X-band
radar supplied by the German Space Agency) with one antenna in the cargo bay and second antenna deployed on a 62m boom. Accuracy assessments based on GPS and other
reference data showed the SRTM DEM has 8.8-m accuracy at
the 90% confidence level (Rodriguez et al., 2006). The
short C-band wavelength does not fully penetrate vegetation,
as was discussed earlier, so that the resulting DEM was a
reflective surface DEM (i.e., mapped elevations depend
on terrain cover). Techniques for exploiting or correcting
for this electromagnetic elevation bias are a continuing
area of research (Sexton et al., 2009; Carabajal and Harding,
2006; Hofton et al., 2006). This data set has become
the standard for world topography and has been used in
many scientific applications, including numerous uses in
geomorphology.
An aspect of repeat-pass differential interferometry relevant to
geomorphology is its ability to measure subcentimeter changes in
topography over large areas (Madsen and Zebker, 1999; Massonnet and Feigl, 1998; Smith, 2009; Simons and Rosen, 2007). The
rapid infusion of differential-radar interferometric techniques into
geomorphology and geophysics has its genesis with the measurement of the 1992 Landers earthquake using repeat pass C-band
radar interferometry from the ERS-1 satellite (Massonnet, 1993).
For the first time it was possible to see the displacement field from
a earthquake at fine resolution and large spatial scales. Figure 46
shows an interferogram of the Landers Earthquake, generated
using two ERS-1 radar passes that spanned the earthquake. Surface
displacement is encoded in the interferometric phase, and each
fringe or color contour represents 2.8 cm of surface displacement
First pass
75
Second pass
D1
D2
D1 Surface deformation causes

D2
Range to change
Surface deformation = D2 D1
Phase =
Surface deformation
Topography
Small baseline facilitates
compensation of this term
Desired signal
Figure 46 Repeat-pass radar interferometry from the ERS-1 satellite shows the surface deformation produced by the Landers Earthquake in
California. This was the first major geophysical event studied by Massonet et al. (1993) that introduced radar-interferometic techniques to the
geophysical community. Each fringe or contour level represents about 2.8 cm of surface displacement. This marked the first time when it was
possible to see wide-scale surface morphology changes due to an earthquake at such fine resolution.
along the radar line of sight. Note that the fringe frequency increases in the vicinity of the fault, and the fringes are absent in
many areas in proximity to the fault, mostly as a result of temporal
decorrelation. Displacements can be measured to millimeter precision using radar-interferometric techniques and can be combined with other sensor information (e.g., GPS measurements to
gain detailed information on the processes involved in surface
modification). An extensive literature exists on the application of
this technique for studying earthquake deformation (e.g., Massonnet et al., 1993), volcano inflation and deflation (e.g.,
Lundgren and Lu, 2006; Masterlark and Lu, 2004; Zebker and
Amelung, 2000; Mouginis-Mark et al., 2000), landslides (e.g.,
Roering et al., 2009), and subsidence (e.g., Bell et al., 2008;
Fielding et al., 1998).
Note that this technique will not work on surfaces that
have changed too much at the sub-resolution scale, such as
sediment deposition and mass-wasting processes, which
would decorrelate the phase returns much like moving vegetation. The decorrelation phenomenon can be turned to an
advantage when change or disruption of a surface is the desired signal. Areas of temporal decorrelation are easily detected
and have been used to map small landslides and damaged
buildings after an earthquake (Fielding et al., 2005) and active
lava flows (Zebker et al., 1996).
We also provide an air-borne surface-deformation example
of anthropogenically induced surface deformation using the
NASA/JPL UAVSAR radar (Hensley et al., 2009, 2010). UAVSAR
collected two repeat passes of 74 km in length over Lost
Hills, CA, on May 6 and July 25 of 2008 flying at an altitude
of 12.5 km. This area is known to be undergoing surface
deformation due to oil pumping based on satellite radarinterferometric and GPS measurement (Fielding et al., 1998).
The area is mostly agricultural interspersed with light urban
development, and thus many areas remain highly correlated for

large temporal baselines at L-band. The two passes were separated by 80 days. Obvious from the radar interferogram, there
are two areas undergoing surface deformation centered over
light urban areas. Figure 47 shows the line-of-sight displacement obtained from the unwrapped interferometric phase, with
a spatial resolution of about 7 m (36 looks). L-band has good
correlation over the urban development and in some fallow
agricultural regions. The maximal displacement is about 7 cm,
corresponding to a deformation rate of about 0.88 mm d1.
Figure 48 shows the plots of the displacement along selected
transects highlighted in Figure 47 that are about 3 km in
length. Note the asymmetric nature of the displacement profiles
along transects A and B that would be difficult to discern in
satellite imagery due to their coarser resolution.
3.3.7
Summary
The remarkable diversity of active microwave sensors and their

applicability to geomorphological applications have been
presented. Microwave sensors have the ability to inform
studies at length scales ranging from planetary scales down to
surface roughness at the centimeter scale. This tremendous
range of applicability of microwave sensors is a result of the
variety of sensor types and the large frequency range over
which these sensors operate.
The type of measurements from various sensor types can be
grouped into four basic categories and include:
Backscatter
data: Backscatter represents the amount of

microwave energy reflected from the surface back to the
sensor. The magnitude of the backscatter signal depends on
the surface composition through the dielectric constant and
76
0.02 0 0.02 0.04 0.06 0.08 0.1

LOS displacement (m)
Figure 47 Oil pumping in the oil fields near Lost Hills, California, produces surface deformation at the rate of about 1 mm d1. This level of
surface deformation is detectable using either space-borne or air-borne radar interferometry. Black areas are regions where there was no
interferometric signal due to temporal decorrelation. Orange-boxed regions of surface deformation are shown zoomed in the images on the right.
12
Transect A
Transect B
Transect C
LOS displacement (cm)
10
8
6
4
2
0
0.5
1.0
1.5
2.0
Distance (km)
2.5
3.0
Figure 48 The color of the arrow corresponds to the color used to plot the displacement along that transect.
3.5
on the wavelength-scale surface roughness within a resolution element. Using polarimetric diversity, incidence-angle
diversity, or modeling, it is sometimes possible to extract
either quantitative or qualitative surface roughness and/or
composition information.
Image data: SAR systems are capable of generating fineresolution imagery of a surface. Using image-interpretation
techniques and knowledge of radar-specific imaging idiosyncrasies (e.g., foreshortening, layover, and shadow), it is
possible to identity and classify many features of geomorphological interest such as faults, lava flows, craters,
alluvial fans, etc.
Topographic data: Altimeters and stereo and interferometric
SARs are capable of making direct measurements of surface
topography. Radar altimeters are nadir-pointing instruments
that are effectively profiling instruments that measure the
elevation beneath the flight track. Stereo and interferometric
SAR systems, on the other hand, measure surface elevations
over an extended swath and are effectively used for large area
fine-resolution topographic map generation. The nature of
the elevation measurement (e.g., whether bare surface or reflective surface elevations are being mapped) depends on the
wavelength of the sensor and how the data are processed.
Surface-deformation data: Repeat-pass radar interferometry
is capable of making millimeter-precision surface-deformation measurements over large areas at fine resolution from
either air-borne or space-borne platforms. Over the last twoand-a-half decades these measurements have become integral to understanding the surficial and geophysical processes
that continue to shape the Earths surface.
Because these sensors operate over a variety of wavelengths
ranging from a few millimeters to a meter, and acquire data
over a range of spatial scales from one to thousands of kilometers, these data are essential for a variety of geomorphological applications on both the Earth and other bodies in the
solar system.
As these sensors continue to proliferate and evolve, along
with computer systems and algorithms, microwave remote
sensing will increasingly make important contributions to
geomorphology.
References
Bamler, R., Hartl, P., 1998. Synthetic aperture radar interferometry. Inverse Problems
14, R154.
Bell, J.W., Amelung, F., Ferretti, A., Bianchi, M., Novali, F., 2008. Permanent
scatterer InSAR reveals seasonal and long-term aquifer-system response to
groundwater pumping and artificial recharge. Water Resources Research, 44. http
://dx.doi.org/10.1029/2007WR006152.
Blom, R., Elachi, C., 1981. Spaceborne and airborne imaging radar observations of
sand dunes. Journal of Geophysical Research 86, 30613073.
Bristow, C.S., Jol, H.M., 2003. Ground penetrating radar in sediments. Geological
Society. Special Publications, London, 211, 328.
Carabajal, C.C., Harding, D.J., 2006. SRTM C-band and ICESat laser altimetry
elevation comparisons as a function of tree cover and relief. Photogrammetric
Engineering and Remote Sensing 72, 287298.
Cloude, S., 2010. Polarisation: Applications in Remote Sensing. Oxford University
Press, Oxford.
Cloude, S.R., Papathanassiou, K.P., 1997. Polarimetric optimization in radar
interferometry. Electronics Letters 33(13), 11761178.
Cochrane, C., Ghail, R., 2006. Topographic constraints on impact crater
morphology on Venus from high-resolution stereo synthetic aperture
77
radar digital elevation models. Journal of Geophysical Research 111,

E04007.
Curlander, J.C., McDonough, R.N., 1991. Synthetic Aperture Radar Systems and
Signal Processing. Wiley, New York, NY.
Farr, T.G., 1992. Microtopographic evolution of lava flows at Cima volcanic field,
Mojave Desert, California. Journal of Geophysical Research 97, 1517115179.
Farr, T.G., Caro, E., Crippen, R., et al., 2007. The Shuttle Radar Topography
Mission. Reviews of Geophysics, 45,doi:1029/2005RG000183.
Farr, T.G., Chadwick, A.O., 1996. Geomorphic processes and remote sensing
signatures of alluvial fans in the Kun Lun Mountains, China. Journal of
Geophysical Research 101, 23,09123,100.
Fielding, E.J., Blom, R., Goldstein, R., 1998. Rapid subsidence over oil fields
measured by SAR interferometry. Geophysical Research Letters 25,
32153218.
Fielding, E.J., Talebian, M., Rosen, P., Nazari, H., Jackson, J., Ghorashi, M.,
Walker, R., 2005. Surface ruptures and building damage of the 2003 Bam, Iran,
earthquake mapped satellite synthetic aperture radar interferometric correlation.
Journal of Geophysical Research, 110. http://dx.doi.org/10.1029/2004JB003299.
Franceschetti, G., Lanari, R., 1999. Synthetic Aperture Radar. CRC Press, New York,
NY.
Ford, P., Pettengill, G., 1992. Venus topography and kilometer-scale shapes.
Journal of Geophysical Research 97, 1310313114.
Goldstein, R.M., 1995. Atmospheric limitations to repeat-track radar interferometry.
Hagberg, J.O., Ulander, L.M.H., Askne, J., 1995. Repeat-pass SAR interferometry
over forested terrain. IEEE Transactions on Geoscience and Remote Sensing
33(2), 331340.
Hensley, S., 2008. A combined methodology For SAR interferometric and
stereometric error modeling. Proceedings of Radarcon 2008.
Hensley, S., Chapin, E., Freedman, A., et al., 2001a. First P-band results using the
GeoSAR mapping system, Geoscience and Remote Sensing Symposium, IGARSS
01. IEEE 2001 International 1, 126128
Hensley, S., Gurrola, E., Harcke, L., et al., 2009. Lunar topographic mapping
using a new high resolution mode for the GSSR radar. Proceedings of Radarcon
2009.
Hensley, S., Klein, J., Rosen, P.A., Chapin, E., Madsen, S.N., Webb, F.H., 1995.
Repeat pass aircraft interferometry results at Portage Lake, Maine, and Innisfail,
Australia. Proceedings of the AIRSAR Science Workshop.
Hensley, S., Muniy, R., Rosen, P., 2001b. Interferometric synthetic aperture radar.
In: Maune, D. (Ed.), Digital Elevation Model Technologies and Applications: The
DEM users Manual. American Society for Photogrammetry and Remote Sensing
(ASPRS), Bethesda, Maryland, Ch. 6.
Hensley, S., Webb, F.H., 1994. Comparison of long valley TOPSAR data with
kinematic GPS measurements. IGARSS Proceedings, Pasadena, CA.
Hensley, S., Zebker, H., Jones, C., Michel, T., Muellerschoen, T., Chapman, B.,
2009. First Deformation Results Using The NASA/JPL UAVSAR Instrument. 2nd
Annual Asia Pacific SAR Conference, Xian, China, October.
Hensley, S., Zebker, H., Jones, C., Michel, T., Muellerschoen, T., Chapman, B.,
2010. Use of airborne SAR interferometry for monitoring deformation of largescale man-made features. Proceeding of the Large Scale Deformation of Manmade Structures, Hong Kong.
Hess, L.L., Melack, J.M., Simonett, D.S., 1990. Radar detection of flooding beneath
the forest canopy: a review. International Journal of Remote Sensing 11,
13131325.
Hofton, M., Dubayah, R., Blair, J.B., Rabine, C., 2006. Validation of SRTM
elevations over vegetated and non-vegetated terrain using medium footprint lidar.
Photogrammetric Engineering and Remote Sensing 72, 279285.
Hokeman, D., Varekamp, C., 2001. Observation of tropical rain forest trees by
airborne high-resolution interferometric radar. IEEE Transactions on Geoscience
and Remote Sensing 39(3), 584594.
Kovaly, J., 1976. Synthetic Aperture Radar. Artech House, Boston, MA.
Leberl, F.W., 1990. Radargrammetric Image Processing. Artech House, Boston, MA.
Lee, J.S., Pottier, E., 2009. Polarimetric Radar Imaging: From Basics to
Applications. CRC Press, New York, NY.
Lee, J.S., Schuler, D.L., Ainsworth, T.L., 2000. Polarimetric SAR data compensation
for Terrain Azimuth slope variation. IEEE Transactions on Geoscience and
Remote Sensing 38(5), 21532163.
Lundgren, P., Lu, Z., 2006. Inflation model of Uzon caldera, Kamchatka, constrained
by satellite radar interferometry observations. Geophysical Research Letters, 33.
http://dx.doi.org/10.1029/2005GL025181.
Madsen, S.N., Martin, J.M., Zebker, H.A., 1993a. Analysis and evaluation of the
NASA/JPL TOPSAR across-track interferometric SAR system. IEEE Transactions
on Geoscience and Remote Sensing 33(2), 383391.
78
Madsen, S.N., Zebker, H.A., 1999. Synthetic aperture radar interferometry: principles
and applications. In: Henderson, F., Lewis, A. (Eds.), Manual of Remote
Sensing. Artech House, Boston, MA, vol. 3, ch. 6.
Madsen, S.N., Zebker, H.A., Martin, J., 1993b. Topographic mapping using radar
interferometry: processing techniques. IEEE Transactions on Geoscience and
Massonnet, D., Feigl, K.L., 1998. Radar interferometry and its application to
changes in the Earths surface. Reviews of Geophysics 36, 441500.
Massonnet, D., Rossi, M., Carmona, C., Adragna, F., Peltzer, G., Feigl, K., Rabaute,
T., 1993. The displacement field of the Landers earthquake mapped by radar
interferometry. Nature 364, 138142.
Masterlark, T., Lu, Z., 2004. Transient volcano deformation sources imaged with
interferometric synthetic aperture radar: Application to Seguam Island, Alaska.
Mott, H., 2007. Remote Sensing with Polarimetric Radar. Wiley, Hoboken, NJ.
Mouginis-Mark, P.J., Crisp, J.A., Fink, J.H., 2000. Remote Sensing of Active
Volcanism. AGU Geophysics Monograph 116, 272.
Nerem, R., Leuliette, E., Cazenave, A., 2006. Present-day sea-level change: a review.
C. R. Geoscience 338, 10771083.
Nerem, R.S., Lerch, F.J., Marshall, J.A., et al., 1994. Gravity model development for
TOPEX/POSEIDON: joint gravity models 1 and 2. Journal of Geophysical
Research 99(C12), 24,42124,447. http://dx.doi.org/10.1029/94JC01376.
Raney, K., 1999. Radar fundamentals: technical perspective. In: Henderson, F.,
Lewis, A. (Eds.), Manual of Remote Sensing. Artech House, Boston, MA, vol. 3,
ch. 2.
Robinson, C.A., 2002. Application of satellite radar data suggest that the Kharga
depression in south-western Egypt is a fracture rock aquifer. International
Journal of Remote Sensing 23, 41014113.
Rodriguez, E., Morris, C., Belz, E., 2006. A global assessment of the SRTM
performance. Photogrammetric Engineering and Remote Sensing 72(3),
249260.
Roering, J.J., Stimely, L.L., Mackey, B.H., Schmidt, D.A., 2009. Using DInSAR,
airborne LiDAR, and archival air photos to quantify landsliding and sediment
transport. Geophysical Research Letters 36, L19402. http://dx.doi.org/10.1029/
2009GL040374.
Rosen, P.A., Hensley, S., Joughin, I.R., Li, F.K., Madsen, S., Rodriguez, E.,
Goldstein, R.M., 2000. Synthetic aperture radar interferometry. Proceedings of
the IEEE 88(3), 333382.
Saatchi, S., Houghton, R., Avala, R., Yu, Y., Soares, J.-V., 2007. Spatial distribution
of live aboveground biomass in Amazon Basin. Global Change Biology 13,
816837.
Schaber, G.G., McCauley, J.F., Breed, C.S., 1997. The use of multifrequency and
polarimetric SIR-C/X-SAR data in geologic studies of Bir Safsaf, Egypt. Remote
Sensing of Environment 59, 337363.
Sexton, J.O., Bax, T., Siqueira, P., Swenson, J.J., Hensley, S., 2009. A comparison
of lidar, radar, and field measurements of canopy height in pine and hardwood
forests of southeastern North America. Forest Ecology Management 257,

11361147. http://dx.doi.org/10.1016/j.foreco.2008.11.022.
Simons, M., Rosen, P.A., 2007. Interferometric Synthetic Aperture Radar Geodesy,
Treatise on Geophysics.
Smith, L.C., 2009. Emerging applications of interferometric synthetic aperture radar
(InSAR) in geomorphology and hydrology. Annals of the Association of
American Geographers 92, 385398.
Stiles, B.W., Hensley, S., Gim, T., et al., 2009. Determining Titan surface topography
from Cassini SAR data. Icarus 202(2), 584598.
Tapley, I.J., 1996. An assessment of airborne synthetic aperture radar (AIRSAR) data
as an aid to iron exploration in the Hamersley Basin, Western Australia. CRC
LEME Restricted Report 12R, CSIRO Exploration and Mining Report 282R.
Tapley, I.J., 1998. Landform, regolith and geological mapping in the
Northeastern Goldfields region, Western Australia, and North Drummond Basin,
Queensland, using AIRSAR polarimetric data, CRC LEME/AMIRA Project 392.
CRC LEME Report 55, CSIRO Exploration and Mining Report 147R, Volumes 1
and 2.
Tapley, I.J., 2000. Geological mapping and terrain analysis in Australia using
AIRSAR polarimetric and interferometric radar data. Report 1 The recognition
of geologic units within the Flinders and Olary Ranges, South Australia, CRC
LEME Report 154R, Volumes 1 and 2.
Tapley, I.J., Craig, M.D., 1995. An evaluation of AIRborne Synthetic Aperture Radar
(AIRSAR) for mapping surface and sub-surface structures in the Telfer region,
Paterson Province, Western Australia. CSIRO/AMIRA Project 392, CSIRO
Exploration and Mining Report 146R, Volumes 1 and 2.
Tarayre, H., Massonnet, D., 1996. Atmospheric propagation heterogeneities
revealed by ERS-1 interferometry. Geophysical Research Letters 23(9),
989992.
Zebker, H.A., Amelung, F., 2000. Jonsson, Remote sensing of volcano surface and
internal processes using radar interferometry. AGU Geophysical Monograph 116,
179205.
Zebker, H.A., Rosen, P.A., Hensley, S., Mouganis-Mark, P.J., 1996. Analysis of
active flows on Kilauea Volcano, Hawaii, using SIR-C radar correlation
measurements. Geology 24(6), 495498.
Zebker, H.A., Stiles, B., Hensley, S., Lorenz, R., Kirk, R.L., Lunine, J., 2009. Size
and shape of Saturns moon Titan. Science 324(5929), 921923.
Zebker, H.A., Villasenor, J., 1992. Decorrelation in interferometric radar echoes. IEEE
Transactions on Geoscience and Remote Sensing 30, 950959.
Zebker, H.A., Werner, C.L., Rosen, P.A., Hensley, S., 1994b. Accuracy of
topographic maps derived from ERS-1 interferometric radar. IEEE Transactions
on Geoscience and Remote Sensing 32, 823836.
Zebker, H.A., Wye, L.C., Janssen, M.A., Cassini Radar Team, 2008. Titans surface
from reconciled Cassini microwave reflectivity and emissivity observations.
Icarus 194(2), 704710.
Biographical Sketch
Scott Hensley received his BS degrees in mathematics and physics from the University of California at Irvine and a
PhD in mathematics from Stony Brook University, where he specialized in the study of differential geometry. In
1992, Dr. Hensley joined the staff of the Jet Propulsion Laboratory where he studied advanced radar techniques
for geophysical applications. He has worked on most of the SAR systems developed at JPL over the past two
decades, including the Magellan and Cassini radars. He was the GeoSAR chief scientist, a simultaneous X-band
and P-band air-borne radar interferometer for mapping above- and beneath-the-vegetation canopies that is now
commercially operated by Earthdata International. He led the SRTM Interferometric Processor Development Team
for a shuttle-based interferometric radar used to map the Earths topography between 7601 latitude. Recently, he
began working with the Earth-based Goldstone Solar System Radar to generate topographic maps of the lunar
surface. He was principal investigator and is currently the chief scientist for the NASA UAVSAR program that uses
an electronically scanned active-array L-band fully polarimetric-designed radar for repeat-pass applications.
79
Tom Farr received BS and MS degrees from Caltech, and a PhD from the University of Washington, all in Geology.
After a short time as an engineering geologist, he joined the Radar Sciences Group at the Jet Propulsion Laboratory, where he has been since 1975. At JPL, he helped develop the first geologic applications of imaging radar
using aircraft, satellites, and the Space Shuttle. He was the deputy project scientist on the Shuttle Radar Topography
Mission, which used interferometric radar to produce a near-global map of the Earth. He has also been a science
investigator on European and Japanese satellite programs, and has assisted in the interpretation of radar images
from Venus and recently from Saturns moon Titan. His scientific research includes the use of remote sensing and
digital topographic data for study of landscapes on Earth and other planets and how they are formed and
modified by climate and tectonic or volcanic activity.
3.4 Remote Sensing of Land Cover Dynamics

TR Allen, Y Wang, and TW Crawford, East Carolina University, Greenville, NC, USA
3.4.1
3.4.2
3.4.2.1
3.4.2.2
3.4.2.3
3.4.2.4
3.4.2.5
3.4.2.6
3.4.2.7
3.4.3
3.4.3.1
3.4.3.2
3.4.3.3
3.4.4
3.4.4.1
3.4.4.2
3.4.4.3
3.4.4.3.1
3.4.4.3.1.1
3.4.4.3.1.2
3.4.4.3.1.3
3.4.5
References
Introduction
Remote Sensing of Land Cover
Discrete Information and Hard Classification
Landscape Metrics and Biophysical Change
Soft and Fuzzy Classification
Change Detection
Spatial and Temporal Scale
Surface Dynamics and Land Cover
Impervious Surfaces and Runoff
Case Studies
Coastal Impervious Runoff
Shoreline Change Analysis using a Multitemporal Radar
Dynamic Feature Extraction: Riverine Flood Inundation
Land-Cover Change Modeling
Emergence and Milestones
Proximate and Ultimate Drivers of Land Change
Modeling Approaches
Spatial modeling approaches
Multicriteria evaluation
Statistical regression
Dynamic spatial simulation
Future Research Directions
80
81
81
82
82
82
83
83
84
84
84
88
88
91
91
95
96
96
97
97
98
98
99
Abstract
Our understanding of land-cover dynamics and geomorphology has advanced with the development of remote-sensing
science and technology. Remote sensing provides data that drive biophysical models of Earth-surface processes, as well as
mapping discrete land cover and land-use change information and landform maps. Land cover remote sensing has also
spurred land-change science, an emergent paradigm linking anthropogenic and natural-system change dynamics using
predictive models. This chapter reviews the state of remote sensing applied to land-cover dynamics, provides selected case
studies, and surveys future challenges and advances in light of increasing anthropogenic changes on the landscape.
3.4.1
Introduction
The significance of land-cover change to the Earths surface,

climate, and life-sustaining systems cannot be understated as a
concern confronting global civilization. Past and ongoing anthropogenic and natural-system changes have wrought grand
challenges to ecological integrity and global biogeochemical
cycles. Fundamental biotic and abiotic processes linking the
Earths surface to climate change are being studied for their
potential to contribute feedbacks to greenhouse warming,
ocean acidification, and climatic variability and extremes. Many
natural environments have been converted from natural vegetation cover into cropland or altered by construction of built
Allen, T.R., Wang, Y., Crawford, T.W., 2013. Remote sensing of land cover
dynamics. In: Shroder, J. (Editor in Chief), Bishop, M.P. (Ed.), Treatise on
Geomorphology. Academic Press, San Diego, CA, vol. 3, Remote Sensing
80
environments including buildings, transportation, and other

impervious surfaces. The resulting microclimates, soilatmospheric interactions, and hydrologic changes now manifest
various geomorphic state and rate changes including erosion,
transportation, and deposition. Less visible, but significant
changes in groundwater hydrology and water quality are exhibited in water-resource developments (dams and reservoirs,
aqueducts, and altered hydro-regimes), extensive soil drainage,
and even residential septic systems and urban storm-water
management. In sum, anthropogenic influences are profoundly
evident on the land cover of the Earth. Geomorphology and
Earth-surface processes, including biotic, soil, and hydrologic
processes, must be factored into understanding of the interactions of human beings and the natural environment. Scientific analyses of these factors have increasingly sought to apply
remote sensing and land-cover change to understand anthropogenic impacts and effects.
Applying systems approaches to investigation of landcover dynamics requires information on surface processes,
http://dx.doi.org/10.1016/B978-0-12-374739-6.00046-4
Remote Sensing of Land Cover Dynamics
landforms, and land use/land cover. The purpose of this

chapter is to review the relevance, limitations, and opportunities of remote sensing of land-cover dynamics for geomorphology, including recent methodological developments
and challenges. Using examples drawn from studies of landcover change and dynamic surface processes, we illustrate
research that captures the state of the art and highlights the
near-term challenges and potential, as new remote-sensing
technologies emerge and our cumulative understanding of
land-cover dynamics increases.
3.4.2
Remote Sensing of Land Cover
Remote-sensing science and technology is very effective for the

production of spatially continuous biophysical and thematic
information that is valuable for geomorphology. Remote
sensing provides for primary data acquisition of continuous
phenomena (e.g., albedo, elevation, surface temperature) as
well as derivative discrete information (e.g., land cover, vegetation, or landforms) using space-borne or aerial platforms.
Despite the wide variety of observations and derivative information, thematic land-cover products vary in detail depending
on four key sensor characteristics including the spatial extent
(image swath and length dimensions) and spatial resolution
(pixel size or field-of-view distance), radiometric resolution
(recording range and sampling of reflectance values), spectral
resolution (number and width of reflectance wavebands in
the electromagnetic spectrum), and temporal resolution (frequency of observation). Land-cover information is primarily
derived from satellite-based, passive sensors that measure reflected solar illumination from the visible to infrared regions
of the spectrum (e.g., Landsat, Ikonos, Quickbird, WorldView
2, moderate resolution imaging spectrometer (MODIS),
ASTER, and SPOT). Land-cover maps are now produced using
image classification methods (i.e., converting spectral pixel
data into discrete information classes), with robust statistics
available for estimating classification accuracy and error.
Nonetheless, there are trends in the production of selected,
fine-scale land-cover products, such as impervious surfaces,
building and infrastructure, shorelines and coastal landforms,
and applications utilizing ultrahigh spatial resolution satellite
or airborne sensors. The latter products, however, remain
costly, less common, and in the domain of isolated projects or
relatively high-value areas or features (e.g., valuable shoreline
erosion monitoring or high-intensity urban development).
Sensors with a spatial resolution of tens of meters, multispectral imaging (46 spectral bands, primarily in the visible
and near-infrared regions), temporal resolution of 516 days
between nadir-viewing, and radiometric resolution of 810
bits produce imagery that remain the most prevalent data
sources for the production of land-cover information. These
data products offer an attractive combination of regional extent, low cost, and relatively frequent observations. Coupled
with the availability of robust, inexpensive, and proven image
analysis and classification software, satellite-derived landcover information can be routinely generated for geomorphological studies. Typical remotely sensed products for
analysis of land-cover dynamics are discrete land-cover maps
(usually having 420 thematic classes, but up to B30) or sets
81
of derived multispectral indices (linearly transformed spectral

data) representing various biophysical characteristics of the
land surface that are useful in monitoring efforts.
Land cover is now well understood to affect many environmental processes. The linkages between land cover and
environmental processes are superbly exhibited in coastal
research linking watersheds and receiving waters. Coastal eutrophication and aquatic habitat degradation, for instance, are
commonly attributed to nonpoint source nutrient pollution.
Satellite remote sensing of land cover provides for routine
characterization for coastal land cover and continues to inform coastal resource management (Thomas, 1995). The National Oceanic and Atmospheric Administration (NOAA)
Coastal Change Analysis Program (CCAP), for example, uses
Landsat Thematic Mapper (TM) and Enhanced Thematic
Mapper (ETM ) satellite data to map coastal land-cover
change (Dobson and Bright, 1991; Jensen et al., 1993a).
Klemas et al. (1993) describe how the classification scheme of
CCAP distinguishes ecologically important features such as
different coastal wetland types. The CCAP program, however,
does not provide a historical perspective based on the use of
the Landsat multispectral scanner (MSS) image archive or
presatellite historical trends. Archived Landsat MSS data have
been analyzed in conjunction with new, higher resolution
sensors (cf., Lee and Marsh, 1995; Jensen et al., 1993b, 1995).
Although the MSS data do not provide highly detailed classifications, the North American Landscape Characterization
(NALC), an archival project of the U.S. Geological Survey
(Lunetta et al., 1998), provides MSS data that are cloud free
and of high radiometric quality. In conjunction with other
Landsat MSS archives, Landsat TM data and robust historical
analyses of land-cover change have been performed. In a study
of urbanization (197282) in Montreal, Charbonneau et al.
(1993) found that automated image classification of a Landsat
MSS time series was 5% to 30% more accurate than traditional
photo-interpretation. Where historical dynamics are rapid and
ground verification is weak, such as dynamic coastal environments, satellite time series are ever more important. Shao
et al. (1998), for example, demonstrated integration of geospatial techniques for assessing shrub thicket distributions in
naturally dynamic barrier islands of the Virginia coast. Clearly,
the integration of geospatial information technology (remote
sensing, GIS, and spatial analysis) and landscape ecology is
still maturing. To more fully understand landscape change,
empirical studies and simulation methods offer a first step to
benchmarking the utility and accuracy of using spatial-analysis
tools for information extraction. Given the geomorphic surface-process dynamism, or degree of permanence of a landform, remote sensing offers numerous image-classification
options that can be matched to the target characteristics.
3.4.2.1
Discrete Information and Hard Classification
Thematic land-cover maps are a staple remote-sensing product

relevant to geomorphology. A handful of national and global
land-cover classification programs are representative of a wide
array of products that typically include a thematic land-cover
classification derived from the hierarchical scheme of Anderson
et al. (1976) and provide a standardized system for a wide
range of uses. This seminal system has subsequently evolved
82
into the Multi-Resolution Landscape Characterization (MRLC)

Consortium formed by the U.S. Geological Survey, Environmental Protection Agency (EPA), NOAA, and U.S. Forest Service
(USFS). The Global Land Cover Characteristics Database
(GLCC) (Loveland et al., 2000) provides a 1 km spatial resolution dataset derived from Advanced Very High Resolution
Radiometer (AVHRR) satellite data over April 1992 to March
1993, with seasonal land-cover classes compiled regionally.
A parallel, national land-cover grid for the U.S. was similarly
compiled (Loveland et al., 1991). In the European Union, the
CO-ordination of Information on the Environment (CORINE)
program has also produced regional NOAA AVHRR-derived
land-cover data. Applications of CORINE data have been wideranging for land cover and agriculture, including the combined
use of thematic cover types and continuous Normalized Difference Vegetation Index (NDVI) time series (Genovese et al.,
2001). In the U.S., efforts to characterize state and regional
land-cover trends for the conservation and preservation of
species and community biodiversity prompted a wide primarily
state-level land-cover program for GAP Analysis Projects in the
1990s (Scott et al., 1996). Other regional land-cover mapping
projects have been implemented by numerous groups. The
Nature Conservancy (TNC) developed a highly floristic vegetation classification system (Grossman et al., 1998). The Albemarle-Pamlico estuarine study also developed a land-use/landcover map from Landsat data as a prelude to development of a
National Estuary Program site for the U.S. Environmental
Protection Agency and State of North Carolina (AlbemarlePamlico Estuary Study (APES), 1994). In the intervening years
since land-cover data have been created, limitations have also
been revealed with so-called hard classification systems, including the necessity of dealing with mixture classes and inherently soft information classes such as transitional
vegetation or ephemeral land use and cover types.
3.4.2.2
Landscape Metrics and Biophysical Change
Thematic land-cover data are highly useful for the landscape

study of ecological and geomorphological relationships (e.g.,
ecogeomorphology or biogeomorphology). Spatial pattern
metrics of the distribution of land cover or habitats have been
increasingly applied for landscape ecological or geobotanical
studies. Numerous indices of landscape spatial pattern have
been the subject of research on land-cover dynamics in landscape ecology (e.g., McGarigal and Marks, 1993; Frohn, 1997;
Gustafson, 1998). These indices primarily measure compositional and spatial patterns of thematic land cover or ecological classes. Predictor variables drawn from terrain (e.g.,
digital elevation models) and biophysical remote-sensing indices such as the NDVI are commonly applied to predict a
particular environmental resource or diagnostic indicator of
environmental change, such as alpine treeline. A wide diversity
and redundancy among the metrics, however, complicates the
task of deriving meaning from quantitative measures of landscape patterns. Further, several metrics are scale-dependent;
hence, the application of semi-variograms to assess scale dependencies of vegetation pattern may be warranted, as illustrated by Bian and Walsh (1993). Yet, scale dependency has
only seldom been compared with measures of spatial pattern
and scale using thematic land cover and landscape pattern

metrics (cf., Gustafson, 1998; Uemaa et al., 2005).
On the coast, remote sensing may identify trends and
patterns of urbanization, agricultural intensification, wetland
loss, or estuarine sedimentation useful to environmental
managers. Among possible change-analysis tools, Markov
models have been used to assess temporal dynamics in ecosystems (Hobbs, 1994) and simulate historical land-cover
changes (Turner, 1988.) They may also be directly derived
from remotely sensed change maps using GIS (Allen and
Walsh, 1993; Li and Reynolds, 1997; Savitzky et al., 1998).
Forest fragmentation, for example, can be characterized by
increasing urbanization or agricultural expansion eroding
forest boundaries and increasing patch isolation. Forest
patch and urbanization patterns have been evaluated in the
northeastern U.S. using landscape-pattern analysis along a
spatial-urbanization gradient (Medley et al., 1995). In sum,
spatial-pattern metrics of remotely sensed land cover have
become widely used for inferring and predicting surface
process influences on habitats and land-cover change.
3.4.2.3
Soft and Fuzzy Classification
Although spatial metrics most commonly exploit pattern analysis of discrete land-cover classification, many instances
occur in geomorphology and landscape ecology where the
phenomena are continuously distributed or exhibit an inherent uncertainty as to thematic classification. Alternatives
have emerged to hard classification schemes with the advent
of techniques such as fuzzy logic, which allow a measure of
pixel similarity (or class membership probability) to be
ascribed to every class (Mather, 1999). For a landscape
undergoing some continuous change, process, or disturbance,
the assignable information class may not be appropriate for a
hard classification, and no objects can be identified or
boundaries delineated. Solutions such as partial membership
or fuzzy classification may be more appropriate in these
cases, where fuzziness may be ascribed to limitations of the
resolution of landscape observation or to mixtures of various
spectral end-members in a pixel. In such situations, pixels can
be decomposed using methods of spectral unmixing. Some
landscapes may lend themselves to discrete hard classification,
whereas other landscapes (or landscapes observed at different
scales) require a soft approach using linear spectral unmixing,
spectral-angle mapping, or reference vectors, which describe
the boundaries between classes in spectral-feature space
(Foody and Mathur, 2004).
Geomorphic examples of fuzzy classification techniques
are most prevalent for inherently continuous or gradational
features and processes. Fuzzy classification has been applied,
for instance, to soil and landform classification (Brabyn, 1997;
Irvin et al., 1997), soil patterns (Odeh et al., 1992), and
geomorphic terrain position (Dragut and Blaschke, 2006;
Arrell et al., 2007; Schmidt and Hewitt, 2004).
3.4.2.4
Change Detection
Oftentimes, the focus of geomorphologic research is the dynamic change or trend rather than the analysis of a static
distribution or relationships. In such research, remote-sensing

change detection offers an array of potential specialized
techniques. In addition, change detection may apply discrete
or continuous approaches to surface dynamics to incorporate
temporal complexity. Change detection can be conducted for
thematic or soft classification, land-cover data as well as more
specialized, hybrid methods. Technical advances in changedetection algorithms and sensor and platform design are receiving major attention with the advent of global environmental change, environmental monitoring and assessment,
and operational needs for natural and technological disasters.
The variety of tools for change detection range from relatively
visual and qualitative interpretation techniques such as
straight-forward, multitemporal composites (e.g., write-function memory displays) and multitemporal principal components analysis (PCA), raster GIS and map algebraic,
postclassification change detection, to abstract spatiotemporal
change vectors for both high spatial and temporal resolution
imagery. Intermediate techniques include band ratioing and
composite indices (e.g., time-series NDVI). The subject of
change detection in remote-sensing research is a major topic.
Among several developments in the change-detection research
arena, phenology, the periodic and seasonal life cycle of
plants and animals, imparts temporal variation between repeat image acquisitions, particularly on an interannual basis.
Although this poses a challenge to the tracking and study of
trends in hard classifications (Moody and Strahler, 1994),
seasonal phenology such as leaf-out, green-up, and senescence
also provides for potential improvement of forest-cover classification (Dymond et al., 2002) or seasonal vegetation variability (Reed et al., 1994). Such temporal spectral changes are
also the foundation of widely used indices, such as the Tasseled Cap for Landsat-derived brightness, greenness, and
wetness (Landsat TM/ETM ) (Crist and Ciccone, 1984).
Sensor temporal frequency of coverage has also been evaluated for its potential impact on change detection. Lunetta et al.
(2004) concluded that a minimum of 34 years of temporal
data-acquisition frequency is necessary to monitor land-cover
change in the northern Piedmont of North Carolina. Another
technique originating in the application of geometric statistics
using multispectral imagery, change-vector analysis (CVA), has
been applied to continuously changing phenomena, ranging
from forest dieback and regeneration (Allen and Kupfer, 2000)
to urbanization (Chen et al., 2003). Emerging platforms with
more flexible orbital-path controls and off-nadir viewing angles also provide for greater temporal frequency of coverage
for monitoring change.
3.4.2.5
Spatial and Temporal Scale
The components of land-cover change necessary for the determination of land-use dynamics are complex. The first dynamic, typical of traditional change detection by map overlay
tabulation, measures aspatial compositional change by quantifying areal amounts and relative proportions of cover types,
without regard to spatial configuration or pattern. Charbonneau et al. (1993) reported primarily on the areal growth of
urban land cover in Montreal (65% increase in 197282).
Second, land-cover changes exhibit explicit spatial patterns
83
such as the number, density, and complexity of patches,

corridors, or mosaic patterns. These are measurable characteristics that may be obtained using a variety of GIS and
pattern statistics software packages (cf., Turner et al., 1991;
Baker and Cai, 1992; McGarigal and Marks, 1993). Because
transition probabilities, such as Markov models, are based on
temporal changes and do not include spatial information
(Barringer and Robinson, 1981), spatial-scale dependence
must be addressed separately. Thus, the third dynamic pertains
to the scale of variation, or scale dependence, of landscape
elements such as grain-size effects on habitat, ecologic transport and flows, and spatial interactions (Forman, 1997). Analysis of semivariograms could indicate the presence of spatial
trends in grain size and test for an ergodic trend, wherein
spatial patterns across a geographic region (e.g., urbanization
front) provide a surrogate or a substitution for temporal
evolution of landscape structure. The assumption or exploration of ergodicity is a fundamental concept of geomorphology, tracing its lineage as early as William Morris Davis Cycle
of Erosion. As a more recent example, space-for-time substitution has been used to conceptually model the transgression
of coastal wetlands in response to climate change and sea-level
rise (Michener et al., 1997). In spacetime dimensions, a
landscape may evolve dynamically through various stages of
spatial, compositional, and scale-dependent structure. Measurement and modeling of the scale dependence may use semivariogram analysis of either thematic land-cover data or continuous biophysical information such as the NDVI (Frank,
1988). For example, toward the rural southern terminus of the
Delmarva Peninsula, one finds a persisting agricultural landscape, whereas northward, landscape structure reflects increasing urbanization and forest fragmentation that should be
evident in NDVI semi-variograms.
3.4.2.6
Surface Dynamics and Land Cover
Numerous dynamic geomorphological processes and features

are investigated using remote sensing, particularly erosion,
deposition, flooding, and natural hazards. Coastal geomorphology and natural hazards, for instance, offer relatively
abundant examples of remote sensing and geomorphology.
Landsat imagery is commonly used for the extraction of landcover data for modeling landslide susceptibility (e.g., Choi
et al., 2010; Pande et al., 2009) and mapping groundwater
recharge zones for aquifers or karst terrain (Nagarajan and
Singh, 2009). Dynamic fluvial landforms and water levels
are also resolvable using airborne light detection and ranging
(LiDAR) and analysis of signal intensity (Hofle et al., 2009).
LiDAR data are classified into floodplain, channel, and
flood-risk zones specific to emergency management, such as
Federal Emergency Management Agency (FEMA) flood-hazard
risk maps. Flood dynamics and vegetation distribution
have been studied using the cloud-penetrating capability of
synthetic aperture radar (SAR) satellite data (e.g., Martinez and
Le Toan, 2007). Regional climates exhibiting prolonged seasonal cloud cover continue to pose challenges to remote
sensing, particularly in tropical and seasonal wetdry and
monsoonal climates. Commonly, a multisensor approach
incorporating aerial as well as satellite data is necessary
84
(e.g., tropical dry deforestation; Galicia et al., 2008). Visual

cues from direct observation of surface conditions are also
feasible for Aeolian geomorphologic modeling (cf., Webb
et al., 2009).
3.4.2.7
Impervious Surfaces and Runoff
Geospatial technologies are accurate and efficient means for

characterizing urban features, impervious surfaces, and providing a variety of useful information for storm-water management and watershed planning (Brown, 1997; Brabec et al.,
2002). The advent of high-resolution satellite imagery and the
increasing environmental problems associated with imperviousness have attracted the attention of remote-sensing researchers for improved characterization and quantification of
impervious surfaces (Weng, 2008). Impervious surfaces affect
energy balance, habitat fragmentation, and especially hydrology (Goetz et al., 2003). Radar imagery has been found to be a
successful tool for characterizing urban features, topography,
and impervious surfaces by several studies (Hepner et al.,
1998) and holds promise for subsurface hydrologic mapping
(Smith, 2002; Becker, 2006). Coastal imperviousness has been
mapped using LiDAR and aerial photography (Hodgson et al.,
2003), as well as IKONOS multispectral data (Goetz et al.,
2003; Sawaya et al., 2003). Urban land-cover and land-use
classification has been historically challenging, but imperviousness mapping in complex urban areas is a priority (Ridd,
1995; Bauer et al., 2002). The literature points to the need for
advancing various techniques using fine-resolution grid-based
models.
A wide variety of spatial rainfall-runoff models have
evolved with the advent of widely available digital data such
as land cover, digital elevation models (DEMs), soils, and
gridded rainfall products from radar or interpolated meteorological-station precipitation. A well-known suite of
TOPographically based hydrologic MODels has been derived
from TOPMODEL (Beven, 1997). The Terrain Analysis Using
Digital Elevation Models (TAUDEM; Tarboton, 1997) has also
been widely applied and improved, such as characterizing
drying in catchments or transport-limited accumulation.
Another example includes the Depth to water, Recharge
rate, Aquifer characteristics, Soil media, Topography, Impact
of vadose zone and hydraulic Conductivity (DRASTIC)
model for predicting vulnerability of groundwater pollution.
DRASTIC has been ported to GIS to build an index value
for the potential of groundwater pollution from surface contamination. The acronym derives from seven input layers
representing DRASTIC. Similar to the AGNPS model (Agricultural Nonpoint Source Pollution) approach, this approach
incorporates topography, vegetation, and soil characteristics
to evaluate field- to watershed-scale runoff characteristics
(Choi and Blood, 1999). The model is demonstrative of a
software tool that has been widely accepted and applied for
aquifer vulnerability. Several other models are currently promoted by the NOAA Coastal Services Center (NOAA
CSC, 2010) and germane to any coastal toolset, such as
the Impervious Surface Analysis Tool (ISAT). Furthermore,
the Nonpoint Source Pollution and Erosion Comparison Tool
(N-SPECT) has also been developed at NOAA CSC for
predicting nonpoint source pollution utilizing integrated

DEMs, hydrology, land cover, and soil-runoff parameters.
Such models are approaching the local scale, but still remain
coarse and more appropriate to regional planning that integrates multiple thematic-surface features and addresses both
groundwater and surface-water interactions. These techniques
have evolved during the prior era of predominantly mediumresolution land-cover data (e.g., Landsat). The N-SPECT and
ISAT software, for instance, principally use the NOAA CCAP
land-use/land-cover data. The application of newly available
LiDAR and ultrahigh resolution satellite data provides challenges and opportunities for the further advancement of these
techniques.
3.4.3
3.4.3.1
Case Studies
Coastal Impervious Runoff
Storm-water runoff arising from impervious surfaces significantly degrades environmental quality, alters stream hydrology and habitats, poses hazards to human health, often
reduces esthetics, and commits a community to capital investment and maintenance. Waters receiving storm-water
runoff accumulate numerous pollutants, including nitrogen,
phosphorus, pesticides and herbicides, oil, lead, grease,
copper, mercury, and pathogens (bacteria and viruses.) Nonpoint-source pollution of surface waters is also affected by
urbanization and is expected to increase unless runoff can be
decreased by diverse methods (Carpenter et al., 1998). Concomitant to expanded estuarine and water-quality monitoring,
North Carolina is among several states now widely utilizing
high-resolution LiDAR laser altimetry that has been collected
in partnership with the FEMA for floodplain mapping. With
spatial resolution finer than 5 m, LiDAR-derived bare-surface
elevation grids provide for improved broad area DEMs for
floodplain delineation and environmental modeling. Integrating these data into a processing workflow provides for
drainage basin delineation and calculation of useful derivatives, including slope, flow direction, and flow accumulation
(e.g., ESRI Hydrologic Analyst tool and TauDEM extensions
for ArcGIS; Tarboton, 2004). These products can be further
augmented with land-cover data and associated runoff and
water-quality models. GIS-based spatial models have demonstrated their ability to provide for accurate runoff, groundwater infiltration, and vulnerability analysis (e.g., GIS-based
DRASTIC groundwater model; Babiker et al., 2005), and such
spatial data are accepted for storm-water permit applications
in model ordinances. In addition to floodplain delineation
(e.g., 100-year recurrence interval flood zones), these data are
used for other hydrologic modeling applications to urban
runoff, water quality and conservation planning, and stormsurge inundation.
A specific coastal example is drawn from a study seeking
to determine the causes of isolated residential street flooding
and to prescribe solutions for storm-water management
and coastal-water protection in the Town of Southern Shores,
located on a barrier island of the northern Outer Banks,
North Carolina. Despite deep, sandy soils with high permeability and infiltration capacity, extensive urbanization resulting
in impervious roofs, driveways, and streets lead to a potential

exceedance of soil infiltration by moderate rainfall. Subsurface
peat deposits, primarily in interdune swales, were also posited
to cause local ponding and perched water tables. Flood-prone
areas (characteristically with connective impervious surfaces)
were mapped by residents and town personnel. The distribution of contributing subwatersheds and underlying peat
were largely unknown, except for field topographic and geomorphic interpretation. Resistivity mapping and reconnaissance investigation with soil-moisture meters, penetrometers,
and augering qualitatively confirmed the interpreted distribution of peat in swales, but extensive mapping was required
for subwatershed delineation, impervious cover mapping, and
analysis of flow accumulation.
The study design provided for a series of remote sensing,
hydrologic GIS modeling, and impervious surface analysis
85
within developed subwatershed units. First, available LiDAR

DEMs from the NC Floodplain Mapping Program (NCFMP)
were acquired and merged. These bare-earth DEMs (50 ft
resolution with hydro-correction) were analyzed using TauDEM, including verifying the absence of sinks and discovering
disconnected topographic lows. Flow accumulation was calculated using the D8 flow-direction algorithm. Figure 1 shows
the distribution of flow accumulation (cell counts) and derived subwatersheds. Flood-prone areas and pixel centroids
(points) are also illustrated, highlighting the strong coincidence of flood-prone areas and topographic flow accumulation for several sites. To characterize impervious cover extent
and runoff contribution within subwatersheds, the NOAA
ISAT tool was applied using 2001 CCAP land cover for the
subwatershed using North Carolina imperviousness-landcover coefficients. Figure 2 shows the CCAP land cover and
Flooding point
Flood-prone areas
Drainage basins
Flow accumulation
High: 12071
Low: 180
0
100 200
400 m
Figure 1 Raster flow-accumulation grid derived from bare-earth LiDAR DEMs with superimposed subwatershed delineation and chronic floodprone sites in a coastal barrier-island community, Town of Southern Shores, Dare County, Outer Banks, North Carolina.
86
Problems
Percent impervious
Problems
0.02.2
Percent ISAT
(a)
(b)
2.33.7
< 10%
3.87.6
1025%
7.716.2
> 25%
16.330.2
(c)
Figure 2 Comparison of (a) Coastal Change Analysis Program (CCAP) land cover and (b) derived ISAT impervious classes and (c) coastal
imperviousness using high-resolution LiDAR DEMs and derived subwatersheds with very high-resolution IKONOS and municipal jurisdiction
impervious data.
two cartographic classifications of the developed imperviousness percentages. Figure 2(b) depicts the default manual
classification intervals, with relatively modest impervious
cover over the small basins, and only six basins with the
highest class of imperviousness (425%); yet, none of these
exhibit runoff-induced flooding problem sites. Out of 31
flood-prone sites, ten occur within the subwatersheds with
moderate impervious cover (1025%), whereas the remainder
are situated in basins with relatively low impervious cover
(o10%). A quantile classification of basin imperviousness
shown in Figure 2(c), however, indicates that two-thirds of
the flood-prone sites are found within basins with impervious
cover 43.8%. Despite having a relatively low impervious
cover, some watersheds are prone to urban flooding, and yet,
higher spatial-resolution data are needed to infer localized
effects arising from imperviousness or the overriding effects of
shallow peat deposits.
In order to identify finer-scale impervious cover linked
to flooding, IKONOS satellite imagery and town cadastral
data were acquired. GeoEyes IKONOSTM 1-m panchromatic
and 4-m multispectral imagery were georeferenced and
merged for a composite 1-m resolution pan-sharpened image,
and analyzed using TRIMBLE eCognitionTM object-based
image-classification software. The context-based algorithm
Figure 3 Coastal impervious surface cover extracted from a hybrid

classification of IKONOS imagery including vegetation index
thresholding and ISODATA unsupervised classification.
provided a high-accuracy delineation of impervious cover,

illustrated in a sample of the town extent in Figure 3. A
notable improvement over prior analyses using only perpixel maximum-likelihood classification, the distribution and
pattern of impervious classes clearly avoids commission errors

over extensive bare-sand areas, despite the sometimes strong
spectral similarity to impervious areas. By mapping the floodprone areas and point-of-maximum flow accumulation within
these zones (Figure 4), it is also possible to further infer
the effective imperviousness of the upslope accumulation
areas (shaded salmon color). Figure 4 reveals that the
flood-prone sites do not always arise from extensive subwatersheds. Further, when a random sample of sites in basins
NOT prone to flooding is drawn and compared, the tally of
total impervious area of the smaller, upslope contributing area
is able to predict the location of flood-prone sites with 67%
accuracy using discriminant analysis. As the town of Southern
Shores grapples with flood hazards and their mitigation, they
are considering the implementation of storm-water management rules and new collection fees to raise capital for costly
87
engineering solutions. Figure 5, however, illustrates the crux

of the dilemma and a possible geomorphically informed
alternative. Figure 5 shows a chronic flooding site, a LiDARderived subwatershed, and a high-resolution digital orthophoto. The small subwatershed consists of two home sites
affected by flooding and vehicular inaccessibility in a culde-sac; yet, the subwatershed shows three additional neighboring parcels that likely contribute runoff. The fine-scale
subwatershed reveals the high imperviousness of this sites
upslope area, despite the relatively moderate parcel impervious cover. Future research in geospatial applications to sustainable design and planning offers the possibility of more
efficient applications, including spatially targeted storm-water
best management practices (BMPs), rainwater collection cisterns, pervious pavers, and the use of remote sensing and GIS
to target subwatersheds for retrofit solutions.
Flood-prone and sample sites,

contributing areas, and sub-basins
Sample
Flood-prone
Sub-basin
Elevation (m)
NAVD 1988
High : 20.0
Low : 0.0
Figure 4 Flood-prone and stratified-random sample site locations and flow-accumulation areas.
88
Figure 5 Flood-prone swale site with highly connected, multiple

parcel, high-connectivity impervious surface area and low-infiltration
soils with subsurface peat. The flood-prone cul-de-sac is delineated
with hachures, flow accumulation low point, and DEM-derived
subwatershed transparently superimposed on a 1:100 color digital
orthophoto.
3.4.3.2
Shoreline Change Analysis using a Multitemporal

Radar
Shorelines and coastal ecosystems exhibit dynamics that are of

high interest to society and geomorphologic research. Remote
sensing of coastal areas provides synoptic observations and
temporal monitoring capabilities. Cloud cover, however,
hampers optical and multispectral coastal remote sensing in
several ways, such as by limiting the observation of shore
features at optimal repeating times (e.g., low tides), acquiring
images during urgent situations such as immediately pre- and
postdisaster (e.g., hurricanes), or repeating observations on
annual calendar dates for environmental monitoring. Radar
sensors, however, utilize microwave radiation emission and
backscatter to actively image the landscape and have allweather imaging capabilities that overcome cloud-cover constraints. For example, Wang and Allen (2008) developed an
edge-extraction filter model of shoreline delineation and
quantified shoreline changes using multidate SAR imagery
over a 13-year period. On the Pamlico Peninsula of North
Carolina, shoreline changes vary from statistically none, to a
maximum landward migration rate of B11 m year1 between
December 1994 and December 2006. Subsequently, Wang
et al. (2009) reported almost no shoreline changes between
December 2006 and June 2008 for their study sites using
Advanced Land Observation Satellite/Phrased Array Type LBand SAR (ALOS/PALSAR) data. One of the revealing insights
is that multisensor imagery can be used effectively to assess
short- and long-term changes. In the case of relatively short
time-periods to capture relative erosion, deposition, or
shoreline stability, SAR has the advantage of being unaffected
by cloudiness, providing for a rich source of observations
compared with patchy or wholly cloud-obscured optical
imagery.
The SAR data in the Wang et al. (2009) study were multilook (ML), geo-referenced, amplitude-only imagery. The
pixel size was 12.5 12.5 m for 1994 JERS1 SAR data,
6.25 6.25 m for the fine-beam single-polarization mode of
PALSAR data, or 12.5 12.5 m for the fine-beam double-polarization mode of PALSAR data. The multilook process reduces speckle of SAR imagery and simplifies the use of the
data, but unfortunately, the multilook fusion procedure degrades the spatial resolution and removes the phase information, compared with single-look complex (SLC) data, a
more standard and widely available SAR data processing
product. For instance, the ML and SLC PALSAR are defined as
level 1.5 and level 1.1, respectively. The PALSAR data at level
1.5 can be used to delineate large changes, such as dozens of
pixels. Fine-scale changes, however, within 71 pixels, cannot
be discerned statistically with confidence. Finally, three SLC
data sets with near anniversary dates (24 September 2007, 26
September 2008, and 14 August 2009) were available (the
anniversary dates minimize the impact of seasonal variation of
individual land-use and land-cover (LULC) types on the coherence of data pairs. The coherence is one of the parameters
used in geo-referencing the pairs.). Therefore, one can reliably
use the SLC datasets to delineate changes in the study area
between a relatively short time span of 2007 and 2009.
After reanalyzing the ML and SLC SAR data between 2007
and 2009, we conclude that in most locations, ML and SLC data
reveal no major shoreline changes. In a small number of isolated places, however, the ML data show no changes, whereas
the SLC data show erosion or accretion. For example, Figure 6
shows Stumpy Point Bay, North Carolina, where no shoreline
changes are detected from 2007 (Figure 6(a)), 2008
(Figure 6(b)), to 2009 (Figure 6(c)) using ML datasets. The
distance measurements, shown along solid reference lines on
Figure 6(a)6(c), are 4448.6, 4448.5, and 4446.9 m, respectively. The absolute difference is B2 m, which is about 4.5% of
the entire distance measured. Because each pixel size is
12.5 12.5 m, the difference cannot be determined or there
was no detectable shoreline change. However, using the SLC
data pair of 2007 (Figure 7(a)) and 2009 (Figure 7(b)), a slight
landward movement occurred near the tip of the peninsula of
the Stumpy Point Bay. The estimated erosion rate here is
13 m year1 between 2007 and 2009. One cannot observe the
change at all using the ML data (Figure 7(c) and 7(d)); however, because the total change of the two-year period is on the
order of or less than half of the pixel size of the ML data, the
delineation of small changes can be attributed to the fine spatial
resolution of the SLC data as compared with the ML data. In
addition, due to the use of phase information in geo-referencing the SLC pair, the RMS error can be as low as 1/10 of a
pixel size. There is no phase information in level 1.5 data. Thus,
the RMS error can be easily up to or more than 71 pixel.
3.4.3.3
Dynamic Feature Extraction: Riverine Flood

Inundation
Delineation of shoreline changes and floodwater/dry land

boundaries present a challenge for remote sensing that requires ultrafine-scale feature extraction. The case studies that
follow focus on shoreline changes of the North Carolina coast
(a)
(b)
(c)
Figure 6 Multilook and multitemporal amplitude data of ALOS/

PALSAR show the Stumpy Point Bay, North Carolina. Acquisition
dates are (a) 24 September 2007, (b) 26 September 2008, and (c) 14
August 2009. The images are geo-rectified. The cross provides georeferencing information, and it is the starting point for the measured
line segment.
using multitemporal SAR data and the delineation of water/

nonwater boundaries at different river stages near Greenville,
NC, using a DEM inundation-overlay model.
Impacts of global climate change and sea-level rise on
coastal zones will accelerate, and coastal vulnerability will
concomitantly increase (Intergovernmental Panel on Climate
Change (IPCC), 2007). Rapid population growth in coastal
communities increases demands on coastal resources and
natural hazards pose additional threats to the communities.
89
As one moves landward, people and communities on

coastal floodplains are also impacted by the change in water
levels in streams. Owing to a gently sloping coastal plain, a
large increase of surface water in streams could produce
a massive flood event. The ability to delineate the water/
nonwater boundary in a timely manner and accurately
during a flood event could mitigate dire consequences.
Thus, several efforts have been made toward resolving this
problem, including the creation of nationwide flood maps
(i.e., flood insurance rate maps (FIRMs) and recently digital
FIRMs for some states) by FEMA. The flood maps are widely
used in flood-hazard mitigation because the floodway, floodway fringe, boundaries of a 100-year floodplain, and a 500year floodplain are delineated. Given the variability of precipitation and climate change, however, the 100-year or
the 500-year floodplain can vary considerably. In addition,
the flood maps are static products that are the result of the
integration of hydraulic and hydrological modeling, parameterization of local topography and LULC types, and stream
flow/discharge. Changes between LULC types and imperviousness on the floodplain can alter the local topography,
drainage, and runoff flow pattern and speed. Such temporal
variations cause variability in flood mapping results. Finally, it
is important to note that flood maps are legal documents.
Therefore, their compilation and revision are a very long and
costly process. More importantly, during the course of a flood
event, the flood maps may not provide timely and accurate
information on the location of floodwater/dry-land boundaries. Other alternatives to determine the boundary locations
must be sought.
Mapping the extent of a flood using remotely sensed
data, DEMs, river-gauge data, and supporting ground observations has received increased attention because of data
availability and the potential to assess flood recurrence intervals, delineate floodplains, and respond to emergencies
(e.g., Colby et al., 2000; Wang et al., 2002; Wang, 2004;
BourgeauChavez et al., 2005; Zheng and Wang, 2007; Lang
et al., 2008). Wang et al. (2002) summarized the advantages
and limitations of using the datasets individually or in various
combinations. The temporal variability of LULC types is
relatively high on and near floodplains because of urban expansion and floodplain development. Updated spatial information about individual LULC types, communities, and
infrastructure is critical for assessing future flooding. Multitemporal and high-resolution imagery is best suited for providing this information. Furthermore, multiple governmental
agencies are obtaining and releasing topographic data and
multitemporal DEMs. For example, the United States Geological Survey (USGS) is currently carrying out a nationwide
LiDAR-DEM initiative to create the next generation of DEMs
for the nation (http://lidar.cr.usgs.gov/). Other topographic
data can be obtained from the NASA Shuttle Radar Topography Mission (SRTM; http://www2.jpl.nasa.gov/srtm/) and
other optical-image sensors such as the Advanced Spaceborne
Thermal Emission and Reflection Radiometer (ASTER; http://
www.ersdac.or.jp/GDEM/ or https://lpdaac.usgs.gov/). Finally,
the surface-water heights of streams are continuously monitored in the U.S. at tens of thousands of river-gauge stations
run by the USGS and the EPA. These measurements and other
data are available and downloadable through the Internet.
90

Range direction
Flight direction
Water
Land
(a)
(b)
800
Meters
0
800
(c)
(d)
Figure 7 PALSAR data for the peninsula near the Stumpy Point Bay, NC. Images in (a) and (b) are slant range, single-look complex data. Their
acquisition dates were 24 September 2007 and 14 August 2009. The cross indicates the landward migration of about one to two pixels from
2007 to 2009. The shoreline change is not noticeable on the geo-rectified multilook data from 2007 (c) and 2009 (d) for the same area.
These field data can be used in combination with a simple

and heuristic DEM inundation-overlay model (Zheng and
Wang, 2007) that maps flood extent. Such simple models can
efficiently and accurately simulate real-time flooding, and the
spatial information considerably facilitates the use of existing
floodplain maps. The approach includes: (1) the delineation
of a stream centerline using a DEM and high-resolution
imagery; (2) the interpolation of a rivers surface-water height
along the centerline using the measured surface-water heights
at the upstream and downstream gauge stations; (3) delineation of the regular stream surface based on the regular
flow or surface-water heights along the channel; and (4)
delineation of the spatial extent using the elevated and
measured surface-water height to simulate a flood event.
Advantages include accounting for topographic changes using
multitemporal DEMs and overlaying flood extent over multitemporal remotely sensed data.
For example, Figure 8 shows DEMs arranged from the
finest to coarsest cell resolution, and depicts a small area near
Greenville, NC. By varying the scale of the DEM cell resolution, the sensitivity of the inundation model to DEM error
can be assessed. The DEMs are displayed in grayscale, with
dark tones representing lower elevations and lighter tones
representing higher elevations. In general, lower altitudes
are located in the north and higher altitudes in the south.
The Tar River is also identified (Figure 8(a) and 8(b)). From
Figure 8(a)8(f), one can note the pixilation or loss of detail

in portraying local topography. It is important to note the
amount of lost detail in the SRTM DEM. Thus, one should be
cautious in using the DEM-inundation-overlay model and
attentive to scale-dependent representation and error propagation when modeling processes such as inundation.
Spatial and temporal changes in inundation arising from
Land-Use and Land-Cover Change (LULCC) are also significant to riverine and coastal-flood hazards. Figure 9(a) depicts
an overlay of the simulated surface-water coverage of Tar River
at regular flow over the 1998 USGS false-color IR digital
orthophoto quartile quads (DOQQ). The same surface-water
coverage is then overlaid on the 2006 USDA natural-color
orthophotos. Significant development occurred during the
19982006 period, which altered the LULC types in the
floodplain and watersheds. This information on landscape
change could be critical to forecasting and emergency response to a future flood event, including terms of runoff
modeling and flood-impact assessment. To illustrate its importance, the surface-water heights at flood stage on the Tar
River and the 1999 flood after Hurricane Floyd (Wang et al.,
2002) were simulated and overlaid on the 2006 image (Figure 10). The water is near the door-steps of the houses when
the river is at its flood stage (Figure 10(a)). Consequently,
most of the area is flooded, and a newly developed area is
completely submerged when the water rises to the level of the
(a)
(b)
(c)
(d)
(e)
(f)
91
Figure 8 Various DEMs near Greenville, North Carolina, arranged from fine to coarse spatial resolutions illustrate scale and resolution
sensitivity. (a) USGS DEM of 3 3 m, (b) NC floodplain mapping program DEM of 6 6 m, (c) USGS DEM of 10 10 m, (d) NC floodplain
mapping program DEM of 15 15 m, (e) USGS NED of 30 30 m, and (f) SRTM DEM of 1 1 arcsecond.
1999 flood (Figure 10(b)). The ground photo (Figure 11)

taken after the 1999 flood looking toward the north illustrates
the disastrous result. In the photo, the flood water line was
clearly visible, and its elevation value is much higher than that
of the ground where the development was underway (completed as shown in a 2006 image).
3.4.4
3.4.4.1
Land-Cover Change Modeling

Emergence and Milestones
Land-cover mapping and analysis enabled by geospatial technologies is a major component of integrative land-change science. This field has emerged as a cornerstone of global
environmental change and attempts to understand the human
and biophysical factors influencing land-cover dynamics,
defined here as temporal variations in land cover and land use
(Rindfuss et al., 2004). Land-change science, as it may be
termed, is of interest for understanding spatial and temporal
dynamics and promotes investigations of humanenvironment
linkages and systems that require integrative contributions

from social, natural, and geographical information sciences.
Goals include the identification and understanding of forcing
factors, spatial and temporal dynamics, and the consequences
of anthropogenic and natural change. Expertise in geospatial
techniques and information theory is a prerequisite for
integration of social and environmental data or the linking
of people-to-pixels at relevant spatial and temporal scales
(Liverman et al., 1998; Rindfuss et al., 2004). Furthermore,
investigation of complex anthropogenic influences on land
cover that affect surface processes and geomorphology requires a multidisciplinary approach, careful research design,
and expertise spanning social science, physical geography, and
geospatial techniques.
Major milestones in the development of land cover-related
science include the LULCC program that was initiated in 1994
as a core project of the International GeosphereBiosphere
Programme to investigate the questions: How do human and
biophysical forces affect land cover, and what are the environmental and social impacts of land-cover dynamics? Affiliated
LULCC projects have applied remote-sensing approaches to a
92
Tar r
i
ver a
t reg
ular
flow
stag
(a)
Tar riv
er at
regula
r flow
stage
(b)
Figure 9 Overlays of modeled surface-water coverage at the regular flow of Tar River on the 1998 USGS false IR DOQQ image (a) and the
2006 USDA natural color image (b). Water appears black. Development has occurred on the south side of the river.
Water from Tar River at

flood stage
(a)
Water from Tar River at surface

height after Hurricane Floyd in 1999
100
meters
0
100
(b)
Figure 10 Modeled surface-water coverage for 2006 natural color image (a) Tar River, NC, during flood stage, and (b) flood flow during 1999
Hurricane Floyd. Water is shown in black.
93
94
Figure 11 Ground photo of the Tar River floodplain after the recession of the 1999 Hurricane Floyd flood. The flood-water line is clearly visible,
with its elevation much higher than that of the ground.
variety of internationally ranging case studies having global

implications, with notable emphases on tropical deforestation
(Galicia et al., 2008), agricultural extensificationintensification, desertification, and grassland dynamics. A comprehensive synthesis of LULCC results since its inception is reported
in Lambin and Geist (2006). Related in scope and focus to
LULCC, in 1996, the U.S. National Aeronautics and Space
Administration (NASA) launched its LCLUC program, which
addresses science questions similar to LULCC, but has an
additional focus on remote sensing applications using NASA
or other earth-observation data products. NASA LCLUC
also regularly sponsors funded regional and thematic-based
research. A synthesis of LCLUC results since its inception is
reported in Gutman et al. (2004).
Recognizing the significance of land-cover change, the
National Research Council (2001) identified land-use dynamics as one of eight grand challenges in its prescriptive
publication Grand Challenges in Environmental Sciences. Although stated as a stand-alone problem domain, it is worth
noting that other identified challenges have clear connections
and feedbacks with land use and cover change. These include
biogeochemical cycles, biodiversity and ecosystem functioning, climate variability, hydrologic forecasting, infectious
disease and the environment, and institutions and resource
use. The impetus and relevance to geomorphology and
Earth-surface processes in these research thrusts are broad. It
has been recognized, for instance, that changes in land cover
affect biodiversity (e.g., tropical rain-forest deforestation;
Michalski et al., 2010), and yet biodiversity owes much of its
community and landscape heterogeneity to underlying topographic and physiographic diversity. Hence, geomorphologists
and ecologists recognize that interactions of biotic and geomorphic change from sources to sinks, including environments ranging from alpine tundra (French, 1986), montane
forests (Allen and Kupfer, 2000), to coasts (Efe and Tagil,
2008). Processes stemming from land-cover change, such as
erosion, are also known to cause riverine and estuarine sediment pollution (Kusimi, 2008). Land cover derived from the
classification of remotely sensed data (digital satellite data or
aerial photography), in particular, governs many parameters
used in modeling environmental processes, including the
Universal Soil Loss Equation (USLE) and soil-runoff curve
numbers found in many models of soil erosion and runoff
potential (Renard et al., 1997). Increasingly fine spatial-resolution sensors have promoted the study of smaller catchments
(Jurgens and Fander, 1993; Schonbrodt et al., 2010, Seeber
et al., 2010).
Table 1 Strategic goals from Geography for a Changing World,

USGS Circular 1281
1. Characterize and quantify land-surface status and trends to provide a
framework for understanding change patterns and processes from
local to global scales.
2. Identify local, regional, national, and global drivers of land change to
forecast plausible land-change scenarios over the next 2050 years.
3. Understand past, present, and future environmental consequences of
land change to support better management of their effect on people,
environment, economy, and resources.
4. Improve the scientific basis for vulnerability and risk assessment,
mitigation, response, and recovery related to the human and
environmental dynamics of land change.
5. Develop credible and accessible geographic research, tools, and
methods to support decision-making related to the human and
environmental consequences of land change.
6. Develop and test hypotheses about the use of geographic regions to
understand the human and environmental dynamics of land change.
7. Observe the Earth at all scales using remote sensing to understand
the human and environmental dynamics of land change.
8. Provide timely, intelligent access to new and archived USGS geographic
data needed to conduct science and support policy decisions.
9. Develop innovative methods of modeling and information synthesis,
fusion, and visualization to improve our ability to explore geographic
data and create new knowledge.
U.S. Geological Survey (USGS) (2005) published a science

strategy for geographic research titled Geography for a Changing
World. Although geography has connotations beyond the sole
topic of land-cover change, this strategic plan parallels related
programmatic developments of land-change science described
above. The science strategy articulated nine goals each with
associated operational objectives (Table 1). A related USGS
initiative was the establishment of the Land Cover Institute in
2006 that functions to address land-cover topics from local to
global scales, and serves as a facilitator for land-cover and
land-use science, applications, and production functions
(U.S. Geological Survey (USGS), 2010).
Initiatives of the U.S. National Science Foundation (NSF)
have contributed to the development of land-change science
that investigates geomorphology science questions. Beginning
in 1999, NSFs Biocomplexity Program has funded interdisciplinary research dealing with complex problems including land use and cover dynamics. The year 2000 program
announcement stated that in order to understand land-cover
change processes, we need to better understand the impact of
the constructed environment, human activities, and social
structures on ecosystem functions. The Biocomplexity program ceased following the 2006 award cycle, but elements of
the program that support land-change science projects have
continued to be present with the 2007 establishment of NSFs
Dynamics of Coupled Natural and Human Systems Program.
NSFs Geomorphology and Land Use Dynamics Program has
funded research since 2003 that examines natural and human
processes shaping geomorphic landscapes over a variety of
space and time scales.
3.4.4.2
Proximate and Ultimate Drivers of Land Change
To date, no unifying formal theory or theories of land use and

cover change exist. Instead, a growing body of regionally
varying case studies has investigated causative factors and associated consequences of change. Although specific contexts
necessarily vary, a commonality in much of the land-change
research is a basic framework that specifies the roles of both
Infrastructure expansion
Demographic factors
Fertility and mortality
Roads, sidewalks, paths
In and out-migration
Public service
Population density
Water lines, electrical grids

sanitation, govt. buildings
Population distribution
Residential expansion
Economic factors
Apertments, condominiums
Golf course communities
Retirement communities
Institutional housing
Life cycle features
Market growth
Economic structure
Urbanization
Special variables
Price increases, comparative
cost advantages
Recreation expansion
Hotels/motels
Policy/institutional factors
Seasonal vacation houses
Formal policies
Camp grounds, RV parks
Planning and land use regulations
Golf course, restaurants, shops
Policy climate
Parks
Non-profit organizations
Commercial expansion
Cultural factors
Office and industrial
Attitudes, values, beliefs
Wholesale and retail
Environment protection
Underlying causes
Proximate causes
proximate and underlying drivers (Turner et al., 1993; Geist

and Lambin, 2001; Xu and Wilkes, 2004). As an example,
Figure 12 illustrates forcing factors relevant to urban and
suburban development in the contemporary U.S. Such changes in urbanizing watersheds may promote the increase in
soil erosion and runoff in small watersheds, contributing to
flashy hydrology, incision in urban watersheds and downstream sediment pollution and exacerbated flood inundation.
Land-cover change impacts on storm hydrographs are
well documented in hydrologic models under current and
projected climate change (Praskievicz and Chang, 2009).
Proximate causes are behaviors that directly alter land uses
and associated land covers. For example, the construction of
new roads, housing subdivisions, and shopping centers
are proximate causes in that they convert natural cover types
(or prior human altered uses) into newly anthropogenized
covers. The pattern of such change is also of interest, as landcover fragmentation from increasing urbanization increases
Transport
Single family homes
95
Pro-growth, anti-growth
Individual and household behavior
Place utility, sense of place
Other factors
Predisposing environmental factors (soil type, topography, etc.)
Biophysical drivers (triggers, floods, hurricanes, sea-level rise)
Social trigger events (war, economic shocks, abrupt policy shifts)
Figure 12 Hypothesized causes of land-use change impacting land cover associated with urban/suburban development in the United States.
96
stream flashiness (McMahon et al., 2003). Such tangible alterations are detectable on the ground and via remote observation using standard data products such as Landsat, MODIS,
and other sensors. They generally involve complex interactions
and typically are more difficult to observe. A case in point may
be the transitional process of agricultural land subdivision for
residential housing. Agricultural land cover can be mistaken
for new or imminent suburbs. For instance, bare agricultural
land recently ploughed may be spectrally similar to newly
graded residential lots. Emergent crops with high greenness
may mimic newly sodded or seeded residential lawns. The
process of land subdivision or parcelization can thus
be difficult to discern in early phases as it is happening.
Nonetheless, mediating the link between underlying and
proximate causes are other factors that can be viewed as influences that may be site-specific and contextual in nature and
may vary across temporal and spatial scales.
Key challenges of land-change science include: (1) observation, measurement, and detection of changing use and cover;
(2) establishment of empirical links between underlying and
proximate causes and their contributions to changing use and
cover; and (3) development of conceptual theories to explain
land-cover dynamics. The first challenge clearly features
the role of remote sensing and related geospatial science
approaches. The subsequent challenges involve contributions
from allied social-science perspectives as well as scientific
modeling techniques. Importantly, Figure 12 indicates a potentially complex array of causative drivers and negates a
more simplistic monocausal explanation. A singular focus
on population growth exemplifies a simplistic monocausal
explanation. Although population growth can surely be a significant driver, Lambin et al. (2001) identified it as one
of several other myths, or what might be considered biases
regarding issues and problems in understanding land use and
cover change (Table 2; see also Heilig, 1994). Major challenges
still remain for the linkage of human interactions and the Earth
system, particularly in the acquisition of socio-economic data
and its linkage to satellite imagery and the forecasting of future
human activities and policies (Liverman and Cuesta, 2008).
3.4.4.3
Modeling Approaches
Modeling of land-cover changes involves manipulating digital

geospatial representations of human and biophysical patterns
and processes. Description and qualitative narrative
methodologies are sometimes used to first build a conceptual

model before quantitative predictive models are constructed.
Before modeling, a remotely sensed time-series data set is
typically required to quantify trends of change describing the
types, magnitudes, and pattern of change. Additionally, spatial-pattern metrics can be used to describe the change trajectories of spatial configurations and patch dynamics
(McGarigal and Marks, 1993). Human-dimensions data collected via detailed social surveys and key informant interviews
are required to describe and understand land-use decisions of
relevant decision makers (i.e., households, firms, land managers, government agencies) and to link these agents to their
respective land parcels (Rindfuss et al., 2004).
A diversity of modeling approaches exists that can be
characterized along different dimensions (Lambin and Geist,
2006). One distinction is between spatial and nonspatial
models. Spatial, or spatially explicit models, represent change
at some level of spatial detail. Spatially explicit models
simulate change in disaggregated, relatively fine spatial units
such as individual pixels within a digital geospatial context.
Clearly, finer spatial resolution yields more detailed model
output, but at an added computational cost. Models may also
be either dynamic or static.
3.4.4.3.1
Spatial modeling approaches
Temporally static modeling approaches typically are in the

form of regression models and are useful to acquire knowledge about the driving factors of land change. They require at
least two observation dates (although more can also be used),
with which to observe change as the dependent variable to be
modeled. They are static in that no dynamic and iterative
updating is implemented, either spatially or temporally, to
reflect change occurring within the modeled time window.
Conversely, dynamic models are constructed to incrementally
represent change at specific locations and at incremental time
steps throughout the modeled time window. A geomorphologic example of a static spatial model using linear regression
is the application of land cover, soil type, and soil-runoff curve
numbers to predict soil erosion. Applications of this approach
may also be spatially explicit (e.g., soil erosion, rainfall infiltration, or runoff per pixel) or alternatively, aggregated spatial
units such as catchments, hillslopes, or land-use zones.
Descriptive models aim to represent the actual functioning
of land use and cover change in a manner that best matches
empirically observed data. Prescriptive models, in contrast,
Table 2 Myths (or biases) of land use and cover change research
Myth (or bias)
Reality
Population growth is the major cause of change.

Change is comprehensive land conversion.
Tropical forest is the most pressing problem of land change.
Agricultural expansion in developing world is among the most pressing
problem of land change.
Prior unaltered land is assumed pristine.
Change is permanent.
Change is homogeneous.
Local factors primarily drive change.
Change is due to multicausal factors.

Change includes incremental land modification.
Many other important types of changes exist.
Urbanization and suburbanization in developed and developing world is
important.
Land has been altered for millennia.
Change can be reversible.
Change is heterogeneous and fragmented.
Remote influences/globalization are important.
Source: Adapted from Heilig, G.K., 1994. Neglected dimensions of global land-use change: reflections and data. Population and Development Review 20(4), 831859, and Lambin,
E.F., Turner, II B.L., Geist, H.J., et al., 2001. The causes of land-use and land-cover change: moving beyond the myths. Global Environmental Change 11, 261269.
aim to represent optimized land use and change patterns according to a defined set of goals and objectives. A further
distinction can be made between deductive and inductive
models. This distinction shares similarities to the descriptive/
prescriptive distinction. Deductive modeling postulates theory
a priori governing change to predict pattern from process,
albeit articulated in the theory. Theory derived from Von
Thunen, for example, has been used to model agricultural
expansion, deforestation, and wetland conversion. Inductive
modeling is more common within the land-change science
community, where model specifications are commonly based
on statistical correlations between land change and a suite of
explanatory variables. Other inductive approaches include
rule-based models informed by empirical observations and
neural-network techniques. Although most studies rely heavily
on inductive approaches to develop quantitative models,
they typically rely on theory and understanding of land-use
decision processes to guide the set of factors to include in
model representation. A rule-based descriptive geomorphologic model example could be sea-level rise resulting in inundation and shifting floodplains (Brown, 2006). In this
sense, land-change modeling commonly blends inductive and
deductive approaches drawing from a range of physical and
social-scientific disciplines, such as anthropology, ecology,
economics, geography, political science, and geography for
both theoretical and methodological guidance.
3.4.4.3.1.1 Multicriteria evaluation
Multicriteria evaluation (MCE) modeling is a prescriptive form
of suitability-site modeling that is used in numerous GIS applications (Malczewski, 2006). Geomorphologic examples of
MCE modeling include management of groundwater-recharge
zones (Murthy and Mamo, 2009), erosion-risk assessment
for railway corridors (Barbieri et al., 2002), site-suitability
modeling for sanitary landfills (Mariolakos et al., 1997), and
landslide-hazard assessment (Barredo et al., 2000). As applied
to land-change science, MCE incorporates multiple GIS data
layers hypothesized to represent relevant factors (criteria) influencing human behavior and future land-use decisions. For
example, factors might include current zoning, future land-use
plan, terrain slope angles, important natural resource areas,
hazard status, road accessibility, and proximity to existing
development. Central to MCE is the assignment of quantitative scores reflecting suitability for a particular land use in the
future. Typically, a numeric scheme of 1 (least suitable) to 9
(most suitable) is applied to each spatial unit in each GIS layer
(vector polygons or raster grid cells) with a weighting scheme
for each input layer. The idea is that criteria (i.e., GIS layers) do
not share the same importance regarding suitable future land
use. A weighted-sum for each spatial land unit is obtained via
GIS overlay that quantifies the MCE output, which is a
quantitative surface with high scores indicating high suitability for conversion to a particular land use (i.e., development) and vice versa. Multiple outputs can be created by using
varying rankings and weighting schemes to assess output
stability.
The use of subjective assessment regarding both the suitability scoring and weighting schemes is a common criticism
of the MCE technique. One advantage is that it includes
local stakeholder input. Including stakeholder input, however,
97
comes with the cost of additional local engagement with local

planners and stakeholders. Another criticism of the MCE approach is that it quantifies a gradient of suitable future land
use, but it does not actually produce future land-use maps for
defined time periods. Methods requiring projections of land
consumption or population growth must be implemented in
concert to project amounts of land to allocate for future
conversion. Rule sets must also be implemented to resolve
situations where competing uses have similarly high suitability. These decisions partly depend on the use of the model
and the resolution of data available during the modeling
process. The LUCIS Model (Land Use Conflict Identification
Strategy) developed by Carr and Zwick (2007) is an example
of a recent and significant MCE model that draws from a rich
tradition of similar modeling techniques.
3.4.4.3.1.2 Statistical regression
Statistical-regression techniques offer another approach to
modeling land-cover dynamics. Spatial-regression applications of land cover in geomorphology are wide ranging,
including mapping applications such as bathymetry (predicting depth from reflectance among different classified bottom types), predicting landslide activity as a function of
topographic variables and land cover (Biswajeet et al. (2011)),
relationships between woody debris, fluvial channels, and
land cover (Wing and Skaugset, 2002), or stream sediment
loads (Phippen and Wohl, 2003). Logistic regression and
multinominal regression have been considerably used for such
purposes within the research community of land-change science (Lesschen et al., 2005). Logistic regression is described
below, although similarities extend to multinominal regression (see Lesschen et al., 2005).
Geomorphologists seeking to predict the occurrence of
patterns or processes, but having categorical or dichotomous,
spatial-variable predictors may use logistic regression models.
Pradhan and Lee (2010), for example, incorporated multinominal land-cover classes into a logistic-regression model to
delineate landslide-hazard zones on Penang Island, Malaysia.
The dependent variable in logistic regression, y, is the presence
(1) or absence (0) of land conversion at the individual pixel
level. This observed change is for a defined historical period
that includes initial and terminal dates. Observed change is
measured via change-detection techniques using multitemporal satellite data. A vector, X, serves as a set of spatially
explicit predictive variables and p is the response probability,
where p Pr(Y 19X). A random sample of pixels can be selected for initial modeling. A second random sample can be
selected to validate the resulting model(s) using standard
techniques such as error matrices, kappa coefficients, and the
receiver operator characteristic (ROC) method. The resulting
logistic-regression equation can be implemented within a GIS
to calculate and map a probability surface (p) of development
ranging from 0 (low) to 1.0 (high).
The resulting probability surfaces are used to project future
land change. A key input for models of anthropogenic land
change is observed population growth and land consumption
for the observed historical data that provide information on the
amount of land consumed (i.e., converted to human use) per
capita. Using the regression-derived probability surfaces, pixels
are ranked according to the highest probability of change and
98
selected in descending rank order to be converted up to a volume consistent with the estimated conversion amounts.
Some limitations and criticisms of this regression-based
approach include the following: (1) regression parameter
estimates from historical data are assumed to remain stationary across space and through time; (2) empirically derived
models based on historical data exhibit poor accuracy
performance as assessed by quantitative accuracy assessment
techniques, and accuracies may vary geographically; and
(3) the temporal scale of the original regression models
(e.g., a 20-year time step) may not match the temporal scales
of the projected output (e.g., projecting out to 40 years).
3.4.4.3.1.3 Dynamic spatial simulation
Cellular automata (CA) and agent-based modeling are rapidly
emerging dynamic spatial-simulation modeling techniques.
CA models typically operate using a gridded raster landscape,
with individual grid cells having discrete land-use/cover categories (i.e., residential, commercial, agriculture, natural, etc.).
A set of locally based transformation rules are applied in
discrete time steps (i.e., annual updates) to model land-use
change. For example, for an individual undeveloped cell, if a
defined number of neighboring cells are in a developed state,
then the cell is converted from undeveloped to developed at
the next annual iteration. Topographic and landform features
can be used as raster inputs to CA models affecting land-use
changes (e.g., elevation, slope, aspect, and curvature; discrete
landforms such as floodplains, valleys, and ridges; or higher
order topographic derivatives such as view-sheds and subwatersheds). A variety of rules may also be incorporated into
such models, such as diffusive growth emanating from existing
roads or the developed fringe. Stochastic (random) growth
such as leapfrog development can be simulated by using
appropriate model parameters. The SLEUTH model (Clarke
and Gaydos, 1998) developed by Keith Clarke at UC Santa
Barbara is perhaps the most well known of such CA modeling
approaches that has been widely implemented.
In addition to the use of terrain or landforms as predictors
of human land-use change, spatial simulation models are increasingly being applied for investigating environmental processes and terrain evolution. For instance, Saco et al. (2006)
developed an enhanced form of the SIBERIA model that
incorporates landform evolution to predict the spatial organization and depth of soils. Three-dimensional landscapepedogenesis models have also emerged such as matrices
ARMOUR 3D (mARM3d) to simulate soil evolution as a
function of soil profile for points on a spatial grid. Markov-like
transition matrices provide for the examination of depthdependent and independent soil-forming factors (Cohen,
2010). Vegetation-geomorphic patterns have also been simulated using spatial models, including montane and subalpine
forests (Brown, 1994; Horsch, 2003), and patterns of Southern Appalachian forest regeneration following invasive
pathogen introduction (Allen and Kupfer, 2000). Spatial dependence has also developed as a valuable consideration in
the prediction of pattern and process. Some spatial-simulation
models incorporate scale dependency explicitly, such as using
semivariograms and geostatistics (e.g., Brown et al., 2002).
Agent-based models are another form of dynamic spatial
simulation and are thoroughly reviewed in Parker et al. (2002).
Similar to CA, agent-based models typically operate on a

gridded, raster landscape. They differ by simulating a large
number of virtual agents (i.e., homebuyers, developers, businesses) that are imbued with decision-making behavior
that alters land use in discrete, typically annual, time steps.
Decision-making behavior is represented by a rule set
hypothesized to represent real-world processes changing the
landscape. Organisms have become a subject of simulation
using agent-based models, because behaviors can be modeled
in coupled external routines (i.e., model code) that incorporate economic strategies such as utility maximization or
bounded rationality. Empirical survey data are frequently
collected from actual agents (i.e., a survey of landowners) to
reflect empirical realities. Interactions between prehistoric
human populations and basic processes of food acquisition
(hunting, gathering, and agriculture), for example, have been
simulated and linked to landform evolution (Wainwright,
2008). Developments in anthropogenic factors modifying
landscapes have further been demonstrated for human impacts on wildfire (the Special Protection Area SIMulator
model, SPASIMv1) and simulation of changes in erosion and
deposition over century scales (CYBEREROSION model;
Wainright and Millington, 2010).
Both approaches to dynamic spatial simulation provide a
virtual laboratory to help replicate and explain observed
change, and explore and probe alternative futures and scenarios. Difficulties with these models include complexities of
both model calibration (i.e., setting appropriate model parameters) and model validation (i.e., checking results against
reference data). Other challenges include obtaining the
technical expertise required for software coding (i.e., Java,
Repast, C ).
3.4.5
Future Research Directions
A treatment of remote sensing of land-cover dynamics has

been presented that accounts for new developments and
challenges in land-change science and geospatial technologies.
Selected examples of new sensors and methodologies have
also been presented. Concluding remarks can thus be summarized as a situation of coevolution of remote-sensing science with scientific paradigms and geographic information
technology.
The rapid advancement of remote sensing is widely known,
with sensor technology and information extraction being
dominant themes. Historically, the first impetus was on the
increasing continuum of information extraction, classification,
and change detection. Each increase in this continuum represents a higher order of information. Most of the information-extraction techniques also engender products that are
informative to science. For instance, we quantified land-cover
change and inferred the increasing impact of land-use conversion on sediment yield, pollution, or habitat degradation.
Having quantified change and verified the relationships between cause and effect and understanding processes, we can
thus predict and simulate alternatives in order to further
understand Earth-surface systems coupled with human alterations. Land-change science has emerged concomitant to
more widely available LULC data. Although the use of hard
classification still predominates, soft and objected-based

classification techniques are attracting research and application interest. The proliferation of available imagery and
growing time series for new change-detection algorithms has
also prompted the expansion of biophysical analysis, such as
the assessment of phenology in global change phenomena
and operational use of remote sensing data for near-real-time
use in hazards and agriculture. Dynamic geomorphic environments such as coasts and urban environments are attracting greater interest in these techniques. Advances in
sensors and platforms provide for new opportunities, such as
illustrated in coastal imperviousness mapping, LiDAR-derived
digital terrain modeling (including subwatershed delineation
and hydrologic modeling), and their integration for predicting
flooding from urban runoff. Passive space-borne SAR sensors
were also illustrated for time-series analysis of shorelines, including coastal erosion and riverine flooding.
Many challenges and great potential remain for the expanded application of remote sensing to a wide array of
geomorphologic research. New technologies are allowing for
questions that could not be addressed in recent decades to be
handily studied by primarily remote-sensing investigations. In
other cases, new observational data have revealed limitations
that require more research and modeling and technology solutions, such as addressing the effects of differential illumination in topographically complex environments (Colby, 1991;
Allen, 2000; Bishop and Colby, 2002). The use of new data is
another example, as LiDAR is becoming essential in the
hydrologic and coastal geomorphology subdisciplines. Yet,
LiDAR for some environments is relatively straightforward
(e.g., topographic LiDAR for beach mapping) versus highly
vegetated, urban, or aquatic environments. The tropics and
perennially cloudy regions also pose great challenges to routine earth observations from passive sensors, which has
spawned a greater interest in active sensors such as SAR.
Land-change science provides a road map and rubric for
integrating remote-sensing derived products with geomorphology. We might expect that land-change science will further
exploit the ever-growing time series of imagery, perhaps expanding beyond discrete, thematic land-cover conditions and
typologies. Beyond the principal human dimensions of environmental change, such as land use and urban growth
modeling, geomorphologists can join in biocomplexity research programs, further investigating the stressor or receptor
systems in environmental processes. Land-cover change studies
in arctic and alpine environments prompt many studies of
glacio-hydrological and geomorphic significance, for example.
The resulting changes in mountain hydrology, although having
human impacts on water resources, also portend significant
changes to stream power, load, and fluvial processes documented elsewhere in this volume. These applications have initiated the use of time-dependent classes, phases, and sequential
Markovian analyses of change. Second, biophysical remote
sensing is increasingly able to integrate and apply data-fusion
techniques for combining multisensor information, such as
passive and active sensors, airborne and satellite, and spectral
and topographic data. The coastal imperviousness case study,
for example, demonstrated the integration of LiDAR and IKONOS satellite data, but the study also used ground-based resistivity mapping by means of an OhmMapperTM system. This
99
example also suggests potential increased application of remote

sensing at ever finer spatial resolutions, taking advantage of
ultrahigh resolution sensors using objected-based image classification. The coastal SAR image analysis case points to a parallel development occurring in time series and multilook-angle
image analysis for feature extraction.
Feature-extraction techniques are also poised to continue
their rapid advancement. High spatial and temporal resolution datasets of SAR, and the taskable, off-nadir viewing
geometry of some satellites allow improved research on shore
erosion processes and disasters. Thus, one can map and
quantify trends of shoreline changes, discover how shoreline
changes and erosion rates may affect infrastructure, security,
and social and economic vulnerabilities, and disseminate the
geospatial data and shore information to stakeholders in
support of improved coastal policy and sustainability.
As these examples and others demonstrate, the remotesensing research community will have numerous opportunities to investigate the information content of ever-increasing
sophisticated sensors that produce spatiotemporal data. The
development and evaluation of new information extraction
techniques and models will also be critical and will permit
new scientific applications. Land-cover dynamics will become
ever more important to geomorphology in order to understand landscape evolution and estimate anthropogenic forcing
on erosion, natural hazards, and ecological and hydrological
systems.
References
Allen, T.R., 2000. Topographic normalization of Landsat Thematic Mapper data in
three mountain environments. Geocarto International 15(2), 1319.
Allen, T.R., Kupfer, J.A., 2000. Application of spherical statistics to change vector
analysis of Landsat data: southern Appalachian Spruce-fir forests. Remote
Allen, T.R., Walsh, S.J., 1993. Characterizing multitemporal alpine snowmelt
patterns for ecological inferences. Photogrammetric Engineering & Remote
Sensing 59(10), 15211529.
Anderson, J.R., Hardy, E.E., Roach, J.T., Witmer, R.E., 1976. A Land Use and Land
Cover Classification System for Use With Remote Sensor Data. US Geological
Survey Professional Paper 964, Washington, 41 pp.
APES Albemarle-Pamlico Estuary Study, 1994. Land Use/Land Cover TM
(APES 1987). North Carolina Center for Geographic Information and
Analysis (NC CGIA). http://www.geodata.gov/metadata/ (last accessed 12 May
2010).
Arrell, K.E., Fisher, P.F., Tate, N.J., Bastin, L., 2007. A fuzzy c-means classification
of elevation derivatives to extract the morphometric classification of landforms in
Snowdonia, Wales. Computers & Geosciences 33, 13661381.
Babiker, I.S., Mohamed, M.A., Hiyama, T., Kato, K., 2005. A GIS-based DRASTIC
model for assessing aquifer vulnerability in Kakamigahara Hights, Gifu
Prefecture, central Japan. Science of the Total Environment 345(1-3), 127140.
Baker, W.L., Cai, Y., 1992. The r.le programs for multiscale analysis of landscape
structure using the GRASS geographical information system. Landscape Ecology
7, 291302.
Barbieri, M., Bonachea, J., Corsini, A., Giusti, C., 2002. A GIS based model for
geomorphological impact assessment of a railway track (province of Modena,
Italy). Geografia Fisica e Dinamica Quaternaria 25, 1122.
Barredo, J.I., Benavides, A., Hervas, J., van Westen, C.J., 2000. Comparing
heuristic landslide hazard assessment techniques using GIS in the Tirajana
basin, Gran Canaria Island, Spain. International Journal of Applied Earth
Observation and Geoinformation 2, 923.
Barringer, T.H., Robinson, V.B., 1981. Stochastic models of cover class dynamics.
Proceedings of Fifth International Symposium on Remote Sensing of
Environment, Ann Arbor, MI, pp. 125-144.
100
Bauer, M.E., Doyle, J.K., Heinert, N.J., 2002. Impervious surface mapping using
satellite remote sensing. Proceedings of IEEE Geoscience and Remote Sensing
Symposium, Toronto, Canada, June 2428, cd-rom, 3 pp.
Becker, M.W., 2006. Potential for satellite remote sensing of ground water. Ground
Water 44(2), 306318.
Beven, K., 1997. TOPMODEL: a critique. Hydrologic Processes 11, 10691085.
Bian, L., Walsh, S.J., 1993. Scale dependencies of vegetation and topography in a
mountainous environment of Montana. The Professional Geographer 45(1), 111.
Bishop, M.P., Colby, J.D., 2002. Anisotropic reflectance correction of SPOT-3 HRV
imagery. International Journal of Remote Sensing 23(10), 21252131.
Biswajeet, P., Mansor, S., Pirasteh, S., Buchroithner, M.F., 2011. Landslide hazard
and risk analyses at a landslide prone catchment area using statistical based
geospatial model. International Journal of Remote Sensing 32(14), 40754087.
BourgeauChavez, L.L., Smith, K.B., Brunzell, S.M., Kasischke, E.S., Romanowicz,
E.A., Richardson, C.J., 2005. Remote monitoring of regional inundation patterns
and hydroperiod in the greater Everglades using Synthetic Aperture Radar.
Wetlands 25(1), 176191.
Brabec, E., Schulte, S., Richards, P.L., 2002. Impervious surfaces and water quality:
a review of current literature and its implications for watershed planning. Journal
of Planning Literature 16(4), 499514.
Brabyn, L.K., 1997. Classification of macro landforms using GIS. ITC Journal
1997(1), 2640.
Brown, D., 1997. Using GIS technology in the development and maintenance
of a stormwater utility. GIS/LIS Annual Conference Proceedings of American
Society for Photogrammetry and Remote Sensing. Cincinnati, OH. (http://
stormwaterfinance.urbancenter.iupui.edu/PDFs/GISLISv3.pdf).
Brown, D.G., 1994. Predicting vegetation types at treeline using topography and
biophysical disturbance variables. Journal of Vegetation Science 5, 641656.
Brown, D.G., Goovaerts, P., Burnicki, A., Li, M.-Y., 2002. Stochastic simulation of
land-cover change using geostatistics and generalized additive models.
Photogrammetric Engineering and Remote Sensing 68(10), 10511061.
Brown, I., 2006. Modelling future landscape change on coastal floodplains using a
rule-based GIS. Environmental Modelling & Software 21(10), 14791490.
Carpenter, S.R., Caraco, N.F., Correll, D.L., Howarth, R.W., Sharpley, A.N., Smith,
V.H., 1998. Nonpoint pollution of surface waters with phosphorus and nitrogen.
Ecological Applications 8(3), 559568.
Carr, M., Zwick, P., 2007. Smart Land Use Analysis: The LUCIS Model. ESRI Press,
Redlands, CA.
Charbonneau, L., Morin, D., Royer, A., 1993. Analysis of different methods for
monitoring the urbanization process. Geocarto International 8(1), 1725.
Chen, J., Gong, P., He, C., Pu, R., Shi, P., 2003. Land-use/land-cover change
detection using improved change-vector analysis. Photogrammetric Engineering
& Remote Sensing 69(4), 369379.
Choi, J., Oh, H.J., Won, J.S., Lee, S., 2010. Validation of an artificial neural
network model for landslide susceptibility mapping. Environmental Earth
Sciences 60(3), 473483.
Choi, K.S., Blood, E., 1999. Modeling developed coastal watersheds with the
agricultural non-point source model. Journal of the American Water Resources
Association 35(2), 233244.
Clarke, K., Gaydos, L., 1998. Loose-coupling a cellular automaton model and GIS:
long term urban growth prediction for San Francisco and Washington/Baltimore.
International Journal of Geographical Information Science 12(7), 699714.
Cohen, S., 2010. The mARM3D spatially distributed soil evolution model:
threedimensional model framework and analysis of hillslope and landform
responses. Journal of Geophysical Research 115, 16.
Colby, J.D., 1991. Topographic normalization in rugged terrain. Photogrammetric
Colby, J.D., Mulcahy, K.A., Wang, Y., 2000. Modeling flooding extent from
Hurricane Floyd in the coastal plains of North Carolina. Global Environmental
Change Part B: Environmental Hazards 2(4), 157168.
Crist, E.P., Ciccone, R.C., 1984. A physically-based transformation of Thematic
Mapper data the TM Tasseled Cap. IEEE Transactions in Geoscience and
Remote Sensing GE-22, 256263.
Dobson, J.D., Bright, E.A., 1991. CoastWatch detecting change in coastal
wetlands. Geological Information Systems 1, 3640.
Dragut, L., Blaschke, T., 2006. Automated classification of landform elements using
object-based image analysis. Geomorphology 81, 330344.
Dymond, C.C., Mladenoff, D.J., Radeloff, V.C., 2002. Phenological differences in
tasseled cap indices improve deciduous forest classification. Remote Sensing of
Efe, R., Tagil, S., 2008. The use of multitemporal and multispectral Landsat data to
determine change detection around Tuz Lake on Seyhan Delta. Fresenius
Environmental Bulletin 17(5), 577583.
Foody, G.M., Mathur, A., 2004. Toward intelligent training of supervised image
classifications: directing training data acquisition for SVM classification. Remote
Forman, R.T.T., 1997. Land Mosaics: The Ecology of Landscapes and Regions.
Cambridge University Press, Cambridge, 632 pp.
Frank, T.D., 1988. Mapping dominant vegetation communities in the Colorado Front
Range with Landsat Thematic Mapper and digital terrain data. Photogrammetric
Engineering & Remote Sensing 54(12), 17271734.
French, N.R., 1986. Hierarchical conceptual model of the alpine geosystem. Arctic
and Alpine Research 18, 133146.
Frohn, R.C., 1997. Remote sensing for landscape ecology: new metric indicators for
monitoring. Modeling, and Assessment of Ecosystems. Lewis Publishers, Boca
Raton, FL, 99 pp.
Galicia, L., Zarco-Arista, A.E., Mendoza-Robles, K.I., Palacio-Prieto, J.L., GarciaRomero, A., 2008. Land use/cover, landforms and fragmentation patterns in a
tropical dry forest in the southern region of Mexico. Singapore Journal of
Tropical Geography 29(2), 137154.
Geist, H.J., Lambin, E.F., 2001. What Drives Tropical Deforestation?. LUCC
International Project Office, Louvain-la-Neuve, Belgium, LUCC Report Series No.4.
Genovese, G., Vignolles, C., Negre, T., Passera, P., 2001. A methodology for a
combined use of normalized vegetation index and CORINE land cover data for
crop yield monitoring and forecasting: A case study on Spain. Agronomie 21(1),
91111.
Goetz, S.J., Wright, R.K., Smith, A.J., Zinecker, E., Schaub, E., 2003. IKONOS
imagery for resource management: Tree cover, impervious surfaces, and riparian
buffer analysis in the mid-Atlantic region. Remote Sensing of Environment 88,
195208.
Grossman, D.H., Faber-Langendoen, D., Weakley, A.S., et al., 1998. International
classification of ecological communities: terrestrial vegetation of the United
States. Volume I. The National Vegetation Classification System: development,
status, and applications. The Nature Conservancy, Arlington, VA.
Gustafson, E.J., 1998. Quantifying landscape spatial pattern: what is the state of the
art? Ecosystems 1, 143156.
Gutman, G., Janetos, A.C., Justice, C.O. (Eds.), 2004. Land Change Science:
Observing, Monitoring and Understanding Trajectories of Change on Earths
Surface. Kluwer Academic Publishers, Dordrecht.
Heilig, G.K., 1994. Neglected dimensions of global land-use change: reflections and
data. Population and Development Review 20(4), 831859.
Hepner, G.E., Houshmand, B., Kulikov, I., Bryant, N., 1998. Investigation of the
integration of AVIRIS and IFSAR for urban analysis. Photogrammetric
Hobbs, R.J., 1994. Dynamics of vegetation mosaics: can we predict responses to
global change? Ecoscience 1, 346.
Hodgson, M.E., Jensen, J.R., Tullis, J.A., Riordan, K.D., Archer, C.M., 2003.
Synergistic use of lidar and color aerial photography for mapping urban parcel
imperviousness. Photogrammetric Engineering and Remote Sensing 69(9),
973980.
Hofle, B., Vetter, M., Pfeiffer, N., Mandlburger, G., Stotter, J., 2009. Water surface
mapping from airborne laser scanning using signal intensity and elevation data.
Earth Surface Processes and Landforms 34(12), 16351649.
Horsch, B., 2003. Modelling the spatial distribution of montane and subalpine
forests in the central Alps using digital elevation models. Ecological Modelling
168(3), 267282.
Intergovernmental Panel on Climate Change (IPCC), 2007. Climate Change of 2007:
Impacts, Adaptation and Vulnerability. Working Group II Contribution to the
Intergovernmental Panel on Climate Change Fourth Assessment Report. 23 p.
Irvin, B.J., Ventura, S.J., Slater, B.K., 1997. Fuzzy and isodata classification of
landform elements from digital terrain data in Pleasant Valley, Wisconsin.
Geoderma 77, 137154.
Jensen, J.R., Cowen, D., Althausen, J.D., Narumalani, S., Weatherbee, O., 1993a.
An evaluation of CoastWatch change detection protocol in South Carolina.
Photogrammetric Engineering & Remote Sensing 59(6), 10391046.
Jensen, J.R., Cowen, D.J., Althausen, J.D., Narumalani, S., Weatherbee, O., 1993b.
The detection and prediction of sea level changes on coastal wetlands using
satellite imagery and a geographic information system. Geocarto International
8(4), 8798.
Jensen, J.R., Rutchey, K., Koch, M.S., Narumalani, S., 1995. Inland wetland change
detection in the Everglades Water Conservation Area 2A using a time series of
normalized remotely sensed data. Photogrammetric Engineering & Remote
Sensing 61(2), 199209.
Jurgens, C., Fander, M., 1993. Soil erosion assessment and simulation by means
of SGEOS and ancillary digital data. International Journal of Remote Sensing
14(15), 28472855.
Klemas, V.V., Dobson, J.E., Ferguson, R.L., Haddad, K.D., 1993. A coastal land
cover classification system for the NOAA Coastwatch Change Analysis Project.
Journal of Coastal Research 9(3), 862872.
Kusimi, J.M., 2008. Stream processes and dynamics in the morphology of the
Densu river channel in Ghana. International Archives of the Photogrammetry,
Remote Sensing, and Spatial Information Sciences 37, 11771182.
Lambin, E.F., Turner, II B.L., Geist, H.J., et al., 2001. The causes of land-use and
land-cover change: moving beyond the myths. Global Environmental Change 11,
261269.
Lambin, E.R., Geist, H.J. (Eds.), 2006. Land-Use and Land-Cover Change, Local
Processes and Global Impacts. Springer, Berlin.
Lang, M.W., Pittman, K.W., Kasischke, E.S., Prince, S.D., 2008. Assessment of
C-band synthetic aperture radar data for mapping coastal plain forested wetlands
in the Mid-Atlantic Region. Remote Sensing of Environment 112, 41204135.
Lee, C.T., Marsh, S.E., 1995. The use of archival Landsat MSS and ancillary data in
a GIS environment to map historical change in an urban riparian habitat.
Lesschen, J.P., Verburg, P.H., Staal, S.J., 2005. Statistical Methods for Analysing
the Spatial Dimension of Changes in Land Use and Farming Systems. LUCC
Focus 3 Office, Wageningen University, Wageningen, The Netherlands, LUCC
Report Series No. 7.
Li, H., Reynolds, J.F., 1997. Modeling effects of spatial pattern, drought, and
grazing on rates of rangeland degradation: a combined Markov and cellular
automaton approach. In: Quattrochi, D.A., Goodchild, M.F. (Eds.), Scale in
Remote Sensing and GIS. Lewis Publishers, Boca Raton, FL, pp. 211230.
Liverman, D., Moran, E.F., Rindfuss, R.R., Stern, P.C. (Eds.), 1998. People and
Pixels: Linking Remote Sensing and Social Science. National Academy Press,
Washington, DC.
Liverman, D.M., Cuesta, R., 2008. Human interactions with the Earth system: people
and pixels revisited. Earth Surface Processes and Landforms 33(9), 14581471.
Loveland, T.R., Merchant, J.W., Ohlen, D.O., Brown, J.F., 1991. Development of a
land-cover characteristics database for the conterminous U.S. Photogrammetric
Engineering and Remote Sensing. 57(11), 14531463.
Loveland, T.R., Reed, B.C., Brown, J.F., Ohlen, D.O., Zhu, J, Yang, L., Merchant,
J.W., 2000. Development of a global land cover characteristics database and
IGBP DISCover from 1-km AVHRR data. International Journal of Remote
Sensing. 21(6/7), 13031330.
Lunetta, R.S., Johnson, D.M., Lyon, J.G., Crotwell, J., 2004. Impacts of temporal
frequency on land-cover change detection monitoring. Remote Sensing of
Environment. 89, 444454.
Lunetta, R.S., Lyon, J.G., Guindon, B., Elvidge, C.D., 1998. North American
landscape characterization dataset development and data diffusion.
Malczewski, J., 2006. GIS-based multicriteria decision analysis: a survey of the
literature. International Journal of Geographical Information Science 20(7), 703726.
Mariolakos, I., Fountoulis, I, Logos, E., 1997. The crucial role of the neotectonic
deformation at the landfill site selection. The case study of Avlona (Attiki,
Greece). Proceedings, International Association for Engineering Geology 2,
Balkema, Rotterdam, pp. 20072010.
Martinez, J.M., Le Toan, T., 2007. Mapping of flood dynamics and spatial
distribution of vegetation in the Amazon floodplain using multitemporal SAR
data. Remote Sensing of Environment 108(3), 209223.
Mather, P.M., 1999. Land cover classification revisited. In: Atkinson, P.M., Tate, N.J.
(Eds.), Advances in GIS Analysis. Wiley and Sons, Chichester, pp. 716.
McGarigal, K., Marks, B.J., 1993. FRAGSTATS: Spatial Pattern Analysis Program for
Quantifying Landscape Structure, Forest Sciences Department, Oregon State
University, Corvallis, OR.
McMahon, G., Bales, J.D., Coles, J.F., Giddings, E.M.P., Zappia, H., 2003. Use of
stage data to characterize hydrologic conditions in an urbanizing environment.
Journal of the American Water Resources Association 39(6), 15291546.
Medley, K., McDonnell, M.J., Pickett, S.T.A., 1995. Forest-landscape structure along
an urban-to-rural gradient. Professional Geographer 47(2), 159168.
Michalski, F., Norris, D., Peres, C.A., 2010. No return from biodiversity loss.
Science 329, 1282.
Michener, W.K., Blood, E.R., Bildstein, K.L., Brinson, M.M., 1997. Climate change,
hurricanes and tropical storms, and rising sea level in coastal wetlands.
Ecological Applications. 7(3), 770801.
Moody, A., Strahler, A.H., 1994. Characteristics of composited AVHRR data and
problems in their classification. International Journal of Remote Sensing 15(17),
34733491.
Murthy, K.S.R., Mamo, A.G., 2009. Multi-criteria decision evaluation in groundwater
zones identification in MoyaleTeltele subbasin, South Ethiopia. International
101
Nagarajan, M., Singh, S., 2009. Assessment of groundwater potential zones using
GIS techniques. Journal of the Indian Society of Remote Sensing 37(1), 6977.
National Research Council, 2001. Grand Challenges in Environment Sciences.
National Academies Press, Washington, DC.
NOAA CSC, 2010. National Oceanographic and Atmospheric Administration, Coastal
Services Center. Nonpoint Source Pollution and Erosion Comparison Tool (NSPECT) and Impervious Surface Analysis Tool (ISAT). http://www.csc.noaa.gov/
digitalcoast/ (last accessed 31 May 2010).
Odeh, I.O.A., McBratney, A.B., Chittleborough, D.J., 1992. Soil pattern recognition
with fuzzy-c-means: application to classification and soillandform
interrelationships. Soil Science Society of America Journal 56(2), 505516.
Pande, R.K., Bunnan, D., Singh, R., 2009. Landslide hazard zonation in Hanuman
Chatti area of Uttarakhand, India. Disaster Prevention and Management 18(4),
410417.
Parker, D.C., Berger, T., Manson, S.M. (Eds.), 2002. Agent-Based Models of LandUse and Land-Cover Change. Indiana University, LUCC Focus 1 Office, LUCC
Report Series No. 6.
Phippen, S.J., Wohl, E., 2003. An assessment of land use and other factors
affecting sediment loads in the Rio Puerco watershed, New Mexico.
Geomorphology 52(34), 269287.
Pradhan, B., Lee, S., 2010. Landslide susceptibility assessment and factor effect
analysis: backpropagation artificial neural networks and their comparison with
frequency ratio and bivariate logistic regression modelling. Environmental
Modeling & Software 25(6), 747759.
Praskievicz, S., Chang, H., 2009. A review of hydrologic modeling of basin-scale
climate change and urban development impacts. Progress in Physical Geography
33(5), 650671.
Reed, B.C., Brown, J.F., VanderZee, D., Loveland, T.R., Merchant, J.W., Ohlen, D.O.,
1994. Variability of land cover phenology in the United States. Journal of
Vegetation Science v. 5, 703714.
Renard, K.G., Foster, G.R., Weesies, G.A., McCool, D.K., Yoder, D.C., 1997.
Predicting Soil Erosion by Water: A Guide to Conservation Planning With the
Revised Universal Soil Loss Equation (RUSLE). U.S. Department of Agriculture,
Washington, D.C, Agricultural Handbook 703.
Ridd, M.K., 1995. Exploring a V-I-S (vegetation-impervious surface-soil) model for
urban ecosystem analysis through remote sensing: comparative anatomy for
cities. International Journal of Remote Sensing 16(12), 21652185.
Rindfuss, R.R., Walsh, S.J., Turner, II B.L., Fox, J., Mishra, V., 2004. Developing a
science of land change: challenges and methodological issues. Proceedings of
the National Academy of Sciences of the United States of America 101(39),
1397613981.
Saco, P.M., Willgoose, G.R., Hancock, G.R., 2006. Spatial organization of soil depths
using a landform evolution model. Journal of Geophysical Research 111, 16.
Savitzky, B.G., Ogneva-Himmelberger, Y., Chowdhury, R., 1998. The use of multitemporal image classification in modeling land use and land cover change: a
case study in the southern Yucatn Peninsular region. Proceedings of Annual
Meeting of the American Society for Photogrammetry and Remote Sensing,
Tampa, FL (cd-rom.).
Sawaya, K.E., Olmanson, L.G., Heinert, N.J., Brezonik, P.L., Bauer, M.E., 2003.
Extending satellite remote sensing to local scales: land and water resource
monitoring using high-resolution imagery. Remote Sensing of Environment 88,
144156.
Schmidt, J., Hewitt, A., 2004. Fuzzy land element classification from DTMs based
on geometry and terrain position. Computers and Geosciences 121, 243256.
Schonbrodt, S., Behrens, T., Imbery, S., Scholten, T., 2010. Soil erosion modeling
in terraced landscapes examples from the Three Gorges area, China. 19th
World Congress of Soil Science, Soil Solutions for a Changing World. Brisbane,
Australia 16, 228231,August.
Scott, J.M., Tear, T.H., Davis, F.W. (Eds.), 1996. Gap Analysis: A Landscape
Approach to Biodiversity Planning. American Society for Photogrammetry and
Remote Sensing, ASPRS Distribution Center. Annapolis Junction, Maryland.
Seeber, C., Hartmann, H., Xiang, W., King, L., 2010. Land use change and causes
in the Xiang catchment, Three Gorges Area, derived from multispectral data.
Journal of Earth Science 21(6), 846855.
Shao, G., Young, D.R., Porter, J.H., Hayden, B.P., 1998. An integration of remote
sensing and GIS to examine the responses of shrub thicket distributions to
shoreline changes on Virginia barrier islands. Journal of Coastal Research
14(1), 299307.
American Geographers 92(3), 385398.
Tarboton, D.G., 2004. Terrain Analysis Using Digital Elevation Models (TauDEM.)
http://hydrology.neng.usu.edu/taudem/ (accessed May 31, 2010).
102
Tarboton, D.G., 1997. A new method for the determination of flow directions and
contributing areas in grid digital elevation models. Water Resources Research
33(2), 309319.
Thomas, J.P., 1995. Remote sensing and relating coastal development to living
marine resources and their habitats. Natural Areas Journal 15, 2136.
Turner, M.G., 1988. A spatial simulation model of land use changes in a piedmont
county in Georgia. Applied Mathematics and Computation 27, 3951.
Turner, S.J., ONeill, R.V., Conley, W., Conley, M.R., Humphries, H.C., 1991. Pattern
and scale: statistics for landscape ecology. In: Turner, M.G., Gardner, R.H. (Eds.),
Quantitative Methods in Landscape Ecology. Springer-Verlag, New York, pp. 1749.
Turner, II B.L., Hyden, G., Kates, R.W. (Eds.), 1993. Population Growth and
Agricultural Change in Africa. University of Florida Press, Gainesville.
Uemaa, E., Roosaare, J., Mander, U., 2005. Scale dependence of landscape metrics
and their indicatory value for nutrient and organic matter losses from
catchments. Ecological Indicators 5(4), 350369.
U.S. Geological Survey (USGS), 2005. Geography for a Changing World: A Science
Strategy for the Geographic Research of the U.S. Geological Survey, 20052015,
Circular 1281. U.S. Department of Interior, U.S. Geological Survey, Reston, VA.
U.S. Geological Survey (USGS), 2010. The USGS Land Cover Institute (LCI). Web
access: http://landcover.usgs.gov/ (last accessed May 11, 2010).
Wainright, J., Millington, J.D.A., 2010. Mind, the gap in landscape-evolution
modelling. Earth Surface Processes and Landforms 35(7), 842855.
Wainwright, J., 2008. Can modeling enable us to understand the role of humans in
landscape evolution? Geoforum 39(2), 659674.
Wang, Y., Colby, J.D., Mulcahy, K.A., 2002. An efficient method for mapping flood
extent in a coastal floodplain by integrating Landsat TM and DEM data.
International Journal of Remote Sensing 23(18), 36813696.
Wang, Y., 2004. Seasonal change in the extent of inundation on floodplains
detected by JERS1 synthetic aperture radar data. International Journal of
Wang, Y., Allen, T.R., 2008. Estuarine shoreline change detection using Japanese
ALOS PALSAR HH and JERS1 LHH SAR data in the AlbemarlePamlico
Sounds, North Carolina, USA. International Journal of Remote Sensing 29(15),
44294442.
Wang, Y., Liao, M., Wang, C., 2009. Applications and analyses of satellite-borne
L-band synthetic aperture radar data in coastal environments. Geography
Compass 3(4), 14651482.
Webb, N.P., Phinn, S.R., McGowan, H.A., 2009. Visual assessment of the Australian
Land Erodibility Model. Journal of Arid Environments. 73(67), 678682.
Weng, Q. (Ed.), 2008. Remote Sensing of Impervious Surfaces. CRC Press, Taylor
& Francis, Boca Raton, FL, 454 pp.
Wing, M.G., Skaugset, A., 2002. Relationships of channel characteristics, land
ownership, and land use patterns to large woody debris in western Oregon
streams. Canadian Journal of Fisheries and Aquatic Sciences 59(5), 796807.
Xu, J.C., Wilkes, A., 2004. Biodiversity impact in northwest Yunnan, southwest
China. Biodiversity and Conservation 13, 959983.
Zheng, T., Wang, Y., 2007. Mapping flood extent using a simple DEM-inundation
model. The North Carolina Geographer 15, 119.
Biographical Sketch
Dr. Thomas R. Allen is Associate Professor of Geography at East Carolina University. Allen holds a BS degree in
geography from Old Dominion University (Norfolk, VA, USA) and a PhD from the University of North Carolina
at Chapel Hill. His research focuses on applied environmental applications of remote sensing and geographic
information systems, emphasizing coastal, biogeographic, and geomorphologic applications and environmental
hazards. He is a past Fulbright Scholar (Finland) and currently chairs the Coastal and Marine Specialty Group of
the Association of American Geographers.
Dr. Yong Wang obtained his BS in electric engineering from the Northwestern Polytechnic University, China, and
MA and PhD in geography with a focus on synthetic aperture radar (SAR) and remote sensing from the University
of California, Santa Barbara, CA, USA. Current research interests include the application of remote sensing and
geographic information systems to coastal areas and natural hazards, and development of imaging and application algorithms for air- and space-borne SAR. He is a professor of Geography, East Carolina University, NC,
USA.
Dr. Thomas W. Crawford is Associate Professor of Geography at East Carolina University. Crawford holds a BS
degree in Mathematical Economics from Wake Forest University (Winston-Salem, NC, USA) and a PhD in
Geography from the University of North Carolina at Chapel Hill. His research focuses on GIScience and applied
human-environment applications of geospatial technologies including land-use analysis, coastal development,
environmental hazards, and spatial accessibility. He is currently the president of the North Carolina Geographical
Society and is a member of the Association of American Geographers.
3.5 Near-Surface Geophysics in Geomorphology

S Kruse, University of South Florida, Tampa, FL, USA
3.5.1
3.5.2
3.5.2.1
3.5.2.2
3.5.2.3
3.5.2.3.1
3.5.2.3.2
3.5.2.3.3
3.5.2.3.4
3.5.2.3.5
3.5.2.3.6
3.5.3
3.5.3.1
3.5.3.2
3.5.3.3
3.5.3.3.1
3.5.3.3.2
3.5.3.3.3
3.5.3.3.4
3.5.4
3.5.4.1
3.5.4.2
3.5.4.3
3.5.4.3.1
3.5.4.3.2
3.5.4.3.3
3.5.4.3.4
3.5.4.3.5
3.5.4.3.6
3.5.4.3.7
3.5.5
3.5.5.1
3.5.5.2
3.5.5.3
3.5.5.3.1
3.5.5.3.2
3.5.5.3.3
3.5.5.3.4
3.5.5.3.5
3.5.5.3.6
3.5.5.3.7
3.5.5.3.8
3.5.6
3.5.6.1
3.5.6.2
3.5.6.3
3.5.6.3.1
3.5.6.3.2
3.5.6.3.3
Introduction
Gravity
Strengths
Limitations
Gravity Case Studies
Faulting
Volcanic processes
Topography and weathering
Slope processes
Coastal processes and sea-level change
Glacial and periglacial processes
Magnetics
Strengths
Limitations
Case Studies
Faulting
Volcanic processes
Resistivity and EM Methods
Strengths
Limitations
Case Studies
Faulting
Volcanic processes
Slope processes
Coastal processes
Fluvial processes
Ground-Penetrating Radar
Strengths
Limitations
Case Studies
Faulting
Volcanic processes
Slope processes
Aeolian processes
Fluvial processes
Seismic Methods
Strengths
Limitations
Case Studies
Faulting
Volcanic processes
104
105
106
106
106
106
106
107
109
109
109
109
109
110
110
110
110
110
110
110
112
112
112
112
112
112
112
113
113
113
113
115
115
115
115
115
115
117
119
119
119
119
119
121
121
121
121
122
122
Kruse, S., 2013. Near-surface geophysics in geomorphology. In: Shroder, J.

(Editor in Chief), Bishop, M.P. (Ed.), Treatise on Geomorphology. Academic
Press, San Diego, CA, vol. 3, Remote Sensing and GIScience in
http://dx.doi.org/10.1016/B978-0-12-374739-6.00047-6
103
104
Near-Surface Geophysics in Geomorphology
3.5.6.3.4
3.5.6.3.5
3.5.6.3.6
3.5.6.3.7
3.5.7
3.5.8
3.5.8.1
3.5.8.2
References
Slope processes
Fluvial processes
Combining Geophysical Methods
Discussion and Conclusions
Practical Considerations
Integrating Geophysics, Remote Sensing, and Geographic Information Systems
Abbrevations
CMP
CSAMT
DEM
dGPS
EM
ERS-1
FEM
Geosat
GOCE
common mid-point
controlled source audio-magnetotellurics
digital elevation model
differential global positioning system
electromagnetic
European remote-sensing satellite 1
frequency-domain electromagnetics
geodetic satellite
GRACE
InSAR
LiDAR
MT
Seasat
TEM, TDEM
TLS
122
122
122
122
122
123
123
125
125
gravity field and steady-state ocean

circulation explorer
gravity recovery and climate experiment
interferometric synthetic aperture radar
light detection and ranging
magnetotelluric
sea satellite
timedomain electromagnetics
terrestrial laser scanning
Abstract
Near-surface geophysical methods can provide information on subsurface structure and stratigraphy that is critical to
understanding surficial processes. Gravity, magnetics, resistivity, electromagnetics, ground-penetrating radar, and various
seismic methods are applied across a range of process domains, including faulting, volcanism, topography and weathering,
hillslope processes, coastal and sea-level change, aeolian and fluvial processes, and glacial and periglacial processes. Gravity
and magnetic methods have long been used to image faults and other tectonic and volcanic features, and they have also
been used to document weathering patterns. With the development of multi-electrode resistivity systems, resistivity profiling has become a staple tool to map the structure and water content of slope deposits and volcanic features. Groundpenetrating radar studies have been crucial to understanding of geologic features as varied as the internal structure of
aeolian dunes, coastal and fluvial deposits, the structure and volume of glaciers, the spatial distribution of sinkholes, and
the geometry of tephra and lahar deposits. Key to successful studies are calibrations against direct observations and/or uses
of multiple complementary methods. Increased efficiencies in geophysical data acquisition and positioning in recent years
have made possible very-high-resolution three-dimensional (3-D) or quasi-3-D imaging of subsurface structures. However,
there is still a mismatch in the typical spatial scales of large remote-sensing-based studies and smaller geophysical surveys.
Bridging this disconnect with new geophysical acquisition techniques and new instrumentation such as terrestrial laser
scanning should improve our understanding of the role that subsurface structure plays in the evolution of topography.
3.5.1
Introduction
Near-surface geophysical methods are invaluable to geomorphology investigations where knowledge of subsurface
structure is important for understanding surface processes and
tectonics. The internal structure of landforms generally provides insight into the evolution of paleo-topography and
paleo-landforms. For example, gravity and magnetic surveys
have been used for over a century both to understand mechanisms of isostatic compensation of topography and to estimate fault offsets. Much more recently (in the past 20 years),
ground-penetrating radar imaging of the internal structure of
various types of dunes has significantly advanced our knowledge of dune formation and evolution. In both cases, nearsurface structures could simply not be adequately resolved
from direct observations, and surficial or borehole data.
Geophysical techniques offer the only practical means of
imaging structures between boreholes and outcrops, and extending imaging into the third dimension.
There is some confusion about the meaning of near
surface in geophysics. To many practitioners, near surface
implies material in the uppermost B100 m. This is a useful definition because numerous methods are only effective
at depths less than B100 m, and, as a result, are simply not
used in kilometer-scale or whole-Earth studies. These methods
include high-frequency electromagnetic (EM) methods,
ground-penetrating radar (except on ice), 12-V batterypowered resistivity systems, and hammer-sourced seismic
refraction. However, some methods applicable to studying the
uppermost 100 ms can, with appropriate survey design, capture information on much deeper structure. In particular,
gravity and magnetic measurements are sensitive to heterogeneities from a few meters to hundreds of kilometers or more
in scale.
3.5.2
Gravity
The strength of the Earths gravity field at a given point on the

surface is sensitive to a weighted sum of density variations in
the near and far subsurface. Thus, gravity surveys can be useful
in settings where high-density materials, such as igneous or
metamorphic basement, are in contact with much lowerdensity materials, such as unconsolidated sediments or ice.
Depth to basement is perhaps the most common target of
gravity surveys. Measuring the gravity field at a given point
simply involves placing a gravity meter on the ground at the
desired site and following the necessary steps to acquire an
instrument reading. Acquiring high-quality gravity data,
however, requires care and ancillary data.
Some key aspects of the gravity method can be illustrated
with the analytical expression for the extremely simplified case
of a buried spherical anomaly. In this case
gsphere
G4pr 2 Dr
3x2 z2
1
Gravity anomaly
The key to successful use of near-surface geophysics in

geomorphology involves verifying the interpretation of geophysical data by using multiple sources of information. This
requires collection of in situ data for calibration or validation
via cores, trenches, outcrops, or geologic field observations, or
the use of multiple geophysical methods that yield complementary information. Successful studies increase data quality
by attempting to achieve high spatial resolution and by
choosing targets that produce strong geophysical signals (the
result of either careful reconnaissance or good luck). Productive strategies include field collection of rock and sediment
samples and subsequent laboratory analyses to determine
relevant physical properties, and development and modification of data-processing routines to reduce site-specific noise
and enhance target signal. Most positive factors add cost, in
terms of either data acquisition (e.g., high-resolution airborne
studies and multiple methods) or post-processing labor
(careful merging of separate data sets and development and
testing of nonstandard processing algorithms).
Because there is no forum in geophysics for describing sites
where methods, that might be expected to work, in fact do not
work, it is difficult to estimate what fraction of attempts to
use a given geophysical method yields results productive
enough to be included in geomorphological research and
publications.
This chapter focuses on the most commonly used geophysical techniques in practical terms, based upon the use of
simple physics, relationships between measured parameters
and soil/rock properties, survey requirements, and inherent
strengths and limitations. Conditions both likely and unlikely
to lead to productive results are discussed. The emphasis on
recent studies gives the reader an easy path to data sets that
exemplify what can be achieved with current equipment and
acquisition and processing methods. The uses of methods are
described on large and small scales, reflecting common usage.
Examples are presented that utilize multiple techniques in
successful combinations.
105
Figure 1 Schematic illustration of gravity anomalies over a 2-D

positive-density anomaly (red) and negative anomaly, such as a void
(orange area). (Top) Cross section of subsurface structure. (Bottom)
Gravity anomalies measured on a profile over these features. Red line
shows gravity anomalies that would be measured in the absence of
the orange void. The Red positive-density anomaly is assumed to
extend to great depth and past the end of the profile. Note the deep
positive anomaly produces a long-wavelength signal with a high over
the mass. The black line represents values that would be measured
over the combined features. Note that the effect of the shallow
negative-density anomaly is to depress gravity readings across a
shorter-wavelength region centered over the density anomaly. The
schematic emphasizes that gravity is sensitive to a range of depths
and that deeper sources produce broader anomalies.
where gsphere is the gravity anomaly measured at a point at the

surface that is a horizontal distance, x, from the center of the
sphere, z is the depth of the sphere center, G is the gravitational constant, r is the radius of the sphere, and Dr is the
density contrast between the sphere and the host rock. Gravity
anomalies are intuitive: excess mass (positive Dr) produces
positive density anomalies that peak over the source mass
the opposite for mass deficits (Figure 1). The gravity anomaly
weakens as 1/distance2 ( 1/(x2 z2)) to the excess mass.
Deeper targets produce lower-amplitude anomalies and
longer-wavelength anomalies. It is important to note that the
horizontal distance x must be on the order of z or greater
before gsphere decays significantly. So, the greater the target
depth, the longer the survey dimensions needed to capture the
gravity signal. In general, a survey dimension of Z5 times the
target depth is desirable.
Gravity signals are also inherently nonunique. The density
contrast Dr and the target dimension r can trade off in eqn [1]:
small, high-contrast anomalies can produce identical signals
to larger, lower-contrast features. Because the gravity anomaly
measured at any one point in the subsurface represents the
sum of the effects of all local and regional density anomalies
(similar to the sum of many gspheres), an infinite number of
density patterns could produce any given gravity profile. Interpretation of gravity data is therefore only useful when other
constraints are applied. Well data, seismic profiles, and field
observations can show where density contrasts do exist.
106
Simplifying assumptions are generally made about density

patterns, including a limited number of layers of features.
Practitioners generate a series of trial-and-error models of
Earth density structures, honing in on plausible geometries
that are compatible with the observations. The method is
called forward modeling, and, to keep the modeling process
tractable, most surveys are fit with two-dimensional (2-D)
models that make simple assumptions about the extent of
features in the third dimension. Consequently, gravity surveys
should be run perpendicular to the expected strike of features
of interest. Advanced software can automate the process of
finding a geometry that best fits the observations, but users are
still required to limit the number of unknown variables in the
problem by making significant assumptions about the starting
model. This latter process is called inverse modeling.
Modern gravimeters are capable of measuring variations in
the Earths gravity field with a precision of 0.0010.01 mGal
(1 mGal 105 m s2). To illustrate the sensitivity, a spherical
void in limestone (density 2000 kg m3) at 10-m depth, with
radius 2.6 m, produces an anomaly of 0.01 mGal directly over
the sphere. A significant number of corrections are needed to
reduce a gravity measurement to a value that reflects only
subsurface density anomalies. Data need to be corrected for
elevation (gravity weakens with increasing distance from the
center of the Earth) and the effects of the gravitational attraction of local and regional topography. These corrections
can be large compared to target anomalies. A difference in
5 cm in elevation corresponds to a correction on the order of
the instrument sensitivity (0.01 mGal); hence, gravity measurements must typically be made at benchmarks or at elevations surveyed with differential global positioning system
(dGPS) equipment. Instrument drift and the absolute value of
instrument readings have to be calibrated by acquisition of
data at known base stations. In practice, several minutes are
required for each high-quality gravity-plus-elevation measurement, and including trips to base stations, at most a few to
tens of data points can typically be collected per day.
Aerogravity surveys can acquire data at much higher rates
and can be extremely useful for reconnaissance surveys in
inaccessible settings. The high cost of such surveys and dataprocessing issues, however, limits widespread applicability.
Satellites such as Seasat (Sea Satellite), Geosat (Geodetic Satellite), ERS-1 (European Remote-Sensing Satellite 1), GRACE
(Gravity Recovery and Climate Experiment), and GOCE
(Gravity Field and Steady-State Ocean Circulation Explorer)
have measured the gravity field over continent and oceanbasin scales.
3.5.2.1
Strengths
Gravity is widely used as a reconnaissance tool, because information about relatively deep structure (hundreds of meters
to kilometers or more) can be derived from surficial observations at much lower cost than seismic methods. Small-scale
near-surface features (uppermost tens of meters) produce
smaller gravity anomalies, requiring extreme care in elevation
measurements and data processing. These high-resolution
surveys are commonly referred to in the literature as microgravity surveys, as they require precision on the order of
microgals (0.001 mGal). Despite the labor required, surveys

targeting these smaller anomalies are yielding fruitful new
information in studies of weathering processes and volcano
evolution.
3.5.2.2
Limitations
The nonuniqueness of the gravity method limits definitive

interpretation of results in the absence of other direct or
sensor observations. Gravity is best used in combination with
other geophysical measurements. Automobile traffic causing
ground vibrations can render data acquisition difficult in
daytime in urban environments. Many local surveys are referenced to a local base station and cannot be easily incorporated into larger databases.
3.5.2.3
3.5.2.3.1
Gravity Case Studies

Faulting
Gravity profiles across faults, particularly those that juxtapose

basement against less dense sediments, have been a staple of
tectonic studies for decades (e.g., Telford et al., 1990). Gravity
data are typically used in conjunction with field observations,
and with magnetic and/or seismic data. Here, we illustrate
gravity methods with results of a combined magnetic, gravity,
and light detection and ranging (LiDAR) topography analysis
by Blakely et al. (2009) for faulting on scales of tens of kilometers (Figure 2). This study demonstrates the utility of
gravity mapping in conjunction with magnetic surveys and
common interpretation techniques. The authors compute
isostatic residual gravity anomalies (i.e., they assume that
topography is isostatically compensated and that remaining
gravity anomalies reflect middle and upper crustal density
heterogeneities). With gravity and magnetic observations, the
attitude of active faults mapped at the surface from LiDAR and
field observations can be projected downward to many kilometers depth. Figure 2(a) and the corresponding cross section
on Figure 3 demonstrate classical techniques of (1) compiling
a gravity map to identify regional strike directions and (2)
developing a cross-strike 2-D model that accounts for geologic
observations, and gravity and magnetic data. It is the contrast
between higher-density Crescent Formation basalts and lowerdensity surrounding sediments that produces an B75-mGal
gravity signature that can be modeled. Figure 3(c) illustrates
the complexities inherent in forward modeling of gravity data,
such as the assumptions that must be made about density
contrasts of rocks at depth, as well as the geometry of contacts.
Most authors present a best-fitting model that is based on a
conceptual structural model, as done by Blakely et al. (2009)
for Figure 3(c) from Tabor and Cady (1978). It can be illuminative when authors also illustrate a range of models, and
the relative quality of their fits.
3.5.2.3.2
Volcanic processes
Gravity, combined with geological observations, can provide

useful constraints on isostatic compensation mechanisms for
volcanic edifices, which influence and are in turn influenced
by erosion processes and deformation histories. Gravity data
complement surface geological information for understanding

12330
48
123
12230 12330
123
107
12230
4845
4830
10 km
20
29
12
4
16
>30
2030
1020
010
28
40
34
396
141
56
136
184
227
334
nT
(a)
Susceptibility, SIU 103
mGal
(b)
Figure 2 Example of combined magnetic and gravity survey. Magnetic and gravity data over the Saddle Mountain Fault area in northwest
Washington, USA. (a) Aeromagnetic anomalies. Colored dots show magnetic susceptibility measurements of outcrop samples. (b) Isostatic
residual gravity anomalies. Red dashed line indicates location of geologic cross section shown in Figure 3. White barbed lines are thrust faults.
White solid lines map outlines of exposures of the basaltic Crescent Formation between sedimentary deposits. White dotted line shows mapped
contact between lower and upper members of the Crescent Formation (see Figure 3). Coincidence of geophysical signatures and mapped faults
indicates fault structures extend to depth. Figure 3 shows forward models of fault structure. Modified from Blakely, R.J., Sherrod, B.L., Hughes,
J.F., Anderson, M.L., Wells, R.E., Weaver, C.S., 2009. Saddle Mountain fault deformation zone, Olympic Peninsula, Washington: western
boundary of the Seattle uplift. Geosphere 5(2), 105125, with permission from GSA.
active volcano evolution (e.g., Brown et al., 1987). For example, surveys on Concepcion volcano in Nicaragua show a
remarkably low bulk density (1700 kg m3) for the edifice as a
whole. This is caused by thick deposits of low-density scoria
that are presumed to be the underlying cause of the frequent
occurrences of lahars (Saballos et al., 2010).
Recent time-lapse gravity studies at volcanoes have generated new constraints on processes underlying surface deformation and unrest. Typically, surface deformation is
initially noted from surface observations or interferometric
synthetic aperture radar (InSAR). Because microgravity data
can be used to constrain mass changes at depth, in several
studies, these data are key to identification of magma as the
driving force for surface deformation (e.g., Rymer, 1994;
Budetta et al., 1999; Carbone et al., 2003; de Zeeuw-van
Dalfsen et al., 2005; Tizzani et al., 2009). Such interpretation
of volcano gravity surveys requires corrections beyond
those typically considered, including temperature and watertable effects (Battaglia et al., 2008). Because deformation
processes at volcanoes are so rapid, investigations are
currently limited by the high costs of automatic gravity data
acquisition with high temporal resolution (Williams-Jones
et al., 2008).
3.5.2.3.3
Temporal changes in topography are observed on scales ranging from sinkholes to continents. Gravity data can provide
insights into the processes driving changes on all these scales.
For instance, microgravity surveys are used to estimate mass
deficiency associated with sinkhole formation and ground
subsidence on lateral scales of meters to tens of meters (e.g.,
Calhoun et al., 1989). Branston and Styles (2003) showed
remarkably good agreement between time-lapse microgravity
surveys and resistivity surveys in a study targeting the source of
subsidence in an area of historical salt mining. The combined
data sets show a zone of low-density ground 34 m below
the subsiding surface. Interestingly, they documented that the
zone of low density is migrating upward over a 3-year period.
On intermediate length scales, circular gravity anomalies,
either positive or negative, can be indicators of impact structures (e.g., McCall, 2009; Dentith et al., 1999). On scales
of hundreds of kilometers or more, gravity data are key
to understanding the causes of continental-scale surface
dynamics. Gravity signatures have been used for many years
to distinguish mantle processes (e.g., Forte et al., 2010),
lithospheric flexure (e.g., Prezzi et al., 2009), and glacial isostatic adjustment (e.g., Steffen et al., 2009).
108
Magnetic Anom. (nT)
1000
NW
SE
500
Observed
0
500
Calculated
1000
Gravity Anom. (mGal)
(a)
(b)
25
0
Calculated
25
Observed
50
Mapped faults
Flight path
= 0.12, = 0.006
= 0.05, = 0
= 0.01, = 0.00
Depth (km)
= 0.20, = 0.014
= 0.05, = 0
10
= 0.25, = 0.088
= 0.05, = 0.025
= 0.25, = 0.050
= 0.20, = 0.020
20
= 0.05, = 0
Vertical exaggeration 1:1
(c)
Hurricane ridge fault
Saddle Mt. fault
Depth (km)
Pliocene and younger

sedimentary deposits
OligoceneMiocene
sedimentary rocks
Eocene crescent formation,
upper member
Eocene crescent formation,
lower member
10
20
0
(d)
10
20
30
Distance (km)
40
50
60
EoceneOligocene
sedimentary rocks
Figure 3 Successful forward models of gravity and magnetic data for the profile shown with red dashed line in Figure 2. Although other models
could be generated that fit the data, this model is shown because it agrees with the proposed regional tectonic structure (Tabor and Cady, 1978).
The models assume density and susceptibility structure shown in (c). (a) Model and observed magnetic anomaly. (b) Model and observed gravity
anomaly. (c) Model densities and susceptibilities. (d) Geologic interpretation of model. Dr density contrast relative to 2670 kg m3.
w magnetic susceptibility in SI units multiplied by 1000. Note that some contacts act as sources of both magnetic anomalies and gravity
anomalies, while others only act as sources for one or the other. The overall attitude of beds proposed fits the form and amplitude of both sets
of anomalies. Reproduced from Blakely, R.J., Sherrod, B.L., Hughes, J.F., Anderson, M.L., Wells, R.E., Weaver, C.S., 2009. Saddle Mountain fault
deformation zone, Olympic Peninsula, Washington: western boundary of the Seattle uplift. Geosphere 5(2), 105125, with permission from GSA.

3.5.2.3.4
Slope processes
Gravity surveys can constrain subsurface structures that are

related to vertical motions. For example, Gottsmann et al.
(2008) used gravity data to construct a model that explains the
position of a landslide in the Central Volcanic Complex of
Tenerife.
Large-scale (tens to hundreds of kilometers) airborne and

satellite gravity data are used to extract information on subglacial geology (e.g., Bell et al., 1998), glacial isostatic adjustment (e.g., Tamisiea et al., 2007), and mass changes in ice
sheets (Riva et al., 2009). Recent satellite gravity data are of
sufficient quality that the relative magnitudes of the latter two
can be distinguished over Antarctica (Riva et al., 2009).
Gravity data have also been used as an ancillary technique for
assessing the existence and thickness of permafrost (Vonder
Muehll et al., 2000), although other geophysical methods are
typically more effective.
3.5.3
Gravity data are key to distinguishing origins of local observed

sea-level changes. Local changes represent a combined response to glacial isostatic adjustment, tectonic motions,
changes in ocean water temperature and salinity, and eustasy
(e.g., Milne and Mitrovica, 2008).
3.5.2.3.6
Magnetics
Magnetic methods share fundamental similarities with gravity

methods. In magnetic surveys, as in gravity surveys, the total
strength of the field is measured at points on or above the
surface, except that the measurements are sensitive to rock
magnetization rather than rock density. Igneous rocks typically have much higher intensities of magnetization than
sedimentary rocks; so, magnetic data are useful for assessing
sediment thickness over basement. Igneous dikes and sills
produce distinctive anomalies; hence, magnetic data can be
particularly useful for understanding pluton emplacement and
associated host rock deformation and erosion patterns.
Interpretation of magnetic anomalies is less intuitive than
that of gravity anomalies for several reasons. Magnetization is
inherently dipolar, and the direction of target rock magnetization may be unknown. If rock magnetization is remnant,
the rock locks in the direction of the magnetic field that
existed at the time it cooled through the Curie temperature,
around 300 1C. If rock magnetization is induced, then the
magnetization direction parallels that of the present-day
Earths field, and the strength of the rocks magnetization is a
function of the rocks magnetic susceptibility. Adding further
complexity, the measured total field strength (Earths main
field plus anomalous field produced by study targets) depends
on both the direction and strength of the main field and the
anomalous field at the time the measurement was made.
Identical buried structures produce different magnetic anomalies depending on their strike, and the inclination and declination of the Earths field at the time of measurement
(Figure 4). Drift corrections must be made when collecting
Magnetic anomaly
3.5.2.3.5
109
Figure 4 Schematic illustration of magnetic anomalies over a 2-D

target with higher magnetic susceptibility than background (red area),
and a target with lower susceptibility than background (orange). Red
line shows the magnetic anomalies that would be measured on an
EW traverse across target structures striking NS in an environment
with an Earths field inclination of 601. Solid black line shows the netmagnetic anomaly from both targets. Dashed black line shows
magnetic anomalies if the scenario were rotated 901, to an NS
profile. The schematic emphasizes that the form of the magnetic
anomalies is more complicated than gravity anomalies (compare with
Figure 1) and the form depends on the orientation of the profile and
the local direction of the Earths field.
magnetic data, not typically for instrument response drift, but

for diurnal variations in the Earths field.
Following data acquisition and processing, users typically
generate models that can fit the observations. As with gravity,
these models are inherently nonunique and must be constrained wherever possible by geologic observations. For
studies on scales of kilometers to hundreds of kilometers,
magnetic data are commonly interpreted in conjunction
with gravity data, and forward models are sought that fit both
data sets.
It should be noted that this chapter discusses field-based
surveys of the total magnetic field strength, and not magnetic
properties of samples measured in the laboratory. Laboratory
measurements associated with rock magnetization, however,
are used in geomorphology, including paleomagnetic measurements that are used to reconstruct the timing of the formation of volcanic deposits. Laboratory measurements of
anisotropy in magnetic susceptibility from field samples are
used to understand development of slip zones in landslides
(Herrero-Bervera et al., 2002; Mamtani and Sengupta, 2009).
Finally, magnetic susceptibility measurements of core samples
are common components of sedimentological studies in
coastal settings and for paleoenvironmental reconstructions.
3.5.3.1
Strengths
Unlike in gravity surveys, high-precision elevation measurements are not required for ground-based acquisition of
110
magnetic data. With modern magnetometers, data can be

collected with instruments carried by a single person at
walking speeds.
3.5.3.2
Limitations
Ground-based magnetic surveys can be very noisy, as natural

lithologic variability, particularly in igneous rocks, can generate large-amplitude, short-wavelength magnetic anomalies.
On the other hand, the short-wavelength variability can be
exploited as an interpretation tool, as some geologic units will
produce much more erratic magnetic data than others. In any
developed setting, cultural noise can be a significant problem
and agricultural areas can be problematic where metal is used
in irrigation systems. Finally, because of the inherent nonuniqueness of magnetic data, other in situ data or sensor observations are essential for meaningful interpretation.
3.5.3.3
3.5.3.3.1
3.5.3.3.3
Many weathering and soil-formation processes enhance

magnetic signals through the formation of magnetic minerals
(Evans and Heller, 2003). Surface measurements of magnetic
signal amplitude and spatial variability can be used to distinguish variability in weathering processes and rates, as well
as to characterize underlying variability in bedrock lithology
(e.g., Dimalanta et al., 2009). In addition, magnetic surveys
can provide insight on the internal structure of impact craters.
For example, Shah et al. (2009) used core-derived magnetic
properties to identify basement-derived megablocks as the
source of significant magnetic anomalies over the Chesapeake
Bay impact structure.
Case Studies
Faulting
To illustrate an effective magnetic survey, we return to the

Blakely et al. (2009) study of active faulting on the Olympic
Peninsula of Washington, USA (Figures 2 and 3). Blakely et al.
(2009) combined aeromagnetic data with more detailed surface-based data acquisition, including a canoe-mounted profile across a lake. The magnetic method exploits the magnetic
susceptibility contrast between basalts and sediments, and
between individual basaltic units. Direct measurements of
susceptibilities at outcrops (Figure 2) provide added confidence to the magnetic forward modeling (Figure 3). It is
interesting to note that the magnetic data provide complementary information to the gravity data. For example, in the
forward model, magnetic anomalies respond to high susceptibilities in the uppermost basalt unit, whereas gravity
anomalies respond to strong density contrasts bounding the
entire basaltic package (Figure 3).
3.5.3.3.2
With recent improvements in the resolution of surveys with

very close line spacings, time-lapse aeromagnetic methods
(repeated surveys) now can detect the subtle changes associated with volcanic activity (e.g., Ueda, 2007; Nakatsuka et al.,
2009).
Volcanic processes
Magnetic surveys are effective at delineating the boundaries of

volcanic units. Because the geometry of deposits sheds light on
their deposition and emplacement mechanisms, magnetics
can be useful for understanding these processes (e.g., Diez
et al., 2009; Wetmore et al., 2009), as well as the resultant
surface topography (e.g., Morgan et al., 2003b). Zones of low
magnetization can be associated with demagnetization caused
by heat or hydrothermal activity (e.g., Okuma et al., 2009).
Identifying areas affected by hydrothermal activity may be of
practical importance because hydrothermally altered rocks can
form weakened zones with potential to form debris flows (e.g.,
Finn et al., 2007). Debris flow deposits or volcanic fans will
also have lower magnetizations than intact rock (Okubo et al.,
2009).
Thermally active volcanic zones can be associated with
temporally variable magnetic fields. Thus, monitoring of
magnetic field strength may be useful for identifying thermally
active sites (Kanda et al., 2010). Various temporal relationships
between magnetic field changes and ground deformation have
also been observed (Hurst et al., 2004; Zlotnicki et al., 2009).
3.5.3.3.4
Magnetic surveys, especially aeromagnetic surveys, are used to

map subglacial lithology (e.g., Ferraccioli et al., 2009). Bell
et al. (1998) used aeromagnetic data, along with airborne
radar and gravity, to demonstrate that underlying geology influences ice-stream dynamics in the West Antarctic ice sheet.
3.5.4
Resistivity and EM Methods
The resistivity, r, of ground (r 1/conductivity) represents the

combined electrical response of matrix and pore waters.
Measurements of ground resistivity can be useful for detecting
lithologic boundaries, degree of saturation, and variations in
pore-water conductivity. Resistivities are also mildly temperature dependent, a characteristic that is exploited in volcano
studies and must be corrected for in other settings. Terrain
resistivities can be quite sensitive to the degree of fracturing
and weathering. Saturated fractured rocks are typically better
conductors than their intact equivalents; hence, their resistivity
is lower. Many clays that are a by-product of weathering
have high conductivities relative to their protoliths. Sediment
basement contacts typically exhibit significant resistivity
contrasts.
Two different types of methods are used to measure terrain
resistivities. The first method sends a direct current through
the ground; this is called resistivity sounding, resistivity profiling, electrical resistivity tomography (ERT), or electrical resistivity imaging (ERI). With the second method, currents in
the ground are induced by time-varying magnetic fields; this
is known as EM methods. This second category includes the
magnetotelluric (MT) method, which exploits naturally occurring time-varying magnetic fields generated over a broad
frequency range by lightning storms and other natural phenomena. Other EM methods require instruments with an active source that generates currents in the ground by creating
alternating currents in a transmitter coil. These AC currents
spawn their own time-varying magnetic field, which in turn
drives currents in the ground. The secondary magnetic field
associated with the ground currents is detected in a second
passive coil. These methods include controlled source

audio-magnetotellurics (CSAMT) and frequency-domain
electromagnetics (FEM or simply EM). Time-domain electromagnetics (TEM or TDEM) work slightly differently, as they
induce currents to flow in the ground at the moment a constant current in an instrument coil is shut off.
Measurements of terrain resistivity via ERT are relatively
labor intensive. A known current is driven through the ground
between a pair of electrodes (metal stakes) and the potential
is measured between nearby electrode pairs. The method is
automated by modern instruments pre-programmed to make
selected combinations of measurements with pairs of electrodes within a typically linear array of electrodes (25 to about
100) set in the ground (Figure 5). A DC current must be
established between a pair of current electrodes, and reestablished in the opposite direction to avoid charge buildup on
electrodes. Because it takes a few seconds to establish each set
of DC currents, completing surveys with many tens of electrodes can take up to several hours after the electrode array is
in place.
The separation of electrodes is related to the overall depth
of current flow; the greater the separation, the greater is the
depth sensitivity of the measurement. Depths of hundreds of
meters can be imaged when power supplies are large enough
to drive current that will produce measureable potential differences over large scales. Twelve-volt (12 V) battery-driven
systems can achieve penetration depths of several tens of
meters or more, depending on terrain conductivity.
111
The potential that arises between a given pair of electrodes

as current flows depends on the electrode positions and on the
subsurface resistivity structure. To recover the resistivity structure requires inverting the potential measurements for a
ground resistivity pattern that best fits the readings. The inversion is typically done with commercial software. Common
practice with modern systems is to acquire and invert data
along profiles, assuming 2-D structure. However, 3-D data can
in theory be acquired and inverted with the same systems.
A variety of traditional survey geometries (different spatial
patterns of current and potential electrodes) are commonly
used (Wenner, dipoledipole, etc.), although a growing body
of literature suggests that resolution can be maximized with
nontraditional geometries (Stummer et al., 2004; Wilkinson
et al., 2006). For most survey geometries, the maximum depth
of resolution will be approximately 1/5 of the longest spacing
between electrodes in any current/potential combination. This
high ratio of survey length to penetration depth makes resistivity surveys vulnerable to off-profile heterogeneities as current
flows radially out from and into electrodes. As almost all results
reported in the literature involve 2-D inversions, it is useful to
be aware of potential 3-D features that could affect a survey.
Surveys produce noisy data in settings where it is difficult
to make good contact between electrodes and ground. Dry
surface sands yield notoriously noisy data and require spraying saltwater around electrodes to lower contact resistances.
EM methods are typically much less labor intensive than
ERT, and are more commonly used in reconnaissance surveys.
Depth
Low resisitivity
High resisitivity
Figure 5 Schematic illustration of resistivity inversion from data collected over 2-D subsurface heterogeneities. In this model, the yellow area is
less resistive (more conductive) than background and the cylindrical orange target is less conductive than the background. This schematic
illustrates features of resistivity profiles, where sharp contrasts appear as gradients, and the resolution is better near the surface compared to the
resolution at greater depth. For example, the dip angle of the semi-vertical contact and the lower extent of the cylindrical target are not correctly
reproduced in the inversion process. The true resistivities of smaller features (such as the orange target) are not typically recovered. The nature
of such artifacts depends also on the inversion algorithms used.
112
Because current is induced in the ground by alternating

magnetic fields, no direct contact between instrument and
ground is required. Airborne EM surveys are used, particularly
in volcanological studies.
Depth of investigation in EM surveys is generally inversely
proportional to the EM frequencies used. MT recordings over
many hours or days can capture low frequencies and gather
information about features that are kilometers in depth. The
highest-frequency portable alternating current EM systems
record at B10 000 kHz and sample only the uppermost few
meters. The geometry and size of the receiver coils that sense
the secondary magnetic fields vary with manufacturer design,
but are generally inversely proportional to instrument frequency. To capture a range of depths may require the use of
multiple instruments. With time-domain electromagnetics,
the secondary currents generated in the ground flow initially at
shallow depths, and propagate deeper with time. Thus, the
resistivity structure over a range of depths can be extracted
from the progressively deeper current flow at a single instrument site.
Solving even the forward problem of induced current flow
for either frequency-domain or time-domain instruments is
difficult for structures more complicated than horizontal
layering. For shallow investigations, data interpretation is
typically done using the horizontal layering assumption at
each reading location. For airborne data, more advanced 2-D
and 3-D inversion schemes are sometimes used.
In comparison with resistivity profiling, EM method survey
footprints can be quite small. A time-domain EM sounding at
a single site can sample to a depth many times of the dimension of the current coil laid out on the ground. However,
unlike DC resistivity, EM methods are incompatible with
nearby sources of cultural noise.
3.5.4.1
Strengths
Ground resistivities are sensitive to changes in porosity and

water saturation, as well as lithology. Resistivity and EM
methods are thus well suited to investigating phenomena that
influence surficial processes, such as the degree of sediment
compaction, presence of fractures, and degree of water saturation. These methods work in high-conductivity environments
(brackish water and clays) where ground-penetrating radar
fails.
3.5.4.2
Limitations
Because numerous factors influence the electrical behavior of

ground, it can be difficult to differentiate pore water, matrix,
and temperature effects. Quantitative resistivity measurements
at one site cannot typically be extrapolated to other sites. For
meaningful interpretation, local in situ measurements are
critical. Inversion algorithms have inherent limitations, as
ground resistivity profiles are smoother than reality, so sharp
contacts are smeared. Furthermore, resistivity contrasts of
small anomalies are significantly underestimated (e.g., DayLewis et al., 2006) and 3-D structures with short dimensions
perpendicular to a 2-D profile are improperly characterized in
1-D and 2-D inversions.
3.5.4.3
3.5.4.3.1
Case Studies
Faulting
Where hanging wall and footwalls have distinct lithologies,

resistivity profiles across faults can help constrain fault attitude
(Diaferia et al., 2006; Lee et al., 2009). On some faults, fault
gouge or other materials within the fault zone have a distinct
low-resistivity signature (e.g., Diaferia et al., 2006; Suski et al.,
2010). Pseudo-3D surveys, which involve a compilation of
multiple parallel resistivity profiles, can help confirm the
identification of fault structure and better establish fault attitudes (e.g., Vanneste et al., 2008).
3.5.4.3.2
Volcanic processes
Hydrothermally active zones are conductive (low resistivity)

and many investigators have exploited this property to better
understand fluid circulation in volcanic processes. Examples
of recent ERT studies include Barde-Cabusson et al. (2009)
and Finizola et al. (2009). Time-domain electromagnetics
studies yield hydrostratigraphy to hundreds of kilometers
depth (MacNeil et al., 2007; Srigutomo et al., 2008). Yamaya
et al. (2009) and Kanda et al. (2010) used audio-frequency
magnetotellurics, the latter combining the EM with magnetics.
Finn et al. (2007) used helicopter-borne EM data together with
magnetic data and mapping to evaluate the extent of hydrothermally altered rocks with avalanche potential.
3.5.4.3.3
Resistivity profiles have been commonly used for detecting

approximate depth to bedrock for geomorphological and
hydrogeological investigations. Bedrock typically has much
higher resistivity than regolith. Recent studies emphasize new
methods for extracting information on this type of contact
from 2-D surveys (e.g., Hsu et al., 2010). Resistivitys sensitivity
to sediment/rock moisture has been used to study relationships between sediment moisture and weathering (Mol and
Viles, 2010). Braun et al. (2009) combined EM (Barbiero et al.,
2007) and resistivity surveys with core sample mineralogy and
geochemistry to derive a long-term mass balance for chemical
weathering of regolith in gneissic terrain (Figure 6). This study
illustrates the utility of reconnaissance surveying with EM
before running more labor-intensive resistivity profiles, and
the importance of field-data acquisition for interpreting resistivity data. The authors estimated regolith thickness based
on a resistivity threshold (Figure 6(b)), in this case 400 O m.
Braun et al. (2009) further assessed the uncertainties in resistivity-based thickness estimates both with simple synthetic
models and by comparing profiles and boreholes.
3.5.4.3.4
Slope processes
Material in landslides, such as intact blocks (Sass et al., 2008)

or earthflows (Schmutz et al., 2009), can be distinguished
from substratum with resistivity methods (Naudet et al.,
2008). Bichler et al. (2004) and Jomard et al. (2009) resolved
stratigraphic units in landslide deposits many tens of meters
deep. Mauritsch et al. (2000) applied EM and resistivity
methods, together with seismic refraction on an unstable zone
of a large landslide. They found the combined methods could
not resolve a distinct slip surface, although EM data are in
good agreement with seismic refraction data in determining
113
9
23.7 m 9
10
16.2 m 8
11
BH6
22.5 m 11
13.2 m 7
12
Weathered rock
Fresh rock
16.8 m 12
19.5 m 5
16.2 m 4
BH1
2
1
ERT calculated resistivity (Ohm.m)
17.4 m
6
BH5
14.1 m 3
BH12
00
30 m
14.2 m 2
13.5 m 1
30 m
6
5
4
19.5 m 10
m
0
50
500 m
500 m
10 26 66 168 429 100 815 206

7
1 2
Estimated regolith thickness (m)
Figure 6 Example of resistivity profiles targeting regolith thickness. Resistivity profiles were run on the small, subhumid, forested Mule Hole
watershed on the margin of the Karnataka Plateau, India. Regolith thicknesses are estimated assuming a threshold resistivity (400 ohm-m)
between weathered and fresh rock. Uncertainties associated with this method were then assessed by comparison of boreholes (marked with BH)
and by modeling of resistivity signatures of simple synthetic models. Modified from Braun, J.J., Descloitres, M., Riotte, J., et al., 2009. Regolith
mass balance inferred from combined mineralogical, geochemical and geophysical studies; Mule Hole gneissic watershed, south India.
Geochimica et Cosmochimica Acta 73(4), 935961, with permission from Springer.
thickness of overburden in some areas. Helicopter-borne EM

surveys over landslides in Japan showed positive correlations
between zones of high conductivity and locations of previous
landslides (Konishi, 1998).
3.5.4.3.5
Coastal processes
Most resistivity surveys in coastal settings have hydrogeological targets. As saltwater is a much better conductor than
freshwater, resistivity is an excellent technique for mapping
the freshwater/saltwater interface. A number of studies, however, have used resistivity methods to target coastal stratigraphy and to link coastal hydrogeology and stratigraphy. For
example, Schneider and Kruse (2005) used ERT and EM surveys to suggest that rapid coastal erosion has left a barrier
island out of equilibrium with its groundwater system. Voett
et al. (2009) used ERT to trace the lateral extent of Holocene
tsunami deposits. Massey and Taylor (2007) used ERT surveys
as an integral part of a paleoenvironmental reconstruction in
the southwest coast of England.
3.5.4.3.6
Fluvial processes
Resistivity data can also be used to extrapolate fluvial architectural elements between boreholes or trenches (Bersezio
et al., 2007; Hickin et al., 2009). Hickin et al. (2009) combined resistivity with ground-penetrating radar, using resistivity as a proxy for sediment grain size. Froese et al. (2005)
used a combination of resistivity, ground penetrating radar
(GPR), and cores to construct a sediment mass balance for a
large river system.
3.5.4.3.7
Resistivity data can be used in a manner similar to magnetic

data for extracting information on subglacial geology (e.g.,
Ahmad et al., 2009). Massive ice and ice-rich sediments can be
distinguished in rock glaciers with resistivity (e.g., Ribolini

et al., 2007). Kristensen et al. (2009) used resistivity profiling
to help distinguish between glaciogenic sediments, buried
glacier ice, and redeposited marine muds on the edge of surge
moraines.
Resistivity profiles can be useful for detecting the presence
of permafrost (e.g., Vonder Muehll et al., 2002), as well
as for characterizing permafrost and rock glaciers (e.g., Maurer
and Hauck, 2007). Ice and permafrost tend to have higher
resistivities than unfrozen sediments or bedrock. Hauck
et al. (2003) showed that one can address the problem of
high-resistivity surfaces in dry permafrost environments by
attaching saltwater-soaked sponges to the electrodes. They
demonstrated the resolution limits of the resistivity method in
these extremely resistive glacial settings. Because resistivity
surveys can help distinguish icedebris mixtures from massive
ice, they provided insight on the evolution of rock glaciers
(Kneisel, 2004; Kneisel and Kaab, 2007; Ribolini et al., 2010).
Recent research has been conducted using 25-m electrode
spacing (multielectrode surveys), achieving depths of penetration on the order of 1040 m (Hauck et al., 2003; Ribolini
et al., 2010; Kneisel and Kaab, 2007).
3.5.5
Ground-Penetrating Radar
Most commercial ground-penetrating radar systems transmit

an EM pulse from one antenna and record returns at a second
adjacent antenna. The antennas are typically placed on or
just above the ground surface. The receiving antenna records
pulses that reflect off of contacts or scatter (diffract) off of
point-like discontinuities in the subsurface. Radar pulses are
reflected or scattered where there are abrupt changes in radarwave velocity. Wave velocity is controlled by the ground
114
cannot be imaged directly because no energy is reflected back

to the surface, and point-like objects and sharp corners appear
as inverted hyperbolas. These distortions can be largely corrected for by using migration (a reflection seismic data processing technique). Many published GPR images, however,
have not been corrected using migration.
Because wave velocity varies with water content, a radar
profile can look quite different in the wet season than in the
dry season, and radar-wave velocities have to be established on
the day profiles are acquired. Knowledge of the velocity
structure at time of survey is essential for converting the arrival
times (Figure 7) into the depths of the corresponding reflectors. Three methods are used to determine velocities. First,
special surveys can be conducted where antenna separation is
expanded around a common mid-point (CMP surveys). Velocities can be readily determined from the increase in arrival
times with increasing antenna separation, in environments
where relatively flat-lying reflectors are present. Second, the
shape of the wings of the diffraction hyperbolas is velocity
dependent, and can be used to determine velocities where
point-like objects produce diffractions. Third, wells, outcrops,
or trenches can be used to directly correlate stratigraphy and
radar returns.
Finally, wave energy that travels through ground and air
directly from the transmitting antenna to the receiving antenna must be interpreted differently than reflected arrivals.
This energy is invariably the first arrival, typically fairly uniform in appearance across a profile. An untrained analyst
might misinterpret these as shallow reflections parallel to the
Time
through which the wave is traveling, and depends primarily on

the electric permittivity (dielectric constant) in the medium.
The permittivity varies with porosity, lithology, and, to a large
extent, the degree of water saturation. In ice, permittivity varies
with ice structure, water content, and rock debris.
As transmitter and receiver are moved across the surface,
and the pattern of returns recorded produces a filtered
and somewhat distorted view of subsurface discontinuities
(Figure 7). The filtering of the subsurface in the GPR profiles
comes from several sources: (1) most systems produce pulses
that are tri-lobed (positivenegativepositive), so that a single
contact to an untrained interpreter may appear as two parallel
contacts (Figure 7); (2) resolution of scattering point-like
objects typically decreases with depth (e.g., Irving and Knight,
2003; Kruse et al., 2010); (3) the GPR wave is polarized, so
returns from dipping contacts can be sensitive to antenna
orientation (e.g., Roberts and Daniels, 1996); (4) strongly
reflecting horizons can produce multiples, in which energy
travels twice round-trip between the surface and the reflecting
horizon; and (5) although layers thinner than B14 wavelength
may be detected, their thickness cannot typically be resolved
(Guha et al., 2005). In general, the attitude of bedding can be
determined while individual contacts within layered units
generally cannot be resolved.
The distortion of the subsurface image stems from the fact
that arrivals recorded at the receiver are plotted on a vertical
trace; so, returns appear to have originated directly below the
receiver. Figure 7 illustrates some of the complexities that arise
as true dips of beds are distorted, very steeply dipping contacts
Figure 7 Schematic illustration of GPR traces collected over 2-D targets with permittivities different from that background. Transmitting and
receiving antennas are typically moved in sync across the survey. Characteristic features of GPR profiles are illustrated. Steep surfaces are not
resolved because no energy is scattered back to the surface (orange dashed line). Sharp edges (green returns) and point-like objects (purple
returns) produce diffraction patterns. Energy traveling directly from transmitting antenna to receiving antenna (red arrows in inset, red pulses)
from the air and ground represent the first arrival. These direct waves are often not filtered from the record. Pulses typically have a tri-lobed
form so that a single contact can appear as a double-layer to the untrained interpreter.
surface. It is also important to note that these direct arrivals

render the earliest reflections undetectable, and, hence, GPR
profiles do not typically yield useful information at the shallowest depths. Care is needed to extract information on
structure at depths less than B1.5 radar wavelengths, which
can be a few centimeters or as much as a few meters, depending on antenna frequency. More recently, several authors
have demonstrated that the first arrival of the ground wave can
in fact yield useful information on the uppermost shallow
layer (Grote et al., 2003; van der Kruk et al., 2006, 2009). Such
surveys require different acquisition geometries than shown in
Figure 7.
3.5.5.1
Strengths
GPR permits rapid acquisition of stratigraphy in a range of

resistive environments. As Table 1 illustrates, GPR has proven
productive in various capacities for imaging most shallow
geomorphological targets. In clay-free sedimentary and volcanic
environments, stratigraphy can be recovered in the uppermost
few meters to, in exceptional cases, tens of meters. Because ice is
very resistive, penetration is excellent in ice and ice/rock mixtures, and glacial stratigraphy and ice thickness can be resolved
to depths of hundreds of meters. GPR systems can also be
floated on freshwater or run on lake floors to obtain excellent
sub-lake floor stratigraphy. Frozen-lake surfaces are similarly
good environments. It is important to note that there is no
penetration in saltwater or saltwater-saturated ground.
GPR data are relatively easy to acquire. Profiles can be
obtained at speeds ranging from hundreds of meters per day
in rough or wooded terrain to automobile speeds on open
roads. Airborne data, useful over glaciers, can be acquired even
more rapidly. With shielded antennas, data can be collected in
the presence of cultural noise, even inside buildings. Modern
systems are linked to global positioning system (GPS) or run
with odometers; therefore, positioning errors are significant
only in very-high-resolution 3-D surveys. Most commercial
systems are capable of operating over high and low temperatures and in moderate rain. A final strength of the method is
that real-time data displays make it possible to rapidly determine whether the method is effective at a given site.
permit simultaneous acquisition with multiple antenna

frequencies.
3.5.5.3
3.5.5.3.1
Limitations
GPR is not effective in conductive environments. Conductive

clay minerals and brackishsaline water reduce penetration
and render the method useless in many settings. Even worse,
when no real subsurface reflections are present, reflections
from aboveground objects and system noise can sometimes
appear to represent geologic features. Some training and care
must be taken to confirm a subsurface origin for interpreted
features. If a soil survey indicates high-conductivity soils, it is
unlikely that GPR will work well in that setting. In the United
States, Doolittle et al. (2007) have generated maps showing
GPR suitability. Even at locations where GPR penetration is
good, there is an inherent trade-off between depth of penetration (better with low-frequency antennas) and spatial
resolution (better with high-frequency antennas). This
trade-off can be partially overcome by modern systems that
Case Studies
Faulting
Some of the most stunning examples of GPR imaging of

shallow (o10 m) fault structure come from 2-D and 3-D
surveys of normal faults within the Taupo Rift of New Zealand
(Tronicke et al., 2006; McClymont et al., 2008, 2009a, 2009b)
(Figure 8). By imaging overlapping fault strands, off-fault
folding, rotations of hanging-wall and footwall blocks, and
along-strike variations in vertical fault displacements, GPR
data provide information that significantly improves estimates
of fault slip rates based solely on topographic mapping and
trenching studies (McClymont et al., 2009a, 2009b). By correlating GPR reflections with well-dated stratigraphic units
identified in a paleoseismic trench, it is possible to establish
short-term (o12.5 ka) variability in slip rates (McClymont
et al., 2009a). The high quality of the Taupo Rift surveys stems
from the presence of three strong and distinguishable subhorizontal marker horizons, high spatial resolution and extensive coverage (1525-cm line spacing over hundreds of
meters), and from the use of advanced processing techniques.
These techniques include appropriate trace-binning methods,
3-D migration that incorporates topography (Lehmann and
Green, 2000; Heincke et al., 2005), and careful corrections for
day-to-day differences in subsurface velocities. The sites potential was recognized prior to acquisition of the large 3-D
GPR data set through reconnaissance work, including a series
of short 2-D profiles and a small 3-D GPR data set (Tronicke
et al., 2006).
Because GPR cannot directly resolve steeply dipping features (Figure 7), faults must be recognized on the basis of
offsets of subhorizontal reflectors. Thus, GPR is not useful at
resolving fault features at sites that lack stratigraphic reflectors
or where relative slip is purely horizontal. Surveys are also
ineffective where weathering of surficial layers obscures
hanging wall/footwall contrasts, and where fault-scarp topography is not properly accounted for in image processing.
3.5.5.3.2
3.5.5.2
115
Volcanic processes
GPR is exceptionally effective at illuminating stratigraphy in

volcanic deposits before the onset of significant weathering
(Russell and Stasiuk, 1997; Cagnoli and Ulrych, 2001). GPR
profiles illuminate the internal structure of pyroclastic flows
(Gomez et al., 2009; Cassidy et al., 2009) and surge and fall
deposits (Kruse et al., 2010). Ettinger and Kruse (2007)
documented the effectiveness of GPR in mapping stratigraphy
of lahar flows.
3.5.5.3.3
GPR is used extensively for the location and identification of

sinkholes, particularly in covered karst where sinkholes may
not have clear topographic expressions (e.g., Kruse et al., 2006;
Gutierrez et al., 2009). In many settings, chemical weathering
processes will increase the conductivity of the substrate
through the development of clays, and thereby reduce the
efficacy of GPR. Beauvais et al. (2004) showed, however, that
116
Table 1
Summary of geomorphic targets accessible with various geophysical methods

Gravity
Magnetics
Resistivity, EM
GPR
Seismic reflection,
refraction, and surface
wave
Schematic
Figure 1
Figure 4
Figure 5
Figure 7
Figure 10
Faulting
Hanging wall/footwall
distinctions, Basinbounding faults, Figures
2 and 3
distinctions, Basinbounding faults, Figures
2 and 3
distinctions, fault
attitudes, low-resistivity
fault zones
Fault geometries, complex

slip patterns, Figure 8
Fault geometries, complex

slip patterns, Figure 11
Volcanics
Volcanic edifice
compensation mechanisms and erosion
patterns Identifying the
role of magma
movement in surface
deformation
Emplacement and
depositional processes
of volcanic units;
mapping debris
deposits; demagnetization associated with
thermal activity
Hydrothermal zones in
volcanoes
Internal structure of
pyroclastic flows and
lahars
Volcanic edifice
compensation mechanisms and erosion
patterns; internal
structure of offshore
volcanic deposits
Topography and
weathering
Causes of large-scale
surface dynamics;
subsidence processes
and causes; impact
structures
Soil formation, thickness,

and variability of
weathered horizons;
impact structures
Depth to bedrock;
sediment moisture;
weathering patterns
Figure 6
Slope processes
Subsurface structures that

drive vertical collapse
Landslide internal
structure, water content,
and contact with
substratum
Coastal processes and

sea-level change
Causes of local observed

sea-level changes
Coastal stratigraphy
Aeolian
Blakely et al. (2009);

Lopes Cardozo et al.
(2005); Peters et al.
(2005)
Quantifying erosion and

deposition, thickness
and variability of
weathered horizons
McCall (2009); Schrott

and Sass (2008);
Ezersky et al. (2009)
Granular flow Figure 9

talus thickness
Landslide structure,
offshore slumping and
landsliding
Bichler et al. (2004); Sass

(2006); Mauritsch et al.
(2000); Schrott and
Sass (2008)
Sea-level change history,

coastal evolution, barrier
island architecture
Offshore coastal
stratigraphy, coastal
evolution, sea-level
change, tidal flat
stratigraphy, Figures 12
and 13
Missiaen et al. (2008);

Gunn et al. (2006)
Fluvial and estuarine

stratigraphy and
sediment transport
Bowling et al. (2007)
Sub-glacial geology,
internal structure in
glaciers, glacier
thickness, stratigraphic
records of ice-sheet
extents
Maurer and Hauck (2007);

Vonder Muehll et al.
(2002); Schrott et al.
(2003a); Hubbard and
Glasser (2005)
Dune architecture and

migration
Fluvial
Glacial and periglacial
Mutiple methods;
comparisons
Map fluvial architectural

elements; grain size
distributions
Glacial isostatic
adjustment, subglacial
geology; permafrost
thickness
Subglacial geology
Permafrost and rock

glacier thickness,
distinguish ice-debris
mixtures from massive
ice
Rightmost column is a summary of comparative studies of various geophysical methods applied to geomorphology.
Fluvial architecture and

evolution
Feature/Process
175 E
Tau
p
rift o
Hav
Maleme
fault zone
duc
tion
zon
e
35 S
180 E
re t
rou
gh
Ker
m
ade
Hik
ura
c r id
ngi
ge
sub
Australian
plate
117
47 mm yr1
40 S
Pacific
plate
10
15
250
200
150
100
50 m
50
S2
50
y (m)
0
0
50
x (m)
9
Z (m)
10
11
12
13
14
Figure 8 Example of 3-D GPR surveying over an active fault. Perspective view of three cross sections extracted from a migrated 3-D groundpenetrating radar volume. Subtle variations of fault slip could not be distinguished without 3-D acquisition and migration. Modified from
McClymont, A.F., Villamor, P., Green, A.G., 2009b. Assessing the contribution of off-fault deformation to slip-rate estimates within the Taupo Rift,
New Zealand, using 3-D ground-penetrating radar surveying and trenching. Terra Nova 21(6), 446451.
in some places weathering horizons developed over crystalline

basement rocks can be mapped with both GPR and resistivity
methods.
3.5.5.3.4
Slope processes
Courtland et al. (2009) resolved grain avalanching packages

up to 3m thick and many tens of meters long in the highly
resistive deposits on Cerro Negro, Nicaragua, a small basaltic
cinder cone (Figure 9). Radar profiles exhibit clear evidence
of a cone built in large part by successive granular flows
(Figure 9(c)). During eruptions, growth of the cone becomes

oversteepened, resulting in grain avalanching, in turn producing reversely graded deposits. The dramatic change in
particle size at the boundary between successive events is believed to cause strong radar reflections (Figure 9).
Sass (2006, 2007) showed that GPR can be used to derive
talus thickness on alpine slopes. He showed that where a reflection from the debrisbedrock contact is absent, the contact
may nevertheless be recognized in a change in the texture of
the radar returns, referred to as a radar facies change. Sass and
118
0
20
40
60
Time (ns)
80
100
120
140
160
180
(a)
200
320
330
340
350
360
370
380
390
400
410
420
575
Caribbean
sea
570
565
Cerro negro
560
Len
Elev (m)
555
Pacific
ocean
550
100 km
545
540
535
530
525
520
320
(b)
1 vert exag
330
340
350
360
370
Distance (m)
380
390
400
410
420
(c)
Figure 9 Example of a GPR survey illustrating granular flow on the slope of the Cerro Negro cinder cone, Nicaragua (Courtland et al., 2009).
(a) GPR profile, no topography correction. (b) GPR profile, corrected for elevation. Horizon marked with red has geometry expected for base of a
granular flow package. (c) Schematic illustration of formation of package above red horizon. Data acquired with 100-MHz antennas.
Krautblatter (2007) used GPR-derived internal structure of

talus slopes to postulate models of talus development.
The more bouldery a deposit, the more chaotic the GPR
returns (e.g., Mills and Speece, 1997). It can be difficult to
make meaningful interpretations of GPR records from deposits with complex internal structure. For example, radar
surveys on pyroclastic flow deposits on Arenal Volcano, Costa
Rica, failed to resolve internal structure (Kruse, unpublished
data).
3.5.5.3.5
For several decades, GPR has been a useful technique in

studies of coastal stratigraphy landward of the swash zone
(Bristow and Jol, 2003; Baker and Jol, 2007). The strong EM
response of small variations in porosity and lithology in
coastal sands makes this method sensitive to stratigraphic
variability that may not be detected even by visual inspection
of cores. Penetration depths range up to 10 m or more. In
south Florida, 50-MHz GPR resolves discontinuity surfaces
that record subaerial exposure during sea-level lowstands, referred to as Q1Q5 in Florida stratigraphy (Kruse et al., 2000).
Studies on the beach system itself address beach ridge architecture (Clemmensen and Nielsen, 2010), barrier island and
barrier spit architecture (Jol et al., 2003; Nielsen et al., 2003),
barrier beach evolution (Bennett et al., 2009), coastal progradation (Moore et al., 2004; Bristow and Pucillo, 2006),
strand plain deposits (Tamura et al., 2008), relict beach scarps
(Buynevich et al., 2007), and overwash deposits (Wang and
Horwitz, 2007).
3.5.5.3.6
Aeolian processes
GPRs utility in aeolian studies is comparable to that in coastal

settings (Bristow, 2009). This method offers unparalleled access to the internal structure of a variety of types of dunes,
including linear (Bristow et al., 2000), crescentic (Derickson
et al., 2008), parabolic (Clemmensen et al., 2007), barchan
(Gomez-Ortiz et al., 2009), and reversing (Bristow et al.,
2010).
3.5.5.3.7
Fluvial processes
GPR is also used to identify the internal architecture of fluvial

deposits, including floodplain deposits (e.g., Leclerc and
Hickin, 1997). Pelpola and Hickin (2004) combined GPR
with aerial photography, bathymetry, and sediment sampling
to derive bed-load transport and fan-delta growth rates over a
52-year period on a steep drainage system. Grant et al. (2004)
used GPR to map buried fluvial channels and assess landscape
evolution.
3.5.5.3.8
GPR has been used for decades in glacial studies on scales

ranging from a few meters to hundreds of meters. It is widely
used as a tool to help calculate glacier volumes (Binder et al.,
2009; Fischer, 2009) and illuminate the internal structure of
glaciers (Fukui et al., 2008; Arcone and Kreutz, 2009; Degenhardt, 2009). These latter studies shed light on glacier-flow
dynamics and debris transport and depositional processes.
Studies relate the internal structure of glacial moraines
(Sadura et al., 2006; Burki et al., 2009) and eskers (Burke
et al., 2009) to their surface morphology. GPR studies have
119
also helped increase understanding of ice-stream dynamics

(e.g., Stokes et al., 2008) and active-layer thickness (e.g.,
Brosten et al., 2009). Brandt et al. (2009) went beyond glacial
stratigraphic interpretation and examined with both theory
and field data the causes of radar returns in firn with ice lenses.
3.5.6
Seismic Methods
Seismic methods, like GPR, are based on generating and recording waves that travel through the ground. As for GPR, any
subsurface heterogeneity associated with a change in the wave
velocity will scatter the waves. Seismic waves (elastic waves)
consist of oscillations of the ground material rather than the
GPR oscillations of EM fields and molecular charges. Hence
the seismic wave velocity is determined by completely different physical parameters than the radar wave, namely the bulk
modulus and shear modulus (measures of compressibility and
resistance to shear), and the density. Of relevance to nearsurface geophysics are the facts that seismic wave velocities are
low in loose dry sediments, higher in saturated sediments,
higher in sedimentary rock, and highest in basement igneous
and metamorphic rock.
Thus, despite the fundamental physical differences, both
GPR and seismic methods are responsive to lithologic variations and the water table. Seismic waves have typically much
longer wavelengths (several meters or more), so the spatial
resolution of seismic studies is lower than for GPR, and it is
difficult (as for GPR) to extract information from depths of
less than a few wavelengths. In practice, the two methods often
provide complementary data, GPR imaging shallowest strata
(uppermost B10 m) and seismic methods imaging deeper
(Sloan et al., 2007).
On land, seismic pulses at the surface generate three distinct wave types of oscillations: P, S, and surface waves. Each
mode travels with a different velocity (P fastest, surface
slowest) and can be analyzed and exploited, depending on the
survey targets. Unless otherwise stated, seismic surveys described here are based on analysis of P waves. Shear (S) waves
do not propagate in fluids and are insensitive to water saturation. S-wave studies can be useful where gases are present that
strongly attenuate P waves. Surface waves, traditionally a
source of noise in reflection surveys, have been the focus of
new method development in recent decades.
Seismic sources for marine use are designed to generate an
explosive or carefully modulated pulse while minimizing
secondary bubbles. Air guns, water guns, and boomer systems
create small explosions or implosions from/into cavities or
between plates. The dominant frequency of pulses used in
near-surface marine surveying (uppermost tens of meters)
ranges from tens of hertz to B4000 Hz. On land, practitioners
use hammer blows against plates or cylinders laid on the
surface, dropped or accelerated weights, or shotgun shells or
small explosives buried to shallow depths and detonated. The
sources offer peak frequencies in the range of 100400 Hz.
S-wave sources require a horizontal motion, such as a hammer
blow against a block held in place on the ground.
Seismic wave energy from a blow or shot that returns to the
surface is recorded on an array of receivers called hydrophones
(in water) or geophones (on land). The source and receiver
120
and significantly more data processing are required than for

refraction surveying. Reflection seismics are the backbone of
petroleum exploration where target depths are many hundreds
of meters to kilometers. Processing and acquisition techniques
need to be modified for near-surface studies, which require
high frequencies for good spatial analysis, careful separation
of early (shallow) reflections from coincident refracted and
surface wave arrivals, and careful corrections of near-surface
velocity variations and topography.
Near-surface marine reflection seismic surveys are relatively
straightforward in that both a high-frequency source and
hydrophones are towed by boat, and easily moved into desired survey positions. Reflections are easily distinguished in
marine environments, and surficiallateral velocity variations
and topography are not problems. However, gassy sediments
and hard bottom reflections (such as exposed limestone) can
severely limit penetration. Reshef et al. (2007) reported good
correlations between high-resolution seismic penetration
depths and grain size distributions.
Recent advances in land-survey reflection processing
now permit resolution of strata as shallow as a few meters
(Baker et al., 2000). However, data processing for shallow
(o30 m) land-based seismic reflection surveys is extensive
and less standardized than for other geophysical methods.
One recent advance that can improve shallow resolution is
migration that incorporates topography (Lehmann and Green,
2000). Examples of effective data acquisition, filtering, and
topographic corrections are given in the case studies cited
below.
Time
array are then moved to the next shot location, and the process
repeated. This is relatively straightforward in marine surveying,
where source and receiver array are towed behind the boat,
but is labor intensive on land. Geophones have to make good
contact with the ground to record motion; most involve
pushing a probe some 10 cm into the ground. In highresolution surveys, many geophones are laid out, and the shot
location is progressively moved through or around the array.
Information on near-surface features can be derived from
waves refracted from subhorizontal contacts, reflected waves,
or spectral analysis of surface waves (Figure 10). Surveys exploiting refracted waves have long been a staple of engineering
geology. Survey geometries are designed so that refracted energy is the first to arrive at geophone arrays, making arrival
times of refracted waves easy to identify. Only contacts in
which a lower layer has a higher wave velocity than the
overlying layer will refract energy back to an array of surface
geophones (Figure 10). Such contacts typically include the
water table and sedimentbasement boundary. Refraction
surveys suffer over sites with large lateral velocity variations,
with low-velocity layers, and with thin layers (Whiteley and
Eccleston, 2006). Increasingly sophisticated analysis techniques include refraction tomography techniques, which attempt to recover velocity structure throughout the strata above
the deepest refractor (Sheehan et al., 2005).
Reflection seismic surveys require analysis of all returns
recorded by geophones or hydrophones. These typically provide a much fuller picture of the subsurface, and to greater
depths than refraction surveys. However, many more shots
Figure 10 Schematic illustration of the first arrivals recorded at geophones in a survey designed for analysis of refracted energy. The seismic
source (hammer blow) generates a number of waves that will be recorded by the geophones: direct-body waves, direct-air waves, refracted-body
waves, reflected-body waves, and surface waves. Only the first wave to arrive at each geophone is shown. The nearest geophones record directbody waves as a first arrival (red line, red pulses). Farther geophones record refracted waves (head waves) as a first arrival (blue lines, blue
pulses). Wave energy is refracted (blue lines) when the velocity of the medium underlying the contact is higher than the overlying medium.
Velocities of the media above and below the contact and contact depth and dip can be reconstructed from arrival times from records of sources
(hammer blows) at both ends of the array of geophones. Information on a point-like object and steeply dipping contacts is unlikely to be
recovered from seismic-refraction surveys. Surveys designed for analysis of reflected energy will involve geometries similar to the GPR
schematic in Figure 7 and can capture more information, but require hammer blows (or other source) at regular intervals, close to the geophone
spacing, along the line.

3.5.6.2
Surface-wave studies can be conducted with geophones

sensitive to low frequencies (B1 Hz) connected to conventional near-surface seismic systems and sources. As lower
frequency waves permeate to greater depths, the dispersion of
the surface waves can be used to estimate the S-wave velocity
structure in the uppermost few tens of meters. The two primary methods of data analysis, spectral analysis of surface
waves (SASW) and multichannel analysis of surface waves
(MASW), are well summarized by Pelton (2005). These
methods are useful for mapping soil, imaging karst structure,
and identifying liquefaction potential.
3.5.6.1
3.5.6.3
3.5.6.3.1
35 S
175 E
Case Studies
Faulting
Seismic reflection imaging of strata offset at faults has been a

standard method of recovering fault displacement and
geometry for many decades. Near-surface capabilities are
highlighted here with a high-quality case study of the transpressive Alpine Fault Zone, New Zealand (Kaiser et al., 2011)
(Figure 11). The imaging of a Pleistocene erosional basement
surface horizon provides the means to estimate late Quaternary
slip rates on a steeply dipping dominant fault strand (AF in
Figure 11). Although only a single fault scarp is visible at the
surface, deformation is inferred to extend over a 60-m-wide
Reflection seismic methods can produce high-resolution images of subsurface stratigraphy over depths ranging from a few
meters to kilometers. The method is unparalleled in its resolution over scales of hundreds of meters to kilometers (where
development has been fostered by the petroleum industry).
Efficient methods for rapid acquisition of near-surface data are
under development (e.g., Tsoflias et al., 2006).
170 E
Limitations
Seismic methods are best suited to capture subhorizontal

strata. Land-based reflection methods tend to be labor intensive in terms of data acquisition and processing. Processing
techniques needed to extract optimal images from reflection
data are sufficiently complex that they may not be accessible to
casual practitioners.
Strengths
165 E
121
180 E
Australian
plate
on
lt z
40 S
e
pin
fau
Al
45 S
SE
Late PleistoceneHolocene
postglacial sediments
NW
AF
Moraine
Schist
MFS
m
38 m
yr
Marble
Pacific
plate
Extent
of seismic data
50 S
(a)
(b)
ine )
ssl
Cro nce (m
50 100
ta
dis
100
200
Inlin
distan e
ce (m
ine ) NW
ssl
Cro nce (m 50 100
a
t
dis
100
SF1
300
400
200
Inli
distan ne
ce (m
300
400
SE
100
200
AF
(c)
Depth (m)
100
Depth (m)
SF2
200
(d)
Figure 11 High-resolution 3-D land-based seismic reflection imaging of the extent of deformation structures within the shallow expression of
active faults in the transpressive Alpine Fault Zone, New Zealand. (a) Location map. (b) Local geology. The authors image a steeply dipping
dominant fault strand (AF) that significantly offsets the late Pleistocene erosional basement surface. (c) Seismic reflection image. (d) Interpreted
seismic reflection image. Although only a single fault scarp was visible at the surface, deformation was inferred to extend over a 60-m-wide zone
within the fault zone, encompassing two minor fault strands (SF1 and SF2) that were first imaged in shallow GPR studies (McClymont et al.,
2009). Modified from Kaiser, A.E., Horstmeyer, H., Green, A.G., Campbell, F.M., Langridge, R.M., McClymont, A.F., 2011. Detailed images of the
shallow Alpine Fault Zone, New Zealand, determined from narrow-azimuth 3D seismic reflection data. Geophysics 76, B19B32, with permission
from AGU and A.E. Kaiser (personal communication).
122
zone within the fault zone, encompassing two minor fault

strands (SF1 and SF2 in Figure 11) that were first imaged in
shallow GPR studies (McClymont et al., 2010). The detailed
seismic imaging of strata truncations, rotations, and tilting was
achieved through dense data coverage and a carefully tailored
processing scheme (Kaiser et al., 2011). Of particular importance in processing were corrections for the severe static shifts
and strong source-generated noise that typically complicate
shallow seismic data. The imaging of complex dipping as well
as diffracted and crossing events required careful velocity analyses, dip-moveout corrections, and 3-D migration.
3.5.6.3.2
Volcanic processes
Seismic reflection profiling is useful in studying deposits of

near-surface volcanic structures when they lie under water, on
lakes (e.g., Johnson et al., 2003) or offshore (e.g., Morgan
et al., 2003a).
3.5.6.3.3
Over tectonic scales (kilometers to hundreds of kilometers), reflection seismic profiles can be used to reconstruct uplift and
regional topographic evolution (e.g., Regard et al., 2009). Refraction surveys can be used to quantify rates of erosion and
deposition. For instance, Hunt and Wu (2004) used refraction
surveys in combination with soil sampling and field observations to estimate Holocene rates of soil erosion on a small hill.
Schrott et al. (2003b) established the volume of valley fill deposits in an alpine basin with an array of refraction surveys.
Refraction surveys with engineering targets yield data on lithologic parameters such as porosity that may be related to degrees
of weathering (e.g., Gay et al., 2006). In addition, both refraction
(e.g., Whiteley and Eccleston, 2006) and reflection methods (e.g.,
Roberts et al., 1992) can resolve depth to basement in a variety of
geologic settings. Surface-wave methods of mapping soil development properties and voids have been, and are being, tested in
a variety of settings (e.g., Miller et al., 1999; Xu and Butt, 2006).
3.5.6.3.4
Slope processes
Marine seismic-reflection surveys can reveal the depositional

record of offshore slumping and landsliding (e.g., Morgan et al.,
2003a) and mass transport in general on passive and active
continental margins (e.g., Reis et al., 2010; Pedley et al., 2010).
In a study that combined both onshore and offshore seismic
surveys, Stucchi and Mazzotti (2009) mapped the extent of a
landslide across the coastline. Their work provides an excellent
illustration of the challenges involved in near-surface seismic
investigations in steep topography and noisy environments.
3.5.6.3.5
Reflection seismic data are readily collected in the shallow

coastal marine environment using boomer seismic sources
with center frequencies of B100010 000 Hz, yielding vertical
resolution (as discussed for GPR) on the order of 20 cm to
2 m. Transitions between gravels, sands, muds, and limestone
tend to produce good seismic reflections. Unconformities,
fluvial valleys, and progradation and aggradation sequences are
commonly recognized in such data sets, permitting reconstructions of coastal evolution. Such data are also valuable
for understanding the degree to which heterogeneities in the
geologic substrate control Quaternary coastal evolution.
Because coastal stratigraphy is typically subhorizontal, regionally extensive strata can be correlated between data gaps, and
data are commonly collected over grids. Mallinson et al. (2010)
combined seismic stratigraphy with amino acid racemization
and strontium-isotope age data from cores to determine the
Quaternary stratigraphic framework and evolution of the coast
near Cape Hatteras, USA (Figures 12 and 13). In such coastal
studies, the distances required between crossover points are
determined by geologic complexity and resources available.
Coring at sites close to seismic lines is critical for data interpretation (Figure 12). In these environments, data processing
is typically limited to band-pass filtering and time-to-depth
conversions done with a constant velocity assumption. More
recently, surface-wave methods have been applied to noninvasively map tidal-flat stratigraphy (Watabe and Sassa, 2008).
3.5.6.3.6
Fluvial processes
Stratigraphy mapped with shallow-marine seismic methods

can be used to trace morphological evolution of fluvial and
estuarine environments (e.g., Ginsberg et al., 2009; Wang
et al., 2007) and estimate fluvial sediment-transport volumes
(e.g., Liu et al., 2007).
3.5.6.3.7
Refraction and reflection seismics can be used to assess subglacial geology (Ahmad et al., 2009) and infer variations in
frictional resistance to ice flow (Anandakrishnan et al., 1998).
Musil et al. (2002) showed that while seismic surveying in an
alpine rock glacier failed to yield information on reflecting
horizons, tomographic inversion of first-arrival times did yield
information on both the internal structure and the base of the
rock glacier. Baker et al. (2003) found strong seismic reflections from contacts between clean ice and underlying
debris-rich ice, as well as the glacier bottom. Wellner et al.
(2006) combined marine seismic lines with multibeam bathymetry and coring to document the geomorphological features and seaward extent of the Antarctic ice sheet during the
Last Glacial Maximum. Andreassen et al. (2008) similarly reconstructed geomorphic features left by former ice sheets
using petroleum industry 3-D reflection seismic surveys. Refraction methods, including refraction tomography, are useful
for mapping top of permafrost and heterogeneities within
permafrost (Musil et al., 2002; Maurer and Hauck, 2007).
3.5.7
Combining Geophysical Methods
Many of the aforementioned studies incorporated coincident

data acquisition using multiple geophysical techniques. The incremental cost of acquiring a second type of geophysical data at
a survey site can be relatively small, because the initial survey
positioning, via GPS or taped grids, can be reused. There
can also be significant advantages associated with the analysis of
data acquired from different methods. For example, if a
resistivity survey shows strata of contrasting resistivity, these
differences could be due to variations in degree of saturation,
pore-water conductivity, or lithology. A seismic-refraction
survey could help distinguish between these possibilities. If a
refracting horizon is observed coincident with a resistivity
123
76450 W 76300 W 76150 W 7600 W 75450 W 75300 W

36300 N
Atlantic
Ocean
36150 N
3600 N
35450 N
35300 N
35150 N
3500 N
W
eastct
Value
34450 N
N
0 5 10
20
30
40
High: 0
0 5 10
Low: 20
E
S
20
30
40
km
50
km
(a)
(b)
Figure 12 High-resolution shallow-marine seismic reflection imaging conducted as part of a study of the stratigraphy and geologic controls on
Quaternary evolution of the Cape Hatteras region, US. (a) Map showing bathymetry of the study area. Depth (gray shading) is meters below
mean sea level. (b) Map showing core locations (black dots) and seismic data transects (black lines) used in this investigation. Cores referred to
in this paper are labeled. Because strata are laterally continuous and subhorizontal, a loose grid of 2-D profiles is sufficient to reconstruct
stratigraphy. (Compare with Figures 8 and 11, where careful resolution of fault structure required full 3-D surveys.) Reproduced from Mallinson,
D.J., Culver, S.J., Riggs, S.R., et al., 2010. Regional seismic stratigraphy and controls on the Quaternary evolution of the Cape Hatteras region of
the Atlantic passive margin, USA. Marine Geology 268, 1633.
gradient, then the resistivity gradient is better explained by a

water table or lithologic boundary than a pore-water conductivity gradient.
Multiple methods can add value to a survey in different
ways. Gravity and magnetic data offer distinct physical information on similar spatial scales. The same is typically true for
resistivity and seismic refraction methods. On the other hand,
GPR and reflection-seismic surveys offer complementary
shallow and deep depths of investigations, respectively. A
combination of resistivity and EM methods offers different
spatial sensitivities of the same physical property, with more
labor-intensive resistivity profiles used to extract local resistivity structure, whereas lateral mapping with more rapid EM
methods can cover areas between profiles. Studies that compare and contrast individual methods at a given site are listed
in the rightmost column of Table 1.
3.5.8
3.5.8.1
Discussion and Conclusions

Practical Considerations
In geophysical studies, the importance of collecting in situ field

observations and reference data cannot be overstated. The best
way to define success of a geophysical method is by examining

results near outcrops or cores. If a geophysical method captures key parameters or features of interest, its utility lies in
being able to extrapolate the geometry of such conditions
away from calibration points. Where the collection of additional reference data is simply impossible, comparisons of
multiple geophysical methods should clarify interpretation of
survey results, and any interpretations must be based on forward or inverse modeling of the data and comparison of recovered geophysical parameters (velocities and resistivity)
with published values for expected lithologies.
Investigators commonly have the capability of acquiring
and processing 3-D data, but have to decide whether the gains
from the significant additional time expense are critical to the
study. For example, the McClymont et al. (2009a, 2009b,
2010) GPR example and the Kaiser et al. (2009, 2011) seismic
surveys resolved structural complexities that would have been
mis-estimated or mis-identified without fully 3-D acquisition
and processing. By contrast, the Braun et al. (2009) resistivity
study and the Mallinson et al. (2010) seismic-reflection study
achieved their research goals with a loose network of 2-D
profiles. Reconnaissance surveys often help clarify data density
needs, as well as giving a practitioner a realistic estimate of
data acquisition rates.
124
Two-way travel time (s)
North
South
L1F2
0.02
L1F2
L1F19
0.04
OBX-17
0.06
Cape
Hatteras
0.08
OBX-17
G S M
Pleistocene Holocene
Meters below sea level
0
Sed/water interface (Q100)
SSU VII
20
Q99
SSU V
40
Open shelf
Q30
Upper shoreface sand
SSU IV
60
Lower shoreface sand
Q20
500
1000
1.4 km
(a)
L1F19
Two-way travel time (s)
West
Eest
0.04
0.05
0.06
0.07
0.08
1 km
40
Q30
TST
Flooding surface
Q30b
FSST
Shoreface sands
50
60
70
Inner shelf and estuarine

Q20
LST
Subaerial unconformity
NR
FR
Conformity
Deepening
Meters below sea level
30
Outer shelf
TST
Inner shelf
1 km
(b)
Figure 13 Selected profiles from the Mallinson et al. (2010) study of the Cape Hatteras Quaternary coastal evolution (location shown in
Figure 12). (a) High-resolution single-channel seismic profile L1F2 (see inset map for location) shows prograding clinoforms characteristic of a
regressive shoreface. Two-way travel time (twtt) and meters below sea level (mbsl) based on an estimated seismic velocity of 1600 m s1
(calculated by correlations to core data). (b) Seismic data with interpretation from L1F19_03 showing the interpretation of the falling stage
systems tract (FSST). Map labels represent normal regression (NR), forced regression (FR) and transgressive systems tract (TST). This study
was conducted with a 175 J boomer source with center frequency B1500-Hz, band-pass filtering 5002000 Hz, 0.083-ms sampling frequency,
and a 0.5-s shot rate yielding B90-cm sampling interval. Reproduced from Mallinson, D.J., Culver, S.J., Riggs, S.R., et al., 2010. Regional
seismic stratigraphy and controls on the Quaternary evolution of the Cape Hatteras region of the Atlantic passive margin, USA. Marine Geology
268, 1633.

3.5.8.2
Integrating Geophysics, Remote Sensing, and

Geographic Information Systems
The availability of high-resolution digital elevation models

(DEMs) has led to an emphasis on analysis of topography in
geomorphology studies. Yet, this chapter demonstrates that
geophysical investigations of the subsurface fundamentally
contributes to our understanding of geomorphological processes. Clearly, better integration of subsurface and surface
data will advance the field. But better synthesis of geophysical,
remote sensing, and geographic information system GIS data
will require data acquisition at overlapping scales, and further
development of 3-D and temporal representational schemes.
Remotely sensed data are generally acquired over larger
spatial scales, with relatively low-to-moderate spatial resolution, compared to higher-resolution near-surface geophysical studies. Such scale variations are being bridged by
advances in imaging sensor characteristics and in geophysical
methodologies. In particular, the development of terrestrial
laser scanning (TLS) brings the measurement scale of DEMs
down to the meter or centimeter scale of high-resolution nearsurface geophysical studies. At the same time, the spatial scale
of geophysical studies can be expanded through the use of
airborne instruments for gravity, magnetic, and EM methods.
Airborne combinations of remote sensing and geophysical
instrumentation such as thermal imagery and magnetic may
prove fruitful. Ground-based geophysical acquisition systems
are also becoming more efficient, making data sets over scales
of hundreds of meters to kilometers more affordable. Advances include capacitively coupled resistivity systems (e.g.,
Hickin et al., 2009), automated geophone installation systems
(Tsoflias et al., 2006), and side-by-side mounting of multiple
ground-penetrating radar antennas.
The temporal scale is very important in geomorphological
research, and studies should involve simultaneous surfaceimaging and geophysical surveying. For example, Carrivick et al.
(2010) combined hydrological measurements acquired over a
period of airborne LiDAR surveys of lahars to model lahar
transport dynamics; pre- and postevent geophysical surveys
would add another level of control on the transport models.
Time-lapse remote-sensing studies synchronized with repeated
geophysical surveys can be productive for understanding geomorphic processes in a variety of rapidly evolving environments. In coastal environments, airborne imagery coupled
with near-shore seismic or onshore ground-penetrating radar
profiling would illuminate sediment-transport processes.
Similarly, on landslides and debris flows, further studies
linking remote sensing to resistivity or ground-penetrating
radar profiles will help connect surficial and subsurface deformation mechanisms.
Better integration of geophysical, remote sensing, and GIS
data is at present largely limited by costs. In addition to the
acquisition issues described above, significant resources can be
required for data processing, visualization, and interpretation.
Many commonly used GIS software platforms do not readily
accommodate subsurface data. Software packages for 4-D
(space and time) visualization of data can be expensive, even
as computer-processing costs shrink. Large data volumes are
most easily understood with specialized visualization tools
such as stereo projectors and wall screens. However, as use of
125
these software and hardware tools becomes more widespread,

their relative costs should fall.
References
Ahmad, J., Schmitt, D.R., Rokosh, C.D., Pawlowicz, J.G., 2009. High-resolution
seismic and resistivity profiling of a buried Quaternary subglacial valley;
northern Alberta, Canada. Geological Society of America Bulletin 121(1112),
15701583.
Anandakrishnan, S., Blankenship, D.D., Alley, R.B., Stoffa, P.L., 1998. Influence of
subglacial geology on the position of a West Antarctic ice stream from seismic
observations. Nature 394, 6265.
Andreassen, K., Laberg, J.S., Vorren, T.O., 2008. Sea-floor geomorphology of the
SW Barents Sea and its glaci-dynamic implications. Geomorphology 97(12),
157177.
Arcone, S.A., Kreutz, K., 2009. GPR reflection profiles of Clark and Commonwealth
Glaciers, Dry Valleys, Antarctica. Annals of Glaciology 50(51), 121129.
Baker, G.S., Jol, H.M., 2007. Stratigraphic Analyses Using GPR. The Geological
Society of America, Special Paper, 432 pp.
Baker, G.S., Steeples, D.W., Schmeissner, C., Spikes, K.T., 2000. Ultrashallow
seismic reflection monitoring of seasonal fluctuations in the water table.
Environmental and Engineering Geoscience 6(3), 271277.
Baker, G.S., Strasser, J.C., Evenson, E.B., Lawson, D.E., Pyke, K., Bigl, R.A., 2003.
Near-surface seismic reflection profiling of the Matanuska Glacier; Alaska.
Geophysics 68(1), 147156.
Barbiero, L., Parate, H.R., Descloitres, M., et al., 2007. Using a structural approach
to identify relationships between soil and erosion in a non-anthropic forested
area, South India. Catena 70, 313329.
Barde-Cabusson, S., Finizola, A., Revil, A., et al., 2009. New geological insights and
structural control on fluid circulation in La Fossa Cone (Vulcano, Aeolian Islands,
Italy). Journal of Volcanology and Geothermal Research 185(3), 231245.
Battaglia, M., Gottsmann, J., Carbone, D., Fernandez, J., 2008. 4D volcano
gravimetry. Geophysics 73(6), WA3WA18.
Beauvais, A., Ritz, M., Parisot, J.-C., Bantsimba, C., Dukhan, M., 2004. Combined
ERT and GPR methods for investigating two-stepped lateritic weathering systems.
Geoderma 119(12), 121132.
Bell, R.E., Blankenship, D.D., Finn, C.A., Morse, D.L., Scambos, T.A., Brozenak,
J.M., Hodge, S.M., 1998. Influence of subglacial geology on the onset of a
West Antarctic ice stream from aerogeophysical observations. Nature 394,
5862.
Bennett, M.R., Cassidy, N.J., Pile, J., 2009. Internal structure of a barrier beach as
revealed by ground penetrating radar (GPR): Chesil beach, UK. Geomorphology
104, 218229.
Bersezio, R., Giudici, M., Mele, M., 2007. Combining sedimentological and
geophysical data for high-resolution 3-D mapping of fluvial architectural
elements in the Quaternary Po Plain (Italy). Sedimentary Geology 202(12),
230248.
Bichler, A., Bobrowsky, P., Best, M., Douma, M., Hunter, J., Calvert, T., Burns, R.,
2004. Three-dimensional mapping of a landslide using a multi-geophysical
approach: the Quesnel Forks landslide. Landslides 1, 2940.
Binder, D., Brueckl, E., Roch, K.H., Behm, M., Schoener, W., Hynek, B., 2009.
Determination of total ice volume and ice-thickness distribution of two glaciers
in the Hohe Tauern region, Eastern Alps, from GPR data. Annals of Glaciology
50(51), 7179.
Blakely, R.J., Sherrod, B.L., Hughes, J.F., Anderson, M.L., Wells, R.E., Weaver, C.S.,
2009. Saddle Mountain fault deformation zone, Olympic Peninsula, Washington:
western boundary of the Seattle uplift. Geosphere 5(2), 105125.
Bowling, J.C., Harry, D.L., Rodriguez, A.B., Zheng, C., 2007. Integrated geophysical
and geological investigation of a heterogeneous fluvial aquifer in Columbus
Mississippi. Journal of Applied Geophysics 62(1), 5873.
Brandt, O., Langley, K., Giannopoulos, A., Svein, E., Kohler, J., 2009. Radar
response of firn exposed to seasonal percolation, validation using cores and
FDTD modeling. IEEE Transactions on Geoscience and Remote Sensing 47(8),
27732786.
Branston, M.W., Styles, P., 2003. The application of time-lapse microgravity for
the investigation and monitoring of subsidence at Northwich, Cheshire.
Quarterly Journal of Engineering Geology and Hydrogeology 36(3),
231244.
Braun, J.J., Descloitres, M., Riotte, J., et al., 2009. Regolith mass balance inferred
from combined mineralogical, geochemical and geophysical studies; Mule Hole
126
gneissic watershed, south India. Geochimica et Cosmochimica Acta 73(4),

935961.
Bristow, C., 2009. Ground penetrating radar in Aeolian dune sands. In: Jol, H.M.
(Ed.), Ground Penetrating Radar Theory and Applications Elsevier, Amsterdam,
The Netherlands 544 pp.
Bristow, C.S., Augustinus, P.C., Wallis, I.C., Jol, H.M., Rhodes, E.J., 2010.
Investigation of the age and migration of reversing dunes in Antarctica using
GPR and OSL, with implications for GPR on Mars. Earth and Planetary Science
Letters 289(12), 3042.
Bristow, C.S., Bailey, S.D., Lancaster, N., 2000. The sedimentary structures of linear
dunes. Nature 406, 5659.
Bristow, C.S., Jol, H.M. (Eds.), 2003. Special Publications, London.
Bristow, C.S., Pucillo, K., 2006. Quantifying rates of coastal progradation from
sediment volume using GPR and OSL, the Holocene fill of Guichen Bay, southeast South Australia. Sedimentology 53, 769788.
Brosten, T.R., Bradford, J.H., McNamara, J.P., Gooseff, M.N., Zarnetske, J.P.,
Bowden, W.B., Johnston, M.E., 2009. Estimating 3D variation in active-layer
thickness beneath Arctic streams using ground-penetrating radar. Journal of
Hydrology 373(34), 479486.
Brown, G.C., Rymer, H., Thorpe, R.S., 1987. The evolution of andesite volcano
structures: new evidence from gravity studies in Costa Rica. Earth and Planetary
Science Letters 82, 323334.
Budetta, G., Carbone, D., Greco, F., 1999. Subsurface mass redistributions at Mount
Etna (Italy) during the 19951996 explosive activity detected by microgravity
studies. Geophysical Journal International 138(1), 7788.
Burke, M.J., Woodward, J., Russell, A.J., Fleisher, P.J., 2009. Structural controls on
englacial esker sedimentation; Skeidararjokull, Iceland. Annals of Glaciology
50(51), 8592.
Burki, V., Larsen, E., Fredin, O., Margreth, A., 2009. The formation of sawtooth
moraine ridges in Bodalen, western Norway. Geomorphology 105(34),
182192.
Buynevich, I.V., FitzGerald, D.M., Goble, R.J., 2007. A 1500 yr record of North
Atlantic storm activity based on optically dated relict beach scarps. Geology
(Boulder) 35(6), 543546.
Cagnoli, B., Ulrych, T., 2001. Ground penetrating radar images of unexposed climbing dune-forms in the Ubahebe hydrovolcanic field (Death
Valley, California). Journal of Volcanology and Geothermal Research 109,
279298.
Calhoun, K., Long, L.T., Schmitt, T.J., 1989. Sinkholes detected by microgravity
survey. Southeastern Geology 30(2), 109119.
Carbone, D., Budetta, G., Greco, F., 2003. Bulk processes prior to the 2001 Mount
Etna eruption highlighted through microgravity studies. Journal Of Geophysical
Research 108(B12), 2556. http://dx.doi.org/10.1029/2003JB002542.
Carrivick, J.L., Manville, V., Graettinger, A., Cronin, S.J., 2010. Coupled fluid
dynamics-sediment transport modeling of a Crater Lake break-out lahar: Mt.
Raupehu, New Zealand. Journal of Hydrology 388, 399413.
Cassidy, N.J., Calder, E.S., Pavez, A., Wooller, L., 2009. GPR-derived facies
architectures; a new perspective on mapping pyroclastic flow deposits.
In: Thordarson, T., Self, S., Larsen, G., Rowland, S.K., Hoskuldsson, A. (Eds.),
Studies in Volcanology; The Legacy of George, Walker. Special Publications of
the International Association of Volcanology and Chemistry of the Earths Interior
Geological Society for IAVCEI, London, vol. 2, pp. 181210.
Clemmensen, L.B., Bjornsen, M., Murray, A., Pedersen, K., 2007. Formation of
aeolian dunes on Anholt, Denmark since AD 1560: a record of deforestation and
increased storminess. Sedimentary Geology 199, 171187.
Clemmensen, L.B., Nielsen, L., 2010. Internal architecture of a raised beach ridge
system (Anholt, Denmark) resolved by ground-penetrating radar investigations.
Sedimentary Geology 223, 281290.
Courtland, L.M., Kruse, S.E., Connor, C., 2009. Into the cone: a ground penetrating
radar investigation of near vent processes at Cerro Nego volcano. Eos
Transactions of AGU 90 (52). Fall Meeting Supplement, abstract V21C-2013,
Day-Lewis, F.D., White, E.A., Johnson, C.D., Lane, Jr. J.W., Belaval, M., 2006.
Continuous resistivity profiling to delineate submarine groundwater discharge:
examples and limitations. Leading Edge 25(6), 727728.
Degenhardt, J.J., 2009. Development of tongue-shaped and multilobate rock
glaciers in alpine environments; interpretations from ground-penetrating radar
surveys. Geomorphology 109(34), 94107.
Dentith, M.C., Bevan, A.W.R., Backhouse, J., Featherstone, W.E., Koeberl, C., 1999.
Yallalie: a buried structure of possible impact origin in the Perth Basin, Western
Australia. Geological Magazine 136, 619632.
Derickson, D., Kocurek, G., Ewing, R.C., Bristow, C., 2008. Origin of a complex and
spatially diverse dune field pattern, Algodones, southeastern California.
Diaferia, I., Barchi, M., Loddo, M., Schiavone, D., Siniscalchi, A., 2006.
Detailed imaging of tectonic structures by multiscale Earth resistivity
tomographies: the Colfiorito normal faults (central Italy). Geophysical Research
Letters 33, L09305.
Diez, M., Connor, C.B., Kruse, S.E., Connor, L., Savov, I.P., 2009. Evidence of
small-volume igneous diapirism in the shallow crust of the Colorado Plateau,
San Rafael desert, Utah. Lithosphere 1, 328336.
Dimalanta, C.B., Ramos, E.G.L., Yumul, Jr. G.P., Bellon, H.N., 2009. New features
from the Romblon Island Group; key to understanding the arc-continent
collision in central Philippines. Tectonophysics 479(12), 120129.
Doolittle, J.A., Minzenmayer, F.E., Waltman, S.W., Benham, E.C., Tuttle, J.W.,
Peaslee, S.D., 2007. Ground-penetrating radar soil suitability map of the
conterminous United States. Geoderma 141, 416421.
Ettinger, S., Kruse, S., 2007. Using ground penetrating radar to help delineate lahar
hazard zones at Cotopaxi Volcano, Ecuador. EOS Transactions of AGU 88(52),
Fall Meeting Supplement, abstract V11D-0815.
Evans, M.E., Heller, F., 2003. Environmental Magnetism. Academic Press,
SanDiego, CA.
Ezersky, M., Legchenko, A., Camerlynck, C., Al-zoubi, A., 2009. Identification of
sinkhole development mechanism based on a combined geophysical study in
Nahal Hever South area (Dead Sea coast of Israel). Environmental Geology 58,
11231141.
Ferraccioli, F., Armadillo, E., Jordan, T., Bozzo, E., Corr, H.F.J., 2009. Aeromagnetic
exploration over the East Antarctic ice sheet; a new view of the Wilkes
subglacial basin. Tectonophysics 478(12), 6277.
Finizola, A., Aubert, M., Revil, A., Schutze, C., Sortino, F., 2009. Importance
of structural history in the summit area of Stromboli during the 2002
2003 eruptive crisis inferred from temperature, soil CO2, self-potential, and
electrical resistivity tomography. Journal of Volcanology and Geothermal
Research 183(34), 213227.
Finn, C.A., Deszcz-Pan, M., Anderson, E.D., John, D.A., 2007. Three-dimensional
geophysical mapping of rock alteration and water content at Mount Adams,
Washington; implications for lahar hazards. Journal of Geophysical Research
112, B10204.
Fischer, A., 2009. Calculation of glacier volume from sparse ice-thickness data,
applied to Schaufelferner, Austria. Journal of Glaciology 55(191), 453460.
Forte, A.M., Moucha, R., Simmons, N.A., Grand, S.P., Mitrovica, J.X., 2010. Deep
mantle contributions to the surface dynamics of the North American continent.
Tectonophysics 481(14), 315.
Froese, D.G., Smith, D.G., Clement, D.T., 2005. Characterizing large river history
with shallow geophysics; middle Yukon River, Yukon Territory and Alaska.
Fukui, K., Sone, T., Strelin, J.A., Cesar, A., Mori, J., Fujii, Y., 2008. Dynamics and
GPR stratigraphy of a polar rock glacier on James Ross Island, Antarctic
Peninsula. Journal of Glaciology 54(186), 445451.
Gay, D.A., Morgan, F.D., Sogade, J.A., Vichabian, Y., Reppert, P., Wharton, A.E.,
2006. Investigations of andesitic volcanic debris terrains; Part I, Geophysical.
Geophysics 71(1), B1B8.
Ginsberg, S.S., Aliotta, S., Lizasoain, Guillermo, O., 2009. Morphodynamics and
seismostratigraphy of a deep hole at tidal channel confluence. Geomorphology
104(34), 253261.
Gomez, C., Lavigne, F., Lespinasse, N., Hadmoko, D.S., Wassmer, P., 2009.
Longitudinal structure of pyroclastic-flow deposits, revealed by GPR survey, at
Merapi Volcano, Java, Indonesia. Journal of Volcanology and Geothermal
Research 174(4), 439447.
Gomez-Ortiz, D., Martin-Crespo, T., Rodriguez, I., Sanchez, M.J., Montoya, L.,
2009. The internal structure of modern barchan dunes of the Ebro River delta
(Spain) from ground penetrating radar. Journal of Applied Geophysics 69(2),
159170.
Gottsmann, J., Camacho, A.G., Marti, J., Wooller, L., Fernandez, J., Garcia, A.,
Rymer, H., 2008. Shallow structure beneath the Central Volcanic Complex of
Tenerife from new gravity data: implications for its evolution and recent
reactivation. Physics of the Earth and Planetary Interiors 168(34), 212230.
Grant, J.A., Maxwell, T.A., Johnston, A.K., Kilani, A., Williams, K.K., 2004.
Documenting drainage evolution in Bir Kiseiba, southern Egypt: constraints from
ground-penetrating radar and implications for Mars. Journal of Geophysical
Research 109, EO9002.
Grote, K., Hubbard, S., Rubin, Y., 2003. Field-scale estimation of volumetric water
content using ground-penetrating radar ground wave techniques. Water
Resources Research 39, 13211333.
Guha, S., Kruse, S.E., Wright, E.E., Kruse, U.E., 2005. Spectral analysis of ground
penetrating radar response to thin sedimentary layers. Geophysical Research
Letters 32, L23304. http://dx.doi.org/10.1029/2005GL023933.
Gunn, D.A., Pearson, S.G., Chambers, J.E., et al., 2006. An evaluation of combined
geophysical and geotechnical methods to characterize beach thickness. Quarterly
Journal of Engineering Geology and Hydrogeology 39(4), 339355.
Gutierrez, F., Galve, J.P., Lucha, P., Bonachea, J., Jorda, L., Jorda, R., 2009.
Investigation of a large collapse sinkhole affecting a multi-storey building by
means of geophysics and the trenching Technique (Zaragoza city, NE Spain).
Environmental Geology 58(5), 11071122.
Hauck, C., Vonder Muhll, D., Maurer, H., 2003. Using DC resistivity tomography to
detect and characterize mountain permafrost. Geophysical Prospecting 51,
273284.
Heincke, B., Green, A.G., van der Kruk, J., Horstmeyer, H., 2005. Acquisition and
processing strategies for 3 D georadar surveying a region characterized by
rugged topography. Geophysics 70(6), K53K61.
Herrero-Bervera, E., Canon-Tapia, E., Walker, G.P.L., Guerrero-Garcia, J.C., 2002.
The Nuuanu and Wailau giant landslides; insights from paleomagnetic and
anisotropy of magnetic susceptibility (AMS) studies. Physics of the Earth and
Planetary Interiors 129(1-2), 8398.
Hickin, A.S., Kerr, B., Barchyn, T.E., Paulen, R.C., 2009. Using ground-penetrating
radar and capacitively coupled resistivity to investigate 3-D fluvial architecture
and grain-size distribution of a gravel floodplain in northeast British Columbia,
Canada. Journal of Sedimentary Research 79(6), 457477.
Hubbard, B., Glasser, N., 2005. Field Techniques in Glaciology and Glacial
Geomorphology. Wiley, England, 400 pp.
Hsu, H.L., Yanites, B.J., Chen, C.C., Chen, Y.-G., 2010. Bedrock detection using 2D
electrical resistivity imaging along the Peikang River, central Taiwan.
Hunt, A.G., Wu, J.Q., 2004. Climatic influences on Holocene variations in soil
erosion rates on a small hill in the Mojave Desert. Geomorphology 58(14),
263289.
Hurst, A.W., Rickerby, P.C., Scott, B.J., Hashimoto, T., 2004. Magnetic field
changes on White Island, New Zealand, and the value of magnetic changes for
eruption forecasting. Journal of Volcanology and Geothermal Research 136,
5370.
Irving, J.D., Knight, R.J., 2003. Removal of wavelet dispersion from groundpenetrating radar data. Geophysics 63(3), 960970.
Johnson, S.Y., Stephenson, W.J., Morgan, L.A., Shanks, III. W.C, Pierce, K.L., 2003.
Hydrothermal and tectonic activity in northern Yellowstone Lake, Wyoming.
Geological Society of America Bulletin 115(8), 954971.
Jol, H.M., Lawton, D.C., Smith, D.G., 2003. Ground penetrating radar: 2-D and 3-D
subsurface imaging of a coastal barrier spit, Long Beach, WA, USA.
Jomard, H., Lebourg, T., Guglielmi, Y., Tric, E., 2009. Electrical imaging of sliding
geometry and fluids associated with a deep seated landslide (La Clapie`re,
France). Earth Surface Processes and Landforms 35(5), 588599.
Kaiser, A.E., Green, A.G., Campbell, F.M., et al., 2009. Ultrahigh-resolution seismic
reflection imaging of the Alpine Fault, New Zealand. Journal of Geophysical
Research 114, B11306. http://dx.doi.org/10.1029/2009JB006338.
Kaiser, A.E., Horstmeyer, H., Green, A.G., Campbell, F.M., Langridge, R.M.,
McClymont, A.F., 2011. Detailed images of the shallow Alpine Fault Zone
New Zealand, determined from narrow-azimuth 3D seismic reflection data.
Geophysics 76, B19B32.
Kanda, W., Utsugi, M., Tanaka, Y., Hashimoto, T., Fujii, I., Hasenaka, T., Shigeno,
N., 2010. A heating process of Kuchi-erabu-jima Volcano, Japan, as inferred
from geomagnetic field variations and electrical structure. Journal of
Volcanology and Geothermal Research 189(12), 158171.
Kristensen, L., Benn, D.I., Hormes, A., Ottesen, D., 2009. Mud aprons in front of
Svalbard surge moraines; evidence of subglacial deforming layers or proglacial
glaciotectonics? Geomorphology 111(34), 206221.
Kruse, S., Grasmueck, M., Weiss, M., Viggiano, D., 2006. Sinkhole structure
imaging in covered karst terrain. Geophysical Research Letters 33, L16405.
Kruse, S., Mora, R., Ramirez, C., Alvarado, G., 2010. Ground penetrating radar
imaging of tephra stratigraphy on Poas and Irazu volcanoes, Costa Rica. Revista
Geologica de America Central 43, 119136.
Kruse, S.E., Schneider, J.C., Campagna, D.J., Inman, J.A., Hickey, T.D., 2000.
Ground penetrating radar imaging of cap rock, caliche and carbonate strata.
Journal of Applied Geophysics 43, 239249.
Kneisel, C., 2004. New insights into mountain permafrost occurrence and
characteristics in glacier forefields at high altitude through the application of 2D
resistivity imaging. Permafrost and Periglacial Processes 15, 221227.
Kneisel, C., Kaab, A., 2007. Mountain permafrost dynamics within a recently
exposed glacier forefield inferred by a combined geomorphological, geophysical
and a photogrammetrical approach. Earth Surface Processes and Landform 32,
17971810.
127
Konishi, N., 1998. Landslide surveys in Tertiary soft rock areas using HEM.
Exploration Geophysics (Melbourne) 29(12), 234239.
Leclerc, R.F., Hickin, E.J., 1997. The internal structure of scrolled floodplain
deposits based on ground-penetrating radar, North Thompson River, British
Columbia. Geomorphology 21(1), 1738.
Lee, C.-C., Tsai, L.L.-Y., Yang, C.-H., Wen, K.-L., Wang, Z.-B., Hsieh, Z.-H., Liu,
H.-C., 2009. The identified origin of a linear slope near chichi earthquake
rupture combining 2D, 3D resistivity image profiling and geological data.
Environmental Geology (Berlin) 58(7), 13971406.
Lehmann, F., Green, A.G., 2000. Topographic migration of georadar data:
implications for acquisition and processing. Geophysics 65, 836848.
Liu, J.P., Xu, K.H., Li, A.C., Milliman, J.D., Velozzi, D.M., Xiao, S.B., Yang, Z.S.,
2007. Flux and fate of Yangtze River sediment delivered to the East China Sea.
Lopes Cardozo, G., Gideon, G.O., Edel, J.B., Granet, M., 2005. Detection of active
crustal structures in the Upper Rhine Graben using local earthquake
tomography, gravimetry and reflection seismic. Quaternary Science Reviews
24(34), 337344.
MacNeil, R.E., Sanford, W.E., Connor, C.B., Sandberg, S.K., Diez, M., 2007.
Investigation of the groundwater system at Masaya Caldera, Nicaragua, using
transient electromagnetics and numerical simulation. Journal of Volcanology and
Geothermal Research 166(34), 217232.
Mallinson, D.J., Culver, S.J., Riggs, S.R., et al., 2010. Regional seismic stratigraphy
and controls on the Quaternary evolution of the Cape Hatteras region of the
Atlantic passive margin, USA. Marine Geology 268, 1633.
Mamtani, M.A., Sengupta, A., 2009. Anisotropy of magnetic susceptibility analysis
of deformed kaolinite: implications for evaluating landslides. International
Journal of Earth Sciences 98(7), 17211725.
Massey, A.C., Taylor, G.K., 2007. Coastal evolution in south-west England, United
Kingdom; an enhanced reconstruction using geophysical surveys. Marine
Geology 245(14), 123140.
Maurer, H., Hauck, C., 2007. Geophysical imaging of alpine rock glaciers. Journal
of Glaciology 53(180), 110120.
Mauritsch, H.J., Seiberl, W., Arndt, R., Roemer, A., Schneiderbauer, K., Sendlhofer,
G.P., 2000. Geophysical investigations of large landslides in the Carnic region
of southern Austria. Engineering Geology 56(34), 373388.
McCall, G.J.H., 2009. Half a century of progress in research on terrestrial impact
structures; a review. Earth-Science Reviews 92(34), 99116.
McClymont, A.F., Green, A.G., Kaiser, A.E., Horstmeyer, H., Langridge, R.M., 2010.
Shallow fault segmentation of the Alpine fault zone, New Zealand revealed
from 2- and 3-D GPR surveying. Journal of Applied Geophysics 70,
343354.
McClymont, A.F., Green, A.G., Villamor, P., Horstmeyer, H., Grass, C., Nobes, D.,
2008. Characterization of the shallow structures of active fault zones using 3-D
GPR data. Journal of Geophysical Research 113, B10315.
McClymont, A.F., Villamor, P., Green, A.G., 2009a. Fault displacement accumulation
and slip-rate variability within the Taupo Rift (New Zealand) based on 3-D
ground-penetrating radar data. Tectonics 28, TC4, 005.
McClymont, A.F., Villamor, P., Green, A.G., 2009b. Assessing the contribution of
off-fault deformation to slip-rate estimates within the Taupo Rift, New Zealand,
using 3-D ground-penetrating radar surveying and trenching. Terra Nova 21(6),
446451.
Miller, R.D., Xia, J., Park, C.B., Ivanov, J.M., 1999. Multichannel analysis of surface
waves to map bedrock. Leading Edge (Tulsa, OK) 18(12), 13921396.
Mills, H.H., Speece, M.A., 1997. Ground-penetrating radar exploration of alluvial
fans in the southern Blue Ridge Province, North Carolina. Environmental and
Engineering Geoscience 3(4), 487499.
Milne, G.A., Mitrovica, J.X., 2008. Searching for eustasy in deglacial sea-level
histories. Quaternary Science Reviews 27(2526), 22922302.
Missiaen, T., Slob, E., Donselaar, M.E., 2008. Comparing different shallow
geophysical methods in a tidal estuary, Verdronken Land van Saeftinge, western
Scheldt, the Netherlands, Geologie en Mijnbouw. Netherlands Journal of
Geosciences 87(2), 151164.
Mol, L., Viles, H.A., 2010. Geoelectric investigations into sandstone moisture
regimes: implications for rock weathering and the deterioration of San Rock Art
in the Golden Gate Reserve, South Africa. Geomorphology 118(34),
280287.
Moore, L.J., Jol, H.M., Kruse, S., Vandenburgh, S., Kaminsky, G., 2004. Annual
layers revealed by GPR in the subsurface of a prograding coastal barrier,
southwest Washington, U.S.A. Journal of Sedimentary Research 74(5), 690696.
Morgan, J.K., Moore, G.F., Clague, D.A., 2003a. Slope failure and volcanic
spreading along the submarine south flank of Kilauea volcano, Hawaii. Journal
of Geophysical Research 108, B9.
128
Morgan, L.A., Shanks, III W.C., Lovalvo, D.A., et al., 2003b. Exploration and
discovery in Yellowstone Lake; results from high-resolution sonar imaging,
seismic reflection profiling, and submersible studies. Journal of Volcanology
and Geothermal Research 122(34), 221242.
Musil, M., Maurer, H., Green, A.G., Horstmeyer, H., Nitsche, F.O., Vonder Muehll,
D., Springman, S., 2002. Shallow seismic surveying of an alpine rock glacier.
Geophysics 67(6), 17011710.
Nakatsuka, T., Utsugi, M., Okuma, S., Tanaka, Y., Hashimoto, T., 2009. Detection
of aeromagnetic anomaly change associated with volcanic activity; an
application of the generalized mis-tie control method. Tectonophysics 478(12),
318.
Naudet, V., Lazzari, M., Perrone, A., Loperte, A., Piscitelli, S., Lapenna, V., 2008.
Integrated geophysical and geomorphological approach to investigate the
snowmelt-triggered landslide of Bosco Piccolo village (Basilicata, southern Italy).
Engineering Geology 98, 156167.
Nielsen, L., Moller, I., Nielsen, L.H., et al., 2003. Integrating ground-penetrating
radar and borehole data from a Wadden Sea barrier island. Journal of Applied
Okubo, A., Kanda, W., Tanaka, Y., et al., 2009. Apparent magnetization intensity
map on Sakura-jima Volcano, Kyushu, Japan, inferred from low altitude, high
density helicopter-borne aeromagnetic surveys. Tectonophysics 478(12),
3442.
Okuma, S., Stotter, C., Supper, R., Nakatsuka, T., Furukawa, R., Motschka, K., 2009.
Aeromagnetic constraints on the subsurface structure of Stromboli Volcano,
Aeolian Islands, Italy. Tectonophysics 478(12), 1933.
Pedley, K.L., Barnes, P.M., Pettinga, J.R., Lewis, K.B., 2010. Seafloor structural
geomorphic evolution of the accretionary frontal wedge in response to seamount
subduction, Poverty Indentation, New Zealand. Marine Geology 270(14),
119138.
Pelpola, C.P., Hickin, E.J., 2004. Long-term bed load transport rate based on aerialphoto and ground penetrating radar surveys of fan-delta growth, Coast
Mountains, British Columbia. Geomorphology 57(34), 169181.
Pelton, J.R., 2005. Near-surface seismology: surface-based methods. In: Butler, D.K.
(Ed.), Near-Surface Geophysics. Society of Exploration Geophysicists, Tulsa, OK,
pp. 219263.
Peters, G., Buchmann, T.J., Connolly, P., van Balen, R.T., Wenzel, F., Cloetingh,
S.A.P.L., 2005. Interplay between tectonic, fluvial and erosional processes along
the Western Border Fault of the northern Upper Rhine Graben, Germany.
Tectonophysics 406(12), 3966.
Prezzi, C.B., Uba, C.E., Goetze, H.J., 2009. Flexural isostasy in the Bolivian Andes;
Chaco foreland basin development. Tectonophysics 474(34), 526543.
Regard, V., Lagnous, R., Espurt, N., et al., 2009. Geomorphic evidence for recent
uplift of the Fitzcarrald Arch (Peru); a response to the Nazca Ridge subduction.
Reshef, M., Ben-Avraham, Z., Tibor, G., Marco, S., 2007. The use of acoustic
imaging to reveal fossil fluvial systems; a case study from the southwestern Sea
of Galilee. Geomorphology 83(12), 5866.
Ribolini, A., Chelli, A., Guglielmin, M., Pappalardo, M., 2007. Relationships
between glacier and rock glacier in the Maritime Alps, Schiantala Valley, Italy.
Quaternary Research 68(3), 353363.
Ribolini, A., Guglielmin, M., Fabre, D., et al., 2010. The internal structure of rock
glaciers and recently deglaciated slopes as revealed by geoelectrical tomography:
insights on permafrost and recent glacial evolution in the Central and Western
Alps (ItalyFrance). Quaternary Science Reviews 29(34), 507521.
Reis, A.T., Perovano, R., Silva, C.G., Vendeville, B.C., Araujo, E., Gorini, C.,
Oliveira, V., 2010. Two-scale gravitational collapse in the Amazon Fan; a
coupled system of gravity tectonics and mass-transport processes. Journal of
the Geological Society of London 167(3), 593604.
Riva, R.E.M., Gunter, B.C., Urban, T.J., et al., 2009. Glacial isostatic adjustment
over Antarctica from combined ICESat and GRACE satellite data. Earth and
Planetary Science Letters 288(34), 516523.
Roberts, M.C., Pullan, S.E., Hunter, J.A., 1992. Applications of land-based high
resolution seismic reflection analysis to Quaternary and geomorphic research.
Quaternary Science Reviews 11(5), 557568.
Roberts, R.L., Daniels, J.J., 1996. Analysis of GPR polarization phenomena. Journal
of Environmental and Engineering Geophysics 1(2), 139157.
Russell, J.K., Stasiuk, M.V., 1997. Characterization of volcanic deposits with
ground-penetrating radar. Bulletin of Volcanology 58, 515527.
Rymer, H., 1994. Microgravity change as a precursor to volcanic activity. Journal of
Saballos, J.A., Malservisi, R., Connor, C., 2010. Gravity and geodetic studies at
Concepcion volcano. American Geophysical Union, Nicaragua, Fall Meeting
2010, abstract G23C-0846.
Sadura, S., Martini, I.P., Endres, A.L., Wolf, K., 2006. Morphology and GPR
stratigraphy of a frontal part of an end moraine of the Laurentide ice
sheet; Paris Moraine near Guelph, ON, Canada. Geomorphology 75(12),
212225.
Sass, O., 2006. Determination of the internal structure of alpine talus deposits
using different geophysical methods (Lechtaler Alps, Austria). Geomorphology
80(12), 4558.
Sass, O., 2007. Bedrock detection and talus thickness assessment in the European
Alps using geophysical methods. Journal of Applied Geophysics 62(3),
254269.
Sass, O., Bell, R., Glade, T., 2008. Comparison of GPR, 2D-resistivity and
traditional techniques for the subsurface exploration of the Oschingen landslide,
Swabian Alb (Germany). Geomorphology 93, 89103.
Sass, O., Krautblatter, M., 2007. Debris flow-dominated and rockfall-dominated talus
slopes; genetic models derived from GPR measurements. Geomorphology
86(12), 176192.
Schmutz, M., Guerin, R., Andrieux, P., Maquaire, O., 2009. Determination of the
3D structure of an Earthflow by geophysical methods: the case of Super
Sauze, in the French southern Alps. Journal of Applied Geophysics 68(4),
500507.
Schneider, J.A., Kruse, S.E., 2005. Assessing natural and anthropogenic impacts on
freshwater lens morphology on small barrier islands: Dog Island and St. George
Island, Florida. Hydrogeology Journal 14, 131145.
Schrott, L., Hordt, A., Dikau, R., 2003a. Geophysical applications in
geomorphology. Zeitschrift fur Geomorphologie Supplementband, Stuttgart and
Berlin Gebruder Borntraeger 132, 190.
Schrott, L., Hufschmidt, G., Hankammer, M., Hoffmann, T., Dikau, R., 2003b. Spatial
distribution of sediment storage types and quantification of valley fill deposits in
an alpine basin, Reintal, Bavarian Alps, Germany. Geomorphology 55(14),
4563.
Schrott, L., Sass, O., 2008. Application of field geophysics in geomorphology;
advances and limitations exemplified by case studies. Geomorphology 93(12),
5573.
Shah, A.K., Daniels, D.L., Kontny, A., Brozena, , John, M., 2009. Megablocks and
melt pockets in the Chesapeake Bay impact structure constrained by magnetic
field measurements and properties of the Eyreville and Cape Charles cores.
Special Paper Geological Society of America 458, 195208.
Sheehan, J.R., Doll, W.E., Mandell, W.A., 2005. An evaluation of methods and
available software for seismic refraction tomography analysis. Journal of
Environmental and Engineering Geophysics 10(1), 2134.
Sloan, S.D., Tsoflias, G.S., Steeples, D.W., Vincent, P.D., 2007. High-resolution
ultra-shallow subsurface imaging by integrating near-surface seismic reflection
and ground-penetrating radar data in the depth domain. Journal of Applied
Steffen, H., Gitlein, O., Denker, H., Mueller, J., Timmen, L., 2009. Tectonophysics.
Present rate of uplift in Fennoscandia from GRACE and absolute gravimetry
474(12), 6977.
Stokes, C.R., Lian, O.B., Tulaczyk, S., Clark, C.D., 2008. Superimposition of ribbed
moraines on a palaeo-ice-stream bed; implications for ice stream dynamics and
shutdown. Earth Surface Processes and Landforms 33(4), 593609.
Srigutomo, W., Kagiyama, T., Kanda, W., et al., 2008. Resistivity structure of Unzen
Volcano derived from time domain electromagnetic (TDEM) survey. Journal of
Stucchi, E., Mazzotti, A., 2009. 2D seismic exploration of the Ancona landslide
(Adriatic Coast, Italy). Geophysics 74(5), B139B151.
Stummer, P., Maurer, H., Green, A.G., 2004. Experimental design; electrical
resistivity data sets that provide optimum subsurface information. Geophysics
69(1), 120139.
Suski, B., Brocard, G., Authemayou, C., Consenza Muralles, B., Teyssier, C.,
Holliger, K., 2010. Localization and characterization of an active fault in an
urbanized area in central Guatemala by means of geoelectrical imaging.
Tectonophysics 480, 8898.
Tabor, R.W., Cady, W.M., 1978. The structure of the Olympic Mountains,
Washington analysis of a subduction zone. U.S. Geological Survey
Professional Paper 1033, 25pp.
Tamisiea, M.E., Mitrovica, J.X., Davis, J.L., 2007. GRACE gravity data constrain
ancient ice geometries and continental dynamics over Laurentia. Science 316,
881883.
Tamura, T., Murakami, F., Nanayama, F., Watanabe, K., Saito, Y., 2008. Groundpenetrating radar profiles of Holocene raised-beach deposits in the Kujukuri
strand plain, Pacific coast of eastern Japan. Marine Geology 248(12), 1127.
Telford, W.M., Geldart, L.P., Sheriff, R.E., 1990. Applied Geophysics, Second ed.
Cambridge University Press, New York, NY, 770 pp.
Tizzani, P., Battaglia, M., Zeni, G., Atzori, S., Berardino, P., Lanari, R., 2009. Uplift
and magma intrusion at Long Valley caldera from InSAR and gravity
measurements. Geology 37(1), 6366.
Tronicke, J., Villamor, P., Green, A.G., 2006. Detailed shallow geometry and vertical
displacement estimates of the Maleme Fault Zone, New Zealand, using 2-D and
3-D georadar. Near Surface Geophysics 4, 155161.
Tsoflias, G.P., Steeples, D.W., Czarnecki, G.P., Sloan, S.D., Eslick, R.C., 2006.
Automatic deployment of a 2-D geophone array for efficient ultra-shallow
seismic imaging. Geophysical Research Letters 33(9), 4.
Ueda, Y., 2007. A 3D magnetic structure of Izu-Oshima Volcano and their changes
after the eruption in 1986 as estimated from repeated airborne magnetic surveys.
Journal of Volcanology and Geothermal Research 164(3), 176192.
van der Kruk, J., Streich, R., Green, A.G., 2006. Properties of surface waveguides
derived from separate and joint inversion of dispersive TE and TM GPR data.
Geophysics 71(1), K19K29.
van der Kruk, J., Vereecken, H., Jacob, R.W., 2009. Identifying dispersive GPR
signals and inverting for surface wave-guide properties. Leading Edge 28(10),
12341239.
Vanneste, K., Verbeeck, K., Petermans, T., 2008. Pseudo-3D imaging of a low-sliprate, active normal fault using shallow geophysical methods: The Geleen fault in
the Belgian Maas River valley. Geophysics 73(1), B1B9.
Voett, A., Brueckner, H., Brockmueller, S., et al., 2009. Traces of Holocene tsunamis
across the Sound of Lefkada, NW Greece. Global and Planetary Change
66(12), 112128.
Vonder Muehll, D., Hauck, C., Gubler, H., 2002. Mapping of mountain permafrost
using geophysical methods. Progress in Physical Geography 26, 643660.
Vonder Muehll, D.S., Hauck, C., Lehmann, F., 2000. Verification of geophysical
models in Alpine permafrost using borehole information. Annals of Glaciology
31(1), 300306.
Wang, P., Horwitz, M.H., 2007. Erosional and depositional characteristics of
regional overwash deposits caused by multiple hurricanes. Sedimentology 54(3),
545564.
Wang, Z., Chen, Z., Chen, J., Wei, Z., 2007. Seismic framework and the Holocene
morphological evolution of the Changjiang River mouth, China. Geomorphology
85(34), 237248.
129
Watabe, Y., Sassa, S., 2008. Application of MASW technology to identification of

tidal flat stratigraphy and its geoenvironmental interpretation. Marine Geology
252(34), 7988.
Wellner, J.S., Heroy, D.C., Anderson, J.B., 2006. The death mask of the Antarctic
ice sheet; comparison of glacial geomorphic features across the continental
shelf. Geomorphology 75(12), 157171.
Wetmore, P.H., Connor, C.B., Kruse, S.E., Callihan, S., Pignotta, G., Stremtan, C.,
Burke, A., 2009. Geometry of the Trachyte Mesa intrusion, Henry Mountains,
Utah: implications for the emplacement of small melt volumes into the upper
crust. Geochemistry, Geophysics, Geosystems 10, 8. http://dx.doi.org/10.1029/
2009GC002469.
Whiteley, R.J., Eccleston, P.J., 2006. Comparison of shallow seismic refraction
interpretation methods for regolith mapping. Exploration Geophysics
(Melbourne) 37(4), 340347.
Wilkinson, P.B., Meldrum, P.I., Chambers, J.E., Kuras, O., Ogilvy, R.D., 2006.
Improved strategies for the automatic selection of optimized sets of electrical
resistivity tomography measurement configurations. Geophysical Journal
International 167(3), 11191126.
Williams-Jones, G., Rymer, H., Mauri, G., Gottsmann, J., Poland, M., Carbone, D.,
2008. Toward continuous 4D microgravity monitoring of volcanoes. Geophysics
73(6), WA19WA28.
Xu, C., Butt, S.D., 2006. Evaluation of MASW techniques to image steeply dipping
cavities in laterally inhomogeneous terrain. Journal of Applied Geophysics 59,
106116.
Yamaya, Y., Mogi, T., Hashimoto, T., Ichihara, H., 2009. Hydrothermal system
beneath the crater of Tarumai Volcano, Japan; 3-D resistivity structure revealed
using audio-magnetotellurics and induction vector. Journal of Volcanology and
de Zeeuw-van Dalfsen, E., Rymer, H., Sigmundsson, F., Sturkell, E., 2005. Net
gravity decrease at Askja Volcano, Iceland; constraints on processes responsible
for continuous caldera deflation, 19882003. Journal of Volcanology and
Zlotnicki, J., Sasai, Y., Toutain, J.P., et al., 2009. Combined electromagnetic,
geochemical and thermal surveys of Taal volcano (Philippines) during the period
20052006. Bulletin of Volcanology 71(1), 2947.
Biographical Sketch
Sarah Kruse is an associate professor in the Geology Department of the University of South Florida. She uses nearsurface geophysics to address hydrogeological, volcanological, and tectonic problems. She has worked on topics
ranging from the hydrogeology of small islands to volcano stratigraphy to the resolution limits of groundpenetrating radar.
3.6 Digital Terrain Modeling

T Wasklewicz, East Carolina University, Greenville, NC, USA
DM Staley, US Geological Survey, Golden, CO, USA
K Reavis, East Carolina University, Greenville, NC, USA
T Oguchi, University of Tokyo, Kashiwa-shi, Japan
3.6.1
3.6.2
3.6.3
3.6.3.1
3.6.3.2
3.6.3.3
3.6.3.3.1
3.6.3.3.2
3.6.3.3.3
3.6.3.3.4
3.6.3.3.5
3.6.4
3.6.4.1
3.6.4.2
3.6.4.3
3.6.4.4
3.6.5
3.6.5.1
3.6.5.1.1
3.6.5.1.2
3.6.5.1.3
3.6.5.2
3.6.6
3.6.6.1
3.6.6.1.1
3.6.6.1.2
3.6.6.1.3
3.6.7
3.6.7.1
3.6.7.2
3.6.7.3
3.6.7.4
3.6.8
References
Introduction
Background
DTM Representation
Land Surface
Scales and Land Surface
Data Structures
Raster (grid)
Triangulated irregular network (TIN)
Contour lines
Breaklines
Mass points
Data Sources
Terrestrial Laser Scanning Techniques
Terrestrial Photogrammetry Techniques (TPT)
Airborne Laser Scanning
Interferometric Synthetic Aperture Radar
Preprocessing
Point-Cloud Reduction Techniques
Decimation
Filtering
Segmentation
Surface Interpolation
DTM Error Assessment
Error Issues
Global errors
Local errors
Systematic errors
High-Resolution DTMs
High-Resolution DTMs and Scale
Data Fusion
Temporal Data Acquisition
Conclusions
Glossary
Accuracy The degree of closeness of a measurement of a
quantity to its actual (true) value.
Airborne laser scanning Laser scanning of the Earths
surface (topographic and bathymetric) conducted from
either a fixed wing aircraft or a helicopter.
Digital elevation model A file or database of
numerical values that approximates topographic/
bathymetric variability at given x, y, and z;
coordinates.
131
131
133
133
134
138
138
140
140
140
140
141
141
143
144
145
146
147
147
147
148
148
150
150
150
151
151
151
152
154
154
154
155
156
Digital terrain model A bare-earth digital elevation

model that is devoid of vegetation and human-built
structures.
Interferometric synthetic aperture radar Interferometric
synthetic aperture radar refers to a rapidly evolving form
of radar data processing that allows for the
characterization of topography using airborne or
spacecraft-mounted radar systems.
Point clouds A large number of points representing
topographic/bathymetric variations that are recorded as x, y,
Wasklewicz, T., Reavis, K., Staley, D.M., Oguchi, T., 2013. Digital terrain
modeling. In: Shroder, J. (Editor in Chief), Bishop, M.P. (Ed.), Treatise on
130
http://dx.doi.org/10.1016/B978-0-12-374739-6.00048-8
Digital Terrain Modeling
and z coordinates most commonly by laser scanners or

photogrammetric approaches.
Terrestrial laser scanning Laser scanning of the Earths
surface conducted with a tripod-mounted high-definition
surveying device.
Terrestrial photogrammetry techniques The
development of high-resolution digital terrain models for
131
geomorphic analysis using digital cameras and specialized

photogrammetric software. Conducted properly, this
complex process allows high-resolution DTMs to be
created from vertical or oblique images of the Earths
surface.
Abstract
Geomorphologists require quantitative information about the land surface. New sensors can now measure elevation
changes at a variety of scales and these data are used to generate digital terrain model (DTM) that accurately characterize
topography. A variety of DTM analytics are used to support a multitude of geomorphological studies. However, there are
numerous issues involving representation, sampling, interpolation, and error assessment and correction, which must be
addressed before using the elevation data. Data reduction, filtering, and accuracy are key aspects to consider. Knowledge of
these issues is critical for terrain analysis and the communication of information derived from a DTM.
3.6.1
Introduction
Quantitative characterization of topography is required to explain the formation, diversity, and dynamics of surface processes
and landforms, as well as unraveling the complexity of anthropogenic forcing on landscape evolution (Tucker and Hancock,
2010). Historically, the generation and use of topographic maps
greatly facilitated terrain analysis and geomorphological mapping. Modern-day terrain analysis, or geomorphometry, requires
digital terrain modeling and the production of digital terrain
models (DTMs) that accurately characterize the topography. The
rapid proliferation of geospatial technologies greatly facilitates
digital terrain modeling and terrain analysis. Scientific and applied research in geomorphology is increasingly making use of
DTMs. Digital terrain modeling and terrain analysis are now
fundamental to geomorphology, and applications span the
spectrum of surface-process domains.
Advances in remote-sensing technology permit multiple
sensors on numerous platforms to acquire digital data that
can be used to generate DTMs. What seems to be an everincreasing capacity to acquire detailed landscape information,
greatly facilitates digital terrain modeling and terrain analysis.
Consequently, new spatio-temporal data permit many new
advances in geomorphometry, involving the development and
evaluation of new algorithms and analysis procedures. Technically, the cost effectiveness (monetary expenses and human
labor), increased spatial extent, accuracy, measurement scales,
and an ability to rapidly mobilize before and after a geomorphic event represent important developments that are
unprecedented in geomorphological research.
The availability of new and improved data, analytical capabilities, and progress in theoretical/conceptual treatments of
geomorphological systems now permit new opportunities to
study process-form relationships. Numerous geomorphometric
parameters have been developed to characterize the multiscale
morphological attributes of the topography. Many of these
same parameters also play a role in specific process mechanics,
as topographic parameters govern many surface processes, and
determine the magnitude of process rates. In some instances,

the rates of erosion and deposition can be accurately estimated
by using multitemporal DTMs. Research reveals, however, that
high-resolution data with increased spatial coverage is required
to account for the operational scale-dependencies of many
surface processes and their linkage to multiscale terrain parameters. Such process-form linkages have already resulted in
new insights into short-term and long-term landscape evolution (e.g., Hancock et al., 2010; Stark et al., 2010). Nevertheless, there are many issues associated with digital terrain
modeling that require knowledge of the process and limitations
associated with analysis and information extraction.
Consequently, the objective of this chapter is to cover the
many developments related to digital terrain modeling and the
use of digital elevation models (DEMs). The focus is on important background information and key issues. The authors
first address terrain representation concepts and discuss scale
and data structure issues. Recent advances in DTM generation are
highlighted with particular emphasis placed on terrestrial laser
scanning, terrestrial photogrammetry techniques, airborne laser
scanning, and interferometric synthetic-aperture radar. The
aforementioned data sources characteristically produce large
volumes of data, and it is important for geomorphologists to
note that not all DTMs represent bare-earth information.
The authors then discuss the numerous types of preprocessing
that are required to address missing data and error via various
algorithms and spatial interpolation procedures. Postprocessing
is also covered and related to DTM error and uncertainty. Finally,
the chapter concludes by considering the importance of DTMs to
current and future studies in geomorphology, with consideration
of undeveloped opportunities and research directions.
3.6.2
Background
A fledgling group of geomorphologists began to present research on landforms in a more systematic and quantitative
fashion during the late 1940s and early 1950s at the onset of
132
the quantitative revolution (Horton, 1945; Mackin, 1948;

Strahler, 1952). These works led to a proliferation of research
that examined the operation of process-mechanics within
geomorphological systems, and became a dominant approach
for almost 50 years of applied and scientific studies in geomorphology (e.g., Dury, 1983; Morisawa, 1988; Strahler,
1992; Rhoads, 2006; Haggett, 2008). This is not to suggest that
research during this period did not consider process-form relationships. Observations of process-oriented information
were linked with two-dimensional (2D) landform shapes
(Horton, 1945; Schumm, 1956; Strahler, 1958; Melton,
1957). Studies linking process and form remained spatially,
temporally, and morphometrically limited and continued to
be underrepresented in the literature until the 1980s. The
limited scope of process-form studies (beyond 2D measurements of landform shape) can be directly related to an inability to produce measures/metrics that describe various
aspects of process and form. Detailed information on particle
entrainment and movement, wind speeds, and fluid discharges were generally not in accordance with 2D measurements of topography and landforms. Early emphases on
quantitative process mechanics had rapidly eclipsed the geomorphologists ability to capture topographic information in a
manner capable of producing comparable measures of form
(Pike, 1995). This was in part a function of rapid advances in
sampling instruments for measuring flows of energy (wind,
water, and waves) and matter (sediment), as well as equations
to approximate the dynamics of energy and matter (e.g., unit
stream power and critical threshold shear velocity), and lack of
geospatial technologies critical to the emergence and evolution of digital terrain modeling.
Attempts were made to quantify terrain and landforms
before the availability of DTMs. A variety of measures and
indices from topographic maps were employed to characterize
basic morphometric characteristics such as length, area, and
relief (Schumm, 1956; Strahler, 1958; Mark, 1974). Many
studies, however, did not follow consistent rules of collecting
or analyzing morphometric data, which further inhibited the
broader application of the research and brought into question
the representativeness of the measures and indices for characterizing specific aspects of terrain and landforms (Evans,
1972, 1980). This work was also hindered by a general lack of
spatial analyses of topographic data during the quantitative
revolution (Gregory, 2000). Much of the work of this period
relied on statistical and later mathematical approaches to
testing hypotheses, where more recent work has focused on
integrating statistics, mathematics, and spatial analyses
(Hayakawa and Oguchi, 2006; Lin and Oguchi, 2006;
Dunning et al., 2009).
Early morphometric and process-mechanic oriented techniques devised during the quantitative revolution led to advances in geomorphology. They also established a foundation
and a need for digital data that could represent the continuous
nature of the land surface and permit topographic change
detection. The emergence of computers in the early- to mid1950s led to the first digital analysis of topography. Although
not a DTM in the sense of representing a continuous surface,
initial models were numerical points used in the development
of topographic profiles (Pike et al., 2009) and used to assess
microrelief (Walls et al., 1954) and elevation spectra
(Jaeger and Schuring, 1966). It is evident from these and other

examples that progress was occurring in geomorphology and
engineering. These approaches would later be adopted to
initiate analysis of the topography using DTMs.
Pike et al. (2009) provided a detailed history of DTM use in
the field of geomorphometry, which is now fundamental to
geomorphological research. Herein, the authors synthesize their
historical interpretation to provide a short-history of DTM incorporation into geomorphology. Digital terrain modeling as a
concept had its early origins in the late 1950s (Miller and
Leflamme, 1958). Practical applications for DTMs were not
appreciated until the development of 63 m gridded data derived from 1:250 000 scale maps of the US during 196372
(Noma and Misulia, 1959; U.S. Army Map Service, 1963), or in
a very few cases through laborious efforts by individuals who
generated DTMs from topographic maps, field surveys, or
photogrammetry techniques (Tobler and Davis, 1968; Sakaguchi, 1969). The visually interpreted DTMs were first used to
formulate mathematical and statistical means for measuring
elevation and its surface derivatives in geomorphology (Tobler,
1969; Evans, 1972). The geomorphological relevance of this
work was expounded on in the context of hillslope geomorphology (Carson and Kirkby, 1972), which provided a
subsequent framework to advance quantitative DTM approaches into other subfields of geomorphology.
Computer processing, visualization, and software with the
capability to handle DTMs remained an issue throughout
much of the 1970s and 1980s. Statistical software packages
produced descriptive, univariate, and multivariate statistics
from tabular elevation and surface derivative data sets during
this time frame. It was not until the development of personal
computers, however, and the advent of software packages with
the capability to process and analyze raster data such as
GRASS (http://grass.osgeo.org/, 1985), MicroDEM (Guth
et al., 1987), IDRISI (Eastman, 1989), and ArcView (http://
www.esri.com/about-esri/index.html, 1991) that there was
broad-scale use of DTMs. Data availability from the 1980s to
the present was also a driving factor in the advancement of
terrain-analysis techniques in geomorphology. The development of DTMs by a variety of governmental agencies across
the globe provided users with free access to data (e.g., USGS
7.5-min DEM, 11 DEM and 2 asec DEM). Users were also able
to purchase instruments capable of producing digital elevation
data (e.g., total stations, Global positioning systems (GPS)
surveying equipment, and terrestrial laser scanners (TLS)), or
had access to a variety of remote-sensing data (e.g., Shuttle
Radar Topography Mission, digital photogrammetry, InSAR/
IfSAR, ASTER, SPOT, and Airborne laser scanning (ALS;
commonly referred to as LiDAR)) that could be used to provide products more tailored to the users needs in terms of
accuracy and acquisition time and frequency.
The combination of advances in sensors, computer technology, and analysis capabilities is clearly documented in the
peer-reviewed literature. A Web-of-Science topics search using
the terms digital and elevation, revealed 33 publications
during 197589, 995 during 199099, and 3602 during
200009. Similar increases, but smaller overall numbers of
articles were noted for a title search using the same terms, the
addition of geomorphology with the previous terms as a
topic and title search, and a topic search with the terms
digital and terrain. It is obvious from these numbers that the

use of DTMs increased exponentially, and the likelihood for
continued use in research and education is very high, given the
current training and job opportunities afforded to students
and scientists in multiple sectors of society.
3.6.3
DTM Representation
Information representation is an important topic in geomorphology and geographic information science (Bishop and
Shroder, 2004). Quantitative characterization of the topography is required if we want to formalize our understanding
of surface processes and landscape evolution. This raises the
complex question of how do we best characterize various aspects of the topography? Do we want to have formal representations of surface processes and landforms, or do we want
to simply represent the altitude field? Furthermore, do we require 3D and 4D representations for modeling and mapping,
or are static cartographic representations suitable? These and
many more representation questions (e.g., indeterminate
boundaries) are related to formalizing our knowledge and
setting the stage for developing and testing new concepts in
geomorphology.
Historically, many concepts have evolved from detailed
fieldwork. However, field investigations sometimes do not
result in a definitive or quantitative statement about the
complexities of the landscape. Many observers do not possess
the ability to recognize subtle features or changes critical to
interpreting landscape or landform dynamics. DTM development is based on a particular representational scheme (e.g.,
entity or field model) that permits a detailed means of mapping, quantifying, and analyzing the topography. Detailed
maps and measurements support existing field observations
and offer new virtual observations about geomorphic change.
Nevertheless, capabilities and the quality of the information
extracted from the DTM is only as good as the representational
scheme.
Similarly, geomorphological research has brought about the
need for quantitative parameters that characterize various concepts of topographic structure and process-form relationships.
This information is commonly combined with process-oriented
data (i.e., rates of erosion) to provide a more definitive view of
landscape or landform dynamics. However, our ability to accurately characterize specific parameters is ultimately related to
representation (parameterization schemes).
Finally, scale is an important topic in geomorphology and
GIScience, and geomorphologists require information at a
variety of spatial and temporal scales. Geospatial technologies
permit the capture and production of information at a
variety of measurement scales, spatial extents, and temporal
frequencies. Nevertheless, in addition to geographic, cartographic, and measurements scales, the operational scaledependencies of surface processes and the hierarchical
scale-dependencies of topographic structure/organization must
also be accounted for. Currently, several aspects of scale can be
addressed via spatial resolution and spatial analysis software
(e.g., Bishop et al., 2003), however, standard cartographic
representational schemes do not formally account for the 2D
133
or 3D spatial structure of the topography that is related to

process domains, landform evolution, and tectonics.
The aforementioned discussion, however, involves accounting for numerous conceptual treatments of what the
landscape actually is. It can be viewed as a complex entity or a
simple entity depending on ontological considerations. Based
on existing conceptual treatments and geospatial technologies,
a DTM is a representation of the land surface. Geomorphologists must be mindful that DTMs are not exact replicas of
the Earths surface, and do not formally characterize surface
processes and landforms. The term representation reflects an
inability to formalize many parameters and linkages that integrate the total complexity associated with the concepts of
topography and land surface. Therefore, DTMs are surface
models developed to best represent the spatial continuity of
the altitude field. These caveats require us to formally define
what the land-surface represents and address scale issues, as
background to DTMs and their associated data structures.
3.6.3.1
Land Surface
The field of geomorphology is concerned with the geometry,

structure (internal and external), coexistence with other forms
(biota and humans), and dynamics (mode of evolution and
processes integral to their existence and evolution) of the land
surface (Bauer, 1996). The land surface is defined here as the
entire solid Earth surface (oceanic and continental crust). Although geomorphologists are conducting studies at the scale
of continents or ocean basins, far more research has been and
continues to be focused on the dynamics of landscapes and
landforms. In general, a landscape is frequently thought of as
a collection of systems that produce a unique landscape
character including climate, geology, topography and landforms, soils, and vegetation. A more restricted interpretation
can be envisaged as a unique composite of landforms. For
example, the landscape of western North Carolina is composed of the Appalachian Mountains (Figure 1). Watersheds
are a major structural landform governing the collection and
drainage of water. Each watershed is composed of some
combination of geomorphic features such as hillslopes, cliffs,
talus, colluvial hollows, rills, gullies, channels, and alluvial
Figure 1 A photograph of a section of the Appalachian Mountains

to the west of Mount Mitchell showing the drainage basins, which
are a common landform on the landscape.
134
fans. Mountainous topography of western North Carolina is

also punctuated by other landforms such as peaks, ridges, and
passes. Each of these landforms possesses their own characteristic morphology and surface expression that are tied to a
complex response to recent and historical processes. The
morphometric and topographic expressions of the landforms
and landscapes are generally what are sought after in digital
terrain modeling and terrain analyses.
Terrain analysis at the landscape-scale has traditionally
focused on the Earths surface at a scale of 101102 m or
greater, dependent on application and the area of interest
(Wilson and Gallant, 2000; Hengl and Evans, 2009). Terrain
analysis at the landscape-scale typically produces information
that captures patterns attributed to spatial and temporal
variations in the topography, which may or may not be related
to specific landforms or processes (Minar and Evans, 2008).
Landforms represent smaller distinct components of a landscape that may be characteristic of a particular process or the
result of polygenetic evolution. They exhibit internal facies,
surface roughness, and 2D and 3D morphological characteristics (Smith and Mark, 2003). Geomorphological analyses at
the landform-scale ranges from 101 to 102 m. Characterization of landforms involves assessing spatial and temporal
variations in topography through a variety of morphometric
measurements, process-based analyses, statistical and mathematical approaches, various modeling approaches, or some
combination of these.
Space, time, and scale variations in the land surface have
been of particular interest to geomorphologists and are generally considered in the conceptual context of dynamism
within geomorphology (Rhoads, 2006). Throughout much of
the history of geomorphology, these items have generally been
investigated conceptually and analytically as separate entities.
Advances in geomorphology beyond the quantitative era
have begun to place more emphasis on integrating these items
in the context of land-surface dynamics (Walker, 2010). Landsurface dynamics involve process-form interactions during an
event that promotes a net spatial and temporal change in the
landscape or landform topography. Operational scales of
processes, feedback mechanisms, and systems vary from event
to event. Thus, landscapes and landforms represent an amalgamation of events that can produce different topographic
signatures (Volker et al., 2007) or might lead to emergence
(Phillips, 1999b). Regardless of the outcome, DTMs are critical
data sets for capturing the topographic changes associated
with land-surface dynamism. Topography reflects both the
external and internal dynamism affecting change of the land
surface, as well as represents a critical forcing factor interacting
with processes during subsequent events.
Land-surface dynamism is not completely driven by natural processes and events. An ever increasing population has
also intensified the role of humans as agents of landscape and
landform change. Impacts of early populations were more
localized as population density was lower and mechanization
was limited (Hooke, 2000). The development of large urban
areas and the use of mechanized equipment brought a shift in
impacts from the local-scale to the regional- and global-scales.
Land-use change associated with human development and
agriculture has in general played a significant role in altering
the processes associated with landscape and landform
evolution (Brierley, 2010). Ground-cover alteration or removal

can initially accelerate the magnitude and timing of sediment
and water flux. The initial increase may diminish through time
as the source material is reduced, urbanization covers surfaces,
or exotic species establish (Houben, 2008). Increasing temperature and in some locales increased rates of precipitation,
associated with human-induced global climate change, are
major factors in changing weathering rates and patterns
(Gislason et al., 2008). Modification of weathering rates and
patterns can lead to significantly different rates of soil production, soil erosion, bedrock erosion, and saprolite production (Dixon et al., 2009). The combination of the internal
and external energy and matter within systems ultimately
alters the magnitude and form of landscape and landform
evolution. Although a consensus exists that humans play a
significant role as agents of land-surface change, major efforts
are required to accurately determine the spatial and temporal
scales at which humans modify processes and forms.
3.6.3.2
Scales and Land Surface
Identifying an objective geographical scale of unique landscapes and landforms through time has been an elusive goal
of science in general, and of geomorphology in particular
(Atkinson and Tate, 2000). Many studies only address the
concept of scale from a computational or analysis point-ofview (Dikau, 1990; Evans, 2003; Bishop, 2013), or only account for a very narrow range of geographic scales (Schmidt
and Andrew, 2005; Shary et al., 2005; Staley et al., 2006),
despite scale being a major issue in geomorphology. Furthermore, these treatments generally exclude micro- to local-scale
surface roughness (Reuter et al., 2009). Surface roughness
represents local topographic variability. Topographic variations result from various objects (e.g., boulder- or cobblesized material on the surface of an alluvial fan) that are part of
the land surface. An adequate assessment of surface roughness
requires an initial understanding of the measurement scale
required to collect the data. The operational and geographic
scales of the processes and structures under investigation, and
the representational resolution of the DTM are important
pieces of information needed to accurately capture landform
and landscape change. Yet, the appropriate levels of scale required to characterize the topography have rarely been
examined or synthesized in a systematic way for studying
landform and landscape evolution.
From a digital terrain modeling perspective, geographical
scale is defined as the spatial extent of a land-surface feature.
Our ability to characterize and map landforms at various scales
is largely controlled by the horizontal and vertical resolution of
the DTM. In general, as the DTM resolution (cell size, triangle
size, or point spacing) becomes coarser, there is a corresponding decrease in the captured land-surface information
(terrain details are lost). A finer resolution DTM possesses more
data and generally more precisely represents the topographic
variability indicative of a landscape (Figure 2). A logical inference from the previous statements is that users want to actively
seek the highest resolution data possible. There are, however,
potential pitfalls with this approach, as higher resolution data
require a greater computational expense (and might be cost
135
Location
of Fig. 2D
NED 10 m
High : 4345.43
0
0.15
0.3 km
(a)
Low : 2396.18
0.15
0.3 km
(c)
Elevation (m) TLS

High: 1002.81
Low: 950.627
3
6m
ALS Data (1m)

High: 4595.84
0
(b)
0.15
0.3 km
Low: 2494.64
(d)
Figure 2 Three different DTM resolutions from the USGS Natural Debris Flow Laboratory at Chalk Cliffs, CO that show the differences in detail
ranging from medium to fine resolution. (a) 10 m DTM from the National Elevation Dataset. (b) 1 m resolution DTM generated with ALS
techniques. (c) Hillshade map from ALS data showing the location of the DTM in Figure 2(d). (d) A 2 cm DTM from a section of the channel in the
intensively monitored section of the Natural Laboratory (work funded by the National Science Foundation-Grant No. 0934131 and EAR-0643240).
prohibitive to produce), and many algorithms are not designed

to address the added variability in the data. Additionally, with
higher resolution data there is a tendency to produce data for a
smaller spatial footprint of the landscape. In this case, the
characteristics of the feature under investigation may be
overwhelmed by roughness elements existing at a finer spatial

scale. Important coarse-scale patterns will be lost when small
footprints of the land surface are examined, and this ultimately
limits a researchers ability to make broader inferences and
connections about landscape or landform evolution.
136
A goal then is to come up with some general guiding

principles with regard to spatial-scale and DTM resolution
before conducting an analysis. An initial step to investigating
the spatial scale of a landform or landscape is through field
observation and field measurement/mapping. A DTM user
must know the spatial scale of the land-surface object(s) of
interest, as well as the operational scales of processes that are
altering and are modified by the object(s). The goal of this
endeavor is to develop a framework for understanding the
DTM resolution required to investigate the feature(s) or process(es). Once the observations have been made and measurements obtained, a good rule of thumb in resolution
selection is that, at a minimum, the user should select a DTM
whose resolution is half of the size of the smallest feature
under investigation (Maune et al., 2007). For example, a scientist concerned with quantifying gullies within a colluvial
hollow would need to have knowledge of the cross-sectional
dimensions of the gullies. If the gullies had an average width
of 1 m, then the user would want to use a square tessellation
with a cell resolution of 0.5 m or less in order to adequately
produce measures that will capture the form of the gully systems. Hengl (2006) presented a more quantitative approach to
determining a suitable grid resolution based on the complexity of the terrain. For sampled elevation data (e.g., contours and elevation), the density of inflection points on a land
surface can be used where the cell size (p) is at least half the
average spacing between the inflection points or finer:
pr
l
2 ndz
1
where l is the length of the transect and n(dz) is the number of

inflection points observed. If the DTM is based on digitized
contours, a suitable raster resolution can be estimated from
the total length of the contours. Inflection points are not
actually mapped; instead they are approximated from contours selected to present planform changes in the surface.
A suitable cell is defined as:
p
A
2 Sl
2
P
where A is the total size of the study area and l is the total
cumulative length of all the digitized contours. Hengl (2006)
also identified a more precise approach than equation [2] that
produces a 5% probability of the smallest spacing between
contours determined from the density of contours in a
given area.
The previous examples highlight how to determine the
representational scale and the horizontal cell resolution, but
one may also want to consider the vertical resolution of the
DTM. In the previous example of measuring the cross-sections
of gully systems, a depth measurement would need to be determined, and this value may be used independently or in
conjunction with the horizontal measures. Determining the
vertical resolution also requires a priori information about the
depth of the gully feature. Initial work in examining the vertical resolution requirements of DTMs began in fluvial
geomorphology with the analysis or determination of hydrologically accurate drainage networks. Gyasi-Agyei et al. (1995)
discovered that accurate hydrological networks could be
extracted from a DTM only if the ratio of average elevation

change per pixel (pixel relief or drop) to vertical resolution
(accuracy of elevation) of the DTM was greater than unity (a
ratio below unity was unacceptable for the development and
analysis of drainage networks). Average pixel drop is calculated from average slope and grid size, and vertical resolution
is measured by the error between topographic data and
benchmark elevation points. The main inference from this
work was that areas of lower relief required greater vertical
accuracy than areas in higher relief. McMaster (2002) designed
a new method to calculate hillslope length based largely on
the ratio test of Gyasi-Agyei et al. (1995) for DTM suitability,
and confirmed earlier qualitative work by Quinn et al. (1991)
and Zhang and Montgomery (1994) that the DTM resolution
must be greater than the average spatial scale of the hillslope
for accurate analyses.
Vertical resolution consideration has also been addressed
in the context of slope measurements. The ratio of the vertical
and horizontal resolution determines the minimum (nonzero) gradient measurement that can be determined from a
DTM (Gruber and Peckham, 2009). Results from this simple
ratio can be used to determine the lower bounds of surface
gradient on hillsides that can be computed with minimal
error. Vertical accuracy, independent of elevation and gradient,
was also shown to be an important measure in governing
horizontal resolution (Walker and Willgoose, 1999). The
vertical accuracy, however, must be consistent throughout
the DTM. If vertical accuracy is consistent, then topography in
the flattest locales of a watershed controls the horizontal
resolution (Walker and Willgoose, 1999).
Determining the spatial scale necessary for examining a
particular form or landscape is not always clear cut, despite the
above guidelines. This endeavor is further confounded by the
fact that each landform or landscape undergoes processes that
operate at different spatial and temporal scales. This complex
spacetime coupling of process(es) and form(s) is critical to
capturing information on landscape evolution (Lane and
Richards, 1997; Phillips, 1999a, b, 2006; Thoms and Parsons,
2003). A key to unraveling the dynamic conditions occurring
in many landscapes is to capture process imprints that are
recorded at different spatial scales within individual landforms (Schmidt and Andrew, 2005). In the previous example
of gully systems, only one spatial scale was considered.
Therefore, it may be necessary to consider using a finer resolution DTM to examine topographic changes at a variety of
spatial scales. This approach permits an assessment of differences that arise within and between landforms. One approach
might be to resample the DTM to the different resolutions to
produce these measures, but there is a concern of decreasing
the accuracy of the original high-resolution data by producing
coarser resolution land-surface measurements (Liu, 2008). An
alternative is to develop an approach that samples scalar differences to the same DTM resolution by measuring standardized values with an increasing window size (Volker et al.,
2007). Another option would be to consider measures from a
variety of different DTM sources with varying scales from fineto coarse-resolution. In fact, given the constraints of current
high-resolution DTMs (discussed more in the following
paragraph), integrating data from multiple sources is required
to capture the landscape and landform complexity associated
with hierarchical scales of the land-surface structure. There is

no simple approach to fusing data sets with varying scales.
Many concerns exist, but a fundamental concern stems from
the varying data quality associated with the different data sets.
At issue is how to assimilate the varying error budgets to
produce reliable results that are comparable from location to
location.
The aforementioned discussion highlights the inherent
spacetime hierarchical connections in landscapes, and the
need for terrain analyses to consider multiple spacetime
scales and multiple-resolution characterization. Space
time hierarchical connections are by no means novel to geomorphology. Kugler (1964) and Dikau (1989, 1990)
conceptualized distinct spacetime hierarchies for surficial features. These scales remain relevant to current geomorphological research (MacMillan and Shary, 2009). The
distinct spacetime hierarchical scales can also serve as a
means for considering the wide-range of DTM data sources
(Table 1). Horizontal and vertical resolutions have significant
impacts on how the landforms are portrayed (Shary et al.,
2002). Table 1 was developed with the general idea of putting

commonly used DTM data sources from geomorphological
studies into the context of fine, intermediate, and coarse
resolution. Fine, intermediate, and coarse scale classifications
were used to categorize the DTM data sources and, in turn,
capture the relevant hierarchical landform levels, as conceptualized by Dikau (1990), applicable for each of the DTM
sources. Table 1 demonstrates that many high-resolution data
sources are limited in terms of their applicability across all
scales, and highlights the previous point that to capture the
full range of scales across broad landscapes, there is a need to
consider integrating DTMs from multiple sources.
The ability to integrate multiple DTM sources is tempered
by a limited amount of research that has accounted for scale
dependencies within geomorphology. Phillips (2012) has
highlighted both operational problems (e.g., determining
conditions of scale dependence, what to feature or process to
represent at a given scale, and developing tools necessary to
extract and integrate the information and analyses) and theoretical issues (e.g., bottomup vs. topdown influences, and
Table 1 DTM data sources in the context of Dikaus (1990) range of scales associated with the land surface
Resolution class
Data source
DTM resolution
Accuracy
Landform
Fine-resolution
Close-range
photogrammetry
Terrestrial laser scanning
(TLS)
GPS-RTK surveying
Surveying
Soft-copy
photogrammetry
Orthophotograpy
Airborne laser swath
mapping (LiDAR)
Interferometric synthetic
aperture radar (IfSAR
or InSAR)
oMillimeter
No theoretical limitation
Pico-microrelief
Millimeter
Millimetercentimeter
Pico-microrelief
Centimeter
Centimeter
Meter
Centimeter
Millimetercentimeter
Decimeter
Pico-microrelief
Pico-microrelief
Micromegarelief
Meter
Meter
Centimeter
Decimeter
Micromegarelief
Micromegarelief
Meter
X-band
H1.252.5 m
V0.33.0 m
P-band
H2.04.0 m
V1.05.0 m
1/9 arcs
V1.0 m
1/3 arcs
V7.0 m
1 arcs
V7.015.0 m
H7.050.0 m
V7.050.0 m
Mesomegarelief
Intermediate-resolution
Coarse-resolution
137
National elevation
dataset (NED)
3.3 m, 10 m, and 30 m
Advanced space-borne
termal emission and
reflection radiometer
(ASTER)
SPOT
30 m
2030 m
SRTM X-Band
30 m
SRTM C-Band
90 m
Global 30 arcs elevation

data ser (GTOPO30)
The Global land 1 km
base elevation project
(GLOBE)
Mesomegarelief
Mesomegarelief
Mesomegarelief
1 km
H15.0 m
V10.0 m
H6.0 m
V10.0 m
H20.0 m
V16.0 m
Variable by location
1 km
Variable by location
Macromegarelief
Mesomegarelief
Mesomegarelief
Macromegarelief
138
multiple scale causality (see Phillips (2005) for more details)).

Future advancement of scale dependence will likely require
further advancements in data structures that account for
multiscale representation and analyses.
3.6.3.3
Data Structures
The land surface and its inherent complexity can be represented by a variety of data models (Bishop, 2012). As many
DTM are developed or analyzed in a GIS, some of the more
common data models include field (layer), entity (object),
and network data models (Goodchild, 1992). Each of the
aforementioned data models can also be linked with a relational data model. Continuous topographic data are commonly represented by a variety of different types of field
models. Traditional field models include sampled points,
contours, polygons, tessellations, and triangular nets. If discrete boundaries are present between landforms or features
within the landscape, vectors can be used to map objects.
Concerns pervade in the literature, however, with regard to
how field and object models can be effectively used when
processes and dynamism enter into the equation (Raper and
Livingstone, 1995). Despite these concerns, the aforementioned data models represent the most widely used and
currently accepted approaches to DTM representation.
The terminology related to digital-elevation data has become somewhat convoluted, as terrain data are used for a
variety of applications, and advances in terrain modeling and
analysis have evolved in a variety of fields. In the broadest
sense, all digital-elevation data can be considered as a digital
elevation model (DEM). A DEM is a collection of numerical
values that can be represented in a file or database to approximate topographic/bathymetric variability at given x- and
y-coordinates (Kidner and Smith, 1993; Kimerling et al.,
2009). Although the importance of bathymetric data to a
variety of research areas is recognized (including geomorphology), subaerial topography is the focus throughout this
chapter. The x- and y-coordinates may be expressed as geographic coordinates (e.g., latitude and longitude, UTM coordinates or other local coordinate system), or they might
simply represent a user-defined Cartesian coordinate system.
Topography is recorded as the z-value and is represented using
the function z f(x,y). A height value z is mapped to a 2D
location (x,y). Elevation represents a height above a vertical
datum (see Maune (2007) for definitions and explanations of
data and various height measures). Throughout this chapter,
elevation refers to orthometric height, which is the height
measured along a plumbline from the surface of a geoid
model to the Earths surface (Maune, 2007), unless otherwise
specified. A bare-earth DEM is devoid of vegetation and
human-built structures and herein is referred to as a DTM.
DTMs commonly contain a combination of ground and
water-surface information, and these two measures can also be
incorporated with bathymetric data. A digital surface model
(DSM) is a DEM that contains vegetation and human-built
structures, and this term is also adopted herein.
Each of the aforementioned terms contains the word
model. The term model evolved from an ability to use computers to automatically map and analyze topographic features
from a database containing x, y, and z values. Current sampling techniques used to capture topographic data are unable
to record all of the heights for a given locale. An inability to
record all spatial information regarding topography means
that the elevation data and any subsequent media produced
within a computer environment represent an approximation
of the land surface, hence the term model. DTMs are commonly represented as raster or triangulated irregular network
(TIN) data structures (Figure 3). More recent work has focused
on developing hybrid data structures that are a combination of
raster and TIN formats (Maune et al., 2007). Other data
structures can also be used to enhance information contained
within DTMs and these include contour lines, breaklines, and
mass points.
3.6.3.3.1
Raster (grid)
Raster files are the most common data structure used by

geomorphologists for terrain analysis. This simple data structure is based on a 2D matrix containing x and y coordinates,
which are gathered at regular intervals to produce a grid
pattern of square-shaped cells (Kidner and Smith, 1993;
El-Sheimy et al., 2005; Maune et al., 2007). Raster data are
commonly used in geomorphology because they require less
physical storage space. Many current software packages possess
algorithms to rapidly produce surface measures from raster
datasets. DTM mapping software stores data in the form of a
hierarchical tile-block structure that uses the grid cell as its
foundation (El-Sheimy et al., 2005). The process of representing data in this structure is called a tessellation and is
commonly represented as a square with similar x and y dimensions. Tessellation shape can vary, however, and the
choice of shape is dependent on a variety of other factors (e.g.,
distance calculations). Many other shapes have been used in
place of squares. The grid is then divided up into smaller
squares called tiles, which are divided up into blocks, and then
cells, creating a grid pattern. Breaking the surface down into
tessellations permits information about a specific cell or area
to be retrieved more efficiently, thereby reducing the processing time (El-Sheimy et al., 2005).
Each grid cell contains its own specific value and the value
only changes at the borders with an adjacent cell (Maune
et al., 2007). A smaller cell size represents a higher resolution,
creating a more detailed depiction of the surface area, compared to a larger cell size (El-Sheimy et al., 2005). A smaller
cell size, however, generates larger file sizes (Maune et al.,
2007). A smaller storage size is ideal as it saves time and
money to store and analyze the data (El-Shrimy et al., 2005).
Although compression techniques can be applied to help
lower the storage volume (Maune et al., 2007), there are no
generally accepted compression methods.
A main disadvantage of the raster data structure is that it
does not conform to the ever-changing characteristics of the
surface topography that reflect surface changes as well as internal structure of the landform or feature. Raster DTMs are
static cartographic representations, which would require a
large number of maps (data sets) to capture even the most
basic changes to surface topography. Even in the case of highresolution (small cell sizes) data sources, the data structure
does not represent numerous attributes of the topography.
This partially results from the fact that a raster file is only
Elevation (m)
14.376 16.342
12.41 14.376
10.444 12.41
8.478 10.444
6.512 8.478
4.546 6.512
2.58 4.546
0.613 2.58
1.353 0.613
139
2.5
10 meters
(a)
N
2 meters
(b)
Figure 3 Example of a raster file (top) and a TIN file (bottom). The images are from a stream reach that was scanned with a terrestrial laser
scanner. The raster file contains 1 1 cm cells. The TIN is a mesh constructed in Cyclone that has been decimated to enhance the triangles of
the TIN points (work funded by the National Science Foundation-Grant No. 0934131).
representative of individual cells or points that have been

captured at one particular time given a data collection technique (Kidner and Smith, 1993). In essence, not all of the
variations in the land surface are represented in the raster file
because of the inability to collect all of the surface topographic
expressions at a location given the limitations in the sensors
and current data collection approaches. Therefore, there is a
need to capture a large number of raster files to characterize
the temporal variation of the topography.
Raster files are best at representing the actual bare surface

of the Earth, because the data structure does not adequately
represent specific features like vegetation or buildings as well
as other data structures. This is also a limiting factor because
these other land-surface features are important for characterizing human alteration and land-use for better representation
and understanding of geomorphic change. Recent work using
voxels (3D pixels), however, has shown better results in
quantifying vegetation (Popescu and Zhao, 2008), as well as
140
quantifying and visualizing true 3D (surface and subsurface)

data (Hickin et al., 2009).
3.6.3.3.2
Triangulated irregular network (TIN)
Triangulation of mass points (a point cloud of x, y, z data) or

grids generally serves as the basis for generating TINs (ElSheimy et al., 2005). If triangulation is performed on a raster
file, a Triangular Regular Network (TRN) is generated, as opposed to irregularly distributed mass points, from which triangulation would generate a TIN (a series of contiguous
triangles of irregular size and shape). For TIN models mass
points are connected by nonoverlapping 3D lines, creating
planes of neighboring triangles (Figure 3). Triangles are
formed from any one of a wide variety of triangulation techniques. Delaunay triangulation (Kidner and Smith, 1993) has
been most commonly used in the geosciences because it meets
three basic requirements for TIN formation (Li et al., 2005):
(1) the resulting TIN from any set of points should be identical if the same algorithm is used, regardless of the starting
point of the algorithm; (2) each triangle should be nearly
equilateral; and (3) nearest-neighbor points must be used to
generate each triangle. Furthermore, the circumscribing circle
of each triangle would not include any other surrounding
points than the three nearest neighbors, hence no other vertex
exists (Kidner and Smith, 1993; Li et al., 2005).
Unlike grids, TINs are adaptable to topographic changes
that are generally associated with the heterogeneity of many
land surfaces. Therefore, TINs are a more reliable source when
it comes to specific surface details (Kidner and Smith, 1993).
For example, the triangles of a TIN define topography by
representing features (vectors) as points, lines, or polygons
(Kidner and Smith, 1993). Points representing the angle of
each triangle are called nodes, and an edge is a line connecting
two nodes (El-Sheimy et al., 2005). As a result, a TIN represents a more complex data structure. This complexity does
have advantages, as TINs can be used to characterize linear
features such as coast-lines and rivers more accurately than
grids. TINs also require fewer points than raster files to represent the land surface (El-Shrimy et al., 2005). Several
problems, however, arise when using the TIN data structure.
One problem is that the TIN size (data volume) increases
exponentially with the complexity of the terrain it represents,
which results from the data storage approach required to
maintain triangles in a TIN (Maune et al., 2007). TIN databases store the topological relations of the triangular data.
Information is stored in a clockwise list of all the neighboring
edges that produces an efficient means of accessing data. A
single triangle uses the equivalent of 10 raster cells of storage
space (Kidner and Smith, 1993).
3.6.3.3.3
Contour lines
One of the most common techniques to represent topography

on a 2D map is the use of contour lines (Figure 4). Contour
lines are one of several forms of isoline maps that record equal
lines of elevation at a specified interval (National Digital
Elevation Program (NDEP), 2004; Kimerling et al., 2009).
Contours were initially derived directly from aerial photography using a photogrammetric stereoplotter, but can also be
interpolated from point data. Spatial interpolation in this
manner results in a less accurate representation of the land
surface (El-Sheimy et al., 2005). Contour lines are commonly

used on topographic maps to reveal features in the landscape
such as mountains, ridges, and valleys. They represent a reasonable approach for representing and visualizing the land
surface, because consistent rules are used to produce contours.
For example, contours will never intersect one another, which
produce a contiguous visual surface to the reader. The contour
interval (vertical distance between contours) also aids in rapid
visual connections to the mapped surface, as closely spaced
contours represent steeper terrain. Contour maps have also
played a critical role in past geomorphologic research where
profiles and cross-sections were used to quantify form (Hack,
1957; Rapp and Nyberg, 1981).
3.6.3.3.4
Breaklines
Land-surface complexity can be represented as variations in

the local relief of a landscape or landform. This might exist as
a break-in-slope, as the landscape transitions from a gradual
slope to a steep cliff. Breaklines are used to assure that dramatic discontinuities are maintained, as terrain conditions can
dramatically change (National Digital Elevation Program
(NDEP), 2004). Two types of breaklines are commonly utilized and include hard and soft breaklines. Hard breaklines
define abrupt disruptions in the continuity and evenness of
surface topography (e.g., rivers, shorelines, dams, and buildings). They are commonly illustrated in 3D, but can also be
illustrated in 2D. This is because there are a number of features that are only defined by x and y points. Soft breaklines
are used to make certain that elevations are accurately
represented along linear features such as roads or pipelines
(National Digital Elevation Program (NDEP), 2004). Breaklines have many important roles in the development of DTMs,
as they force a data structure to adhere to the heterogeneity of
the land surface (Maune et al., 2007). This would include
transitions to flatter surfaces (e.g., lake surface) or sharp
breaks-in-slope such as from the flooplain to a channel. If
breaklines were not present, many of these features would be
lost or generalized in the final DTM.
3.6.3.3.5
Mass points
Mass points represent topography as x, y, and z values at

random locations over a land surface (National Digital Elevation Program (NDEP), 2004). Each point represents a slight
change in the morphology of the terrain. The use of mass
points, however, is not a completely accurate representation of
the surface, because the spaces in between the points are not
taken into account (El-Shrimy et al., 2005). Recent developments in ALS and TLS provide dense point clouds (often
oversampling a site) whereby over small areas mass points can
be used to accurately record the morphometry of a feature and
the morphometric changes through time (Wasklewicz and
Hattanji, 2009). Dense mass points also provide information
on the location of vegetation and other nontopographic objects. The removal of these objects can be accomplished by
knowing a specific location and extracting the feature, or by
applying any one of a variety of data reduction approaches to
the mass points.
141
5 m contours
N
12.5
25
50 meters
Figure 4 A hillshade map overlain by contours. The area is a small, recently burned basin near Goleta, CA. The hillshade was developed from a
1 cm DTM and the contours represent relative elevation in meters (work funded by the National Geographic Society-CRE Grant #8590-08).
3.6.4
Data Sources
A wide variety of DTM data sources exists, and many of these

have been covered in detail in a variety of books and review
articles. Here, the authors focus on four DTM data sources that
have been recently adopted in geomorphology. In particular,
the authors focus on TLS and terrestrial photogrammetry
techniques (TPT), ALS, and interferometric synthetic aperture
radar (InSAR). The basic principles associated with each are
presented.
3.6.4.1
Terrestrial Laser Scanning Techniques
At present there are two types of TLS instruments in use: static

and dynamic. Static scanners are stationary and are generally
mounted on a surveying tripod or mast (Figure 5). Dynamic
scanners are vehicle-mounted and motion-compensated
through the use of an inertial measuring unit (controls roll,
pitch, and yaw) and a GPS to record the location of the
scanner during the sampling campaign. A majority of the TLS
review will focus on the static scanners, as they are more frequently used in geomorphology, although usage of motioncompensated scanners is growing in coastal geomorphology.
For more detailed information see Petrie and Toth (2009a, b).
Static TLS can be classified in a number of ways, but the

authors focus on a general spatial categorization of short-,
medium-, and long-range scanners. Short-range scanners are
generally limited to less than 100 m maximum scanning
range, but commonly the practical limits are 4060 m. A vast
majority of the short-range scanners are suitable for indoor
and outdoor use. Short-range scanners frequently employ
phase-measuring methods with a Class 3R continuous wave
laser that operates in the near-infrared part of the spectrum at
780790 nm. A few manufacturers employ pulse-ranging
methods using Class R lasers that emit the pulsed light at
658 nm. Short-range scanners with phase-measuring capabilities can possess a manufacturers specified accuracy of
36 mm and can record up to 100 000200 000 points
per second. Many of these scanners record point-cloud data to
internal hard-drives or to a computer via Ethernet or FireWire
connections. Medium-range scanners employ pulsed-ranging
methods and can measure objects at maximum distances
ranging from 150 to 350 m dependent on the objects surface
reflectivity (Figure 5). These scanners also use a Class 3R at
350 nm. This style of scanner collects data at the 500050 000
points per second range at a manufacturers specified accuracy
of 36 mm at 50 m. Long-range scanners also use pulseranging techniques and can measure up to distances of
142
sinusoidal wave pattern of the transmitted and received light

signals. Phase measurements are generally performed using a
digital pulse-counting technique, which records a fractional
part of the total distance (Dl). The amplitude of the laser
radiation is modulated with a sinusoidal signal and by changing the modulation pattern; the integer number of the
wavelength can be determined and added to the fractional
distance to provide a slant range, such that
R Ml Dl=2,
Figure 5 Terrestrial-laser scanner set-up in the Front Range of the

Rocky Mountains, CO (work funded by the National Science
Foundation Grant No. 0502343 and 0239749).
500 m1.5 km. Although these scanners are considered longrange, they do possess the capability to measure objects as
close as 3 m. A caveat with long-range scanners is that in order
to attain long-distance measurements, the object must be
highly reflective. A majority of the long-range scanners use a
Class 1 laser rangefinder emitting its infrared radiation at
9051550 nm. However, some also use a Class R laser at
905 nm. In general, this range of TLS records 400012 000
points per second with accuracies in the range of 57 mm at
100 m.
Recording topographic data with TLS is conducted in a
similar manner regardless of the scanners range. All of these
instruments are using laser-based ranging to accurately record
distance. This is accomplished by measuring the time-of-flight
(TOF) of the laser light or the phase variation between the
transmitted and received signal. A TOF instrument emits a
series of short bursts of light from the laser ranger to the object
and back to the instrument. The distance is resolved by accurately knowing the speed of light through the following
equation:
R v t=2,
3
where R is the slant range, v is the speed of the electromagnetic

radiation, and t is the measured time interval. The second
ranging option is referred to as phase-measuring, where the
laser transmits a continuous beam of radiation instead of a
pulse. Range is measured via the phase difference of the
4
where M is the integer number of wavelengths, l is the known

value of the wavelength, Dl is the fractional part of the
wavelength (j/2p) l, and j is the phase angle.
In addition to distance measurements, each instrument
also possesses a scanning mechanism (e.g., rotating mirror or
prism) that permits vertical measurements and a motor drive
or robotic servo to capture horizontal (azimuthal) direction.
In essence, the scanner measures a series of profiles around the
scanner.
A majority of the TLS instruments are also single return.
That is, once an object reflects radiation from the pulsed or
phase-based laser light, any object behind that object is not
recorded. This can produce shadowing in the point cloud
(Figure 6). Shadowing can be reduced or removed by moving
the TLS to a position that will fill-in the data void. This approach, however, produces multiple point clouds that must be
merged into a single point cloud.
Development of a single point cloud, or what is termed as
registration of the point cloud, commonly requires a minimum of three common points between each cloud. The three
common points can be selected manually within a manufacturers registration program. This approach is often used in
built environments where corners of structures and other
known points are quite easily picked. This is often difficult to
apply, however, in natural topographic settings that are most
frequently used in DTM analyses. Another alternative is to use
a series of control points established before scanning a location (Figure 7). These control points can be referenced to
real-world coordinates via any number of surveying techniques, or they can just represent fixed points in 3D Cartesian
coordinate space, which has an origin at the point where the
laser is emitted from the scanner (points are recorded as
above, below, left, and right of this point of origin). Control
points are used to set-up targets or objects of known dimensions or volumes. The target and objects of known dimensions
are highly reflective, so they produce quality laser returns
during the scanning process. The targets or objects of known
dimensions serve as a surface from which a series of vertices
are established in the manufacturers software or after-market
software. Each vertex represents a common point that can be
used to accurately link together multiple point clouds
(Figure 7). The number of targets or objects in the study site is
generally dependent on the complexity of the terrain. In
general, the more complex the terrain, the more targets or
objects will be required. Target numbers also increase with
increasing area, especially when short- and medium-range
scanners are employed in the data acquisition process. Both
manual and control-point registration approaches produce a
registration-error budget. Registration-error budgets must also
be considered in the context of ground measurements, objects

placed in the field-of-view, or spot elevation checks from other
accurately defined positions. Targets or objects of known dimension have been proven to provide the most accurate and
rapid way to register the data (Hetherington et al., 2005;
Heritage and Hetherington, 2007; Wasklewicz and Hattanji,
2009).
Figure 6 An image of a TLS point cloud collected from one of the

2009 USGS Landslide Hazards Programs intensive research site in
the Station Fire located in the San Gabriel Mountains. Note the
shadowing (data void) in the gullies from a single scan location (xyz
indicator in the upper right-hand corner). The shadowing on the
right-hand side of the gullies is a result of topographic complexity,
scanner location, and the single-return nature of the scanner. The
shadows were filled by scanning from other locations within the
basin (work funded by the National Science Foundation-Grant No.
0934131).
3.6.4.2
143
Terrestrial Photogrammetry Techniques (TPT)
Photogrammetry, the extraction of quantitative 3D data from

stereo-imagery, is an important tool for the collection of DTM
data (Lane et al., 2000). Traditionally, photogrammetric data
of the Earths surface have been derived from airborne imagery
sources, where the camera is oriented vertically in relation to
the ground surface, and there is sufficient overlap (typically
60%) between sequential photographs to replicate the optical
parallax needed to view images in three dimensions (Matthews, 2008). More recently, advancements in camera technology and photogrammetric software have increased the
popularity of TPT (Chandler, 1999; Chandler et al., 2005;
Matthews, 2008). TPT permits the development of highresolution DTMs for geomorphic analysis using digital cameras and specialized photogrammetric software. Conducted
properly, this complex process allows high-resolution DTMs to
be created from vertical or oblique images of the ground
surface (Clarke, 1995; Lane et al., 2001; Redmondino and
El-Hakim, 2006; Matthews, 2008). DTM derivation from TPT
requires careful consideration of three phases of the process
including: (1) image acquisition; (2) stereo-matching of
photographs; and (3) manual or automated DTM generation.
Image acquisition refers to the process of camera selection,
calibration, and the recording of the images. Traditional
photogrammetry techniques relied on metric cameras for the
acquisition of images. Metric cameras are more easily implemented in photogrammetric work, as the geometry of the
camera is stable, the coordinate system is included in the
images as fiducial marks, and the lens distortion is well defined (Peipe and Stephani, 2003). Although these highly
Figure 7 Terrestrial-laser scanner targets, as seen from a digital photograph on the left. The right image is the same targets (t7, t8) as seen in
a point cloud with colors from scanner gathered with a terrestrial laser scanner (work funded by the National Science Foundation-Grant No.
0934131).
144
accurate cameras are ideal for this type of data collection, the
camera costs are prohibitive to most researchers. In recent
years, geoscientists have begun to rely on consumer-grade
nonmetric digital cameras to acquire images for photogrammetry and DTM creation (Chandler et al., 2005).
Since the mid-1990s, the cost of consumer-grade digital
cameras has decreased dramatically whereas sensor and lens
quality have increased. Chandler et al. (2005) compared the
photogrammetric capabilities of low-cost digital cameras. They
found that these types of cameras are suitable for the production of high-quality DTMs provided that errors associated
with lens imperfections are considered (Chandler et al., 2005;
Peipe and Tecklenburg, 2006). The process of determining
camera parameters and lens distortions is known as camera
calibration (Fraser, 1997; Chandler et al., 2005; Remondino
and Fraser, 2006; Peipe and Tecklenburg, 2006). Stand-alone
software is available that allows users to calibrate their cameras and lenses. These self-calibration bundle adjustments
(SCBA) provide users with distortion and camera parameters
for digital photogrammetry systems (Fraser, 1997; Peipe and
Tecklenburg, 2006). In addition, many photogrammetric
software packages now include SCBAs (Chandler et al., 2005;
Matthews, 2008; Sturzenegger and Stead, 2009). Once cameras have been properly calibrated, images of the research area
need to be acquired. Although there is some flexibility in the
methods used to take pictures, extracting the best possible
DTM from photogrammetric techniques requires careful attention to internal and external control, camera position, and
image processing.
Internal control of a TPT study area refers to the insertion
of objects of known length into the acquired images. These
objects are used, in combination with the lens parameters
obtained from camera calibration, to produce a highly accurate and precise local coordinate system (Matthews, 2008).
Local coordinate systems can be projected to geographic coordinates by surveying control objects using a total station or
GPS survey, though these methods may degrade the relative
accuracy of the photogrammetrically derived DTM.
TPT images can be obtained from strategic vantage points
at ground level, at elevated positions (e.g., from ladders or
platforms), or from low-altitude airborne positions, such as
unmanned aerial vehicles (UAVs). Regardless of the platform
from which images are acquired, the derivation of highly accurate DTMs requires a minimum overlap of 60% between
images and a camera angle as close to nadir (perpendicular to
the object of interest) as possible (Matthews, 2008). Where
possible, the use of a tripod or some other stable base is recommended (Haneberg, 2008). In addition, the axes for each
camera image must not cross (i.e., converge) between consecutive images or the software may not be able to obtain a
stereographic model.
Once the study site is photographed, photogrammetric
software is used to produce DTMs. This process requires three
steps: pixel matching, feature matching, and surface creation
(Fraser and Cronk, 2009). Pixel matching refers to the process
by which similar points are identified in multiple images.
These points may be obtained from both control objects in the
scene, or from points with other distinct characteristics. Feature matching follows the same process, but matches similar
areas between images. Both the pixel- and feature-matching
processes are automated in most current generation photogrammetric software (Haneberg, 2008; Matthews, 2008). Once
pixels and features have been matched, DTMs are automatically created by triangulation (for matched points and areas)
and interpolation (unmatched points and areas) (Brasington
and Smart, 2003).
3.6.4.3
Airborne Laser Scanning
ALS is conducted from either a fixed-wing aircraft or a helicopter. A basic ALS system is composed of a laser-ranging unit
(mostly pulsed time-of-flight measurement systems) that includes a transmitter and receiver optics. This is mounted over a
hole in the fuselage of an aircraft. There is also an opticalscanning mechanism, such as a rotating mirror (with angular
encoder) that directs the pulsed-laser light in a cross-track
direction at known angles. It should be mentioned that there
are other types of optical-scanning mechanisms that use dual
mirrors, a spinning optical polygon, a nutating (rocking,
swaying, or nodding motion) mirror producing an ellipticalscan pattern (called a Palmer scan), or a pair of linear fiberoptics arrays. The forward motion of the plane allows a series
of ranges to be gathered along the flight line. The combination
of a series of ranges along successive lines and angular measurements permit large swaths of the land-surface elevation to
be recorded. When overlapping swaths are combined, the
measurement densities are often in the range 0.2 to 50 points
per meter square. The range of values depends on the aircraft
height and velocity (Beraldin et al., 2010).
An airborne GPS antenna is mounted on top of the aircraft
and data are differentially corrected with a ground-based reference GPS. Ground-based reference information can be
provided from a GPS receiver operating at a base station, a
wide-area augmentation system, or satellite-based augmentation system (e.g., the Japanese Multifunctional Satellite
Augmentation System or commercial systems such as StarFire
or OmniSTAR), continuously operation reference stations
(CORS), or by employing precise point positioning (PPP)
technology. An inertial measurement unit (IMU) is used to
record acceleration data and orientation of the aircraft (Petrie
and Toth, 2009c). A typical sampling rate for the IMU is
B200 Hz. Data from the GPS and IMU are combined to reconstruct the air trajectory of the aircraft to an accuracy of
o10 cm. The scanner, GPS, and IMU are synchronized by an
onboard control and data-recording unit. The synchronization
is keyed by the GPS pulse-per-second signal. This signal triggers the internal clocks in the IMU and scanner to provide a
time stamp for each range measurement and trajectory positions (Beraldin et al., 2010).
Many commercial ALS systems operate at wavelengths of
8001550 nm. The wavelength is important because different
wavelengths will perform differently depending on the object
being recorded on the land surface. For example, in alpine
settings where snow or glacial ice is present, it would be
preferable to use an ALS system that makes use of near infrared wavelengths (8001300 nm), as snow and ice have low
reflectivity at 1550 nm (Beraldin et al., 2010). Laser-pulse
repetition rates can be measured at up to 150 Khz, but many
ALS systems have a scan rate in the 3050 Khz range.
ALS systems have the ability to record multiple echos from

a single pulse. For example, in a forested ecosystem, the beam
is wide enough that it could hit the edge of a leaf or branch in
the upper canopy and the remaining portion of the beam
would continue downward either striking the ground or
additional branches or leaves. The elevation at the top of the
canopy is called the first return, whereas the last return could
potentially be the ground. The terms terrain (ground) and offterrain (above the ground) are also commonly used terms to
classify the multiple echoes. Thus, a coarse (in the sense that it
is not recording every branch and leave) DSM could be created
incorporating the multiple echoes (in earlier ALS systems this
ranged from 35 echoes). Alternatively, if a data-reduction
method was employed to only consider the last returns, a
DTM could be developed from this same dataset.
Recent developments (B2004 to present) in ALS technology have allowed the adoption of small footprint (recording a
diameter of up to 100 cm) ALS systems that record the full
waveform (FW) of a laser pulse. FW-scanners permit the
complete waveform of each backscattered pulse to be digitized. A major benefit of this approach to ALS is that it provides
the user with the ability to improve the classification of terrain
and off-terrain objects. The major advantages of FW ALS sensors to geomorphological research are the production of
more accurate DTMs, an ability to consider vegetation as a
roughness component, and the ability to measure the spatial
extent or spread of vegetation (Devereux and Amable, 2009).
Classification of terrain and off-terrain points, however,
is reliant on proper segmentation or filtering of the FW data.
There has not been a great deal of application of FW to
date within geomorphology because of the novelty of the
technology.
After each ALS surveying mission significant data processing is required to reach a final product. At the heart of this
processing is the merging of the three datasets, GPS ground
data, navigation data (GPS/IMU), and the ranging data, from
each mission. A variety of other information also includes logs
of aircraft behavior, instrument behavior, and recording errors
that may be included at some point in the processing step
(Devereux and Amable, 2009). These data streams are integrated into a database.
Systematic errors also are commonly associated with ALS
instrumentation. Several conditions produced during operation of a light aircraft lead to systematic errors and they include: high vibration; variable pressure and temperature;
impacts from rough landings; changing orientation; and inconsistent electrical supply. Each of the aforementioned conditions is a major factor leading to varying degrees of
systematic error (Shrestha et al., 2007). Furthermore, the
natural aging process of the ALS system can also cause variations in measurements over time, and this is particularly
important to spacetime measurements in geomorphology.
The NSF funded National Center for Airborne Laser Mapping
(NCALM) has presented a number of scenarios by which the
calibration processes varies from mission to mission (Shrestha
et al., 2007), and Katzenbeisser (2003) has also provided detailed examples of calibration issues and techniques. One
common approach, used by NCALM has been point-by-point
comparisons of kinematic GPS data at a permanent test location (100010 000 points). Before each extended mission or
145
when the ALS instrument is deployed, field crews establish

kinematic GPS lines on hard surfaces such as roads and land
strips. These point-to-point comparisons are used to accurately
identify the zero point and scale of the scanner to within a few
centimeters (Shrestha et al., 2007). Vertical accuracies of ALS
derived data range from 710 to 715 cm, but this is
dependent on careful calibration and processing. Horizontal
accuracies are typically worse than the vertical accuracies and
range from 15 to 100 cm (Baltsavias, 1999; Maas, 2002;
Maune, 2007).
3.6.4.4
Interferometric Synthetic Aperture Radar
Synthetic aperture radar (SAR) refers to a rapidly evolving

form of radar data processing that allows for the characterization of topography using airborne or spacecraft-mounted
radar systems (Gabriel et al., 1989; Gens and Van Genderen,
1996; Massonnet and Feigl, 1998; Burgmann et al., 2000;
Rosen et al., 2000; Smith, 2002; Zhou et al., 2009). SAR systems use an antenna to produce a narrow band of electromagnetic energy directed at the surface. SAR systems both
transmit and receive electromagnetic waves at wavelengths
commonly ranging from 3 cm (X band) to 24 cm (C band)
(Massonnet and Feigl, 1998), although, shorter wavelength
sensors in the Ku band (roughly 1.7 cm) have been recently
used (Okada et al., 2007). At these wavelengths, the energy is
able to travel through the atmosphere with minimal degradation in all weather conditions (Okada et al., 2007). The phase
(time delay) and amplitude (energy intensity) of the radar
echoes are then used to derive both topographic and positional data from the SAR image, as well as information about
the surface (Smith, 2002; Zhou et al., 2009).
Interferometric SAR (InSAR) refers to the process of using
information from two or more SAR images of the same location to improve the accuracy of elevation data or calculate
changes in surface elevation over the study area. The product
of the two images is known as an interferogram. The interferogram can be produced using: (1) two (or more) SAR antennae mounted perpendicular to the flight path or orbit on
the same platform (single pass InSAR); or (2) the same SAR
antenna during sequential flight paths or orbits (multipass
InSAR) (Gabriel et al., 1989; Smith, 2002; Zhou et al., 2009).
The length of time between images used to create the interferogram, referred to as the temporal baseline, may range from
o1 s (for single pass InSAR) to days, months or years (for
multipass InSAR) (Zhou et al., 2009). The temporal baseline
for multipass is generally dictated by the orbital frequency
of the satellite or the time-scale of interest for the particular
application. Readers seeking a more thorough review of InSAR
theory and methods are referred to Gabriel et al. (1989),
Gens and Van Genderen (1996), Massonnet and Feigl (1998),
Rosen et al. (2000), Smith (2002), and Zhou et al. (2009).
Depending on aircraft height, sensor size and atmospheric
conditions, InSAR is able to produce DTMs ranging in resolution from B30 m (Roering et al., 2009) to less than 1 m
(Okada et al., 2007). The ground resolution of the pixel is
dependent on numerous factors, including the wavelength of
radiation emitted by the sensor, the distance of the aircraft or
satellite from the surface, and the dielectric properties of the
146
surface (e.g., the amount of scatter and absorption) (Zhou

et al., 2009). DEMs produced from InSAR are generally of
inferior quality when compared to those derived from ALSM
or TLS methods. The higher level of inaccuracy associated with
a single set of topographic measurements from InSAR is related to the nature of satellite navigation and tracking, errors
in the estimation of interferometric phase and atmospheric
conditions (e.g., water vapor content) (Smith, 2002). These
factors combine to produce elevation measurements that may
contain uncertainties on the order of meters (Smith, 2002).
The strength of InSAR data lies in its ability to detect finescale changes in surface elevation or characteristics associated
with both natural and anthropogenic processes (Gabriel et al.,
1989; Smith, 2002; Zhou et al., 2009). A temporal sequence of
data collection allows for the differencing of interferograms
and calculation of phase changes in the spatial or dialectic
properties of the surface (Zhou et al., 2009). The accuracy of
such an analysis is highly dependent on the proper selection
of a control area, where surface deformation and dielectric
changes have not occurred between image acquisitions.
A perfect control area would be topographically homogenous,
low gradient and stable. In addition, optimum quantification
of the differences in phase between images would be achieved
when the sensor is in the same orbital configuration and observational direction, and all postprocessing was conducted
using the same procedures (Zhou et al., 2009). Differences in
phase recorded at the control area are then used to correct for
the interferogram so that differences in dielectric properties
and surface deformation may be calculated (Gabriel et al.,
1989; Zhou et al., 2009).
InSAR data have been used in a wide variety of applications
in geomorphology. Although these data were originally used
to produce models of surface topography at a single point in
time (Graham, 1974; Zebker and Goldstein, 1986), the recent
trend has been to analyze phase differences in the interferograms to quantify landscape change. Dielectric differences
related to changes in soil moisture content (Nolan et al.,
2009), depth to bedrock (Elsherbini and Sarabandi, 2010),
wetland analysis (Hong et al., 2010), and forest canopy

structure (Kellndorfer et al., 2010; Solberg et al., 2010) are all
recent examples of analysis of dielectric differences in the
interferogram. Phase differences have also been used to
quantify surface deformations (changes in elevation) associated with both natural and anthropogenic processes. Recent
examples include deformation associated landsliding and
sediment transport (Catani et al., 2005; Roering et al., 2009;
Lauknes et al., 2010), tectonic activity (Jo et al., 2010; Poland,
2010; Wen and Ma, 2010; Xu et al., 2010), soil consolidation
(Kim et al., 2010), glacial dynamics (Capps et al., 2010; Shugar
et al., 2010) and mine collapse (Plattner et al., 2010).
3.6.5
Preprocessing
Ground and off-terrain data (e.g., vegetation and components

of the built environment) are simultaneously gathered as x, y,
z data in a point cloud (Figure 8). This may not be a problem
in studies employing a DSM, but most geomorphological
studies require bare-earth models. The issue of classifying
ground and off-terrain points must be considered at the onset
of preprocessing the DTM source data. Of particular importance is the removal of off-terrain points from the point cloud
before DTM production. Removal of off-terrain points is a
form of data reduction and is a critical step to producing a
hydrologically sound DTM. Even after removal of the offterrain points, the point cloud may remain exceedingly large,
and further reduction of ground points might be required to
permit data analysis or representation to occur in an efficient
manner. Larger point clouds might also be reduced regardless
of vegetation removal. Extracting ground data from and reducing data from point clouds is an operational activity, but it
is particularly important for geomorphologists to understand
these processes to either critically evaluate the already preprocessed data or to determine a starting point for preprocessing on her/his own.
Y
Y X
0.5
Figure 8 Ground (brown) and off-terrain (highlighted in green) data collected with a TLS. The image is from a point cloud with a picture
superimposed over the points (work funded by the National Science Foundation-Grant No. 0934131).

3.6.5.1
Point-Cloud Reduction Techniques
Reducing the complexity of point clouds is not a trivial matter.

Pfeifer and Mandlburger (2009) conveyed a number of reasons why this is the case: (1) the geometrical structure of
objects located only slightly above the ground surface can be
very similar to topographic features, making it quite difficult
to differentiate between the two; (2) other attribute information is commonly recorded with the locational data (e.g.,
echo number, intensity, rgb values from digital imagery, or
echo-width extracted from full-waveform) and cannot be used
to classify off-ground points because items such as last echoes
from multiple echoes (in ALS these represent multiple return
data classifiers) can contain multipath errors that would place
the point well below the land surface; (3) it is difficult to code
and incorporate an algorithm that accounts for contextual
knowledge (a common form of human interpretation) into a
feature-extraction program; and (4) several data-reduction
techniques are highly influenced by systematic and random
measurement errors (Mandlburger et al., 2008). At a minimum, data quality must be determined from both field
measurements and from information gathered from the
digital data set. Once this is determined, errors must be removed or minimized through statistical or modeling approaches to proceed forward with the data reduction
component. Each of these steps increases the potential for
more data storage as well as increasing the labor and computational time involved in the data reduction process. Three
types of data reduction techniques are highlighted and include
decimation, filtering, and segmentation.
3.6.5.1.1
Decimation
A variety of fields such as geodesy, computer vision, computer

graphics, computer science, and applied engineers are developing techniques that thin or reduce the number of points
used in the production of DTMs. The thinning of data is
commonly termed decimation. Decimation can be accomplished at a variety of different levels, and involves the reduction of the original point cloud via point removal,
refinement, or cloud segmentation. Point-based simplification
is more rapid and less demanding of computer memory because unlike meshes (another term for a type of TIN) topological and geometric features are not preserved (Dey et al.,
2001; Moenning and Dodgson, 2003). The greater efficiency
afforded point-based approaches indicates a potential for
further advances in the near future. A more detailed review can
be found in Pauly et al. (2002).
Three main types of point-based simplification occur: (1)
clustering; (2) particle simulation; and (3) iterative simplification methods (Pauly et al., 2002). Clustering is an approach
that splits the point clouds into subsets and replaces them
with representative points (Rossignac and Borrel, 1993;
Brodsky and Watson, 2000; Shaffer and Garland, 2001; Pauly
et al., 2002). Particle simulation uses interparticle repeling
forces to compute sampling positions based on a desired
number of points, but have not been widely applied (Turk,
1992). The iterative simplification method iteratively reduces
the number of points along a local plane or surface by removing the least important points at each of the iterations
(Garland and Heckbert, 1997; Alexa et al., 2001; Linsen, 2001;
147
Fleishman et al., 2003). Iterative simplification is the most

promising approach for geomorphology because it relies on
point extraction based on curvature, distance between a point
and a specified target, and locality within a neighborhood as a
means of defining the points to remove. This approach should
only be used in the case where all of the systematic errors in
the point cloud have been removed through sensor calibration. Random errors should also be removed by an interpolation strategy with measurement noise filtering.
Despite the promise of this technique, concern exists at
present with the current algorithms used to decimate point
clouds. Mandlburger et al. (2008) suggested not using decimation methods for large high-resolution DTM data because
they work from a fine-to-coarse, which involves triangulation
of the entire point cloud. This is computationally complex and
some GISs would not be able to handle the number of triangles required to conduct this approach. Therefore, decimation is currently feasible for high-resolution DTMs with
small spatial coverage, as has been applied by Dupuis et al.
(2006). Decimation was used to develop small, but accurate
TIN data for use with autonomous long-range rover navigation. Dupuis et al. (2006) particular application required
only a narrow DTM (small area), measuring slightly larger
than the width of the rover and was therefore, computationally feasible.
Refinement methods represent multipass algorithms that
resample from coarse-to-fine resolution starting with a minimal initial approximation (Heckbert and Garland, 1997;
Pfeifer and Mandlburger, 2009). Each subsequent pass incorporates one or more additional points as vertices for triangulation until the desired tolerance is achieved (Pfeifer and
Mandlburger, 2009). Mandlburger et al. (2008) presented a
modified refinement approach using a hybrid DTM (regular
grid and breaklines, structure lines, and spot heights). A
maximum height tolerance and a maximum planimetric point
distance were used as the basic parameters for this data thinning approach. An initial approximation of the DEM is generated using Constrained Delaunay Triangulation, and each
cell is subsequently refined by iteratively inserting additional
points until the height tolerance is achieved. Mandlburger
et al. (2008) reduced the original data set to B8% of its original size. A comparison of this simplified data set with lowresolution hydrological modeling found similarities in terms
of water height, but drastic differences in a variety of other
hydrological measurements (e.g., flooded area and flow velocity). Refinement and feature approaches, in conjunction
with calibrated and corrected high-resolution DTMs, are important to geomorphological research where data reduction is
needed. Furthermore, Mandlburger et al. (2008) made a
strong case for the need of high-resolution DTMs in hydrology
research.
3.6.5.1.2
Filtering
Filtering is a process designed to extract terrain points from

off-terrain points that commonly occur in many DTM data
sources. A vast majority of the ground-filtering techniques for
point clouds assume gradual variations in topography (e.g.,
smooth variation in slope from point to point). Therefore,
terrain points are detected through a combination of slope,
elevation difference, and local elevations (Vosselman, 2000;
148
Zhang and Whitman, 2005). The more complex the topography, however, the more difficult this process becomes, and
is further complicated by other natural and human-made
objects on the surface. Silvan-Cardenas and Wang (2006),
subsequently reviewed by Liu (2008) and Pfeifer and Mandlburger (2009), have identified three generally accepted
classes of filtering processes: interpolation-based (Kraus and
Pfeifer, 1998), slope-based (Vosselman, 2000; Roggero, 2001;
Sithole, 2001; Shan and Sampath, 2005), and morphological
(Zhang et al., 2003; Chen et al., 2007).
Interpolation-based filters use weighted linear least-squares
interpolation to iteratively estimate the topography under
consideration (Chen et al., 2007). Residuals from the approximated surface are assigned to the elevation values
(negative for terrain points and positive for off-terrain points).
Negative residuals are more heavily weighted, whereas positive
residuals are weighted less. Residuals with higher weights are
kept in the iterative process until the surface gets closer and
closer to the ground (Crosilla et al., 2004). This technique
does not work well in steep topography, but Lee and Younan
(2003) developed a combined modified linear-prediction filter and an adaptive filter to produce better results.
Slope-based point-cloud filters assume distinct differences
between the slope of the topography and the slope of an offterrain object (Zhang et al., 2003). If the maximum value of
the slope between a point and any other point for a given
neighborhood is less than a threshold value, it is classified as a
terrain point (Vosselman, 2000). Threshold development is
highly dependent on knowledge of the topographic variability
in a study area. Slope-based filters perform well in gentle
topography, but accuracy declines as the slope of the terrain
increases (Sithole and Vosselman, 2004). Sithole (2001) improved the ability to produce accurate data in steep topography by developing a varying threshold with slope.
Morphological filters are derived from mathematical
morphology (Harlick and Shapiro, 1992). These point-cloud
filters are based on an admissible height difference that is a
function of planimetric distance (Pfeifer and Mandlburger,
2009). Points are identified as off-terrain if one or more height
differences between it and its surrounding points are above
the admissible height difference. The neighborhood contributing to the height difference is limited to a maximum distance. Other variants of this style of filter have expanded its
capabilities. For information on these variants, the reader is
referred to Zhang et al. (2003) and Kobler et al. (2007).
3.6.5.1.3
Segmentation
Point-cloud segmentation is a type of data reduction whereby

point clouds are aggregated into homogeneous segments (e.g.,
roads, buildings, motor vehicles, and topography) and subsequent analyses are performed within the context of the
segmented cloud. This process can be accomplished in variety
of ways. Typically, smooth surfaces are segmented using
region-growing techniques. A normal vector or its change is
used to produce a homogeneity criterion to segment out planar surfaces and topographically varying surfaces (Pfeifer and
Mandlburger, 2009). A cloud segment is generated by using a
threshold value. If the threshold value of the angle between
the surface normal and normal at the nearest surface point to
be expanded is below the threshold, the point is accepted into
the segment (Vosselman et al., 2004). Other criteria might be

employed such as proximity of points (considering only
points close to one another) and locally planar (a candidate
point is accepted based on its orthogonal distance to a plane).
Cloud segmentation may be most advantageous to geomorphologists for filtering and extracting objects. Large
buildings and other objects might be removed to limit data for
analyses. The potential also exists to use segmentation to extract surficial features of varying topographic complexity, but
to date, this has not been fully applied in geomorphology.
3.6.5.2
Surface Interpolation
In general, interpolation is required to generate a higher

resolution DTM, or to utilize spot elevations to generate a
continuous gridded surface that is typically required for
morphometric analyses or modeling efforts (El-Sheimy et al.,
2005). The need for interpolation techniques stems from a
variety of reasons. Shadows or data voids might result from
radar shadowing. Radar illumination is absent during surveying missions because the topography, either concave or
convex relief features, reflects the energy, which does not
permit objects behind or below the terrain to be recorded. It is
also the case that some objects absorb the energy associated
with the radar illumination to generate a data void. Shadows
are also associated with TLS techniques (Figure 6). Another
reason to interpolate data stems from the data source not
being at the desired sampling interval. A DTM user might have
two data sets that are of different sampling intervals and to
make these data sets comparable, one or both might be interpolated to the same resolution. Sampling intervals might
also be discontinuous or at a scale that does not meet the
demands of the desired representation or analysis. Interpolation could be used in both cases to improve the resolution
of a DTM. Off-terrain data might produce another scenario
where interpolation is needed. For example, filtering the
vegetation from the point cloud would leave discontinuous
data (larger intervals between terrain points) or data voids.
Interpolation algorithms could be used to minimize the effects
of these voids in the final DTM.
Point heights are commonly used to generate DTMs with the
aid of any number of interpolation techniques. Interpolation is
a process whereby an unknown elevation is estimated from
known elevations of neighboring points. Interpolation methods
are categorized into three dominant approaches (Table 2).
Local neighborhood methods use surrounding points to estimate unknown elevation values. Two algorithmic approaches
include fitted functions and weighted averages. The fittedfunction algorithms examine small patches of data to determine the coefficients of selected analytical bivariate functions
based on sample points inside the window. This is frequently
accomplished using least-squared criterion. Weighted average
algorithms use a weighted summation of elevation values from
a running subset of surrounding elevation values, without
the aid of an intermediate parametric surface. Geostatistical
interpolation approaches involve understanding the inherent
spatial structure of the data (local, regional, and global spatial
structures) and then predicting missing values based on the
spatial structure. Kriging techniques are a common geostatistical
149
Table 2 Examples of methods employed for DTM generation from each of the three major categories of spatial interpolation techniques (Mitas
and Mitasova, 1999)
Local
Nearest neighbors
Natural neighbors
Inverse distance weighting
Triangular irregular network related interpolations
Geostatistical
Spline
Univariate
Multivariate
Simple kriging
Ordinary kriging
Block kriging
Factorial kriging
Universal kriging
Kriging with an external drift
Simple cokriging
Ordinary cokriging
approach. Splines are another common gridding or interpolation technique that have also been referred to as variational
approaches (Mitas and Mitasova, 1999). Spatial interpolation
via this approach assumes missing values should be closely
related to, or pass through existing data points and should be as
smooth as possible (Hutchinson, 1989; Mitasova and Hofierka,
1993). It is clear from this very brief overview of the different
interpolation algorithms and approaches, that each technique is
not equivalent and does not produce the same result. Therefore,
an analyst/scientist must have a clear understanding of the
mathematical underpinnings of interpolation algorithms that
are used to generate a DTM, or what is needed to produce a
high-quality DTM from the point cloud. Individuals requiring a
mathematical and statistical treatment of spatial interpolation
should refer to Li et al. (2005), El-Sheimy et al. (2005), and
Maune (2007).
Interpolation techniques estimate unknown elevations and
have different levels of accuracy. Several comparative studies
have examined how different interpolation techniques perform under a variety of test scenarios (Carrara et al., 1997;
Robeson, 1997; Burrough and McDonnell, 1998; Zimmerman
et al., 1999; Wilson and Gallant, 2000). In general, geostatistical approaches have performed better than other interpolation approaches in terms of interpolating elevation values
accurately (Zimmerman et al., 1999; Wilson and Gallant,
2000, Erdogan, 2010). In some situations, however, neighborhood approaches such as the inverse distance weighted
(IDW) algorithm and radial-basis functions can match or
outperform results from geostatistical approaches (Weber and
Englund, 1992; Brus et al., 1996; Declercq, 1996; Aguilar
et al., 2005).
These disparate conclusions have more recently been reconciled in studies conducted by Chaplot et al. (2006) and
Heritage et al. (2009). Chaplot et al. (2006) used a nested
scales approach (DTMs at the microplot, hillslope, and
catchment scales) to evaluate the accuracy of several interpolation techniques (IDW; ordinary kriging (OK); universal
kriging (UK); multiquadratic radial basis function (MRBF);
and regularized spline with tension (RST)). Heritage et al.
(2009) tested five different interpolation algorithms under five
different sampling strategies that included cross section, bar
outline only, bar and chute outline, bar and chute outline with
spot heights, and aerial ALS equivalent derived from degraded
TLS data. Key findings from both studies indicated data
sources with greater point densities and those that contained a
greater number of survey point positions relative to the
breaks-in-slope of the surficial feature(s) of interest significantly reduced the error differences between interpolation
Regular thin plate spline

Regular spline with tension
Thin plate spline
Thin plate spline tension
techniques. The basic premise from this work is that with

higher resolution data, the differences between the various
interpolation approaches are reduced. Chaplot et al. (2006)
also identified where particular interpolation techniques
would be most effectively used: (1) kriging worked particularly
well where sample data densities were low; (2) IDW and RST
performed better in environments where structure of the
height was weak (microplots); (3) OK was better suited in
topographically smooth conditions and interpolation across
multiple scales; (4) MRBF provided better results in mountainous topography; and (5) RST was found to perform well at
hillslope and catchment scales.
Beyond existing interpolation techniques, there is a need to
advance and test new interpolation algorithms that will improve our ability to address topographic complexity and permit the generation of high-resolution DTMs. A variety of new
approaches have been proposed over the past decade. Almansa
et al. (2002) attempted to apply an image interpolation approach called the absolutely minimizing Lipschitz extension
(AMLE) model to DTM development. The main advantage of
this approach was that it did not produce artifacts in the interpolated surface, which satisfies a maximum principle
(Almansa et al., 2002). A major disadvantage of AMLE to
geomorphologists is that it cannot interpolate slopes and
other land-surface parameters. Furthermore, AMLE requires
use with another interpolation technique, which further
complicates the use of this promising approach. Almansa et al.
(2002) suggested combing AMLE with Kriging or thin-plate
spline to overcome this issue. A relatively recent novel advance
was a hybrid interpolation method, which integrates linear
(bilinear) and nonlinear (bicubic) interpolation methods (Shi
and Tian, 2006). Root mean square error (RMSE) values from
a hybrid interpolation method were found to be significantly
lower (greater accuracy) than the other bilinear or bicubic
methods. This approach has promise for areas with heterogeneous topography. A final example of advances in new interpolation approaches is an approach specifically designed to
handle high-resolution data sets. Isenburg et al. (2006) combined finalized point streams and streaming meshes to link
several streaming software modules, which collectively used
large data sets of ALS points, produced giant TINs, and generated a high-resolution raster DTM. This approach generated
a 50 394 30 500 raster DTM with 6 m post spacing in 16 bit
binary BIL format (3 GB), using less than 100 MB of main
memory and 300 MB of temporary disk space in just over an
hour on a standard laptop (Isenburg et al., 2006). This approach provides an opportunity to convert large mass point
data structures to a DTM with minimal computer processing
150
power and less time than traditional techniques. This type of

interpolation technique has been under developed to this
point. We believe that further advances in this type of interpolation technique would greatly benefit geomorphologists
interested in high-resolution DTMs.
3.6.6
DTM Error Assessment
Error assessment is used to evaluate the accuracy of a DTM.

Geomorphologists rely on the accurate representation of landsurface heights to generate land-surface parameters, assess
topographic patterns, and analyze terrain changes (Reuter
et al., 2009). An initial assessment of DTM errors is an
evaluation of elevation values compared to corresponding
field measurements that are more accurate. This is not always
practical, however, because field measurements also exhibit
error (Florinsky, 1998), and field data may be difficult to
obtain. Unfortunately, acceptable guidelines for assessing
DTM accuracy for geomorphological applications do not exist
(Reuter et al., 2009), although numerous approaches can be
used to assess the quality of a DTM. These include visual,
manual, mathematical, statistical, or modeling approaches to
error assessment.
DTM under investigation (Reuter et al., 2009). The RMSE is

often expressed as follows:
s
Pn
2
i 1 zsi zREF si
5
RMSE
n
where z(si) is the DTM elevation at the spot location (si), and
zREF(si) is the reference elevation at the same location.
The RMSE metric is the most common metric used to
characterize global error. For example, we used it to examine
the global error associated with TLS measurements of the
topography in a recently burned watershed (Figure 9). Two
TLS surveying missions were conducted with seven control
points (Leica Geosystems HDS targets), which produced a
total of 187 reference points. The RMSE for the first survey was
1.38 mm and 2.94 mm for the second survey. These errors are
below the centimeter-scale change of measurement used to
detect changes in sediment transport after a rainfall event.
They also highlight the importance of targets in reducing error
associated with TLS techniques. Inferences from error
TLS survey parameter summary (09/2729/08)

n
192
RMSE (m)
3.6.6.1
Error Issues
Quantiles (m)
Wise (2000) grouped errors into three categories that include

artifacts, systematic errors, and random errors. Artifacts are
obvious spurious data that are commonly visually detected
with the aid of viewing hillshade maps and 3D perspectives.
They are generally the result of timing errors, which ultimately
produces erroneous positioning of data, or may be holes in
the DTM data that are the result of improperly coded data
(Reuter et al., 2009). Systematic errors result from inherent
bias in the instrumentation that causes the data to be distorted
in the x, y, or z direction, physical causes (i.e., temperature,
clouds, and air quality), or result from human-observer limitations (Li et al., 2005). Systematic errors are not as visually
evident as artifacts and require the use of mathematical or
statistical techniques to ascertain the level to which these
impact a particular DTM. Random errors are commonly random noise and may be normally distributed in the data. A
major goal of digital terrain modeling is to identify, reduce,
remove, or at least quantify these errors (Reuter et al., 2009).
Other sources of error can arise from issues such as data
density, data distribution, and interpolation algorithms (Gong
et al., 2000).
3.6.6.1.1
0.001
90
0.002
75
0.001
50
0.000
25
0.001
10
0.002
Global errors
A standard way to assess the accuracy of a DTM is analysis of

the entire data set to produce a global summary statistic that
characterizes the average error in the altitude field (e.g.,
RMSE). The RMSE represents a direct comparison of DTM
elevations to reference elevations obtained from GPS, maps,
and field survey (e.g., GPS real-time kinematic surveying).
Reference points should be widely and systematically distributed across the surface to capture the complexity of the terrain
and be measured at a much higher degree of accuracy than the
Control point location

TLS instrument location
0
10
20
40
meters
Figure 9 A control network of high-definition targets from the 2008

USGS Landslide Hazards Programs intensive research site located
outside of Santa Barbara, CA. The targets were used to calculate
global measures of error for the DTM (work funded by the National
Geographic Society - CRE Grant #8590-08).
assessment indicated DTM errors are spatially variable and are

linked to topographic complexity (Chaplot et al., 2006).
Many studies also utilize the mean error (ME) and the
standard deviation (SD) as highlighted by Fisher and Tate
(2006):
Pn
zsi zREF si 2
ME i 1
,
6
n
s
Pn
2
i 1 zsi zREF si ME
SD
:
n1
7
Equations [5][7] represent global measures of DTM elevation accuracy. They do not account for localized patterns of
errors that are commonly tied to the heterogeneous nature of
the topography in many geomophic settings.
3.6.6.1.2
Local errors
Recent studies have begun to focus on assessing accuracy at

local scales (Kraus et al., 2006; Erdogan, 2010). Kraus et al.
(2006) presented an approach that considers the accuracy of
every grid cell in a DTM with particular emphasis on the
number, alignment, and distance of the neighboring original
points, the terrain curvature of the neighborhood of each grid
point, and the height accuracy of the original points. The
approach provides a good starting point to assess local accuracy in geomorphological studies, but should not be the
only approach considered in the assessment of local accuracy
issues. Kraus et al. (2006) highlighted the need to properly
qualify the DTM errors by visually depicting points that are
either unusable or contain larger errors. Erdogan (2010) used
an error mapping approach to assess the spatial patterns of
DTM errors. A combination of local spatial statistics and
geographically weighted regression was used to reveal that
topographic complexity and point spacing were linked to localized DTM errors.
The importance of local versus global error measures to
accurately assess DTMs requires additional research. Other
approaches have also been presented, and surely new approaches will emerge. There is no consensus on the validity of
accuracy measures, as they characterize different aspects of
error. An example of a new research direction was presented by
the work of Hohle and Hohle (2009) who assessed errors in
laser scanning and photogrammetry DTM source data to determine if outliers would significantly degrade the effectiveness of the RMSE and other approaches that assume that error
is normally distributed. Their findings led to the conclusion
that if the DTM error distribution departs from normality,
other approaches such as sample quantiles of error, median,
and the normalized absolute deviation should be used to assess DTM error. Aguilar et al. (2007, 2010), Aguilar and Mills
(2008), and Habib (2009) also provided important reviews of
new techniques for assessing the accuracy of ALS-derived DTM
data sets.
3.6.6.1.3
Systematic errors
Systematic errors are commonly manifested as systematic

biases associated with sensor systems and procedures used in
the DTM generation process (Albani and Klinkenberg, 2003).
Reuter et al. (2009) also showed systematic error could be
151
introduced or enhanced by geomorphometric algorithms. This

is significant and indicates that error can be introduced and
amplified in terrain analysis. Systematic errors are not easily
detectable and generally not recognized via the visualization
process, as biases generally represent lower frequency variation
(Wechsler, 1999). Statistical or mathematical analyses are required to identify and assess these types of error (Reuter et al.,
2009; Blanchard et al., 2010). If the systematic error can be
identified, it can be modeled and significantly reduced to
improve the quality of the DTM. At a minimum, there should
be qualifying statements about the quality of the DTM that
accurately quantifies the systematic error (Daniel and Tennant,
2007).
There are a variety of ways to assess or reduce systematic
errors. The NCALMs calibration procedure is an example of
how careful preparation of the data collection mission and
independent survey data can be used to reduce the potential
for systematic error. Field conditions, however, may not be
optimal for calibration or the field conditions may change
during the calibration process, thereby making the collected
data unreliable in terms accuracy assessment. In these scenarios, the lack of calibrated elevation data or the unreliable
calibrated elevation data should be replaced with reliable data
from other sources (i.e., benchmarks, prior surveys, or other
high-quality DTMs). For example, Rignot et al. (2001) reported systematic errors in SRTM data due to microwave
penetration into snow cover on alpine glaciers, thereby impacting the accuracy of glacial volumetric estimates based on
analysis. Knowing the location of such systematic biases permits the use of other data sources to be used in those locations
for assessing elevation or volumetric changes.
Although the term reliable data is used here, each data set
should be independently evaluated for systematic biases. This
is particularly important to studies that examine land-surface
changes with data from different sensors. Systematic errors
associated with different sensors are generally hard to detect,
which makes it difficult to draw conclusive inferences from the
results. Mills et al. (2005) identified systematic over- or
underestimation of the actual land-surface altitudes, and in
other locations a significant increase in noise. A lack of
knowledge of the existence of these types of errors would have
led to misinterpretation of the changes in the coastal landscape investigated by these researchers. Furthermore, it is
critical to consider systematic error when conducting DTM
differencing, as systematic error propagation between data sets
can occur. The systematic error can be either linear or nonlinear in nature.
3.6.7
The use of DTMs in geomorphological research has risen significantly over the last three decades, and applications extend
over all process domains. Numerous advances have resulted
from the rapid proliferation of geospatial technologies, including remote sensing, GPS, ALS, TLS, and TPT. Recent advances in geospatial technologies have provided new
opportunities to map landscapes and landforms without
issues traditionally associated with field mapping (Jones et al.,
2007). Issues commonly associated with field mapping
152
include: (1) access to locales (e.g., private property, road

closures, or hazardous trekking to the site); (2) time to conduct the survey and cost; and (3) an inability to capture subtle
features in the topography (topographic expressions missed by
the human eye). Furthermore, multitemporal data offer the
ability to capture topographic changes through repeat surveys
and provide critical information regarding land-surface
evolution.
The ability to capture repeat high-resolution data and generate multitemporal DTMs provides much better spatial and
temporal information regarding surface roughness and topographic change. Therefore, the use of geospatial technologies
have directly led to new and innovative findings that could not
have been generated using traditional techniques. For example,
Volker et al. (2007) used local relief (a proxy of surface
roughness) to identify scalar differences in alluvial fan morphometry from fans with disparate formative processes. The
results revealed a distinct topographic signature associated with
debris flow and fluvial dominated fans. Frankel and Dolan
(2007) also used surface roughness to differentiate the ages of
fan units, as they were found to exhibit statistically unique
topography. They found a smoothing trend in the topography
of fan units up to B70 ka, with older units exhibiting greater
surface roughness, as secondary erosional processes begin to
dominate the process regime on older units.
Advances are also being made by integrating surface
roughness with other landscape ecology metrics to unravel
ecosystem function and structure (Hoechstetter et al., 2008).
In particular, Hoechstetter et al. (2008) found that area and
distance landscape metrics were sensitive to surface roughness.
Forest-island distributions associated with fire in the central
Andes Mountains were found to be significantly correlated to
locations that exhibit higher topographic roughness (Coblentz
and Keating, 2008). The integration of surface roughness and
other land-surface and ecological parameters provided greater
insight into landscape and landform evolution.
Despite many new insights into how geomorphic systems
operate, more terrain-analysis research is required. Opportunities exist for incorporating high-resolution DTMs with
spatially and temporally continuous process-oriented data
(e.g., rates of weathering, sediment flux, surface and subsurface hydrology, ablation, etc.). The potential to fuse data
from a variety of remote-sensing techniques also exists to enhance geomorphologists ability to extract more information
about features and processes. This will become more important as existing and emerging instruments provide the capability to conduct repeat surveys in a timely and cost efficient
manner. This provides geomorphologists with the ability to
push-the-envelope in terms of accurately estimating processrates from assessment of morphological changes. The increased frequency in the temporal resolution of acquiring
digital elevation data has the potential to advance our knowledge of shorter term landscape and landform changes. Such
change detection capabilities are also relevant from resource
management and policy perspectives. Nevertheless, more
progress is required to effectively acquire, transfer, process, and
analyze large data sets that represent computationally intensive endeavors. Technically, there is also a need for improved
storage capabilities and data/processing quality control. Such
developments will further advance geomorphological
research. With these technical or methodological issues in

mind, we take the liberty to speculate about the possible research pathways involving the use of elevation data and terrain
analysis in geomorphology. These are based on our field experiences and an extensive literature review.
3.6.7.1
High-Resolution DTMs
One of the important aspects of high-resolution DTMs is the

ability to extract information at the length-scale of particle
roughness in many environmental settings (Heritage and
Milan, 2009; Hodge et al., 2009a,b). This level of detail is
critical, as it provides an accurate depiction of the boundary
conditions that commonly influence, and in turn, are modified by process-mechanics. For example, particle roughness
has a direct impact on fluid mechanics, creating frictional drag
that generally reduces velocity or generates turbulent flow.
Flow over particles exerts drag that is important to sediment
entrainment and subsequent sediment transport. Therefore,
particle dimension variability is critical for understanding
fluid-particle interactions. TLS and TPT can be used to characterize these conditions, but with some limitations. TLS work
has commonly been constrained to studies of specific events.
There are limitations to capturing within-event temporal data
because scanning rates do not permit rapid scans of an area
(generally only very small areas can be scanned rapidly
enough and therefore a large portion of the landscape is not
assessed), the laser does not penetrate water or solid materials
in the case of thick granular flows, and inclement weather
conditions can hinder scanning or at least make it dangerous
to the user. TPT can be used to obtain higher temporal resolution data, but there are limitations with it in terms of
lighting conditions, particularly for events that occur during
darkness, unless a significant light source were added to illuminate the site. TPT, much like TLS, is limited in its ability to
obtain data during an event, as the sediment movement impedes accurately capturing temporal variability of the surface.
It is also the case that TPT is commonly hampered by inclement weather conditions thereby further limiting the ability
to capture all of the temporally relevant topographic changes.
The novelty of obtaining high-resolution topographic data
is also tempered by an inability to measure process-related
data at similar spatial extents. For example, TLS surveying can
be used to accurately measure topographic changes in a
channel as the direct result of sediment transport associated
with runoff-generated debris flow (McCoy et al., 2010). Even
in the previously mentioned study, however, the processrelated data (gathered with pressure transducers, a load cell,
rainfall gauges, laser and sonic stage devices, soil moisture
probes, digital videography, etc.) were limited to crosssectional process-oriented information (stage, pore water
pressure, velocity, force, etc.) or point-based process-oriented
data (rain fall, soil moisture, etc.). The ability to correlate
energy and matter dynamics with an accurate depiction of the
topographic change (obtained with the TLS) is hindered by
limitations in the spatial extent of process-oriented data and
the temporal extent of the TLS data. The TLS was used to only
sample before and after each event. Topographic changes between individual debris-flow surges that are captured in the
digital videography are not obtained in our TLS sampling

scheme. Therefore, our temporal resolution is inconsistent
with the permanent field instruments that are sampling during
the rainfall event on the order of 10s to 100s of Hz. Despite
these high sampling rates from the permanent field instruments, the measurements occurred at two cross-sections and
two hillslope locations along the B150 m extent of channel
scanned with repeat TLS surveys. Topographic changes from
2 cm DTM data gathered along the B150 m channel segment
were spatially more contiguous than the cross-section and
point information gathered by the permanent field
instruments.
Point and cross-sectional data are generally not sufficient
to estimate local-scale erosion or for considering the local and
nonlocal interactions that speak to connectivity within a variety of environmental settings. For example, studies of erosional and depositional processes have been hindered by the
uncertainties associated with spatial and temporal lumping of
data to attain results on sediment delivery, yield, transport,
and causal mechanisms (Walling, 1983; de Vente et al., 2007).
This lack of information has led to the development of spatially distributed sediment transport models (Merritt et al.,
2003; de Vente and Poesen, 2005), and transport models
based on the conservation of mass and energy (see Dietrich
et al., 2003). Wainwright et al. (2008) have shown that many
of the spatially distributed models are significantly over simplified, whereas other researchers (Tucker and Bras, 1998;
Roering et al., 1999) have found these models to be based on
misinterpretations of the literature. Studies examining geomorphic transport laws have the potential to characterize
causal mechanisms for long-term landscape evolution, but
assume local transport. A reasonable assumption for some
processes (e.g., grain detachment by raindrop impact) can be
expressed as an at-a-point measurement (i.e., local surface
gradient). However, transport laws do not significantly integrate process with form parameters across the landscape or
landform. This limitation has led to a recent proliferation of
papers focused on local and nonlocal sediment flux (e.g.,
Tucker and Bradley, 2010; Foufoula-Georgiou et al., 2010;
Foufoula-Georgiou and Stark, 2010).
Spatially and temporally continuous process-related data
are required to make connections between local and nonlocal
sediment fluxes in the field. Ultimately, spatially and temporally continuous process-related data can be made available
from numerical models. The models, however, need to be
based on and continuously calibrated with field measurements, but not field measurements as they are currently obtained. Denser sampling networks are required to capture the
variability within geomorphic systems. In the bestcase scenario, increased spatially and temporally continuous
data sets could be developed in conjunction with highresolution digital elevation measurements. This would allow
geomorphologists to gain a better appreciation the connectivity of local and nonlocal sediment fluxes as they relate to
landscape and landform change. At a minimum, this additional data might be used to validate or extrapolate modeling
approaches to bring a new perspective on the issue of local
and nonlocal fluxes as they govern the magnitude of landscape
and landform change. Alternatively, process-related data might
also provide a greater understanding of diffusion processes,
153
that when linked with high-resolution DTM data, may provide

additional information and insight required to unravel the
complexities of landscape and landform evolution. This
would seemingly make the local and nonlocal sources
issue moot.
Several new field sampling approaches and instruments are
likely necessary to increase the amount of process data that
need to be collected in order to obtain spatial coverage associated with DTM coverage. These data must also be integrated
with data collected at-a-point (i.e., cross-sectional or pointbased measurements) and with data and information generated
from various modeling approaches (Tucker and Bradley, 2010;
Hancock et al., 2010) to provide spatially and temporally continuous measures of hydrologic and sedimentologic fluxes. The
lack of spatial integration has continuously plagued sediment
and hydrological transport modeling by reducing the ability to
validate model results with field data. Therefore, a logical goal
might be to develop new instrumentation that can be used in
conjunction with measurements from high-resolution DTM
data to obtain field measurements. This approach, although
likely expensive, would produce the best results moving forward and likely become less cost prohibitive over time.
Another fundamental geomorphological research topic is
landscape evolution or change. High-resolution DTM data
from TLS and TPT offer the ability to obtain a time-series of
topographic conditions that can be compared, such that landsurface change can be characterized, as opposed to relying
solely on numerical modeling approaches. Wheaton et al.
(2010a) have formalized this approach producing DEM of
Difference (DoD) maps, based on the fundamental work of
Lane et al. (1994). The same approach can be used with
coarser resolution DTMs. However, ALS, TLS, and TPT provide
users with the ability to conduct repeat surveys with significantly greater temporal frequency. Indeed, TLS and TPT can be
automated to acquire continuous data at a locale. These new
advances in repeat high-resolution topographical mapping
techniques have improved morphological measurements of
sediment budgets that were traditionally done with repeat
surveys at cross-sections and longitudinal profiles (Lane, 1998;
Stott, 2002; Bezzi et al., 2009; Martin-Vide et al., 2010). Highresolution DoD approaches can increase our understanding of
spacetime relationships by providing accurate measures of
sediment loading (i.e., dry raveling and rock fall; Lim et al.,
2010), sediment entrainment, and event-by-event, sedimenttransport rates (Oppikofer et al., 2009). This information can
be derived at a variety of scales from the original data to
provide information on important concepts in geomorphology. Examples include: (1) local and nonlocal variations in
sediment flux (Gabet, 2003; Tucker and Bradley, 2010); (2)
connectivity within the landscape (Brierley et al., 2006;
Bracken and Croke, 2007); and (3) multiple-scale causality
(Phillips, 2005).
High-resolution geomorphological maps that capture the
spacetime details of morphological features can be used to
estimate rates of slow-moving geomorphic events (Prokop and
Panholzer, 2009). Rates of movement on an event-by-event
basis can also improve estimates of the extent to which a
specific feature spatially evolves on the landscape given variations in erosion and deposition. For example, a dune crest
(Rubin et al., 2008) or sediment waves (James, 2006) might
154
be mapped before and after a wind or flood event to determine the extent that these features have migrated during that
event. Rates of movement from geomorphological mapping
are not only important to understanding process mechanisms,
but they also have significant implications for habitat changes
and management (Wheaton et al., 2010b).
Surface roughness can provide topographic boundary
conditions that alter a variety of flow properties and in turn
influence subsequent flow events (Frankel and Dolan, 2007).
High-resolution topographic data can be used to represent
initial boundary conditions for process-form interactions
(Heritage and Milan, 2009). This source of information can
establish baseline topographic conditions for model initiation
as well as an end form to calibrate results from model simulations. The integration of models with high-resolution
topographic data might serve as potential solutions to concerns with the lack of spatially continuous process-related data
and model validation. A practical issue in moving forward,
however, is that high-resolution DTM data require substantial
computational resources such as memory and fast processors
given the data volume. Although this is a definite shortcoming, the use of high-resolution data in numerical process
models offers significant potential for improved parameterization schemes and validation of model results.
3.6.7.2
High-Resolution DTMs and Scale
High-resolution DTMs also show promise for formalizing

some of the previously mentioned shortcomings in addressing
issues of scale, such as the scale at which processes and
landforms interact, and hierarchical connections that may
exist between scales. The use of TLS and TPT offer the potential
to develop millimeter to centimeter DTMs, a scale that approximates particle roughness and the operational scale of
many processes. Variations in the computational scale of analysis can then be used to determine what scales are best for
assessing topographic variability and landscape change. Such
scale-dependent analysis is of interest because multiple processes operate on landforms, and each process may have a
unique spacetime signature that could potentially be characterized (Chorley et al., 1984). As Phillips (2012) has outlined, analyses of varying spatial scales in geomorphology
should identify: (1) characteristic scales (Schmidt and
Andrews, 2005; Perron et al., 2008); (2) conditions of scale
independence (Pelletier, 1999; Southgate and Moller, 2000);
and (3) tools for visualization and data exchange between
scales (Murzyn et al., 2006). Data exchanges and analyses
between scales are critical for interpreting hierarchical connections between scales. Although a great deal of literature in
geomorphology has highlighted scale invariance (RodrguezIturbe and Rinaldo, 2001), scale linkages have been identified
in the landscape (Phillips, 2012). High-resolution DTMs can
aid in the assessment of fine-scale fluxes in mass and energy
(as represented by spatio-temporal morphologic variations
that govern those fluxes) and for identifying and mapping
process-pattern relationships that are manifested at these
scales. In turn, these results can be investigated in the context
of geomorphic responses at coarser scales within a system that
produces different patterns at different spatial and temporal
scales. This information could ultimately serve as an
important starting point for formalizing our interpretations

and predictions of landscape evolution.
A limiting factor with high-resolution DTMs is the coarsest
scale that can be analyzed given their limited spatial extent.
Techniques used to generate high-resolution topographic data,
such as TLS and TPT, record small areas (100 s to 1000 s of m2).
Topographic wavelengths and complexity of larger landforms
are generally not represented in the data, and this limits the use
of the data for selected geomorphological applications. Opportunities exist, however, to use coarser resolution DTMs from
other imaging sensors to provide additional information to
address the spatial structure and hierarchical organization of
the topography at regional and global scales. New conceptual
and technical approaches need to be developed that integrate a
wide variety of information from numerous DTM sources. This
will permit a better assessment and mapping of the inherent
hierarchical organization of the land surface.
3.6.7.3
Data Fusion
Very few attempts have been made to fuse other remotely sensed
data with DTMs within the constructs of geomorphological
methodologies. As an example, ALS and TLS both capture intensity data. Intensity data are underutilized for geomorphic
research and could be used in conjunction with elevation to aid
in mapping surficial features. Intensity offers a means of capturing information about the reflectivity of an object and phenomena that are both visible and invisible, with some modest
fine-tuning of the energy returns (Wasklewicz et al., 2005,
2007). Intensity values provide geomorphology with an increased ability to map objects, place topographic variability in
context of the mapped objects, and put the discipline in a
position to better understand human impacts on the natural
geomorphic system. The later is possible because humanmodified landscapes generally have different spectral signatures
than the materials that make up geomorphic features.
Additional information can be obtained from imaging
sensors that include spectrometers and radar systems. For example, high-resolution multispectral imagery has been combined with ALS data to map the distribution of marsh features
across the PIE-LTER site in northeastern Massachusetts (Millette et al., 2010). The majority of the work involves the integration of hyperspectral data, multispectral imagery, and
thermal imagery with a focus on classification and thematic
mapping (e.g., Gilvear et al., 2004). Further developments are
needed that consider landscape and landform dynamics. Work
conducted in a coastal setting to examine sediment dynamics
provides a snapshot of what could be attempted to advance
geomorphological research (e.g., Deronde et al., 2008). They
used hyperspectral data to classify and map different natural
and artificial berm and dune deposits. Spectral information
was combined with ALS data from multiple surveys to investigate sediment transport amounts and directions. This
fusion of data was found to be particularly useful to capturing
berm dynamics associated with beach renourishment projects.
3.6.7.4
Temporal Data Acquisition
The advances in instruments such as TLS, ALS, and TPT have the
potential to increase the temporal frequency of data acquisition
and increase DTM inventories. As instrument weight and cost

continue to decrease, it has become more feasible to mobilize
equipment rapidly or set-up these instruments at various locations for extended periods of time. Instruments such as TLS
are presently capable of being set-up and programmed at fixed
locations to sample periodically throughout the day and night.
This particular application has scarcely been applied in geomorphology, but is used in the mining industry to monitor rock
fall and volumes of mined material. A similar approach should
be used to increase our temporal understanding of landscape
evolution and supplement existing information gathered at
long-term sampling sites (e.g., stream gauges, weather stations,
coastal-wave monitoring sites). Marzolff and Poesen (2009)
demonstrated how using multiple DTMs can improve our
space-time understanding of morpho-dynamics within gully
systems. They examined changes to two different bank gully
types using TPT. A major finding from their research was the
large degree of storage of sediment within the gully network
after headward erosion of the bank gullies. This was not anticipated, as most hypotheses would suggest that the gullies
erode material into other parts of the drainage network with
little deposition taking place within the gully system. Furthermore, TPT has been used to quantify the rate of parabolic dune
migration (e.g., Arteaga et al., 2008).
A decrease in instrument weight and cost may also make it
feasible to mount instruments on UAVs. This would greatly
enhance the geomorphologists ability to capture more data
and provide greater flexibility in planning field-sampling campaigns. Missions (ALS) commonly take a great deal of time to
mobilize because equipment is not located near relevant field
sites. A UAV is much easier to mobilize and would permit
measurement of time-sensitive data within a matter of hours.
The ability to respond rapidly would permit the user to obtain
digital elevation data immediately before and after an event.
This flexibility greatly facilitates fieldwork and the assessment of
geomorphological hazards. For example, multitemporal DTMs
can be used to calibrate debris-flow sediment yield and prediction models after wildfires. This information is important for
hazard mitigation schemes (Gartner et al., 2009; Cannon et al.,
2010a,b). Furthermore, the ability to rapidly respond to events
or forcing factors is critical, and DTMs can be combined with
rainfall intensity data to study the timing, location, and run-out
distances of debris flow-prone areas (Cannon et al., 2010a,b).
3.6.8
Conclusions
Advances in sensors, computer technology, and information technology have provided many new insights into
geomorphological systems. This is clearly demonstrated by the
ever-increasing usage of DTMs and geomorphometric research
over the last four decades. This phenomenon has arisen from a
shift in how geomorphologists conduct research, whereby
greater emphasis has been placed on enumerating topographic
complexity to capture spatial, temporal, and scalar differences
in the land surface. There are, however, growing pains with
implementing geospatial technologies, as many technical
issues must be addressed. The objective of this chapter has
been to reduce some of these difficulties by providing a definitive introduction to digital terrain modeling and examples
155
of the incorporation of DTMs in geomorphological research.

Topics covered include the basic history of digital terrain
modeling, recent data sources, preprocessing issues, DTM
production, DTM error analysis, and the use of DTMs in the
current context of geomorphology.
Recent advances in sensors and methods include light detection and ranging, radio detection and ranging systems, and
progress in terrestrial photogrammetry techniques. These instruments are capable of producing moderate to highresolution elevation data over a large range of spatial extents,
with LiDAR and terrestrial methods producing data with much
greater temporal frequency than previously possible. A concern with instruments that capture high-resolution data is the
nondiscrete nature of obtaining elevation data. The point
clouds captured by many instruments contain both terrain
and off-terrain points. A variety of methods can be used to
remove vegetation and other off-terrain objects from the point
cloud, and geomorphologists must be cognizant of landsurface concepts, sensor characteristics, preprocessing, error
analysis and postprocessing, and the limitations associated
with utilizing any DTM for geomorphological research. These
stages in digital terrain modeling are critical to understand, as
the quality of DTMs is a major issue, and high quality bareearth models are most often sought after in geomorphology.
The generation of DTMs generally requires selecting a
spatial interpolation algorithm. Spatial interpolation of the
cleaned point clouds is an essential step in preprocessing. Each
interpolation technique has advantages and disadvantages.
The advantages and limitations associated with various interpolation approaches were examined in the context of environments with different topographic complexity.
Another key aspect of digital terrain modeling involves
error assessment, although this is not always accounted for in
many studies. It is important to note that inherent error is due
to sensor systems, the data acquisition approach, the representational scheme, and issues related to preprocessing approaches. In general, the error can be classified as random or
systematic, although global, local, and systematic errors
should be evaluated. Global measures are well established and
are commonly used in engineering, statistics, and cartography.
Systematic errors tend to be associated with field methods,
instrumentation, and choice of algorithms in the production
of the DTM. Local error analysis represents a relatively new
research area that assesses variations in the local error that
greatly influences terrain analysis results. Change detection
using multitemporal DTMs requires the evaluation of all error
in or to assess topographic changes. Further research is required to determine the extent to which these types of error
influence geomorphological research, although numerous
studies have demonstrated that error propagation does occur,
stemming from DTM error that influences the accuracy of
geomorphometric parameters and modeling results, as many
models make use of specific parameters.
Digital terrain modeling is now critical in geomorphological research. The availability of high-resolution DTMs has
greatly improved our ability to assess geomorphological systems and maps landforms at a variety of scales. These data also
greatly facilitate change detection and the generation of process-related data to investigate the linkages of process and
form, and process and patterns. Such data and the derived
156
information are important for process modeling, as we attempt to improve our process parameterization schemes,
calibrate numerical models and validate results from simulations. Integrating process-oriented data from the field with
terrain analysis and simulation results enable new investigations of land-surface evolution, evaluation of the impact of
scalar variations on landscape change, and promote better
understandings of specific landform evolution. Digital terrain
modeling and geomorphometry have the potential to provide
us with many more new understandings of complex geomorphological systems and landscape evolution.
References
Aguilar, F.J., Aguera, F., Aguilar, M.A., Carvajal, F., 2005. Effects of terrain
morphology, sampling density, and interpolation methods on grid DEM
accuracy. Photogrammetric Engineering and Remote Sensing 71, 805816.
Aguilar, F.J., Aguilar, M.A., Aguera, F., 2007. Accuracy assessment of digital
elevation models using a non-parametric approach. International Journal of
Geographical Information Science 21, 667686.
Aguilar, F.J., Mills, J.P., 2008. Accuracy assessment of LiDAR-derived digital
elevation models. Photogrammetric Record 23, 148169.
Aguilar, F.J., Mills, J.P., Delgado, J., Aguilar, M.A., Negreiros, J.G., Perez, J.L.,
2010. Modelling vertical error in LiDAR-derived digital elevation models. ISPRS
Journal of Photogrammetry and Remote Sensing 65, 103110.
Albani, M., Klinkenberg, B., 2003. A spatial filter for the removal of striping artifacts
in digital elevation models. Photogrammetric Engineering and Remote Sensing
69, 755765.
Alexa, M., Behr, J., Cohen-Or, D., Fleishman, S., Levin, D., Silva, T., 2001. Point
set surfaces. In Proceeding of the 12th IEEE Visualization Conference, San
Diego, USA, pp. 2128.
Almansa, A., Cao, F., Gousseau, Y., Rouge, B., 2002. Interpolation of digital
elevation models using AMLE and related methods. IEEE Transactions on
Geoscience and Remote Sensing 40, 314325.
Arteaga, C., Juan de Sanjose, J., Serrano, E., 2008. Terrestrial photogrammetric
techniques applied to the control of a parabolic dune in the Liencres dune system,
Cantabria (Spain). Earth Surface Processes and Landforms 33, 22012210.
Atkinson, P.M., Tate, N.J., 2000. Spatial scale problems and geostatistical solutions:
a review. Professional Geographer 54, 607623.
Baltsavias, E.P., 1999. A comparison between photogrammetry and laser scanning.
ISPRS Journal of Photogrammetry and Remote Sensing 54, 8394.
Bauer, B., 1996. Geomorphology, geography and science. In: Rhoads, B.L., Thorn,
C.E. (Eds.), The Scientific Nature of Geomorphology: Proceedings of the 27th
Binghamton Symposium in Geomorphology. Wiley, New York, pp. 381413.
Beraldin, J.-A., Blais, F., Lohr, U., 2010. Laser scanning technology. In: Vosselman,
G., Maas, H-G. (Eds.), Airborne and Terrestrial Laser Scanning. Whittles
Publishing, Dunbeath, Scotland, UK, pp. 139.
Bezzi, A., Fontolan, G., Nordstrom, K.F., Carrer, D., Jackson, N.L., 2009. Beach
nourishment and foredune restoration: practices and constraints along the
Venetian shoreline, Italy. Journal of Coastal Research 56, 287291.
Bishop, M.P., 2013. Remote sensing and GIScience in geomorphology: introduction
and overview. In: Shroder, J. (Editor in Chief), Bishop, M.P. (Eds.), Treatise on
Geomorphology. Academic Press, San Diego, CA, vol. 3, pp. 124.
Bishop, Jr. M.P., Shroder, Jr. J.F. (Eds.), 2004. Geographic Information Science and
Blanchard, S.D., Rogan, J., Woodcock, D.W., 2010. Geomorphic change analysis
using ASTER and SRTM digital elevation models in Central Massachusetts,
USA. GIScience and Remote Sensing 47, 124.
Bracken, L.J., Croke, J., 2007. The concept of hydrological connectivity and its
contribution to understanding runoff-dominated geomorphic systems.
Hydrological Processes 21, 17491763.
Brasington, J., Smart, R.M.A., 2003. Close range digital photogrammetric analysis
of experimental drainage basin evolution. Earth Surface Processes and
Landforms 28, 231247.
Brierley, G.J., 2010. Landscape memory: the imprint of the past on contemporary
landscape forms and processes. Area 42, 7685.
Brierley, G.J., Fryirs, K., Jain, V., 2006. Landscape connectivity: the geographic
basis of geomorphic applications. Area 38.2, 165174.
Brodsky, D., Watson, B., 2000. Model simplification through refinement.
In: MacKenzie, S., Stewart, J. (Eds.), Proceedings of Graphics Interface 2000.
Morgan Kaufmann Publishers, Montreal, Canada, pp. 221228.
Brus, D.J., de Gruijter, J.J., Marsman, B.A., Visschers, R., Bregt, A.K., Breeuwsma,
A., 1996. The performance of spatial interpolation methods and choropleth maps
to estimate properties at points: a soil survey case study. Environmetrics 7,
116.
Burrough, P.A., McDonnell, R.A., 1998. Principles of Geographical Information
Systems. Oxford University Press, New York, 333 pp.
Burgmann, R., Rosen, P.A., Fielding, E.J., 2000. Synthetic aperture radar
interferometry to measure Earths surface topography and its deformation. Annual
Review of Earth and Planetary Sciences 28, 169209.
Cannon, S.H., Boldt, E.M., Kean, J.W., Laber, J., Staley, D.M., 2010a. Relations
between rainfall and postfire debris-flow and flood magnitudes for emergencyresponse planning, San Gabriel Mountains, Southern U.S. U.S. Geological
Survey Open-File Report #20101039: 31 pp.
Cannon, S.H., Gartner, J.E., Rupert, M.G., Michael, J.A., Rea, A.H., Parrett, C.,
2010b. Predicting the probability and volume of postwildfire debris flows in the
intermountain western United States. Geological Society of America Bulletin 122,
127144.
Capps, D.M., Rabus, B., Clague, J.J., Shugar, D.H., 2010. Identification and
characterization of alpine subglacial lakes using interferometric synthetic aperture
radar (InSAR): Brady Glacier, Alaska, USA. Journal of Glaciology 56, 861870.
Carrara, A., Bitelli, G., Carla, R., 1997. Comparison of techniques for generating
digital terrain models from contour lines. International Journal of Geographical
Information Science 11, 451473.
Carson, M.A., Kirkby, M.J., 1972. Hillslope Form and Process. Cambridge
University Press, Cambridge.
Catani, F., Farina, P., Moretti, S., Nico, G., Strozzi, T., 2005. On the application of
SAR interferometry to geomorphological studies: estimation of landform
attributes and mass movements. Geomorphology 66, 119131.
Chandler, J.H., 1999. Effective application of automated digital photogrammetry for
geomorphological research. Earth Surface Processes and Landforms 24, 5163.
Chandler, J.H., Fryer, J.G., Jack, A., 2005. Metric capabilities of low cost digital
cameras for close range surface measurement. The Photogrammetric Record 20,
1226.
Chaplot, V., Darboux, F., Bourennane, H., Leguedois, S., Silvera, N., Phachomphon,
K., 2006. Accuracy of interpolation techniques for the derivation of digital
elevation models in relation to landform types and data density. Geomorphology
77, 126141.
Chen, Q., Gong, P., Baldocchi, D., Xin, G., 2007. Filtering airborne laser scanning
data with morphological methods. Photogrammetric Engineering and Remote
Sensing 73, 175185.
Chorley, R.J., Schumm, S.A., Sugden, D.E., 1984. Geomorphology. Methuen,
London, 605 pp.
Clarke, T.A., 1995. An analysis of the prospects for digital close-range
photogrammetry. ISPRS Journal of Photogrammetry and Remote Sensing 50,
47.
Coblentz, D., Keating, P.L., 2008. Topographic controls on the distribution of tree
islands in the high Andes of south-western Ecuador. Journal of Biogeography
35, 20262038.
Crosilla, F., Visintini, D., Prearo, G., 2004. A robust method for filtering nonground
measurements from airborne LiDAR data. International Archives of
Photogrammetry, Remote Sensing and Spatial Information Sciences 35,
196201.
Daniel, C., Tennant, K., 2007. DEM quality assessment. In: Maune, D.F. (Ed.),
Digital Elevation Model Technologies and Applications: The Dem Users Manual.
American Society for Photogrammetry and Remote Sensing, Bethesda, MD,
pp. 395440.
Declercq, F.A.N., 1996. Interpolation methods for scattered sample data: accuracy,
spatial patterns, processing time. Cartography and Geographic Information
Systems 23, 128144.
Deronde, B., Houthuys, R., Henriet, J.P., van Lancker, V., 2008. Monitoring of the
sediment dynamics along a sandy shoreline by means of airborne hyperspectral
remote sensing and LIDAR: a case study in Belgium. Earth Surface Processes
and Landforms 33, 280294.
Devereux, B., Amable, G., 2009. Aiborne LiDAR: instrumentation, data acquisition
and handling. In: Heritage, G.L., Large, A.R.G. (Eds.), Laser Scanning for the
Environmental Sciences. Blackwell Publishing Ltd., Chicester, UK, pp. 4966.
Dey, T.K., Giesen, J., Hudson, J., 2001. Decimating samples for mesh
simplification. Proceedings of the 13th Canadial Conference on Computational
Geometry, Waterloo, Canada, pp. 8588.
Dietrich, W.E., Bellugi, D., Heimsath, A.M., Roering, J.J., Sklar, L., Stock, J.D.,
2003. Geomorphic transport laws for predicting landscape form and dynamics.
In: Wilcock, P., Iverson, R. (Eds.), Prediction in Geomorphology, Geophysical
Monograph Series. AGU, Washington, D.C., vol. 136, pp. 103123.
Dikau, R., 1989. Application of a digital relief model to landform analysis in
geomorphology. In: Raper, J. (Ed.), Three Dimensional Application in
Geographic Information Systems. Taylor and Francis, London, pp. 5177.
Dikau, R. 1990. Geomorphic landform modelling based on Hierarchy Theory. In:
Brassel K., Kishimoto H. (Eds.). Proceedings 4th International Symposium on
Spatial Data Handling, 23rd27th July, Zurich, pp. 230239.
Dixon, J.L., Heimsath, A.M., Kaste, J.M., Amundson, R., 2009. Climate driven
processes of hillslope weathering. Geology 37, 975978.
Dunning, S.A., Massey, C.I., Rosser, N.K., 2009. Structural and geomorphological
features of landslides in the Bhutan Himalaya derived from terrestrial laser
scanning. Geomorphology 108, 1729.
Dupuis, E., Bakambu, J.N., Rekleitis, I., Bedwani, J., Gemme, S., Rivest-Caissy, J.,
2006. Autonomous long-range rover navigation experimental results.
Proceedings of the 9th ESA Workshop on Advanced Space Technologies for
Robotics and Automation, Noordwijk, The Netherlands, 8 pp.
Dury, G.H., 1983. Geography and geomorphology: the last fifty years. Transactions,
Institute of British Geographers, NS 8, 9099.
Eastman, J.R., 1989. IDRISI: A Grid-based Geographical Information System. Clark
University, Worcester, MA.
El-Sheimy, N., Valeo, C., Habib, A., 2005. Digital Terrain Modeling: Acquisition,
Manipulation and Applications. Artech House Publishers, Boston, MA, 257 pp.
Elsherbini, A., Sarabandi, K., 2010. Mapping of sand layer thickness in deserts
using SAR Interferometry. IEEE Transactions on Geoscience and Remote Sensing
48, 35503559.
Erdogan, S., 2010. Modelling the spatial distribution of DEM error with
geographically weighted regression: an experimental study. Computers and
Evans, I.S., 1972. General geomorphometry, derivatives of altitude, and description
statistics. In: Chorley, R.J. (Ed.), Spatial Analysis in Geomorphology. Methuen &
Co., London, pp. 1790.
Evans, I.S., 1980. An integrated system of terrain analysis and slope mapping.
Zeitschrift fur Geomorphologie N.F. Supplementband 36, 274295.
Evans, I.S., 2003. Scale-specific landforms and aspects of the land surface.
In: Evans, I.S., Dikau, R., Tokunaga, E., Ohmori, H., Hirano, M. (Eds.), Concepts
and Modelling in Geomorphology: international Perspectives. Tokyo, Terrapub,
pp. 6184.
Fisher, P.F., Tate, N.J., 2006. Causes and consequences of error in digital elevation
models. Progress in Physical Geography 30, 467489.
Fleishman, S., Cohen-Or, D., Alexa, M., Silva, C.T., 2003. Progressive point set
surfaces. ACM Transactions on Graphics 22, 9971011.
Florinsky, I.V., 1998. Accuracy of local topographic variables derived from digital elevation
models. International Journal of Geographical Information Science 12, 4761.
Foufoula-Georgiou, E., Stark, C., 2010. Rethinking geomorphic transport: stochastic
theories, broad scales of motion and non-locality. Journal of Geophysical
Research Earth Surface 115, F00A01.
Foufoula-Georgiou, E., Ganti, V., Dietrich, W.E., 2010. A non-local theory for
sediment transport on hillslopes. Journal of Geophysical Research - Earth
Surface 115, F00A16.
Frankel, K.L., Dolan, J.F., 2007. Characterizing arid region alluvial fan surface
roughness with airborne laser swath mapping digital topographic data. Journal
of Geophysical Research 112, F02025.
Fraser, C.S., 1997. Digital camera self-calibration. ISPRS Journal of Photogrammetry
and Remote Sensing 52, 149159.
Fraser, C.S., Cronk, S., 2009. A hybrid measurement approach for close-range
photogrammetry. ISPRS Journal of Photogrammety and Remote Sensing 64,
328333.
Gabet, E.J., 2003. Sediment transport by dry ravel. Journal of Geophysical Research
108, 20492057.
Gabriel, A.K., Goldstein, R.M., Zebker, H.A., 1989. Mapping small elevation changes
over large areas; differential radar interferometry. Journal of Geophysical
Research 94, 91839191.
Garland, M., Heckbert, P.S., 1997. Surface simplification using quadratic error
metrics. SIGGRAPH 97 97, 209216.
Gartner, J.E., Cannon, S.H., Helsel, D.R., Bandurraga M., 2009. Multivariate
Statistical models for predicting sediment yields from Southern California
watersheds. US Geological Survey Open-File Report #2009-1200, 42 pp.
157
Gens, R., Van Genderen, J.L., 1996. Review Article SAR interferometry issues,
techniques, applications. International Journal of Remote Sensing 17,
18031835.
Gilvear, D., Tyler, A., Davids, C., 2004. Detection of estuarine and tidal river
hydromorphology using hyper-spectral and LiDAR data: Forth Estuary, Scotland.
Estuarine Coastal and Shelf Science 61, 379392.
Gislason, S.R., Oelkers, E.H., Eiriksdottir, E.S., et al., 2008. Direct evidence of the
feedback between climate and weathering. Earth and Planetary Science Letters
277, 213222.
Gong, J., Zhilin, L., Zhu, Q., Sui, H.G., Zhou, Y., 2000. Effect of various factors on
the accuracy of DEMs: an intensive experimental investigation. Photogrammetric
Goodchild, M., 1992. Geographical information science. International Journal of
Geographical Information Systems 6, 3145.
Graham, L.C., 1974. Synthetic interferometer radar for topographic mapping.
Proceedings of the IEEE 62, 763788.
Gregory, K.J., 2000. The Changing Nature of Physical Geography. Arnold, New
York, 368 pp.
Gruber, S., Peckham, S., 2009. Land surface parameters and objects specfic to
hydrology. In: Hengl, T., Reuter, H.I. (Eds.), Geomorphometry: Geomorphometry:
Concepts, Software, Applications. Developments in Soil Science. Elsevier,
Amsterdam, vol. 33, pp. 171194.
Guth, P.L., Ressler, E.K., Bacastow, T.S., 1987. Microcomputer program for
manipulating large digital terrain models. Computers and Geosciences 13,
209213.
Gyasi-Agyei, Y., Willgoose, G.R., De Troch, F.P., 1995. Effects of vertical resolution
and map scale of digital elevation models on geomorphological parameters used
in hydrology. Hydrological Processes 9, 363382.
Habib, A., 2009. Accuracy, quality assurance, and quality control of LiDAR data.
In: Shan, J., Toth, C.K. (Eds.), Topograpic Laser Ranging and Scanning. CRC
Press, Boca Raton, pp. 269294.
Hack, J.,1957. Studies of Longitudinal Stream Profiles in Virginia and Maryland.
United States Geological Survey Professional Paper 294-B: 45-97.
Haggett, P., 2008. The local shape of revolution: reflections on quantitative
geography at Cambridge in the fifties and sixties. Geographical Analysis 40,
336352.
Hancock, G.R., Lowry, J.B.C., Coulthard, T.J., Evans, K.G., Moliere, D.R., 2010. A
catchment scale evaluation of the SIBERIA and CAESAR landscape evolution
models. Earth Surface Processes and Landforms 35, 863876.
Haneberg, W.C., 2008. Using close-range digital photogrammetry for 3-D rock slope
modeling and discontinuity mapping in the United States. Bulletin of
Engineering Geology and the Environment 67, 457469.
Harlick, R.M., Shapiro, L.G., 1992. Computer and Robot Vision: vol. 1 and vol. 2.
Addison-Wesley, Reading, MA, 1302 pp.
Hayakawa, Y.S., Oguchi, T., 2006. DEM-based identification of fluvial knickzones
and its application to Japanese mountain rivers. Geomorphology 78, 90106.
Heckbert, P. and Garland, M., 1997. Survey of polygonal surface simplification
algorithms. multi-resolution surface modeling course, ACM Siggraph Course notes.
Hengl, T., 2006. Finding the right pixel size. Computers and Geosciences 32,
12831298.
Hengl, T., Evans, I.S., 2009. Mathematical and digital models of the land surface.
In: Hengl, T., Reuter, H.I. (Eds.), Geomorphometry Concepts, Software,
Applications: Developments in Soil Science. Elsevier, Amsterdam, vol. 33, pp.
3163.
Heritage, G.L., Hetherington, D., 2007. Toward a protocol for laser scanning in
fluvial geomorphology. Earth Surface Processes and Landforms 32, 6674.
Heritage, G.L., Milan, D.J., 2009. Terrestrial laser scanning of grain roughness in a
gravel-bed river. Geomorphology 113, 411.
Heritage, G.L., Milan, D.J., Large, A.R.G., Fuller, I., 2009. Influence of survey strategy
and interpolation model upon DEM quality. Geomorphology 112, 334344.
Hetherington, D., Heritage, G.L., Milan, D.J., 2005. Reach scale sub-bar dynamics
elucidated through oblique LiDAR survey. International Association of
Hydrological Scientists Red Book Publication, IAHS Publication 291, 278284.
Hickin, A.S., Kerr, B., Barchyn, T.E., Paulen, R.C., 2009. Using ground-penetrating
radar and capacitively coupled resistivity to investigate 3-D fluvial architecture
and grain-size distribution of a gravel floodplain in Northeast British Columbia,
Canada. Journal of Sedimentary Research 79, 457477.
Hodge, R., Brasington, J., Richards, K., 2009a. In situ characterization of grainscale fluvial morphology using Terrestrial Laser Scanning. Earth Surface
Hodge, R., Brasington, J., Richards, K., 2009b. Analysing laser-scanned digital
terrain models of gravel bed surfaces: linking morphology to sediment transport
processes and hydraulics. Sedimentology 56, 20242044.
158
Hoechstetter, S., Walz, U., Dang, L.H., Thinh, N.X., 2008. Effects of topography and
surface roughness in analyses of landscape structurea proposal to modify the
existing set of landscape metrics. Landscape Online 1, 114.
Hohle, J., Hohle, M., 2009. Accuracy assessment of digital elevation models by
means of robust statistical methods. ISPRS Journal of Photogrammetry and
Remote Sensing 64, 398406.
Hong, S.-H., Wdowinski, S., Kim, S.-W., 2010. Evaluation of TerraSAR-X
observations for wetland InSAR application. IEEE Transactions on Geoscience
Hooke, R., Le, B., 2000. On the history of humans as geomorphic agents. Geology
28, 843846.
Horton, R.E., 1945. Erosional development of streams and their drainage basins:
Hydrophysical approach to quantitative morphology. Geological Society of
America Bulletin 56, 275370.
Houben, P., 2008. Scale linkage and contingency effects of field-scale and hillslopescale controls of long-term soil erosion: anthropogeomorphic sediment flux in
agricultural loess watersheds of southern Germany. Geomorphology 101,
172191.
Hutchinson, M.F., 1989. A new procedure for gridding elevation and stream line
data with automatic removal of spurious pits. Journal of Hydrology 106,
211232.
Isenburg, M., Liu, Y., Shewchuk, J., Snoeyink, J., Thirion, T., 2006, Generating
raster DEM from mass points via TIN streaming. In Raubal, H.J. Miller, A.U.
Frank, M.F. Goodchild (Eds.). Proceedings of the Geographic Information
Science, 4th International Conference, GIScience, Munster, Germany, September
2006M. 4197 of Lecture Notes on Computer Science, Springer, Berlin,
pp. 186198.
Jaeger, R.M., Schuring, D.J., 1966. Spectrum analysis of terrain of Mare Cognitum
(Lunar surface roughness in Mare Cognitum analyzed spectrally and compared
with two terrestrial areas). Journal of Geophysical Research 71, 20232028.
James, L.A., 2006. Bed waves at the basin scale: implications for river management
and restoration. Earth Surface Processes and Landforms 31, 16921706.
Jo, M.-J., Won, J.-S., Kim, S.-W., Jung, H.-S., 2010. A time-series SAR
observation of surface deformation at the southern end of the San Andreas Fault
Zone. Geosciences Journal 14, 277287.
Jones, A.F., Brewer, P.A., Johnstone, E., Macklin, M.G., 2007. High-resolution
interpretative geomorphological mapping of river valley environments using
airborne LiDAR data. Earth Surface Processes and Landforms 32, 15741592.
Katzenbeisser, R., 2003. About the calibration of LIDAR sensors. Proceedings of
ISPRS Workshop 3D Reconstruction from Airborne Laser-Scanner and InSAR
data, Dresden, Germany. http://www.isprs.org/publications/archives.html
Kellndorfer, J.M., Walker, W.S., LaPoint, E., Kirsch, K., Bishop, J., Fiske, G., 2010.
Statistical fusion of lidar, InSAR, and optical remote sensing data for forest
stand height characterization: a regional-scale method based on LVIS, SRTM,
Landsat ETM , and ancillary data sets. Journal of Geophysical Research. 115,
G00E08.
Kidner, D.B., Smith, D.H., 1993. Terrain modelling and ground cover data for
propagation studies, IEE Colloquium on Terrain Modelling and Ground Cover
Data for Propagation Studies. The Institution of Electrical Engineers, London,
7/1-7/9.
Kim, S.-W., Wdowinski, S., Dixon, T.H., Amelung, F., Kim, J.W., Won, J.-S., 2010.
Measurements and predictions of subsidence induced by soil consolidation
using persistent scatterer InSAR and a hyperbolic model. Geophysical Research
Letters 37, L05304.
Kimerling, A.J., Buckley, A.R., Muehrcke, P.C., Muehrcke, J.O., 2009. Map Use:
Reading and Analysis. ESRI Press, Redlands, CA, 493 pp.
Kobler, A., Pfeifer, N., Ogrinc, P., Todorovski, L., Ostir, K., Dzeroski, S., 2007.
Repetitive interpolation: a robust algorithm for DTM generation from aerial laser
scanner data in forested terrain. Remote Sensing of Environment 108, 923.
Kraus, K., Pfeifer, N., 1998. Determination of terrain models in wooded areas with
airborne laser scanner data. ISPRS Journal of Photogrammetry and Remote
Sensing 53, 193203.
Kraus, K., Karel, W., Briese, C., Mandlburger, G., 2006. Local accuracy measures
for digital terrain models. The Photogrammetric Record 21, 342354.
Kugler, H., 1964. Die geomorphologische Reiefanalyse als Grundlage
grossmassstabiger geomorphologischer Kartierung. Wissenschaft
Veroffentlichungen der Deustsche Institut fur. Landerkunde N.F. 21/22, 541655.
Lane, SN., 1998. The use of digital terrain modelling in the understanding of
dynamic river channel systems. In: Lane, SN, Richards, K, Chandler, J (Eds.),
Landform Monitoring, Modelling and Analysis. Wiley, Chichester, pp. 311342.
Lane, S.N., Chandler, J.H., Richards, K.S., 1994. Developments in monitoring and
modeling small-scale river bed topography. Earth Surface Processes and
Lane, S.N., Chandler, J.H., Porfiri, K., 2001. Monitoring river channel and flume
surfaces with digital photogrammetry. Journal of Hydraulic Engineering 127,
871877.
Lane, S.N., James, T.D., Crowell, M.D., 2000. Application of digital photogrammetry
to complex topography for geomorphological research. Photogrammetric Record
16, 793821.
Lane, S.N., Richards, K.S., 1997. Linking river channel form and process: time,
space and causality revisited. Earth Surface Processes and Landforms 22,
249260.
Lauknes, T.R., Piyush Shanker, A., Dehls, J.F., Zebker, H.A., Henderson, I.H.C.,
Larsen, Y., 2010. Detailed rockslide mapping in northern Norway with small
baseline and persistent scatterer interferometric SAR time series methods.
Remote Sensing of Environment 114, 20972109.
Lee, H.S., Younan, N.H., 2003. DTM extraction of LiDAR returns via adaptive
processing. IEEE Transactions on Geoscience and Remote Sensing 41,
20632069.
Li, Z., Zhu, Q., Gold, C., 2005. Digital Terrain Modeling: Principles and
Methodology. CRC Press, Boca Raton, FL, 323 pp.
Lim, M., Rosser, N.J., Allison, R.J., Petley, D.N., 2010. Erosional processes in the
hard rock coastal cliffs at Staithes, North Yorkshire. Geomorphology 114,
1221.
Lin, Z., Oguchi, T., 2006. DEM analysis on longitudinal and transverse profiles of
steep mountainous watersheds. Geomorphology 78, 7789.
Linsen, L., 2001. Point cloud representation. Technical report 2001-3, Science, FOC
University of Karlsruhe, Germany, 18 pp.
Liu, X., 2008. Airborne LiDAR for DEM generation: some critical issues. Progress
in Physical Geography 32, 3149.
Maas, H., 2002. Methods for measuring height and planimetry discrepancies in
airborne laser scanner data. Phogogrammetric Engineering and Remote Sensing
68, 933940.
MacMillan, R.A., Shary, P.A., 2009. Landforms and landform elements in
Software, and Applications. Elsevier, Oxford, U.K., pp. 227254.
Mackin, J.H., 1948. Concept of the graded river. Geological Society of America
Bulletin 59, 463512.
Mandlburger, G., Hauer, C., Hofle, B., Habersack, H., Pfeifer, N, 2008. Optimisation
of LiDAR derived terrain models for river flow modelling, Hydrology and Earth
System Sciences Discussions 5, 36053638.
Mark, D.M., 1974. Line intersection method for estimating drainage density.
Geology 2, 235236.
Martin-Vide, J.P., Ferrer-Boix, C., Ollero, A., 2010. Incision due to gravel mining:
modeling a case study from the Gallego River, Spain. Geomorphology 117,
261271.
Marzolff, I., Poesen, J., 2009. The potential of 3D gully monitoring with GIS using
high-resolution aerial photography and a digital photogrammetry system.
changes in the Earths surface. Reviews in Geophysics 36, 441500.
Matthews, N.A., 2008. Aerial and close-range photogrammetric technology:
providing resource documentation, interpretation, and preservation. Technical
Note 428. US Department of the Interior, Bureau of Land Management, National
Operations Center, Denver, Colorado, 42 pp.
Maune, D.F., 2007. Digital Elevation Model Technologies and Applications: The
Dem Users Manual. American Society for Photogrammetry and Remote Sensing,
Bethesda, MD.
Maune, D.F., Kopp, S.M., Crawford, C.A., Zervas, C.E., 2007. Introduction.
In: Maune, D.F. (Ed.), Digital Elevation Model Technologies and Applications:
The Dem Users Manual. American Society for Photogrammetry and Remote
Sensing, Bethesda, MD, pp. 135.
McCoy, S.W., Kean, J.W., Coe, J.A., Staley, D.M., Wasklewicz, T.A., Tucker, G.E.,
2010. Evolution of a natural debris flow: in situ measurements of flow
dynamics, video imagery, and terrestrial laser scanning. Geology 38, 735738.
McMaster, K.J., 2002. Effects of digital elevation model resolution on derived
stream network positions. Water Resources Research 38, 13-113-8.
Melton, M.A., 1957. An analysis of the relations among elements of climate, surface
properties and geomorphology. Columbia University, Office of Naval Research.
Project NR 389042, Technical Report 11.
Merritt, W.S., Letcher, R.A., Jakeman, A.J., 2003. A review of erosion and sediment
transport models. Environmental Modelling and Software 18, 761799.
Miller, C.L., Leflamme, R.A., 1958. The digital terrain model - Theory and
application. Photogrammetric Engineering 24, 433442.
Millette, T.L., Argow, B.A., Marcano, E., Hayward, C., Hopkinson, C.S., Valentine, V.,
2010. Salt marsh geomorphological analyses via integration of multitemporal
multispectral remote sensing with LIDAR and GIS. Journal of Coastal Research
26, 809816.
Mills, J.P., Buckley, S.J., Mitchell, H.L., Clarke, P.J., Edwards, S.J., 2005. A
geomatics data integration technique for coastal change monitoring. Earth
Surface Processes and Landforms 30, 651664.
Minar, J., Evans, I.S., 2008. Elementary forms for land surface segmentation: the
theoretical basis of terrain analysis and geomorphological mapping.
Mitas, L., Mitasova, H., 1999. Spatial interpolation. In: Longley, P.A., Goodchild,
M.F., Maguire, D.J., Rhind, D.W. (Eds.), Geographical Information Systems.
Wiley, New York, pp. 481492.
Mitasova, H., Hofierka, J., 1993. Interpolation by regularized spline with tension: II.
Application to terrain modeling and surface geometry analysis. Mathematical
Geology 25, 657669.
Moenning, C., Dodgson, N., 2003. A new point cloud simplification algorithm.
Proceedings of the 3rd IASTED International Conference on Visualization,
Imaging and Image Processing. Benalmadena, Spain, pp. 10271033.
Morisawa, M., 1988. The Geological Society of America Bulletin and the
development of quantitative geomorphology. Geological Society of America
Bulletin 100, 10161022.
Murzyn, F., Mouaze., D., Chaplin, J.R., 2006. Flow visualization and free surface
length scales measurements in a horizontal jet beneath a free surface.
Experimental Thermal and Fluid Science 30, 703710.
Nolan, M., Fatland, D.R., Hinzman, L., 2009. DInSAR measurement of soil moisture.
IEEE Transactions on Geoscience and Remote Sensing 41, 28022813.
Noma, A.A., Misulia, M.G., 1959. Programming topographic maps for automatic
terrain model construction. Surveying Mapping 19, 355366.
National Digital Elevation Program (NDEP), 2004. Guidelines for Digital Elevation
Data v.1. www.ndep.gov/NDEP_Elevation_Guidelines_Ver1_10May2004.pdf
(visited 6-2-10), 93 pp.
Okada, Y., Hirao, C., Horiuchi, T., Hara, Y., Yedidia, J.S., Azarbayejani, A., Oishi, N.,
2007. Highly accurate DSM reconstruction using Ku-band Airborne InSAR, IEEE
International Geoscience and Remote Sensing Society Symposium (IGARSS),
July 2007, pp. 50495052.
Oppikofer, T., Jaboyedoff, M., Blikra, L., Derron, M.-H., Metzger, R., 2009.
Characterization and monitoring of the Aknes rockslide using terrestrial laser
scanning. Natural Hazards and Earth System Science 9, 10031019.
Pauly, M., Gross, M., Kobbelt, L.P., 2002. Efficient simplification of point-sampled
surfaces. Proceedings of the 13th IEEE Visualization Conference, Boston, MA,
USA, pp. 163170.
Peipe, J., Stephani, M., 2003. Performance evaluation of a 5 megapixel digital
metric camera for use in architectural photogrammetry. The International
Archives of the Photogrammetry, Remote Sensing and Spatial Information
Sciences 34, 259261.
Peipe, J. and Tecklenburg, W., 2006. Photogrammetric camera calibration software a comparison. Proceedings of the ISPRS Commission V Symposium Image
Engineering and Vision Metrology, Dresden, Germany, 4 pp.
Pelletier, J.D., 1999. Self-organization and scaling relationships of evolving river
networks. Journal of Geophysical Research 104, 73597375.
Perron, J.T., Kirchner, J.W., Dietrich, W.E., 2008. Spectral evidence of characteristic
spatial scales and nonfractal structure in landscapes. Journal of Geophysical
Research 113, F04003.
Petrie, G., Toth, C.K., 2009a. Introduction to laser ranging, profiling and
scanning. In: Shan, J., Toth, C.K. (Eds.), Topographic Laser Ranging
and Scanning: Principles and Processing. CRC Press, Boca Raton, FL,
pp. 127.
Petrie, G., Toth, C.K., 2009b. Terrestrial laser scanners. In: Shan, J., Toth, C.K.
(Eds.), Topographic Laser Ranging and Scanning: Principles and Processing.
CRC Press, Boca Raton, FL, pp. 87127.
Petrie, G., Toth, C.K., 2009c. Airborne and spaceborne laser profilers and scanners.
In: Shan, J., Toth, C.K. (Eds.), Topographic Laser Ranging and Scanning:
Principles and Processing. CRC Press, Boca Raton, FL, pp. 2985.
Pfeifer, N., Mandlburger, G., 2009. LiDAR data filtering and DTM generation.
In: Shan, J., Toth, C.K. (Eds.), Topographic Laser Ranging and Scanning. CRC
Press, Boca Raton, FL, USA, pp. 308333.
Phillips, J.D., 1999a. Earth surface systems. Complexity, order and scale. Basil
Blackwell, Oxford.
Phillips, J.D., 1999b. Divergence, convergence, and self-organization in landscapes.
Annals of the Association of American Geographers 89, 466488.
Phillips, J.D., 2005. Entropy analysis of multiple scale causality and qualitative
causal shifts in spatial systems. The Professional Geographer 57, 8393.
Phillips, J.D., 2006. Deterministic chaos and historical geomorphology: a review
and look forward. Geomorphology 76, 109121.
159
Phillips, J.D., 2012. Synchronization and scale in geomorphic systems.

Pike, R.J., 1995. Geomorphometry progress, practice, and prospect. Zeitschrift fur
Geomorphologie, Supplement band 101, 221238.
Pike, R.J., Evans, I.S., Hengl, T., 2009. Geomorphometry: a brief guide. In: Hengl,
T., Reuter, H.I. (Eds.), Geomorphometry: Concepts, Software, and Applications.
Elsevier, Oxford, U.K., pp. 330.
Plattner, C., Wdowinski, S., Dixon, T.H., Biggs, J., 2010. Surface subsidence
induced by the Crandall Canyon Mine (Utah) collapse: InSAR observations and
elasto-plastic modelling. Geophysical Journal International 183, 10891096.
Poland, M., 2010. Localized surface disruptions observed by InSAR during strong
earthquakes in Java and Hawaii. Bulletin of the Seismological Society of
America 100, 532540.
Popescu, S.C., Zhao, K., 2008. A voxel-based LiDAR method for estimating crown
base height for deciduous and pine trees. Remote Sensing of Environment 112,
761781.
Prokop, A., Panholzer, H., 2009. Assessing the capability of terrestrial laser
scanning for monitoring slow moving landslides. Natural Hazards and Earth
System Science 9, 19211928.
Quinn, P., Beven, K., Chevallier, P., Planchon, O., 1991. The prediction of hillslope
flow paths for distributed hydrological modelling using digital terrain models.
Raper, J., Livingstone, D., 1995. Development of a geomorphological spatial model
using object-oriented design. International Journal of Geographical Information
Systems 9, 359383.
Rapp, A., Nyberg, R., 1981. Alpine debris flows in northern Scandinavia:
morphology and dating by lichenometry. Geografiska Annaler Series A, Physical
Geography 63, 183196.
Redmondino, F., El-Hakim, S., 2006. Image-based 3D modelling: a review. The
Photogrammetric Record 21, 269291.
Remondino, F., Fraser, C., 2006. Digital camera calibration methods: consideration
and comparisons. The International Archives of the Photogrammetry and Remote
Sensing Commission V Symposium Image Engineering and Vision Metrology,
Dresden, Germany, vol. 36, pp. 266272.
Reuter, H.I., Hengl, T., Gessler, P., Soille, P., 2009. Preparation of DEMs for
geomorphometric analyses. In: Hengl, T., Reuter, H.I. (Eds.), Geomorphometry:
Geomorphometry: Concepts, Software, Applications. Developments in Soil
Science. Elsevier, Amsterdam, vol. 33, pp. 87120.
Rhoads, B.L., 2006. The Dynamic Basis of Geomorphology Reenvisioned. Annals of
the Association of American Geographers 96, 1430.
Rignot, E., Echelmeyer, K., Krabill, W., 2001. Penetration depth of interferometric
synthetic-aperture radar signals in snow and ice. Geophysical Research Letters
28, 35013504.
Robeson, S.M., 1997. Spherical methods for spatial interpolation: review and
evaluation. Cartography and Geographic Information Systems 24, 320.
Rodrguez-Iturbe, I., Rinaldo, A., 2001. Fractal River Basins: Chance and SelfOrganization. Cambridge University Press, Cambridge, U.K., 564 pp.
Roering, J., Kirchner, J., Dietrich, W.E., 1999. Evidence for nonlinear, diffusive
sediment transport on hillslopes and implications for landscape morphology.
Water Resources Research 36, 853870.
Roering, J.J., Stimely, L.L., Mackey, B.H., Schmidt, D.A., 2009. Using DInSAR,
airborne LiDAR, and archival air photos to quantify landsliding and sediment
transport. Geophysical Research Letters 36, L19402.
Roggero, M., 2001. Airborne laser scanning: clustering in raw data. International
Archives of Photogrammetry, Remote Sensing and Spatial Information Sciences
34, 227232.
Rosen, P.A., Hensley, S., Joughin, I.R., Li, F.K., Madsen, S.N., Rodriguez, E.,
Goldstein, R.M., 2000. Synthetic aperture radar interferometry. Proceedings of
the IEEE 88, 333382.
Rossignac, J., Borrel, P., 1993. Multiresolution 3D approximations for rendering
complex scenes. Modeling in Computer Graphics: Methods and Applications.
Springer, New York, 455-465.
Rubin, D.M., Tsoar, H., Blumberg, D.G., 2008. A second look at western Sinai seif
dunes and their lateral migration. Geomorphology 93, 335342.
Sakaguchi, Y., 1969. A theory of relief forming. Bulletin Department of Geography
University of Tokyo 1, 3366.
Schmidt, J., Andrew, R., 2005. Multi-scale landform characterization. Area 37,
341350.
Schumm, S.A., 1956. Evolution of drainage basins and slopes in badlands at Perth
Amboy, New Jersey. Bulletin of Geological Society of America 67, 597646.
Shaffer, E., Garland, M., 2001. Efficient adaptive simplification of massive meshes.
Proceedings of the Conference on Visualization 01, Washington DC.,
pp. 127134.
160
Shan, J., Sampath, A., 2005. Urban DEM generation from raw LiDAR data: a
labeling algorthm and its performance. Photogrammetric Engineering and
Shary, P.A., Sharaya, L.S., Mitusov, A.V., 2002. Fundamental quantitative methods of
land surface analysis. Geoderma 107, 132.
Shary, P.A., Sharaya, L.S., Mitusov, A.V., 2005. The problem of scale-specific and
scale-free approaches in geomorphometry. Geografia Fisica e Dinamica
Quaternaria 28, 81101.
Shi, W.Z., Tian, Y., 2006. A hybrid interpolation method for the refinement of a
regular grid digital elevation model. International Journal of Geographical
Shrestha, R., Cather, W., Slatton, C., Dietrich, W., 2007. Research-Quality Airborne
Laser Swath Mapping: The Defining Factors. National Center for Airborne Laser
Mapping White Paper (v. 1.2), http://www.ncalm.cive.uh.edu/assets/
publication_pdf/NCALM_WhitePaper_v1.2.pdf.
Shugar, D.H., Rabus, B., Clague, J.J., 2010. Elevation changes (19491995) of
Black Rapids Glacier, Alaska, derived from a multi-baseline InSAR DEM and
historical maps. Journal of Glaciology 56, 625634.
Silvan-Cardenas, J.L., Wang, L., 2006. A multiresolution approach for filtering
LiDAR altimetry data. ISPRS Journal of Photogrammetry and Remote Sensing
61, 1122.
Sithole, G., 2001. Filtering of laser altimetry data using a slope adaptive filter.
International Archives of Photogrammetry, Remote Sensing and Spatial
Information Sciences 34, 203210.
Sithole, G., Vosselman, G., 2004. Experiomental comparison of filter algorithms for
for bare-Earth extraction from airborne laser scanning point clouds. ISPRS
Journal of Photogrammetry and Remote Sensing 59, 85101.
Smith, B., Mark, D.M., 2003. Do mountains exist? Toward an ontology of
landforms. Environment and Planning B: Planning and Design 30, 411427.
Solberg, S., Astrup, R., Gobakken, T., Nsset, E., Weydahl, D.J., 2010. Estimating
spruce and pine biomass with interferometric X-band SAR. Remote Sensing of
Southgate, H.N., Moller, N.K., 2000. Fractal properties of coastal profile evolution at
Duck, North Carolina. Journal of Geophysical Research 105, 1148911507.
Staley, D.M., Wasklewicz, T.A., Blaszczynksi, J.S., 2006. Surficial patterns of debris
flow deposition on alluvial fans in Death Valley, CA using airborne laser swath
mapping data. Geomorphology 74, 152163.
Stark, C.P., Barbour, J.R., Hayakawa, Y., et al., 2010. The climatic signature of
incised river meanders. Science 327, 14971501.
Stott, T., 2002. Bedload transport and channel bed changes in the proglacial Skeldal
River, Northeast Greenland. Arctic, Antarctic, and Alpine Research 34, 334344.
Strahler, A.N., 1952. Dynamic basis of geomorphology. Bulletin Geological Society
of America 63, 923938.
Strahler, A.N., 1958. Dimensional analysis applied to fluvially eroded landforms.
Geological Society of America Bulletin 69, 279300.
Strahler, A.N., 1992. Quantitative/dynamic geomorphology at Columbia 1945-60: a
retrospective. Progress in Physical Geography 16, 6584.
Sturzenegger, M., Stead, D., 2009. Close-range terrestrial digital photogrammetry
and terrestrial laser scanning for discontinuity characterization on rock cuts.
Thoms, M.C., Parsons, M., 2003. Identifying spatial and temporal patterns in the
hydrological character of the Condamine-Balonne River, Australia, using
multivariate statistics. River Research and Applications 19, 443458.
Tobler, W.R., 1969. An analysis of a digitized surface. In: Davis, C.M. (Ed.), A Study
of the Land Type, Final Report on Contract No. DA-31-124-ARO-D-456 to the
U.S. Army Research Office, Durham, NC, Department of Geography, University of
Michigan, Ann Arbor, pp. 5983.
Tobler, W.R., Davis, C.M., 1968. A digital terrain library. Technical Report ORA
08055, Department of Geography, University of Michigan, Ann Arbor, 28 pp.
Tucker, G.E., Bradley, D.N., 2010. Trouble with diffusion: reassessing hillslope
erosion laws with a particle based model. Journal of Geophysical Research 115,
F00A10.
Tucker, G.E., Bras, R.L., 1998. Hillslope processes, drainage density, and landscape
morphology. Water Resources Research 34, 27512764.
Tucker, G.E., Hancock, G.R., 2010. Modelling landscape evolution. Earth Surface
Turk, G., 1992. Re-tiling polygonal surfaces. SIGGRAPH 92 Conference Proceedings
26, 5564.
U.S. Army Map Service, 1963. Computer production of terrain models.
Communication of the Association for Computing Machinery 6, 190191.
de Vente, J., Poesen, J., 2005. Predicting soil erosion and sediment yield at the
basin scale: scale issues and semi-quantitative models. Earth-Science Reviews
71, 95125.
de Vente, J., Poesen, J., Arabkhedri, M., Verstraeten, G., 2007. The sediment
delivery problem revisited. Progress in Physical Geography 31, 155178.
Volker, H.X., Wasklewicz, T.A., Ellis, M.A., 2007. A topographic fingerprint to
distinguish alluvial fan formative processes. Geomorphology 88, 3445.
Vosselman, G., 2000. Slope based filtering of laser altimetry data. International
Archives of Photogrammetry, Remote Sensing and Spatial Information Sciences
33, 958964.
Vosselman, G., Gorte, B., Sithole, G., Rabbani, T., 2004. Recognising structure in
laser scanner point clouds. International Archives of Photogrammetry, Remote
Sensing and Spatial Information Sciences 46, 3338.
Wainwright, J., Parsons, A.J., Muller, E.N., Brazier, R.E., Powell, M., Fenti, B., 2008.
A transport-distance approach to scaling erosion rates: 1. Background and
model development. Earth Surface Processes and Landforms 33, 813826.
Walker, I.J., 2010. Changing views in Canadian Geomorphology: are we seeing the
landscape for the processes? The Canadian Geographer 54, 261276.
Walker, J.P., Willgoose, G.R., 1999. On the effect of digital elevation model
accuracy on hydrology and geomorphology. Water Resources Research 35,
22592268.
Walling, D.E., 1983. The sediment delivery problem. Journal of Hydrology 65,
209237.
Walls, J.H., Houbolt, J.V., Press, H., 1954. Some measurements and power spectra
of runway roughness. NACA Technical Note 3305, Washington, DC.
Wasklewicz, T., Whitley, D., Volker, H., Staley, D., 2005. 3D Laser Scanning and
Intensity Mapping: A New Methodology for Mapping and Monitoring Rock Art
Sites. International Newsletter on Rock Art 41, 1625.
Wasklewicz, T.A., Hattanji, T., 2009. High-resolution Analysis of Channel Change in
a Headwater Stream after a Debris Flow, Ashio Mountains, Japan. Professional
Geographer 61, 231249.
Wasklewicz, T.A., Staley, D.M., Mihir, M., Seruntine, L., 2007. Virtual recording of
lichen species: integrating terrestrial laser scanning and GIS techniques.
Physical Geography 28, 183192.
Weber, D., Englund, E., 1992. Evaluation and comparison of spatial interpolators.
Mathematical Geology 24, 381391.
Wechsler, S.P., 1999. Digital Elevation Model (DEM) uncertainty: evaluation and
effect on topographic parameters. In Proceedings of the 1999 ESRI User
Conference, San Diego, California, http://gis.esri.com/library/userconf/proc99/
proceed/papers/pap262/p262.htm.
Wen, Y.-Y., Ma, K.-F., 2010. Fault geometry and distribution of asperities
of the 1997 Manyi, China (Mw 7.5), earthquake: integrated analysis
from seismological and InSAR data. Geophysical Research Letters 37, L05303.
Wheaton, J.M., Brasington, J., Darby, S.E., Merz, J.E., Pasternack, G.B., Sear, D.A.,
Vericat, D., 2010b. Linking Geomorphic Changes to Salmonid Habitat at a Scale
Relevant to Fish. River Research and Applications 26, 469486.
Wheaton, J.M., Brasington, J., Darby, S.E., Sear, D., 2010a. Accounting for
Uncertainty in DEMs from Repeat Topographic Surveys: Improved Sediment
Budgets. Earth Surface Processes and Landforms 35, 136156.
Wilson, J.P., Gallant, J.C., 2000. Terrain Analysis. Principles and Applications.
Wiley, New York, 479 pp.
Wise, S., 2000. Assessing the quality for hydrological applications of digital elevation
models derived from contours. Hydrological Processes 14, 19091929.
Xu, C., Liu, Y., Wen, Y., Wang, R., 2010. Coseismic slip distribution of the 2008
Mw 7.9 Wenchuan Earthquake from joint inversion of GPS and InSAR data.
Bulletin of the Seismological Society of America 100, 27362749.
Zebker, H.A., Goldstein, R.M., 1986. Topographic mapping from Interferometric
Synthetic Aperture Radar observations. Journal of Geophysical Research 91,
49934999.
Zhang, K., Whitman, D., 2005. Comparison of three algorithms for filtering airborne
LiDAR data. Photogrammetric Engineering and Remote Sensing 71, 313324.
Zhang, K.Q., Chen, S.C., Whitman, D., Shyu, M.L., Yan, J.H., Zhang, C.C., 2003.
A progressive morphological filter for removing nonground measurements from
airborne LiDAR data. IEEE Transactions on Geoscience and Remote Sensing 41,
872882.
Zhang, W., Montgomery, D.R., 1994. Digital elevation model grid size, landscape
representation, and hydrologic simulations. Water Resources Research 30,
10191028.
Zhou, X., Chang, N.-B., Li, S., 2009. Applications of SAR interferometry in Earth
and environmental science research. Sensors 9, 18761912.
Zimmerman, D., Pavlik, C., Ruggles, A., Armstrong, M., 1999. An experimental
comparison of ordinary and universal kriging and inverse distance weighting.
Mathematical Geology 31, 375390.
161
Biographical Sketch
Thad Wasklewicz is an Associate Professor in the Department of Geography at East Carolina University. Wasklewicz is a geomorphologist interested in debris-flow processes as they relate to the evolution of hillslopes,
channels, and alluvial fans. Research and teaching experiences include the application of GIS, ALS, TLS, TPT, and
GPS to geomorphic studies.
Dennis Staley (PhD) is a research scientist with the United States Geological Survey, Central Region Geologic
Hazards Team in Denver, CO, USA. Staley has research experience in geomorphology with special interest on mass
wasting and expertize in GIS, ALS, TLS, TPT, and GPS. His current research involves postwildfire debris flow
susceptibility, hazard assessments, and warning system development.
Takashi Oguchi is the Vice-Director and a Prof. at the Center for Spatial Information Science at the University of
Tokyo. Prof. Oguchi has been working on the application of GIS to various fields of earth science, especially
geomorphology, hydrology, and Quaternary science. Oguchi is also Co-Editor-in-Chief of the journal
Geomorphology.
Kathryn Reavis is an undergraduate student in the Department of Geography at East Carolina University. Reaviss
research-educational experiences focus on geomorphology with applications of GIS, TLS, TPT, and GPS.
3.7 Geomorphometry
JP Wilson, University of Southern California, Los Angeles, CA, USA
MP Bishop, Texas A&M University College Station, TX, USA
3.7.1
3.7.2
3.7.2.1
3.7.2.2
3.7.2.3
3.7.2.4
3.7.3
3.7.3.1
3.7.3.2
3.7.3.2.1
3.7.3.2.2
3.7.4
3.7.5
References
Introduction
Representation
Data Capture
Data Preprocessing and DEM Construction
Error and Artifacts
Land-Surface Parameters
Primary Parameters
Secondary Land-Surface Parameters
Hydrology
Climatology
Land-Surface Objects and Landforms
Conclusions
163
164
164
165
167
168
169
169
173
173
175
178
180
181
Abstract
The study of surface processes and landforms requires quantitative characterization of the topography. New theoretical/
conceptual and practical advances in understanding and mapping various aspects of geomorphological systems have
emerged from new geospatial data and analysis of the topography. This chapter describes geomorphometry, or the science
of quantitative land-surface characterization, and how it can be used to represent and sample the land surface, generate
digital elevation models (DEMs), correct errors and artifacts from surface models, compute land-surface parameters and
objects, and use various forms of quantitative information in different application domains to address or solve problems.
Glossary
portray the altitude field of the topography. In
geomorphology, it commonly takes the form of a raster data
layer representing a field of square tessellations. The
resolution of the grid cells is usually determined based on
the source data utilized and the desired scale for
representing the topography.
Digital terrain modeling Digital terrain modeling (DTM)
refers to a workflow process of acquiring data that samples
the altitude field, preprocessing the data to generate a digital
elevation model, and error and uncertainty analysis to
identify and remove systematic and random errors.
Error propagation Digital elevation models exhibit an
inherent error due to digital terrain modeling that includes
choice of sensor in data capture, algorithm and methods
selection in DEM generation, and effective implementation
of DEM error correction, given DEM error and uncertainty
analysis. Such errors are propagated in terrain analysis in

the computation of land-surface parameters and objects
and in modeling efforts that utilize parameters and objects.
Geomorphometry The discipline that is concerned with
the science of quantitative land-surface characterization.
Land-surface objects Land-surface objects are spatial
entities that represent a meaningful segmentation of the
Earths surface. They are generated from land-surface
parameters using a variety of algorithms and methods, and
generally relate to the morphology of the topography in
terms of landform elements, features, or functional units.
Terrain segmentation can also relate to topographic
position and structure, surface material, or process
domains, if properly defined and delineated.
Land-surface parameters Land-surface parameters are
also called geomorphometric parameters, and they attempt
to quantitatively characterize various aspects of the
topography. They can be defined and classified based on
geometry, scale, and by surface-process characterization.
Wilson, J.P., Bishop, M.P., 2013. Geomorphometry. In: Shroder, J. (Editor in

Chief), Bishop, M.P. (Ed.), Treatise on Geomorphology. Academic Press,
San Diego, CA, vol. 3, Remote Sensing and GIScience in Geomorphology,
pp. 162186.
162
http://dx.doi.org/10.1016/B978-0-12-374739-6.00049-X
Geomorphometry
They are used to generate land-surface objects and

characterize process mechanics in surface-process modeling.
A variety of parameters such as slope, slope azimuth,
curvature, surface roughness, and relief are used for
studying geomorphological systems and for
geomorphological mapping.
Segmentation Segmentation refers to the partitioning of
the land surface into meaningful spatial entities.
3.7.1
Introduction
The topography plays a fundamental role in modulating several

components of Earths dynamic systems including atmospheric,
geomorphic, hydrologic, ecologic, and geological processes.
The topography constrains the operational scale of surface
processes, and partially governs both climate and tectonic forcing (Molnar and England, 1990; Bishop et al., 2010; Koons
et al., 2012). The strength of the linkage between form and
process can range from weak to strong, and may or may not be
inherently visible on the landscape depending on the complexity of the topography. Nevertheless, moderate to strong
linkages have been observed, such that an understanding of the
nature of the land surface can provide insights and understanding of the nature and magnitude of several processes
(Hutchinson and Gallant, 2000). Consequently, it is necessary
to quantitatively characterize the land surface and segment the
topography into fundamental spatial units, as the topography
inherently represents the results of the interplay between various systems, and records an imprint of landscape dynamics
(over a limited time). Therefore, the utility of digital elevation
models (DEMs) and the analysis of topography (geomorphometry) have and will continue to revolutionize the field of
geomorphology, as critical information regarding process
mechanics, process domains, feedback mechanisms, polygenetic evolution, and landforms continues to be investigated
and generated. Furthermore, with the rapid proliferation of
geographic information technologies, new data, algorithms,
and analysis/modeling techniques allow new capabilities. These
capabilities represent the evolution of the field of geomorphometry, which, in its broadest sense, refers to the science of
quantitative land-surface characterization (Pike, 1995, 2000) or
digital terrain analysis. For more details regarding the definition
and terminologies used in geomorphometry, see Wilson and
Gallant (2000a), Li et al. (2005), Zhou et al. (2008), Pike et al.
(2009), Hengl and Reuter (2009), and Wilson (2012).
Modern geomorphometry focuses on the extraction of
land-surface parameters and the segmentation of the landscape into spatial entities/features (land-surface objects) from
digital topography. This characterization relies on the so-called
specific and general modes of geomorphometric analysis that
were first defined by Evans (1972). The specific mode of analysis attempts to describe discrete surface features (i.e.,
landforms), whereas the general mode attempts to describe
the continuous land surface. Pike et al. (2009) have since
updated these definitions, such that a land-surface parameter
is a descriptive measure of surface form (e.g., slope, slope
azimuth, curvature) and a land-surface object is a discrete
163
Segmentation is required for object-oriented analysis of

land-surface objects.
Surface-process modeling Surface-process modeling
refers to physics-based deterministic modeling that
attempts to characterize process mechanics and how
topography governs various process rates, and the
influence of various processes on topographic landscape
evolution.
surface feature (e.g., watershed, cirque, alluvial fan, drainage

network). Although this definition represents an improvement, it is worth noting that this is a somewhat arbitrary
distinction, and there are already research examples that
demonstrate that these two perspectives are closely linked
to one another (e.g., Gallant and Dowling, 2003; Deng and
Wilson, 2008). Furthermore, it is clear that formalizing such
linkages is necessary in geomorphological research to address
concepts such as surface-process overprinting and polygenetic
evolution and to address multiple perspectives in geomorphological mapping (Bishop et al., 2012).
Collectively, geomorphometry is a rapidly evolving and
complicated field. This is in part due to its multidisciplinary
nature and the inclusion of information technology. Similar to
the field of geographic information science, it is based on developments in a variety of fields including source and end-user
disciplines. It not only attempts to deal with theoretical/conceptual issues involving representation and spatio-temporal
variation, but also includes issues of data collection and analysis, numerical modeling, and the utilization of other domain
knowledge for conceptual and practical problem solving.
Consequently, the field of geomorphometry is based on a scientific treatment of land surface and its characterization that
accounts for surface processes and morphology. Rapid
evolution has been facilitated by geographic information
technology and the widespread availability of DEMs. Geomorphologists now have many new capabilities to manipulate
and extract information from a variety of data sources. Nevertheless, it is important to recognize the empirical nature of
many forms of spatial analysis and modeling, and many issues
raise important questions about the assumptions and validity
of various approaches (Bishop et al., 2001).
Many questions still remain, and geomorphologists must
be aware of the advantages and limitations associated with
various representations and data structures, metrics/indices,
spatial modeling approaches, and their utility for geomorphological investigations. Furthermore, investigators must
be familiar with the mathematical underpinnings of geomorphometric analysis in order to adequately use information
and to interpret the results (Bishop and Shroder, 2004a).
Collectively, many issues point to a series of key questions that
in general include: (1) How should the land surface be represented? (2) What preprocessing is required to produce a
useable DEM? (3) What approach to error and uncertainty
analysis is required? (4) What algorithms are best for producing land-surface parameters? (5) What methods are best
for producing land-surface objects? (6) Is there a need to
develop new parameters and objects to address a particular
164
Geomorphometry
problem? (7) What algorithms and approaches are best suited

for a particular mapping application or do methods even
exist? (8) Does an adequate model exist or is there a need to
develop or modify one?
Many of these questions relate to data sources, issues, and
capabilities, although it is important to note that considerable
research is still required to address a whole host of issues in
geomorphology. In many cases, the answers to these questions
may not be clear, as software-tool development has focused on
the tool-box approach, foregoing formal scientific treatments
to mapping and geographic information system (GIS)-based
analysis/modeling. These challenges, however, can also be
attributed to the steady growth in the development and sheer
number of parameters and algorithms for processing DEMs.
Consequently, the values of parameters depend on a variety of
factors, including the parameterization scheme, measurement
scale of data, computational scale of analysis, and many empirical parameters that are used to address several conceptual
issues. This represents a serious issue in geomorphology, as the
spatial patterns associated with many metrics/indices may not
represent real-world phenomena, and the use of empirical
parameters reduces the ability to formalize important aspects
of the gemorphological system (Bishop et al., 2012).
Notwithstanding these challenges, land-surface parameters
and objects have been adopted in a variety of applications and
environmental settings. Geomorphometry has been used to
predict the distribution of soil properties (e.g., Zhu et al.,
1997; Bishop and Minasny, 2005), model soil redistribution
(i.e., erosion and deposition) processes (e.g., Mitasova et al.,
1995), assess the likelihood of slope hazards (e.g., Guzzetti
et al., 2005; Kheir et al., 2007), model solar-radiation potential (e.g., Reuter et al., 2005), improve vegetation mapping
(e.g., Bolstad and Lillesand, 1992; Antonic et al., 2003),
analyze wildfire propagation (e.g., Hernandez Encinas et al.,
2007), assess the role of surface processes in mountain topographic evolution (Burbank et al., 1996; Bishop and Shroder,
2000; Bishop et al., 2003), and to predict water flow, drainage,
and flooding in many hydrological applications. More
applications are rapidly emerging, and geomorphologists play
an important role in the development and evaluation of approaches that are based on geomorphological concepts.
Collectively, this chapter examines the data sources and
methods used to generate DEMs, and how land-surface parameters and objects can be used in geomorphology. A typical
digital terrain analysis workflow is presented that includes basic
data capture, data preprocessing, DEM generation, computation of land-surface parameters and objects, landform classification, and surface-process modeling. Throughout, examples
are provided of how parameters and geomorphometric analyses can provide new opportunities for geomorphological research. The focus is on presenting current capabilities and
issues associated with the quantitative characterization of the
topography.
3.7.2
The generation of DEMs involves data modeling or representation choices, sampling the land surface, representing
and creating a surface model from the sampled heights, and
correcting the errors and artifacts in the surface model (Hengl

and Reuter, 2009). Each phase in the production of a DEM is
critical for determining its utility, and in assessing the amount
of error that will propagate through the analysis phase. The
general workflow is depicted in Figure 1.
3.7.2.1
Representation
Many topics in geomorphology and geomorphometry are inherently related to the spacetime representation of topography (Bishop and Shroder, 2004b). This topic is complex,
and a variety of philosophical, cognitive, and natural-science
perspectives exist. The current use of representation is dominated by static cartographic representations. Although this
approach provides many advantages in terms of spatial overlay, management of data, basic spatial analysis, and information distribution, it does not address many issues related to
surface processes and landforms.
Topographic variation can be represented in a variety of
ways using data models. The common data models are the
field (layer), entity (object), and network data models, which
can be linked to a relational data model (Goodchild, 1992).
These data models are represented in a computer using data
structures (i.e., raster and vector). Consequently, topography
can be represented by many field models (sampled points,
contours, polygons, tessellations, triangular nets) to characterize the continuous spatial variation in altitude. Object
models are used to define well-defined features, assuming that
discrete boundaries actually exist, whereas indeterminant
boundaries have been recognized to pose a unique challenge,
as environmental gradients or zones of homogeneous and
Applications
Interpretations
Visualization
DEM
DEM generation
Data capture
Reality
Figure 1 The main tasks associated with digital terrain modeling.

Reproduced from Hutchinson, M.F., Gallant, J.C., 2000. Digital
elevation models and representation of terrain shape. In: Wilson, J.P.,
Gallant, J.C. (Eds.), Terrain Analysis: Principles and Applications.
John Wiley and Sons, New York, pp. 2950.
Geomorphometry
heterogeneous surface properties can effectively represent

boundaries or limits to the spatial distribution of phenomena
(Burrough, 1996; Usery, 1996; Lagacherie et al., 1996). Earth
scientists have noted the advantages and disadvantages of such
data models and have recognized that these representations do
not effectively address process mechanics or dynamics (Raper
and Livingstone, 1995). It is also important to note that field
and object models do not formally represent the complex nature of landforms, as issues of scale, organization, composition,
and age must be taken into consideration. Furthermore, the
degree to which qualitative and quantitative analysis should be
used to characterize the land surface needs to be determined, as
qualitative analysis is supported by the entity view, whereas the
science of studying process mechanics, feedback mechanisms,
geodynamics and landscape evolution tends to focus on continuous space (Raper and Livingstone, 1995).
Clearly, determining how to best represent the land surface
is a complex issue. Dikau (1989) indicated that a digital relief
model involving the parameterization of relief units could be
used to represent topography that is hierarchically organized.
Relief is scale dependent, and the concept of homogeneous
relief can be defined based on distance and direction. Nevertheless, relief and many other land-surface parameters are
scale dependent. Does this mean that the scale dependence
and the anisotropic nature of the topography need to be
represented? Should the hierarchical spatial structure of the
topography be represented? Furthermore, how should process
mechanics, processform relationships, and temporal dynamics be characterized?
An intriguing proposition for geomorphometry has been
presented by Cova and Goodchild (2002) that involves the
extension of spatial representation to include fields of spatial
objects. This effectively represents the linking of continuous
space with object representation. It also allows considerable
flexibility in terms of representing the complexity associated
with landforms, as the issues of homogeneity, heterogeneity,
complexity, and other concepts can be addressed, as a tessellation can have more than one object and the objects can have
discrete or fuzzy boundaries. In addition, an object hierarchy
can be developed to address issues associated with scale.
Furthermore, it also allows the representation of process via
process objects, wherein a multitude of process objects can
simultaneously alter the topography at fundamentally different
scales. This allows the integration of process modeling and
mapping in a seamless way, and raises the important issue
of parameterization schemes for characterizing the process
mechanics and specific processform relationships. Such formal representations of the topography in geomorphology are
required to validate the results obtained via empirical analysis
using geospatial technologies. In addition, such a representation can handle temporally changing spatial patterns by using
a dynamic representational scheme that results from the process dynamics. Consequently, spatio-temporal relationships are
inherently represented. Several complexities associated with
3-D and temporal representation, however, remain.
3.7.2.2
Data Capture
The data sources and processing methods for generating DEMs

have evolved rapidly over the past 2030 years. Data collection
165
approaches (Table 1) have been categorized by Nelson

et al. (2009) and include: (1) ground-survey techniques
(electronic theodolites, total stations, electronic distance
measuring (EDM), and global positioning system (GPS)
units); (2) existing topographic maps (derivation of contours,
streams, lakes, and spot heights from existing hardcopy
topographic maps); and (3) remote sensing (both air and
space-borne optical, radar, and Light Detection and Ranging
(LiDAR) sensors).
Modern-day data collection is based on remote sensing for
the production of DEMs, as rapid progress has occurred over
the past two decades. Global DEM data products from the
Shuttle Radar Topographic Mission (SRTM) and the Advanced
Spaceborne Thermal Emission and Reflectance Radiometer
Global Digital Elevation Model (ASTER GDEM) were released
in 2000 and 2009, respectively. The 3 arc-second SRTM DEM
covers a large fraction of the globe (from 601 N to 581 S) and
has already emerged as one of the most consistent, complete,
and popular environmental datasets in the world (Zandbergen, 2008; Nelson et al., 2009). The 3 arc-second (B90 m)
grid spacing is much better than the 1 km spacing of the
worldwide GTOPO30 DEM, and an accuracy assessment using
kinematic GPS data showed good absolute height accuracy,
with 90% of the errors o5 m (Rodriguez et al., 2006). This
product must nevertheless be used carefully because: (1) it
represents a digital surface model (DSM) (not a bare-earth
model); (2) surface characteristics may affect accuracy; (3)
voids generally occur at landwater margins; (4) problems
may occur in desert and mountain areas due to foreshadowing
and shadowing effects (Rodriguez et al., 2005); and (5) the
current 90-m resolution provided by the global SRTM DEM is
not fine enough for accurate surface characterization and the
mapping of soils, vegetation, and many landforms (Gessler
et al., 2009). It represents, however, an excellent dataset for
geomorphometric analysis of mountain environments to
study the influence of glaciations and tectonics on landscape
evolution (Figure 2).
Some, but probably not all of the aforementioned problems, may be addressed by the ASTER GDEM. This relatively
new product offers better resolution (1 vs. 3 arc-seconds),
improved spatial coverage (831 N to 831 S), as well as comparable vertical and horizontal accuracy (Hiranoa et al., 2003;
Nelson et al., 2009; Slater et al., 2009). The release of this
product is so recent that there are few published reports
documenting the strengths and weaknesses of this new data
source for specific locations and applications, although the
improvement in the measurement scale should greatly facilitate geomorphological investigations (Figure 3).
Recently, LiDAR surveys have been conducted, resulting in
the generation of DEMs with improved resolution (Figure 4).
Consequently, the reliance on LiDAR surveys has increased
quickly, and this source now dominates local and regional
projects everywhere. Belgium and the Netherlands, for example, have produced national LiDAR DSMs at resolutions of
25 m, and much finer-resolution DEMs have been produced
for many smaller areas as well (Nelson et al., 2009). The advantages of using LiDAR include the high density of sampling,
high vertical accuracy, and the opportunity to derive a set
of surface models, given that some laser-scanning systems
can already provide at least two versions of the surface: the
166
Geomorphometry
Table 1 Key characteristics of data sources

Source
Resolution (m)
Accuracy
Footprint (km2)
Postprocessing
requirements
Elevation/
surface
Ground survey
Low
Elevation
Variable, but usually small
Low
Elevation
Depends on map footprint
Medium
Elevation
Medium
Elevation
High
Elevation
High
Surface
LiDAR
13
3050 h 1
High
Surface
InSAR/IfSAR
2.55
Very high vertical and

horizontal
Medium vertical and
horizontal
Medium vertical and
horizontal
Medium vertical and
horizontal
Medium vertical and
horizontal
Very high vertical and
horizontal
0.150.11 m vertical, 1 m
horizontal
12 m vertical, 2.510 m
horizontal
Variable, but usually small
Ortho photography
Variable but usually

o5 m
Variable but usually
o5 m
Depends on map scale
and contour interval
o1
Depends on method of
acquisition
High
Surface
SRTM, Band C
90 (30)
Almost global 601 N581 S
Potentially high
Surface
SRTM, Band X
30
16 m vertical, 20 m
horizontal
16 m vertical, 6 m
horizontal
Potentially high
Surface
ASTER
30
Similar to Band C, but only

every second path is
available
3600
Medium
Surface
SPOT
30
72 000 per swath
Medium
Surface
GPS
Table digitizing
On-screen digitizing
Scanned topo-map
750 m vertical, 750 m

horizontal
10 m vertical, 15 m
horizontal
Source: Modified from Nelson, A., Reuter, H.I., Gessler, P., 2009. DEM production methods and sources. In: Hengl, T., Reuter, H.I. (Eds.), Geomorphometry: Concepts, Software, and
Applications. Elsevier, Amsterdam, pp. 6585.
Figure 2 Shuttle Radar Topographic Mission (SRTM) 3 arc-second

DEM for the Shimshal Valley in northern Pakistan. The 90 m
resolution allows a relatively accurate geomorphometric
characterization of the region for studying surface processes and
tectonics.
vegetation canopy (first returns) and ground surface (last

returns), which should help with the modeling of erosion
and water budgets. The smaller footprint and measurement
challenges encountered in areas with tall buildings, densevegetation canopies, and water surfaces make this a relatively
Figure 3 Advanced Spaceborne Thermal Emission and Reflectance

Radiometer Global Digital Elevation Model (ASTER GDEM) for the Mt.
Everest region in Nepal. The 30 m resolution allows a more detailed
assessment of the mountains, although several errors and artifacts
are found in such high-mountain environments. ASTER GDEMs in a
less complex topography are usually of higher quality. The displayed
x and y dimensions are 216 km, respectively.
Geomorphometry
Figure 4 Five-meter resolution digital elevation model (DEM)

generated from bare-earth LiDAR data from the North Carolina
Floodplain Mapping Program. The area is located in the Blue-Ridge
Province of western North Carolina, and is approximately 232 km2 in
size. The region exhibits an elevation range of 867 m (1454587 m)
and has a mean slope of 171. The incised terrain in the southeast
portion of the DEM represents the beginning of the Blue-Ridge
Escarpment. DEM courtesy of Jeffrey Colby, Appalachian State
University.
expensive option, and several studies have documented how

the elevation accuracy of LiDAR data varies with both the
sensor system and the land-cover characteristics (e.g., Dowman, 2004; Hodgson et al., 2005).
3.7.2.3
Data Preprocessing and DEM Construction
Regardless of the representational scheme, preprocessing of

the elevation data for analysis can be difficult, as altitude is but
one of the land-surface characteristics that are required for
various applications. The morphological accuracy can only be
assessed by measuring surface parameters and objects such as
the slope angle, slope azimuth, curvature, drainage lines, and
landforms in the field, and then comparing their magnitudes,
shapes, distributions, and locations with the values obtained
by geomorphometric analysis (e.g., Fisher, 1998; Wilson et al.,
2008). Reuter et al. (2009) suggested that the true utility of
DEMs for geomorphometric analysis can only be assessed by
determining the following: (1) accuracy of surface-roughness
representation; (2) accuracy of land-surface morphology; (3)
accurate detection of actual ridge- and stream-lines; and (4)
spatial consistency of altitude measurements. Such issues are
interrelated, and the frequency and magnitude of errors will
depend on the technologies and methods used to collect the
source data, the preprocessing algorithms that are applied, and
the complexity of the land surface itself.
Not surprisingly, the horizontal and vertical resolution of the
elevation data used to portray a terrain surface will have a significant influence on the level of detail and the accuracy of the
167
DEM, and on the analysis results (MacMillan and Shary, 2009).

Several authors have documented the effects of grid spacing on
the value and accuracy of land-surface parameters and landform
objects (e.g., Zhang and Montgomery, 1994; Florinsky, 1998;
Jones, 1998; Wilson et al., 2000; Thompson et al., 2001; Shary
et al., 2002; Tang et al., 2002; Kienzle, 2004; Warren et al., 2004;
Zhou and Liu, 2004; Raaflaub and Collins, 2006). The increasing
interest in various forms of multiscale analysis (e.g., Bishop et al.,
2003; Gallant and Dowling, 2003; Sulebak and Hjelle, 2003;
Deng and Wilson, 2008; Bishop et al., 2012) and the need to be
able to move seamlessly across scales indicate that more research
on scale and its effects is required.
Other decisions made about unwanted depressions (i.e.,
spurious pits or sinks) will also have an impact on subsequent
geomorphometric analysis and interpretation of the results. Two
approaches have been utilized and include: (1) progressively
filling the sinks by increasing their elevation values until the
elevation of their lowest outflow point is reached (e.g., Jenson
and Domingue, 1988; Martz and de Jong, 1988; Soille and
Gratin, 1994; Planchon and Darboux, 2001; Wang and Liu,
2006) and (2) creating a descending path from the bottom of the
sink by carving the terrain along this path until the nearest point
is reached, which has an elevation lower than the bottom of the
sink (e.g., Reiger, 1992; Morris and Heerdegen, 1988; Martz and
Garbrecht, 1999; Soille et al., 2003; Soille, 2004). Reuter et al.
(2009) recently used both of these approaches along with one
that combined sink filling and carving, such that the sum of the
differences in elevation between the input and the output DEMs
that did not have sinks was minimized. Lindsay and Creed
(2005a, b, 2006) have also combined elements of the aforementioned approaches and used them to distinguish artifact and
real depressions in digital elevation data, and propose a minimum-impact approach for removing artifact depressions in
relatively flat landscapes. Grimaldi et al. (2007) have proposed
an alternative physically-based approach to remove spurious pits
as well.
At least two other related challenges may need to be addressed. The first concerns the problem of unresolved flow
directions on flat terrain, because the assignment of flow directions relies on the presence of elevation differences between
adjacent cells to drive the flow. The presence of lakes and
reservoirs, and reliance on the first of the aforementioned
approaches for filling sinks, may exacerbate this challenge by
creating artificial flat regions. Whatever the cause, one of two
approaches is typically used to remove or minimize these
kinds of problems. The first relies on an iterative procedure to
assign a single flow direction to a neighboring cell without
alteration of the elevation values (e.g., Jenson and Domingue,
1988), whereas the second method makes small alterations to
the elevation of the flat cell(s) in order to create a small artificial gradient (e.g., Garbrecht and Martz, 1997). The solutions
obtained with these approaches will vary slightly from one to
the other, and an in-depth knowledge of field conditions will
generally be required to know whether one approach produces superior results in most landscape settings.
The second challenge is the need to reconcile the DEM and
drainage lines acquired from some other dataset (Lindsay
et al., 2008). One approach relies on stream burning, where
the local topography is altered to provide consistency with
some existing vector hydrography dataset (Saunders and
168
Geomorphometry
Maidment, 1996) and the second method utilizes the stream

network as a part of the surface-fitting approach used to
generate square-grid DEMs (Hutchinson, 1989). The latter has
been used here almost exclusively in the authors hydrological
work, in part because of existing software tools that tackle the
past three problems (i.e., unwanted depressions, unresolved
flow directions on flat terrain, and reconciliation of elevation
and hydrography datasets) simultaneously.
The rapid growth in sources of mass-produced, remotesensing-derived DEMs during the past two decades has demanded new forms of DEM preprocessing. For example, Reuter
et al. (2009) and Webster and Dias (2006) described various
approaches and opportunities for ortho-rectifying DEMs, reducing local outliers and noise, filtering water surfaces, filtering
pure noise, filtering forests in SRTM DEMs, reducing padi terraces
(i.e., areas with closed contours where all the surrounding pixels
show the same value), filling voids and sinks, mosaicking adjacent DEMs, and filtering LiDAR DEMs. Some problems are
more difficult to fix than others, and several authors have noted
the presence of systematic and random errors that are not so easy
to detect and correct in LiDAR datasets (e.g., Filin, 2003; Katzenbeisser, 2003; LaLonde et al., 2010). The detection of varying
deflection errors and of time delays between measurements is
especially difficult and specific to the composition of the LiDAR
sensor system, and the large number of parameters that were
assigned when the individual sensor systems were manufactured
(Dowman, 2004). Reuter et al. (2009) also noted that two
additional trends have emerged in recent years. The first is the
integration of topographic and auxiliary information (such that
the location of lakes, streams, ridges, and breaks will be identified from satellites and incorporated into the DEM processing
chain) and the second is the increased use of fully data-driven
simulation methods that reduce some or all of the aforementioned errors by calculating the average value of the landsurface parameter from multiple equi-probable realizations of
the DEM (e.g., Burrough et al., 2000b; Hengl et al., 2004; Raaflaub and Collins, 2006).
3.7.2.4
Error and Artifacts
With new data sources and information technologies, DEMs

can now be produced in a variety of ways. Many subtleties are
embedded in the various data sources and methods, and
errors can be introduced at many stages of the production
process. Some can be attributed to the fact that these errors
may vary with the choice of the sensor and a specific application (i.e., method of deployment). Consequently, DEM
differencing cannot be accurately used to detect altitude variations for assessing change, erosion, deposition, etc. (Burns
et al., 2010). A second set of challenges concerns the propagation of the elevation errors in primary and secondary
land-surface parameters, and the considerable effort that is
generally required to identify them. The general approach
for propagating errors incorporates statistical modeling of the
error in the DEM (which is generally only partially known)
and running a Monte Carlo analysis (Temme et al., 2009).
Digital terrain modeling will utilize various techniques to
detect and remove some or all of these errors. Data source
errors, however, cannot always be eliminated, and those
interested in using land-surface parameters calculated from

DEMs must be cognizant of these errors and how they may
affect the analysis and interpretation of results.
It is worth reviewing what is known about the accuracy of
DEM elevation values and the land-surface parameters calculated from these elevations. Several approaches have been
proposed to assess the accuracy of DEM elevation values (e.g.,
Hutchinson, 2008; Temme et al., 2009). Many researchers
have compared DEM-derived altitudes with elevation values
taken from a more accurate source of topographic data,
computing the root-mean-square error (Wise, 2000). One
problem with this approach is that it ignores both the presence of systematic bias and the spatial pattern of errors, which
is critical for those land-surface parameters that are heavily
influenced by the shape of the land surface (Hutchinson and
Gallant, 2000; Deng et al., 2008). Carara et al. (1997) suggested simple criteria to evaluate DEM quality when the DEM
is constructed from contours: (1) the DEM should have the
same values as contours close to the contour lines; (2) the
DEM values must be in the range given by the bounding
contour lines; (3) the DEM values should vary almost linearly
between the values of the bounding contour lines; (4) the
DEM patterns must reflect realistic shapes in flat areas; and (5)
the artifacts must be limited to a small proportion of the data
set. Hutchinson and Gallant (2000) have suggested a larger
and more diverse list of simple metrics for measuring quality
for DEMs constructed from surface-specific point elevation
and contour- and stream-line data that incorporate some of
the same ideas, and a rapidly growing literature is documenting the quality of DEMs constructed from remotely
sensed sources (e.g., Carabajal and Harding, 2006; Hoften
et al., 2006; Rodriguez et al., 2006; Shortridge, 2006; Berry
et al., 2007; Bhang et al., 2007).
Whatever the source of the elevation data, a high-resolution
DEM may still have greater uncertainty than a low-resolution
DEM. Furthermore, the errors in DEMs may propagate to the
land-surface parameters and modeling results in ways that are
not easily predicted. See Bolstad and Stowe (1994), Band et al.
(1995), Desmet (1997), Hunter and Goodchild (1997), Wood
et al. (1997), Wise (1998), Holmes et al. (2000), Endreny and
Wood (2001), Aerts et al. (2003), Van Niel et al. (2004),
Lindsay and Creed (2005b), Fisher and Tate (2006), Lindsay
(2006), Lindsay and Evans (2006), and Chow and Hodgson
(2009), for examples spanning multiple DEM data sources and
land-surface parameters.
In one particularly impressive study, Temme et al. (2009)
examined the propagation of errors from DEMs in the
computation of the slope (a local parameter), the topographic-wetness index (a regional parameter), and the soil
redistribution resulting from water erosion (a complex model
output) in the Baranja Hill watershed in Croatia. The DEM
errors propagated strongly to slope (the mean coefficient
of variation across 100 Monte Carlo simulations was 42%
for unfilled DEMs and 49% for filled DEMs), but only moderately for the wetness index (the mean coefficient of variation
was 10% for unfilled and 16% for filled DEMs), although
the coefficient of variation for the index varied more spatially
than that of slope. These results show that the wetness index
values were less sensitive than slope to the input DEM, but
this may have been influenced by the flow-routing algorithm
Geomorphometry
(Holmgren, 1994) that was used to calculate upslope contributing areas.

Temme et al. (2009) next used the water-erosion module
of the LAPSUS landscape-evolution model (Schoorl et al.,
2000) to simulate erosion and deposition in the Baranja Hill
study area for 10 years. The model utilizes water flow and slope
to calculate a sediment-transport capacity, and calculates erosion and deposition by comparing this transport capacity with
the predicted amount of sediment in transport. The latter was
simulated with the same multiple flow-direction algorithm
used to calculate the wetness index, and the approach
of Temme et al. (2006) was used to handle the flows of water
and sediment into sinks. The latter capability was very important because it meant that the model could simulate erosion
and deposition using both unfilled and filled DEMs. The
results, at first glance, might be viewed as reassuring because the
general erosion and deposition patterns were similar for unfilled and filled DEMs, with erosion occurring in the upper
valleys and deposition occurring in flat areas. The mean
soil-redistribution maps of the 100 simulations on unfilled
DEMs, however, showed considerably more deposition and less
erosion than the filled DEMs (in part because the depressions
were filled before the model runs in the latter case), and the
results in both sets of model runs were very sensitive to errors in
the DEM (the mean coefficient of variation of soil redistribution was 4600% for unfilled and 1000% for filled DEMs).
Hence, the coefficients of variation were larger and more spatially variable for soil redistribution than they were for the
wetness index and slope, because the LAPSUS model results
were sensitive to three forms of error in the input DEM
those associated with the error in the DEM and those that were
introduced into the slope and topographic wetness index terms
by the same source error. These kinds of dependencies are likely
to be embedded in many of the analysis and modeling applications that incorporate one or more of the aforementioned
primary and secondary terrain parameters. Consequently, error
analysis is critical in geomorphometry, and preprocessing errorremoval techniques are usually required.
3.7.3
most frequently used parameters represent the first and second

derivatives of the altitude field. The two main geometric
properties are the average slope gradient, S, and the slope
azimuth, f. The average slope gradient accounts for orthogonal directions such that

qz qz
,
1
S
qx qy
where z represents the altitude, and x and y represent the
directions. Several algorithms or modeling approaches may
require the use of S, or the slope angle b, which is defined as
b arctanS
As
Acell
cos b
3
where Acell is the area of a grid cell. Slope is routinely used in

sediment-transport modeling, landform mapping, surfaceenergy-budget studies, and for characterizing various aspects
of process mechanics related to fluvial, mass movement, and
glacier-erosion dynamics (Figure 5).
The direction of the slope, or slope azimuth, is another
critical geometric property that governs water and sediment
flows, while also reflecting the orientation structure of the
topography as governed by lithology and structural influences.
Gallant and Wilson (1996) defined it as

q
p
90
f 180 arctan
4
p
9p9
Land-Surface Parameters
Primary Parameters
Many local parameters are calculated by moving a threeby-three window across a grid and computing land-surface
parameters for the target cell (i.e., the central cell in the threeby-three window). There are special rules on how to handle
the edges, and this approach produces a new grid or GIS layer
with the same dimensions as the DEM for each parameter. The
2
Slope information can also be expressed as slope percentage and surface area, As, which can be approximated by
Computations of land-surface parameters attempt to characterize various multiscale properties of the terrain, and are used
to extract land-surface objects. Consequently, many parameters can be classified based on geometry considerations,
scale, or use in process mechanics and numerical modeling.
Wilson and Gallant (2000a) classified them as primary and
secondary, given that they constitute the basic building blocks
for landform classification and other forms of more sophisticated analysis and modeling.
3.7.3.1
169
Figure 5 Slope-angle map generated from 5 m LiDAR DEM

displayed in Figure 4. Slope variations are significant in mountain
environments and can reflect variations in erosion, lithology, and
tectonics. Slope information is also critical for evaluating naturalhazard potential and for planning purposes.
170
Geomorphometry
where
p
qz
,
qx
qz
qy
5
Slope azimuth is a circular parameter and is frequently

transformed using cos f and sin f to examine linear trends
in northerness and easterness, respectively. Slope azimuth is also
important is solar-radiation modeling. As with slope, various
algorithms can be used for computation, and each will produce
slightly different estimates across a range of land-surface conditions. Details about the performance of various equations can
be found in Skidmore (1989) and Hengl and Evans (2009).
Terrain curvature is also frequently used to estimate the
magnitude of concavity and convexity of the land surface. The
convention followed in the Earth sciences is for a positive
curvature value to represent a convex surface shape, whereas a
negative value represents a concave surface shape (Olaya, 2009).
The profile (or vertical) curvature, Cprofile, and tangential (horizontal) curvature, Ctangential, are generally used to distinguish
locally convex and concave shapes. They are defined as
Cprofile
p2 r 2pqrs q2 t
q
p2 q2 1 p2 q2 3
6
q2 r 2pqs p2 t
p
1 p2 q2
7
Ctangential
p2 q2
Figure 6 Profile-curvature map generated from the 5 m LiDAR DEM

displayed in Figure 4.
where r, s, and t represent

r
q 2z
,
q x2
q 2z
,
q xq y
q 2z
q y2
8
Convex profile curvature indicates acceleration of flows

and a local increase in potential energy, whereas concave
profile curvature indicates a flattening of the slope and
therefore a decline in potential energy (Figure 6). Concave
tangential curvature indicates convergence and convex tangential curvature indicates divergence of flow lines, which may
in turn influence overland flow paths, soil moisture distribution, and soil redistribution (Figure 7). Planform curvature
is sometimes used to describe the curvature of contour
lines and should yield results similar to tangential curvature,
as long as the contour lines describe the shape of the land
surface (Gallant and Wilson, 2000). Other forms of curvature
can also be utilized, and these include mean curvature,
unsphericity curvature (Figure 8), Gaussian curvature, and
curvature of flow lines. See Olaya (2009) for the details on
their computation and their potential significance in the Earth
sciences.
Other local-statistical parameters can also be used to
characterize key aspects of the surface. Relief and surface
roughness represent two important parameters valuable
in geomorphology. Local relief is highly correlated to slope,
although a nonlinear relationship has been found in extreme
mountain environments that can be useful for differentiating
processform relationships, and the nature of the relationship
may be related to the magnitude of erosion and the rate of
relief production. It simply represents the range in the altitude
Figure 7 Tangential-curvature map generated from the 5 m LiDAR

DEM displayed in Figure 4.
values over the spatial extent of the computational window.

It is important to note that relief is scale dependent, and
examining relief variation at different scales highlights different aspects of the geomorphological system (i.e., tectonics at
larger scales). Similarly, surface roughness can be estimated in
a variety of ways. An interesting characterization makes use of
the vector approach such that
Xi sinb cosf,
Yi sinb sinf,
Zi cosb
9
Geomorphometry
171
Figure 9 Negative-openness map generated from the 5 m LiDAR

DEM displayed in Figure 4. A window search radius of 1 km was
used to highlight the scale-dependent hydrological network. This
parameter, when combined with other geomorphometric parameters,
can greatly facilitate the assessment of local and regional
hydrological conditions.
Figure 8 Unsphericity-curvature map generated from the 5 m LiDAR
DEM displayed in Figure 4. The unsphericity parameter highlights
land-surface conditions that are planar in nature.
where Xi, Yi, and Zi are the components of the unit vector
normal to the land surface. The surface-roughness factor (SRF)
can then be computed as
SRF
P
P
P
ni Xi 2 ni Yi 2 ni Zi 2
n
10
where n is the number of cells in a window. Other approaches

to characterizing surface roughness include semivariogram
analysis and the use of the fractal dimension to characterize
terrain complexity. Terrain roughness has utility related to
weathering studies, estimating aerodynamic drag, and other
surface applications.
The regional (i.e., nonlocal) land-surface parameters are
mainly concerned with the climatic, geomorphic, hydrologic,
or visual properties of landscapes. The first category relies
on the accurate delineation of the shadowing, sky-view, and
terrain-view nature of the surrounding topography as it influences irradiance, temperature, and precipitation. The geomorphic and hydrological parameters focus on the movement
of water and sediment and, as such, rely on the accurate delineation of flow paths, watersheds, and scale-dependent slope
and relief. The most common parameters are sediment flux,
upslope contributing area, flow-path length, and a variety
of statistical measures. For example, positive and negative
openness (Yokoyama et al., 2002) can be extremely valuable
to geomorphologists, as positive openness highlights areas of
extreme relief and has been used to discover active erosion
zones in the Himalaya (Bishop et al., 2010). Negative openness (Figure 9) highlights the scale-dependent hydrological
network, and has potential for use in mapping valley bottoms
and hydrological modeling. For the final category, the visibility (i.e., from what other points can a single point be seen
or the reverse, what other points can be seen from a single

point) can be calculated by drawing the line of sight from the
point of interest to all other points, and checking whether or
not the relief forms that occur between them block visibility.
From here, various measures of visual exposure, such as the
number of cells that can be seen from each cell, can be calculated. See Fisher (1991, 1992, 1993, 1995, 1996) and Ruiz
(1997) for examples of these types of applications and some
of the pitfalls that should be avoided.
Last but not the least, the complex issue of scale as it relates
to representing and characterizing the topography must be
discussed. Unfortunately, key issues associated with conceptual
and practical treatments of scale have not been appropriately
addressed (Bishop et al., 2012). These issues are related to
measurement, cartographic, geographical, computational, and
operational scales, coupled with hierarchical organization, and
the anisotropic nature of the topography. First, the local terrain
shape, which is generally thought of as the continuous variation of the altitude field from point to point, has an enormous impact on local and regional terrain parameters, but this
role is primarily influenced by cartographic and computational
scale. Florinsky (1998) suggested that local parameters, such as
slope gradient, slope azimuth, and curvatures, are mathematical variables rather than real-world characteristics. This statement may be extended to all local terrain parameters for two
reasons. First, local terrain shape may rely on different mathematical descriptions, so that the local parameters calculated
depend on algorithm selection. Furthermore, the terrain shape
portrayed by DEMs is a function of cartographic scale, combining the complexity of the terrain (geographic scale of features), computational scale, and measurement scale at which
the terrain surface was sampled (e.g., Deng et al., 2008). Thus,
it is possible to use the same local parameter to describe terrain
shape at different scales (resolutions, distances, and directions). The special feature of nonlocal primary attributes is
that they rely on the terrain shape of a larger, non-neighbor
area (computational scale) and need to be defined with
172
Geomorphometry
reference to other nonlocal points. Therefore, calculation of

nonlocal attributes is more difficult because it incurs additional
efforts in constructing point to-point connections over the
landscape and involves more complex algorithms and scale
considerations (e.g., Desmet and Govers, 1996; Gallant and
Wilson, 2000).
Finally, the computation of geomorphometric parameters
does not usually account for a meaningful computational
scale corresponding to the geographic scale of landforms,
to the operational scale of surface processes, or to coupled
systems dynamics. This disconnect significantly affects the
magnitude and interpretation of a land-surface parameter,
such that it does not accurately characterize the topography, as
governed by theory and practical utility. Such spatial constraint problems of computational scale are related to surface
materials and the hierarchical organization of the topography,
where hierarchically organized landforms and features effectively represent complex topographic patterns. Such scale
dependencies and organizational structure have yet to be
formally addressed, and are rarely accounted for by many
land-surface parameters. The partial exception to this is basic
semivariogram analysis, as it effectively accounts for spatial
complexity, although a spatial limit based on hierarchical
structure has yet to be rigorously evaluated. Furthermore, the
anisotropic nature of the topography at a multitude of distances must also be accounted for. This is demonstrated at the
local level using slope computed from the eight directions
within a 3 3 window. Principal component analysis of
the entire slope data set reveals fundamentally important
anisotropic information that can be used to highlight topographic structure related to surface processes, lithology, and
tectonics (Figure 10). The second principal component image
highlights the local basin structure, such that many spatial
characteristics can be used for geomorphological mapping and

system characterization. This aspect of scale is very important
in geomorphology, as it may improve the ability to assess
new aspects of geomorphological systems. Another example
at a larger computation scale is presented by Koons et al.
(2012), as they reveal that rock strength may be related to the
anisotropic nature of scale-dependent relief. Land-surface
parameters and objects that characterize operational scales
and structural constraints are urgently needed, and may aid in
establishing new theories about landscape evolution. Table 2
lists the most commonly used primary land-surface parameters and their significance.
Figure 10 Second principal-component image representing

anisotropic slope information extracted from local slope data over the
Shimshal Valley in northern Pakistan. The slope data set was
generated from a Shuttle Radar Topographic Mission (SRTM) 3 arcsecond DEM. Spatial variation patterns are related to lithological units
and tectonics. Less resistant metasedimetary rocks in the
northeastern region and less precipitation generate a higher spatialfrequency pattern, compared with a lower spatial-frequency pattern in
the southwestern region associated with more resistant lithologic
units, active uplift, and greater relief.
Table 2 Select list of primary and secondary land-surface parameters and their significance
Parameters
Type
Significance
Elevation
Slope
Slope azimuth
Profile curvature
Tangential curvature
Roughness
Elevation percentile
Flow width
Upslope contributing area
Flow-path length
Upslope height, elevationrelief ratio, hypsometric
curve, etc.
Mean slope of upslope area
Mean slope of dispersal area
Visual exposure
Topographic wetness index
Local
Local
Local
Local
Local
Local
Local
Local
Regional
Regional
Regional
Climate, vegetation, potential energy

Precipitation, overland and subsurface flow velocity and runoff rate, soil-water content
Flow direction, solar insolation, evapotranspiration, flora and fauna distribution and abundance
Flow acceleration and deceleration, soil erosion and deposition rates
Local flow convergence and divergence
Terrain complexity
Relative landscape position, flora and fauna distribution and abundance
Flow velocity, runoff rate, and sediment load
Runoff volume, soil water content, soil redistribution
Runoff volume, soil water content, soil redistribution
Distribution of height values, potential energy, flow characteristics
Regional
Regional
Regional
Regional
Stream-power index
Regional
Runoff velocity and possibly other flow characteristics

Rate of soil drainage
Exposure, solar insolation, wind patterns
Spatial distribution and extent of zones of saturation (i.e., variable source areas) for runoff generation as
a function of upslope contributing area, soil transmissivity, and slope
Erosive power of flowing water (based on the assumption that discharge is proportional to the specific
catchment area)
Source: Modified from Wilson, J.P., Gallant, J.C., 2000a. Digital terrain analysis. In: Wilson, J.P., Gallant, J.C. (Eds.), Terrain Analysis: Principles and Applications. John Wiley and
Sons, New York, pp. 127, and Olaya, V., 2009. Basic land-surface parameters. In: Hengl, T., Reuter, H.I. (Eds.), Geomorphometry: Concepts, Software, Applications. Elsevier,
Amsterdam, pp. 141169.
Geomorphometry
3.7.3.2
Secondary Land-Surface Parameters
In general, there are two basic sets of secondary land-surface

parameters. The first is for hydrologic characterization related to
quantifying water flow and related surface processes, and the
second is a series of climatology parameters that are related to
multiscale topographic influences on radiation, temperature, and
precipitation. Collectively, these parameters attempt to quantify
the interactions between the atmosphere and surface processes.
The underlying theory for both is well established, and the
parameterization schemes and computational methods have
evolved continuously over the past 20 years.
3.7.3.2.1
173
Contributing
area (ha)
200
20
2
Hydrology
The movement of water is primarily driven by gravity and, to

some degree, modified by the properties of the material it
flows through or over (Gruber and Peckham, 2009). The effect
of gravity can be approximated easily with a DEM, but the
surface and subsurface properties and conditions are cumbersome to describe and treat. There are steadily improving
regional and national databases describing the spatial variability of selected land-surface and soil characteristics (e.g.,
Miller and White, 1998; Feng et al., 2009), but these have a
much coarser resolution than the authors DEMs, and seldom
include the various properties needed for specific applications.
The typical approach relies on a series of parameter estimation
equations (e.g., Rawls, 1983; Saxton et al., 1986; Abdulla and
Lettenmier, 1997; Homann et al., 1998; Waltman et al., 2003;
Saxton and Rawls, 2006), although these will undoubtedly
introduce some additional uncertainty and error into the
analytical workflow or modeling application at hand (e.g.,
Band, 1993; Wilson et al., 1996; Zhu and Mackay, 2001;
Quinn et al., 2005). Given this state of affairs, it can be assumed that the DEM-based parameters will do better where
the relative importance of gravity is greatest (i.e., in headwater
areas and on steep slopes).
The initial development and use of flow-based land-surface
parameters can be traced to the introduction of the D8 algorithm (OCallaghan and Mark, 1984). This is now but one of
more than a dozen flow-routing algorithms, however, and
a distinction is usually drawn between single- and multipleflow direction algorithms. The single flow-routing algorithms,
which direct flow to just one downslope or neighboring cell,
include the Rho8 (Fairfield and Leymarie, 1991) and aspectdriven kinematic routing (Lea, 1992) algorithms in addition
to the D8 algorithm. The multiple flow-routing algorithms,
which are capable of directing flow to two or more downslope
or neighboring cells, include the FD8 (Freeman, 1991), TOPMODEL (Quinn et al., 1991, 1995), DEMON (Costa-Cabral
and Burges, 1994), DN (Tarboton, 1997), and Mass-Flux
algorithms (Gruber and Peckham, 2009). These flow-routing
algorithms will generally generate very different results
(Figures 11 and 12).
The performance of many of these flow-routing algorithms
has been compared across a variety of landscapes (e.g., Wolock
and McCabe, 1995; Desmet and Govers, 1996; Zhou and Liu,
2002; Endreny and Wood, 2003; Chirico et al., 2005; Wilson
et al., 2000, 2007). Wilson et al. (2008) recently evaluated
the performance of several algorithms including ANSWERS
(Beasley and Huggins, 1978), D8 (OCallaghan and
0.2
0.02
N
300
600
(m)
Figure 11 Gridded contributing area (ha) map derived from the

Cottonwood Creek, Montana DEM, using the D8 single flow direction
algorithm, with the catchment boundary overlaid. Reproduced from
Wilson, J.P., Gallant, J.C., 2000b. Secondary topographic attributes.
In: Wilson, J.P., Gallant, J.C. (Eds.), Terrain Analysis: Principles and
Applications. John Wiley and Sons, New York, pp. 5185.
Mark, 1984), Rho8 (Fairfield and Leymarie, 1991), FD8/

TOPMODEL (Freeman, 1991; Quinn et al., 1991, 1995), an
aspect-driven kinematic routing algorithm (Lea, 1992),
DEMON (Costa-Cabral and Burges, 1994), a flow decomposition algorithm (Desmet and Govers, 1996), DN
(Tarboton, 1997), and MFD-md (Qin et al., 2007). They
showed that various algorithms can be expected to generate
different patterns of flow (based on upslope contributing
areas) on different parts of a hill slope or a watershed. The
gridded contributing area map reproduced in Figure 11 shows
how the D8 algorithm often generates many parallel flow
lines, which of course do not match the dendritic patterns that
characterize most surface and channel flow systems, whereas
the DEMON algorithm (Figure 12) produces more realistic
patterns, but is computationally slow when used with large
DEMs and prone to failure in flat areas. The final choice of the
flow-routing algorithm should aim to minimize the most
important of these tradeoffs for the particular study area and
application at hand.
Notwithstanding the presence of these kinds of tradeoffs, the
multiple flow-routing algorithms have grown in popularity over
time. The fundamental goal with this class of algorithms is to
find a method or a sequence of methods that move water into
one or more downslope cells, and Gruber and Peckham (2009)
174
Geomorphometry
Contributing
area (ha)
200
20
2
0.2
0.02
and delineate the drainage networks along with the basin

boundaries. See Band (1986, 1989), Montgomery and Dietrich
(1989, 1992) and Peckham (1998) for examples of methods for
delineating drainage networks with single flow-direction algorithms. The topographic wetness and stream-power indices are
among the most popular of various stream attributes, and unlike
the catchment (basin) boundaries and some other attributes, this
pair of attributes can be calculated with both the single and the
multiple flow-direction algorithms.
The typical form of the topographic wetness index (TWI)
assumes steady-state conditions and describes the spatial distribution and extent of zones of saturation (i.e., variable source
areas for runoff generation) as a function of upslope contributing
area Ac, slope, and occasionally soil transmissivity (this last term
is often excluded because the transmissivity is assumed to be
constant throughout the catchment), such that

Ac
TWI ln
11
tanb
300
600
(m)
Figure 12 Gridded contributing area (ha) map derived from the

Cottonwood Creek, Montana DEM, using the DEMON stream-tube
algorithm, with the catchment boundary overlaid. Reproduced from
Wilson, J.P., Gallant, J.C., 2000b. Secondary topographic attributes.
In: Wilson, J.P., Gallant, J.C. (Eds.), Terrain Analysis: Principles and
have noted how this approach might be justified by actual divergence (i.e., the need to treat flows across convergent and
divergent land surfaces) and/or the attempt to overcome the
limits of having only eight adjacent cells (i.e., methods to
overcome the limitations generated when a continuous flow
field is represented with a regular grid that has only eight possible directions in multiples of 451).
This pair of explanations helps to explain why the final
choice of a flow-routing method for a specific application will be
a compromise. The single flow-direction algorithms cannot represent divergent flow, but for the same reason, have no
problem with overdispersal (i.e., the dispersal of the available
flow over too many cells or too large an area). The multiple
flow-direction algorithms, however, can represent divergent flow,
but usually also suffer from some overdispersal. In addition, the
subtleties and outcomes of the methods are concerned with the
need to: (1) treat ambiguous flow directions (as for example
occurs along ridgelines or saddles and across flat plains or valley
bottoms); and (2) reconcile the DEM-delineated flow lines and
the drainage lines acquired from some other source that are also
likely to influence the results generated with these different flowrouting algorithms.
Therefore, the flow directions are generally computed to calculate upslope contributing areas (i.e., flow accumulation areas)
The steady-state form of the TWI predicts zones of saturation

where the specific catchment area is large (which typically occurs
in converging areas of the landscape), the slope is small (which
typically occurs at the base of concave slopes), and soil transmissivity is low (which is often characteristic of areas with
shallow soils). This index has been used successfully in a variety
of hydrological applications because the aforementioned conditions are commonly encountered along drainage paths and in
zones of water concentration in many landscapes (e.g., Beven
and Kirkby, 1979; Burt and Butcher, 1985; Moore and Burch,
1986; OLoughlin, 1986; Sivapalan et al., 1987; Moore et al.,
1988; Phillips, 1990; Montgomery and Dietrich, 1994; Moore
and Wilson, 1992, 1994; Fried et al., 2000; Kheir et al., 2007).
These types of static indices, however, must be used carefully to predict the distribution of dynamic phenomena like
soil-water content because surface saturation is a threshold
process, the presence of hysteric effects, and the reliance
on one or more assumptions. The two most important assumptions in this case are that: (1) the gradient of the
piezometric head, which governs the direction of subsurface
flow, is parallel to the land surface; and (2) there is
sufficient time between rainstorms for the subsurface flow to
achieve a steady state (Moore et al., 1993a). Several authors
have described the pitfalls of using these kinds of indices in
inappropriate ways. For example, Jones (1986, 1987) documented some of the advantages and limitations of using
wetness indices to describe the spatial patterns of soil-water
content and drainage, and Quinn et al. (1995) summarized
the various problems and described how the steady-state TWI
can be calculated and used effectively as part of the TOPMODEL hydrologic-modeling framework. Several variants
of the original equation have also been proposed. Barling
(1992) proposed a quasi-dynamic topographic wetness index
(QD-TWI) to overcome the limitations of the steady-state
assumption and used it to show how the topographic
hollows, and not the drainage channels themselves, determined the response of a semiarid catchment in New South
Wales, Australia (see Barling et al. (1994) for additional details). Wood et al. (1997) later proposed an alternative index
to predict the saturated-zone thickness that incorporated both
Geomorphometry
spatial and temporal variation in recharge. Both Fried et al.

(2000) and Nguyen and Wilson (2010) calculated QD-TWI
using a variety of flow-routing algorithms and showed how
the results varied depending on the flow-routing algorithm
that was utilized.
3.7.3.2.2
Climatology
The topography also governs microclimate, and surface-energy

conditions influence the magnitude of various surface processes. Bohner and Antonic (2009) reviewed topoclimatology
parameters, as the land surface controls the spatial variability
of near-ground atmospheric processes and meso-scale climatic
variations. The regionalization approaches that use kriging,
universal kriging, and splines to map the climate variables
measured at climate stations are ignored, and the focus is on
the land-surface parameters that are used to assess the variability of the short- and long-wave radiation fluxes, as these
influence surface temperature, evapotranspiration, air movement, and other surface processes.
Understanding and predicting the magnitude of solar and
surface irradiance is of primary concern. Solar irradiance varies
as a function of: (1) the changing distance from the Sun to the
Earth; (2) intrinsic variation in the output of the Sun; and (3)
changes in the radiation field from the Sun toward the Earth.
Annual changes in irradiance can be B6%, independent
of wavelength. A nonuniform distribution of active regions
also occurs on the solar disk that are modulated by a 27-day
rotation period, which results in irradiance variations that are
wavelength (l) dependent. Irradiance variations are also
caused by solar magnetic activity (22-year cycle). These variations are caused by eruptive phenomena such as flares, and
range in temporal scale from minutes, to months, to years,
and include the 11-year sunspot cycle.
The exo-atmospheric irradiance, E0, is therefore a function
of orbital parameters that determine its magnitude. Most GISbased solar-radiation models do not account for orbitalparameter variation of obliquity and eccentricity; therefore,
they cannot be used effectively for paleo-climate and futurescenario studies too far into the past or future. For example,
such models cannot be used to study radiative forcing in the
Holocene. Rather, such models make use of standard irradiance spectra and account for the annual variation in the
EarthSun distance to portray seasonal and diurnal variations
in E0.
Atmospheric conditions then determine the amount
of atmospheric attenuation, and atmospheric-transmission
functions are wavelength dependent. Atmospheric conditions
are generally prescribed based on modeled atmospheric conditions. For solar-radiation modeling and estimation of surface-process rates, there is a need to accurately estimate the
surface spectral irradiance (E) which is a composite of three
downward irradiance components
El Eb l Ed l Et l
12
The direct/beam irradiance (Eb) is typically dominant,

followed by the diffuse-skylight irradiance (Ed) and the
adjacent terrain irradiance (Et). Variations in atmospheric,
topographic, and land-cover conditions determine the sequential dominance of irradiance partitioning.
175
Under cloudless skies, Eb is the dominant term in eqn [12].

Consequently, considerable research has focused on modeling
the direct atmospheric-transmittance functions accurately.
The atmosphere attenuates the direct irradiance primarily by
gaseous absorption and molecular and aerosol scattering
(Chavez, 1996). These atmospheric processes are wavelength
dependent, and spatially and temporally controlled by changing atmospheric and landscape conditions. The total downward atmospheric transmission (Tk) is a function of the total
optical depth of the atmosphere, which varies with solarzenith angle and altitude, and can be represented as
T k l Tr lTa lTO3 lTgas lTH2 O l
13
where Tr is the Rayleigh transmittance, Ta is the aerosol

transmittance, TO3 is the ozone transmittance, Tgas is the
transmittance for miscellaneous well-mixed gases, and TH2 O is
the water-vapor transmittance. Atmospheric attenuation is
highly variable with wavelength, with Rayleigh and aerosol
scattering dominating at shorter wavelengths and water vapor
dominating at longer wavelengths.
The direct irradiance is also governed by multiscale topographic parameters. Local or microscale topographic variation
is represented by the incidence angle of illumination between
the Sun and the vector normal to the ground, such that
cos i cos yi cos b sinyi sin b cosf fi
14
where yi is the incident solar-zenith angle and fi is the incident solar-azimuth angle.
It is possible to estimate cos i using a DEM, and uncertainty
in the estimate is related to the resolution, as subpixel-scale
topographic variation is not accounted for. Values of cos i can
be r0.0, indicating no direct irradiance due to the orientation
of the topography. It is important to note that the incident
solar geometry varies across the landscape, although this is
usually assumed to be constant when working with image
scenes (i.e., small-angle approximation). In addition, the
meso-scale topographic relief in the direction of fi determines
whether a pixel is in shadow (Sc). This can be accounted for by
ray tracing, shadow detection, and shadow interpolation algorithms that alter cos i values appropriately (Dozier et al.,
1981; Rossi et al., 1994; Giles, 2001). The local and meso-scale
topographic influences on the direct irradiance are significant
over annual and diurnal time scales (Figure 13). Consequently, direct irradiance exhibits a high degree of spatiotemporal variability. The direct irradiance component can be
estimated as
Eb l E0 lT k lcos iSc
15
Atmospheric scattering also generates a hemispherical

source of irradiance that should be calculated as an integration
of the total sky irradiance. This source can be simplistically
represented as a composite of a Rayleigh-scattered component
(Er), an aerosol-scattered component (Ea), and the groundbackscattered component (Eg) that represent inter-reflections
between the land surface and the atmosphere, where
Ed l Er l Ea l Eg l
16
176
Geomorphometry
(a)
(b)
(c)
(d)
Figure 13 Simulated topographic influences on the direct irradiance (cos iSc) over the Mt. Everest region in Nepal. Darker tones represent less
direct irradiance whereas light tones represent greater direct irradiance. Simulations based on an Advanced Spaceborne Thermal Emission and
Reflectance Radiometer Global Digital Elevation Model of the region. Simulations account for local topographic conditions including slope and slope
azimuth, as well as meso-scale topographic shielding that casts shadows. The solar azimuth was held constant at 135.01: (a) Simulation with a
solar-zenith angle of 0.01; (b) Simulation with a solar-zenith angle of 45.01; (c) Simulation with a solar-zenith angle of 70.01; and (d) Simulation with
a solar-zenith angle of 85.01. Assuming relatively constant atmospheric conditions, direct irradiance in this region exhibits a high degree of
spatiotemporal variability.
Its accurate estimation is complicated by the fact that an

anisotropic parameterization scheme is required. In general,
the irradiance decreases with angular distance from the Sun.
In addition, this irradiance component is also influenced
by meso-scale hemispherical shielding of the topography.
Consequently, only a solid angle of the sky will contribute
to Ed, and this angle will change as a function of pixel location
and direction. In general, the solid angle will increase with

altitude. It is frequently referred to as the sky-view factor (Vf )
in the remote-sensing and energy-balance literature, and can
be estimated using a DEM (Figure 14) such that
Vf
360
X
f0
cos2 ymax f,d
Df
360
17
Geomorphometry
177
relief areas can exhibit a strong adjacent-terrain irradiance

component due to highly reflective features such as snow and
vegetation.
The magnitude of the reflected and emitted radiance at the
surface is determined by the conservation of energy, such that
rl aa l Tl 1:0
19
where, r, aa, and T represent the reflectance, absorption, and

transmission, respectively. For opaque objects, T 0.0. It is
commonly assumed that reflectance is isotropic (surface reflects radiation equally in all directions) and the surface
spectral radiance (L) can be computed as follows
Figure 14 Sky-view factor over northern Pakistan. This parameter
was computed from a Shuttle Radar Topographic Mission (SRTM)
3 arc-second DEM. Darker tones represent those areas that do not
receive as much diffuse-skylight irradiance, whereas lighter tone
areas receive more. The spatial patterns reveal that meso-scale relief
conditions are highly variable in the Himalaya. Examination of the
magnitude and spatial patterns reveals that some glaciers are
receiving more diffuse-skylight irradiance than others, whereas
different glaciers exhibit different trends in their sky-view-factor
altitude profile. Such topographic variation regulates ablation rates
and governs the sensitivity of glacier response to climate forcing.
This is turn regulates melt-water production and glacial and fluvial
erosion.
where ymax is the maximum local horizon angle at a given

azimuth, f, over a radial distance of d.
In mountain environments exhibiting extreme relief and
deep valleys, topographic shielding of the skylight-diffuse irradiance can be significant (Proy et al., 1989). Furthermore,
difficulties in accurately predicting the bidirectional reflectance distribution function (BRDF) of land-cover characteristics surrounding individual pixels generate uncertainty in the
estimation of the ground-backscattered component.
The irradiance components Eb and Ed interact with the
terrain and land-cover biophysical characteristics to generate
an adjacent-terrain irradiance component. This irradiance
component is not generally considered in remote-sensing,
GIS-based radiation modeling, and energy-balance studies because it is assumed that its magnitude is relatively minor, and it
is a difficult parameter to estimate accurately. A first-order approximation has been formulated by Proy et al. (1989) and
assumes that surface radiance is Lambertian. It is then possible
to estimate the radiance received at any pixel, by accounting for
the geometry between two pixels (p1 and p2) such that

Ap
L12 cosy1 L2 cosy2 2
d
18
where L12 represents the radiance received at p1 from the luminance of p2 (L2), y1 and y2 are the angles between the
normal to the terrain and the direct line of sight from p1p2,
Ap is the pixel area (p2), and d represents the distance between
p1 and p2.
This equation can be used to estimate Et for any pixel by
integrating over all of the pixels whose slopes are oriented
toward a pixel of interest and where the line of sight is not
blocked by the topography. High-altitude and extreme local

Ll rl
El
p

20
The Lambert assumption, however, is not an accurate

characterization of the anisotropic nature of surface reflection.
The BRDF describes such reflectance variations and can be
used to estimate the surface albedo, a key parameter in surface
energy-budget modeling needed to compute the net shortwave radiant flux. Several parameterization schemes exist for
the long-wave irradiance and net long-wave radiant flux, and
other surface energy-budget components are also dependent
on topographic parameters including surface roughness.
Several GIS-based solar-radiation models and surface energy-budget models can be used to produce maps of various
surface irradiance and energy parameters. Such models can be
used over user-specified periods ranging from 1 day to a year
in length. SRAD (Moore et al., 1993b; Wilson and Gallant,
2000b) is but one of a number of models that have been
proposed for calculating the radiation fluxes and it incorporates the effects of cloudiness into the calculations. Others include r.sun (Hofierka, 1997), Solar Analyst (Fu and Rich,
2000), Solar Flux (Dubyah and Rich, 1995; Hetrick et al.,
1993a, b), and Solei (Meszaros, 1998; Miklanek, 1993). All
the aforementioned models document how spatial variability
in elevation, slope, slope azimuth, cast shadows, sky-view
factor, and other variables can create very strong local gradients in solar radiation and surface temperature, and thereby
exert a large influence on surface processes including photosynthesis, evapotranspiration, ablation, weathering, as well as
influencing vegetation diversity and biomass production at
specific locations on the land surface.
Analysis of the topography is also required for the characterization of precipitation, air flow, and wind speed, as
rainfall is governed by local slopes and meso-scale altitude
variations, and air-flow direction and speed is governed by
regional relief structure, deformation orientation patterns, and
local and meso-scale surface roughness. For more details, see
Bohner and Antonic (2009).
Finally, it is important to note that it is difficult to verify
some parameter estimates because it is not easy to accurately
measure the spatio-temporal variability of key variables (e.g.,
albedo, BRDF, Et). Furthermore, climate stations are not generally established in complex terrain, but occur in low-altitude
areas in flat terrain, and may not collect data representative
of higher-altitude areas or more complex topography. One
possible way around this problem is to use satellite data for
Geomorphometry
estimating incoming solar radiation and precipitation. Consequently, there is a need to develop and validate new parameterization schemes that address process mechanics and the
spacetime issues connected to data, analysis, and modeling.
Table 2 also lists the most frequently used secondary landsurface parameters and their significance.
The use of land-surface parameters to segment the landscape

into terrain features or landform classes (i.e., objects) can be
traced to the pioneering work of Speight (1968) and Dikau
(1989). Recent developments have included the use of
automated fuzzy-classification algorithms to detect landform
elements (e.g., Burrough et al., 2000a; Schmidt and Hewitt,
2004). The focus of these kinds of applications may range
from the identification of specific landforms (e.g., mountains,
valleys, glaciers, alluvial fans) to landform elements (i.e.,
geometric shapes that constitute part or all of a specific
landform) and repeating landforms types (e.g., a series of
rolling hills and valleys). Here, the focus is on the extraction
and classification of landform elements, since these constitute
the basic building blocks for segmenting the landscape into
landform classes.
The early landform classification approaches relied on
various representations of landscape position and the shape of
the land surface itself. For example, Dikau (1989) divided the
landscape into combinations of concave, straight, and convex
planform curvatures, on the one hand, and concave, straight,
and convex profile curvatures on the other. This approach (like
many subsequent ones) relies on the inferred relationship
between surface shape (i.e., local curvature) and the accumulation of surface flow and consequently that of surface deposition through two accumulation mechanisms. The first
mechanism reflects the divergence and convergence of flow
across a hill slope, whereas the second reflects the relative
deceleration of flow in the downslope direction, as influenced
by changes in profile curvature (Moore et al., 1991; MacMillan
and Shary, 2009).
Shary and his colleagues have criticized Dikaus (1989)
original approach on two levels, and proposed a more robust
and predictable classification based on curvatures. Their first
suggestion was to use tangential curvature in place of planform curvature in the classification of basic form elements
(Figure 15), because both tangential and profile curvatures are
curvatures of normal sections and both exhibit similar statistical distributions, unlike planform curvature (MacMillan
and Shary, 2009). The second criticism concerned the contradiction in Dikaus (1989) original premise that this approach differentiated form elements with a homogeneous
plan and profile curvature, because these facets invariably
contain even more homogeneous form facets with similar
gradients, aspects, and curvatures. Shary (1995) and Shary
et al. (2005), in turn, have proposed an objective, local, scalespecific classification of elemental landform features based
only on the consideration of the signs of the tangential, profile, mean, difference, and total Guassian curvatures, as a way
of avoiding both of these sets of problems (Figure 16).
Concave
X/X
SF/X
V/X
Tangentialstraight
Land-Surface Objects and Landforms
Profile-straight
X/SL
SF/SL
V/SL
X/V
Convex
Convex
Tangential curvature
3.7.4
Profile curvature
Concave
178
SF/V
V/V
Figure 15 The modified Dikau (1989) classification of form

elements based on the profile and tangential (i.e., across slope)
curvatures. The elements have been further classified as positive or
negative based on the radius of curvatures (4600 or o600 m) and
the planform curvature in the original classification was replaced by
tangential curvature based on Shary and Stepanov (1991).
Reproduced from MacMillan, R.A., Shary, P.A., 2009. Landforms and
landform elements in geomorphometry. In: Hengl, T., Reuter, H.I.
(Eds.), Geomorphometry: Concepts, Software, Applications. Elsevier,
None of the aforementioned approaches, however, incorporates the contextual position of specific facets that make
up the landscape. Most of the automated classifications that
incorporate context build on the conceptual classifications of
hillslopes like those of Ruhe (1960), which divided hill slopes
into five units (summits, shoulders, back slopes, foot slopes,
and toe slopes). Many other such classifications have been
proposed, and Conacher and Dalrymple (1977) and Speight
(1990) divided hill slopes into 9 and 10 units, respectively.
This was accomplished by delineating finer-resolution hill
slope facets or by extending the hill slope to include the
channel or various parts of the channel at the bottom of the
hill slope. Taken as a whole, these conceptual classes consider
slope gradient and relative-slope position along a toposequence from divide to channel. In addition to curvatures
and automated classification, various approaches have included absolute and relative horizontal and vertical distance
to ridge lines or channels (e.g., Skidmore et al., 1991) and
position in the landscape relative to the order of the nearest
stream channel below a hillslope (e.g., Schmidt et al., 1998;
Schmidt and Dikau, 1999).
The adoption and use of fuzzy-classification algorithms to
detect landform elements marked an important step forward.
The importance of this innovation can be traced to the fact
that each of the aforementioned approaches will work in
some instances and not others, and they will seldom produce
satisfactory answers to questions linked to the locations of
a specific mountain or valley (Fisher et al., 2004). Indeed,
there are many phenomena that are difficult to locate or delineate because their meaning is not well defined or because of
the subjectivity, vagueness, and ambiguity that have often
characterized the ways in which the world is described
Geomorphometry
179
Difference
curvature
Total gaussian
curvature
Total gaussian
curvature
4
5
11
12
Mean
curvature
9
10
Tangential
curvature
Profile
curvature
Figure 16 Sharys complete system of classification of landform elements based on signs of tangential, profile, mean, difference, and total
Guassian curvatures. Reproduced from Shary, P.A., Sharaya, L.S., Mitusov, A.V., 2005. The problem of scale-specific and scale-free approaches
to geomorphometry. Geografia Fisica e Dinamica Quaternaria 28, 81101.
(Burrough, 1996; Wilson and Burrough, 1999). Fuzzy-set

theory represents an alternative approach to classic set theory
(Burrough and McDonnell, 1998; Robinson, 2003) and has
been used in many environmental domains to solve these
kinds of problems. Hence, one or more forms of fuzzy classification have been used to describe soil variability (e.g.,
McBratney and Odeh, 1997; Zhu, 1997a, b, 1999; Ahn et al.,
1999), land cover (e.g., Fisher and Pathirana, 1994; Foody,
1996, Brown, 1998; DeBruin, 2000), site selection and multicriteria evaluation (Charnpratheep et al., 1997; Jiang and
Eastman, 2000), and the parameterization of land-surface
models (e.g., Mackay et al., 2003).
The above mentioned examples indicate how many
authors have used fuzzy sets and fuzzy-logic operators with
land-surface parameters to generate partial and multiple
memberships of spatial objects of various kinds during the past
15 years. The various membership functions that have been
described in considerable detail by Robinson (2003) represent
the core of the method because they allow the expression
of irreducible observation and measurement uncertainties in
their various manifestations, and make these uncertainties intrinsic to the classification (i.e., using grades of membership).
In this way, the fuzzy-logic approach will associate a fuzzy
likelihood of each output class with each value or class on each
input map (Figure 17). This means that when fuzzy data are
processed, their intrinsic uncertainties are processed as well,
and their results are more meaningful than their counterparts
obtained by processing the usual crisp data (Klir and Yuan,
1995; Robinson, 2003). The advantages of using such an
approach are evident in the landform elements specified by
MacMillan et al. (2000) on the basis of a combination of
measures of local-surface shape (convexity/concavity) plus
the relative-slope position for a 64 ha site in Alberta, Canada.
Two basic fuzzy-set approaches exist. The first relies on

expert knowledge and is often referred to as the SemanticImport model, whereas the second approach relies on the
identification and characterization of feature-space clusters
(e.g., Bezdek et al., 1984). The fuzzy sets generated in both of
these instances will often provide tremendous expressive
power (Robinson, 2003). For example, Zhu (1997a, b, 1999)
work clearly demonstrates the utility of the fuzzy paradigm for
capturing and representing spatially explicit soils knowledge
from human experts. Zhu and his colleagues used knowledge
of observable environmental inputs including several landsurface parameters, and targeted outputs (soil-map units), and
adopted a limiting-factor approach for computing the overall
similarity. The value for the overall similarity between an
unclassified entity and a reference entity was obtained by
simply selecting the smallest similarity value from among all
similarity values computed for all attributes for an unclassified
entity in this application.
Both the expert- and data-driven fuzzy-classification approaches will result in multiple maps of membership, where
each class will be represented on a separate map. The advantages of using memberships are threefold since it is possible
to: (1) determine which classes are connected with which
land-surface parameters; (2) assess the confusion between
the classes; and thereby (3) detect the areas where the confusion between two or more classes is high (Burrough and
McDonnell, 1998; Hengl et al., 2004; Shi et al., 2005; Evans
et al., 2009).
There are many subtleties, however, connected to these
fuzzy-classification methods. The knowledge in the aforementioned approach could have been acquired via several
methods, and each would have varied in the degree to which it
is theoretically, empirically, or statistically valid. See Qi and
0.5
not A
0.0
(a)
Threshold value
Land-surface parameter
(b)
0.8
0.
0.6
0.8
6
0.
Membership
1.0
Land-surface parameter #2
Geomorphometry
0.
180
Class centre
Land-surface parameter #1
Figure 17 Schematic showing the derivation of fuzzy memberships using (a) the definition of thresholds and (b) the definition of class centers.
Reproduced from Hengl, T., Reuter, H.I. (Eds.), 2009. Geomorphometry: Concepts, Software, and Applications. Developments in Soil Science.
Elsevier, Amsterdam.
Zhu (2003), Qi (2004), and Qi et al. (2006) for an extended

discussion of these issues. Similarly, the results can be expected to vary with the method used to compute the overall
similarity scores. MacMillan et al. (2000), in their implementation of the Semantic-Import model, relied on a weighted-average method to calculate the overall similarity of an
unclassified site to a reference entity, based on the assumption
that all input variables should be included in calculating the
similarity of a site to a reference entity. However, a case can
also be made, that some inputs may deserve to be afforded
a greater importance or weight than others (Hengl and
MacMillan, 2009). Indeed, the sensitivity of the fuzzy c-means
approach to the choice of input variables and weights assigned
to them when calculating the overall similarity values has
been explored extensively (e.g., Deng et al., 2006; Deng and
Wilson, 2006).
There is also the need to choose an appropriate scale for
both the Semantic-Import model and the fuzzy c-means fuzzyclassification approaches. The fact that there is no single true
or fixed value for local land-surface parameters such as slope
or curvature at a point, but rather a whole range of values that
are dependent on the horizontal and vertical resolution has
already been noted. Not unexpectedly, no best resolution
can be singled out at which to compute local land-surface
parameters to portray and classify terrain (Hengl, 2006; Smith
et al., 2006; Deng et al., 2007), and the final scale that is
chosen should be appropriate for capturing and describing the
surface features of interest for a particular application (Deng
et al., 2008). The size or the extent of the study area needs to
be added to this list of sensitive variables since some landsurface parameters will vary in systematic ways across the
landscape, and may generate locally specific results when the
fuzzy classification is implemented for a limited area (Evans
et al., 2009).
The aforementioned discussion gives some sense of the
great progress that has been made with automated landform
classification during the past quarter century. The successful
deployment of these techniques, however, requires considerable knowledge and experience with the techniques themselves, and of the study area to which they are to be applied.
Furthermore, the generation of land-surface objects requires
better formalization that links process and form. Addressing
various issues associated with the formal characterization of

topographic structure and the use of object-oriented technology also has the potential to lead to significant progress. Unfortunately, new representational schemes and the use of
Earth-science concepts in analysis and modeling have not kept
pace with empirical exploration (Bishop et al., 2012). Both are
required for formalizing the generation of land-surface objects
that will facilitate diagnostic geomorphological mapping
efforts.
3.7.5
Conclusions
There have been tremendous advances in DEM data sources,

digital terrain modeling techniques, new algorithms for landsurface parameterization, and new geomorphological applications driven by geomorphometry. Nevertheless, new theoretical/conceptual and information technology advances must
also occur that formalize the understanding of geomorphological systems and topographic complexity. There are at least
four research paths that can be expected to yield substantial
benefits.
1. Knowledge of the presence of, and propagation of, errors in
both the current and the new remote-sensing data sources
that emerge needs to be improved. This is a challenging
task because many of the systematic and random errors in
the current data streams are specific to the sensor used and
the specific protocols and methods that have been used in
individual projects (Dowman, 2004). This suggests that
ways to clarify and publish information related to data
quality need to be established, since much of this is proprietary information of firms that have built and deployed
the aforementioned technologies.
2. Field observations and the development and testing of new
analytical methods are required. Taking the modeling of
flow directions and upslope contributing areas as examples, there is an urgent need to learn more about the
ways in which the land surface and the interactions with
the underlying soil and regolith influence rainfallrunoff
relationships and the growth and contraction of flow
networks in specific environments. Lindsay and colleagues
Geomorphometry
at the University of Guelph in Ontario, Canada, for

example, are exploring the spatial pattern and timing of
ephemeral flows in headwater channels that may provide
some important new insights. These kinds of projects are
time-consuming, but are vital for the development of DEM
datasets and analytical methods that support the representation of the key hydrologic and geomorphic processes
(i.e., those influencing nonpoint source pollution) operating in specific landscapes (e.g., Mitasova et al., 1995).
3. The critical evaluation and adoption of key algorithms and
best approaches for solving specific problems is required. This
is similar to the second path, although it is likely to yield
faster returns. The goal here would be to combine and integrate best practices, as exemplified by the following example.
The QD-TWI model proposed by Barling (1992) would appear to have considerable merit, given what is known about
the distribution of soil moisture along with both the surface
and subsurface flow patterns in a variety of semiarid and arid
landscapes. The original QD-TWI model, however, incorporated the D8 flow-routing algorithm, and several studies
have demonstrated that D8 generates many undesirable artifacts. The DEMON flow-routing algorithm, however, offers
many advantages but sometimes fails in areas with flat terrain, and shows slow performance when applied to relatively
fine-resolution DEMs covering large areas. Nguyen (2011) has
exploited this opportunity by building a faster and more robust version of DEMON along with a flexible version of the
QD-TWI model that allows the user to choose from a variety
of flow-routing algorithms. Consequently, a terrain analyst
can work with both approaches simultaneously. There are
many opportunities like this that can be exploited in the
immediate future.
4. The final research path concerns issues of scale. The rapid
advent and adoption of fine-resolution remote-sensing
data sources, and the need to characterize coupled geomorphological systems, means that there is an urgent need
to address multiple issues of scale that affect the ability to:
(1) collect, represent, and integrate data and information
across multiple scales; (2) characterize land-surface parameters; (3) generate meaningful land-surface objects that
are based on scientific principles and concepts; and (4)
develop and refine techniques that allow multiscale characterization and visualization to address a variety of
problems. These topics have been recognized by a variety of
researchers as key agendas (Gallant et al., 2000; Gallant
and Dowling, 2003; Sulebak and Hjelle, 2003; Fisher et al.,
2005; Deng, 2007; Deng and Wilson, 2008; Bishop et al.,
2012).
Finally, various equations have been included for those
interested in calculating one or more of the aforementioned
land-surface parameters, as these can be used for segmenting
the topography into land-surface objects that are useful for
geomorphological mapping. Furthermore, many of the secondary parameters can be used in geomorphological research
related to assessment of hydrological, glaciological, and geological conditions. The current state of the art suggests that the
present-day terrain analyst will need to choose wisely among
the extreme multitude of options (i.e., data, algorithms, analysis approaches, and models), while paying special attention
181
to their own project goals, the advantages and disadvantages

of different data sources and digital terrain modeling techniques, the characteristics of their study area(s), and how errors
might have been introduced and propagated, and the likely
significance of these errors, given the results that are produced.
References
Abdulla, F.A., Lettenmier, D.P., 1997. Development of regional parameter estimation
equations for a macroscale hydrologic model. Journal of Hydrology 197,
230257.
Aerts, C.J.H., Heuvelink, G.B.M., Goodchild, M.F., 2003. Accounting for spatial
uncertainty in optimization with spatial decision support systems. Transactions in
GIS 7, 211230.
Ahn, C.W., Baumgartner, M.F., Biehl, I.I., 1999. Delineation of soil variability using
geostatistics and fuzzy clustering analysis of hypospectral data. Soil Science
Society of America Journal 63, 142150.
Antonic, O., Pernar, N., Jelaska, S.D., 2003. Spatial distribution of main forest soil
groups in Croatia as a function of basic pedogenetic factors. Ecological
Modeling 170, 363371.
Band, L.E., 1986. Topographic partition of watersheds with digital elevation models.
Band, L.E., 1989. A terrain based, watershed information system. Hydrological
Processes 3, 151162.
Band, L.E., 1993. Effect of land surface representation on forest water and carbon
budgets. Journal of Hydrology 150, 749772.
Band, L.E., Vertessey, R., Lammers, R.B., 1995. The effect of different terrain
representation schemes and resolution on simulated watershed processes.
Zeitschrift fur Geomorphologie, Suppl-Bd 101, 187199.
Barling, R.D., 1992. Saturation zones and ephemeral gullies on arable land in
Southeastern Australia. University of Melbourne, Unpublished PhD dissertation.
Barling, R.D., Moore, I.D., Grayson, R.B., 1994. A quasi-dynamic wetness index for
characterizing the spatial distribution of zones of surface saturation and soil
water content. Water Resources Research 30, 10291044.
Beasley, D.G., Huggins, L.F., 1978. ANSWERS: a model for watershed planning.
Proceedings of the Tenth Conference on Water Simulation Volume 2, Miami
Beach, FL. 503515
Berry, P.A.M., Garlick, J.D., Smith, R.G., 2007. Near-global validation of the SRTM
DEM using satellite radar altimetry. Remote Sensing of Environment 106,
1727.
Beven, K.J., Kirkby, M.J., 1979. A physically-based, variable contributing area
model of basin hydrology. Hydrological Sciences Bulletin 24, 4369.
Bezdek, J.C., Ehrlick, R., Full, W., 1984. FCM: the fuzzy c-means clustering
algorithm. Computers and Geosciences 10, 191203.
Bhang, K.J., Schwartz, F.W., Braun, A., 2007. Verification of the vertical error in CBand SRTM DEM using ICESat and Landsat-7, Otter Tail County, MN. IEEE
Transactions on Geoscience and Remote Sensing 45, 3644.
Bishop, M.P., Bonk, R., Kamp, Jr. U., Shroder, Jr. J.F., 2001. Terrain analysis and
data modeling for alpine glacier mapping. Polar Geography 24, 257276.
Shroder, J.F., 2010. Climate change and mountain topographic evolution in the
Central Karakoram, Pakistan. Annals of the Association of American Geographers
100(4), 772793.
and digital geomorphological mapping: Concepts, issues and research.
Bishop, M.P., Shroder, Jr. J.F., 2000. Remote sensing and geomorphometric
assessment of topographic complexity and erosion dynamics in the Nanga
Parbat massif. In: Khan, M.A., Treloar, P.J., Searle, M.P., Jan, M.Q. (Eds.),
Tectonics of the Nanga Parbat Syntaxis and the Western Himalaya. Geological
Society, London, No. 170 in Special Publication, pp. 181200.
Bishop, M.P., Shroder, Jr. J.F. (Eds.), 2004a. Geographic Information Science and
Bishop, M.P., Shroder, Jr. J.F., 2004b. GIScience and mountain geomorphology:
Overview, feedbacks, and research directions. In: Bishop, M.P., Shroder, Jr. J.F.
(Eds.), Geographic Information Science and Mountain Geomorphology. SpringerPraxis, Chichester, pp. 131.
182
Geomorphometry
Bishop, T.F.A., Minasny, B., 2005. Environmental soil-terrain modeling: the

predictive potential and uncertainty. In: Grunwald, S. (Ed.), Environmental SoilLandscape Modeling: Geographic Information Technologies and Pedometrics.
CRC Press, Boca Raton, FL, pp. 185213.
Bohner, J., Antonic, O., 2009. Land-surface parameters specific to topo-climatology.
In: Hengl, T., Reuter, H.I. (Eds.), Geomorphometry: Concepts, Software,
Bolstad, P.V., Lillesand, T.M., 1992. Improved classification of forest vegetation in
northern Wisconsin through a rule-based combination of soils, terrain, and
Landsat TM data. Forest Science 38, 520.
Bolstad, P.V., Stowe, T., 1994. An evaluation of DEM accuracy: elevation, slope and
aspect. Photogrammetric Engineering and Remote Sensing 60, 13271332.
Brown, D.G., 1998. Classification and boundary vagueness in mapping presettlement forest types. International Journal of Geographical Information Science
12, 105129.
Burbank, D., Leland, J., Fielding, E., Anderson, R.S., Brozovik, N., Reid, M.R.,
Duncan, C., 1996. Bedrock incision, rock uplift and threshold hillslopes in the
northwestern Himalaya. Nature 379, 505510.
Burns, W.J., Coe, J.A., Kay, B.S., Ma, L., 2010. Analysis of elevation changes
detected from multi-temporal LiDAR surveys in forested landslide terrain in
western Oregon. Environmental and Engineering Geoscience 16, 315341.
Burrough, P., 1996. Natural objects with indeterminate boundaries. In: Burrough,
P.A., Frank, A.U. (Eds.)., Geographic Objects with Indeterminate Boundaries.
Taylor and Francis, London, pp. 328.
Burrough, P.A., McDonnell, R., 1998. Principles of Geographical Information
Science. Oxford University Press, New York.
Burrough, P.A., van Gaans, P.F.M., MacMillan, R.A., 2000a. High-resolution
landform classification using fuzzy k-means. Fuzzy Sets and Systems 113,
3752.
Burrough, P.A., Wilson, J.P., van Gaans, P.F.M., Hansen, A.J., 2000b. Fuzzy kmeans classification of topo-climatic data as an aid to forest mapping in the
Greater Yellowstone Area, USA. Landscape Ecology 16, 523546.
Burt, T.P., Butcher, D.P., 1985. Topographic controls of soil moisture distribution.
Journal of Soil Science 36, 469486.
Carabajal, C., Harding, D., 2006. SRTM C-Band and ICESat laser altimetry elevation
comparisons as a function of tree cover and relief. Photogrammetric Engineering
Carara, A., Bitelli, G., Carla, R., 1997. Comparison of techniques for generating
digital terrain models from contour lines. International Journal of Geographical
Charnpratheep, K., Zhou, Q., Garner, B., 1997. Preliminary landfill site screening
using fuzzy geographical information systems. Waste Management and Research
15, 197215.
Chavez, Jr. P.S., 1996. Image-based atmospheric corrections: Revisited and improved.
Chirico, G.B., Western, A.W., Grayson, R.B., Gunter, B., 2005. On the definition of
the flow width for calculating specific catchment area patterns from gridded
elevation data. Hydrological Processes 19, 25392556.
Chow, T.E., Hodgson, M.E., 2009. Effects of LiDAR post-spacing and DEM
resolution to mean slope estimation. International Journal of Geographic
Conacher, A.J., Dalrymple, J.B., 1977. The nine-unit land surface model: an
approach to pedogeomorphic research. Geoderma 18, 1154.
Costa-Cabral, M., Burges, S.J., 1994. Digital Elevation Model Networks (DEMON), a
model of flow over hillslopes for computation of contributing and dispersal
areas. Water Resources Research 30, 16811692.
Cova, T.J., Goodchild, M.F., 2002. Extending geographical representation to include
fields of spatial objects. International Journal of Geographical Information
Science 16, 509532.
DeBruin, S., 2000. Querying probabilistic land cover data using fuzzy set theory.
International Journal of Geographical Information Science 14, 359372.
Deng, Y.X., 2007. New trends in digital terrain analysis: landform definition,
representation, and classification. Progress in Physical Geography 31,
405419.
Deng, Y.X., Wilson, J.P., 2006. The role of attribute selection in GIS representations
of the biophysical environment. Annals of the Association of American
Geographers 96, 4763.
Deng, Y.X., Wilson, J.P., 2008. Multi-scale and multi-criteria mapping of mountain
peaks as fuzzy entities. International Journal of Geographical Information Science
22, 205218.
Deng, Y.X., Wilson, J.P., Bauer, B.O., 2007. DEM resolution dependencies of terrain
attributes across a landscape. International Journal of Geographical Information
Science 21, 187213.
Deng, Y.X., Wilson, J.P., Gallant, J.C., 2008. Terrain analysis. In: Wilson, J.P.,
Fotheringham, A.S. (Eds.), Handbook of Geographic Information Science.
Blackwell Publishers, Oxford, pp. 417435.
Deng, Y.X., Wilson, J.P., Sheng, J., 2006. The sensitivity of fuzzy landform
classification to variable attribute weights. Earth Surface Processes and
Landforms 31, 14521462.
Desmet, P.J.J., 1997. Effects of interpolation errors on the analysis of DEMs. Earth
Desmet, P.J.J., Govers, G., 1996. Comparison of routing algorithms for
digital elevation models and their implications for predicting ephemeral
gullies. International Journal of Geographical Information Systems 10,
311331.
Dikau, R., 1989. The application of a digital relief model to landform analysis in
geomorphology. In: Raper, J. (Ed.), Three Dimensional Applications in
Geographical Information Systems. Taylor and Francis, London, pp. 5177.
Dowman, I.J., 2004. Integration of LIDAR and IFSAR for mapping. International
Archives of the Photogrammetry, Remote Sensing and Spatial Information
Sciences 35, 90100.
Dozier, J., Bruno, J., Downey, P., 1981. A faster solution to the horizon problem.
Dubyah, R., Rich, P.M., 1995. Topographic solar radiation models for GIS.
International Journal of Geographical Information Systems 9, 405419.
Endreny, T.A., Wood, E.F., 2001. Representing elevation uncertainty in runoff
modeling and flowpath mapping. Hydrological Processes 15, 22232236.
Endreny, T.A., Wood, E.F., 2003. Maximizing spatial congruence of observed and
DEM-delineated overland flow networks. International Journal of Geographical
Evans, I.S., 1972. General geomorphometry, derivatives of altitude, and descriptive
statistics. In: Chorley, R.J. (Ed.), Spatial Analysis in Geomorphology. Harper and
Row, London, pp. 1790.
Evans, I.S., Hengl, T., Gorsevski, P., 2009. Applications in geomorphology.
Fairfield, J., Leymarie, P., 1991. Drainage networks from grid digital elevation
models. Water Resources Research 27, 709717.
Feng, G., Sharratt, B., Vaughan, J., Lamb, B., 2009. A multiscale database of soil
properties for regional environmental quality modeling in the western United
States. Journal of Soil and Water Conservation 64, 363373.
Filin, S., 2003. Recovery of systematic biases in laser altimetry data using natural
surfaces. Photogrammetric Engineering and Remote Sensing 57, 12351242.
Fisher, P.F., 1991. First experiments in viewshed uncertainty: the accuracy of the
viewshed area. Photogrammetric Engineering and Remote Sensing 57,
13211327.
Fisher, P.F., 1992. First experiments in viewshed uncertainty: simulating fuzzy
viewsheds. Photogrammetric Engineering and Remote Sensing 58, 345352.
Fisher, P.F., 1993. Algorithm and implementation uncertainty in viewshed analysis.
International Journal of Geographical Information Systems 7, 331347.
Fisher, P.F., 1995. An exploration of probable viewsheds in landscape planning.
Environment and Planning B 22, 527546.
Fisher, P.F., 1996. Reconsideration of the viewshed function in terrain modeling.
Geographical Systems 3, 3358.
Fisher, P.F., 1998. Improved modeling of elevation error with geostatistics.
Geoinformatica 2, 215233.
Fisher, P.F., Pathirana, S., 1994. The evaluation of fuzzy membership of land cover
classes in the suburban zone. Remote Sensing of the Environment 34, 121132.
Fisher, P.F., Tate, N.J., 2006. Causes and consequences of error in digital elevation
models. Progress in Physical Geography 30, 467489.
Fisher, P.F., Wood, J., Cheng, T., 2004. Where is Helvellyn? Fuzziness of multiscale landscape morphometry. Transactions of the Institute of British
Fisher, P.F., Wood, J., Cheng, T., 2005. Fuzziness and ambiguity in multi-scale
analysis of landscape morphometry. In: Petry, F.E., Robinson, V.B., Cobb, M.A.
(Eds.), Fuzzy Modeling with Spatial Information for Geographic Problems.
Springer-Verlag, New York, pp. 209232.
Florinsky, I.V., 1998. Accuracy of local topographic variables derived from digital
elevation models. International Journal of Geographical Information Science 12,
4762.
Foody, G.M., 1996. Fuzzy modeling of vegetation from remotely sensed imagery.
Ecological Modeling 85, 312.
Freeman, G.T., 1991. Calculating catchment area with divergent flow based on a
regular grid. Computers and Geosciences 17, 413422.
Fried, J.S., Brown, D.G., Zweifler, M.O., Gold, M.A., 2000. Mapping contributing
areas for stormwater discharge to streams using terrain analysis. In: Wilson,
Geomorphometry
J.P., Gallant, J.C. (Eds.), Terrain Analysis: Principles and Applications. John
Wiley and Sons, New York, pp. 183203.
Fu, P., Rich, P.M., 2000. A geometric solar radiation model with applications
in agriculture and forestry. Computers and Electronics in Agriculture 37,
2535.
Gallant, J.C., Dowling, T.I., 2003. A multi-resolution index of valley bottom flatness
for mapping depositional areas. Water Resources Research 39, 13471360.
Gallant, J.C., Hutchinson, M.F., Wilson., J.P., 2000. Future directions for terrain
analysis. In: Wilson, J.P., Gallant, J.C. (Eds.), Terrain Analysis: Principles and
Gallant, J.C., Wilson, J.P., 1996. TAPES-G: a grid-based terrain analysis program
for the environmental sciences. Computers and Geosciences 22, 713722.
Gallant, J.C., Wilson, J.P., 2000. Primary topographic attributes. In: Wilson, J.P.,
Gallant, J.C. (Eds.), Terrain Analysis: Principles and Applications. John Wiley
and Sons, New York, pp. 5185.
Garbrecht., J., Martz, L.W., 1997. The assignment of drainage direction over
flat surfaces in raster digital elevation models. Journal of Hydrology 193,
204213.
Gessler, P., Pike, R., MacMillan, R.A., Hengl, T., Reuter, H.I., 2009. The future of
Software, Applications. Elsevier, Amsterdam, pp. 3163.
Giles, P.T., 2001. Remote sensing and cast shadows in mountainous terrain.
Grimaldi, S., Nardi, F., Di Benedetto, F., Istanbulluoglu, E., Bras, R.L., 2007. A
physically-based method for removing pits in digital elevation models. Advances
in Water Resources 30, 21512158.
Goodchild, M.F., 1992. Geographical data modeling. Computers and Geosciences
18, 401408.
Gruber, S., Peckham, S., 2009. Land-surface parameters and objects in hydrology.
Guzzetti, F., Reichenbach, P., Cardinali, M., Galli, M., Ardizzone, F., 2005.
Probabilistic landslide hazard assessment at the basin scale. Geomorphology 72,
272299.
Hengl, T., 2006. Finding the right pixel size. Computers and Geosciences 32,
12831298.
Hengl, T., Evans, I.S., 2009. Mathematical and digital models of the land surface.
Hengl, T., Gruber, S., Shrestha, D.P., 2004. Reduction of errors in digital terrain
parameters used in soil-landscape modeling. International Journal of Applied
Earth Observation and Geoinformation 5, 97112.
Hengl, T., MacMillan, R.A., 2009. Geomorphometry: a key to landscape mapping
and modeling. In: Hengl, T., Reuter, H.I. (Eds.), Geomorphometry: Concepts,
Hengl, T., Reuter, H. I. (Eds.), 2009. Geomorphometry: Concepts, Software, and
Applications. Developments in Soil Science. Elsevier, Amsterdam.
Hernandez Encinas, A., Hernandez Encinas, L., Hoya White, S., Martin del Rey, A.,
Rodriquez Sanchez, G., 2007. Simulation of forest fire fronts using cellular
automata. Advances in Engineering Software 38, 372378.
Hetrick, W.A., Rich, P.M., Barnes, F.J., Weiss, S.B., 1993a. GISbased solar
radiation flux models. In: Proceedings of the ASPRS-ACSM Annual Convention,
Volume 3. American Society for Photogrammetry and Remote Sensing,
Bethesda, Maryland, pp. 132143.
Hetrick, W.A., Rich, P.M., Weiss, S.B., 1993b. Modeling insolation on complex
surfaces. Proceedings of the Thirteenth Annual ESRI International User
Conference, Volume 2. Esri, Redlands, CA, pp. 447458.
Hiranoa, A., Welch, R., Lang, H., 2003. Mapping from ASTER stereo image data:
DEM validation and accuracy assessment. ISPRS Journal of Photogrammetry
Hodgson, M.E., Jensen, J.R., Raber, G., Tullis, J., Davis, B., Schuckman, K.,
Thompson, G., 2005. An evaluation of LiDAR-derived elevation and terrain slope
in leaf-off conditions. Photogrammetric Engineering and Remote Sensing 71,
817823.
Hofierka, J., 1997. Direct solar radiation within an open GIS environment.
Proceedings of the 1997 Joint European GI Conference, Vienna, Austria,
pp. 575584.
Hofton, M., Dubayah, R., Blair, J.B., Rabine, D., 2006. Validation of SRTM
elevations over vegetated and non-vegetated terrain using medium footprinting
LiDAR. Photogrammetric Engineering and Remote Sensing 72, 279285.
Holmes, K.W., Chadwick, O.A., Kyriakidis, P.C., 2000. Error in a USGS 30 m digital
elevation model and its impact on digital terrain modeling. Journal of Hydrology
233, 154173.
183
Holmgren, P., 1994. Multiple flow direction algorithms for runoff modeling in grid
based elevation models: an empirical evaluation. Hydrological Processes 8,
327334.
Homann, P.S., Sollins, P., Fiorella, M., Thorson, T., Kern, J.S., 1998. Regional soil
organic carbon storage estimates for western Oregon by multiple approaches.
Soil Science Society of America Journal 62, 789796.
Hunter, G.J., Goodchild, M.F., 1997. Modeling the uncertainty of slope and
aspect estimates derived from spatial databases. Geographical Analysis 29,
3549.
Hutchinson, M.F., 1989. A new procedure for gridding elevation and stream line
data with automatic removal of spurious pits. Journal of Hydrology 106,
211232.
Hutchinson, M.F., 2008. Adding the Z-dimension. In: Wilson, J.P., Fotheringham,
A.S. (Eds.), Handbook of Geographic Information Science. Blackwell Publishers,
Oxford, pp. 144168.
Hutchinson, M.F., Gallant, J.C., 2000. Digital elevation models and representation of
terrain shape. In: Wilson, J.P., Gallant, J.C. (Eds.), Terrain Analysis: Principles
and Applications. John Wiley and Sons, New York, pp. 2950.
Jenson, S.K., Domingue, J.O., 1988. Extracting topographic attributes from digital
elevation data for geographical information system analysis. Photogrammetric
Jiang, H., Eastman, J.R., 2000. Application of fuzzy measures in multi-criteria
evaluation in GIS. International Journal of Geographical Information Science 14,
173184.
Jones, J.A., 1986. Some limitations to the a/s index for predicting basin-wide
patterns of soil water drainage. Zeitschrift fur Geomorphologie 60, 720.
Jones, J.A., 1987. The initiation of natural drainage networks. Progress in Physical
Jones, K.H., 1998. A comparison of algorithms used to compute hill slope as a
property of the DEM. Computers and Geosciences 24, 315323.
Katzenbeisser, R., 2003. On the calibration of LiDAR sensors. In: Maas, H-G.,
Vosselman, G., Streilein, A. (Eds.), 3-D Reconstruction from Airborne
Laserscanner and InSAR Data. Institute of Photogrammetry and Remote Sensing.
University of Twente, Enschede, The Netherlands, pp. 5964.
Kheir, R.B., Wilson, J.P., Deng, Y.X., 2007. Use of terrain variables for mapping
gully erosion susceptibility in Lebanon. Earth Surface Processes and Landforms
32, 17701782.
Kienzle, S., 2004. The effect of DEM raster resolution on first order, second order,
and compound terrain derivatives. Transactions in GIS 8, 83111.
Klir, G.J., Yuan, B., 1995. Fuzzy Sets and Fuzzy Logic: Theory and Applications.
Prentice-Hall, Upper Saddle Creek, NJ.
Koons, P., Upton, P., Barker, A.D., 2012. The influence of mechanical properties
on the link between tectonic and topographic evolution. Geomorphology 137,
168180.
Lagacherie, P., Andrieux, P., Bouzigues, R., 1996. Fuzziness and uncertainty of soil
boundaries: From reality to coding in GIS. In: Burrough, P.A., Frank A.U. (Eds.),
Geographic Objects with Indeterminate Boundaries. Taylor and Francis, London,
pp. 275286.
LaLonde, T., Shortridge, A., Messina, J., 2010. The influence of land cover on
Shuttle Radar Topography Mission (SRTM) elevations in low-relief areas.
Transactions in GIS 14, 461479.
Lea, N.L., 1992. An aspect driven kinematic routing algorithm. In: Parsons, A.J.,
Abrahams, A.D. (Eds.), Overland Flow: Hydraulics and Erosion Mechanics.
Chapman and Hall, New York, pp. 147175.
Li, Z., Zhu, Q., Gold, C., 2005. Digital Terrain Modeling: Principles and
Methodology. CRC Press, Boca Raton, FL.
Lindsay, J.B., 2006. Sensitivity of channel mapping techniques to uncertainty in
digital elevation data. International Journal of Geographical Information Science
20, 669692.
Lindsay, J.B., Creed, I.F., 2005a. Removal of artifact depressions from digital
elevation models: towards a minimum impact approach. Hydrological Processes
19, 31133126.
Lindsay, J.B., Creed, I.F., 2005b. Sensitivity of digital landscapes to artifact
depressions in remotely-sensed DEMs. Photogrammetric Engineering and
Lindsay, J.B., Creed, I.F., 2006. Distinguishing between artifact and real depressions
in digital elevation data. Computers and Geosciences 32, 11941204.
Lindsay, J.B., Evans, M.G., 2006. The influence of elevation error on the
morphometrics of channel networks extracted from DEMs and the implications
for hydrological modeling. Hydrological Processes 22, 15881603.
Lindsay, J.B., Rothwell, J.J., Davies, H., 2008. Mapping outlet points used for
watershed delineation onto DEM-derived stream networks. Water Resources
Research 44, W08442.
184
Geomorphometry
Mackay, D.S., Samanta, S., Ahl, D.E., Ewers, B.E., Gower, S.T., Burrows, S.N., 2003.
Automated parameterization of land surface process models using fuzzy logic.
Transactions in GIS 7, 139153.
MacMillan, R.A., Shary, P.A., 2009. Landforms and landform elements in
MacMillan, R.A., Pettapiece, W.W., Nolan, S.C., Goddard, T.W., 2000. A generic
procedure for automatically segmenting landforms into landform elements using
DEMs, heuristic rules and fuzzy logic. Fuzzy Sets and Systems 113, 81109.
Martz, L.W., Garbrecht, J., 1999. An outlet breaching algorithm for the treatment of
closed depressions in a raster DEM. Computers and Geosciences 25, 835844.
Martz, L.W., de Jong, E., 1988. CATCH: a Fortran program for measuring catchment
area from digital elevation models. Computers and Geosciences 14, 627640.
McBratney, J.A.B., Odeh, I.O.A., 1997. Application of fuzzy sets in soil science:
fuzzy logic, fuzzy measurements, and fuzzy decisions. Geoderma 77, 85113.
Meszaros, I., 1998. Modelovanie prkonu slnecnej energie na horske povodie. Acta
Hydrologica Slovaca 1, 6875.
Miklanek, P., 1993. The Estimation of Energy Income in Grid Points Over the Basin
Using a Simple Digital Elevation Model. Springer, Berlin, European Geophysical
Society, Annales Geophysicae 11, Suppl. II, Berlin.
Miller, A.M., White, R.A., 1998. A conterminous United States multilayer soil
characteristics database for regional climate and hydrology modeling. Earth
Interactions 2, 126.
Mitasova, H., Mitas, L., Brown, W.M., Gerdes, D.P., Kosinovsky, I., Baker, T., 1995.
Modeling spatially and temporally distributed phenomena: new methods and
tools for GRASS GIS. International Journal of Geographical Information Systems
9, 433446.
climate change: Chicken or egg? Nature 346, 2934.
Montgomery, D.R., Dietrich, W.E., 1989. Source areas, drainage density, and
channel initiation. Water Resources Research 25, 19071918.
Montgomery, D.R., Dietrich, W.E., 1992. Channel initiation and the problem of
landscape scale. Science 255, 826830.
Montgomery, D.R., Dietrich, W.E., 1994. A physically-based model for the
topographic control on shallow landsliding. Water Resources Research 30,
11531171.
Moore, I.D., Burch, G.J., 1986. Modeling erosion and deposition: topographic
effects. Transactions of the American Society of Agricultural Engineers 29,
16241630, 1640.
Moore, I.D., Burch, G.J., MacKenzie, D.H., 1988. Topographic effects on the
distribution of surface soil water and the location of ephemeral gullies.
Transactions of the American Society of Agricultural Engineers 31, 10981117.
Moore, I.D., Grayson, R.B., Ladson, A.R., 1991. Digital terrain modeling: a review of
hydrological, geomorphological, and biological applications. Hydrological
Processes 5, 330.
Moore, I.D., Lewis, A., Gallant, J.C., 1993a. Terrain attributes: estimation
methods and scale effects. In: Jakeman, A.J., Beck, M.B., McAleer, M.J. (Eds.),
Modeling Change in Environmental Systems. John Wiley and Sons, New York,
pp. 189214.
Moore, I.D., Norton, T.W., Williams, J.E., 1993b. Modeling environmental
heterogeneity in forested landscapes. Journal of Hydrology 150, 717747.
Moore, I.D., Wilson, J.P., 1992. Length-slope factors for the Revised Universal Soil
Loss Equation: simplified method of estimation. Journal of Soil and Water
Conservation 47, 423428.
Moore, I.D., Wilson, J.P., 1994. Reply to Comment on length-slope factors for the
Revised Universal Soil Loss Equation: Simplified method of estimation by G.R.
Foster. Journal of Soil and Water Conservation 49, 174180.
Morris, D., Heerdegen, R., 1988. Automatically derived catchment boundaries and
channel networks and their hydrological applications. Geomorphology 1,
131141.
Nelson, A., Reuter, H.I., Gessler, P., 2009. DEM production methods and sources.
In: Hengl, T., Reuter, H.I. (Eds.), Geomorphometry: Concepts, Software, and
Nguyen, T.M., 2011. Uncertainty in geomorphometry. University of Southern
California, Unpublished PhD dissertation.
Nguyen, T.M., Wilson, J.P., 2010. Sensitivity of the quasi-dynamic topographic
wetness index to choice of DEM resolution, flow routing algorithm and soil
variability. Proceedings of the Ninth International Symposium on Spatial
Accuracy Assessment in Natural Resources and the Environmental Sciences
(Accuracy 2010), Leicester, UK.
OCallaghan, J.F., Mark, D.M., 1984. The extraction of drainage networks from
digital elevation data. Computer Vision, Graphics and Image Processing 28,
323344.
OLoughlin, E.M., 1986. Prediction of surface saturation zones in natural catchments

by topographic analysis. Water Resources Research 22, 794804.
Olaya, V., 2009. Basic land-surface parameters. In: Hengl, T., Reuter, H.I. (Eds.),
Geomorphometry: Concepts, Software, Applications. Elsevier, Amsterdam, pp.
141169.
Peckham, S.D., 1998. Efficient extraction of river networks and hydrologic
measurements from digital elevation data. In: Barndorff-Nielsen, O.E., Gupta,
V.K., Perez-Abreu, V., Waymire, E. (Eds.), Stochastic Methods in Hydrology:
Rain, Landforms, and Floods. World Scientific, Singapore, pp. 173203.
Phillips, J.D., 1990. A saturation-based model of relative wetness for wetland
identification. Water Resources Bulletin 26, 333342.
Pike, R.J., 1995. Geomorphometry: progress, practise and prospect. Zeitschrift fur
Geomorphologie 101, 221238.
Pike, R.J., 2000. Geomorphometry: diversity in quantitative surface analysis.
Progress in Physical Geography 24, 120.
T., Reuter, H.I. (Eds.), Geomorphometry: Concepts, Software, Applications.
Elsevier, Amsterdam, pp. 330.
Planchon, O., Darboux, F., 2001. A fast, simple and versatile algorithm to fill the
depressions of digital elevation models. Catena 46, 159176.
Proy, C., Tanre, D., Deschamps, P.Y., 1989. Evaluation of topographic effects in
remotely sensed data. Remote Sensing of the Environment 30, 2132.
Qi, F., 2004. Knowledge discovery from area-class resource maps: data
preprocessing for noise reduction. Transactions in GIS 8, 297308.
Qi, F., Zhu, A.-X., 2003. Knowledge discovery from soil maps using inductive
learning. International Journal of Geographical Information Science 17,
771795.
Qi, F., Zhu, A.-X., Harrower, M., Burt, J.E., 2006. Fuzzy soil mapping based on
prototype category theory. Geoderma 136, 774787.
Qin, C., Zhu, A-X., Pei, T., Li, B., Zhou, C., Yang, L., 2007. An adaptive approach
to selecting the flow partition exponent for multiple flow direction algorithms.
Quinn, P.F., Beven, K.J., Chevallier, P., Planchon, O., 1991. The prediction of
hillslope paths for distributed hydrological modeling using digital terrain
models. Hydrological Processes 5, 5979.
Quinn, P.F., Beven, K.J., Lamb, R., 1995. The ln(a/tan b) index: how to calculate it
and how to use it within the TOPMODEL framework. Hydrological Processes 9,
161182.
Quinn, T., Zhu, A-X., Burt, J.E., 2005. Effects of detailed soil spatial information on
watershed modeling across different model scales. International Journal of
Applied Earth Observation and Geoinformation 7, 324338.
Raaflaub, L.D., Collins, M.J., 2006. The effect of error in gridded digital elevation
models on the estimation of topographic parameters. Environmental Modeling
and Software 21, 710732.
Raper, J., Livingstone, D., 1995. Development of a geomorphological spatial model
using object-oriented design. International Journal of Geographical Information
Systems 9, 359383.
Rawls, W.J., 1983. Estimating soil bulk density from particle size analysis and
organic matter. Soil Science 135, 123125.
Reuter, H.I., Hengl, T., Gessler, P., Soille, P., 2009. Preparation of DEMs for
geomorphometric analysis. In: Hengl, T., Reuter, H.I. (Eds.), Geomorphometry:
Concepts, Software, and Applications. Elsevier, Amsterdam, pp. 87120.
Reuter, H.I., Kersebaum, K.C., Wendroth, O., 2005. Modeling of solar radiation
influenced by topographic shading: evaluation and application for precision
farming. Physics and Chemistry of the Earth 30, 139149.
Reiger, W., 1992. Automated river line and catchment area extraction for DEM data.
Proceedings of the Seventeenth International Society for Photogrammetry and
Remote Sensing Congress, Washington, DC, pp. 642649.
Robinson, V.B., 2003. A perspective on the fundamentals of fuzzy sets and their use
in Geographical Information Systems. Transactions in GIS 7, 330.
Rodriguez, E., Morris, C., Belz, J.E., 2006. A global assessment of the SRTM
performance. Photogrammetric Engineering and Remote Sensing 72, 249260.
Rodriguez, E., Morris, C.S., Belz, J.E., Chapin, E.C., Martin, J.M., Daffer, W.,
Hensley, S., 2005. An Assessment of the SRTM Topographic Products. Jet
Propulsion Laboratory, Pasadena, California.
Rossi, R.E., Dungan, J.L., Beck, L.R., 1994. Kriging in the shadows: Geostatistical
interpolation for remote sensing. Remote Sensing of the Environment 49,
3240.
Ruhe, R.V., 1960. Elements of the soil landscape. Proceedings of the Ninth
Congress of the International Society of Soil Science, Madison, WI, pp.
165170.
Ruiz, M., 1997. A causal analysis of error in viewsheds from USGS digital elevation
models. Transactions in GIS 2, 8594.
Geomorphometry
Saunders, W.K., Maidment, D.R., 1996. A GIS Assessment of Nonpoint Source

Pollution in the San Antonio-Nueces Coastal Basin. University of Texas Center
for Research in Water Resources Report No. 96-1, Austin, TX.
Saxton, K.E., Rawls, W.J., 2006. Soil water characteristic estimates by texture and
organic matter for hydrologic solutions. Soil Science Society of America
Agronomy Journal 70, 15691578.
Saxton, K.E., Rawls, W.J., Romberger, J.S., Papendick, R.I., 1986. Estimating
generalized soil-water characteristics from texture. Transactions of the American
Society of Agricultural Engineers 50, 10311035.
Schmidt, J., Dikau, R., 1999. Extracting geomorphometric attributes and objects
from digital elevation models: semantics, methods, and future needs. In: Dikau,
R., Saurer, H. (Eds.), GIS for Earth Science Systems: Analysis and Modeling of
the Natural Environment. Schweizbartsche Verlagsbuchhandlung, Stuttgart,
Germany, pp. 153173.
Schmidt, J., Hennrich, K., Dikau, R., 1998. Scales and similarities in runoff
processes with respect to geomorphometry. Proceedings of the Third
International Conference on GeoComputation, Bristol, UK.
Schmidt, J., Hewitt, A., 2004. Fuzzy land element classification from DTMs based
on geometry and terrain position. Geoderma 121, 243256.
Schoorl, J.M., Sonneveld, M.P.W., Veldkamp, A., 2000. Three-dimensional
landscape process modeling: the effect of DEM resolution. Earth Surface
Shary, P.A., 1995. Land surface in gravity points classification by a complete
system of curvatures. Mathematical Geology 27, 373390.
Shary, P.A., Sharaya, L.S., Mitusov, A.V., 2002. Fundamental quantitative methods of
land surface analysis. Geoderma 107, 132.
Shary, P.A., Sharaya, L.S., Mitusov, A.V., 2005. The problem of scale-specific and
scale-free approaches to geomorphometry. Geografia Fisica e Dinamica
Quaternaria 28, 81101.
Shary, P.A., Stepanov, I.N., 1991. On the second derivative method in geology.
Doklady AN SSSR 319, 456460, (in Russian).
Shi, X., Zhu, A-X., Wang, R-X., 2005. Deriving fuzzy representations of some
special terrain features based on their typical locations. In: Petry, F.E., Robinson,
V.B., Cobb, M.A. (Eds.), Fuzzy Modeling with Spatial Information for Geographic
Problems. Springer-Verlag, New York, pp. 233252.
Shortridge, A., 2006. Shuttle Radar Topography Mission elevation data error and its
relationship to land cover. Cartography and Geographic Information Science 33,
6575.
Sivapalan, M., Beven, K.J., Wood, E.F., 1987. On hydrologic similarity: 2, a
scaled model of storm runoff production. Water Resources Research 23,
22662278.
Skidmore, A.K., 1989. A comparison of techniques for calculating gradient and
aspect from a gridded digital elevation model. International Journal of
Geographical Information Systems 3, 323334.
Skidmore, A.K., Ryan, P.J., Short, D., Dawes, W., 1991. Forest soil type mapping
using an expert system with Landsat Thematic Mapper data and a digital
elevation model. International Journal of Geographical Information Systems 5,
431445.
Slater, J.A., Heady, B., Kroenung, G., et al., 2009. Evaluation of the New ASTER
Global Digital Elevation Model. National Geospatial-Intelligence Agency,
Washington, DC.
Smith, M.P., Zhu, A.X., Burt, J.E., Stiles, C., 2006. The effects of DEM resolution
and neighborhood size on digital soil survey. Geoderma 137, 5868.
Soille, P., 2004. Optimal removal of spurious pits in grid digital elevation models.
Water Resources Research 40, W12509.
Soille, P., Gratin, C., 1994. An efficient algorithm for drainage networks extraction
on DEMs. Journal of Visual Communication and Image Representation 5,
181189.
Soille, P., Vogt, J., Colombo, R., 2003. Carving and adaptive drainage enforcement
of grid digital elevation models. Water Resources Research 39, 13661378.
Speight, J.G., 1968. Parametric description of land form. In: Stewart, G.A. (Ed.),
Land Evaluation. Papers of a CSIRO Symposium, MacMillan, Melbourne,
pp. 239250.
Speight, J.G., 1990. Landform. In: MacDonald, R.C., Isbell, R.F., Speight, J.G.,
Walker, J., Hop, M.S. (Eds.), Australian Soil and Land Survey Field Handbook.
Inkata Press, Melbourne, pp. 957.
Sulebak, J.R., Hjelle, ., 2003. Multi-resolution spline models and their
applications in geomorphology. In: Evans, I.S., Dikau, R., Tokunaga, R., Ohmori,
H., Hirano, M. (Eds.), Concepts and Modeling in Geomorphology: International
Perspectives. Terra Publications, Tokyo, pp. 221237.
Tang, G., Shi, W., Zhao, M., 2002. Evaluation of the accuracy of hydrologic data
derived from DEMs of different spatial resolution. In: Hunter, G.J., Lowell, K.
(Eds.), Proceedings of the Fifth International Symposium on Spatial Accuracy
185
Assessment in Natural Resources and Environmental Sciences (Accuracy 2000).

RMIT University, Melbourne, pp. 204213.
Tarboton, D.G., 1997. A new method for the determination of flow directions and
upslope areas in grid digital elevation models. Water Resources Research 33,
309319.
Temme, A.J.A.M., Heuvelink, G.B.M., Schoorl, J.M., Claessens, L., 2009.
Geostatistical simulation and error propagation in geomorphometry. In: Hengl,
T., Reuter, H.I. (Eds.), Geomorphometry: Concepts, Software, and Applications.
Temme, A.J.A.M., Schoorl, J.M., Veldkamp, A., 2006. An algorithm for dealing with
depressions in dynamic landscape evolution models. Computers and
Thompson, J.A., Bell, J.C., Butler, C.A., 2001. Digital elevation model resolution:
effects on terrain attribute calculation and quantitative soil-landscape modeling.
Geoderma 100, 6789.
Usery, E.L., 1996. A conceptual framework and fuzzy set implementation for
geographic features. In: Burrough, P.A., Frank, A.U. (Eds.)., Geographic Objects
with Indeterminate Boundaries. Taylor and Francis, London, pp. 7185.
Van Niel, K., Laffan, S.W., Lees, B.G., 2004. Effect of error in the DEM on
environmental variables for predictive vegetation modeling. Journal of Vegetation
Science 15, 747756.
Waltman, W.J., Waltman, S.W., Palecki, M.A., 2003. A GIS-based spatial pattern
analysis model for eco region mapping and characterization. International
Journal of Geographical Information Science 17, 445462.
Wang, L., Liu, H., 2006. An efficient method for identifying and filling surface
depressions in digital elevation models for hydrologic analysis and modeling.
Warren, S.D., Hohmann, M.G., Auerswald, K., Mitasova, H., 2004. An evaluation of
methods to determine slope using digital elevation data. Catena 58, 215233.
Webster, T.L., Dias, G., 2006. An automated GIS procedure for comparing GPS and
proximal LIDAR elevations. Computers and Geosciences 32, 713726.
Wilson, J.P., 2012. Digital terrain modeling. Geomorphometry 137, 107121.
Wilson, J.P., Aggett, G.R., Deng, Y.X., Lam, C.S., 2008. Water in the landscape: a
review of contemporary flow routing algorithms. In: Zhou, Q., Lees, B.G., Tang,
G-A. (Eds.), Advances in Digital Terrain Analysis. Springer Lecture Notes in
Geoinformation and Cartography, Berlin, pp. 213236.
Wilson, J.P., Burrough, P.A., 1999. Dynamic modeling, geostatistics, and fuzzy
classification: new sneakers for a new geography? Annals of the Association of
Wilson, J.P., Gallant, J.C., 2000a. Digital terrain analysis. In: Wilson, J.P., Gallant,
J.C. (Eds.), Terrain Analysis: Principles and Applications. John Wiley and Sons,
New York, pp. 127.
Wilson, J.P., Gallant, J.C., 2000b. Secondary topographic attributes. In: Wilson, J.P.,
Wilson, J.P., Inskeep, W.P., Wraith, J.M., Snyder, R.D., 1996. GIS-based solute
transport modeling applications: scale effects of soil and climate input data.
Journal of Environmental Quality 25, 445453.
Wilson, J.P., Lam, C.S., Deng, Y.X., 2007. Comparison of performance of flow
routing algorithms used in Geographic Information Systems. Hydrological
Processes 21, 10261044.
Wilson, J.P., Repetto, P.L., Snyder, R.D., 2000. Effect of data source, grid resolution,
and flow-routing method on computed topographic attributes. In: Wilson, J.P.,
Wise, S.M., 1998. The effect of GIS interpolation errors on the use of digital
elevation models in geomorphology. In: Lane, S.N., Richards, K.S., Chandler,
J.H. (Eds.), Landform Monitoring, Modeling, and Analysis. John Wiley and
Sons, Chichester, pp. 139164.
Wise, S.M., 2000. Assessing the quality of hydrological applications of digital
elevation models derived from contours. Hydrological Processes 14,
19091929.
Wolock, D.M., McCabe, G.J., 1995. Comparison of single and multiple flow
direction algorithms for computing topographic parameters in TOPMODEL.
Wood, R., Sivapalan, M., Robinson, J., 1997. Modeling the spatial variability of
surface runoff using a topographic index. Water Resources Research 33,
10611073.
Yokoyama, R., Shirasawa, M., Pike, R.J., 2002. Visualizing topography by
openness: a new application of image processing to digital elevation models.
Zandbergen, P., 2008. Applications of Shuttle radar topography mission elevation
data. Geography Compass 2, 14041431.
186
Geomorphometry
Zhang, W.H., Montgomery, D.R., 1994. Digital elevation model grid size, landscape
representation, and hydrological simulations. Water Resources Research 30,
10191028.
Zhou, Q., Lees, B.G., Tang, G-A. (Eds.), 2008. Advances in Digital Terrain Analysis.
Springer Lecture Notes in Geoinformation and Cartography, Berlin.
Zhou, Q., Liu, X., 2002. Error assessment of grid-based flow routing algorithms
used in hydrological models. International Journal of Geographical Information
Science 16, 819842.
Zhou, Q., Liu, X., 2004. Analysis of errors of derived slope and aspect related to
DEM data properties. Computers and Geosciences 30, 369378.
Zhu, A-X., 1997a. A similarity model for representing soil spatial information.
Zhu, A-X., 1997b. Measuring uncertainty in class assignment for natural resource
maps under fuzzy logic. Photogrammetric Engineering and Remote Sensing 63,
11951202.
Zhu, A-X., 1999. A personal construct-based knowledge acquisition for natural resource
mapping. International Journal of Geographical Information Science 13, 119141.
Zhu, A-X., Band, L.E., Vertessy, R., Dutton, B., 1997. Derivation of soil properties
using a soil land inference model (SoLIM). Soil Science Society of America
Journal 61, 523533.
Zhu, A-X., Mackay, D.S., 2001. Effects of spatial detail of soil information on
watershed modeling. Journal of Hydrology 248, 5477.
Biographical Sketch
Dr. John P Wilson is a professor of spatial sciences at the University of Southern California (USC), where he heads
the Spatial Sciences Institute as well as the Geographic Information Science & Technology Graduate Programs and
GIS Research Laboratory, and also holds appointments as professor in the Dornsife Colleges Department of
Sociology, the School of Architecture and in the Viterbi School of Engineerings Departments of Computer Science
and Civil & Environmental Engineering. His current research is focused on the use of geospatial data and tools for
characterizing hydrologic systems and the relationships linking environmental exposures, societal forces, and
human health outcomes.
Dr. Michael P Bishop is a professor and Haynes Chair of Geosciences in the Department of Geography at Texas
A&M University. He received his PhD at Indiana State University (1987) in physical geography with a focus on
remote sensing and geographic information science. His areas of expertise are in remote sensing, geographic
information systems (GIS), geomorphometry, numerical modeling, and mountain geomorphology. He has
published over 30 articles in scientific journals, three books, and numerous book chapters on topics including
radiation transfer, image and terrain analysis, surface processes and landforms, climate and glacier change, and
landscape evolution modeling. Furthermore, he has presented over 200 national and international professional
papers on various remote sensing, GIS, and mountain geomorphology-related topics. Financial support for his
research has been obtained through the National Geographic Society, National Science Foundation, NASA, DOE,
USGS, and numerous foundations. His current research is focused on the use of satellite imagery and terrain
analysis for characterizing surface processes and landforms in complex mountain environments.
3.8 Remote Sensing and GIScience in Geomorphological Mapping

J Napieralski, University of Michigan-Dearborn, Dearborn, MI, USA
I Barr, Queen Mary University of London, London, UK
U Kamp, The University of Montana, Missoula, MT, USA
M Kervyn, Vrije Universiteit Brussel, Brussels, Belgium
3.8.1
3.8.2
3.8.2.1
3.8.2.1.1
3.8.2.1.2
3.8.2.1.3
3.8.2.1.4
3.8.2.2
3.8.2.2.1
3.8.3
3.8.3.1
3.8.3.2
3.8.3.2.1
3.8.3.2.2
3.8.3.2.3
3.8.3.3
3.8.4
3.8.4.1
3.8.4.2
3.8.4.3
3.8.5
3.8.5.1
3.8.5.2
3.8.5.3
3.8.5.3.1
3.8.5.3.2
3.8.5.3.3
3.8.5.3.4
3.8.5.4
3.8.5.5
3.8.5.6
3.8.6
3.8.6.1
3.8.6.2
3.8.6.3
3.8.6.4
3.8.7
References
Introduction
Background
Remotely Sensed Data
Aerial photographs
Satellite images
Digital elevation models
Supplementary data
GIScience
Mapping approaches
Glacial Landscapes and Landforms
Glacial Landforms
Data Processing and Visualization
Aerial photographs
Satellite images
GIS-based Mapping
Volcanic Terrain and Landforms
Satellite Images
Digital Elevation Models
GIS-Based Mapping
Landslide Mapping
Terrestrial and Airborne Photography
Satellite Imagery
LiDAR
Radar
ASTER
Geomorphometric analysis
GIS Modeling
GIS-based Landslide Inventories
GIS-based Landslide Assessment Maps
Fluvial Landscapes and Landforms
Aerial Photographs
Airborne And Satellite Sensors
Active Sensors
Conclusion
Glossary
Aerial photography The act of photographing Earths
surface from an elevated position. Aerial photographs are
188
190
190
190
190
191
192
192
194
194
195
195
195
196
197
198
199
201
203
206
207
207
208
209
209
209
211
211
212
212
212
213
213
214
215
215
218
219
most commonly acquired using visible light from either

vertical (e.g., useful for photogrammetry) or oblique
perspectives.
Napieralski, J., Barr, I., Kamp, U., Kervyn, M., 2013. Remote sensing and
GIScience in geomorphological mapping. In: Shroder, J. (Editor in Chief),
http://dx.doi.org/10.1016/B978-0-12-374739-6.00050-6
187
188
Remote Sensing and GIScience in Geomorphological Mapping

portray the altitude of the topography. In geomorphology,
it commonly takes the form of a raster data layer
representing a field of square tessellations. The resolution of
the grid cells is generally determined based upon the source
data utilized and the desired scale for representing the
topography.
Fuzzy logic Fuzzy theory deals with approximations
rather than exactness. Fuzzy logic is used in
geomorphological mapping because of the difficulty
associated with mapping exact boundaries of landforms.
Geographic information science The discipline of
GIScience is concerned with the theory behind the
development, implementation, and application of GIS.
Geographic information systems A geographic
information system (GIS) is used to store, manipulate,
analyze, and display geographically referenced data.
Geomorphology An Earth-science discipline that focuses
on studying and understanding surface processes and
the morphological characteristics of the topography, via
the classification, measurement, and description of
landforms.
Geomorphometry The discipline that is concerned
with the science of quantitative land-surface
characterization.
Photogrammetry The science and technology of
obtaining quantitative measurements and other
geometrically reliable information from photographs.
An advanced version of photogrammetry is
stereophotogrammetry, which allows for calculations of

relief (a valuable landform mapping parameter).
Raster data A data structure that is used to represent a
field data model to characterize surface attributes. It
represents a field of tessellations organized into rows and
columns where each cell/pixel contains an attribute value
(e.g., elevation, reflectance).
Remote sensing The science and technology of obtaining
information about objects and planetary surfaces using
sensors that record variations in surface matter/energy
interactions. It also involves the symbolic transformation of
data into useable information that can be used to study the
Earth and solve problems.
Scale The concept of scale can be defined in many ways.
It is most commonly used to characterize the geographic
coverage and level of detail associated with maps,
photographs, and imagery. A large scale indicates a
relatively large representative fraction (e.g., relatively
more detail and less spatial coverage compared to small
scale).
Semantics The relationship between computer
representations and the corresponding real-world feature
within an application domain. Semantic models can be
used for landform mapping, and they formalize the
landform attributes and characteristics that permit GISbased landform mapping.
Vector data Spatial cartographic primitives are used to
characterize discrete features (i.e., points, lines, and
polygons). Vector data also include attribute information
that is referenced to the spatial information.
Abstract
The rapid proliferation of remote sensing and geographic information systems (GIS) into geomorphologic mapping
has increased the objectivity and efficiency of landform segmentation, measurement, and classification. The near
ubiquitous presence of Earth-observing satellites provides an array of perspectives to visualize the biophysical characteristics
of landscapes, access inhospitable terrain on a predictable schedule, and study landscape processes when conditions
are hazardous. GIS technology has altered the analysis, visualization, and dissemination of landform data due to the
shared theoretical concepts that are fundamental to geomorphology and GIScience. The authors review geospatial technology applications in landform mapping (including emerging issues) within glacial, volcanic, landslide, and fluvial
research.
3.8.1
Introduction
Geomorphological studies serve to improve our understanding and problem-solving capabilities involving surface processes, landscape evolution, landform mapping, hazard
prediction, inventory, planning/management, and ecological
analyses (Dragut- and Blaschke, 2006). Surveying and mapping technologies are critical to geomorphological and mapping studies, as they provide a primary source of data to
represent bare-Earth terrain, and permit information extraction for product generation. In particular, geomorphologists frequently view landforms as geometric signatures, or
evidence that specific processes modified the shape of the
Earths surface. Therefore, landform mapping can be used to

study dynamic processes based upon landform distributions,
terrain morphology, and terrain composition. The techniques
and approaches by which landforms are delineated, measured,
and classified have evolved rapidly, from point observations,
surveys, and qualitative interpretations, to a dependence on
digital data from airborne- and space-based sensors, advanced
geostatistical analysis and spatial modeling, and scientific
visualization techniques. Although field-based surveying and
qualitative methods continue to support geomorphic studies,
contemporary geomorphologists rely on an ever-expanding
range and profusion of digital data because of the relative ease
by which inaccessible terrain can be studied, the availability
and low-cost of topographic and spectral data, and the

access to automation and efficient quantitative assessment
(Millington and Townshend, 1987; Walsh et al., 1998; Smith
and Pain, 2009). Geographic information technologies (GIT)
that include digital image processing, geographic information
systems (GIS), and global positioning systems (GPS), have
equipped geomorphologists with the capability to solve more
sophisticated and interdisciplinary problems. This includes
revisiting more traditional concepts and problems, addressing
new concepts and issues (Hynek and Phillips, 2003; Mullen
and Kellet, 2007), developing algorithms and analysis procedures (Chandler et al., 2005), and evaluating new algorithms and emerging technologies in landform studies (Lejot
et al., 2007; Wilford, 2009).
With the rapid proliferation of GIT, geomorphologists are
routinely utilizing GIS to analyze the topography and map
and study landforms. The use of GIT, however, will not
completely replace traditional field methods of mapping, as
first-hand knowledge of a study area can yield information
that is not readily accessible solely from spatial data and GISbased analysis. Rather, the complexity of landscape mapping
and geomorphology requires a comprehensive approach to
fully understand landform morphology and landform-process
linkages. Figure 1 illustrates the value of combining multiple
datasets, along with field knowledge, to identify landforms
and understand geomorphic processes. Aerial photos provide
an easy-to-interpret perspective of the Western Michigan
landscape (Figure 1(a)). Topographic variations are difficult
to detect in the photograph, so a contoured elevation model
(Figure 1(b)), derived from the U.S. National Elevation
Dataset (NED), illustrates the lay of the land, which includes
topographic highs (e.g., south and north of the Herring Lakes,
and Lake Michigan shoreline) and topographic lows (Herring
Lakes). Segmenting this landscape into individual units
or landforms is significantly enhanced when using multispectral imagery (i.e., regions of the spectrum beyond visible
wavelengths). In particular, landforms such as sand dunes,
water bodies, and sediment-filled depressions are enhanced
using a Landsat Thematic Mapper (TM) image of the area
(Figure 1(c)). It is only when information derived from these
datasets are combined, along with a contextualization of
geologic history based upon fieldwork and local knowledge,
however, that one can relate landform characteristics to
geomorphic processes. The ridges (topographic highs in
Figure 1(b)) are actually glacial moraines, formed during the
retreat of the Last Glacial Maximum (LGM). Due to sediment
transportation and deposition from the longshore current, the
bay mouth has shifted from the filled depression adjacent to
Upper Herring Lake (dark red in Figure 1(c)), to Upper
Herring and then to Lower Herring Lake, where geologic, climatic, and biophysical processes continue to modify the
western coast of Michigan. This simple, qualitative assessment
using spatial data typifies many of the spatial problems (and
solutions) in geomorphology, as GIT offers countless perspectives and methods to map and study landforms, although
limitations can exist regarding data, methodological design,
and user experience.
Although a vast array of landform mapping applications
rely on remote sensing, GIS, and GPS (e.g., hazards assessment, glacier mapping, impact craters, aeolian dunes, coastal
189
Figure 1 Western Michigan viewed from different data sources. (a)

Aerial photography from the US National Agriculture Imagery Program
(NAIP). (b) Contoured digital elevation model (DEM) from the U.S.
NED (30 m). (c) Landsat ETM false-color composite (FCC; 4, 3, 2).
structures), this chapter summarizes the use of GIT to map

and assess glacial, volcanic, fluvial, and mass movement
landforms. The integration of GIT in glacial geomorphology
has enhanced the ability to map the surface topography
190
and the ice-velocity fields of glaciers and ice sheets (Smith

et al., 2001; Greenwood and Clark, 2008; Barr and Clark,
2009; Hughes et al., 2010). Spatial information can be further
combined with numerical models to better understand landscape evolution in glaciated terrain (Napieralski et al., 2007a).
Remote sensing and GIS have also improved efforts to map
and distinguish between volcanic lava flows (Kahle et al.,
1988; Wadge et al., 1995), characterize the slope and dynamic
growth of volcanic edifices (Kervyn et al., 2008; Favalli et al.,
2009; Grosse et al., 2009), discriminate between lava flows
and different lithologies, predict rates of erosion, and generate
hazard maps (Marti and Felpeto, 2010). Geospatial technologies are also required to continuously monitor active geological areas characterized by slope stability, as inventory of
landform structures are needed (Owen et al., 2008) to predict
mass movements (Xie et al., 2004; Refice and Capolongo,
2002), risk assessment is required (Davis et al., 2000; Dahal
et al., 2008; Hasegawa et al., 2009), and rapid dissemination
of data and information to scientists and the general public is
critical (Kim et al., 2004; Chang and Park, 2004). Furthermore, dynamic river environments exhibit a wide range of
biophysical characteristics that make them a challenge to map,
but the integration of geospatial technologies in fluvial geomorphology has improved our understanding of channel
migration, sinuosity, and migration (e.g., Winterbottom and
Gilvear, 2000; OConnor et al., 2003; Jordan and Fonstad,
2005), floodplain extent (Marston et al., 1995; Thoma et al.,
2005; Hamilton et al., 2006), stream-bank stability and erosion (Micheli and Kirchner, 2002), sediment entrainment, and
facilitates paleochannel reconstruction (Gurnell et al., 1994;
Gurnell, 1997).
3.8.2
Background
3.8.2.1
Remotely Sensed Data
Remote sensing involves the acquisition and analysis of information about the landscape through the use of sensor
systems that record the magnitude of matterenergy interactions. Biophysical and thematic information can be produced that greatly facilitates geomorphological assessment
and landform mapping. Key landscape information includes
spectral data in the form of aerial photographs, satellite images, ground photography, and videography, as well as key
biophysical information of the near-surface environment, including topographic, surface temperature, albedo, and subsurface information. These sources can be used alone or, more
commonly, in conjunction with one another or with fieldbased data, to enable real-time or comprehensive geomorphological mapping (e.g., Heyman et al., 2008; Hughes
et al., 2010; Karatson et al., 2010b). In addition, GPS receivers
can be used to survey on-the-fly in the field or to remotely
monitor rugged landscapes affected by tectonic forces (e.g.,
Higgitt and Warburton, 1999; Burbank and Anderson, 2001;
Baldi et al., 2008).
3.8.2.1.1
Aerial photographs
The earliest application of remote-sensing methods for landform mapping involved the qualitative (visual) analysis of
aerial photographs (e.g., Prest et al., 1968). Such photographs
Figure 2 Oblique aerial photo (15 October 2010) of lava streams

flowing from the Piton de la Fournaise volcano on Reunion Island.
Aerial photography offers a relative safe vantage point during
volcanic eruptions, floods, and landslides. Note helicopter and
people along right of photograph for scale. Photograph from
Richard Bouhet/AFP/Getty Images.
remain a valuable source to study local-scale landforms,

and are commonly used to augment field mapping, or used
in conjunction with other data (Hall et al., 2009). Aerial
photographs are a reliable data source for detailed mapping
and are frequently used to produce baseline information
for conducting change-detection studies. Visual analyses of
aerial photographs also support assessments of flooding
events, lava flows (Figure 2), and landslide hazards. Largescale information (i.e., high spatial resolution), however, is
counterbalanced by the limited spatial coverage of individual
photographs, limiting their utility for local to regional
mapping. The analysis of large areas (e.g., mountains, glaciated, extensive stream and watershed systems) generally
requires numerous photographs, and is therefore both
time-consuming and expensive (Smith et al., 2006). Nevertheless, digital cameras and video systems are rapidly replacing ground and aerial photography to provide additional
benefits including spectral analysis and quantitative analysis/
modeling.
3.8.2.1.2
Satellite images
Satellite images are available from a multitude of sensors

and can be used for geomorphological mapping, although
their formal utility for specific types of landform mapping
has not yet been formally defined in digital geomorphological
mapping (Bishop et al., 2012). In comparison to aerial
photographs, they have the advantage of covering large areas
(typically 100 100 km), allowing rapid, systematic, and lowcost mapping (Millington and Townshend, 1987; Walsh et al.,
1998; Smith and Pain, 2009). They are ideal for mapping large
areas at a wide-range of scales, which can elucidate landforms
(commonly large-scale features) that may be unidentifiable
using traditional field methods or aerial photographs
(Figure 3). Some satellite data are now easily obtained from
on-line sources (e.g., http://www.landcover.org).
The potential of using satellite images for geomorphological mapping was first realized through the initiation of the
Landsat program in the 1970s, as Landsat images remain one
of the most commonly used sources for mapping geomorphological features (Figure 4). Various platforms have
been launched since the inaugural Landsat mission and sensors vary with respect to their spatial, spectral, and radiometric
resolutions. Such parameters determine the utility of imagery
for mapping landforms (Table 1). Satellite images typically
have a spatial resolution from 5120 m, with some imagery
comparable to aerial photographs (e.g., IKONOS, Quickbird,
WorldView, and GeoEye), though coverage and cost often
limit their use for detailed landform mapping. Image characteristics have different merits and limitations, so to ensure
comprehensive mapping, more than one type of imagery is
commonly used within a single study (i.e., multisensor
imagery).
Figure 3 Advanced Spaceborne Thermal Emission and Reflection

Radiometer (ASTER) False-color composite image showing the linear
dunes in Rub al Khali, or the Empty Quarter, of Saudi Arabi. The
dunes appear yellow due to the presence of iron oxide minerals,
whereas the inter-dune areas are comprised of clays and silts, which
appear bright due to a high reflectance in ASTERs band 1. Image
covers an area of 37 km by 28 km and is from NASA/GSFC/METI/
ERSDAC/JAROS, and U.S./Japan ASTER Science Team (http://
asterweb.jpl.nasa.gov).
3.8.2.1.3
191
The Earths topography is optimally represented using digital

elevation models (DEMs), that are derived from a variety of
data sources, including terrestrial surveying, laser scanning,
topographic maps, interpretations from stereographic imagery,
and radar measurements (Raseman et al., 2004). Morphological information (e.g., slope, slope azimuth, curvature) is
fundamental for identifying and mapping landforms, and
many new terrain analysis approaches permit unique characterization of the topography. Consequently, geomorphometry
is considered by some to be superior to satellite image interpretation and analysis (e.g., Clark, 1997), although surface
compositional information is best obtained from spectral
analysis. Such assessments are the result of the need for inferential interpretation strategies due to the disconnect between land-cover patterns and landforms, and our inability to
effectively utilize biophysical information to describe landform compositional variations in some environments (Bishop
et al., 2012). Their use has flourished partly because of the
development of new methods of analyzing the elevation of the
Earths surface; in fact, the increased availability and accessibility of DEMs has perhaps been the greatest advance in data
collection within the field of mountain geomorphology over
the past decade. With world-wide coverage and rapidly improving measurement scales, the production, analysis, and
visualization of DEMs have become a standard practice in
landform mapping.
In general, geomorphologists use DEMs to analyze continuous land-surface morphometric conditions (i.e., land-surface
parameters) or partition landscapes into discrete surface features
(Pike et al., 2009; i.e., land-surface object). This quantitative
characterization of the topography is referred to as geomorphometry, a multidisciplinary field of study that has produced
an immense amount of data and information regarding
processes, landform morphometry, and landscape evolution
(Figure 5). The very rapid development of sensors, GIS, and new
advanced algorithms to characterize land-surface parameters and
objects complicates landform research, but also offers endless
opportunities to merge knowledge from different domains
(e.g., software, mathematics, engineering, geomorphology) to
study landscapes in a more scientific manner.
Figure 4 Mount McKinley (Alaska) viewed using a Landsat ETM false-color composites (FCC) using bands 5, 4, 3 (a) and 4, 3, 2 (b). Ice
surface area, crevasses, lateral moraines, and valley characteristics are easily detected from FCC images of glaciated terrain.
192
Table 1 Summary of available satellite sensors used in landform mapping

Satellite
Sensor
Bands
Spatial resolution (m)
Spectral range (mm)
Landsat 13
Landsat 4, 5
MSS
TM
Landsat 7
ETM
EO-1
Hyperion
Ali
SPOT 13
HRV
SPOT 4
HRVIR
SPOT 5
HRG
IKONOS
Terra
Panchromatic
MSS
Panchromatic
MSS
ASTER
Geoeye
HRG
Envisat
ERSI/2
TerraSAR-X
ASAR
4
6
1
1
6
1
220
1
9
1
3
1
4
1
3
1
4
1
4
3
6
5
1
4
2
1
1
80
30
120
15
30
60
30
10
30
10
20
10
20
5, 2.5
10
1
4
0.61
2.44
15
30
90
0.4
1.65
30
30
1,3,18
VNIR
VNIR, SWIR
TIR
VNIR
VNIR, SWIR
TIR
Hyperspectral
Panchromatic
VNIR
VNIR
VNIR
VIS
VNIR, SWIR
VIS, VNIR
VNIR
VNIR
VNIR
VNIR
VNIR
VNIR
SWIR
TIR
Panchromatic
VNIR
Microwave
Microwave
Microwave
Quickbird
geophysical data into this category, as gravity, aeromagnetic,

electrical, and seismic information can provide valuable information about near-surface composition and structure.
These datasets may be compared to mapping results using
spatial overlay techniques to ensure that each feature has been
interpreted correctly with respect to identification and
delineation.
3.8.2.2
Figure 5 Negative-openness image generated from Shuttle Radar

Topography Mission (SRTM) elevation data for the Karakoram
Himalaya that highlights the hydrological network. Terrain openness
is an angular measure of the relation between surface relief and
horizontal distance (in negative openness, bright areas correspond to
valley floors).
3.8.2.1.4
Supplementary data
When generating landform maps from aerial photographs,

satellite images, and DEMs, it is generally necessary to consider additional data sources to provide some means of
evaluating mapping accuracy (e.g., Greenwood and Clark,
2008; Hughes et al., 2010). Such supplementary data might
include topographic maps and Quaternary geology maps that
distinguish sediment types from exposed bedrock, solid, and
structural geological datasets, permafrost maps, and landcover and land-use data. It is also important to include
GIScience
Geographic information science refers to the conceptual and

technological issues associated with the representation, analysis, modeling, and visualization of space-time data and
information (Goodchild, 1992a, b; Bishop and Shroder, 2004;
Bishop et al., 2012). In this context, GIScience represents
much more than just a cartographic tool to map landforms
(Wright et al., 1997). Rather, it represents an interdisciplinary
field of study focused on the development of the philosophical, cognitive, and scientific treatments of spatial theory
and concepts that can be used to address information scienceand Earth- science-related problems (Bishop and Shroder,
2004). These objectives align with many key concepts and
issues in geomorphology and landform mapping, including
space-time and multidimensional representation, preservation
of spatial properties and relationships (i.e., topology), and the
development of empirical and predictive models (Napieralski
et al., 2007a, b). It is not unreasonable to assume that the
theoretical developments occurring in the rapidly growing
field of GIScience will contribute to a more efficient and integrated treatment of spatial theory, advance data assimilation,
and analytical capabilities, and facilitate new approaches to
visualization that will permit more scientific treatments of
geomorphological mapping, as described in Bishop et al.
(2012).
Conducting terrain analyses and delineating landforms
into spatial entities is not without challenges, especially when
relying on GIT. The demarcation of a landform boundary
should theoretically rely on the principle of homogeneity of
spatial entities, which states: borders are expressions of
maximizing internal homogeneity while maximizing external
heterogeneity (Eisank et al., 2010). These circumstances,
however, are not always realistic, as there are a number of
factors that influence the capability to identify landform
boundaries, including topographic relief (i.e., landforms tend
to be easier to extract with higher topographic relief), landform properties (e.g., size, shape, relative relief), scale, differences between landform orientation and illumination
orientation associated with remote-sensing-derived data (i.e.,
azimuth bias), and the application of knowledge-based classification rules (Smith et al., 2001; Eisank et al., 2010). In fact,
landform boundaries are not characteristically discrete; rather
they are transitional. So overcoming these limitations require
an established, universally accepted set of classification rules,
in addition to an interdisciplinary integration of spatial information designed to approach landform mapping from a
rigorous, but also realistic, perspective. As an example, fuzzy
logic can be used to represent landform boundaries as spatial
fuzzy uncertainty approximations, rather than discrete, exact
boundaries. Many fuzzy techniques have recently contributed
to landform segmentation and classification (Burrough et al.,
2000; MacMillian et al., 2000; Fisher et al., 2004; Mark and
Smith, 2004; Moody and Katz, 2004). They are considered
effective for mapping landforms that have transitional
boundaries. Conversely, object-orientated analysis involves
characterizing each landform as a discrete component in a
complex system, and this is ideal for investigations related to
process mechanics and space-time representation (Dragut- and
Blaschke, 2006; Dragut- et al., 2009; Anders et al., 2009;
193
Ghosh et al., 2010; Eisank et al., 2010), although fuzzy

boundaries could be represented as a discrete object. The
framework for characterizing and measuring landforms continues to evolve, but it progresses faster when geomorphologists integrate principles and concepts that are central to
the underpinnings of GIScience theory.
The ability to resolve landform entities and understand
landscape evolution is largely controlled by operational scales
(i.e., spatio-temporal scales at which surface processes operate). Spatial scales range from picorelief forms (e.g., glacial
striations) that measure millimeters in length and width to
megarelief (e.g., continental) features (Table 2; Dikau, 1989).
The spatial scale-dependency of landform entities within a
landscape signifies the presence of nested-form assemblages
that are organized in a spatial hierarchy (Raseman et al.,
2004). Therefore, scale-dependent analysis is vital in determining landform boundaries (based on the principle of
homogeneity). Software tools, such as the Estimation of Scale
Parameters (ESP), are designed specifically for finding the
appropriate segmentation scales (Dragut- et al., 2009). Further
complicating landform identification and extraction is the fact
that landforms evolve over varying lengths of time, which
frequently leaves behind a palimpset (i.e., smaller landforms
superimposed on larger features, younger modifying older)
landscape (Chorley et al., 1984; Brunsden, 1993; Raseman
et al., 2004). Therefore, the spatio-temporal characterization
of landforms require objective analysis with repeatable
results (Table 2). This can potentially be accomplished using
advances in GIScience and well-defined taxonomic schemes
developed by geomorphologists to reduce subjectivity and
facilitate landform classification.
The importance of understanding landform patterns is reflected in the wide interest in segmenting landscapes into distinct units. Landform patterns influence the flow of sediment,
nutrients, and water (e.g., Dalrymple et al., 1968), create corridors that impact biodiversity and environmental quality
(Hunsaker et al., 1992; Gordon et al., 1994; Lopez-Blanco and
Villers-Ruiz, 1995; ONeill et al., 1997; Butler, 2001), and limit
the spatial distribution of geologic resources. Of interest to
geomorphologists is the linkage between landform patterns and
Table 2 A classification of geomorphological features by spatio-temporal characteristics

Order
0
1
2
3
4
5
6
7
8
9
10
11
Classification
Megarelief
Macrorelief
Mesorelief
Microrelief
Nanorelief
Picorelief
Area (km2)
Time (yr)
8
5.098 10
107108
107105
105103
103102
102101
101101
101103
103105
105107
107109
1091012
5 10
109108
108
108107
107
107106
106105
105104
104103
103102
o102
Landforms
Lithosphere, cryosphere, hydrosphere, atmosphere, ecosphere
Continents, ocean basins, tectonic plates
Physiographic provinces, shields, large volcanoes, mountain ranges
Medium-scale tectonic units, mountain massifs, fault blocks, grabens
Small-scale tectonic units, fault blocks, sackung
Large-scale erosional/depositional landforms, major valleys, piedmonts, deltas, landslides
Medium-scale erosional/depositional landforms, cirques, moraines, floodplains, alluvial fans
Small-scale erosion/depositional landforms, ridges, terraces, dunes, slump blocks, talus
Larger geomorphic process units, hillslopes, stream channel reaches, talus, small debris flows
Medium geomorphic process units, riffles/pools, river bars, slope facets, solution pits, gullies
Smaller geomorphic process units, ripple marks, glacial striae, rills, raindrop impact pits
Clast grain morphologies, clay mineral structure
Source: Modified from Shroder, Jr. J.F., Bishop, M.P., 2004. Mountain geomorphic systems. In: Bishop, M.P., Shroder, J.F. (Eds.), Geographic Information Science and Mountain
Geomorphology. Springer-Praxis, Berlin Heidelberg, pp. 3366, and Raseman, S., Schimdt, J., Schrott, L., Dikau, R., 2004. Geomorphometry in mountain terrain. In: Bishop, M.P.,
Shroder, J.F. (Eds.), Geographic Information Science and Mountain Geomorphology. Springer-Praxis, Berlin, pp. 101137.
194
process mechanisms or landscape evolution. Pattern recognition techniques range from more tedious manual and semiautomated approaches (digitizing boundaries of images before
comparing properties to neighboring features), to more advanced methods that are fundamental in computer science. In
particular, artificial intelligence (AI) has improved the decisionmaking process for detecting patterns and spatial correlations
between landforms. The process by which landforms are
identified or patterns recognized (i.e., decision rules) in a virtual system can be designed to simulate the process used by the
human brain (i.e., artificial neural networks). Input data may
include landform geometry, size, texture, curvature, or even
spatial topological characteristics. Thus the semantic modeling
of landforms (i.e., the relationship between computer representations of landforms and real-world features) requires
comprehensive formalization of landform properties to enable
landform classification (Bishr, 1998; Eisank et al., 2010). For
example, Eisank et al. (2010) provide five steps used for semantics-based landform classification: (1) identification of
landforms observable in DEM; (2) conceptualization of landforms based on geomorphological knowledge; (3) scale detection to define the threshold of the maximum increase in
heterogeneity (recall the principle of homogeneity of spatial
entities); (4) semantic modeling, which is how the geomorphic
concepts are represented within the computer; and (5) software
implementation, such as the use of object-based image classification (OBIA).
Once landform morphology and pattern are described and
understood, these results are commonly linked to surface processes and feedback mechanisms to gain knowledge of landscape evolution. The physical characteristics of landforms can be
classified according to the dominant surface processes (Volker
et al., 2007; Perron et al., 2008). For example, Volker et al.
(2007) classified alluvial landforms as either dry debris flows, or
fluvial (wet) sediment transport based on local-relief, proving
that higher local-relief most likely resulted from a dry debris
flow. These results can be used to link topographic features
(i.e., fingerprints) to dominant surface process. Finally, geomorphologists have evaluated the factors that potentially control landscape evolution (Bishop et al., 2003; Walcott and
Summerfield, 2008). As expected, landscapes (e.g., mountain,
hillslopes) are formed by scale-dependent interactions between
surface processes, climate, and tectonics. Eliciting these interactions is key to understanding some of the most classic problems pursued by geomorphologists, including quantifying the
driving and resisting forces behind topographic production and
modification over varying temporal and spatial scales.
Finally, Walsh et al. (1998), Bishop and Shroder (2004),
and Bishop et al. (2012) summarized several GIScience issues
and geospatial concepts relevant to landform and geomorphic
studies that include:
Viewing landscapes through an array of spatial, spectral, and

temporal perspectives.
Linkage
to causal and multivariate models and pattern

indices.
Efficient display and dissemination of data and maps (e.g.,
cartographic modeling).
3.8.2.2.1
Mapping approaches
Traditionally, landform mapping required intensive fieldwork,

qualitative interpretations of aerial photography and topographic maps, or quantitative assessments using photogrammetric techniques. For example, some early landform
mapping techniques relied on extracting and measuring
landforms using contour lines off topographic maps, frequently to indicate the boundary of a landform (e.g., lowest
enclosed contour indicated boundary of glacial landform).
However, this was time-intensive, limited to map scale and the
horizontal and vertical error, and was subjective and relatively
inconsistent. The rapid proliferation of GIT has created an
opportunity to integrate data from various sources, to conduct
objective, reproducible studies, and to increase the efficiency
of landform identification, extraction, and measurement.
Clark and Wilson (1994) noted that automated techniques
associated with computer programming and GIS could drastically increase the efficiency of landform mapping relative to
traditional methods (e.g., they could measure 4800 features
using automated methods, and acknowledged that in the
same time, only 501600 might be measured using traditional
techniques).
Landforms can be mapped using either field or entity data
models. Raster-based DEMs can be used to generate landsurface parameters, and topographic information can be
overlayed with other raster layers to assign raster cells to a
particular landform class. In contrast, landforms can be
mapped using points (indicating landform location), lines
(the location and orientation of linear or nonlinear landforms), or polygons (demarking the landform boundaries that
permit efficient calculations of area or perimeter). The type of
data structure used to represent a landscape or individual
landform units is influenced by a desire to have boundarybased representation (i.e., vector) or accept topographic
gradation (Plewe, 1997). This ontological predicament is
critical in geomorphology. As an example, Mark and Smith
(2004) noted that mountain peaks are typically marked at the
highest point, but rarely are they illustrated with boundaries
due to the challenge of identifying where mountains and
peaks begin and end. It is clear that spatial scale and
perception (e.g., structural geologist vs. geomorphologist) influence how and where landforms are delimited (Deng, 2007),
and one way around this problem is to develop multiscale,
quantitative rules for landform identification. As an example,
Lucieer and Stein (2005) defined mountain peak memberships by characterizing multiscale topographic variance, leading to the definition of peak prototypes and multiscale peak
memberships (Deng, 2007).
Issues of scale related to surface processes, feedback mechanisms, system dynamics, and landform features.
Exploration of data through a variety of attribute and geo-
3.8.3
graphic operations.
Conversion of data to information, including thematic
layers and composites.
Glacial Landscapes and Landforms
Investigations of glacier and ice-sheet extent, volume, and

behavior frequently require spatial data covering large areas,
high latitudes, or high altitudes. Additionally, the surface

topography of glaciers adjusts relative to the glacial mass
balance, which can occur rapidly in some situations, thus requiring the acquisition of spatial data on a relatively short
time interval. As a result, the acquisition and dissemination of
spatio-temporal glacial data is central to the productivity and
effectiveness of local and global glacial surface and landform
studies. As an example, the Global Land Ice Measurements
from Space (GLIMS) project (Bishop and Shroder, 2004;
Kargel et al., 2005) involves assessing and mapping the Earths
glaciers and provides access to spatio-temporal data through
either a web-based GIS server or a traditional text-based
interface that allows for more advanced searches. Organizations such as GLIMS increase interdisciplinary collaborations
to monitor and map glacier surface characteristics, but they
also reduce discrepancies in data format, minimize geometric
distortion between datasets, and augment the development of
universally accepted taxonomic schemes.
Mapping glaciers and glacial landforms is one of the key
approaches to reconstructing the extent and dynamics of former ice masses (Andrews, 1982), and has been the focus of
research for well over a century (Close, 1867; Charlesworth,
1928). Studies have traditionally involved field-based analyses
of landforms such as glaciers, drumlins, moraines, and eskers,
but are now commonly based on remote-sensing studies
(Boulton and Clark, 1990; Bishop and Shroder, 2004; Smith
et al., 2006). In fact, remote sensing investigations have been
so successful that some consider them to have largely superseded the traditional (field-based) approach (e.g., Smith
et al., 2006). Remote sensing has been particularly instrumental in acquiring large inventories of glacial landforms to
reconstruct ice-sheet extent and regional flow characteristics
in the British Isles (Clark and Meehan, 2001; Jansson and
Glasser, 2005; Bradwell et al., 2008), North America (Boulton
and Clark, 1990; Clark et al., 2000; Jansson et al., 2002;
Kleman et al., 2002; Stokes and Clark, 2003), and Scandinavia
(Punkari, 1993; Kleman and Borgstrom, 1996; Kleman et al.,
1997; Hattestrand and Stroeven, 2002).
Some of the limitations inherent to the field-based approaches are that it is often expensive, time-consuming, and
the data produced generally relate to small areas leading to
difficulties with assimilating information to derive a countryor region-wide assessment or overview (see Clark et al., 2004;
Evans et al., 2005). The remote-sensing approach, by contrast,
provides a time- and cost-effective way for a single user or user
group to systematically map large areas, some of which are
remote and inhospitable, at a range of scales, while maintaining consistency in mapping style (Punkari, 1982; Clark,
1997). It is then possible to collate these data in a GIS together
with other spatial data (e.g., soils, geological maps). The
synoptic coverage of remote-sensing data also reveals some
landforms (e.g., megascale glacial lineations (MSGLs)) and
patterns that are not easily identifiable by traditional means
(Clark, 1993, 1997).
The above advantages clearly make remote sensing critical
for mapping glacial geomorphology, and one that has developed rapidly over recent years. The intention here is to focus
upon the data sources and data processing and visualization
methods most widely used when employing a remote sensing
and GIS approach to the mapping of glacial landforms.
3.8.3.1
195
Glacial Landforms
Mapping landforms with the intention of understanding

glacierization and glaciation relies upon the identification and
interpretation of field evidence indicating ice geometry
and thickness. Ice-marginal landforms such as terminal, end,
and lateral moraines are commonly used to infer the horizontal dimensions of former ice masses, though the extent of
till may also be used for this task. The vertical dimensions
of former ice masses are generally more difficult to constrain
(see Ng et al., 2010), though trimlines, where identifiable
(and this is characteristically difficult), are used in some cases,
but not unambiguously (e.g., Ballantyne, 1997). Bedforms
(i.e., features formed beneath an ice mass) such as drumlins,
crag-and-tails, flutes, and ribbed moraines are used as indicators of former ice-flow direction and velocity, and also
provide information on former ice mass extent and evolution.
Robust paleo-ice sheet reconstructions therefore require a GIS
to integrate and map cross-cutting sets of bedforms of varying
scales and from a wide range of sources before any interpretations can be considered.
3.8.3.2
Data Processing and Visualization
Techniques for processing and visualizing remotely sensed

datasets in order to maximize their effectiveness for mapping
glacial landforms have been reviewed by Clark (1997) and
Smith and Clark (2005). These relatively simple techniques
are generally implemented in a GIS and, when judiciously
applied, can increase the detectability of glacial landforms
significantly (Smith et al., 2000; Jansson and Glasser, 2005).
3.8.3.2.1
Aerial photographs
Methods of mapping glacial landforms from aerial photographs have varied very little since the first use of these
data (Woodruff and Evenden, 1962; Aniya and Welch, 1981;
Lindner et al., 1985), and a qualitative approach, whereby
features are manually mapped either in a GIS or directly onto
aerial photographs, remains widely used. Relative to satellite
imagery, the revisit period, swath width, spatial resolution,
and variable flying altitude associated with aerial photography
is more flexible and allows for a customized perspective of
small, localized glaciers, and glacial landscapes (Gao and Liu,
2001). Despite the advantages of using aerial photography,
there are still few established rules for interpreting glacial
landforms and features, especially since some landforms may
be extracted from physical copies of the aerial photo (or in
combination with a topographic map), whereas others may
use on-screen digitizing with a GIS (Hubbard and Glasser,
2005). Additionally, interpretations derived from analog
photographs (e.g., scanning, then digitizing) can be laborious
and the issue of distortion in high relief areas is far more
prevalent with aerial photography than satellite imagery
(Sidjak and Wheate, 1999).
Identifying glacial landforms in such photographs depends
upon obtaining largely cloud-free, high-quality images, acquired during periods of low solar elevation, as many glacial
features (e.g., drumlins) are identified by a break-of-slope,
which is easier to identify with relatively high solar zenith
angles (Clark, 1997). Glacial landforms mapped and extracted
196
from mosaics of aerial photos lend support to paleoglaciological reconstructions of basal conditions, flow direction, and
extent, and include erosion (e.g., polished bedrock, melt-water
channels, cirques) and depositional (e.g., moraines, glaciofluvial deposits) features (Sahlin and Glasser, 2008). A common approach to managing such large assemblages of data is
to segregate the landform data into flow sets that act as
building blocks in the reconstruction of previous ice sheets
and glaciers.
Although some of the advantages of using aerial photography in glacial studies discussed earlier have diminished due
to improved satellite sensor capabilities, aerial photos are
valuable when monitoring and mapping glacier extent and
glacial landform distribution. Fluctuations in glacier extent
(e.g., Field, 1966; Meier, 1966; Petrie and Price, 1966; Espizua,
1986), surface motion (Evans, 2000), and crevasse migration
(Haefeli, 1966; Harper et al., 1998; Aniya et al., 2002) have
relied on repeat aerial photography due to relatively high
spatial and temporal resolutions. Aniya and Welch (1981)
utilized photogrammetric techniques to analyze glacial cirques, such that the crest line and lowermost limit, gradient,
aspect, and shape of cirques were compiled and compared
between glaciated regions to link the topographic, climatic,
and tectonic conditions that influence the evolution of glacial
cirques (Federici and Spagnolo, 2004). Mosaics of aerial
photos also have offered confirmation or rejection of interpretations derived from satellite data. Clark (1993) compiled
glacial lineation data from over 75 000 aerial photographs and
Landsat MSS images to analyze the spatial pattern of glacial
landforms (i.e., flutes, drumlins, megaflutes) to assess possibilities for subglacial landform evolution.
3.8.3.2.2
Satellite images
As with aerial photographs, satellite images should be cloudfree, and acquired with comparatively low solar elevations,
ideally below 201 (Smith et al., 2006). In some regions, this
necessitates obtaining images with a winter acquisition date,
because sensors such as Landsat always acquire data for a
given region at a particular time of day. Obtaining images with
low solar elevations is easier in the high-Arctic, as any images
with a summer acquisition date are likely to meet this criterion, but at lower latitudes obtaining cloud-free data captured during periods of low solar elevation is difficult (partly
because satellite images cover large areas), and concessions
must commonly be made (e.g., using inferior data). Another
approach employed to enhance glacial landform identification
from satellite images is to view data acquired during a period
of uniform snow cover (e.g., Eyton, 1989; Skoye and Eyton,
1992). As emphasized by Clark (1997), however, this is an
approach rarely and opportunistically applied, as cloud-free
images, with uniform snow-cover, but which also have sufficient illumination levels, are scarce.
Once suitable satellite images have been acquired, processing prior to visualization is a comparatively straightforward task. The approach advocated by Clark (1997) is to
select a subset of an image, covering a region where the distribution of glacial landforms is well known, then experiment
with image-enhancement techniques, such as contrast
stretching and histogram equalization, until the most useful
enhancements for landform detectability are found. These
techniques can then be applied to the remainder of the satellite image if the global probability distribution does not
adversely affect the results. Some have also applied high-pass
filtering to further enhance landform boundary clarity, but
with varying degrees of success (Jansson and Glasser, 2005),
as this can also produce image artifacts.
The most widely used satellite images for glacial landform
mapping, Landsat 7 ETM , come in a variety of spectral
bands, ranging in wavelength from 0.45 mm (visible blue) to
12.50 mm (thermal infrared). These images are typically
viewed as false-color composites (FCCs) using a variety of
band combinations (Jansson and Glasser, 2005), or as hybrid
(pansharpened) images with information in a color composite
transferred onto the higher resolution panchromatic image
(e.g., Greenwood and Clark, 2008; Heyman et al., 2008).
The optimum combination of bands within a FCC varies
with season and location, but as vegetation (type and quantity) is generally an important factor in distinguishing glacial
ice and landforms from the surrounding landscape, ETM
band 4 is generally used, as this is the near-infrared waveband
which allows differences in vegetation to be best detected, and
has limited atmospheric effects (Figure 6; Clark, 1997).
Using Landsat 7 ETM images, Jansson and Glasser
(2005) compared the panchromatic band 8, different FCC
combinations (RGB bands 4-3-2, 7-4-2, and 4-5-6), and the
Ordnance Survey Panoramas digital terrain model (DTM,
50 m spatial resolution) to assess which is best suited to the
compilation of regional-scale maps of glacial lineaments in
Wales and similar environments. They displayed all FCC images, and establish band combinations on initial experiments
to determine the suitability of different combinations. They
found that only 9% of lineaments within their study area were
identified when the panchromatic band alone was considered;
25% were identified when the FCC combination 4-3-2 alone
was considered; 22% with the FCC combination 7-4-2 alone;
36% with the FCC combination 4-5-6 alone; and 29% when
the DEM alone was considered. The number of features
identified increased significantly when FCC images were
overlaid on the DEM, with the FCC combination 4-5-6
proving the most effective. On this basis, Jansson and Glasser
(2005) recommended that to obtain the best possible results
when mapping glacial lineaments in this type of environment,
a FCC Landsat 7 ETM combination 4-3-2 should be draped
over a DEM. In addition, they suggested viewing a non DEMdraped FCC combination 4-5-6. This work demonstrates the
value of image enhancement, data fusion, and visualization
for mapping, as it has been found to be effective elsewhere
(Greenwood and Clark, 2008).
Satellite imagery permits the calculation of surface area
(Aniya et al., 1996; Sidjak and Wheate, 1999), measurements
of surface motion (Bindschadler and Scambos, 1991; Scambos
et al., 1992; Whillans and Tsend, 1995), and ice-snow surface
classifications (Bronge and Bronge, 1999). Image processing
techniques such as principal component analysis (PCA) and
the normalized difference snow index, reduce data redundancy, and increase capabilities to distinguish changes in ice
surface characteristics, respectively. Sidjak and Wheate (1999)
analyzed Landsat TM scenes of the Illecillewaet icefield
(British Columbia) using PCA, image ratioing, and image
differencing to produce a land-cover classification map
(a)
17700E
17800E
6600N
17800E
(b)
6500N
6500N
6600N
17700E
197
17700E
17800E
17700E
17800E
Figure 6 Moraines in the Anadyr Lowland of Eastern Siberia, depicted (a) in a Landsat ETM false-color composite image (FCC; 5, 4, and 2),
and (b) using ASTER GDEM data. This is a region outside that covered by SRTM data.
separating the icefield into subsets of glaciated and nonglaciated terrain. It was also ascertained that the lower order
components of the analysis likely indicated ice-surface topography. The assessment of sequential satellite images can also
reveal the ice-flow velocity fields, especially with surging ice or
active alpine glaciers (e.g., Scambos and Bindschadler, 1993).
3.8.3.2.3
When mapping using topographic information, the user

generally has little choice regarding acquisition date or conditions. This is less important than with aerial photographs or
satellite images, however, as GIS-based simulations of solar
geometry can be used to produce shaded-relief maps that
highlight the topographic structure. For example, it has been
found that such illumination variations improve feature
detectability (see Clark and Meehan, 2001; Smith and Clark,
2005). More specifically, varying the solar azimuth can highlight features that might otherwise remain undetected (e.g.,
Clark and Meehan, 2001; Greenwood and Clark, 2008), with
low (o201) solar elevations typically maximizing detectability
(Smith et al., 2001; Jansson and Glasser, 2005). Thus, DEMs
have been successfully used to visualize glacial landforms
based upon relief shading, with solar illumination geometry
defined by the user (e.g., Pike, 1992). Other visualization
methods include local contrast stretching and enhancing the
vertical exaggeration of altitudes, as well as examining slope
gradient and various curvature metric variations (Clark and
Meehan, 2001; Smith and Clark, 2005; Smith et al., 2006).
Smith and Clark (2005) analyzed 12 separate DEM visualization techniques and found no single method to allow
complete and unbiased mapping of glacial lineaments. They
recommended that initial mapping should be conducted by
viewing relief-shaded visualizations illuminated from two

different azimuths (preferably orthogonal to one another). In
fact, in order to minimize illumination bias some have used
images illuminated from three different locations, commonly
including one with the sun directly overhead (e.g., Hughes
et al., 2010).
There are numerous DEM analyses that would be of
interest to glacial geomorphologists, including some that
have been developed for other purposes, but modified specifically for use in glacial landform mapping. For example,
Duncan et al. (1998) modified drainage extraction techniques to estimate modern and historic glacier extents in
central Nepal. They computed hillslope angles to generate
links to define network channels and determine morphologic parameters of valley form (e.g., distinguishing between
U- and V-shaped valleys using flatness of valley floor or
curvedness of valley walls). This was estimated in part using
the flow accumulation scheme frequently used to delineate
watersheds and stream networks and to simulate runoff. In
other circumstances, DEMs contribute to landform recognition and identification, even if it is a reevaluation of
previously identified features. For example, Moraines have
been historically mapped with some subjectivity and combining high-resolution DEMs and supplemental data can
reduce the amount of bias in detecting moraine crests versus
moraine outlines (e.g., Clark et al., 2004). Morphological
features can also be highlighted using terrain exaggeration
of shaded relief derived from LiDAR, which filters out
vegetation to produce a high-resolution DEM (Salcher
et al., 2010). Mapping minor topographic variations in alpine environments using LiDAR produces a more complete
picture of landform morphology, which can potentially be
used to generate information regarding landform genesis
198
(e.g., sequence of proglacial outwash plains subsequently

covered by ice due to glacial advances).
Mapping the topographic characteristics of glacier surfaces also necessitates DEMs derived from satellites with
worldwide coverage, or those designed specifically for ice
surface characterization. The Ice, Cloud, and Land Elevation
Satellite (ICESat) (National Aeronautics and Space Administration NASA) and CryoSat (European Space Agency ESA)
both provide radar altimetry data of Earths polar and glaciated regions that can be used to map ice sheet topography
(Zwally et al., 2002; Drinkwater et al., 2003). Changes in landsurface elevation reflect changes in glacial mass balance conditions, and multitemporal DEM analysis facilitates estimates
of regional mass balance. Data from ICESat, CryoSat, ASTER,
and the Shuttle Radar Topography Mission (SRTM) can be
used. For example, Sauber et al. (2005) correlated changes in
surface ice elevation with rates of ice thinning for the
Malaspina Glacier, Alaska using ICESat and SRTM data. Furthermore, multitemporal synthetic-aperture radar (SAR)
interferometry techniques are also used to measure surface iceflow velocities (Goldstein et al., 1993; Joughin et al.,
1995,1996; Bamber et al., 2000; Li et al., 2008a, b; Koehl
et al., 2009), and even estimate three-dimensional (3D) velocity fields (Joughin et al., 1998; Mohr et al., 1998). In both
approaches, data representing surface elevation and ice
thickness are vital to the production of accurate surface iceflow velocity maps (Hgda et al., 2010). Even airborne laser
scanning can produce surface elevation approximations and
velocity fields, detect crevasses, distinguish between glacier
surface types, and map glacier surface roughness to use as
input for surface energy-balance modeling (e.g., Geist and
Sttter, 2010).
The influence of DEM grid-cell size on glacial landform
mapping accuracy has recently been tested. Using statistical
process control charts (SPCC), Napieralski and Nalepa (2010)
found an optimum grid cell size to delineate and measure
drumlin length, width, and orientation (and thus area,
elongation, Rose curve). Although the drumlin sample size
was relatively small, the results indicated that the use of a
30 m resolution DEMs caused a significant loss in drumlin
count, size, and shape (Figure 7). The authors warned, however, that higher resolution does not necessarily equate to
increased accuracy, as there was no statistically significant
difference between morphological characteristics calculated
from a 10 m versus a 1 m DEM. Therefore, the combination of
DEM resolution and landform size and shape represents a
mapping issue.
3D perspectives represent another way of visualizing image
and topographic information. Visualizations based on elevation allow the topographic-context of landforms to be better
accounted for, although direct digitizing from oblique views is
difficult, and the process does not lend itself to rapid mapping
(Smith and Clark, 2005). Moren et al. (2011) used Google
TM
Earth to visualize 3D glacial landforms (i.e., valleys, marginal moraines, lineations, and hummocky terrain) generated
from a SRTM DEM and supplemental data (Figure 8). Their
ability to distinguish between glaciation and nonglaciated
terrain around the periphery of the Tibetan Plateau contributes to the belief that glaciers have had limited extent, and
thus influence, on landscape evolution there.
3.8.3.3
GIS-based Mapping
The mapping of glacial landforms from remotely sensed

sources typically involves on-screen digitization of features
using a digital-image processing system or a GIS. Features are
usually mapped as points, lines or polygons depending
upon the feature type and the scale of the resulting map
relative to the size of the mapped features (Smith et al., 2001;
Greenwood and Clark, 2008). Emphasis is commonly placed
on mapping the crest-lines of individual landforms or outlining the break-of-slope around their margins (e.g., Hughes
et al., 2010). Mapping is conducted at a range of scales, again
depending upon the data- and landform-type in question.
A repeat-pass approach is commonly adopted, whereby each
region is viewed on numerous occasions at a range of scales,
preferably using various data sources (Greenwood and Clark,
2008; Hughes et al., 2010). Systematic and comprehensive
mapping may be performed in a comparatively rapid and
cost-effective manner. This approach, however, depends to a
large extent on the observers ability and experience (Clark,
1997; Smith et al. 2001), representing a subjective and timeconsuming approach.
GIS-based research focuses on the development and
evaluation of automated, or semiautomated techniques.
Automated techniques, including pattern-recognition approaches, have been used to extract geomorphological information directly from remotely sensed data and DEMs, but
despite improvements in the sophistication of pattern-recognition algorithms and performance, these methods do not
formalize landform classification and are not universally applicable, as results typically depended upon data selection and
the generation of an appropriate feature space, given the
landforms of interest. Consequently, reliance on the humanvisualization system and digitization continue to be favored.
Landform patterns are generally distinctive in previously
glaciation terrain, as analogous landforms tend to cluster as a
result of topography, subglacial hydrology and thermal conditions, and sedimentological characteristics. Similar landforms are commonly grouped according to morphological
criteria (e.g., size, orientation, shape indices) and these landform sets typically reflect subglacial conditions during formation, acting as building blocks for inverse reconstructions
of previous glaciations (Figure 9). Many of these landform
sets contain digitized landforms that display distinct patterns,
such that their relative alignment within the set (i.e., parallel
conformity), the number of landforms per unit area (landform density), and surface area of landforms per unit area (i.e.,
packing) can be calculated and established as criteria for
identifying similar landforms elsewhere (Stokes and Clark,
2003). Dunlop and Clark (2006) digitized over 33 000 landforms with the objective of developing a clear definition of a
ribbed moraine using morphological characteristics (i.e., ridge
width, length, height, wavelength, shape), moraine accordance (e.g., ridge summit heights within field, jigsaw matching), and position of ridges relative to ice-sheet margin or
divide and topography. In particular, they repositioned a series
of digitized ridges (polygons) to test the theory that the
landforms were once connected, and subsequently split due to
overlying ice motion (Figure 10). The capability to move and
rotate polygons is efficiently accomplished using GIS and
80 m
50 m
30 m
20 m
10 m
5m
199
1m
500
Meters
Figure 7 Drumlin shape and size are influenced by the resolution of a digital elevation model. Note only subtle changes occur between 1 and
20 m, but larger grid cells then influence landform characteristics at 30 m (e.g., Landsat TM) and coarser. Reproduced from Napieralski, J.A.,
Nalepa, N., 2010. The role of resolution and contour interval in drumlin delineation and measurement. Computers & Geosciences 36, 222230.
results from morphological studies of glacial landforms, such

as this, will contribute to our understanding of landformprocess dynamics.
The dimensions and temporal nature of paleoice sheets is
determined through subglacial landform characterization.
Napieralski et al. (2007a) developed spatial analytical GIS
tools to assist in the systematic improvement and refinement
of numerical ice-sheet models through a comparison between
landform sets (i.e., lineations, end moraines) and model
output for the LGM in Scandinavia. The Automated Proximity
and Conformity Analysis (APCA) (Li et al., 2008a, b) was
designed to measure agreement between lineations and model
output indicating ice extent (commonly mapped and analyzed
in vector format), whereas the Automated Flow Direction
Analysis (AFDA) (Li et al., 2007) compared the orientation of
streamlined landforms (i.e., surrogate for ice-flow direction)
with simulated ice-flow direction (landform data converted to
raster format). In this example, the landform maps were developed in GIS and then compared against model data to elicit
the model parameters that generated ice-sheet configurations
that best matched field evidence. In addition to merging field
and model data to reconstruct ice-sheet extent and behavior,
geostatistical tools within GIS have also been utilized to reconstruct Quaternary landscapes using strandline (i.e., raised
shorelines) evidence (Leverington et al., 2002). Strandline
data can be interpolated from isobase data and then subtracted from modern topographic and bathymetric data
to generate a rasterized approximation of paleotopography.
Leverington et al. (2002) noted the efficiency at which the
datasets could be generated using relatively basic GIS functionalities, as well as the ability to integrate and visualize
extensive data for a large area of the Canadian Arctic.
3.8.4
Volcanic Terrain and Landforms
Volcanic landscapes are rapidly evolving geomorphologic

systems. During their growth, volcanoes undergo repeated
phases of eruptive activity associated with the emplacement of
different types of volcanic deposits such lava flows, ash-fall
deposits, and pyroclastic-flow deposits, causing progressive
evolution of edifices with varying shapes and heights. Concurrently, emplacement of magmatic intrusions at depth
contributes to the deformation and uplift of the surface
topography. Eruptive activity and other volcano-tectonic
processes also lead at times to a rapid destruction of part of a
volcano through caldera formation or flank collapses. Gravitational deformation and erosional processes will act at reducing and dissecting the volcanic relief during and following
the period of active volcano growth. Changes in volcano
200
Figure 8 Landform category examples used to map Tibetan Plateau, including SRTM panels (left), Landsat 7 ETM imagery (middle), and
mapped landforms (right). All satellite images are false-color composites of bands 5, 4, and 2, draped with a semitransparent band 8. The DEM
panels are draped by a semitransparent slope image in panel (a) and by semitransparent hill-shade images in panels (b), (c), and (d). North is
toward the top of the maps and each map is 9 9 km. (a) Glacial valley in the upper left. (b) Marginal moraine deposited beyond the reaches of
a glacial valley (lower left corner) and an integral part of the hummocky terrain area. (c) Glacial lineations on the floor of glacial valleys. (d)
Hummocky terrain area. Reproduced from Moren, B., Heyman, J., Stroeven, A.P., 2011. Glacial geomorphology of the central Tibetan Plateau.
Journal of Maps v2011, 115125. doi: 10.4113/jom.2011.1161, with permission from Taylor and Francis.
geomorphology are associated with, and control the spatial

distribution of volcanic hazards. Remote sensing and GIS offer
a wide range of techniques to document volcano morphology,
its temporal evolution, and the associated hazards.
Individual subaerial volcanoes on Earth range in diameter
from 102 (e.g., spatter or scoria cones) to 105 m (i.e., Hawaiian
shield volcanoes), and from a few meters to B4 km in relief
(e.g., Kilimanjaro). The synoptic view provided by satellite

imagery is used to assess essential geomorphologic questions
such as: (1) how does the size and shape of volcanoes vary
with the type of volcanic activity and the tectonic setting? (2)
what is the spatial distribution and geometric characteristics of
contrasted volcanic deposits within a volcano? (3) what geomorphologic, tectonic, and geologic factors contribute to the
(a)
201
Spacing Length
Spacing Length
(b)
(c)
(d)
Figure 9 A spatial distribution of glacial bedforms reveals a distinct difference that can be used to discriminate between flow events: (a)
hypothetical lineation pattern; (b) an interpretation that assumes all of the flow evidence is of the same age; (c) an alternative interpretation that
accounts for cross-cutting lineations; and (d) flow event discrimination based on spatial pattern and morphometry. Reproduced from Clark, C.D.,
1997. Reconstructing the evolutionary dynamics of former ice sheets using multitemporal evidence, remote sensing and GIS. Quaternary Science
Reviews 16, 10671092.
destabilization of volcanoes? and (4) can the morphology of

eroded volcanoes be used to estimate their age or their
initial shape?
The geomorphology of volcanoes is characteristically difficult to study in the field due to their large scale, their
inhospitable and remote locations, or due to the hazards associated with ongoing eruptive activity. Remote-sensing investigations are therefore essential for assessing and
documenting the rapid evolution of volcanic surface features,
and for characterizing the surface morphology of edifices or
structural features, such as calderas. Great advances in the
study of volcanic processes and the mapping of related hazards were also achieved within the last two decades thanks to
the increased availability of accurate DEMs and the integration
of datasets and models into GIS.
3.8.4.1
Satellite Images
A variety of satellite sensors are readily used to study eruptive

processes, including thermal emissions, identification of volcanic clouds, and quantification of ash and gas emitted during
explosive eruptions (Ernst et al., 2008). Newly emplaced volcanic material, such as lava or pyroclastic flows, can be easily
mapped based upon thermal emissions/surface temperature.
After cooling, lava flows generally preserve a spectral contrast
relative to older surface features given weathering, making it
possible to discriminate flows using spectral data. Simple
techniques such as spectral-feature extraction, thresholding,
and various other pattern-recognition techniques can be used
to produce thematic maps.
Real-time mapping of newly emplaced material is essential

for the identification of the hazard/risk areas, but also to
rapidly assess the potential of secondary hazards, such as
those caused by pyroclastic flows. Detailed mapping of eruptive material is crucial at remote volcanoes where direct observations and monitoring data are generally lacking, as it
provides key constraints on the eruption style and the volume
of erupted magma. Whereas such mapping is now routine for
lava flows, which cause a drastic and complete change in
surface cover, a future research challenge will be to use satellite
images to map spatial areas affected by ash deposition, and to
derive information on the variation in ash thickness from its
impact on surface spectra. The spatial distribution of ash is an
essential constraint on eruption dynamics, and has implications related to health hazards (Figure 11).
Optical sensors are of limited utility to assess posteruptive geomorphologic changes in tropical countries, due to
the frequent cloud coverage. In these regions, microwave
sensors, with all-weather capabilities, facilitate assessment of
newly emplaced material. For example, the 2002 lava flow of
Nyiragongo caused infrastructure damage and casualties.
Rapid acquisition of a range of satellite data, through the
implementation of the U.N. International Charter for Space
and Major Disasters, highlighted that Synthetic Aperture
Radar (SAR) data were the most useful in characterizing
the extent of the area affected by the lava flow, providing
support to relief operations (www.disasterscharter.org).
Microwave SAR data are generally useful for identifying and
studying volcano-tectonic structures and deposits in tropical
regions like Indonesia (Carn, 1999).
202
Ribbed moraine ridge in

lac naococane, quebec
40 km
(a)
(b)
Figure 10 (a) A generalized overview of the Lac Naococane ribbed

moraine ridge in Central Que`bec (Canada) and estimated ice-flow
direction. (b) The connectivity of moraines is determined by sliding
moraines along the axis of ice-flow direction (i.e., solid line) and
rotating the moraines in GIS to achieve a better fit (the shaded
moraines were rotated). Modified from Dunlop, P., Clark, C.D., 2006.
The morphological characteristics of ribbed moraines. Quaternary
Science Reviews 25, 16681691.
Detailed mapping of eruptive material using satellite imagery is also used to reconstruct the eruptive history of a volcano, and to constrain the spatial variability in eruption
dynamics. Different lava-flow surfaces, associated with contrasted flow dynamics and eruption rate, can be discriminated
based upon spectra or surface roughness derived from SAR
backscatter data. Using SPOT and AirSAR data, Rowland
(1996) produced a detailed map of the lava flows of the San
Fernandina volcano in the Galapagos. Radar images proved
especially useful to discriminate between the smooth pahoehoe and the rough as lava surfaces. This map enabled the
identification of two contrasting types of eruption characterized by vent type and location, eruption volumes, and lava
morphology. These differences accounted for variations in
slope segments along the volcano flank. Smets et al. (2010)
used a time series of multispectral and radar data to document
the multiple overlap of lava flows at Nyamulagira volcano in
the Democratic Republic of Congo over the last century. The
loss of coherence between successive SAR phase images was
Figure 11 Satellite images captured by the High Resolution

Geometric sensor onboard SPOT 4 satellite over Oldoinyo Lengai
volcano (Tanzania). (a) 10 m resolution false-color composite
acquired on 15 June 2007, highlighting the active crater and the
overflow of natrocarbonatite lava on the North flank, as well as a
fresh 3 km long lava flow on the West flank, emplaced in March
2006; (b) 5 m resolution black and white image acquired on 29
October 2007, after the onset of explosive activity within the active
crater. Recently burned areas are observed on the NW flank and
significant ash deposits affect the entire region west of the summit.
Opposing satellite-viewing angles on the two images account for the
varying topographic distortions.
used to identify newly emplaced flows in zones where spectral

contrast between recent flows was insufficient in the visible
and near-infrared wavelengths.
In a complementary approach, Kahle et al. (1988) analyzed the evolution of the spectral characteristics of lava flows
in the thermal wavelengths, in relationship to their emplacement age. They showed that the weathering of superficial
lava crust was associated with a detectable variation in their
thermal-infrared emittance spectra. Such an approach opens
interesting perspectives for the systematic mapping and relative dating of lava flows at volcanoes for which the eruption
history still remains to be documented.
Spectral variations of volcanic materials were also used to
discriminate between lithologies (i.e., dacite lava blocks,
ignimbrite) observed in the deposits of a large debris avalanche at Socompa volcano in Chile (Wadge et al., 1995).
These rock types could be associated with different sectors
of the volcanic edifice prior to collapse, which enabled further
constraints to be put on the sector collapse dynamics.
Hyperspectral data were also used to document the occurrence

and type of weathered volcanic rocks at active volcanoes like
Vulcano (Mazzarini et al., 2001). At Mt. Rainier in Washington
state, U.S., the identification of weathered rocks provided essential constraints on the potential location of future flank
instabilities (Crowley and Zimbelman, 1997). Destabilization
of these weathered materials have produced repetitive sector
collapses in the past, and the probability of a future event
represents a potential hazard for the densely populated area of
Tacoma, around the volcano.
Multitemporal imagery of volcanic landscapes that document eruption events is of great interest to assess erosional
processes rapidly reshaping the newly formed landscape, and
can characterize vegetation colonization, and succession over
new volcanic surfaces. This was documented after the Plinianstyle explosive eruption of Mt. Pinatubo in the Philippines in
1991, which deposited an ash layer, over 200 m thick in places, on the volcano flanks. This loose ash was rapidly remobilized during the successive rainy seasons following the
eruption, causing deadly mudflows or lahars in inhabited
plains at the volcanos base and secondary ash avalanches due
to detachment of poorly indurated ash blocks in the deep
erosion canyons (Torres et al., 2004). Vegetation growth on
this ash layer caused a significant decrease in ash mobilization
and canyon incision within 10 years of the eruption.
Finally, the nature of planetary volcanism (e.g., Mars, Venus,
Io) has been a matter of debate for a long time. One has to rely
almost exclusively on imagery and topographic datasets collected by space missions. Based on these constraints, volcanologists, and planetary scientists are looking for analogous
topographic features on Earth to identify the processes responsible for the formation of extra-terrestrial morphologies.
This approach applies to a variety of landforms, and was specifically followed to investigate the volcanogenic nature of
cones on Mars, to identify large-scale gravitational instabilities
on Olympus Mons (e.g., McGovern and Morgan, 2009), or to
suggest the occurrence of magma-water interaction processes,
and therefore the presence of water in the subsurface of Mars
(e.g., Murray et al., 2010).
3.8.4.2
DEMs are now widely available and used for: (1) overall
visualization of the geomorphology of a volcanic area; (2)
production of background maps to locate field observations
and sampling sites; (3) morphostructural interpretation of
volcano-tectonic interactions; (4) geomorphometric characterization of volcanic features; (5) geometric characteristics of
volcanic deposits or features; and (6) modeling of volcanic
flow processes and assessment of related hazards.
Since geologists recognized the wide range of sizes and
shapes of edifices produced by volcanic activities, classification
schemes are required that account for morphological variations related to magma composition, eruption dynamics,
or tectonic settings. Systematic quantification of terrestrial
volcano morphologies is also required to compare the
morphologies of potential volcanic structures observed on
other planets (e.g., Pike and Clow, 1981).
Using DEMs, it is now possible to systematically document
the spatial distribution and morphometry of many volcanic
203
structures within a given region or on a worldwide scale.

Kervyn et al. (2008) and Wright et al. (2006) showed that the
grid-cell size of the DEM needs to be adapted to the average
size of the features of interest. A SRTM DEM at 90 m resolution is sufficient to characterize polygenetic stratovolcanoes
(Wright et al., 2006), but a finer resolution is required to accurately characterize the morphometry of pyroclastic cones
which are a few hundred meters wide (Kervyn et al., 2008).
Using GIS-based techniques, a large set of morphometric
data characterizing the size, height, and shape of volcanic
cones can be derived. Such statistical datasets provide key
constraints to interpret the relationship between the relative
crater size and the intensity of the eruptive activity, or the
decrease in the relative cone height caused by erosion. Using
LiDAR data, Favalli et al. (2009) documented all the cones on
the flank of Etna in Italy, and highlighted the great variability
in morphometric ratios relative to previously defined average
values (Figure 12). The spatial distribution of volcanic
vents within a volcanic field can be easily mapped with a
high-resolution DEM, putting constraints on the structures
controlling their spatial distribution and the associated distribution of hazard. The alignment of these vents and the
morphology of the surrounding volcanic cones can moreover
be used as the main constraint to infer the orientation of
feeding dykes which are clear indicators of the stress fields
within volcanic edifices (e.g., Corazzato and Tibaldi, 2006).
The ASTER GDEM has also provided thematic information
(e.g., caldera floor, incised hydrographic network, various
deposits and flows, flow paths and morphologies such as
ogives, landslide scars) derived from geomorphometric analyses of cones, thus revealing new volcano-tectonic features
and offering a reconnaissance tool for landform mapping in a
volcanic environment (Figure 13; Zouzia et al., 2011).
Grosse et al. (2009) recently developed the MORVOLC
method, with the objective of characterizing all terrestrial volcanoes using a standardized methodology. This method is
based on a set of quantitative parameters describing the size
and shape of an edifice from a planimetric (e.g., ellipticity, irregularity) and cross-sectional perspective (e.g., height/width,
summit/basal width, slope). Based on these parameters, volcanoes were classified as cones, subcones, and massifs. It can be
applied to any volcanic structures represented in a DEM, given
an appropriate spatial resolution. Application of this methodology to 145 strato-volcanoes from the Andes suggested that
this quantitative method provided strong evidence for different
evolutionary trends of volcano growth, which can be interpreted in terms of magma supply, spatial distribution of vents,
edifice strength, gravitational deformation processes, and tectonic controls. A comparable methodology has recently been
applied to the study of a selection of symmetrical stratovolcanoes, which highlighted two groups with contrasting slope
profiles around the summit, attributed to different material
characteristics and thus different dominant eruptive dynamics
(Karatson et al., 2010a). The key limitation of such systematic
quantification of volcano morphologies is the definition of the
volcano base. So far the delimitation of the volcano base depends upon visual interpretation, introducing subjectivity and
low reproducibility. The definition of common repeated criteria
or the development of automated algorithms to identify the
volcano base is crucial for the production of consistent results.
204
(a)
(d)
(b)
(e)
(i)
0
(c)
400 km
N
(f)
(g)
200
(h)
Figure 12 Examples of Etnean scoria cones showing typical shape variations. (a) a pre-122 BCE well formed cone (minimal modification in
two millennia); (b) a less regular cone (the cone shape is obvious, but the base is irregular); (c) the majority of Etnean cones are even more
poorly formed (cone can be recognized but the base is irregular due to lava flows; (d, e) strongly eroded cones (if there is a well-developed gully
system, these are referred to as heavily dissected cones); (f) many cones show only little evidence of gully dissection; (g, h) example of how
gully formation is controlled by vegetation, which is related to altitude and exposure; (i) cone shape can change during emplacement due to
eruptive dynamics (e.g., structurally deformed cone). Reproduced from Favalli, M., Karatson, D., Mazzarini, F., Pareschi, M.T., Boschi, E., 2009.
Morphometry of scoria cones located on a volcano flank: a case study from Mt. Etna (Italy), based on high-resolution LiDAR data. Journal of
Volcanology and Geothermal Research 186, 320330.
Morphological analysis of volcanoes has also proved useful

to reconstruct the past shape of eroded edifices. Combining
terrain analysis with field observations and geochronological
data, Karatson et al. (2010b) proposed a model for the different stages of evolution for Mt. San Francisco in Arizona,
U.S. Rodriguez-Gonzalez et al. (2010) also illustrated how
paleogeomorphologic reconstruction can be conducted
using interpolation to estimate the original volume of tephra
and lava emitted by an eroded volcanic center. Such

reconstruction, when sufficiently well constrained with complementary data sources, provides valuable insights on the
volcanic history of a region.
It is now widely recognized that volcanoes are not simply
growing by accumulation of volcanic material from one or
several vents. They are also affected by internal deformation
due to magmatic, tectonic, or gravitational forces, leading to
780000
786000
783000
LLP
4056000
Emporios
AF
(1)
(2)
BR
(3)
(5)
(1)
(4)
(5)
(7)
(2)
(6)
(a)
MPE
(4)
PE
KV
PV
4053000
ST
TR
(3)
205
Nikia
Avlaki
0
4
7
11
15
18
22
26
30
33
37
41
44
48
52
55
59
Degrees
4050000
Nikia lava flows
Pyroclastic flows
Avlaki lava flows
Lava flow front and limits
Post caldera domes
Dome
N-NE-E-SE-S calders wall
Number and relative age
(1) of lava flows and domes
Caldera floor
Caldera rim and
Landslide
phreatic crater
Magma feeder systems
River
Dome scree deposits
Fault
(b)
Lava flows
Trapdoor fault (PRF)
Figure 13 Slope image of the Nisyros caldera (Aegean Volcanic Arc). The pixels range between 01 and 58.81 (the darker the pixel, the higher
the slope. Inset (a) displays volcano-morphological features resulting from ash and lava flows, lave flow paths and feeder systems, ogive
structures, younging, and contact relationships, height of lava fronts and width of flow. Inset (b) displays volcano-tectonic and tectonic faults and
resultant features such as radial hydrographic network along radial caldera faults, the Nisyros caldera ring fault, caldera rim, and the Perigussa
trapdoor (PRF). Reproduced from Zouzias, D., Miliaresis, G.C., Seymour, K.S., 2011. Interpretation of Nisyros volcanic terrain using land surface
parameters generated from the ASTER global digial elevation model. Journal of Volcanology and Geothermal Research 200, 159170. Site names
are provided as acronyms. Consult article for site names.
long-term incremental deformation, or catastrophic and largescale ground displacements such as caldera or flank collapses.
All these processes, which are potential sources of hazards, are
currently being studied at many volcanoes, using a variety of
techniques including field-based structural assessment, realtime ground and space-based monitoring techniques, and
analogue and numerical modeling efforts. In many regions,
application of field monitoring and geological mapping is
restricted by access or cost limitations. DEMs provide essential
data to explore the distribution of volcano-tectonic structures,
generally indicated by km-scale topographic lineaments or
break-in-slopes. Lagmay and Valdivia (2006) illustrated how
the morpho-structural features (e.g., breached crater) identified on the SRTM DEM for Philipinian strato-volcanoes could
be related to the orientation of the main regional principal
stresses and known strike-slip faults. These observations were
consistent with results from sand-cone analogue models
simulating the effect of strike-slip tectonic faults on the deformation of a conical edifice. In the same way, the natural
applicability of analogue models is generally validated primarily based on morpho-structural comparison of the models
with natural volcanoes using terrain analysis. The synoptic
coverage of imagery and topographic information offers the
unique advantage to identify large scale but subdued features,
such as some calderas or debris avalanche deposits that cannot be readily identified by local field observations.
Multitemporal DEMs can be used to characterize the volume and change in geometry of newly emplaced volcanic
materials. Advances in technologies such as SAR interferometry, ground, air- or space-borne stereoscopic imaging and
LiDAR, have the potential to rapidly and repetitively produce
DEMs over an evolving eruption area. Lu et al. (2003) used
SAR-based DEMs to estimate the volume of a lava flow at
Okmok Volcano in the Aleutian Islands, U.S., with unprecedented accuracy (15% error). Bastero and Lagmay (2006)
showed that essential information such as lavas silicate content and yield strength can be estimated from high-resolution
DEMs. Such applications are, however, limited by the insufficient spatial resolution and accuracy of freely available DEMs,
resolving the volume of newly emplaced materials only a few
meters thick. High-resolution LiDAR data proved efficient in
mapping individual lava-flow units, characterizing the channel geometry at Etna. This opens the potential for more detailed characterization of the geometry and volume of newly
emplaced volcanic materials, as repeated acquisition of LiDAR
data become more readily available and affordable in the
future.
Differential InSAR is another relevant technique to document topographic changes at volcanoes, as it is capable of
mapping mm-scale surface displacement related to coseismic
ground displacement or crustal-magma movement. This
technique is based on the calculation of the phase shift between the radar backscatter registered by a radar antenna in
two successive flights over the same area. In order to isolate the
phase shift attributed to displacement occurring between the
two data acquisitions, an accurate DEM has to be available to
206
retrieve the topographic effect on phase differences. Documentation of these displacements in volcanic regions provide
essential constraints on the location, volume, and geometries
of intrusive bodies, as well as the location of active faults.
Interpolated over repetitive individual intrusive or seismic
events, the deformation field calculated through differential
InSAR provides an indication on the long-term morphological
evolution that can be attributed to intrusive or tectonic
processes.
Topographic information also contributes to the better
understanding of volcanological processes and volcano
growth, as terrain conditions are required for hazard modeling. Some of the greatest hazards posed by volcanic activity are
associated with flow-dominated processes such as pyroclastic
density currents, debris avalanches, lava flows or mud flows
(lahars). Although the physics of each of these processes is
different and often poorly understood, topography always
represents a key constraint on the flow path and the extent of
the affected area. Stevens et al. (2002) illustrated how the
accuracy and resolution of DEMs was crucial in controlling the
results of a lahar-simulation model. Some of the most easily
implemented lava-flow simulation models use topography as
the main control on lava-flow invasion probability. The spatial
variation in the probability of invasion is modeled by integrating the results of multiple iterations of flow-path simulations, that implement a random-height variation of chosen
amplitude into the initial DEM to account for the ability of
lava flows to overflow limited relief (e.g., Felpeto et al., 2001).
The difference between the reconstructed pre- and postcollapse volcano topography is also a principal constraint of
debris-avalanche models. These examples show that although
it is crucial to constrain the rheology and dynamics of hazardous volcanic flows to accurately model the area they might
affect, much attention should also be paid to the use of accurate and up-to-date topographic datasets with the appropriate spatial resolution to obtain reliable hazard maps.
3.8.4.3
GIS-Based Mapping
GIT is used in volcano geomorphology and volcanology to

develop, integrate and manage spatial databases, conduct
spatial analysis and modeling, and to produce a variety of
information products to communicate the results of scientific
research to a nonexpert audience.
Geospatial analytical software tools are used to study the
controlling factors that govern the spatial distribution of
point- or linear-type features in volcanic environments. The
distribution of volcanic cones in flat volcanic provinces can be
analyzed to highlight the orientation of the underlying structural pattern controlling the propagation of magma through
the crust. For example, Hamilton et al. (2010) analyzed the
nearest-neighboring distance within a field of rootless cones in
Iceland. Through this geospatial analysis, it was shown that
these cones were characterized by a repelled distribution
maximizing the distance to the nearest neighbor, indicating
that the formation of these structures was controlled by a
competitive process for a limited resource, here water.
The study of active volcanic regions for the prediction of
short-term hazards or the understanding of long-term
evolution always requires the integration of a variety of data

and observations including topographic and geological data,
imagery, field observations, geophysical data, and numerical
model results. Prior to an eruption at a large shield volcano,
it is essential to be able to locate the location of earthquake
epicenters, the direction and amplitude of surface deformation recorded by GPS stations, and to examine this information contextually in relation to existing fractures, dike
networks, or local topographic lows, to obtain an indication
on the probable location of the eruption outbreak. Analysis
of slope orientation, flow paths, or more advanced flow
simulations cross-correlated with maps of human infrastructure and settlements, enable elements-at-risk to be rapidly
identified.
Detailed geological maps of active volcanoes can be rapidly
updated once they are integrated in a GIS. During the eruption
of Nyamulagira in the Democratic Republic of Congo in
January 2010, mapping of the advancing lava flow using
thermal imagery was combined with lava-front localization by
helicopter flights in the field to constrain lava flow-path
simulation models. Such near real-time integration of information, even in the absence of advanced monitoring techniques, lead to rapid production of updated maps that serve to
communicate to the authorities about an evolving eruptive
crisis, and in refining the risk assessment for the local population. At the end of the eruption, the existing GIS-based
geological map directly included the newly emplaced lava
flow, providing key information about its geometry and the
impacted region. Making assumptions on the lava thickness,
its volume can be estimated and the DEM updated accordingly, in order to account for this morphological feature in the
prediction of flow path during future eruptions.
GIS is also essential in the assessment of the spatial distribution of long-term hazard and risks around a volcano. A
first key step in such assessment, especially for a large composite volcano, is estimating the spatial distribution of the
probability of eruption outbreaks. Marti and Felpeto (2010)
recently proposed a GIS method that integrates several indicators such as the distribution of past eruptive vents, fractures,
and the documented volcano stress field in estimating outbreak probability. Using this probability of occurrence,
plausible eruption scenarios, varying in the eruption dynamics
and intensity can be simulated with appropriate models and
results displayed in a GIS. Separate hazard maps are generally
produced for different eruptive processes such as fallout and
pyroclastic flows, as they respond to contrasting dispersion
dynamics, but hazards are associated with different intensity
of one process (e.g., increasing lava effusion rates) and are
summarized on a single output map. Correlation of these
hazard maps with the vulnerability of the population and
infrastructures in the region leads to the production of synthesized maps localizing the probability of risk from volcanic
eruptions.
A compromise has to be found between the precision of
the data provided by such hazard or risk maps, and its readability for the stakeholders and population that will base their
decisions on it. Using GIS, great efforts can be focused on the
production of the most informative and understandable endproducts: different levels of hazard should be coded in a
few clearly labeled classes and should be displayed on a
background map or image on which people can easily locate

themselves. The effectiveness of various visual outputs to
convey information about risks was evaluated by Haynes et al.
(2007) showing that 3D perspective views of aerial photos
revealing the areas at risk were judged to be the most effective
by the targeted population. Research on effective dissemination of scientific spatial information to the general public is
only in its early stage in volcanology, and much more attention needs to be paid to these issues in the future.
With the rapid development of interactive web-based GIS
platforms, several new developments will become possible.
First, existing process-simulation models will be fully integrated into GIS and made more available. This will make it
possible for the end-user to evaluate the impact of eruption
scenarios and to produce hazard and risk maps most appropriate for their needs and area of interest. The integration of
several eruption simulation models into a GIS offers the opportunity to simulate the long-term evolution of the morphology of a volcano according to defined scenarios of long-term
eruptive behavior and eruption rates evolution; it also allows
volcanologists to test the effectiveness of engineering mitigation measures (e.g., Chirico et al., 2009).
3.8.5
Landslide Mapping
Landslides occur frequently and can be very destructive.

Hence, it is imperative to understand what the controlling and
triggering factors are, and be able to conduct hazard and risk
assessments. The use of GIT greatly facilitates the assessment
and mapping of landslides, and in predicting the potential
impacts on the environment and society. Numerous forms of
spatial data are collected and used for local and regional scale
assessments (Alexander, 2008). Detailed maps can be generated to show past events and predict likely future events.
GIS-based landslide analysis is a relatively new field and the
majority of such GIS-based investigations have occurred since
the 1990s. Carrara and Guzzetti (1995) noted that few branches of hazard studies have been transformed as completely by
information technology as has landslide research. The study of
geospatial technologies for landslide analysis has rapidly
evolved, and Alexander (2008) acknowledged that, as a result,
landslide occurrence and susceptibility are now generally
known in many areas of the world. Good overviews, reviews,
and edited volumes on GIS and natural hazards and landslides
are those by Varnes (1984), Carrara and Guzzetti (1995),
Coppock (1995), Wohl and Oguchi (2004), Chacon et al.
(2006), Alexander (2008), and Carrara and Pike (2008).
Alexander (2008) revealed that there is a considerable redundancy in landslide research results, and a tendency for
scientists to use GIS in an exploratory, rather than hypothesissolving mode. Alexander (2008) argued that the reason for
this misguided trend is due to the use of GIS as mainly a
deductive, rather than an inductive method. Carrara and Pike
(2008) warned that although the increasing user-friendliness
of GIS-software and the widespread use of it reflect a positive
development, it also encourages the perception that GIS-based
landslide assessment is a point-and-click undertaking, accomplished rapidly with limited resources. Although a large
number of assessment methods are available today, there is a
207
lack of standardization in the use of analysis protocols and

information products (Chacon et al., 2006). Nevertheless, GIS
software tools are routinely used in sustainable land management and development.
Slope instability in a cultural landscape is a function of
direct and indirect natural and human factors, the so called
controlling or triggering parameters. These generally include
lithology, structure, tectonics, geomorphology, topography,
precipitation, temperature, infiltration, runoff, land cover, and
road construction. Others exist, although they have not been
widely utilized and include topographic stress, rock strength,
and rate of erosion, as system dynamics control slope stability.
Unfortunately, many of these variables effectively represent
associations with the potential for slope instability. Terrain
analysis greatly facilitates an assessment of land-surface morphometry and processes that are related to slope instability,
and provides a better understanding of how and under what
circumstances, events, states, dynamics, or processes trigger
landslides. Chacon et al. (2006) described the types of data
and the methods that are included in these analyses which
include: (1) geotechnical data and slope-hydrological models;
(2) temporal data on failure and mass activity; (3) spatial and
temporal data on stages of development; (4) rainfall and
seismic events; and (6) slope-stability analysis based on
probabilistic-reliability index assessment.
Today, there is a wide range of data that can be used in
landslide research including photography and movies (terrestrial, aerial, repeat), surveying data, images from passive
and active sensor systems (LiDAR and Radar), and morphometric parameters and indices based upon terrain analysis of
DEMs. Although terrestrial data sources are mainly employed
to study discrete landslides or smaller landslide complexes,
satellite imagers are used for examining entire landslide regions. Digital imagery from airborne or satellite sensors deliver important information about the spatial distribution of
landslides in affected regions (inventory), landslide parameters such as area and run-out distance, and environmental
conditions leading to the failure (event-controlling parameters) such as lithology, vegetation, or roads. Multitemporal
data sets allow for studying and monitoring changes and for
understanding geomorphic processes.
3.8.5.1
Terrestrial and Airborne Photography
The qualitative interpretation and quantitative analysis of

terrestrial photographs and movies from discrete landslides or
landslide complexes are used to reveal causes and processes of
landsliding, and to automatically monitor landslide displacement over longer time periods. Photographs proved to be
useful in documenting landslide movement (e.g., Goguel,
1989) and for rapid assessment of landslide displacement
(e.g., Kalaugher and Grainger, 1997). Iovine and Parise (2002)
developed a graphic method to classify landslide damage in
urban structures. Movies can be taken using stationary 35 mm
cameras for monitoring (e.g., Ballantyne et al., 1987) and
time-lapse studies of landslide movement (e.g., Belknap and
Gilmore, 1987).
Airborne photographic surveys offer a synoptic view of entire
landslide areas, and stereophotogrammetry allows for estimating
208
the 3D coordinates of points on a landslide. Airborne photographs can be used to study landslide behavior, displacement,
and kinematics (Yoshizawa et al., 1991; Brueckl et al., 2006;
Baldi et al., 2008), and for monitoring landslides (Chandler and
Cooper, 1988; Kajiyama et al., 2005; Walstra et al., 2007). Using
aerial photogrammetry (including aerial photos taken at a variety of flying altitudes), Dvigalo and Melekestsev (2009) mapped
deposit thickness and surface characteristics with the objective of
qualitatively and quantitatively characterizing post-landslide
landforms and forecasting future events in Geyser Valley, Kamchatka (Figure 14). Furthermore, repeat terrestrial photographs
can add information about how the landscape changed after an
event (e.g., if the landsliding continued or the affected slope
recovered). Extensive multitemporal landslide inventories can be
used to draw general conclusions for entire landslide-effected
regions (Khattak et al., 2010).
3.8.5.2
component. Recent technological developments led to the

introduction of very high resolution (VHR) sensors with spatial resolution o0.5 m that allow for the creation of large scale
maps of discrete landslides (Figure 15).
Over the last 40 years, Landsat images have been an important dataset in landslide research and landslide-risk assessment. For example, Strom and Korup (2006) combined
Landsat ETM imagery with SRTM DEMs and declassified
KFA-1000 satellite imagery to study extremely large rockfalls
and rock avalanches. Seismic-triggered landsliding was mapped by Evans and Bent (2004), Leroy et al. (2010), and Willige
(2010). Dunning et al. (2006) draped an ETM image over a
90 m SRTM DEM to investigate the formation and failure
of a landslide dam. Alcantara-Ayala et al. (2006) related
landsliding to land-cover change by using vegetation fragmentation determined from decadal Landsat images (TM
and ETM ). MacDonald and Grubbs (1975) predicted landslide-prone areas for highway construction, and Aniya et al.
(1985) identified landslides for watershed management.
Other examples include Bencardino (2005) who assessed
landslide hazard and risk in river basins, and HaberlerWeber et al. (2009) who conceptualized multiscale landslide
monitoring and evaluation. Landslide susceptibility and
Satellite Imagery
The utilization of satellite imagery in analyzing landslides

started after the launch of the first Landsat satellite in the
1970s. Today, satellite imagery is frequently used in landslide
research and landslide-risk assessment that has a spatial
Zhemchuzhnyi
Goryashchii
Rozovyi konus
Velikan
Ge
ize
rn
ay
aR
Neprotayannyi
Grot
New
Fountain Fountain
shchel
Dammed
lake
Bolshov
Office of
kronotskii
reserve
Line of detachment of
landslide rocks
Helicopter field
Thermal patches arising
after the landslide
Stream beds overlain by landslide

rocks and by the dammed lake
Lakes
Preserved geysers in the

central part of the valley
Sestrenka R.
Lines of equal landslide

deposit thickness (5 m)
Remnants of rocks partly or

completely buried by landslide
Thermal anomalies
Triumfalnye
vorota
Vdp
um
Sh
ya
na
Pervenets
.
250
500 Meters
Figure 14 Map of landslide thickness (in meters) for the Geyser Valley in Kamchatka, derived from a 2007 aerial photography survey (flying
altitude varied to produce three different scales) and subsequent creation of three-dimensional terrain models. The goal was to measure
thickness, but also to map the transportation path of material. Modified from Dvigalo, V.N., Melekestsev, I.V., 2009. The geological and
geomorphic impact of catastrophic landslides in the geyser valley of Kamchatka: aerial photogrammetry. Journal of Volcanology and Seismology
3, 314325, with permission from Springer.
209
(Davis et al., 2000; Dahal et al., 2008; Hasegawa et al., 2009).

They can be generated from a variety of data sources, and
special care must go into the preprocessing that is typically
required before conducting an analysis. We highlight a few of
the data sources and types of analysis commonly used.
3.8.5.3.1
Figure 15 Very high-resolution image of landslides in Nova

Friburgo, Brazil. The image is from the GeoEye-1 satellite (0.5 m
resolution), acquired on 20 January 2011 following heavy rains that
triggered numerous deadly landslides in Nova Friburgo. Note the
intensity of deforestation that is known to increase the probability of
landsliding (image downloaded from www.geoeye.com).
hazard mapping was also conducted by Samarakoon et al.

(1993), Lee and Dan (2005), and Pradhan (2010).
Higher resolution SPOT imagery has also been used in
landslide and hazard studies. For example, Zhou et al. (2001)
and Yang et al. (2004) used imagery to study specific landslides. Aste and Girault (1995) and Yamaguchi et al. (1999)
used imagery to detected landslide movement, and Leroi et al.
(1992) and Ostir et al. (2003) monitored landslides. Roessner
et al. (2006) also studied landslides with the potential for dam
formation. Lee and Pradhan (2006), Dymond et al. (2006),
and Schneider (2009) used SPOT data for landslide-risk assessment. The data were utilized to assess the relationship
between tectonics and landslides (e.g., Morelli et al., 2001;
Wang et al., 2003). Vegetation recovery after an earthquake can
also be investigated (e.g., Chou et al., 2009).
The VHR sensors on IKONOS and QuickBird are useful
for large-scale landslide mapping. IKONOS data were used
for slope instability analysis for pyroclastic deposits (Baeza
et al., 2006), extraction of landslide traces (Ochoa-Tejeda and
Parrot, 2007), retrogressive thaw slump activity studies (Lantuit and Pollard, 2008), mapping of earthquake-related landsliding (Rathje et al., 2006), investigating urban landslides
(Wang, 2006), and landslide-risk assessment (Nichol et al.,
2006). Using QuickBird data, Delacourt et al. (2004) and
Chadwick et al. (2005) carried out landslide displacement
measurements, and Granica et al. (2007) derived indicators
for landslide detection.
3.8.5.3
DEMs are now widely used in landslide research (Iwahashi

et al., 2001; Barlow et al., 2003) and landslide-risk assessment
LiDAR
Light detection and ranging (LiDAR) systems have led to significant advances in geomorphology, particularly through
their abilities to detect subtle topographic features and penetration through vegetation canopies. The use of LiDAR data in
landslide mapping is a substantial improvement over more
traditional types of data that are used for landslide inventories
(Haugerud et al., 2003; Schulz, 2007; Van den Eeckhaut et al.,
2007). Nevertheless, there is still a degree of subjectivity associated with landslide mapping based upon this data, as
there is no consensus on which topographic features are part
of, or uniquely characterize landslides (Guzzetti et al., 2000;
Van den Eeckhaut et al., 2005, 2007; Booth et al., 2009).
Terrestrial LiDAR has been used to map morphological
changes on rock falls (Mikos et al., 2005), describe geomorphic features of large landslides in mountains (Dunning
et al., 2009), characterize retrogressive landslides in sensitive
clay (Figure 16) and rotational landslides in river banks
(Jaboyedoff et al., 2009), and monitor unstable slopes in
urban environments (Pieraccini et al., 2006). Obviously the
utility of terrestrial LiDAR is limited in spatial extent, as the
primary focus tends to be on discrete slope failures or smaller
failure complexes.
Alternatively, airborne LiDAR offers a high-resolution
synoptic view of entire landslide areas to track landslide
displacements (Dewitte et al., 2008), reveal topographical
changes in earthquake regions (Chen et al., 2006), estimate
mass-wasting processes in active Earth slides (Figure 17;
Corsini et al., 2009), map landslides under forest (Van den
Eeckhaut et al., 2007) and in urban areas (Madin and Burns,
2006), and carry out landslide-susceptibility mapping (Schulz,
2007). McKean and Roering (2004) used DEMs derived from
LiDAR to characterize surface roughness of terrain near
Christchurch, New Zealand. Bedrock and earthflow landslides
tend to be rougher than adjacent unfailed terrain thereby
permitting landslide-boundary mapping. They concluded that
older mass movements are smoothed by surface processes
over time, allowing them to differentiate between active zones
of movement and older, inactive failures. Glenn et al. (2006)
also relied on surface roughness, in addition to using slope,
semivariograms, and the fractal dimension to characterize
morphology. The landslide surface characteristics were related
to underlying material type and local motion of the landslide,
although landslide size will influence semivariance values (i.e.,
small scale landslides show less roughness and semivariance
than large slides).
3.8.5.3.2
Radar
Most radar systems are able to detect surface elevation beneath

vegetation cover and be used during any season. Current airborne SAR sensors provide ground resolutions to about 10 cm,
and permit the monitoring of subtle landslide changes.
Interferometric SAR (InSAR) uses the scattering properties
210
Figure 16 Othophoto obtained from aerial photograph taken on 9

May 2006 and the terrestrial laser scanning (TLS) high-resolution
DEM at Nicolet landslide, Quebec, Canada. Contour lines with a
vertical spacing of 1 m (white lines) were derived from the TLS data.
The main and secondary landslide scarps are shown as red and
yellow lines, respectively Reproduced from Jaboyedoff, M., Demers,
D., Locat, J., et al., 2009. Use of terrestrial laser scanning for the
characterization of retrogressive landslides in sensitive clay and
rotational landslides in river banks. Canadian Geotechnical Journal
46, 13791390, with permission from Canadian Science Publishing.
from two or more SAR images to generate maps of surface

deformation and digital elevation of large-scale terrain (Carnec et al., 1996; Fruneau et al., 1996; Singhroy et al., 1998;
Rott et al., 1999). For example, Singhroy et al. (1998) integrated SAR and Landsat TM images to characterize and inventory landslides in Canada. They concluded that InSAR
techniques provide better representations of elevation and
slope changes (improving landslide detection capabilities)
than RADARSAT and that a combination of SAR and TM can
characterize retrogressive slope failures and flow features in
regions with relatively low relief. Polarimetric SAR is also
useful in multitemporal landslide analysis.
Numerous radar-based landslide applications exist, including earthquake-triggered landsliding (Czuchlewski et al.,
2003), landslide displacement (Tarchi et al., 2003; Noferini
et al., 2006), surface deformation mapping (Finnegan et al.,
2008), monitoring unstable slopes (Colesanti and Wasowski,
2006), and modeling rainfall-induced landslides (Chiang and
Chang, 2009). Bozzano et al. (2008) installed a monitoring
system that continuously measured a landslide interacting
with an under-construction infrastructure, whereas Rott and
Nagler (2006) evaluated the use of radar in landslide-risk
assessment. Tarchi et al. (2003) combined ground-based
InSAR with field instrumentation to derive multitemporal
surface deformation maps of the depletion zone of the Tessina
Elevation difference 20031973

Mass waste in river
Accumulation
250
500
750
1000 m
Depletion
Depletion
Depletion
TOE zone
TRANSIT zone
SOURCE zone
Depletion
20 15 10 5
(a)
3 0.5+0.5 +3 +5 +10 +15 +20
Elevation difference (m)

Depletion
10 m contour lines: 2006 topography
Accumulation
Accumulation
Depletion
Accumulation
TOE zone
(b)
TRANSIT zone
Elevation difference 20062003
SOURCE zone
0
250
500
750
1000 m
Figure 17 Digital elevation model difference images, computed from LiDAR data for the Valoria landslide in the Northern Apennines of Italy:
(a) 20031973; (b) 200603. Reproduced from Corsini, A., Borgatti, L., Cervi, F., Dahne, A., Ronchetti, F., Sterzai, P., 2009. Estimating masswasting processes in active earth slides: earth flows with time-series of high-resolution DEMs from photogrammetry and airborne LiDAR. Natural
Hazards and Earth System Sciences 9, 433439.
landslide (Italy) that was used to interpret landslide kinematics and evolution (short term). The use of InSAR, however, has several limitations related to landslide scale and
viewing geometry, which influence landslide mapping studies
(Catani et al., 2005; Glenn et al., 2006).
3.8.5.3.3
ASTER
The Advanced Spaceborne Thermal Emission and Reflection

Radiometer (ASTER) sensor can also be used to generate DEMs
(Kaab, 2002). ASTER-derived DEMs offer fine geometric detail,
relatively high spatial and low temporal coverage, and are costeffective for landslide studies at the global scale (Santini et al.,
2009). These DEMs are commonly used to map landslide
susceptibility and hazard assessment in areas with high relief
(Kaab, 2002; Liu et al., 2004; Havenith et al., 2006; Fourniadis
et al., 2007a, b). Haeberli et al. (2004) studied a massive rockice slide and subsequent mud-flows on the Kazbek massif
(Russia) to put the event in a historical context, mapped surface features indicative of unstable conditions, and inferred
the impact of a valley glacier. Domakinis et al. (2008) mapped
coastal landslides and Giardino et al. (2008) analyzed deepseated slope deformation. Examples of susceptibility and
hazard assessments using ASTER DEMs include the work of
Ercanoglu (2005), Fourniadis et al. (2007a, b), and Kamp
et al. (2008).
3.8.5.3.4
Geomorphometric analysis
Although much research has focused on DEM generation

and evaluation in general, the utility of terrain analysis in
landslide research has tremendous potential. Claessens et al.
(2005) showed that DEM resolution effects landslide-hazard
modeling. Fabris and Pesci (2005) described precision
and validation of automated DEM extraction using digital
aerial photogrammetry. Iwahashi et al. (2009) addressed
computational scale issues for calculation of slope using highresolution LiDAR DEMs, and Liu et al. (2007) simulated the
formation of landslides using a DEM. Although new capabilities exist, it is advisable to use caution in landslide mapping,
as researchers do not generally address issues such as digital
terrain modeling techniques, DEM error and uncertainty
issues, as well as the propagation of errors in terrain analysis.
Carrara and Pike (2008) criticized that despite a veritable
cottage industry of current DEM-based hazard mapping,
many of the predictions are either based on unsuitable data or
are not properly modeled.
Topographic parameters such as slope angle and azimuth,
profile, various curvatures, upslope contributing area, and
catchment area can be event-controlling parameters and as
a result, they are important for mapping landslide features
and analyzing causes of slope instability. In particular, highresolution DEMs offer detailed geomorphometric analyses to
ascertain scale-dependent processes in landslide susceptible
areas (in contrast to slide mechanics or model constraints
developed from InSAR) (Glenn et al., 2006). Thus, landslide
mapping now relies heavily on rigorous geomorphometric
analyses using high-resolution DEMs to produce new perspectives on the morphology and triggering mechanisms
associated with landslides. Terrain analyses in high-relief
areas include hypsometric, clinometry, and profile analysis
211
Table 3 Geomorphometric parameters and objects commonly used

in mountain terrain
Parameter/object
Application/Mountain landforms
Local geomorphic parameters

Slope angle
Insolation/snow cover, permafrost distribution
Curvature
Classification of form elements
Flow accumulation
Delineation of drainage basins
Linear geomorphic parameters
Hillslope profile
Hillslopes, toposequences
Valley long profile
Valley floor (e.g., steps)
Ridgelines
Landform/mountain classification
Areal geomorphic parameters
Form elements
Hillslopes, talus
Basins
Drainage system
Landform units
Aggregated geomorphic objects
Landform units
Toposequence
Hillslopes types
Representative geomorphic parameters: dimension/form
Length
Cirques
Elongation
Cirques
Circularity
Cirques
Representative geomorphic parameters: internal characteristics
Relief
Cirques
Floor elevation
Cirques
Steepest gradient
Cirques
Hypsometric integral
Valleys
Representative geomorphic parameters: internal/external topology
Surface structure
Rock glacier
Toposequence
Hillslopes
Source: Reproduced from Raseman, S., Schimdt, J., Schrott, L., Dikau, R., 2004.
Geomorphometry in mountain terrain. In: Bishop, M.P., Shroder, J.F. (Eds.),
Geographic Information Science and Mountain Geomorphology. Springer-Praxis,
Berlin Heidelberg, pp. 101137.
(Table 3). Hypsometric curves characterize the area/altitude

function (e.g., elevation relief ratio), whereas clinometry describes the slope angle/altitude function. Slope and curvature
profiles can include longitudinal, cross-section, and even
tangential perspectives on topographic variation that reflect
multiple process domains (i.e., erosion, transportation, and
deposition), and are thus of interest to landslide scientists
(Raseman et al., 2004). Cross-sectional assessment of valleys
can distinguish between U-shape and V-shaped valleys, map
the extent of talus, and indicate valley wall stability.
A basic mapping approach that can be used is DEM differencing, where multitemporal DEMS are used to estimate
the differences in altitude caused by landslide material depletion and accumulation. This helps to identify source,
transit, and toe zones (Corsini et al., 2009). This relatively
simple approach can highlight changes in surface characteristics over time, or following a landslide event (Figure 17).
Alternatively, the manner in which DEMs are visualized can
elucidate older landslides that have been modified by surficial
processes. Van den Eeckhaut et al. (2005) combined a series of
hill-shaded DEMs (either a sun azimuth angle of 451 or 3151)
212
with expert knowledge to test the reliability of hill-shading as

an identification tool. Some landslides did not appear on both
hill-shaded DEMs, proving that relying solely on the visualization capabilities of DEMs can be misleading. Developing,
shading, and analyzing the DEMs took less than 1 day, however, whereas field surveying lasted 100 days.
Most geomorphometric analyses use slope and curvature
parameters to map landslide structures in mountainous regions
(McKean and Roering, 2004; Chadwick et al., 2005; Glenn
et al., 2006; Van den Eeckhaut et al., 2007). Landslide surfaces
tend to exhibit more surface roughness (at local scale of meters)
than unfailed slopes (McKean and Roering, 2004), and the
extent of older, inactive landslides can be mapped based on
surface roughness, as surface processes will reduce overall
topographic variability and the variability of inactive landslide
morphological components (Glenn et al., 2006). Glenn et al.
(2006) generated semivariograms from LiDAR-derived DEMs to
study the morphological and scale-dependent nature of the
topography, and used the fractal dimension to contrast scaledependent topographic variability of different landslide components. McKean and Roering (2004) also used surface
roughness as a surrogate for detecting landslides by measuring
variability in slope and azimuth in local patches of a LiDARderived DEM. They suggested accounting for resolution/feature
scale issues, in order to increase the probability of detection and
high-quality mapping. Several studies have provided insight
into the use of semivariograms and spatial autocorrelation to
analyze topographic variability and surface roughness (Bishop
et al., 1998, 2003; Walsh et al., 2003; Miska and Hjort, 2005).
In addition, signal processing techniques, such as Fourier
transforms and wavelet transforms can quantify landslide
morphologic characteristics and highlight patterns of past slope
instability over large areas (Booth et al., 2009).
Much like landslides, rockfalls have been studied using
GIS and DEMs at varying scales (e.g., Guzzetti et al., 2002;
Baillifard et al., 2003; Gunther, 2003; Jaboyedoff et al., 2004;
Derron et al., 2005). Mapping rockfall potential based on
threshold angles at which slopes maintain stability is a relatively simple approach. Angles 4601 (Guzzetti et al., 2003),
4451 (Jaboyedoff and Labiouse, 2003), and 4371 (Frattini
et al., 2008) have been documented, although this depends on
a multitude of variables (e.g., lithology, relief, tectonic setting).
Foliation orientation and major discontinuities can be derived
directly from DEMs, and the spatial arrangement of these
features can be labeled as hazardous structures with a high
probability of falling (Derron et al., 2005). This process has
been enhanced with the development of geomorphometric
software tools that are designed to expedite the analysis of cliff
face and structure (e.g., Gunther, 2003). Finally, GIS-based
rockfall prediction is readily being adopted to forecast where a
rockfall might originate. Spatial modeling approaches included the use of weighting schemes that make use of topographic, geomorphological, and geological criteria involving
faults, scree slopes within a short distance, rocky cliffs, steep
slopes, and road cuts (Baillifard et al., 2003).
3.8.5.4
GIS Modeling
Different spatial modeling approaches can be used for

landslide mapping and analysis. Perhaps the most common
type of analysis is landslide susceptibility (or the relative

spatial probability of occurrence) that relies on probability
theory and fuzzy-set theory (Remondo et al., 2003; Lu
and Rosenbaum, 2003; Tangestani, 2004), and bivariate or
multivariate statistical analyses (Santacana et al., 2003;
Fernandez et al., 2003; Suzen and Doyuran, 2004). The use
of GIS spatial analysis in landslide studies has shifted from
inventorying and mapping to predicting the time, place, and
extent of a mass movement. This includes the trajectory (i.e.,
runout) of an event (Corominas et al., 2003a, b) or the instability of slopes (e.g., slope-safety factor) (Esaki et al., 2001;
Bhattacharya et al., 2003). For example, the rockfall-activity
indicator metric can be used to estimate rockfall-activity
susceptibility, based on climate variables, morphological
factors, and lithological properties (Marqunez et al., 2003).
Artifical neural networks (ANN) have also been used for
mapping (Rumelhardt et al., 1986; Fernandez-Steeger et al.,
2002). Specifically, Ermini et al. (2005) utilized nominally
classified lithology, slope angle, profile curvature, land cover,
and upslope contributing area as input into two different
ANNs (e.g., multilayered perceptron and a probabilistic
neural network). Both ANNs produced satisfactory results,
though validation of the outcome was difficult, as the only
options were a comparison against an inventory map, or to
simply wait and see if the models properly predicted slope
failure. Unfortunately, the complexity of modeling massmovement processes means that many GIS-based models still
oversimplify the terrain characteristics and mechanics that
trigger landslides.
3.8.5.5
GIS-based Landslide Inventories
Landslide-inventory maps provide the occurrence and outlines

of landslides and, in the case of larger scale maps, also classify
landslide types (Carrara and Merenda, 1976; Spiker and Gori,
2000; Owen et al., 2008). In particular, geomorphological
inventory maps can be either a landslide single-event inventory, or a historical landslide inventory, which summarizes
a palimpset landscape comprised of many mass movements
(Malamud et al., 2004). Landslide inventories vary in
scale, including small scale (o1:200 000), medium scale
(1:25 0001:200 000), and large scale (41:25 000; see
Guzzetti et al., 2003). Most inventories produce discrete
boundaries (e.g., crown, depletion zones) of different failures
types (e.g., rockfalls, deep-seated slides, debris flows, etc.) so
that the basic dimensions, such as area or perimeter, can be
automatically calculated. Malamud et al. (2004) warned of the
limitations associated with compiling landslide inventories
using a GIS, including the quality and scale of imagery, morphological and geological complexity, the expertise of the
geomorphologist involved, and the calculations of planar
approximations of 3D features.
3.8.5.6
GIS-based Landslide Assessment Maps
The landslide-hazard map presents the probability (likelihood) of landslide occurrence throughout an area within a
specified period of time (landslide frequency). Such maps
were generated, for example, for rainfall-induced landslides
(Terlien et al., 1995; Coe et al., 2000; Xie et al., 2004) and
earthquake-triggered landslides (Jibson et al., 1998; Refice
and Capolongo, 2002). In contrast, landslide-risk assessment
maps describe the expected annual cost of landslide damage
throughout an area based on the analysis of all possible
consequences. These maps typically require large-scale datasets
(41:25 000) and are displayed in easy to read formats for
dissemination to the general public. Fell (1994) presented
classifications results where colored zones (from green to
red) represent the level of landslide hazard in a specific delineated area (from low to high). It is commonly accepted
that the use of such linguistic expressions creates ambiguity
about the results.
The use of GIT in landslide studies has contributed much
to the knowledge of slope instability, landslide processes, and
landslide-risk assessment. Furthermore, GIT is used to map
morphology, occurrence, and susceptibility in landslide regions in many countries. The rapid developments in geospatial technologies permit new analytical and dissemination
capabilities. For example, field data from landslide sites can be
collected using mobile GIT technology and transferred online
in real time using existing web facilities (Kim et al., 2004;
Chang and Park, 2004), and 3D GIS can help us to better
understand specific landslide types such as rotational slides
(Esaki et al., 2001). Although GIS-based analysis and mapping
can be iterative, the data typically represent cartographic
snapshots in time, and do not characterize dynamic processes
or account for nonlinear variations of change. Furthermore,
existing GIT methods and products can be very subjective
(Hudson, 1992; Budetta et al., 2008; Kamp et al., 2010).
Product quality depends on the skill and experience of the
investigator, the complexity of the study area, and the completeness and reliability of the available data (Galli et al.,
2008).
3.8.6
Fluvial Landscapes and Landforms
Streams and fluvial environments are complex and dynamic

systems with fluctuating flow properties such as discharge,
competence, and power, and landform assemblages such as
channel form, levees, floodplains, and meanders. Geospatial
technologies can help facilitate river restoration and maintenance, inventory aquatic biodiversity and floodplain ecosystems, and quantify river form, process, and rates of
morphological change. Remote surveying techniques and
digital data are valuable for the contemporary fluvial geomorphologist where analyzing and mapping the temporal and
spatial scaling and organization of fluvial processes and
landforms because of the far-reaching capabilities in river
science and rapid development of GIT. As in other fields in
geomorphology, remote sensing investigations are supplanting field-based methods (Carbonneau, 2005) to estimate
stream-water depth (Lyon et al., 1992; Winterbottom and
Gilvear, 1997; Legleiter and Roberts, 2005; Carbonneau,
2005), assess stream habitats (Wright et al., 2000; Whited
et al., 2002; Legleiter et al., 2002; Legleiter and Goodchild,
2005), and estimate the amount of suspended sediment
(Mertes, 2002). Although this shift toward digital data is
progressive, it should not be seen as a replacement for field-
213
based data acquisition. It is important to note that field data

are essential to support or augment imagery and topographic
information (Gilvear and Bryant, 2005).
Many remote sensing studies that investigate inchannel
water properties such as water-light interaction originate from
research in lacustrine or near-shore marine environments.
Within the last decade or two, fluvial geomorphologists have
accelerated their development and application of GIT to
study landscape and water properties, inundation patterns,
and channel form and flow characteristics using a variety of
spectral and spatial perspectives (Mertes, 2002). This includes
the early use of analytical photogrammetry for terrain modeling and the extraction of fluvial features (Lane et al., 1994;
Barker et al., 1997; Dixon et al., 1998; Heritage et al., 1998),
to using more advanced techniques of image processing
(Lyon et al., 1992; Gilvear et al., 1995, 1998; Winterbottom
and Gilvear, 1997; Whited et al., 2002; Marcus et al., 2003;
Westaway et al., 2003; Legleiter et al., 2004; Legleiter and
Roberts, 2005) and laser scanning (Hicks et al., 2002; Thoma
et al., 2005).
Numerous challenges exist, however, when using imagery to
extract information that accurately characterizes the river environment. Within one photograph or image, a river environment will exhibit a wide range of soil moisture conditions and
image textures, including vegetation and land cover variations,
and topographic conditions, all of which influence surface reflectance. Gilvear and Bryant (2005) and Mertes (2002) provide
guidelines on selecting appropriate sensors (Table 1) and discuss the diverse matter/energy interactions associated with
stream-system surfaces. The challenges associated with remote
sensing water-rich environments are amplified when trying to
penetrate water to study stream bathymetry (e.g., reflectance of
shallow, clear water may be driven by reflectance of bed sediments and surface roughness) and substream landforms, or
when attempting to link stream dynamics with landforms, such
as distinguishing between pools, glides, and riffles, extracting
data about sediment concentration, amount of organics,
and surface roughness. The objective here is to briefly discuss
data sources, processing, and visualization techniques, along
with recent applications that are most widely used in fluvial
geomorphology.
3.8.6.1
Aerial Photographs
Aerial photograph interpretation in geomorphology requires

an understanding of photographic scale. Small-scale aerial
photographs lack detail but cover large areas, whereas largescale photos offer great detail with limited spatial coverage. As
a result, many high-resolution stream studies that depend on
remote sensing only cover limited areas or stream reaches,
which tend to characterize streams as discontinuous systems,
even though it is understood that streams vary continuously at
different scales (Marcus and Fonstad, 2008). Therefore, the
potential of aerial photography is maximized when, among
other things, sensor resolution and spatial scale of landform
and fluvial process are matched. For example, small stream
reaches (stream channel width o20 m) are best mapped and
studied by combining field-based techniques with large-scale
aerial photography (1:2500 or better), or even hand-held
214
cameras. Most analyses of medium sized rivers (channel width

between 20 and 200 m) tend to use airborne remote
sensors and relatively large-scale aerial photography (1:5000
to 1:25 000), whereas the characteristics of large rivers (width
4200 m) should be observed using space-borne sensors
(Gilvear and Bryant, 2005).
Photographic and digital stereoimagery can produce 3D
perspectives on fluvial environments, revealing topographic
variation. Although it is relatively straightforward to interpret
above-water landforms and channel characteristics based on
data derived from aerial photographs, such as studies of
braided, gravel-bed rivers (Westaway et al., 2003), and twoand 3D channel morphology (Winterbottom, 2000), the development of refraction-correction algorithms have opened
new opportunities to study the ongoing evolution of substream landforms, as it relates to sediment characteristics and
channel shape (Lane et al., 2000; Westaway et al., 2000; Butler
et al., 2002). If stereophotographs are merged with calibrating
camera data and ground-control points, the results are highaccuracy DEMs, which are most effective in open (i.e., limited
shadow interference), relatively low-relief, shallow and
clear-watered, and coarse-grained braided stream systems
(Westaway et al., 2001, 2003). Clear water allows for greater
penetration of electromagnetic energy, as the percent of organic material increases energy absorption, and gravel-sized
substrate increases reflectivity. This results in an accurate
rendering of streambed depth below water, although these
characteristics are associated with stream discharge and sediment concentration (which fluctuate over time and along
stream length). A sequence of these high-resolution DEMs
that are at times subdecimeter could prove useful when
quantifying streambed morphology and channel migration
(Winterbottom and Gilvear, 1997). Again, successful use of
aerial imagery requires a clean line of site (e.g., avoiding
bridges, shadows), thus photogrammetry is somewhat limited
in headwater streams or on streams that have densely vegetated riparian zones with obstructed views of the water surface,
or even shadows.
Qualitative interpretations derived from single or mosaicked photos are a simple, but effective, method to measure
sinuosity and changes in channel width, so long as the photo
scale is known and consistent. Most applications of aerial
photography include plan-view measurements of bankfull
width, river sinuosity, and lateral channel migration (e.g.,
Brooks and Luft, 1987; Whitesell et al., 1988; Marston et al.,
1995; Leys and Werrity, 1999; Winterbottom and Gilvear,
2000; OConnor et al., 2003; Mount and Louis, 2005; Hughes
et al., 2006). Temporal sequences of aerial photos, frequently
spanning decades, reveal stream channel change (e.g., meander migration) over time (Brice, 1977; Gurnell et al., 1994;
Gurnell, 1997), potentially providing baseline data for stream
monitoring and management (Micheli and Kirchner, 2002).
For example, digitized stream centerlines, taken from a sequence of aerial photos, can be overlayed in GIS to analyze
shifts in channel location, and then relate this to zones of
wetness to predict stream-bank erosion potential. Micheli and
Kirchner (2002) calculated eroded area polygons of aerial
photos in Arc/Info to quantify the distance the channel migrated perpendicular to a channel centerline. They concluded
stream-banks with dry vegetation (i.e., vegetation that thrives
in relatively dry conditions) migrate faster than those with wet

vegetation, but also emphasized the value of sequential aerial
photos and geoprocessing tools in GIS in quantifications
of stream-channel migration over time. Photogrammetry of
stream-banks also offers an objective method for calculating
volume of sediment loss due to erosion, as a means of
monitoring stream-bank stability, but also to determine the
relative importance of stream-banks as sediment sources
(Barker et al., 1997).
Due to the desire for fine-resolution, fluvial geomorphologists now consider geospatial error as a necessary aspect of
analysis for the successful application of aerial photography to
study fluvial environments. The positional errors associated
with aerial images and DEMs are just now being recognized as
a critical influence on lateral two-dimensional measurements
from aerial photographs. Error analysis of data is now being
rigorously investigated to provide a conceptual foundation for
future stream studies. This includes testing the number and
type of ground-control points, interpolation issues, the practicality of using simple metrics like root-mean-square error
(RMSE), and the impact that positional error has on stream
studies (e.g., Mount et al., 2003; Mount and Louis, 2005;
Hughes et al., 2006). Finally, many countries or agencies have
extensive libraries of aerial photographs, and using modern
photogrammetric methods, some of these archived images can
still be used to reevaluate river dynamics and change over the
past 50 years or more (Marcus and Fonstad, 2008; Lane et al.,
2010).
3.8.6.2
Airborne And Satellite Sensors
Early applications of multispectral data on river systems focused on estimating suspended sediment concentrations in
relatively large rivers using Landsat, Coastal Zone Color
Scanner (CZCS), and AVHRR data (e.g., Amos and Alfoldi,
1979; Li, 1993; Mertes et al., 1993). For example, Landsat TM
has a water penetration band (0.450.54 mm) that can
penetrate 1020 m of clear water. This allows for bathymetric
mapping of lakes and rivers (the optimum wavelength is
0.48 mm), but the utility of Landsat data is limited by spatial
resolution, which is too large to adequately sample small
and many medium-sized streams. With specific wavelengths,
suspended-sediment texture in rivers can be distinguished
(0.580.69 mm), whereas it is possible to use near infrared
wavelengths (0.7140.880 mm) to estimate the amount of
suspended sediments (Han and Rundquist, 1994). Therefore,
advances in sensor technology (i.e., hyperspectral sensors) and
data availability have exponentially increased the possibilities
to generate deeper insights into the relationships between
stream dynamics and landform evolution. Although digitally
acquired data is used to characterize stream properties and
landforms, remote-sensing applications are still limited to case
studies, rather than operational assessment, or comprehensive
assessment of entire watershed systems.
Hyperspectral data, collected from satellites or aircraft,
cameras suspended over rivers, or hand-held spectrometers, can
be used to elucidate the physical conditions and processes in
stream environments, as biophysical conditions and processes
influence matter/energy interactions. Legleiter et al. (2004)
combined field-derived spectral data with radiative-transfer

modeling to illustrate the effect stream depth and substrate
characteristics, water-surface roughness, and suspended-sediment concentration have on spectra. Fonstad and Marcus
(2005) merged airborne hyperspectral images with local
stream-gage data and hydrologic principles such as Mannings
equation, to develop the hydraulically assisted bathymetry
(HAB) model. The HAB model generates stream-depth maps
based on image brightness, and the authors demonstrated the
capability to integrate a wide range of sensor data, including
airborne multispectral and space-borne imagery such as
IKONOS. They also demonstrate the use of archived imagery
to map 3D changes in river bathymetry. As is the case with
aerial photography, the use of these images is limited to water
turbidity conditions and a clear line-of-sight.
3.8.6.3
Active Sensors
Although the use of hyperspectral data has been explored in

more detail in river science, a number of surface hydrologists
and fluvial geomorphologists are using active sensors to map
stream bottom, stage height, and discharge (Spicer et al., 1997;
Costa et al., 2000; Jones et al., 2007). For example, Jung et al.
(2010) characterized and compared floodplains in the
Amazon and Congo using InSAR from the Japanese Earth
Resources Satellite-1 (JERS-1). They used the spatial and
temporal variations in water-surface levels to describe complex
fluvial processes and describe floodplain-building features
between two different fluvial systems. One reason for this
work was to illustrate the value of, and contribute to the design of a future radar satellite system that aims to improve
global water monitoring (e.g., Surface Water and Ocean
Topography (SWOT) Mission). Similar research has also relied
on radar altimetry to estimate stream discharge, slope, and
water-surface topography (Koblinsky et al., 1993; Birkett et al.,
2002; Kouraev et al., 2004; Frappart et al., 2008; Calmant and
Seyler, 2006; Leon et al., 2006), even though it is recommended that field-based data still support interpretations from
remote sensing investigations (e.g., cross-sectional profiles).
Laser ranging systems have proven to be a valuable technology that can be used to monitor subtle landform changes
and measure ongoing fluvial processes such as stream-bank
erosion (Thoma et al., 2005; Notebaert et al., 2009). The detection of depressions and subtle changes in topography, especially with regards to small channel forms, is facilitated by
the fine resolution typical of LiDAR data (Figure 18). Much
like aerial photography, sequential LiDAR data provides a
temporal perspective on stream processes, including change
migration and stream-bank erosion (Notebaert et al., 2009).
Although LiDAR data are not readily available for many
watersheds, and data-collection missions are expensive, LiDAR
accuracy permits mapping of paleochannels and other lowrelief landforms, and facilitates hydraulic modeling efforts
(e.g., Cobby et al., 2001; French, 2003; Dal Cin et al., 2005)
and studies on morphological change (e.g., Lohani and
Mason, 2001; Challis, 2006; Challis and Howard, 2006;
Nelson et al., 2006). Furthermore, the signal penetrates vegetation and LiDAR data can be acquired at all times of the year
in some of the most inaccessible fluvial environments in the
215
world. Despite the high level of topographic detail produced

from LiDAR data, field-based data such as coring and groundpenetrating radar are still frequently required to properly interpret the importance of mapped features (Notebaert et al.,
2009).
Floodplains, especially those adjacent to large stream
systems (e.g., Amazon, Mississippi), contain an array of
landforms indicative of fluctuations in stream flow and watertable elevation (i.e., hydroperiod). Yet, despite being dynamic
Earth surface features, the relative relief within a floodplain
can be quite small. To address this limitation, and considering
the dynamic nature of floodplains, geomorphic studies now
integrate data from various sources, at various scales, to systematically delineate landforms in floodplain environments.
Floodplain features have been distinguished using satellite
imagery (FCC 5, 4, 3 from Landsat ETM and the L-band of
JERS-1) and topographic information (Figure 19; Hamilton
et al., 2006). The high backscatter on radar imagery clearly
denoted flooded forests, whereas Landsat imagery characterized the meander melts. An object-based analysis then simultaneously analyzed image elements and topography to
produce homogeneous terrain objects, which were then hierarchically classified. Unfortunately, most of the imagery did
not capture flooding events, meaning that some parts of the
classification (i.e., flooded terrain) relied more on hypothesized classes.
Mapping by Marston et al. (1995) focused on the interaction between the stream channel, floodplain, and biodiversity. They produced a landscape change matrix that
captured channel migration and floodplain disturbance along
the Ain River, France between 1945 and 1991 (Figure 20).
Although their goal was more focused on understanding
vegetation development, the spatial distribution of river deposits and landforms (i.e., landscape units) reflected the impact humans have on channel migration (e.g., dams, increased
summer discharge to support salmon) and floodplain disturbance. Changes in channel location and behavior then
caused changes in the distribution and diversity of vegetation,
as a decrease in floodplain disturbance increased the growth of
larger shrubs and trees, whereas areas once dominated by
hydrophytic plants were replaced with mixed forests and
hardwood trees.
3.8.6.4
Fluvial geomorpholgists have integrated topographic, soil,

lithology, and precipitation data into GIS databases to systematically evaluate the driving forces behind stream processes, or to piece together the evolution of fluvial landforms.
Jordan and Fonstad (2005) estimated water-surface slope and
average depth within a GIS so they could use Mannings
equation to estimate stream velocity, discharge, and power for
the Brazos River in Texas, U.S. Finlayson and Montgomery
(2003) analyzed the influence of DEM resolution and map
projection on predictions of erosion in large, mountainous
stream systems (e.g., Olympic Mountain Range, U.S.). They
concluded that DEM resolution and map projection can
distort area and stream length (typically minimizing), areadischarge relationships may be relatively weak when involving
216
169 000
169 200
169 000
169 200
169 000
169 200
0.1 km
162 200
162 200
> 35 m
162 000
162 000
30 m
(b)
162 000
162 000
162 200
162 200
(a)
(c)
169 000
169 200
(d)
Figure 18 Comparison of two LiDAR datasets in Belgium (Dijle and Amble`ve). (a) Flemish LiDAR at a resolution of 5 m; (b) Walloon LiDAR at a
resolution of 1 m; (c) Hill-shade map of F-LiDAR data (azimuth: 901); (d) Hill-shade map of the W-LiDAR data (azimuth: 901). Modified from
Notebaert, B., Verstraeten, G., Govers, G., Poesen, J., 2009. Qualitative and quantitative applications of LiDAR imagery in fluvial geomorphology.
Earth Surface Processes and Landforms 34, 217231.
multiple climatic zones, and the stream-power law is sensitive

to DEM resolution (e.g., stream-power per unit area decreased
as DEM resolution decreased). Therefore, numerous methodological issues must be considered before using GIS-based
erosion models to predict denudation for large mountainous
stream systems, which are complicated by topographic variability, changes in lithology, and abrupt changes in climate.
Nevertheless, such investigations have revealed the value of
integrating data to elicit key factors that influence topographic
evolution in large stream systems, a task that would otherwise
be difficult to complete using more traditional mapping
methods.
Geomorphologists can derive and extract fluvial features
from DEMs, such as stream networks, watershed boundaries,
flow accumulation grids, and nodes or confluences of tributaries. These characteristics are based upon the use of geomorphometric parameters that include slope azimuth and
curvature, whereas connectivity, or topology, predict flow
paths and relationships between stream channels. So common
are these applications that many software programs offer

functions that are specific to stream research (e.g., the Hydro
toolbox in ArcGIS, IDRISI, River Tools).
Advances in remote sensing and GIS provide the ability to
measure river-system behavior at various spatial and temporal
scales. Relative to most field-based techniques, remote surveying is the only practical method to monitor and map fluvial environments, especially considering the high-resolution
imagery and DEMs, the temporal frequency of airborne
and spaceborne imagery, and the numerous regions of the
electromagnetic spectrum that can be used to assess the biophysical characteristics of inundated or floodplain environments. Furthermore, topographic information is now being
integrated with spectral data, significantly altering how river
scientists view river-system components and behavior, which
in turn impacts our dependence on GIT for studying fluvial
environments.
The contributions of remote sensing in river science are
substantial. New data and information are being used as input
217
25 km
River channels (open water)
Meander belt (early succession herbaceous vegetation)
Palm swamp (dry)
Lakes (mostly oxbow)
Meander belt (late successional broadleaf canopy)
Palm swamp (flooded)
River bars (seasonally exposed gravel, sand or mud)
Backswamp (low-lying forest outside meander belts)
Clouds & shadows
Figure 19 Image analysis of the Madre de Dios river system in Peru based on a Landsat ETM FCC (5, 4, 3) (left), and a floodplain
classification based on SRTM, Landsat ETM , and JERS-1 (right). Modified from Hamilton, S.K., Kellndorfer, J., Lehner, B., Tobler, M., 2006.
Remote sensing of flood plain geomorphology as a surrogate for biodiversity in a tropical river system (Madre de Dios, Peru). Geomorphology
89, 2338.
Unvegetated gravel-cobble deposits

Sand-gravel-cobble deposits (low shrubs)
Sand-gravel-cobble deposits (high shrubs)
Old channels
Dense, diverse, mesophytic shrubs
Mixed forest, dominated by hardwoods
Dry grassland on sandy soils (few shrubs)
Dry grassland on gravel-cobble deposits
Cleared land for mines, campgrounds, agriculture
Water
0
0.5
1.0 km
Figure 20 Landscape unit maps for a portion of the Ain River 100 year floodplain near Mollon, France between 1945 (left) and 1991 (right).
Note reduction in meandering and increase in vegetation that requires drier conditions (i.e., reduced floodplain disturbance). Modified from
Marston, R.A., Girel, J., Pautou, G., Piegay, H., Bravard, J.P., Arneson, C., 1995. Channel metamorphosis, floodplain disturbance, and vegetation
development: Ain River, France. Geomorphology 13, 121131.
218
for hydrological modeling (spatial hydrology), permit detection and mapping of previously undetected landforms in
densely vegetation floodplains, enable the generation of
quantitative estimates of stream characteristics such as discharge and power, permit the monitoring of stream-channel
migration, and enable the characterization of habitat biodiversity based upon stream conditions. Based on these advances in river science, some have argued for permanent
remote-sensing river observatories (Marcus and Fonstad,
2008), which would provide standardized imagery for a host
of users, while minimizing the impact of ongoing privatization of remote-sensing technology and data that is of interest
to geomorphologists.
3.8.7
Conclusion
Rapid advances in remote sensing and GIT have now significantly altered the way that geomorphologists analyze data and
study processes and system dynamics. Continued multidisciplinary investigations of landform mapping that combine
the theoretical underpinnings that are fundamental to geomorphology and GIScience will speed-up and advance our
capabilities for improved mapping and assessment of geomorphological systems. The relative ubiquity of spectral data
from unique regions of the electromagnetic spectrum provide
us with a variety of biophysical information that permit innovative perspectives to map, extract, and measure landforms.
Advances in representation, geomorphometry, and AI continue to alter how geomorphic concepts are implemented in
the software domain, and there are mutual challenges common to geomorphology and GIScience, such as space-time
and multidimensional representation, characterization of
spatial properties and topological relationships, and predictive
modeling and mapping, that ensure that GIT will continue
contributing to the field of geomorphology for years to come.
Although there are other developments in geospatial
technologies related to landform mapping not covered in this
review (e.g., aeolian, coastal, and tectonic geomorphology),
several issues and future directions have emerged that include:
Scale is critical in landform identification and classification,

including the scaling relationships between land-surface
geometric properties and landform size (Walsh et al., 1998;
Evans, 2000; Bishop et al., 2012). It is now possible to estimate the scale at which the spatial organization of the
topography best characterizes landforms (e.g., Eisank et al.,
2010), which can be used to guide computational scale selection and methodological design and interpretation. In
particular, multiscale analysis approaches can be used to
determine the amount and patterns of spatial autocorrelation for various landscape properties. Moreover, it is important to consider temporal scales as well, given that many
studies are focusing on characterizing the operational scale
of processes and landform evolution.
Geomorphologists are also beginning to recognize the importance of semantics in landform classification. Eisank et al.
(2010) stressed that semantic modeling offers one possible
solution for relating a particular landform concept to representation, analysis, and visualization using GIT. Furthermore,
semantic modeling requires a solid knowledge-based classification scheme, as noted by Raseman et al. (2004), Dunlop
and Clark (2006), and Grosse et al. (2009). This will require a
rigorous, universal, and quantitative treatment of landform
properties that includes morphology in order to formalize
landform classification relationships and hierarchies (Deng,
2007). Thus, the merging of GIScience into geomorphology is
forcing geomorphologists to revisit landform taxonomic
schemes, as GIScience investigations attempt to standardize
geomorphological mapping approaches based upon a scientific treatment of geomorphological concepts and information requirements.
While it is acknowledged that automating landform identification, extraction, and classification is an ultimate goal in
geomorphology, this continues to be a challenge for geomorphologists. Klingseisen et al. (2008) customized a GIS to
develop LANDFORM, a semiautomated classification of
landform elements (e.g., crest, flats, depressions). Based on
the semiautomated derivation of these elements, a landscape
could potentially be divided into individual morphological types. Although this particular automation was designed for hydrological structures and drainage patterns,
landform characterization based on basic landform element
properties most likely will contribute to the automation of
landform identification and classification in other fields of
geomorphology.
Distinguishing spatial and temporal patterns has and continues to dominate landform-mapping research. This usually
requires the design of spatial metrics or indices that quantify
morphological conditions (Bishop et al., 2012), and composition and spatial patterns of landscape units (Walsh et al.,
1998). Many of these can be found in the literature, and some
are built into customized programs or software packages.
Landform patterns, however, are usually a remnant of a particular process (e.g., subglacial conditions and drumlins),
such that pattern identification and measurement contribute
to our understanding of the dominant processes responsible
for landscape evolution.
The need to access various forms of data and disseminate
information has grown rapidly within the past decade. Webbased software tools can now be used for sharing data (e.g.,
TM
Google Earth , NASA World Wind) and permit advanced
cartographic functionalities. The capability to distribute preand postprocessed images and data means that reporting
sources, methods, and positional accuracies should now
become mandatory, so as to reduce error propagation between studies (Napieralski et al., 2007b) and to develop
universally-accepted methodological protocols for landform
mapping.
Remote sensing and GIT have greatly increased objectivity
and efficiency of geomorphological investigations, but just as
important, it is forcing us to merge different theoretical perspectives on space, time, processes, and systems. As illustrated
in Figure 1, the ability to formally segment landscapes into
landforms, and then align the morphological characteristics
with process mechanics, depends on the quality of data, range
of perspectives (e.g., topography, spectral, temporal), and a
much needed geologic contextualization of the area. Landform classifications continue to be modified, as the inclusion
of more modern data and the use of new analytical techniques

(e.g., fuzzy logic, semivariance) reveal previously unrecognized characteristics. Such progress will continue, although the use of GIScience in geomorphology requires the
user to have experience with an ever-expanding profusion of
digital data, software interfaces, geostatistical tools, spatial
models, visualization approaches, and now space-time theory.
Regardless, it is now almost standard practice to characterize
landforms using geomorphometry and GIS, and this has made
significant contributions to our understanding of the topographic, climatic, biological, and geologic processes responsible for shaping Earths surface.
References
Alcantara-Ayala, I., Esteban-Chavez, O., Parrot, J.F., 2006. Landsliding related to
land-cover change: a diachronic analysis of hillslope instability distribution in
the Sierra Norte, Puebla, Mexico. Catena 65, 152165.
Alexander, D.E., 2008. A brief survey of GIS in mass-movement studies, with
reflections on theory and methods. Geomorphology 94, 261267.
Amos, C.L., Alfoldi, T.T., 1979. The determination of suspended sediment
concentration in a macrotidal system using Landsat data. Journal of Sedimentary
Petrology 49, 159174.
Anders, N., Seijmonsbergen, A., Bouten, W., 2009. Multi-scale and object-oriented
image analysis of high-red LiDAR data for geomorphological mapping in Alpine
mountains. In: Purves, R., Gruber, S., Straumann, R., Hengl, T. (Eds.),
Proceedings of Geomorphometry 2009. University of Zurich, Zurich, pp. 6165.
Andrews, J.T., 1982. On the reconstruction of Pleistocene ice sheets: a review.
Quaternary Science Reviews 1, 130.
Aniya, M., Etaya, M., Shimoda, H., 1985. Evaluation of Landsat data for landslide
identification as a means for watershed management. Journal of the Japan
Society of Photogrammetry and Remote Sensing 24, 1721.
Aniya, M., Naruse, R., Yamaguchi, S., 2002. Utilization of 6 6 cm format vertical
aerial photographs for repetitive mapping of surface morphology and
measurement of flow velocities of a small glacier in a remote area: Glacier Soler,
Hielo Patagonico Norte, Chile. Annals of Glaciology 34, 385390.
Aniya, M., Sato, H., Naruse, R., Skvarca, P., Casassa, G., 1996. The use of satellite
and airborne imagery to inventory outlet glaciers of the Southern Patagonia
Icefield, South America. Photogrammetric Engineering & Remote Sensing 62,
13611369.
Aniya, M., Welch, R., 1981. Morphometric analyses of Antarctic cirques from
photogrammeric measurements. Geografiska Annaler 63A, 4153.
Aste, J.P., Girault, F., 1995. GIS, SPOT, DEM and morphology of major land
movements. Proceedings of the International Symposium on Landslides 6,
15391545.
Baeza, R.C., Esteban-Chavez, O., Marcos-Lopez, J., Pena-Garnica, R.J., AlcantaraAyala, I., 2006. Slope instability on pyroclastic deposits: landslide distribution
and risk mapping in Zacapoaxtla, Sierra Norte de Puebla, Mexico. Journal of
Mountain Science 3, 119.
Baillifard, F., Jaboyedoff, M., Sartori, M., 2003. Rockfall hazard mapping along a
mountainous road in Switzerland using a GIS-based parameter rating approach.
Natural Hazards in Earth System Science 3, 435442.
Baldi, P., Cenni, N., Fabris, M., Zanutta, A., 2008. Kinematics of a landslide derived
from archival photogrammetry and GPS data. Geomorphology 102, 435444.
Ballantyne, C.K., 1997. Periglacial trimlines in the Scottish Highlands. Quaternary
International 38/39, 119136.
Ballantyne, J.D., Dean, Jr. D.R., Thompson, B.L., 1987. Monitoring landslide
movement with a 35-mm camera. Transportation Research Record 1119, 4754.
Bamber, J.L., Hardy, R.J., Joughin, I., 2000. An analysis of balance velocities over
the Greenland ice sheet and comparison with synthetic aperture radar
interferometry. Journal of Glaciology 46, 6774.
Barker, R., Dixon, L., Hooke, J., 1997. Use of terrestrial photogrammetry for
monitoring and measuring bank erosion. Earth Surface Processes and
Landforms 22, 12171227.
Barlow, J., Martin, Y., Franklin, S.E., 2003. Detecting translational landslide scars
using segmentation of Landsat ETM and DEM data in the northern Cascade
Mountains, British Columbia. Canadian Journal of Remote Sensing 29,
510517.
219
Barr, I.D., Clark, C.D., 2009. Distribution and pattern of moraines in Far NE Russia
reveal former glacial extent. Journal of Maps 2009, 186193.
Bastero, C.F., Lagmay, A., 2006. Estimating SiO2 content of lava deposits in the
humid tropics using remotely sensed imagery. Journal of Volcanology and
Belknap, E.P., Gilmore, J.B., 1987. Use of time-lapse movie photography in
landslide monitoring. Transportation Research Record 1119, 3946.
Bencardino, M., 2005. Landslide hazards and risk assessment at the Basento river
basin using Landsat satellite imagery. Bollettino Della Societa Geografica Italiana
12, 115130.
Bhattacharya, G., Jana, D., Ojha, S., Chakraborty, S., 2003. Direct search for
minimum reliability index of earth slopes. Computer and Geotechnics 30,
455462.
Bindschandler, R.A., Scambos, T.A., 1991. Satellite-image-derived velocity field of
an Antarctic ice stream. Science 252, 242246.
Birkett, C.M., Mertes, A.A.K., Dunne, T., Costa, M.H., Jasinski, M.J., 2002. Surface
water dynamics in the Amazon Basin: application of satellite radar altimetry.
Journal of Geophysical Research 107. http://dx.doi.org/10.1029/2001JD000609.
and digital geomorphological mapping: concepts, issues and research.
Bishop, M.P., Shroder, Jr. J.F., 2004. GIScience and mountain geomorphology:
overview, feedbacks, and research directions. In: Bishop, M.P., Shroder, J.F.
(Eds.), Geographic Information Science and Mountain Geomorphology. SpringerPraxis, New York, pp. 131.
Bishop, M.P, Shroder, Jr. J.F., Colby, J.D., 2003. Remote sensing and
Bishop, M.P, Shroder, Jr. J.F., Hickman, B.L., Copland, L., 1998. Scale-dependent
analysis of satellite imagery for characterization of glacier surfaces in the
Karakoram Himalaya. Geomorphology 21, 217232.
Bishr, Y., 1998. Overcoming the semantic and other barriers to GIS interoperability.
Booth, A.M., Roering, J.J., Perron, J.T., 2009. Automated landslide mapping using
spectral analysis and high-resolution topographic data: Puget Sound lowlands,
Washington, and Portland Hills, Oregon. Geomorphology 109, 132147.
Boulton, G.S., Clark, C.D., 1990. A highly mobile Laurentide ice sheet revealed by
satellite images of glacial lineations. Nature 346, 813817.
Bozzano, F., Mazzanti, P., Prestininzi, A., 2008. A radar platform for continuous
monitoring of a landslide interacting with an under-construction infrastructure.
Italian Journal of Engineering Geology and Environment 8, 7187. (In Italian).
Bradwell, T., Stoker, M., Krabbendam, M., 2008. Megagrooves and streamlined
bedrock in NW Scotland: the role of ice streams in landscape evolution.
Brice, J., 1977. Lateral Migration of the Middle Sacramento River, California.
U.S. Geological Survey, Water Resources Investigation, Sacramento, CA,
pp. 7743.
Bronge, L.B., Bronge, C., 1999. Ice and snow-type classification in the Vestfold
Hills, East Antarctica, using Landsat-TM data and ground radiometer
measurements. International Journal of Remote Sensing 20, 225240.
Brook, G.A., Luft, E.R., 1987. Channel pattern changes along the lower Oconee
River, Georgia, 1805/7 to 1949. Physical Geography 8, 191209.
Brueckl, E., Brunner, F.K., Kraus, K., 2006. Kinematics of a deep-seated landslide
derived from photogrammetric, GPS and geophysical data; remote sensing and
ground-based geophysical techniques for recognition, characterisation and
monitoring of unstable slopes. Engineering Geology 88, 149159.
Brunsden, D., 1993. The persistence of landforms. Zeitschrift fur Geomorphologie
Supplementband 93, 1328.
Budetta, P., Santo, A., Vivenzio, F., 2008. Landslide hazard mapping along the
coastline of the Cilento region (Italy) by means of a GIS-based parameter rating
approach. Geomorphology 94, 340352.
Burbank, D.W., Anderson, R.S., 2001. Tectonic Geomorphology. Wiley-Blackwell,
Hoboken, NJ, pp. 1274.
Burrough, P.A., van Gaans, P.F.M., MacMillan, R.A., 2000. High-resolution
landform classification using fuzzy k-means. Fuzzy Sets and Systems 113,
3752.
Butler, D., 2001. Geomorphic process-disturbance corridors: a variation on a
principle of landscape ecology. Progress in Physical Geography 25, 237248.
Butler, J.B., Lane, S.N., Chandler, J.H., Porfiri, K., 2002. Through-water close range
digital photogrammetry in flume and field environments. Photogrammetric
Record 17, 419439.
Calmant, S., Seyler, F., 2006. Continental surface waters from satellite altimetry.
Comptes Rendus Geoscience 338, 111122.
220
Carbonneau, P.E., 2005. The threshold effect of image resolution on image-based

automated grain size mapping in fluvial environments. Earth Surface Processes
Carn, S.A., 1999. Application of Synthetic Aperture Radar (SAR) imagery to volcano
mapping in the humid tropics: a case study in East Java, Indonesia. Bulletin of
Volcanology 61, 92105.
Carnec, C., Massonet, D., King, C., 1996. Two examples of the use of SAR
interferometry on displacement fields of small spatial extent. Geophysical
Research Letters 23, 35793582.
Carrara, A., Guzzetti, F. (Eds.), 1995. Geographical Information Systems in
Assessing Natural Hazards. Kluwer, Dordrecht, 353 pp.
Carrara, A., Merenda, L., 1976. Landslides inventory in northern Calabria, southern
Italy. Geological Society America Bulletin 87, 12291246.
Carrara, A., Pike, R.J., 2008. GIS technology and models for assessing landslide
hazard and risk. Geomorphology 94, 257260.
Catani, F., Farina, P., Moretti, S., Nico, G., Strozzi, T., 2005. On the application of
SAR interferometry to geomorphological studies: estimatation of landform
attributes and mass movements. Geomorphology 66, 119131.
Chacon, J., Irigaray, C., Fernandez, T., El Hamdouni, R., 2006. Engineering geology
maps: landslides and geographical information systems. Bulletin Engineering
Geology and Environment 65, 341411.
Chadwick, J., Dorsch, S., Glenn, N., Thackray, G.D., Shilling, K., 2005. Application
of multitemporal high-resolution imagery and GPS in a study of the motion of a
canyon rim landslide: remote sensing and geospatial information for natural
hazards characterization. Journal of Photogrammetry and Remote Sensing 59,
212221.
Challis, K., 2006. Airborne laser altimetry in alluviated landscapes. Archaeological
Prospection 13(2), 103127.
Challis, K., Howard, A.J., 2006. A review of trends within archaeological remote
sensing in alluvial environment. Archaeological Prospection 13(4), 231240.
Chandler, J., Cooper, M., 1988. Monitoring the development of landslides using
archival photography and analytical photogrammetry. Land and Minerals
Surveying 6, 576584.
Chandler, J.H., Fryer, J.G., Jack, A., 2005. Metric capabilities of low-cost digital
cameras for close range surface measurement. Photogrammetric Record 20,
1226.
Chang, Y.S., Park, H.D., 2004. Development of a web-based geographic information
system for the management of borehole and geological data. Computer and
Charlesworth, J.K., 1928. The glacial geology of North Mayo and West Sligo.
Proceedings of the Royal Irish Academy 38, 100115.
Chen, R.-F., Chang, K.-J., Angelier, J., Chan, Y.-C., Deffontaines, B., Lee, C.-T., Lin,
M.-L., 2006. Topographical changes revealed by high-resolution airborne LiDAR
data: the 1999 Tsaoling landslide induced by the Chi-Chi earthquake.
Chiang, S.-H., Chang, K.-T., 2009. Application of radar data to modeling rainfallinduced landslides. Geomorphology 103, 299309.
Chirico, G.D., et al., 2009. Lava flow hazard at Nyiragongo Volcano, DRC. Bulletin
of Volcanology 71(4), 375387.
Chorley, R.J., Schumm, S.A., Sudgen, D.E., 1984. Geomorphology. Methuen,
London.
Chou, W.-C., Lin, W.-T., Lin, C.-Y., 2009. Vegetation recovery patterns assessment
at landslides caused by catastrophic earthquake: a case study in central Taiwan.
Environmental Monitoring and Assessment 152, 245257.
Claessens, L., Heuvelink, G.B.M., Schoorl, J.M., Veldkamp, A., 2005. DEM
resolution effects on shallow landslide hazard and soil redistribution modelling.
Clark, C.D., 1993. Mega-scale glacial lineations and cross-cutting ice-flow
landforms. Earth Surface Processes and Landforms 18, 129.
Clark, C.D., 1997. Reconstructing the evolutionary dynamics of former ice sheets
using multitemporal evidence, remote sensing and GIS. Quaternary Science
Reviews 16, 10671092.
Clark, C.D., Evans, D.J., Khatwa, A., et al., 2004. Map and GIS database of
glacial landforms and features related to the last British Ice Sheet. Boreas 33,
359375.
Clark, C.D., Knight, J.K., Gray, J.T., 2000. Geomorphological reconstruction of the
Labrador sector of the Laurentide Ice Sheet. Quaternary Science Reviews 19,
13431366.
Clark, C.D., Meehan, R.T., 2001. Subglacial bedform geomorphology of the Irish Ice
Sheet reveals major configuration changes during growth and decay. Journal of
Quaternary Science 16, 483496.
Clark, C.D., Wilson, C., 1994. Spatial analysis of lineaments. Computers &
Close, M.H., 1867. Notes on the general glaciation of Ireland. Journal of the Royal
Geological Society of Ireland 1, 207242.
Cobby, D.M., Mason, D.C., Davenport, I.J., 2001. Image processing of airborne
scanning laser altimetry data for improved river flood modeling. ISPRS Journal
of Photogrammetry and Remote Sensing 56(2), 121138.
Coe, J.A., Michael, J.A., Crovelli, R.A., Zavage, W.A., 2000. Preliminary map
showing landslides densities, mean recurrence intervals, and exceedance
probabilities as determined from historic records, Seattle, Washington. USGS
Open-File report 00-303.
Colesanti, C., Wasowski, J., 2006. Investigating landslides with space-borne
synthetic aperture radar (SAR) interferometry: remote sensing and ground-based
geophysical techniques for recognition, characterisation, and monitoring of
unstable slopes. Engineering Geology 88, 173199.
Coppock, J.T., 1995. GIS and natural hazards: an overview from a GIS perspective.
In: Carrara, A., Guzzetti, F. (Eds.), Geographical Information Systems in
Assessing Natural Hazards. Kluwer, Dordrecht, pp. 2134.
Corazzato, C., Tibaldi, A., 2006. Fracture control on type, morphology and
distribution of parasitic volcanic cones: an example from Mt. Etna. Italy Journal
of Volcanology and Geothermal Research 158(12), 177194.
Corominas, J., Copons, R., Vilaplana, J.M., Altimir, J., Amigo, J., 2003b. Integrated
landslide susceptibility analysis and hazard assessment in the principality of
Andorra. Natural Hazards 58, 421435.
Corominas, J., Santacana, N., 2003a. Stability of the Vallcebre translational
slide, Eastern Pyrenees (Spain) by means of a GIS. Natural Hazards 58,
473485.
Corsini, A., Borgatti, L., Cervi, F., Dahne, A., Ronchetti, F., Sterzai, P., 2009.
Estimating mass-wasting processes in active earth slides: earth flows with timeseries of high-resolution DEMs from photogrammetry and airborne LiDAR.
Natural Hazards and Earth System Sciences 9, 433439.
Costa, J.E., Spicer, K.R., Cheng, R.T., et al., 2000. Measuring stream discharge by
non-contact methods: a proof of concept experiment. Geophysical Research
Letters 27(4), 553556.
Crowley, J., Zimbelman, D., 1997. Mapping hydrothermally altered rocks on Mount
Rainier, Washington, with Airborne Visible Infrared Imaging Spectrometer
(AVIRIS) data. Geology 25, 559569.
Czuchlewski, K.R., Weissel, J.K., Kim, Y., 2003. Polarimetric synthetic aperture radar
study of the Tsaoling landslide generated by the 1999 Chi-Chi earthquake,
Taiwan. Journal of Geophysical Research 108, 6006.
Dahal, R.K., Hasegawa, S., Nonomura, A., Yamanaka, M., Dhakal, S., 2008. DEMbased deterministic landslide hazard analysis in the lesser Himalaya of Nepal.
Georisk 2, 161178.
Dal Cin, C., Moens, L., Dierickx, P., Bastin, G., Zech, Y., 2005. An integrated
approach for realtime floodmap forecasting on the Belgian Meuse River. Natural
Hazards 36(12), 237256.
Dalrymple, J.B., Blong, R.J., Conacher, A.J., 1968. A hypothetical nine unit land
surface model. Zeitschrift fur Geomorphologie 12, 6076.
Davis, A.M., Moore, J.M., Mason, P.J., 2000. A study of landslide hazard in S.E.
Spain using integrated DEM, orthophotography and Landsat ETM imagery.
Proceedings of the Fourteenth International Conference on Applied Geologic
Delacourt, C., Allemand, P., Casson, B., Vadon, H., 2004. Velocity field of the La
Clapiere landslide measured by the correlation of aerial and QuickBird satellite
images. Geophysical Research Letters 31, 5.
Deng, Y., 2007. New trends in digital terrain analysis: landform definition,
representation, and classification. Progress in Physical Geography 31,
405419.
Derron, M.-H., Jaboyedoff, M., Blikra, L.H., 2005. Preliminary assessment of
rockslide and rockfall using a DEM (Oppstadhornet, Norway). Natural hazards in
Earth System Science 5, 285292.
Dewitte, O., Jasselette, J.C., Cornet, Y., van den Eeckhaut, M., Collignon, A.,
Poesen, J., Demoulin, A., 2008. Tracking landslide displacements by multitemporal DTMs: a combined aerial stereophotogrammetric and LIDAR approach
in western Belgium. Engineering Geology 99, 1122.
Dikau, R., 1989. The application of a digital relief model to landform analysis in
geomorphology. In: Raper, J. (Ed.), Three Dimensional Applications in
Geographic Information Systems. Taylor & Francis, London, pp. 5177.
Dixon, L.F.J., Barker, R., Bray, M., et al., 1998. Analytical photogrammetry for
geomorphological research. In: Lane, S., Richards, K., Chandler, J. (Eds.),
Landform Mapping, Modelling, and Analysis. Wiley, Chichester, pp. 4963.
Domakinis, C., Oikonomidis, D., Astaras, T., 2008. Landslide mapping in the
coastal area between the Strymonic Gulf and Kavala (Macedonia, Greece) with
the aid of remote sensing and geographical information systems. International
Drinkwater, M.R., Francis, R., Ratier, G., Wingham, D.J., 2003. The European Space
Agencys Earth Explorer Mission CryoSat: measuring variability in the
cryosphere. Annals of Glaciology 39, 313320.
Dragut- , L., Blaschke, T., 2006. Automated classification of landform elements using
object-based image analysis. Geomorphology 81, 330344.
Dragut- , L., Schauppenlehner, T., Muhar, A., Strobl, J., Blaschke, T., 2009.
Optimization of scale and parametrization for terrain segmentation: an
application to soil-landscape modeling. Computers & Geosciences 35,
18751883.
Duncan, C.C., Klein, A.J., Masek, J.G., Isacks, B.L., 1998. Comparison of late
Pleistocene and modern glacier extent in central Nepal based on digital
elevation data and satellite imagery. Quaternary Research 49, 241254.
Dunlop, P., Clark, C.D., 2006. The morphological characteristics of ribbed
moraines. Quaternary Science Reviews 25, 16681691.
Dunning, S.A., Massey, C.I., Rosser, N.J., 2009. Structural and geomorphological
features of landslides in the Bhutan Himalaya derived from terrestrial laser
scanning. Geomorphology 103, 1729.
Dunning, S.A., Rosser, N.J., Petley, D.N., Massey, C.R., 2006. Formation and failure
of the Tsatichhu landslide dam, Bhutan. Landslides 3, 107113.
Dvigalo, V.N., Melekestsev, I.V., 2009. The geological and geomorphic impact of
catastrophic landslides in the geyser valley of Kamchatka: aerial
photogrammetry. Journal of Volcanology and Seismology 3, 314325.
Dymond, J.R., Ausseil, A.-G., Shepherd, J.D., Buettner, J., 2006. Validation of a
region-wide model of landslide susceptibility in the Manawatu-Wanganui region
of New Zealand. Geomorphology 74, 7079.
Eisank, C., Dragut- , L., Gotz, J., Blaschke, T., 2010. Developing a semantic model of
glacial landforms for object-based terrain classification the example of glacial
cirques. In: Addink, E., Van Coillie, F.M.B. (Eds.), GEOBIA 2010- Geographic
Object-Based Image Analysis. Ghent University, Ghent, Belgium, 29 June to
June 2 July. ISPRS XXXVIII-4/C7, Archives ISSN 1682-1777.
Ercanoglu, M., 2005. Landslide susceptibility assessment of SE Bartin (west Black
Sea region, Turkey) by artificial neural networks. Natural Hazards and Earth
System Sciences 5, 979992.
Ermini, L., Catani, F., Casagli, N., 2005. Artificial neural networks applied to
landslide susceptibility assessment. Geomorphology 66, 327343.
Ernst, G.G.J., Kervyn, M., Teeuw, R.M., 2008. Advances in the remote sensing of
volcanic activity and hazards, with special consideration for applications in
developing countries. International Journal of Remote Sensing 29, 66876723.
Esaki, T., Xie, M.W., Zhou, G.Y., Mitani, Y., 2001. Monte Carlo method for locating
and evaluating 3D critical slope slip based on GIS Database. In: Sijing, W.,
Bingjunm, F., Zhongkui, L. (Eds.), Frontiers on Rock Mechanics and Sustainable
Development in 21st Century. Balkema, Rotterdam, pp. 1721.
Espizua, L.E., 1986. Fluctuations of the Rio del Plomo Glaciers. Geografiska
Annaler 68, 317327.
Evans, A.N., 2000. Glacier surface motion computation from digital image
sequences. IEEE Transactions on Geoscience and Remote Sensing 38,
10641072.
Evans, D.J.A., Clark, C.D., Mitchell, W.A., 2005. The last British Ice Sheet: a review
of the evidence utilised in the compilation of the Glacial Map of Britain. EarthScience Reviews 70, 253312.
Evans, S.G., Bent, A.L., 2004. The Las Colinas landslide, Santa Tecla: a highly
destructive flowslide triggered by the January 13, 2001, El Salvador earthquake.
Geological Society of America Special Paper 375, 2537.
Eyton, J.R., 1989. Low relief topographic enhancement in a Landsat snow cover
scene. Remote Sensing Environment 27, 105118.
Fabris, M., Pesci, A., 2005. Automated DEM extraction in digital aerial
photogrammetry: precisions and validation for mass movement monitoring.
Annals of Geophysics 48, 973988.
Favalli, M., Karatson, D., Mazzarini, F., Pareschi, M.T., Boschi, E., 2009.
Morphometry of scoria cones located on a volcano flank: a case study from Mt.
Etna (Italy), based on high-resolution LiDAR data. Journal of Volcanology and
Geothermal Research 186, 320330.
Federici, P.R., Spagnolo, M., 2004. Morphometric analysis on the size, shape and
areal distribution of glacial cirques in the maritime Alps (Western French-Italian
Alps). Geografiska Annaler 86, 235248.
Fell, R., 1994. Landslide risk assessment and acceptable risk. Canadian
Geotechnical Journal 31, 261272.
Felpeto, A., Arana, V., Ortiz, R., Astiz, M., Garcia, A., 2001. Assessment and
modelling of lava flow hazard on Lanzarote (Canary islands). Natural Hazards
23, 247257.
Fernandez, T., Irigaray, C., El Handouni, R., Chacon, J., 2003. Methodology for
landslide susceptibility mapping by means of a GIS. Application to the
Contraviesa area (Granada, Spain). Natural Hazards 58.2897-308.
221
Fernandez-Steeger, T.M., Rohn, J., Czurda, K., 2002. Identification of landslide areas
with neural nets for hazard analysis. In: Rybar, J., Stemberk, J., Wagner, P.
(Eds.), Proceedings of the 1st European Conference on Landslides, Prague.
Balkema, Rotterdam, pp. 163168.
Field, W.O., 1966. Mapping glacier termini in southern Alaska, 19311964.
Canadian Journal of Earth Sciences 6, 819825.
Finlayson, D.P., Montgomery, D.R., 2003. Modeling large-scale fluvial erosion in
geographic information systems. Geomorphology 53, 147164.
Finnegan, N.J., Pritchard, M.E., Lohman, R.B., Lundgren, P.R., 2008. Constraints on
surface deformation in the Seattle, WA, urban corridor from satellite radar
interferometry time-series analysis. Geophysical Journal International 174,
2941.
Fisher, P., Wood, J., Cheng, T., 2004. Where is Helvellyn? Fuzziness of multi-scale
landscape morphometry. Transactions of the Institute of British Geographers 29,
106128.
Fonstad, M.A., Marcus, W.A., 2005. Remote sensing of stream depths with
hydraulically assisted bathymetry (HAB) models. Geomorphology 72, 320339.
Fourniadis, I.G., Liu, J.G., Mason, P.J., 2007a. Regional assessment of landslide
impact in the Three Gorges area, China, using ASTER imagery: Wushan-Zigui.
Landslides 4, 267278.
Fourniadis, I.G., Liu, J.G., Mason, P.J., 2007b. Landslide hazard assessment in the
Three Gorges area, China, using ASTER imagery: Wushan-Badong.
Frappart, F., Papa, F., Famiglietti, J.S., Prigent, C., Rossow, W.B., Seyler, F., 2008.
Interannual variations of river water storage from a multiple satellite approach: a
case study for the Rio Negro River basin. Journal of Geophysical Research 113.
http://dx.doi.org/10.1029/2007JD009438.
Frattini, P., Crosta, G., Carrar, A., Agliardi, F., 2008. Assessment of rockfall
susceptibility by integrating statistical and physically-based approaches.
French, J.R., 2003. Airborne LiDAR in support of geomorphological and hydraulic
modeling. Earth Surface Processes and Landforms 28(3), 321335.
Fruneau, B., Achache, J., Delacourt, C., 1996. Observation and modeling of the
Saint-Etienne-de Trinee landslide using SAR interferometry. Tectonophysics 265,
181190.
Galli, M., Ardizzone, F., Cardinali, M., Guzetti, F., Reichenbach, P., 2008. Comparing
inventory maps. Geomorphology 94, 268289.
Gao, J., Liu, Y., 2001. Applications of remote sensing, GIS, and GPS in glaciology:
a review. Progress in Physical Geography 25, 520540.
Geist, T., Sttter, J., 2010. Airborne laser scanning in glacier studies. In: Pellikka,
P., Rees, W.G. (Eds.), Techniques for Topographic, Spatial, and Thematic
Mapping of Glaciers. Taylor & Francis, New York, NY, pp. 179194.
Ghosh, S., Stepniski, T., Vilalta, R., 2010. Automatic annotation of planetary
surfaces with geomorphic labels. IEEE Transactions on Geoscience and Remote
Sensing 48, 175185.
Giardino, M., Alberto, W., Perotti, L., 2008. Geomatics and geomorphology
integration for the analysis of morphostructures and deepseated gravitational
slope deformations: application in the Susa Valley (western Alps). Memorie
Descrittive Della Carta Geologica dItalia 78, 105112.
Gilvear, D., Bryant, R., 2005. Analysis of aerial photography and other remotely
sensed data. In: Kondolf, G.M., Piegay, H. (Eds.), Tools in Fluvial
Geomorphology. Wiley, Chichester. http://dx.doi.org/10.1002/0470868333.
Gilvear, D.J., Waters, T., Milner, A., 1998. Image analysis of aerial photography to
quantify the effect of gold place mining on channel morphology, Interior Alaska.
In: Lane, S., Richards, K., Chandler, J. (Eds.), Landform Mapping, Modelling,
and Analysis. Wiley, Chichester, pp. 195216.
Gilvear, D.J., Waters, T.M., Milner, A.M., 1995. Image analysis of aerial photography
to quantify changes in channel morphology and instream habitat following place
mining in interior Alaska. Freshwater Biology 34, 389398.
Glenn, N.F., Streutker, D.R., Chadwick, D.J., Thackray, G.D., Dorsch, S.J., 2006.
Analysis of LiDAR-derived topographic information for characterizing and
differentiating landslide morphology and activity. Geomorphology 73, 131148.
Goguel, B., 1989. A mass movement detected by photography: the Friolin landslide,
Savoie. Revue Francaise De Geotechnique 48, 5564.
Goldstein, R.M., Engelhardt, H., Kamb, B., Frolich, R.M., 1993. Satellite radar
interferometry for monitoring ice sheet motion: application to an Antarctic Ice
Stream. Science 262, 15251530.
Goodchild, M.F., 1992a. Geographical data modeling. Computers & Geosciences
18.401-4-8.
Goodchild, M.F., 1992b. Geographical information science. International Journal of
Geographical Information Systems 6(1), 3145.
Gordon, J.E., Brazier, V., Lees, G., 1994. Geomorphological systems: developing
fundamental principles for sustainable landscape management. In: OHalloran, D.,
222
Green, J., Harley, M., Stanley, M., Knill, J. (Eds.), Geological landscape
conservation. Geological Society, London, pp. 185189.
Granica, K., Hirschmugl, M., Proske, H., Schardt, M., Wallner, H., Wurm, M., 2007.
The derivation of indicators for landslide detection using very high resolution
(VHR) satellite imagery. Proceedings of the 9th International Symposium on
High Mountain Remote Sensing Cartography, Grazer Schriften der Geographie
und Raumforschung 43, 5966.
Greenwood, S.L., Clark, C.D., 2008. Subglacial bedforms of the Irish Ice Sheet.
Journal of Maps 2008, 332357.
Grosse, P., de Vries, B.V., Petrinovic, I.A., Euillades, P.A., Alvarado, G.E., 2009.
Morphometry and evolution of arc volcanoes. Geology 37, 651654.
Gurnell, A.M., 1997. Channel change on the River Dee meanders, 19461992, from
the analysis of air photographs. Regulated Rivers: Research and Management
13, 1326.
Gurnell, A.M., Downward, S.R., James, R., 1994. Channel planform change on the
River Dee meanders. Regulated Rivers: Research and Management 9, 187204.
Guzzetti, F., Cardinali, M., Reichenbach, P., Carrarra, A., 2000. Comparing landslide
maps: a case study in the upper Tiber River basin. Environmental Management
25, 247263.
Guzzetti, F., Crosta, G., Detti, R., Agliardi, F., 2002. STONE: a computer program for
the three-dimensional simulation of rock-falls. Computers & Geosciences 28,
10791093.
Guzzetti, F., Reichenbach, P., Wieczorek, G.F., 2003. Rockfall hazard and risk
assessment in the Yosemite Valley, California, USA. Natural Hazards in Earth
System Science 3, 491503.
Gunther, A., 2003. SLOPEMAP: programs for automated mapping of geometrical
and kinematical properties of hard rock hill slopes. Computers & Geosciences
29, 865875.
Haberler-Weber, M., Eichhorn, A., Kahmen, H., Theilen-Willige, B., 2009. Concept
of a multi-scale monitoring and evaluation system for landslide disaster
prediction. Proceedings of the 2007 IAG General Assembly, International
Association of Geodesy Symposia 133, 669677.
Haeberli, W., Huggel, C., Kaeaeb, A., et al., 2004. The Kolka-Karmadon rock/ice
slide of 20 September 2002: an extraordinary event of historical dimensions in
north Ossetia, Russian Caucasus. Journal of Glaciology 50, 533546.
Haefeli, R., 1966. Some notes on glacier mapping and ice movement. Canadian
Journal of Earth Sciences 3, 863876.
Hall, S.R., Farber, D.L., Ramage, J.M., et al., 2009. Geochronology of Quaternary
glaciations from the tropical Cordillera Huayhuash, Peru. Quaternary Science
Reviews 28(2526), 29913009.
Hamilton, C.W., Fagents, S.A., Thordarson, T., 2010. Explosive lava-water
interactions II: self-organization processes among volcanic rootless eruption
sites in the 17831784 Laki lava flow, Iceland. Bulletin of Volcanology 72(4),
469485.
Hamilton, S.K., Kellndorfer, J., Lehner, B., Tobler, M., 2006. Remote sensing of
floodplain geomorphology as a surrogate for biodiversity in a tropical river
system (Madre de Dios, Peru). Geomorphology 89, 2338.
Han, L., Rundquist, D.C., 1994. The response of both surface reflectance and the
underwater light field to various levels of suspended sediments: preliminary
results. Photogrammetric Engineering & Remote Sensing 60, 14631471.
Harper, J.T., Humphrey, N.F., Pfeffer, W.T., 1998. Crevasse patterns and the strainrate tensor: a high resolution comparison. Journal of Glaciology 44, 6876.
Hasegawa, S., Dahal, R.K., Nishimura, T., Nonomura, A., Yamanaka, M., 2009.
DEM-based analysis of earthquake-induced shallow landslide susceptibility.
Geotechnical and Geological Engineering 27, 419430.
Hattestrand, C., Stroeven, A.P., 2002. A relict landscape in the centre of
Fennoscandian glaciation: geomorphological evidence of minimal Quaternary
glacial erosion. Geomorphology 44, 127143.
Haugerud, R.A., Harding, D.J., Johnson, S.Y., Harless, J.L., Weaver, C.S., Sherrod,
B.L., 2003. High resolution Lidar topography of the Puget Lowland, Washington
a bonanza for earth science. GSA Today 13, 9.
Havenith, H.B., Strom, A., Caceres, F., Pirard, E., 2006. Analysis of landslide
susceptibility in the Suusamyr region. Tien Shan: Statistical and Geotechnical
Approach. Landslides 3, 3950.
Haynes, K., Barclay, J., Pidgeon, N., 2007. Volcanic hazard communication
using maps: an evaluation of their effectiveness. Bulletin of Volcanology 70,
123138.
Heritage, G.L., Fuller, I.C., Charlton, M.E., Brewer, P.A., Passmore, D.P., 1998. CDW
photogrammetry of low relief fluvial features: accuracy and implications for
reach-scale sediment budgeting. Earth Surface Processes and Landforms 23,
12191233.
Heyman, J., Hattestrand, C., Stroeven, A.P., 2008. Glacial geomorphology of the
Bayan Har sector of the NE Tibetan Plateau. Journal of Maps 2008, 4262.
Hicks, D.M., Duncan, M.J., Walsh, J.M., Westaway, R.M., Lane, S.N., 2002. New
views of the morphodynamics of large braided rivers from high-resolution
topographic surveys and time-lapse video. In: The Structure, Function and
Management Implications of Fluvial Sedimentary Systems, International
Association of Hydrological Sciences (IAHS) Publication 276, pp. 373380.
Higgitt, D.L., Warburton, J., 1999. Applications of differential GPS in upland fluvial
geomorphology. Geomorphology 29, 121134.
Hubbard, B., Glasser, N., 2005. Field Techniques in Glaciology and Glacial
Geomorphology. Wiley, Chichester, 400 pp.
Hudson, J.A., 1992. Rock Engineering Systems: Theory and Practice. High Plains
Press (JAH), Chichester, 185 pp.
Hughes, L.C., Clark, C.D., Jordan, C.J., 2010. Subglacial bedforms of the last
British Ice Sheet. Journal of Maps 2010, 543563.
Hughes, M.L., McDowell, P.F., Marcus, W.A., 2006. Accuracy assessment of
georectified aerial photographs: implications for measuring lateral channel
movement in GIS. Geomorphology 74, 116.
Hunsaker, C.T., Levine, D.A., Timmins, S.P., Jackson, B.L., ONeill, R.V., 1992.
Landscape characterization for assessing regional water quality. In: McKenzie,
D.H., Hyatt, D.E., McDonald, V.J. (Eds.), Ecological Indicators. Elsevier, New
York, pp. 9971006.
Hynek, B.M., Phillips, R.J., 2003. New data reveal mature, integrated drainage
systems on Mars indicative of past precipitation. Geology 31, 757760.
Hgda, K.A., Storvold, R., Lauknes, T.R., 2010. SAR imaging of glaciers.
In: Pellikka, P., Rees, W.G. (Eds.), Techniques for Topographic, Spatial, and
Thematic Mapping of Glaciers. Taylor & Francis, New York, NY, pp. 153178.
Iovine, G., Parise, M., 2002. Graphic method to classify and survey landslide
damage in buildings of urban areas. Memorie Della Societa Geologica Italiana
57, 595603. (In Italian).
Iwahashi, J., Kamiya, I., Yamagishi, H., 2009. Estimation of the most
suitable window size of the slope gradient and convexo-concave index for the
assessment of shallow landslides using high resolution Lidar DEM. Transactions
Japanese Geomorphological Union 30, 1527.
Iwahashi, J., Watanabe, S., Furuya, T., 2001. Landform analysis of slope
movements using DEM in Higashikubiki area, Japan. Computers and
Jaboyedoff, M., Baillifard, F., Couture, R., Locat, J., Locat, P., 2004. New insight of
geomorphology and landslide prone area detection using DEM. In: Lacerda, W.,
Ehrlich, M., Fontoura, A., Sayao, A. (Eds.), Landslides: Evaluation and
Stabilization. Taylor & Francis, London, pp. 191198.
Jaboyedoff, M., Demers, D., Locat, J., et al., 2009. Use of terrestrial laser scanning
for the characterization of retrogressive landslides in sensitive clay and rotational
landslides in river banks. Canadian Geotechnical Journal 46, 13791390.
Jaboyedoff, M., Labiouse, V., 2003. Preliminary assessment of rockfall hazards
based on GIS data. 10th International Congress on Rock Mechnics ISRM 2003Technology roadmap for rock mechanics, South African Institute of Mining and
Metallurgy, Johannesburgh, South Africa, pp. 575578.
Jansson, K.N., Glasser, N.F., 2005. Using Landsat 7 ETM imagery and digital
terrain models for mapping glacial lineaments on former ice sheet beds.
International Journal of Remote Sensing 26, 39313941.
Jansson, K.N., Kleman, J., Marchant, D., 2002. The succession of ice-flow patterns
in north-central Quebec-Labrador, Canada. Quaternary Science Reviews 21,
503523.
Jibson, R.W., Harp, E.L., Michael, J.A., 1998. A method for producing digital
probabilistic seismic landslide hazard maps. USGS Open-file report 98113,
17 pp.
Jones, A.F., Brewer, P.A., Johnstone, E., Macklin, M.G., 2007. High-resolution
interpretative geomorphological mapping of river valley environments using
airborne LiDAR data. Earth Surface Processes and Landforms 32(10),
15741592.
Jordan, D.C., Fonstad, M.A., 2005. Two dimensional mapping of river bathymetry
and power using aerial photography and GIS on the Brazos River, Texas.
Geocarto International 20(3), 18.
Joughin, I., Kwok, R., Fahnestock, M., 1996. Estimation of ice-sheet motion using
satellite radar interferometry: method and error analysis with application to
Humboldt Glacier, Greenland. Journal of Glaciology 42, 564575.
Joughin, I.R., Kwok, R., Fahnestock, M.A., 1998. Interferometric estimation of threedimensional ice-slow using ascending and descending passes. IEEE
Transactions on Geoscience and Remote Sensing GE 36, 2537.
Joughin, I.R., Winebrenner, D.P., Fahnestock, M.A., 1995. Observations of ice-sheet
motion in Ggreenland using satellite radar interferometry. Geophysical Research
Letters 22, 571574.
Jung, H.C., Hamski, J., Durand, M., et al., 2010. Characterization of complex fluvial
systems using remote sensing of spatial and temporal water level variations in
the Amazon, Congo, and Brahmaputra Rivers. Earth Surface Processes and
Kaab, A., 2002. Monitoring high-mountain terrain deformation from repeated
air and spaceborne optical data: examples using digital aerial imagery and
ASTER data. ISPRS Journal of Photogrammetry and Remote Sensing 57,
3952.
Kahle, A.B., Gillespie, A.R., Abbott, E.A., Abrams, M.J., Walker, R.E., Hoover, G.,
Lockwood, J.P., 1988. Relative dating of Hawaiian lava flows using multispectral
thermal infrared images : a new tool for geologic mapping of young volcanic
terranes. Journal of Geophysical Research 93(B12), 1523915251.
Kajiyama, A., Mitamura, M., Luan, T.X., Okudaira, T., Ha, N.V., 2005. Landslide
monitoring by photogrammetry in Mongset area, northwestern Vietnam.
International Journal of Geoinformatics 1, 125132.
Kalaugher, P.G., Grainger, P., 1997. A photographic method for rapid assessment of
landslide displacement. The Quarterly Journal of Engineering Geology 30,
193196.
Kamp, U., Growley, B.J., Khattak, G.A., Owen, L.A., 2008. GIS-based landslide
susceptibility mapping for the 2005 Kashmir earthquake region. Geomorphology
101, 631642.
Kamp, U., Owen, L.A., Growley, B.J., Khattak, G.A., 2010. Back analysis of landslide
susceptibility zonation mapping for the 2005 Kashmir earthquake: an assessment
of the reliability of susceptibility zoning maps. Natural Hazards 54, 125.
Karatson, D., Favalli, M., Tarquini, S., Fornaciai, A., Worner, G., 2010a. The regular
shape of stratovolcanoes: a DEM-based morphometrical approach. Journal of
Karatson, D., Telbisz, T., Singer, B., 2010b. Late-stage volcano geomorphic
evolution of the Pleistocene San Francisco Mountain, Arizona (USA), based on
high-resolution DEM analysis and Ar-40/Ar-39 chronology. Bulletin of
Kargel, J.S., Abrams, M.J., Bishop, M.P., et al., 2005. Multispectral imaging
contributions to global land ice measurements from space. Remote Sensing of
Environment 99(12), 187219.
Kervyn, M., Ernst, G.G.J., Goossens, R., Jacobs, P., 2008. Mapping volcano
topography with remote sensing: ASTER vs SRTM. International Journal of
Khattak, G.A., Owen, L.A., Kamp, U., Harp, E.L., 2010. Evolution of earthquaketriggered landslides in the Kashmir Himalaya, northern Pakistan. Geomorphology
115, 102108.
Kim, J.W., Park, S.S., Kim, C.S., Lee, Y., 2004. The efficient web-based mobile GIS
service system through reduction of digital map. In: Lagana`, A., Gavrilova, M.L.,
Kumar, V., Mun, Y., Tan, C.J.K., Gervasi, O. (Eds.), ICCSA 2004, LNCS 3043.
Springer, Berlin, pp. 410417.
Kleman, J., Borgstrom, I., 1996. Reconstruction of palaeo-ice sheets: the use of
geomorphological data. Earth Surface Processes and Landforms 21, 893909.
Kleman, J., Fastook, J., Stroeven, A.P., 2002. Geologically and geomorphologically
constrained numerical model of Laurentide Ice Sheet inception and build-up.
Quaternary International 9596, 8798.
Kleman, J., Hattestran, C., Borgstrom, I., Stroeven, A., 1997. Fennoscandian
palaeoglaciology reconstructed using a glacial geological inversion model.
Journal of Glaciology 43, 283299.
Klingseisen, B., Metternicht, G., Paulus, G., 2008. Geomorphometric landscape
analysis using a smie-automated GIS-approach. Environmental Modelling &
Software 23, 109121.
Koblinsky, C.J., Clarke, R.T., Brenner, A.C., Frey, H., 1993. Measurement of river
level variations with satellite altimetry. Water Resources Research 29(6),
18391848.
Koehl, M., Mombrun, L., Gay, M., et al., 2009. SAR interferometry and GPS
measurements on Argentieres glacier: comparisons and validation. Traitement
Du Signal 26, 109126.
Kouraev, A.V., Zakharova, E.A., Samain, O., Mognard, N.M., Cazenave, C., 2004. Ob
river discharge from TOPEX/Poseidon satellite altimetry (1992-2002). Remote
Sensing of the Environment 93, 238245.
Lagmay, A.M.F., Valdivia, W., 2006. Regional stress influence on the opening
direction of crater amphitheaters in Southeast Asian volcanoes. Journal of
Volcanology and Geothermal Research 158(1-2), 139150.
Lane, S.N., 2000. The measurement of river channel morphology using digital
photogrammetry. Photogrammetric Record 16(96), 937957.
Lane, S.N., Chandler, J.H., Richards, K.S., 1994. Developments in monitoring and
modeling small-scale river bed topography. Earth Surface Processes and
Lane, S.N., Widdison, P.E., Thomas, R.E., et al., 2010. Quantification of braided
river channel change using archival digital image analysis. Earth Surface
223
Lantuit, H., Pollard, W.H., 2008. Fifty years of coastal erosion and retrogressive
thaw slump activity on Herschel Island, southern Beaufort Sea, Yukon Territory,
Canada. Geomorphology 95, 84102.
Lee, S., Dan, N.T., 2005. Probabilistic landslide susceptibility mapping in the Lai
Chau province of Vietnam: focus on the relationship between tectonic fractures
and landslides. Environmental Geology 48, 778787.
Lee, S., Pradhan, B., 2006. Probabilistic landslide hazards and risk mapping of
Penang Island, Malaysia. Journal of Earth System Science 115, 661672.
Legleiter, C.J., Goodchild, M.F., 2005. Alternative representations of in-stream
habitat: classification using remote sensing, hydraulic modeling, and fuzzy logic.
International Journal of Geographic Information Science 19(1), 2950.
Legleiter, C.J., Marcus, W.A., Lawrence, R., 2002. Effects of sensor resolution on
mapping in-stream habitats. Photogrammetric Engineering and Remote Sensing
68(8), 801807.
Legleiter, C.J., Roberts, D.A., 2005. Effects of channel morphology and sensor
spatial representation on image-derived depth estimates. Remote Sensing of
Legleiter, C.J., Roberts, D.A., Marcus, W.A., Fonstad, M.A., 2004. Passive optical
remote sensing of river channel morphology and in-stream habitat: Physical
basis and feasibility. Remote Sensing of Environment 93, 493510.
Lejot, J., Delacourt, C., Piegay, H., Fournier, T., Termelo, M.-L., Allemand, P., 2007.
Very high spatial resolution imagery for channel bathymetry and topography
from an unmanned mapping controlled platform. Earth Surface Processes and
Landforms 33, 424.
Leon, J.G., Calmant, S., Seyler, F., et al., 2006. Rating curves and estimation of
average water depth at the upper Negro River based on satellite altimeter data
and modeled discharges. Journal of Hydrology 328, 481496.
Leroi, E., Rouzeau, O., Scanvic, J.Y., Weber, C.C., Vargas Cuervo, G., 1992. Remote
sensing and GIS technology in landslide hazard mapping in the Colombian
Andes. Episodes 15, 3235.
Leroy, S., Markus, A.G., Schwab, J., Costa, P.J.M., 2010. Seismic influence on the
last 1500-year infill history of Lake Sapanca (north Anatolian Fault, NW Turkey).
Leverington, D.W., Teller, J.T., Mann, J.D., 2002. A GIS method for reconstruction
of late Quaternary landscapes from isobase data and modern topography.
Computers & Geosciences 28, 631639.
Leys, K.F., Werrity, A., 1999. River channel planform change: software for historical
analysis. Geomorphology 29, 107120.
Li, S.S., Benson, C., Gens, R., Lingle, C., 2008a. Motion patterns of Nabesna
Glacier (Alaska) revealed by interferometric SAR techniques. Remote Sensing of
the Environment 112, 36283638.
Li, X., 1993. A united model for quantitative remote sensing of suspended sediment
concentration. International Journal of Remote sensing 14(14), 26652676.
Li, Y., Napieralski, J., Harbor, J., 2008b. A revised automated proximity and
conformity analysis method to compare predicted and observed spatial
boundaries of geologic phenomena. Computers & Geosciences 34, 18061814.
Li, Y., Napieralski, J., Harbor, J., Hubbard, A., 2007. Identifying patterns of
correspondence between modeled flow directions and field evidence: an
automated flow direction analysis. Computers & Geosciences 33, 141150.
Lindner, L., Marks, L., Ostaficzuk, S., Pekala, K., Szczesny, R., 1985. Application of
photogeological mapping to studies of glacial history of south Spitsbergen.
Earth Surface Process and Landforms 10, 387399.
Liu, G.-R., Shi, Y.-C., Feng, W.-K., 2007. Simulation of the formation of Jinjiang
landslide by DEM. Hydrogeology and Engineering Geology 34, 747680.
Lohani, B., Mason, D.C., 2001. Application of airborne scanning laser altimetry to
the study of tidal channel geomorphology. ISPRS Journal of Photogrammetry
and Remote Sensing 56(2), 100120.
Lopez-Blanco, J., Villers-Ruiz, L., 1995. Delineating boundaries of environmental
units for land management using a geomorphological approach and GIS a
study in Baja California, Mexico. Remote Sensing of the Environment 53,
109117.
Lu, P., Rosenbaum, M.S., 2003. Artificial neural networks and grey systems for the
prediction of slope stability. Natural Hazards 30, 383398.
Lu, Z., Fielding, E., Patrick, M., Trautwein, C., 2003. Estimating lava volume by
precision combination of multiple baseline spaceborne and airborne
interferometric synthetic aperture radar: the 1997 eruption of Okmok volcano.
IEEE Transactions on Geoscience & Remote Sensing 41, 14281436.
Lucieer, A., Stein, A., 2005. Texture-based landform segmentation of LiDAR imagery.
International Journal of Applied Earth Observation and Geoinformation 6,
261270.
Lui, J.G., Mason, P.J., Clerici, N., et al., 2004. Landslide hazard assessment in the
Three Gorges area of the Yangtze river using ASTER imagery. Geomorphology
61, 171187.
224
Lyon, J.G., Lunetta, R.S., Williams, D.C., 1992. Airborne multispectral scanner data
for evaluating bottom sediment types and water depths of the St. Marys River,
Michigan. Photogrammetric Engineering and Remote Sensing 61(2), 7182.
MacDonald, H.C., Grubbs, R.S., 1975. Landsat imagery analysis; an aid for
predicting landslide-prone areas for highway construction. NASA Technical
Memorandum, X-58168 Vol.I-B, 769778.
MacMillan, R.A., Pettapiece, W.W., Nolan, S.C., Goddard, T.W., 2000. A generic
procedure for automatically segmenting landforms in elements using DEMs,
heuristic rules and fuzzy logic. Fuzzy Sets and Systems 113, 81109.
Madin, I.P., Burns, W.J., 2006. Map of the Landslide Geomorphology of Oregon
City, Oregon and Vicinity Interpreted from Lidar Imagery and Aerial
Photographs. Oregon Department of Geology and Mineral Industries, Portland,
USA.
Malamud, B.D., Turcotte, D.L., Guzzetti, F., Reichenbach, P., 2004. Landslide
inventories and their statistical properties. Earth Surface Processes and
Marcus, W.A., Fonstad, M.A., 2008. Optical remote mapping of rivers at sub-meter
resolutions and watershed extents. Earth Surface Processes and Landforms 33,
324.
Marcus, W.A., Legleiter, C.J., Aspinall, R.J., Boardman, J.W., Crabtree, R.L., 2003.
High spatial resolution hyperspectral mapping of in-stream habitats, depths, and
woody debris in mountain streams. Geomorphology 55, 363380.
Mark, D., Smith, B., 2004. A science of topography: From qualitative ontology to
digital representations. In: Bishop, M.P., Shroder, J.F. (Eds.), Geographic
Information Science and Mountain Geomorphology. Springer, Berlin Heidelberg,
pp. 75100.
Marqunez, J., Duarte, R.M., Farias, P., Sanchez, M.J., 2003. Predictive GIS-based
model of rockfall activity in mountain cliffs. Natural Hazards 58, 341360.
Marston, R.A., Girel, J., Pautou, G., Piegay, H., Bravard, J.P., Arneson, C., 1995.
Channel metamorphosis, floodplain disturbance, and vegetation development:
Ain River, France. Geomorphology 13, 121131.
Marti, J., Felpeto, A., 2010. Methodology for the computation of volcanic
susceptibilityan example for mafic and felsic eruptions on Tenerife (Canary
Islands). Journal of Volcanology and Geothermal Research 195(1), 6977.
Mazzarini, F., Pareschi, M.T., Sbrana, A., Favalli, M., Fulignati, P., 2001. Surface
hydrothermal alteration mapping at Vulcano island using MIVIS data.
McGovern, P.J., Morgan, J.K., 2009. Volcanic spreading and lateral variations in the
structure of Olympus Mons, Mars. Geology 37(2), 139142.
McKean, J., Roering, J., 2004. Objective landslide detection and surface
morphology mapping using high-resolution airborne laser altimetry.
Meier, M., 1966. Some glaciological interpretations of remapping programs of
South Cascade, Nisqually, and Klawatti Glaciers, Washington. Canadian Journal
of Earth Sciences 6, 811818.
Mertes, L.A.K., 2002. Remote sensing of riverine landscapes. Freshwater Biology
47, 799816.
Mertes, L.A.K., Smith, M.O., Adams, J.B., 1993. Estimating suspsended sediment
concentrations in surface waters of the Amazon River wetlands from Landsat
images. Remote Sensing of the Environment 43, 281301.
Micheli, E.R., Kirchner, J.W., 2002. Effects of wet meadow riparian vegetation on
streambank erosion. 1. Remote Sensing Measurements of Streambank Migration
and Erodibility. Earth Surface Processes and Landforms 27, 627639.
Mikos, M., Vidmar, A., Brilly, M., 2005. Using a laser measurement system for
monitoring morphological changes on the Strug rock fall, Slovenia. Natural
Hazards and Earth System Sciences 5, 143153.
Millington, A.C., Townshend, J.R.G., 1987. The potential of satellite remote sensing
for geomorphological investigations: an overview. In: Gardiner, V. (Ed.),
International Geomorphology. Wiley, Chichester, pp. 331342.
Miska, L., Hjort, J., 2005. Evaluation of current statistical approaches for predictive
geomorphological mapping. Geomorphology 67, 299315.
Mohr, J.J., Reeh, N., Madsen, S.N., 1998. Three-dimensional glacial flow and
surface elevation measured with radar interferometry. Nature 391, 273276.
Moody, A., Katz, D.B., 2004. Artificial intelligence in the study of mountain
landscapes. In: Bishop, M.P., Shroder, J.F. (Eds.), Geographic Information Science
and Mountain Geomorphology. Springer, Berlin Heidelberg, pp. 219247.
Morelli, M., Piana, F., Polino, R., 2001. Comparison between remotely sensed
lineament intersection density and landslide distribution in the Langhe area
(southern Piemonte). Geoingegneria Ambientale e Mineraria 38, 251256. (In
Italian).
Moren, B., Heyman, J., Stroeven, A.P., 2011. Glacial geomorphology of the central
Tibetan Plateau. Journal of Maps v2011, 115125. http://dx.doi.org/10.4113/
jom.2011.1161.
Mount, N.J., Louis, J., 2005. Estimation and propogation of error in measurements
of river channel movement from aerial imagery. Earth Surface Processes and
Mount, N.J., Louis, J., Teeuw, R.M., Zukiski, P.M., Stott, T., 2003. Estimation of
error in bankfull width comparisons from temporally sequences raw and
corrected aerial photographs. Geomorphology 56, 6577.
Mullen, I., Kellet, J., 2007. Groundwater salinity mapping using airborne
electromagnetics and borehole data within the lower Balonne catchment,
Queensland, Australia. International Journal of Applied Earth Observation and
Geoinformation 9, 116123.
Murray, J.B., van Wyk de Vries, B., Marquez, A., Williams, D.A., Byrne, P., Muller, J.-P.,
Kim, J.-R., 2010. Late-stage water eruptions from Ascraeus Mons volcano, Mars:
implications for its structure and history. Earth and Planetary Science Letters
294(34), 479491.
Napieralski, J., Harbor, J., Li, Y., 2007b. Glacial geomorphology and geographic
information systems. Earth Science Reviews 85, 122.
Napieralski, J., Hubbard, A., Li, Y., et al., 2007a. Toward a GIS assessment of
numerical ice sheet model performance using geomorphological data. Journal of
Glaciology 53, 7183.
Napieralski, J.A., Nalepa, N., 2010. The role of resolution and contour interval in
drumlin delineation and measurement. Computers & Geosciences 36, 222230.
Nelson, P.A., Smith, J.A., Miller, A.J., 2006. Evolution of channel morphology and
hydrologic response in an urbanizing drainage basin. Earth Surface Processes
and Landforms 31(9), 10631079.
Ng, F., Barr, I.D., Clark, C.D., 2010. Using the surface profiles of modern ice
masses to inform palaeo-glacier reconstructions. Quaternary Science Reviews
29(2324), 32403255.
Nichol, J.E., Shaker, A., Wong, M.-S., 2006. Application of high-resolution stereo
satellite images to detailed landslide hazard assessment. Geomorphology 76,
6875.
Noferini, L., Pieraccini, M., Mecatti, D., Macaluso, G., Luzi, G., Atzeni, C., 2006.
Long term landslide monitoring by ground-based synthetic aperture radar
interferometer. International Journal of Remote Sensing 27, 18931905.
Notebaert, B., Verstraeten, G., Govers, G., Poesen, J., 2009. Qualitative and
quantitative applications of LiDAR imagery in fluvial geomorphology. Earth
Ochoa-Tejeda, V., Parrot, J.-F., 2007. Automatic extraction of landslide traces
utilizing digital terrain models and high-resolution IKONOS satellite imagery:
example of the Sierra Norte de Puebla, Mexico. Revista Mexicana De Ciencias
Geologicas 24, 354367.
Ostir, K., Veljanovski, T., Podobnikar, T., Stancic, Z., 2003. Application of remote
sensing and GIS in Mt. Mangart landslide observation (Slovenia). Geologija 46,
385390.
Owen, L.A., Kamp, U., Khattak, G., Harp, E., Keefer, D.K., Bauer, M., 2008.
Landslides triggered by the October 8, 2005, Kashmir earthquake.
OConnor, J.E., Jones, M.A., Haluska, T.L., 2003. Flood plain and channel dynamics
of the Quinault and Queets Rivers, Washington, USA. Geomorphology 51,
3159.
ONeill, R.V., Hunsaker, C., Jones, K.B., et al., 1997. Monitoring environmental
quality at the landscape scale. Bioscience 47, 513519.
Perron, J.T., Kirchner, J.W., Dietrich, W.E., 2008. Spectral signatures of
characteristic spatial scales and nonfractal structure in landscapes. Journal of
Geophysical Research 113, F04003. http://dx.doi.org/10.1029/2007JF000866.
Petrie, G., Price, R.J., 1966. Photogrammetric measurements of the ice wastage and
morphological changes near the Casement Glacier, Alaska. Canadian Journal of
Earth Sciences 6, 827840.
Pieraccini, M., Noferini, L., Mecatti, D., Atzeni, C., Teza, G., Galgaro, A., Zaltron, N.,
2006. Integration of radar interferometry and laser scanning for remote
monitoring of an urban site built on a sliding slope. IEEE Transactions on
Geoscience & Remote Sensing 44, 23352342.
Pike, R., 1992. Machine visualization of synoptic topography by digital image
processing. In: Wiltshire, D.A. (Ed.), Selected Papers in the Applied Computer
Sciences. US Geological Survey Bulletin, vol. 2016, pp. 112.
Pike, R.J., Clow, G.D., 1981. Revised classification of terrestrial volcanoes and a
catalog of topographic dimensions with new results on edifice volume. U.S.
Geological Survey Open-File Report, OF 81-1038.
T., Reuter, H.I. (Eds.), GeomorphometryConcepts, Software, Applications.
Developments in Soil Science. Elsevier, Amsterdam, vol. 33, pp. 330.
Plewe, B., 1997. A representation-oreinted taxonomy of gradation. In: Hirtle, S.C.,
Frank, A.U. (Eds.), Spatial information theory: a theoretical basis for GIS,
Lecture Notes in Computer Sciences 1329. Spring-Verlag, Berlin, pp. 121136.
Pradhan, B., 2010. Remote sensing and GIS-based landslide hazard analysis and
cross-validation using multivariate logistic regression model on three test areas
in Malaysia. Advances in Space Research 45, 12441256.
Prest, V.K., Grant, D.R., Rampton, V.N., 1968. The Glacial Map of Canada.
Geological Survey of Canada, Ottawa.
Punkari, M., 1982. Glacial geomorphology and dynamics in the eastern parts of the
Baltic Shield interpreted using landsat imagery. Photogrammetric Journal
Finland 9, 7793.
Punkari, M., 1993. Modelling of the dynamics of the Scandinavian ice sheet using
remote sensing and GIS methods. In: Aber, J. (Ed.), Glaciotectonics and
Mapping Glacial Deposits. Proceedings of the INQUA commission of formation
and properties of glacial deposits. Canadian Plains Research Center, University
of Regina, Canada, pp. 232250.
Raseman, S., Schimdt, J., Schrott, L., Dikau, R., 2004. Geomorphometry in
mountain terrain. In: Bishop, M.P., Shroder, J.F. (Eds.), Geographic
Information Science and Mountain Geomorphology. Springer-Praxis, Berlin,
pp. 101137.
Rathje, E.M., Kayen, R., Woo, K.-S., Towhata, I., 2006. Remote sensing
observations of landslides and ground deformation from the 2004 Niigata Ken
Chuetsu earthquake. Soils and Foundations 46, 831842.
Refice, A., Capolongo, D., 2002. Probabilistic modelling of uncertainties in
earthquake-induced landslide hazard assessment. Computers and Geosciences
28, 735749.
Remondo, J., Gonzalez-Dez, Daz de Teran, J.R., Cendrero, A., 2003. Landslide
susceptibility models utilizing spatial data analysis techniques. A case study
from the Lower Deba Valley, Guipuzcoa (Spain). Natural Hazards 58, 267279.
Rodriguez-Gonzalez, A., Fernandez-Turiel, J.L., Perez-Torrado, F.J., Gimeno, D.,
Aulinas, M., 2010. Geomorphological reconstruction and morphometric
modelling applied to past volcanism. International Journal of Earth Sciences 99,
645660.
Roessner, S., Wetzel, H.U., Kaufmann, H., Sarnagoev, A., 2006. Satellite remote
sensing and GIS for analysis of mass movements with potential for dam
formation. Italian Journal of Engineering Geology and Environment, Special
Issue 2006, 103114.
Rott, H., Nagler, T., 2006. The contribution of radar interferometry to the assessment
of landslide hazards. Advances in Space Research 37, 710719.
Rott, H., Scheuchl, B., Siegel, A., Grasemann, B., 1999. Monitoring very slow slope
movements by means of SAR interferometry: a case study from a mass waste
above a reservoir in the Otztal Alps, Austria. Geophysical Research Letters
26(11), 16291632.
Rowland, S.K., 1996. Slope, lava flow volumes, and vent distribution on Volcano
Fernandina, Galapagos Islands. Journal of Geophysical Research 101,
2765727672.
Rumelhart, H., Hinton, G.E., Williams, R.J., 1986. Learning internal representation
by error propagation. In: Rumelhart, H., McClelland, J.L. (Eds.), Parallel
Distributed Processing. MIT Press, Cambridge, pp. 318362.
Sahlin, E.A.U., Glasser, N.F., 2008. A geomorphological map of Cadair Idris, Wales.
Journal of Maps v2008, 299314. http://dx.doi.org/10.4113/jom.2008.1011.
Salcher, B.C., Hinsch, R., Wagreich, M., 2010. High-resolution mapping of glacial
landforms in the North Alpine Foreland, Austraia. Geomorphology 122,
283293.
Samarakoon, L., Ogawa, S., Ebisu, N., Lapitan, R., Kohki, Z., 1993. Inference of
landslide susceptible areas by landsat thematic mapper data. IAHS-AISH
Publication 217, 8390.
Santacana, N., Baeza, B., Corominas, J., de Paz, A., Marturia, J., 2003. A GISbased multivariate statistical analysis for shallow landslide susceptibility
mapping in La Pobla de Lillet area (Eastern Pyrenees), Spain. Natural Hazards
58, 281295.
Santini, M., Grimaldi, S., Nardi, F., Petroselli, A., Rulli, M.C., 2009. Pre-processing
algorithms and landslide modeling on remotely sensed DEMs. Geomorphology
113, 110125.
Sauber, J., Molnia, B., Carabajal, C., Luthcke, S., Muskett, R., 2005. Ice elevations
and surface change on the Malaspina Glacier, Alaska. Geohysical Research
Letters 32, L23S01. http://dx.doi.org/10.1029/2005GL023943.
Scambos, T.A., Bindschadler, R.A., 1993. Complex ice stream flow revealed by
sequential satellite imagery. Annals of Glaciology 17, 177182.
Scambos, T.A., Dutkiewicz, M.J., Wilson, J.C., Bindschadler, R.A., 1992. Application
of image cross-correlation to the measurement of glacier velocity using satellite
image data. Remote Sensing of the Environment 42, 177186.
Schneider, J.F., 2009. Seismically reactivated Hattian slide in Kashmir, Northern
Pakistan. Journal of Seismology 13, 387398.
Schulz, W.H., 2007. Landslide susceptibility revealed by LIDAR imagery and
historical records, Seattle, Washington. Engineering Geology 89, 6787.
225
Shroder, Jr. J.F., Bishop, M.P., 2004. Mountain geomorphic systems. In: Bishop,
M.P., Shroder, J.F. (Eds.), Geographic Information Science and Mountain
Geomorphology. Springer-Praxis, Berlin Heidelberg, pp. 3366.
Sidjak, R.W., Wheate, R.D., 1999. Glacier mapping of the Illecillewaet icefield,
British Columbia, Canada, using Landsat TM and digital elevation data.
International Journal of Remote Sensing 20, 273284.
Singhroy, V., Mattar, K.E., Gray, A.L., 1998. Landslide characterization in Canada
using interferometric SAR and combined SAR and TM images. Advances in
Space Research 21, 465476.
Skoye, K.R., Eyton, J.R., 1992. Digital analysis of low relief topography in a
Landsat snow-cover scene in south central Alberta. Canadian Journal of Remote
Sensing 18, 143150.
Smets, B., Wauthier, C., dOreye, N., 2010. A new map of the lava flow field of
Nyamulagira (D.R. Congo) from satellite imagery. Journal of African Earth
Smith, G.R., Woodward, J.C., Heywood, D.I., Gibbard, P.L., 2000. Interpreting
Pleistocene glacial features from SPOT HRV data using fuzzy techniques.
Smith, M.J., Clark, C.D., 2005. Methods for the visualisation of digital elevation
models for landform mapping. Earth Surface Processes and Landforms 30,
885900.
Smith, M.J., Clark, C.D., Wise, S.M., 2001. Mapping glacial lineaments from
satellite imagery: an assessment of the problems and development of best
procedure. Slovak Geological Magazine 7, 263274.
Smith, M.J., Pain, C.F., 2009. Applications of remote sensing in geomorphology.
Smith, M.J., Rose, J., Booth, S., 2006. Geomorphological mapping of glacial
landforms from remotely sensed data: an evaluation of the principal
data sources and an assessment of their quality. Geomorphology 76,
148165.
Spicer, K.R., Costa, J.E., Placzek, G., 1997. Mesuring flood discharge in unstable
stream channels using ground penetrating radar. Geology 25(5), 423426.
Spiker, E.C., Gori, P.L., 2000. National landslide hazards mitigation strategy: a
framework for loss reduction. Department of Interior, USGS, Open-file Report
00-450, 49 pp.
Stevens, N.F., Manville, V., Heron, D.W., 2002. The sensitivity of a volcanic flow
model to digital elevation model accuracy: experiments with digitised map
contours and interferometric SAR at Ruapehu and Taranaki volcanoes, New
Zealand. Journal of Volcanology and Geothermal Research 119, 89105.
Stokes, C.R., Clark, C.D., 2003. Giant glacial grooves detected on Landsat ETM
satellite imagery. International Journal of Remote Sensing 24, 905910.
Strom, A.L., Korup, O., 2006. Extremely large rockslides and rock avalanches in the
Tien Shan mountains, Kyrgyzstan. Landslides 3, 125136.
Suzen, M.L., Doyuran, V., 2004. Data driven bivariate landslide susceptibility
assessment using geographical information system: a method and application to
Asarsuyu catchment, Turkey. Engineering Geology 71, 303321.
Tangestani, M.H., 2004. Landslide susceptibility mapping using the fuzzy gamma
approach in a GIS, Kahan catchment area, southwest Iran. Australian Journal of
Earth Sciences 51, 439450.
Tarchi, D., Casagli, N., Fanti, R., et al., 2003. Landslide monitoring by using
ground-based SAR interferometry: an example of application to the Tessina
landslide in Italy. Engineering Geology 68, 1530.
Terlien, M.T.J., Van Westen, C.J., Van Asch, T.W.J., 1995. Deterministic modelling
in GIS-based landslide hazards assessment. In: Carrara, A., Guzzetti, F. (Eds.),
Geographical Information Systems in Assessing Natural Hazards. Kluwer,
Dordrecht, pp. 5777.
Thoma, D.P., Satish, T., Gupta, C., Bauer, M.E., Kirchoff, C.E., 2005. Airborne laser
scanning for riverbank erosion assessment. Remote Sensing of Environment 95,
493501.
Torres, R., et al., 2004. Monitoring the evolution of the Pasig-Potrero alluvial fan,
Pinatubo Volcano, using a decade of remote sensing data. Journal of
Van den Eeckhaut, M., Poesen, J., Verstraeten, G., Vanacker, V., Moeyersons, J.,
Nyssen, J., van Beek, L.P.H., 2005. The effectiveness of hillshade maps and
expert knowledge in mapping old deep-seated landslides. Geomorphology 67,
351363.
Van den Eeckhaut, M., Poesen, J., Verstraeten, G., et al., 2007. Use of LIDARderived images for mapping old landslides under forest. Earth Surface Processes
Varnes, D.J. 1984. Landslide Hazard Zonation: a Review of Principles and Practice.
UNESCO Natural Hazards Series 3, Paris, 63 pp.
Volker, H.X., Wasklewicz, T.A., Ellis, M.A., 2007. A topographic fingerprint to
distinguish alluvial fan formative processes. Geomorphology 88, 3445.
226
Wadge, G., Francis, P.W., Ramirez, C.F., 1995. The Socompa collapse and
avalanche event. Journal of Volcanology and Geothermal Research 66, 309336.
Walcott, R.C., Summerfield, M.A., 2008. Scale dependence of hypsometric integrals:
an analysis of southeast African basins. Geomorphology 96, 174186.
Walsh, S.J., Bian, L., McKnight, S., Brown, D.G., Hammer, E.S., 2003. Solifluction
Stepps and risers, Lee Ridge, Glacier National Park, Montana, USA: a scale and
pattern analysis. Geomorphology 55, 381398.
Walsh, S.J., Butler, D.R., Malanson, G.P., 1998. An overview of scale, pattern,
process relationships in geomorphology: a remote sensing and GIS persepctive.
Walstra, J., Chandler, J.H., Dixon, N., Dijkstra, T.A., 2007. Aerial Photography and
Digital Photogrammetry for Landslide Monitoring; Mapping Hazardous Terrain
Using Remote Sensing. Geological Society, London, Special Publication 283,
53-63.
Wang, W., Wu, H., Nakamura, H., Wu, S., Ouyang, S., Yu, M., 2003. Mass
movements caused by recent tectonic activity: the 1999 Chi-Chi earthquake in
central Taiwan. Island Arc 12, 325334.
Wang, Z.-H., 2006. Digital landslide technology and its application in investigation
of Tiantai town landslide. Chinese Journal of Geotechnical Engineering 28,
516520.
Westaway, R.M., Lane, S.N., Hicks, D.M., 2000. Development of an automated
correction procedure for digital photogrammetry for the study of wide, shallow
gravel-bed rivers. Earth Surface Processes and Landforms 25(2), 200226.
Westaway, R.M., Lane, S.N., Hicks, D.M., 2001. Airborne remote sensing of clear
water, shallow, gravel-bed rivers using digital photogrammetry and image
analysis. Photogrammetric Engineering and Remote Sensing 67(11), 12711281.
Westaway, R.M., Lane, S.N., Hicks, D.M., 2003. Remote survey of large-scale
braided, gravel-bed rivers using digital photogrammetry and image analysis.
Whillans, I.M., Tsend, Y.-H., 1995. Automatic tracking of crevasses on satellite
images. Cold Regions Science Technology 23, 201421.
Whited, D., Stanford, J.A., Kinball, J.S., 2002. Application of airborne multispectral
digital imagery to quantify riverine habitats at different base flows. River
Research and Applications 18(6), 583594.
Whitesell, B.L., Vitek, J.D., Butler, D.R., 1988. Changes in the planform of the Red River,
McCurtain County, Oklahoma, 1938-84. Oklahoma Geology Notes 48, 194211.
Wilford, J., 2009. Using airborne geophysics to define the 3D distribution and
landscape evolution of Quaternary valley-fill deposits around the Jamestown
area, South Australia. Australian Journal of Earth Sciences 56, 567588.
Willige, B., 2010. Detection of local site conditions influencing earthquake shaking
and secondary effects in southwest-Haiti using remote sensing and GISmethods. Natural Hazards and Earth System Sciences 10, 11831196.
Winterbottom, S.J., 2000. Medium and short-term channel planform changes on the
Rivers Tay and Tummel, Scotland. Geomorphology 34, 195208.
Winterbottom, S.J., Gilvear, D.J., 1997. Quantification of channel-bed morphology

in gravel-bed rivers using airborne multispectral imagery and aerial
photography. Regulated Rivers: Research and Management 13(6), 489499.
Winterbottom, S.J., Gilvear, D.J., 2000. A GIS-based approach to mapping
probabilities of river bank erosion: regulated River Tummel, Scotland. Regulated
Rivers: Research and Management 16, 127140.
Wohl, E., Oguchi, T., 2004. Geographic information systems and mountain hazards.
In: Bishop, M.P., Shroder, J.F. (Eds.), Geographic Information Science and
Mountain Geomorphology. Springer-Praxis, Chichester, pp. 309342.
Woodruff, J.F., Evenden, L.J., 1962. Geomorphic measurement from aerial photos.
Professional Geographer 14, 2326.
Wright, A., Marcus, W.A., Aspinall, R.J., 2000. Applications and limiations of using
multispectral digital imagery to map geomorphic stream units in a lower order
stream. Geomorphology 33(1/2), 107120.
Wright, D.J., Goodchild, M.F., Proctor, J.D., 1997. Demystifying the persistent
ambiguity of GIS as tool versus science. Annals of the Association of
American Geographers 87(2), 346362.
Wright, R., Garbeil, H., Baloga, S.M., Mouginis-Mark, P.J., 2006. An assessment of
shuttle radar topography mission digital elevation data for studies of volcano
morphology. Remote Sensing of Environment 105(1), 4153.
Xie, M., Esaki, T., Cai, M., 2004. A time-space based approach for mapping
rainfall-induced shallow landslide hazard. Environmental Geology 46,
840850.
Yamaguchi, Y., Tanaka, S., Odajima, T., Tsuchida, S., Kamai, T., ERIM International,
1999. Detection of landslide movement using SPOT HRV and ADEOS AVNIR
data in Azumayama area, Japan. Proceedings of the 13th International
Conference on Applied Geologic Remote Sensing 13, II.97-II.103.
Yang, M.D., Yang, Y.F., Hsu, S.C., 2004. Application of remotely sensed data to the
assessment of terrain factors affecting the Tsao-Ling landslide. Canadian Journal
of Remote Sensing 30, 593603.
Yoshizawa, N., Nishizawa, S., Misawa, T., Negishi, R., 1991. Application of
photogrammetry to analysis of landslide behavior. Journal of the Japan Society
of Photogrammetry and Remote Sensing 30, 820.
Zhou, C.H., Yue, Z.Q., Lee, C.F., Zhu, B.Q., Wang, Z.H., 2001. Satellite image
analysis of a huge landslide at Yi Gong, Tibet, China. Quarterly Journal of
Engineering Geology and Hydrogeology 34, 325332.
Zouzias, D., Miliaresis, G.C., Seymour, K.S., 2011. Interpretation of Nisyros volcanic
terrain using land surface parameters generated from the ASTER global digial
elevation model. Journal of Volcanology and Geothermal Research 200,
159170.
Zwally, H.J., Schutz, B., Abdalati, W., et al., 2002. ICESats laser measurements of
polar ice, atmosphere, ocean, and land. Journal of Geodynamics 34, 405445.
Biographical Sketch
Associate Professor Napieralski completed his PhD in Earth and Atmospheric Sciences at Purdue University, West
Lafayette, IN, USA in 2005 with a dissertation focused on the development of geospatial tools to link numerical
ice-sheet model output with landform assemblages. After completing his PhD, he joined the Department of
Natural Sciences at the University of MichiganDearborn, MI, USA as an Assistant Professor of Geology, where he
continues to integrate spatial theory and applications into geomorphology both in research and in the classroom.
He was recently a US Fulbright Scholar to Cyprus to study soil and water conservation and was awarded the
Distinguished Teaching Award at the University of MichiganDearborn. His research includes alpine geography,
fluvial and glacial geomorphology, environmental science, GIScience, and field-based science education.
227
Dr. Barr is currently a lecturer in Physical Geography at Queen Mary University of London, UK, having obtained
his PhD in Palaeoglaciology from the University of Sheffield, UK, in 2009. Dr. Barrs research interests are in the
use of remote sensing and geographic information systems methods to better understand the distribution, dimensions, and dynamics of both modern and paleoglaciers. To date, his research has largely focused on the glacial
history of Eastern Siberia and Kamchatka.
Associate Professor Kamp received his PhD in Geography with an emphasis on Geomorphology from the
Technische Universitat Berlin, Germany, in 1999. In the following years, he had positions at Freie Universitat
Berlin, University of NebraskaOmaha, and DePaul University in Chicago before joining The University of
Montana in 2005. Dr. Kamps research is on alpine geomorphology and glaciology, in particular, Quaternary and
recent glacial change. He carried out fieldwork in the mountains of South and Central Asia, and in the Rocky
Mountains and the Andes.
Dr. Kervyn holds a Masters in Geography from the University Catholique de Louvain, Belgium, and a PhD in
Geology from Ghent University, Belgium. He is currently working as a lecturer in physical geography, geomorphology, and natural hazards at the Vrije Universiteit Brussel. His research focuses on the geomorphology and
spatial distribution of volcanic edifices, on the effect of volcano-tectonic, gravitational, and erosion processes on
volcano morphology, and on monitoring and assessing hazards in active volcano-tectonic regions of Africa. His
research combines observations from digital elevation model and satellite images with analog experiments and
field observations.
3.9 GIS-Based Soil Erosion Modeling

H Mitasova, North Carolina State University, Raleigh, NC, USA
M Barton and I Ullah, Arizona State University, Tempe, AZ, USA
J Hofierka, Pavol Jozef Safarik University, Kosice, Slovakia
RS Harmon, North Carolina State University, Raleigh, NC, USA
3.9.1
Introduction
3.9.2
Background
3.9.2.1
Erosion Processes
3.9.2.2
Spatial Variability
3.9.2.3
Temporal Variability
3.9.2.4
GIS-Based Erosion Modeling
3.9.3
Foundations in Erosion Modeling
3.9.3.1
Sediment Transport and Net Erosion/Deposition Equations
3.9.3.2
Detachment and Sediment Transport Capacities
3.9.4
Simplified Models of Erosion Processes
3.9.4.1
Detachment Capacity Limited Case
3.9.4.2
Transport Capacity Limited Case
3.9.4.3
Process-Form Relationship
3.9.4.4
Path-Sampling Transport Modeling
3.9.4.5
Gully Erosion
3.9.4.6
Statistical Modeling
3.9.4.7
Landscape Evolution Modeling
3.9.5
GIS Implementation
3.9.5.1
Coupling GIS and Models
3.9.5.2
Derived Model Parameters
3.9.5.3
Analysis and Visualization
3.9.6
Case Studies
3.9.6.1
North Carolina Piedmont
3.9.6.2
Mediterranean Landscape Evolution
3.9.7
Conclusion and Future Directions
Acknowledgments
References
r (m1)
Symbols
(x, y) (m)
t (s)
qs(x, y, t) (kg m3)
h(x, y, t) (m)
qs(x, y, t) (kg (ms)1)
qs 9qs9
=
ds(x, y, t) (kg m2s1)
q(x, y, t) (m2 s1)
Tc (kg (ms)1)
Dc (kg m2 s1)
georeferenced coordinates
time
sediment mass density
water flow depth
direction and rate of sediment
flow per unit width
magnitude of sediment flow rate
per unit width
divergence of a flow vector field
net erosion or deposition rate
direction and rate of water flow
per unit width
sediment transport capacity
detachment capacity
Kt(s)
Kd (s m1)
s(Pa kg m2)
s0 (Pa)
a, b, m, n
b (deg)
S tanb
gw qwg
g 9.81 (m s2)
229
230
230
230
230
232
234
234
235
237
237
238
239
240
241
242
242
243
243
244
244
244
245
248
253
254
254
first-order reaction term

dependent on land cover
properties
effective sediment transport
capacity coefficient
effective erodibility (detachment
capacity coefficient)
shear stress
critical shear stress
empirical exponents
surface slope angle
surface slope (rise over run)
hydrostatic pressure of water with
the unit height
gravitational acceleration
Mitasova, H., Barton, M., Ullah, I., Hofierka, J., Harmon, R.S., 2013. GISbased soil erosion modeling. In: Shroder, J. (Editor in Chief), Bishop, M.P.
Remote Sensing and GIScience in Geomorphology, pp. 228258
228
http://dx.doi.org/10.1016/B978-0-12-374739-6.00052-X
GIS-Based Soil Erosion Modeling
qw 103 (kg m3)

mass density of water
v (m s1)
stream velocity
n
Mannings coefficient
the unit stream power
x (J (ms)1)
average annual soil loss
D (kg m2 yr1)
R (M J mm (ha hr yr)1) rainfall factor
soil erodibility factor
K (ton ha hr
(ha M J mm)1)
LS
dimensionless topographic
(length-slope) factor
L (m)
hillslope length
length of the standard USLE
L0 22.23 m
experimental plot
C
dimensionless land-cover factor
P
dimensionless prevention
measures factor
upslope contributing area per
U (m2 m1)
unit width
upslope contributing area
A (m2)
unit vector in the steepest slope
s0 (cosa, sina)
direction
Glossary
Detachment capacity Maximum potential soil
detachment by overland flow.
Erosion by overland flow Detachment of soil particles by
raindrop impact and their removal downslope by water
flowing overland as a sheet or in small concentrated flow
channels called rills.
Gully erosion Detachment of soil by surface water flow
concentrated in rapidly developing channel via headwall or
knickpoint migration.
Landscape evolution The change in the altitude and
morphology of the topography over time given variations in
erosion and deposition caused by numerous surface
processes.
Net erosion and deposition Soil mass that is lost or
gained at a unit area for a unit time due to removal,
transport, or deposition by water flow.
a (deg)
jp (m1)
jt (m1)
jh (m1)
Pe
w
z (m)
c (m2 s1)
Kg (m(1000 yr)1)
r (m)
229
aspect angle equivalent to the

direction of flow
profile curvature (normal
curvature in gradient direction,
rate of change in slope)
tangential curvature (normal
curvature in the direction
perpendicular to the gradient, rate
of change in aspect)
plan curvature (tangential
curvature projected to
horizontal plane)
channel erosion index
topographic wetness index
elevation
sediment transport diffusion
constant
long term diffusion coefficient
rainfall excess
Rill erosion Removal of soil particles by water flowing in

small concentrated flow channels called rills.
Sediment transport capacity Maximum potential
sediment transport by overland flow.
Shear stress Force of moving water against the bed of the
channel.
Stream power Rate of energy dissipation against the bed
and banks of a stream per unit downstream length.
Topographic erosion factor Measure of terrain impact on
erosion rates, computed as a function of water flow proxy
(hillslope length or contributing area) and slope angle.
Universal Soil Loss Equation Simple empirical equation
for estimation of annual soil loss rate for hillslopes with
simple geometry.
Water Erosion Prediction Project Process-based
continuous time modeling system for prediction of
sediment yield and erosion/deposition in small watersheds.
Abstract
This chapter explains the theory and methods for GIS-based modeling of soil erosion, sediment transport, and deposition
by surface water flow. The mathematical foundations of erosion models are introduced and simplified equations,
suitable for GIS implementation, are derived. The presented methods cover modeling of hillslope erosion and deposition,
gully formation, and landscape evolution processes. Coupling of erosion models with GIS is discussed, followed by
examples of GIS implementation of simple and advanced models. The concepts and methods are illustrated using two case
studies, that focus on feedbacks between the human activity and landscape processes.
3.9.1
Introduction
The Earths surface, exposed to gravitational forces, wind,

water, and ice action, continuously evolves over wide range of
spatial and temporal scales. Erosion processes that form the

land surface are extremely complex, poorly understood, and
hard to predict quantitatively over large landscapes (Finlayson
and Montgomery, 2003). Remote sensing provides important
230
data that allow us to gain insights into interactions between

physical processes and environmental conditions that control
erosion and landform evolution. Recent advances in mapping
technologies, such as Light Detection and Ranging (LiDAR),
hyperspectral imaging, and ground penetrating radar have
dramatically increased the spatial and temporal resolution of
Earth surface and shallow subsurface monitoring. The new,
more detailed data indicate that fundamental changes in the
underlaying theory of erosion processes may be needed to
align it with the new observations. Geospatial information
science (GISc)-based analysis and modeling plays an important role in integrating observations and models, and improves
understanding and prediction capabilities aimed at minimizing negative impacts of erosion and sedimentation.
3.9.2
3.9.2.1
Background
Erosion Processes
Erosion encompasses a broad range of processes that involve

soil detachment and transport due to forces that act upon
Earth surface. Remote sensing and field-based geospatial
technology provides effective tools for detecting and mapping
specific landforms that are created by different driving forces,
including (Figure 1):
soil erosion by overland flow (sheet, rill, gully),

channel erosion by fluvial processes,
gravitational erosion, landslides, debris flow,
wind erosion,
coastal erosion by surge and waves,
glacial erosion.
Fundamentals of erosion processes and quantitative modeling are covered in the relevant volumes of this series. Although all types of erosion have been, at least to some extent,
analyzed or modeled using remote sensing data and geographic
information system (GIS) tools, most focus has been on soil
erosion by overland flow, due to its impact on agricultural
productivity, water quality, and sustainable land management.
Therefore, in this chapter, the emphasis is on erosion, sediment
transport, and deposition driven by surface water flow.
3.9.2.2
Spatial Variability
GIS-based modeling of soil erosion investigates spatial patterns of soil detachment, transport and deposition, and their
impact on landscape evolution. Several approaches are used to
develop models of erosion processes: Empirical or statistical
models derive the governing equations from monitoring data
or field experiments using statistical methods such as regression. Rule-based models are constructed from general
observations that relate a combination of inputs to observed
result, implemented as logical operations such as overlays.
Physics-based models use equations derived from natural
laws, such as continuity of mass and energy conservation.
Specific model implementations commonly combine approaches, for example, physics-based, sediment transport
routing with empirically derived parameters (Renschler, 2003;
Mitasova et al., 2005b). Spatial variability in landscape-scale,
soil-erosion processes requires discretization of landscape

representation in GIS-based models.
Spatially averaged hydrologic units are commonly employed in rule-based models (indexes and conditional overlays) or spatially averaged, process-based models. For example,
when modeling erosion by surface water flow, units with
simple geometry (tilted planes or polylines) represent hillslope segments, watershed hierarchies, channels, and stream
networks (Figure 2). The processes are then described by unitto-unit transport rules or by ordinary differential equations for
quasiunivariate transport. This approach is very effective for
systems that include anthropogenic features (ditches, sediment control structures), but selection of suitable spatial units,
their network topology and hierarchies require substantial
expertise and can significantly influence the results (Arabi
et al., 2006). This is true especially for complex, natural
environments that cannot be easily described by simple geometrical features and where spatially averaged models have
limited capabilities to identify precise location of sediment
sources and sinks, and the pattern of their propagation
through landscapes. By reducing the size of the discrete units,
especially in locations with complex topography, spatial pattern of erosion can be captured at the level of detail that approaches the fully distributed models. For example, the
combined Water Erosion Prediction Project (WEPP) and
GeoWEPP model supports modeling at three different levels of
spatial detail watershed, user-defined hillslope segments,
and flow-paths (Renschler, 2003).
Spatially distributed models represent the input variables and
modeled values as continuous fields, usually discretized as
regular grids or irregular meshes. As opposed to the spatially
averaged models that predict erosion rates for an area of a discrete spatial unit, the distributed models predict the modeled
values for any point in space and provide detailed spatial patterns
of sediment sources and sinks (Figure 2). On the other hand,
these models are complex in terms of data inputs and calibration,
and the quantitative summary predictions (e.g., total sediment
loads at a watershed outlet) are not necessarily more accurate
than the predictions by simpler, spatially averaged models.
Erosion and sediment transport is a multiscale process and
different processes require modeling at different spatial resolutions. At hillslope scale, sheet and rill erosion dominate and
their accurate spatial modeling requires submeter resolution.
Watershed scale uses averaged sheet and rill erosion representation and generally captures large gullies and channels.
Depending on the size of the modeled watershed and its
channels, 110 m resolution is required. Regional scale modeling of large watersheds (thousands of square km), averages
sheet, rill and gully erosion, and uses simplified channel
representations and spatially averaged hydrologic units. The
resolutions of input data and results range from 30 m to
hundreds of meters. Irregular meshes and hydrologic units
provide means to adjust the level of detail to the complexity of
landscape features and support spatially variable scale.
3.9.2.3
Temporal Variability
Simple empirical erosion models assume relatively steady

erosion over time and estimate long-term, average annual
231
(a)
(b)
(c)
Figure 1 Erosion caused by surface water flow: (a) rills, (b) gullies, (c) stream bank erosion.
erosion rates based on empirical factors derived from longterm observations. In reality, erosion rates change rapidly due
to the highly dynamic nature of the driving forces, such as
rainfall and water flow, over multiple temporal scales. The full
dynamics of erosion processes are difficult to capture due to
complex, multiscale interactions and limited temporal resolution of the input data. Therefore, modeling is commonly
performed for the steady state case associated with peak flows
during individual erosion events. Sophisticated modeling
systems then perform continuous time simulations of several
interacting processes that incorporate steady-state impacts of
rainfall events over several years. For example, the WEPP
model simulates daily changes in soil and vegetation (crops)

and when rainfall occurs, the plant and soil characteristics are
used to determine whether surface runoff will occur or not. If
runoff is predicted, sheet, rill, and channel sediment detachment and deposition are computed. The model thus simulates
impacts of climate (rainfall, temperature, solar radiation),
hydrology (infiltration, depressional storage and runoff),
water balance (evapotranspiration, percolation, and drainage), vegetation growth (cropland, rangeland, and forest),
erosion (interrill, rill, channel), and deposition (in rills,
channels, and impoundments). In addition to continuous
time simulations, WEPP can be used to assess impact of a
232
(a)
3.2
5.8
2.5
1.4
(b)
500 m
Deposition
20
10
0
(c)
Erosion
(d)
Figure 2 Representation of spatial variability: (a) spatially averaged representation: study area is partitioned into hydrologic units; (b) complex
topography within hydrologic units is replaced by simple geometry of a tilted plane with uniform slope and soil erosion is estimated as uniform
value within each unit; (c) spatially distributed representation estimates topographic erosion index for each grid cell using raster-based flow
routing; (d) spatially distributed erosion and deposition rates estimated by GeoWEPP, red area shows high erosion rates due to vegetation
removal (Moore et al., 2007).
single storm as well as an average long-term impact over

several years.
3.9.2.4
GIS-Based Erosion Modeling
GIS implementation of erosion models supports efficient

management of georeferenced data, computation of input
parameters for different scenarios, spatial analysis of the
modeling results, and effective visualization. GIS also incorporates tools for statistical analysis and modeling of erosion processes captured by remotely sensed data.
In early 1990s, Geographic Resources Analysis and Support
System (GRASS) (Neteler and Mitasova, 2008) provided an
environment for pioneering work in integrating GIS and
hydrologic as well as erosion modeling (Rewerts and Engel,
1991; Mitchell et al., 1993). Most of the geospatial erosion
modeling development and applications focused on agriculture,
soil conservation, sediment pollution control, sustainable
military land management (Harmon and Doe, 2001; Gaffer
et al., 2008), and forestry, especially post-wildfire impacts.
Among the first implementations of hillslope erosion
modeling within GIS was the computation of the Universal
Soil Loss Equation (USLE) (Wischmeier and Smith, 1978),
and derivation of its topographic parameters from digital
elevation models (DEMs). Moore and Burch (1986) and later
Moore and Wilson (1992) paved the way for the USLE applications for landscapes with complex topography by deriving the relationship between the unit stream power theory
and USLE. Although the work generated heated discussions, it
stimulated development in the direction that resulted in numerous implementations of USLE with GIS support for complex topographic conditions (Mitasova et al., 1996; Desmet
and Govers, 1996). More recent GIS applications of USLE
cover wide range of scales including large watersheds with
land cover mapped from remote sensing imagery (Suri et al.,
2002; Cebecauer and Hofierka, 2008; Pandey et al., 2009a, b;
Jain and Das, 2010).
USLE and its updated, improved versions (Revised Universal
Soil Loss Equation RUSLE and RUSLE2) have been used to estimate soil detachment within watershed-scale models that focus
on nonpoint source pollution in agricultural, environmental,
and engineering applications, including well known models
such as ANSWERS, AGNPS, and SWAT (Rewerts and Engel, 1991;
Mitchell et al., 1993; Lim et al., 2005; Kim et al., 2009). These
models have been coupled with proprietary and open source
GIS to process and manage modeling inputs and analyze the
outputs (Table 1). Several watershed modeling systems have
been recently implemented as on-line tools (Park et al., 2009).
Although USLE was developed as a simple tool for farmers
to encourage soil conservation and thus its representation of
erosion processes has been greatly simplified, the needs for
research level modeling stimulated development of more
complex models and their coupling with GIS. A process-based,
continuous time model, based on the work by Foster (1982),
was developed as WEPP (Flanagan and Nearing, 1995;
Flanagan et al., 2007). Adaptation of WEPP to GIS, which led
to the development of Geo WEPP (Renschler, 2003), traces the
flow of sediment along flow paths on hillslopes generated
from a DEM. Integration of water and wind erosion models
233
Table 1 Selected soil erosion models and their coupling with GIS; s, r sheet and rill erosion; g gullies; ch channel; fl fluvial; dz
elevation change; see also US Department of Agriculture, Agricultural Research Service (USDA ARS) at http://www.ars.usda.gov/Research/
docs.htmdocid=5971, SWAT at http://swatmodel.tamu.edu/, EUROSEM at http://www.eurosem-soil-erosion.org/, LISEM at http://www.itc.nl/lisem/
Models
Processes
Spatial scale and representation
Temporal scale
GIS implementation
Reference
USLE
RUSLE
GeoWEPP
AnnAGNPS
ArcSWAT
MapWinSWAT
openLisem
SIBERIA
CHILD
USPED
SIMWE
r.landscape.evol
s,
s,
s,
s,
s,
s,
s,
s,
s,
s,
s,
s,
hillslope profile
hillslope profile
small watershed raster
watershed custom grid
watershed hydrologic unit
watershed hydrologic unit
watershed raster
landscape raster
landscape mesh
landscape raster
annual
annual event
continuous time
continuous time
continuous time
continuous time
event
continuous time
continuous time
annual event
event
continuous time
map algebra
map algebra
extension, Web
custom GIS
extension
custom Web GIS
PCRaster
data input
data input
map algebra
GIS module
GIS module
USDA ARS
USDA ARS USDA
ARS, NSERL USDA
ARS
USDA ARS, TAMU
USDA ARS, TAMU
De Roo et al. (1996)
CSMDS (2011)
CSMDS (2011)
Neteler and Mitasova (2008)
Neteler and Mitasova (2008)
Barton et al. (2010a)
r
r
r,
r,
r,
r,
r,
r,
r,
r
r
r,
ch
ch
ch
ch
ch
ch, dz
ch, dz
ch, dz
within the WEPP modeling framework is under development.

It represents an important step toward understanding of
combined impacts of water and wind erosion on soil conservation and agricultural sustainability.
Several erosion modeling systems with strong geospatial
components have been developed in Europe. European Soil
Erosion Model (EUROSEM) (Morgan et al., 1998) introduced
a dynamic approach for predicting sediment transport from
small watersheds. The LImburg Soil Erosion Model (LISEM)
(De Roo et al., 1996; Sheikh et al., 2010) is a physically based
runoff and erosion model for research, planning, and conservation purposes. It simulates the spatial effects of rainfall
events on small watersheds and it uses the free GIS PCRaster
Environmental Software (Karssenberg et al., 2001). As one of
the most comprehensive geospatial erosion modeling tools,
LISEM incorporates rainfall, interception, surface storage in
microdepressions, infiltration, vertical movement of water in
the soil, overland flow, channel flow in man-made ditches,
detachment by rainfall and throughfall, transport capacity,
and detachment by overland flow. Influence of compaction on
the hydrological and soil erosion processes and gully incision
are also included.
The SIMulation of Water Erosion (SIMWE) model (Mitas
and Mitasova, 1998; Mitasova et al., 2005b) was developed as
a bivariate generalization of hillslope erosion model used by
WEPP to capture the spatial pattern of erosion, sediment
transport, and deposition under spatially diverse conditions.
Its robust path sampling algorithm and integration within GIS
supports simulations of water and sediment flow patterns at
high spatial resolutions, including impacts of small terrain
variations such as depressions, ditches, or, checkdams.
Although several GISc-based implementations of erosion
models predict locations of high erosion due to concentrated
water flow, specific models that focus on ephemeral gully erosion were also developed (DeRose et al., 1998; Woodward,
1999). These models rely on field observations and modified
channel sediment transport equations to improve the quantitative estimates of sediment eroded by ephemeral gully formation.
Time series of aerial photography combined with multitemporal
DEMs were also used to map the location, evolution and erosion
rates of gullies (Martnez-Casasnovas et al., 2004).
event
event
event
event
In addition to the models focused on short term erosion

(single event to multiple years), landscape evolution models
have been developed to capture impact of erosion on landforms over hundreds or thousands of years. These models
commonly use GIS data as inputs, but they have been developed and used mostly outside GIS and are covered in respective volumes of this series. Well known models include
SIBERIA (Willgoose et al., 1991; Hancock et al., 2002;
Willgoose, 2004), CHILD (Tucker et al., 2001a, b), or
rule-based cellular automata models for simulation of landforms (Luo, 2001), braided rivers evolution (Murray and
Paola, 1994) and sand dunes (Pelletier et al., 2009). The
WILSIM landform evolution model (Luo et al., 2004) was
implemented as a web-based simulation tool. Recently, the
development of landscape evolution models has been coordinated by the Community Surface Dynamics Modeling
System (CSMDS, 2011) program.
Statistical approaches have been commonly combined with
remote sensing techniques to improve the results of erosion
mapping. For example, a logistic regression-based erosion
index was developed to map probability of erosion resulting
from concentrated water flow from high resolution DEMs
(Pike et al., 2009). Tree-based regression models were used to
identify the topographic parameters that explain the variability
in field gully measurements (Kheir et al., 2007). The most recent GIS-based models study pattern of erosion and deposition
rates at very high resolutions using repeated scans of landscape
by terrestrial LiDAR and by differencing the resulting DEMs.
Several recent papers indicate that there are major weaknesses in the current understanding and data underpinning
existing models (Govers et al., 2007; Wainwright et al., 2008;
Polyakov et al., 2004; Finlayson and Montgomery, 2003; Jetten
et al., 2003). For example, Van Oost et al. (2005) concluded
that performance of process-based erosion and hydrological
models is extremely sensitive to parameter estimations and that
predictions are generally poor. The difficulties associated with
accuracy of spatially distributed soil erosion models are commonly due to the spatial and temporal variability of erosion
processes and uncertainty associated with the model parameters. Jetten et al. (2003) suggest that model performance
may be improved by using more complete spatial information
234
for model calibration and validation instead of the data restricted to the measurements at watersheds outlets. Spatially
and temporally distributed observations may provide the data
to elucidate the poorly understood interactions between sediment load and detachment rates, and provide insights into
scaling of erosional processes and mechanisms. Therefore, adjustments to the theoretical foundations of erosion modeling
presented in the next section can be expected in future as new
monitoring and sensing technologies provide more complete
experimental data and field observations.
3.9.3
Foundations in Erosion Modeling
Spatial and temporal patterns of erosion and deposition

are results of complex interactions between the Earth
system processes such as rainfall, surface, subsurface and
ground water flow, vegetation growth, soil detachment,
transport, and deposition. The focus of this section is on soil
and sediment erosion and transport while treating rainfall,
runoff, vegetation cover, and soil properties as inputs without
going into detail about the estimation of these parameters
values. Here, general mathematical representation of erosion
and sediment transport processes is described and relation
between models with different levels of complexity and
backgrounds is derived, highlighting the common principles.
3.9.3.1
Sediment Transport and Net Erosion/Deposition

Equations
Soil, detached by raindrop impact and shearing force of overland

flow is transported by flowing water until its transport capacity
decreases, leading to sediment deposition (Haan et al., 1994).
The fundamental relationship that governs sediment transport
by overland flow is continuity of mass. It can be described by the
sediment continuity equation, which relates the change in
sediment storage over time, and the change in sediment flow rate
along 3D hillslope to effective sources and sinks (Haan et al.,
1994; Govindaraju and Kavvas, 1991; Foster and Meyer, 1972).
To capture the spatial pattern of sediment transport over complex topography the bivariate form of the continuity of sediment
mass equation (Hong and Mostaghimi, 1997) is used
q rs x,y,thx,y,t
r qs x,y,t sources sinks
qt
ds x,y,t,
1
where (x, y) are georeferenced coordinates, t (s) is time,

rs(x, y, t) (kg m3) is sediment mass density, h(x, y, t) (m) is
water flow depth, qs(x, y, t) (kg (ms)1) is the vector that
represents the direction and rate of sediment flow per unit
width (unit sediment load), r denotes divergence of the
sediment flow vector field, and ds(x, y, t) (kg m2 s1) is the
net erosion or deposition rate. The sediment flow rate is a
function of water flow and sediment concentration:
qs x,y,t rs x,y,tqx,y,t,
2
where q(x, y, t) (m2 s1) represents the direction and rate of

water flow per unit width (unit flow discharge vector) that can
be expressed as a function of water depth h(x, y, t), where the

form of this function depends on the channel cross-section
and flow conditions (Haan et al., 1994).
A simplified, steady state form of the continuity equation
can be derived for the conditions when, at the given location,
the change in water flow and sediment concentrations over
time is close to zero. The net erosion or deposition rate ds(x, y)
is then computed as a divergence of steady state sediment flow
rate per unit width qs(x, y):
q rs x,y,thx,y,t
0 - r qs x,y ds x,y:
qt
3
Assuming steady rainfall excess rates, eqn [3] applies to

the peak flow and peak concentration conditions when
the water depth and sediment concentrations remain
constant over time. To simplify equations, the (x, y) notation
indicating spatially distributed variables represented by
bivariate continuous functions is omitted in the rest of this
chapter.
The sources and sinks term in eqn [1] is derived from the
assumption that the detachment and deposition rates are proportional to the difference between the sediment transport capacity and the actual sediment flow rate (Foster and Meyer, 1972):
ds sTc qs ,
4
where Tc (kg(ms)1) is the sediment transport capacity,

qs 9qs9 is the magnitude of sediment flow rate per unit width,
and s (m1) is the first-order reaction term dependent on soil
and cover properties. The expression for s can be obtained
from the following relationship (Foster and Meyer, 1972):
ds qs
1
Dc Tc
5
which states that the ratio of the erosion rate to the detachment capacity Dc(kg m2 s1) plus the ratio of the sediment
flow rate to the sediment transport capacity is a conserved
quantity (unity). Equation [5] is based on the observed relationship between soil detachment and transport when the finite
amount of energy available in the overland flow is applied
proportionally to detaching and to transporting the sediment so
that the total relative available energy is unity (Haan et al.,
1994). The s coefficient thus controls how close is the sediment
transport to the detachment limited (erosion only) or transport
limited (maximum extent of deposition) regime (Figure 3). The
detachment capacity Dc is then proportional to the transport
capacity Tc:
Dc sTc :
6
This leads to the expression of net erosion and deposition as:

qs
:
ds Dc 1
Tc
7
This concept is used in several erosion models including

WEPP (Haan et al., 1994; Flanagan and Nearing, 1995) and
SIMWE (Mitas and Mitasova, 1998). To estimate deposition
of particles with different settling velocities s may be
235
= 0.001
0
1.0
10.0 100.0
Sediment flow rate
Erosion Deposition
(a)
= 100.0
300 m
(b)
Figure 3 Impact of s on spatial distribution of sediment flow rate (visualized as a surface) and erosion and deposition (draped as a color map
over DEM): (a) low value of s 0.001, with Dc{Tc leads to prevailing erosion, close to the detachment limited regime, (b) high value of s 100,
DccTc leads to large extent of deposition, close to the transport capacity limited regime.
approximated by (Foster, 1982):
Vs
,
2q
8
where Vs (m s1) is settling velocity for the given particle size

and q (m2 s1) is flow discharge per unit width. It is possible
to use other frameworks for estimation of s. For example,
Govers et al. (2007) proposed to express the detachment DL as
function of unit length sheer force G:

qs
,
9
DL aG 1
Tc
where a is an empirical coefficient.
It is important to note, however, that the relationship between the sediment detachment and transport capacities and
the actual detachment and sediment loads is not fully
understood (Govers et al., 2007; Nearing et al., 1997). Further
experimental research is needed to elucidate this relationship
and develop the underlying theory and equations.
3.9.3.2
Detachment and Sediment Transport Capacities
To solve the equation for sediment transport rate, it is necessary to estimate the transport and detachment capacities. The
sediment transport capacity Tc and detachment capacity Dc
represent the maximum potential sediment flow rate and the
maximum potential detachment rate, respectively. Numerous

simplified empirical equations represent these rates under
different conditions (Julien and Simons, 1985). They are often
expressed as functions of shear stress (Foster and Meyer,
1972):
10
Tc Kt ta ,
Dc Kd t t0 b ,
11
where Kt(s) is the effective sediment transport capacity

coefficient, Kd(s m1) is the effective erodibility (detachment
capacity coefficient), t(Pa kg m2) is the shear stress,
t0(Pa) is the critical shear stress, and a and b are empirical
exponents. The shear stress t is function of water depth h and
surface slope angle b(deg) such that
t gw RS;
12
where gw rwg is the hydrostatic pressure of water with the

unit height, g 9.81 (m s2) is the gravitational acceleration,
rw 103 (kg m3) is the mass density of water, S tanb is
surface slope and b (deg) is the steepest slope angle (see Haan
et al. (1994), Moore and Burch (1986) for discussion on relationship between water depth and wetted perimeter used in
the shear stress and stream power equations).
Transport capacity, especially in channels, streams and
rivers, can be expressed as a function of unit stream power
236
o (J (ms)1) (Moore and Burch, 1986) that represents the rate

of energy dissipation against the bed and banks of a channel.
If the stream velocity v (ms1) is estimated using Mannings
equation (Haan et al., 1994; Dingman, 2002), then
v n1 h0:6 S0:5 ;
13
where n is Mannings coefficient, then the unit stream power

can be expressed as:
o tv gw n1 h1:6 S1:5 ;
14
Sediment transport capacity is then computed as

Tc Ks o Ks n1 gw hm Sn ,
15
where Ks is transport capacity coefficient. Several studies indicate that stream power is a better hydraulic predictor variable for detachment and sediment yield than is shear stress
(Nearing et al., 1997, 1999). The equation for sediment
transport can significantly influence the magnitude and spatial
pattern of sediment flow rates as well as the predicted erosion
and deposition (Figure 4). Field observations are generally
needed for selection of the sediment transport equation and
its parameters that adequately represents the specific modeled
landscape.
Julien and Simons (1985) analyzed numerous sediment
transport equations and derived the following general equation
qs jid qm Sn 1 t=t0 e ,
16
where q is unit water discharge, i is rainfall intensity, and j, d,

e, m, n are experimental or physically based coefficients that
depend on type of flow.
When the critical shear stress t0 is negligible and for
channels where d 0 and qs is not dependent on rainfall intensity the equation simplifies to
qs jqm Sn :
17
As presented above, the stream power, shear stress, and

consequently transport and detachment capacities are generally expressed as power functions of water flow properties and
slope. For a given rainfall excess rate and surface roughness,
these two variables can be derived from DEM using the GIS
functions for topographic analysis to compute the slope, and
the flow routing tools to compute the upslope contributing
area as an input for estimating unit water flow and
water depth.
The parameters Kt, Kd, t0 used for the estimation of Dc and
Tc are functions of soil and land-cover properties, and are
much more difficult to estimate accurately than the topographic parameters. They have been derived using empirical
equations or directly from experiments for a wide range of
soils, cover, agricultural, and erosion prevention practices
within the WEPP model (Flanagan and Nearing, 1995), but
the values are valid only for the conditions similar to those in
the experiments. Moreover, for landscape scale modeling, the
soil data are rarely available at the level of detail comparable
with DEMs, and temporal changes in soil properties are generally neglected. Land-cover data can be obtained at very high
resolutions from airborne or even satellite imagery, however,
the relationships between the type and density of vegetation
cover and the soil detachment and transport parameters are
not very well understood.
In addition to the difficulties of accurate estimation of
detachment and transport capacity parameters, the values of
exponents a, b in the eqns [10] and [11] or m, n in eqn [16]
depend on the type of flow (Julien and Simons, 1985) and
substantially influence model behavior. For example, the
WEPP model uses a 1 and b 1.5 which means that with
increasing water flow, transport capacity increases faster than
detachment, which is not always the case. Uniform values of
these exponents are commonly not general enough for erosion modeling at landscape scale with different types of flow
due to variability in land-cover properties and spatially variable exponents may be needed (Mitas and Mitasova, 1998).
The solution of eqn [1] can be obtained by several partial
differential equation solvers, including finite element methods
(Hong and Mostaghimi, 1997), finite difference methods
(Desmet and Govers, 1995), or path sampling (Mitas and
Mitasova, 1998; Mitasova et al., 2005b). Detailed description
15
10
5
0
N
(a)
(b)
(c)
500 m
Figure 4 Impact of the sediment transport equation on spatial distribution of sediment flow rate: (a) shear stress eqn [10] with a 0.6 predicts
sediment transport increasing slowly with concentrated flow; (b) with a 1.5 sediment transport increases rapidly with concentrated water flow;
(c) stream power eqn [15] further increases the difference between sediment transport by sheet and concentrated flow.
of these methods is beyond the scope of this chapter, however,

in the following sections, various simplifications of the full
solution useful for GIS-based soil erosion modeling are
described.
3.9.4
Simplified Models of Erosion Processes
To satisfy the need for models which are easy to compute and
for which data are readily available, equations for simplified
models of erosion processes can be derived. Depending on
interactions between rainfall, runoff, and local soil and cover
conditions, erosion and sediment transport varies between
two limiting cases (Foster and Meyer, 1972; Tucker et al.,
2001a):
detachment capacity limited,

sediment transport capacity limited.
Deriving equations for these limiting cases by simplifying
the general sediment transport equation, leads to well-known
soil erosion models that are useful for estimates of average
annual soil erosion rates or total sediment loads at watershed
outlets. In addition to the limiting cases of sheet and rill
erosion, gully formation and a landscape evolution model
that integrates several erosion processes are also discussed.
3.9.4.1
the original USLE has the form (Wischmeier and Smith,

1978):
LS L=L0 m 64:5 sin2 b 4:56 sin b 0:0654,
ds EDc Kd tb Kd gw hSb :
18
This equation can then be used to compute the detachment

limited, steady state erosion with water depth approximated,
for example, from upslope contributing area. If more accurate
estimate is needed, water depth distribution can be computed
by a hydrologic model.
If the estimate of water depth is further simplified by assuming a planar hillslope with no water flow divergence or
convergence, upslope contributing area per unit width can be
replaced by hillslope length, leading to expression that is the
basis for the most common erosion model USLE and its
revised version RUSLE
D RKLSCP,
19
where D (kg m2 yr1) is average annual soil loss, R (M J mm

(ha hr yr)1) is rainfall factor, K (ton ha hr (ha M J mm)1) is
soil erodibility factor, LS is a dimensionless topographic
(length-slope) factor, C is a dimensionless land-cover factor,
and P is a dimensionless prevention measures factor. The
length-slope factor has been derived from experiments and in
20
where L (m) is the hillslope length, L0 22.1 m is the length of

the standard USLE experimental plot, m is exponent with
values ranging between 0.2 for slopes less than 1% and 0.5 for
slopes steeper than 5%. RUSLE expands the applicability of
this equation by introducing additional empirical equations
for computation of LS factor for different slope steepness
intervals and provides updated values and equations for
computation of all factors, including the R-factor for individual storms (Renard et al., 1994).
The USLE/RUSLE has been developed for estimation of
average sheet and rill erosion on uniform fields with simple,
relatively planar geometry. Numerous efforts and implementations have extended its applications to large watersheds and
fields with complex topography. For example, USLE is used in
several nonpoint source pollution models for estimation of
averaged soil detachment in hydrologic units, with the detached soil then routed through the watershed.
An important modification of USLE/RUSLE has been derived by Moore and Burch (1986) and applied within GRASS
GIS (Mitasova et al., 1996). This improvement replaced the
hillslope length by upslope contributing area (Figure 5),
leading to the following equation:
Detachment Capacity Limited Case
When transport capacity of overland flow exceeds its detachment capacity over the entire studied landscape, erosion, and
sediment transport is detachment capacity limited (i.e., no
deposition occurs). For example, this case is typical for sediment transported by large amount of water over compacted
soil. The detachment capacity limited case is represented by
TccDc leading to s - 0. Assuming that the critical shear stress
is negligible t0 0 the net erosion will be equal to the detachment capacity:
237
D RKCPm 1
m

U
sinb n
,
L0
S0
21
where D(kg m2 yr1) is the average annual soil detachment

(soil loss) rate, U(m2 m1) is the upslope contributing area
per unit width (a proxy for discharge), and S0 0.09 is the
slope of the standard USLE plot. Single storm and monthly R
are also available, making eqn [21] suitable for estimation of
soil loss for single storms and for modeling of monthly soil
loss distribution over a year (Haan et al., 1994).
Exponents m, n depend on the prevailing type of erosion
(sheet, rill) and the typical values are m 0.40.6 and
n 1.01.3. Replacement of slope length, used in the original
formulation of USLE/RUSLE, by the upslope area predicts
increased erosion due to the concentrated flow without the
need to a priori define these locations as inputs for the model.
It is important to note that the USLE/RUSLE model is
commonly applied within GIS for conditions for which it has
not been originally designed and at scales at which its validity
is uncertain. For example, the original values of the USLE/
RUSLE factors have been derived from experiments using plots
only 22.13 m long, however, many published GIS applications
of USLE used the equations at resolutions with much larger
grid cell size and at locations with much steeper topography
than the 9% slope of the USLE plots (Pandey et al., 2009a).
The authors generally justify such applications by the fact that
comparable experiments were not performed for the studied
conditions and USLE was the best option available to them.
Moreover, in complex topography, detachment limited erosion is rather rare because concave areas at toes of hillslopes
reduce transport capacity, leading to sediment deposition,
further limiting the applicability of the original USLE/RUSLE
238
t ha1 yr1
0
6
12
18
(a)
Modeled soil erosion
Observed sediment deposition
300 m
(b)
Figure 5 Spatial pattern of topographic erosion factor based on: (a) flowpath length (eqn [19]), which does not capture increased erosion due
to convergent water flow, (b) flow accumulation (eqn [21]), which incorporates impact of water flow convergence. The topographic factor maps
are draped over elevation surface. The cross-section shows observed deposited material, indicating locations where these topographic erosion
factors are not applicable for modeling long term net erosion/deposition pattern. Images based on data provided by Prof. K. Auerswald,
Technische University Muenchen.
(Figure 5). In spite of its limitations, the USLE/RUSLE models,

as well as their modifications and applications in GIS have
played an important role in soil conservation efforts and
sustainable land management (Renard et al., 1994; Haan
et al., 1994; Hammad et al., 2004; Cebecauer and Hofierka,
2008).
3.9.4.2
qs ETc ERKCPU m sinbn :
Transport Capacity Limited Case
If the soil detachment significantly exceeds the sediment

transport capacity of overland flow, the sediment flow rate will
be at the sediment transport capacity. This means that the
magnitude of sediment flow can be approximated by Tc and
net erosion/deposition rate can be computed as a change in
the sediment transport capacity derived from eqn [3]:
ds r qs Er Tc s0
qTc cosa qTc sina
,
qx
qy
The transport capacity limited case with spatially variable

soil and cover properties can be modeled using the idea originally proposed by Moore and Burch (1986). It combines the
USLE/RUSLE parameters and upslope contributing area per
unit width to estimate the sediment flow at sediment transport
capacity:
22
where s0 (cos a, sin a) is the unit vector in the steepest slope

direction given by a(deg), the aspect angle of the terrain surface equivalent to the direction of flow. As opposed to the
detachment capacity limited case, the transport capacity limited formulation can predict spatial pattern of both erosion
and deposition (Figure 6).
23
The net erosion/deposition ds(kg m2 s1) is then computed as a divergence of sediment flow vector field given by
eqn [22].
The exponents m, n control the relative influence of water
and slope terms and reflect the impact of different types of
flow. The observed extent of colluvial deposits indicated that
an exponent m 1 reflects the pattern of compounded, long
term impact of both rill and sheet erosion (Mitas and Mitasova, 1998). Averaging over a long-term sequence of large and
small events was important for capturing the observed spatial
extent of deposition as well as channel incision due to concentrated flow (Warren et al., 2005). Spatially variable m(x, y),
n(x, y), approximated by a continuous function, can be used
to account for the differences in surface flow (Figure 7).
Models representing limiting cases of erosion are simple
to compute in a GIS using the standard flow routing,
239
t ha1 yr1
6
Deposition
0
Erosion
6
N
0
300 m
Observed sediment deposition

Figure 6 Topographic potential for net erosion and deposition, estimated as a change in sediment transport capacity.
Erosion
Deposition
(a)
(b)
500 m
(c)
Figure 7 Impact of exponents m, n in eqn [23] on resulting erosion and deposition rates pattern: (a) m 1.0, n1.0 predicts erosion in
87% and deposition in 13% area, including deposition in concave forms of headwaters; (b) m 1.6, n1.0 predicts erosion in 92% area
whereas limiting deposition to 8% area mostly in low, broader concave forms; (c) spatially variable m 1 at lower and m 1.6 at higher
elevations.
topographic analysis modules combined with map algebra.

They can be used to estimate spatial distribution of soil
detachment and net erosion/deposition rates for a single
storm, as well as monthly and annual averages. Caution
should be used when interpreting the results from modifications of USLE/RUSLE that incorporate water flow convergence/divergence and erosion/deposition, because the
soil and cover parameters were developed for simple planar
fields and detachment limited erosion. To accurately predict
erosion and deposition rates for complex terrain and spatially
variable land-cover conditions these models need to be
calibrated for a specific geographic area.
deposition as a difference between sediment inflow and outflow along the segments of this path (Moore and Wilson,
1992; Desmet and Govers, 1995; Mitasova et al., 1996). Assuming uniform rainfall, soil and cover conditions, and a
transport capacity limiting case with qsETc, the net erosion/
deposition rate along a flow path (univariate model) can be
derived as directional derivative of sediment transport capacity
in the direction of steepest slope (aspect):
ds
dTc
rTc s0
ds
ds Ke rh s0 sinb hkp ,
3.9.4.3
Process-Form Relationship
As shown in the section 3.9.4.2 topography controls the spatial pattern of sediment transport and erosion/deposition
through upslope contributing area and slope angle. In this
section, the role of terrain curvature in distribution of erosion
and deposition is demonstrated.
Most sediment transport models route sediment in single
direction along flow paths and compute the net erosion and
24
25
where kp(m1) is profile curvature that measures the rate

of change in slope in the gradient direction, Ke Kt rw g, and
ds40 represents the net erosion rate and dso0 represents
the net deposition rate. The univariate formulation includes
the impact of water flow convergence/divergence through the
water depth term and flow acceleration/deceleration through
the profile curvature.
Within the bivariate formulation, given by eqn [22], the
net erosion and deposition rate is estimated as a divergence of
240
the sediment flow vector field qs (Mitas and Mitasova, 1998):

ds r qs r Tc s0
26
ds Ke rh s0 sinb hkp kt ,
27
where kt (m1) is the tangential curvature that measures the

change in aspect (curvature in the direction perpendicular to
the gradient, i.e., the direction tangential to a contour line
projected to the normal plane).
When eqns [25] and [27] are compared, the fundamental
difference is in incorporation of tangential curvature. According to the bivariate eqn [27], the spatial distribution of
erosion/deposition is controlled by the change in the overland
flow depth rh and by the local geometry of terrain including
both profile and tangential curvatures. Equation [27] thus
demonstrates that the local acceleration of flow in both the
gradient and tangential directions play equally important roles
in spatial distribution of erosion/deposition.
The impact of the tangential curvature is therefore twofold
(Figure 8). First, kt influences the water depth through its control
of water flow convergence and divergence, with tangential concavity leading to rapid increase in water depth and increase in
erosion rates. Second, kt causes a local change in sediment flow
velocity with an opposite effect, with tangential concavity reducing the sediment transport, creating conditions for deposition. Therefore, it is the interplay between the magnitude of
water flow change and both terrain curvatures in eqn [27] which
determines whether erosion or deposition will occur.
When the results of the univariate and the bivariate models
were compared with the observed pattern of deposits (Mitas
and Mitasova, 1998), the model that computes the sediment
(a)
load change along a flowline (eqn [25]) failed to predict

deposition observed in areas where kpE0 but kto0.
The prediction by eqn [27] was in better agreement with
the observed pattern of deposition in these areas (Figure 8). It
is important to note, however, that the total sediment load at
the outlet is the same for both formulations and it is only the
spatial pattern of sediment redistribution within the watershed that is influenced by the univariate versus bivariate
solution.
3.9.4.4
Path-Sampling Transport Modeling
Evolution of sediment transport over complex landscapes can

be simulated using a path-sampling approach (Mitas and
Mitasova, 1998; Mitasova et al., 2005b) that is based on the
concept of duality between discrete particles and continuous
fields. The approach solves the steady state sediment flow
eqn [3] that is rewritten to include a small diffusion term
pr2r:
g
r2 r r rv rs9v9 sTc ,
2
28
where r rsh(kg m2) is the mass of sediment carried by

water per unit cross-section area, g(m2 s1) is the diffusion
constant, and v(m s1) is the water flow velocity estimated
by Manning or Chezy equation (Dingman, 2002). On the left
hand side of eqn [28] the first term describes local diffusion,
the second term is a drift driven by the water flow whereas the
third term represents a velocity dependent potential acting
on r. The size of the diffusion constant is about one order of
magnitude smaller than the reciprocal Mannings constant so
that the impact of the diffusion term is relatively small.
(b)
(c)
80.00
Erosion Deposition
50.00
25.00
0.00
(d)
300 m
(e)
(f)
Figure 8 Computing net erosion and deposition using univariate and bivariate formulation (eqns [25] and [27]): (a) term representing change in
water-flow depth (rh s0 sinb); (b) term that combines water-flow depth with profile curvature (hkp); (c) term that combines water-flow depth with
tangential curvature (hkt); (d) net erosion/deposition computed as a change in sediment flow along flow-path is a sum of (a) and (b); (e) net erosion/
deposition as sediment flow divergence is a sum of (a), (b) and (c); (f) observed deposited sediment with highlighted locations where divergence
improves the modeling result. Figure based on Mitas and Mitasova (1998) using data provided by Prof. Auserwald, Technical University Muenchen.
241
0.2
0.1
0.02
0.001
(a)
(b)
1 km
t ha1 yr1
Depo>30
sition
0
Erosion
<30
kg m1s1
10
1
0.1
0.01
(c)
(d)
Figure 9 GIS-based erosion modeling using path sampling method: (a) C-factor at the study site with disturbed land shown in red; (b) particle
representation of soil detachment (only 1% of particles is displayed), particle density is higher in locations with disturbed land; (c) raster
representation of sediment flow rate computed as a function of particle density per grid cell; (d) net erosion and deposition computed as a
divergence of sediment flow.
It represents local dispersion processes caused by microtopography which is not captured by the DEM.
The sediment flow described by eqn [28] can be solved by
the path-sampling stochastic method (Mitasova et al., 2005b)
by propagation of particles according to the continuity equation. Sediment concentrations are then computed based on
particle density (Figure 9). This approach has been used to
develop the SIMWE model which was implemented in GRASS
GIS as a module r.sim.sediment (Neteler and Mitasova, 2008).
3.9.4.5
Gully Erosion
Traditionally, modeling of gully and channel erosion has been

performed outside of a GIS because most of the required data,
especially the DEMs, did not have sufficient resolution to
capture the gully or channel geometry. With high resolution
DEMs (submeter grid cell size) now available, gully and
channel erosion models can be implemented and applied
using GIS-based approaches.
Gullies develop under specific conditions when the tractive
force of the water flow exceeds a certain threshold value and
shallow flow starts to concentrate in the channel. This transition occurs abruptly at a knickpoint or headwall. In fact,
several such transitions may occur, eventually merging the
knickpoints and forming one incised channel (Haan et al.,
1994). The gully then develops via headwall migration and
channel widening. The process of initiation requires a relatively large rainfall event so that the rainfall excess generates
water flow exceeding the threshold value needed to detach and
transport large amounts of sediment. The immediate morphological effect of the gullying process is a substantial change
in topography with a sharp change in slope in a headwall and
channel banks. This contributes to a further gully development until a new equilibrium is found.
The difference between modeling the rill and gully erosion
lies in the treatment of elevation represented by a DEM. In the
case of landscape-scale rill erosion modeling, the elevation
surface is considered as fixed throughout the erosion event.
Actual net erosion/deposition rates change the elevation surface only marginally and the morphological effect can be
observed only after many similar events. Gully erosion, however, has an immediate effect on topography and the elevation
change must be taken into account during the simulation.
242
0.4
0.2
0
0.2
0.4
(a)
350
320
190 m
120
120
300 m
(b)
Figure 10 Temporal evolution of a gully, modeled in GRASS GIS: (a) elevation surface with evolving gully, color indicates spatial pattern of
erosion/deposition rates in (kg m 2 s 1), (b) evolving elevation surface: creation of multiple knickpoints, incision of the channel, upslope
headwall migration and channel widening, color represents the elevation values in (m).
This inherently requires a dynamical erosion model reflecting

the changes in topography. To simulate the gully formation,
it is assumed that erosion starts before water flow reaches
steady state, the sediment transport is close to detachment
limited conditions, and all eroded material is transported
outside the gully. Using eqn [1] to estimate erosion rates at
time t(s) driven by the unit water flow qt (at nonsteady state
for the time steps t0, t1,y, ts, where ts is time of concentration)
the change in elevation Dz(x, y, t) (m) due to net erosion can
be estimated as
Dzx,y,t Dt ds x,y,t=rs ;
29
where rs(kg m3) is soil density and Dt ti ti1 is the time

interval. Using the map algebra in a GIS, the initial DEM0 is
modified using the changes in elevations Dz(x, y, t). A new set
of topographic parameters is then derived from the updated
DEMi to reflect new flow gradients used by the water flow and
soil erosion model in the next iteration. Figure 10 shows a
time-series of a developing gully modeled by the SIMWE
model (Mitasova et al., 2005b) in a GRASS GIS environment
(Koco, 2009). It illustrates creation of multiple knickpoints,
incision of the channel and then an upslope headwall
migration and channel widening.
3.9.4.6
Statistical Modeling
As previously shown, even the physics-based erosion models

require a set of empirical parameters that are derived from
field or laboratory experiments using statistical methods.
Statistical modeling can also be used to derive empirical
models or indices that relate landscape surface attributes to
erosion rates. The resulting models require field or remotely
sensed data and are applicable only to locations with conditions similar to those for which they have been developed.
For example, Pike et al. (2009) used logistic regression and
neural networks to derive a probabilistic model that predicts
the occurrence of channel erosion using the index Pe:
Pe 1 exp3:631:11LS0:217w12:1kh 1 ,
30
where LS is the estimated length-slope factor, w is the topographic wetness index, and kh is the plan curvature (tangential
curvature projected to horizontal plane). Because this is an
empirical model, the authors make it clear that the model
should only be applied if the same methods were used to
create the DEMs and calculate terrain attributes as in their
application.
An alternative statistical approach explored the capabilities
of three tree-based regression models to explain gully erosion
field measurements using topographic parameters. The investigated parameters included elevation, upslope contributing area, aspect, slope, plan, profile, and tangential
curvature; flow direction, flow width, flow path length, rate of
change of specific catchment area along the direction of flow,
steady-state, quasidynamic topographic wetness; and sediment
transport capacity (Kheir et al., 2007). The best regression tree
model combined the steady-state topographic wetness and
sediment transport capacity indices, which explained 80% of
the variability in field gully measurements.
3.9.4.7
Landscape Evolution Modeling
Most GIS-based erosion models focus on mapping the spatial

pattern of erosion and deposition without taking into account
changes in elevation surface. Landscape evolution models
simulate erosion processes including their impact on elevation
surface and their feedback on erosion and deposition patterns.

Landscape evolution process can be described by partial differential equations (Willgoose, 2004) and most models are
implemented outside a GIS using customized code or numerical modeling environment such as MATLAB. An exception is a
landscape evolution model that simulates impact of land use
and climate change on landscape evolution r.landscape.evol that
has been fully integrated with GRASS GIS (Barton et al., 2010a).
Simulation of landscape evolution over long periods of
time requires integration of several erosion processes. These
include change in elevation surface due to gravitational hillslope erosion modeled using diffusion equation, processes
induced by surface water flow, including hillslope erosion and
deposition, gully erosion, channel incision, and meandering
in regions with low topography.
An approach inverse to the forward landscape evolution
modeling was used to reconstruct ancient topography (Peeters
et al., 2006) by computing soil erosion and deposition based
on change in sediment flow rates estimated from slope and
unit contributing area. Then, the topography was iteratively
updated back in time by adding the elevation associated with
erosion rates and subtracting elevation in areas with deposition. Comparison with field measurements of historical
soil erosion and sediment deposition volumes shows that it is
possible to simulate realistic soil redistribution patterns. Further research is necessary, however, to address simulation artifacts produced by the routing and elevation change
computation used in this study (Peeters et al., 2006).
Several papers suggest the importance of large events in
landscape evolution (Peeters et al., 2006; Mitasova et al.,
1999) when long periods of relatively small events, dominated
by transport capacity limited erosion and deposition, are
disrupted by rare large events that cause flushing of deposits
and transport of sediment over large distances. These large
events are generally associated with gully and channel incision
leading to significant changes in elevation surface.
3.9.5
GIS Implementation
Implementation of the simple erosion models in GIS is relatively straightforward and involves creating a workflow that
includes input data processing, model computation, and analysis of results. The workflows can be developed using model
builders available in GIS and saved in the form of scripts that
support fully automated simulations. Complex physics-based
models usually require partial differential equation solvers
and can be fully integrated with GIS using the systems libraries for managing the geospatial data. More often, such
models are only loosely linked to GIS.
3.9.5.1
Coupling GIS and Models
Erosion modeling can be coupled with GIS through data exchange, graphical user interface (GUI), or it can be fully integrated as set of modules or scripts (Mitasova and Mitas, 2002).
Loose coupling links a model with GIS through import
and export of data. The model is developed independently
from a GIS and different GIS software packages can be used for
243
data preprocessing, analysis, and visualization of modeling

results. Standardization in georeferenced data formats and
improvements in software interoperability have made loose
coupling a routine procedure. Several landscape evolution and
erosion models are available with this type of GIS coupling
(CSMDS, 2011).
Tight coupling integrates a model and a GIS using shared
GUI which guides the user through input data processing,
modeling, and analysis. The interface also allows the user to
visualize the results using both the GIS display tools and
specialized graphical and numerical outputs. This type of integration has proven to be effective for hydrologic and nonpoint source pollution modeling systems (Rewerts and Engel,
1991; Renschler, 2003; Di Luzio et al., 2002) and several
models have been coupled with more than one GIS. Increasingly, the interface is provided through on-line tools via the
Web Processing Service (WPS).
Full integration involves spatial models which are useful
for a wide range of GIS applications. These models are developed and implemented within a GIS, using its programming tools such as Application Programming Interface (API),
scripting language, or map-algebra operations. The spatial
model is then run as a GIS function or command, with the
inputs and outputs stored in a GIS database (no data transfer
is needed). Portability of the model is restricted, and the enhancements as well as the maintenance of the model is
dependent on the GIS. This type of model development is
further supported by customization and application of development tools, extensions to map algebra (Wesseling et al.,
1996) and visual modeling tools.
Full GIS integration of complex models involving solutions
of coupled partial differential equations has been limited, in
spite of several successful implementations (Mitas and Mitasova, 1998). Even with all the necessary capabilities available
in the GIS, the biggest disadvantage of full integration of
complex models is that the models become too dependent on
the development and fate of a particular GIS. Changes in the
GIS data structures, functionality, interface, libraries or programming tools, may require time consuming changes in the
models or the models become incompatible with the latest
version of the GIS software. Also, the fully integrated model is
less portable and users have to install entire GIS even if they
need the model only for a one-time application. Some of these
issues have been addressed by open source software and
community systems development.
Large, professional modeling systems, most commonly
aimed at engineering applications, use both loose coupling
with a GIS and their own, specialized GIS capabilities. An
external GIS is generally used for storing, managing, and
processing of basic topographic data and for generating the
cartographic output. The modeling system itself includes
support for GIS functions where tight coupling with the model
is necessary, such as the design of a conceptual model for the
given site, adjustment of finite-element grids and meshes, as
well as modifications of the model parameters (conditions of
simulations) based on the simulation results.
With the explosive growth of the Internet, erosion models
useful for a wider range of users, such as farmers, land owners,
city planners, or public land managers are being implemented
as web-based applications. The successful applications include
244
not only the modeling tools, but also the databases with input
data and model parameters so that the user does not have to
deal with the time consuming tasks of finding, processing, and
submitting the input data for the model runs. Generally, only
selection of the location and land-use management scenario is
needed from the given set of options. Spatial data are stored in a
GIS on the server and the digital maps or animations, representing the inputs and model results, are served using the online map serving technology (Park et al., 2009; Luo et al., 2004).
3.9.5.2
Derived Model Parameters
Finding, importing, integrating, and processing the data necessary for erosion modeling can be a time-consuming task.
Therefore, a well maintained, consistent GIS database makes
data preparation for modeling and generating alternative
scenarios highly efficient. The following input data are generally required for GIS-based erosion modeling.
Elevation data are widely available as raster DEMs. Accurate
flow routing and slope estimation requires resolutions with
grid size of 10 m and smaller and vertical precision at least in
centimeters. If a raster DEM is not available, contours, or point
data, such as LiDAR point clouds, can be interpolated to create
the DEM (see Chapter 3.6). Elevation data are used to compute
slope, direction of flow, and upslope contributing area for
physics-based models as well as numerous other parameters
such as curvatures or wetness indices for statistical models (see
Chapter 3.7).
Land cover data are generally derived from aerial or satellite imagery as a raster map layer (see Chapter 3.4). If raster
data are not available, polygon areas can be transformed to
raster at desired resolution. Land cover data are used to estimate the land-cover C-factor for USLE and for surface roughness, detachment capacity and sediment transport capacity
coefficients in process-based models. These parameters have
empirically derived values and are commonly available in
literature and reference tables (Haan et al., 1994).
Soil data are generally available as polygons and have to be
transformed to raster or hydrologic units using standard GIS
data model transformations. If soil samples are available,
spatial interpolation using geostatistics methods is applied to
compute the raster representation of soil properties that are
the basis for computation of erosion related factors. Soil data
are used to derive the soil erodibility K-factor for USLE and
detachment capacity and sediment transport capacity coefficients in process-based models.
Rainfall data are available in the form of isoline maps for
R-factor or from databases used for RUSLE. For field and small
watershed scales, a single value is sufficient; for regional
modeling the rainfall factor should be given as raster map and
can be obtained using Tropical Rainfall Mapping Mission data
(see Chapter 3.1). Rainfall factors can also be derived from
measured rainfall data (Haan et al., 1994).
3.9.5.3
Analysis and Visualization
A GIS provides powerful tools for analysis and communication of modeling results, and it is commonly combined with
advanced statistical packages and specialized tools to perform
model calibration and validation. Erosion and sediment

transport modeling involves substantially more uncertainty
than modeling in disciplines where most experiments
are done in laboratory under controlled conditions. Model
calibration and validation are, therefore, an important component of the modeling effort and are generally handled by
external statistical tools loosely coupled with models and GIS
(Poeter et al., 2005). Calibration of erosion models and validation of the modeling results has been commonly limited by
difficulties of data collection. Recent advances in temporal
acquisition of high-resolution airborne or terrestrial LIDAR
data provide new opportunities to test the accuracy of predicted patterns of erosion and deposition based on DEM
differencing. This, in turn, can be used to evaluate the sensitivity of model parameters for obtaining accurate erosion
magnitudes and distributions.
In addition to summary statistics, maps representing spatial distribution of soil detachment, net erosion and deposition, and sediment transport rate are among the most
important outputs of erosion modeling that communicate the
spatial pattern of sediment sources, highlight high erosion risk
areas and locations with damaging rates of deposition. Perspective views of erosion modeling results draped over DEMs
are especially effective in highlighting the relationship between landforms and the spatial pattern of erosion and deposition (Figures 3 and 8). Although the perspective 3D views
of topography do not replace an accurate 2D map, they are
useful in interactive mode for visual analysis of complex
spatial relations. As static images, they can be used to improve
the perception of 3D features and their impact on erosion that
may be difficult to capture by 2D maps. Animations are
commonly used to represent the process dynamics and this
can include animated 2D images as well as dynamic surfaces
in perspective 3D views (Mitas et al., 1997).
Recent advances in humancomputer interaction technology provide environments for enhanced, collaborative exploration of landscape models. In addition to multitouch
screens and immersive 3D environments, interaction with 3D
objects is emerging as a powerful tool for investigation of
landscape properties and processes. The Tangible Geospatial
Modeling System (TanGeoMS) couples a 3D laboratory laser
scanner and a physical landscape model with a video projector
and GIS to support exploration of terrain change impacts on
topographic parameters and land surface processes (Mitasova
et al., 2006; Tateosian et al., 2010). The flexible clay landscape model can be modified by multiple users. When a
modified landscape is scanned, the selected terrain parameter
or process simulation is recomputed based on the scanned
data and the result is projected over the modified surface
(Figure 11). The projected image or animation provides the
users with feedback on impact of their terrain modifications
and guides the future exploration.
3.9.6
Case Studies
Two case studies were selected to illustrate the GIS-based

erosion modeling and highlight its focus on spatial patterns of
erosion and deposition.
90
245
20
0
(a)
(b)
(c)
Figure 11 Tangible Geospatial Modeling System: (a) flexible landscape model M is scanned by an overhead 3D laser scanner S while GIS data
(e.g., orthphoto) and simulation results are projected over the model using a projector P; (b) several users can modify the landscape model by
adding buildings or creating road tracks; (c) perspective views of initial and modified landscape models; slope maps draped over the DEMs
provide feedback on the modification impact on slope angle values.
Grass
Dirt path
Wetland
Crop field
Vineyard
Dirt road
Paved road
Bare soil
Buildings
Forest
Pond
North Carolina
USA
ntic
Atla
Location
of study site
an
Oce
100 m
(a)
Watershed boundary
(b)
(c)
Figure 12 Case study site: (a) location in the central North Carolina, (b) aerial photo draped over DEM with watershed boundaries and two
monitoring sites, and (c) simplified land cover used in simulations.
3.9.6.1
North Carolina Piedmont
A small watershed at the North Carolina State University

Sediment and Erosion Control Research and Education Facility
(SECREF) (McLaughlin et al., 2001) has been used for erosion
model testing and sediment transport monitoring (Figures 12
and 13). This site is used to demonstrate advantages and
limitations of various approaches to GIS-based erosion modeling. The 9.4 ha watershed has a mean slope of 3.41 and a
maximum flow-path length of 530 m. The soil type is mostly
Cecil, and land use includes agricultural fields, grass, small
forested area, vineyard, buildings, and experimental ponds.
Disturbances are limited to tilling at a 6 ha agricultural field.
The area was mapped by airborne LiDAR in the year 2001
with point density of 1 point for each 2 m resolution grid cell.
Bare-ground points were used to compute a 1 m resolution
DEM, a slope, and a direction of flow maps using regularized
smoothing spline with tension (RST) method (Mitasova et al.,
2005a). Several approaches were tested for mapping of flow
accumulation and estimation of unit water flow, including
single direction flow routing with 8 directions (SDR-D8 implemented in the GRASS module r.watershed), single direction
flow routing with infinite number of directions (SDR-Dinf,
module r.flow), and multiple direction flow routing with 8
directions (MDR-D8 module r.watershed and r.terraflow).
Additional parameters for the erosion models were derived
from land-cover, soil and rainfall data stored in a GIS and in
the related WEPP databases.
Baseline erosion modeling was performed for simplified
land cover (Figure 12(c)) to obtain estimates of soil erosion
and deposition using the following models:
on-line
version of GeoWEPP to estimate runoff and sediment yield;

GIS based version of detachment limited erosion model to
estimate soil detachment;
GIS based erosion/deposition model to estimate net erosion
and deposition rates;
overland water flow and sediment transport model SIMWE
to estimate runoff and sediment flow rates;
246
>0.1 m
0.01
0.001
1.0
0.5
N
0.01 kg m1s1
300 m
(a)
(b)
Figure 13 Study site after high intensity storms: (a) runoff with high concentration of sediment, (b) flooding in depressions and over a service
road. GIS-based simulation of sediment flow rate (a) and overland flow depth (b) using a LiDAR-based DEM and the SIMWE model implemented
in GRASS GIS reflects the observed water and sediment flow patterns. Photo courtesy R. McLaughlin, North Carolina State University.
landscape evolution model r.landevol to simulate change in

elevation.
The GeoWEPP on-line tool was the easiest to use because
the GIS data were provided along with the model interface.
The results were limited by resolution of the available data
(in this case 30 m) and only annual average soil loss and
sediment yield were computed (Figure 14). Using the default
parameters for central North Carolina, total annual runoff
from the watershed was estimated as 2372 m3 from hillslopes,
and sediment yield as 4.9 ton yr1 with erosion rate of 0.5 t
(ha yr)1. Deposition was predicted in less than 10% of the
area, most likely due to the low resolution of the DEM that did
not adequately represent shallow concave landforms.
Annual soil loss and net erosion/deposition was also estimated using the detachment limited and transport capacity
limited models combined with the USLE rainfall, soil, and
land-cover parameters (R 220 and soil erodibility factors
ranging between K 0.150.28 were used). Both models were
run in GRASS GIS using map algebra applied to 1 m resolution raster data. The results depicted the spatial pattern of
soil detachment (Figure 15(a)) and net erosion and deposition (Figure 15(b)). The results show significant potential
for large rills to develop in the agricultural fields and deposition to occur in the lower section of the valley. Subtle
terraces with alternating erosion and deposition in the western, agricultural areas required closer inspection to ensure that
they are not artifacts of point cloud measurements or
interpolation. The fact that the patterns do not follow the
data sampling distribution indicates that the terraces are not
data-processing artifacts. The results from terrestrial LiDAR

surveys confirm that the subtle terracing is due to the configuration of agricultural fields and direction of tillage.
The detachment capacity limited model (eqn [21]) predicted average annual soil detachment rate of 3.95 t (ha yr)1
for exponent values m 1.1 and n 0.6. These estimates are in
line with the USLE-based Natural Resources Conservation
Service (NRCS) erosion rates published for North Carolina,
but they are much higher than the values predicted by the
GeoWEPP model and those actually measured at the observed
values at the watershed outlet. This is due to the fact that the
detachment capacity limited models represent soil loss without considering sediment deposition. The estimates from the
transport capacity limited model indicate that most of the
eroded soil can be deposited within the watershed (i.e., 6.3 ha
has a potential for net erosion and deposition will occur over
3.1 ha). The predicted mean deposition rates are higher than
mean erosion rates, compensating for the smaller area with
potential for deposition. The mean net erosion rate was estimated as 1.4 t (ha yr)1 and mean net deposition rate was
3.9 t (ha yr)1 when m 1 and n 1 were used. The net soil
loss (total erosion minus total deposition) was estimated at
0.48 t (yr1) assuming vegetation cover, indicating that relatively small amount of sediment will leave the watershed
under transport capacity limited conditions.
The SIMWE model was used for single-storm simulations
to estimate runoff and sediment transport. The model correctly predicted ponding in the depression created by the road
(Figure 13). It also provides more realistic pattern of erosion/
deposition along the boundaries between the tilled field and
247
Figure 14 Simulation of hillslope erosion and deposition at 30 m resolution using the Web interface of the GeoWEPP model.
t ha1 yr1
t ha1 yr1
>30
Depo-
>30
sition
15
Erosion
0
0
<30
(a)
100 m
(b)
Figure 15 Spatial distribution of (a) soil detachment rate estimated from eqn [21] that includes prediction of high erosion rates due to
concentrated water flow; (b) net erosion and deposition rates indicating formation of gullies and subtle terraces predicted from eqn [23]. Higher
rates are predicted in areas with crops, dirt roads, and bare soil (see Figure 12).
vegetated strip, by allowing gradual deposition over longer

flow path, rather than instant deposition, as is the case with
the steady-state models. However, the model requires more
development and research to support rainfall time-series
input, additional channel cross-section geometries, conversion

of outputs into commonly used units and measures, and
validation with spatially and temporally distributed field data
that are only now becoming available from terrestrial laser
248
scanning technology. It is also computationally much more

demanding than the steady-state models.
Although only small changes in elevation are observed
during the years with average precipitation, severe storms have
occasionally caused major erosion and sediment transport in
this area, altering the topography. The landscape evolution
model was used to simulate the change in elevation over time
due to erosion and deposition by overland flow and diffusive
processes between the rainfall events. It was applied to variable land-cover conditions (Figure 16) with 40% area in
agricultural fields. The results show potential for developing
large rills in the fields with subsequent deposition at the edge
of the fields and in concave terrain. Under bare ground conditions, the formation of a large gully is predicted along the
valley due to concentrated water flow. The simulated erosion
and deposition pattern is in general agreement with observations. As in the case of the SIMWE model, repeat terrestriallaser surveys will provide opportunity to calibrate the model
and evaluate its capability to predict the magnitudes and
patterns of erosion and deposition.
TanGeoMS was used to explore the impact of various
land-design alternatives on surface runoff and soil erosion
(Figure 17). GIS layers, such as orthophotography, footprints
of structures, and flow-accumulation maps were projected
over the 1:1200 scale flexible clay terrain model to guide the
model modifications. Then various sediment control alternatives were created by modifying the plasticine surface by
hand. The modified model was scanned, runoff and erosion
130 m
were computed in GRASS GIS and resulting animations of

water and sediment flow were projected back onto the model
to provide rapid feedback on the effects of the proposed
change and to guide the next modification (Figure 17).
3.9.6.2
Mediterranean Landscape Evolution
The goal of Mediterranean Landscape Dynamics (MedLanD)

project was to build a sophisticated modeling laboratory to
carry out virtual experiments on the long-term, recursive
interactions between society, land-use, and environmental
change. Because it aims to simulate high-resolution, realworld landscape dynamics and land-use practices, the
MedLanD Modeling Laboratory (MML) is a hybrid modeling
environment that tightly couples different modeling approaches (Mayer and Sarjoughian, 2008; Mayer, 2009; Mayer
et al., 2006). These include: (1) a GIS-based landscape evolution model that simulates hydrology, erosion/deposition,
and vegetation succession; (2) stochastic GIS-based models
and dynamic agent-based models (ABM) of farming households and their land-use practices; and (3) regression-based
paleoclimate and paleovegetation models (Figure 18). Details
of the MML are published elsewhere (Barton et al., 2010a, b;
Ullah, 2011; Ullah and Bergin, 2011) and only summarized
here with emphasis on the landscape evolution modeling
component.
In the MML, agropastoral land-use can be modeled stochastically in a GIS or as rule-based actions of independent
Deposition
120
Erosion
110
0.3 m
0
0.3
(a)
100 m
Deposition
Erosion
1m
0
1
(b)
Figure 16 Landscape evolution under different conditions: (a) during small rainfall events erosion rates are low and the valley is filling with
deposits, (b) during high intensity rainfall gullies form in locations with concentrated flow. Perspective views of resulting DEMs with elevation
color maps are on the left, erosion/deposition pattern is on the right.
130 m
>0.1 m
30
120
0.01
15
110
0.001
249
(a)
N
300 m
(b)
Figure 17 Exploring terrain modification impact on water flow and sediment transport: (a) DEM, water flow depth and topographic erosion
potential for the initial model, (b) adding buildings and bio-swales to the model, resulting water flow depth, and topographic erosion potential for
the modified model. Model design by B. Harmon (Tateosian et al. 2010).
Vegetation
model
(regression)
Agropastoral
socioecology
model
Communities
& land use
model (ABM)
Climate model
(regression)
Landscape
evolution
model
(GIS/CA)
Figure 18 Schematic of the component structure and model couplings for the MedLand Modeling Laboratory.
computer agents in an ABM platform. When modeled stochastically, farming and grazing patches are randomly distributed within catchments, calculated using GIS routines to
account for terrain and suitability for farming or herding,
around communities (Barton et al., 2010a; Ullah, 2011). Alternatively, individual house-holds can be simulated as virtual
agents, organized into villages. Agents choose land to farm
or graze on the basis of their need for farming returns (calories
that affect birth and death rate), the suitability of land
for particular agricultural activities, and costs to use the
land, including access on foot and clearance of vegetation
(Barton et al., 2010a; Mayer, 2009; Mayer et al., 2006; Ullah and
Bergin, 2011). Agents can also collect fuel wood, the amount of
which varies according to household size and activities.
Whether modeled stochastically or as agent behavior, household land-use can alter the vegetation cover and soil characteristics. These anthropogenic changes to landscapes in turn
impact the results of other surface processes on landscapes
location and intensity of erosion and deposition. These
are simulated in the MML with a landscape evolution model
implemented in GRASS GIS as a Python-based script
r.landscape.evol.py developed by the MedLanD project team.
250
The model iteratively calculates the amount of erosion

and deposition that occurs across the landscape over time.
Because each of the process equations described perform
better under different topographic conditions, the script uses
different process equations for different landforms, and implements them in a manner that optimizes the ratio of
model run-time to accuracy of erosion/deposition calculations. The module implements a diffusion equation for areas
near drainage divides, a three-dimensional transport capacity
limited method for hillslopes and gully heads (eqn [22]), and
an equation based on the reach-average shear stress for
channels (eqn [10]). It is important to choose the optimal
locations on the terrain for the transition between surface
process models to ensure that the most appropriate process
equation is used for each cell of the DEM. Although these
transition points can vary with overall watershed geometry,
area, and topographic relief, and also can change during a
hydrologic event (e.g., during a storm) as a function of
rainfall intensity and duration, they can be estimated in a GIS
on the basis of upslope contributing area and topographic
profile curvature.
The model employs the advanced flow accumulation and
basin analysis module r.watershed to calculate upslope contributing area using a multiple flow direction (MFD) algorithm that produces much smoother and more realistic
patterns of flow convergence and divergence than does a
simpler single flow direction (SFD) D-8 algorithm. Upslope
contributing area is then plotted against the profile curvature
for each cell (Figure 19). Positive values of profile curvature
indicate a marked increase in slope (a convex profile), negative
values indicate a marked decrease in slope (a concave profile),
whereas values close to zero represent cells where there is little
change in slope. Drainage divides have little accumulation and
little change of slope, plotting near 0 on the x and y axes;

hillslopes also have little change in slope, but have higher
accumulation, plotting near 0 on the x-axis but higher on the
y-axis. The transition between the drainage divide and hillslopes has the maximum convex profile curvature and relatively low values of upslope contributing area, whereas the
transition between hillslopes and gully heads at the base of
slopes has concave profile curvature and higher contributing
area values. Hence, for a given landscape and hydrologic regime, the authors use the value of contributing area for the
maximum value of profile curvature for the transition from
the diffusion equation to transport capacity limited erosion by
overland flow with exponents m, n for sheet wash processes.
The value of contributing area the minimum value of profile
curvature is used for the transition from sheet wash to transport capacity limited erosion with higher exponents m, n for
rill/gully flow (Figure 19(a)).
In a similar way, contributing area can be plotted against
the tangential curvature of each cell to identify the accumulation value for the transition from the transport capacity
limited erosion by overland flow to a shear-stress equation for
channelized flow. The beginning of channelized flow can be
identified as the location where very low negative values of
tangential curvature (concave) are associated with high values
of contributing area. Negative values of tangential curvature
that are associated with lower values of contributing area
represent the larger gullies and gully-heads (i.e., that occur
higher in the drainage network than the real stream channels),
and even higher values of contributing area where tangential
curvature has decreased indicate a widening channel carrying
more water (Figure 19(b)). Figure 20 shows the locations of
the transition points identified in Figure 19 on the DEM for
which they were derived.
Upslope accumulated area (A)
100 000
10 000
1000
Transition point 3
100
Transition point 2
Transition point 1
10
1
0.004
(a)
0.002
0.002
Profile curvature (pc)
0.004
0.004
(b)
0.002
0.002
0.004
Tangential curvature (tc)
Figure 19 Plot of upslope contributing area vs. profile curvature (pc). Transition point 1 indicates the value of contributing area where the
transition from diffusion to sheetwash occurs, and transition point 2 indicates the value of contributing area where transition from sheet wash to
rilling/gullying occurs; (b) a plot of contributing area versus tangential curvature (tc). Transition point 3 indicates the value of contributing area
where the transition from rilling/gullying to channel flow occurs.
251
Diffusion
Sheetwash
Rill/Gully
Channel
Figure 20 Map showing the location of the process transition points (1, 2, and 3) derived from Figure 19 draped over a DEM. Actual gullies
and channels will be located within the outlined respective areas.
The diffusion equation used by the MML is well-known

and simulates soil creep, the movement of soil downslope due
to the effect of gravity and particle movement from rainsplash,
bioturbation, and other local factors, on portions of the
landscape where there is not enough accumulated runoff for
overland flow (Tucker and Bradley, 2010):
qz
Kg r2 z;
qt
31
where the rate of elevation change is controlled by the diffusion coefficient Kg(m(1000 yr)1), for a soil density of 1,
which is an empirically derived constant for different climate
and vegetation regimes.
The transport capacity limited erosion eqns [22] and [23]
used in the MML hillslope model were derived from concepts
described by Kirkby (1971), adapted for complex topography
by Moore and Burch (1986), and operationalized in GRASS
GIS (Mitasova et al., 1996; Mitas and Mitasova, 1998). The
approach combines the USLE/RUSLE parameters, slope and
upslope contributing area per unit contour width to estimate
sediment flow at sediment transport capacity, and calculate net
erosion and deposition at the center of each grid cell. Implementation in a GRASS script combines the results of GIS
modules for calculating slope, aspect, and flow accumulation
using map algebra. Input data for the script include a raster
DEM of initial surface topography, soil erodibility (K-factor as
a constant for uniform soil or a raster map for variable soil),
vegetation cover (C-factor as a constant or raster map), and
rainfall intensity (R-factor as a constant only). An underlying
bedrock topography DEM is also created to provide a limit
the total depth of unconsolidated sediment that can be eroded; when bedrock is reached, soil erodibility is set to near
zero. K-factor, C-factor, and R-factor values for RUSLE have
been calculated empirically for a variety of settings in the
Mediterranean (Boellstorff and Benito, 2005; Essa, 2004;
Hammad et al., 2004; Martinez-Casasnovas and SanchezBosch, 2000; Renard et al., 1997; Renard and Freimund,
1994).
For flow in channels, the MML employs a variation of
eqn [10], where the shear stress is approximated using eqn
[12], but unit stream power eqns [14] and [15] are also
considered. In these equations, transport capacity in channels

is largely a function of the local slope and the depth of flow.
However, depth of flow is difficult to estimate accurately because it is dependent on the local channel geometry and
changes over the course of a rainfall event. Therefore, MML
estimates flow depth in each cell from an idealized unit
hydrograph for that cell using the following equation:
h 2rA=1:25t,
32
where r (m) is the excess rainfall (precipitation minus infiltration) during a hydrologic event, and A(m2) is the upslope
contributing area. Then rA(m3) is the volume of water that
passed over the cell during the simulation interval and t(s) is
the number of hydrologic instants in the simulation interval.
The hydrologic instant is the time it takes water to cross one
cell of a raster DEM, which can be determined by multiplying
the average velocity of flowing water in the watershed (e.g., as
derived with Mannings eqn [13]) by the cell resolution.
The MML then computes net erosion and deposition rates
as divergence in sediment transport capacity (eqn [22]). To
approximate the transition between the zones dominated by
different processes, sediment flow divergence is computed
independently for the zones in which each process is modeled, and the results are then combined into integrated spatial
representation of erosion and deposition rates.
For the overland flow driven processes, the net erosion and
deposition rates are calculated as units of mass per area per
year. To model terrain evolution over time, these rates must be
converted to the depth of lost or gained sediment per cell
Dz(x, y, t) (m) for a given time period Dt based on soil density
(eqn [29]). Soil density is approximated using the method
outlined by Rawls (1983) combining the percentages of sand,
silt, clay, and organic matter, and estimated for Terra Rossa
soils (i.e., for Mediterranean landscapes) on the basis of empirical studies by Onori et al. (2006). Similarly as in the gully
modeling, the estimated change in elevation is then added to
(for deposition) or subtracted from (for erosion) the initial
DEM0, to create a new DEM1 after a cycle of land-use and
terrain change. This process can be iterated repeatedly to
simulate decades to millennia of landscape evolution. In
addition to the terrain evolution component, other aspects of
252
the MML simulate vegetation regrowth and recovery of soil

fertility for fallowed patches of the landscape. Together, these
provide feedbacks to the land-use ABM, affecting subsequent
decisions by households about which land to farm and graze.
Initial results using the MML to study the dynamics of
ancient land-use and terrain evolution have been promising.
In one set of modeling experiments, the MML was parameterized with values derived from ancient farming communities of northern Jordan (Figure 21) and used for
experiments on the consequences of varying land-use (intensive vs. shifting cereal cultivation, and ovicaprine grazing vs.
no grazing) and community size on soil loss and vegetation
over the course of two and ten generations, 40 and 200 years
respectively (Barton et al., 2010a, b). Some of the modeling
results confirm widely held expectations about such land-use:
shifting cultivation can cause greater soil loss than intensive,
repeated cultivation of a few plots (e.g., with manuring); extensive ovicaprine grazing will cause more erosion than
farming without associated grazing; larger communities with

more people farming and grazing more land will have a
greater impact on the landscape than smaller communities
(Figure 22). Other results were less intuitively obvious because of the complex interactions between land-use and
landscape dynamics. Notably, when community size is below
a threshold (whose value is determined by local environmental conditions) the amount of soil loss can be substantially offset by soil accumulation (i.e., by the redeposition
of sediments eroded from other parts of a catchment), so that
the economic impacts of mixed agropastoral land-use can be
negligible or even beneficial. However, if communities pass
this threshold size, the consequences change qualitatively such
that soil loss greatly exceeds soil accumulation within a landuse catchment. This imbalance continues over the long-term,
with the potential for leaving a catchment unsuitable for
farming. One mitigating strategy is, not surprisingly, to reduce
community size through emigration or fissioning. Another
France
Spain
Turkey
Jordan
N
Egypt
1000 km
Figure 21 Location of Mediterranean landscapes used in this case study.
>3 m
0m
<3 m
5 km
Figure 22 Forty years of cumulative hillslope erosion/deposition around modeled prehistoric farming village (location marked by blue star) in
the Wadi Ziqlab drainage of northern Jordan. Dark line marks maximum extent of ovicaprine grazing catchment; grey line marks maximum extent
of zone of shifting cultivation around village.
253
4
3
2
1400
1200
1000
800
N
600
400 m
10 km
Figure 23 Map of the Rio Penaguila and Rio Serpis valleys showing the different locations and topographic settings of village for each set of
experiments: (1) in an alluvial plain (for easy access to land for farming and grazing), (2) in a canyon bottom (for seclusion), (3) at the base of a
cliff (for defensibility), and (4) on a topographic prominence (for maximum visibility).
less obvious solution discovered in these experiments is to

increase the area devoted to grazing relative to cultivation,
moving zones of soil loss into uncultivated uplands and
providing more sediment for redeposition in the areas around
farmed fields. Conservation measures, like terracing, could
also be instituted but would require some degree of social
reorganization to ensure the availability of sufficient labor for
terrace construction and long-term maintenance. This kind of
investment in landesque capital and intensification of land
use has often been accompanied by the growth of inequalities
in social power and prestige. Comparison with the known
archaeological record of this region indicates good agreement
between modeling results and both the long-term impacts of
Neolithic farming and the social responses to these impacts
(Kuijt and Goring-Morris, 2002; Legge and Harris, 1996;
Martin, 1999; Quintero et al., 2004; Rollefson and KohlerRollefson, 1992; Rosen et al., 2008; Simmons, 2007; Twiss,
2007).
A second set of experiments involved studying the results
of situating a small farming village in different topographic
contexts in the Rio Penaguila and upper Rio Serpis Valleys of
eastern Spain (Figure 21), the location of one of the earliest
known farming communities in the Iberian Peninsula
(Bernabeu et al., 2003; Bernabeu and Kohler, 2005). In
four different experimental runs (Figure 23): a simulated village, populated by household agents, was set alternatively in
an alluvial plain (for easy access to land for farming and
grazing); in a canyon bottom (for seclusion); at the base of a
cliff (for defensibility); and on a topographic prominence (for
maximum visibility).
In each locale, all other initializing parameters besides
geographic setting were kept the same for the village. The
agents farmed and grazed the land around each site for
100 years and resulting data were collected on population
size, economy, vegetation cover, and erosion/deposition
(Figure 24). When the village was located in the alluvial plain
it was more successful initially than when placed in the other

settings, as measured in terms of population growth and
agricultural returns. However, this success also led to a recursive, self-amplifying growth cycle of increasing population,
expanding cultivation and grazing, soil degradation and loss,
and even more expansion of cultivation and grazing. When
situated in the other locales, the same village grew more
slowly and experienced more variable economic returns.
However, the smaller and more stable population also had
much less detrimental impact on the surrounding landscape
(Ullah and Bergin, 2011).
3.9.7
Conclusion and Future Directions
This chapter has presented the underlying theory and methods

for GIS-based modeling of soil erosion, sediment transport,
and deposition by surface water flow. Although the dynamic
models of sediment transport are complex, the topographic
controls for most approaches can be approximated using a
power function of upslope contributing area and slope. These
topographic parameters can be readily extracted from DEMs
using GIS tools and they constitute the foundation of GISbased erosion modeling.
Multitemporal elevation data are becoming available at
ever increasing accuracies and at levels of detail that can capture submeter features over large areas, thus providing significant opportunities for improvements in modeling and
mapping of soil erosion and deposition. In spite of advances
in sensor technologies, spatial representations of soil properties, land cover, and rainfall excess remain weak components
of erosion modeling. Uncertainty in these critical input parameters impacts the accuracy of process-based erosion models
and the practical solutions require combination of simplified
empirical and physics-based models. These models can be
calibrated to provide acceptable quantitative summary
254
10 km
Bare
Shrubs
Open woodland
Woodland
Figure 24 Landcover maps for each modeling experiment with village locations indicated in Figure 23.
estimates, such as total volumes or spatially averaged erosion

rates, as well as depict the spatial patterns of erosion and deposition. Nevertheless, the accuracy at any point in space and
time is often limited. The GIS implementation extends the
value and usefulness of these approaches by simulating spatial
patterns of erosion and deposition under different existing
and future conditions. Such information is critical for erosion
and sediment control and invaluable for development of
sustainable land-use management practices.
The case studies presented here highlight the use of
GIS-based erosion modeling to study the interactions and
feedbacks between human activities and erosion processes
at different spatial and temporal scales. The first study illustrated the differences in outputs from models at annual
and event-based temporal scales. It also introduced an interactive environment - TanGeoMS - for exploration of terrain
change impacts on landscape processes with applications for
runoff and sediment control design. The second study represents a unique application of erosion and landscape evolution
modeling to investigation of relationship between geomorphology and evolution of prehistoric agricultural communities. The simulations highlighted the complexity of
feedback mechanisms between the location and size of communities, landforms and their evolution. These examples
demonstrate the role of GIS-based erosion modeling in practical problem-solving and in research involving anthropogenic
forcing.
Further improvements in the quantitative prediction of

sediment transport in space and time may require probabilistic rather than deterministic modeling, as well as a wider use
of data assimilation techniques and model ensembles. Spatial
analysis and modeling, coupled with advances in laser scanning and hyperspectral technologies provide unique opportunities to address the gaps in our understanding of erosion
processes and fundamentally improve both the theory, models, and their GIS implementation.
Acknowledgments
Research results presented in this paper were supported by the
US Army Research Office, grant W911NF-04-D-0003-0011,
National Science Foundation, grant BCS-410269 North Carolina Water Resources Research Institute. Assistance with data
and modeling at the SECREF location by Robert Austin and
Richard McLaughlin is greatly appreciated.
References
Arabi, M., Govindaraju, R., Hantush, M., Engel, B., 2006. Role of watershed
subdivision on modeling the effectiveness of best management practices with
SWAT1. JAWRA Journal of the American Water Resources Association 42(2),
513528.
Barton, C., Ullah, I., Mitasova, H., 2010a. Computational modeling and neolithic
socioecological dynamics: a case study from southwest Asia. American Antiquity
75(2), 364386.
Barton, C.M., Ullah, I.I., Bergin, S., 2010b. Land use, water and mediterranean
landscapes: modeling long-term dynamics of complex socio-ecological systems.
Philosophical Transactions of the Royal Society A: Mathematical, Physical and
Engineering Sciences 368(1931), 52755297.
Bernabeu, J., Kohler, T., 2005. Mas dIs (Penaguila, alicante): un recinto
monumental del VI milenio cal AC. In: Arias, P., Ontanon, R., Garcia- Monco, C.
(Eds.), III Congreso sobre el Neolitico en la Peninsula Iberica (Santander, 2003).
Instituto Internacional de Investigaciones Prehistoricas de Cantabria, Santander,
pp. 485495.
Bernabeu, J., Kohler, T., Diez, A., Gomez, M., 2003. Mas dIs (Penaguila, alicante):
aldeas y recintos monumentales del neoltico antiguo en el valle del serpis.
Trabajos de Prehistoria 60(2), 3959.
Boellstorff, D., Benito, G., 2005. Impacts of set-aside policy on the risk of soil
erosion in central spain. Agriculture, Ecosystems & Environment 107(23),
231243.
Cebecauer, T., Hofierka, J., 2008. The consequences of land-cover changes on soil
erosion distribution in Slovakia. Geomorphology 98(34), 187198.
CSMDS, 2011. Community surface dynamics modeling system.
/http://csdms.colorado.edu/S.
De Roo, A., Jetten, V., Wesseling, C., Ritsema, C., 1996. LISEM: a physically based
hydrological and soil erosion model incorporated in a GIS. IAHS Publication
235, 395403.
DeRose, R., Gomez, B., Marden, M., Trustrum, N., 1998. Gully erosion in Mangatu
Forest, New Zealand, estimated from digital elevation models. Earth Surface
Processes and Landforms 23(11), 10451053.
Desmet, P., Govers, G., 1995. GIS-based simulation of erosion and deposition
patterns in an agricultural landscape: a comparison of model results with soil
map information. Catena 25(14), 389401.
Desmet, P., Govers, G., 1996. A GIS procedure for automatically calculating the
USLE LS factor on topographically complex landscape units. Journal of Soil and
Water Conservation 51(5), 427433.
Di Luzio, M., Srinivasan, R., Arnold, J., 2002. Integration of watershed tools and
SWAT model into BASINS. Journal of the American Water Resources
Association 38(4), 11271142.
Dingman, S., 2002. Physical Hydrology. Prentice Hall, Upper Saddle River, NJ. vol.
646.
Essa, S., 2004. GIS modeling of land degradation in Northern-Jordan using landsat
imagery. In: Proceedings of XXth ISPRS Congress. Istanbul.
Finlayson, D., Montgomery, D., 2003. Modeling large-scale fluvial erosion in
geographic information systems. Geomorphology 53(12), 147164.
Flanagan, D., Gilley, J., Franti, T., 2007. Water erosion prediction project(WEPP):
development history, model capabilities, and future enhancements. Transactions
of the ASAE 50(5), 16031612.
Flanagan, D., Nearing, M., 1995. USDA-water erosion prediction project: technical
documentation. NSERL Rep 10.
Foster, G., Meyer, L., 1972. A closed form soil erosion equation for upland areas.
In: Shen (Ed.), Sedimentation: Symposium to Honor Prof. H.A. Einstein.
pp. 12.112.19.
Foster, G.R., 1982. Modeling the erosion processes. In: Haan, C.T., Johnson, H.D.,
Brakensiek, D.L. (Eds.), Hydrologic modeling of small watersheds. ASAE
Monograph No. 5, ASAE, St. Joseph, MI, pp. 197380.
Gaffer, R., Flanagan, D., Denight, M., Engel, B., 2008. Geographical information
system erosion assessment at a military training site. Journal of Soil and Water
Conservation 63(1), 1.
Govers, G., Gimenez, R., Van Oost, K., 2007. Rill erosion: exploring the relationship
between experiments, modeling and field observations. Earth- Science Reviews
84(3-4), 87102.
Govindaraju, R., Kavvas, M., 1991. Modeling the erosion process over steep
slopes: approximate analytical solutions. Journal of Hydrology 127(14),
279305.
Haan, C., Barfield, B., Hayes, J., 1994. Design Hydrology and Sedimentology for
Small Catchments. Academic Press, New York.
Hammad, A.A., Lundekvam, H., Barresen, T., 2004. Adaptation of RUSLE in the
eastern part of the mediterranean region. Environmental Management 34(6),
829841. http://dx.doi.org/10.1007/s00267-003-0296-7.
Hancock, G., Willgoose, G., Evans, K., 2002. Testing of the SIBERIA land-scape
evolution model using the Tin Camp Creek, Northern Territory, Australia, field
catchment. Earth Surface Processes and Landforms 27(2), 125143.
Harmon, R., Doe, W., 2001. Landscape Erosion and Evolution Modeling. Kluwer/
Plenum Publishers, New York.
255
Hong, S., Mostaghimi, S., 1997. Comparison of 1-D and 2-D modeling of overland
runoff and sediment transport. JAWRA Journal of the American Water Resources
Association 33(5), 11031116.
Jain, M., Das, D., 2010. Estimation of sediment yield and areas of soil erosion and
deposition for watershed prioritization using GIS and remote sensing. Water
Resources Management 24(10), 20912112.
Jetten, V., Govers, G., Hessel, R., 2003. Erosion models: quality of spatial
predictions. Hydrological Processes 17(5), 887900.
Julien, P., Simons, D., 1985. Sediment transport capacity of overland flow.
Transactions of the ASAE-American Society of Agricultural Engineers (USA)
28(3), 755762.
Karssenberg, D., Burrough, P., Sluiter, R., De Jong, K., 2001. The PCRaster
software and course materials for teaching numerical modeling in the
environmental sciences. Transactions in GIS 5(2), 99110.
Kheir, R., Wilson, J., Deng, Y., 2007. Use of terrain variables for mapping gully
erosion susceptibility in Lebanon. Earth Surface Processes and Land-Forms
32(12), 17701782.
Kim, J., Park, Y., Yoo, D., et al., 2009. Development of a SWAT patch for better
estimation of sediment yield in steep sloping watersheds1. JAWRA Journal of
the American Water Resources Association 45(4), 963972.
Kirkby, M.J., 1971. Hillslope process-response models based on the continuity
equation. Institute of British Geographers Special Pub 3, 1530.
Koco, S. 2009. The simulation of gully erosion by geographic information system.
Ph.D. thesis, University of Presov, Department of Geography and Regional
Development.
Kuijt, I., Goring-Morris, N., 2002. Foraging, farming, and social complexity in the
Pre-Pottery neolithic of the southern levant: a review and synthesis. Journal of
World Prehistory 16(4), 361440.
Legge, T., Harris, D., 1996. The beginnings of caprine domestication in south-west
asia. The Origins and Spread of Agriculture and Pastoralism in Eurasia. UCL
Press, London, pp. 238262.
Lim, K., Sagong, M., Engel, B., Tang, Z., Choi, J., Kim, K., 2005. GIS-based
sediment assessment tool. Catena 64(1), 6180.
Luo, W., 2001. LANDSAP: a coupled surface and subsurface cellular automata
model for landform simulation. Computers and Geosciences 27, 363367.
Luo, W., Duffin, K., Peronja, E., Stravers, J., Henryb, G., 2004. A web-based
interactive landform simulation. Computers & Geosciences 30, 215220.
Martin, L., 1999. Mammal remains from the eastern jordanian neolithic, and the
nature of caprine herding in the steppe. Palorient 25(2), 87104.
Martnez-Casasnovas, J., Ramos, M., Poesen, J., 2004. Assessment of sidewall
erosion in large gullies using multi-temporal DEMs and logistic regression
analysis. Geomorphology 58(14), 305321.
Martinez-Casasnovas, J.A., Sanchez-Bosch, I., 2000. Impact assessment of changes
in land use/conservation practices on soil erosion in the Peneds-Anoia vineyard
region (NE spain). Soil and Tillage Research 57(12), 101106.
Mayer, G.R., Jun. 2009. Composing hybrid discrete event system and cellular
automata models. Ph.D. dissertation, Arizona State University.
Mayer, G.R., Sarjoughian, H.S., 2008. A composable discrete-time cellular
automaton formalism. First International Workshop on Social Computing,
Behavioral Modeling, and Prediction. Springer, Phoenix, AZ, pp. 187196.
Mayer, G.R., Sarjoughian, H.S., Allen, E.K., Falconer, S.E., Barton, C.M.,
2006. Simulation modeling for human community and agricultural landuse.
In: Agent-Directed Simulation, Proceedings of the Agent-Directed Simulation
Multi-Conference, Huntsville, Alabama. Huntsville, AL, pp. 6572, April 24,
2006.
McLaughlin, R.A., Hunt, W.F., Rajbhandari, N., Ferrante, D.S., Sheffield, R.E., 2001.
The sediment and erosion control research and education facility at NC State
University. In: Ascough, II J.C., Flanagan, D.C. (Eds.), Soil Erosion Research for
the 21st Century. Am. Soc. Ag. Eng., St. Josepth, MI, 2001.
Mitas, L., Brown, W., Mitasova, H., 1997. Role of dynamic cartography in
simulations of landscape processes based on multi-variate fields. Computers
and Geoscience 23, 437446.
Mitas, L., Mitasova, H., 1998. Distributed soil erosion simulation for effective
erosion prevention. Water Resources Research 34(3), 505516.
Mitasova, H., Hofierka, J., Zlocha, M., Iverson, L., 1996. Modeling topographic
potential for erosion and deposition using GIS. International Journal of
Geographical Information Science 10(5), 629641.
Mitasova, H., Mitas, L., 2002. Modeling physical systems. In: Clarke, K.C., Parks,
B.O., Crane, M.P. (Eds.), Geographic Information Systems and Environmental
Modeling. Prentice Hall, pp. 189210.
Mitasova, H., Mitas, L., Brown, W., Johnston, D., 1999. Terrain modeling and soil
erosion simulations for Fort Hood and Fort Polk test areas. Research report,
University of Illinois, Urbana-Champaign, IL.
256
Mitasova, H., Mitas, L., Harmon, R., 2005a. Simultaneous spline interpolation and
topographic analysis for LiDAR elevation data: methods for open source GIS.
IEEE GRSL 2(4), 375379.
Mitasova, H., Mitas, L., Ratti, C., Ishii, H., Alonso, J., Harmon, R., 2006. Real-time
landscape model interaction using a tangible geospatial modeling environment.
IEEE Computer Graphics and Applications 26(4), 5563.
Mitasova, H., Thaxton, C., Hofierka, J., McLaughlin, R., Moore, A., Mitas, L.,
2005b. Path sampling method for modeling overland water flow, sediment
transport, and short term terrain evolution in Open Source GIS. In: Miller, C.,
Farthing, M., Gray, V., Pinder, G. (Eds.), Proceedings, XVth International
Conference on Computational Methods in Water Resources (CMWR XV).
Elsevier, pp. 14791490.
Mitchell, J., Engel, B., Srinivasan, R., Wang, S., 1993. Validation of AGNPS for
small watersheds using an integrated AGNPS/GIS system. JAWRA Journal of the
American Water Resources Association 29(5), 833842.
Moore, A., McLaughlin, R., Mitasova, H., Line, D., 2007. Calibrating wepp model
parameters for erosion prediction on construction sites. Transactions of the
ASABE 50(2), 507516.
Moore, I., Burch, G., 1986. Modeling erosion and deposition: topographic effects.
Transactions ASAE 29, 16241640.
Moore, I., Wilson, J., 1992. Length-slope factors for the revised universal soil loss
equation: simplified method of estimation. Journal of Soil and Water
Conservation 47, 423428.
Morgan, R., Quinton, J., Smith, R., et al., 1998. The European Soil Erosion Model
(EUROSEM): a dynamic approach for predicting sediment transport from fields
and small catchments. Earth Surface Processes and Landforms 23(6), 527544.
Murray, A., Paola, C., 1994. A cellular model of braided rivers. Nature 371, 5457.
Nearing, M., Norton, L., Bulgakov, D., Larionov, G., West, L., Dontsova, K., 1997.
Hydraulics and erosion in eroding rills. Water Resources Research 33(4),
865876.
Nearing, M., Simanton, J., Norton, L., Bulygin, S., Stone, J., 1999. Soil erosion by
surface water flow on a stony, semiarid hillslope. Earth Surface Processes and
Landforms 24(8), 677686.
Neteler, M., Mitasova, H., 2008. Open Source GIS: A GRASS GIS Approach.
Springer Verlag, New York.
Onori, F., De Bonis, P., Grauso, S., 2006. Soil erosion prediction at the basin scale
using the revised universal soil loss equation (RUSLE) in a catchment of Sicily
(southern Italy). Environmental Geology 50(8), 11291140.
Pandey, A., Chowdary, V., Mal, B., 2009a. Sediment yield modeling of an
agricultural watershed using MUSLE, remote sensing and GIS. Paddy and Water
Pandey, A., Mathur, A., Mishra, S., Mal, B., 2009b. Soil erosion modeling of a
Himalayan watershed using RS and GIS. Environmental Earth Sciences 59(2),
399410.
Park, Y., Kim, J., Engel, B., et al., 2009. Development of web-based SWAT System.
Proceedings of ASABE, Reno, Nevada, June 2009.
Peeters, I., Rommens, T., Verstraeten, G., Govers, G., Van Rompaey, A., Poesen, J.,
Van Oost, K., 2006. Reconstructing ancient topography through erosion
modeling. Geomorphology 78(34), 250264.
Pelletier, J., Mitasova, H., Harmon, R., Overton, M., 2009. The effects of interdune
vegetation changes on eolian dune field evolution: a numerical-modeling case
study at Jockeys Ridge, North Carolina, USA. Earth Surface Processes and
Landforms 34(9), 12451254.
Pike, A., Mueller, T., Schorgendorfer, A., Shearer, S., Karathanasis, A., 2009.
Erosion index derived from terrain attributes using logistic regression and neural
networks. Agronomy Journal 101(5), 10681079.
Poeter, E., Hill, M., Banta, E., Mehl, S., Christensen, S., 2005. UCODE 2005 and
six other computer codes for Universal Sensitivity Analysis, calibration, and
uncertainty evaluation, technical report, USGS.
Polyakov, V., Nearing, M., Shipitalo, M., 2004. Tracking sediment redistribution in a
small watershed: implications for agro-landscape evolution. Earth Surface
Quintero, L.A., Wilke, P.J., Rollefson, G.O., Bienert, H., Gebel, H.G., Neef, R., 2004.
Highland towns and desert settlements: origins of nomadic pastoralism in the
jordanian neolithic. Central Settlements in Neolithic Jordan. ex oriente, Berlin,
pp. 201213.
Rawls, W.J., 1983. Estimating soil bulk denisty from particle size analysis and
organic matter content. Soil Science 135(2), 123.
Renard, K., Foster, G., Yoder, D., McCool, D., 1994. RUSLE revisited: status,
questions, answers, and the future. Journal of Soil and Water Conservation
49(3), 213.
Renard, K.G., Foster, G.R., Weesies, G.A., McCool, D.K., Yoder, D.C., 1997.
Predicting soil erosion by water: a guide to conservation planning with the
revised universal soil loss equation (RUSLE). Agriculture Handbook. US
Department of Agriculture, Washington, DC, vol. 703, p. 1251.
Renard, K.G., Freimund, J.R., 1994. Using monthly precipitation data to estimate
the r-factor in the revised USLE. Journal of Hydrology 157(14), 287306.
Renschler, C., 2003. Designing geo-spatial interfaces to scale process models: the
GeoWEPP approach. Hydrological Processes 17(5), 10051017.
Rewerts, C., Engel, B., 1991. ANSWERS on GRASS: integrating a watershed
simulation with a GIS. ASAE paper, pp. 912621.
Rollefson, G.O., Kohler-Rollefson, I., 1992. Early neolithic exploitation patterns in
the levant: cultural impact on the environment. Population and Environment 13,
243254, NSF Biocomplexity 2002.
Rosen, S.A., Barnard, H., Wendrich, W., 2008. Desert pastoral nomadism in the
longue dure: a case study from the negev and the southern levantine deserts.
The Archaeology of Mobility: Old World and New World Nomadism. Cotsen
Institute of Archaeology. University Of California, Los Angeles, pp. 115140.
Sheikh, V., van Loon, E., Hessel, R., Jetten, V., 2010. Sensitivity of LISEM predicted
catchment discharge to initial soil moisture content of soil profile. Journal of
Hydrology 393(34), 174185.
Simmons, A.H., 2007. The Neolithic Revolution in the Near East: Transforming the
Human Landscape. University of Arizona Press, Tucson.
Suri, M., Cebecauer, T., Hofierka, J., Fulajtar, E, 2002. Erosion assessment of
Slovakia at a regional scale using GIS. Ecology (Bratislava) 21(4), 404422.
Tateosian, L., Mitasova, H., Harmon, B., Fogleman, B., Weaver, K., Harmon, R.,
2010. TanGeoMS: tangible geospatial modeling system. Visualization and
Computer Graphics, IEEE Transactions on 16(6), 16051612.
Tucker, G., Bradley, D., 2010. Trouble with diffusion: reassessing hillslope erosion
laws with a particle-based model. Journal of Geophysical Research 115,
F00A10.
Tucker, G., Lancaster, S., Gasparini, N., Bras, R., 2001a. The channel-hillslope
integrated landscape development model (CHILD). In: Harmon, R.S., Doe, W.W.
(Eds.), Landscape Erosion and Evolution Modeling. Kluwer/Plenum Publishers,
New York, pp. 349388.
Tucker, G., Lancaster, S., Gasparini, N., Bras, R., Rybarczyk, S., 2001b. An objectoriented framework for distributed hydrologic and geomorphic modeling using
triangulated irregular networks. Computers & Geosciences 27(8), 959973.
Twiss, K.C., 2007. The neolithic of the southern levant. Evolutionary Anthropology
1, 2435.
Ullah, I., 2011. A GIS method for assessing the zone of human-environmental
impact around archaeological sites: a test case from the late neolithic of wadi
ziqab, jordan. Journal of Archaeological Science 38(3), 623632.
Ullah, I.I., Bergin, S., 2011. Modeling the consequences of village site location:
least cost path modeling in a coupled GIS and Agent-Based model of village
agropastoralism in eastern spain. In: White, D.A., Surface-Evans, S.L. (Eds.),
Least Cost Analysis of Social Landscapes: Archaeological Case Studies for
Beginners and Experts Alike. University of Utah Press, Salt Lake City.
Van Oost, K., Govers, G., Cerdan, O., et al., 2005. Spatially distributed data for
erosion model calibration and validation: The Ganspoel and Kinderveld datasets.
Catena 61(23), 105121.
Wainwright, J., Parsons, A., Muller, E., Brazier, R., Powell, D., Fenti, B., 2008. A
transport-distance approach to scaling erosion rates: 3. Evaluating scaling
characteristics of Mahleran. Earth Surface Processes and Land-forms 33(7),
11131128.
Warren, S., Mitasova, H., Hohmann, M., Landsberger, S., Iskander, F., Ruzy-cki, T.,
Senseman, G., 2005. Validation of a 3-D enhancement of the Universal Soil
Loss Equation for prediction of soil erosion and sediment deposition. Catena
64(23), 281296.
Wesseling, C.G., Karssenberg, D., Van Deursen, W.P.A., Burrough, P.A., 1996.
Integrating dynamic environmental models in GIS: the development of a dynamic
modelling language. Transactions in GIS 1, 4048.
Willgoose, G., 2004. Mathematical modeling of whole landscape evolution. Annual
Review of Earth and Planetary Sciences 33(1), 443.
Willgoose, G., Bras, R., Rodriguez-Iturbe, I., 1991. A coupled channel network
growth and hillslope evolution model: 2. Nondimensionalization and
applications. Water Resources Research 27(7), 16851696.
Wischmeier, W., Smith, D., 1978. Predicting Rainfall Erosion Losses: A Guide to
Conservation Planning United States. Dept. of Agriculture. Agriculture Handbook,
Washington, DC.
Woodward, D., 1999. Method to predict cropland ephemeral gully erosion. Catena
37(34), 393399.
257
Biographical Sketch
Helena Mitasova is an Associate Professor at the Department of Marine, Earth, and Atmospheric Sciences, North
Carolina State University in Raleigh, NC, USA. She co-authored the first book on open source GRASS GIS and
published more than 50 papers on methods and applications of topographic` analysis, erosion modeling, coastal
dynamics, and visualization. Her PhD is from the Slovak Technical University, Bratislava, Slovakia.
C. Michael Barton is a Director of the Center for Social Dynamics and Complexity and a Professor of Anthropology in the School of Human Evolution and Social Change at Arizona State University. Bartons research
interests center around long-term human ecology and the dynamics of Quaternary landscapes. He heads an
international, multidisciplinary team studying the long-term socioecology of the beginnings of agriculture. He
also is active in the development and application of spatial technologies in archaeology and serving on the open
source GRASS GIS development team. He holds a PhD from the University of Arizona in anthropology and
geosciences.
Isaac Ullah is a PhD candidate in the School of Human Evolution and Social Change at Arizona State University,
Tempe, USA. He is a member of the Mediterranean Land-scape Dynamics project, and specializes in geological
applications in archaeology using GIS and modeling techniques. He studies the long term effects of human land
use decisions and how potential anthropogenic environmental degradation may have affected settlement patterns.
He holds an MA in Anthropology from the University of Toronto.
Jaroslav Hofierka is an Associate Professor at the Institute of Geography, Faculty of Sciences, Pavol Jozef Safarik
University, Kosice, Slovakia. His research activities include digital terrain modeling, geomorphometry, erosion risk
analysis as well as solar radiation and 3D city modeling. He has been active in open source GIS development and
has contributed several core modules to GRASS GIS. He received a PhD degree in Cartography and GIScience from
the Comenius University, Bratislava, Slovakia.
258
Russell S. Harmon is an Adjunct Associate Professor at North Carolina State University and he has been a Program
Manager at the Army Research Office, where he is responsible for the ARO extramural basic research program in
terrestrial sciences. He is a Fellow of the Army Research Laboratory, Geological Society of America, and National
Speleological Society and is current Past-President of the International Association of GeoChemistry. Dr. Harmon
is actively engaged in research in landscape dynamics and geospatial analysis, riverine geo-chemistry, and laserinduced breakdown spectroscopy. He holds a MS from the Pennsylvania State University, and PhD from the McMaster University.
3.10 Remote Sensing and GIS for Natural Hazards Assessment and Disaster
Risk Management
CJ Van Westen, Faculty ITC, University of Twente, The Netherlands
3.10.1
Introduction
3.10.2
Background
3.10.2.1
Trends in Disaster Statistics
3.10.2.2
Disaster-Risk Management Framework
3.10.2.3
Risk-Analysis Framework
3.10.3
Hazard Assessment
3.10.3.1
Scale and Hazard Assessment
3.10.3.2
Spatial Data for Hazard Assessment
3.10.3.2.1
Hazard inventories and triggering events
3.10.3.2.2
Environmental factors
3.10.3.3
Hazard Assessment Examples
3.10.3.3.1
Global hazard assessment
3.10.3.3.2
(Inter)National hazard assessment
3.10.3.3.3
Provincial and municipal level
3.10.3.3.4
Community level
3.10.4
Elements-At-Risk and Vulnerability
3.10.4.1
Elements-At-Risk Information
3.10.4.1.1
Collaborative mapping and Mobile-GIS
3.10.4.1.2
Population data
3.10.4.1.3
Building data
3.10.4.2
Vulnerability
3.10.5
Multi-Hazard Risk Assessment
3.10.5.1
Qualitative Approaches
3.10.5.2
Quantitative Approaches
3.10.5.3
Spatial Risk Visualization
3.10.6
Conclusions
Acknowledgements
References
Glossary
AAL Average annual loss, or expected annual loss
(EAL), is the long-term loss rate that can be obtained
by summing the product of each discrete loss state with
its annual frequency of occurrence over all loss
states.
ADPC The Asian Disaster Preparedness Center is a
non-profit organization from Bangkok, supporting the
advancement of safer communities and sustainable
development through implementing programs and projects
that reduce the impact of disasters on countries and
communities in Asia and the Pacific.
ADRC The Asian Disaster Reduction Center in Kobe,
Japan has a mission to enhance disaster resilience of the
264
266
266
268
270
271
272
273
273
276
278
278
279
279
280
281
281
281
281
283
283
286
287
288
289
290
293
293
member countries, to build safe communities, and to create

a society where sustainable development is possible.
AEP Annual exceedance probability is the estimated
probability that an event of specified magnitude will be
exceeded in any year.
ALOS Advanced Land Observing Satellite is a land
observing satellite launched in January 2006 by Japan
Aerospace Exploration Agency (JAXA).
APRSAF Asia-Pacific Regional Space Agency Forum was
established in 1993 to enhance space activities in the
Asia-Pacific region.
ArcGIS GIS software by ESRI.
ArcPad GIS software developed by ESRI for use on
Mobile-GIS devices.
Van Westen, C.J., 2013. Remote sensing and GIS for natural hazards
assessment and disaster risk management. In: Shroder, J. (Editor in Chief),
http://dx.doi.org/10.1016/B978-0-12-374739-6.00051-8
259
260
Remote Sensing and GIS for Natural Hazards Assessment and Disaster Risk Management
ASTER Advanced Spaceborne Thermal Emission and

Reflection Radiometer is a Japanese sensor that is one of
five remote-sensing devices on board of the Terra satellite launched by NASA in 1999. ASTER provides
high-resolution images of the Earth in 15 different bands of
the electromagnetic spectrum, ranging from visible to
thermal infrared light.
AVHRR Advanced Very High Resolution Radiometer is a
space-borne sensor on the National Oceanic and
Atmospheric Administration (NOAA) family of polar
orbiting platforms.
Capacity The positive managerial capabilities of
individuals, households, and communities to confront the
threat of disasters (e.g., through awareness raising, early
warning, and preparedness planning).
CAPRA Central American Probabilistic Risk Assessment,
initiative developed by the Worldbank.
Cartosat A stereoscopic Earth observation satellite in a
sun-synchronous orbit. The satellite was built, launched,
and maintained by the Indian Space Research Organisation
(ISRO).
CLC CORINE Land Cover. See CORINE
Consequence The expected losses in a given area as a
result of a given hazard scenario.
CORINE The Corine Land Cover project is a joint
initiative of the EU Commission and EU Environment
Agency, which developed an available database now
including ortho-rectified Landsat 7 ETM satellite images of
the European territory, Land Cover, and Change
definition maps.
CRED The Centre for Research on the Epidemiology of
Disasters, of the Universite catholique de Louvain, Belgium,
specializes in the fields of international disaster and conflict
health studies, with research and training activities linking
relief, rehabilitation, and development, and maintains the
EM-DAT database.
CVA The Capacity and Vulnerability Assessment is used as
a diagnostic tool to understand problems and their
underlying causes, related to the vulnerability and capacity
of local communities to natural hazards.
DEM A digital elevation model is a digital model or 3-D
representation of a terrains surface created from terrain
elevation data.
DGPS A differential global positioning system is an
enhancement to the global positioning system that
uses a network of fixed, ground-based reference stations
to broadcast the difference between the positions
indicated by the satellite systems and the known fixed
positions.
Disaster A serious disruption of the functioning of a
community or a society causing widespread human,
material, economic, or environmental losses that exceed the
ability of the affected community or society to cope using its
own resources.
DNCA Damage, needs, and capacity assessment involves
a participatory analysis of the disaster event focussing on
the damages caused, the immediate needs and priorities of
the affected community, and of the remaining capacities
people use to cope with the adverse effects.
DRI Disaster Risk Index produced a model of factors

influencing levels of human losses from natural hazards at
the global scale, by the United Nations Development
Programme.
DRM Disaster-risk management is the systematic process
of using administrative directives, organizations, and
operational skills and capacities to implement strategies,
policies, and improved coping capacities in order to lessen
the adverse impacts of hazards and the possibility of
disaster.
DRR Disaster-risk reduction is the concept and practice of
reducing disaster risks through systematic efforts to analyse
and manage the causal factors of disasters, including
through reduced exposure to hazards, lessened vulnerability
of people and property, wise management of land and the
environment, and improved preparedness for adverse
events.
DSM A digital surface model is a digital model or 3-D
representation of the Earths surface and includes all
objects on it.
DTM Digital terrain model is a digital model or 3-D
representation of the bare ground surface without any
objects like plants and buildings.
EDM Electronic distance measurement is an electronic
theodolite integrated with an electronic distance meter to
read slope distances from the instrument to a
particular point.
EFD European Flood Directive requires EU Member States
to engage their government departments, agencies, and
other bodies to draw up a Preliminary Flood Risk
Assessment.
EFFIS The European Forest Fire Information System
supports the services in charge of the protection of forests
against fires in the EU countries and provides the European
Commission services and the European Parliament with
updated and reliable information on wildland fires in
Europe.
EIA An environmental impact assessment is an assessment
of the possible positive or negative impacts that a proposed
project may have on the environment, together consisting
of the natural, social, and economic aspects.
Elements-at-risk Population, properties, economic
activities, including public services, or any other defined
values exposed to hazards in a given area. Also referred to as
assets. The amount of elements-at-risk can be quantified
either in numbers (of buildings, people, etc.), in monetary
value (replacement costs, market costs, etc.), area, or
perception (importance of elements-at-risk).
EM-DAT Emergency Events Database, the international
database on natural hazards, maintained by CRED.
ENVISAT Polar orbiting spacecraft operated by the
European Space Agency possessing several instruments
used to monitor the Earths environment. Parameters
measured include ozone concentration, aerosols,
surface stress for earthquake potential, sea-level heights,
and fires.
EO Earth observation is the gathering of information
about planet Earths physical, chemical, and biological
systems, using remote sensing. It is used to monitor and
assess the status of, and changes in, the natural

environment and the built environment.
ERS European radar satellite, developed and maintained
by the European Space Agency (ESA).
ESRI The worlds largest GIS company that develops
geographic information systems (GIS) solutions that
function as an integral component in nearly every type of
organization.
ETOPO Topographic dataset from the NOAA NGDC,
of topography and bathymetry for the entire Earths
surface.
EWS Early Warning System is the set of capacities needed
to generate and disseminate timely and meaningful warning
information to enable individuals, communities, and
organizations threatened by a hazard to prepare and to act
appropriately and in sufficient time to reduce the possibility
of harm or loss.
Exposure Exposure indicates the degree to which
the elements-at-risk are exposed to a particular hazard.
The spatial interaction between the elements-at-risk and the
hazard footprints are depicted in a GIS by simple map
overlaying of the hazard map with the elementsat-risk map.
FEMA Federal Emergency Management Agency, the
national organization for disaster-risk management in
the USA.
Formosat Earth observation satellite operated by the
National Space Organization (NSPO) of the Republic of
Taiwan.
Frequency A measure of likelihood expressed as the
number of occurrences of an event in a given time.
GDACS Global Disaster Alert and Coordination System,
by Joint Research Center (JRC) of the European
Commission.
GeoEye A commercial satellite imagery company based in
Dulles, Virginia, USA.
GEM The Global Earthquake Model is a public-private
partnership initiated in 2006 by the Global Science Forum
of the OECD to develop global, open-source risk
assessment software and tools.
Geological hazard A geological process or phenomenon
that may cause loss of life, injury or other health impacts,
property damage, loss of livelihoods and services, social and
economic disruption, or environmental damage.
GIS A geographic information system, or geospatial
information system is a system designed to capture, store,
manipulate, analyze, manage, and present all types of
geographically referenced data.
GMES Global Monitoring for Environment and Security
initiative of the European Commission and the European
Space Agency (ESA).
G-MOSAIC GMES services for Management of
Operations, Situation Awareness, and Intelligence for
regional Crises.
GMOSS Global Monitoring for Security and Stability is a
project in the aeronautics and space priority of the EU 6th to
integrate Europes civil security research and to develop and
maintain an effective capacity for global monitoring using
satellite earth observation.
261
GPS Global positioning systems is a space-based global

navigation satellite system (GNSS) that provides location
and time information in all weather anywhere on or near
the Earth, where there is an unobstructed line of sight to
four or more GPS satellites.
GRUMP Global Rural-Urban Mapping Project is a project
for generation of a gridded model with a grid cell resolution
of 30 arc-seconds for global population distribution
developed by the Socioeconomic Data and Applications
Center (SEDAC).
GSHAP The Global Seismic Hazard Assessment Project
was launched in 1992 by the International Lithosphere
Program (ILP) with the support of the International
Council of Scientific Unions (ICSU), and produced smallscale standardized seismic hazard maps for all seismic
regions of the world.
Hazard A dangerous phenomenon, substance, human
activity, or condition that may cause loss of life, injury or
other health impacts, property damage, loss of livelihoods
and services, social and economic disruption, or
environmental damage.
Hazard zoning The subdivision of the terrain in zones
that are characterized by the temporal probability of
occurrence of hazardous events of a particular size and
volume within a given period of time.
HAZUS Hazards US is the loss estimation software
developed by the Federal Emergency Management Agency
(FEMA) together with the National Institute of Building
Sciences (NIBS).
HVCA Hazards, vulnerability, and capacity assessment is a
method that involves a participatory analysis of historical
hazard trends and present threats undertaken at the level of
the community (hazard assessment). It is combined with an
understanding of the underlying reasons why hazards
become disasters (vulnerability assessment) and of the
available resources an affected community uses to cope
(capacity assessment).
Hydrometeorological hazard Process or phenomenon of
atmospheric, hydrological, or oceanographic nature that
may cause loss of life, injury or other health impacts,
property damage, loss of livelihoods and services, social and
economic disruption, or environmental damage.
ICG International Centre for Geohazards, Norway.
IFRC International Federation of Red Cross and Red
Crescent Societies.
IKONOS A commercial Earth-observation satellite, and
was the first to collect publicly available high-resolution
imagery at 1- and 4-meter resolution. It offers multispectral
(MS) and panchromatic (PAN) imagery.
ILWIS The Integrated Land and Water Information
System is an open-source combined GIS and image
processing software developed by ITC, and maintained by
53North.
InSAR Interferometric SAR is a radar technique used in
geodesy and remote sensing. This geodetic method uses two
or more synthetic aperture radar (SAR) images to generate
maps of surface deformation or digital elevation, using
differences in the phase of the waves returning to the
satellite.
262
IRS Indian Remote Sensing satellites are a series of Earth

Observation satellites, built, launched, and maintained by
Indian Space Research Organisation. The IRS series provides
many remote sensing services to India.
ITC The Faculty of Geo-Information Science and
Observation of the University of Twente is an organization
that provides international postgraduate education,
research, and project services in the field of geo-information
science and earth observation using remote sensing and
GIS, located in Enschede, the Netherlands.
LANDSAT The longest running enterprise for acquisition
of imagery of Earth from space. The first Landsat satellite
was launched in 1972; the most recent, Landsat 7, was
launched on April 15, 1999. Landsat 7 data has eight
spectral bands with spatial resolutions ranging from 15 to
60 m; the temporal resolution is 16 day.
LiDAR Light Detection And Ranging is an optical remote
sensing technology that can measure the distance to, or
other properties of, a target by illuminating the target with
light, often using pulses from a laser.
LIMES Land and Sea Integrated Monitoring for
Environment and Security is an Integrated Project cofunded by the European Commission within the 6th
Framework Programme Aeronautics&Space/GMES
Security.
MARSOP MARS Crop Yield Forecasting System (MCYFS)
carried out by AGRI4CAST and FOODSEC actions within
the Institute for the Protection and the Security of the
Citizen (EU JRC).
MERIS Medium Resolution Imaging Spectrometer is one
of the main instruments on board of the European Space
Agency (ESA)s Envisat platform, to observe the color of the
ocean, both in the open ocean and in coastal zones.
Mobile-GIS A GIS that is running on a mobile, hand held
device, which is linked to a GPS for collecting spatial and
attribute data in digital format directly in the field.
MODIS Moderate Resolution Imaging Spectro-radiometer
is a sensor launched into Earths orbit by NASA in 1999 on
board of the Terra (EOS AM) Satellite, and in 2002 on
board of the Aqua (EOS PM) satellite, designed to provide
measurements in large-scale global dynamics including
changes in Earths cloud cover, radiation budget, and
processes occurring in the oceans, on land, and in the lower
atmosphere.
Natural hazard A potentially damaging physical event,
phenomenon, or human activity that may cause loss of life
or injury, property damage, social and economic disruption,
or environmental degradation. This event has a probability
of occurrence within a specified period of time and within a
given area, and has a given intensity.
NDVI Normalized Differential Vegetation Index is a
simple numerical indicator that can be used to analyze
remote sensing measurements, typically but not necessarily
from a space platform, and assess whether the target being
observed contains live green vegetation or not.
NOAA National Oceanic and Atmospheric
Administration, USA.
OneGeology It is an international initiative of the
geological surveys of the world, launched in 2007 as a
contribution to the International Year of Planet Earth, with

the aim to create a web-based geological map of the world.
ORCHESTRA One of the European Unions major
research and innovation projects for risk management, to
help national and local governments predict and react to
natural disasters by joining up national and local
information systems and applications.
OSM Open Street Map is a collaborative project to create a
free editable map of the world. The maps are created using
data from portable GPS devices, aerial photography, other
free sources, or simply from local knowledge.
PAGER Prompt Assessment of Global Earthquakes for
Response, a tool for the rapid assessment of peoples
exposed after an earthquake, developed by the USGS.
PALSAR Phased Array type L-band Synthetic Aperture
Radar sensor on board of the Advanced Land Observing
Satellite (ALOS) developed by the National Space
Development Agency of Japan (NASDA).
PGA Peak ground acceleration is a measure of earthquake
acceleration on the ground and an important input
parameter for earthquake engineering.
PGIS Participatory GIS is encompassing participatory
approaches to planning of spatial information and
communication management using mobile GIS.
PML Probable maximum loss is the anticipated value of
the largest loss that could result from the destruction and
the loss of use of property.
Preparedness The knowledge and capacities developed by
governments, professional response and recovery
organizations, communities, and individuals to effectively
anticipate, respond to, and recover from the impacts of
likely, imminent, or current hazard events or conditions.
Prevention The avoidance of adverse impacts of hazards
and related disasters.
PREVIEW Prevention, Information, and Early Warning
pre-operational services to support the management
of risks.
PRISM Panchromatic Remote-sensing Instrument for
Stereo Mapping, on board of the Advanced Land Observing
Satellite (ALOS).
Probability A measure of the degree of certainty. This
measure has a value between zero (impossibility) and 1.0
(certainty). It is an estimate of the likelihood of the
magnitude of the uncertain quantity, or the likelihood of
the occurrence of the uncertain future event.
PSInSAR Permanent Scatterers Radar Interferometry is a
relatively recent development from conventional InSAR,
and relies on studying pixels that remain coherent over a
sequence of interferograms.
Qualitative risk analysis An analysis that uses word form
and descriptive or numeric rating scales to describe the
magnitude of potential consequences and the likelihood
that those consequences will occur.
Quantitative risk analysis An analysis based on
numerical values of the probability, vulnerability, and
consequences, and resulting in a numerical value of the risk.
QuickBird A high-resolution commercial Earthobservation satellite, owned by DigitalGlobe and launched
in 2001.
RADARSAT The RADARSAT constellation is a pair of

Canadian remote-sensing satellites. The constellation
consists of RADARSAT-1 launched in 1995 and RADARSAT2 launched in 2007.
RADIUS Risk assessment tools for diagnosis of urban
areas against seismic disasters.
Remote sensing The acquisition of information about an
object or phenomenon without making physical contact
with the object. In modern usage, the term generally refers
to the use of aerial sensor technologies to detect and classify
objects on Earth (both on the surface and in the
atmosphere and oceans) by means of propagated signals
(e.g., electromagnetic radiation emitted from aircraft or
satellites).
Resilience The ability of a system, community, or society
exposed to hazards to resist, absorb, accommodate to, and
recover from the effects of a hazard in a timely and efficient
manner, including through the preservation and restoration
of its essential basic structures and functions.
Resourcesat The Indian Remote Sensing (IRS) P6 satellite
operated by ISRO, and includes the two remote sensing
instruments known as LISS-III and AWiFS.
Response The provision of emergency services and public
assistance during, or immediately after, a disaster in order to
save lives, reduce health impacts, ensure public safety, and
meet the basic subsistence needs of the people affected.
Risk The probability of harmful consequences or expected
losses (deaths, injuries, property, livelihoods, economic
activity disrupted, or environment damaged), resulting
from interactions between various types of hazards and
vulnerable conditions in a given area and over a specific
time period.
Risk analysis The use of available information to estimate
the risk to individuals or populations, property, or the
environment from hazards. Risk analysis generally contains
the following steps: hazard identification, hazard
assessment, elements-at-risk/exposure analysis,
vulnerability assessment, and risk estimation.
Risk assessment The process of risk analysis and risk
evaluation.
Risk control or risk treatment The process of decision
making for managing risks, and the implementation, or
enforcement of risk-mitigation measures and the reevaluation of its effectiveness from time to time, using the
results of risk assessment as one input.
Risk evaluation The stage at which values and judgements
enter the decision process, explicitly or implicitly, by
including consideration of the importance of the estimated
risks and the associated social, environmental, and
economic consequences, in order to identify a range of
alternatives for managing the risks.
Risk management The complete process of risk
assessment and risk control (or risk treatment).
Risk perception The way people/communities/authorities
judge the severity of the risk based on their personal
situation, social, political, cultural, and religious
background, economic level, their level of awareness, the
information they have received regarding the risk, and the
way they rate the risk in relation with other problems.
263
RiskScape New Zealand software and methodology for

multi-hazard risk assessment.
Risk transfer The process of formally or informally
shifting the financial consequences of particular risks from
one party to another whereby a household, community,
enterprise, or state authority will obtain resources from the
other party after a disaster occurs, in exchange for on-going
or compensatory social or financial benefits provided to
that other party.
SAFER Services and applications for emergency
Response, an EU GMES Emergency Response Service, with
the aim to provide space-based products in support of
European decision-makers facing natural and technological
disasters.
SAR Synthetic aperture radar is a form of radar whose
defining characteristic is its use of relative motion between
an antenna and its target region to provide distinctive longterm coherent-signal variations that are exploited to obtain
finer spatial resolution than is possible with conventional
beam-scanning means. It originated as an advanced form of
side-looking airborne radar (SLAR).
SDI Spatial data infrastructure is a framework of spatial
data, metadata, users, and tools that are interactively
connected in order to use spatial data in an efficient and
flexible way.
SELENA SEimic Loss EstimatioN using a logic tree
Approach, an open risk assessment package consisting of
the two separate software tools SELENA (Seismic Loss
Estimation using a Logic Tree Approach) and RISe (Risk
Illustrator for SELENA), developed by the University of
Alicante, NOSAR, and ICG.
ShakeMaps A GIS-based tool for earthquake hazard
assessment, developed by the USGS in cooperation with
regional seismic-network operators. ShakeMaps provides
near-real-time maps of ground motion and shaking
intensity after important earthquakes.
SMCE Spatial Multi-Criteria Evaluation, a tool for
decision-aid and a mathematical tool allowing the
comparison of different alternatives or scenarios according
to many criteria, often conflicting, in order to guide the
decision maker toward a judicious choice.
Societal risk The risk of multiple fatalities or injuries
in society as a whole: one where society would have to
carry the burden of a landslide causing a number of
deaths, injuries, financial, environmental, and other
losses.
SPOT Satellite Pour lObservation de la Terre, a highresolution, optical imaging Earth observation satellite
system from France.
SRTM Shuttle Radar Topography Mission (SRTM), an
international research effort that obtained digital elevation
models on a near-global scale from 561S to 601N to
generate a high-resolution digital topographic database of
the Earth.
Technological hazard A hazard originating from
technological or industrial conditions, including accidents,
dangerous procedures, infrastructure failures, or specific
human activities that may cause loss of life, injury, illness or
other health impacts, property damage, loss of livelihoods
264
and services, social and economic disruption, or

environmental damage.
TMPA TRMM Multi-satellite Precipitation Analysis.
TRMM Tropical Rainfall Measuring Mission, a joint NASA
and JAXA mission to monitor and study tropical rainfall
from satellite imagery.
UN-ISDR United Nations International Strategy for
Disaster Risk Reduction.
UN-SPIDER United Nations Platform for Space-based
Information for Disaster Management and Emergency
Response.
USGS United States Geological Survey.
Ushahidi Free open-source software for co-ordinating
citizen reports on a national crisis.
VDV Virtual Disaster Viewer, a crowdsourcing tool for
disaster-damage assessment developed by ImageCat.
Vulnerability The conditions determined by physical,
social, economic, and environmental factors or processes,
which increase the susceptibility of a community to the

impact of hazards. Can be subdivided in physical, social,
economical, and environmental vulnerability.
WASP Weighted Anomaly of Standardized Precipitation
gives an estimate of the relative deficit or surplus of
precipitation for different time intervals ranging from 1- to
6-months.
Web-GIS It is the process of designing, implementing,
generating, and delivering maps on the World Wide Web
with an emphasis on analysis and processing of project
specific geodata and exploratory aspects.
WorldView Commercial high-resolution panchromatic
satellite owned by DigitalGlobe. WorldView-2 provides
commercially available panchromatic imagery of 0.5 m
resolution, and eight-band multispectral imagery with
1.8 m resolution.
Abstract
The world has experienced an increasing impact of disasters in the past decades. Many regions are exposed to natural
hazards, each with unique characteristics. The main causes for this increase can be attributed to a higher frequency of
extreme hydro-meteorological events, most probably related to climate change and an increase in a vulnerable population.
To reduce disaster losses, more efforts should be applied toward disaster-risk management, with a focus on hazard assessment, elements-at-risk mapping, and vulnerability and risk assessment, all of which have an important spatial component. Multi-hazard assessment involves the assessment of relationships between different hazards, especially
concatenated or cascading hazards. The use of Earth observation (EO) products and geographic information systems (GIS)
has become an integrated approach in disaster-risk management. Hazard and risk assessments are carried out at multiple
scales, ranging from global to community levels. These levels have their own objectives and spatial data requirements for
hazard inventories, environmental data, triggering or causal factors, and elements-at-risk. This chapter provides an overview
of various forms of spatial data, and examines the approaches used for hazard and risk assessment. Specifically, hazard
examples include earthquakes, windstorms, drought, floods, volcanic eruptions, landslides, and forest fires. Several approaches are also treated that have been developed to generate elements-at-risk databases with emphasis on population and
building information, as these are the most used categories for loss estimation. Furthermore, vulnerability approaches are
discussed, with emphasis on the methods used to define physical vulnerability of buildings and population, and indicatorbased approaches used for a holistic approach, also incorporating social, economic, and environmental vulnerability, and
capacity. Finally, multi-hazard risk approaches and spatial-risk visualization are addressed. Multi-hazard risk assessment is a
complicated procedure, which requires spatial data on many different aspects and a multi-disciplinary approach. Within
this procedure, geographers and in particular geomorphologists can play a key role, as they are able to integrate spatial
information from various disciplines. The research challenge for geomorphologists is to focus on the integrated modeling
of multi-hazards that share the same triggering event or occur as cascading hazards.
3.10.1
Introduction
Disasters are headline news almost every day. Most happen in

far-away places and are rapidly forgotten. Others keep the
attention of the world media for a longer period of time. The
events that receive maximum media attention are those that
hit instantaneously and cause widespread losses and human
suffering, such as earthquakes, tsunamis, hurricanes, and
floods. Recent examples are the Indian Ocean tsunami (2004),
the earthquakes in Pakistan (2005), Indonesia (2006), China
(2008), Haiti (2010), and Japan (2011), and the hurricanes in
the Caribbean and the USA (2005, 2008). However, there
are many serious geomorphologic hazards that have a slow
onset, such as the recent drought in the Horn of Africa (2011),
soil erosion, land degradation, desertification, glacial retreat,
sea-level rise, loss of biodiversity, etc. These processes and

related events may cause local, regional, and global impacts in
the long run, but generally receive less attention.
Disasters are defined by the United Nations International
Strategy for Disaster Risk Reduction (UN-ISDR, 2004) as a
serious disruption of the functioning of a community or a
society causing widespread human, material, economic or
environmental losses which exceed the ability of the affected
community or society to cope using its own resources. Although the term natural disasters in its strict sense is not
correct, as disasters are a consequence of the interaction
between hazards and vulnerable societies, the term is used
extensively in the literature and also in daily use.
A hazard is defined as a potentially damaging physical
event, phenomenon or human activity that may cause the loss
of life or injury, property damage, social and economic disruption or environmental degradation. This event has a
probability of occurrence within a specified period of time and
within a given area, and has a given intensity (UN-ISDR,
2004). Hazards can be single, sequential, or combined in their
origin and effects. Each hazard is characterized by its location,
area affected (size or magnitude), intensity, speed of onset,
duration, and frequency. Hazards can be classified in several
ways. A possible subdivision is between natural, humaninduced, and human-made hazards. Natural hazards are
natural processes or phenomena in the Earths system (lithosphere, hydrosphere, biosphere, or atmosphere) that may
constitute a damaging event (e.g., earthquakes, volcanic
eruptions, and hurricanes). A subdivision of natural hazards
relates to the main controlling factors of the hazards leading
to a disaster. They may be hydro-meteorological (including
floods and wave surges, storms, droughts, and related disasters
such as extreme temperatures and forest/scrub fires, landslides,
and snow avalanches), geophysical hazards (resulting from
anomalies in the Earths surface or subsurface, such as
earthquakes, tsunamis, and volcanic eruptions), or biological
hazards (related to epidemics and insect infestations).
Human-induced hazards are those resulting from modifications of natural processes in the Earths system caused by
human activities that accelerate/aggravate the damage potential (e.g., land degradation, landslides, and forest fires).
Human-made hazards originate from technological or industrial accidents, dangerous procedures, infrastructure failures,
or certain human activities, which may cause the loss of life
or injury, property damage, social and economic disruption,
or environmental degradation (e.g., industrial pollution,
nuclear activities and radioactivity, toxic wastes, dam failures,
and transport, industrial, or technological accidents such as
explosions, fires, and oil spills).
The objective of a hazard assessment is to identify the
various types of hazards that may threaten a territory, and to
partition the landscape in spatial zones that are characterized
by different expected intensities and frequencies of hazardous
processes. In a hazard assessment, several aspects should be
evaluated: 1) the triggering event; 2) the areas where hazards
are likely to initiate; 3) the areas where the hazards are likely
to spread; and 4) the expected intensity of the hazard and its
associated frequency or probability of occurrence. Hazards
may impact vulnerable societies, and may result in physical
and other types of damage to so called elements-at-risk. With
certain exceptions (e.g., ash clouds affecting air-traffic or oilspills affecting marine flora and fauna), these impacted
elements-at-risk are mostly located on the Earths surface.
Therefore, the relationship between the hazard events and
surface processes, landforms, and materials is a key component to study. Therefore, geomorphology plays a key role in
hazard and risk analysis. Geomorphology is the study of
landforms and surface materials and of the processes that have
formed or reshaped them. Geomorphologists investigate the
evolution of landscapes, and study the history and the dynamics of landforms and the processes responsible for creating or modifying them, through a combination of field
observations, physical experiments, and numerical modelling.
Numerous processes that have shaped the Earths surface can
be potentially dangerous if they exceed a certain threshold
265
(e.g., they may result in instability and erosion on slopes and

catastrophic flooding in river or coastal areas).
The study of the endogenic (volcanic eruptions or earthquakes) and exogenic (extreme meteorological) triggering
events causing hazardous processes may sometimes be outside
of the scope of geomorphology, as this is the domain of
seismologists, volcanologists, meteorologists, and climatologists, although geomorphology is becoming more multidisciplinary. Nevertheless, geomorphologists contribute to the
reconstruction of the frequency, extent, and intensity of past
events by studying their imprints on landforms and surface
materials. Geomorphologists play a key role in the analysis of
how these triggering events result in hazardous processes on
the Earth surface. They are specialized in analyzing the possible spatial extent of future hazardous processes related to a
set of environmental factors (geomorphology, topography,
geology, soils, land cover, etc.). Examples of this are the study
of potential landslide areas or the analysis of the effects of
surface materials and landforms in the amplification of seismic waves or liquefaction. Geomorphologists are also good in
modeling the extent of the spreading of hazardous processes
over the topography (e.g., lahar flows, landslide runout, and
flood extent) based on the use of digital elevation models
(DEMs) and terrain analysis.
The study of hazards and risks has a very important spatial
component, as certain types of hazards are restricted to certain
geographical regions. Earthquakes occur along active tectonicplate margins and volcanos occur along subduction zones (e.g.,
around the margins of the Pacific plate, so-called Ring of Fire).
Tsunamis occur in the neighborhood of active-plate margins,
but their effects can be felt at considerable distances from their
origin as the waves can travel long distances. Tropical cyclones
(in North America called hurricanes and in Asia called
typhoons) occur in particular coastal zones. Landslides occur
in hilly and mountainous regions. In the analysis of hazard and
risk, geospatial technologies and Earth observation via remote
sensing are increasingly playing important roles. Remote sensing science and technology are now essential in monitoring
changes related to the Earths surface, oceans, and atmosphere,
and are increasingly used as the basis for early warning for
hazardous events. Furthermore, remote sensing enables thematic information extraction that represents the input data for
hazard modeling (e.g., topography, lithology, and land cover).
The use of spatio-temporal data and geographic information systems (GIS) has now become a part of an integrated
approach to disaster risk management. New GIS algorithms
and analysis/modeling techniques are revolutionizing the
potential capacity to analyse hazards, vulnerability, and risks.
These spatial information systems are used for storage, situation analysis, modeling, and visualization (Twigg, 2004).
Disaster-risk management benefits greatly from the use of
geospatial technologies because spatial and temporal variation
can be accounted for, and new methodologies can be developed and fully explored. One of the key advantages of using
GIS-based software tools for the risk decision-making process
is the possibility to explore what if analysis by varying parameters and generating alternative scenarios in a spatial context (Longley et al., 2005). Earlier publications on this topic
can be found in Wadge et al. (1993), Coppock (1995), Emani
(1996), and Kaiser et al. (2003).
266
Table 1 Summary of definitions related to disasters, hazards, and vulnerability

Term
Definition
Disaster
A serious disruption of the functioning of a community or a society causing widespread human, material, economic, or
environmental losses that exceed the ability of the affected community or society to cope using its own resources.
A potentially damaging physical event, phenomenon, or human activity that may cause loss of life or injury, property
damage, social and economic disruption, or environmental degradation. This event has a probability of occurrence
within a specified period of time and within a given area, and has a given intensity.
Population, properties, economic activities, including public services, or any other defined values exposed to hazards in a
given area. Also referred to as assets. The amount of elements-at-risk can be quantified either in numbers (of
buildings, people, etc.), monetary value (replacement costs, market costs, etc.), area, or perception (importance of
elements-at-risk).
Exposure indicates the degree to which the elements-at-risk are exposed to a particular hazard. The spatial interaction
between the elements-at-risk and the hazard footprints are depicted in a GIS by a simple map overlaying of the hazard
map with the elements-at-risk map.
The conditions determined by physical, social, economic, and environmental factors or processes, which increase the
susceptibility of a community to the impact of hazards. Can be subdivided in physical, social, economical, and
environmental vulnerability.
The positive managerial capabilities of individuals, households, and communities to confront the threat of disasters (e.g.,
through awareness raising, early warning, and preparedness planning).
The expected losses in a given area as a result of a given hazard scenario.
The probability of harmful consequences or expected losses (deaths, injuries, property, livelihoods, economic activity
disrupted, or environment damage) resulting from interactions between (natural, human-induced, or man-made)
hazards and vulnerable conditions in a given area and time period.
Natural hazard
Elements-at-risk
Exposure
Vulnerability
Capacity
Consequence
Risk
Source: Based on UN-ISDR, 2004. Terminology of disaster risk reduction. United Nations, International Strategy for Disaster Reduction, Geneva, Switzerland. http://www.unisdr.org/
eng/library/lib-terminology-eng%20home.htm
The objective of this chapter is to give an overview of the

application of geo-information science and Earth observation
in the analysis of hazards and risk. The chapter starts with a
background section discussing disaster trends and the framework for disasterrisk management. Subsequent sections discuss the use of GIS and remote sensing in the generation of
inventories of past hazardous events, the modelling of hazards, the generation of elements-at risk databases, and the
integration of these data in the modeling of potential losses.
3.10.2
3.10.2.1
Trends in Disaster Statistics
Data on disaster occurrences, their effect on people, and

their cost to countries are very important for disaster-risk
Risk
Hazards
Disaster
Hazards
Risk
Vulnerable
society
(a)
Background
Table 1 provides a summary of the various terms that are

relevant in the context of this chapter, relating to disasters,
disaster risk, and its various components (UN-ISDR,2004). It
is important to distinguish between the terms disaster, hazard,
and risk. Risk results from the combination of hazards, conditions of vulnerability, and insufficient capacity or measures
to reduce the potential negative consequences of risk (OKeefe
et al., 1976). When the hazard or threat becomes a reality (i.e.,
when it materializes), the risk becomes a disaster. For example,
a certain river valley may be prone to flooding. There is risk if a
vulnerable society/community or property is located within
this flood-prone area. If the hazard materializes, that is, if the
flood actually occurs, it will cause losses to the vulnerable
society or property, thus creating a disaster (Figure 1).
Hazards
Vulnerable
society
(b)
Vulnerable
society
(c)
Figure 1 Schematic representation of the relation between hazards,

vulnerable society, risk, and disasters. (a) Risk indicates the expected
losses to a vulnerable society as a result of hazards. (b) A disaster
occurs when the threat of a hazard becomes reality and impacts on a
vulnerable society. (c) Future trends of increasing hazards and
increasing vulnerability will lead to increasing risk.
management. There are now a number of organizations that

collect information on disasters at different scales and with
different objectives.
Since 1988, the Centre for Research on the Epidemiology of

Disasters (CRED) has been maintaining an Emergency
Events Database (EM-DAT, 2009). Disasters have to fulfill
certain criteria in order to be included in the EM-DAT
database: They have to cause at least 10 casualties; 100 or
more people should be affected; it should result in a declaration of emergency; or it should lead to a call for external
assistance.
Data on disaster impacts are also collected by reinsurance
companies. For instance, the MunichRe database for natural
catastrophes (NatCatSERVICE) includes more than 28 000
entries on material and human-loss events worldwide
(MunichRe, 2010). A similar disaster-event database
(SIGMA) is maintained by SwissRe. These data, however, are
not publicly available.
The Asian Disaster Reduction Center (ADRC) has initiated a
new disaster database, called Glidenumber (2010). The
specific feature of this database is that each disaster receives a
unique identifier and a number of relevant attributes.
At a local level, disaster data have been collected by an initiative of NGOs, called LaRed, initially in Latin America, but
later on expanding also to other regions. They generated a
tool called DesInventar (2010), which allows local authorities, communities, and NGOs to collect disaster information. Recently, the DesInventar database has become
available online.
There are also many disaster databases collected at the national level, or that are related to a specific type of hazard.
The Global Risk Identification Program (GRIP) and the
Centre for Research in Epidemiology of Disasters (CRED)
have initiated a service, called DisDAT, which brings together
all publicly available disaster databases from different
267
countries (GRIP, 2010). It contains 60 registered disaster

databases, of which 13 are global.
When examining the reported disasters in these databases,
there is a clear increase in hazardous events over the past
decades (Figure 2). The number of natural disasters in the last
decade has increased by a factor of 9 compared to the decade
19501959 (EM-DAT, 2009), which is mainly caused by an
increase in hydro-meteorological disasters. In terms of losses,
earthquakes resulted in the largest amount of losses (35% of
all losses), followed by floods (30%), windstorms (28%), and
others (7%). Earthquakes are also the main cause of fatalities,
which is estimated on the order of 1.4 million during the
period 19502000 (47%), followed by windstorms (45%),
floods (7%), and others (1%) (MunichRe, 2010; EM-DAT,
2009). It is interesting to note that human fatalities due to
natural disasters show a decreasing trend, which may be due
to better warning systems and improved disaster management,
but the number of people affected follows the increasing trend
of the number of events (see Figure 2).
Disaster information collected at the local level (e.g.,
DesInventar) is more complete, as it also includes smallmagnitude/high-frequency events, but the coverage of such
databases is limited worldwide. One of the major problems
with the use of disaster databases for natural hazard and risk
assessment is that they normally lack proper georeferencing of
the reported events (Verelst, 1999). A comparative study of the
EM-DAT, Sigma, and NATCAT databases carried out for four
countries showed that these databases differed significantly
(Guha-Sapir and Below, 2002).
The increase in the number of disasters, the losses, and
people affected cannot be explained only by better reporting
methods and media coverage of disasters, lack of which
probably made the number too low for the first part of the last
century. A number of factors influence the increase in the
number of disasters that can be subdivided as those leading to
500
Million
300
300 000
200
200 000
Nr. disasters reported
Nr. people killed
Nr. people affected
100 000
150
100
100
50
19
00
19
10
19
20
19
30
19
40
19
50
19
60
19
70
19
80
19
90
20
00
20
09
Nr. disasters reported
400 000
Nr. people killed
200
400
Nr. people affected (millions)
500 000
Year
Figure 2 Summary of natural disasters, showing the number of reported disasters, the number of people killed, and the number of people
affected over the period 19002009. Reproduced from EM-DAT, 2009. The OFDA/CRED International Disaster Database, Universite Catholique de
Louvain, Brussels, Belgium. http://www.emdat.be
268
a larger vulnerability, and those leading to a higher occurrence

of hazardous events.
The increased vulnerability is due to a number of reasons
including the rapid increase of the world population, which
has doubled in size from 3 billion in the 1960s to 7 billion in
2011 (World Bank, 2011). Depending on the expected growth
rates, world population is estimated to be between 7.9 and
11.0 billion by the year 2050 (UNPD, 2010a). The increase in
disaster impact, however, is higher than the increase in
population, which indicates that other important factors are
involved that increase the overall vulnerability of the world
population. One of the main factors is the large urbanization
rate. According to United Nation figures (UNPD, 2010b), the
worldwide urbanization percentage has increased from 29%
in 1950 to 50% in 2010, and is expected to rise to 69% in
2050. Another factor related to the population growth is that
areas become settled, that were previously avoided due to their
susceptibility to natural hazards. Many of the largest cities in
the world, the so-called Megacities, are located in hazardous
regions, either in coastal zones or in seismically active regions
(Smith and Petley, 2008; Kraas, 2008).
The increasing impact of natural disasters is also related
with the development of highly sensitive technologies, and the
growing susceptibility of modern industrial societies to
breakdowns in their infrastructure. Data from MunichRe
(2010) show that the economic losses have increased by a
factor of 8 over the past 50 years, and insured losses by a factor
of 15. There is a rapid increase in the insured losses, which are
mainly related to losses occurring in developed countries.
Windstorms clearly dominate the category of insured losses
(US $90 billion), followed by earthquakes (US $ 25 billion).
Insured losses to flooding are remarkably less (US $ 10 billion) due to the fact that they are most severe in developing
countries with lower insurance coverage (MunichRe, 2010).
It is not only the increased exposure of the population to
hazards, however, that can explain the increase in natural
disasters. The frequency of destructive events related to atmospheric extremes (such as floods, drought, cyclones, and
landslides) is also increasing (EM-DAT, 2009). During the past
10 years a total of 3 750 windstorms and floods were recorded,
accounting for two-thirds of all events. The number of catastrophes due to earthquakes and volcanic activity (approximately 100 per year) has remained constant (MunichRe,
2010). Although the time-span is still not long enough to indicate it with certainty, these data suggest that climate change
is related to the increased occurrence of natural disasters
(IPCC, 2007).
An inverse relationship exists between the level of development and loss of human lives in the case of disasters. Approximately 85 % of the disaster-related casualties occur in
less developed countries, where more than 4.7 billion people
live. The greater loss of lives is due to the lower quality of
buildings, lack of building codes or lack of enforcement,
construction of buildings in hazardous areas due to lack of
land-use planning, lower awareness and disaster preparedness,
less accurate or missing early-warning systems, lack of evacuation planning, and lack of facilities for search-and-rescues
and medical attention. Although 65% of the overall losses
occur in high-income countries (with GNI US$ 412 000 per
capita) (World Bank, 2010), and only 3% in low-income
countries (GNI US$ o1000 per capita), the effect in the latter
group is devastating, as they may represent as much as 100%
of their gross national income (UN-ISDR, 2009). Economic
losses in absolute terms (billions of dollars) show an increase
with the level of development, as the absolute value of
elements-at-risk that might be damaged during a disaster increases with increasing level of development. In relative terms,
however, the trend is reversed, showing a decrease in the losses
expressed as percentage of GDP with increasing level of
development (MunichRe, 2010).
3.10.2.2
Disaster-Risk Management Framework
Disaster Risk Management (DRM) is defined as the systematic process of using administrative decisions, organization,
operational skills, and capacities to implement policies,
strategies, and coping capacities of the society and communities to lessen the impacts of natural hazards and related
environmental and technological disasters. This comprises all
forms of activities, including structural and non-structural
measures to avoid (prevention) or limit (mitigation and preparedness) adverse effects of hazards (UN-ISDR, 2004). Disaster-risk management is aimed at disaster-risk reduction,
which refers to the conceptual framework of elements considered with the possibilities to minimize vulnerabilities and
disaster risks within the broad context of sustainable development (UN-ISDR, 2004).
The past decades have witnessed a shift in focus from
disaster recovery and response to risk management and mitigation. This change was also from an approach that focused
primarily on the hazard as the main causal factor for risk,
and the reduction of the risk by physical-protection measures,
to a focus on vulnerability of communities and ways to
reduce those through preparedness and early warning. Later,
more focus was given to the strengthening of the capacity
of local communities and the development of communitybased coping strategies (Blaikie et al., 1994; Lavel, 2000;
Pelling, 2003). The Yokohama conference in 1994 put into
perspective the socio-economic aspects as a component of
effective disaster prevention. It was recognized that social
factors such as cultural tradition, religious values, economic
standing, and trust in political accountability are essential
in the determination of societal vulnerability. In order to
reduce societal vulnerability, and therewith decrease the
consequences of natural disasters, these factors need to be
addressed (Hillhorst, 2004). The ability to address socioeconomic factors requires knowledge and understanding of
local conditions, which can, in most cases, only be provided
by local participants.
The decade from 1990 to 2000 was declared by the United
Nations as the International Decade for Natural Disaster Reduction (IDNDR). As the impact of disasters increased dramatically during this decade, the international community
decided to continue this effort after 2000 in the form of an
International Strategy for Disaster Reduction (ISDR). The
ISDR stressed the need to move from a top-down management of disasters and a cycle that focuses on reconstruction
and preparedness, toward a more comprehensive approach
that tries to avoid or mitigate the risk before disasters occur,
and at the same time fosters more awareness, public commitment, knowledge sharing, and partnerships to implement
various risk reduction strategies at all levels (UN-ISDR,
2005b).
This more positive concept has been referred to as the
risk-management cycle, or better spiral, in which learning
from a disaster can stimulate adaptation and modification in
development planning, rather than a simple reconstruction of
pre-existing social and physical conditions. This is illustrated
in Figure 3, by showing the disaster cycle and various components (relief, recovery, reconstruction, prevention, and
preparedness), and how these changed through time. Initially,
most emphasis was given to disaster relief, recovery, and reconstruction, thereby getting into a cycle where the next disaster was going to cause the same effects or worse (Figure 3A).
Later on, more attention was given to disaster preparedness
(Figure C), by developing warning systems and disaster
awareness programs (Figure 3B). Now, the efforts are focusing
on disaster prevention and preparedness Figure 3C, thus increasing the time gap between individual disasters, and reducing their effects, requiring less emphasis in relief, recovery,
and reconstruction. The eventual aim of disaster-risk management is to enlarge this cycle, and only reach the response
phase for extreme events with very low frequency.
Disaster prevention is achieved through risk management.

Figure 4 presents the general risk-management framework,
which is composed of a risk-assessment block and a block in
which risk-reduction strategies are defined. A summary of the
terminology used in risk management is given in Table 2.
Central to the procedure is risk analysis, in which the available
information is used to estimate the risk to individuals or
populations and property or the environment from various
hazards. Risk analysis generally contains the following steps:
1) hazard identification; 2) hazard assessment; 3) elements-atrisk/exposure analysis; 4) vulnerability assessment; and 5) risk
estimation. Risk evaluation is the stage at which values and
judgments enter the decision process, explicitly or implicitly,
by including consideration of the importance of the estimated
risks and the associated social, environmental, and economic
consequences, in order to identify a range of alternatives for
reducing the risks (UN-ISDR, 2004).
Risk assessment is the combination of risk analysis and
risk evaluation. It is more than a purely scientific enterprise
and should be seen as a collaborative activity that brings
professionals, authorized disaster managers, local authorities,
and the people living in the exposed areas together (OBrien,
2000; Plapp, 2001; Montague, 2004). Risk governance is
therefore an integral component. The final goal, reduction of
Prevention
Reconstruction
Preparedness
Recovery
B
Relief
Disaster
event
Reconstruction
Prevention
Recovery
Preparedness
Relief
Disaster
event
Recovery
Reconstruction
Prevention
Relief
Preparedness
Disaster
event
Figure 3 Disaster cycle and its development through time.
269
Ti
270
Risk management
Risk assessment
Risk analysis
Hazard
identification
Risk monitoring and updating
Vulnerability
assessment
Risk estimation
Risk evaluation
Risk visualization
Non-structural
Risk transfer
Emergency planning
Awareness/training
Early warning
Landuse planning
EIA/SEA
Structural
Codes
Standard
Reinforcing
Protective measures
Risk governance/Risk communication
Elements at risk/
exposure
(Spatial) Information management
Hazard
assessment
Cost-benefit
assessment
Risk reduction
Figure 4 Risk-management framework.
Table 2 Summary of definitions related to risk management

Term
Definition
Risk analysis
The use of available information to estimate the risk to individuals or populations, property, or the environment from
hazards. Risk analysis generally contains the following steps: hazard identification, hazard assessment, elements-at-risk/
exposure analysis, vulnerability assessment, and risk estimation.
The stage at which values and judgements enter the decision process, explicitly or implicitly, by including consideration of
the importance of the estimated risks and the associated social, environmental, and economic consequences, in order to
identify a range of alternatives for managing the risks.
The process of risk analysis and risk evaluation.
The process of decision making for managing risks, and the implementation, or enforcement of risk-mitigation measures
and the re-evaluation of its effectiveness from time to time, using the results of risk assessment as one input.
The complete process of risk assessment and risk control (or risk treatment).
Risk evaluation
Risk assessment
Risk control or risk
treatment
Risk management
Source: Based on UN-ISDR, 2004. Terminology of disaster risk reduction. United Nations, International Strategy for Disaster Reduction, Geneva, Switzerland. http://www.unisdr.org/
eng/library/lib-terminology-eng%20home.htm
disaster risk, should be achieved by combining structural and

non-structural measures that focuses on emergency preparedness (e.g., awareness raising, early-warning systems, etc.),
inclusion of risk information in long term land-use planning,
and evaluation of the most cost-effective risk-reduction
measures (see Figure 4). In the entire risk-management
framework, spatial information plays a crucial role, as the
hazards are spatially distributed, as well as the vulnerable

elements-at-risk.
3.10.2.3
Risk-Analysis Framework
As illustrated in Figure 5, there are three important components in risk analysis: 1) hazards; 2) vulnerability; and 3)
Risk =
Probability of losses occurring
Risk =
Hazard
Temporal
probability
Temporal
probability
Degree of loss to
elements at risk
Vulnerability
Hazard intensity
Vulnerability
function
% Damage
Duration
Time of onset
Type of hazard
Amount
Consequences or
losses
Hazard
Temporal probability
Vulnerability
271
Quantification of
elements at risk
Elements at
risk
Type of elements
at risk
Type 1
Type 2
Spatial extent
Initiation
Spreading/runout
Harard
intensity
Temporal variation
quantification
Number
Economic value
Location
Exposure:
Spatial overlay
of hazard and
elements at
risk
Figure 5 Risk analysis and its components.
elements-at-risk (Van Westen et al., 2008). They are characterized by both spatial and non-spatial attributes. Hazards are
characterized by their temporal probability and intensity, derived from frequency-magnitude analysis. Intensity expresses
the severity of the hazard, for example, water depth, flow
velocity, and duration in the case of flooding. The hazard
component in the equation actually refers to the probability of
occurrence of a hazardous phenomenon with a given intensity
within a specified period of time (e.g., annual probability).
Hazards also have an important spatial component, both related to the initiation of the hazard (e.g., a volcano) and the
spreading of the hazardous phenomena (e.g., the areas affected by volcanic products such as lava flows) (Van Westen,
2009).
Elements-at-risk or assets are the population, properties,
economic activities, including public services, or any other
defined values exposed to hazards in a given area (UN-ISDR,
2004). Elements-at-risk also have spatial and non-spatial
characteristics. There are many different types of elements-atrisk and they can be classified in various ways. The way in
which the amount of the elements-at-risk is characterized (e.g.,
as number of buildings, number of people, economic value, or
qualitative rating according to their importance) also defines
the way in which the risk is presented. The interaction of
elements-at-risk and hazard defines the exposure and the
vulnerability of the elements-at-risk. Exposure indicates the
degree to which the elements-at-risk are actually located in an

area affected by a particular hazard. The spatial interaction
between the elements-at-risk and the hazard footprints are
depicted in a GIS by spatial overlay of the hazard map with the
elements-at-risk map (Van Westen, 2009).
Vulnerability refers to the conditions determined by physical, social, economic, and environmental factors or processes,
which increase the susceptibility of a community to the impact
of hazards (UN-ISDR, 2004). The vulnerability of communities
and households can be analysed in a holistic qualitative manner using a large number of criteria that characterize the
physical, social, economic, and environmental vulnerability.
The importance of each of these indicators is evaluated by assigning weights and combining them using spatial multicriteria evaluation. Physical vulnerability is evaluated as the
interaction between the intensity of the hazard and the type of
element-at-risk, making use of so-called vulnerability curves.
For further explanations on hazard and risk assessment, see
Alexander (1993), Okuyama and Chang (2004), Smith and
Petley (2008), and Alcantara-Ayala and Goudie (2010).
3.10.3
Hazard Assessment
Figure 6 presents a schematic overview of a set of

natural hazards and their causeeffect relationships. The
272
Volcanic
eruptions
Meteorological
extremes
Earthquakes
Lava
flows
Fault
rupture
Wind
storms
Pyroclastic
flows
Tsunamis
Strong
winds
Drought
Cold wave
Ash fall
Ground
shaking
Storm
surge
Precipitat
ion
Heat wave
Lahars
Rainfall
extremes
Temperature
extremes
Flooding
Forest fires
Mass
movements
Soil erosion
Snow
avalanches
Land
degradation
Figure 6 Examples of multi-hazards and their interactions required for multi-hazard risk assessment. Partly based on CAPRA, 2009. Central
American Probabilistic Risk Assessment (CAPRA). World Bank. www.ecapra.org
upper row in the figure consists of the triggering events,

which are the endogenic or exogenic events that precipitate
other events. They may cause direct effects, such as ground
shaking resulting from an earthquake (Jimenez et al., 2000),
drought caused by deficiency in precipitation (Karnieli
and DallOlmo, 2003), pyroclastic flows and ash fall following
a volcanic eruption (Zuccaro et al., 2008), or wind speeds
caused by tropical cyclones (Holland, 1980; Emanuel et al.,
2006). The direct effects may trigger indirect effects, or
secondary hazards, such as landslides caused by ground
shaking in mountainous areas (Jibson et al., 1998), landslides
and floods occurring in recently burned areas (Cannon
et al., 2008), or tsunamis caused by earthquake-induced surface displacement in the sea (Priest et al., 2001; Ioualalen
et al., 2007). Secondary hazards that are caused by other
hazards are also referred to as concatenated hazards or
cascading hazards. Figure 6 aims to depict the interrelationships between the triggering factors, the primary hazards
and secondary hazards. These relationships can be very complex, for instance, the occurrence of floods as a result of
the breaking of earthquake-induced landslide dams (Korup,
2002). Given this complexity, a multi-hazard assessment,
which forms the basis for subsequent risk assessment, should
always lead to some sort of simplification in terms of the
cause-effect relationships.
There are relatively few examples in literature on such
complete multi-hazard assessments, and most studies focus
on the evaluation of individual hazard types. Some of the
best examples of a multi-hazard assessment approach
and subsequent risk assessment will be discussed later in
Section 3.10.5.
3.10.3.1
Scale and Hazard Assessment
Hazard assessment using GIS can be carried out at different

geographical scales. Although it is possible to use a range of
spatial resolutions of the input data for GIS analysis (measurement scale), in practice the geographic scale determines the
size of the study area. This in turn restricts the scale of the
input data and the computational scale of analysis. The geographical scale also effectively determines the cartographic
scale at which information will be mapped. A number of
factors play a role in deciding what is the appropriate scale for
hazard and risk assessment (Fell et al., 2008; Van Westen et al.,
2008). These are related first of all to the aim of the hazard
assessment, the type of hazard, and the operational scale at
which these hazard processes are triggered and manifest
themselves. They also relate to the size and characteristics of
the study area, the available data and resources, and the required accuracy. Table 3 provides an overview of scales and
levels for different hazard types.
Hazard assessments that are carried out for the entire
Earth (global scale) focus on global problems, such as climate
change, or are aimed at displaying the distributions of a
particular hazard worldwide (e.g., land degradation). Risk
assessment at this scale is mainly intended to generate
risk indices for individual countries, to link them to indices
related to socio-economic development, and to make prioritizations for support by international organizations, such as
the World Bank, ADB, WHO, UNDP, FAO, etc. (Cardona,
2005; Peduzzi et al., 2009). The input data have a scale less
than 1:10 million and spatial resolutions on the order of
15 km. Under the umbrella of the ProVention Consortium
273
Table 3 Scales/levels for hazard assessment, with indication of basic mapping units and the optimal scale for displaying different types of
hazards (EQEarthquakes, VOVolcanic hazards, DRDrought, WSWindstorms, FLFloods, CO Coastal, LS Landslides, WF Wildfire).
Indicated is the applicability ( highly applicable, moderately applicable, and Less applicable)
Scale
Level
Cartographic
scale
(million)
Spatial
resolution
Area covered
(km2)
EQ
VO
DR
WS
FL
CO
LS
WF
Global
Very small
o1:5
15
15 km
1
148 million
520 million

Small
Global
Continental / large
countries
National
0.11
0.11 km
Regional
Medium
Large
Provincial
Municipal
Community
0.050.1
0.0250.05
40.025
100 m
10 m
15 m
30600
thousand
100010 000
100
10

staff from the Hazard Management Unit of the World Bank,

the Development Economics Research Group (DECRG) and
Columbia University carried out a global-scale multi-hazard
risk analysis that focused on identifying key hotspots where
the risks of natural disasters are particularly high (Dilley et al.,
2005). The project resulted in a series of global hazard and
risk maps, which can be downloaded from a website (CIESIN,
2005).
For individual continents or regions covering several
countries, hazard applications focus on understanding the
triggering mechanism(s) of hazards that cover vast areas (e.g.,
millions of km2) including tropical cyclones, earthquakes, or
drought. Applications also include evaluation of hazards
crossing national boundaries (e.g., flood hazard in large
catchments like the Rhine, Ganges, etc.), or those critical to
reduction policies at the international level (e.g., for the entire
European Union). The hazard maps are generated using
standardized methodologies, and used for risk assessment,
early warning (De Roo et al., 2007), and post-disaster damage
assessment. The geographic areas that are evaluated vary in
size, as some countries like China, India, or the USA are as
large as continents like Europe, but under one administrative
setup. The scale of the input maps can range between
1:100 000 and 1:5 million, and spatial resolutions may vary
from 90 m to 1 km, depending on the application.
Hazard and risk assessment at the national scale cover
areas ranging from tens to several hundred thousand km2,
depending on the size of the country. Hazard assessment
is carried out at a national scale for national planning purposes, implementation of national disaster-risk reduction
policies, early-warning systems, disaster preparedness, and
insurance. The applications in planning become more concrete when zooming in on larger scales such as the provincial
level. For instance, hazard and risk assessment become
an integral component of regional development plans and
environmental impact assessments for infrastructure developments. At the municipal level, hazard and risk assessment
are carried out as a basis for land-use zoning and for the design of non-structural risk-reduction measures. At a community level, hazard and risk assessment are carried out in
participation with local communities and local authorities, as
a means to obtain commitment for disaster-risk reduction
programmes.
3.10.3.2
Spatial Data for Hazard Assessment
The assessment of multi-hazards and the subsequent risk assessment is a very data intensive procedure. The availability of
certain types of spatio-temporal data can be one of the main
limitations for carrying out specific types of analyses. Table 4
provides a schematic overview of the main GIS data layers
required for hazard and risk assessment for different hazard
types. These can be subdivided into hazard inventory data and
triggering events, and environmental factors.
3.10.3.2.1
Hazard inventories and triggering events
The hazard-inventory data are by far the most important, as

they should provide insight into the distribution of past hazardous phenomena, their types, mechanisms, causal factors,
frequency of occurrence, intensities, and the damage that has
been caused. The most straightforward way of generating
hazard inventories is through direct measurements of the
phenomena. Sensor-station networks are used to collect such
measurements (e.g., earthquake strong-motion data, flooddischarge stations, meteorological stations, coastal-tide gauging stations, or wave-measurement buoys).
Seismic networks have been formed globally (NERIES,
2009; ANSS, 2009; GSN, 2009), and the data can be managed
centrally and distributed using web-mapping applications
(e.g., the United States Geological Survey (USGS)). In the US,
a similar network has been established for recording streamdischarge data for nearly 10 000 sites in a central database
linked with a web-mapping service (NWIS, 2010). Although a
tsunami warning system has been operational in the Pacific
Ocean for a number of decades, the 2004 Indian Ocean tsunami has prompted the international community to implement such systems worldwide. For these sensor networks, the
spatial coverage is important so that potentially hazardous
areas are monitored. The density of observations required for
the monitoring networks differs strongly for various hazard
types. This is more problematic for flood-discharge stations, as
each potential hazardous river needs to be monitored,
whereas for seismic stations, the required density can be much
less. Also, the spacing between the individual stations is of
importance given the variability of specific parameters (e.g.,
rainfall varies significantly over relatively short distances in
mountain regions). The period for which measurements are
274
Table 4 Overview of spatial data for hazard assessment, and their relevance for different types of hazards ( highly relevant,
moderately relevant, and less relevant). (EQ Earthquakes, VO Volcanic hazards, DRDrought, WSWindstorms, FLFloods,
CO Coastal, LSLandslides, WF Wildfire)
Group
Data layer and types
EQ
VO
DR
WS
FL
CO
LS
WF
Hazard inventories
Satellite-based monitoring
Ground-based networks
Archive studies
Visual image interpretation
Field mapping
Participatory approaches
Dating methods
Relief
Altitude difference (in time)
Slope steepness
Slope direction
Flow accumulation
Rock types
Weathering
Faults
Structural geology
Soil types
Soil depth
Geotechnical properties
Hydrological properties
Discharge
Ground water tables
Soil moisture
Runoff
Physiographic units
Origin/genesis
Landforms
Active processes
Natural vegetation
Land use
Vegetation changes
Land-use changes
Linear infrastructures
Built-up areas
Rainfall
Temperature
Wind speed & direction
Wave height
Tides
Earthquakes
Volcanic eruptions

Topography
Geology
Soils
Hydrology
Geomorphology
Land use
Triggering factors
available, and the spatio-temporal continuity of the measurements, play an important role, as generally the period for
which measurements are available is not sufficiently large to
capture major events from the past. Catalogs from the measurement networks should be carefully analysed before being
used in a hazard assessment. The monitoring networks located
on the ground or in the oceans are supported by a number of
satellite systems that are used for transmitting information
to data centres. There are also a large variety of satellite-based
monitoring systems that can measure characteristics of
hazards over larger areas on a regular basis, such as seasurface temperature, rainfall, altitude, clouds, and vegetation
indices.
For larger areas, lack of data from meteorological stations
can be substituted with rainfall estimates from satellite
imagery, such as from the Tropical Rainfall Measuring Mission

(TRMM) and Multi-Satellite Precipitation Analysis (TMPA),
which are used to issue landslide and flood warnings based on
a threshold value derived from earlier published intensityduration-frequency relationships for different countries (Hong
et al., 2007b). As another example, GEONETCast is a global
network of satellite-based data dissemination systems providing environmental data to the world-wide user community.
Products include meteorological satellites (Meteosat, GOES,
and FengYun), and vegetation monitoring using SPOT-Vegetation data. This information is made available to many users,
with low cost receiving stations and open-source software
(Mannaerts et al., 2009). Another example is the Sentinel Asia
program that is an initiative supported by JAXA and the
APRSAF (Asia-Pacific Regional Space Agency Forum) to share
disaster information in the Asia-Pacific region using a Digital

Asia (Web-GIS) platform, and to make the best use of Earth
observation satellites data for disaster management in the
Asia-Pacific region (Sentinel Asia, 2010).
An important initiative that is focused on the provision of
space-based information for disaster response is the international charter Space and Major Disasters (Disaster Charter, 2010). A number of organizations and program are
involved in rapid mapping activities after major disasters including UNOSAT (2010), DLR-ZKI (2010), SERTIT (2010),
GDACS (2010), and the Dartmouth Flood Observatory (2010).
In Europe, the Global Monitoring for Environment and Security (GMES) initiative of the European Commission and the
European Space Agency (ESA) is actively supporting the use of
satellite technology in disaster management, with projects
such as PREVIEW (Prevention, Information and Early Warning
pre-operational services to support the management of risks),
LIMES (Land and Sea Integrated Monitoring for Environment
and Security), GMOSS (Global Monitoring for Security and
Stability), SAFER (Services and Applications For Emergency
Response), and G-MOSAIC (GMES services for Management
of Operations, Situation Awareness and Intelligence for regional Crises) (GMES, 2010). The United Nations Platform for
Space-based Information for Disaster Management and
Emergency Response (UN-SPIDER, 2010) has been established
by the UN to ensure that all countries have access to and
develop the capacity to use space-based information to support the disaster management cycle. They are working on a
space application matrix that will provide the satellite-based
approaches for each type of hazard and each phase of the
disaster management cycle. Overviews on the use of spacebased information in hazard inventory assessment can be
found in CEOS (2003), Tralli et al. (2005), IGOS (2007), and
Joyce et al. (2009).
For a number of hazards, satellite-based information is the
major source for generating hazard inventories and hazard
monitoring (e.g., tropical cyclones, forest fires, and drought).
For other hazards, such data support ground-based measurements (e.g., earthquakes, volcanic eruptions, and coastal hazards). Some hazard types cannot be recorded by a network of
measurement stations, as these do not have specific measurable
characteristics (e.g., landslides, forest fires, and snow avalanches). There are also many areas where recorded information is not available. Thus, the identification of hazardous
phenomena may require techniques such as automatic classification or expert visual interpretation of remotely-sensed data.
Automated classification methods make use of reflectance
variations in different parts of the electromagnetic spectrum,
and by data collected from active microwave and light detection and ranging (LiDAR) sensors. For flooding applications, Earth-observation satellites can be used for mapping
historical events and sequential inundation phases, including
duration, depth of inundation, and direction of current
(Smith, 1997). Geomorphological information can be obtained using optical (LANDSAT, SPOT, IRS, and ASTER) and
microwave (ERS, RADARSAT, ENVISAT, and PALSAR) data
(Marcus and Fonstad, 2008). The use of optical satellite data is
commonly hampered by the presence of clouds, and hazard
mapping is also hampered in areas with vegetation cover. The
use of synthetic-aperture radar (SAR) data are therefore better
275
for mapping hazard events such as floods (Schumann et al.,

2007).
Mapping of forest fires with satellite information is done
by mapping the fires themselves using thermal sensors (Giglio
and Kendall, 2001), or through the mapping of burnt areas
using MODIS or AVHRR, which have a high temporal resolution (Trigg et al., 2005), or with SAR (Bourgeau-Chavez and
Kasischke, 2002).
To facilitate visual interpretation of hazard phenomena
that cannot be automatically obtained from satellite images
(e.g., landslides) and for many geomorphological interpretations of hilly and mountainous areas, stereoscopic imagery
with a high to very-high resolution is required (Metternicht
et al., 2005). Very-high resolution imagery (QuickBird, IKONOS, WorldView, GeoEye, SPOT-5, Resourcesat, Cartosat,
Formosat, and ALOS-PRISM) have currently become the best
option for visual mapping from satellite imagery, and the
number of operational sensors with similar characteristics is
growing year by year as more countries are launching Earthobservation satellites with stereo capabilities and spatial
resolutions of 3 m or better. The high costs may still be a
limitation for obtaining high-resolution images for particular
study areas, especially for multiple dates after the occurrence
of main triggering events such as tropical storms or cyclones.
Automatic classification of landslides using digital airphotos
and very-high resolution satellite images has been applied
successfully by Hervas et al. (2003), Barlow et al. (2006), and
Martha et al. (2010).
Hazard inventory databases should contain information
for extended periods of time so that magnitude/frequency
relationships can be analysed. This requires the inclusion of
high-frequency/low-magnitude events for estimating hazards
with a high probability of occurrence, but should also contain
sufficient low-frequency/high-magnitude events to evaluate
the hazard for extreme events as well. Therefore, apart from
measuring, observing, and mapping recent hazard events, it is
very important to carry out extensive archive studies. For example, one of the most comprehensive projects for landslide
and flood inventory mapping has been the AVI project in Italy
(Guzzetti et al., 1994). Another example is from China where
an analysis was made on extreme precipitation events based
on datasets derived from Chinese historical documents over
eastern China for the past 1500 years (Zheng et al., 2006).
Hazard inventories can also be produced using participatory
mapping and participatory GIS (PGIS). PGIS involves communities in the production of spatial data and spatial decision-making. For example, local people can interpret the
outputs from a GIS or contribute by integrating participatory
mapping of hazardous events to modify or update information in a GIS database. Capturing local knowledge and
combining it with other spatial information is a central objective. This process may assist communities to look at their
environment and explore alternative scenarios based on
understanding of their own goals, constraints, and preferences
(McCall, 2003; Peters Guarin et al., 2005).
The data and techniques described above are intended to
support the generation of hazard inventory databases. Such
databases may have a very large degree of uncertainty, which
can be related to the incompleteness of historical information
with respect to the exact location, time of occurrence, and type
276
Table 5 Global data sources for inventory of hazardous events and hazard assessment used in the PREVIEW project
Hazard type
Historic events
Hazards
Cyclones
Cyclones storm surges
Droughts
UNEP/GRID-Europe, based on various raw data sources

UNEP/GRID-Europe, based on Cyclones winds data
UNEP/GRID-Europe based on Climate Research Unit (CRU)
precipitation data
United States Geological Survey (USGS) ShakeMap Atlas
UNEP/GRID-Europe
UNEP/GRID-Europe
International Research Institute for Climate
Prediction (IRI), Columbia University
UNEP/GRID-Europe, USGS, and GSHAP
(Global Seismic Hazard Assessment Project)
IONA Fire Atlas
Earthquakes
Fires
Floods
Tsunamis
Volcanoes
Landslides
European Space Agency (ESA-ESRIN) and World Fires Atlas Program

(ATSR).
Dartmouth Flood Observatory (DFO).
National Geophysical Data Center (NGDC) Tsunami database, NOAA
Smithsonian Institution Volcanoes of the world
Not available
UNEP/GRID-Europe
Norwegian Geotechnical Institute (NGI)
Hotspots project, International Centre for
Geohazards (ICG/NGI)
Source: Reproduced from UNEP/DEWA/GRID, 2010. PREVIEW Global Risk Data Platform. United Nations Environment Programme, Global Resource Information Database, Europe,
Geneva. http://www.grid.unep.ch/activities/earlywarning/preview/index.php
of hazard. Table 5 lists a number of sources for global-hazard

inventories that have been used in the PREVIEW project
(Peduzzi et al., 2009).
3.10.3.2.2
The environmental factors are a collection of GIS data layers

that are expected to have an effect on the occurrence of the
hazardous phenomena, and can be utilized as causal factors in
the prediction of future events (given numerous assumptions).
The list of environmental factors in Table 4 is not exhaustive,
and it is important to select those factors that are related to a
specific type of hazard in each particular environment. They
represent, however, an example of data types related to
topography, geology, soils, hydrology, geomorphology, and
land use. The basic data can be subdivided into those that are
more or less static, and those that are dynamic and need to be
updated regularly. Examples of static data sets are related to
geology, soil types, geomorphology, and topography. The time
frame for the updating of dynamic data may range from hours
to days (e.g., meteorological data and its effect on hydrology),
to months and years for land-cover and land-use data. Landuse information should be evaluated with care, as this is both
an environmental factor, which determines the occurrence of
new events (such as forest fires, landslides, and soil erosion),
as well as an element-at-risk, which may be affected by the
hazards. Table 4 provides an indication on the relevance of
these factors for hazard assessment for different types of hazards (Van Westen, 2009).
As topography is one of the major factors in most types of
hazard analysis, the generation of a DEM and geomorphometric analysis plays a critical role. Elevation data can be obtained through a variety of techniques, such as digitizing
contours from existing topographic maps, topographic levelling, EDM (electronic distance measurement), differential
Global Positioning System (GPS) measurements, digital
photogrammetry, Interferometric Synthetic-Aperture Radar
(InSAR), and LiDAR. Data source selection depends on a
variety of factors that include data availability, price, and application. Many topographic parameters can be produced
from DEMs using fairly simple GIS operations. For more
details regarding digital terrain modeling and geomorphometry, see these chapters in this volume.
The main sources for global DEMs used in hazard and risk
analysis are GTOPO30 (USGS, 1997; Hastings and Dunbar,
1998) and the Shuttle Radar Topographic Mission (SRTM; Farr
and Kobrick, 2000; CGIAR-CSI, 2008). The relatively low
vertical accuracy of SRTM data (Falorni et al., 2005) does not
make it suitable for large-scale hazard assessments. It is extensively used, however, for many small-scale applications in
areas where other sources of DEM are not available, such as in
tsunami-hazard assessment (Blumberg et al., 2005). ASTERderived DEMs are also commonly used in hazard assessments
for (parts of) countries in the absence of more detailed data
(Fujisada et al., 2005). For smaller areas, the best option is to
generate DEMs from very-high resolution images (e.g.,
Quickbird, IKONOS, ALOS PRISM, and Cartosat).
Microwave (InSAR) data can also be used for the generation of DEMs, but in practice it is mostly used for detecting
changes in topographic heights related to different hazardous
geological processes, such as land subsidence, slow-moving
landslides, tectonic motions, ice movement, and volcanic activity (Massonnet and Feigl, 1998; Ferretti et al., 2001; Hilley
et al., 2004; Salvi et al., 2004; Burgmann et al., 2006). Multitemporal InSAR analyses using techniques such as the permanent scatterers (PSInSAR; Ferretti et al., 2001) can be used
to measure displacement of permanent scatterers (e.g., buildings) with millimetre accuracy, and allow the reconstruction of
the deformation history (Farina et al., 2008).
For detailed measurement of displacements, networks of
differential global positioning system (DGPS) receivers at
fixed points are used extensively for mapping strain rates,
tectonic plate movements (Vigni et al., 2005), volcanic
movements (Bonforte and Puglisi, 2003), and landslides (Gili
et al., 2000).
More detailed DEMs derived using LiDAR are used extensively for geomorphologic mapping and terrain classification
(Asselen and Seijmonsbergen, 2006). Airborne LiDAR data
can be applied to glacial hazards (Favey et al., 2002), coastal
hazards (Miller et al., 2008), flood modeling (Cobby et al.,
2001; French, 2003), and landslide-hazard assessment
(Haugerud et al., 2003). Multi-temporal LiDAR can also be
used to model changes and quantify rates of active fluvial

processes such as river-bank erosion (Thoma et al., 2005).
Topographic information is used for heuristic hazard analysis at small scales (e.g., hillshading images, physiographic
classification, internal relief, and drainage density), regional
scales (e.g., altitude zones, slope gradient, slope direction,
contributing area, plan curvature, profile curvature, and slope
length), and in physically-based modeling at local scales (e.g.,
local drainage direction, flow path, and slope gradient), and
spread modeling (Moore et al., 2001). The use of slope-gradient
maps in hazard assessment is greatly affected by the resolution
of the DEM (Zhou and Liu, 2004). As a general rule of thumb,
the use of slope-gradient maps is not advisable for small-scale
studies (Van Westen et al., 2008), although some have used
1 km resolution DEMs to calculate slope-angle distribution
(Hong et al., 2007a). In larger-scale studies, slope maps and
other DEM derivatives can be used as input factors for heuristic
or statistical analysis. In local-scale hazard assessment, DEMs
are used in hydrology modeling and slope maps are used for
physics-based modeling (Kuriakose et al., 2009a).
Geological maps represent a standard information component in the hazard assessment of many hazard types (e.g.,
seismic, volcanic, landslides, and soil erosion). A geological
map of the world (CGMW) was developed in 2009 with maps
at the 1:5 million and 1:25 million scales. OneGeology is an
international initiative of the geological surveys of the world,
launched in 2007 as a contribution to the International Year of
Planet Earth, with the aim to create a web-based geological
map of the world (OneGeology, 2010). Digital geological
maps of chronostratigraphy, lithostratigraphy, faults, tectonic
lineaments, tectonic units, and other themes are available
online with scales ranging from 1:250 000 (for certain countries) to 1:50 million. For individual countries, geological information is generally digitally available at much larger scales.
For example, through the web-portal of the USGS, scanned
geological maps, as well as GIS data that can be downloaded
(USGS, 2010). The subdivision of geological formations into
meaningful units for hazard assessment is generally problematic at small scales of analysis. In detailed hazard studies,
specific engineering geological maps are collected and rock
types are characterized using field tests and laboratory measurements. For detailed analysis, 3-D geological maps have also
been used, although the amount of outcrop and borehole
information collected will make it difficult to use this information on a scale smaller than 1:5000 and its use is restricted
mostly to a site investigation level (e.g., Xie et al., 2003). Apart
from lithological information, structural information is very
important for hazard assessment (e.g., earthquakes, landslides,
and volcanic eruptions). At medium and large scales, attempts
have been made to generate maps indicating dip direction and
dip amount based on field measurements, but the success of
this depends very strongly on the amount of measurements
and the complexity of the geological structure (Gunther,
2003).
Soil information is also required for hazard assessment.
This includes soil types with associated geotechnical and
hydrological properties, and soil sequences with depth information. These data layers are essential components for any
physical numerical-modeling approach (e.g., earthquake
amplification studies, landslides, and soil erosion). Pedologic
277
soil maps usually portray soil classes based on the upper soil
horizons with rather complicated legends, and are relevant for
soil erosion, drought, and forest-fire hazard assessment. Engineering soil maps describe all loose materials on top of the
bedrock and portray classes according to geotechnical characteristics. They are based on outcrops, borehole information,
and geophysical studies. The soil depth is very difficult to map
over large areas, as it may significantly vary over a relatively
small area. Soil thickness can be modeled using an interpolation technique that incorporates factors such as land use
and slope (Kuriakose et al., 2009b). Digital soil information is
available worldwide (FAO, 1981), and includes soil-type
classification, clay mineralogy, soil depth, soil-moisture capacity, bulk density, soil compaction, and other variables. This
product is not based on satellite information directly, but
primarily on ground surveys and national databases.
Geomorphological maps are made at various scales to
show land units based on their shape, material, processes, and
genesis (e.g., Klimaszewski, 1982). No generally accepted legend exists for geomorphological maps, and there may be
large variations in the information content based on the experience of the geomorphologist. An applied geomorphological mapping working group has been formed as part of the
International Association of Geomomorphologists (IAG) to
set guidelines for geomorphological mapping and develop a
digital atlas of geomorphological maps. Detailed geomorphological maps contain a wealth of information, but
require extensive field mapping and are very difficult to convert into digital format (Gustavson et al., 2006). Unfortunately, traditional geomorphological mapping seems to have
nearly disappeared with the developments of digital geospatial
techniques, and relatively few publications on hazard and risk
still focus on it (Carton et al., 2005; Castellanos and Van
Westen, 2007), or replace it by using morphometric information. An important new field within geomorphology is the
quantitative analysis of the topography, called geomorphometry or digital terrain analysis, which combines elements of
the Earth sciences, engineering, mathematics, statistics, and
computer science (Pike, 2000; Dragut- and Blaschke, 2006).
Part of the work focuses on the segmentation of the topography into land-surface objects or geomorphological land
units based on morphometric characteristics at multiple scales
(Giles and Franklin, 1998; Miliaresis, 2001), and on the extraction of landform elements (Carrara et al., 1995). Digital
geomorphological maps are available for some parts of the
world including Germany (GMK, 2010), Austria, and New
Zealand (GNS, 2010).
Land cover can be considered as a static factor in some
hazard studies, although most types of hazard assessments
actually focus on the detection of land-cover changes in relation to hazard phenomena. Changes in land cover and land
use resulting from human activities, such as deforestation,
forest logging, road construction, fire, drought, and cultivation
on steep slopes, can have an important impact on hazards.
An example is the evaluation of the effect of logging and
deforestation on landslides (e.g., Furbish and Rice, 1983).
Land-use maps are made on a routine basis from medium
resolution satellite imagery such as LANDSAT, SPOT, ASTER,
and IRS1-D. Other sources for land-cover data with higher
temporal and lower spatial resolution include MODIS
278
(Moderate Resolution Imaging Spectroradiometer; Friedl et al.,

2002), MERIS (Medium Resolution Imaging Spectrometer),
NOAA-AVHRR, Global Imager (GLI), and SPOT-Vegetation
imagery with varying resolutions (250 m1 km). They are
used on a routine basis for monitoring the global distribution
of land-cover types (e.g., 10-day basis; Cihlar, 2000). Specific
change-detection algorithms that can be used for land-cover
studies are reviewed by Coppin et al. (2004). Seasonal and
inter-annual variations in land cover that may be caused by
natural disasters and land-use changes can be detected using
high temporal frequency satellite data.
Several initiatives have produced global land-cover maps
for different time periods. For example, the CORINE Land
Cover 2000 dataset (CLC2000) has been produced using remotely sensed imagery to produce a land-cover database at a
scale of 1:100 000, a positional accuracy of 150 m, and a
minimum mapping unit of 25 ha in Europe, and a resolution
of 1 km globally. A more recent map is the ESA Globcover
global land-cover map based on MERIS data acquired between
mid-2005 and mid-2006 (Arino et al., 2007). For individual
continents, more detailed land-cover information is available
(e.g., Africover, 2010).
Hazard and risk assessments require a multitude of data
from different sources. Therefore, it is important to have a
strategy for data availability. Important information concerns
include data quality, metadata, and multi-user databases.
Many project-specific data sets can be used for various purposes (e.g., resource management and risk assessment).
This requires that the potential users know what data exist and
have access. Spatial-risk information requires the use of
a spatial data infrastructure, where through the internet basic
GIS data can be shared among the different technical
and scientific organizations involved in hazard and risk assessment. A spatial data infrastructure is the foundation
or basic framework (e.g., a system or organizational scheme)
with policies, resources, and structures to make spatial information available to decision makers when they need it,
where they need it, and in a form where they can use it (almost) immediately. The website where the data are actually
exchanged is called a clearinghouse. A good example of that
is the European ORCHESTRA project (ORCHESTRA, 2009),
which designed and implemented the specifications for a
service-oriented spatial data infrastructure for improved
interoperability among risk-management authorities in
Europe. In the framework of the CAPRA project of the
World Bank (CAPRA, 2009), the GeoNode was developed
as an open-source platform that facilitates the creation, sharing, and collaborative use of geospatial data for risk assessment (GeoNode, 2010). Examples of initiatives that focus
on spatial-data infrastructure for disaster relief include
Reliefweb (2010), Alernet (2010), HEWSweb (2010), and
GDACS (2010).
3.10.3.3
Hazard Assessment Examples
As indicated in Table 3, hazard assessment is carried out at

various scales, and the methods for hazard assessment are
determined by the type of hazard and by the availability of
input data (Table 4). This section discusses the main
approaches for hazard assessment used at various geographic

scales.
3.10.3.3.1
Global hazard assessment
Hazardous events can encompass large areas such as windstorms, drought, earthquakes, and tsunamis. Therefore, global
or international hazard assessment is required. For example,
the Global Seismic Hazard Mapping Project (GSHAP, 1999), a
demonstration project of the UN/International Decade of
Natural Disaster Reduction, was conducted in the 19921998
period with the goal of improving global standards in seismichazard assessment. The GSHAP produced regional seismichazard maps for most parts of the world that display the
global seismic hazard as peak ground acceleration (PGA) with
a 10% chance of exceedance in 50 years, corresponding to a
return period of 475 years. The procedure involves the identification of seismo-tectonic zones in which earthquake characteristics were analyzed from historic earthquake databases.
For each point, seismic hazard is then analyzed using modules
(Arnold, 1989).
For windstorms, international databases exist for tropical
cyclones in different parts of the world. For the North Atlantic
region, for example, the HURDAT database (Jarvinen et al.,
1984) contains all historic Hurricane tracks. Windstorm-hazard models generate a set of stochastic events based on historical and modeled windstorm tracks, with parameters on
intensity, size, and shape. For each simulated track, data are
calculated for wind velocity together with associated levels of
storm surge, and rainfall intensities using empirical relations
(Mouton and Nordbeck, 2003). Areas that may inundate due
to tidal changes are mapped using a DEM in coastal zones
(Lavelle et al., 2003). Drought-hazard assessment at an international level is carried out using monthly average precipitation data (e.g., the Weighted Anomaly of Standardized
Precipitation (WASP)) developed by the International Research Institute for Climate and Society (IRI), computed on a
2.5 2.5 1 grid (Lyon and Barnston, 2005).
For other hazards, such as floods and landslides, information at international levels is too general for the estimation
of hazards, as the hazard events are too localized and require
more detailed information. Nadim et al. (2006, 2009) made an
attempt to generate a global landslide-hazard map, making use
of general spatial data sets with global coverage, such as an
SRTM-derived DEM with 1 km spatial resolution, the geological
map of the world at 1:25 million scale, a soil-moisture index,
monthly precipitation data, and the Global Seismic Hazard
Mapping Programme (GSHAP) results. Given the poor resolution of the data as compared to the specific conditions in
which landslides occur, however, the results are only a general
indication of landslide susceptibility. Hong et al. (2007a) presented a qualitative method for a global landslide-susceptibility
map using GIS-based map overlay techniques, combining several layers of different parameters (e.g., elevation, slope, land
use, etc.). Recently, an attempt to provide global-scale landslide
early warnings in near real time using stochastic models combining a global landslide database, TMPA rainfall estimates,
SRTM DEM, and MODIS land-cover products was conducted at
Columbia University, the success of which was mainly limited
by the lack of completeness of the landslide database and the
quality of the rainfall estimates from Tropical Rainfall
Measuring Mission and Multi-satellite Precipitation Analysis

(TMPA) (Kirschbaum et al., 2009). Global flood-hazard studies
are difficult to carry out, as the DEMs available at global scale
are generally not of sufficient detail for flood modeling applications. One example of an approach used for flood-hazard
mapping over very large areas is based on an inventory of past
flood events (e.g., from Dartmouth Flood Observatory), coupled with a very simple flood model based on the HYDRO1k
Elevation Derivative Database (USGS, 1996; Verdin and
Greenlee, 1996). HYDRO1k is a geographic database developed
to provide comprehensive and consistent global coverage of
topographically derived data sets, including streams, drainage
basins, and ancillary layers derived from the USGS 30 arcsecond DEM of the world.
At the global scale, few approaches have been carried out
for multi-hazard assessment that attempt to provide general
indicators or risk indices for countries, or for parts of countries, mainly for comparison of risk levels between countries.
Dilley et al. (2005) developed a methodology for global
hazard and risk assessment for the main hazard types in
Table 3. Peduzzi et al. (2009) presented a model designed for
the United Nations Development Programme as a component
of the Disaster Risk Index (DRI), which aims at monitoring
the evolution of risk. Four hazards (droughts, floods, cyclones,
and earthquakes) were modeled using a GIS based on the
datasets shown in Table 5.
3.10.3.3.2
(Inter)National hazard assessment
Hazard assessment is commonly carried out for individual

continents or countries, as they are related to the same
administrative area and controlled by national or international governments (e.g., EU, USA, and China). The methodology for hazard and risk assessment is standardized, and
mostly follows established guidelines that are requested by
governments (e.g., the European Floods Directive). The applications at the (inter)national level are more refined
than those carried out globally, and require higher-resolution
data. For example, the European Flood Directive (EFD) indicated that preliminary flood-risk assessments in Europe
should be completed by 2011, flood hazard and risk maps
should be available by 2013, and flood-management
plans should be completed by 2015 (EFD, 2007). In order to
accomplish these goals, standardized methods, datasets, and
GIS-based tools are to be used for the assessment and monitoring of flood risk for the whole of Europe. Flood-hazard
maps are generated based on DEMs with a resolution ranging
between 100 m and 1 km. The hazard factor is estimated by
using hydrological modeling (e.g., LISFLOOD) at different
scales and for many return periods (Barredo, 2007; van der
Knijff et al., 2010). Modeling of extreme precipitation and
resulting river discharge is calculated in real time, and flood
forecasts are made for the whole of Europe. In the USA, the
Federal Emergency Management Agency (FEMA) has established a national flood-hazard mapping project with the
Federal Insurance and Mitigation Administrations Hazard
Mapping Division, through their national flood insurance
program (FEMA, 2010).
Similar initiatives in Europe are in the field of forest fires.
The European Forest Fire Information System (EFFIS) makes a
rapid assessment of the burned areas through a series of daily
279
images from the MODIS instruments on board the TERRA and

AQUA satellites, and displays fires with burned area of approximately 40 ha or larger from a web-GIS (Ayanza et al.,
2003). A third example that is implemented at both the
European level as well as globally is the MARSOP-3 project on
crop yield forecasting, carried out by the Joint Research Centre
(JRC) of the EC, with other partners. This system includes the
management of a meteorological database, an agro-meteorological model and database, low-resolution satellite information, statistical analyses of data, crop-yield forecasting, and
publishing of bulletins containing analysis, and forecasts and
thematic maps on crop-yield expectations using a Web-GIS
application (Reidsma et al., 2009). An overview on the use of
satellite data for drought monitoring and hazard assessment
can be found in Henricksen and Durkin (1986), Peters et al.
(2002), and White and Walcott (2009). The aforementioned
software tools are used for early warning as well as for hazard
assessment for the whole of Europe.
In the USA, a GIS-based tool for earthquake hazard assessment, ShakeMaps, was developed by the USGS in cooperation
with regional seismic-network operators. ShakeMaps provides
near-real-time maps of ground motion and shaking intensity
after important earthquakes. It can also be used to generate
hazard maps using scenario earthquakes (Wald et al., 1999).
Later, a methodology was developed for modeling of seismic
site conditions using topographic slope as a proxy, using the
SRTM30 database, which provided the average shear-velocity
down to 30 m (Wald et al., 2004). Initiatives to incorporate
open-source software in seismic-hazard assessment have been
taken by OpenSHA (2010) and by the Global Earthquake
Model (GEM, 2010), an international initiative to develop
uniform and open standards and platforms for calculating
earthquake risk worldwide. The GEM brings together all major
players in the earthquake risk assessment field, including partners from the insurance sector, international organizations,
public organizations, and research centers from all over
the world.
In terms of landslide-hazard assessment, this scale is still
too general to be able to map individual landslide phenomena. The analysis of landslide hazards at this scale is still done
by weighting a number of input maps (e.g., Malet et al., 2009;
Castellanos and Van Westen, 2007).
3.10.3.3.3
Provincial and municipal level
At local and municipal scales, spatial information is generally

of sufficient quality to run more sophisticated models, which
can be either empirical (e.g., statistical) or deterministic
(physically-based). A flood-hazard assessment example follows. The first step is to transform catchment characteristics
like topography, relief, and land cover, complemented with
hydrological boundary conditions, into estimates of the discharge at various locations along the river downstream. This
can be done with (distributed) 1-dimensional models. These
kinds of models are very useful to assess the response of the
river to extreme events and to changes in the topography and
land cover. Typical models to do this are HEC-HMS and
HEC-RAS of the US Army Corps of Engineers, MIKE-SHE
(Refsgaard and Storm, 1995), IHDM (Beven et al., 1987),
LISFLOOD (De Roo et al., 2000), and HEC-RAS (Brunner,
2002). They require the characterization of the terrain through
280
a series of cross-sections perpendicular to the direction of flow

for which the average water depth and flow-velocity are calculated. This type of modeling is typically applied for catchment analysis, and the underlying assumption is that all flow
is parallel to a predefined river-network. In near-flat terrain
with complex topography, it cannot be assumed that all flow
will be parallel to the main river. Also in urban environments
and in areas with a dominant presence of man-made structures, models are required that calculate flow in both X- and
Y-directions. Such models like SOBEK (Stelling et al., 1998;
Hesselink et al., 2003), Telemac 2D (Hervouet and Van Haren,
1996), and MIKE21 can also be applied in the case of
diverging flow at a dike breach. They require high-quality
DEMs, which ideally are generated using LiDAR data (Dal Cin
et al., 2005; Alkema and Middelkoop, 2005). The flood
modeling is generally carried out at a municipal to provincial
scale for a selected stretch of the river. These models provide
information on how fast the water will flow and how it
propagates through the area. They are very suitable for assessing the effects of the surface topography, like embanked
roads and different land-cover types on the flood behavior
(Stelling et al., 1998).
Furthermore, for landslide-hazard assessment, the provincial
and municipal scales offer many more possibilities, as sufficient
information can be collected on hazard inventories and
the factors that control the location of landslides (Dai et al.,
2002). Dai et al. (2002) differentiated between statistical
methods and physically based models. Guzetti et al. (2005)
provided an overview of the various statistical methods that can
be applied, focusing on the use of multivariate statistical
methods in which landslide inventories for different periods are
used in combination with environmental factors for predicting
landslide activity within slope units that are defined from a
DEM. Van Asch et al. (2007) provided an overview of the
physically based modeling approaches. Most of the physically
based landslide models make use of the infinite-slope model,
and are therefore only applicable to modeling shallow landslides. They can be subdivided into static models that do not
include a time component, and dynamic models, which use the
output of one time step as input for the next time step. Physically-based models for shallow landslides account for the
transient groundwater response of the slopes to rainfall, and the
effect of earthquake acceleration (Van Beek and van Asch,
2004).
The provincial and municipal scales are also the most appropriate for volcanic hazard assessment, as a lot of this work
depends on the determination of the eruptive history on the
basis of geological investigation and age dating (Tilling,
1989). Given different volcanic eruption scenarios, several
modeling techniques can be carried out for the various volcanic hazards (ash fall, lava flow, pyroclastic flow, and lahars).
Most of these hazard-assessment methods require some sort of
spread modeling, where the volcanic products are distributed
over the terrain away from the vent. This requires the use of
dynamic models (Zuccaro et al., 2008). The evaluation of
volcanic hazards from tephra fallout is determined by utilizing
volcanic ash volumes, eruption height, and wind information
(Connor et al., 2001). Remote sensing also plays an important
role in volcanic-hazard assessment (e.g., Kerle and Oppenheimer, 2002).
3.10.3.3.4
Community level
Approaches based on local knowledge and experiences may be

useful in developing countries, where detailed information
required for conventional model-based risk analyses facilitated by GIS is generally not available. For instance, historical
records on river discharges and rainfall are commonly missing, whereas knowledge about hazardous events is generally
available within the local communities (Ferrier and Haque,
2003). A vast quantity exists of undocumented local
knowledge on disaster occurrences in the field, which usually
remains untapped because of the lack of funding, a format to
systematically collect it, and a low commitment to do so
(Hordijk and Baud, 2006). Anderson and Woodrow (1998)
stated that much of the information needed for risk assessment and mitigation can be obtained from local people who
generally already know what the situation is, but do not always have the skills for understanding and organizing what
they know. Several organizations, such as the International
Federation of Red Cross and Red Crescent Societies (IFRC),
have developed community-based assessment instruments for
analyzing disaster situations at the grassroots level, and for
improving the communitys expertize in identifying and articulating its needs and reducing its vulnerabilities. Some examples of these community-based methods are named
Capacity and Vulnerability Assessment (CVA), Hazards,
Vulnerability and Capacity Assessment (HVCA), and Damage, Needs and Capacity Assessment methods (DNCA)
(Provention Consortium, 2010). These methods aim at eliciting tacit local knowledge within communities on historic
disaster events, the perception of hazards, characterization of
elements-at-risk, identifying the main factors of vulnerability,
coping mechanisms, and disaster reduction scenarios. The
application of such collaborative approaches is not common
in many developing countries, and decision-making about risk
is commonly done in a top-down approach by local authorities where specialists diagnose problems, formulate alternatives, and determine options without a meaningful
consultation with communities (UN-ESCAP, 2003). Hazard
specialists commonly consider that community participation
is difficult to achieve, and the information perceived is
unscientific, not always easy to retrieve, difficult to be expressed in quantitative terms, or needs to be converted into
spatial formats (Peters Guarin, 2008).
The integration of geo-information systems and localcommunity knowledge relevant to hazards, vulnerability, and
risk modeling is still in an initial stage (Maskrey, 1998; Ferrier
and Haque, 2003; Zerger and Smith, 2003). Very often the
sketches, paper maps, historical profiles, and other results
obtained through participatory mapping are not kept after a
risk project has finished, leading to a loss of valuable information. As Cannon et al. (2003) advised, these products need
to be converted from raw data into useful spatial information
that allows the community and other participants to develop
analytical processes for risk analysis and exploration of management alternatives. Several authors have shown that local
communities are indeed the primary sources of information
for flood depths, time of occurrence, severity measured in
terms of damage, and the like (Whitehouse, 2001; AlcantaraAyala, 2004; Rautela, 2005). Systematic collection of data from
significant events using public participation can provide a very
useful component for the development of data-sets that can be

used as input for risk studies at the community level, and as a
basis for risk management and community planning (Ireland,
2001). Information from local communities can also be useful
in calibrating and verifying risk and disaster scenarios (Bassole
et al., 2001; Peters Guarin, 2008).
3.10.4
Elements-At-Risk and Vulnerability
The next step in risk assessment, after analyzing the hazard, is

to evaluate the elements-at-risk. There are many different types
of elements-at-risk and they can be classified in various ways.
In this section, several types of elements-at-risk and their data
sources are evaluated, followed by a discussion on how these
are used in vulnerability assessment.
3.10.4.1
Elements-At-Risk Information
Elements-at-risk inventories can be carried out at various levels

depending on the requirement of the study. Table 6 provides
a more detailed description. Elements-at-risk data should
be collected for basic spatial units, which may be grid-cells
on a global scale (see Tables 3 and 6), administrative units
(countries, provinces, municipalities, neighbourhoods, and
census tracts), or so-called homogeneous units with similar
characteristics in terms of type and density of elements-at-risk.
Risk can also be analyzed for linear features (e.g., transportation lines) and specific sites (e.g., a dam site). The risk
assessment will be done for these spatial units of the
elements-at-risk, rather than those used in the hazard assessment. In the HAZUS methodology (FEMA, 2004), the loss
estimation is done based on census tracts.
Digital information on coastlines, international boundaries, cities, airports, elevations, roads, railroads, water features, and cultural landmarks are available from different
sources such as the Geonetwork established by FAO (2010),
with available data comprising base layers (e.g., boundaries,
roads, and rivers), thematic layers (e.g., protected areas), or a
backdrop image (e.g., World Forest, 2000).
One of the most important spatial attributes of the mapping units for an elements-at-risk inventory is land use. The
land use determines to a large extent the type of buildings that
can be expected in the unit, the economic activities that are
carried out, and the density of the population for different
time periods during the day. Land-cover and land-use maps
are prepared by image classification at small scales or through
visual interpretation at larger scales. Ebert et al. (2009) developed a method using object-oriented image classification
methods for the automatic characterization of land-use types
in urban areas.
3.10.4.1.1
Collaborative mapping and Mobile-GIS
Elements-at-risk information is collected from a wide variety

of sources. There are also many areas in the world for which
no detailed digital data are available on elements-at-risk. In
such situations, data should be digitized from analog maps, or
in case these also do not exist, be mapped in the field, for
instance, using mobile GIS. With the use of mobile GIS, it is
281
possible to directly collect the spatial information based on a

high-resolution image that can be uploaded into a palmtop
computer or smart phone, and link it with attribute information that is collected in the field. Some of the most used
tools for mobile GIS in urban elements-at-risk mapping are
ArcPad (Montoya, 2003) and Cybertracker (McCall, 2008).
Several initiatives have been established for collaborative
mapping of topographic features, also referred to as crowdsourcing. For example, OpenStreetMap is a free editable map
of the whole world, which is made using collaborative mapping by volunteers. It allows users to collect, view, edit, and
use geographical data in a collaborative way from anywhere
on Earth (OpenStreetMap, 2010). Another crowdsourcing example is Ushahidi (Ushahidi, 2010), which means testimony in Swahili, which was initially developed to map
reports of violence in Kenya in 2008, and which has been used
later in many disaster events to rapidly collect and visualize
spatial information. Other applications that are specifically
directed to post-disaster relief coordination are Sahana
(Sahana, 2010) and Virtual Disaster Viewer (Virtual Disaster
Viewer, 2009). Sahana is a free web-based disaster management system developed after the Indian Ocean tsunami as a
collaboration tool that addresses the common coordination
problems during a disaster. The Virtual Disaster Viewer is a
crowd-sourcing tool for collaborative disaster impact and
damage assessment, which has proven to be effective after the
Haiti earthquake in 2010. Hundreds of earthquake and remote-sensing experts were assigned specific areas (tiles) of the
affected areas to review and provide their assessment by
comparing before and after high-resolution satellite images,
which became available on Google EarthTM immediately after
the disaster, and which served as the basis for the collaborative
mapping. Such collaborative-mapping applications might
become a very important tool in the future.
3.10.4.1.2
Population data
People are the most important elements-at-risk, and population data exhibit a static and dynamic component. The static
component relates to the number of inhabitants per mapping
unit and their characteristics, whereas the dynamic component refers to their activity patterns and their distribution in
space and time. Population distribution can be expressed as
either the absolute number of people per mapping unit or as
population density. The way population data are collected and
represented in a risk assessment depends on the scale of analysis (see Table 3) and the availability of information (Rhind,
1991).
Census data are the obvious source for demographic data.
They are used as benchmark data for studying population
changes, and are key input for making projections concerning
population, households, labour force, and employment.
Census data are costly to collect, and updating of population
information is carried out on average every 10 years. Census
data are aggregated to census tracts, and normally data at an
individual household level are confidential. This is also the
reason why risk assessment is normally carried out at the
census tract level (FEMA, 2004). Census tracts are divisions of
land that are designed to contain 25008000 inhabitants with
relatively homogeneous population characteristics, economic
status, and living conditions. Census data may also contain
Table 6 Main elements-at-risk and how they can be spatially represented at various mapping scales
282
Scale
Continental
National
Provincial/Municipal
Community
Basic unit
Population
1 km grid or countries
Gridded population map
901 km grid and countries

Gridded population map
3090 m and municipality

By municipality
Population density
Census tract
By Census tract
Population density
Daytime/Night-time
Groups of buildings
People per building
Daytime/Night-time
Gender
Age
Education, etc.
Buildings
N.A.
Gridded building density map
Transportation
networks
N.A.
Main roads, railroads, harbours,

and airports
Lifelines
N.A.
Main powerlines
Essential
facilities
N.A.
Agricultural data
Gridded main land cover

types, crops
Ecological data
Economic data
Main land cover types,

crops
GDP
By municipality
By Census tract
Building footprints
Number of Buildings
Generalized use
Height
Building types
Detailed use
Height
Building types
Construction type
Quality/Age
Foundation
Road & railway networks, with

general traffic density
information
All transportation networks with

detailed classification, including
viaducts etc. & traffic data
All transportation networks with

detailed engineering works &
detailed dynamic traffic data
Only main networks
Detailed networks:
Water supply
Electricity
Water supply
Waste water
Electricity
Communication
Gas
Detailed networks and related

facilities:
Water supply
Waste water
Electricity
Communication
Gas
By Municipality
As points
Individual building footprints
Individual building footprints
Number of essential facilities
General characterization
Buildings as groups
Normal characterization
Buildings as groups
Detailed characterization
Each building separately
Gridded maps:
By homogeneous unit
By cadastral parcel
Crop types
Yield information
Crop types
Yield information
Crop types
Crop rotation
Yield information
Agricultural buildings
By cadastral parcel, for a given period

of the year
Crop types
Crop rotation and time
Yield information
Natural protected areas with

international approval
By region: Economic production,
import/export, type of economic
activities
Natural protected area with

national relevance
By Municipality
Economic production
Import/export
Type of economic activities
General flora and fauna data per

cadastral parcel.
By Mapping unit
Employment rate
Socio-economic level
Main income types plus larger
scale data
Detailed flora and fauna data per

cadastral parcel
By household
Employment
Income
Type of business plus larger
scale data
Source: Adapted from Van Westen, C.J., Castellanos Abella, E.A., Sekhar, L.K., 2008. Spatial data for landslide susceptibility, hazards and vulnerability assessment: an overview. Engineering Geology 102(34), 112131.
Global
283
other relevant characteristics that are used in risk assessment,

such as information on age, gender, income, education, and
migration.
For larger areas, census data may be aggregated into larger
administrative units. For large parts of the world, however,
census data are not available, out dated, or unreliable.
Therefore, other approaches have been used to model population distribution with remote sensing and GIS, based on
a number of factors, such as land cover, roads, slopes, and
night-time illumination. The use of remote-sensing data in
combination with other data to redistribute population information over smaller areas based on general population
data for large administrative units is also referred to as
dasymetric mapping (Balk et al., 2006). Global population
data are available from the LandScan Global Population
Database (Bhaduri et al., 2007; LandScan, 2010) that provide
the average population over 24 h in a 1 km resolution grid.
The Global Rural-Urban Mapping Project (GRUMP) is another example of modeling human populations in a common
geo-referenced framework (GRUMP, 2004), as is the African
Population Database (APD, 2010). Higher-resolution population databases have also been developed for specific areas,
especially in low-income countries where limited information
is available and there is a need to generate population information using satellite data. Tatem et al. (2007) made a comparison between semi-automated population distribution
mapping for several countries in East Africa, based on 30 m
Landsat ETM data, and concluded that these produced more
accurate results than existing products at a cost of $0.01 per km2.
For risk assessment at municipal or community level,
population is required at a high spatial resolution for every
census tract or even for each building. In the absence of census
data, static-population information can be derived directly
using high-resolution satellite imagery (e.g., Harvey, 2002) or
through a building-footprint map, where the land-use type
and the floor-space are used to estimate the number of people
present in a particular building (Chen et al., 2004; Lwin and
Murayama, 2009).
are important for evaluating the damaging effects, such as

structural type, construction materials, application of building
code, age, maintenance, roof type, height, floor space, volume,
shape, proximity to other buildings, proximity to hazard
source, proximity to vegetation, and openings (FEMA, 2004;
Jones et al., 2005; Grunthal et al., 2006; Douglas, 2007).
For risk maps that express losses in economic terms, an
estimation of building costs is also required. Several sources of
information can be used, such as data on house prices from
real-estate agencies, information from cadastres that indicate
the value used as the basis for taxation, engineering societies
that calculate the replacement costs, or insurance companies
(Grunthal et al., 2006). It is often difficult to get hold of the
building values used by the cadastres, whereas it is easier to
use the values from real-estate agencies. Samples are taken
from each type of building in the various land-use classes. In
some countries, building societies produce a monthly index
that permits an update of property prices. Cost estimation can
be based on using the replacement or market value. Apart
from building costs, content costs are also very relevant, especially for those hazards that have less structural damage
such as flooding.
Building information can be obtained in several ways.
Idealy, data are available on the number and types of buildings
per mapping unit, or even in the form of building-footprint
maps. If such data are not available, building-footprint maps
can be generated using screen digitizing from high-resolution
images (Van Westen et al., 2002). Automated building mapping
has also been carried out using high-resolution satellite images
(Fraser et al., 2002), InSAR (Stilla et al., 2003), and specifically
using LiDAR (Priestnall et al., 2000; Brenner, 2005; Oude
Elberink and Vosselman, 2009). LiDAR data also allow the
extraction of other relevant features and the calculation of
shapes, building height, and volumes that are needed in risk
assessment.
3.10.4.1.3
Vulnerability is the most complicated component of risk

assessment, as indicated in Figure 5, because the concept
of vulnerability has a wide range of interpretations. The
concept originated from the social sciences in response to
the pure hazard-oriented perception of disaster risk in the
1970s. Since that time, different disciplines have developed
their own concepts. Multiple definitions and different conceptual frameworks of vulnerability exist (e.g., Blaikie et al.,
1994; Pelling, 2003). For more details, see the overview
by Birkmann (2006). The definition of vulnerability (used in
Table 1) indicates that vulnerability can be viewed as multidimensional (physical, social, economic, environmental, institutional, and human factors define vulnerability), dynamic
(it changes over time), scale-dependent (it can be expressed at
different scales from individuals to countries), and site-specific
(each location might need its own approach) (Bankoff
et al., 2003). Risk-assessment methods can be differentiated
based on quantitative and qualitative approaches. Figure 8
presents a framework for multi-hazard risk assessment that
will be subsequently explained. It is important to note that
quantitative methods focus mostly on physical vulnerability,
Building data
After population, buildings are the second most important

group of elements-at-risk. They house the population, and the
behaviour of a building under a hazard event determines
whether the people in the building might be injured or killed.
In order to assess the potential losses and degree of damage of
buildings, it is important to analyze the type of negative effects
that the event might have on the building exposed to it, and
the characteristics of the building. The negative effects of
hazardous events on buildings can be classified into a number
of groups depending on the type of hazard (Blong, 2003;
Hollenstein, 2005). Figure 7 depicts a schematic overview of
the various hazard processes that may occur and that have a
different effect on buildings. For instance, a building may
be impacted by a mass, and the damaging effects would be
determined by the volume of the mass, speed of impact, and
the medium, such as rocks, soil, debris, snow, water, or air.
Buildings are also affected by undercutting (erosion or
landslides), shaking (earthquakes), inundation, fires, loss-ofsupport (subsidence), gasses, or loading (e.g., volcanic ashes).
In each of these situations, particular building characteristics
3.10.4.2
Vulnerability
Rock fall
Snow
avalanche
Landslide
Airplane crash
Lava flow
Pyroclastic flow
Tornado
Cyclone
Explosion
Erosion
Landslide
Earth quake
Flooding
Tsunami
Volcanic lahar
Fire
Bush fire
Chemical accident
Liquefaction
Landslide
Ash fall
Covering
Loss of
support
Fire
Gasses
Inundation
Shaking
Under
cutting
Wind
impact
Impact by mass
284
Figure 7 Examples of the types of hazardous processes to which buildings can be exposed. Each type of processes will have different effects.
whereas qualitative methods also incorporate the other definitional aspects.

Physical vulnerability is the potential for physical impact
on the built environment and population. It is defined as the
degree of loss to a given element-at-risk or set of elements-atrisk resulting from the occurrence of a natural phenomenon of
a given magnitude, and expressed on a scale from 0 (no
damage) to 1 (total damage). As can be seen from Figures 3
and 7, vulnerability is related to the characteristics of the
elements-at-risk and to the hazard intensity. Physical vulnerability as such is therefore not a spatial component, but is
determined by the spatial overlay of exposed elements-at-risk

and hazard footprints (Van Westen et al., 2009). Economic
vulnerability is defined as the potential impact of hazards on
economic assets and processes (i.e., business interruption,
secondary effects such as increased poverty and job loss). Social vulnerability is the potential impact of events on groups
within the society (such as the poor, single parent households,
pregnant or lactating women, the handicapped, children, and
elderly), and it considers public awareness of risk, ability of
groups to self-cope with catastrophes, and the status of institutional structures designed to help them cope. Environmental
Geology
Triggering factors
Land cover (t 0)
Geomorphology
Soils
Topography
Landslides
Roads
Land use
Floods
Windstorms
Hydrology (t 0)
Elements at risk
Buildings
Earthquakes
Land cover (t n)
Etc.
Hazard inventory
Rainfall
Soil erosion
Essential facilities
Forest fires
Cadastral data
Census data
Hydrology (t n)
(a)
Initiation analysis
(Inventory, heuristic,
statistical, physical
modeling)
Spatial probability
Event-based
hazard inventories
Temporal probability
Magnitude - loss
relationships
Magnitude probability
Run-out analysis
(Empirical, analytical,
numerical models)
(b)
Hazard assessment
Pt, Ps and Pm for different return
periods of triggering events and
hazard types
(c)
Hazard Vulnerability
Loss (t
Total risk
Loss (t
(d)
Amount
ype)/tim
Specific
risk
Vulnerability curve
Vulnerability matrix
V=1
ype)/tim
All hazard types

All hazard intensities
All return periods
All triggering events
All elements at risk
(e)
Quantitative risk assessment
Qualitative risk assessment
Total risk
Max
Min
Probability
Probability
Combining specific risk curves
Loss
Hazard index
F-N
curves
Risk index
Spatial multi criteria evaluation
Number (N) of fatalities
(f)
Vulnerability index
(g)
Cost-benefit analysis
Early warning systems
Spatial planning
Mitigation measures
Preparedness planning
Environmental
impact assessment
(h)
Figure 8 Framework of the use of GIS for multi-hazard risk assessment. Based on Van Westen, C.J., Van Asch, T.W.J., Soeters, R., 2005.
Landslide hazard and risk zonation; why is it still so difficult? Bulletin of Engineering Geology and the Environment 650(2), 167184, and Van
Westen, C.J., Castellanos Abella, E.A., Sekhar, L.K., 2008. Spatial data for landslide susceptibility, hazards and vulnerability assessment: an
overview. Engineering Geology 102(3-4), 112131.
285
286
vulnerability evaluates the potential impacts of events on the

environment (flora, fauna, ecosystems, and biodiversity)
(Birkmann, 2006).
Vulnerability can be expressed or presented in various ways
(Calvi et al., 2006). Vulnerability indices are based on indicators of vulnerability and are mostly used for holistic
vulnerability and capacity and resilience assessment. Vulnerability tables show the relation between hazard intensity and
degree of damage in the form of a table. Vulnerability curves
display the relation between hazard intensity and degree of
damage for a group of elements-at-risk (e.g., a certain building
type) ranging from 0 to 1. Different types of elements-at-risk
will show different levels of damage given the same intensity
of hazard (see Figure 3). Vulnerability curves can be relative
(showing the percentage of property value damaged) or absolute (showing the absolute amount of damage). Fragility
curves provide the probability for a particular group of
element-at-risk to be in or exceeding a certain damage state
(e.g., complete destruction, extensive damage, moderate
damage, and slight damage) under a given hazard intensity
(FEMA, 2004). A damage probability matrix (DPM) indicates
the probability that a given structural typology will be in a
given damage state for a given intensity.
Measuring physical vulnerability is a complicated process,
and can be done using either empirical or analytical methods
(Lang, 2002). Empirical methods are either based on damage
data from historical hazard events, or on expert opinion. For
events that are relatively frequent and widespread, it is possible to collect information on the degree of physical damage
to buildings or infrastructure after the event has occurred (e.g.,
Reese et al., 2007). This method is particularly suited for
flooding and for earthquakes, which normally affect many
buildings that are of the same type, and allow generating large
enough samples in order to make a correlation between the
hazard intensity (e.g., modified Mercalli intensity, ground acceleration, water depth, etc.) and the degree of damage. The
result is either a DPM or a vulnerability curve. In many situations, expert opinion will be the most feasible option for
obtaining vulnerability information, either because there is no
prior damage information and not enough funding to apply
analytical methods, or because building classifications used
elsewhere do not reflect the local building stock (Douglas,
2007). This method involves the consultation of a group of
experts on vulnerability to give their opinion (e.g., on the
percentage damage they expect for the different structural
types with different intensities of hazard).
Analytical methods are used to study the behavior of
buildings and structures based on engineering design criteria,
analyzing the seismic load to derive the likelihood of failure,
using physical modeling tests (e.g., shake tables or wind tunnels), as well as computer simulation techniques. Analytical
methods are able to model the relation between the intensity
of the hazard and the level of damage of objects. They require,
however, detailed numerical input data. For instance, in the
case of earthquake vulnerability analysis of buildings, it is
important to have geotechnical reports to establish the value
of the effective peak acceleration coefficient, the value of the
effective peak-velocity-related acceleration coefficient, and the
soil-profile type. Spectral acceleration should also be obtained.
One of the common tests is using a shake table. This is a
device for shaking structural models or building components

with a wide range of simulated ground motions, including
reproductions of recorded earthquake time-histories (Calvi
et al., 2006).
Most of the work on the measurement of physical vulnerability is done for earthquakes, floods, and windstorms
(FEMA, 2004). Even though flood vulnerability has been defined in a rather detailed manner (Moel et al., 2009), there are
still many uncertainties involved. For volcanic hazards, much
progress in defining vulnerability has been made in recent
years (Spence et al., 2004, 2005). For mass movement,
less work has been done on defining vulnerability (Glade,
2003), partly due to the large variation in mass-movement
processes, the difficulty in expressing landslide intensity versus
the degree of damage, and limited amount of landslidedamage data. Some approaches exist for single landslide
types such as debris flows (e.g., Fuchs et al., 2007), but an
integrated methodology is still lacking. Hollenstein (2005)
developed an approach for multi-hazard vulnerability assessment by definition of hazards using a common set of parameters (e.g., acceleration, pressure, and temperature change)
and fragility functions, so that they were applicable to
all risks.
Population vulnerability can be subdivided into direct
physical population vulnerability (injury, casualties, and
homelessness) and indirect social vulnerability and capacity.
Physical population vulnerability is mostly carried out after a
building-vulnerability study by analyzing the effect of the
building damage on the population inside, using different
injury severity classes. Empirical relations exist for different
types of hazards, although most information is available for
earthquakes (Coburn and Spence, 2002; FEMA, 2004). For
example, such relations are available for volcanic hazards
(Spence et al., 2005), landslides (Glade et al., 2005), drought
(Wilhite, 2000), flooding, and windstorms (FEMA, 2004).
The methods described above aim at quantifying physical
vulnerability to natural hazards, and mostly follow an engineering approach that is restricted to quantifying the physical effects of disasters on buildings and other infrastructure,
and secondary effects of these related to casualties and economic losses. Other approaches also exist that look at vulnerability in a holistic way and try to incorporate all the
components of vulnerability using an indicator approach
(qualitative).
3.10.5
Multi-Hazard Risk Assessment
A framework of multi-hazard risk assessment is presented in

Figure 8 that identifies various components. The first component (a) represents the input data, which can be subdivided
in datasets required for generating susceptibility maps, triggering factors, multi-temporal inventories, and elementsat-risk. The second component (b) focuses on susceptibility
assessment, and is divided into two parts. The first one dealing
with the modeling of areas where the hazard may initiate (e.g.,
earthquakes, landslide initiation, hydrological modelling, soil
erosion, and volcanic eruptions), which can make use of a
variety of different methods (inventory based, heuristic, statistical, and physically-based models). The resulting maps form
the input as source areas in the modeling of potential

spreading of the phenomena (e.g., spreading of volcanic
deposits, landslide run-out, flood extent modelling, seismic
amplification, and forest-fire spreading).
The third component (c) deals with hazard assessment,
which heavily depends on the availability of magnitudefrequency information. The susceptibility maps together with
the magnitude-frequency relations of the triggering events
are used to determine three components that are needed for
the hazard assessment: 1) the spatial probability (indicating
the probability that a given area will be affected by the hazard
of a given intensity); 2) the temporal probability (indicating
the probability of the event to happen in time); and 3) the
magnitude probability (indicating the probability that the
hazard event will have a given magnitude) (Corominas and
Moya, 2008).
The fourth component (d) focuses on vulnerability assessment and indicates the various types of vulnerability assessment approaches that can be used. Component (e) in
Figure 8 portrays the concept of risk assessment that integrates
the hazard, vulnerability, and amount of elements-at-risk. The
specific risk is calculated for many different situations related
to hazard type, hazard intensity, return period of the triggering
event, and type of element-at-risk. The integration of hazard,
vulnerability, and risk can be done in two ways: quantitative or
qualitative.
Component (f) presents the quantitative risk approach in
which the results are shown in risk curves plotting the expected losses against the probability of occurrence for each
hazard type individually, and also expressing the uncertainty,
by generating two loss curves expressing the minimum and
maximum losses for each return period of triggering events, or
associated annual probability. The individual risks curves can
be integrated into total risk curves for a particular area, and the
population loss can be expressed as F-N curves. The risk curves
can be made for different basic units (e.g., administrative units
such as individual slopes, road sections, census tracts, settlements, municipalities, regions, or provinces). Component (g)
deals with methods for qualitative risk assessment, which are
mostly based on integrating a hazard index, and a vulnerability index, using spatial multi-criteria evaluation. The last
component (h) deals with the use of risk information in
various stages of disaster-risk management.
Hazards will impact different types of elements-at-risk, and
it is therefore important to calculate the risk for different
sectors/environments (e.g., housing, agriculture, transportation, education, health, tourism, protected areas, forests,
wetlands, etc.). Risk assessment should involve the relevant
stakeholders (i.e., individuals, businesses, organizations, and
authorities). The methodology for conducting risk assessments can be broadly classified into qualitative and quantitative approaches.
3.10.5.1
Qualitative Approaches
Qualitative methods for risk assessment are useful as an initial

screening process to identify hazards and risks. They are also
used when the assumed level of risk does not justify the time
and effort of collecting the vast amount of data needed for a
287
quantitative risk assessment, and where the possibility of obtaining numerical data is limited.
The simplest form of qualitative-risk analysis is to combine
hazard maps with elements-at-risk maps in GIS, using a simple-risk matrix in which the classes are qualitatively defined
(AGS, 2000). This method is widely applied, mostly at
(inter)national or provincial scales where the quantitative
variables are not available or they need to be generalized.
Qualitative approaches consider a number of factors that have
an influence on the risk. The approaches are mostly based on
the development of so-called risk indices, and on the use of
spatial multicriteria evaluation. One of the first attempts to
develop global-risk indicators was done through the Hotspots
project (Dilley et al., 2005). In a report for the Inter-American
Development Bank, Cardona (2005) proposed different sets
of complex indicators for benchmarking countries in different
periods (e.g., from 1980 to 2000), and to make cross-national
comparisons. Four components or composite indicators reflect
the principal elements that represent vulnerability and show
the advances of different countries in risk management: 1)
Disaster Deficit Index (DDI); 2) Local Disaster Index (LDI); 3)
Prevalent Vulnerability Index (PVI); and 4) Risk Management
Index (RMI). Each composite index is generated on the basis
of a number of indicators. For instance, the DDI can be considered as an indicator of a countrys economic vulnerability
to disaster. The method has been applied thus far only in Latin
America and the Caribbean. Peduzzi et al. (2005, 2009) have
developed global indicators, not on the basis of administrative
units but based on gridded maps. The Disaster Risk Index
(DRI) (UN-ISDR, 2005b) combines both the total number
and the percentage of killed people per country in large- and
medium-scale disasters associated with droughts, floods, cyclones, and earthquakes based on data from 1980 to 2000. In
the DRI, countries are indexed for each hazard type according
to their degree of physical exposure, their degree of relative
vulnerability, and their degree of risk.
At local scales, risk indices are also used, often in combination with spatial multi-criteria evaluation (SMCE). Castellanos and Van Westen (2007) present an example of the use
of SMCE for the generation of a landslide-risk index for the
country of Cuba, generated by combining a hazard index and
a vulnerability index. The hazard index is computed using
indicator maps related to event-triggering factors (earthquakes
and rainfall) and environmental factors. The vulnerability
index was made using five key indicators including housing
condition and transportation (physical-vulnerability indicators), population (social-vulnerability indicator), production (economic-vulnerability indicator), and protected
areas (environmental-vulnerability indicator). The indicators
were based on polygons related to political-administrative
areas, which are mostly at the municipal level. Each indicator
was processed, analysed, and standardized according to its
contribution to hazard and vulnerability. The indicators were
weighted using direct pair-wise comparison and rank-ordering
weighting methods, and weights were combined to obtain the
final landslide risk-index map. The results were analyzed per
physiographic region and administrative units at provincial
and municipal levels. At the local level, Villagran de Leon
(2006) incorporated 3 dimensions of vulnerability that included the scale/level (from human being to national level),
288
the various sectors of society, and six components of vulnerability. The method uses matrices to calculate a vulnerability
index that are grouped in qualitative classes (high, medium,
and low).
3.10.5.2
Quantitative Approaches
Quantitative approaches aim at expressing the risk in quantitative terms, either as probabilities or expected losses. They can
be deterministic (scenario-based) or probabilistic (taking into
account the effect of all possible scenarios and uncertainties).
Quantitative risk assessment (QRA) follows an engineering
approach and focus on the evaluation of the direct physical
losses resulting directly from the impact of the hazard (e.g.,
flooded buildings and collapsed buildings). Some also
analyze indirect losses due to loss of function (e.g., disruption
of transport, business losses, or clean up costs). The focus is on
tangible losses that have a monetary (replacement) value.
Disasters also cause a large amount of intangible losses, for
example, lives and injuries, cultural heritage, environmental
quality, and biodiversity. Quantitative risk assessment aims at
quantifying the risk according to the equation given in
Figure 8. Several approaches, however, differ in how they
calculate hazards and vulnerability consequences. For a
number of different hazard scenarios, the consequences are
plotted against the temporal probability of occurrence of the
hazard events. Through these points a curve is fitted, the socalled risk curve, and the area below the curve presents the
total risk. This procedure is carried out for all individual
hazard types, and care should be taken to evaluate interrelations between hazards. Since the risk is normalized into
annual risk, it is then possible to evaluate the multi-hazard
risk, and use the risk curves as the basis for disaster-risk reduction. The (epistemic and aleatory) uncertainties are incorporated in the modeling and used to calculate exceedance
probability curves, average annual losses (AAL), and probable
maximum losses (PML).
Loss estimation has been carried out initially from the early
days of insurance, and has evolved to computer-based catastrophe modeling since the late 1980s using advanced information technology and GIS (Grossi et al., 2005). Since the end
of the 1980s, risk-modeling firms such as AIR Worldwide, Risk
Management Solutions (RMS), EQECAT, and others have led
the industry of probabilistic-risk modeling. Numerous integrated numerical-models were developed for simulating catastrophic hazards that account for different types of hazards.
For instance, EQECAT developed a system (WORLDCATenterprise) that includes 181 natural-hazard models from 95
countries (EQECAT, 2010). These software tools are proprietary, however, and were used for the insurance market.
Software tools developed by the scientific community also
exist for disaster-risk management. One of the earliest publicly
available software methods for loss estimation was RADIUS
(Risk Assessment Tools for Diagnosis of Urban Areas against
Seismic Disasters). The very simple tool enabled users to
perform an aggregated loss estimation using a gridded mesh
and estimate the number of damaged buildings, length of
damaged lifelines, and the number of casualties and injured
people (RADIUS, 1999). The best publicly available software
tool for loss estimation to-date is HAZUS (which stands for

Hazards US), developed by FEMA together with the National Institute of Building Sciences (NIBS). The goal of FEMA
was to create a methodology that was the standard nationalloss methodology for assessing losses from natural hazards
(FEMA, 2004). The first version of HAZUS was released in
1997 with a seismic loss estimation focus, and was extended
to multi-hazard losses in 2004, incorporating losses from
floods and windstorms. HAZUS was developed as a software
tool under ArcView and later ArcGIS. Although the HAZUS
methodology has been very well documented, the tool was
primarily developed for the US, and all data formats, building
types, fragility curves, and empirical relationships cannot be
exported easily to other countries.
Several other countries have adapted the HAZUS methodology to their own situation (e.g., in Taiwan (Yeh et al., 2006)
and Bangladesh (Sarkar et al., 2010)). The HAZUS methodology has also been the basis for the development of several
other open-source software tools for loss estimation. One is
called SELENA (SEimic Loss EstimatioN using a logic tree
Approach), and was developed by the International Centre for
Geohazards (ICG), NORSAR (Norway) and the University of
Alicante (Molina et al., 2010). Although most of the above
mentioned GIS-based loss-estimation tools focus on seismic
hazard, the Central American Probabilistic Risk Assessment
Initiative (CAPRA, 2009) has a true multi-hazard risk focus.
The aim of CAPRA is to develop a system that utilizes GIS,
Web-GIS, and catastrophe models in an open platform for
disaster-risk assessment, which allows users from the Central
American countries to analyze the risk in their areas, and be
able to make informed decisions on disaster-risk reduction.
The methodology focuses on the development and use of
probabilistic hazard-assessment modules for earthquakes,
hurricanes, extreme rainfall, and volcanic hazards, and the
hazards triggered by them, such as flooding, windstorms,
landslides, and tsunamis. These are based on event databases
with historical and simulated events. This information is
combined with elements-at-risk data focusing on buildings
and population. For the classes of elements-at-risk, vulnerability data can be generated using a vulnerability module. The
main product of CAPRA is a software tool, called CAPRA-SIG,
which combines the hazard scenarios, elements-at-risk, and
vulnerability data to calculate loss-exceedance curves.
In New Zealand, a comparable effort includes the RiskScape methodology for multi-hazard risk assessment (Reese
et al., 2007; Schmidt et al., 2011). Another example of multihazard risk assessment is the Cities project in Australia, which
is coordinated by Geoscience, Australia. Studies have been
made for six cities of which the Perth study is the latest
(Durham, 2003; Jones et al., 2005). Also in Europe, several
projects have developed multi-hazard loss estimations systems, such as the ARMAGEDOM system in France (Sedan and
Mirgon, 2003) and in Germany (Grunthal et al., 2006).
The aforementioned systems focus on the assessment of
losses before events, whereas other systems aim at providing
fast assessments of damage directly after the occurrence of
major events. For instance, the PAGER (Prompt Assessment of
Global Earthquakes for Response) system, developed by the
USGS is an automated system that rapidly assesses earthquake
impacts by comparing the population exposed to each level of
289
Table 7 Relationship between stakeholders in risk management and risk visualization options
Stakeholder
Purpose
Type of risk visualization
General public
General information on risks

over large areas
Awareness raising
Community-based DRR projects
Businesses
Investment policies and location

planning
Land use regulation/zoning
Basic WebGIS applications in which they can overlay the location of major
hazard types with high-resolution imagery or topographic maps.
Animations (what if scenarios)
Simple maps of the neighborhood with risk class, buildings, evacuation
routes, and other features.
General information about hazards and risks in both graphical and map
format.
Map with simple legend including construction restricted, construction
allowed, further investigation required.
Maps indicating the types of building allowed (building type, number of
floors).
Hazard maps, with simple legends related to probabilities and possible
consequences.
Maps and possible loss figures for future scenarios.
Technical staff of (local)

authorities
Building codes
Spatial planning
Environmental Impact
Assessment
Disaster preparedness
Decision makers / local
authorities
NGOs
Scientists / technical staff of

hazard data producers
Insurance industry
Media
Decision making on risk

reduction measures
Investments
Strategic Environmental
Assessment
Influence political decisions in
favor of environment and
sustainable development
Hazard information exchange to
public and other agencies
Exchange of basic information
for hazard and risk
assessment
Development of insurance policy
Risk communication to public
shaking intensity with models of economic and fatality losses

based on past earthquakes in each country or region of the
world (PAGER, 2010).
3.10.5.3
Spatial Risk Visualization
Risk management cannot take place without proper risk governance. Risk governance has been promoted in the ISDR,
Hyogo framework for action to: Promote and improve dialogue and cooperation among scientific communities and
practitioners working on disaster-risk reduction, and encourage partnerships among stakeholders, including those
working on the socio-economic dimensions of disaster-risk
reduction (UN-ISDR, 2005a). Governance depends on the
level of political commitment and strong institutions. Good
governance is identified in the ISDR Framework for disaster
reduction as a key area for the success of effective and sustained disaster-risk reduction (IRGC, 2005). One of the important processes in risk governance is risk communication,
which is the interactive exchange of information about risks
among risk assessors, managers, news media, interested
groups, and the general public. An important component of
that is the visualization of risk. Since risk is a spatially varying
phenomenon, GIS technology is now the standard approach
for the production and presentation of risk information. Risk
Real time simple and concise Web-based information in both map and
graphical forms.
Statistical information, loss-exceedance curves, F-N curves, and maps.
Economic losses, projected economic losses for future scenarios.
General statistical information for administrative units.
This can vary from simple maps to Web-based applications, depending
on the objectives of the NGO.
WebGIS applications where they can access the basic information.
Spatial Data Infrastructure / Clearinghouse for exchanging information.
Loss-Exceedance Curves of economic losses, F-N curves.

Animations of hazard phenomena that clearly illustrate the problems.
can be presented in the form of statistical information per

administrative unit, such as a risk index value resulting from
qualitative risk assessment, the probable maximum loss or the
average annual loss, loss-exceedance curve for economic risk,
or F-N curves for societal population risk. Risk can also be
visualized in map form that depicts the spatial variation
of risk.
The type of risk (qualitative/quantitative, direct/indirect,
societal risk/individual risk, etc.) and the visualization technique used depends on the stakeholder to which the risk information is communicated. Table 7 provides an overview of
the relation between stakeholders and the type of risk
visualization.
Internet-based GIS systems have been developed in which
all the individual layers are separated (multi-tier approach),
thus allowing many clients to access and visualize the geo-data
at the same time. A WebGIS is a special GIS tool that uses the
Internet as a means to access and transmit remote data, conduct analysis, and present GIS results. WebGIS applications for
risk visualization have been developed for different purposes.
At the global level, the PREVIEW Global Risk Data Platform is
the result of efforts of UNEP, UNISDR, UNDP, and World
Bank, to share spatial data and information on global risk
from natural hazards through the internet. Users can visualize,
download, or extract data on past hazardous events, human
and economical hazard exposure, and risk from natural
290
Figure 9 Global Risk Data Platform, PREVIEW. Reproduced from UNEP/DEWA/GRID, 2010. PREVIEW Global Risk Data Platform. United Nations
Environment Programme, Global Resource Information Database, Europe, Geneva. http://www.grid.unep.ch/activities/earlywarning/preview/
index.php
hazards on a platform compliant with OGC Web Services

(OWS). It covers tropical cyclones and related storm surges,
drought, earthquakes, biomass fires, floods, landslides, tsunamis, and volcanic eruptions (see Figure 9). The collection of
data is made via a wide range of partners (UNEP/DEWA/
GRID, 2010).
An example of risk visualization at the international level is
the multi-hazard risk atlas for the Andean region (Communidad Andina, 2009) that is available in paper atlas and Webbased versions. This atlas provides a comprehensive overview
of the elements-at-risk in the region (population, production,
and infrastructure), the hazard phenomena (earthquakes,
tsunami, volcanic eruptions, landslides, flooding, cold waves,
and drought), and the risks in a very well designed manner.
Examples of different approaches for visualizing flood hazard
and risk maps from 19 European countries, USA, and Japan
are presented in EXCIMAP (2007). Many countries are also
developing their own Web-based risk maps. For example, the
CEDIM Risk Explorer, Germany, is a web-based map viewer
that interactively presents the results of the CEDIM project
"Riskmap Germany" (Muller et al., 2006). A more complicated Web-GIS system has been developed in the Netherlands,
which can be accessed by the general public as part of the
national risk communication strategy. A secured section of the
same system can be accessed by professionals involved in risk
management, allowing them to get more detailed information
required for emergency response planning. National-scale risk

mapping in the Netherlands was carried out after the occurrence of major technical and flood disasters in the past
decades. The Web-GIS application (see Figure 10) shows information on natural hazards (flooding, natural fires, and
earthquakes), technological hazards (transportation accidents,
hazardous substances, and nuclear), and vulnerable objects
(Risicokaart, 2008). The flood-prone areas are defined by
more than 1 m flooding depth with a frequency larger than
1/4000 per year.
3.10.6
Conclusions
This chapter provides a framework for understanding hazards

assessment and disaster-risk management. Spatial data requirements and techniques for multi-hazard risk assessment
have been addressed. It should be emphasized that data collection, analysis/modeling, and information production as
part of this process is a complex task, because risk is a dynamic
concept and has many facets. This is illustrated in Figure 11. It
is evident that the world undergoes rapid changes in terms of
population growth, urbanization, economic development,
and socio-political structures. Furthermore, there is convincing
evidence that greenhouse gas forcing may be causing changes
in the Earths climate that are expected to lead to an increase in
291
Figure 10 The national risk atlas of the Netherlands is publicly available on the internet (www.risicokaart.nl).
hazardous events due to a hydro-meteorological trigger (IPCC,

2007).
The difficulty in predicting the magnitude of these changes
and the frequency of occurrence of extreme events reiterates
the need for a thorough change in our adaptation management of hydro-meteorological risks (EEA, 2004). According to
recent European studies, the projected impact of flooding in
Europe will increase dramatically in the coming decades. By
2080, it is estimated that between 250 000 and 400 000
people will be affected each year by flooding, and the total
annual expected flood damage will range between 7.7 and 15
billion Euros. These values are more than double of those in
the period 19611990 (Ciscar, 2009). Very limited work has
been carried out up to now to include the cascading or conjoint (also called domino) effects in the analysis of future
impacts of environmental changes to hydro-meteorological
hazards. The exposure of elements-at-risk also increases and
therefore the risk of natural hazards is constantly growing.
Land-use changes will occur as a result of technological, socioeconomic, and political developments, as well as global environmental change. The nature and rate of change will
strongly depend on policy decisions. Many environmental
problems are caused by unplanned urban expansion. By 2050,
approximately 70% of all people will be living in urban areas,
whereas in several countries the proportion will be 90% or
more. Some of the drivers of change to the urban environment
292
Change analysis
Analysis of uncertainty
Climate change
Changing
hazards
Risk management
Risk assessment
Changing
elements at risk
Changing
vulnerability
Vulnerability
assessment
Risk analysis
Scenario
development
Risk evaluation
Scenario A
Risk reduction measures
Scenario B
Emergency
preparedness
Scenario C
Land use planning
Scenario ..
Cost-benefit
assessment
(Spatial) information management
Socio-economic
change
Elements at risk
/exposure
Risk governance/Risk communication/Risk visualization
Land use
change
Multi-hazard
assessment
Figure 11 Framework of the implementation of environmental change scenarios in risk management.
are the global economy, cross-border transport networks,

large-scale societal, economic and demographic changes, and
differences in national planning laws. As the level of uncertainty of the components used in the risk equation (hazard,
vulnerability, and quantification of the exposed elements-atrisk) is very high, the analysis of the changes in future risk
should incorporate these uncertainties in a probabilistic
manner.
Impacts of natural hazards on the environment and on the
society are still tackled by mono-disciplinary approaches. The
focus is reflected in the domains of scientific research (single
approach and tools for each type of hazard), in the existing
management tools, and in the legislative basis of these activities. Management tools, models, and local-to-regional technical solutions have been proposed by numerous projects for
single hazards. Only a few of them have tackled the issue of
risk assessment and management, however, from a multihazard perspective, especially including possible combined
and domino effects. Probabilistic tools for multi-hazard risk
assessment are not available to stakeholders at the local level.
Insurance companies and specialized risk-assessment consultants have developed models, but these are not open for
public use. The implementation of risk-management measures
such as disaster-preparedness program, land-use planning,
regulatory zoning and early warning systems are considered
essential. Fleischauer et al. (2006) concluded that spatial

planning is only one of many aspects in risk management and
that it is, in general, not properly implemented. Furthermore,
multi-risk assessment approaches are not used in planning
practice: Risk indicators are hardly used and vulnerability indicators are not at all.
Therefore, approaches are needed for integrating disasterrisk assessment in long-term resource allocation and land-use
planning at all levels of administration. Additionally, scientific
advances in hazard and risk assessment and demands of
stakeholders/end-users are still not well connected. In many
cases, the scientific outcomes remain rooted solely within the
scientific community, or new knowledge is not fabricated
enough to be implemented by stakeholders and end-users
(IRGC, 2005). A key cause of the gap between the science
community and stakeholders/end-users is in the complexity of
human-environment interactions. This has led to the development of a diversity of approaches, commonly not easy to
implement by the end-user community. A need exists for the
development of a harmonized decision-making structure for
applying hazard and risk mitigation through spatial planning
in risk-prone areas. A need also exists for capacity building in
the field of multi-hazard risk assessment, and the transfer of
the knowledge from developed countries to developing
countries using open-source software tools and methods
adapted to the data availabilities in these countries (Van

Westen et al., 2009). The Hyogo framework of action
20052015 of the UN-ISDR indicates risk assessment and
education as two of the key areas for the development of action in the coming years.
In conclusion, it is clear that geomorphology, geographic
information science, and Earth observation (remote sensing)
have made significant contributions to the understanding of
natural hazard processes, and the way these can be monitored
and assessed/modeled at various scales using a range of
geospatial technologies. Such capabilities have also been used
successfully in analyzing the risk to vulnerable societies, and
the results have been communicated to stakeholders that have
often used them in appropriate disaster-risk reduction strategies. Nevertheless, as extreme events and the number of
vulnerable people are expected to increase in the future, more
work is required to improve our ability to accurately estimate
future hazards and risks. Geomorphologists should further
develop their scientific work into practical applications that
can be used to save lives and reduce economic losses due to
natural disasters.
Acknowledgements
This research was supported by the United Nations University
ITC School for Disaster Geo-Information Management. The
author would like to thank Sekhar Lukose Kuriakose, Norman
Kerle and Michael Bishop for their comments on the draft.
References
Africover, 2010. Africover Land cover database for Africa. Food and Agricultural
Organization of the United Nations. FAO.
AGS, 2000. Landslide risk management concepts and guidelines, Australian
Geomechanics Society (AGS), Sub-committee on landslide risk management.
Alcantara-Ayala, I., 2004. Flowing Mountains in Mexico, Incorporating Local
Knowledge and Initiatives to Confront Disaster and Promote Prevention. In
Mountain Research and Development 24(1), 1043.
Alcantara-Ayala, I., Goudie, A.S., 2010. Geomorphological Hazards and Disaster
Prevention. Cambridge University Press, Cambridge, 291 pp.
Alernet, 2010. Interactive map. AlertNet, Alerting humanitarians to emergencies.
Thomas Reuters Foundation, http://www.alertnet.org/map/index.htm
Alexander, D., 1993. Natural disasters. UCL Press Ltd., University College, London.
Alkema, D., Middelkoop, H., 2005. The influence of floodplain compartmentalization
on flood risk within the Rhine - Meuse delta. Natural hazards: journal of
the international society for the prevention and mitigation of natural hazards
36(1-2), 125145.
Anderson, M.B., Woodrow, P.J., 1998. Rising from the Ashes: Development
Strategies in Times of Disaster. London, IT Publications.
ANSS, 2009. Advanced National Seismic System. http://earthquake.usgs.gov/
monitoring/anss/
APD, 2010. African Population Database. United Nations Environmental Programme
(UNEP). http://www.na.unep.net/globalpop/africa/
Arino, O., Gross, D., Ranera, F., et al., 2007. GlobCover: ESA service for global
land cover from MERIS. Geoscience and Remote Sensing Symposium, 2007.
IGARSS 2007. IEEE International : 24122415
Arnold, E., 1989. Program SEISRISK III Adapted to Personal computers. Open-File
Report, 89557.
Asselen, S.V., Seijmonsbergen, A.C., 2006. Expert-driven semi-automated
geomorphological mapping for a mountainous area using a laser DTM.
293
Ayanza, J.S.M., Barbosa, P.M., Schmuck, G., Liberta, G., (2003). The European
Forest Fire Information System (EFFIS). European Commission. JIR Institute for
Environment and sustainability.
Balk, D.L., Deichmann, U., Yetman, G., Pozzi, F., Hay, S.I., 2006. Determining
global population distribution: methods, applications and data. Advances in
Parasitology 62, 119156.
Bankoff, G., Frerks, G., Hillhorst, D. (Eds.), 2003. Mapping Vulnerability: Disasters,
Development and People. Earthscan, London.
Barlow, J., Franklin, S., Martin, Y., 2006. High spatial resolution satellite imagery,
DEM derivatives, and image segmentation for the detection of mass wasting
processes. Photogrammetric Engineering and Remote Sensing 72(6), 687692.
Barredo, J.I., 2007. Major flood disasters in Europe: 19502005. Natural Hazards
42(1), 125148.
Bassole, A., Brunner, J., Tunstall, D., 2001. GIS: supporting environmental planning
and management in West Africa. A report of the joint USAID/World Resources
Institute Information Group for Africa.
Beven, K., Calver, A. and Morris, E.M., 1987. The Institute of Hydrology distributed
model. Institute of Hydrology report 98, Wallingford UK.
Bhaduri, B., Bright, E., Coleman, P., Urban, M., 2007. LandScan USA: A High
Resolution Geospatial and Temporal Modeling Approach for Population
Distribution and Dynamics. GeoJournal 69, 103117.
Birkmann, J. (Ed.), 2006. United Nations University Press, Tokyo, 524pp524pp.
Blaikie, P., Cannon, T., Davis., I., Wisner, B., 1994. At risk: natural hazards, peoples
vulnerability and disasters. Routledge, London.
Blong, R., 2003. A new damage index, Nat. Hazards 30(1), 123.
Blumberg, D., Bach, D., Weissel, J., Gorokhovich, Y., Small, C., & Balk, D. 2005.
The 2004 Sumatra tsunami event: Contribution of SRTM data to the analysis of
devastation. The Shuttle Radar Topography Mission Data Validation and
Applications, Workshop, June 1416, 2005, Reston, Virginia.
Bonforte, A., Puglisi, G., 2003. Magma uprising and flank dynamics on
Mount Etna volcano, studied using GPS data (19941995). J. Geophys. Res
108(B3), 2153.
Bourgeau-Chavez, L., Kasischke, E., 2002. Mapping fire scars in global boreal
forest using imaging radar data. International Journal of Remote Sensing 23,
42114234.
Brenner, C., 2005. Building reconstruction from images and laser scanning.
International Journal of Applied Earth Observation and Geoinformation 6(34),
187198.
Brunner, G.W., 2002. HEC-RAS River Analysis System; Hydraulic Reference Manual.
US Army Corps of Engineers, Davis, USA.
Burgmann, R., Hilley, G., Ferretti, A., Novali, F., 2006. Resolving vertical tectonics
in the San Francisco Bay Area from Permanent Scatterer InSAR and GPS
Analysis. Geology 34, 221224.
Calvi, G.M., Pinho, R., Magenes, G., Bommer, J.J., Restrepo-Velez, L.F., Crowley,
H., 2006. Development of Seismic Vulnerability Assessment Methodologies Over
The Past 30 Years. ISET Journal of Earthquake Technology, Paper No. 472
43(3), 75104.
Cannon, S.H., Gartner, J.E., Wilson, R.C., Bowers, J.C., Laber, J.L., 2008. Storm
rainfall conditions for floods and debris flows from recently burned areas in
southwestern Colorado and southern California. Geomorphology 96(3-4),
250269.
Cannon, T., Twigg, J., Rowell, J., 2003. Social vulnerability, sustainable Livelihoods
and disasters. Greenwich, DFID, 63 p.
CAPRA, 2009. Central American Probabilistic Risk Assessment (CAPRA). World
Bank. www.ecapra.org
Cardona, O.D., 2005. Indicators for Disaster Risk and Risk Management. Instituto
de Estudios Ambientales, Universidad Nacional de Columbia, Manizales,
Columbia, Program for Latin America and the Caribbean: Summary Report.
Carrara, A., Cardinali, M., Guzzetti, F., Reichenbach, P., 1995. GIS-based techniques
for mapping landslide hazard. In: Carrara, A., Guzzetti, F. (Eds.), Geographical
Information Systems in Assessing Natural Hazards. Kluwer Academic
Publications, Dordrecht, The Netherlands, 135-176135-176.
Carton, A., Coratza, P. and Marchetti, M., 2005. Guidelines for geomorphological
sites mapping: examples from Italy , Geomorphologie : relief, processus,
environnement, 3/2005.
Castellanos, E., Van Westen, C.J., 2007. Qualitative landslide susceptibility
assessment by multicriteria analysis; a case study from San Antonio del Sur,
Guantanamo, Cuba. Geomorphology 94(3-4), 453466.
CEOS, 2003. Geohazards Team Report. Committee on Earth Observation Satellites
(CEOS). Integrated Global Observing Strategy (IGOS) http://www.igospartners.
org/Geohazards.htm
CGIAR-CSI, 2008. SRTM 90m Digital Elevation Data. The CGIAR Consortium for
Spatial Information (CGIAR-CSI) http://srtm.csi.cgiar.org/
294
Chen, K., McAneney, J., Blong, R., Leigh, R., Hunter, L., Magill, C., 2004. Defining
area at risk and its effect in catastrophe loss estimation: a dasymetric mapping
approach. Applied Geography 24, 97117.
CIESIN, 2005. Global Multihazard Proportional Economic Loss Risk Deciles. Center
for Hazards and Risk Research (CHRR); Center for International Earth Science
Information Network (CIESIN). Columbia University, Palisades, NY. http://
www.ldeo.columbia.edu/chrr/research/hotspots/
Cihlar, J., 2000. Land cover mapping of large areas from satellites: status and
research priorities. International Journal of Remote Sensing, 13665901 21(6),
10931114.
Ciscar, J.C. (Ed), 2009. Climate change impacts in Europe. Final report of the
research project. European Commission Joint Research Centre. EUR 24093 EN
Cobby, D.M., Mason, D.C., Davenport, I.J., 2001. Image processing of airborne
scanning laser altimetry data for improved river flood modeling. ISPRS Journal
of Photogrammetry and Remote Sensing 56(2), pp. 121138.
Coburn, A.W., Spence, R.J.S., 2002. Earthquake Protection, 2nd ed. John Wiley &
Sons, Chichester, UK.
Communidad Andina 2009. Atlas de las dinamicas del Territorio Andino.
Poblacion y bienes expuestos a amenazas naturales. Communidad Andina,
Comite Andino para la Prevencion y Atencion de Desastres, CAPRAD.
http://www.comunidadandina.org/predecan/atlasweb/index.html
Connor, C.B., Hill, B.E., Winfrey, B., Franklin, N.M., La Femina, P.C., 2001. Estimation
of volcanic hazards from tephra fallout. Natural Hazards Review 2, 3342.
Coppin, P., Jonckheere, I., Nackaerts, K., Muys, B., Lambin, E., 2004. Digital
change detection methods in ecosystem monitoring: a review. International
Journal of Remote Sensing 25(9), 15651596.
Coppock, J.T., 1995. GIS and natural hazards: an overview from a GIS perspective.
In: Carrara, A., Guzzetti, F. (Eds.), Geographical information systems in
assessing natural hazards. Kluwer, Dordrecht, pp. 2134.
Corominas, J., Moya, J., 2008. A review of assessing landslide frequency for
hazard zoning purposes. Engineering geology 102, 193213.
Dai, F.C., Lee, C.F., Ngai, Y.Y., 2002. Landslide risk assessment and management:
an overview. Engineering Geology 64, 6587.
Dal Cin, C., Moens, L., Dierickx, P., Bastin, G., Zech, Y., 2005. An integrated
approach for realtime floodmap forecasting on the Belgian Meuse river. Natural
Hazards 36(12), 237256.
Dartmouth Flood Observatory, 2010. Space based measurement of surface water.
http://floodobservatory.colorado.edu/
De Roo, A., Barredo, J., Lavalle, C., Bodis, K., Bonk, R., 2007. Potential Flood
Hazard and Risk Mapping at Pan-European Scale. Digital Terrain Modelling.
Lecture Notes in Geoinformation and Cartography 2007, 183202.
De Roo, A.P.J., Wesseling, C.G., Van Deursen, W.P.A., 2000. Physically based river
basin modelling within a GIS: the LISFLOOD model. Hydrological Processes 14,
19811992.
DesInventar, 2010. Inventory system of the effects of disasters. http://
online.desinventar.org
Dilley, M., Chen, R.S., Deichmann, U., et al., 2005. Natural Disaster Hotspots: A
Global Risk Analysis. The World Bank, Hazard, Management Unit, Washington,
D.C. Disaster Risk Management Series No. 5, 132 pp.
Disaster Charter (2010). The International Charter on Space and Major Disasters.
http://www.disasterscharter.org/home
DLR-ZKI, 2010. Center for Satellite Based Crisis Information (ZKI) , German Remote
Sensing Data Center (DFD) of DLR. http://www.zki.dlr.de/mission
Douglas, J., 2007. Physical vulnerability modelling in natural hazard risk
assessment. Nat. Hazards Earth Syst. Sci 7, 283288.
Dragut- , L., Blaschke, T., 2006. Automated classification of landform elements using
object-based image analysis. Geomorphology 81(3-4), 330344.
Durham, K., 2003. Treating the Risks in Cairns. Natural Hazards 30(2), 251261.
Ebert, A., Kerle, N., Stein, A., 2009. Urban social vulnerability assessment with
physical proxies and spatial metrics derived from air- and spaceborne imagery
and GIS data. Natural hazards 48(2), 275294.
EEA, 2004. "Mapping the impacts of recent natural disasters and technological
accidents in Europe", Environmental Issue Report No. 35, European Environment
Agency (EEA), Copenhagen, Denmark.
EFD, 2007. European Floods Directive. European Commission. http://ec.europa.eu/
environment/water/flood_risk/index.htm
EM-DAT, 2009. The OFDA/CRED International Disaster Database, Universite
Catholique de Louvain, Brussels, Belgium. http://www.emdat.be
Emani, S., 1996. Applications in hazard assessment and management. Explorations
in Geographic Information Systems Technology, Vol. 6. Worcester (USA), United
Nations Institute for Training and Research (UNITAR).
Emanuel, K., Ravela, S., Vivant, E., Risi, C., 2006. A statistical deterministic
approach to hurricane risk assessment. Bull. Am. Met. Soc 87, 299314.
EQECAT, 2010. WORLDCAT enterpriseTM, EQECATs catastrophe risk modeling

software platform. http://www.eqecat.com
EXCIMAP, 2007. Atlas of flood maps. European Commission.
Falorni, G., Teles, V., Vivoni, E.R., Bras, R.L., Amaratunga, K.S., 2005. Analysis and
characterization of the vertical accuracy of digital elevation models from the
shuttle radar topography mission. Jounal of Geophysical Research 110.
FAO, 1981. Soil Map of the World, United Nations Food and Agriculture
Organization, Vols. 1-10 (1:5M scale maps and accompanying texts). UNESCO,
Paris.
FAO, 2010. Geonetwork. Food and Agriculture Organisation of the United Nationa
(FAO). http://www.fao.org/geonetwork/srv/en/main.home
Farina, P., Casagli, N., Ferretti, A., 2008. Radar-interpretation of InSAR
measurements for landslide investigations in civil protection practices. Proc. 1st
North American Landslide Conference. Vail, Colorado, 272283.
Farr, T.G., Kobrick, M., 2000. Shuttle Radar Topography Mission produces a wealth
of data. EOS Trans 81, 583585.
Favey, E., Wehr, A., Geiger, A., Kahle, H.-G., 2002. Some examples of European
activities in airborne laser techniques and an application in glaciology. Journal
of Geodynamics 34, 347355.
Fell, R., Corominas, J., Bonnard, C., Cascini, L., Leroi, E., Savage, W.Z., 2008.
Guidelines for landslide susceptibility, hazard and risk zoning for land use
planning. Engineering Geology 102, 8598.
FEMA, 2004. HAZUS-MH. FEMAs Methodology for Estimating Potential Losses
from Disasters. US Federal Emergency Management Agency. http://
www.fema.gov/plan/prevent/hazus/index.shtm
FEMA, 2010. National Flood Insurance Programme. US Federal Ergency
Management Agency. http://www.fema.gov/business/nfip/
Ferretti, A., Prati, C., Rocca, F., 2001. Permanent Scatterers in SAR interferometry.
IEEE Trans. Geosci. Remote Sens 39(1), 820.
Ferrier, N., Haque, C.E., 2003. Hazards Risk Assessment Methodology for
Emergency Managers: A Standardized Framework for Application. Natural
Hazards 28(2-3), 271290.
Fleischhauer, M., Greiving, S., Wanczura, S. (Eds.), 2006. Natural Hazards and
Spatial Planning in Europe. Dortmund 2006, 206 pp.
Fraser, C.S., Baltsavias, E., Grune, A., 2002. Processing of IKONOS imagery for
submetre 3D positioning and building extraction. ISPRS Journal of
Photogrammetry and Remote Sensing 56(3), 177194.
French, J.R., 2003. Airborne LiDAR in support of geomorphological and hydraulic
modeling. Earth Surface Processes and Landforms vol. 28, 3.
Friedl, M.A., McIver, D.K., Hodges, J.C.F., et al., 2002. Global land cover mapping
from MODIS: algorithms and early results. Remote Sens Environ 83(12),
287302.
Fuchs, S., Heiss, K., Hubl, J., 2007. Towards an empirical vulnerability function for
use in debris flow risk assessment. Natural Hazards and Earth System Sciences
7, 495506.
Fujisada, H., Bailey, G.B., Kelly, G.G., Hara, S., Abrams, M.J., 2005. ASTER DEM
performance. IEEE Transactions on Geoscience and Remote Sensing 43(12),
27072714.
Furbish, D.J., Rice, R.M., 1983. Predicting landslides related to clearcut logging,
Northwestern California, USA. Mountain Research and Development 3, 253259.
GDACS, 2010. Global Disaster Alert and Coordination System. Joint Research
Centre of the European Commission, and United Nations Office for Coordination
of Humanitarian Affairs. http://www.gdacs.org/
GEM, 2010. Global Earthquake Model. http://www.globalquakemodel.org/
GeoNode, 2010. GeoNode. Initiative of OpenGeo and the World Bank. http://
www.geonode.org/
Giglio, L., Kendall, J.D., 2001. Application of the Dozier retrieval to wildfire
characterization: a sensitivity analysis. Remote Sensing of Environment 77,
3449.
Giles, P.T., Franklin, S.E., 1998. An automated approach to the classification of the
slope units using digital data. Geomorphology 21, 251264.
Gili, J.A., Corominas, L., Rius, J., 2000. Using Global Positioning System
techniques in landslide monitoring. Engineering Geology 55(3), 167192.
Glade, T., 2003. Vulnerability assessment in landslide risk analysis. Die Erde 134,
121138.
Glade, T., Anderson, M., Crozier, M.J., 2005. Landslide Hazard and Risk. John
Wiley & Sons, Ltd., Chichester, England, 802 pp.
Glidenumber, 2010. Disaster database. ADRC. See www.glidenumber.net
GMES, 2010. Global Monitoring for Environment and Security. European
Commission.
GMK, 2010. Geomorphologische Karten im Netz. Bonn University.
GNS, 2010. Central South Island Glacial Geomorphology. Institute of Geological
and Nuclear Sciences. http://maps.gns.cri.nz/website/csigg/
GRIP, 2010. DISDAT the disaster data collection initiative portal. Global Risk
Identification Program (GRIP). http://www.gripweb.org/
Grossi, P., Kunreuther, H., Patel, C.C., 2005. Catastrophe Modeling: A New
Approach to Managing Risk. Springer, New York.
GRUMP, 2004. Center for International Earth Science Information Network (CIESIN),
Columbia University. Global Rural-Urban Mapping Project (GRUMP). CIESIN,
Columbia University, Palisades, NY. http://sedac.ciesin.columbia.edu/gpw
Grunthal, G., Thieken, A.H., Schwarz, J., Radtke, K.S., Smolka, A., Merz, B., 2006.
Comparative Risk Assessments for the City of Cologne Storms, Floods,
Earthquakes. Natural Hazards 38(1-2), 2144.
GSHAP, 1999. The Global Seismic Hazard Assessment Project. International
Lithosphere Program (ILP) with the support of the International Council of
Scientific Unions (ICSU). http://www.seismo.ethz.ch/static/GSHAP/
GSN, 2009. Global Seismographic Network. http://earthquake.usgs.gov/monitoring/
gsn/
Guha-Sapir, D. and Below, R., 2002. The quality and accuracy of disaster data: a
comparative analyses of three global data sets, technical document. The
ProVention Consortium and the Disaster Management Facility, The World Bank,
WHO Centre for Research on the Epidemiology of Disasters, University of
Louvain School of Medicine.
Gunther, A., 2003. SLOPEMAP: programs for automated mapping of geometrical
and kinematical properties of hard rock hill slopes. Computers and Geoscience,
865875.
Gustavsson, M., Kolstrup, E., Seijmonsbergen, A.C., 2006. A new symbol and GIS
based detailed geomorphological mapping system: renewal of a scientific
discipline for understanding landscape development. Geomorphology 77,
90111.
Guzzetti, F., Cardinali, M., Reichenbach, P., 1994. The AVI project: a bibliographical
and archive inventory of landslides and floods in Italy. Environmental
Management 18(4), 623633.
Guzzetti, F., Reichenbach, P., Cardinali, M., Galli, M., Ardizzone, F., 2005.
Probabilistic landslide hazard assessment at the basin scale. Geomorphology 72,
272299.
Harvey, J.T., 2002. Estimating census district populations from satellite imagery:
Some approaches and limitations. International Journal of Remote Sensing 23,
20712095.
Hastings, D.A., Dunbar, P.K., 1998. Development and assessment of the global land
1 km base elevation digital elevation model (GLOBE). International Archives of
Photogrammetry and Remote Sensing 32(4), 218221.
Haugerud, R.A., Harding, D.J., Johnson, S.Y., Harless, J.L., Weaver, C.S., Sherrod,
B.L., 2003. High-resolution LiDAR topography of the Puget Lowland,
Washington - a bonanza for earth science. GSA Today 13, 410.
Henricksen, B.L., Durkin, J.W., 1986. Growing period and drought early warning in
Africa using satellite data. International Journal of Remote Sensing 7(11),
15831608.
Hervas, J., Barredo, J.I., Rosin, P.L., Pasuto, A., Mantovani, F., Silvano, S., 2003.
Monitoring landslides from optical remotely sensed imagery: the case history of
Tessina landslide, Italy. Geomorphology 54, 6375.
Hervouet, J.M., Van Haren, L., 1996. Recent advances in numerical methods for
fluid flows. In: Anderson, M.G., Walling, D.E., Bates, P.D. (Eds.), Floodplain
Processes. John Wiley & Sons Ltd., England, pp. 183214.
Hesselink, A.W., Stelling, G.S., Kwadijk, J.C.J., Middelkoop, H., 2003. Inundation of
a Dutch river polder, sensitivity analysis of a physically based inundation model
using historic data. Water Resources Research 39, 9, art. no. 1234.
HEWSweb, 2010. Humanitarian Early Warning Service. Inter Agency Standing
Committee, World Food Programme. http://www.hewsweb.org
Hilhorst, D., 2004. Unlocking domains of disaster response. In: Bankoff, G., Frerks,
G., Hilhorst, D. (Eds.), Mapping Vulnerability: Disasters, Development and
People. Earthscan Publications, London, pp. 5266.
Hilley, G.E., Burgmann, R., Ferretti, A., Novali, F., Rocca, F., 2004. Dynamics of
slow-moving landslides from Permanent Scatterer analysis. Science 304(5679),
19521955.
Holland, G.J., 1980. An analytic model of the wind and pressure profiles in
hurricanes, Mon. Weather Rev 108, 12121218.
Hollenstein, K., 2005. Reconsidering the risk assessment concept: Standardizing the
impact description as a building block for vulnerability assessment. Nat. Hazards
Earth Syst. Sci., 5, 301307.
Hong, Y., Adler, R., Huffman, G., 2007a. Use of satellite remote sensing
data in the mapping of global landslide susceptibility. Natural Hazards 43,
245256.
Hong, Y., Adler, R.F., Negri, A., Huffman, G.J., 2007b. Flood and Landslide
Applications of Near Real-time Satellite Rainfall Estimation. Natural Hazards
43(2), 285294.
295
Hordijk, M.A., Baud, I.S.A., 2006. The role of research and knowledge generation in
collective action and urban governance: How can researchers act as catalysts?
to Habitat International. Habitat International 30, 668689.
IGOS 2007. Geohazards Earth Observation Requirements. Integrated Global
Observing Strategy (IGOS). Geohazards initiative. BRGM/RP 55719-FR. http://
igosg.brgm.fr/documents.asp
Ioualalen, M., Asavanant, J., Kaewbanjak, N., Grilli, S.T., Kirby, J.T., Watts, P., 2007.
Modeling the 26 December 2004 Indian Ocean tsunami: Case study of impact in
Thailand. Journal of Geophysical Research 112(c07024), 21 pp.
IPCC, 2007. Impacts, Adaptation and Vulnerability. Intergovernmental Panel on
Climate Change, Fourth Assessment Report, Working Group II. http://
www.ipcc.ch/ipccreports/ar4-wg2.htm
Ireland, D., 2001. Disaster Risk Management from a Remote Shires Perspective.
Centre for Disasters Studies, JCU, Australia, 57 p.
IRGC, 2005. Risk Governance: Towards an Integrative Approach. International Risk
Governance Council, White Paper 1 Geneva, http://www.irgc.org/
Publications.html
Jarvinen, B.R., Neumann, C.J., Davis, M.A.S., 1984. A tropical cyclone data tape for
the North Atlantic basin, 18861983, contents, limitations and uses. Miami,
Florida, NOAA Tech Memo, NWS NHC 22.
Jones, T., Middelmann, M. and Corby, N. 2005. Natural Hazard Risk in Perth,
Western Australia. The Cities Project, Geoscience Australia. http://www.ga.gov.au/
hazards/reports/perth/
Joyce, K.E., Belliss, S.E., Samsonov, S.V., McNeill, S.J., Glassey, P.J., 2009. A
review of the status of satellite remote sensing and image processing techniques
for mapping natural hazards and disasters. Progress in Physical Geography
33(2), 183207.
Jibson, R.W., Harp, E.L. and Michael, J.A., 1998, A Method for Producing Digital
Probabilistic Seismic Landslide Hazard Maps: An Example from the Los
Angeles, California, Area, U.S. Geological Survey Open File Report 98-113, 16.
Jimenez, M.J., Garcia-Fernandez, M., Zonno, G., Cella, F., 2000. Mapping soil
effects in Barcelona, Spain, through an integrated GIS environment. Soil
Dynamics and Earthquake Engineering 19(4), 289301.
Kaiser, R., Spiegel, P.B., Henderson, A.K., Gerber, M.L., 2003. The application of
Geographic Information Systems and Global Positioning Systems in
humanitarian emergencies: lessons learned, programme implications and future
research. Disasters 27, 127140.
Karnieli, A., DallOlmo, G., 2003. Remote sensing monitoring of desertification,
phenology, and droughts. Management of Environmental Quality 14(1), 2238.
Kerle, N., Oppenheimer, C., 2002. Satellite remote sensing as a tool in lahar
disaster management. Disasters 26, 140160.
Kirschbaum, D.B., Adler, R., Hong, Y., Lerner-Lam, A., 2009. Evaluation of a
preliminary satellite-based landslide hazard algorithm using global landslide
inventories. Natural Hazards and Earth System Sciences 9, 673686.
Klimaszewski, M., 1982. Detailed geomorphological maps. ITC Journal 3, 265271.
Korup, O., 2002. Recent research on landslide dams-a literature review with special
attention to New Zealand. Progress in Physical Geography 26, 206235.
Kraas, F., 2008. Megacities as Global Risk Areas. Urban Ecology 2008 V, 583596.
Kuriakose, S.L., van Beek, L.P.H., Van Westen, C.J., 2009a. Parameterizing a
physically based shallow landslide model in a data poor region. Earth Surface
Kuriakose, S.L., Devkota, S., Rossiter, D.G., Jetten, V.G., 2009b. Prediction of soil
depth using environmental variables in an anthropogenic landscape, a case
study in the Western Ghats of Kerala, India. Catena 79(1), 2738.
Lang, K. 2002. Seismic Vulnerability of Existing Buildings, IBK Report No. 273,
Swiss Federal Institute of Technology (ETH), Zurich, Switzerland.
LandScan, 2010. LandScanTM Global Population Database. Oak Ridge National
Laboratory, Oak Ridge, TN. http://www.ornl.gov/landscan
Lavell, A., 2000. An approach to concepts and definitions in risk management
terminology and practice. ERD-UNDP, Geneva, 32 pp.
Lavelle, F., Vickery, P.J., Schauer, B., Twisdale, L.A., Laatsch, E. 2003. HAZUS-MH
Hurricane Model, Proceedings, 11th International Conference on Wind
Engineering, Texas Tech University, June 2003, 10151022.
Longley, M.F., Goodchild, D.J., Maguire, Rhind, D.W., 2005. Geographical
Information Systems: Principles, Techniques, Management and Applications.
John Wiley & Sons, Hoboken, NJ.
Lwin, K., Murayama, Y., 2009. A GIS Approach to Estimation of Building Population
for Micro-spatial Analysis. Transactions in GIS 13, 401414.
Lyon, B., Barnston, A.G., 2005. ENSO and the Spatial Extent of Interannual
Precipitation Extremes in Tropical Land Areas. Journal of Climate 18,
50955109.
Malet, J.-P., Thiery, Y., Puissant, A., Hervas, J., Gunther, A., Grandjean, G., 2009.
Landslide susceptibility mapping at 1:1M scale over France: exploratory results
296
with a heuristic model. In: Malet, J.-P., Rematre, A., Boogard, T. (Eds.), Proc.
International Conference on Landslide Processes: from Geomorphologic
Mapping to Dynamic Modelling, 67 February 2009. Strasbourg, France, CERG
Editions, Strasbourg, pp. 315320CERG Editions, Strasbourg.
Mannaerts, C.M., Maathuis, B.H.P., Molenaar, M. and Lemmens, R.L.G. 2009. The
ITC GEONETCast toolbox : a geo capacity building component for education
and training in global earth observation and geo information provision to
society. In: IGARSS 2009: Proceedings of the 2009 IEEE international
geoscience and remote sensing symposium : Earth observation, origins and
applications, July 1217, 2009, Cape Town, South Africa, 385388.
Marcus, W.A., Fonstad, M.A., 2008. Optical remote mapping of rivers at
sub-meter resolutions and watershed extents. Earth Surf. Process. Landf. 33(1),
424.
Martha, T.R., Kerle, N., Jetten, V., Van Westen, C.J., Kumar, K.V., 2010.
Characterising spectral, spatial and morphometric properties of landslides for
semi - automatic detection using object - oriented methods. Geomorphology
116(1-2), 2436.
Maskrey, A., 1998. Navegando entre brumas. La aplicacion de los sistemas de
informacion geografica al analisis del riesgo en America Latina. ITDG-Peru. LA
RED.
changes in the Earths surface. Rev. Geophys 36, 441500.
McCall, M.K., 2003. Seeking good governance in participatory-GIS. Habitat
International 27, 549573.
McCall, M.K., 2008. Participatory Mapping and Participatory GIS (PGIS) for CRA,
Community DRR and Hazard Assessment. Provention Consortium. http://
www.proventionconsortium.org/themes/default/pdfs/CRA/PGIS_Sept08.pdf
Metternicht, G., Hurni, L., Gogu, R., 2005. Remote sensing of landslides: An
analysis of the potential contribution to geo-spatial systems for hazard
assessment in mountainous environments. Remote Sensing of Environment
98(23), 284303.
Miller, P., Mills, J., Edwards, S., Bryan, P., Marsh, S., Mitchell, H., Hobb, P., 2008.
A robust surface matching technique for coastal geohazard assessment and
management. ISPRS Journal of Photogrammetry & Remote Sensing 63,
529542.
Miliaresis, G.C., 2001. Geomorphometric mapping of Zagros Ranges at regional
scale. Computers & Geosciences 27, 775786.
Moel, H., de, van Alphen, J., Aerts, J.C.J.H., 2009. Flood maps in Europe
methods, availability and use. Nat. Hazards Earth Syst. Sci. 9, 289301.
Molina, S., Lang, D.H., Lindholm, C.D., 2010. SELENA An open-source tool for
seismic risk and loss assessment using a logic tree computation procedure.
Computers & Geosciences 36(3), 257269.
Montague, P., 2004. Reducing the harms associated with risk assessments.
Environmental Impact Assessment Review 24, 733748.
Montoya, L., 2003. Geo-data acquisition through mobile GIS and digital video: an
urban disaster management perspective. Environmental Modelling & software
18(10), 869876.
Moore, I.D., Grayson, R.B., Ladson, A.R., 2001. Digital terrain modelling: a review
of hydrological, geomorphological, and biological applications. Hydrological
Processes 5(1), 330.
Mouton, F., Nordbeck, O., 2003. Cyclone database manager: a tool for converting
point data from cyclone observations into tracks and wind speed profiles in a
GIS. University of Geneva.
Muller, M., Vorogushyn, S., Maier, P., et al., 2006. CEDIM Risk Explorer - a map
server solution in the project "Risk Map Germany". Natural Hazards and Earth
System Sciences 6, 711720.
MunichRe, 2010. NATCATSERVICE. Downloadcenter for statistics on natural
catastrophes. Munchener Ruckversicherungs-Gesellschaft, Geo Risks Research,
Munich, Germany. http://www.munichre.com/en/reinsurance/business/non-life/
georisks/natcatservice/default.aspx
Nadim, F., Kjekstad, O., 2009. Assessment of Global High-Risk Landslide Disaster
Hotspots. In: Sassa, K., Canuti, P. (Eds.), Landslides - Diaster Risk Reduction.
Springer, pp. 213221.
Nadim, F., Kjekstad, O., Peduzzi, P., Herold, C., Jaedicke, C., 2006. Global landslide
and avalanche hotspots. Landslides 3(2), 159174.
NERIES, 2009. Network of Research Infrastructures for European Seismology. http://
neries.knmi.nl
NWIS, 2010. National Water Information system. United States Geological Survey.
http://nwis.waterdata.usgs.gov/nwis
OBrien, M., 2000. Making better Environmental Decisions; An Alternative to Risk
Assessment. MIT Press, Cambridge, Massachusets, USA.
OKeefe, P., Westgate, K., Wisner, B., 1976. Taking the naturalness out of natural
disasters. Nature 260, 566567.
Okuyama, Y., Chang, S.E. (Eds.), 2004. Modeling spatial and economic impacts of
disasters. Springer, Advances in spatial science, 329 pp329 pp.
OneGeology, 2010. OneGeology Portal with worldwide geological data. http://
portal.onegeology.org/
OpenSHA, 2010. OpenSHA open source platform. Southern California Earthuake
Center (SCEC) at the University of Southern California (USC) and the United
States Geological Survey (USGS). http://www.opensha.org/
OpenStreetMap, 2010. Open Street Map. http://www.openstreetmap.org/
ORCHESTRA, 2009. Reference ModelORCHESTRA Architecture (RM-OA). EC
ORCHESTRA project. http://www.eu-orchestra.org/documents.shtml
Oude Elberink, S., vossleman, G., 2009. Building Reconstruction by Target Based
Graph Matching on Incomplete Laser Data: Analysis and Limitations. Sensors
2009 9, 61016118.
PAGER, 2010. PAGER - Prompt Assessment of Global Earthquakes for Response.
United States Geological Survey. http://earthquake.usgs.gov/earthquakes/pager/
Peduzzi, P., Dao, H., Herold, C., 2005. Mapping Disastrous. Natural Hazards Using
Global Datasets, Natural Hazards 35(2), 265289.
Peduzzi, P., Dao, H., Herold, C., Mouton, F., 2009. Assessing global exposure and
vulnerability towards natural hazards: the Disaster Risk Index. Nat. Hazards Earth
Syst. Sci. 9, 11491159.
Pelling, M., 2003. The Vulnerability of Cities. Natural disasters and Social
Resilience. Earthscan Publications, London, 256 p.
Peters, A.J., Walter Shea, E.A., JI, L., Vliia, A., Hayes, M., Svoboda, M.D., 2002.
Drought Monitoring with NDVI-Based Standardized Vegetation Index.
Peters Guarin, G., 2008 Integrating local knowledge into GIS based flood risk
assessment, Naga city, The Philippines. Wageningen, Enschede, Wageningen
University, ITC, 2008. ITC Dissertation 157, 352 p.
Peters Guarin, G., Van Westen, C.J., Montoya, A.L., 2005. Community-based flood
risk assessment using GIS for the town of San Sebastian, Guatemala. J. of
Human Security and Development 1, 2949.
Pike, R.J., 2000. Geomorphometry diversity in quantitative surface analysis.
Plapp, T., 2001. Understanding risk perception from natural hazards. In: Amman,
Dannenmann, Vulliet (Eds.), RISK 21 Coping with Risks due to Natural
Hazards in the 21st Century. Taylor & Francis Group, London, pp. 101108.
Priest, G., A. Baptista, E. Myers III, R. Kamphaus 2001. Tsunami hazard assessment
in Oregon. ITS 2001 Proceedings, NTHMP Review Session, R-3, 5565.
Priestnall, G., Jaafar, J., Duncan, A., 2000. Extracting urban features from LiDAR
digital surface models. Computers, Environment and Urban Systems 24(2),
6578.
Provention Consortium, 2010. Community Risk Assessment methodologies and
case studies. http://www.proventionconsortium.org/pageid=43
RADIUS, 1999. RADIUS method (Risk Assessment Tools for Diagnosis of Urban
Areas against Seismic Disasters). http://www.geohaz.org/news/images/
publications/RADIUS_RiskAssessment.pdf
Rautela, P., 2005. Indigenous Technical Knowledge Inputs for Effective Disaster
Management in the Fragile Himalayan Ecosystem. In Disaster Prevention and
Management 14(2), 233241.
Reese, S., Bell, R.G., King, A.B., 2007. RiskScape: a new tool for comparing risk
from natural hazards. Water & Atmosphere 15(3), 2425.
Reese, S., Cousins, W.J., Power, W.L., Palmer, N.G., Tejakusuma, I.G., Nugrahadi,
S., 2007. Tsunami vulnerability of buildings and people in South Java field
observations after the July 2006 Java tsunami. Nat. Hazards Earth Syst. Sci., 7,
573589.
Refsgaard, J.C., Storm, B., 1995. MIKE-SHE. In: Singh, V.J. (Ed.), Computer
models in watershed hydrology. Water Resources Publications.
Reidsma, P., Ewert, F., Boogaard, H., Van Diepen, K., 2009. Regional crop
modelling in Europe: The impact of climatic conditions and farm characteristics
on maize yields. Agricultural Systems 100(1-3), 5160.
Reliefweb, 2010. ReliefWeb, serving the needs of the humanitarian relief community.
http://www.reliefweb.int
Rhind, D.W., 1991. Counting the people: The role of GIS. In: Maguire, D.J.,
Goodchild, M.F., Rhind, D.W. (Eds.), Geographical Information Systems.
Longman, London, pp. 127137.
Risicokaart, 2008. National hazard and risk map of the Netherlands. Landelijke
Beheer Organisatie Risicokaart, Interprovinciaal Overleg (IPO). http://
www.risicokaart.nl
Sahana, 2010. Sahana Free and Open Source Disaster Management System. http://
sahanafoundation.org/
Salvi, S., Atzori, S., Tolomei, C., et al., 2004. Inflation rate of the Colli Albani
volcanic complex retrieved by the permanent scatterers SAR interferometry
technique. Geophys. Res. Lett. 31, 21.
Sarker, J.K., Ansary, M.A., Rahman, M.S., Safiullah, M.M., 2010. Seismic hazard
assessment for Mymensingh, Bangladesh. Environ Earth Sci., 60, 643653.
Schmidt, J., Matcham, I., Reese, S., et al., 2011. Quantitative multi-risk analysis for
natural hazards: a framework for multi-risk modeling. Natural Hazards 58,
11691192.
Schumann, G., Matgen, P., Pappenberger, F., Hostache, R., Puech, C., Hoffmann, L.,
Pfister, L., 2007. High-resolution 3-D flood information from radar for effective
flood hazard management. IEEE Trans. Geosci. Remote Sens 45(6), 17151725.
Sedan, O. and Mirgon, C. (2003). Application ARMAGEDOM Notice
utilisateur,BRGM open file BRGM/RP-52759-FR.
Sentinel Asia, 2010. Sentinel Asia initiative for space-based information and
disaster management in the Asia-Pacific region. Web site: http://
dmss.tksc.jaxa.jp/sentinel/
SERTIT, 2010. SErvice Regional de Traitement dImage et de Teledetection.
University of Strasbourg, http://sertit.u-strasbg.fr
Smith, L.C., 1997. Satellite remote sensing of river inundation area, stage, and
discharge: a review. Hydrological Processes 11, 14271439.
Smith, K., Petley, D.N., 2008. Environmental hazards. Assessing risk and reducing
disaster. Taylor & Francis, London.
Spence, R.J.S., Baxter, P.J., Zuccaro, G., 2004. Building vulnerability and human
casualty estimation for a pyroclastic flow: a model and its application to
Vesuvius. J. Volcanol 133, 321343.
Spence, R.J.S., Kelman, I., Calogero, E., Toyos, G., Baxter, P.J., Komorowski, J.C.,
2005. Modelling expected physical impacts and human casualties from
explosive volcanic eruptions. Natural Hazards and Earth System Sciences 5,
10031015.
Stelling, G.S., Kernkamp, H.W.J., Laguzzi, M.M., 1998. Delft Flooding System, a
powerful tool for inundation assessment based upon a positive flow simulation.
In: Babovic, Larsen (Eds.), Hydro-informatics 1998. Balkema, Rotterdam the
Netherlands, pp. 449456.
Stilla, U., Soergel, U., Thoennessen, U., 2003. Potential and limits of InSAR data
for building reconstruction in built-up areas. ISPRS Journal of Photogrammetry
and Remote Sensing 58(1-2), 113123.
Tatem, A.J., Noor, A.M., von Hagen, C., Di Gregorio, A., Hay, S.I., 2007. High
Resolution Population Maps for Low Income Nations: Combining Land Cover
and Census in East Africa. PLoS ONE 2(12), e1298. http://dx.doi.org/10.1371/
journal.pone.0001298.
Thoma, D.P., Gupta, S.C., Bauer, M.E., Kirchoff, C.E., 2005. Airborne laser scanning
for riverbank erosion assessment. Remote Sensing of Environment 95, 493501.
Tilling, R.I., 1989. Volcanic Hazards, Amer. Geophys. Union, Washington DC, 123 pp.
Tralli, D.M., Blom, R.G., Zlotnicki, V., Donnellan, A., Evans, D.L., 2005. Satellite
remote sensing of earthquake, volcano, flood, landslide and coastal
inundation hazards. ISPRS Journal of Photogrammetry and Remote Sensing 59,
185198.
Trigg, S.N., Roy, D.P., Flasse, S., 2005. An in situ study of the effects of surface
anisotropy on the remote sensing of burned savannah. International Journal of
Twigg, J., 2004. Disaster Risk Reduction: mitigation and preparedness in
development and emergency programming. Overseas Development Institute,
Humanitarian Practice Network, London, Good Practice Review no. 9.
UNEP/DEWA/GRID, 2010. PREVIEW Global Risk Data Platform. United Nations
Environment Programme, Global Resource Information Database, Europe,
Geneva. http://www.grid.unep.ch/activities/earlywarning/preview/index.php
UN-ESCAP, 2003. Guidelines on Participatory Planning and Management for Flood
Mitigation and Preparedness, United Nations- Economic and Social Commission
for Asia and the Pacific (ESCAP). Water Resources Series No. 82. UN.
New York.
UN-ISDR, 2004. Terminology of disaster risk reduction. United Nations, International
Strategy for Disaster Reduction, Geneva, Switzerland. http://www.unisdr.org/eng/
library/lib-terminology-eng%20home.htm
UN-ISDR, 2005a. Hyogo framework for Action 2005-2015. Building the Resilience
of Nations and Communities to Disasters. United Nations, International Strategy
for Disaster Reduction, Geneva, Switzerland. http://www.unisdr.org/eng/hfa/
hfa.htm
UN-ISDR, 2005b. Reducing Disaster Risk: A Challenge for Development. United
Nations, International Strategy for Disaster Reduction, Geneva, Switzerland http://
www.undp.org/bcpr/disred/rdr.htm
UN-ISDR, 2009. Economic damages: share of GDP, by natural disaster and country.
United Nations, International Strategy for Disaster Reduction, Geneva,
Switzerland. http://www.unisdr.org/disaster-statistics/top50.htm
UNPD, 2010a. World Population Prospects: the 2008 Revision. United Nations
Population Division. United Nations, Department of Economic and Social Affairs,
New York. http://esa.un.org/unpd/wpp2008/index.htm
297
UNPD, 2010b. World Urbanization Prospects:the 2009 Revision. United Nations

Population Division. United Nations, Department of Economic and Social Affairs,
New York, http://esa.un.org/unpd/wup/index.htm
UNOSAT, 2010. Institute for Training and Research (UNITAR) Operational Satellite
Applications Programme. http://unosat.web.cern.ch/unosat/
UN-SPIDER, 2010. United Nations Platform for Space-based Information for
Disaster Management and Emergency Response - UN-SPIDER. http://www.unspider.org/
USGS, 1996. HYDRO1k Elevation Derivative Database. U.S. Geological Survey,
Sioux Falls, South Dakota, USA. http://eros.usgs.gov/#/Find_Data/
Products_and_Data_Available/gtopo30/hydro
USGS, 1997. GTOPO30. Global 30-Arc-Second Elevation Data Set. U.S. Geological
Survey, Sioux Falls, South Dakota, USA. http://gcmd.nasa.gov/records/
GCMD_GTOPO30.html
USGS, 2010. National Geological Map Database (NGMB). United States Geological
Survey. http://ngmdb.usgs.gov/
Ushahidi, 2010. Ushahidi http://www.ushahidi.com/
Van Asch, T.W.J., Malet, J.P., van Beek, L.P.H., Amittrano, D., 2007. Techniques,
issues and advances in numerical modelling of landslide hazard. Bulletin de la
Societe Geologique de France 178, 6588.
Van Beek, L.P.H., van Asch, T.W.J., 2004. Regional assessment of the effects of
land-use change and landslide hazard by means of physically based modeling.
Natural Hazards 30(3), 289304.
Van der Knijff, J.M., Younis, J., De Roo, A.P.J., 2010. LISFLOOD: a GIS-based
distributed model for river-basin scale water balance and flood simulation.
International Journal of Geographical Information Science 24(2), 189212.
Van Westen, C.J. (Ed.), 2009. Distance Education course on the use of spatial
information in Multi-hazard risk assessment.
Van Westen, C.J., Castellanos Abella, E.A., Sekhar, L.K., 2008. Spatial data for
landslide susceptibility, hazards and vulnerability assessment: an overview.
Engineering Geology 102(34), 112131.
Van Westen, C.J., Montoya, L., and Boerboom, L. 2002. Multi-Hazard Risk
Assessment using GIS in urban areas: A case study for the city of Turrialba,
Costa-Rica, In: Proc. Regional workshop on Best Practise in Disaster Mitigation,
Bali, 120136.
Verdin, K.L., and Greenlee, S.K., 1996. Development of continental scale digital
elevation models and extraction of hydrographic features. In: Proceedings, Third
International Conference/Workshop on Integrating GIS and Environmental
Modeling, Santa Fe, New Mexico, January 21-26, 1996. National Center for
Geographic Information and Analysis, Santa Barbara, California.
Verelst, L., 1999. Recommendations for Geo-referencing EM-DAT Data Base,
Consultancy Report prepared for the Asia FIVIMS Project and the CRED, Rome.
Vigni, C., et al., 2005. Insight into the 2004 SumatraAndaman earthquake from
GPS measurements in southeast Asia. Nature 436, 201206.
Villagran de Leon, J.C., 2006. Vulnerability: A Conceptual and Methodological
Review. UNU-EHS. UNU. No 4/2006.
Virtual Disaster Viewer (2009). Virtual disaster viewer. ImageCat. http://
virtualdisasterviewer.com
Wadge, G., Wislocki, A.P., Pearson, E.J., 1993. Spatial analysis in GIS for natural
hazard assessment. In: Goodchild, M.F., Parks, B.O., Steyaert, L.T. (Eds.),
Environmental Modeling with GIS. Oxford University Press, New York, pp.
332338.
Wald, D.J., Earle, P.S., Quitoriano, V., 2004. Topographic Slope as a Proxy
for Seismic Site Correction and Amplification. EOS. Trans. AGU 85(47),
F1424.
Wald, D.J., Quitoriano, V., Heaton, T.H., Kanamori, H., Scrivner, C.W., Worden, B.C.,
1999. TriNet "ShakeMaps": Rapid generation of peak ground-motion and
intensity maps for earthquakes in southern California. Earthquake Spectra 15(3),
537556.
White, D.H., Walcott, J.J., 2009. The role of seasonal indices in monitoring and
assessing agricultural and other droughts: a review. Crop and Pasture Science
60, 599616.
Whitehouse, G., 2001. Community involvement in flood and floodplain
management: the Australian Scene. Best practice guidelines for floodplain
management in Australia. Draft report. Canberra. 17 p.
Wilhite, D.A., 2000. Drought: A Global Assessment. Routledge Publishers, U.K,
Natural Hazards and Disasters Series. World Bank, 2010. GDP per capita. http://
data.worldbank.org/indicator/NY.GDP.PCAP.CD
World Bank, 2011. Population data. http://data.worldbank.org/indicator/SP.POP.TOTL
Xie, M., Tetsuro, E., Zhou, G., Mitani, Y., 2003. Geographic Information Systems
based three-dimensional critical slope stability analysis and landslide hazard
assessment. Journal of Geotechnical and Geoenvironmental Engineering
129(12), 11091118.
298
Yeh, C.H., Loh, C.H., Tsai, K.C., 2006. Overview of Taiwan Earthquake Loss
Estimation System. Natural Hazards 37(1-2), 2337.
Zerger, A., Smith, D., 2003. Impediments to using GIS for real-time disaster
support. Computers, Environmental and Urban Systems 27, 123141.
Zheng, J., Wang, W.C., Ge, Q., Man, Z., Zhang, P., 2006. Precipitation Variability
and Extreme Events in Eastern China during the Past 1500 Years. Terr. Atmos.
Ocean. Sci. 17(3), 579592.
Zhou, Q., Liu, X., 2004. Analysis of errors of derived slope and aspect related to
DEM data properties. Computers & Geosciences 30(4), 369378.
Zuccaro, G., Cacace, F., Spence, R., Baxter, P., 2008. Impact of explosive eruption
scenarios at Vesuvius. Journal of Volcanological and Geothermal Research
178(3), 416453.
Biographical Sketch
Dr. Cees Van Westen is an associate professor in the Department of Earth Systems Analysis, and on the Faculty of
Geo-Information Science and Earth Observation (ITC), University of Twente, the Netherlands. He obtained his
M.S. in Physical Geography from the University of Amsterdam (1988) and his Ph.D. in Engineering Geology from
the Technical University of Delft (1993), with emphasis on geographic information systems for landslide hazard
zonation. Since 2005, he is the Director of the United Nations University - ITC School on Geo-information for
Disaster Risk Management. He received the ITC research award in 1993 and the Richard Wolters Prize of the
International Association of Engineering geology (IAEG) in 1996. He is specialized on topics related with the use
of spatial information for landslide hazard and risk assessment, participatory GIS for flood-risk assessment,
volcanic-hazard assessment, seismic hazard and risk assessment, technological risk assessment, and multi-hazard
risk assessment. Most of his research is in the field of landslides, dealing with topics including generation of eventbased landslide inventories using remote sensing, historical records and field mapping, combination of heuristic
and statistical models for landslide susceptibility analysis, dynamic modelling of landslide initiation, landslide
run-out analysis, and different approaches for landslide-risk assessment. He has published more than 45 papers in
ISI journals, 10 book chapters, and numerous papers in conference proceedings. He has been involved in many
projects funded by the EU (FP6, FP7), World Bank, ADB, Dutch government, and US-AID. He is currently the
project coordinator of the CHANGES project, an EU FP7 Marie Curie International Training Network.
3.11 Geovisualization
MJ Smith, Kingston University, Surrey, UK
JK Hillier, Loughborough University, Loughborough, UK
J-C Otto and M Geilhausen, University of Salzburg, Salzburg, Austria
3.11.1
3.11.2
3.11.2.1
3.11.2.2
3.11.2.3
3.11.3
3.11.3.1
3.11.3.2
3.11.3.3
3.11.3.3.1
3.11.3.3.2
3.11.3.4
3.11.4
3.11.4.1
3.11.4.2
3.11.4.3
3.11.4.4
3.11.4.5
3.11.5
3.11.5.1
3.11.5.1.1
3.11.5.1.2
3.11.5.2
3.11.5.2.1
3.11.5.2.2
3.11.5.2.3
3.11.6
References
Introduction
Background
Historical Context
Geomorphology and Geovisualization
Applications and Emergent Technologies
Visual Processing
Detection of Landforms
Image Enhancement
Enhancement of DEMs
Regionalresidual separation
Land-surface parameters
Recommendations for Terrain Visualization
Visual Interaction
Display
Digitization and Overlay
2D to 2.5D in Space
3D in Space
Virtual Globes
Visual Outputs
Geomorphological Maps
Legend systems
Map design
Digital Mapping
Open standards
GeoPDF
Principles of web mapping and WebGIS
Conclusions
Glossary
ALS Airborne laser scanning is used to collect highresolution data for generation of a digital surface or digital
terrain model.
Bathymetry Depth of the Earths surface in water-covered
areas.
Decluttering Regionalresidual separation performed on
a digital surface model with the intention of removing
superficial clutter to leave bare earth heights for the
generation of a digital terrain model.
DEM A digital elevation model refers to any digital
representation of the topography that characterizes the
altitude field. See also DTM and DSM.
300
301
301
303
304
305
306
307
309
309
310
312
313
313
313
314
315
315
316
316
316
316
318
319
319
319
321
322
Detectability How readily a landform may be

distinguished or correctly identified by an interpreter in an
image that is displayed. The concept may be extended to
automated approaches for identification and mapping of
landforms.
DSM A digital surface model represents the highest, upper
surface of all visible objects on the landscape, as recorded
using a particular measurement technique.
DTM A digital terrain model is a representation of the
Earths surface free from anthropogenic features such as
buildings and vegetation.
Elevation Elevation is the height above a reference datum,
historically sea level.
Smith, M.J., Hillier, J.K., Otto, J.-C., Geilhausen, M. 2013. Geovisualization.

In: Shroder, J. (Editor in Chief), Bishop, M.P. (Ed.), Treatise on
http://dx.doi.org/10.1016/B978-0-12-374739-6.00054-3
299
300
Geovisualization
Geomorphology The study of the form and processes that

shape the Earth or other planetary surfaces, and how they
have evolved through time.
Geovisualization This generically refers to the depiction
of spatial data. In geomorphology, it facilitates geospatial
analysis and processing. See the International Cartographic
Association (ICA) Commission on Geovisualization
(http://geoanalytics.net/ica/).
Heuristics Using experience-based rules to solve a
problem.
Hyperspectral imagery Imagery produced from an
instrument with sensors sufficient to create observations
across a continuous region of the spectrum for each pixel.
Kernel In image processing, the kernel refers to the width
of a spatial filter and is synonymous with the window size.
Landform A shape or morphology of a particular part of
the landscape that is identified as a distinct spatial entity.
Multispectral imagery Imagery created from an
instrument with sensors sensitive to different regions of the
electromagnetic spectrum from a relatively small number of
spectral bands that exhibit a relatively large bandwidth.
Panchromatic (image) A single image generated from a
sensor that records the entire region of the visible spectrum.
Point cloud Clustering of point data in a 3D coordinate
space. Data consist of irregularly spaced (x,y,z) triplets that
can be used to generate a DSM or DTM.
Raster A data format where attributes are represented in
tessellations that usually form a regularized spatial grid or
layer. For satellite imagery and DEMs, reflectance and

surface elevation, values are stored in each grid cell.
Raw data Data as measured from sensor systems,
although significant processing has usually been applied
before presentation to the user community.
Regionalresidual (relief) separation (RRS) Numerical
operation to isolate a layer of spatially varying vertical
thickness related to features of interest from a DEM.
Signal-to-noise ratio Ratio of information about the
landforms of interest (i.e., topographic signal) to all other
features in an image or DEM. The ratio may be increased by
filtering the data before viewing, or by how the data are
displayed.
Sounding Measurement of depth. See bathymetry.
TLS Terrestrial laser scanning is used to produce highresolution digital elevation models. It represents a fieldbased approach.
Topographic signal The presence of landforms of interest
in a particular study as expressed in datasets such as height.
Topography The shape of the Earths surface. This can
include landscape features such as vegetation and manmade features.
Visual processing The act of visual exploration of the
terrain or landscape for the purpose of landform
identification, or in order to generate other derived
observational data sets.
VNIR The visible and near infra-red regions of the
electromagnetic spectrum.
Abstract
Geovisualization involves the depiction of spatial data in an attempt to facilitate the interpretation of observational and
simulated datasets through which Earths surface and solid Earth processes may be understood. Numerous techniques can
be applied to imagery, digital elevation models, and other geographic information system data layers to explore for patterns
and depict landscape characteristics. Given the rapid proliferation of remotely sensed data and high-resolution digital
elevation models, the focus is on the visualization of satellite imagery and terrain morphology, where manual human
interpretation plays a fundamental role in the study of geomorphic processes and the mapping of landforms. A treatment of
some techniques is provided that can be used to enhance satellite imagery and the visualization of the topography to
improve landform identification as part of geomorphological mapping. Visual interaction with spatial data is an important
part of exploring and understanding geomorphological datasets, and a variety of methods exist ranging across simple
overlay, panning and zooming, 2.5D, 3D, and temporal analyses. Specific visualization outputs are also covered that focus
on static and interactive methods of dissemination. Geomorphological mapping legends and the cartographic principles for
map design are discussed, followed by details of dynamic web-based mapping systems that allow for greater immersive use
by end users and the effective dissemination of data.
3.11.1
Introduction
Earth scientists aim to study and understand the Earths

dynamics systems. The interaction of atmospheric, geologic,
and ecological processes greatly influences surface processes and topographic evolution. Geomorphological studies focus on understanding various forcing factors, process
mechanics, and the role of surface processes in landscape
evolution. Fundamental to geomorphology is the linking
of process to form and the identification of landforms
that have genetic, morphological, and chronological
characteristics. Therefore, geomorphology is fundamentally
concerned with geomorphological mapping of various aspects of the system including process domains, materials
and structure, morphology, landforms, and chronology.
Visualization of the landscape is a primary means to study
geomorphological conditions, and this is accomplished
through the use of field studies and Earth observation that
focus on collecting spatio-temporal land-surface information.
To study landforms, however, it is necessary to be able to
characterize the topography and specifically identify various
landforms and features/conditions that are the result of specific processes, feedback mechanisms, or system couplings.
Consequently, it is essential to collect such data and analyze
Geovisualization
patterns and trends so that new insights, relationships, and

understandings advance the discipline. Quantitative analysis
of data can be used and numerous analytical spatial and
temporal techniques exist (e.g., geostatistics and Fourier analysis). The power of the human visualization system must
also be considered, as accurate, detailed and reliable visualization can significantly augment and facilitate the interpretation of quantitative analysis results, and can also be of
primary importance to the geomorphological study of processes shaping the Earth and other planetary landscapes, given
that completely automated spatial analysis and mapping
models are still being developed and evaluated for a wide
range of geomorphological applications.
In geomorphology, the shape of the land is so fundamental, that although various forms of spatial data exist to
characterize different aspects of the land surface, topographic
information is of central importance to geovisualization.
Other forms of landscape information such as imagery, biophysical surface properties, land cover, and land use, as well as
other thematic information can easily be integrated with
topographic information to facilitate effective geovisualization
of the landscape. Consequently, many techniques for the extraction of information from spatial data sets also represent
visualization techniques. This includes the extensive and
complex range of image processing and spatial analysis,
techniques that can be applied to a variety of datasets, although they may not be visually displayed in an effective
manner.
Therefore, the scope of this chapter focuses on visual processing, visual interaction and visual outputs, all of which
form part of the geovisualization of terrain. This chapter does
not cover all aspects or techniques of scientific visualization for
a variety of different datasets, but instead focuses on the use of
spatial data (e.g., imagery and topography) and geovisualization for geomorphological studies and mapping. It provides
an historical perspective, formal definitions of geovisualization
and an outline of emergent technologies and applications. The
chapter then addresses visual processing and the optimization
of imagery for landform visualization. The mechanics of various methods are then examined by which a user can interact
with data, given that it has been prepared for viewing. Finally,
a discussion follows on the dissemination of data in both
traditional paper and electronic formats, highlighting various
visualization products.
3.11.2
3.11.2.1
Background
Historical Context
The visual interpretation of data and numerous conceptual

ideas have been a key aspect of human understanding
(e.g., Kraak and Ormeling, 2010). Historically, maps have
commonly been created to assist in interpretation and understanding, as attribute information is linked to spatial coordinates or location, depicting spatial patterns. Figure 1(a)
illustrates the utility of creating maps for the investigation of
geomorphic processes by displaying the distribution of glacial
landforms and their composition (indicated by the symbol fill
color), something that is difficult to convey using vertical
301
Figure 1 Examples of common methods of visualization, Bowridge,

Scotland. (a) Geomorphological map of glacial landforms, utilizing
Ordnance Survey 1:25 000 base mapping with 25-ft contour lines.
Reproduced from Google Maps. (b) Aerial photo (A notes the location
of Bowrdige drumlin depicted in (a) and (c). Reproduced from Rose,
J., Smith, M.J., 2008. Glacial geomorphological maps of the Glasgow
region, western central Scotland. Journal of Maps v2008, 399416,
with permission from JOM. Note the photo is as composite as
presented in Google EarthTM). (c) Oblique photo (J. Rose). The arrow
indicates the approximate photo orientation.
(Figure 1(b)) or oblique (Figure 1(c)) aerial photos. Alongside

maps, alternative portrayals of spatial data have been routinely
used (e.g., Kraak and Ormeling, 2010; Bonham-Carter, 1994)
including field sketches (Figure 2), conceptual diagrams
(Figure 3), vertical/oblique aerial photos (Figure 1(b) or 1(c)),
satellite imagery, digital elevation models (DEMs), and augmented reality (e.g., Reitmayr et al., 2005).
Early geomorphological maps were generally produced for
military and engineering purposes and designed for use in the
field (Klimaszewski, 1982). With the onset of a morphological
view of landscapes and the description of their physiography
at the turn of the twentieth century, landform analysis
and morphological description became new objectives for
geomorphological mapping (e.g., Passarge, 1912, 1914). From
this standpoint, maps were more than just a medium for
visualization but represented a research tool for landscape
analysis providing a generalized inventory of landforms, surface structures, geomorphological processes, surface and subsurface materials, and genetic information. The applications
302
Geovisualization
Narrow
rock
outcrop
10 m high, 45
bare soil slope
Level vegetated
rock shelf, 1.5 m
deep
2 m high rock face

B
Appr
ox. 8
Back and side

slopes of basin
all about 45
m
s,
Rocky step
ing plane
dd
be
t
ns
ai
ag
2m
Distance AB approx. 18 m
A
Figure 2 Field sketch of a slope profile highlighting morphological features on the eastern slope of Sg. Belalong, northwest Borneo. Reproduced
from Dykes, A.P., 1995. Hydrological controls on shallow mass movements and characteristic slope forms in the tropical rainforest of
Temburong District, Brunei, Department of Geography. Unpublished PhD thesis, Kings College, University of London.
of geomorphological maps range from simple descriptions of

a field site (e.g., accompanying a journal publication or construction site report), to land system analyses (Bennett et al.,
2010), land surveys, land management, or natural hazard
assessment (Brunsden et al., 1975; Seijmonsbergen and de
Graaff, 2006).
Maps remained the fundamental geomorphological output
through to the 1980s, as they provided both 2D visualization
and an effective data storage paradigm for spatial data. Geomorphological mapping subsequently declined due to a preoccupation with cartographic symbolization and a move to
field-scale experimentation. Since the 1990s, widely available
remotely sensed data, progress in computing power, and
rapid improvements of geographic information systems (GIS;
e.g., Wessel and Smith, 1998; ESRI, 2003) have permitted the
combination of field-scale and regional approaches that have
led to a resurgence in mapping. In particular, the emergence of
GIS through the wider field of geographic information science
or geomatics has provided software tools that can be used to
store, manipulate, analyze, and display spatial data. The latest
developments are also permitting more temporal analysis
capabilities. Such information technology provides a powerful

methodology to combine diverse datasets including totalstation data, global positioning system (GPS) measurements,
satellite imagery, postcodes, and historic data forms. Specifically, GIS can be used to prevent data overload, facilitating a
clear and carefully constructed approach for analysis, interpretation, and higher level use of the data and information.
Historically, geovisualization has been a paper-based analog technique; however, it now involves the routine collection
of large digital datasets, detailed analytical analysis, and
computer-based geovisualization. The largest datasets are
generated by sensors aboard mobile platforms, and numerous
datasets can be used for geomorphological applications
(Smith and Pain, 2009). In terms of data volume, satellite
imagery remains the single most important product, although
the issues associated with spectral, spatial, and temporal
resolution can be problematic for specific applications. In
terms of impact upon geomorphology, however, the generation of DEMs has arguably had the largest influence. This is
profound because a variety of elevation data can be synthesized or fused into a DEM and, once in raster format,
Geovisualization
Rock face with rockfall

Degraded sideslope Talus/scree
of former river valley
Taluvium
Taluvium cover to
weathered bedrock
2
Colluvium
Colluvium
High altitude glacial

and periglacial
Degraded debris slide
1
Boulder field
Ancient high level terrace

or erosion surface
Incised gully
Active lower slopes
2
Rotational slide
Tributary
valley
3
4
303
Principal
valley
5
4
Floodplain deposits
Residual soil or
weathered rock
Debris slide
Braided channel
with bar deposits
Rockslope
Rapids
Fan
Alluvium
Young river
terrace deposits
Figure 3 Conceptual block diagram of high-mountain environments showing landforms and processes of five major zones: 1 high-altitude
glacial and periglacial; 2 free rock faces and associated slopes; 3 degraded middle slopes and ancient valley floors; 4 active lower slopes;
and 5 valley floors. Reproduced from Fookes, P.G., Sweeney, M., Manby, C.N.D., Martin, R.P., 1985. Geological and geotechnical engineering
aspects of low-cost roads in mountainous terrain. Engineering Geology 21, 1152.
image-processing or spatial analysis techniques, common in

geovisualization, can be applied. Contemporary mapping is
therefore computer based, reliant upon the input of digital
datasets, with analysis and output performed using GIS.
3.11.2.2
Geomorphology and Geovisualization
Geovisualization is widely utilized in geomorphology for the

exploration and analysis of spatio-temporal data. Although a
spatial framework is not a requirement for geomorphological
study (e.g., Smith and McClung, 1997), it is natural in many
studies to use space as the organizing paradigm (e.g., Benetti
et al., 2010). A better understanding of many geomorphic
phenomena can therefore be gained through the recording
(i.e., mapping) and analysis of their spatial distribution. In
the past, the observed distribution and form (i.e., morphology) would typically have been communicated using a
geomorphological map (e.g., Rose and Smith, 2008).
The term geovisualization is a contraction of geographic
visualization. This was first mooted by MacEachren et al.
(2004), who defined geovisualization as a process for leveraging y data resources to meet scientific and societal needs
and a research field that develops visual methods and tools
to support a wide array of geospatial data applications.
This definition extends geovisualization beyond simply
communication using the presentation of an image or images.

The definition proposes that the term visualization include
four primary functions (Figure 4): (1) exploration of datasets
in order to find patterns and information; (2) analysis of
patterns and relationships between data; (3) synthesis, involving the generation of an overview and understanding of
the inherent information; and (4) presentation of the findings. Within this definition, geovisualization incorporates a
wide gamut of activities and capabilities. In contrast, Kraak
(2008) suggested that the term geovisualization was being
used increasingly widely and indiscriminately, becoming
synonymous with mapping, and therefore increasingly less
useful as a term. Consequently, he favors the use of the term
geovisual analytics (Thomas and Cook, 2005; Andrienko
et al., 2007) for the range of activities in and around the
geovisualization cube, implicitly retaining a more focused
definition of geovisualization. Discussion exists about the
terminology surrounding geographic visualization; yet, at its
simplest, geovisualization is simply a synthesis of the longdeveloped visual communication of cartography with current
digital analytical technologies, principally GIS. Indeed, it
could be argued that in the past cartography was geovisualization; that is, cartography embodied the sum of geovisual
techniques that were possible before practical, pervasive
desktop computing. The introduction of computer technologies in the 1960s, however, saw a split in the discipline of
Geovisualization
ts
s
eciali
... sp
Info
.
sha ..task
ring ...kn
con owle
stru dge
ctio
n
sers
c ... u
Publi
- Explore - analyze - synthesize - present -
304
n
ctio
...
low
a
ter
.. in
.
igh
Figure 4 Four functions of geovisualization modified from

MacEachren, A.M., Gahegan, M., Pike, W., Brewer, I., Cai, G.,
Lengerich, E., Hardistry, F., 2004. Geovisualization for knowledge
construction and decision support. Computer Graphics and
Applications 24(1, 24), 1317, that include exploration, analysis,
synthesis, and presentation. Geovisualization feature-space is defined
by the types of task, user, and interactivity.
geographic visualization with users either focused principally upon design and communication, or data handling.
The former we now think of as cartography, whereas the
latter became geographic information science. The meaning
of the term geovisualization is therefore debated, but in its
widest, intuitive, sense it is the visual depiction of spatial data
and information. It is a convenient term to employ within
geomorphology and is used here with this latter broad
meaning.
3.11.2.3
Applications and Emergent Technologies
Some applications of geovisualization are depicted in

Figure 4, with the axes of the cube illustrating the task being
performed, the user doing the task, and the degree of interaction with the data being visualized. Within this, the four
functions proposed by MacEachren et al. (2004) move sequentially from highly interactive exploration of the data by
specialists to generate basic observational knowledge, to presenting synthesized information to the public involving
little interaction with the data. Knowledge construction at the
start of this sequence requires a specialist with a high degree
of interaction, and can be considered a research-intensive
application. At the opposite end of the spectrum, information
sharing generally requires lower levels of interaction by
non-specialists, and is an application for the public and decision makers. This latter area is now an important aspect of
research, as funding bodies are aware of their accountability,
and therefore desire further downstream application, as well
as interaction with the general public or, more generally,

public relations.
Two trends not well represented by the four functions
proposed within the geovisualization cube are the growth in
both public interaction with data and data distribution to the
widest possible audience. The latter facilitates the former, and
with both specialists and the public exploring data, the face of
the cube representing high levels of interaction is becoming
increasingly occupied. Specifically, there are three requirements for this occupation to be achieved:
freely or easily available data;

nonproprietary or open formats for geospatial data; and
free or easily available software to download, visualize, and
analyze data.
The public release of data is not new (e.g., SYNBAPS; Van

Wyckhouse, 1973), but there has been increasing political
pressure to do so. In the USA, the National Geophysical Data
Center (http://www.ngdc.noaa.gov/) makes data publically
available, and it is a requirement of funding that data be released. The UK government has released significant quantities
of data (http://data.gov.uk), and Research Councils UK also
require scientists to place their data in repositories. Data
sharing and archival requires a move away from proprietary
data formats, and this has been achieved through the development of industry-standard, open, formats (e.g., Keyhole
Markup Language (KML); see Section 3.11.5.2.1) and openly
specified proprietary formats (e.g., the shapefile format; ESRI,
2003).
The ability to select base data and easily overlay specialist
datasets is fundamental across a spectrum of activities, ranging
from the development to the use of software applications. For
example, Generic Mapping Tools (Wessel and Smith, 1998),
GRASS and QGIS allow scripting to enable advanced users to
create wider dynamic access to data. The Seamount Catalog
(http://earthref.org/SBN/; Koppers et al., 2010) exemplifies a
scripted front end to geospatial data. For both technical and
nonspecialist users, GeoMapApp (http://www.geomapapp.org/;
Carbotte et al., 2004) is a free, easy-to-use software package that
contains datasets, displays (e.g., maps and profiles), and permits importing and overlays. Google Earth performs similar
tasks in an accessible way, encouraging map-making by the
public at large. The public use of geovisualization itself is no
better displayed than in the culture of mashups (e.g., Wood
et al., 2007), and by the ease of use of a site that allows anyone
to make maps (e.g., https://arcgis.com).
Emergent technologies are driving research capabilities in
geomorphology and are being developed to meet new challenges. Innovations in on-screen visualization and public engagement have led to the development of technologies that
allow existing forms of display to be added to, or augmented.
By enhancing visualization, augmentation presents new opportunities for researchers to interact with, and demonstrate,
their work. This is particularly useful for field trips, interaction
with the general public, and consultancy work. Augmentation
takes a base information layer, such as a real-time view of a
landscape, and allows users to overlay specific information on
top. This information could be pre-identified landforms,
outlines of historic spatial extents, or information on the
process that formed them and the material they consist of.
Geovisualization
The base layer is commonly a live video stream, for

which the location and orientation of the camera must be
known. The technology is now well developed and implemented on many smartphones. For example, Layar (http://
www.layar.com/) allows the development of data layers that
can be made available to end users. This technology is known
as augmented reality.
In addition to augmented reality, geomorphologists have
new research opportunities by using high-resolution satellite
imagery. For example, the development of low-altitude imaging systems (e.g., Hardin and Jensen, 2011) allow for the
capture of submeter resolution data (o10 cm). A variety of
innovative platforms are available (e.g., Smith et al., 2009;
Laliberte and Rango, 2011) that enable geomorphologists to
examine the landscape at high spatial and temporal frequencies. Likewise, terrestrial-laser scanning (TLS; Heritage and
Large, 2009) can be used to dramatically increase the data
volume and precision of topographic data. Examples of new
approaches based upon such data include Williams et al.
(2011) who outline methods for monitoring channel changes
to gravel-bed rivers before and after flood events, and Pesci
et al. (2011) who monitor landslide deformation using TLS
data over three epochs. Perhaps the greatest excitement has
been generated around real-time sensors that relay measurements of physical processes as they occur. For example, Hart
et al. (2006) utilized subglacial sensors to monitor changes to
till through time. Linked clusters of sensors, or sensor networks, are a powerful methodology to remotely monitor
processes at very high spatial and temporal resolutions, particularly in areas that are hazardous or remote (http://glacsweb.org/). Importantly they have mission-critical capabilities,
where communication and data storage are replicated across
probes so that sensor failure does not cause network failure.
Overall, the increasing amounts of data, both spatially and
temporally, present new challenges for geovisualization where
high data storage and multidimensional datasets require innovative solutions.
3.11.3
Visual Processing
Geomorphologists are interested in the geovisualization of

spatial data in order to study and understand the Earths surface. Given their objectives, they generate an observational
data set that provides them with information that they can
analyze, and one that facilitates the identification of landforms. The observational data are representations of the location and characteristics of the landscape, and representation
of the topography may be termed digital terrain modeling
(DTM).
An intermediate output product of terrain analysis (i.e.,
geomorphometry) may be an enhanced data layer for further
visual inspection, analysis, and/or numerical modeling. Ultimate output products are typically maps for distribution,
analysis, and archival. Although researchers may conduct
fieldwork for geomorphological mapping, we only consider
terrain analysis using VNIR (visible and near infra-red) imagery and topographic information in this section. Specifically, the concepts of topographic signal (Hillier and
Smith, 2008) and detectability (Smith and Wise, 2007) are
305
introduced. These concepts can be used as a basis for evaluation of image processing and terrain analysis techniques
that enhance visibility of specific landforms by increasing the
ratio of (topographic) signal to noise of an image. The section concludes with recommendations for enhancement and
cautions about limitations in geomorphological mapping
using digital data.
The process of landform identification can be performed
using manual or automated/semi-automated techniques.
Manual mapping techniques require a skilled interpreter to
identify and outline landforms of interest. Skill, how well
landforms are identified, is based upon many factors and
includes the basic expertise and experience of the analyst
(Colwell, 1983). The process is qualitative and relies upon the
development of ordered relationships between features in an
image using complex visual heuristics in order to identify
object types. Although many have focused on using panchromatic aerial photos, geomorphologists now have access to
multispectral, thermal, and radar imagery. Such data permit
better interpretation using physics (matterenergy interactions) as a basis for understanding reflectance, emission,
and backscatter variations. The basic image elements remain
the same: that is, the identification of landforms via the
assessment of tone, texture, shape, size, shadow, pattern, location, and association. For further details on the image
elements and basic image-interpretation strategies, see general
textbooks on remote sensing (e.g., Gibson and Power, 2000;
Campbell, 2007; Lillesand et al., 2008).
Automated and semi-automated mapping techniques use a
variety of algorithmic approaches to identify landforms (e.g.,
Hillier and Watts, 2004; Van Asselen and Seijmonsbergen,
2006; Seijmonsbergen et al., 2011). Such approaches have the
benefit of being numerical, repeatable, and not reliant on an
individual interpreter or systematic variation during processing. Any inherent biases tend to originate from the ultimately
subjective calibration underpinning the method. A significant
number of these approaches are primarily statistical, having
been developed from clustering methods used more generally
in remote sensing. The increasing availability of DEMs, however, has allowed the incorporation of elevation and derived
land-surface parameters, so that morphological information
can be used in the classification process. The wealth of
remotely sensed data now available has opened considerable
opportunities for developing the processing of extensive
datasets, as well as integrating them into new types of analyses. For example, Camargo et al. (2009) integrated reflectance data into a geomorphometric object-based analysis, and
this approach is likely to become more common in an attempt
to evaluate surface materials from surface spectra. Further insights can be gained through the integration of subsurface data
from active and passive sensors. At sea, the subsurface structure has been commonly probed using sound (e.g., seismic
reflection), electrical (e.g., resistivity), magnetic and gravity
fields. See Jones (1999) for further details. On land, passive
airborne systems, such as aeromagnetics (detecting subsurface
magnetic features), and active systems, such as airborne electromagnetics (3D conductivity), provide subsurface data
(Smith and Pain, 2009). The use of automated and semiautomated techniques that integrate a variety of datasets is
therefore expected to become increasingly prominent,
Geovisualization
170 000
175 000
180 000
308 000
302 000
302 000
296 000
296 000
290 000
290 000
284 000
165 000
170 000
175 000
165 000
170 000
175 000
308 000
165 000
284 000
(a) 160 000
160 000
302 000
180 000
296 000
175 000
290 000
170 000
(b) 160 000
284 000
165 000
308 000
160 000
308 000
Manual mapping requires the visual detection of individual

landforms and the recording of their spatial distribution onto
some kind of a base map. Detection may be accomplished
using aerial photography or satellite imagery, or may require
additional topographic information to delineate the boundaries of landform features. The use of specific regions of
the electromagnetic spectrum may make detection easier, as
variations in reflectance between the landform and the
surrounding landscape depends on landform composition
versus landscape composition. Detection may also relate
to surface structure variations that may be characterized by
image texture. Smith and Wise (2007) outline the concept of
detectability; that is, the ability for an interpreter to accurately
identify the true dimensions of a feature of interest. This is
dependent on the data source, image pre-processing, and enhancement, and the skill of the interpreter. Interpreter skill is a
difficult capability to quantify and can be highly variable
between individuals, although there are procedures that can
be implemented to minimize this influence. Fundamentally,
302 000
Detection of Landforms
1. Relative size. The minimum resolvable size of a landform

that is visible in an image is a function of the spatial
resolution (Figure 5). The spatial resolution must be high
enough, or the landform large enough, that it contains a
sufficient number of pixels to be recognizable. Small features may be under-represented in any population, and
where spatial clustering of small landforms exists, systematic errors in interpretation may consequently result.
2. Azimuth biasing. The orientation angle of an elongated
landform (e.g., drumlin); this causes their appearance to
differ in an image depending on the relationship between
the orientation angle and the solar geometry (Figure 6).
This becomes more pronounced with greater elongation
(Smith and Clark, 2005).
3. Landform signal strength. The amount of tonal and textural
information that is available to visually distinguish individual landforms; also termed the topographic signal
(Figure 7; Hillier and Smith, 2008). For DEMs, the signal
strength is related to altitude variation with respect to the
landforms of interest and the surrounding terrain. For
imagery, topography is inferred from associations between
landforms and land cover (e.g., Punkari, 1982) and tonal
296 000
3.11.3.1
however, the detectability of a landform will be primarily

dependent on sensor system characteristics. The following
three object characteristics in relation to sensor system characteristics combine to affect landform detectability and should
be considered by an interpreter prior to selecting data:
290 000
however, for some mapping applications, manual mapping

remains the most prevalent technique for recording the
shape and distribution of landforms. Therefore, visualization
techniques that can be used to facilitate this objective are
presented.
284 000
306
Figure 5 Illustration of the effects of resolution on the detectability of drumlins. (a) 50 m DEM. (b) 150 m DEM. Note that small landforms in
the north of the area would be difficult to detect from the 150 m DEM. Irish Grid coordinates in meters, relief shaded with a solar elevation of
451 and a solar azimuth of 451.
Geovisualization
170 000
175 000
308 000
302 000
296 000
296 000
290 000
284 000
165 000
290 000
308 000
302 000
302 000
296 000
290 000
175 000
284 000
175 000
170 000
308 000
170 000
165 000
302 000
165 000
284 000
(a) 160 000
160 000
296 000
175 000
290 000
170 000
284 000
165 000
308 000
160 000
307
(b) 160 000
Figure 6 Illustration of the effects of solar azimuth angle on the detectability of drumlins from a relief-shaded DEM. (a) Azimuth angle parallel
to the dominant drumlin orientation. (b) Azimuth angle orthogonal to the principal drumlin orientation. Arrows indicate azimuth angle. Irish Grid
coordinates in meters.
and textural variations from shadows (Slaney, 1981), particularly where solar elevation angles are relatively low.
Furthermore, synthetic-aperture radar (SAR) systems use an
oblique (rather than nadir) viewing geometry (Ford, 1984;
Vencatasawmy et al., 1998), which enhances the topographic expression in the imagery.
The minimum-resolvable plan form size for detecting
landforms can only be reduced by increasing the spatial
resolution of the data (Smith et al., 2006). Although such data
may be available, there will almost certainly be a trade-off
between financial cost and spatial coverage, and project objectives must be tailored to available resources (e.g., Punkari,
1982). For satellite imagery, routine repeat coverage from
moderate-resolution sensors (e.g., Landsat) mitigates this
somewhat, as images at optimal illumination angles and
azimuths may be available in existing archives.
Azimuth biasing is perhaps the most significant problem
for satellite imagery, although other issues associated with
landform mapping from imagery still exist. Bias is minimized
through the acquisition of imagery with relatively high solar
elevation (i.e., illumination is basically overhead); however,
this minimizes the landform signal strength so that
features are not as clearly observable (Figure 7). Smith and
Wise (2007) recommend solar elevation angles o201; however, polar-orbiting satellites operate with fixed overpass times,
which means that it is not possible to specify acquisition time
at a site. With a reduced ability to obtain ideal satellite
imagery, ephemeral landforms are commensurately more

difficult to identify and monitor. If solar geometry conditions
cannot be adequately controlled, interpreters should be aware
of the possibility of introducing systematic biases through the
exclusion of landforms suites. Azimuthal bias does not affect
DEMs unless relief shading is generated using low-angle illumination. Using both satellite imagery and DEMs, it is possible to enhance the landform signal strength to make
landforms more detectable, thereby increasing the accuracy of
manual mapping.
3.11.3.2
Image Enhancement
With the acquisition of satellite imagery and appropriate preprocessing, it is sometimes necessary to enhance the imagery
to provide the best possible visualization of landforms. Enhancement is largely based on a subset of the standard imageprocessing techniques that include spectral and spatial
enhancements that are commonly used in remote sensing
(e.g., Mather, 2004; Lillesand et al., 2008), and have proven
useful for geomorphological visualization. Clark (1997) noted
that brightness/color variations are most effectively detected
by humans, and therefore recommends the use of images
rendered in gray scale. For multispectral imagery, it has been
common practice to generate image ratio images that effectively portray spectral differences that are associated with surface-compositional variations. The geological community
Geovisualization
160 000
165 000
170 000
175 000
180 000
155 000
160 000
165 000
170 000
175 000
180 000
155 000
160 000
165 000
170 000
175 000
180 000
155 000
160 000
165 000
170 000
175 000
180 000
311 000
155 000
296 000
296 000
299 000
299 000
302 000
302 000
305 000
305 000
308 000
308 000
311 000
(a)
311 000
296 000
296 000
299 000
299 000
302 000
302 000
305 000
305 000
308 000
308 000
311 000
308
(b)
Figure 7 Illustration of the effects of landform signal strength using Landsat TM imagery acquired on contrasting dates (B1801 solar azimuth).
(a) Imagery with low solar elevation (111). (b) Imagery with high solar elevation (481). The images overlap with the top half of Figures 4, 6, 8, 9,
10, and 11. Irish Grid coordinates in meters.
commonly uses image ratios and examines false-color composites of ratio images. Standard contrast enhancements such
as linear, piece-wise, and other histogram-modifying techniques should be utilized to maximize the tonal contrast for
original imagery and spectral features such as ratios. Principal
component analysis (PCA) is another standard technique for
spectral-feature extraction that isolates the fundamental linear
variation from the entire multispectral dataset, thereby producing spectral features (principal component images) that
highlight different aspects of the landscape. The first principalcomponent image contains the majority of the information
from the original dataset. It is possible that vegetation
landform, topographylandform, and moisture landform
relationships can be ascertained by interpreting these images.
Geovisualization
3.11.3.3
Enhancement of DEMs
DEMs represent the altitude field, and they can be used to

derive morphological information and be displayed in a variety
of ways. Numerous spatial-analysis algorithms exist that characterize different aspects of the topography, and this information can be easily integrated and draped over the altitude
field. By default, most image-processing software can be used to
display DEMs as a gray-scale image with a palette of gray tones
representing altitude that linearly varies between the maximum
and minimum values in the dataset (Figure 8). Displaying the
topography in this way has its disadvantages because subtle
variations in the altitude field cannot be easily seen, and low
amplitude landform features cannot be identified against larger
scale higher amplitude relief variations in the landscape. It is
therefore necessary to focus on specific areas of the landscape
containing features of interest and enhancing the landform
signal strength to facilitate identification.
3.11.3.3.1
Regionalresidual separation
As an aid to better understanding of surface processes, landforms can be divided into classes or categories, based on some
similarity of a topographic characteristic or a priori knowledge
that they relate to a physical process. These classes of features
can constitute a component of a landscape. If H is elevation,
the decomposition of a DEM into components may be
expressed as HDEM H1 H2 yHn, where n is the number of
components, and HDEM the observed relief (e.g., Hillier, 2011).
Hn is the thickness of a component, a layer that varies spatially
in thickness; i.e., the values of H1, H2 and so on to Hn vary for
each location (x,y) of the DEM. Thicknesses are always greater
than or equal to zero. Summing the components recreates the
165 000
170 000
175 000
180 000
165 000
170 000
175 000
180 000
290 000
290 000
296 000
296 000
302 000
302 000
308 000
308 000
160 000
160 000
284 000
284 000
Finally, spatial-feature enhancement techniques can also be

applied. Examples include texture enhancement, spatial convolution (kernel) filtering, and more advanced techniques
including fractal analysis and Fourier analysis. Edge-detection
techniques and high-pass filters can be used to detect subtle
variations in reflectance or topography that may help to delineate landforms of interest. They are commonly used for
mapping lithological units and faulting, and play an important role in tectonic geomorphology.
The work of Punkari (1982) illustrates how multispectral imagery is important for identifying landforms. In
his work, inter-drumlin regions exhibited greater surface
moisture conditions. These areas were enhanced for mapping
by using those regions of the spectrum that were most sensitive to moisture conditions (i.e., NIR, SWIR), thereby permitting the identification of drumlins. Other techniques
could also be used including the use of the wetness index and
the principal component image that enhances surface moisture conditions. In a second example, Jansson and Glasser
(2005) used a simple false-color composite to effectively
integrate near infra-red, thermal infra-red, and topographic
information (shaded relief) to permit landform identification. With respect to topographic information, it is important to note that many image enhancement algorithms
can also be applied to terrain parameter images generated
from terrain analysis.
309
Figure 8 A DEM gray-scale image. Irish Grid coordinates in meters.
terrain. Isolating landforms into a component of the landscape

so that they may be studied independently is termed regionalresidual (relief) separation (RRS). Wessel (1998) coined the
term as applied to DEMs, and Hillier and Smith (2008)
introduced it to subaerial geomorphology. The approach is
more difficult to apply where classes of landform are morphologically similar to each other or to the underlying
topography.
Figure 9 illustrates one RRS method that was used to
isolate drumlins and hills from one another. Other application examples of RRS include quantifying the large-scale
subsidence trend across the oceans to understand how tectonic plates evolve (e.g., Sclater et al., 1975; Calcagno and
Cazenave, 1994), isolating underwater volcanoes (White,
1993; Hillier and Watts, 2007), removing trees and other
surface features to declutter digital surface models and create
DTMs for tasks such as flood modeling (e.g., Sithole and
Vosselman, 2004; Mason et al., 2007), or archeological prospecting (Hillier et al., 2007; Hesse, 2010). It is important to
note that the spatial components are not terrain units that
exhibit similar properties such as those portrayed in a conventional geomorphological map.
By convention, the regional component contains features
of a larger width-scale than the residual components because,
in older methods (e.g., Watts, 1976; McKenzie et al., 1980;
Hillier and Smith, 2008), the altitude trend representative of a
region was calculated first and then subtracted from the elevation to obtain the residual. Alternatively, it is possible to
directly isolate (i.e., identifying and determining spatial limits
for) all individual features of a certain type within landscapes
(e.g., Hillier and Watts, 2004; Hillier, 2008) leaving the regional trend. But, the same naming convention is still used.
310
Geovisualization
Elevation (m)
(a)
50
North
20 000
South
Elevation (m)
30 0000
100
0
(c)
Low
Relative
elevation
290 000
20
10
0
10
(b) 20
Relative elevation
150
10 000
High
10 000
20 000
Distance along profile (m)
(d)
160 000
170 000
180 000
Figure 9 Regionalresidual separation (RSS) applied to a profile across the DEM (Hillier and Smith, 2008). (a) 1 1 km median sliding-window
filter is used to approximate regional relief (hills; regions indicated by the bold line). (b) Subtraction of regional relief from the DEM to leave
topographic signal (drumlins; thin line). Envelope (gray shade) is based upon filters returning maximum and minimum residual values within a
1 1 km kernel. (c) Normalization of heights, equivalent to a local contrast stretch, so that landforms of differing amplitude may be effectively
displayed using a single color palette. (d) Display of relative elevations using gray scale. Irish Grid coordinates in meters. Profile is along
white line.
The skill in performing a successful RRS is always determining

a property that makes the features you wish to isolate distinctively different.
The earliest RRSs were manual (e.g., Menard, 1973; Sclater
et al., 1975), and the quantitative differences among the
components were not known. Computationally, early methods used efficient frequency-domain filters (e.g., Watts and
Daly, 1981; Cazenave et al., 1986) to obtain the regional
component (e.g., 4004000 km wavelength topography), and
high-pass filters to emphasize the small-scale residual. This
approach depends upon nonoverlapping spatial frequencies.
This so-called spectral overlap in the frequency domain is
therefore a problem (Wessel, 1998).
Alternatively, a valid regional elevation can be estimated using low-pass spatial filtering (e.g., Watts, 1976;
McKenzie et al., 1980; Wessel and Smith, 1998). If the DEM
contains many statistical outliers (e.g., anomalously high
points), bias can occur in the regional (Smith, 1990). Use of
other, robust statistics such as the median and mode can
mitigate this substantially (Smith, 1990; Crosby et al., 2006;
Kim and Wessel, 2008), as can iterative statistical approaches
(Marty and Cazenave, 1989; Wang et al., 2001).
These older techniques are good first approximations, and
remain useful in some situations (e.g., Hillier and Smith,
2008). The main difficulties include:
Spectral overlap between classes of landform (e.g., Wessel,

1998).
Spatial overlap of landforms.
High spatial density of smaller landforms obscuring the
regional trend (e.g., Smith, 1990).
Complex high-amplitude regional trends.

Widely ranging shapes and sizes within a class of feature
(e.g., seamounts, drumlins).
A number of complex algorithms have focused on overcoming these problems to create a regional digital surface model
when de-cluttering LiDAR data (e.g., Sithole and Vosselman,
2004). For geomorphic features, several algorithms also exist
that identify the smaller features directly (Hillier and Watts,
2004; Hillier, 2008). These seek to avoid the problems experienced by the older approaches above, and have the advantage
that features are implicitly mapped during the separation but,
like the de-cluttering algorithms, are somewhat task specific.
Lastly, if the locations of features are known a priori, this information may be used (e.g., Smith et al., 2009). Once a component is isolated through RRS, the image can be enhanced and
displayed to facilitate landform mapping.
3.11.3.3.2
Land-surface parameters
To facilitate landform mapping, land-surface parameters

(LSPs) (Hengl and Reuter, 2008) are commonly used, as they
characterize unique morphometric properties of the topography. These may then be displayed separately, or in unique
combinations to identify landforms and delineate their
boundaries. Such parameters may represent physical attributes
of the topography, components of biophysical variables, or
empirical proxies related to potential or position.
Relief shading is the most common visualization technique
that can be used to enhance the topography (Kraak and
Ormeling, 2010). The landscape is illuminated from a specific
solar azimuth and elevation angle, and one can assume
Geovisualization
296 000
290 000
284 000
165 000
170 000
175 000
180 000
308 000
308 000
302 000
302 000
296 000
290 000
180 000
302 000
180 000
175 000
296 000
175 000
170 000
290 000
170 000
165 000
(b) 160 000
284 000
165 000
284 000
(a) 160 000
160 000
308 000
180 000
302 000
175 000
296 000
170 000
290 000
165 000
308 000
160 000
used as an LSP for mapping, as many landforms have steep or

shallow slopes, although users generally find it difficult to
interpret and must become familiar with the topography. For
visualization purposes, gray-scale display of a slope image is
best when flat areas are represented using light gray tones,
giving an image a shaded-relief quality that does not exhibit
azimuth biasing (Figure 10(a)). Terrain curvature is a measure of the rate of change of gradient, and may be quantified
as planform, profile, and tangential curvature (Schmidt
et al., 2003). Profile curvature measures downslope curvature
(the derivative of gradient) and highlights breaks in slope
(Figure 10(b)). It is therefore most appropriate for landform mapping, but like gradient, some find it difficult to
interpret.
Another parameter useful for mapping is surface roughness.
It represents the variability of elevation of a topographic surface
at a given scale (Figure 11(a)). For example, McKean and
Roering (2004) separated landslide debris from different
time periods based on decreasing surface roughness due to
subaerial erosion. There are numerous metrics to characterize
surface roughness and these are reviewed by Grohmann et al.
(2010), who classified them as area ratio, vector dispersion, and
standard deviation methods. The first approach calculates the
ratio between the area of a dipping surface and its area when
projected onto a horizontal plane (Hobson, 1972), with values
approaching one indicating flat surfaces. The second approach
calculates vectors normal to local tangents to the land surface,
and then characterizes how concentrated or dispersed these
vectors are (Hobson, 1972; Guth, 2001; McKean and Roering,
284 000
perfectly specular (i.e., mirror-like) reflection from the surface.

Relief shading, however, highlights a break-in-slope as seen
from a particular direction, which does not necessarily characterize the surface morphology, and suffers from azimuth
biasing. Figure 6 illustrates the problem with a DEM that is
illuminated parallel and orthogonal to the principal landform
orientation. It is better illustrated by using an animation,
where the solar azimuth angle is rotated through 3601 at 51
intervals. Such an animation reveals landforms appearing,
disappearing, and changing shape. It is therefore common
to utilize relief-shaded images generated from multiple
solar-azimuth angles. An alternative solution is to use PCA to
produce optimal combinations of relief-shaded images (Smith
and Clark, 2005).
Supplementary material related to this article can be found
online at doi:10.1016/B978-0-12-374739-6.00054-3.
The above discussion illustrates that even if a DEM accurately represents the topography, the method used to display
topographic information can introduce significant bias. It is
therefore preferable to use LSPs that are designed to numerically (rather than for display) characterize terrain properties,
as they provide the basis for progress toward reproducible
visualization and mapping methodologies.
The slope of the terrain is a fundamental parameter
(Evans, 1972; Olaya, 2008) as it controls the component of
gravitational force available for geomorphic processes and is
defined at a point as a plane tangential to the land surface.
That plane has a steepness (gradient) and orientation (slope
azimuth). The slope gradient, or slope angle, is commonly
311
Figure 10 DEM-derived land-surface parameters. (a) Slope gradient image. (b) Profile curvature image.
Geovisualization
296 000
290 000
284 000
170 000
175 000
308 000
165 000
302 000
296 000
302 000
308 000
302 000
296 000
290 000
175 000
290 000
175 000
170 000
284 000
170 000
284 000
(a) 160 000
165 000
308 000
165 000
160 000
302 000
175 000
296 000
170 000
290 000
165 000
308 000
160 000
284 000
312
(b) 160 000
Figure 11 DEM-derived land-surface parameter and enhancement. (a) Surface roughness image. (b) Local contrast-stretch enhanced image.
2004). For flat terrain, vectors are near vertical and dispersion is
low. Finally, standard deviation calculations of elevation, slope,
and residual relief are commonly used.
Landforms can also be highlighted using the local-relative
elevation (e.g., Pike and Wilson, 1971; Hillier et al., 2007).
This approach highlights local altitudinal variations. Smith
and Clark (2005) refer to this as a local contrast stretch, where
a linear stretch is applied across a user-specified window or
kernel of appropriate width (Figure 11(b)), and a value is
assigned to the pixel according to the relative position between the lowest and highest point in the locality.
Yokoyama et al. (2002) proposed a terrain parameter
called openness. Openness (or the opposite, enclosure) is
calculated using line-of-sight methods that emphasize convexity and concavity in terrain, capturing the degree of geometric dominance. It utilizes standard viewshed algorithms,
but calculates these at nadir and zenith for eight compass
directions. The parameter is sensitive to changes in local relief,
where multiscalar expressions of openness (i.e., distance) will
have an impact. Conceptually it is related to measures of
both roughness and curvature, although the interpretation is
different.
3.11.3.4
Recommendations for Terrain Visualization
A sample of techniques designed to enhance satellite imagery

and DEMs for landform detection have been presented,
although many other techniques and approaches exist. The
selection of the most appropriate visualization technique(s)
will depend on the objectives of the study, the type of analysis

that is required, and the information that must be produced.
There are, however, some approaches that can be generally
recommended as useful.
Simple image display of DEM data is rarely useful as a
visualization technique for studying the landscape or landform mapping. It is therefore necessary to utilize techniques
that are designed to highlight landforms and features of
interest. Although appropriate RRS methods are not necessarily trivial to design, simple methods can be effective in
highlighting individual landforms of interest, relegating all
other aspects of terrain to noise (e.g., Hillier and Smith,
2008). An appropriate RRS to isolate the landforms of interest
will always tend to improve the output of any subsequently
applied technique.
The technique of relief shading highlights subtle topographic features well; it is widely implemented in software, fast
to compute, and very useful when appropriate care is taken to
allow for azimuthal bias. Azimuthal bias from the artificial
illumination source alters the position of breaks-in-slope,
thereby giving the impression of a landform changing shape
or disappearing. The use of multiple relief-shaded images at
least partially mitigates this problem. Nevertheless, techniques
that do not involve illumination are preferable. In general, the
use of LSPs can greatly facilitate the visualization of the
landscape, as a multitude of topographic characteristics can be
examined. Numerous enhancement techniques can be used
on multispectral and terrain images to facilitate mapping efforts. Smith and Clark (2005) and Hillier and Smith (2008)
reviewed many methods, concluding that no single technique
Geovisualization
is ideally suited to terrain visualization. Therefore, the best

practice for landform visualization and mapping is likely to
involve the integration of terrain information using visualization techniques applicable to specific objectives.
3.11.4
Section 3.11.3 presented a variety of techniques that could be

used for static visualization of terrain. It is frequently necessary, however, to combine and interact dynamically with
data to explore for patterns in order to make associations.
This section introduces common techniques to dynamically
interact with spatial data, and then explores increasingly sophisticated methods involving 2D planes and 2.5D surfaces,
before outlining how surface data can complement true 3D
volumetric data.
3.11.4.1
kernel). Reddish colors represent high slope and low openness

(e.g., at steep and enclosed hillslopes), whereas greenish/
bluish colors depict low slopes and relatively high openness
(e.g., in flat and open areas).
3.11.4.2
Visual Interaction
Display
Raster outputs of terrain are usually displayed and inspected

on a video display unit (VDU) using the red, green, and blue
(RGB) additive model of mixing colors (Figure 12). Given the
sensitivity of the human eye, the color cube allows the mixing
of the three primary colors at different intensities to provide
the full gamut of possible viewable colors. It is therefore
possible to display up to three separate images of the same
area at any one time, by encoding the pixel values as a 3D
coordinate position within the color cube. This is known as a
false-color composite (unless it represents the actual object
color as perceived by the human eye in which case it is a
true-color composite). The technique is widely used in remote sensing and, within geovisualization, allows the interactive inspection of multiple terrain datasets.
Figure 13 presents an example of how a false-color composite can be used to interpret the landscape, as it combines
slope angle (red: 3 3 kernel), topographic openness (green:
51 51 kernel), and topographic openness (blue: 501 501
313
Digitization and Overlay
The main output of visual processing is a raster dataset, or

datasets, generated using a variety of techniques. These can
then be interpreted by geomorphologists for the identification
of individual landforms and the recording of their position
and extent. The process is generically known as digitization
and can form a body of work for the production of geomorphological maps (e.g., Hughes and Clark, 2010), or serve
as input layers for further quantitative processing. Digitization
is the electronic creation of geographic coordinates, usually
through a mouse click at the position identified on screen.
When the coordinates are used individually (points) or combined together (lines or polygons), they are known as feature
types, with each feature type forming a separate layer.
Layers can be stacked (e.g., overlain) and reordered,
allowing the interpreter to interact with different data layers
and visually inspect their interaction, as well as removing or
adding layers. This simple organizational framework provides a
powerful paradigm through which geomorphologists can work
digitally, and is conceptually the same as overlaying tracing
paper on an aerial photo and tracing the outlines of landforms.
The process in nondestructive and can involve modifying other
interactive features such as layer symbolization. Raster layers
(such as DEMs or satellite images) are typically rendered as
opaque, obscuring any data they are overlaid onto, although a
simple solution is to change the opacity of a specific data layer.
Interactive methods include switching layers on and off so that
DN (red)
Yellow
Magenta
White
Gray line
Black
DN (green)
Cyan
DN (blue)
Figure 12 A red, green, and blue (RGB) color cube. The cube
defines the intensity of each primary color, allowing the 3D plotting
of all colors within the cube. The gray line shows how colors are
plotted for a gray scale each where each RGB component has equal
intensity.
Figure 13 False-color composite generated using topographic

openness (green: 51 51 kernel), topographic openness (blue: 501
501 kernel) and slope (red: 3 3 kernel). Reddish colors represent
high slope and low openness and greenish/bluish colors depict low
slopes and relatively high openness (e.g., in flat and open areas).
314
Geovisualization
Figure 14 Integrated topographic map, geomorphological map, and aerial photo for Bruchi and Massa River valleys, Switzerland. Reproduced from
Otto, J.-C., Gustavsson, M., Geilhausen, M., 2011. Cartography: design, symbolisation and visualisation of geomorphological maps. In: Smith, M.J.,
Paron, P., Griffiths, J. (Eds.), Geomorphological Mapping: A Handbook of Techniques and Applications. Elsevier, London.
the interpreter can rapidly move between them, either manually or automatically at a prescribed frequency. Alternatively,
users can operate a slider to move a new layer horizontally or
vertically across the data viewer. Simultaneous viewing of a
raster data grid and vector data is simpler, as the vector data can
simply be overlain. The ability to manage large datasets and
interact with them, allowing detailed inspection is an important, although simple, feature of a digital workflow. Figure 14
depicts how multiple datasets, including a topographic map,
geomorphological map, and aerial photo, can be integrated
and viewed simultaneously.
The challenges in digital workflows come not through the
use of technology, but in planning and organizing the digitization process. It is important to predetermine the features that
are of interest (e.g., Sahlin and Glasser, 2008), and group them
thematically (e.g., fluvial, glacial, periglacial, and mass movements). The choice of feature type used to represent the landform will depend on the landform being digitized and the scale
of mapping, although it is important to remember that feature
types are simplified categorizations (or abstractions) of reality,
and that definitions change over time. Polygons approximate
outlines of 2D features, or area features, such as drumlins or
landslides. Over larger areas, however, lines and points can also
be used to represent some area features. For example, regionalscale mapping may represent drumlins as lines and landslides
as points. Further detail on the overarching rational for geomorphological mapping is discussed by Dramis et al. (2011),
with Smith (2011) detailing the digital workflows involved.
3.11.4.3
2D to 2.5D in Space
A simple way to interact with data is through activities such as

zooming, panning, rotating, and performing simple contrast
enhancements. There is a risk that these techniques become to
seem trivial, as a result of the ease with which the user is able to
perform them. Rudimentary visual exploration, however, is an

essential task. Google EarthTM, for example, provides an elegant
interface for the navigation of a single set of imagery for anywhere on Earth, and this is a remarkable achievement. Indeed
the scale of their data licensing means that Google EarthTM is
generally not only the first data source queried, but also the
most up-to-date and highest spatial resolution (e.g., Dykes,
2008). Most commercial image processing and GIS software
supports basic interactive visualization, commonly augmented
with aspatial statistical output limited to summarizing the
whole dataset. Although somewhat restricted, these global
measures can be used to explore features of individual datasets.
For example, histograms of DEMs can highlight artifacts resulting from the conversion of contours (e.g., Smith, 1993;
Wise, 2000). Wood and Fisher (1993) introduced spatial
measures, but highlighted the benefits of visualization in
identifying error in DEMs (Gousie and Smith, 2010).
Animations provide a powerful methodology for data
visualization by depicting attribute change to vector or raster
data. Figure 6 demonstrates that effect of changing illumination azimuth when relief shading is used to display a DEM,
and in this instance, each frame of the animation is rendered
using a different solar azimuth angle. Perhaps more commonly, temporal data can be animated where each frame
represents a successive time slice. Frames can progress at a
fixed rate, or be manipulated manually using a time slider
where users can step through a historical archive of satellite
imagery and aerial photography. Given various landscape
evolution models, the DEM at each time-stage of the simulation could be similarly displayed. This type of functionality
is now incorporated into many desktop GISs. For quantitative
work, change detection results could be used based on a timeseries of raster data using simple raster differencing (e.g.,
Etzellmuller, 2000; Chen et al., 2004). TIMESAT (Jonsson and
Eklundh, 2004) extends imagery analysis to multiepoch data,
and similar results are achievable in geomorphology.
Geovisualization
Figure 15 Perspective view of the Lough Gara region, Ireland. The

surface has been color-coded according to elevation, relief-shaded,
and then draped over a 3D rendition of the DEM. Generated using a
solar azimuth of 3451, with a vertical exaggeration of 10x. Long edge
is 30 km. GeoEye Satellite image.
The discussion so far has focused on the 2D manipulation

of spatial data. DEMs form a special case, as they are a 2D
array representing the upper surface of a 3D volume, but are
not 3D as they do not formally define a volume. They are also
not 2D in the sense of a plane or map, so are sometimes
referred to as 2.5D. A variety of GIS applications allow a user
to navigate a 3D perspective, examining the DEM surface
vertically and horizontally through rotation and panning
(Figure 15; e.g., Jordan et al., 2009). Data can not only be
viewed on traditional 2D screens, but also on 3D display
systems allowing the user to interact with perspective views of
the terrain, overlay aerial photos, and satellite imagery from
different time periods, and add digitized vector line-work that
can subsequently be edited. This enables virtual field reconnaissance and is particularly important for remote or inaccessible sites. The penetration of Google EarthTM into the
geomorphological community has enabled many users to
become familiar with this method of navigation and the creation of flight paths that can be automatically navigated or
animated as fly-throughs.
3.11.4.4
315
structure (e.g., ECHIDNA) to, for example, estimate biomass

(Lovell et al., 2003). Integrating such geophysical data with
DEMs to understand geomorphic processes has been the focus
of much recent work (e.g., Johnson et al., 2006; Hillier et al.,
2008; Jordan et al., 2009; Otto et al., 2009). Processing and
display to achieve insightful interrogation of 3D data are exemplified in tools for the oil industry for processing seismic
reflection data and integrating this with other geospatial information such as well-log measurements (e.g., GeoVisionary,
GeoWall, ProMAX, GeoProbe, and KINGDOM). Such 3D
visualization packages integrate data such as outcropsedimentary logs, strike-and-dip measurements, and horizon
interpretations with photography draped over LiDAR DEMs to
characterize subaerial conditions (e.g., Fabuel Perez et al.,
2010). These packages typically display one or many planar
slices through a 3D volume, or render the 3D volume onto a
screen. Visualization of such data can be enhanced by true 3D
display, usually through the application of stereoscopy. A
variety of systems are currently available, with the general
popularity of 3D video films leading to the development of
low-cost home systems that utilize polarized or shutter-based
techniques. In polarized systems, a single monitor can display
polarized images that are interlaced on the screen. Through
the use of polarized glasses, the human visual system assembles the display into a 3D image. The use of interlacing
effectively halves the resolution of the image, although some
manufacturers use a twin monitor set-up to mitigate this effect. Shutter-based systems use glasses that are actively controlled by the host computer that synchronizes the use of a
blocking shutter on each lens of the glasses with the display of
left and right stereoscopic images and can utilize the full
resolution of the display. Projection systems (e.g., GeoWall)
are better developed as twin projectors, using polarized filters
and screen to allow very large images to be displayed at full
resolution. We therefore see integration of surface, surface
structure, and subsurface information as being integral to the
future building of detailed models that can be used for a wide
range of Earth-science applications.
3D in Space
It is natural to move from 2.5D to full 3D visual interaction

with data, either in terms of estimating volumes, or being displayed in conjunction with true 3D volumetric data (e.g., seismic reflection data). As noted in Section 3.11.4.3, DEMs do not
provide volumetric data, however, where it is possible to define
a lower boundary through an understanding of a geological or
geomorphological basal surface (e.g., Sclater et al., 1975; White,
1993; Wessel, 1998; Hillier and Watts, 2004; Smith et al., 2009),
volume can be estimated. In this context, volume is computed
as the height or thickness between the lower and upper
bounding surfaces of a landform. These calculations may be the
only estimate of volume possible in a study, but uncertainty
exists, as the basal surface is frequently based on inference.
Where subsurface and additional surface data can be collected, a DEM can be seen as just one dataset contributing to
an understanding of feature or landscape volumes. Airborne
radiometric, magnetic, and electromagnetic measurements of
the subsurface, for instance, could be integrated. Further
datasets derived from laser scanning can give information
about supra-surface features such as vegetation canopy
3.11.4.5
Virtual Globes
Virtual globes and online maps have become an important

aspect of the Internet, since the first release of NASAs World
Wind (http://worldwind.arc.nasa.gov). The seamless integration of satellite imagery and DEMs has changed our perception of the Earth within a few years, placing more public
emphasis on geomorphology (Tooth, 2006). The geoscience
community has embraced the use of virtual globes with
emerging applications including simple terrain inspection and
feature mapping (Sato and Harp, 2009; Welsh and Davies,
2010), data visualization, and geo-data exchange (http://
www.usgs.gov). Many national research agencies visualize
their data using virtual globes. For example, the United States
Geological Survey (USGS) provides stream-flow, watershed, or
earthquake data for display in Google EarthTM. The
National Snow and Ice Data Centre (NSIDC) also offers information about ICEsat data, NASAs Ice, Cloud and Land
Elevation satellite (http://nsidc.org/data/virtual_globes/glas/
anchorage.kml). The application localizes each laser footprint
316
Geovisualization
acquired by the satellite and provides further information on

physical parameters (e.g., waveforms of laser reflection) of the
data that help researchers choose the right data set for their
objectives (Ballagh et al., 2011). Land cover and landform
changes are ideal information for display on virtual globes, as
overlays can be used to depict different multitemporal imagery. Terrascope, for example, allows the rapid comparison of
Landsat imagery at different acquisition times that document
land-cover changes in the tropics (http://www.ambiotek.com/
terrascope). Landform and hazardous-process monitoring and
inventories make use of virtual globes to visualize their data,
such as the database of glacier and permafrost disasters compiled by the Glacier and Permafrost Hazards Group (GAPHAZ,
http://www.geo.uio.no/remotesensing/gaphaz/home.html).
Virtual globes not only enhance spatial thinking, which is
regarded as an educational necessity, but also augment science in general, far beyond simple data visualization (Ballagh
et al., 2011). Geomorphologists use virtual globes like Google
EarthTM or NASA World Wind for field-work planning, and the
identification of study sites in presentations. It should be
noted, however, that virtual globes are increasingly applied to
convey scientific data and research results. Tooth (2006) stated
that virtual globes help to address key geomorphological
questions such as scale-dependency of form, simplified by
techniques like zooming. For example, tectonic geomorphology and planetary remote sensing studies strongly benefit
from the large-scale visualization of a planet surface.
Virtual globes make use of data overlays using an existing
database. Thus, visualization via virtual globes is restricted to
the existing data, and does not allow for more complex data
analysis provided by GIS applications. The standard protocol
for data overlay is the KML. This (and KMZ, the compressed
version of KML) is a data format that displays geo-registered
place-marks, annotations, geometries, imagery, and 3D surface
objects on virtual globes, comparable to Hypertext Markup
Language (HTML) that displays text and imagery in a web
browser. The data format is human-readable and can be
written using a simple text editor (Werneke, 2008).
3.11.5
Visual Outputs
A variety of visual-output products can be produced for research or public consumption. Although visualization workflows remain almost entirely digital and interactive, specific
outputs need to be targeted toward the intended audience and
publication medium. For spatial information dissemination,
static maps can be made available in printed or digital form.
Interactive spatial-output products are increasingly becoming common as part of research outcomes, and as a
method designed to engage stakeholders to make research
results more widely available for downstream applications.
Interactive maps have become a standard component on
many websites, and can currently be considered a state-of-theart method for the dissemination of geomorphological data
and information. Virtual globes provide the 3D visualization
of satellite imagery and topographic data of the Earth and are
used to locate the sites where photographs and scientific data
have been collected in the field. A wide range of GIS applications and tools have been developed to generate Internet
maps for the visualization of research results, and to enhance

the search for, or effective distribution of, scientific data. This
section summarizes static and interactive visualization products used in geomorphology. It provides a review of the legend
systems commonly used and basic cartographic principles in
map design. Details regarding digital-map production, web
mapping, and WebGIS are then covered.
3.11.5.1
3.11.5.1.1
Geomorphological Maps
Legend systems
Traditional geomorphological maps differ from other thematic maps in that qualitative information usually prevails
over quantitative data. Consequently, most geomorphological
maps are compiled using descriptive symbols to represent
landforms and processes. Many different symbol sets and
mapping systems have evolved during the twentieth century in
different countries with different thematic emphases and
varying usage of symbols and color (Figure 16). Even though
attempts were made to create a general legend for geomorphological maps by the International Geographical Union
(IGU) in the 1960s (Demek et al., 1972), no universally applicable legend system has been established. Two main styles
of geomorphological maps include: (1) multilayer maps depicting morphology, landform genesis, current processes,
surface and subsurface material, and chronology and (2)
simpler maps focusing on landform relationships and
morphology (Evans, 2012). The former style results in multicolored maps with several stacked layers of data that tend to
be overloaded with information and may be difficult to interpret (e.g., the German system; Barsch and Liedtke, 1980),
whilst the latter is usually presented in black and white and
represents a reduction of information for practical purposes
(e.g., the British system; Evans, 1990). Furthermore, in addition to more universal legend systems that provide symbols
for the entire breadth of geomorphological landscapes
(Demek et al., 1972; Barsch and Liedtke, 1980; Cooke and
Doornkamp, 1990), specialized legends for high mountains
(Kneisel et al., 1998; De Graaff et al., 1987) or hazardous
processes exist (Kienholz, 1978; Kienholz and Krummenacher,
1995). See Otto et al. (2011) for an overview of selected geomorphological legend/mapping systems.
3.11.5.1.2
Map design
Geomorphological maps are complex thematic maps that

place demands on cartographic visualization techniques to
provide a comprehensible and readable map. It is therefore
beneficial to consider some underlying principles of cartography and map design. The basic representations of objects
on maps are the symbol primitives: point, line, and area
(Robinson et al., 1995). Whether a linear feature in nature is
represented by a line symbol on a map is primarily a question
of scale. For example, a blue line can be used to depict a river,
whereas on larger scale maps a river can be depicted using an
area symbol, covering the space between the riverbanks. The
map scale also determines if a landform is depicted by a point
symbol, or if it is split up into its morphological components.
Rock glaciers, for example, could be represented by a single
point symbol on small-scale maps, or by the assemblage of
line and area symbols that differentiate the step height of the
Geovisualization
Legend system
Landform example
Moraine ridge
Fluvial terrace
Emphasis
IGU unified key

(Demek et al., 1972)
Morphogenesis
ITC (NL)
(Verstappen & Zuidam, 1968)
Process/genesis
GMK 25/100 (GER)

(Barsch, 1978)
British Geomorphological Maps
(Evans, 1990)
Genesis
Form/genesis
ARAG (NL)
(DeGraaf et al., 1987)
Genesis/
surface material
IGUL (CH)
(Schoeneich et al., 1998)
Morphogenesis/
landforms
Mapping system by Gustavsson

(Gustavsson et al., 2006)
Morphogenesis
BUWAL/BAFU (CH)
(Kienholz, 1976)
317
Process/landform
Figure 16 Symbols for moraine ridge and fluvial terrace from different geomorphological legend systems. Reproduced from Otto, J.-C.,
Gustavsson, M., Geilhausen, M., 2011. Cartography: design, symbolisation and visualisation of geomorphological maps. In: Smith, M.J., Paron,
P., Griffiths, J. (Eds.), Geomorphological Mapping: A Handbook of Techniques and Applications. Elsevier, London.
rock-glacier front, furrows, and ridges, and the accumulation

of boulders and blocks on top of the rock glacier, if the
map scale increases. Note that this is different and independent of the feature types used when digitizing data. For
specific projects, it is often the case that landforms may be
well aligned (e.g., drumlins digitized as lines are presented
on a geomorphological map as lines); however, it may be
necessary to use the same base data for the production of
different scale maps. This would require the generalization of
feature types.
A differentiation of these basic representations, to express
relationships among or differences between the data, can be
achieved by variations of the basic visual variables: color,
texture, size, shape, or orientation (Robinson et al., 1995;
Kraak and Ormeling, 2010). For example, different classes of
steps or breaks in slopes, depicting various levels of river terraces, could be visualized using different sizes or variations of
the same line symbol (Figure 17). Variable symbol shapes
demonstrate qualitative differences, and this is the most
commonly applied visual variable in geomorphological maps,
because of the great number of different symbols for different
landforms and processes.
Color is an important visual variable, mainly used to depict
qualitative differences. Geomorphological maps, however, are
commonly produced in black and white especially when they
are part of a journal publication to keep production costs
low. If color is used, variation of color characteristics, hue
(color variation), value (lightness), and chroma (saturation)
can be used to emphasize certain aspects of the map. Within
cartography, some color conventions exist that should be
acknowledged to avoid confusion. For example, on topographic maps, blue is used for objects related to water, like
rivers, springs, or lakes, and green generally represents
areas covered by vegetation. Color decisions can be facilitated
using the online software tool Colorbrewer (http://colorbrewer2.org/). The software assists users in choosing the right
composition of colors by displaying different color schemes.
Color combinations can be tested on a complex map sample
that enables the designer to experience the differentiation and
perception of the colors used. In many geomorphological legend systems, colors are applied to represent variations in
landform genesis (Barsch and Liedtke, 1980), process domains (Gustavsson et al., 2006), or lithology (Pasuto et al.,
1999).
Map layout consists of the arrangement of the map components into a functional composition and a meaningful and
aesthetically pleasing design to facilitate visual communication (GITTA, 2006). Geomorphological maps commonly
include the following elements surrounding the main map:
title, legend, scale, directional indicator (north arrow), coordinate grid, or border, information on coordinate system
and map projection, and author credits. Commonly inset
maps are included that show the location of the mapped
area (essential for large-scale maps), an overview of the
geological situation, or other additional information on the
study area (e.g., a slope map). These items need to be arranged
carefully to guide the viewers eyes toward the focus of
the map.
Most geomorphological maps are now produced using
graphics or GIS software that provide tools to facilitate map
318
Geovisualization
25
50 Meters
18
1850
50
184
40
18
Legend
Step height >5 m
Step height 25 m
Step height < 2 m
Creek
Contours 5 m
Figure 17 Mapping different levels of fluvial terraces using classified line symbols.
creation. The underlying cartographic principles outlined

above, however, remain important to produce maps that are
fit for purpose. On a map, all information is spatially related
and needs to be considered holistically. The composition of
map items decides if and how the reader understands the
message, with perception and understanding occurring subconsciously. To engage map-users and enable them to develop an understanding of the meaning of the map, a visual
sense of the symbols and their attributes that correspond to
the intention of the cartographer is required (Robinson et al.,
1995).
3.11.5.2
Digital Mapping
Digital map creation is performed using either vector graphics

software (e.g., Adobe Illustrator) or GIS software. The main
advantage of graphical software with respect to the generation
of geomorphological maps is the great number of tools that
are available for the creation and modification of graphic
objects, as well as advanced printing options. Generally, these
can be adjusted and customized to the users needs and the
specific requirements of the legend system applied. These
functions still exceed the cartographic capabilities of GIS
software.
The primary advantage of a GIS is that it enables the
geographical reference system of the map to be retained, with
data stored in geographical databases for later analysis. Thus,
a GIS offers the ability to combine basic information on
landforms such as process/material type and geometry with

secondary data, for example, from physical sampling, laboratory analyses, geophysical investigations, or the results
from geospatial analyses within the GIS (Minar et al., 2005;
Gustavsson et al., 2008). The results of GIS analyses are generally compiled into maps and consequently, GIS software
includes mapping facilities and some graphic-design capabilities. Among these are automatic tools to generate the legend,
scale bar, north arrow, or coordinate grid, and functions
managing text labels and symbols. These map elements are
automatically and self-consistently updated when changes
are made in scale or symbol type. One challenge in the
process of map production is the generation of reusable,
standardized digital symbols. Although many symbols for
various objects or thematic purposes exist, only a few specialized geomorphological symbol sets are available for GIS
software (Bundesamt fur Wasser und Geologie, 2002; Otto
and Dikau, 2004; Otto, 2008; IGUL, 2010). Given the existing
cartographic restrictions of many GIS, however, geomorphological legends and symbols need to be simplified for GISbased map production (Gustavsson et al., 2006; Gustavsson
and Kolstrup, 2009). Special symbol editors are provided to
compose and define the symbol set for the map (e.g., in
ArcGIS). As with graphic software, GIS software offers tools to
digitize vectors (points, line, polygons) with a high degree of
accuracy and provides the ability to modify single vector
nodes. An optimized graphical output is reached by transferring the GIS map to the graphics software to finalize the
geomorphological map.
Geovisualization
Generating geomorphological maps using a GIS enables

numerous possibilities for the dissemination of research
outputs extending beyond simple paper products. Internet
technologies can contribute to both the dissemination of
geomorphological maps and access to geomorphologic data
and help to make geomorphological knowledge available
to the general public. In contrast to static digital maps
(i.e., simple images of maps, for example (http: http://gidimap.
giub.uni-bonn.de/gmk.digital/home_en.htm)), dynamic-web
maps are characterized by interactive capabilities, as the user
can interact with the map by zooming, panning, querying, or
adding thematic layers (Mitchell, 2005). Geo-registered map
data can be transferred and published in several digital ways
including GeoPDF, dynamic-web maps (e.g., WebGIS), and
virtual globes (e.g., Google EarthTM). These techniques are
outlined in the following sections.
3.11.5.2.1
Open standards
Data distribution and access in distributed web-based geospatial infrastructures need to be specified to achieve interoperability in a way that different applications (e.g., spatial
databases, Internet web servers, or web browsers) on various
platforms (e.g., Linux and Microsoft Windows) can interact
and communicate with each other. The specific needs for
interoperable geospatial technologies are implemented in
specifications or standards describing the basic data models to
represent different geographical features. They are important
because they contribute to both the interoperability for users
mashing up different applications and allowing software/
technology developers to make complex spatial information
and services universally accessible.
The standards are specified by the Open Geospatial Consortium (OGC), a non-profit international standards organization, with members from commercial, governmental, and
research organizations (http://www.opengeospatial.org). It
leads the development of standards to establish interoperability
and ensures platform- and software-independent usability of
geospatial services and data sharing. The standards or specifications are the main outcomes of the OGC and appear as
openly published technical documents that detail interfaces or
encodings. These enable software developers to build support
for the interfaces or encodings into their products and services.
There are currently more than 30 standards defined, the most
prominent of which are web services, specifically:
Web Map Service (WMS), which provides maps as raster

images;
Web Feature Service (WFS), which allows the retrieval of
feature types (i.e., point, line, and polygon vector features);
Web Coverage Service (WCS), which allows the retrieval of
tessellated surfaces, most commonly raster images. This
could be satellite imagery, DEMs, or aerial photography.
For data description and storage, Extensible Markup Language (XML)-based languages such as GML (Geography
Markup Language), GeoSciML (GeoScience Markup Language), or KML have been developed. XML defines a set of
rules for document encoding, comparable to HTML, used for
creating web pages. The OGC specifications and standards
have greatly influenced the direction of web-based GIS
319
developments making it much easier to publish, visualize, and

exchange any geospatial data over the Internet (Mooney and
Winstanley, 2009). Basic functionality, advantages, and limitations of WMS and KML are discussed in Section 3.11.5.2.3
and exemplified by case studies. In addition to specifications
and standards, the OGC publishes several White and Discussion Papers or Best Practice Guides (e.g., GeoPDF Encoding
Best Practice Version 2.2; Graves and Carl, 2009). Section 3.11.5.2.2 introduces the GeoPDF, a merging of geospatial
data with the portable document format (PDF) file format.
3.11.5.2.2
GeoPDF
A GeoPDF, an OGC standard, includes one or multiple map

frames within a PDF page associated with a coordinate reference system (Graves and Carl, 2009). It enables the sharing of
geospatially referenced maps and data in PDF documents.
Multiple, independent map frames with individual spatial
reference systems are possible within a GeoPDF for map
overlays or insets. Geospatial functionality includes scalable
map display, layer visibility control, access to attribute data,
coordinate queries, and spatial measurements. This allows
similar functionality to simple interactive queries that would
be performed in a GIS. Adobe Reader (starting with Version
9.0) supports the geospatial functions of GeoPDFs; however,
full GeoPDF functionality requires the free TerraGo plug-in for
Adobe Reader (http://www.terrago.com). GeoPDFs can be
created either directly from a GIS (e.g., ESRI ArcGIS 9.3) or
using bespoke software (e.g., TerraGo Publisher or Map2PDF).
A GeoPDF enables fundamental GIS functionality outside
specialized GIS documents, turning the formerly static PDF
maps into interactive, portable, geo-referenced maps. It is an
interesting and valuable way of dissemination for geomorphological maps. Some geospatial data providers, such as
the USGS or the Australian Hydrographic Service (AHS), have
already started publishing interactive maps using the GeoPDF
format (http://store.usgs.gov).
3.11.5.2.3
Principles of web mapping and WebGIS
Web mapping is a common way of presenting dynamic

maps online. It links the Internet with GIS technology and
enables visualization, localization, and dissemination of georegistered data. Web-mapping applications such as Google
Maps or Bing Maps, which provide access to street data or
aerial imagery, are very popular and widespread and have increased the interest and access to mapping. Mooney and
Winstanley (2009) point out that web mapping and WebGIS
applications are key components in the distribution of geospatial data and information.
Depending on the software system components, advanced
symbols, map overlays from different applications, and their
integration into a Desktop GIS are possible. Interoperability is
achieved through the use of OGC web services that include
mechanisms for the integration and visualization of information from multiple sources. The terms web mapping
and WebGIS are generally used synonymously, although they
do not necessarily refer to the same technologies. When
analytical GIS functionality is provided, the term WebGIS is
generally used (Mitchell, 2005; Mooney and Winstanley,
2009). GIS processing is performed online using a GIS server
and maps are visualized in interactive web viewers (e.g.,
320
Geovisualization
Server
(a) WebGIS
Forwarding
Map request
Browser/
web client
Access
Web server
Geodata
Mapserver
Map
Map
(b) Web service
Local client
(e.g. Google earth/
desktop GIS)
Map request
Geodata
Map
Server
Data
server
(optional)
Mapserver
Web server
Geodata
Geodata
KML file
Data
server
KML file
(optional)
Figure 18 Simplified scheme of information and data transfer of a WebGIS and Web-Service application.
OpenLayers www.openlayers.org, MapBender www.

mapbender.org, ka-Map! www.ka-map.maptools.org, MapBuilder www.mapbuilder.net, Google Maps maps.google.
com, or Bing Maps www.bing.com/maps). Although there are
many ways to develop a WebGIS or to access web maps, depending on the software components used, most applications
are based on the same principles (Figure 18).
The principles are as follows. The user operates a web
viewer inside a web browser and this provides selected
GIS functionality (e.g., zooming or panning, data query, layer
selection). The software compiles the user requests and forwards them to the application server (Figure 18(a)). The
server passes the map requests to the map-server, where the
central software performs the GIS processing. The map-server,
having access to the spatial data, executes the map requests
and returns the maps as images to the web server, which
finally serves them back to the users web viewer. The application acts as a web-based information system. One popular
package is Maptools MapServer for Windows (http://
www.maptools. org/ms4w/), which uses open-source components to provide a map-server environment including libraries for data input and output. MapServer is GIS software
running on a web server that enables interaction with GIS data
over the Internet and generates cartographic output of geographic content. For more information on common WebGIS
tools, see Mitchell (2005).
Most map-servers provide standardized web services like
WMS for accessing maps online. The WMS contains the map
request and parameters specifying GIS processing for the mapserver (e.g., choice of layers or spatial extent). Desktop GIS
(e.g., ArcGIS, MapInfo, and Global Mapper) as well as Internet

map viewers compile WMS data providing direct access to map
data from Internet servers. WMS technology permits users to
visualize entire GIS projects, such as a geomorphological map
on the Internet. Thematic layers of the map can be provided as
well as full analysis functionality, depending on the server-side
GIS software (e.g., Mapserver and GeoServer). Figure 19 shows
a WebGIS that visualizes the results of a geomorphological field
mapping campaign in the Turtmann valley, Switzerland (http://
www.geomorphology.at). The application employs MapServer
generating the maps as WMS, the spatial database management
system PostgreSQL (http://www.postgresql.org) maintaining
the geometries, and the web-mapping client Mapbender (http://
www.mapbender.org). Aerial images and a relief shaded map
are provided as base layers and several thematic layers present
information on process domains, surface materials, landforms,
and single processes. Due to MapServers powerful cartographic
engine, complex geomorphological symbols can be designed
and displayed. Symbols based on the legend for high-mountain
geomorphological systems established by Kneisel et al. (1998)
have been implemented. The WebGIS map thus uses the same
symbols as the printed map of the same area (Otto and Dikau,
2004). MapServer uses one symbol file that defines the composition of symbols for all types of vector geometries. Point
information, such as individual landforms, is displayed using a
geomorphological True Type font (Otto and Dikau, 2004) and
the spatial orientation of each character is achieved by providing the rotation angle as attribute data. Line features such as
crests and ridges are constructed using multilevel symbols, and
advanced polygon symbols are supported by hatchuring or
Geovisualization
321
Figure 19 The graphical user interface (GUI) of the geomorphological WebGIS application Turtmanntal (Universities of Salzburg and Bonn,
available at http://www.geomorphology.at).
image fills. The Turtmanntal WebGIS offers simple functionality

of a Desktop GIS such as spatial navigation, coordinate queries,
length and area calculations, as well as the selection of single
layers of information. The composed image of the map frame
can be exported as a high-resolution PDF (300 dpi) in A4 and
A3 landscape or portrait orientation. For educational purposes,
a glossary defines geomorphological terms.
Usually no restrictions exist concerning the number of
WMS services included within a WebGIS application. Thus,
WebGIS applications are powerful tools to disseminate geospatial information to users from different organizations (e.g.,
local authorities and environmental agencies).
It is important to note that Desktop-GIS software benefits
from the processing power of the local computer, web-based
applications perform all operations online in real time, and
factors such as bandwidth capacity, network latency, browser
type, and system performance need to be considered. In addition, users expect rapid applications and instantaneous responses to their spatial queries (Mooney and Winstanley, 2009).
3.11.6
Conclusions
Geovisualization has garnered considerable interest, as a term

covering a wide swath of activities ranging from exploration,
through to analysis, synthesis, and presentation. An intuitive,
useful definition is the visual depiction of spatial data, and
this chapter has focused more on the use of geovisualization

in geomorphology under this definition. There is considerable
overlap between some of the techniques outlined here with
analysis techniques covered in other chapters in this volume,
and this demonstrates the important role that visualization
plays in the analysis of spatial data for geomorphological
applications. It also clearly demonstrates the very important
role that visualization plays in the exploration of patterns, as
we seek to formalize concepts and theories, and better
understand landscape evolution.
Advances in geospatial technologies have led to the widespread availability of high-resolution terrain data that have facilitated advances in geomorphometry and geovisualization,
collectively revolutionizing geomorphology. The generation of
new observational geospatial data sets, through the use of
geovisualization and mapping, may be functionally termed
visual processing. In geomorphology, researchers are specifically
interested in attempting to visualize the topography in new
ways through quantitative characterization and qualitative
visualization. Numerous techniques can be used to describe
various characteristics of the topography that can be visualized,
including location, extent, variation, height, volume, and many
other parameters. This is still predominantly performed by
manual mapping efforts using image interpretation, although
new automated approaches also exist for the generation of
land-surface parameters and mapping, although they must be
carefully evaluated and ultimately formalized.
322
Geovisualization
For manual mapping, interpreters should be aware of

issues pertaining to the identification of landforms and specifically the minimum resolvable landform size, the effect of
azimuth biasing, and the concept of landform signal strength.
Satellite imagery rarely provides ideal datasets for accurate
landform mapping, and care should be taken in the selection
of imagery to account for those regions of the electromagnetic
spectrum that best differentiate landforms versus the surrounding landscape. Image-processing techniques facilitate
visualization and range from simple to complex algorithms
and approaches. Many produce similar results, so it is critical
to utilize the best algorithms for enhancement.
Similarly, there are a variety of terrain visualization techniques. Examples that were presented include relief shaded,
regionalresidual separation to isolate landforms of interest,
and the use of multiscale geomorphometric parameters that
highlight or enhance specific topographic characteristics.
Combining terrain information in innovative ways that facilitate geomorphological mapping and visualization of the
topography are required to standardize visualization products
that characterize specific aspects of geomorphological systems.
Geovisualization also extends to interacting with spatiotemporal data, a process that is now far more common across
disciplines. This can incorporate simple methods for spatial
overlay, panning, and zooming of 2D data, as well as extending to 2.5D visualization and exploration of surface data,
as well as true 3D and temporal analysis. Although the development and evaluation of many of these techniques and
approaches are research intensive and require specialist
knowledge, simple functions such as panning, zooming, and
rotation illustrate the real power of interacting with, and
understanding, the nature of spatial data. There is no better
demonstration of this than in web-based interactive mapping
applications, which essentially represent a stripped down GIS
interface for users to explore the characteristics of their data
within a managed setting. This straddles analysis and product
generation, and visualization products that also include traditional paper-based cartographic outputs. Although such
interactive maps might seem to be radically different products,
the collective power of human visualization and spatial
overlay for the organization and management of geospatial
data lies in the end-to-end digital workflows. That is, satellite
imagery and DEMs represent the primary data inputs and,
after subsequent processing, are used as the basis for generating geomorphological maps. The system can be used for the
generation of a variety of output that is appropriate for the
intended end-user, be that specific data layers, intermediate
processing results, or a printed map. Regardless of the medium
employed, care should be taken during the design phase in
order to synthesize a comprehensible and readable map.
Given that geomorphological maps are complex thematic
maps, there are high demands on cartographic visualization
techniques. Although effective geospatial communication may
not be the primary objective compared with information
production or science objectives, stakeholder engagement has
become a significant part of the research process. Researchers
are now required to engage the general public in terms of
outcomes, and relevance, of their work, whilst also providing
products that maximize downstream application and use.
Web-mapping systems provide the flexibility to target different
end-user groups and geomorphologists should make full use

of the facilities available.
It is also pertinent to consider the potential for new developments in geovisualization. It is very likely that more
automated approaches to landform mapping will dominate in
the future, as new analysis and visualization techniques permit
better characterization of topographic features. Dynamic
visualization techniques, and particularly in the areas of 3D
and temporal analysis, will gradually improve, and we can
expect to see new output products that better engage people
and make use of augmented reality. Such technologies already
exist in limited ways, and new developments for assessing
subsurface and material variations in the near-surface
environment are sorely needed.
References
Andrienko, G., Andrienko, N., Jankowski, P., Keim, D., Kraak, M.J., MacEachren,
A.M., Wrobel, S., 2007. Visual analytics for spatial decision support: setting the
research agenda. International Journal of Geographic Information Science 8,
839857.
Ballagh, L.M., Raup, B.H., Duerr, R.E., Khalsa, S.J.S., Helm, C., Fowler, D., Gupte,
A., 2011. Representing scientific data sets in KML: methods and challenges.
Computers & Geosciences 37(1), 5764.
Barsch, D., Liedtke, H., 1980. Principles, scientific value and practical applicability
of the geomorphological map of the Federal Republic of Germany at the scale of
1:25,000 (GMK 25) and 1:100,000 (GMK 100). Zeitschrift fur Geomorphologie
Suppl 36, 296313.
Benetti, S., Dunlop, P., Cofaigh, C.O., 2010. Glacial and glacially-related features on
the continental margin of northwest Ireland mapped from marine geophysical
data. Journal of Maps v2010, 1429.
Bennett, G.L., Evans, D.J.A., Carbonneau, P., Twigg, D.R., 2010. Evolution of a
debris-charged glacier landsystem, Kviarjokull, Iceland. Journal of Maps v2010,
4067.
Bonham-Carter, G.F., 1994. GIS for Geoscientists. Elsevier, Oxford.
Brunsden, D., Doornkamp, J.C., Fookes, P.G., Jones, D.K.C., Kelly, J.M.H., 1975.
Large scale geomorphological mapping and highway engineering design.
Quarterly Journal of Engineering Geology 8, 227253.
Bundesamt fur Wasser und Geologie, B., 2002. Naturgefahren Symbolbaukasten zur
Kartierung der Phanomene EDV-Legende fur die digitale Karthographie.
Bundesamt fur Umwelt, Switzerland, Bern.
Calcagno, P., Cazenave, A., 1994. Subsidence of the sea-floor in the Atlantic and
Pacific Oceans regional and large-scale variations. Earth and Planetary
Camargo, F.F., Florenzano, T.G., de Almeida, C.M., de Oliveira, C.G., 2009.
Geomorphological mapping using object-based analysis and ASTER DEM in the
Paraba do Sul Valley, Brazil. International Journal of Remote Sensing 30,
66136620.
Campbell, J.B., 2007. Introduction to Remote Sensing. Taylor & Francis, London.
Carbotte, S.M., Arko, R., Chayes, D.N., et al., 2004. New integrated data
management system for Ridge2000 and MARGINS research. EOS Transactions
85, 553.
Cazenave, A., Dominh, K., Alle`gre, C.J., Marsh, J.G., 1986. Global relationship
between oceanic geoid and topography. Journal of Geophysical Research
91(B11), 11,43911,450.
Chen, J., Jonsson, P., Tamura, M., Gu, Z., Matsushita, B., Eklundh, L., 2004. A
simple method for reconstructing a high-quality NDVI time-series data set based
on the SavitzkyGolay filter. Remote Sensing of Environment 91, 332344.
Clark, C.D., 1997. Reconstructing the evolutionary dynamics of palaeo-ice sheets
using multi-temporal evidence, remote sensing and GIS. Quaternary Science
Reviews 16, 10671092.
Colwell, R.N., 1983. Manual of Remote Sensing. American Society of
Photogrammetry, Falls Church, VA.
Cooke, R.U., Doornkamp, J.C., 1990. Geomorphology in environmental
management. A New Introduction, 2nd ed. Clarendon Press, Oxford.
Crosby, A.G., McKenzie, D., Sclater, J.G., 2006. The relationship between depth,
age and gravity in the oceans. Geophysical Journal International 166, 553573.
Geovisualization
De Graaff, L.W.S., De Jong, M.G.G., Rupke, J., Verhofstad, J., 1987. A

geomorphological mapping system at scale 1:10 000 for mountainous areas
(Austria). Zeitschrift fur Geomorphologie 31, 229242.
Demek, J., Embleton, C., Gellert, J.F., Verstappen, H.T., 1972. Manual of detailed
geomorphological mapping. International Geographical Union Commission on
Geomorphological Survey and Mapping. Academia, Prague.
Dramis, F., Guida, D., Cestari, A., 2011. Nature and aims of geomorphological
mapping. In: Smith, M.J., Paron, P., Griffiths, J. (Eds.), Geomorphological
Mapping: Methods and Applications. Elsevier, London, chap. 3.
Dykes, A.P., 2008. Geomorphological maps of Irish peat landslides created using
hand-held GPS. Journal of Maps 2008, 258276.
ESRI, 2003. Spatial Data Standards and GIS Interoperability. ESRI white paper.
http://www.esri.com/library/whitepapers/pdfs/spatial-data-standards.pdf (accessed
12/07/2010).
Etzellmuller, B., 2000. On the quantification of surface changes using grid-based
digital elevation models (DEMs). Transactions in Geographical Information
Systems 4, 129143.
Evans, I.S., 1972. General Geomorphometry, derivatives of altitude and descriptive
statistics. In: Chorley, R.J. (Ed.), Spatial Analysis in Geomorphology. Harper and
Row, New York, pp. 1790.
Evans, I.S., 1990. Cartographic techniques in geomorphology. In: Goudie, A. (Ed.),
Geomorphological Techniques. Unwin Hyman, London, pp. 97108.
Evans, I.S., 2012. Geomorphometry and landform mapping: What is a landform?
Fabuel Perez, I., Hodgetts, D., Redfern, J., 2010. Integration of digital outcrop
models (DOMs) and high resolution sedimentology workflow and implications
for geological modelling. Petroleum Geoscience 16, 123.
Ford, J.P., 1984. Mapping of glacial landforms from Seasat radar images.
Quaternary Research 22, 314327.
Gibson, P., Power, C., 2000. Introductory Remote Sensing Digital Image Processing
and Applications. Routledge, London.
GITTA, 2006. Layout Design Settings/Graphical Semiology. Geographic Information
Technology Training Alliance. http://www.gitta.info/LayoutDesign/en/html/
index.html
Gousie, M.B., Smith, M.J., 2010. DEMView: 3D visualization of DEM error,
Accuracy 2010, Leicester, July 2010, pp. 165168.
Graves, P., Carl, R., 2009. GeoPDF Encoding Best Practice Version 2.2, OGC
project document: OGC 08-139r2, Version 0.5.1. Open Geospatial Consortium
Inc.
Grohmann, C.H., Smith, M.J., Riccomini, C., 2010. Multi-scale analysis of
surface roughness. IEEE Transactions on Geoscience and Remote Sensing 99,
114.
Gustavsson, M., Kolstrup, E., 2009. New geomorphological mapping system used
at different scales in a Swedish glaciated area. Geomorphology 110, 3744.
Gustavsson, M., Kolstrup, E., Seijmonsbergen, A.C., 2006. A new symbol-and-GIS
based detailed geomorphological mapping system: renewal of a scientific
discipline for understanding landscape development. Geomorphology 77, 90111.
Gustavsson, M., Seijmonsbergen, A.C., Kolstrup, E., 2008. Structure and contents
of a new geomorphological GIS database linked to a geomorphological map With an example from Liden, central Sweden. Geomorphology 95, 335349.
Guth, P.L., 2001. Quantifying terrain fabric in digital elevation models. In: Ehlen, J.,
Harmon, R. (Eds.), The Environmental Legacy of Military Operations. Geological
Society of America Reviews in Engineering Geology, Boulder, CO, pp. 1325.
Hardin, P.J., Jensen, R.R., 2011. Introduction small-scale unmanned aerial
systems for environmental remote sensing. GIScience and Remote Sensing
48(1), 13.
Hart, J.K., Martinez, K., Ong, R., Riddoch, A., Rose, K.C., Padhay, P., 2006. A
wireless multi-sensor subglacial probe: design and preliminary results. Journal
of Glaciology 52, 389397.
Hengl, T., Reuter, H.I., 2008. Geomorphometry: Concepts, Software, Applications,
Elsevier.
Heritage, G., Large, A., 2009. Laser Scanning for the Environmental Sciences.
Wiley-Blackwell, Chichester.
Hesse, R., 2010. LiDAR-derived local relief models - a new tool for archaeological
prospection. Archaeological Prospection 17, 6772.
Hillier, J., Bunbury, J., Graham, A., 2007. Monuments on a Migrating Nile. Journal
of Archaeological Science 34, 10111015.
Hillier, J., Smith, M.J., 2008. Residual relief separation: DEM enhancement for
geomorphological mapping. Earth Surface Processes and Landforms 33, 22662276.
Hillier, J., Tilmann, F., Hovius, N., 2008. Submarine geomorphology: new views on
an unseen landscape. Basin Research 20, 467472.
Hillier, J.K., 2008. Seamount detection and isolation with a modified wavelet
transform. Basin Research 20, 555573.
323
Hillier, J.K., 2011. Submarine geomorphology: quantitative methods illustrated with the
Hawaiian volcanoes. In: Smith, M.J., Paron, P., Griffiths, J. (Eds.),
Geomorphological Mapping Methods and Applications. Elsevier, London, chap. 12.
Hillier, J.K., Watts, A.B., 2004. Plate-like subsidence of the east Pacific rise
south Pacific superswell system. Journal of Geophysical Research 109. http://d
x.doi.org/10.1029/2004JB003041.
Hillier, J.K., Watts, A.B., 2007. Global distribution of seamounts from ship-track
bathymetry data. Geophysical Research Letters 34(13), L13304. http://dx.doi.org/
10.1029/2007GL029874.
Hobson, R.D., 1972. Surface roughness in topography: quantitative approach.
In: Chorley, R.J. (Ed.), Spatial analysis in geomorphology. Methuen, London,
pp. 225245.
Hughes, A., Clark, C.D., 2010. Subglacial bedform map of Britain. Journal of Maps
v2010, 543563.
IGUL, 2010. La legende geomorphologique de lIGUL. http://www.unil.ch/igul/
page19238.html
Jansson, K.N., Glasser, N.F., 2005. Using landsat 7 ETM imagery and Digital
Terrain Models for mapping lineations on former ice sheet beds. International
Johnson, A., Leigh, J., Morin, P., Van Keken, P., 2006. GeoWall: stereoscopic
visualization for geoscience research and education. IEEE Computer Graphics
and Applications 26, 1014.
Jones, E.J.W., 1999. Marine Geophysics. Wiley, Chichester.
Jonsson, P., Eklundh, L., 2004. TIMESAT a program for analysing time-series of
satellite sensor data. Computers and Geosciences 30, 833845.
Jordan, C., Bateson, L., Bow, J., Newell, A., Napier, B., Sabine, R.J., 2009.
Geovisionary Software for 3D Visualisation and Petroleum Exploration in
Southern Tajikistan, RSPSoc 2009: New Dimensions in Earth Observation.
Remote Sensing Society, Leicester, UK, pp. 637-642.
Kienholz, H., 1978. Maps of geomorphology and natural hazards of Grindelwald,
Switzerland: scale 1:10000. Arctic and Alpine Research 10, 169184.
Kienholz, H., Krummenacher, B., 1995. Symbolbaukasten zur Kartierung der
Phanomene. In: Bundesamt fur Umwelt, W.u.L.B. (Ed.), Mitteilungen des
Bundesamtes fur Wasser und Geologie Nr. 6. Bundesamt fur Umwelt, Wald und
Landschaft (BUWAL), Bern.
Kim, S., Wessel, P., 2008. Directional median filtering for the regionalresidual
separation of bathymetry. Geochemistry Geophysics Geosystems 9, Q03005.
Klimaszewski, M., 1982. Detailed geomorphological maps. ITC Journal 3,
265271.
Kneisel, C., Lehmkuhl, F., Winkler, S., 1998. Legende fur geomorphologische
Kartierungen in Hochgebirgen (GMK Hochgebirge). Trierer Geographische
Studien 18, 12.24.
Koppers, A.A.P., Staudigel, H., Minnett, R., 2010. Seamount catalog: seamount
morphology, maps, and data files. Oceanography 23, 3637.
Kraak, M.-J., 2008. Editorial: from geovisualisation toward geovisual analytics.
Cartographic Journal 45, 163164.
Kraak, M.-J., Ormeling, F., 2010. Cartography: Visualisation of Spatial Data, Third
ed. Prentice Hall, Harlow, Essex.
Laliberte, A.S., Rango, A., 2011. Image processing and classification procedures for
analysis of sub-decimeter imagery acquired with an unmanned aircraft over arid
rangelands. GIScience and Remote Sensing 48(1), 423.
Lillesand, T.M., Kiefer, R.W., Chipman, J.W., 2008. Remote Sensing and Image
Interpretation, Sixth ed. Wiley, New York.
Lovell, J.L., Jupp, D.L.B., Culvenor, D.S., Coops, N.C., 2003. Using airborne and
ground based ranging Lidar to measure canopy structure in Australian forests.
Canadian Journal of Remote Sensing 29, 607622.
MacEachren, A.M., Gahegan, M., Pike, W., Brewer, I., Cai, G., Lengerich, E.,
Hardistry, F., 2004. Geovisualization for knowledge construction and decision
support. Computer Graphics and Applications 24(1, 24), 1317.
Marty, J.C., Cazenave, A., 1989. Regional variations in subsidence rate of oceanic
plates: a global analysis. Earth and Planetary Science Letters 94, 301315.
Mason, D.C., Horritt, M.S., Hunter, N.M., Bates, P.D., 2007. Use of fused airborne
scanning laser altimetry and digital map data for urban flood modeling.
Mather, P.M., 2004. Computer Processing of Remotely-Sensed Images. Wiley
Blackwell, London.
McKean, J., Roering, J.J., 2004. Objective landslide detection and surface
morphology mapping using high-resolution airborne laser altimetry.
McKenzie, D.P., Watts, A.B., Parsons, B., Roufosse, M., 1980. Planform of mantle
convection beneath the Pacific Ocean. Nature 288, 442446.
Menard, H.W., 1973. Depth anomalies and the bobbing motion of drifting islands.
324
Geovisualization
Minar, J., Mentlk, P., Jedlicka, K., Barka, I., 2005. Geomorphological information
system idea and options of practical implementation. Geograficky Casopis
57(3), 247266.
Mitchell, T., 2005. Web Mapping Illustrated. OReilly, Sebastopol.
Mooney, P., Winstanley, A.C., 2009. Representation of environmental data in webbased GIS. In: Mount, N., Harvey, G., Aplin, P., Priestnall, G. (Eds.),
Representing, Modelling, and Visualising the Natural Environment. CRC Press,
Boca Raton, London, New York.
Olaya, V., 2008. Basic Land-Surface Parameters. In: Hengl, T., Reuter, H.I. (Eds.),
Geomorphometry: Concepts, Software, Applications. Elsevier, Amsterdam, pp.
330.
Otto, J.-C., 2008. Symbols for geomorphologic mapping in high mountains for
ArcGIS. http://www.geomorphology.at/index.phpoption=com_content&task=
view&id=133&Itemid=176.
Otto, J.-C., Dikau, R., 2004. Geomorphic system analysis of a high mountain valley
in the Swiss Alps. Zeitschrift fur Geomorphologie N.F 48, 323341.
Otto, J.-C., Gustavsson, M., Geilhausen, M., 2011. Cartography: design,
symbolisation and visualisation of geomorphological maps. In: Smith, M.J.,
Paron, P., Griffiths, J. (Eds.), Geomorphological Mapping: A Handbook of
Techniques and Applications. Elsevier, London.
Otto, J.C., Schrott, L., Jaboyedoff, M., Dikau, R., 2009. Quantifying sediment
storage in a high alpine valley (Turtmanntal, Switzerland). Earth Surface
Passarge, S., 1912. Physiologische morphologie. Mitt. Geogr. Ges. Hamburg 26,
135337.
Passarge, S., 1914. Morphologie des MeXtischblattes Stadtremda. Mittheilungen der
Geographischen Gesellschaft in Hamburg 28, 1221.
Pasuto, A., Soldati, M., Siorpaes, C., 1999. Geomorphological Map of the
Surroundings of Cortina dAmpezzo (Dolomites, Italy): Scale 1: 20,000. Dipartimento
di Scienze della Terra, Universita` di Modena e Reggio Emilia, Modena.
Pesci, A., Teza, G., Casula, G., et al., 2011. Multitemporal laser scanner-based
observation of the Mt. Vesuvius crater: characterization of overall geometry and
recognition of landslide events. ISPRS Journal of Photogrammetry and Remote
Sensing 66, 327336.
Pike, R.J., Wilson, S.E., 1971. Elevation-relief ratio, hypsometric integral, and
geomorphic area-altitude analysis. GSA Bulletin 82(4), 10791084.
Punkari, M., 1982. Glacial geomorphology and dynamics in the eastern parts of the
Baltic Shield interpreted using Landsat imagery. Photogrammetric Journal
Finland 9, 7793.
Reitmayr, G., Eade, E., Drummond, T., 2005. Localisation and interaction for
augmented maps. Proceedings of the 4th IEEE/ACM International Symposium on
Mixed and Augmented Reality. IEEE Computer Society, Washington DC, USA,
pp. 120129.
Robinson, A.H., Morrison, J.L., Muehrcke, P.C., Kimerling, A.J., Guptill, S.C., 1995.
Elements of Cartography, Sixth ed. Wiley, Chichester.
Rose, J., Smith, M.J., 2008. Glacial geomorphological maps of the Glasgow region,
western central Scotland. Journal of Maps v2008, 399416.
Sahlin, E.A.U., Glasser, N.F., 2008. Geomorphological map of Cadair Idris,
Wales. Journal of Maps 2008, 299314. http://dx.doi.org/10.4113/
jom.2008.1011.
Sato, H.P., Harp, E.L., 2009. Interpretation of earthquake-induced landslides
triggered by the 12 May 2008, M7.9 Wenchuan earthquake in the Beichuan
area, Sichuan Province, China using satellite imagery and Google Earth.
Schmidt, J., Evans, I.S., Brinkmann, J., 2003. Comparison of polynomial models
for land surface curvature calculation. International Journal of Geographic
Sclater, J.G., Lawver, L.A., Parsons, B., 1975. Comparison of long-wavelength
residual elevation and free air gravity anomalies in the north Atlantic and
possible implications for the thickness of the lithospheric plate. Journal of
Geophysical Research 80, 10311042.
Seijmonsbergen, A.C., de Graaff, L.W.S., 2006. Geomorphological mapping and
geophysical profiling for the evaluation of natural hazards in an alpine
catchment. Natural Hazards and Earth System Science 6, 185193.
Seijmonsbergen, A.C., Hengl, T., Anders, N.S., 2011. Semi-automated identification
and extraction of geomorphological features using digital elevation data.
In: Smith, M.J., Paron, P., Griffiths, J. (Eds.), Geomorphological Mapping:
Methods and Applications. Elsevier, London, chap. 11.
Sithole, G., Vosselman, G., 2004. Experimental comparison of filter algorithms for
bare-Earth extraction from airborne laser scanning point clouds. ISPRS Journal
of Photogrammetry and Remote Sensing 59, 85101.
Slaney, V.R., 1981. Landsat images of Canada a geological appraisal. Geological
Survey of Canada Paper 80-5, p. 102.
Smith, M.J., 2011. Digital Mapping: visualisation, interpretation and quantification

of landforms. In: Smith, M.J., Paron, P., Griffiths, J. (Eds.), Geomorphological
Mapping: Methods and Applications. Elsevier, London, chap. 8.
Smith, M.J., Chandler, J., Rose, J., 2009. High spatial resolution data acquisition
for the geosciences: kite aerial photography. Earth Surface Processes and
Smith, M.J., Clark, C.D., 2005. Methods for the visualisation of digital elevation
models for landform mapping. Earth Surface Processes and Landforms 30,
885900.
Smith, M.J., McClung, D.M., 1997. Avalanche frequency and terrain
characteristics at Rogers Pass, British Columbia, Canada. Journal of Glaciology
43, 165171.
Smith, M.J., Pain, C., 2009. Applications of remote sensing in geomorphology.
Smith, M.J., Rose, J., Booth, S., 2006. Geomorphological mapping of glacial
landforms from remotely sensed data: an evaluation of the principal data
sources and an assessment of their quality. Geomorphology 76, 148165.
Smith, M.J., Rose, J., Gousie, M.B., 2009. The cookie cutter: a method for
obtaining a quantitative 3D description of glacial bedforms. Geomorphology
108, 209218.
Smith, M.J., Wise, S.M., 2007. Problems of bias in mapping linear landforms from
satellite imagery. International Journal of Applied Earth Observation and
Geoinformation 9, 6578.
Smith, W.H.F., 1990. Marine geophysical studies of seamounts in the Pacific Ocean
Basin. Unpublished PhD thesis, Columbia University.
Smith, W.H.F., 1993. On the accuracy of digital bathymetric data. Journal of
Thomas, J.J., Cook, C.A., 2005. Illuminating the Path: The Research and
Development Agenda for Visual Analytics. IEEE Press, Washington.
Tooth, S., 2006. Virtual globes: a catalyst for the re-enchantment of
geomorphology? Earth Surface Processes and Landforms 31, 11921194.
Van Asselen, S., Seijmonsbergen, A.C., 2006. Expert-driven semi-automated
geomorphological mapping using a laser DTM. Geomorphology 78,
309320.
Van Wyckhouse, R., 1973. SYNBAPS. Technical Report TR-233, National
Oceanographic Office, Washington D.C.
Vencatasawmy, C.P., Clark, C.D., Martin, R.J., 1998. Landform and lineament
mapping using radar remote sensing. In: Lane, S.N., Richards, K.S., Chandler,
J.H. (Eds.), Landform Monitoring and Analysis. Wiley, New York.
Wang, Y., Mercer, B., Tao, V.C., Sharma, J., Crawford, S., 2001. Automatic
generation of bald earth digital elevation models from digital surface models
created using airborne IFSAR, Proceedings of 2001 ASPRS Annual Conference,
St. Louis, Missouri, USA.
Watts, A.B., 1976. Gravity and bathymetry in the central Pacific Ocean. Journal of
Watts, A.B., Daly, S.F., 1981. Long wavelength gravity and topography anomalies.
Annual Review of Earth and Planetary Sciences 9, 415448.
Welsh, A., Davies, T., 2010. Identification of alluvial fans susceptible to debris-flow
hazards. Landslides 8(2), 183194.
Werneke, J., 2008. The KML Handbook: Geographic Visualization for the Web.
Addison-Wesley Professional, San Francisco.
Wessel, P., 1998. An empirical method for optimal robust regionalresidual
separation of geophysical data. Mathematical Geology 30, 391408.
Wessel, P., Smith, W.H.F., 1998. New, improved version of generic mapping tools
released. EOS Transactions of the American Geophysical Union 79, 579.
White, R.S., 1993. Melt production-rates in mantle plumes. Philosophical
Transactions of the Royal Society of London Series A-Mathematical Physical and
Engineering Sciences 342, 137153.
Williams, R., Brassington, J., Vericat, D., Hicks, M., Labrosse, F., Neal, M., 2011.
Monitoring braided river change using terrestrial laser scanning and optical
bathymetric mapping. In: Smith, M.J., Paron, P., Griffiths, J. (Eds.),
Geomorphological Mapping: Methods and Applications. Elsevier, London,
chap. 20.
Wise, S., 2000. Assessing the quality for hydrological applications of digital
elevation models derived from contours. Hydrological Processes 14, 19091929.
Wood, J., Dykes, J., Slingsby, A., Clarke, K., 2007. Interactive visual exploration of
a large spatio-temporal dataset: reflections on a geovisualization mashup. IEEE
Transactions on Visualization and Computer Graphics 13, 11761183.
Wood, J.D., Fisher, P.F., 1993. Assessing interpolation accuracy in elevation
models. IEEE Computer Graphics and Applications 13, 4856.
Yokoyama, R., Shirasawa, M., Pike, R.J., 2002. Visualizing topography by
Openness: a new application of image processing to digital elevation models.
Geovisualization
325
Biographical Sketch
Mike J. Smith received a BS (Hons.) in geography from the University of Wales, Aberystwyth, an MS degree in
geography from the University of British Columbia, and a PhD degree in paleo-glaciology from the University of
Sheffield. He is currently a senior lecturer in GIS in the School of Geography, Geology and the Environment at
Kingston University, where he contributes to bachelor and masters programs on the application of remote sensing
in the geosciences. His research interests are based on the application of digital elevation models in geomorphology and specifically focused upon the visualization and geomorphometric modeling of glacial landscapes. Recent interests also include field spectroscopy of loess. He is the founder and editor of the Journal
of Maps.
John K. Hillier received BA and MS degrees in Natural Sciences (Geology) from the University of Cambridge, and
a PhD degree in Marine Geophysics from the University of Oxford. He was awarded a research fellowship at St.
Catharines College Cambridge and has worked for Zurich Insurance as a catastrophe modeler. He is currently a
lecturer at Loughborough University, with research interests in geohazards and quantitative geomorphology.
Recent interests include drumlin formation, volcano formation, evolution of oceanic tectonic plates, and the
correlation between natural perils.
Jan-Christoph Otto has studied Geography, Geology, and Soil Science at the Universities of Bonn and Grenoble.
He received his Diploma and PhD in Geography from the University of Bonn.
He is currently a postdoctoral researcher in the Department of Geography and Geology, University of Salzburg,
within the Geomorphology and Environmental Systems Research Group. His main research interests include
landscape change in high-mountain environments with special focus on sediment budgets and the dynamics and
consequences of permafrost changes in mountain areas. His recent interests include the creation of digital geomorphological maps and the combination of high-resolution surface and subsurface data using GIS technology.
Martin Geilhausen studied Geography at the University of Bonn and received a diploma investigating fluvial and
glacial deposits using geophysics. After working for a geoinformation company, where he developed webGIS and
web-mapping applications, he enrolled as a PhD student in the Department of Geography and Geology, University of Salzburg (Geomorphology and Environmental Systems Research Group). His research involves the
quantification of recent and postglacial sediment dynamics in glacier fore-fields, Austrian Alps. Martin is applying
a suite of field techniques ranging from fluvial erosion measurement and geophysics (ground penetrating radar,
seismic refraction, and resistive tomography) to terrestrial laser scanning.
4.1 Overview of Weathering and Soils Geomorphology

GA Pope, Montclair State University, Montclair, NJ, USA
4.1.1
4.1.1.1
4.1.2
4.1.2.1
4.1.2.2
4.1.2.3
4.1.2.4
4.1.3
4.1.3.1
4.1.3.2
4.1.3.3
4.1.3.4
4.1.4
References
Previous Major Works in Weathering and Soils Geomorphology

Relevant Topics not Covered in this Text
What Constitutes Weathering Geomorphology?
Weathering Voids
Weathering-Resistant Landforms
Weathering Residua: Soils and Sediments
Weathered Landscapes
Major Themes, Current Trends, and Overview of the Text
Synergistic Systems
Environmental Regions
Processes at Different Scales
Soils Geomorphology, Regolith, and Weathering Byproducts
Conclusion
1
3
5
5
6
6
6
6
7
7
8
8
9
9
Abstract
Weathering and soil geomorphology constitutes a specific subfield of earth surface processes, equally important in the
process system in creating the surface landscape. While weathering and soils are of interest to a broad range of sciences, their
role in geomorphology is specific and important. Weathering is the precursor to both chemical and mechanical erosion,
thus important to the generation of sediments carried by other surface processes. Weathering is essential to the creation of
soils. Weathering in all the cases is relevant to erosional and depositional landforms. However, specific in this volume are
the landforms and landform processes dominated by weathering.
yThe fundamental control on landscape evolution in erosional

landscapes is weathering.
(Phillips, 2005)
Weathering is the precursor to erosion, and further, all sediments in depositional landscapes derive from weathering. It is
not out of line to therefore conclude that all landforms are
weathering related. Alluvial, aeolian, glacial, and marine
sediments cannot exist without being liberated from their
parent rocks by weathering. Erosion is eased along by the
weakening of rock through weathering processes. Slopes are
defined by the rate of exposure due to weathering and removal
of material, and by the colluvial mantle derived from in situ
weathering and downslope movement. Landforms that resist
erosion also tend to resist weathering. However, there is a
specific class of landforms that are particularly dependent on
weathering processes, and these are the subject of weathering
geomorphology (Figure 1).
Pope, G.A., 2013. Overview of weathering and soils geomorphology. In:

Shroder, J. (Editor in chief), Pope, G.A. (Ed.), Treatise on Geomorphology,
Academic Press, San Diego, CA, vol. 4, Weathering and Soils
Soils are a direct byproduct of weathering, so are included

with weathering geomorphology. Different authors present
different definitions for what encompasses soil geomorphology. Daniels and Hammer (1992: 1) used a methodological
definition, ythe application of geologic field techniques and
ideas to soil investigations; essentially, an extension of sedimentology to soil science. Birkeland (1999) was more encompassing, and somewhat reverse of Daniels and Hammer,
applying soils and pedogenesis toward the study of landform
evolution and dynamics. Gerrards (1992: 2) definition, also
adopted by Schaetzl and Anderson (2005), stated simply that
soil geomorphology strived to explain the genetic relationships between soils and landforms, a two-way view toward
both pedogenesis and landform evolution.
4.1.1
Previous Major Works in Weathering and Soils

Geomorphology
This edited text builds on a heritage of important publications

in the fields of weathering and soils geomorphology. First, it is
germane to define what this volume is, and is not. The aim here
is to focus on the science of geomorphology, consistent with
the intent for the Treatise on Geomorphology series. The study of
weathering is not necessarily limited to geomorphology. Weathering is also pertinent to studies in mineralogy and petrology,
http://dx.doi.org/10.1016/B978-0-12-374739-6.00056-7
Overview of Weathering and Soils Geomorphology
Figure 1 A weathering landform. This granite tor, near Divide,

Colorado (in the Pikes Peak Massif), is a resistant remnant of deep
weathering. The landform owes its existence primarily to long-term
weathering processes. Genesis involved the two-stage saprolite
stripping model (see Chapter 4.8), first with long-term attack of
chemical weathering agents progressively deeper into the bedrock,
then later stripping of weathered material, revealing the resistant
corestone boulders. Subsequent to exposure, mechanical weathering
by way of ice, pressure release, and possible thermal extremes
further shaped the rock as chemical weathering continued. Endogenic
factors influenced the outcome: composition and joint density of the
rock, regional tectonic uplift modifying base level (erosion) as well as
climate. As with most weathering landforms, other geomorphic
agents played a role as well, in this case slope colluvium movement
and fluvial erosion removed weathered material, each responding
differently in warm, mild, or cold climates known to exist in this
region over the past several million years.
sedimentology, biogeochemistry, and mechanical engineering,

to name several fields. Naturally, there is overlap to be found.
Presently, there is considerable activity concerning weathering
and biogeochemistry (cf. Drever, 2005 for a good reference,
part of the Treatise in Geochemistry series). Some of this activity
pertains directly to geomorphology, while the remaining research trends toward topics such as soil processes, hydrology,
and chemical cycling. Likewise in soil science, not all studies
pertain to geomorphology. In the very broad field of soil
science, geomorphology sustains relative minority interest,
with the much larger emphasis trending toward agronomy
and ecology, hydrology, pollution, and engineering. The present volume strikes a course more specifically relevant to
geomorphology. Though, in the interest of cross-fertilization,
readers are encouraged to explore these allied fields. A
quick survey of the references cited within this volume
reveals diverse sources of information and ideas, proof of the
interdisciplinarity expected within weathering and soil
geomorphology.
Soil science has a long heritage (covered well in Schaetzl
and Anderson, 2005), and the soil aspect of geomorphology
has been recognized at least as far back as Dokuchaev and
Glinka (late nineteenth and early twentieth centuries), who
included a topographic factor in soil genesis (refer to Marbuts
translation, Glinka and Marbut, 1927). Later authors such as
Jenny (1941), Hole (1953), and Ruhe (1956) further advanced

the ideas of soil geomorphic relationships.
Weathering geomorphology was recognized early on as one
of the several key earth surface processes. Strabo and Herodotus observed the weathering processes in building stones
(Camuffo, 1992). Said (1950) reported sophisticated concepts
of weathering processes from a tenth century Arabic writing,
The Discourses of the Brothers of Purity. By the time of the scientific revolution in geology in the late eighteenth and early
nineteenth centuries, writers incorporated weathering, though
briefly mentioned, as part of the rock cycle (as best understood at that time). Playfair (1802) was foremost among
these pioneers to first elaborate on the system of geomorphology, including the role of weathering. Scattered examples
of weathering research followed (e.g., Yates, 1830; Bischof,
18541859; Kinnahan, 1866; Gilbert, 1877). Van den Broeks
(1881) and Merrills (1897) works were the first major treatises
published concerning the geomorphic and geologic aspects of
weathering.
Still, weathering geomorphology has held a fraction of the
interest compared to the other subfields of geomorphology.
According to Yatsu (1988), in his comprehensive review of the
foundation of weathering science, early twentieth century
geomorphologists were preoccupied with erosion cycles to the
exclusion of other aspects of landform modification (such as
weathering). Evident in Table 1 below is a skew toward more
recent writings, demonstration of the growing interest in
weathering geomorphology since the 1960s. Momentum
launched by the successful British Geomorphological Research
Group (BGRG) symposium in 1992 (Robinson and Williams,
1994) led to two successful symposia in recent years, attracting
a broad international audience. Weathering 2000 (held in
Belfast, Northern Ireland), and the 2005 Binghamton Symposium in Geomorphology (in Lexington, Kentucky) both
featured weathering and landscape evolution. On an ongoing
basis annually, the Goldschmidt Conference attracts several
themed sessions concerning weathering geochemistry, and the
weathering focus on stone conservation produces serial
meetings, such as those of the Stone Weathering and Atmospheric Pollution Network (SWAPNet). The national and specialized organizations in Europe, Australia, Asia, and North
America regularly see themed sessions on weathering geomorphology at their annual meetings as well. The field has
attracted new participants, but credit is also due to collaborative efforts such as SWAPNet, the Critical Zone Exploration Network, and interdisciplinary symposia to connect
already active researchers in allied and disparate fields, like
minds working on similar ideas worth sharing.
Table 1 presents an overview of the major previous works
in weathering and soils geomorphology. While thorough, it is
not necessarily exhaustive (no slight is intended for any
omissions), but provides a reasonable cross-section over time.
In addition to these texts, research articles on soils and
weathering geomorphology are regularly found in the main
geomorphology journals; Geomorphology, Catena, and Geoderma (Elsevier), Earth Surface Processes and Landforms (Wiley),
Zeitschrift fur Geomorphologie (Schweizerbart) receiving the
most coverage. Still, dozens of periodicals beyond these listed
are apt to find weathering and soils geomorphology within
their interest. These include many subfields in geology,
Table 1 Major works in weathering and soils geomorphology over the last half century. A separate and large literature concerning soils and
paleoenvironments is not included here apart from several examples under the category of other
Author and date
Title
Broad themes or topics
Texts specifically concerning weathering

Bland and Rolls (1998)
Boyer (1971)
Carroll (1970)
Colman and Dethier (1986)
Delvigne (1998)
Drever (1984)
Drever (2005)
Goudie and Viles (1997)
Kittrick (1986)
Lerman and Meybeck (1988)
Loughnan (1969)
Marshall (1964, 1977)
Martini and Chesworth (1992)
Nahon (1991)
Ollier (1984)
Ollier and Pain (1996)
Reiche (1950)
Taylor and Eggleton (2001)
Yatsu (1988)
Weathering: An Introduction to the Scientific Principles

Field Guide to Rock Weathering
Rock Weathering
Rates of Chemical Weathering of Rocks and Minerals
Atlas of Micromorphology of Mineral Alteration and Weathering
The Chemistry of Weathering
Surface and Ground Water, Weathering, and Soils
Salt Weathering Hazards
Soil Mineral Weathering
Physical and Chemical Weathering in Geochemical Cycles
Chemical Weathering of the Silicate Minerals
The Physical Chemistry and Mineralogy of Soils, vols. 1 and 2
Weathering, Soils, and Paleosols
Introduction to the Petrology of Soils and Chemical Weathering
Weatheringa
Regolith, Soils, and Landforms
A Survey of Weathering Processes and Products
Regolith Geology and Geomorphology
The Nature of Weathering
1
1
1
1
2
2
1
1
1
3
1
2
2
1
1
1
1
1
2
Conference proceedings or special journal issues devoted to weathering and/or soils geomorphology
Kneupfer and McFadden (1990)
Soils and Landscape Evolution (Binghamton Symposium)
Robinson and Williams (1994)
Rock Weathering and Landform Evolution (from BGRG 1992)
Deriving from the Weathering 2000 conference:
Smith and Turkington, 2004
Stone decay: its causes and controls (from Weathering 2000)
Warke, 2001
Weathering 2000 (Special issue, ESP and L)
Whalley and Turkington, 2001
Weathering and Geomorphology (Special Issue, Geomorphology)
Turkington et al. (2005)
Weathering and Landscape Evolution (Binghamton Symposium)
2
3
2
2
34
7
3
3456
3
245
23
2
4
4
3
2
2
2
2
2
3
5
4
3
3
3
3
3
4
5
4
5
6
4
4
5
6
56
5
6
1356
123457
1
1
1
1
2
2
2
2
3
3
3
3
67
7
56
567
Texts specifically concerning soil geomorphology

Birkeland (1999)
Soils and Geomorphology, 3rd ed.a
Catt (1986)
Soils and Quaternary Geology
Daniels and Hammer (1992)
Soil Geomorphology
Gerrard (1992)
Soil Geomorphologya
Jungenius (1985)
Soil and Geomorphology (Catena supplement)
Schaetzl and Anderson (2005)
Soils: Genesis and Geomorphology
1
1
1
1
1
1
2
4
4
3
3
3
34567
567
57
Other texts with substantial coverage of weathering or soils geomorphology

Bourke and Viles (2007)
A Photographic Atlas of Rock Breakdown Featuresy
Brunsden (1979)
Process in Geomorphology (Chapter 4)
Buol et al. (2003)
Soil Genesis and Classificationa
Dorn (1998)
Rock Coatings
Gerrard (1988)
Rocks and Landforms
Migon (2006)
Granite Landscapes of the World
Nash and McLaren (2007)
Geochemical Sediments and Landscapes
Scott and Pain (2008)
Regolith Science
Small (1982)
Slopes and Weathering
Thomas (1994)
Geomorphology in the Tropicsa
Twidale (1982)
Granite Landforms
Yatsu (1966)
Rock Control in Geomorphology
Young et al. (2009)
Sandstone Landforms
1
1
1
1
1
1
1
1
1
1
1
1
1
2
3
3
2
2
2
2
2
3
3
2
2
2
3
5
4
3
3
3
3
4
456
456
5
46
456
56
56
3
3
Indicates that earlier editions were published, but not included here.
Broad themes and topics: 1. Geomorphology. 2. Minerals or petrology. 3. Weathering factors. 4. Geochemistry and hydrochemistry. 5. Pedogenesis. 6. Palaeoenvironments.
7. Education/Field Methods. All the above include weathering or soil processes as a topic.
physical geography, soil science, hydrology, environmental

change, ecology, material science, and heritage conservation,
as well as the general science venues such as Nature and
American Journal of Science.
4.1.1.1
Relevant Topics not Covered in this Text
Several topics relevant to weathering geomorphology are not

covered in this text. Fortunately, there are excellent current
(a)
(b)
(c)
(e)
(d)
works on the topics of karst geomorphology, biogeochemical

landscapes, and weathering dating methods, which readers
may avail beyond this text.
Karst geomorphology is a direct result of weathering processes: the solution weathering of susceptible rocks (such
as limestone, marble, dolomite, sometimes sandstone, or
quartzite) to produce holes, caverns, and depressions of various scale. Authors in this volume do discuss solution as a
process. But, the breadth and importance of karst geomorphology is sufficient to warrant full coverage in a dedicated text. Hence, the Treatise on Geomorphology series does
include a separate volume.
Weathering plays an important role in biogeochemical
cycles, and in defining the geochemical landscape. This
geochemical landscape has geomorphic expression (such as
in pedogenesis, weathering mantles, and chemical denudation). The geochemical landscape can also be regarded
without geomorphology, as some research in hydrochemistry,
mineralogy, pedology, and ecology does. Several chapters
herein make mention of geochemical cycles related to
soils and weathering, but there is no specific chapter on
geochemical landscapes. These concepts are well covered in
Drever (2005).
A number of dating methods are dependent on weathering,
if not entirely based on weathering: weathering rind thickness,
rock coating properties, degree of mineral etching, mineral
depletion and sediment maturity, rock face recession rates,
rock integrity, morphology of mesoscale weathering features
(such as tafoni), and soil profile development. All are at least
capable of relative dating of geomorphic features or surfaces,
some are useful as calibrated quantitative dating methods. The
use of weathering dating methods dates back to Blackwelder
(1931) and Kay (1931), a larger proliferation of quantitative
and semiquantitative methods began in the late 1960s and
early 1970s (cf. Brookes, 1981 for review). Continuing studies
advance the utility of weathering dating methods wedded to
modern technology. The basis for some of these dating
methods are covered in chapters in this volume by Dorn
(Chapter 4.5), Oguchi (Chapter 4.6), Paradise (Chapter 4.7),
Dixon (Chapter 4.3), and Schmid (Chapter 4.16). Readers are

encouraged to reference works by Birkeland (1999), Martini
and Chesworth (1992), Dorn and Phillips (1991), Catt
(1986), Mahaney (1984), Drever (1984), and Brookes (1982)
for a more complete foundation on dating methods.
Paleosols, like soils, are originally pedogenic and thus
weathering related. Paleosols are, however, more stratigraphic
than geomorphic, though useful in establishing time and past
environments useful for geomorphology. They can be altered
by diagenetic processes similar to weathering when buried in
deeper geologic context. Works by Wright (1986), Martini and
Chesworth (1992), and Retallack (2008) provide background
in this active field.
4.1.2
What Constitutes Weathering Geomorphology?
It is convenient to parse out the individual processes or individual environments for purposes of discussion. Weathering
landforms can be defined in various ways, depending on the
frame of reference and how much erosion is paired with the
weathering process. The discussion of weathering-related
landforms and soil-based geomorphology may be classified by
means of generalized morphology: weathering voids, weathering resistance, weathering residua (including soils), and
weathered landscapes. This classification has an inherent scalar and temporal organization (Figure 2). At increasing spatial
and temporal scales, specific individual weathering processes
diminish in importance replaced by the works of the entire
weathering system.
4.1.2.1
Weathering Voids
Weathering, with erosion, creates voids, in other words, places

where rock used to be. To borrow from visual arts, this is the
negative space of geomorphology, and sculptor Henry Moores
assertion that a hole can itself have as much shape-meaning
as a solid mass (Friedenthal, 1963: 251) is as relevant to
Figure 2 Weathering morphology at increasing scale. (a) A sand grain, seen in cross section with backscatter scanning electron microscope,
approximately 1000 mm (1 mm) in diameter. The grain was sampled from an in situ soil profile with granitic parent material, on a marine terrace
near Carmel, CA, USA. The grain exhibits differential weathering along grain and crystal weaknesses. The lower half of the grain is relatively
unweathered quartz, the upper half is plagioclase, riddled with secondary porosity from dissolution (Pope, 1995). (b) Rock coatings, case
hardening, and differential weathering of aeolian-bedded, Triassic Aztec Sandstone, Valley of Fire State Park, NV, USA. Iron and manganese rock
coatings, derived from weathering and dust, impart a dark color to the outer surface. Cementation of an outer crust or rind, also a product of
weathering, protects parts of the rock face, though more rapid weathering occurs when this crust is breached, particularly vigorous along
joints and bedding planes. (c) Tafoni and cavernous weathering in Cambrian Remarkable Granite (Fairclough, 2008), Remarkable Rocks,
Kangaroo Island, South Australia. The rock pictured is approximately 10 m high. Proximity to the ocean shoreline adds moisture and salt,
both aggressive weathering agents. The cavernous weathering forms are partially governed by internal rock weaknesses and surface induration
(case hardening), but also a regulated pattern of intersecting surface declivities that positively enhance the weathering environment by
retaining moisture and surface organisms. (Photo courtesy of Dr. P. Beyer, Bloomsburg University.) (d) Granite pinnacles and tors in Cathedral
Park (Pikes Peak batholiths), near Victor, CO, USA (cf. Blair, 1976). Vertical relief of the rock pinnacles is B60 m, based on the proportion of
the trees on the slopes. Weathering is joint controlled, and a covering of saprolite developed along the tertiary erosion surface has been
exhumed during more recent uplift. Fluting from subaerial weathering and erosion is visible, as well as pressure-release jointing parallel to
rock surfaces. (e) A weathered landscape, centering on Kata Tjuta (The Olgas), NT, Australia. The Kata Tjuta inselbergs, composed of
Cambrian Mt. Currie Conglomerate, are remnant and resistant landforms in an exhumed etch plain, in which the surrounding weathered
regolith has been stripped (Twidale, 2010). Fracture control is evident. The inselberg group rises 500 m above the plain, and spans about 8 km;
the left-to-right dimension of this image is approximately 15.5 km. (Image source: NASA Earth Observatory (http:earthobservatory.nasa.gov),
acquired by Landsat 7 Enhanced Thematic Mapper plus (ETM ), October 16, 1999, false-color composite (bands 7, 4, and 1) image with
panchromatic.)
weathering as it is to art. Lack of weathering product (residua,

particles, solutes) implies that erosion necessarily follows
weathering to create these landforms. What remains are holes
or declivities by many names: rillenkarren, tafoni, cavernous
weathering, alveolar weathering (see Chapter 4.7). Most of
these are small- and medium-scale features. Even so,
weathering voids are apparent from submillimeter (see
Chapter 4.5) to landscape scale. At the scale of landscapes,
weathering paired with erosion derives the concept of denudation. Lowering of land surface can be observed in small
scale, as in small drainage basins, but this is the one form of
weathering that can assume regional or even continental relevance. It is quantified more often as dissolved load carried by
rivers over years.
4.1.2.2
A second class of weathering landform is actually lack of

weathering effectiveness, or resistance to weathering. There is
variability in weathering and erosion, and areas that are resistant stand out prominently. Landforms that are resistant to
erosion are first resistant to weathering; hence, it is possible to
classify resistant bodies as weathering-related landforms. Von
Engeln (1948: 276) considered this tenet to be the Law
of Weathering: differential weathering causes resistant beds
and structures to stand out in relief (his use of italics), as
valid as any of the laws of geomorphology (cf. Rhodes and
Thorn, 1996). A wide variety of resistant forms permeate the
literature and language of geomorphology: bornhardts, inselbergs, monadnocks, tors, castle kopjes, hoodoos, towers, fins,
and others. (Olliers, 1969 original text provided a comprehensive description of the myriad weathering landform
names.) These tend to be medium to large scale, the largest
being entire plateaus of resistant rock, such as the tepuis
of the highland region at the BrazilGuianaVenezuela
frontier. Even so, small-scale rock knobs and pedestals
are equivalent examples of this process, and weathering-resistant forms are evident even at microscopic scales (see
Chapter 4.5).
4.1.2.3
the weathering residuum (cf. Phillips et al., 2005), the outer

skin of the earthatmospherebiosphere interface (Figure 3).
Most soil, essentially any soil containing mineral matter, is a
weathering product at least partially. Mineral soils are either
in situ residua of bedrock weathering, or formed in the
transported sediment deriving from previously weathered
material (which still continues to weather as part of pedogenic
processes).
4.1.2.4
Weathered Landscapes
Any terrestrial surface is exposed to weathering, though some

landscapes retain a more obvious weathering signature because of aggressive weathering, lithology particularly vulnerable to weathering, long-term exposure, and relative lack of
disturbance from other geomorphic processes. Multiple weathering agents, defining factors, and weathering morphologies
are apparent at the landscape scale.
Regions with thick weathering mantles as well as areas
stripped of weathering mantle to expose the basal weathering
front (etchplains) are examples of weathered landscapes. The
classic definition of weathering-limited and transport-limited
landscapes is often used to define the presence (or absence) of
regolith or exposed bedrock. The most simplistic illustration is
that in areas where weathering is less rapid or extensive, erosive processes are more efficient, cleaning off the weathering
material to reveal more bedrock. Transport limitation is
the opposite, in which erosion processes are less efficient in
removing weathered material, allowing more of a residual
mantle to accumulate with the soil. The presumption is that
aggressive weathering, e.g., in the tropics by whatever accelerated means, should allow for increased regolith mantle,
while less intense or extensive weathering (i.e., in areas defined by early researcher as weakly weathered) allows regolith
stripping to dominate. Of course, weathering intensity is but
one of several factors that allow regolith accumulation or
stripping, the others being slope, uplift with respect to the base
level, vegetation cover, and time. In any event, weathering is an
important factor, so that it is possible to discuss regolith
mantle accumulation as a weathering-related landform. Weathering mantles can be seen as medium to large scale.
Weathering Residua: Soils and Sediments
Byproducts of weathering become part of the geomorphic

environment. Weathering byproducts occur as secondary
minerals such as clays and iron and aluminum oxides (in
which original mineralogy has been altered), as smaller particles (essentially, pieces of larger mineral crystals or rocks),
and as dissolved elements in solution with ground and surface
waters. All three byproducts may exist in situ within a weathering and soil profile, or may be transported. Transported
particles become part of the detrital sediment system. Transported secondary minerals and solutes can become cementing
agents in rock and soil.
Soils are influenced by geomorphology (among other factors, Jenny, 1941), a fact well covered by several existing texts
(cf. Schaetzl and Anderson, 2005; Birkeland, 1999; Ollier and
Pain, 1996; Gerrard, 1992). More relevant in to this volume,
soils are also part of the geomorphology, and a pathway to
understanding geomorphic processes. The soil is a segment of
4.1.3
Major Themes, Current Trends, and Overview

of the Text
Of the myriad themes incorporating weathering and soils

geomorphology, several trends can be traced, and these form
the basis of organization for this book: Synergistic systems,
environmental regions, processes and morphology at different
scales, and soils and other weathering residua.
This volume is divided into the four general themes,
allowing a bit of connective overlap between themes and
chapters. The aim of this organization is to provide a system of
content while capturing the current state of knowledge within
the field. The first three chapters, the present one included,
provide an introduction to weathering and soils geomorphology, emphasizing the synergistic nature of weathering and
pedogenic processes. Following this introduction, five chapters
Atmosphere
Lithosphere
Weathering system
Soil system
Biotic system
Rocks
Soil
Regolith
weathering
products
Minerals
Organic
matter
Weathering
processes
Pedogenic
processes
Decomposition
processes
Geomorphic system
Hydrosphere
Figure 3 The weathering and soil systems reside within the interface between the hydrosphere, lithosphere, and atmosphere at Earths surface.
The biosphere also exists at this interface, and contributes to both weathering and pedogenic processes. Soil and weathering systems are part of
the geomorphic system, the focus of interest for this text. Precise interactions within the geosystem interface are considerably more complex
than indicated in this figure, but encompassed within the chapters herein. Modified from White, I.D., Mottershead, D.N., Harrison, S.J., 1992.
Environmental Systems: An Introductory Text, Second ed. Chapman and Hall, London, 616 pp.
elaborate on weathering processes and factors in different

environmental zones on the planet: tropical regions, cold
regions, arid regions, and coastal environments. Next, a series
of chapters detail the processes responsible for different weathering morphologies. These chapters are arranged in increasing spatial scale, beginning at the submicron scale,
continuing to the scale of rock coatings and weathering rinds,
then to the scale of rock depressions. Slopes and weathering
mantles finalize the upper range of scale morphologies. The
text completes with three chapters focusing on soils and
sediments: the geomorphic processes of soil catenas, the utility
of soil chronosequences in geomorphology, and the role of
weathering in producing sediments.
4.1.3.1
this text concerns geomorphology, hence the focus. The weathering/soil interface has also been termed the Critical Zone
(Anderson et al., 2004; Brantley et al., 2006), an apt term
given its importance to life and biogeochemical cycles. As
defined by its authors, the Critical Zone extends from the
vegetation canopy to groundwater, so includes nonlithosphere
materials. Alternately, regolith also encompasses the interface
from a strictly lithosphere skin perspective, and this term
maintains preference in some circles. Regardless, the omnifarious scope afforded by either term has become both impetus and focus for new research. The introductory chapters in
this text by Heather Viles (Chapter 4.2) and John Dixon
(Chapter 4.3) on weathering and soil geomorphic processes,
respectively, provide the needed sense of interacting systems
relevant to the chapters that follow.
Synergistic Systems
A dominant premise throughout research in the last two

decades is the realization of interactive, synergistic systems in
the environment, and the soil and weathering realm is one of
the best examples of synergism on the planet. Every author
presenting in this text embraces the concept that weathering
and soils are complex interacting systems residing within the
lithospherehydrosphereatmosphere interface (Figure 3; see
also Chesworth, 1993). Weathering and soils are not always
part and parcel with geomorphology, though the impetus for
4.1.3.2
Environmental Regions
From the early pioneering studies to present, the science of

weathering and soils geomorphology has been rooted in an
environmental context: the environment is one of the dominant if not most dominant determinants in process and
outcome. The environmental context is obvious: weathering
processes and pedogenesis require exogenic factors. Climate is
often and readily recognized as the overriding factor, in turn
influencing hydrology and biota. As with the greater field of

geomorphology, there is a risk of assigning too much influence of any single factor. Other influences (such as lithology
and tectonic history) are equally if not sometimes more important. That said, this volume maintains a discussion of environmental factors classified by environment, in Chapters
4.11 (the tropics, by Pope), 4.12 (arid regions, by Warke), 4.15
and 4.14 (cold regions, by Hall and Dixon, respectively), and
4.13 (coastal areas, by Mottershead). Why perpetuate this organization The first reason harkens to the first premise of the
book: interactive systems. The holistic environmental system is
easily approached in the context of specific and unique environments. The second reason is to embrace an extant literature that is already organized by means of global
environmental regions. To reclassify the deep heritage of, say,
arid lands geomorphology, or tropical geomorphology, or
cold regions geomorphology, would involve an effort beyond
the scope of this volume, but may prove to be an interesting
exercise for future researchers.
Readers will note a lack of temperate regions set aside
with a unique chapter. These midlatitude regions are susceptible to frequent and noticeable climate change, such that
significant weathering environments are subsumed under
discussions of tropical, arid, or cold conditions over geologic
time spans, presuming that these extremes dominate the
weathering environment (see also discussions in Chapter 4.11). There are two unique additions of this text not addressed extensively in previous works. The first involves
separate chapters for chemical and mechanical weathering in
cold regions. While difficult to discuss one without the other,
advances in both subfields are significant and groundbreaking
enough to warrant separate treatments. The second original
distinction in this text is the inclusion of coastal environments
as a separate environment, limited in areal extent but distinct
in the weathering agents and factors, and significant to coastal
geomorphology.
4.1.3.3
Processes at Different Scales
Geomorphology readily recognizes scale-dependent (and

sometimes scale-independent) processes, weathering and
pedogenesis included. The detailed processes responsible for
weathering and pedogenesis have seen increased investigation,
particularly at the extremes of microscale and regional scale.
The third section of this book engages the analysis of weathering and soil geomorphic processes to explain morphologies from the smallest to largest scales.
New and widespread availability of research technology
(such as electron microprobes, scanning electron microscopes,
atomic force microscopes, confocal laser microscopes, tomography, cf. Lee, 2010; Ip et al., 2010; Ersoy et al., 2010; Krinsley
et al., 1998; Dorn, 1995) afford the opportunity to visually
and chemically observe weathering phenomena at small
scales. Processes of dissolution, microfracturing, rind formation, and case hardening, are apparent at the mineral grain
scale and as deep as the molecular and atomic scale. Dorn,
Gordon, Krinsley, and Langworthy (Chapter 4.4) extend the
frontier to unprecedented fine extremes at the nanoscale, at
the scale of crystal lattices and molecules; the phenomena
observed here directly explain and encompass the wider appreciation of microscale weathering. One byproduct of nanoand microscale weathering is the production of rock coatings,
discussed by Dorn (Chapter 4.5). Rock coatings have their
own morphology and dynamics, with many of the same
considerations as larger scale stratigraphy, but also the unique
chemical and environmental interactions at the microscale.
Higher in scale dimensions, weathering rinds (Oguchi,
Chapter 4.6) are the product of weathering processes at the
smaller scales, but are visible in a range from millimeters to
centimeters (B103101 m).
Mesoscale features measure on the order of meters
(B101101 m), which can include residual boulders (left
from in situ weathering) or talus (derived from weathering,
discussed in Section 4.1.3.4, and to a lesser degree by Hall,
Chapter 4.15), unloading sheets from domes, as well as
weathering void features such as rillenkarren and tafoni
(Figure 2(c)). Examples of the mesoscale are covered by
Paradise (Chapter 4.7), process and factors in tafoni and other
rock basins, while Mottershead (Chapter 4.13) also discusses
tafoni in the coastal context.
Finally, at the macroscale, weathering features assume the
dimensions of slope surfaces and landscapes (B101104 m),
including vertical relief and subsurface depths of many meters.
Large weathering-related structures such as tors (Figure 1),
domes (Figure 2(d)), and inselbergs (Figure 2(e)); weathering-controlled slopes; weathering mantles; and etchplains
and exhumed weathering profiles are found at this scale.
Migon covers macroscale features of weathering mantles and
landscapes (Chapter 4.8) and hill slope formation (Chapter 4.10) while Pope (Chapter 4.11) and Dixon (Chapter 4.3)
include examples of macroscale forms in their respective
chapters on tropical weathering and pedogenesis.
4.1.3.4
Soils Geomorphology, Regolith, and Weathering

Byproducts
One of the key approaches to understanding geomorphology

is through soils. The soil is the living shell of the deeper
geology and the interactive conduit between the external atmosphere/hydrosphere/biosphere (Figure 2) to the deeper
geology. As such, soil is sensitive to geology as well as surface
processes, including the living environment. Soil development, being time dependent, is directly indicative of the rate
and dynamism of surface processes, certainly at a relative
comparison if not in more concrete and quantitative way.
Thus, soils tell a geomorphic story. And, as mineral soils are
nearly inseparable from weathering processes, continuous
with the greater scale of regolith, soil geomorphology also
overlaps with weathering geomorphology.
Soils have a diffuse and ill-defined transition into nonsoil
regolith. (The term, regolith, most often includes soils, but for
lack of a distinguishing term, use of regolith in this instance
extends from the soil.) Regolith provides investigators a larger
scope on landscape dynamics, as well as interaction with
geohydrologic systems and greater contact with geology.
Regolith science has been highly active in recent years in a
variety of themes. The aforementioned impetus for research in
the biogeochemical systems of the critical zone includes
dedicated long-term studies from a range of environments

(Brantley et al., 2006). As a long-term feature, regolith is a
focus of studies on landscape evolution, aided by the capabilities of dating control (cf. Taylor and Eggleton, 2001;
Phillips, 2005; Pillans, 2008). Applications in mineral resource exploration extend from these landscape evolution
studies (Jones, 1998; Butt et al., 2000). As earth-based geomorphology is applied to other worlds, so too are studies of
regolith landscapes on Mars, the Moon, and other celestial
bodies (Lindsay, 1992; Vanimen and Chipera, 2006; Clark,
2008).
Addressing soil geomorphology and weathering byproducts in this volume, Dixon (Chapter 4.3) provides an overview of soil dynamics and pedogenesis from the geomorphic
perspective. The specific controls of soil profiles modified by
and influencing slope processes, the catena, are covered by
Schaetzl (Chapter 4.9), while Migon (Chapter 4.10) incorporates soils and deeper regolith in his discussion of slope
geomorphology. Dixon (Chapter 4.14), Warke (Chapter 4.12)
and Pope (Chapter 4.11) discuss soils in their respective environmental region chapters. Schmid (Chapter 4.16) explains
the important relationships and geomorphic tools provided
by soil chronosequences. Last, as addendum to the weathering
geomorphology system, Pope (Chapter 2.6) relates the role of
weathering in the production of detrital sediments, relevant in
the erosion, transport, and deposition systems of geomorphic
processes, and a connection to the field of sedimentology.
4.1.4
Conclusion
Given the connection to erosional and depositional processes,

weathering and soils interface with every aspect of geomorphology, and with any branch of earth science pertaining
to surface processes. This volume on weathering and soil
geomorphology thus provides some essential new material of
potential interest to those interested in a foundation in the
field of geomorphology. Some examples of areas where less is
known than perhaps could be researched are also provided,
which will enable development of future new research.
Overall, the subdisciplines of weathering geomorphology and
soil geomorphology continue to provide fertile ground for
those interested in pursuing a difficult but fruitful area of
knowledge.
References
Anderson, S.P., Blum, J., Brantley, S., et al., 2004. Proposed initiative would study
Earths weathering engine. Eos 85(28), 265272.
Birkeland, P.W., 1999. Soils and Geomorphology. Oxford University Press, Oxford,
430 pp.
Bischof, G., 18541859. Elements of Chemical and Physical Geology (volumes
13). The Cavendish Society, London, 2832 pp.
Blackwelder, E., 1931. Pleistocene glaciation of the Sierra Nevada and Basin
Ranges. Bulletin of the Geological Society of America 42, 865922.
Blair, Jr. R.W., 1976. Weathering and geomorphology of the Pikes Peak Granite in
the southern Rampart Range, Colorado. In: Epis, R.C., Weimer, R.J. (Eds.),
Studies in Colorado Field Geology. Professional Contributions of Colorado
School of Mines, Colorado, vol. 8, pp. 6872.
Bland, W., Rolls, D., 1998. Weathering: An Introduction to the Scientific Principles.
Arnold, London, 271 pp.
Bourke, M.C., Viles, H.A. (Eds.), 2007. A Photographic Atlas of Rock Breakdown
Features in Geomorphic Environment. Planetary Science Institute, Tucson,
79 pp.
Boyer, R.E., 1971. Field guide to rock weathering. Earth Science Curriculum Project
Pamphlet Series, PS-1. Houghton Miflin, Boston, 38 pp.
Brantley, S.L., White. T.S., White, A.F., et al., 2006. Frontiers in Exploration of the
Critical Zone: Report of a Workshop Sponsored by the National Science
Foundation (NSF), 2426 October 2005, Newark, Delaware, USA, 30 pp.
Brookes, I.A., 1982. Dating methods of pleistocene deposits and their problems:
VIII, weathering. Geoscience Canada 9(4), 188199.
Brunsden, D., 1979. Weathering. In: Embleton, C., Thornes, J. (Eds.), Process in
Geomorphology. Edward Arnold, London, pp. 73129.
Buol, S.W., Southard, R.J., Graham, R.C., McDaniel, P.A., 2003. Soil Genesis and
Classification, Fifth ed. Iowa State University Press, Ames, Iowa, 499 pp.
Butt, C.R.M., Lintern, M.J., Anand, R.R., 2000. Evolution of regoliths and
landscapes in deeply weathered terrain implications for geochemical
exploration. Ore Geology Reviews 16(34), 167183.
Camuffo, D., 1992. Acid rain and deterioration of monuments: how old is the
phenomenon? Atmospheric Environment 26B, 241247.
Carroll, D., 1970. Rock Weathering. Plenum Press, New York, 203 pp.
Catt, J.A., 1986. Soils and Quaternary Geology: A Handbook for Field Scientists.
Clarendon Press, Oxford, 267 pp.
Chesworth, W., 1993. Weathering systems. In: Martini, I.P., Chesworth, W. (Eds.),
Weathering, Soils, and Paleosols. Elsevier, Amsterdam, pp. 1939.
Clark, J.D.A., 2008. Extraterrestrial regolith. In: Scott, K.M., Pain, C.F. (Eds.),
Regolith Science. Springer Science and CSIRO, Dordrecht, The Netherlands, and
Collingwood, Australia, pp. 377407.
Colman, S.M., Dethier, D.P. (Eds.), 1986. Rates of Chemical Weathering of Rocks
and Minerals. Academic Press, Orlando, 603 pp.
Daniels, R.B., Hammer, R.D., 1992. Soil Geomorphology. Wiley Interscience, New
York, 236 pp.
Delvigne, J.E., 1998. Atlas of Micromorphology of Mineral Alteration and
Weathering. Mineralogical Association of Canada, Ottawa, 494 pp.
Dorn, R.I., 1995. Digital processing of back-scatter electron imagery: a microscopic
approach to quantifying chemical weathering. Geological Society of America
Bulletin 107(6), 725741.
Dorn, R.I., 1998. Rock coatings. Developments in Earth Surface Process. Elsevier,
Amsterdam, vol. 6, 429 pp.
Dorn, R.I., Phillips, F.M., 1991. Surface exposure dating: review and critical
evaluation. Physical Geography 12, 303333.
Drever, J.I. (Ed.), 1984. The Chemistry of Weathering. D. Reidel Publishing
Company, Dordrecht, The Netherlands, 324 pp.
Drever, J.I. (Ed.), 2005. Surface and Ground Water, Weathering, and Soils. Elsevier,
Amsterdam, 626 pp.
Ersoy, O., Sen, E., Erkan, A., Tatar, I., Celik, H.H., 2010. Surface area and volume
measurements of volcanic ash particles using micro-computed tomography
(micro-CT): a comparison with scanning electron microscope (SEM)
stereoscopic imaging and geometric considerations. Journal of Volcanology and
Fairclough, M.C., 2008. Kingscote Special Map Sheet. Geological Atlas 1:250,000
series, sheet SH 53-16. South Australia Geological Survey, Adelaide.
Friedenthal, R., 1963. Letters of the Great Artists: from Blake to Pollock (vol. 2).
Random House, New York.
Gerrard, A.J., 1992. Soil Geomorphology: An Integration of Pedology and
Geomorphology. Chapman and Hall, London, 269 pp.
Gerrard, A.J., 1988. Rocks and Landforms. Unwin Hyman, London, 319 pp.
Gilbert, G.K., 1877. Report on the Geology of the Henry Mountains (Utah). United
States Government Printing Office, Washington, DC, 160 pp.
Glinka, K.D., Marbut, C.F., 1927. The Great Soil Groups of the World and Their
Development. Edwards Brothers, Ann Arbor, Michigan, 235 pp.
Goudie, A., Viles, H.A., 1997. Salt Weathering Hazards. John Wiley, Chichester, UK,
241 pp.
Hole, F.D., 1953. Suggested terminology for describing soils as three-dimensional
bodies. Soil Science Society of America Journal 17, 131135.
Ip, K.H., Stuart, B., Ray, A., Thomas, P., 2010. ESEM-EDS investigation of
weathering of a heritage Sydney sandstone. Microscopy and Microanalysis
17(2), 292295.
Jenny, H., 1941. Factors of Soil Formation: A System of Quantitative Pedology.
McGraw-Hill, New York, 281 pp.
Jones, M., 1998. Regolith an ally to new mineral exploration in eastern
Queensland. In: Taylor, G., Pain, C. (Eds.), Regolith 98: New Approaches to an
Old Continent. Cooperative Research Centre for Landscape Environments and
Mineral Exploration, Canberra, pp. 311318.
10
Jungenius, P.D. (Ed.), 1985. Catena Supplement 6. Catena Verlag, Reiskirchen,

Germany, 174 pp.
Kay, G.F., 1931. Classification and duration of the pleistocene period. Bulletin of the
Geological Society of America 42, 425466.
Kinnahan, G.H., 1866. The effects of weathering on rocks. Geological Magazine 3,
8688.
Kittrick, J.A. (Ed.), 1986. Soil Mineral Weathering. Van Nostrand Reinhold, New
York, 271 pp.
Kneupfer, P.L.K., McFadden, L.D., 1990. Soils and Landscape Evolution:
Proceedings of the 21st Binghamton Symposium in Geomorphology, 67
October 1990. Elsevier, Amsterdam, (Reprinted from Geomorphology 3(3/4),
197575).
Krinsley, D.H., Pye, K., Boggs, Jr. S., Tovey, N.K., 1998. Backscattered Scanning
Electron Microscopy and Image Analysis of Sediments and Sedimentary Rocks.
Lee, M.R., 2010. Transmission electron microscopy (TEM) of Earth and planetary
materials: a review. Mineralogical Magazine 74(1), 127.
Lerman, A., Meybeck, M. (Eds.), 1988. Physical and Chemical Weathering in
Geochemical Cycles. Kluwer Academic Publishers, Dordrecht, The Netherlands,
375 pp.
Lindsay, J.F., 1992. Extraterrestrial soils the Lunar experience. In: Martini, I.P.,
Chesworth, W. (Eds.), Weathering, Soils, and Paleosols. Elsevier, Amsterdam,
pp. 4170.
Loughnan, F.C., 1969. Chemical Weathering of the Silicate Minerals. American
Elsevier Publishing Company, New York, 154 pp.
Mahaney, W.C. (Ed.), 1984. Quaternary Dating Methods. Elsevier, Amsterdam,
431 pp.
Marshall, C.E., 1964. The Physical Chemistry and Mineralogy of Soils. Volume 1:
Soil Materials. Wiley, New York, 388 pp.
Marshall, C.E., 1977. The Physical Chemistry and Mineralogy of Soils. Volume 2:
Soils in Place. Wiley, New York, 313 pp.
Martini, I.P., Chesworth, W. (Eds.), 1992. Weathering, Soils and Paleosols. Elsevier,
Amsterdam, 618 pp.
Merrill, G.P., 1897. A Treatise on Rocks, Rock-Weathering, and Soils. The
Macmillan Company, New York, 411 pp.
Migon, P., 2006. Granite Landscapes of the World. Oxford University Press, Oxford,
393 pp.
Nahon, D.B., 1991. Introduction to the Petrology of Soils and Chemical Weathering.
John Wiley and Sons, New York, 313 pp.
Nash, D.J., McLaren, S.J. (Eds.), 2007. Geochemical Sediments and Landscapes.
Wiley-Blackwell, Malden, MA, 488 pp.
Ollier, C.D., 1969. Weathering, First ed. American Elsevier, New York, 304 pp.
Ollier, C.D., 1984. Weathering, Second ed. Longman, London, 270 pp.
Ollier, C.D., Pain, C., 1996. Regolith, Soils, and Landform. John Wiley, Chichester,
UK, 316 pp.
Phillips, J.D., 2005. Weathering instability and landscape evolution. Geomorphology
67(12), 255272 (Special issues, 35th Annual Binghamton Geomorphology
Symposium on Weathering and Landscape Evolution).
Phillips, J.D., Marion, D.A., Luckow, K., Adams, K.R., 2005. Nonequilibrium
regolith thickness in the Ouachita Mountains. Journal of Geology 113,
325340.
Pillans, B., 2008. Regolith through time. In: Scott, K.M., Pain, C.F. (Eds.), Regolith
Science. Springer Science and CSIRO, Dordrecht, The Netherlands, and
Collingwood, Australia, pp. 730.
Playfair, J., 1802. Illustrations of the Huttonian Theory of Earth. Cadell and Davies,
London (reprinted, 1964, by Dover Publications, New York, 528 pp).
Pope, G.A., 1995. Internal weathering in quartz grains. Physical Geography 16(4),
315338.
Reiche, P., 1950. A survey of weathering processes and products. University of New
Mexico Publications in Geology, No. 3. University of New Mexico Press,
Albuquerque, 95 pp.
Retallack, G.J., 2008. Soils of the Past: An Introduction to Paleopedology. Blackwell
Science, Oxford, 416 pp.
Rhodes, B.L., Thorn, C.E., 1996. Toward a philosophy of geomorphology. In:
Rhodes, B.L.,Thorn, C.E. (Eds.), The Scientific Nature of Geomorphology:
Proceedings of the 27th Binghamton Symposium in Geomorphology, 2729
September, pp. 115143.
Robinson, D.A., Williams, R.B.G. (Eds.), 1994. Rock Weathering and Landform
Evolution. British Geomorphological Research Group Symposia Series, No. 10.
Wiley, Chichester, UK, 519 pp.
Ruhe, R.V., 1956. Geomorphic surfaces and the natures of soils. Soil Science 82,
441455.
Said, R., 1950. Geology in tenth century Arabic literature. American Journal of
Science 248(1), 6366.
Schaetzl, R.J., Anderson, S., 2005. Soils: Genesis and Geomorphology. Cambridge
Scott, K.M., Pain, C.F. (Eds.), 2008. Regolith Science. Springer Science and CSIRO,
Dordrecht, The Netherlands, and Collingwood, Australia.
Small, R.J., 1982. Slopes and Weathering. Cambridge University Press, Cambridge,
112 pp.
Smith, B., Turkington, A. (Eds.), 2004. Stone Decay: Its Causes and Controls:
Proceedings of Weathering 2000, An International Symposium held in Belfast
2630 June 2000. Donhead, Shaftesbury, UK, 306 pp.
Taylor, G., Eggleton, R.A., 2001. Regolith Geology and Geomorphology. John Wiley
and Sons, Chichester, UK, 375 pp.
Thomas, M.F., 1994. Geomorphology in the Tropics. John Wiley and Sons,
Chichester, UK, 460 pp.
Turkington, A., Phillips, J., Campbell, S., 2005. Weathering and Landscape
Evolution: Proceedings of the 35th Binghamton Symposium in Geomorphology,
13 October 2004, Lexington, Kentucky. Elsevier, Oxford.
Twidale, C.R., 2010. Uluru (Ayers Rock) and Kata Tjuta (The Olgas): Inselbergs of
Central Australia. In: Migon, P. (Ed.), Geomorphological Landscapes of the
World. Springer, Dordrecht, pp. 321333.
Van den Broek, E., 1881. Memoire sur les Phenome`nes dAlteration des Depots
Superficiels par lInfiltration de Eaux Meteoriques. F. Hayez, Imprimeur de
lAcademie royale de Belgique, Brussels, 180 pp.
Vanimen, D.T., Chipera, S.J., 2006. Transformations of Mg- and Ca-sulfate hydrates
in Mars regolith. American Mineralogist 91(10), 16281642.
Von Engeln, O.D., 1948. Geomorphology: Systematic and Regional. The Macmillan
Company, New York, 655 pp.
Warke, P.A. (Ed.), 2001. Special issue: weathering 2000. Earth Surface Processes
and Landforms 26(8), 809910.
Whalley, W.B., Turkington, A.V., 2001. Weathering and geomorphology: special
issue. Geomorphology 41(1), 172.
White, I.D., Mottershead, D.N., Harrison, S.J., 1992. Environmental Systems:
An Introductory Text, Second ed. Chapman and Hall, London, 616 pp.
Wright, V.P., 1986. Paleosols: Their Recognition and Interpretation. Princeton
University Press, Princeton, 315 pp.
Yates, J., 1830. Remarks on the formation of alluvial deposits. Proceedings of the
Geological Society of London 1, 237239.
Yatsu, E., 1988. The Nature of Weathering. Sozosha, Tokyo, 624 pp.
Yatsu, E., 1966. Rock Control in Geomorphology. Sozosha, Tokyo, 135 pp.
11
Biographical Sketch
Gregory Pope is an associate professor in Earth and Environmental Studies at Montclair State University in New
Jersey. He mentors and teaches undergraduate through doctoral students, undergraduate research and advising
foremost in this role. His research interests span soils geomorphology, Quaternary environmental change, and
geoarchaeology, working in China, Latin America, Western Europe, and the Western and Northeastern United
States. He is active in both the Geological Society of America and Association of American Geographers, and
served as chair of the Geomorphology Specialty Group and a Regional Councilor for the AAG.
4.2 Synergistic Weathering Processes

HA Viles, University of Oxford, Oxford, UK
4.2.1
4.2.1.1
4.2.2
4.2.3
4.2.3.1
4.2.3.2
4.2.4
4.2.4.1
4.2.4.2
4.2.4.3
4.2.5
4.2.5.1
4.2.5.2
4.2.5.3
4.2.5.4
4.2.5.5
4.2.5.6
References
Introduction
Definitions of Weathering and the Synergy Issues They Raise
Getting to the Heart of Weathering and Its Synergies
Scale Issues and Understanding Weathering Synergies
Timescales and Weathering
Spatial Scales and Weathering
Concepts to Help Understand Weathering Synergies across Scales
Nonlinear Weathering Systems
Weathering- and Transport-Limited Systems
The Critical Zone: An Aid to Understanding Weathering Synergies?
Weathering Process Synergies
Microscale Synergies between Weathering Mechanisms
Synergies between Weathering Agents and Processes
Synergies between Weathering Processes and Weathering Landform Evolution
Synergies Linking Weathering Processes Across Temporal and Spatial Scales
Synergies between Weathering and Erosion
Synergistic Weathering Processes: Toward the Future
Glossary
Biofilm A mixed microorganic community forming a thin
film or layer on a surface, such as on a rock or soil.
Duricrust A hard crust or layer formed near or at the
Earths surface through the action of weathering processes.
12
12
14
15
15
16
18
18
19
21
21
21
22
22
23
24
24
25
Endolithic A habitat within the outer layers of a rock.

Usually, organisms actively bore into the rock to inhabit
this niche.
Epilithic A habitat growing on the surface of a rock.
Tafoni Large cavernous weathering features.
Abstract
Weathering processes rarely act alone, and weathering is highly intertwined with other geomorphological processes. This
chapter reviews the nature of synergies between weathering agents and processes, between weathering and the factors which
control it, and between weathering and erosion. A wide range of interactions occur across many different scales, which can
best be viewed within a nonlinear dynamical systems framework.
The system of weathering is open and y non-equilibriumy The

nature of weathering is utterly intricatey all of the physical,
chemical and biotic reactions on the Earth are proceeding concurrently (Yatsu, 1988: p. 505).
4.2.1
Introduction
The interconnectedness of weathering processes, both with

each other and with other parts of the geomorphic system, has
been recognized for many decades, but it is only in recent
years that the combination of technological advances and
development in ideas have permitted a closer examination of
these interactions. The purpose of this chapter is to examine
Viles, H.A., 2013. Synergistic weathering processes. In: Shroder, J. (Editor in
Chief), Pope, G.A. (Ed.), Treatise on Geomorphology. Academic Press, San
Diego, CA, vol. 4, Weathering and Soils Geomorphology, pp. 1226.
12
the current state of the art of knowledge on the nature and

importance of the main interactions involving weathering
processes and geomorphology. Although the chapter title Synergistic weathering processes implies a mutually reinforcing
relationship between two or more processes, other relationships
between processes and other parts of the weathering system
may occur, such as where one process dampens or mutes the
effects of another. Because of this, the term synergy in the
following sections should be interpreted broadly to refer to all
types of weathering feedbacks.
4.2.1.1
Definitions of Weathering and the Synergy Issues

They Raise
Some initial aspects of the interconnectedness of weathering

emerge from a consideration of the definition of weathering.
Although weathering has been defined in many ways over the
http://dx.doi.org/10.1016/B978-0-12-374739-6.00057-9
Synergistic Weathering Processes
past 150 years, Yatsu (1988) put it in simple but effective

terms as:
Weathering is the alteration of rock or minerals in situ, at or near
the surface of the earth and under the conditions which prevail
there (Yatsu, 1988: p. 2).
Dixon (2004) provided a fuller explanation as follows:

Weathering refers to a group of processes collectively responsible
for the breakdown of materials at or near the Earths surface.
Weathering occurs because the environmental conditions under
which most rock materials formed differ substantially from those
which prevail near the Earths surface. Consequently, they undergo
a variety of modifications which result in more stable products
under the newly imposed conditions of temperature, pressure,
moisture and gaseous environment (Dixon, 2004: p. 1108).
Three main issues relevant to weathering synergies arise

from a consideration of these definitions. First, in looking at
the phrases in situ and breakdown of materials can we make a
clear separation between weathering and erosion? Second,
looking at the phrase at or near the surface of the Earth can we
also make a clear distinction between weathering and diagenesis? Finally, can we make a clear separation between
weathering and its controlling factors, or, to put it another
way, how does weathering interact with the controlling environmental conditions? These three issues are illustrated
diagrammatically in Figure 1.
Looking at the first issue, there has been long-term confusion over the relationship between weathering and erosion,
and whether a sharp line can be drawn between them (as also
reviewed recently by Dixon and Thorn, 2005). Gilbert (1877),
in an oft-quoted passage, made a clear and simple conceptualization of weathering as the first of three divisions of
erosion (with the others being transportation and corrasion).
Weathering, in Gilberts view, involved the disintegration and
loosening of rock and was clearly a component part of erosion. In more recent writings, denudation is commonly used
as a broader umbrella term to apply to all processes that wear
down the surface of the Earth including weathering and a
more narrowly defined erosion (see, e.g., Kennedy (2000) and
Gregory (2010: p. 7677)). There is clearly, in reality, some
overlap between weathering and erosion, however it is defined. Many processes can cause rock breakdown, which are
not conventionally thought of as being part of weathering,
such as aeolian and fluvial abrasion (or corrasion, as G.K.
Geology
Climate
Diagenesis
Weathering
Erosion
Key:
Controls
Feedbacks/synergies
Continuum
Figure 1 Conceptual diagram showing the controls, feedbacks, and

continua within the climategeologyweathering system.
13
Gilbert called it). In turn, many processes that are thought of

as classic examples of weathering (e.g., the dissolution of
limestone) are also often thought of as part of the chemical
erosion of the landscape. Sometimes a process will be identified as erosional in one environment, and categorized as
weathering in another (e.g., boring by photosynthetic microorganisms is generally called bioerosion on coastal rocky
shores and bioweathering in terrestrial environments). In essence, there is in reality a blurring of boundaries between
processes whose names imply that they are distinctively different. Any consideration of weathering process synergies
needs to acknowledge such definitional uncertainties and the
fact that a continuum of types of processes exists between
weathering and erosion (Figure 1).
Turning to the second issue, questions also arise over the
distinction between weathering and diagenesis. Weathering
has been defined as alterations at or near the Earths surface,
whereas diagenesis may be defined as all the physical, chemical, and biological changes experienced by a sediment or rock
after its initial deposition, but while buried. Several of the
processes of diagenesis, such as cementation, leaching, and
hydration, are also weathering processes, and so the only clear
dividing line between the two sets of processes should be one
of depth of burial. In reality, there is no clear divide between
diagenesis and weathering as it is difficult to decide what is at
or near the Earths surface in many situations. Although
weathering is usually seen to operate on rock surfaces, in soil,
within saprolite and at the saprolite-rock boundary, in karst
and other environments weathering may occur much lower.
Bourdon et al. (2009), for example, used novel U-series
methods to investigate weathering in a chalk aquifer in eastern
France and find that some 3050% of dissolved load in surface waters comes from weathering within lower parts of the
aquifer (down to 45 m below the surface). Conversely, diagenesis can occur within the soil profile as calcretes and other
duricrusts develop. More realistically we might conceptualize a
gradient between weathering and diagenetic processes, and
synergies between them (as diagenetic alterations may then
influence subsequent weathering, whereas weathering produces reactants subsequently involved in diagenesis of chemical sediments) as portrayed in Figure 1.
Finally, it is important to consider the relationship between
weathering and the environmental factors controlling it. As
Yatsu suggested, weathering is in reality a nonequilibrium
system because the environment that controls weathering
processes is constantly changing over a wide range of timescales. Indeed, many workers now conceptualize the weathering system in terms of being a nonlinear dynamical system
with complex interactions between system components (e.g.,
Phillips, 2005). The major environmental factors that control
weathering are climate, geology, and biota, and, as Pope et al.
(1995) pointed out, these operate in different ways over a
wide range of scales. They control not only the overall rate of
weathering but also the balance of weathering processes operating. However, at all scales it is clear that weathering and its
controlling factors are interconnected in complex ways such
that change in one can bring about change in the other and so
on as depicted in Figure 1. At the small scale, for example, the
weathering impact of biofilms on a rock surface can, in turn,
lead to ecological succession with a new array of species
14
colonizing, with further impacts on weathering. At the medium scale, many positive feedbacks have been identified between weathering and changing surface rock properties and
microclimates as weathering-controlled landforms such as
tafoni develop. At the very large scale, on the other hand, the
interrelationships between silicate weathering, uplift, erosion,
carbon cycling, and climate change have been well documented but remain under much debate. Thus, we can identify
important synergies between weathering and the factors that
control it across a wide range of scales, which are consequences of the nonlinear dynamical nature of the weathering system.
4.2.2
Getting to the Heart of Weathering and Its

Synergies
In addition to the external synergies identified in the previous

section, there are, of course, synergies operating between
weathering processes. Before we can start to review these in
detail, it is necessary to sharpen up our conceptualization of
what weathering really involves. Figure 2 depicts one attempt
to conceptualize the major components of weathering. We
tend to think of weathering as being an entity (manifested as
an overall weathering rate or intensity/degree of weathering).
Weathering as an entity is seen as being highly important
geomorphologically, ecologically, and to the Earth system in
general because of its capacity to produce sediment, contribute
to soil development, release elements for biogeochemical
cycling, and initiate and contribute to relief development
across a wide range of scales. The entity of weathering can be
broken down into a range of processes (typically bracketed
Linkages
Effects
Relief generation
under headings of physical or mechanical, chemical, and

biological). For example, we identify salt weathering as one
key type of physical weathering process, and dissolution as
one type of chemical weathering process. These processes, in
turn, are seen as being caused by a series of agents of weathering (such as water, salts, temperature cycles, and organisms). Furthermore, these processes produce observable effects
such as flaking, exfoliation, and pitting. Although all these
sound straightforward in theory, in practice there is much
confusion and lack of clarity caused essentially by the difficulties faced in systematizing a very complex reality that can
only be observed partially. Understanding these complexities
is vital in attempts to understand weathering synergies.
Looking first at weathering processes as shown in Figure 2,
it is clear that not all processes are straightforward and that in
many cases the naming and identification of an individual
process came before a clear understanding of what was really
involved. Detailed investigation of some weathering processes
has revealed that there are in fact many different mechanisms
involved. For example, salt weathering operates in at least
three different ways, depending on the circumstances (pressures exerted by crystallization, hydration, and temperature
changes of different salts as discussed more fully in Goudie
and Viles, 1997). In some environments, more than one of
these mechanisms of salt weathering may operate, perhaps
acting synergistically. Some types of weathering process are
very difficult to identify and categorize including many
biological processes that have only relatively recently been
studied in any detail. Is it more appropriate to name biological
weathering processes according to the organism involved (e.g.,
lichen weathering) or in terms of the more detailed mechanism operating (e.g., chelation) or adopt some other scheme?
Biogeochemical cycling
Flaking
Weathering rinds
Pitting
Regolith and soil production
Exfoliation
Granular disintegration
Weathering
Categories
Agents
Physical
Temperature cycles
Biological
Salts
Organisms
Biochemical
Processes
Freezethaw
Chemical
Exudates
Water
Dissolved ions
Biophysical
Hydrolysis
Salt weathering
Chelation
Hydration
Dissolution
Mechanisms
Crystallization
Hydration
Thermal expansion
Figure 2 Conceptual diagram showing the linkages, effects, categories, agents, processes and mechanisms involved in weathering. See text for
explanation of terms.
There is as yet no key agreement and thus a wide range of very

disparate biological weathering process names appear in the
literature. Overall, we can conclude that weathering processes
are generally vaguely defined and commonly can be broken
down into a number of different mechanisms that may have
synergistic relations with other mechanisms and processes.
Turning to issues surrounding the bracketing of weathering
processes into different categories, it becomes clear that here,
too, are complexities and that the tripartite division into
physical, chemical, and biological is inadequate and many
cross-category interactions may occur. Although salt weathering is classified as a physical weathering process, mechanisms of crystallization and hydration involve chemical
transformations and so in reality represent physico-chemical
weathering. Furthermore, salts in solution can also have a
purely chemical influence on some minerals surely a form of
salt weathering, but clearly also something that should be
defined as a chemical weathering mechanism. Looking at
another example, most chemical weathering processes involve
a biological dimension either directly or indirectly, as, for
example, biota play key roles in generating soil CO2 and organic acids, which then cause much chemical weathering at
the soil:rock interface. So, there are many synergistic relationships and linkages between the three major categories of
weathering processes.
Our understanding of the agents and effects of weathering
(both portrayed in Figure 2) also illustrates many overlaps,
complexities, and interactions. For example, lichens are agents
of weathering that can cause both biochemical and biophysical weathering processes, commonly within the same
organism. Although we may refer to these together as lichen
weathering, it is clear that such weathering involves a range of
mechanisms and probably several processes. In terms of the
effects of weathering, authors generally refer to exfoliation
weathering as a process, whereas, more precisely, it is the result
a number of different weathering processes acting alone or
together. Exfoliation is manifested by flaking off of surface
layers of rock, but this effect can be caused by any number of
weathering processes that set up stresses within the nearsurface layers of a rock and thus initiate flaking. Because of the
varying interplay between geological and climatic conditions
at different locations, it is likely that the weathering processes
involved in exfoliation will vary hugely from place to place. In
cold, dry environments such as Antarctica, freezethaw may be
the dominant cause of flaking in combination with the growth
of endolithic biofilms and the presence of salts, whereas,
in hotter climates, insolation weathering in the presence of
moisture and salts may be critical.
Although the above-mentioned points might seem slightly
arcane and nit-picking, they are important as they reveal
generalized lack of clarity over what we mean by different
aspects of the weathering system as depicted in Figure 2. Much
of this confusion has arisen because of the increasingly multiscalar nature of investigations into weathering. In the early
days, when it was difficult to study weathering in any detail,
weathering processes were recognized through their visible
and surficial effects and also by the obvious agents which
seemed to be active. Thus, it was sensible to name and categorize weathering in these terms. New methods have allowed
us to identify hitherto unidentified processes, including a vast
15
array of biological weathering processes. Alongside this, as

techniques become available to study weathering at ever finer
scales, we have also started to classify weathering processes at
similarly finer scales, with some very small-scale mechanisms
now seen to be of fundamental importance. On the other
hand, some very large questions in both geomorphology and
Earth system science now require much larger-scale information on, and understanding of, weathering.
The material covered in this introductory section illustrates
that there are many complex aspects to synergies within
weathering and between weathering and other aspects of
geomorphology. At the heart of many of these issues is what
we might call terminological confusion especially in terms
of what we actually mean by weathering processes. In essence,
what we mean by a weathering process is itself scale
dependent at the microscopic level, hydration and crystallization of a salt can be identified as individual processes,
whereas, at larger scales, salt weathering can be identified as an
individual process. At larger scales again, granular disintegration (or the visible manifestation of the operation of a
range of processes) might be reasonably identified as a process. At the very largest scale, the net effect of a whole range of
chemical and biochemical processes can be lumped together
to define an overall chemical weathering process. It is thus
clear that scale issues underpin all our discussions and
understanding of weathering synergies, as elucidated more
fully in the next section.
4.2.3
Scale Issues and Understanding Weathering

Synergies
As discussed previously by Pope et al. (1995) and Viles (2001),

there are many scale issues that face studies of weathering and
attempts to link weathering processes to geomorphology and
Earth system science. Several of these issues are particularly
important to be borne in mind when trying to understand
weathering synergies as they may help to reduce the level of
complexity portrayed in Figure 2 by indicating the key synergies relevant to any particular scale of enquiry. It is helpful to
simplify the discussion by separating out scale issues related to
time from those related to space, although in reality, as we will
see later, the two are intertwined.
4.2.3.1
Timescales and Weathering
Two major temporal-scale issues affect our understanding of

weathering synergies, that is, weathering timescales and the
variability of weathering over time. Weathering occurs over a
very wide range of timescales, from the instantaneous failure
of an intergranular cement causing the loss of grains from a
rock surface, to the tens of millions of years of dissolution it
takes to produce tropical tower karst landscapes. In turn, different methods to study weathering are able to pick up change
over very different time spans and there may be a mismatch
between the method used and the process of interest. For example, laboratory experiments on chemical reaction rates tell
us something about the short term tempo of weathering, but
extrapolating this to tens of millions of years in order to
16
calibrate models of long-term landform evolution may be

extremely misleading. For such long timescales, indirect assessment of weathering rates from information on denudation
and sedimentation gained from cosmogenic dating studies
may be more relevant. This means that one key need in investigating weathering synergies is for the scale at which data
can be collected to match as near as possible the scale of the
question being asked.
Another aspect of the timescale issue in relation to weathering synergies concerns the depth of knowledge required
about weathering in relation to the question being asked. If we
wish to understand what causes weathering at the annual scale
(perhaps in relation to understanding how millimeter- to
centimeter-scale flaking of rock surfaces occurs), then synergies between very detailed weathering mechanisms and
processes may be important. If, however, we are looking to
explain the development of hillslopes, then synergies between
weathering and removal of material through soil creep, landsliding, etc. are much more important. At even longer timescales, attempts to link chemical weathering of silicate
minerals to the long-term carbon cycle require us to think
carefully about synergies and interactions between weathering,
erosion, and lithification at the global scale. What we conceive
of, and what we can measure, as a weathering synergy thus
depends upon the timescales of interest.
The variability of rates and types of weathering over time is
also an important issue relating to timescales and our understanding of weathering synergies, and helps to explain the
problems of extrapolating short-term weathering rate measurements to the longer term. Because, for most situations, the
weathering system is not in equilibrium, rates of change vary
over all timescales. One cause of such variations in rate is
climate variability and change. For example, short-term data
collected in the dry season may not reflect the nature and rate
of weathering over the annual timescale in highly seasonal
environments. Similarly, measurements of weathering rates
during El Nino years may vary hugely from those collected in
normal or La Nina conditions in places where the El Nino
Southern Oscillation (ENSO) exerts a strong control over key
climatic variables such as rainfall. Measurements of weathering in some locations taken within recent centuries (under
conditions found in the Anthropocene) may be very different
to those occurring under entirely natural conditions earlier in
the Holocene. Measurements taken within paraglacial phases
may not be reflective of longer-term weathering rates and
processes in many environments close to retreating glaciers.
Climate is not the only important boundary condition to
control of weathering that varies over time, and furthermore
weathering rates (and processes) may also vary over time as a
result of the nonlinear nature of most weathering systems.
Changes in tectonic processes may influence weathering rates
and processes, as in areas prone to the development of fault
scarps in which zones of fresh rock surfaces are exposed to
weathering within complex hillslope profiles. Inherent nonlinearities within weathering systems can also be important,
and, in many situations, weathering rates have been shown to
decay exponentially over time even given constant boundary
conditions because initially fast rates decline as weathering
products build up, shielding the surface from subsequent
disintegration. Such situations have been recorded for the
buildup of weathering rinds on boulder surfaces on moraines

(Colman, 1981) and the thickening of regolith on soil-covered
slopes (Dietrich et al., 1995), among other circumstances. The
nature and importance of weathering synergies will thus also
vary over time as a result of (and often causing) such variations in the weathering system over different timescales.
4.2.3.2
Spatial Scales and Weathering
Linked to the above temporal-scale concerns, there are two key

issues surrounding spatial scales and weathering that influence
our understanding of weathering synergies: first, the variation
in weathering across spatial scales, and, second, the patchiness
of weathering across space. As mentioned in the introduction,
weathering occurs at all scales, but its nature and geomorphological role vary hugely with spatial scale. Similarly,
the nature of synergies among weathering processes and between them and other Earth surface processes varies hugely at
different spatial scales. What happens in terms of synergies in
a small-scale weathering pit (millimeters to centimeters in
dimensions) will be completely different to what happens
across a whole soil-covered landscape. As with our discussion
of timescales, it is also important to acknowledge that our
measurement abilities also vary across these spatial scales.
Upscaling from a small micro-erosion meter plot (which
measures weathering rate by assessing surface lowering at a
number of points within a triangular plot tens of centimeters
in size) to the rate and pattern of weathering across a whole
coastal platform, for example, is fraught with problems.
Measuring weathering rates and patterns across the whole
platform, until the advent of ground-based light detection and
ranging (LiDAR), would have been impossible however, and
techniques such as LiDAR produce extremely large data sets
that can be difficult to summarize and simplify. In a similar
vein, Matsuoka (2001) pointed out that most laboratory
freezethaw weathering experiments study microgelivation on
porous rocks, whereas in the field what is most important in
many environments is macrogelivation within hard, jointed
rocks. Extrapolation from the laboratory to the field cannot be
done easily in this case. However, recent research on basalt
weathering has shown that using a roughness fractal dimension to account for active surface area it is possible to develop
a single equation to predict weathering rates across a wide
range of scales (Navarre-Sitchler and Brantley, 2007). Therefore, upscaling and downscaling may be possible for a range of
lithologies. However, at each scale, the questions we might
want to ask about a landscape and the contribution of weathering may also vary hugely and the challenge is to match
data to the relevant process synergies and to the correct scale
of question. What we conceive of, what is geomorphologically
important, and what we can measure, as a weathering synergy,
will thus vary hugely over different spatial scales.
There may also be important links between weathering
synergies stretching across different spatial scales. Associations
of small-scale weathering processes on a shore platform, for
example, may weaken blocks and predispose them to erosion
by wave action. Indeed, cracks are weathering features which
show a huge range in characteristic size distributions (as
illustrated in Viles, 2001, Table 1) and may act as effective
conduits to link weathering processes and effects at different

scales as microcracks become enlarged to form visible cracks,
which are then opened to form joints. At a larger scale, Goudie
and Migon (1997) suggested that the vertical and lateral enlargement of weathering pits on granite surfaces on inselbergs
in central Namibia may play an important role in surface
lowering.
Weathering rates and process synergies will also vary
hugely as a result of the variability of conditions over space as
environments are not isotropic in any way. Thus, for example,
across the central Namib Desert plain surface, which stretches
nearly 100 km from the Atlantic Coast to the base of the Great
Escarpment, weatheringprocess associations vary hugely
from salt-weathering-dominated coastal salt pans, through
areas in the fog belt affected by lichen weathering and to inland areas where temperature cycling is most effective (see
Figure 3, and Viles and Goudie, 2007; Viles, 2005a). Such
patchiness also works on smaller spatial scales, as within the
fog belt very wet areas are likely to be dominated by salt
weathering, creating highly unstable and rapidly weathering
surfaces; west-facing boulders above the salty ground surface
are dominated by lichen weathering, whereas east-facing
boulder surfaces are dominated by aeolian abrasion (Viles and
Goudie, 2000). As Goudie (1996) proposed more generally,
the variation in weathering process synergies across space
leads to the idea of weathering hot spots which, if long term
17
in nature, may be of geomorphological importance. Hartmann et al. (2009) found such chemical weathering hot spots
in their global analysis of CO2 consumption by weathering,
and their model suggested that just over 28% of CO2 consumption by weathering today occurs in hot spots. Such
weathering hot spots can be seen as manifestations of nonlinearity and instability within weathering systems.
Figure 4 draws together some of the above-mentioned
ideas on temporal and spatial scales and their impact on
weathering synergies in order to provide a simple threefold
representation of weathering in the context of geomorphological systems. This approach builds on earlier work by Viles
(2001) and Phillips (2005). One of the most difficult aspects
of producing such a diagram is the huge range of temporal
and spatial scales covered by geomorphology and the impossibility of compressing them into a manageable number of
categories. We all have different perspectives over where to
place the boundaries between categories, based on our own
research experiences. The author has taken the approach in
Figure 4 of simplifying things down to the smallest number of
categories to illustrate the different types of synergies, and
problems with studying synergies, without too much repetition or overlap. One interesting theme to emerge is that the
interconnectedness of biotic and abiotic processes crosscuts all
scale categories and is clearly a fundamental aspect of weathering synergies at all scales.
Figure 3 Spatial variation in weathering process associations in the northern Namib Desert, Namibia. (a) Lichens on west-facing side of basalt
clast, wind abrasion on east-facing side in a near-coastal environment; (b) salt weathering of boulder in coastal salt pan; (c) wind abrasion flutes
on basalt clast; (d) thermal cracking on basalt cobble further inland.
18

Key focus
Problems
Key questions
Features on rock
faces/in soil
profiles
Lack of
detailed
observations
e.g., pits, rinds,

karren, tafoni,
Complexity
of processes
Which coassociations of
processes
produce these
features?
Hillslope profiles
and processes
Spatial scale =
ms to kms
Timescale =
103105 years
Lack of long
term
monitoring
e.g., Inselbergs
and pediments
Complexity
of linkages
Mega
Landscapes
Spatial scale =
10s to 100s km
Timescale =
> 106 years
e.g., passive
margin
escarpment and
plain
Difficulty of
constructing
and
interpreting
weathering
histories
Scale
Micro
Spatial scale =
<mm to m
Timescale =
100 102 years
Macro
Example
How are
weatheing and
erosion linked
across
hillslopes?
How does
weathering
contribute to
landscape
evolution?
4.2.4
Concepts to Help Understand Weathering

Synergies across Scales
High
Figure 4 Conceptual diagram of key questions and problems in studying weathering synergies at different spatial and temporal scales.
Freezethaw
weathering
rate
Biological
weathering
rate
Low
Although the preceding section has illustrated some scale

differences in weathering synergies, and hinted that some
process associations might link across scales, concepts of
nonlinear weathering systems, weathering vs. transportlimited environmental settings, and the critical zone (the
weathered profile from rock to soil) provide some frameworks
for bringing things together.
Overall
weathering rate
Ice cover
4.2.4.1
Nonlinear Weathering Systems
Several authors recently have discussed the application of

nonlinear dynamical systems ideas to weathering systems
(Turkington and Phillips, 2004; Viles, 2001; Viles, 2005b;
Phillips, 2005) building on ideas of the importance of
thresholds to geomorphology. Nonlinear dynamical systems,
by their very nature, experience thresholds or bifurcations in
their behavior and can oscillate between two or more
stable states. Many geomorphic systems have been shown to
exhibit nonlinear dynamical system behavior. It can be helpful
to view weathering process synergies in the light of such
nonlinear behavior. For example, two weathering processes (or
weathering and another geomorphic process) may act together
in one way before a threshold is reached, but after that
threshold the relationship may be quite different. In a weathering system prone to episodic flaking, for example, freezethaw and fungal biochemical weathering processes may
operate synergistically to produce the detachment of a flake.
Subsequently other processes then may dominate until fungi
recolonize as a weathering rind develops and microenvironmental conditions alter so as to facilitate further flaking
Time/warming
Figure 5 Conceptual diagram of the changing interactions between
freezethaw and biological weathering during paraglacial conditions.
(Etienne, 2002). In this case, detachment of the flake represents breaching a threshold.
Similar patterns of process associations may also be seen
over larger temporal and spatial scales as, for example, proposed by Viles et al. (2008) in weathering systems within
paraglacial environments. Figure 5 illustrates some possible
alterations in weathering process synergies over time during a
paraglacial period as colonization of rock-surface-dwelling
microorganisms in the form of biofilms follows and enhances
weathering by freezethaw. More generally, inherited weaknesses from long periods of past weathering may predispose
weathering systems to accelerated weathering under current
environmental conditions (Warke, 2007). Inheritance effects
like these are regarded as classic manifestations of dynamical
instability and self-organizing behavior in geomorphic systems by Phillips (1999). Nonlinear approaches can also
be used to investigate scale linkages as demonstrated by
19
Surface wetting period
Solutional
weathering
Microorganic
growths
Biological weathering
Hollow initiation and

deepening
Mineral debris
Organic debris
(a)
Rate of production of
weathered debris
Regolith thickness
Solid removal (erosion,

mass wasting)
Solutional removal
(b)
Figure 6 (a) A weathering system producing weathering pits on an impure limestone. Reproduced from Viles, H.A., 2001. Scale issues in
weathering studies. Geomorphology 41, 6372. (b) Weathering system interactions at the hillslope scale. In all system diagrams, thick,
solid arrows change in one component causes a change in the same direction in the other component ( ve links), whereas thin, dashed
arrows change in one component causes a change in the opposite direction in the other component (ve links). Reproduced from Phillips, J.D.,
2005. Weathering instability and landscape evolution. Geomorphology 67, 255272.
Navarre-Sitchler and Brantley (2007) who showed that, from

mineral to drainage basin scale, the weathering advance rate of
basalt can be viewed as a fractal property. As Phillips (2005)
articulated in detail, nonlinear weathering systems may exhibit
stable or unstable behavior depending on the scale of analysis.
Many different types of process associations and scale linkages
can fruitfully be understood within a nonlinear system
framework, as shown in the examples given in Figure 6.
4.2.4.2
Weathering- and Transport-Limited Systems
The concept of weathering- and transport-limited systems was

introduced into the geomorphological literature by Carson
and Kirkby (1972) in relation to hillslope development and
building on the pioneering ideas of Gilbert (1877). Whereas
most geomorphologists apply these concepts at the hillslope
scale, the basic ideas are applicable at many scales within
geomorphology. The same sort of concept (i.e., reaction-rate
vs. transport-rate control) is also used in geochemical literature as a way of understanding the rates of chemical
weathering in different situations. Looking at weathering systems in terms of whether they are transport- or weatheringlimited can be fruitful and helps to illustrate how different
process interactions can occur in these contrasting situations.
Figure 7 provides a simple view of the two situations. In
weathering-limited circumstances, there is highly efficient
transport of weathered products (sediments and dissolved
material) away from the surface, thus precluding the development of any more than thin and patchy soils. Bare rock
surfaces may be covered with biofilms, and as well as the
removal of material weathering may produce weathering rinds
or case-hardened zones. Intermittent rock falls may occur, in
some cases switching all or part of the slope from weatheringlimited to transport-limited conditions for some time afterward until the debris is removed. The dissolved products of
weathering as transported elsewhere may be reprecipitated in
the form of a range of geochemical sediments (such as speleothem and tufas). Characteristically, such weatheringlimited situations are dominated by physical (mediated by
biological) weathering processes (except on particularly
20
Weathering on largely bare

rock surface
Weathering in soil and regolith

and at rock: regolith interface
Rapid removal of
weathering products
Transport >>
weathering
Weathering limited
Weathering >>
transport
Slow/episodic removal
of weathering products
Transport limited
Bare rock surfaces often covered

with biofilm with direct influence on
weathering
Soil and regolith cover most rock

surfaces
Weathering rinds and case

hardening can occur
Widespread indirect biological

influences on weathering
Differential resistance to weathering

a major control of slope profile
Aeolian and other processes can also

add to sediment cover
Positive feedbacks also create

complex slope profiles
Duricrusts may form in soils from

weathering products
Figure 7 Weathering-limited and transport-limited situations and their associated weathering system characteristics.
soluble rocks such as limestone). By contrast, transportlimited situations are dominated by the buildup of debris and
soil cover over the bare rock surface, as there is only slow or
intermittent removal of the products of weathering. Another
way of producing transport-limited conditions is by deposition on the surface of allochthonous material from material transported in from elsewhere (e.g., as a result of
colluvial, alluvial, glacial, or aeolian sedimentation). The
products of weathering can also be mobilized within the
overlying soil profile, producing duricrusts (such as laterites
and calcretes), sometimes accentuated by material derived also
from atmospheric sources (e.g., gypcrete in many hyper-arid
environments). Characteristically, such transport-limited situations are dominated by chemical (commonly biochemical)
weathering processes occurring in soil, saprolite, and at the
bedrock weathering front (although Heimsath et al., (1999)
report physical and biophysical weathering to be important
here).
The types of synergies among weathering processes and
between weathering and other geomorphic processes vary
hugely depending on whether we are looking at weatheringor transport-limited situations. Under weathering-limited
conditions, for example, short-term synergies between individual chemical, biological, and physical weathering processes
and the resultant small-scale surface weathering features are
generally of note. In transport-limited situations, complex
feedbacks between weathering and the development of duricrusts and the erosion of mobile regolith are of crucial importance. Of course, most large-scale geomorphic landscapes
are formed as a result of the interplay between weatheringlimited and transport-limited circumstances over long periods
of time, making things even more complex. For example, in
order to understand the development of many tors, it may be
necessary to consider both how weathering under soils under
transport-limited conditions may have initiated the features

and how subsequently weathering under weathering-limited
conditions (after soils and regolith were removed by erosion)
may have acted upon inherited weaknesses to produce the tor
landforms. Within arid environments, recent modeling suggests that tors may form as a result of instabilities within the
linked weathering and erosion system where regolith depth
influences weathering rate (Strudley et al., 2006) and systems
can flip from weathering-limited to transport-limited and vice
versa. Linking currently observable weathering processes on
weathering-limited tor surfaces to the genesis of the features
without considering these longer-term synergies and complexities will, in most cases, prove unsuccessful.
The distinction between weathering-limited and transportlimited systems is helpful in explaining the apparent isolation
of some weathering research from the rest of geomorphology.
For many practical reasons, it is easier to study weathering on
bare rock surfaces. Such environments provide many examples
of landforms largely produced by weathering (characteristically quite small scale), which have proved to be a rich area for
process-form linkage studies. Bare rock surfaces occur predominantly in coastal, mountain, cold, and hot desert environments, and in areas that exhibit particular geological
conditions (e.g., karst, active production of lavas, tectonically
produced scarps, ruiniform sandstone relief, ancient granite
landscapes, and the Danxia landscapes recognized in China
on continental redbeds). Analogs to such bare rock-dominated environments are also found in the built environment,
where surfaces are commonly datable thus explaining why
many weathering geomorphologists also work on problems
on buildings and monuments and why there is an effective
transfer of ideas and technologies between the fields of geomorphology and heritage science. By contrast, studies focusing
on weathering in transport-limited environments have tended
to be carried out by geomorphologists and geochemists (and

increasingly biogeochemists) interested in large-scale links
between nutrient cycles and landscape evolution. The expansion of interest in modeling long-term landscape evolution
has also fueled more interest in weathering under transportlimited situations, or in situations which involve the interplay
between weathering-limited and transport-limited situations,
particularly in terms of understanding and quantifying process
rates in order to develop meaningful transport rules. The advent of cosmogenic dating methods has allowed meaningful
measurement of weathering and denudation rates over long
timescales. Links between the two weathering communities
are not strong, largely because of the focus on very different
questions and methodologies.
4.2.4.3
The Critical Zone: An Aid to Understanding

Weathering Synergies?
The recent identification of the critical zone as a key focus for

geomorphological and biogeochemical research may help to
bring the two weathering communities together and improve
the development of our knowledge of synergistic weathering
processes. The critical zone is defined by the US National
Research Council as a heterogeneous, near-surface environment in which complex interactions involving rock, soil,
water, air and living organisms regulate the natural habitat
and determine availability of life-sustaining resources
(National Research Council, 2001, p. 2). The weathered profile
from soils to underlying rock has been seen as providing the
physical structure for the critical zone (Anderson and Anderson, 2010: p. 164), and, more broadly, it is the combination of
transport-limited and weathering-limited situations that
together provide the physical structure for the critical zone.
Thinking more creatively about the links between the two will
help improve our understanding of the critical zone, illustrating the importance of grappling with synergies between
different types and locales of weathering processes. Using
nonlinear dynamical system approaches, landscapes characterized by alternating patches of bare rock slopes (weatheringlimited) and soil-covered slopes (transport-limited) may be
interpreted as a critical zone exhibiting instability and selforganization.
4.2.5
4.2.5.1
Microscale Synergies between Weathering

Mechanisms
Some weathering processes by definition involve synergisms

between different small-scale mechanisms such as stress corrosion which, as discussed by Yatsu (1988: p. 49), refers to
cracking caused by the simultaneous action of a corrosive
agent and stress. Some microscope-based weathering studies,
such as that of Pye (1985) on the weathering of intrusive rocks
in Kenya, have found some evidence that stress corrosion and
other mechanisms are involved in the widening of microcracks during granular disintegration. Many biological weathering mechanisms also seem to operate in synergies, with
evidence that microorganism boreholes are produced by a
combination of biochemical action (chelation and dissolution
by organic acids and CO2) and biophysical mechanisms
(growth pressures, stresses exerted as microbes expand on
taking up water) to produce weathering effects such as shown
in Figure 8. Duane (2006) managed to separate out the biochemical and biophysical weathering effects within a mixed
lichen, moss, and fungi biofilm on calcareous sandstone
outcrops near the Moroccan coast, and illustrated how they
work synergistically together to cause rock breakdown. In
other circumstances, individual mechanisms may occur within
the same organism: rock interface zone, as, for example, noted
by Paradise (1997) who observed zones under one lichen
thallus where biochemical weathering dominated, and zones
where biophysical mechanisms dominated. As ever-more detailed observations can be made using high-powered microscopy, and as real-time monitoring of processes can be
observed with tools such as the environmental scanning
electron microscopy (ESEM), it is likely that many further
examples of such fundamental synergies will emerge linking
both mechanisms usually bracketed under the same process,
and those that straddle different types of process.
Research at this very fundamental level is advancing very
quickly, with studies carried out by a wide range of scientists
including microbiologists, geochemists, materials scientists,
physicists, and geomorphologists. We now know an increasing
amount about what goes on in terms of the chemical, physical, and biological processes operating within pore spaces and
Weathering Process Synergies
As is apparent from the preceding sections, many different

ways occur in which one can view synergies among weathering
processes as well as between weathering and other geomorphic/Earth surface processes. The term synergies is used
very broadly here, as process associations are not always
strictly synergistic, and, in some instances, one process can
retard rather than enhance the operation of another. The following subsections review a range of examples of weathering
process synergies, roughly organized in order of increasing
scale. Such a review cannot hope to be comprehensive, and it
should be regarded more as indicative of the sorts of interesting and innovative research currently being carried out and
the key questions being asked.
21
Figure 8 Cyanobacteria forming a mixed epilithic and endolithic

biofilm with boreholes on a bare limestone surface, Aldabra Atoll,
Indian Ocean.
22
on the surface. We can detect nanometer-scale transformations

of minerals and movement of moisture, and much improvement in our basic theoretical understanding has been made
(e.g., in terms of the fundamentals of salt weathering by
Scherer, 2004). However, even at this small-scale level, it has
proved very difficult to design experiments, make observations, or develop theories of how very different processes
work together.
4.2.5.2
Synergies between Weathering Agents and

Processes
There are vast numbers of examples of groups of different

weathering agents producing, in as yet often poorly understood co-associations, the breakdown of rocks and minerals.
Such information has usually come from carefully designed
laboratory experiments that investigate the relative efficacies of
single, paired, and multiple agents in producing breakdown
under highly controlled conditions. For example, Wells et al.
(2005) carried out simple but effective experiments on the
effectiveness of wetting and drying in weathering schist and
found that the incorporation of temperature variation into
their experimental design led to enhanced weathering rates
suggesting a synergistic association. Some advances have also
been made using field experiments which capture the multivariate nature of the weathering system in real environments.
Salts have, for example, long been thought to accelerate frost
weathering but as Williams and Robinson (1991: p. 352)
pointed out The search for a single theory of frost and salt
weathering would seem to be doomed to failure. The mechanisms will vary from site to site according to the freezing
regime, the salts present and their concentration, the rock type
and possibly other factors.
Another well-documented series of co-associations of weathering agents involves that between the wide range of different
salts that occur in nature. Some experiments, most fully
those of Goudie (1996), have investigated the relative weathering effectiveness of different combinations of salts on different rock types. He experimented using a range of mixed and
single salt solutions on York sandstone to assess the relative
efficacies of different concentrations and combinations of
salts. He found Na2CO3, Na2SO4, and MgSO4 both singly and
in combination to be more effective than NaCl, NaNO3, and
CaSO4 either singly or together. In many arid environments,
field studies indicate that NaCl and CaSO4 are present in
mixed forms and highly effective at weathering; so, caution
needs to be taken in extrapolating laboratory results to field
settings. As with interactions between salt and frost weathering
processes, the particularities of each location make it unlikely
that a general theory of how they interact can be found unless
we look more to nonlinear dynamical system approaches.
Inferences have also been drawn about the interdependence of biotic and abiotic weathering agents. For example, Gugliemin et al. (2005), from a combination of field
observations and microscope work, suggested that in ice-free
areas of Antarctica granite weathering within tafoni may be
largely caused by the combination of endolithic lichens, algae,
and salt. In laboratory experimental studies, Papida et al.
(2000) showed that salt weathering of dolomite and
limestone samples was enhanced in the presence of mixed

microbial communities, and similar effects might be observable in the field. Within soil research there has also been much
interest in the interplay between chemical and biological
weathering, with many workers arguing that biochemical
processes within the rhizosphere, including direct fungal
weathering of minerals, are of large importance. However,
Sverdrup (2009) suggested that biological influences, although vital to the soil weathering system, are more systemic
and indirect rather than localized and direct.
Several studies have looked at the influence of water on the
effectiveness of weathering by other agents, such as insolation
and salt weathering (Moores et al., 2008; Goudie et al., 2002).
Goudie et al. (2002), for example, experimented on various
mixtures of salt (NaNO3) and fog applications in an experiment on the weathering of chalk blocks. They found that the
amount of fog controlled the rate of breakdown, whereas salt
availability controls the timing of the response, and the style
of breakdown was the same regardless of the combination of
salt and fog used. In general, experiments (such as the classic
ones of Griggs in the 1930s looking at insolation weathering
with and without moisture) reveal that the presence of
moisture facilitates other weathering processes, but there is
little evidence as yet of exactly how such interactions work.
Moisture has also been shown to enhance the bioprotective
role (i.e., the way in which biofilms and other surface growths
can reduce weathering from other processes) in experiments
on the way in which lichens ameliorate thermal stresses on
limestone surfaces (Carter and Viles, 2003). This example illustrates how two agents (moisture and lichen cover) come
together to reduce the threat of weathering from another agent
(temperature cycling). Carter and Viles (2003) called this
bioprotection and such effects may be widely felt. Just to give
one example, Mustoe (2010) has recently shown that the
bioprotective role of endolithic biofilms can be an important
contributor to the development of tafoni.
Interactions between a variety of weathering agents and
processes are acknowledged to exist and thought to be important. It is the exception rather than the norm for weathering processes to operate singularly. Many recent
observations have indicated that biological agents and processes are critical mediators of weathering processes, which
were seen in the past to be entirely physical or chemical in
nature, because of the ubiquity of biofilm coverage on most
bare rock surfaces and of biota in soils. However, there have
been relatively few attempts to get to the heart of such interactions and quantify the linkages, largely because most studies
still concentrate on one or two processes rather than synergistic associations.
4.2.5.3
Synergies between Weathering Processes and

Weathering Landform Evolution
A suite of studies focusing on the development of tafoni

(cavernous weathering features of centimeter-to-meter dimensions) exemplify well how several processes may interact
with a developing topographic profile to produce positive
feedbacks (see, e.g., McBride and Picard, 2000; Turkington and
Phillips, 2004; Mustoe, 2010). Work on such synergies is often
15
14
13
12
11
10
9
8
7
6
Outer
Under overhang
Backwall
Lower
Base
07:15
11:15
15:15
Time of day
19:15
instabilities and self-organization within these nonlinear dynamical systems and thus it may not be possible to develop a
general theory of tafoni development at the detailed level of
synergies between individual processes, but rather one that
focuses on synergies between process and evolving topography. However, knowledge of which specific weathering
processes and agents are operating will be of immense value in
testing such general theories through realistic models.
As Viles (2005c) argued, many different types of studies are
being carried out on tafoni, characteristically asking very different types of questions and therefore focusing on different
types of synergies. There is a clear need now for more integrated studies to work on how associations of weathering
processes operate synergistically with the evolving topography
to produce tafoni and, indeed, the networks of cavernous
features that cover entire rock faces in some places. Similar
investigations are now needed on other, similar weathering
systems, which also seem to involve self-organizing behavior
such as rillenkarren, gnammas, phytokarst, solution grooves
on granite, and polygonal weathering of sandstone.
4.2.5.4
Synergies Linking Weathering Processes Across

Temporal and Spatial Scales
The past is generally the key to understanding the present in

terms of the behavior of weathering systems, because of a
range of inheritance effects. Exposure of rocks to weathering
and other processes in the past characteristically influences
Surface temperature (C)
Surface moisture
(% wood moisture equivalent)
based around careful field observations of morphology,

microenvironmental conditions, and weathering agents, with
some laboratory-based, experimental studies also contributing, as well as one or two examples of numerical modeling
(e.g., Huinink et al., 2004). The development of tafoni has
been seen in a range of studies over many different environments to involve at least two of the following agents: moisture,
temperature fluctuations, salts, biofilms, and wind. Under the
influence of the developing morphology of the tafoni, these
agents produce case hardening and/or bioprotection of the
outer surfaces, and core softening, granular disintegration, and
flaking and spalling of the interior surfaces, coupled with removal of the products of weathering. As the tafoni enlarge,
new conditions are established inside the hollow, becoming
increasingly divergent from conditions occurring on the surrounding rock surface (Figure 9). These conditions within the
hollow promote accelerated weathering, probably through
alterations to the evaporation rate, which has knock-on effects
on salt weathering and other processes. Meanwhile, the more
stable conditions on the surrounding rock face are amplified
by the development of case hardening and/or bioprotection,
thus reinforcing the difference between inner and outer environmental conditions and weathering regimes.
Which combination of process dominates in different parts
of the tafoni and surrounding rock depends on the local
conditions of geology and climate. The occurrence of positive
feedbacks as the cavern enlarges is symptomatic of nonlinear
system behavior. The net result reflects the ongoing
23
35
33
31
29
27
25
23
21
19
17
15
Outer
Under overhang
Backwall
Lower
Base
07:15
11:15
15:15
19:15
Time of day
Figure 9 Synergies between weathering micro-environments and the enlargement of tafoni at Gobabeb, central Namib Desert, Namibia. Circles
show measurement sites for data collected in July 2010. Graph on the left-hand side shows surface moisture (measured with a protimeter) and
graph on the right-hand side shows surface temperature (measured with IR thermometer).
24
how they weather today, showing the importance of initial

conditions to such self-organizing systems. Such inheritance
effects can occur over many different timescales, from decades
to millennia at least. Several studies have simulated varying
weathering histories on cut blocks of different rock types in
the laboratory and then subjected the blocks to experimental
weathering or field exposure trials. Warke (2007), for example,
pretreated blocks cut from limestone, sandstone, granite, and
basalt with varying numbers of heating and freezing cycles and
tracked their differing responses to weathering when exposed
on hillslopes in Death Valley, California. Pretreated blocks
were found to weather more quickly than untreated blocks,
confirming the importance of inherited weaknesses. Similarly,
Viles et al. (2010) pretreated basalt blocks with heating and
cooling and salt immersion and then simulated them in laboratory experiments to hot arid and Martian-type conditions.
Pretreated blocks behaved quite differently to those without
pretreatment, and, furthermore, the nature of pretreatment
had a significant effect on breakdown rates. The well-designed
experiments of Nicholson and Nicholson (2000) revealed
complex inheritance effects of flaws (cracks) within blocks
cut from 10 types of sedimentary rocks and subjected to
freezethaw weathering simulations. The coupled relationship
between rock flaws, rock strength, and textural properties of
the rock samples was found to exert the most important
control on breakdown.
Larger-scale processes may also have an influence on weathering at the small scale, showing synergies across spatial
scales. Such associations have been suggested for flaking and
other small-scale weathering within rock shelters in sandstones in the Drakensberg, South Africa, by Mol and Viles
(2010). Most explanations of the development of these
superficial weathering features focus on surface moisture and
temperature regimes, but they found that wider moisture
regimes deeper within the sandstones provided a key to
understanding the location and nature of weathering within
the rock shelters. Moisture moving through the sandstone
comes out at the surface in places within the rock shelters and,
coupled with the developing micro-environmental conditions
with the shelters, facilitates weathering processes, which produce flaking and other types of rock breakdown. Of course,
this does not exclude the importance of small-scale processes
and agents in the development of flaking, but reminds us that
larger-scale factors are also important. An interesting paper by
Ryne et al. (2008) looked at the reverse situation, where
small-scale processes interact to produce larger-scale effects.
Their study illustrates, for dolerites subjected to spheroidal
weathering in South Africa, how the intimate coupling between physical and chemical weathering processes generates
large reactive surface area by reaction-induced hierarchical
fracturing which then provides a major control on the development of spheroidal weathering.
4.2.5.5
monitoring studies with mathematical modeling. In smallscale, weathering-limited systems, weathering loosens rock
material, which then may become detached by mass movements or eroded by wave action, wind, and runoff, or river flow.
Matsuoka et al. (1998) carried out a careful field-based study of
the relative contributions of freezethaw, salt weathering, and
aeolian abrasion to the breakdown of rocks and development
of tafoni in Antarctica. They found that lack of moisture and
suitable abrasive material reduced the likely impacts of
freezethaw and abrasion, but that salt weathering, in combination with the removal of fine weathered debris by wind,
was the key process. Matsuoka and Sakai (1999) provided one
detailed example of the links between freezethaw weathering
and the nature and timing of rock fall in mountainous hillslopes, showing complex coupling between weathering and
mass movement processes. Similar co-associations occur on
jointed shore platforms where a range of weathering processes
(including wetting and drying, salt and biological processes)
may reduce the resistance of blocks to detachment by storm
events (Trenhaile, 2008). Studies from different environments
show fairly different results in terms of the relative importance
of weathering versus wave erosion on coastal platforms
(Stephenson and Kirk, 2000; Trenhaile and Porter, 2007), but
everyone acknowledges that there is some synergistic relationship between the two sets of processes. Interesting modeling
exercises carried out by Sharmeen and Willgoose (2006) illustrated for semi-arid hillslopes over 100-year periods that the
complex interactions between different styles of clast weathering and surface armoring influence erosion rates. They
further noted that such systems can be weathering- or transportlimited or, in some cases, a combination of the two.
At larger spatial scales, some intriguing relationships have
also been found between weathering and erosion across hillslopes, drainage basins, and large landscapes. Anderson
(2005), for example, reported on the close coupling between
glacial erosion and chemical weathering within glacierized
drainage basins, with glacial erosion producing very high reactive surface areas. The higher the erosion rate, the higher the
rate of chemical weathering. Dixon et al. (2009) illustrated the
key role that saprolite weathering plays in modulating
the balance between weathering and erosion within the Sierra
Nevada Mountains, California. Data collected from different
areas in different ways have shown variable relationships between erosion and chemical weathering, depending on whether weathering- or transport-limited conditions prevail.
However, the recent model of Gabet and Mudd (2009) provides one way to reconcile these. Their model suggests that
chemical weathering rates increase rapidly with increasing
erosion up to about 102 t km2 yr1, but that at higher rates of
denudation chemical weathering shows less of an increase
(Figure 10). Their model also predicts that drainage basins
with average regolith thickness of around 50 cm will experience the greatest amounts of chemical weathering.
Synergies between Weathering and Erosion
Across many scales and in weathering- and transport-limited

situations it is clear that weathering and erosion operate
in often quite complex co-associations. Studies of these
co-associations often combine medium- to long-term field
4.2.5.6
Synergistic Weathering Processes: Toward the

Future
As Yatsu (1988: p. 505) said The nature of weathering is utterly

intricate. Although he was thinking particularly about the small
Regolith thickness (m)
1600
800
6
0
0
Rego
1
2
Regolith thickness (m)
3
4
lith th
ickne
ss
2
athering rate
Volumetric we
0
1
25
100
10
1000
0
10 000
Volumetric weathering rate ( 105 t km3 yr 1)
Chemical weathering rate

(t km2 yr1)
Denudation (t km2 yr1)

Figure 10 Relationships between chemical weathering rate, regolith thickness, and denudation. Reproduced with permission from Figure 3 in
Gabet, E.J., Mudd, S.M., 2009. A theoretical model coupling chemical weathering rates with denudation rates. Geology 37, 151154.
scale, this statement is equally true at larger scales. The ideas put
forward, and the empirical and modeling studies discussed, in
this chapter all point to the need for geomorphologists to tackle
such intricacy by considering many different types of weathering
process synergies across a wide range of scales. A nonlinear
dynamical system framework provides an ideal way of approaching these synergies. Different questions will require focus
on rather different key synergies, but they are all important in
furthering our understanding of the complex interrelationships
between processes and landforms on the Earths surface. It is
clear that the time is now right for concerted, interdisciplinary
effort on improving our understanding of these synergies in
order to strengthen our knowledge on weathering and how it
interacts with landscape evolution at all scales. In order to make
progress in this area, we need to ensure better collection of data
at appropriate scales for the questions we are asking, as well as
enhanced communication between scientists working on very
different scales of weathering systems.
References
Anderson, S.P., 2005. Glaciers show direct linkage between erosion rate and
chemical weathering fluxes. Geomorphology 67, 147157.
Anderson, R.S., Anderson, S.P., 2010. Geomorphology: The Mechanics and
Chemistry of Landscapes. Cambridge University Press, Cambridge, pp. 637.
Bourdon, B., Bureau, S., Andersen, M.B., Pili, E., Hubert, A., 2009. Weathering rates
from top to bottom in a carbonate environment. Chemical Geology 258, 275287.
Carter, N.E.A., Viles, H.A., 2003. Experimental investigations into the interactions
between moisture, rock surface temperatures and an epilithic lichen cover in the
bioprotection of limestone. Building and Environment 38, 12251234.
Colman, S., 1981. Rock weathering rates as functions of time. Quaternary Research
15, 250264.
Dietrich, W.E., Reiss, R., Hsu, M., Montogmery, D.R., 1995. A process-based model
for colluvial soil depth and shallow landsliding using digital elevation data.
Dixon, J.C., 2004. Weathering. In: Goudie, A.S. (Ed.), Encyclopedia of

Geomorphology. Routledge, London, pp. 11081112.
Dixon, J.C., Thorn, C.E., 2005. Chemical weathering and landscape development in
mid-latitude alpine environments. Geomorphology 67, 127145.
Dixon, J.L., Heimsath, A.M., Amundson, R., 2009. The critical role of climate and
saprolite weathering in landscape evolution. Earth Surface Processes and
Landforms 34, 15071521.
Duane, M., 2006. Coeval biochemical and biophysical weathering processes on
Quaternary sandstone terraces south of rabat (Temara), northwest Morocco. Earth
Etienne, S., 2002. The role of biological weathering in periglacial areas: a study of
weathering rinds in South Iceland. Geomorphology 47, 7586.
Gabet, E.J., Mudd, S.M., 2009. A theoretical model coupling chemical weathering
rates with denudation rates. Geology 37, 151154.
Gilbert, G.K., 1877. Report on the geology of the Henry Mountains. Government
Printing Office, Washington, 160 pp.
Goudie, A.S., 1996. Geomorphological hotspots and global warming.
Interdisciplinary Science Reviews 21(3), 253259.
Goudie, A.S., Migon, P., 1997. Weathering pits in the Spitzkoppe area, central
Namib Desert. Zeitschrift fur Geomorphologie 41, 417444.
Goudie, A.S., Viles, H.A., 1997. Salt Weathering Hazards. Wiley, Chichester,
241 pp.
Goudie, A.S., Wright, E., Viles, H.A., 2002. The roles of salt (sodium nitrate) and
fog in weathering: a laboratory simulation of conditions in the northern Atacama
Desert, Chile. Catena 48, 255266.
Gregory, K.J., 2010. The Earths Land Surface. Sage, London, 348 pp.
Gugliemin, M., Cannone, N., Strini, A., Lewkowicz, A.G., 2005. Biotic and
abiotic processes on granite weathering landforms in a cryotic environment,
northern Victoria Land, Antarctica. Permafrost and Periglacial Processes 16,
6985.
Hartmann, J., Jansen, N., Durr, H.H., Kempe, S., Kohler, P., 2009. Global CO2
consumption by chemical weathering: what is the contribution of highly active
weathering regions? Global and Planetary Change 69, 185194.
Heimsath, A.M., Dietrich, W.E., Nishiizumi, K., Finkel, R.C., 1999. Cosmogenic
nuclides, topography, and the spatial variation of soil depth. Geomorphology 27,
151172.
Huinink, H.P., Pel, L., Kopinga, K., 2004. Simulating the growth of tafoni. Earth
Kennedy, B.A., 2000. Denudation. In: Thomas, D.S.G., Goudie, A.S. (Eds.), The
Dictionary of Physical Geography, Third edn. Blackwell, Oxford, pp. 127128.
Matsuoka, N., 2001. Microgelivation versus macrogelivation: towards bridging the
gap between laboratory and field frost weathering. Permafrost and Periglacial
26
Matsuoka, N., Morikawa, K., Hirakawa, K., 1998. Field experiments on physical
weathering and wind erosion in an Antarctic cold desert. Earth Surface
Matsuoka, N., Sakai, H., 1999. Rockfall activity from an alpine cliff during thawing
periods. Geomorphology 28, 309328.
McBride, E.F., Picard, M.D., 2000. Origin and development of tafoni in Tunnel
Spring Tuff, Crystal Peak, Utah, USA. Earth Surface Processes and Landforms
25, 869879.
Mol, L., Viles, H.A., 2010. Geoeletric investigations into sandstone moisture
regimes: implications for rock weathering and the deterioration of San rock art
in the Golden gate reserve, South Africa. Geomorphology 118, 280287.
Moores, J.E., Pelletier, J.D., Smith, P.H., 2008. Crack propagation by differential
insolation on desert surface clasts. Geomorphology 102, 472481.
Mustoe, G.E., 2010. Biogenic origin of coastal honeycomb weathering. Earth
National Research Council, 2001. Basic Research Opportunities in Earth Science.
National Academy Press, Washington, DC, 168 pp.
Navarre-Sitchler, A., Brantley, S., 2007. Basalt weathering across scales. Earth and
Nicholson, D.T., Nicholson, F.H., 2000. Physical deterioration of sedimentary rocks
subjected to experimental freezethaw weathering. Earth Surface Processes and
Landforms 25, 12951307.
Papida, S., Murphy, W., May, E., 2000. Enhancement of physical weathering of
building stones by microbial populations. International Biodeterioration and
Biodegradation 46, 305317.
Paradise, T.R., 1997. Disparate sandstone weathering beneath lichens, Red
Mountain, Arizona. Geografiska Annaler 70A, 177184.
Phillips, J.D., 1999. Earth Surface Systems: Complexity, Order and Scale. Blackwell,
Oxford, 180 pp.
67, 255272.
Pope, G.A., Dorn, R.I., Dixon, J.C., 1995. A new conceptual model for
understanding geographical variations in weathering. Annals, Association of
Pye, K., 1985. Granular disintegration of gneiss and migmatites. Catena 12,
191199.
Ryne, A., Jamtveit, B., Mathiesen, J., Malther-Srenssen, A., 2008. Controls on
rock weathering rates by reaction-induced hierarchical fracturing. Earth and
Scherer, G.W., 2004. Stress from the crystallization of salt. Cement and Concrete
Research 34, 16131624.
Sharmeen, S., Willgoose, G.R., 2006. The interaction between armouring and
particle weathering for eroding landscapes. Earth Surface Processes and
Landforms 31, 11951210.
Stephenson, W.J., Kirk, R.M., 2000. Development of shore platforms on Kaikoura

Peninsula, South Island, New Zealand: II, the role of subaerial weathering.
Strudley, M.W., Murray, A.B., Haff, P.K., 2006. Regolith thickness instability and the
formation of tors in arid environments. Journal of Geophysical Research 111,
F03010. 2006. http://dx.doi.org/10.1029/2005JF000405,? 2006.
Sverdrup, H., 2009. Chemical weathering of soil minerals and the role of biological
processes. Fungal Biology Reviews 23, 94100.
Trenhaile, A.S., 2008. Modelling the role of weathering on shore platform
development. Geomorphology 94, 2439.
Trenhaile, A.S., Porter, N.J., 2007. Can shore platforms be produced solely by
weathering processes? Marine Geology 241, 7992.
Turkington, A.V., Phillips, J.D., 2004. Cavernous weathering, dynamical
instability and self-organization. Earth Surface Processes and Landforms 29,
665675.
Viles, H.A., 2001. Scale issues in weathering studies. Geomorphology 41, 6372.
Viles, H.A., 2005a. Microclimate and weathering in the central Namib Desert,
Namibia. Geomorphology 67, 189209.
Viles, H.A., 2005b. Can stone decay be chaotic? In: Turkington, A.V. (Ed.), Stone
Decay in the Architectural Environment. Geological Society of America Special
Paper 390, Boulder, CO, pp. 1116.
Viles, H.A., 2005c. Self-organized or disorganised? Towards a general explanation
of cavernous weathering. Earth Surface Processes and Landforms 30,
14711473.
Viles, H.A., Ehlmann, B., Wilson, C.F., Cebula, T., Page, M., Bourke, M., 2010.
Simulating weathering of basalt on Mars and Earth by thermal cycling.
Geophysical Research Letters 37, LXXXXX, 2010. http://dx.doi.org/10.1029/
2010GL043522,? 2010.
Viles, H.A., Goudie, A.S., 2000. Weathering, geomorphology and climatic variability
in the central Namib Desert. In: McLaren, S.J., Kniveton, D.R. (Eds.), Linking
Climate Change to Land Surface Change. Kluwer, Dordecht, pp. 6582.
Viles, H.A., Goudie, A.S., 2007. Rapid salt weathering in the coastal Namib Desert:
implications for landscape development. Geomorphology 85, 4962.
Viles, H.A., Naylor, L.A., Carter, N.E.A., Chaput, D., 2008. Biogeomorphological
disturbance regimes: progress in linking ecological and geomorphological
systems. Earth Surface Processes and Landforms 33, 14191435.
Warke, P.A., 2007. Complex weathering in drylands: implications of stress history
for rock debris breakdown and sediment release. Geomorphology 85, 3048.
Wells, T., Binning, P., Willgoose, G., 2005. The role of moisture cycling in the
weathering of a quartz chlorite schist in a tropical environment: findings of a
laboratory simulations. Earth Surface Processes and Landforms 30, 413428.
Williams, R.B.G., Robinson, D.A., 1991. Frost weathering of rocks in the presence
of salts a review. Permafrost and Periglacial Processes 2, 347353.
Yatsu, E., 1988. The Nature of Weathering. Sozosha, Tokyo, 624. pp.
Biographical Sketch
Heather Viles is professor of biogeomorphology and heritage conservation in the School of Geography and the
Environment, University of Oxford, UK, and visiting research fellow in the School of Geography, Archaeology and
Environmental Sciences, University of the Witwatersrand, South Africa. Her research focuses on biogeomorphology, weathering (mainly in extreme environments), and the deterioration and conservation of building stone.
She has carried out field research in many places, including Essex, UK, NW Australia, Germany, Namibia, South
Africa, Libya, Washington State, USA, and Aladabra Atoll, Seychelles. Her publications include over 115 papers in
refereed journals and edited volumes, and she has edited nine books and collections of papers, including Biogeomorphology. She is the author of a number of books including Salt Weathering Hazards and A Very Short
Introduction to Landscapes and Geomorphology, both with Andrew Goudie. She is currently vice-president (expeditions and fieldwork) of the Royal Geographical Society (with IBG), and vice chair of the British Society for
Geomorphology. She holds an MA in geography from the University of Cambridge and a DPhil in geography
from the University of Oxford.
4.3 Pedogenesis with Respect to Geomorphology

JC Dixon, University of Arkansas, Fayetteville, AR, USA
4.3.1
4.3.2
4.3.3
4.3.3.1
4.3.3.2
4.3.3.3
4.3.3.4
4.3.3.5
4.3.4
4.3.5
4.3.5.1
4.3.6
4.3.7
4.3.7.1
4.3.7.2
4.3.8
4.3.9
4.3.9.1
4.3.9.2
4.3.10
References
Introduction
Pedogenic Processes
Pedogenesis and Landscape Evolution
Ferricrete and Laterite
Calcrete and Dolocrete
Gypcrete
Silcrete
Duricrusts and Landscape Evolution
Soil Chronosequences
Soils as Indicators of Landscape Stability
Landscape Evolution in the Midwestern US
Soils and Climate Change
Soil-Slope Relationships
The Catena Concept
Catenas in Different Climates
Hillslope/Soil Process Interaction
Soils and Sedimentation
Aeolian Sedimentation
Fluvial Sedimentation
Conclusions
Glossary
Alkalization (solonization) The accumulation of sodium
ions in the soil.
Allitization Accumulation of aluminum in the soil
profile.
Braunification Release of iron from primary minerals by
weathering and progressive browning of the soil.
Calcification Accumulation of calcium carbonate in soils.
Calcrete Terrestrial materials cemented primarily by
calcium carbonate.
Catena A sequence of soils along a slope with different
properties due to differences in relief.
Cumulization Addition of aeolian-derived materials and
water to the soil solum.
Dealkalization (solodization) Leaching of sodium ions
and salts from nitric soil horizons.
Decalcification Processes that remove calcium carbonate
from soil horizons.
Desalinization Removal of soluble salts from soils.
Dolocrete Terrestrial materials cemented primarily by
dolomite.
Eluviation Removal of materials from one horizon to
another lower horizon in a soil profile.
28
29
29
30
30
30
30
31
31
32
34
35
35
35
36
38
38
38
39
40
40
Feralization (ferritization) Accumulation of sesquioxides

in soil profiles.
Ferricrete Terrestrial materials cemented principally
by iron.
Gleization Reduction of iron under waterlogged
conditions to produce bluish to greenish gray matrix colors.
Gypcrete Terrestrial materials cemented primarily by
gypsum.
Illuviation Movement of material from one soil horizon
into a generally lower soil horizon.
Laterite Iron-cemented terrestrial materials resulting from
the removal of mobile elements and silica.
Laterization Process of laterite formation, usually
associated with the development of a deep-weathering profile.
Lessivage Mechanical migration of particles from A and E
horizons to soil B horizons.
Leucinization The lightening of soil horizons due to
removal of organic material, clay, and/or sesquioxides.
Melanization The darkening of soil horizons principally
by the addition or transformation of organic materials.
Pedoturbation Mixing of soil materials by a diversity of
agents including animals, plants, ice, and swelling and
shrinking of clays.
Dixon, J.C., 2013. Pedogenesis with respect to geomorphology. In: Shroder J.

(Editor in Chief), Pope, G.A. (Ed.), Treatise on Geomorphology. Academic
Press, San Diego, CA, vol. 4, Weathering and Soils Geomorphology,
pp. 2743.
http://dx.doi.org/10.1016/B978-0-12-374739-6.00058-0
27
28
Pedogenesis with Respect to Geomorphology
Podsolization Migration of organic materials, clays, and

sesquioxides in a soil profile resulting in an accumulation of
silica in the eluvial horizon.
Rubification The development of red colors in soils.
Salinization Accumulation of soluble salts in soils.

Silcrete Terrestrial materials cemented primarily by silica.
Soil geomorphology Integration of the study of soils and
the study of landscapes.
Abstract
The integration of soil and geomorphological studies constitute the scientific field known as soil geomorphology. Soil
geomorphological studies consist of six principal components, including the development of chronosequences, soils as
indicators of landscape stability, soils as indicators of climate change, soil development and drainage basin evolution, soils
as indicators of erosion and deposition processes, and soils as indicators of Quaternary parent materials and stratigraphy.
Land surface morphology fundamentally controls patterns of water movement across and through the landscape. These
patterns in turn affect the way in which solid and dissolved materials move, giving rise to the formation of soils with
distinctive properties. Soils that have been indurated and formed duricrusts exert profound influences on landscape
morphology.
4.3.1
Introduction
The term Soil Geomorphology was introduced into the literature about four decades ago by Ruhe (1974) in a paper
entitled Holocene Environments and Soil Geomorphology in
midwestern United States published in Quaternary Research.
Although Ruhe was probably amongst the first geomorphologist to use the term, he was certainly not the first worker to
recognize the interdependence of soils and geomorphology.
As early as the mid-1930s, Milne (1935) recognized that
particular slope forms were associated with particular soil sequences and formulated the concept of the catena, a sequence
of soils linked by conditions of topography. Jenny (1941)
clearly recognized the importance of geomorphology in his
identification of relief as one of the state factors in soil formation. In the 1940s, Butler (1959) proposed the concept of
the K-Cycle to aid in the interpretation of soils as indicators
of periods of landscape stability and instability. Subsequently,
the relationship between soils and geomorphology became
more formalized in the development of soil-slope relationships. Concepts such as soil-geomorphic relations (Daniels
et al., 1971) and pedogeomorphology (Conacher and Dalrymple, 1977) subsequently emerged.
Early studies in soil geomorphology stressed the interrelationships between soil development and hillslope processes. In the 1980s, however, soil geomorphology broadened
in perspective to stress a greater diversity of relationships between soils and landscapes (Dixon, 1986a). This broadening
particularly reflected in the appearance of Birkelands (1974)
Pedology, Weathering and Geomorphological Research, which
was subsequently followed by several collections of papers including Soils and Landforms by Gerrard (1981), Soil Geomorphology (Gerrard, 1992), Geomorphology and Soils
edited by Richards et al. (1985), Soils and Quaternary Landscape Evolution edited by Boardman (1985), and Soils and
Quaternary Geology by Catt (1986). The breadth of research in
the field is further exemplified by the publication of the proceedings of the 21st Binghampton Geomorphology Symposium
on the same topic (Knuepfer and McFadden, 1990).
Soil geomorphology is the study of the interaction of
pedogenic and geomorphic processes for interpreting present
and past landscapes. It is an assessment of the genetic relationship between soils and landforms (Gerrard, 1992, 1993),
and provides a critical framework for the study of soil genesis
as well as the evolution and/or the stability of landforms and
landscapes (McFadden and Knuepfer, 1990). Olson (1989)
defined soil geomorphology as the study of the landscape and
the influence of that landscape on the formation of soils. She
suggested that soil geomorphology studies consist of three
essential components: (1) an assessment of the surficial stratigraphy and parent materials of an area; (2) the definition of
geomorphic surfaces in space and time; and (3) the correlation of soil properties with landscape features. Integration of
pedology and geomorphology occurs in a great diversity of
ways. As indicated above, early emphasis was on the relationship between soils and hillslopes, especially as it relates to
patterns of water movement. However, as soils are a palimpsest of natural events and processes (Johnson and Rockwell,
1982), they record long-term history of erosion, transportation, and deposition of materials on the Earths surface and
thus the history of landscape development.
Birkeland (1990) suggested that studies of soil geomorphology can be classified into one of the four areas:
1. The development of soil chronosequence frameworks that
can be used to establish the age of surficial deposits.
2. The use of soils as indicators of long-term or short-term
landscape stability. To do this, it is necessary to know how
long it takes to form key soil properties in different
environments.
3. The determination of soil property relations that indicate
climatic change.
4. The examination of the interaction of soil development,
rainfall, infiltration, runoff, and erosion on hillslopes.
Schaetzl and Anderson (2005) added two additional
themes to the original four of Birkeland. These authors suggest
that soil-geomorphic studies also often incorporate the
examination of soils as indicators of past or ongoing sedimentological and depositional processes. In addition, they
suggested that soil-geomorphic studies commonly incorporate
the study of soils as indicators of Quaternary stratigraphy and
parent materials. Gerrard (1993) suggested that the challenge
for soil geomorphology is to integrate elements from all of

these areas to produce a conceptual framework for landscape
evolution.
4.3.2
Pedogenic Processes
Soil formation is the result of a great variety of organic and

inorganic processes operating synergistically to alter consolidated or unconsolidated parent materials. These processes are
fundamental processes that contribute to the development of
soil horizons. Variability in the nature and rate at which soil
forming processes operate is fundamentally controlled by the
soil forming factors: climate, organic activity, relief, parent
material, and time (Jenny, 1941).
Schaetzl and Anderson (2005) identified approximately
three dozen individual processes responsible for soil formation, built on a slightly shorter list developed by Buol et al.
(1997). All of these processes, however, can be collapsed into
four principal categories following Simonson (1978): additions, removals, translocations, and transformations.
Additions, also referred to as enrichment, are those processes that fundamentally involve the addition of material to
the soil body from external sources (Simonson, 1978; Buol
et al., 1997; Schaetzl and Anderson, 2005). This group of
processes includes cumulization or the addition of aeolian
and hydrologically derived mineral particles to the soil solum,
melanization or darkening of mineral horizons by the addition of organic matter and humus, and littering or the
addition of organic matter on the mineral soil horizon surface.
Removals collectively refer to those pedogenic processes
that completely remove organic and inorganic materials from
the soil profile. They principally involve the processes of
leaching that refers to the washing out of soluble materials
from the soil solum, and surface erosion that removes organic
and inorganic material from the soil surface (Simonson, 1978;
Buol et al., 1997).
Translocation processes involve movement of materials
within the soil profile. Although translocations are predominantly vertically downward, they may be vertically
upward (upward capillary flow) and also horizontal in
movement. These processes are substantially responsible for
horizon differentiation. The downward and lateral movement
of materials from one portion of a soil profile to another is
referred to as eluviation. The movement of materials into a
portion of a soil profile from overlying horizons is referred to
as illuviation. Eluviation and illuviation collectively refer to
the movement of salts, particulates, and organic matter.
Accumulation of calcium carbonate is referred to as calcification, whereas removal of calcium salts from a soil horizon is
referred to as decalcification. Accumulation of soluble salts is
referred to as salinization, whereas movement of soluble salts
from a soil horizon is referred to as desalinization. Accumulation of sodium ions in a soil horizon is referred to alkalinization or solonization, whereas the removal of sodium
ions from a soil profile is referred to as dealkalinization or
solodization (Schaetzl and Anderson, 2005).
The mechanical migration of mineral matter from the A
and E horizons into the B horizon of the soil is referred to as
lessivage. This process results in the formation of horizons
29
such as Bt horizons as well as the development of silt bulges in

soil profiles. The mixing of soil materials, both inorganic and
organic, is referred to as pedoturbation. Pedoturbation may
result in either amplification of horizonation or in the homogenization of the profile (Johnson et al., 1987). The mixing
of soil materials can occur in a great variety of manners including, but not limited to, cryoturbation, faunalturbation,
and bioturbation (see Wood and Johnson (1978) for a more
comprehensive list). This group of processes is particularly
important in their formation of stone lines in soils (Johnson
and Hester, 1972; Johnson and Balek, 1991).
The migration of aluminum, iron, and/or organic matter
resulting in the concentration of silica in the soil profile is
referred to as podsolization (Buol et al., 1997). Conversely, the migration of silica out of the solum with the accompanying concentration of sesquioxides (oxides of Fe and
Al) within the profile is referred to as laterization. This process is also occasionally referred to as desilication or desilicification with accompanying feralization, ferritization, and
allitization (Widdowson, 2007).
Processes collectively responsible for the darkening of soil
profiles are referred to as melanization, whereas those processes collectively responsible for the lightening of the soil
profile or soil horizons are referred to as leucinization (Buol
et al., 1997; Schaetzl and Anderson, 2005).
The fourth group of pedogenic processes identified by
Simonson are those collectively referred to as transformation
processes. This group of processes result in the formation
of new materials as a result of the breakdown of primary
mineral and organic matter. They might simply be referred to
as chemical and biochemical weathering processes. Decomposition refers to the breakdown of organic and mineral
matter into more stable secondary products. The formation of
new materials in place is referred to as synthesis or neoformation (Buol et al., 1997).
Some processes obviously do not fit nicely into just one of
these categories. For instance, they may involve a combination
of transformation and translocation processes. For example,
processes such as braunification, rubifaction, and ferruginization involve the release of iron from primary minerals and
the dispersion of particles of iron oxide in progressively increasing amounts. Progressive oxidation eventually gives a soil
progressively brownish, reddish brown, or red color, respectively. Similarly, the reduction of iron under waterlogged or
impeded drainage conditions results in gleization. This process results in the production of bluish and greenish matrix
colors with or without mottles of red, brown, and yellow hue,
and the development of iron and manganese nodules (Buol
et al., 1997; Schaetzl and Anderson, 2005).
4.3.3
Pedogenesis and Landscape Evolution
Many of the pedogenic processes discussed above, when operating over long periods of time, at rapid rates, or in especially favorable landscape or climatic settings, ultimately result
in the development of positive relief forms in the landscape.
This is especially true for those processes that result in the
formation of a group of soil materials collectively referred to
as duricrusts (Woolnough, 1927, 1930). Duricrusts consist of
30
a group of chemical sediments dominated by iron (ferricrete),

calcium carbonate (calcrete), calcium sulfate (gypcrete),
dolomite (dolocrete), and silica (silcrete) that are for the most
part strongly indurated or cemented and that result primarily
(but not exclusively) from pediogenic processes. In addition,
many intergrade varieties are identified. On induration, they
represent resistant elements in the landscape and frequently
manifest their presence in the landscape as cappings on
positive relief elements and as inverted relief elements (Nash
and McLaren, 2007).
4.3.3.1
Ferricrete and Laterite
Ferricrete and laterite are ferruginous duricrusts cemented by

iron oxides occurring as indurated continuous crusts and soil
horizons in the landscape (Lamplugh, 1902; Bourman, 1993).
Fundamental to the formation of ferricrete and laterite is the
formation and accumulation of insoluble ferric iron in soils
and regolith. Laterites form principally as a result of the removal of mobile elements associated with prolonged and intense chemical weathering (Aleva, 1994; Widdowson, 2007).
Accumulation of iron can, and frequently does, also occur as a
result of migration of iron into soil or regolith from outside
the profile. Iron-rich crusts formed in this manner are referred
to as ferricrete (Aleva, 1994; Widdowson, 2007). However, in
keeping with the original definition of ferricretes (Lamplugh,
1902), there is a strong movement to abandon the term laterite in preference for a nongenetic term ferricrete (Hamming,
1968; Bourman, 1993; Taylor and Eggleton, 2002). Ferricrete,
therefore, can be the result of a variety of formational processes including pedogenic formation, in-situ formation,
groundwater fluctuations, fluvial, lacustrine, and secondary
reprecipitation (Goudie, 1973). They are thus the result of the
synergistic operation of numerous iron-concentrating processes.
4.3.3.2
Calcrete and Dolocrete
Calcrete refers to terrestrial materials that have been cemented

or replaced predominantly by calcium carbonate. It occurs
in stages ranging from powdery to strongly indurated. The
mechanism of calcification may be pedogenic or nonpedogenic, although pedogenic processes are widely viewed to
be the most widespread mechanisms of calcrete formation
(Wright, 2007; Dixon and McLaren, 2009). Goudie (1973,
1983) suggests several pedogenic models for calcrete formation
including the concentration of carbonate by downward translocation in percolating groundwaters, the concentration of
carbonate by capillary rise waters, in-situ case hardening by
carbonate, and cementation by carbonate. Nonpedogenic calcretes are principally associated with carbonate accumulation
by groundwater (Nash and McLaren, 2003; McLaren, 2004).
Calcretes occur in a variety of morphologies that have been
interpreted to represent stages of calcrete formation. One of the
earliest and most internationally used evolutionary models is
that of Netterberg (1969) who identified six fundamental types
(forms) of calcrete: calcified soil, powder calcrete, nodular
calcrete, honeycomb calcrete, hardpan calcrete, and boulder
calcrete. In North America, the widely used evolutionary model
of calcrete formation is that of Gile et al. (1966), modified by
Machette (1985). This model identifies numerous stages of

calcrete formation depending on the degree of induration,
which range from Stage I consisting of thin filament coatings
and discontinuous pebble coatings, to Stage VI consisting of
massive, multilaminar, and brecciated carbonate with pisoliths
and case hardening (Figure 1).
Where the dominant carbonate cement is dolomite
(CaMgCO3), the duricrusts are referred to as dolocretes
(Goudie, 1973, 1983). As with calcretes, dolocrete may be
pedogenic or nonpedogenic in origin. Dixon (1978) and
Hutton and Dixon (1981) working on pedogenic calcretes in
southern South Australia identified dolomite-dominated
horizons at depth within profiles irrespective of the degree of
induration, and proposed a fundamentally pedogenic model
for their formation. More recently, Dixon (2010) investigating
carbonate profiles in the St. Vincent Basin of South Australia
reported dolocrete dominating the lowest horizons in soil
profiles in low topographic settings. Nonpedogenic dolocrete,
predominantly groundwater dolocretes, have been described
from the Middle East (Khalaf, 1990, 2007) and from Europe
(Colson and Cojan, 1996).
4.3.3.3
Gypcrete
Gypsum-dominated crusts develop in environments receiving

less than 200250 mm of rainfall annually (Watson, 1983).
These are duricrusts formed by the cementation of soils and
regolith by gypsum. As with other duricrust types, they may be
of pedogenic or nonpedogenic origin. Pedogenic models of
gypcrete formation, however, have been most widely embraced (Watson, 1979, 1983, 1985; Chen, 1997). Most studies
of pedogenic gypcretes stress the importance of aeolianderived gypsum (Dan et al., 1982; Watson, 1985; Amit and
Gerson, 1986; Reheis, 1987a). Reheis (1987a) and Chen et al.
(1991) suggest that reworking of hydromorphic gypsum crusts
is also an important process.
Nonpedogenic models of gypcrete formation are principally related to gypsum deposition by groundwater and in
lacustrine settings. Extensive gypcrete occurrences are reported
from playa lakes in central Australia by Jacobson et al. (1988).
Lacustrine gypcretes are reported by Watson (1985, 1988)
from Tunisia and the Namib Desert. Deposition in groundwater systems involves complex interactions between vadose
and phreatic waters, whereas deposition in lacustrine systems
is largely due to evaporation.
4.3.3.4
Silcrete
Silcrete is a surface or near-surface deposit of soil, saprolite,

or sediment that has been cemented by secondary silica to
form an indurated mass (Milnes and Twidale, 1983; Nash and
Ullyot, 2007). Many morphological classifications of silcrete
exist based on macro- (Smale, 1973; Wopfner, 1978, 1983)
and micromorphological (Summerfield, 1983) characteristics.
Thiry (1999) has proposed a genetic classification of silcretes
in which he recognized pedogenic calcretes associated
with silicification in the soil profile, and nonpedogenic
silcretes encompassing groundwater and evaporitic modes of
formation.
31
Carbonate stage
I
Depth (m)
II
A
III
A
Bk
Bkm
BC
IV
VI
Bkm1
Bkm
Bkm1
Bkm2
Bkm2
Bkm2
BCk
BCk
BCk
BCk
Bkm1
(a)
Bkm1
Bkm1
Bkm1
Bkm
Depth (m)
Bk
Bk
Bkm2
Bkm2
Bkm2
BC
c
BCk
BCk
BCk
BCk
(b)
Figure 1 Model of calcrete development in coarse-grained soils (a) and fine-grained soils (b). Stages IVI represent progressively increasing
amounts and degrees of induration by carbonate. Adapted from Gile, L.H., Peterson, F.F., Grossman, R.B., 1966. Morphological and genetic
sequences of carbonate accumulation in desert soils. Soil Science 101, 347360, and Machette, M.N., 1985. Calcic soils of the southwestern
United States. Geological Society of America Special Paper, vol. 203, pp. 121.
Silcrete formation involves three elements: a silica source,

silica transport, and silica deposition. Silica sources include
weathering of quartz and aluminosilicate minerals by acid
weathering. Silica may also be derived from aeolian addition
and organic silica (Milnes and Thiry, 1992; Nash and Ullyot,
2007; Dixon and McLaren, 2009).
Perhaps the most widely applicable models of silcrete
formation are those associated with pedogenesis. These
models are of two fundamental types: vertical rise and illuviation models. Early explanations of silcrete formation invoked principally vertical-rise models (Woonough, 1927;
Frankel and Kent, 1938). These models, however, suffer from
several limitations associated with the distance of capillary
transport and the efficiency of evaporative precipitation at
depth (Summerfield, 1982, 1983). Of broader applicability
are illuvial pedogenic and groundwater models. Illuvial
models postulate the downward movement of silica-rich
waters under the influence of gravity eventually resulting in
the deposition of silica as a result of loss of transport capacity
(Thiry and Milnes, 1991; Simon-Coinc- on et al., 1996; Thiry
et al., 2006).
Intergrade duricrusts consisting of both calcium and silica
cements have been reported from a variety of locations including Africa, Australia, and North America (Dixon and
McLaren, 2009). Similarly, intergrades of silcrete and ferricrete
are reported from Australia (Lee and Gilkes, 2005).
4.3.3.5
Duricrusts and Landscape Evolution
Duricrusts exert a significant influence on topography in tropical

and subtropical regions of the world. Duricrusts developed on
paleosurfaces are commonly preserved as horizontal to subhorizontal caprocks on plateau and mesa surfaces (Mabbutt,
1967; Goudie, 1984, 1985). Paleosurfaces are widely preserved
in the Gondwana surfaces of Australia (Hays, 1967) and
southern Africa (Summerfield, 1982, 1983; Marker et al., 2002).
Where duricrust formation occurs in topographic lows,
relief inversion commonly occurs as a result of cementation
and erosion of softer adjacent unindurated sediments (Ollier
et al., 1988; Ollier and Pain, 1996). In Australia and southern
Africa, silcretes developed in paleochannels crop out in inverted relief (Alley et al., 1999; Hill et al., 2003). The duricrusts act as elements of resistance in the landscape and thus
exert strong controls on slope development and hydrology.
Undercutting and caprock collapse result in the development
of breakaways and provide debris for desert pavement formation (Nash and McLaren, 2007).
4.3.4
Much soil-geomorphological research involves investigating

the development of soils with time, generally referred to as soil
32
chronosequences. The rationale of chronosequences is that the

time factor is the variable under consideration and other soil
forming factors are considered constant. A chronosequence
can be defined as a genetically related suite of soils in which
vegetation, topography, and climate are similar (Gerrard,
1993; Harden, 1982). But when a soil property is plotted
against soil age, changes with age may or may not signify the
result of a single pedogenic process acting on soils through
time. Processes or changes in conditions in the history of the
soils are not always recorded or preserved in morphology.
Most chronosequence studies have been undertaken on
landforms underlain by Quaternary deposits. In general, there
is a predictable variation in both soil horizon sequences and
abundances of constituents such as clay, iron, and calcium
carbonate. With time, the rate at which these constituents
accumulate in the soil is a function of parent material, bioclimate, and atmospheric dust influx (Birkeland, 1990). Profile properties can be used to estimate ages of deposits as well
as faulting events, but there may be frequently considerable
error in the estimate. Birkeland (1990) discussed chronofunctions from a wide variety of climatic settings and pointed
out considerable variability depending on the nature of controlling factors. Soil chronosequences and problems associated with their development are discussed by Stevens and
Walker (1970), Yaalon (1975), and Bockheim (1980).
Soil development can be either progressive or regressive
with time (Johnson and Watson-Stegner, 1987). Under progressive development, soils become more strongly differentiated by horizons, and horizon contrasts become stronger. If
material is added to the surface (such as aeolian materials), it
is assimilated into the profile and profile thickness increases.
Many, if not most chronosequences, are of this type (see
Birkeland (1990) for examples from contrasting climates). In
contrast, examples of regressive pedogenesis are haploidization (decreasing horizon contrasts within profiles), addition
of material to the surface at a rate that suppresses pedogenesis
or suppression of pedogenesis by surface erosion. Detection
of regressive pedogenesis is difficult but it would result in poor
or no correlation of soil or soil properties with time. Some
regressive processes might occur randomly with time or be
triggered by climate change (Birkeland, 1990).
Vreeken (1975) categorized chronosequences into four
different types. A post-incisive chronosequence implies that
each soil is formed in sequence but at successive times. It is
assumed that younger soils progress to older soils in sequence,
so a plot of soil properties versus time tracks soil development
(Birkeland, 1990). A pre-incisive chronosequence involves
soils that began forming simultaneously but were buried at
successive, more recent, times. One of them may still be at the
surface. Soils developing on a fresh glacial till that has then
been gradually buried would lead to such a sequence. Soils
that both began forming and were buried at different times are
called fully time-transgressive chronosequences (Gerrard,
1992). The third type of chronosequence is a time-transgressive sequence with historic overlap. In these chronosequences,
because of erosion and deposition, a sequence of both buried
and relict soils exists. Because most soil and geomorphic
processes are space and time transgressive, the majority of
buried soil-landscapes have originated in this way. Timetransgressive chronosequences without historic overlap are
represented by a vertical sequence of soil-landscapes such as

those occurring between successive depositional units. Because an unknown amount of potential soil-formation time is
taken up by deposition, it is difficult to construct plots of soil
properties versus time with these situations (Birkeland, 1990).
Dating deposits that make up the chronosequence is commonly a major problem, as numerical dates are few. Most
dates are age estimates based on chemical, biological, geomorphic, and correlation dating methods.
Birkeland (1990) pointed out that there may be limits
on the maximum development of a chronosequence due to
landform development with time. He cites the example
(Peterson, 1981) of sequential development of stream terrace
morphology that begins as a flat surface ideal for soil formation, but is subsequently eroded. As the drainage network
develops, soils are subject to erosion along drainage channels
and terrace edges. Eventually, the terrace surface is transformed
into a series of linear ridges. At this stage, all original surfaces
have been obliterated, and because soils are less developed
than those on younger, adjacent flat surfaces the direction of
soil development is apparently reversed.
Progressive pedogenesis is nicely exemplified by the development of soils on progressively older cirque deposits in
the Colorado Front Range (Dixon, 1983; Birkeland et al.,
1987). Here, soils have developed on relatively young Holocene deposits ranging in age from 100 to 12 000 years. The
soils display strongly contrasting profile characteristics with
increasing age of the deposits. Soil development trends with
age include increasing organic carbon, total nitrogen, organicbound phosphorous, clay, extractable iron and aluminum,
and higher cation exchange capacity. Soil horizon development also displays strong time-related trends. Post Gannett
Peak soils (100350 years) display A/Cox horizon sequences,
post-Audubon (9502400 years) soils display A/AC/Cox
horizons, post Tripple Lakes (30005200 years) soils display
A/Bw/Cox horizon sequences, and post Satanta Peak
(10 00012 000 years) soils display A/Bw or Bt/Cox profiles.
The two youngest soils are classified as Pergelic Cryumbrepts,
the post Tripple Lakes soil is a Pergelic Cryumbrept or a Pergelic Cryoboroll, and the post Satanta Peak soil is a Pergelic
Croboroll or a Pergelic Cryoboralf.
Further elaboration of soil chronosequences and their
applications in geomorphology are discussed by Schmid in
Chapter 2.3. Geomorphic settings described therein include
marine and fluvial terraces, glacial and periglacial surfaces, fire
disturbance and vegetation change, and climate change and
carbon sequestration.
4.3.5
Soils as Indicators of Landscape Stability
Soil development is fundamentally a reflection of landscape

stability. Where landscape surfaces are unstable or become
unstable, soils are either eroded or are buried. Once landscape stability returns, then soil formation resumes. If the
unstable surface is buried, then a paleosol is preserved as
evidence of the past period of stability. Landscape stability is
exemplified by studies in Australia and in the Midwestern US.
The idea that all landscapes experience periods of stability
and instability led to the development of the K-cycle concept
(Butler, 1959) (Figure 2). The K-cycle concept proposes that

during periods of landscape instability hillslope processes
overwhelm pedogenesis, whereas during periods of stability
pedogensis dominates.
In Butlers (1959) model, a soil cycle begins with soil development on a stable ground surface and concludes when
that soil is either eroded or buried. One cycle leads immediately to the next, although a cycle need not go through to
completion. An incomplete cycle is a partial cycle because it
did not achieve full development. The soil cycle concept is
most useful when it is used as a time unit. The K-cycle is
defined as the interval of time covering the formation of the
new geomorphic surface, the period in which soils develop
on that surface, and ends when the surface becomes buried
or eroded (Butler, 1959). The chronologic order of the ground
surface is determined by its stratigraphic order and is
K1
33
designated K1, K2, K3, etc., from the uppermost, most recent,
to the consecutively lower and older land surfaces at a given
locality. Thus, a chronological framework for a given locality
can be established consisting of stable and unstable phases of
each cycle (e.g., K1s, K1u, K2s, K2u, etc.).
The K-cycle concept has been applied widely in Australia
since its development by Butler (1959). Van Dijk (1959) was
one of the first to apply it to land surface and soil studies in the
Canberra region. He recognized a series of tablelands extending
back in age to the Cretaceous that are separated by nick points
and sit above the modern drainage system. Within the landscapes, he identified three geomorphic process zones: essentially a crest slope zone of erosion, a shoulder slope zone of
erosion and deposition, and a toe-slope zone of deposition.
Associated with each of these zones was a distinctive assemblage of buried soils corresponding to numerous K cycles. One
K3
K2
K2
K1
Fine to coarse alluvium
Pebbly colluvium
C
C
Shale
C
Torrent bedding
K1
K2
K3
K4
Ground surfaces and soil horizons

Figure 2 Butlers (1959) K-cycle model illustrating spatial relationships between hillslope position and preservation of soils associated with
previous periods (cycles) of soil formation. Adapted from Butler, B.E., 1967. Soil periodicity in relation to landform development in southeastern
Australia. In: Jennings, J.N., Mabbutt, J.A. (Eds.), Landform Studies from Australia and New Guinea. Australian National University Press,
Canberra, pp. 231255.
34
of the most comprehensive applications of the K-cycle was

by Walker (1962a, 1962b) on the south coast of New South
Wales in southeastern Australia. Here, Walker identified hillsides mantled with unsorted and unbedded sediments. On the
steeper upper slopes, Walker identified the K1 ground surface
with residuals of K2 at depth. Further downslope, the K2 surface dominated with the K3 surface at depth, in the toe-slope
positions K3 dominates. Terraces of K1, K2, and K3 age occur
along the stream valleys. These terraces were dated by Walker
and determined to be 443 750 and 29 000 years in age, respectively. The sequence of ground surfaces and associated soilgeomorphic processes identified by Walker are as follows. The
current ground surface (K0) is characterized by erosion and
deposition with no active soil formation. Erosion is cutting
channels into underlying K1 and K2 land surface deposits. The
K1s cycle is characterized by soil development on depositional
and erosional surfaces. Soils are weakly developed with A/C
profiles. The K1 cycle is characterized by limited deposition of a
thin mantle over slopes and deposition of weakly sorted and
nonbedded sediment in upper stream channels. The K1u cycle
is characterized by limited truncation of K2 soils and slight
stream incision. The K2s cycle is characterized by soil development on depositional and eroded surfaces. Soils consist of
A/B/C horizon sequences suggesting that this stable cycle was
somewhat longer than the K1s cycle as reflected in the more
complex soil profile development. K2 is characterized by deposition of a relatively uniform mantle of debris over eroded
hillsides with thick deposits in upper stream channels. The
sediments are poorly sorted and nonbedded. In the K2u cycle,
truncation of K3 soils occurs and stream incision is observed.
Cycle K3s is characterized by soil development on depositional
surfaces with the development of strongly differentiated profiles. Cycle K3 is characterized by the deposition of a thick
mantle of poorly sorted, nonbedded sediment over eroded
hillsides with thick deposits in upper stream channels. The K3u
cycle involves the complete truncation of pre-K3 soils to bedrock and very deep stream incision. Walker attributed these
three alternations between erosion/deposition and soil formation to climate change cycles.
Butler (1950, 1958) developed the concept of the K-cycle
as a result of detailed studies of soil/landscape relationships
on the Riverine Plain. The Riverine Plain is a vast depositional
plain in southeastern Australia formed by the deposition of
alluvium by the precursors of the present Murray, Murrumbidgee, and Goulburn Rivers (Butler, 1961; Pels, 1964, 1966;
Schumm, 1968; Langford-Smith, 1960). Four principal K
cycles are recognized in the regional soils and in individual
dunes and associated swales across the plain. More recently,
Bowler (1978) identified three major fluvial complexes on the
basis of numerous stratigraphic criteria, but exhaustively
relying on datable buried soils. In a more recent synthesis,
Butler (1967) argued that there were consistent trends in soil
development in the similarly numbered ground surfaces
across the region encompassed by the studies of Butler (1950,
1958, 1959), Van Dijk (1959), and Walker (1962a, b).
On a much longer timescale, periods of landscape stability
and associated land surfaces and soils have been identified in
the contemporary lateritized landscapes of the Northern Territory and southwest Western Australia (Mulcahy, 1967) on the
Australian continent. In the Northern Territory, Hays (1967)
identified four spatially expansive erosion surfaces: Ashburton,

Tennant Creek, Wave Hill, and Koolpinyah surfaces, which
represented surfaces extending back to the mid-Cretaceous.
Each of these surfaces was identified by the development of
distinctive laterite profiles. Obviously, as the land surfaces were
all exposed, the lateritic profiles were all time transgressive. The
oldest, Ashburton surface is characterized by the presence of a
remnant lateritic profile as much of the original surface has
been eroded. The Tenant Creek surface is characterized by the
presence of a standard lateritic profile consisting of a laterite
surface horizon underlain successively by a mottled and pallid
horizon or zone over the parent bedrock. The Wave Hill
surface is characterized by the presence of detrital laterite on
truncated standard laterite profiles. The lowest and youngest
Koolpinyah surface is characterized by detrital laterite on
detrital laterite profiles. A more recent study from the same
area (Williams, 1978), however, argued that these surfaces were
still undergoing pedogenic modification, and that landscape
evolution in the region must be evaluated in terms of present
day erosional, pedogenic, and zoogenic processes as well as
inheritance.
4.3.5.1
Landscape Evolution in the Midwestern US
Soils are generally interpreted to represent the location of

former land surfaces in both space and time. They represent
periods of stability on land surfaces that are long enough for
soil to develop in new deposits on landscapes or on newly
evolved landscapes. Ruhe (1956) identified the relationship
between geomorphic (landscape) surfaces and the nature of
soils. He then subsequently applied this relationship to an
understanding of landscape evolution in the midwestern US
(Ruhe, 1969) and elsewhere (1967). One of Ruhes major
contributions was the identification and genetic interpretation
of the erosion surfaces of southern Iowa. Using soil properties
and landscape position, he pieced together the geomorphic
history of the Late Sangamon erosion surface and established
the chronology of the constructional and erosional events.
More recently, Follmer (1982) examined the geomorphology
of the Sangamon surface in Illinois. Follmer was able to elucidate the Sangamon landscape surface using a combination
of soil properties and correlation studies of parent materials.
Using this approach, Follmer found that compared to the
modern land surface, the Sangamon displayed greater relief
and a greater abundance of better-drained and very poorly
drained soils. The Sangamon surface was dominated by soils
developed in Illinoian till and ice contact deposits. Only local
erosion and deposition occurred on flat upland portions of
the landscape. The principal geomorphic process that operated
on the Sangamon surface and continues to the present is
stream headward erosion and entrenchment. This is focused
on very poorly drained, accretionary soils that formed in depressions on the original Sangamon surface.
The evolution of the Iowan erosion surface has long been a
topic of interest to soil geomorphologists in the Midwest
(Olson, 1989). It was once thought to have developed during
a separate ice advance (Olson, 1989), but is now interpreted
to be the result of long-term erosion of a glaciated landscape
(Schaetzl and Anderson, 2005). Elucidation of the evolution
of this land surface involved the integration of studies of

remnant soils and associated stone lines, as well as the presence of relict ice wedges to elucidate a landscape affected
by periglacial processes and associated mass movements and
slope wash.
4.3.6
Soils and Climate Change
Climate is amongst the factors controlling soil formation

(Jenny, 1941). Its influence is primarily driven by moisture
and temperature components that affect the rates of decomposition of organic and inorganic material, moisture
availability, and patterns and processes of organic activity.
Soils, therefore, integrate climate signals in their fundamental
physical, chemical, and organic properties (Birkeland, 1999).
Further examples of soil and climate change pertaining to
more recent, human-induced environmental change are discussed by Schmid (Chapter 2.3).
Numerous studies using soil properties as indicators of
climate change have been undertaken in the Rocky Mountains.
Benedict (1981) identified a series of five different parent
materials in the Fourth of July Valley, which he attributed to
substantial changes in prevailing climates. These parent material changes reflected the deposition of glacial tills, formation and subsequent reworking of periglacial deposits, and
fluvial and lacustrine deposition. Benedict (1970) also demonstrated how soils may be used to decipher hillslope processes and changes in dominant processes. He identified the
association of turf-banked lobes and terraces with solifluction
and stone-banked lobes and terraces with frost creep. On
Niwot Ridge in the Colorado Rocky Mountains, he observed
that stone-banked lobes and terraces commonly overlie turfbanked lobes and terraces, separated by a paleosol, and argued
that there had been a change in hillslope processes. He argued
that there had been a change in available moisture and that
solifluction had been replaced by frost creep after the middle
Pleistocene. During the Holocene glacial advances, solifluction
dominated on lower slopes where moisture was more abundant and frost creep dominated on drier upper slopes.
The depth of accumulation of soluble salts and patterns of
clay mineral distribution with depth are frequently used to
determine the impact of climate change on soils and concurrently the signs of climate change within soils (Karlstrom,
1988; Birkeland, 1990). Reider (1983) and Reider et al. (1988)
used soil grain size distributions, clay mineralogy, and carbonate contents to identify the presence of paleosols in the
Medicine Bow Mountains and the Absaroka Mountains of
Wyoming. From these paleosols, they suggested that the region experienced a shift from a semiarid climate with associated grassland vegetation, in which calcium carbonate
accumulated, to humid climates supporting forest cover.
Similarly, Reheis (1987a, b) employed a variety of soil properties to identify alternating periods of wet glacial and dry
interglacial climates in south central Montana. She identified
wet glacial periods by the presence of illuvial clays in the lower
parts of B horizons. Dry interglacial periods were identified
by accumulations of calcium carbonate and gypsum below
the present depth of water infiltration. Clay mineral distribution with depth can also indicate depths of soil-moisture
35
penetration during glacial and interglacial times. Foscolas

et al. (1977) suggested that clay mineral alteration occurred to
greater depth in the past than at present in soils in Yukon
Territory due to greater moisture availability. Similar interpretations have been made from soils in the Rocky Mountains
of the US and Canada by Karlstrom (1988).
Shroba and Birkeland (1973) used clay mineral assemblages in soils in the Colorado Rocky Mountains to investigate
possible migrations of alpine tree line. They suggested that
alpine soils dominated by vermiculite and hydrobiotite may
have, in fact, formed under subalpine forest rather than under
modern tundra vegetation. Thorn et al. (1989) examined soil
development in a snowpatch hollow on Niwot Ridge in the
Colorado Front Range, and identified a strong clay mineral
contrast between surface and subsurface horizons. The surface
horizons were dominated by smectite whereas deeper horizons were dominated by vermiculite. The authors suggested
that the mineralogical hiatus corresponded to the thaw depth
of the snowpatch soils and associated greater intensity of
weathering due to greater moisture availability.
Dixon (1983, 1991) examined alpine and subalpine soils
in the Colorado Rocky Mountains, and identified a strong
pedogenic contrast between surface horizons and deeper
horizons. This study focused on soils developed on glacial and
periglacial debris in cirques above tree line. For the most part,
the parent materials of the soils consist of gravelly sandy clay
loams with silt-dominated surface horizons. In addition to
strong textural contrasts between A horizons and deeper B and
C horizons, Dixon also identified strong chemical and mineralogical contrasts. He suggested that the finer grained, less
chemically altered surface horizons represented aeolian parent
material associated with cold, dry climate regimes in contrast
to the wet, cold regimes that produced the glacial and periglacial parent materials of the deeper soil horizons. Other
workers identified similar contrasts but attributed the contrast
to frost action processes rather than changes in climate and
associated parent materials (Munn and Spackman, 1990).
Similarly, studies of soils as indicators of environmental
change have been undertaken in the arid and semiarid environments of the southwestern US. Numerous authors
identified significant differences in aeolian dust addition to
soils between the Pleistocene and the Holocene, with clear
implications for rates of soil development (Machette, 1985;
McFadden et al., 1984, 1986, 1987; Wells et al., 1985, 1987;
Chadwick and Davis, 1990; Reheis, 1987a, b; Bull, 1991).
Maximum carbonate accumulation in soil appears to have
occurred at the Pleistocene/Holocene boundary, and was accompanied by a reduction in the depth of carbonate infiltration. The polygenetic nature of arid soils has been widely
reported (Reheis, 1987a, b; Nettleton et al., 1989; Bull, 1991).
The intimate relationship between aeolian dust soil formation
has been strongly demonstrated by McFadden et al. (1998).
4.3.7
4.3.7.1
Soil-Slope Relationships
The Catena Concept
The association of soil property development with slope

position was first formalized by Milne (1935, 1936) in the
36
catena concept, and subsequently amplified by Jenny (1941)

in his identification of soil forming factors. Milnes defined a
catena as a unit of mapping conveniencey., a grouping of
soils which while they fall wide apart in a natural system of
classification on account of fundamental and morphological
differences, are yet linked in their occurrence by conditions of
topography and are repeated in the same relationships to each
other wherever the same conditions are met with (Milne,
1935, p. 197). Detailed discussion of catenas and catena
processes are in Chapter 4.9.
The importance of catenary studies in soil geomorphology
is that they link soil forming processes and landscape-forming
processes (Gerrard, 1992). Although topography has long
been recognized as a significant component of soil formation,
soil formation variability along a hillslope is also a significant
contributor to hillslope morphology. For instance, soil thickening due to slope wash and creep fills in concavities in basal
slopes, thus lessening slope angle. Variation in chemical
weathering intensity along a slope results in variability of
hillslope chemical denudation rates. The differences in soil
properties and classification along a catena are generally related to differences in their position and drainage characteristics, so that emphasis is placed on the difference between
freely drained upper parts of slopes and imperfectly to poorly
drained lower portions (Aandahl, 1948; Tonkin et al., 1977;
Young et al., 1977).
4.3.7.2
Catenas in Different Climates
The physical, chemical, and clay mineralogical characteristics

of soil catenary relationships vary depending on climate
(Ollier, 1973; Birkeland, 1984). These variations result primarily from the variations in the availability of moisture in the
worlds major climate zones and hence the mobility of
chemical constituents. For this reason, it is appropriate to

consider catenas developing under different climatic regimes.
Birkeland (1984) reported that on Baffin Island in the east
Canadian Arctic, soils developed on well-drained topographic
highs, such as moraine crests, were typically well drained and
oxidized, whereas soils in lower parts of the landscape were
typically more reduced. This change he attributed to deeper
permafrost and better-drained conditions on moraine crests
with progressively greater amounts of tundra vegetation
downslope and shallower permafrost with moisture associated
environments. Similar relationships were found in subarctic
and alpine environments (Darmody et al., 2005). In subarctic
Norway, Glazovskaya (1968) demonstrated progressive thickening of A and Bh horizons from summit/shoulder slope to
footslope, as well as increases in humus, fulvic/humic acid,
Fe2O3, and Al2O3 percentages in the Bh horizon.
Dixon (1986b) examined two catenas developed on glacial
tills in the Front Range of the Rocky Mountains of Colorado,
and found an increase in silt and clay in footslope/toe-slope
positions compared to higher topographic positions, and an
increase in oxide molar ratios in B2 horizons in toe slope
compared to backslope/midslope positions. However, there
appeared to be limited removal of metal oxides from crestlope
positions. Secondary clay mineral formation was found to
be most intense in the toe-slope position compared to
higher topographic positions. Also in Colorado, Litaor (1992)
found similar patterns of aluminum migration. At the regional
scale, Burns and Tonkin (1982) developed a Synthetic
Alpine Slope (SAS) Model (Figure 3) to account for variations
in soil properties across the alpine of the Colorado Rocky
Mountains. Soil variability in the model fundamentally reflected spatial variation in snowcover duration (moisture) and
loess thickness. The model identified seven distinct alpine
soil classes based on 21 soil properties. Overall, the soil development sequence (from high to low rank) was minimal
Extremely windblown (EWB)

Windblown (WB)
Minimal snow cover (MSC)
Early melting snowbank (EMS)

Late-melting snowbank (LMS)
Semipermanent snowbank (SPS)
Wet meadow (WM)

Figure 3 Synthetic Alpine Slope Model (Burns, 1980) illustrating principal environmental sites. Dominant soils at each site are as follows: EWB,
Dystric cryochrepts; WB, Dystric cryochrepts; MSC, Pergelic cryumbrepts; EMS, Typic cryumbrepts; LMS, Dystrict cryochrept; SPS, LIthic
cryorthent at headwalls and Pergelic Cryochrept in hollows; and WM, Pergelic Cryaquept. Modified from Burns, S.F., 1980. Alpine soil distribution
and development, Indian Peaks, Colorado Front Range. Ph.D. Dissertation. University of Colorado, Boulder.
37
Table 1 Variables used to develop the Synthetic Alpine Slope Model

Characteristics of surface horizon
Characteristics of below A horizon

8.
9.
10.
()
()
()
11.
()
12.
()
13.
14.
()
()
1.
()
Depth of surface horizon to chroma 43 cm

Depth of Loess (cm)
Percent clay
2.
()
Percent silt
3.
4.
()
()
Percent clay plus silt

Percent organic matter exchange acidity (meq/100 g)
Exchange acidity (meq/100 g)
Calcium plus magnesium (meq/100 g)
Cation exchange capacity (meq/100g)
pH-water
Free iron (% metal)
Free aluminum (% metal)

5.
6.
7.

()
()
()

Maximum percent clay

Maximum percent silt
max:% clay A
Ratio
max:% clay B
max:% silt A
Ratio
max:% silt B
Characteristics of whole profile-weighted mean
Percent clay plus percent silt
Cation exchange capacity (meq/100 q)
pH-water
Free iron (% metal)
Free aluminum (% metal)
Variables 114 and those marked with an represent Burns original 21 variables. Modified variables 114 were those used in the study by Thorn et al. (1989). Variables expected
to increase in value as soil develops are indicated with a ( ). Variables expected to decrease in value as soil develops are indicated with a ( ). Modified from Thorn, C.E., Dixon,
J.C., Darmody, R.G., Rissing, J.M., 1989. Weathering trends in fine debris beneath a snowpatch, Niwot Ridge, Front Range, Colorado. Physical Geography 10, 307321.
snow cover (MSC)4wet meadow (WM)4windblown

(WB) early melting snowbank (EMS)4extremely windblown (EWB)4late-melting snowbank (LMS)4semipermanent snowbank (SPS).
Thorn et al. (1989) tested the SAS model at a smaller
snowpatch site on Niwot Ridge in the Colorado Front Range
and showed that using 14 of Burns and Tonkins 21 soil
property measures (Table 1), the developmental sequence
(from high to low) of seven pits down the slope was WM,
EMS, MSC, LMS, and SPS. In general, the snowpatch site
supported the regionally developed SAS model, but with some
localized deviations as is to be expected as a result of scale
linkage considerations.
McKeague (1965) examined a catena in eastern Canada
dominated by Spodosols, where water table fluctuations were
a significant component of soil formation. Cryorthods occurred in upper slope positions in oxidizing conditions where
the water table was generally deep year-round. In the toe-slope
position in fluctuating redox conditions where the water
table was high, the soils were predominantly Cryaquents. Iron
abundances doubled from summit to footslope positions and
secondary clay mineral assemblages changed from smectite
and chlorite-vermiculite in upper slope positions to vermiculite, smectite, and mixed layer clays in the toe slope.
Soil catenas in warm humid environments typically display
similar trends (Hussain and Swindale, 1974; Kantor and
Schwertmann, 1974; Nye, 1954; Radwanski and Ollier, 1959;
Watson, 1964a, b; Nettleton et al., 1968). Soils in higher
topographic/slope positions are oxidized and well drained
whereas those in lower slope positions are mottled to gleyed
and contain iron and manganese concretions. Soils in higher
slope positions are dominated by kaolinite whereas those in
lower topographic positions are dominated by smectite. In a
study of clay mineral variation on granitic hillslopes, Tardy
et al. (1973) showed the development of gibbsite and residual
iron oxyhydroxides on upper slopes and gibbsite with
most elements removed on lower slopes in humid tropical
environments, and kaolinite in footslopes under slightly drier
conditions.
Muhs (1982) described a catena from San Clemente Island

in southern California and summarized several other studies
from Mediterranean climates including mainland southern
California (Nettleton et al., 1970), the Golan Heights (Dan
and Singer, 1973), southern Lebanon (Verheye, 1974), and
Andalusia, Spain (Taboadela, 1953). Based on these studies,
Muhs developed a conceptual model of catena development in
Mediterranean climates that involved illuviation and in-situ
grain replacement by clay on crest and shoulder soils. In
addition, the shoulder soil undergoes overland flow and
throughflow. Pedoturbation dominates toe-slope soils. Aeolian
dustfall complicates soil development in all slope positions.
In humid tropical environments with a marked dry season,
Tardy et al. (1973) identified upper slopes on granitic catenas
being dominated by kaolinite and lower slopes dominated by
kaolinite in association with iron oxyhydroxides. In more arid
settings, the upper slopes were dominated by kaolinite with
lower slopes being dominated by montmorillonite in association with Si and Mg accumulation.
In arid environments such as those of the Negev Desert of
Israel, catenas display patterns of soil development that relate
to patterns of water movement laterally and vertically. Soils
developed on upper slopes are typically rocky and thin with
limited weathering (Kadmon et al., 1989). Soils on upper
slopes also commonly display accumulation of salts and carbonates. Soils on lower slopes are typically finer grained due
to slope wash from higher topographic positions. Soils in
footslope and toe-slope positions are typically thick, complex,
and display the development of calcic and gypsic horizons
(Wieder et al., 1985). Soils developed on the piedmont beyond the base of the colluvial/alluvial slopes are typically thin
and dominated by gypsum.
On the coastal plain of Israel, Dan et al. (1968) reported
on a catena developed on sand dunes with substantial aeolian
additions. Here, soils developed on upper slopes were typically relatively thin compared to lower slopes, display translocation of loess, forming textural B horizons. The soils
displayed the development of eluvial horizons with low pH.
Soils in midslope positions were typically the least strongly
38
developed as runoff was typically generated on this part of the

catena. Soils at the base of the slope were thick as a result of
slope wash and loess transportation from higher slopes. Soils
were also typically clay-rich in their upper parts but sandy at
depth. Clay mineralogy in these locations was dominated by
smectites due to poor drainage. All of the soils lacked accumulation of salts or carbonates due to winter wet season
leaching.
Catenas developed on granitic slopes in tropical arid environments typically display a concentration of kaolinite on
upper slopes and montmorillonite on lower slopes in association with calcium carbonate accumulation. With increasing
aridity, montmorillonite dominates in association with carbonates whereas zeolites dominate lower slopes in association
with sodium salts (Tardy et al., 1973).
4.3.8
Hillslope/Soil Process Interaction
The preceding discussion focuses on the impact of slope

position on soil properties with obvious implications for
feedback effects on water movement on hillslopes. In addition, there are strong interactions between geomorphic process and hillslope form with accompanying impacts on soil
development. These interactions are most effectively displayed
in the Nine-Unit Slope Model developed by Conacher and
Dalrymple (1977) and Dalrymple et al. (1968) (Figure 4).
The model consists of nine hillslope components, each with a
distinctive, associated set of geomorphic and pedogenic
processes:
Hillslope Unit 1 is the interfluve and is dominated by

pedogenic processes with dominantly vertical movement
of water and in-situ soil formation. Geomorphically, this is
a very stable part of the landscape with slope angles of
o11.
Unit 2 is a seepage slope with a slope angle of 241.
Geomorphically, this slope is dominated by mechanical
and chemical eluviations with lateral subsurface movement
of soil water. Pedogenically, this unit is characterized by
gleying over iron pans or other indurations. Reduced
porosity and increased compaction in the E compared to
the B horizon and in the upper compared to the lower B
horizons. Mottling also develops with associated manganese concentration and concretions.
Unit 3 is a convex creep slope with slope angles up to 451.
It is dominated by soil creep with associated development
of terracettes associated with subsurface soil-water movement. Pedogenic processes are characterized by the transportation of soil material from upslope by creep and
probably surface and subsurface water movement. Soils are
better-drained soils than on the unit above.
Unit 4 is the Fall Face with slope angles from 45 to 4651.
Geomorphically, this unit is dominated by falls and slides
and both physical and chemical weathering. Pedogenically,
the unit is characterized by shallow soils on bedrock, and
are typically Entisols.
Unit 5 constitutes the transportational midslope occurring
at angles between 26 and 351. The unit is dominated by the
transportation of material by mass movements including
flows, slides, slumps, and creep. In addition, there is terracette formation and the surface and subsurface movement of water. Pedogenically, the slope unit is divided into
two sections. The upper portion of the slope unit where
wash and creep dominate, A horizons are of relatively
consistent thickness, not varying by more than 10% in a
downslope direction. In the lower part of the slope where
the other processes of mass wasting dominate, soil thickness variability is much greater.
Unit 6 constitutes a colluvial toe slope. Here, the dominant
geomorphic processes involve redeposition of material
from unit 5 by a variety of mass movement processes and
some surface wash resulting in fan formation. In addition,
material is transported across the slope by creep and subsurface wash. Pedogenically, the unit is characterized by a
heterogeneous soil mantle containing additions from
upslope. Buried soils are common.
Unit 7 constitutes the alluvial toe slope with a slope of
041. The unit is dominated by alluvial deposition and
processes resulting from subsurface water movement.
Pedogenic processes are dominated by poor drainage and
the formation of thick A horizons.
Unit 8 constitutes the drainage channel wall and is dominated by corrosion, slumping, and falls. Pedogenesis is
limited to Entisol formation.
Unit 9 is the stream channel bed and is dominated by
fluvial transport of material downvalley by surface water
with periodic aggregation and corrosion. Overall, there is
an absence of soil formation in this slope unit.
4.3.9
4.3.9.1
Soils and Sedimentation

Aeolian Sedimentation
Aeolian dust, consisting of carbonates, clays, and salts, has

been shown to play a crucial role in the development of desert
soils as well as associated desert pavements in arid environments. Mabbutt (1977, 1979) first proposed that desert pavement development in Australia formed as a result of upward
sorting above a fine textured aeolian mantle. He suggested that
the rough desert surface trapped aeolian dust and thus dust
was washed through the surface gravels to progressively accumulate and lift the pavement. More recently, McFadden et al.
(1987), Wells et al. (1995), and Haff and Werner (1996) suggested a similar mechanism for the formation of pavements in
the Cima volcanic field of California. Their model envisioned
generation of coarse clasts by mechanical weathering of basalt
on uplands with subsequent downslope movement. The
transported coarse debris was then raised in place as a result of
aeolian infall and subsequent raising of the pavement as a
result of loess infiltration. Intimately associated with the loess
infiltration is the development of a soil vesicular horizon.
McFadden et al. (1998) and Anderson et al. (2002) proposed a three-stage model of pavement development as a
result of cumulic pedogenesis. Stage 1 involves the formation
of a laminar crust adhered to the bottom of surface clasts. Stage
2 involves the formation of a vesicular horizon associated with
aeolian infall and associated clast lifting above the crust. Finally, Stage 3 involves the development of strong soil structure
Additions of alluvium. Generally

poorly drained with deep A horizons
Entisols
Absence of soil formation
Corrasion, slumping and fall
Transportation of material
downvalley by streams,
periodic aggradation and
corrasion
Heterogeneous soil mantle

containing additions from upslope
Where other processes of mass

movement dominate, contrasting
areas of deep and shallow soils
Where wash and creep dominate,

A horizon does not differ in
thickness by >10% and does not
thicken much in a downslope direction
Soils shallow to bedrock, entisols
39
3
4
Colluvial redeposition by mass

movement and surface water
action transportation processes
resulting from subsurface soil
water movement
Alluvial redeposition processes

resulting from subsurface soil
and ground water movement
Transportation of material by rapid

mass movement and/or surface water
action processes resulting from
subsurface soil water movement
Falls and slides; chemical and physical weathering
Soil creep processes resulting from subsurface soil water movement
Mechanical and chemical eluviation by

lateral subsurface soil water movement
Vertical pedogenetic processes
Gleying above iron pans. Reduced

porosity and increased compaction in E
as compared with B horizons and in the
upper as compared with lower parts
of B horizons. Motting, Mn
concentrations and concretions
Substitution of soil material from
upslope by creep and probably surface
and subsurface water movement.
Better drained than upslope unit
Soil development in situ
Figure 4 Nine-unit slope model illustrating relationship between hillslope position and dominant geomorphic and pedogenic processes.
Dominant geomorphic processes are indicated along the bottom, and pedogenic processes are summarized at the top of the figure. Adapted from
Dalrymple, J.B., Blong, R.J., Conacher, A.J. 1968. A hypothetical nine unit landsurface model. Zeitschrift fur Geomorphologie 12, 6076, and
Conacher, A.J., Dalrymple, J.B., 1977. The nine unit landsurface model: an approach to pedogeomorphic research. Geoderma 18, 1154.
that permits the thickening and further development of the

vesicular horizon.
Closely associated with desert pavement and stony deserts,
as well as playa surfaces, is the development of gilgai. Gilgai
refers to a group of patterned groundforms widely developed
in inland Australia and in more limited settings on other
continents in association with strong texture-contrast soils
with clay-rich subsoils. A variety of formational patterns are
identified: normal, melon-hole, stony, lattice, linear, and tank
(Ollier, 1966; Mabbutt, 1977; Hubble et al., 1983). These
forms are attributed to four principal formational processes:
heave between cracks, heave over cracks, contraction over
cracks, and heave due to loading (Knight, 1980). In all cases,
gilgai formation is related to horizontal stresses developed in
clay-rich soil materials.
4.3.9.2
Fluvial Sedimentation
Detailed studies by Walker in the coastal valleys of southeastern Australia (Walker 1962a, b; Walker and Green, 1976)
revealed similar patterns of alluvial sedimentation and pedogenic stages at a regional scale. Soils developed on low alluvial
benches (channel bars) represent a stratic stage in which soil
horizonation is minimal and soils are dominated by unaltered
alluvium with prominent sedimentary structures. Cumulic
soils dominate floodplain surfaces in which depositional features are weakly expressed and soil horizonation is minimal.
The soils display strong organic coloration, with extensive
evidence of bioturbation. Soils developed on low river terraces
display strongly expressed A horizons with gradational changes to either weak color B or textural B horizons. High terrace
40
soils display one of the two stages of development. The first of

these is referred to as a high-contrast solum stage. They display
strongly differentiated A and B horizons. Cation saturation is
low, with Mg and Na ions most abundant. Clay mineralogy is
dominated by kaolinite. The second stage observed on high
terraces is an extended solum stage. These soils display strong
weathering beneath the solum, including thick illuvial clay
and silicified and ferruginous zones. The solum properties are
similar on all high terraces (Walker and Coventry, 1976;
Walker and Green, 1976).
4.3.10
Conclusions
Soil geomorphology is the study of the interactions between

pedogenic and geomorphological processes for the purpose of
understanding landscape controls on soil formation and soil
formation as a reflection of landscape processes. The principal
components of this interactive study consist of the study
of soils as indicators of climate and environmental change;
soils as indicators of landscape stability and instability; soil/
hillslope interactions as reflections of the pattern of movement of water on the landscape; soils as indicators of depositional processes; and soils as indicators of Quaternary
history and parent materials. A complete understanding of the
soil formation cannot be obtained without an understanding
of the role that relief plays in the movement of water, materials, and chemical elements. Similarly, soils serve as a record
of geomorphological processes such as erosion, transportation, and deposition in the landscape in space and time.
Without this record, the history of landscape evolution would
be generally incomplete. Gerrard (1990) suggested that the
challenging task before soil geomorphologists is to integrate
the principal components of soil/landscape interactions discussed in this chapter.
References
Aandahl, A.R., 1948. The characterization of slope positions and their influence on
the total nitrogen content of a few virgin soils of western Iowa. Soil Science
Society of America Proceedings 13, 449454.
Aleva, G.J.J., 1994. Laterites: Concepts, Geology, Morphology and Chemistry.
Wageningen (Netherlands). International Soil Reference and Information Center.
Alley, N.F., Clarke, J.D.A., MacPhail, M., Truswell, E.M., 1999. Sedimentary infillings
and development of major Tertiary palaeodrainage systems of south-central
Australia. In: Thiry, M., Simon-Coincon, R. (Eds.), Palaeoweathering,
Palaeosurfaces and Related Continental Deposits, International Association of
Sedimentologists Special Publication 27. Blackwell Science, Oxford, pp. 337366.
Amit, R., Gerson, R., 1986. The evolution of Holocene reg (gravelly) soils in
deserts an example from the Dead Sea region. Catena 13, 5979.
Anderson, K., Wells, S., Graham, R., 2002. Pedogenesis of vesicular horizons, Cima
volcanic field, Mojave Desert, California. Soil Science Society of America
Journal 66, 878887.
Benedict, J.B., 1970. Downslope soil movement in a Colorado alpine region: rates,
processes, and climatic significance. Arctic and Alpine Research 2, 165226.
Benedict, J.B., 1981. The Fourth of July Valley: glacial geology and archeology of
the timberline ecotone. Research Report 2. Center for Mountain Archeology,
Ward, Colorado.
Birkeland, P.W., 1974. Pedology. Weathering, and Geomorphological Research.
Oxford University Press, New York.
Birkeland, P.W., 1984. Soils and Geomorphology. Oxford University Press, New York.
Birkeland, P.W., 1990. Soil geomorphic research a selected overview.
Birkeland, P.W., 1999. Soils and Geomorphology, Third ed. Oxford University Press,
New York.
Birkeland, P.W., Burke, R.M., Shroba, R.R., 1987. Holocene Alpine Soils in Gneissic
Cirque Deposits, Colorado Front Range. U.S. Government Printing Office,
Washington, DC, U.S. Geological Survey Bulletin, 1590E.
Boardman, J. (Ed.), 1985. Soils and Quaternary Landscape Evolution. John Wiley
and Sons, Chichester.
Bockheim, J.G., 1980. Solution and the use of chronofunctions in studying soil
development. Geoderma 24, 7185.
Bourman, R.P., 1993. Modes of ferricrete genesis: evidence from southeastern
Australia. Zeitschrift fur Geomorphologie 37, 77101.
Bowler, J.W., 1978. Quaternary climate and tectonics in the evolution of the
Riverine Plain, southeast Australia. In: Davies, J.L., Williams, M.A.J. (Eds.),
Landform Evolution in Australasia. Australian National University Press,
Bull, W.B., 1991. Geomorphic Responses to Climate Change. Oxford University
Press, New York.
Buol, S.W., Hole, F.D., McCracken, R.J., 1997. Soil Genesis and Classification,
fourth ed. Iowa State University Press, Ames.
Burns, S.F., 1980. Alpine soil distribution and development, Indian Peaks, Colorado
Front Range. Ph.D. Dissertation. University of Colorado, Boulder.
Burns, S.F., Tonkin, P.J., 1982. Soil-geomorphic models and the spatial distribution
and development of alpine soils. In: Thorn, C.E. (Ed.), Space and Time in
Geomorphology. Allen and Unwin, London, pp. 2543.
Butler, B.E., 1950. A theory of prior streams as a causal factor of soil occurrence in
the Riverine Plain of southeastern Australia. Australian Journal of Agricultural
Research 1, 231252.
Butler, B.E. 1958. Depositional Systems of the Riverine Plain of South Eastern
Australia in Relation to Soils. CSIRO Australia. Soil Publication No. 10.
Butler, B.E. 1959. Periodic Phenomena in Landscapes as a Basis for Soil Studies.
CSIRO Australia. Soil Publication No. 14.
Butler, B.E., 1961. Groundsurfaces and the history of the Riverine Plain. Australian
Journal of Science 24, 139.
Butler, B.E., 1967. Soil periodicity in relation to landform development in
southeastern Australia. In: Jennings, J.N., Mabbutt, J.A. (Eds.), Landform
Studies from Australia and New Guinea. Australian National University Press,
Catt, J.A., 1986. Soils and Quaternary Geology: A Handbook for Field Scientists.
Clarenden Press, Oxford.
Chadwick, O.A., Davis, J.O., 1990. Soil forming intervals caused by eolian sediment
pulses in the Lahonton Basin, northwestern Nevada. Geology 18, 243246.
Chen, X.Y., 1997. Pedogenic gypcrete formation in central Australia. Geoderma 77,
3961.
Chen, X.Y., Bowler, J.M., Magee, J.W., 1991. Gypsum ground: a new occurrence of
gypsum sediment in playas of central Australia. Sedimentary Geology 72,
7995.
Colson, J., Cojan, I., 1996. Groundwater dolocretes in a lake-marginal environment:
an alternative model for dolocrete formation in continental settings (Danian of
the Provence Basin, France). Sedimentology 43, 175188.
Conacher, A.J., Dalrymple, J.B., 1977. The nine unit landsurface model: an
approach to pedogeomorphic research. Geoderma 18, 1154.
Dalrymple, J.B., Blong, R.J., Conacher, A.J., 1968. A hypothetical nine unit
landsurface model. Zeitschrift fur Geomorphologie 12, 6076.
Dan, J., Singer, A., 1973. Soil evolution on basalt and basic pyroclastic materials in
the Golan Heights. Geoderma 9, 165192.
Dan, J., Yaalon, D.H., Koyumdjisky, H., 1968. Catenary soil relationships in
Israel. 1. The Netanya catena on coastal dunes of the Sharon. Geoderma 2,
95120.
Dan, J., Yaalon, D.H., Moshe, R., Nissim, S., 1982. Evolution of reg soils in
southern Israel and Sinai. Geoderma 28, 173202.
Daniels, R.B., Gamble, E.E., Nelsom, L.A., 1971. Relations between soil morphology
and water-table levels on a North Carolina Coastal Plain surface. Proceedings of
the Soil Science Society of America 35, 781784.
Darmody, R.G., Allen, C.E., Thorn, C.E., 2005. Soil toposequences at
Storbreen, Jotunheimen, Norway. Soil Science Society of America Journal 69,
12751287.
Dixon, J.C., 1978. Morphology and genesis of calcrete in South Australia with
special reference to the southern Murray Basin and Yorke Peninsula. M.A.
Thesis. University of Adelaide, South Australia.
Dixon, J.C., 1983. Chemical weathering of late Quaternary cirque deposits in the
Colorado front range. Ph.D. Dissertation. University of Colorado, Boulder.
Dixon, J.C., 1986a. Soil geomorphology as an integrative approach in physical
geography: a review of Four Recent Books. Physical Geography 7, 334343.
Dixon, J.C., 1986b. Solute movement on hillslopes in the alpine environment of the
Colorado Front Range. In: Abrahams, A.D. (Ed.), Hillslope Processes. Allen and
Unwin, London, pp. 139159.
Dixon, J.C., 1991. Alpine and subalpine soil properties as paleoenvironmental
indicators. Physical Geography 12, 370384.
Dixon, J.C., 2010. Origin of calcrete and dolocrete in the carbonate mantle of St.
Vincent Basin, southern South Australia. Cadernos Laboratoir Xeoloxico de Laxe,
109122
Dixon, J.C., McLaren, S.J., 2009. Duricrusts. In: Parsons, A.J., Abrahams, A.D.
(Eds.), Geomorphology of Desert Environments. Springer, London, pp. 123152.
Follmer, L.R., 1982. The geomorphology of the Sangamon surface: its spatial and
temporal attributes. In: Thorn, C.E. (Ed.), Space and Time in Geomorphology.
Allen and Unwin, London, pp. 117146.
Foscolas, A.E., Rutter, N.W., Hughes, O.L., 1977. The use of pedological studies in
interpreting the Quaternary history of central Yukon Territory. Geological Survey
of Canada Bulletin 271. 48 pp.
Frankel, J.J., Kent, L.E., 1938. Grahamstown surface quartzites (silcrete).
Transactions of the Geological Society of South Africa 15, 142.
Gerrard, A.J., 1981. Soils and Landforms: An Integration of Geomorphology and
Pedology. Allen and Unwin, London, 219 pp.
Gerrard, A.J., 1990. Soil variations on hillslopes in huimid temperate climates.
Gerrard, J., 1992. Soil Geomorphology: An Integration of Pedology and
Geomorphology. Chapman and Hall, London, 269 pp.
Gerrard, J., 1993. Soil geomorphology present dilemmas and future challenges.
Gile, L.H., Peterson, F.F., Grossman, R.B., 1966. Morphological and genetic
sequences of carbonate accumulation in desert soils. Soil Science 101,
347360.
Glazovskaya, M.A., 1968. Geochemical landscapes and types of geochemical soil
sequences. Transactions of the International Congress of Soil Science 4,
303312.
Goudie, A., 1973. Duricrusts in Tropical and Subtropical Landscapes. Clarendon
Press, Oxford.
Goudie, A., 1983. Calcrete. In: Goudie, A.S., Pye, K. (Eds.), Chemical Sediments
and Geomorphology: Precipitates and Residua in the Near Surface Environment.
Academic Press, New York, pp. 93131.
Goudie, A.S., 1985. Duricrusts and landforms. In: Richards, K.S., Arnett, R.R., Ellis,
S. (Eds.), Geomorphology and Soils. Allen and Unwin, London, pp. 3757.
Haff, P.K., Werner, B.T., 1996. Dynamical processes on desert pavements and the
healing of surficial disturbances. Quaternary Research 45, 3846.
Hamming, E., 1968. On laterite and latosols. Professional Geographer 20, 238241.
Harden, J.W., 1982. A quantitative index of soil development from field
descriptions: examples from a chronosequence in central California. Geoderma
28, 128.
Hays, J., 1967. Land surfaces and laterites in the north of the Northern Territory.
In: Jennings, J.N., Mabbutt, J.A. (Eds.), Landform Studies from Australia and
New Guinea. Australian National University Press, Canberra, pp. 182210.
Hill, S.M., Eggleton, R.A., Taylor, G., 2003. Neotectonic disruption of silicified
paleovalley systems in an intraplate tectonic landscape: regolith and landscape
evolution of the Mulculca front-range, broken Hill Domain, New South Wales.
Australian Journal of Earth Sciences 50, 691707.
Hussain, M.S., Swindale, L.D., 1974. The physical and chemical properties of the
gray hydromorphic soils of the Hawaiian Islands. Soil Science Society of
America Proceedings 38, 935941.
Hutton, J.T., Dixon, J.C., 1981. The chemistry and mineralogy of some South
Australian calcretes and associated soft carbonates and their dolomitization.
Journal of the Geological Society of Australia 19, 3139.
Jacobson, G., Arakel, A.V., Yijian, C., 1988. The central Australian groundwater
discharge zone: evolution of associated calcrete and gypcrete deposits.
Jenny, H., 1941. Factors of Soil Formation. McGraw Hill, New York.
Johnson, D.L., Balek, C.L., 1991. The genesis of Quaternary Landscapes with
stonelines. Physical Geography 12, 385395.
Johnson, D.L., Hester, N.C., 1972. Origin of stone pavements on Pleistocene marine
terraces in California. Proceedings of the Association of American Geographers
4, 5053.
Johnson, D.L., Rockwell, T.K., 1982. Soil geomorphology: theory, concepts and
principles with examples and applications on alluvial and marine terraces in
coastal California. Geological Society of America, programs with Abstracts 14,
176.
Johnson, D.L., Watson-Stegner, D., 1987. Evolution model of pedogenesis. Soil
Science 143, 349366.
41
Johnson, D.L., Watson-Stegner, D., Johnson, D.N., 1987. Proisotropic

and proanisotropic processes of pedoturbation. Soil Science 143,
278292.
Kantor, W., Schwertmann, U., 1974. Mineralogy and genesis of clays in red-black
soil toposequences on basic igneos rocks in Kenya. Soil Science 25, 6778.
Karlstrom, E.T., 1988. Multiple paleosols in pre-Wisconsinan drift, northwestern
Montana and southwestern Alberta. Catena 15, 147178.
Khalaf, F.I., 1990. Occurrence of phreatic dolocrete within Tertiary clastic deposits of
Kuwait, Arabian Gulf. Sedimentary Geology 68, 223239.
Khalaf, F.I., 2007. Occurrences and genesis of calcrete and dolocrete in the MioPleistocene fluvial sequence in Kuwait, northeast Arabian Peninsula. Sedimentary
Geology 199, 129139.
Knight, M.J., 1980. Structural analysis and mechanical origins of gilgai at Boorook,
Victoria, Australia. Geoderma 23, 245283.
Knuepfer, P.L.K., McFadden, L.D. (Eds.), 1990. Soils and landscape evolution.
Proceedings of the 21st Binghamton Symposium in Geomorphology. Elsevier,
Amsterdam, 575 pp.
Lamplugh, G.H., 1902. Calcrete. Geological Magazine 9, 575.
Langford-Smith, T., 1960. The dead river systems of the Murrumbidgee.
Geographical Review 50, 368389.
Lee, S.Y., Gilkes, R.J., 2005. Groundwater geochemistry and composition of
hardpans in southwestern Australian regolith. Geoderma 126, 5984.
Litaor, M.I., 1992. Aluminum mobility along a geochemical catena in an alpine
watershed, Front Range, Colorado. Catena 19, 116.
Mabbutt, J.A., 1967. Denudation chronology in central Australia: Structure, climate
and landform inheritance in the Alice Springs area. In: Jennings, J.N., Mabbutt,
J.A. (Eds.), Landform Studies from Australia and New Guinea. Australian
National University Press, Canberra, pp. 144181.
Mabbutt, J.A., 1977. Desert Landforms. MIT Press, Cambridge.
Mabbutt, J.A., 1979. Pavements and patterned ground in the Australian stony
deserts. Stuttgarter Geographische Studen 93.197123.
Machette, M.N., 1985. Calcic soils of the southwestern United States. Geological
Society of America, Special Paper 203, 121.
Marker, M.E., McFarlane, M.J., Wormald, R.J., 2002. A laterite profile near
Albertinia, Southern Cape, South Africa: its significance in the evolution of the
African Surface. South African Journal of Geology 105, 6774.
McFadden, L.D., Knuepfer, P.L.K., 1990. Soil geomorphology: the link between
pedology and surficial processes. Geomorphology 3, 197205.
McFadden, L.D., McDonald, E.V., Wells, S.G., Anderson, K., Quade, J., Forman,
S.L., 1998. The vesicular layer and carbonate collars of desert soils and
pavements: formation, age, and relation to climate change. Geomorphology 24,
101145.
McFadden, L.D., Wells, S.G., Dohrenwend, J.C., 1986. Influence of climate changes
on processes of soil development on desert loess deposits of the Cima volcanic
field, California. Catena 13, 361389.
McFadden, L.D., Wells, S.G., Dohrenwend, J.C., Turrin, B.D., 1984. Cumulic soils
deformed in the eolian parent materials on flows of the late Cenozoic Cima
volcanic field, California. In: Dohrenwend, J.C. (Ed.), Surficial Geology of the
Eastern Mojave Desert, California. Field Trip Guide, 97th Annual Meeting of the
Geological Society of America, pp 134149.
McFadden, L.D., Wells, S.G., Jercinovich, M.J., 1987. Influences of eolian and
pedogenic processes on the origin and evolution of desert pavements. Geology
15, 504508.
McKeague, J.A., 1965. Properties and genesis of three members of the Uplands
catena. Canadian Journal of Soil Science 45, 6377.
McLaren, S.J., 2004. Evolution and distribution of Quaternary channel calcretes,
southern Jordan. Earth Surface Processes and Landforms 29, 14871508.
Milne, G., 1935. Some suggested units for classification and mapping, particularly
for east African soils. Soil Research 4, 183198.
Milne, G., 1936. Normal erosion as a factor in soil profile development. Nature
138, 548.
Milnes, A.R., Thiry, M., 1992. Silcrete. In: Martini, I.P., Chesworth, W. (Eds.),
Weathering, Soils, and Palaeosols. Developments in Earth Surface Processes,
vol. 2. Elsevier, Amsterdam, pp. 349377.
Milnes, A.R., Twidale, C.R., 1983. An overview of silicification in Cainozoic
landscapes of arid central and southern Australia. Australian Journal of Earth
Muhs, D.R., 1982. The influence of topography on the spatial variability of soils in
Mediterranean climates. In: Thorn, C.E. (Ed.), Space and Time in
Mulcahy, M.J., 1967. Landscapes, laterites and soils in southwestern Australia.
In: Jennings, J.N., Mabbutt, J.A. (Eds.), Landform Studies from Australia and
New Guinea. ANU Press, Canberra, pp. 211230.
42
Munn, L.C., Spackman, L.K., 1990. Origin of silt enriched alpine surface mantles
in the Indian Basin, Wyoming. Soil Science Society of America Journal 54,
16701676.
Nash, D.J., McLaren, S.J., 2003. Kalahari valley calcretes: their nature, origins and
environmental significance. Quaternary International 111, 322.
Nash, D.J., McLaren, S.J., 2007. Introduction: geochemical sediments in
landscapes. In: Nash, D.J., McLaren, S.J. (Eds.), Geochemical Sediments and
Landscapes. Blackwell Publishing, London, pp. 19.
Nash, D.J., Ullyot, S., 2007. Silcrete. In: Nash, D.J., McLaren, S.J. (Eds.),
Geochemical Sediments and Landscapes. Blackwell Publishing, London, pp.
95143.
Netterberg, F., 1969. Ages of calcretes in southern Africa. South African
Archeological Bulletin 24, 117122.
Nettleton, W.D., Daniels, R.B., McCracken, R.J., 1968. Two North Carolina coastal
plain catenas. 1. Morphology and fragipan development. Soil Science Society of
America Proceedings 32, 577582.
Nettleton, W.D., Flach, K.W., Nelson, R.E., 1970. Pedogenic weathering of tonalite in
southern California. Geoderma 4, 387402.
Nettleton, W.D., Gamble, E.E., Allen, B.L., Borst, G., 1989. Relief soils of subtropical
regions of the United States. Catena Supplement 16, 5993.
Nye, P.H., 1954. Some soil-forming processes in the humid tropics. Journal of Soil
Science 5, 721.
Ollier, C.D., 1966. Desert gilgai. Nature 212, 581583.
Ollier, C.D., 1973. Catenas in different climates. In: Derbyshire, E. (Ed.),
Geomorphology and Climate. John Wiley and Sons, New York, pp. 137169.
Ollier, C.D., Chan, R.A., Craig, M.A., Gibson, D.L., 1988. Aspects of landscape
history and regolith in the Kalgoorlie region, Western Australia. BMR Journal of
Australian Geology and Geophysics 10, 309321.
Ollier, C.D., Pain, C., 1996. Regolith. Soils and Landforms. John Wiley and Sons,
Chichester.
Olson, C.G., 1989. Soil geomorphic research and the importance of paleosol
stratigraphy to Quaternary investigations, Midwestern USA. Catena Supplement
16, 129142.
Pels, S., 1964. The present and ancesteral Murray River system. Australian
Geographical Studies 2, 111119.
Pels, S., 1966. Late Quaternary Chtonology of the Riverine Plain of southeastern
Australia. Journal of the Geological Societ of Australia 13, 2740.
Peterson, F.F., 1981. Landforms of the Basin and Range Province. Nevada
Agricultural Experiment Station, Technical Bulletin 28.
Radwanski, S.A., Ollier, C.D., 1959. A study of an east African catena. Journal of
Soil Science 10, 149168.
Reheis, M.C., 1987a. Gypsic soils of the Kane alluvial fans. US Geological Survey
Bulletin, 159-C.
Reheis, M.C., 1987b. Climate implications of alternating clay and carbonate
formation in semi-arid soils of south-central Montana. Quaternary Research 29,
270282.
Reider, R.G., 1983. A soil catena in the Medicine Bow Mountains, Wyoming, USA
with reference to paleoenvironmental influences. Arctic and Alpine Research 15,
181192.
Reider, R.G., Huckleberry, G.A., Frieson, G.C., 1988. Soil evidence for postglacial
forest-grassland fluctuations in the Absaroka Mountains of northwestern
Wyoming, USA. Arctic and Alpine Research 20, 188198.
Richards, K.S., Arnett, R.R., Ellis, S., 1985. Geomorphology and Soils. Allen and
Unwin, London.
Ruhe, R.V., 1956. Geomorphic surfaces and the nature of soils. Soil Science 82,
441455.
Ruhe, R.V., 1969. Quaternary Landscapes in Iowa. Iowa State University Press,
Ames, IA.
Ruhe, R.V., 1974. Holocene environments and soil geomorphology in Midwestern
United States. Quaternary Research 4, 487495.
Schaetzl, R., Anderson, S., 2005. Soils: Genesis and Geomorphology. Cambridge
University Press, New York.
Schumm, S.A., 1968. River adjustment to altered hydrologic regimen-Murrumbidgee
River and paleochannels, Australia. United States Geological Survey Professional
Paper 598.
Shroba, R.R., Birkeland, P.W., 1973. Trends in Late Quaternary soil development in
the Rocky Mountains and Sierra Nevada of the western United States.
In: Wright, Jr. H.E. (Ed.), Late Quaternary Environments of the United States.
Volume 1: The Late Pleistocene, pp. 145156.
Simon-Coinc- on, R., Milnes, A.R., Thiry, M., Wright, M.J., 1996. Evolution of
landscapes in northern South Australia in relation to the distribution and
formation of silcretes. Journal of the Geological Society, London 153,
467480.
Simonson, R.W., 1978. A multi-process model of soil genesis. In: Mahaney, W.C.
(Ed.), Quaternary Soils. GeoAbstracts. Norwich, UK, pp. 125.
Smale, D., 1973. Silcretes and associated silica diagenesis in southern Africa and
Australia. Journal of Sedimentary Petrology 43, 10771089.
Stevens, P.R., Walker, T.W., 1970. The chronosequence concept and soil formation.
Quarterly Review of Biology 45, 333350.
Summerfield, M.A., 1982. Distribution, nature and genesis of silcrete in arid and
semi-arid southern Africa. Catens Supplement 1, 3765.
Summerfield, M.A., 1983. Silcrete. In: Goudie, A.S., Pye, K. (Eds.), Chemical
Sediments and Geomorphology. Academic Press, London, pp. 5992.
Taboadela, M.M., 1953. The clay mineralogy of some soils from Spain and Rio
Muni (West Africa). Journal of Soil Science 4, 4955.
Tardy, T., Bocquier, G., Paquet, H., Millot, G., 1973. Formation of clay from granite
and its distribution in relation to climate and topography. Geoderma 10,
271284.
and Sons, Chichester.
Thiry, M., 1999. Diversity of continental silicification features: examples from the
Cenozoic deposits in the Paris Basin and neighbouring basement. In: Thiry, M.,
Simon-CoinBon, R. (Eds.), Palaeoweathering, Palaeosurfaces, and Related
Continentakl Deposits. International Association of Sedimentologists. Special
Publication No. 27. Blackwell Science, London, pp. 87127.
Thiry, M., Milnes, A.R., 1991. Pedogenic and groundwater silcretes at Sturt
Creek opal field, South Australia. Journal of Sedimentary Petrology 61,
111127.
Thiry, M., Milnes, A.R., Rayot, V., Simon-CoinBon, R., 2006. Interpretation of
palaeoweathering features and successive silicification in the Tertiary regolith of
inland Australia. Journal of the Geological Society, London 163, 723736.
Thorn, C.E., Dixon, J.C., Darmody, R.G., Rissing, J.M., 1989. Weathering trends in
fine debris beneath a snowpatch, Niwot Ridge, Front Range, Colorado. Physical
Tonkin, P.J., Young, A.W., McKie, D.A., Campbell, A.S., 1977. Conceptual models
of soil development and soil distribution in hill country, central South Island,
New Zealand. Part 1. The analysis of the changes in soil pattern. New Zealand
Society of Soil Science. 25th Jubilee Conference, Summary of presented papers,
pp. 2527.
Van Dijk, D.C., 1959. Soil Features in Relation to Erosional History in the Vicinity
of Canberra. CSIRO, Australia, Soil Publication No. 13.
Verheye, W., 1974. Soils and soil evolution on limestones in the Mediterranean
environment. Transactions of the 10th International Congress of Soil Science,
Moscow 6, 387393.
Vreeken, W.J., 1975. Principal kinds of chronosequences and their significance in
soil history. Journal of Soil Science 26, 378394.
Walker, P.H., 1962a. Soil layers on hillslopes: a study at Nowra NSW, Australia.
Journal of Soil Science 13, 167177.
Walker, P.H., 1962b. Terrace chronology and soil formation on the south coast of
NSW. Journal of Soil Science 13, 178186.
Walker, P.H., Coventry, R.J., 1976. Soil profile development in some alluvial
deposits of eastern New South Wales. Australian Journal of Soil Research 14,
305317.
Walker, P.H., Green, P., 1976. Soil trends in two valley fill sequences. Australian
Journal of Soil Research 14, 291303.
Watson, A., 1979. Gypsum crusts in deserts. Journal of Arid Environments 2,
320.
Watson, A., 1983. Gypsum crusts. In: Goudie, A.S., Pye, K. (Eds.), Chemical
Sediments and Geomorphology: Precipitates and Residua in the Near Surface
Environment. Academic Press, New York, pp. 133136.
Watson, A., 1985. Structure, chemistry and origins of gypsum crusts in southern
Tunisia and the central Namib Desert. Sedimentology 32, 855875.
Watson, A., 1988. Desert gypsum crusts as palaeoenvironmental indicators: a
micropetrographic study of crusts from southern Tunisia and the central Namib
Desert. Journal of Arid Environments 15, 1942.
Watson, J.P., 1964a. A soil catena on granite in Southern Rhodesia. I. Field
observations. Journal of Soil Science 15, 238250.
Watson, J.P., 1964b. A soil catena on granite II. Analytical data. Journal of Soil
Science 15, 251257.
Wells, S., Dohrenwend, J.C., McFadden, L.D., Turrin, B.D., 1985. Late Cenozoic
landscape evolution on lava flow surfaces of the Cima volcanic field,
Mojave Desert, California. Bulletin of the Geological Society of America 96,
15181529.
Wells, S.G., McFadden, L.D., Dohrenwend, J.C., 1987. Influence of late Quaternary
climatic changes on geomorphic and pedogenic processes on a desert piedmont
eastern Mojave Desert, California. Quaternary Research 27, 130146.
Wells, S.G., McFadden, L.D., Poths, J., Olinger, C.T., 1995. Cosmogenic 3He
surface exposure dating of stone pavements: implications for landscape
evolution in deserts. Geomorphology 23, 613616.
Widdowson, M., 2007. Laterite and ferricrete. In: Nash, D.J., McLaren, S.J. (Eds.),
Geochemical Sediments and Landscapes. Blackwell Publishing, Oxford, pp.
4694.
Wieder, M., Yair, A., Arzi, A., 1985. Catenary soil relationships on arid hillslopes.
Catena Supplement 6, 4157.
Wood, W.R., Johnson, D.L., 1978. A survey of disturbance processes in
archaeological site formation. In: Schiffer, M.B. (Ed.), Advances in
Archaeological Method and Theory. vol. 1. Academic Press, New York, pp.
315381.
Woolnough, W.G., 1927. The duricrust of Australia. Journal and Proceedings of the
Royal Society of New South Wales 61, 2453.
Woolnough, W.G., 1930. Influence of climate and topography in the formation
and distribution of products of weathering. Geological Magazine 67,
123132.
43
Wopfner, H., 1978. Silcretes of northern South Australia and adjacent regions.
In: Langford-Smith, T. (Ed.), Silcrete in Australia. University of New England.
Armidale, Australia, pp. 93141.
Wopfner, H., 1983. Environment of silcrete formation: a comparison of examples
from Australia and the Cologne Embayment, West Germany. In: Wilson, R.C.L.
(Ed.), Residual Deposits: Surface and Related Weathering Processes
and Materials. Geological Society of London, Special Publication 11. pp.
151158.
Wright, V.P., 2007. Calcretes. In: Nash, D.J., McLaren, S.J. (Eds.), Geochemical
Sediments and Landscapes. Blackwell Publishing, Oxford, pp. 1045.
Yaalon, D.H., 1975. Conceptual models in pedogenesis: can soil forming factors be
resolved? Geoderma 14, 189205.
Young, A.W., Tonkin, P.J., McKie, D.A., Campbell, A.S., 1977. Conceptual models
of soil development and soil distribution in hill country, central South Island,
New Zealand. Part II. Chemical and mineralogical properties. New Zealand
Society for Soil Science, 25th Jubilee Conference, Summaries of presented
papers, pp. 2830.
Biographical Sketch
John C. Dixon is a Professor in the Department of Geosciences at the University of Arkansas, Fayetteville, Arkansas
USA. His research interests lie primarily in the area of geochemical processes and landscape evolution in cold
climates. He has conducted research in Swedish Lapland over the past fifteen years, as well as in the Jotenheimen
of Norway, the Juneau Icefield of Alaska, and the Rocky Mountains of Colorado. He and his research colleagues
have published more than 70 papers in the major geomorphology journals and in specialized monographic
volumes.
4.4 Nanoscale: Mineral Weathering Boundary

RI Dorn, Arizona State University, Tempe, AZ, USA
SJ Gordon, United States Air Force Academy, Colorado Springs, CO, USA
D Krinsley and K Langworthy, University of Oregon, Eugene, OR, USA
Published by Elsevier Inc.
4.4.1
4.4.2
4.4.2.1
4.4.2.2
4.4.3
4.4.3.1
4.4.3.2
4.4.3.3
4.4.3.4
4.4.3.5
4.4.3.6
4.4.3.7
4.4.3.8
4.4.4
References
Introduction to Nanoscale Weathering

Nanoscale Techniques for Geomorphologists
Nanoscale Resolution Electron Microscopy
Linking Scales through Digital Image Processing
Applying Nanoscale Strategies to Contemporary Issues in Geomorphic Weathering
Biotic Weathering
Crossing the Nanoscale to Micron-Scale Threshold
Connecting Etching to Weathering Forms
Rock-Surface Alternation of Dust
Silica Mobility in Rock Coatings and Case Hardening
Thermal Stresses
Silica Glaze Formation on Mars by Water Vapor Deposition
Nanoscale View of Rock Polishing
Conclusion
Glossary
Back-scattered electron (BSE) microscopy An accelerated
electron beam in an electron microscope produces collisions
between electrons and atoms, where the largest atoms with
the higher atomic number (Z) generate a brighter intensity
when imaged with a back-scattered detector.
Biotic weathering Mineral weathering caused by life,
including bacteria, fungi, algae, plants, and animals.
Case hardening The outer shell of a rock that has been
hardened (indurated) through the addition of elements
such as silica or iron.
Etching of minerals Mineral dissolution is not an even
produce; areas of more intense dissolution are seen as pits
on the scale of micrometers.
Heavy metal scavenging Iron and manganese oxides and
hydroxides scavenge heavy metals such as zinc, copper, and
lead.
44
45
45
47
48
49
51
52
55
58
59
60
61
63
66
High-resolution transmission electron microscopy

(HRTEM) Two-dimensional spatial imaging of very thin
samples able to image mineral lattices with a spatial
resolution of 40.08 nm.
Microfractures Breaks in minerals that can carry capillary
water.
Nanoscale Features between 1 nm (109 m) and 100 nm
(107 m) or 0.1 mm.
Splintering Rock fracturing in a pattern of subparallel
fractures that resembles a book that has been thrown in
water and then dried.
Thermal fracturing Breaking of minerals from the process
of heating and cooling, such as the passage of a wildfire over
rock surfaces.
Abstract
This chapter presents the first overview of the connection between nanoscale weathering and geomorphology, where three
overarching themes recur. First, nanoscale processes are on one side of a fundamental threshold between the coarser
microscale (micrometers and up) and the finer nanoscale with its dramatically different molecular dynamics. Second,
nanoscale processes do impact a variety of prior geomorphic research, including threads related to ongoing instability in
mineral weathering, silt production, rock coating behavior, geochemical pollution, thermal weathering from wildfires, and
biotic weathering as an explanation for deviations from Goldichs weathering series. Third, it is possible to link the
nanoscale to more classic geomorphic concerns through scaling up quantitatively by digital image processing of microscope
imagery and conceptually through connections to weathering forms such as rock splintering.
Dorn, R.I., Gordon, S.J., Krinsley, D., Langworthy, K., 2013. Nanoscale:
mineral weathering boundary. In: Shroder, J. (Editor in Chief), Pope, G.A.
Weathering and Soils Geomorphology, pp. 4469.
44
http://dx.doi.org/10.1016/B978-0-12-374739-6.00065-8
Nanoscale: Mineral Weathering Boundary
4.4.1
Introduction to Nanoscale Weathering
Scale is a vitally important concern in the development

of geomorphic weathering theory (Phillips, 2000; Viles, 2001;
Turkington et al., 2005; Hall, 2006a, 2006b). Scale is a key
variable in the boundary layer model designed to interpret
spatial variability in weathering (Pope et al., 1995). Scale issues
have consumed much thought and research trying to link the
largely disparate weathering subfields of controlled laboratory
studies and field-based investigations (Brantley and Velbel,
1993; Casey et al., 1993; Swoboda-Colberg and Drever, 1993;
Banfield and Barker, 1994; Brantley and Mellott, 2000; Brantley,
2005; Turkington and Paradise, 2005; White, 2005; Zhu et al.,
2006; Meunier et al., 2007; Navarre-Sitchler and Brantley, 2007).
The solution to the discrepancy between field and laboratory
studies may rest in weathering interactions with nanoscale
particles (Emmanuel et al., 2010; Emmanuel and Ague, 2011).
Thresholds are also an important focus in geomorphic
weathering research (Paradise, 1995; Pope et al., 2002). A key
scale threshold exists between nanoscale processes and those
operating even at the micron scale. In his 2001 presidential
address to the Geochemical Society, Michael Hocehlla, Jr.
emphasized:
nanoscience is based on the premise that materials properties in the
bulk do not simply scale into the nanodomain, but property
modification, and in some cases entirely different properties, are to
be expected y In the nanoscale size range, physical, electrical,
magnetic, thermal, kinetic, and other properties can be altered
dramatically simply due to the physical dimensions of the material.
(Hochella, 2002b: 738).
Interactions at the micron scale simply do not reflect molecular dynamics in the nanoscale chemical environment, in
part because processes at the nanoscale can undergo substantial changes when exposed to water (Zhang et al., 2003;
Wang et al., 2006; Kalinichev et al., 2007; Baram et al., 2011).
Wang et al. (2006: 579) summarize the findings of a molecular modeling approach:
The atomic density profiles of water perpendicular to the surface are
largely controlled by the mineral surface structure. The orientations
of these molecules, however, are dominantly influenced by surface
hydrophobicity, surface charge distribution, and the ability to form
H-bonds with adsorbed water molecules. The first molecular layers
of water at all the surfaces are well ordered parallel to the surface,
reflecting the substrate crystal structure and composition. This
structure, however, is different from that of ice. The mobility of
adsorbed water molecules and the enthalpy of surface hydration are
controlled by mineral surface charge and hydrophobicity.
Nanoscale is typically defined as examining features between 1 nm (10 9 m) and 100 nm (10 7 m) or 0.1 mm
(Figure 1). One way to visualize the nanoscale is to look at your
hand; the scale difference between the entire Earth and your
hand is the same dimensional scale difference as your hand and
a nanoparticle. Nanoscale processes are not included in the
microscale, defined by Pope et al. (1995: 220) as being submillimeter, because a very different realm of weathering processes operate below 0.1 mm. Research on nanoscale weathering
is still in its infancy, having grown up with progressive
improvements in high-resolution transmission electron
45
microscopy (HRTEM) since the 1980s. The first of this research

tended to focus on the very different watermineral interaction
that occurs in the nanodomain within minerals, in that even
microfracture capillary water (Meunier et al., 2007) behaves
differently from water held within mineral internal surfaces
(Hochella, 2002b). The first decade in the twenty-first century
has seen a diversification of nanoscale weathering research,
including the beginnings of attention paid to connections between geomorphology and nanoscale processes.
This chapter focuses on case studies of nanoscale
weathering from a geomorphic perspective. Nanoweathering
alters minerals and also produces nanoparticles through
mechanisms such as chemical weathering, abiotic precipitation, and microbial processes. Although geomorphic research on spatial variability in nanoscale weathering is in its
infancy, nanoparticles do have spatially specific point sources.
They can derive from human pollution. They can be the result
of anthropogenic alterations of a biogeochemical environment. Spatially explicit geochemical pathways alter geochemical cycles in soils, in phreatic water, and in rivers. Exactly
how the study of nanoscale weathering will alter the field of
geomorphology is not possible to predict at this point. Not
enough research has been carried out at the nanoscale to
understand links to landforms. Thus, a primary goal of this
chapter simply rests in opening a dialog between the nanoscale and the landform scale through case studies of nanoscale
weathering in different geomorphic contexts.
4.4.2
4.4.2.1
Nanoscale Techniques for Geomorphologists

Nanoscale Resolution Electron Microscopy
The study of earth materials has been enhanced in the past

decade through the use of HRTEM and other high precision
techniques. Ongoing technique developments over the past
three decades have led to an array of imaging and analysis
tools available to examine weathering processes and products
(Table 1). Each technique has its advantages and limitations
that are reviewed in detail elsewhere (Lee, 2010). This section
overviews a few options that could be used in geomorphological studies of nanoscale weathering.
The basic idea behind high-resolution microscopy is to
thin a sample enough so that electrons can pass through the
sample upon irradiation with a 80300 kV electron beam.
Given the nature of the material and the particular techniques
used, the thickness for high-resolution imaging can vary from
B10 to 50 nm. Today, most HRTEM specimens are prepared
using a coupled dual-beam focused ion beam electron microscopy (FIB-EM). The FIB-EM can site specifically ion mill
and deposit metals; coupled with a micromanipulator, the
FIB-EM can extract features/regions on the order of 1030 mm
wide and carry out nanoscale tomography (Schiffbauer and
Xiao, 2009). Once extracted and mounted to a transmission
electron microscopy (TEM) grid in situ, the FIB-EM is used to
thin the extracted region until it becomes electron transparent
and usable in the TEM (generally referred to as a lamella)
(Brown and Lee, 2007). FIB-EM-prepared samples have many
advantages, such as allowing investigators to see the context of
a lamella in back-scattered electron detector (BSE), secondary
46
(c)
(h)
(b)
(g)
(d)
(e)
(a)
1010
Nano
Micron
108
106
(f)
Rinds/coatings
104
102
Forms
100
Landscapes
102
104
106 m
Figure 1 Visualization of nanoscale weathering placed within broader spatial scales of weathering phenomenon. Examples presented from nano
to landscape scales are: (a) nanoscale silica spheoids a few tens of nm across from silica glaze in Tibet (high-resolution transmission electron
microscopy (HRTEM) image); (b) micron-scale silt formation from quartz weathering in Arizona (back-scattered electron (BSE) image); (c)
millimeter to centimeter-scale rock coatings and weathering rinds illustrated from Wyoming (BSE image) and (d) Death Valley (case hardened
rock shelter); (e) meter-scale weathering forms of a mushroom rock, Arizona and (f) limestone karst stone forest, Kunming, China; and (g)
kilometer-scale weathering landscapes of a salt-encrusted marine terrace, Peru, and (h) varnish-coated alluvial fan, western China, courtesy of
NASA.
electron (SE), scanning/transmission electron microscopy

(STEM), and energy dispersive X-ray spectroscopy (EDS)
modes prior to breaking vacuum and moving the specimen to
a TEM for higher-resolution analysis. Other techniques used to
prepare samples for TEM include mechanical crushing of
minerals, which can generate edges that are sufficiently electron transparent, and ultramicrotoming, which can prepare
electron transparent slices of material using a diamond knife.
Tripod polishing and precision ion polishing (PIPS) are also
useful for obtaining electron transparent areas from bulk
specimens for TEM analysis.
Use of TEMs/STEMs begins with placement of the TEM grid

into a sample holder and insertion of the holder through an
airlock into a high vacuum electron column. HRTEMs use
much higher electron beam voltages (typically from 80 to
300 kV) than scanning electron microscopies (SEMs). The
high voltage, coupled with a thin sample, decreases interaction
volume and allows EDS measurements to be made with spot
sizes of o10 nm, in contrast to typical SEM interaction volumes on the order of B1 mm at 15 kV with thicker/bulk
samples. Many times the spot size can be seen as beam
damage on thin samples, leaving behind a series of holes or
47
Table 1 Overview of nanoscale microscopy techniques useful in weathering research

Technique (and acronym)
Information obtained
Spatial resolution
Limitations
Coupled dual-beam focused ion beam

electron microscopy (FIB-EM)
Used to create and image crosssections in situ, widely used to

extract sections for TEM
analysis. Real time imaging in
SEM mode during ion milling.
2-D spatial imaging, lattice
imaging.
4B1 nm
Maximum sample and scan size,

requires vacuum, Ga ion
implantation.
40.08 nm
Small area, sample preparation

challenges, sample thickness
o50 nm.
Detection limit varies, B0.2 wt.%.
Difficult for light element
detection.
Best results with light elements,
requires high vacuum and thin
specimens.
Requires vacuum, thin specimen.
High-resolution transmission electron

microscopy (HRTEM)
Energy dispersive X-ray analysis (EDS)
Elemental composition, X-ray

mapping of elements.
41 nm (HRTEM);
420 nm (SEM)
Energy-filtered TEM (EFTEM)
Mapping of elements
detectable from EELS spectrum.
o1 nm
Scanning transmission electron

microscopy (STEM)
Scanning electron microscopy (SEM) with
back-scattered electron detector (BSE)
2-D special imaging, high

contrast.
Average atomic number (Z)
revealed through contrast I
grayscale image.
2-D spatial imaging.
o1 nm
41 nm
3-D surface microtopography.
o1 nm in Z, o10 nm X,Y
Elemental and isotopic

composition.
50 nm with NanoSIMS
Scanning electron microscopy (SEM) with

secondary electrons (SEs)
Atomic force microscopy (AFM or FM)
Secondary ion mass spectroscopy (SIMS)
grid patterns (e.g., Lee, 2010: 10). The high spatial resolution
of STEM and the ability to utilize atomic number (Z) contrast
makes it a particularly powerful tool in weathering studies
(Brown and Lee, 2007).
Nanoscale studies of weathering samples generate a variety
of different types of imagery formed by filtering scattered
electrons. Bright-field images are generated from a direct beam
that contains unscattered and low-angle forward scattered
electrons. Dark-field images, by contrast, are generated only
from forward-scattered electrons. The thickness, atomic
number, and Bragg diffraction (in crystalline material) will
control the angle of scattering and the intensity of scattering
all creating the HRTEM image. Since mineralogy is characteristically important in weathering, electron diffraction patterns
record the angular distribution of electrons where minerals
can be identified by the spacing of spots using selected-area
electron diffraction (SED), precession electron diffraction
(PED), or convergent-beam electron diffraction (CBED). It is
also possible to generate images of specific elements at the
subnanoscale using energy-filtered TEM imaging that takes a
specific region of an electron energy loss spectroscopy (EELS)
spectrum to form an image.
Compositional analysis is typically carried out by EDS
for higher atomic number elements and EELS for lighter
elements. EELS measures energy lost through ionization of
sample atoms, separating electrons according to the energy
they have lost. EELS can measure elemental abundance
and also determine valence states at spatial resolutions less
than 1 nm.
45 nm
4.4.2.2
Generally requires vacuum (dry

samples), sample size
dependent on chamber size.
Requires vacuum, sample size
dependent on chamber size
Maximum scanning area of about
150 150 microns. Flat
samples ideal.
Requires vacuum, damages
analysis region.
Linking Scales through Digital Image Processing
An increasingly important weathering literature focuses on explaining offsets between field-based quantification of weathering
and laboratory studies of weathering rates (Brantley and Velbel,
1993; Casey et al., 1993; Swoboda-Colberg and Drever, 1993;
Banfield and Barker, 1994; Brantley and Mellott, 2000; Brantley,
2005; Turkington and Paradise, 2005; White, 2005; Zhu et al.,
2006; Meunier et al., 2007; Navarre-Sitchler and Brantley, 2007;
Emmanuel et al., 2010; Emmanuel and Ague, 2011). Brantley
(2005: 108) concisely summarized some key issues:
y extrapolating from one scale to another (scaling up) is often not
quantitatively successful. For example, quantitative extrapolation of
laboratory rates to field systems remains difficult, and we now
recognize that multiple factors contribute to this discrepancy. For
example, intrinsic factors related to differences in mineral samples
prepared in the laboratory or weathered in the field contribute to
the laboratory-field discrepancy. Of particular importance is the
reactive surface area of dissolving minerals: this term must be investigated and understood more thoroughly. In addition, laboratory rates are generally measured far from equilibrium, whereas
natural weathering often occurs closer to equilibrium where dissolution is slower. This difference in chemical affinity is a consequence of the hydrological complexity of natural systems wherein
fluids at mineral interfaces may approach equilibrium. Thus, dissolution of laboratory samples may generally be rate-limited by the
interface reaction while dissolution in field systems may at times
become partially rate-limited by transport. The scaling-up problem
inherent in the laboratory-field discrepancy may well be solved
through increased understanding developed as we bridge scales of
analysis from the nano to global scaley
48

Columns
4
1 2 3
1 54 50 30 11
2 50 34 11 11
3 32 10 10 5
n
Micron-scale
BSE image of
polished section
Cut out
area of
focus
100 m
Plagioclase
Cut out
Digitalimage
processing
Number of pixels
0
10
Dissolved area
Unweathered area
Acquire HRTEM
imagery from
randomly
sampled areas
10 nm
100 m
Weathering rind
Digitalimage
processing
quantifies
dissolution
Ro
ws
Acquired
image
10 m
10 nm
Brightness value
0
255
Figure 2 Connecting field and lab results, as well as rescaling weathering rates is feasible using digital-image processing of dissolved minerals.
Originally developed for measuring weathering rates over thousands of years in field samples at the micron scale with BSE imagery, this approach
can also directly compare laboratory with field samples and the same samples at scales ranging from square nanometers to square millimeters.
Meunier et al. (2007: 432) explicitly advocates that the

missing link in bridging scales rests at the scale of hand
specimens:
As weathering is a multi-scale phenomenon, theoretically any
model needs to integrate the solid-fluid exchanges from the atomic
interactions at the very surface of primary and secondary minerals
(nanometre), to the rock sample (decimetre) and finally to the
watershed (kilometre). However, such integration is currently beyond our calculation abilities. Hence we have to focus on the
strategic aspect of weathering processes, although deciding where
and on which variables to focus our attention remains difficult. It is
remarkable that our understanding of the physicochemical weathering processes is much more advanced at the extreme scales of
nanometres to tens/hundreds of kilometres, that at intermediate
scales. At the nanometre level, the experimental and theoretical
studies cover the observation of natural samples. The extensive use
of high-resolution transmission electron microscopy (HRTEM),
atomic force microscopy (AFM) and spectroscopy have significantly
improved our knowledge of fluid-mineral reactions and their related interactions. Conversely, the large-scale studies based on
isotope mass balance methods provide denudation rates for provinces or continents. Surprisingly, it is at the hand-specimen scale
where efforts should now be directed.
One strategy that could help bridge the discrepancy at the

hand-sample level is in situ quantitative analysis of mineral
dissolution (Dorn, 1995) using digital image processing (e.g.,
Figure 2). This direct measurement of mineral dissolution has
led to an understanding of field relationships on the temperature dependence of weathering (Dorn and Brady, 1995),
and understanding of the effects of rainfall and temperature
on mineral weathering (Brady et al., 1999), long-term rates of
glass dissolution (Gordon and Brady, 2002), and the role of
lichens and rock coatings on field mineral weathering rates

(Gordon and Dorn, 2005a, 2005b).
Although digital image processing of BSE imagery has only
been used to analyze field weathering that has taken place over
101105 years and at the micron scale, there is no reason why
laboratory samples or nanoscale samples could not be similarly analyzed. For example, the subnanometer conduits that
carry aqueous fluids (Banfield and Barker, 1994) or the
nanoscale etch pits could be prepared for digital image processing to quantify nanoscale porosity (Figure 2).
Previous research on digital image processing to quantify
weathering has focused on the micron scale, typically measuring 105 mm2 of mineral cross sections imaged with BSE per
field site. New advances in electron microscopy have increased
the resolution of back-scatter images. It is now possible to
analyze mineral dissolution at the nanoscale with BSE (e.g.,
Figure 3). Scaling up (cf. Brantley, 2005: 108) and then scaling
down becomes a task of taking lower-resolution BSE and
higher-resolution BSE images using STEM at different scales
and measuring in situ dissolution. Measurement of field and
laboratory-weathered samples alike at different scales, thus,
enables the integration of data to the hand-sample level as
advocated by Meunier et al. (2007).
4.4.3
Applying Nanoscale Strategies to

Contemporary Issues in Geomorphic
Weathering
The goal of this section rests in highlighting different areas of

research in geomorphology where attention to the nanoscale
Figure 3 High-resolution STEM image in dark field mode, which

provides contrast similar to BSE. The dust particles, resting on a rock
coating from the Akesu Volcanic Field, were lifted out, thinned using
the ion beam, and then mounted perpendicular to the beam to allow
electrons to transmit through the sample onto a STEM detector. Prior
to digital-image processing, each mineral would be cut out separately
and made into its own image. Then, the void spaces would be made
its own image and counted digitally given an area of the mineral that
has dissolved (cf. Dorn, 1995).
has either yielded new insights or has the potential to generate

new theory about weathering. This section starts by illustrating
the potential of nanoscale research in studies of biotic weathering. The second case study presents a cautionary tale of
where a lack of attention to the nanoscale/micron-scale
threshold has created problems in interpretation. The other case
studies explore different facets of weathering including etching,
dust weathering, case hardening, thermal weathering, silica glaze
formation on Mars, and the importance of nanoscale processes
in understanding lead contamination of rock surfaces.
4.4.3.1
Biotic Weathering
Research on biotic weathering is taking on increasing importance with the general observation that bacteria and other
microorganisms occur at great depth in the Earth (Reith, 2011).
Biotic weathering occurs at the micron scale and the nanoscale.
This section exemplifies research at both scales, starting with the
micron scale and a classic mystery in the field of weathering
surrounding weathering of olivine, one of the first minerals to
crystallize as a magma cools. Then, this section moves into the
nanoscale where the higher resolution provides new insights on
how microorganisms play a vital role in mineral weathering.
The generally accepted view has been that mineral weathering is the reverse of Bowens reaction series that explains
mineral crystallization. The weathering sequence published by
Goldich (1938) is the opposite of Bowens sequence. It starts
erroneously with the notion of the weathering of olivine, since
olivine formed at the hottest temperatures, and ends with
quartz being one of the most stable minerals, because it
formed last at the coolest temperatures.
49
A flaw in this thermodynamic interpretation was discovered through field-based research using BSE to study the
surfaces of lava flows of Hualalai, Hawaii. This research revealed that olivine does not weather first (Wasklewicz et al.,
1993; Wasklewicz, 1994), as predicted by the Goldich weathering sequence (Goldich, 1938). The field sites where
olivine weathered last, not first, had a paucity of organic
acids being located in the rainshadow of both Mauna Loa
and Hualalai. Olivine was much less weathered than plagioclase and clinopyroxenes in those lava-flow surfaces devoid of
acid-producing lichens and fungi and distant from vegetation.
By contrast, olivine weathered much faster at lava-flow
surfaces that hosted acid-producing organisms. Nanoscale research may hold the key to explain the apparent contradiction
between Goldichs (1938) sequence and the observations
from the rainshadow of Hawaii.
Goldichs (1938) original field sites were located in a humid
environment that contained abundant organic acids and organisms. Other workers have confirmed Goldichs sequence in
similar humid environmental settings (Velbel, 1993). A working hypothesis is that Goldichs weathering sequence could be
more a function of biotic weathering environment than the
thermodynamics of mineral formation. Furthermore, olivine
has a lower elemental bond strength than quartz.
An important biotic influence on the sequence of mineral
weathering could be microbial weathering. Laboratory research
on bacterial weathering at the micron scale determined that
Goldichs weathering sequence is in accord with the vulnerability
of minerals to bacteria (Song et al., 2010). Figure 4 illustrates
some of the evidence presented by Song et al. (2010) in their
controlled laboratory experiment, illustrating how bacteria can
promote dissolution of plagioclase through pitting. One possible
reason for microbial activity causing the Goldich sequence could
be the nutrient needs of microbes (Bennett et al., 2001):
A basic tenet of sediment diagenesis, the Goldich Weathering Sequence (Goldich, 1938), states that the most unstable silicate mineral will weather (dissolve) first, with more resistant silicates taking
progressively longer to dissolve (from least to most stable), olivineoplagioclaseoalbiteoanorthoclaseomicroclineoquartz.
The
observed weathering sequence of minerals in an anaerobic, microbially controlled system, however, is almost opposite, with olivine
stable with respect to microcline, and the relationship between microbial colonization and weathering rate almost perfectly correlated.
This suggests that, in some environments, the indigenous microorganisms may significantly alter weathering patterns as they aggressively scavenge limiting nutrients. (Bennett et al., 2001: 16)
One possible nanoscale explanation for the discrepancy

between Goldichs sequence and weathering in abiotic contexts (Barker and Banfield, 1996: 55) follows:
Biologically mediated weathering involves a complex dissolution/
selective transport/recrystallization mechanism occurring within
the acidic extracellular gels coating all mineral surfaces. A specialized weathering microenvironment around each mineral grain
initially produces minute phyllosilicate crystallites. A rind of clay
minerals forms around the dissolving parent phase, eventually
culminating in abundant 510 mm diameter polymer-bound aggregates of face-to-face oriented clay minerals of homogeneous
composition. Physiochemical weathering of ferrohastingsite produces topotactically oriented smectite and goethite. The cleavagecontrolled reaction is neither isochemical nor isovolumetric.
50
Figure 4 A granite cross-section viewed in secondary electrons before and after being subject to bacterial weathering. An extensive microscope
study analyzed weathering in the laboratory abiotically and biotically. This type of micron-scale controlled laboratory study is an ideal precursor
to a nanoscale investigation of exactly how bacteria generate increased pitting. We thank Dr. Wonsuh Song of the University of Tsukuba for
permission to use these micrographs.
A sub-40-nm scale study of microbial weathering of a

MgFepyroxene in a meteorite that was exposed to weathering for about 70 years (Benzerara et al., 2005a, 2005b)
provides additional insight. Microbial interactions over seven
decades led to carbonate precipitation in the form of rodshaped nanocrystals of calcite and the development of an
amorphous Al-rich layer all leading to the conclusion that
microorganisms create nanoscale weathering environments
that dramatically alter weathering. Micron- and nanoscale
studies of fungal and bacterial weathering of an exposed
granitic pegmatite similarly stress the importance of microbial
ecology in subaerial settings (Gleeson et al., 2005, 2006).
One possible way that microorganisms might affect weathering is through interactions with fungal hyphae. Nanoscale
in situ observations of a soil fungus interacting with biotite
over 3 months revealed a nanoscale attachment to the mineral. Biomechanical forcing altered interlayer spacings, and
microbial processes leached potassium, all leading to the
formation of vermiculate and clusters of oxidized iron (Bonneville et al., 2009; Smits et al., 2009). These clusters of iron
oxides are features commonly seen in BSE images of weathering rinds at the micron scale (Pope et al., 1995), as illustrated in Figure 5. One nanoscale hypothesis for this
ubiquitous weathering-rind phenomenon of submicron fragments of iron oxide could be the ubiquitous presence of
fungal hyphae.
Linking micrometer and nanoscales at the hand-sample
level requires more than examination of just a few samples.
However, pilot investigations of randomly selected samples do
have the potential to suggest future research directions. A
single sample from the Hawaii BSE study of olivine weathering
(Wasklewicz et al., 1993; Wasklewicz, 1994) was reevaluated
Figure 5 Cross section of the weathering rind of a metamorphic

schist in the Phoenix Mountains, central Arizona. The bright mineral
on the far left side is magnetite. Muscovite is the middle mineral, and
hornblende is on the right side. Submicron clusters of iron oxides
have precipitated throughout the muscovite and even into the
hornblende. Iron mobilized from biotite has been observed to fracture
rocks in Karkevagge (Dixon et al., 2002).
with HRTEM. This sample was collected from a Hualalai

Volcano basalt flow f5d c8.2 (B2000 years old) and was
covered with lichens. As in prior research (Bonneville et al.,
2009), hypha appear to be able to exert enough pressure to
split apart smectite clays in the weathering rind (Figure 6(a)).
As in prior research on bacteria weathering (Hiebert and
Bennett, 1992), fungal hypha appear to be able to extract
particular elements. Instead of potassium extracted from bioite
51
Figure 6 Fungal hyphae appear to be extracting calcium from olivine in the weathering rind of a Hualalai basalt flow covered by lichen. (a)
Overview of the interface of a hypha (granular appearance), smectite clay (sm), and an olivine mineral (dark, thicker, and, hence, electron opaque
at the bottom). Arrows show nanoscale strings of darker material that show a spike in calcium in energy dispersive measurements. Given that
the lower end of typical hypha diameter is B2 mm, it is possible that this cross section shows an undulating hypha surface surrounded by
smectite clay. (b) Close-up of the boundary between hyphae nanoscale strings of calcium and the underlying olivine (dark, thicker, and electron
opaque at lower left).
(Bonneville et al., 2009), there is an uptake in calcium

(Figure 6). EDS analyses reveal that there are strings in the
imagery in Figure 6. These linear strings appear to be rich in
calcium. The apparent source of the calcium appears to be an
olivine grain. The reason for this uptake is uncertain. The
calcium might be used in the formation of calcium oxalates
produced by many lichen (Wadsten and Moberg, 1985;
Bjelland and Thorseth, 2002) and fungi (Smits et al., 2009), or
calcium might be used in another microbial process. Regardless of the need, nanoscale fungal weathering of olivine could
provide yet one more explanation why Goldichs weathering
sequence depends upon the presence of organisms and is incorrect where there exists minimal contact between minerals
and biotic activity.
4.4.3.2
Crossing the Nanoscale to Micron-Scale Threshold
There is a danger involved in extrapolating nanoscale weathering observations to smaller scales. Misunderstanding of
the threshold between processes operating at the micron scale
and above versus the nanoscale environment can lead to
confusion and misinterpretation. An example comes from the
rock varnish literature where nanoscale instability was simply
assumed to invalidate (Garvie et al., 2008, 2009) observations
at the micron scale of over 10 000 microlamination sequences
(see Figure 7) recording paleoclimatic fluctuations (Liu and
Broecker, 2000; Liu et al., 2000; Zhou et al., 2000; Liu, 2003;
Liu and Broecker, 2007, 2008a, 2008b; Liu, 2010). In summary, Garvie et al. (2008, 2009) analyzed three samples with
HRTEM at the nanoscale, finding evidence for nanoscale instability of manganese oxides. They then made the scale jump
to the micron scale and simply assumed that nanoscale instability meant that micron-scale paleoclimatic interpretations
of over a decade rock varnish research by T. Liu and colleagues
must be invalid (Dorn and Krinsley, 2011).
The issue here is not problems with the basic observation
of nanoscale instability. Prior to Garvie et al. (2008, 2009),
others noted that nanoscale instability of Mn-oxides is the key
to explaining how rock varnish forms (Potter, 1979; Dorn,

1998; Krinsley, 1998; McKeown and Post, 2001; Dorn, 2007).
In one model of varnish formation, micron-scale microbial
processes fix Mn and Fe oxides, followed by nanoscale instability that moves Mn and Fe oxides between bacterial casts
and clay minerals. Ongoing shuffling of MnFe oxides reoccurs at the nanoscale in the polygenetic model of varnish
formation (Figure 8) (Krinsley, 1998; Dorn, 2007).
Thus, nanoscale instability is a key to varnish formation,
but it does not invalidate micron-scale processes that lay down
regular patterns at the micron scale. As Krinsley (1998: 721)
explained:
at the spatial scale of micronsy some varnish appears quite
stable with distinct layering patternsy At the spatial scale of
nanometers as viewed with HRTEM, varnish appears to be in an
open systemy This is analogous to automobiles moving in a
crowded parking lot, with oxides moving around until they find a
suitable parking space in clay minerals.
Cars moving around inside a parking lot do not imply that the
parking lot itself is unstable (Dorn and Krinsley, 2011).
The broader issue here is that the geochemistry of rock
varnish instability at the nanoscale involves a great number of
unknown issues. Every investigator who has explored varnish
at the nanoscale agrees that Mn and Fe oxide instability is
common (Krinsley et al., 1990, 1995; Dorn, 1998; Krinsley,
1998; McKeown and Post, 2001; Dorn, 2007; Garvie et al.,
2008, 2009), but none of the HRTEM nanoscale observations
infer that the extensively replicated micron-scale stratigraphy
(Liu and Broecker, 2000; Liu et al., 2000; Zhou et al., 2000;
Liu, 2003; Liu and Broecker, 2007, 2008a, 2008b; Liu, 2010)
is invalid. This discussion highlights the danger of extrapolating nanoscale weathering observations to micron-scale
phenomena.
In contrast to the above example of not recognizing a
scale threshold, research that treats nanoscale processes as
dramatically different can provide powerful explanations of
weathering features. An example linking the soil profile scale
52
Coyote Lake, CA (13,020 330)
(a)
Lone pine creek fault scrap, CA (ca. 17,200)
(b)
Owens River Dry Falls, CA (ca. 15,700)
Summer Lake, OR
(d)
(c)
Figure 7 Microsedimentary basins with high rates of varnish accumulation show varnish microlaminations recording Holocene climatic changes.
Black bands formed during wet periods in the Holocene termed WH (for wet Holocene) and multiple black bands in the terminal Pleistocene
black layers WPO (Younger Dryas, wet Pleistocene black layer zero) and WP1 (wet Pleistocene black layer 1). Under the right conditions of fast
accumulating varnish (images (c) and (d)) and proper thin section procedures, millennial and submillenial wet periods can be used successfully
in geomorphic research (Liu and Broecker, 2008a). Although these varnishes started to form about the same time, faster accumulation rates for
images (c) and (d) allow a higher resolution record than images (a) and (b). Images courtesy of T. Liu.
to the nanoscale is an investigation by Kaufhold et al. (2009)

of an allophane unit in Ecuador that is about 20 000 14C years
old and covers an area of over 4000 km2. The unit is over 5 m
thick and is up to 80% allophane by weight. In addition to
documenting the geography and dimensions of the unit,
this study analyzed the micrometer and nanometer scale aspects of the 15 nm scale allophane particles to inform on the
volcanic and weathering formative settings of this unique
deposit (Kaufhold et al., 2009). The large allophane deposit
started with a particular volcanic eruption that was weathered
in situ under an Andosol soil under conditions that developed
a gradient of alkali earth metals sufficient to form the
allophane.
4.4.3.3
Connecting Etching to Weathering Forms
Etching is a term used at almost all scales (Figure 1) in

geomorphology, from etch plains that are erosion surfaces
(Campbell and Twidale, 1991; Twidale, 2002), to etching that
generates karst weathering forms in sandstone (Young, 1988;

Young and Young, 1992), to etching of minerals in soils used as
a relative dating method (Locke, 1979, 1986), to nanoscale etch
pits generated by bacteria and analyzed by atomic force microscopy (Buss et al., 2007). In the context of mineral weathering,
however, analyses of etch pit morphologies are traditionally
carried out with secondary electrons and sometimes backscattered electrons (Hall and Michaud, 1988; Lee and Parsons,
1995; White, 2005; Song et al., 2010), although they are sometimes studied at both the micrometer and nanometer scale
(Hochella and Banfield, 1995; Brandt et al., 2003; Buss et al.,
2007; Lee et al., 2007). As one would expect, isolated etch pits
occur at the nanometer scale that are typically only a few
nanometers in depth and therefore not visible with SEM
(Brandt et al., 2003: 1457). Higher-resolution nanoscale research
reveals that the nanotexture of minerals, produced by defects
dislocation densities and patterns, plays a key role in how
minerals weather (Banfield and Barker, 1994; Hochella and
Banfield, 1995; Brown and Lee, 2007; Lee et al., 2007).
53
Micron scale: Mn and Fe

fixation in bacteria sheaths
Si
1 m
Varnish
Al
K Ca
Mn Fe
Mn
Si
Al
0.2 m
KCa
Ti
Fe
Spot on bacteria
Nanoscale
dissolution from
20 nm
20 nm
8 nm
cell wall (granular

fragments)
and
Nanoscale
physiochemical
barrier: fixation
in clay matrix
cementing
10 nm
10 nm
20 nm
20 nm
mixed-layered
clays
Figure 8 Nanoscale instability of manganese and iron is a key to the formation of rock varnish. HRTEM and secondary electron microscopy
images of cocci bacterial forms from Antarctic varnish and SEM/EDS reveals bacterial forms that concentrate manganese and iron at the micron
scale. HRTEM of samples from Peru, Death Valley, and Antarctica illustrate nanoscale instability where oxides mobilize from bacterial sheath
fragments, are transported a few nanometers, and then are fixed into mixed-layered illite-montmorillonite clays. These nanometer-scale Mnoxides can be remobilized and refixed over and over again in a shifting nanoscale mineral structure (McKeown and Post, 2001: 712).
Studies of mineral etching at the nanoscale have raised a

critically important conundrum with respect to clay mineral
production by chemical weathering. Decades of thought about
mineral weathering have generated a paradigm that feldspar
chemical weathering always leads to clay mineral production,
as supported by some of the first nanoscale weathering studies
of feldspars (Banfield and Eggleton, 1990). More recent research, however, reveals that water-to-rock ratios in the weathering environment may be a key determinant in whether or
not clay mineral alteration products occur at the nanometer

scale (Lee et al., 2007). In other words, clay minerals do not
always occur next to feldspar minerals; the nanoscale weathering environment is an important player, especially capillary scale and nanoscale water.
The scale jump connection between mineral nanoetching
and much larger forms has not been considered in the literature, according to our understanding. Even linking handsample weathering forms and nanoetching has also not been
54
explored previously. However, we believe that making such a

linkage is possible and propose here a speculative connection
between nanoetching and the weathering form termed
splintering (Fitzner et al., 1997; Stoppato and Bini, 2003;
Fitzner and Heinrichs, 2004; Dorn et al., 2008). Splintering is
where all different types of rocks, from basalt to sandstone to
schist to granite, have the appearance of pages of a book that
has been wetted and then dried (Figure 9).
In the case of a foliated metamorphic rock, where grains
are aligned, the weathering form of splintering is likely a
consequence of differential weathering of more susceptible
minerals at the micrometer to millimeter scale. Splintering
simply reflects the millimeter to centimeter expression of micron-scale processes (Figure 10). However, splintering is a
ubiquitous weathering form in virtually all rock types, even
those with no obvious alignment of minerals. Explaining how
this takes place at the hand-sample scale has not been a topic
of focus in the geomorphic weathering literature.
A hypothesis that relates splintering forms (Figure 9) to the
nanoscale alignment of etching is that subparallel alignment
of mineral etching at the nanoscale can lead to alignment of
capillary water conduits at the micrometer and then millimeter scales. These capillary water conduits appear to be key
agents for the propagation of mineral weathering (Meunier
et al., 2007).
Evidence in support of this hypothesis starts with a
micrometer-scale image of feldspar weathering (Figure 11).
From the perspective of Meunier et al. (2007), microfractures
carry capillary water, and these fractures connect weathering
pits recognized as initial porosity containing resident
fluid. Microfractures propogate into the reaction area of
these areas of initial porosity, just as streams propogate
headward.
If splintering occurs when the enlarging microfractures
(Meunier et al., 2007) encounter nanoscale etch features that
are aligned, then the test of this hypothesis would be that
splintering would not occur where nanoscale etching is not
aligned. This appears to be the case, as evidenced by nanoscale
lava weathering that does not display evidence of splintering
(Figure 12).
Figure 10 BSE image of biotite from a gneiss clast in a desert

pavement near Florence, Arizona, undergoing a mix of splintering and
grussification. The main aspect of this image feature is that the
biotite itself has expanded volumetrically. Some of this expansion is
due to the formation of secondary weathering products, such as the
bright iron oxides or the darker material rich in silicon and aluminum.
The inset image shows the broader context of biotite next to quartz,
and how the biotite weathering breaks bonds between different
minerals in the rock.
Figure 11 BSE image of a plagioclase in a hand sample showing

splintering, Phoenix Mountains, Arizona. Horizontal parallel etching
could relate to lattice imperfections, and there is also etching along
perpendicular fractures. The diagonal marks are an artifact of sample
polishing.
Figure 9 The weathering form of splintering seen at a hand-specimen scale (left, where the flower width is 2 cm) and where the talus boulder
is 2 m tall (right). Left image is of hornfel at Black Canyon, Arizona; right image is of sandstone at Bryce National Park, Utah.
55
Figure 12 HRTEM image of plagioclase minerals illustrates etching

at two scales. A finer scale of etching about 20 nm across could
relate to lattice imperfections. This finer etching is perpendicular to a
much larger etch feature that could relate to a larger mineral
dislocation glide. The misalignment of etching patterns running in
different directions could inhibit the growth of capillary water-bearing
fractures in a subparallel alignment. The sample is from
trachyandesite lava in the Akesu Volcanic field, Tibet (Wei et al.,
2003). This sample shows no evidence of splintering.
Figure 13 presents a case study of a silicified dolomite from

South Australia, where the alignment of splintering at the
hand-sample scale corresponds with dissolution of microfractures carrying capillary water at the micron scale, and then
also with alignment of connected weathered pores at the
nanoscale. Nanoscale alignments appear to enable propogation of capillary water-bearing fractures in subparallel alignments that in turn generated the splintering weathering form
seen at the hand-sample scale. By way of analogy, this conceptual model is similar to the way a parallel drainage pattern
develops through taking advantage of fracture patterns.
4.4.3.4
Rock-Surface Alternation of Dust
Silt in the form of dust is ubiquitous on Earth (Goudie, 1978;

Bullard and Livingston, 2009), Earths moon (Gaier, 2005),
Mars (Israel et al., 1997), and likely elsewhere. Weathering
interactions within dust and between dust and underlying
mineral surfaces are taking on increasing importance in
understanding the decay of building stones (Sharma and
Gupta, 1993; McAlister et al., 2006; Smith et al., 2007), formation of rock coatings on Earth (Dorn, 2009; Krinsley et al.,
2009) and Mars (Johnson et al., 2002; Kraft et al., 2004), in
understanding historic and prehistoric archaeology (Ganor
et al., 2009), and in Quaternary research (Yaalon and Ganor,
1973; Kleber, 1997). Nanoscale research informs on dust
production (Pye, 1987; Pye, 1989; Hochella, 2002a) and also
Figure 13 The weathering form of splintering of a silicified

dolomite, South Australia ((a) hand-sample scale), appears to
correspond with the development of aligned micron-scale fractures
that carry capillary water ((b) micron-scale BSE image) and, in
turn, with the alignment of nanoscale pores that are connected ((c)
nanoscale HRTEM image) and could be responsible for fostering the
growth of micron-scale fractures. Locations of sample collection for
analysis at progressively finer scales is indicated by the lines.
56
Figure 14 Images of quartz weathering observed in a desert pavement near Florence Junction, Arizona. These images show formation of siltsized quartz fragments that appear to depend on both the existence of a fracture inside the quartz grain (white arrows) and also fragmenting
along angles that are not aligned with the fracture (black arrows). For example, the subparallel fracturing seen in the right image likely relates to
aligned crystal defects. However, the jigsaw puzzle appearance of quartz fragments on the edge of the fracture (upper left white arrow) in the left
image is not easily explained by aligned defects.
what happens when dust deposits on rock surfaces as dust

films and components of rock coatings.
Evidence exists that silt production is a function of inherent weaknesses in minerals such as quartz (Moss and
Green, 1998; Smalley et al., 2005; Kumar et al., 2006), as well
as external processes such as salt weathering, abrasion, frost
weathering (Smalley and Krinsley, 1978; Whalley et al., 1982;
Wright et al., 1998), microbial weathering of quartz surfaces
(Brehm et al., 2005), and nanoscale interactions between
quartz and water (Pope, 1995). The new generation of higherresolution-BSE detectors is able to generate nanoscale resolution to illustrate both internal and external modes of quartz
silt production (Figure 14).
The transition between wind-transported dust and dust as
an integrated component of rock coatings has seen surprisingly little study. Even though there is a general consensus
in the rock-coating literature that aeolian dust contributes to
the formation of rock varnish, phosphate skins, and silica
glaze (Potter and Rossman, 1977; Dorn, 1998, 2009; Langworthy et al., 2010), we do not know of a systematic study of
nanoscale alternations of dust in the transition from rocksurface loess to rock-coating constituent.
We present here a nanoscale study of the transition from
dust to rock coating in the Ashikule Basin, Tibet (Figure 15).
The 47004800-m-high graben is a dusty, sulfate-rich, highaltitude, and high-ultraviolet (UV) flux environment. This
field site offers the opportunity to study dust/substrate interactions distant from industrial pollution. The cold, dry, lowerair-pressure nature of the field suggests potential for a Mars
analog site in part because the substrate is a lava flow, a trachyandesite of Ashishan Volcano in the Akesu Volcanic field
(Wei et al., 2003).
While nanoscale mineral etching is of key importance as a
process in chemical weathering, etched mineral surfaces could
also play a role in the initial stages of the physical attachment

of dust particles to mineral surfaces. Dust accretion on mineral
surfaces is generally thought to occur from electrostatic or
physical forces that hold dust particles together (Jordan, 1954;
Bishop et al., 2002; Ganor et al., 2009). Nanoscale mineral
etching creates an irregular surface that could play a key role in
the initial attraction (Figure 16).
The particular sample selected for a detailed elemental
nanoscale analysis is dust that accumulated on top of a rock
coating that is a mixture of silica glaze and rock varnish. Silica
glaze dominates this rock coating, but pods of rock varnish
240 mm across exist and are enveloped by the silica glaze
(Krinsley et al., 2009; Langworthy et al., 2010). We selected a
form with two oval shapes in the dust on top of the rock
coating. This pod form has a texture suggestive of a biological
origin (Figure 17), in that it has a granular texture similar to
microbial hyphae observed elsewhere (Bonneville et al., 2009).
Our nanoscale investigation of grain alternation generated
EDS data in a pattern of 300 points in a 10 30 matrix, where
measurements are taken approximately 1.3 nm apart. Beam
damage allows visualization (Lee, 2010) of the clear grid pattern (Figure 18). The grid serves as ground control points,
allowing mapping of digital EDS data with geographic information system (GIS) software and subsequent visualization of
nanoscale weathering.
Figure 19 displays a visualization of the 300-point grid of
EDS data using a Kriging algorithm to map out elemental
patterns. The dominant peaks in the EDS analyses consisted of
the inorganic components of silicon and oxygen, and potential organic components of carbon and phosphorus. The
highest concentrations of Si and O occur underneath the
granular oval; this particle has a parallel structure suggestive of
a clay mineral composition that would be consistent with the
higher abundance of Si and O. By contrast, the highest
57
Figure 15 Ashishan Volcano study site in the Ashikule Basin, Tibet at 35.6988 N, 81.57623 E. The length of Urukele Lake is about 7 km,
Ashikule Lake about 5.5 km, and the width of the graben at the location of the Ashishan Volcano is about 22 km. Modified from a raw image
designed by William Bowen.
Figure 16 Mineral etching creates nanotopographic irregularities on

the scale of less than 10 nm, as seen in this HRTEM image of cross
section of a dust particle attracted to the underlying mineral surface
of a rock. Nanoscale topographic irregularities likely contributed to
the physical attachment of dust to this mineral surface, collected
from the Akesu volcanic field, Tibet.
concentration of carbon occurs where the oval form displays a

gray granular texture that transitions to an area of greater
porosity with the lowest carbon concentrations. The area
richest in C does appear to correspond with high P concentrations, but some areas of high P do have low concentrations
of C. Banfield et al. (1999: 3407) found potential complex
mixtures of clay minerals and complex organic polymers associated with microbial weathering, and this is possible for the
analyzed oval in Figures 18 and 19.
Visualization of nanoscale spatial geochemistry with EDS
grids has been accomplished in mineralogical research (Lee,
2010), but not analyzing rock coatings. Thus, the following
interpretation of this single pilot analysis is speculative.
Starting with the assumption that the oval particle had an
organic origin, it appears as though Si and O (likely silica)
Figure 17 HRTEM image of dust particles on top of silica glaze and

rock varnish. Many dust particles appear to be clay minerals,
whereas others are fragments of other minerals. We selected a
particle with two oval forms that could have a biotic origin (arrow).
These oval forms appear to be undergoing post-depositional
modification, as indicated by the uneven texture of bright and dark
areas that suggest differential movement of elements.
have begun to migrate from the underlying clay inward a

possibility deduced by the gradient of Si and O from the clay
at the bottom into the granular particle. The lower levels of C
at the top of the grid could reflect the electron transparency of
the middle of the oval form; picture sectioning a highly desiccated bacterial cell (or cross section of a desiccated fungal
hypha). The thickest portions would be on the margins,
58
whereas the center would be thinnest explaining why the

lowest carbon concentrations occur in the middle of a possible
desiccated microbial form. The behavior of phosphorus appears to link with a brighter area of the HRTEM image, and
this bright area sends stringers toward the upper left. Like Si
and O, the spatial pattern of phosphorus suggests nanoscale
movement and hence instability. Our summary speculative
interpretation of nanoscale instability is not unique to rock
coatings such as rock varnish (Dorn, 1998; Krinsley, 1998;
McKeown and Post, 2001) or silica glaze (Dorn, 1998; Gordon
and Dorn, 2005b; Langworthy et al., 2010). However, it appears as though nanoscale movement of inorganic and possible organic components starts in dust deposits on rock
surfaces prior to envelopment inside a rock coating or incorporation into soils.
4.4.3.5
Figure 18 EDS data were collected in a grid pattern, where X-rays

were generated every 1.3 nm apart. The center right part of the
particle appears to be the most porous. The beam damage (seen in
upper right inset image) creates a grid on the sample, and, thus, the
precise locations of data are known and can be analyzed using
geostatistical methods.
Silica Mobility in Rock Coatings and Case

Hardening
Case hardening of the outermost shell of a rock commonly

involves the migration of rock coating and other weathered
products into pore spaces in the weathering rind of the
rock (Conca and Rossman, 1982; Gordon and Dorn, 2005a;
Dorn, 2009). Nanoscale studies into the processes by which
this occurs in migration reveals new insights into how silica
behaves on the surfaces of rocks. The first nanoscale investigation of silica glaze formation revealed that silica precipitates
in the form of spheroids with diameters between 20 and 70
nm (Langworthy et al., 2010) as seen in Figure 20. EDS analyses by Langworthy et al. (2010) indicate that these spheroids
are composed of just Si and O.
The spheroid diameters revealed by HRTEM are particularly
significant for interpreting the behavior of water at the
nanoscale. There is a transition between complete and partial
wetting on silica surfaces between 20 and 70 nm (Zorin et al.,
1992; Churaev, 2003). This is the size range of the observed
Si
Figure 19 Nanoscale EDS mapping of a 300-point grid. Each grid point is approximately 1.3 nm apart, and the grid is 10 columns and 30 rows.
The visual grid corresponds in size with maps of the four most common elements. Figure 18 shows the larger context and a HRTEM prebeam-damaged image of the analyzed area. The Kriging algorithm used to map out elemental patterns presents an intuitive color scale where
lowest values are dark green and highest values are brown to pink.
59
Figure 20 HRTEM imagery of silica spheroids that have precipitated inside a pore space in the weathering rind (left). The silica spheroids
composed of Si and O (Langworthy et al., 2010) are between 20 and 70 nm in diameter (right). Bright spots are artifacts.
spheroids. Dorn (1998) originally suggested that crossing this

transition ruptures the metastable wetting film (Zorin et al.,
1992) resulting in silica precipitation a hypothesis that is
consistent with the presence of these spheroids (Langworthy
et al., 2010).
Silica is an important case-hardening agent (Wilhelmy,
1964; Washburn, 1969; Conca and Rossman, 1982; Robinson
and Williams, 1992; Mottershead and Pye, 1994; Dorn, 2004;
Tratebas et al., 2004; Gordon and Dorn, 2005a). If the postulated model of silica glaze formation by spheroid deposition
is found to occur in locations other than Tibet (Figure 20),
there is every reason to suspect that silica would continue to
remobilize and reprecipitate at the nanoscale, filling in pore
spaces. HRTEM imagery (Langworthy et al., 2010) shows that
the contacts between spheroids appear fused (Figure 20).
Furthermore, there is every reason to believe that silica
movement should shift silica from rock coatings into the
underlying weathering rind, hence case hardening the outer
shell of a rock surface. Thus, it is possible that silica fills in
voids in the weathering rind by growing new spheroids into
pore spaces, since droplets do not cling to a surface immediately after formation, but move somewhat before they attach
to the [pre-existing] solid (Koopal et al., 1999: 24). Thus,
nanoscale silica instability could explain silica-impregnation
and subsequent case hardening of weathering rinds.
4.4.3.6
Thermal Stresses
Studies of thermal stress at the micron scale reveal the need for
higher resolution imagery to refine understanding of weathering processes (Mahaney and Milner, 2011). There is an
awareness that fractures from thermal stresses should exist at
the nanoscale (Van der Giessen and Needleman, 2002; Hall,
2006a, 2006b). However, we are not aware of nanoscale observations of frost weathering, insolation weathering, or fire
weathering. Thus, we present here a few initial observations of

nanoscale features that relate to physical weathering from fire
weathering and perhaps frost weathering.
An AprilMay 2000 wildfire burned about 37.5 km2 of the
Sierra Ancha Mountains, Arizona. Sandstone and diorite
boulders originally surveyed in 1989 were resurveyed after the
spring fire, after the summer monsoon season, and then after
the winter 2001 season (Dorn, 2003). The most extensive
postfire erosion of sandstone boulders did not take place after
summer rains, but after winter precipitation. Figure 21 presents an electron microscope image of quartz and plagioclase,
where winter detachment took place within the weathered
porous regions of the feldspar.
The original interpretation (Dorn, 2003) was that dissolution of feldspar took place in the weathering rinds over
thousands of years before the fire, where moisture held in
these pores may have played a role in accelerating fire spalling
(Allison and Goudie, 1994).
A subsequent HRTEM investigation of this same sample
indicates textural evidence that the mineral grains were subject
to shear stress prior to spalling. Experiments on shearing associated with confined freezing and melting show lateral
movement at the nanoscale (Christenson, 2001). An examination of lower-resolution imagery reveals that the feldspar
grains have a ground-out appearance (middle image of Figure 21), and HRTEM images at the erosional surface reveal
lattice displacement tens of nanometers wide (lower image of
Figure 21). Although not in the original interpretation (Dorn,
2003), this new evidence of shearing at the eroded surface
suggests that frost weathering may have contributed to accelerated erosion rates during the winter.
A 2001 wildfire at Whoopup Canyon, Wyoming, spalled
sandstone joint faces that host rock art panels. A comparison
of prefire samples collected in 1991 and postfire samples
collected in 2003 (Tratebas et al., 2004) reveals that the fire
60
a region of unfractured quartz appears just a little darker, a

nanoscale view reveals jagged fracturing. We speculate that
a detailed study of fire-shattered rock would reveal an array of
twisted and distorted mineral fragments similar to this single
sample.
One conclusion of micron-scale studies of fire weathering
in the Sierra Ancha of Arizona (Dorn, 2003) and Whoopup
Canyon in Wyoming (Tratebas et al., 2004) was that case
hardening helps minimize erosion associated with fire weathering. Locations that did not undergo detachment from fire
weathering at both study sites were rock surfaces where weathering-rind pore spaces had been infilled with silica (Figure 23), and this infilling of silica case hardened the
weathering rind. At scales finer than the 20 mm diameter silica
spheroids (Figure 20), before the transition between complete
and partial wetting on silica surfaces (Zorin et al., 1992;
Churaev, 2003), silica appears to be leached out of quartz in
channels only a few nanometers wide (Figure 23). These silica
flows into the pore spaces help cement grains together.
Detachments after the fire took place underneath this casehardened zone.
4.4.3.7
Figure 21 A spalled grain from sandstone Boulder 1.2 collected

after the winter season in May 2001 from the Sierra Ancha
Mountains, Arizona, after a wildfire (Dorn, 2003). The top image
presents a scanning electron microscope view of quartz (less
weathered upper part of the image) and plagioglase (lower
porous weathered grain). The winter 2001 spall face cut across the
heavily weathered plagioclase. The middle image of a lower resolution
TEM view (location indicated by the telescoping lines) of the erosional
surface presenting a chaotic mix of clays and less weathered feldspar
fragments. The appearance appears jumbled and not well ordered,
consistent with a shearing event. The bottom HRTEM image provides a
nanoscale view of the spalled surface of an illitesmectite mixed-layer
clay; disruptions would be consistent with shearing of the clay.
generated thermal fracturing in the quartz grains of the

sandstone (Figure 22). A sample was extracted from a seemingly unfractured quartz fragment for HRTEM study; however,
at the nanoscale the quartz was fractured into jagged pinnacles
and other forms (Figure 22). Although light microscopy shows
Silica Glaze Formation on Mars by Water Vapor

Deposition
Nanoscale studies of rock coatings are certainly not limited to

terrestrial samples. High-resolution STEM reveals the presence
of iron-rich coatings on lunar samples. An iron-rich accretion
less than 10 nm thick is chemically distinct from the underlying rock, perhaps vapor deposited (Noble and Keller, 2006).
The planetary geology literature is replete with speculation on
the occurrence of rock (or desert) varnish (Greeley, 1987;
DiGregorio, 2002; Boston et al., 2008). However, a working
hypothesis among some investigators is that coatings of silica
glaze should be the most common rock coating on Mars
(Dorn, 1998; Kraft and Greeley, 2000).
The pre-Noachian period of Mars is the period of accretion
about four billion years ago. The Noachian period is represented by the oldest surfaces with the greatest density of impact craters and also with the greatest density of surfaces
impacted by liquid water producing river valley networks.
Silica glaze on Mars might relate to Noachian rainfall events
(McSween et al., 1999).
An alternative is that silica glaze relates to liquid
water films on mineral surfaces (Boynton et al., 2009) or
perhaps water vapor interactions around springs (Allen and
Oehler, 2008). Phoenix Landing Site research reveals that H2O
ice and vapor interacts with soil particles (Smith et al., 2009)
and that water vapor concentrations can vary considerably
(Whiteway et al., 2009). The presence of water vapor near and
at the surface of Mars makes reasonable the hypothesis that
billions of years of tiny amounts of water vapor interactions
might play a role in generating accumulations of amorphous
silica on Mars. This hypothesis was tested in a 20-year experiment (Dorn, 2012).
For 20 years, basalt rock chips were placed in a closed
reaction vessel with circulating air containing 80% relative
humidity. At no point did liquid water come into contact
with basalt surfaces. Basalt rock chips that lacked any rock
61
Figure 22 Sample WC-03-12 was collected from a sandstone joint face that acted like a fire chimney, resulting in fire erosion. The left image
shows a light microscope thin section view of thermal fracturing in the quartz, exemplified by the arrows. Although fractures did occur in 1991
prefire samples, the abundance increased dramatically. The right HRTEM image reveals nanoscale fractures in segments of the quartz that
appeared to be unfractured at the micrometer scale.
coatings at the start of the experiment had coatings of

silica glaze formed entirely by water vapor by the end
of the experiment. Figure 24 shows the progressive accumulation of silica glaze on basalt chips removed from the reaction
vessel after 5, 10, 15, and 20 years in response to only water
vapor. The thickness appears to slowly increase in the first
decade and then there is jump in thickness after 15 years of
exposure to water vapor. By 20 years, the water-vaporproduced silica glaze was thick enough to image at the micron
scale with BSE.
In addition to demonstrating that silica glaze could form
without liquid water interactions, through nanoscale processes, this experiment also offers an explanation why silica
glaze formation is favored over rock varnish and other rock
coatings in humid environments like the rainshadows of
Hawaiian volcanoes (Farr and Adams, 1984; Gordon and
Dorn, 2005b); the abundance of humidity and the paucity of
dust combines to favor silica glaze formation. The notion of
nanoscape vapor deposition of rock coatings is not limited to
Mars; space weathering has occurred on the surface of Itokawa
through the accretion of a sulfur-bearing iron-rich 5 to 15 nm
thick coating (Noguchi et al., 2011).
The source of the silica must have derived from the
basalt chips. There is no other choice. Thus, there must be
some nanoscale movement of silicon and oxygen from the
underlying minerals toward the surface. Prior research on
alphaquartz reveals that hydration energies of water vapor
absorption range from 90 to 28 kJ mol 1 (de Leeuw et al.,
1999). Artificial silica powder samples are known to alter
more under higher relative humidity (Morel et al., 2009). It is
possible that silicawater vapor interactions occurred through
nanoparticle dissolution (Rimer et al., 2007) in the underlying

mineral and then nanoparticle assembly on the surface
(Rimer et al., 2008). The preliminary HRTEM observations
(Figure 24), however, are insufficient to infer nanoscale processes. Still, the absence of a clear explanation of process does
not deny the fundamental observation that silica glaze can
accumulate on basalt mineral surfaces in just the presence of
atmospheric water vapor.
While 80% is a relative humidity and a room temperature
an absolute humidity far greater than found at the Mars-soil
interface, this experiment reveals that liquid water is not
needed to form coatings of silica glaze. Silica glaze could be
forming on Mars now and over the billions of years where
mineral surfaces have been exposed to water vapor near or on
the surface of Mars.
4.4.3.8
Nanoscale View of Rock Polishing
This last case study explores something as simple as rock

polishing. Those involved in weathering research regularly
prepare cross sections and thin sections to examine samples in
electron microscopes. Extremely smooth and well-polished
surfaces are often a key for obtaining clear imagery using
techniques such as BSE microscopy. Polishing is done in
progressive stages of increasingly finer grits. Thus, from the
perspective of sample preparation, polishing is a result of
smoothing a surface.
However, micron-scale research on actively abraded surfaces described to have been polished is associated most
typically with the presence of silica glazes in settings such as
gibbers in Australia (Dorn, 1998: 282), abraded river cobbles
62
Figure 23 Samples collected from a Big Elk Petroglyph on Panel 6.2, collected from Whoopup Canyon, Wyoming (Tratebas et al., 2004), reveal
how silica glaze infiltrates into the weathering rind. The three images were acquired at successively finer scales, where the letter C indicates the
location of successive close-up imagery. The top two BSE images show the micron-scale context of how silica glaze infiltrates into pore spaces
and cements sandstone quartz grains. The lower HRTEM image reveals what happens at the contact of the quartz grain and the pore space. The
pore space has some clay minerals (e.g., upper right in lower HRTEM image), but much of the material at the quartz boundary appears to be
amorphous. EDS analyses reveals that the silica appears to be mobilized from the quartz and moves into this amorphous material in what
appears to be darker channels that are only a few nanometers across. This material could be a precursor to the silica spheroids presented in
Figure 20.
(Dorn, 1998: 292), and glacially polished rock in Nepal and

Mexico (Dorn, 1998: 303). In other circumstances, iron films
are associated with active glacial polishing (Dorn, 1998: 96,
162) or iron films together with silica glaze (Dorn, 1998:
343). Although rock coatings may be responsible for some of
the polished appearance, nanoscape perspectives paint a different picture.
Figure 25 illustrates the role of weathering in creating a
polish. By weathering, we mean physical abrasion that fragments the rock (physical weathering), detaches rock fragments, and then erodes those fragments. Consider a sample of
glacially polished plagioclase phenocryst collected next to the

Khumbu Glacier, Nepal (Figure 25(a)). The polishing appears
to derive from a very smooth surface, where the only irregularities are undulations on the scale of a couple of
nanometers. The polishing, therefore, represents physical
weathering by abrasion and then eroding plagioclase
fragments.
Rock polishing can also occur from the combination of
physical abrasion and the addition of external materials. A
slickenside collected from near Badwater in Death Valley had a
distinct shiny polish. The polishing appears to be an effect
4.4.4
Figure 24 Silica glaze forms on basalt rock chips after exposure to

only air with a relative humidity of 80% at 18 1C in a 20-year-long
laboratory experiment. The label for each micrograph indicates how
many years the basalt chip was exposed to the water vapor. Arrows
point to the boundary between overlying silica glaze and the
underlying mineral (plagioclase in for the three HRTEM images for
years 5, 10, and 15). The very thin lighter band at the outer (upper)
edge of each HRTEM image is an artifact of sample-beam interaction.
The sample was not prepared in such a way that presence of
spheroids could be imaged. In the case of the 20-year-BSE image,
arrows show the contact between the darker (lower Z) silica glaze
and the brighter (higher Z) underlying basalt minerals.
created by smearing of thin clay coating that is predominantly

smectite mineralogy, similar to nanocoatings occurring along
the San Andreas Fault (Schleicher et al., 2010) and in experimental studies (Han et al., 2011). Fragments of plagioclase
weathered from the host rock were embedded in the smeared
smectite, suggesting that some clay mineral weathering had
taken place along the fault surface. Ongoing movements
mixed the plagioclase from the rock and the smectite.
This last case study of polishing is intended to whet the
appetite of the reader. Collect samples polished by active
geomorphic processes coastal, fluvial, aeolian, or glacial.
Examine the sample with nanoscale tools (Figure 1). Something as simple as rock polishing may turn out to be complex
and fascinating.
63
Conclusion
Weathering is a multidisciplinary arena explored by geochemists, soil scientists, geomorphologists, hydrologists, civil
and environmental engineers, archaeologists, planetary
geologists, and others interested in the breakdown and decay
of rocks. Geomorphologists were among the first to explore
the importance of nanoscale processes in weathering
(Eggleton, 1980; Krinsley et al., 1995; Pope, 1995), even
though nanoscale research in weathering has been dominated
by geochemists over the past two decades (Banfield and
Eggleton, 1988; Banfield et al., 1999; Hochella, 2002a;
Waychunas et al., 2005; Hochella et al., 2008). A consequence
of a general lack of attention to the nanoscale by geomorphologists has resulted in a situation where few connections exist
between Earths landforms and the nanoscale.
Certainly, weathering theory in geomorphology is concerned with scale (Pope et al., 1995; Phillips, 2000; Viles,
2001; Turkington et al., 2005; Hall, 2006b). However, the
microscale in geomorphic weathering theory has been the
subject of research at micrometer and higher scales. Nanoscale
processes operating between 1 nm (10 9 m) and 100 nm
(10 7 m) or 0.1 mm (Figure 1) do not reflect microscale conditions or processes.
Physical, electrical, magnetic, thermal, and kinetic properties differ dramatically across the nanomicron threshold
(Hochella, 2002b; Zhang et al., 2003). The behavior of water
in microfractures supporting capillary flow at the micron scale
(Dorn, 2003; Dixon and Thorn, 2005; Meunier et al., 2007)
does not reflect the behavior of nanoscale water (Lower et al.,
2001; Zhang et al., 2003; Wang et al., 2006; Kalinichev et al.,
2007; Langworthy et al., 2010). An example comes from brucite, a magnesium hydroxide that is a common weathering
product (Kalinichev et al., 2007). Kalinichev et al. (2007)
modeled three well-defined layers with respect to the behavior
of water within 15 A of a brucite surface. There is an inner
layer of a few angstroms that is highly structured and contains
water molecules with a high atomic density coordinated to the
brucite surface. There is a middle transition layer about 4 A
from the surface with a low atomic density, and the outer layer
that becomes similar to bulk nanoscale water about 515 A
from the surface. A similar result comes from studies of how
water adheres to mica (Xu et al., 2010), where the first and
second 0.4 nm layers of water behave like ice and accumulate
at surface defects, but thicker accumulations of water are
liquidlike.
A large realm of considerable challenge in nanoscale weathering research rests in explaining weathering forms such as
those displayed in Figure 26. A biotite grain that is undergoing
splintering has split into subparallel fragments as a result of
nanoscale weathering processes by water accumulating at
surface defects. The next generation of weathering researchers
will link nanoprocesses to micron-scale forms with technology
such as coupled dual-beam FIB-EM (Table 1).
Linking the nano- and micron scales will involve exceptionally different and hard-to-predict behavior such as
biologically generated particles. In many cases, nanobiominerals are more stable (Tang et al., 2004) than would be
expected of inorganic counterparts. In other cases, they
are less stable (Dorn, 1998; Krinsley, 1998) than inorganic
64
Figure 25 Lattice fringe HRTEM nanoscale imagery of geomorphologically abraded surfaces that display a polish in hand samples. Image (a) is
a cross section of a glacially polished surface of a plagioclase phenocryst collected from the terminus of the Khumbu Glacier, Nepal. Image (b) is
a cross section of polishing on a slickenside collected from an unnamed fault inside the Black Mountains near Badwater, Death Valley, California.
Figure 26 Weathering of biotite mineral, Adirondack Mountains,

New York. The subparallel fracturing called splintering is a result of
weathering that is taking place at the nanoscale.
counterparts, for reasons. Given the dominance of microorganisms on Earth (Reith, 2011), biologically generated particles will be a persistent issue in all aspects of nanoscale
weathering.
Further development of weathering theory as it applies
to understanding weathering forms needs to cross the
nanomicron threshold. Unfortunately, very little research
exists connecting nanoscale processes and geomorphic
weathering forms. For example, Hall (2006a: 388) argued

that little or no recognition has been given to either the
processes or the measuring of attributes influencing the
[freezethaw] processes or the measuring of attributes influencing the weathering processes, at the micro-scale, let alone
the nano-scale. Until nanoscale observations are woven into
the fabric of understanding basic processes such as frost
weathering, it would be premature to develop geomorphic
weathering theory further. We do not believe that there exists a
sufficient understanding to be able to link the nanomicron
threshold with a new theoretical understanding of landforms.
Instead, this chapter represents an attempt to provide empirical case studies exemplifying why nanoscale matters to
geomorphology.
We started by re-examining Goldichs (1938) classic weathering sequence, which argues that weathering is driven by
the thermodynamics of mineral crystallization. In Goldichs
sequence, olivine weathers first because it is most unstable in
the low-temperature of Earths surface. Although Goldichs
sequence has been confirmed in humid study sites with an
abundance of organisms and organic acid, some have found
that olivine weathers last not first at locales that have
experienced a minimum amount of organic activity (Wasklewicz et al., 1993; Wasklewicz, 1994). Support for a biotic explanation for olivine weathering, and hence the sequence of
expected mineral instability has been found in studies of
microorganisms (Barker, 1994; Barker and Banfield, 1996;
Song et al., 2010). We present here additional nanoscale evidence for the importance of microbial processes through a
study of the fungi in lichens where fungi uptake calcium from
olivine in nanoscale veins (Figure 6). The broader geomorphic implication of invalidating Goldichs (1938) general
sequence of mineral weathering rests in connecting mineral
weathering sequences to landscape geochemistry (Perelman,
1966; Fortescue, 1980) variability in the microscale and
nanoscale processes that appear to control the order of mineral weathering.

Nanoscale mineral etching could play a key role in determining weathering form at the hand-sample scale. An example
comes from the splintering weathering form of parallel fracturing (Figure 9). Splintering occurs in all rock types, even
those like basalt with millimeter-scale and micron-scale textures that do not appear to be well aligned with a subparallel
orientation. One explanation, supported by a pilot study of
splintered dolomite, is that nanoscale etching with a subparallel orientation allows micron-scale microfractures to extend with a subparallel geometry. These microfractures allow
the movement of capillary water that then further develops
these subparallel weaknesses into hand-scale splintered rock
surfaces. We also propose that if nanoscale etching does not
display a subparallel alignment, micron-scale microfractures
do not propagate with a subparallel orientation and splintering does not occur.
Nanoscale processes are important in the geomorphic
study of dust and dust deposition. Nanoscale defects are recognized as important in the production of silt (Smalley et al.,
2005; Kumar et al., 2006), along with external weathering
processes. Nanoscale topographic irregularities help dust adhere to mineral surfaces (Figure 16). We present here the first
nanoscale mapping of geochemical alterations of a dust particle, probably biologically generated, adhered to a rock
coating. Using a grid of 300 points 1.4 nm apart, we used EDS
to map Si, O, C, and P. Silicon and oxygen appear to be migrating into the bioparticle from an external silicate, while
phosphorus appears to be mobilizing from the outer to the
inner portions of the particle (Figure 19). If we are correct that
the particle has a biological origin, these findings present
evidence for ongoing instability of biomineral deposits even as
dust particles adhered to rock coatings.
Nanoscale processes help explain case hardening, an important reason why many rock surfaces remain stable
even as the internal portions of a weathered rock erode
away (Washburn, 1969; Conca and Rossman, 1982; Dorn,
2004). Case hardening occurs where the porosity in the weathering rind has been filled in with material, typically mobilized from the overlying rock coating (Dorn, 1998; Gordon
and Dorn, 2005a; Dorn, 2009). Silica is probably the most
common and hence most important agent of case hardening.
The first nanoscale study of silica glaze reveals that silica
precipitates as 2070 nm diameter spheroids, perhaps in response to crossing a transition between partial and complete
wetting that occurs between 20 and 70 nm (Zorin et al., 1992;
Churaev, 2003). One implication of these findings is that this
silica is quite mobile, moving from surficial deposits of silica
glaze down into pores in the underlying weathering rind
(Figure 20).
This chapter is necessarily filled with speculation, trying to
connect a few observations using HRTEM to research
problems in weathering geomorphology. For example, a logical speculation is that the silica in the weathering rind
could then be remobilized out of the weathering rind
where core softening results in a cavern. Cavernous weathering
forms typically have an outer shell of case-hardened rock
that erodes slower than the core-softened rock. Formation
of tafoni, for example, provides a gradient for capillary
65
water flow, and there is no reason why the silica precipitated

in weathering-rind pores (Figures 20 and 23) would not then
be remobilized again toward cavern side walls. Thus, the
nanoscale mobility and deposition of silica could be important in rock-coating formation, case hardening, and then
stabilization of the inside walls of a cavernous weathering
feature. The reason why such issues remain speculative
is because the general problem of explaining differential
weathering has not been a concern of geochemists who
have pioneered nanoscale weathering research, but of
geomorphologists.
The last theme of this chapter is the danger and potential
of crossing the nanomicron threshold. The danger is illustrated by the type of thinking illustrated by Garvie et al. (2008,
2009), who analyzed three samples of rock varnish and concluded that nanoscale instability invalidated all prior research
connecting rock varnish microlaminations to climatic
change research not based on three samples, but based on
observations of over 10 000 microlamination sequences (Liu
and Broecker, 2000; Liu et al., 2000; Zhou et al., 2000; Liu,
2003; Liu and Broecker, 2007, 2008a, 2008b; Liu, 2010).
Thinking that observations on three samples invalidate research based on three orders of magnitude, more samples
could be considered a prime case of scientific hubris. However,
there is nothing wrong scientifically with nanoscale data obtained and presented by Garvie et al. (2008, 2009). In fact,
prior research not mentioned by Garvie et al. (2008, 2009)
similarly found nanoscale instability of varnish constituents
(Figure 8) and concluded that this instability was a key in rock
varnish formation (Dorn, 1998; Krinsley, 1998; Dorn, 2007).
Rather, the fundamental danger illustrated by Garvie et al.
(2008, 2009) was in a false assumption that the micron-scale
processes associated with rock varnish microlaminations are
the same as the nanoscale processes. Nanoscale movement
does occur within laterally consistent varnish microlaminations, but this shifting does not invalidate the consistency of
climatically controlled pattern so extensively documented in
the literature (Liu and Broecker, 2000; Liu et al., 2000; Zhou
et al., 2000; Liu, 2003; Liu and Broecker, 2007, 2008a, 2008b;
Liu, 2010).
Despite the danger of assuming that nanoscale processes
operate at micron or higher scales, there is considerable
potential in crossing the nanomicron threshold. The potential is illustrated in Figure 13, where it is possible to connect
quantitative rates of dissolution from the nanoscale to the
hand-sample scale, and in Figures 21 and 22 where thermal
stresses from wildfires generate nanoscale erosion. BSE imagery at different scales has been successfully processed to
measure long-term rates of dissolution from field hand samples (Dorn, 1995; Dorn and Brady, 1995; Brady et al., 1999;
Gordon and Brady, 2002). Similar data could be collected
from HRTEM imagery and from samples originally analyzed
through BSE. Scaling up and down becomes an issue of
sampling procedure and the basic mechanics of preparing
micrographs for digital-image processing. The benefits of
making such a connection across scales have implications for
explaining discrepancies between field and laboratory findings
and also in developing a far deeper understanding of connections between geomorphology and landscape geochemistry (Perelman, 1966; Fortescue, 1980).
66
References
Allen, C.C., Oehler, D.Z., 2008. A case for ancient springs in Arabia Terra, Mars.
Astrobiology 8, 10931112.
Allison, R.J., Goudie, A.S., 1994. The effects of fire on rock weathering: an
experimental study. In: Robinson, D.A., Williams, R.B.G. (Eds.), Rock weathering
and landform evolution. Wiley, Chichester, pp. 4156.
Banfield, J.F., Barker, W.W., 1994. Direct observation of reactantproduct interfaces
formed in natural weathering of exsolved, defective amphibole to smectite:
evidence for episodic, isovolumetric reactions involving structural inheritance.
Geochimica Cosmochimica Acta 58, 14191429.
Banfield, J.F., Barker, W.W., Welch, S.A., Taunton, A., 1999. Biological impact on
mineral dissolution: application of the lichen model to understanding mineral
weathering in the rhizosphere. Proceedings National Academy of Sciences 96,
34043411.
Banfield, J.F., Eggleton, R.A., 1988. Transmission electron microscopy study of
biotite weathering. Clays and Clay Minerals 36, 4760.
Banfield, J.F., Eggleton, R.A., 1990. Analytical transmission electron microscope
studies of plagioclase, muscovite and K-feldsapr weathering. Clay and Clay
Minerals 38, 7789.
Baram, M., Chatain, D., Kaplan, W.D., 2011. Nanometer-thick equilibrium films: the
interface between thermodynamics and atomistics. Science 332, 206209.
Barker, W.W., 1994. Biogeochemical weathering of hornblende syenite. EOS 74,
281.
Barker, W.W., Banfield, J.F., 1996. Biologically versus inorganically mediated
weathering reactions: relationships between minerals and extracellular microbial
polymers in lithobiontic communities. Chemical Geology 132, 5569.
Bennett, P.C., Rogers, J.R., Choi, W.J., Hiebert, F.K., 2001. Silicates, silicate
weathering, and microbial ecology. Geomicrobiology Journal 18, 319.
Benzerara, K., Menguy, N., Guyot, F., Vanni, C., Gillet, P., 2005a. TEM study of a
silicatecarbonatemicrobe interface prepared by focused ion beam milling.
Geochimica Cosmochimica Acta 69, 14131422.
Benzerara, K., Yoon, T.H., Menguy, N., Tyliszczak, T., Brown, G.E., 2005b.
Nanoscale environments associated with bioweathering of a MgFepyroxene.
Proceedings of the National Academy of Sciences 102, 979982.
Bishop, J.L., Murchie, S.L., Pieters, C.M., Zent, A.P., 2002. A model for formation
of dust, soil, and rock coatings on Mars: physical and chemical processes on
the Martian surface. Journal of Geophysical Research 107(E11), http://dx.doi.org
/10.1029/2001JE001581.
Bjelland, T., Thorseth, I.H., 2002. Comparative studies of the lichenrock
interface of four lichens in Vingen, western Norway. Chemical Geology 192,
8198.
Bonneville, S., Smits, M.M., Brown, A., Harrington, J., Leake, J.R., Brydson, R.,
Benning, L.G., 2009. Plant-driven fungal weathering: early stages of mineral
alteration at the nanometer scale. Geology 37, 615618.
Boston, P.J., Spilde, M.N., Northup, D.E., Dichosa, A., 2008. Biogenic Fe/Mn
oxides in caves and surface desert varnish: potential biosignatures for Earth and
Mars. Astrobiology 8, 448.
Boynton, W.V., Ming, D.W., Kounaves, S.P., et al., 2009. Evidence for calcium
carbonate at the Mars Phoenix Landing Site. Science 325, 6164.
Brady, P.V., Dorn, R.I., Brazel, A.J., Clark, J., Moore, R.B., Glidewell, T., 1999.
Direct measurement of the combined effects of lichen, rainfall, and
temperature on silicate weathering. Geochimica et Cosmochimica Acta 63,
32933300.
Brandt, F., Bosbach, D., Krawczyk Barsch, E., Arnold, T., Bernhard, G., 2003.
Chlorite dissolution the acid pH-range: a combined microscopic and
macroscopic approach. Geochimica Cosmochimica Acta 7, 14511462.
Brantley, S.L., 2005. Reaction kinetics of primary rock-forming minerals under
ambient conditions. In: Drever, J.I. (Ed.), Surface and ground water, weathering,
and soils. Treastise on Geochemistry. Elsevier, Amsterdam, vol. 5, pp. 73117.
Brantley, S.L., Mellott, N.P., 2000. Surface area and porosity of primary silicate
minerals. American Mineralogist 85, 17671783.
Brantley, S.L., Velbel, M.A., 1993. Preface to special issue on geochemical kinetics
of mineralwater reactions in the field and the laboratory. Chemical Geology
105, viiix.
Brehm, U., Gorbushina, A.A., Mottershead, D.N., 2005. The role of microorganisms
and biofilms in the breakdown and dissolution of quartz and glass.
Palaeogeography, Palaeoclimatology, Palaeoecology 219, 117129.
Brown, D.J., Lee, M.R., 2007. From microscopic minerals to global climate change?
Geology Today 23(5), 172177.
Bullard, J.E., Livingston, I., 2009. Dust. In: Parsons, A.J., Abrahams, A.D. (Eds.),
Geomorphology of Desert Environments. Springer, New Work, NY,
pp. 629654.
Buss, H.L., Luttge, A., Brantley, S.L., 2007. Etch pit formation on iron silicate
surfaces during sideophore-promoted dissolution. Chemical Geology 240,
326342.
Campbell, E.M., Twidale, C.R., 1991. The evolution of bornhardts in silicic volcanic
rocks in the Gawler Ranges. Australian Journal of Earth Sciences 38, 7993.
Casey, W.H., Banfield, J.F., Westrich, H.R., McLaughlin, L., 1993. What do
dissolution experiments tell us about natural weathering? Chemical Geology
105, 115.
Christenson, H.K., 2001. Confinement effects on freezing and melting. Journal of
Physics: Condensed Matter 13, R95R133.
Churaev, N.V., 2003. Surface forces in wetting films. Advances in Colloid and
Interface Science 103, 197218.
Conca, J.L., Rossman, G.R., 1982. Case hardening of sandstone. Geology 10,
520525.
DiGregorio, B.E., 2002. Rock varnish as a habitat for extant life on Mars.
In: Hoover, R.B., Levin, G.V., Paepe, R.R., Rozanov, A.Y. (Eds.), Instruments,
Methods, and Missions for Astrobiology IV. NASA, SPIE, Washington, DC, Vol.
4495, pp. 120130.
alpine environments. Geomorphology 67, 127145.
Dixon, J.C., Thorn, C.E., Darmody, R.G., Campbell, S.W., 2002. Weathering rinds
and rock coatings from an Arctic alpine environment, northern Scandinavia.
Dorn, R.I., 1995. Digital processing of back-scatter electron imagery: a microscopic
approach to quantifying chemical weathering. Geological Society of America
Bulletin 107, 725741.
Dorn, R.I., 1998. Rock coatings. Elsevier, Amsterdam, 429 pp.
Dorn, R.I., 2003. Boulder weathering and erosion associated with a wildfire, Sierra
Ancha Mountains, Arizona. Geomorphology 55, 155171.
Dorn, R.I., 2004. Case hardening. In: Goudie, A.S. (Ed.), Encyclopedia of
Dorn, R.I., 2007. Rock varnish. In: Nash, D.J., McLaren, S.J. (Eds.), Geochemical
Sediments and Landscapes. Blackwell, London, pp. 246297.
Dorn, R.I., 2009. Desert rock coatings. In: Parsons, A.J., Abrahams, A. (Eds.),
Geomorphology of desert environments. Springer, Amsterdam, pp. 153186.
Dorn, R.I., 2012. Formation of silica glaze rock coatings through water vapor
interactions. Physical Geography 33(1), 2131.
Dorn, R.I., Brady, P.V., 1995. Rock-based measurement of temperature-dependent
plagioclase weathering. Geochimica et Cosmochimica Acta 59, 28472852.
Dorn, R.I., Krinsley, D.H., 2011. Spatial, temporal and geographic considerations of
the problem of rock varnish diagenesis. Geomorphology 130, 9199.
Dorn, R.I., Whitley, D.S., Cerveny, N.C., Gordon, S.J., Allen, C., Gutbrod, E., 2008.
The rock art stability index: a new strategy for maximizing the sustainability of
rock art as a heritage resource. Heritage Management 1, 3570.
Eggleton, R.A., 1980. High resolution electron microscopy of feldspar weathering.
Clays and Clay Minerals 28, 173178.
Emmanuel, S., Ague, J.J., 2011. Impact of nano-size weathering products on
dissolution rates of primary minerals. Chemical Geology 282, 1118.
Emmanuel, S., Ague, J.J., Walderhaug, O., 2010. Interfacial energy effects and the
evolution of pore size distributions during quartz precipitation in sandstone.
Geochimica et Cosmochimica Acta 74, 35393552.
Farr, T., Adams, J.B., 1984. Rock coatings in Hawaii. Geological Society of America
Bulletin 95, 10771083.
Fitzner, B., Heinrichs, K., 2004. Photo atlas of weathering forms on stone
monuments. http://www.stone.rwth-aachen.de/ (accessed June 2011).
Fitzner, B., Heinrichs, K., Kownatski, R., 1997. Weathering forms at natural stone
monuments classification, mapping and evaluation. International Journal for
Restoration of Buildings and Monuments 3(2), 105124.
Fortescue, J.A.C., 1980. Environmental geochemistry. A holistic approach. Springer,
New York, NY, 347 pp.
Gaier, J.R., 2005. The Effects of Lunar Dust on EVA Systems during the Apollo
Missions. NASA Center for Aerospace Information, Hanover, MD. NASA/
TMF2005-213610.
Ganor, E., Kronfeld, J., Feldman, H.R., Rosenfeld, A., Ilani, S., 2009. Environmental
dust: a tool to study the patina of ancient artifacts. Journal of Arid Environments
73, 11701176.
Garvie, L.A.J., Burt, D.M., Buseck, P.R., 2008. Nanometer-scale complexity, growth,
and diagenesis in desert varnish. Geology 36, 215218.
Garvie, L.A.J., Burt, D.M., Buseck, P.R., 2009. A microscopists view of desert varnish
from the Sonoran Desert. Lunar and Planetary Science Conference 40, 1344.
Gleeson, D.B., Clipson, N., Melville, K., Gadd, G.M., McDermott, F.P., 2005.
Characterization of fungal community structure on a weathered pegmatitic
granite. Microbial Ecology 50, 360368.
Gleeson, D.B., Kennedy, N.M., Clipson, N., Melville, K., Gadd, G.M., McDermott,
F.P., 2006. Characterization of bacterial community structure on a weathered
pegmatitic granite. Microbial Ecology 51, 526534.
Goldich, S.S., 1938. A study of rock weathering. Journal of Geology 46, 1758.
Gordon, S.J., Brady, P.V., 2002. In situ determination of long-term basaltic glass
dissolution in the unsaturated zone. Chemical Geology 90, 115124.
Gordon, S.J., Dorn, R.I., 2005a. In situ weathering rind erosion. Geomorphology
67, 97113.
Gordon, S.J., Dorn, R.I., 2005b. Localized weathering: implications for theoretical
and applied studies. Professional Geographer 57, 2843.
Goudie, A.S., 1978. Dust storms and their geomorphological implications. Journal
of Arid Environments 1, 291310.
Greeley, R., 1987. Desert varnish; Earth, Mars, and Venus. International Union of
Geodesy and Geophysics (IUGG), XIX general assembly; abstracts 1, 96.
Hall, K., 2006a. Monitoring of thermal conditions in building stone with particular
reference to freezethaw events. In: Kourkoulis, S.K. (Ed.), Fracture and Failure
of Natural Building Stones. Springer, Amsterdam, pp. 373394.
Hall, K., 2006b. Perceptions of rock weathering in cold regions: a discussion on
space and time attributes of scale. Geomorphologie: relief, processus,
environnement 3, http://geomorphologie.revues.org/index146.html (accessed
June 2011).
Hall, R.D., Michaud, D., 1988. The use of hornblende etching, clast weathering and
soils to date alpine glacial and periglacial deposits: a study from southwestern
Montana. Geological Society of America Bulletin 100, 458467.
Han, R., Hirose, T., Shimamoto, T., Lee, Y., Ando, J., 2011. Granular nanoparticles
lubricate faults during seismic slip. Geology 39, 599602.
Hiebert, F.K., Bennett, P.C., 1992. Microbial control of silicate weathering in
organic-rich ground water. Science 258, 278281.
Hochella, M.F., 2002a. Nanoscience and technology: the next revolution in the Earth
sciences. Earth and Planetary Science Letters 203, 593605.
Hochella, M.F., 2002b. Theres plenty of room at the bottom: nanoscience in
geochemistry. Geochimica Cosmochimica Acta 66, 735743.
Hochella, M.F., Banfield, J.F., 1995. Chemical weathering of silicates in nature: a
microscopic perspective with theoretical considerations. In: White, A.F., Brantley,
S.L. (Eds.), Chemical weathering rates of silicate minerals. Mineralogical Society
of America, Washington, DC, pp. 353406.
Hochella, M.F., Lower, S.K., Maurice, P.A., Penn, R.L., Sahai, N., Sparks, D.L.,
Twining, B.S., 2008. Nanominerals, mineral nanoparticles, and earth systems.
Science 319, 16311635.
Israel, E.J., Arvidson, R.E., Wang, A., Pasteris, J.D., Jolliff, B.L., 1997. Laser Raman
spectroscopy of varnished basalt and implications for in situ measurements of
Martian rocks. Journal of Geophysical Research Planets 102(E12),
2870528716.
Johnson, J.R., Christensen, P.R., Lucey, P.G., 2002. Dust coatings on basaltic rocks
and implications for thermal infrared spectroscopy of Mars. Journal of
Geophysical Research Planets 107(E6), Art No. 5035.
Jordan, D.W., 1954. The adhesion of dust particles. British Journal of Applied
Physics 5, S194S198.
Kalinichev, A.G., Wang, J., Kirkpatrick, R.J., 2007. Molecular dynamics modeling of
the structure, dynamics and energetics of mineralwater interfaces: applications
to cement materials. Cement and Concrete Research 37, 337347.
Kaufhold, S., Kaufhold, A., John, R., et al., 2009. A new massive deposit of
allophane raw material in Ecuador. Clays and Clay Minerals 57, 7281.
Kleber, A., 1997. Cover-beds as soil parent materials in midlatitude regions. Catena
30, 197213.
Koopal, L.K., Goloub, T., de Keizer, A., Sidorova, M.P., 1999. The effect of cationic
surfactants on wetting, colloid stability and flotation of silica. Colloids and
Surfaces A: Physicochemical and Engineering Aspects 151, 1525.
Kraft, M.D., Greeley, R., 2000. Rock coatings and aeolian abrasion on Mars:
application to the Pathfinder landing site. Journal of Geophysical Research
Planets 105(E6), 1510715116.
Kraft, M.D., Michalski, J.R., Sharp, T.G., 2004. High-silica rock coatings: TES
surface-type 2 and chemical weathering on Mars. Lunar and Planetary Science
35, 1936.
Krinsley, D., 1998. Models of rock varnish formation constrained by high resolution
transmission electron microscopy. Sedimentology 45, 711725.
Krinsley, D., Dorn, R.I., Anderson, S., 1990. Factors that may interfere with the
dating of rock varnish. Physical Geography 11, 97119.
Krinsley, D., Dorn, R.I., DiGregorio, B.E., 2009. Astrobiological implications of rock
varnish in Tibet. Astrobiology 9, 551562.
Krinsley, D.H., Dorn, R.I., Tovey, N.K., 1995. Nanometer-scale layering in rock
varnish: implications for genesis and paleoenvironmental interpretation. Journal
of Geology 103, 106113.
67
Kumar, R., Jefferson, I.F., OHara-Dhand, K., Smalley, I.J., 2006. Controls on quartz
silt formation by crystalline defects. Naturwissenschaften 93, 185188.
Langworthy, K., Krinsley, D., Dorn, R.I., 2010. High resolution transmission electron
microscopy evaluation of silica glaze reveals new textures. Earth Surface
Lee, M.R., 2010. Transmission electron microscopy (TEM) of Earth and planetary
materials: a review. Mineralogical Magazine 74, 127.
Lee, M.R., Brown, D.J., Smith, C.L., Hodson, M.E., Mackenzie, M., Hellman, R.,
2007. Characterization of mineral surfaces using FIB and TEM: a case study of
naturally-weathered alkali feldspars. American Mineralogist 92, 13831394.
Lee, M.R., Parsons, I., 1995. Microtextural controls of weathering of perthitic alkali
feldspars. Geochimica et Cosmochimica Acta 59, 44654488.
de Leeuw, N.H., Higgins, F.M., Parker, S.C., 1999. Modeling the surface structure
and stability of a-quartz. Journal of Physical Chemistry B 103, 12701277.
Liu, T., 2003. Blind testing of rock varnish microstratigraphy as a chronometric
indicator: results on late Quaternary lava flows in the Mojave Desert, California.
Liu, T., 2010. VML Dating Lab. http://www.vmldating.com/ (accessed June 2011).
Liu, T., Broecker, W.S., 2000. How fast does rock varnish grow? Geology 28,
183186.
Liu, T., Broecker, W.S., 2007. Holocene rock varnish microstratigraphy and its
chronometric application in drylands of western USA. Geomorphology 84, 121.
Liu, T., Broecker, W.S., 2008a. Rock varish evidence for latest Pleistocene
millennial-scale wet events in the drylands of western United States. Geology
36, 403406.
Liu, T., Broecker, W.S., 2008b. Rock varnish microlamination dating of late
Quaternary geomorphic features in the drylands of the western USA.
Liu, T., Broecker, W.S., Bell, J.W., Mandeville, C., 2000. Terminal Pleistocene wet
event recorded in rock varnish from the Las Vegas Valley, southern Nevada.
Locke, W.W., 1979. Etching of Hornblende Grains in Arctic Soils: an indicator of
relative age and paleoclimate. Quaternary Research 11, 197212.
Locke, W.W., 1986. Rates of Hornblende Etching in Soils on Glacial Deposits.
Baffin Island, Canada. Academic Press, New York, NY, pp. 129145.
Lower, S.K., Hochella, M.F., Beveridge, T.J., 2001. Bacterial recognition of mineral
surfaces: nanoscale interactions between Shewanella and alpha-FeOOH. Science
292, 13601363.
Mahaney, W.C., Milner, M.W., 2011. Lightning-induced changes in red pine (Pinus
resinosa). Palaeogeography, Palaeoclimatology, Palaeoecology 309, 367373.
McAlister, J.J., Smith, B.J., Torok, A., 2006. Element partitioning and potential
mobility within surface dusts on buildings in a polluted urban environment,
Budapest. Atmospheric Environment 40, 67806790.
McKeown, D.A., Post, J.E., 2001. Characterization of manganese oxide mineralogy
in rock varnish and dendrites using X-ray absorption spectroscopy. American
Mineralogist 86, 701713.
McSween, Jr. H.Y., Murchie, S.L., Britt, D., et al., 1999. Chemical, multispectral,
and textural constraints on the composition and origin of rocks at the
Mars Pathfinder landing site. Journal of Geophysical Research 104(E4),
86798716.
Meunier, A., Sardini, P., Robinet, J.C., Pret, D., 2007. The petrography of
weathering processes: facts and outlooks. Clay Minerals 42, 415435.
Morel, B., Autissier, L., Autissier, D., Lemordant, D., Yrieix, B., Quenard, D., 2009.
Pyrogenic silica ageing under humid atmosphere. Powder Technology 190,
225229.
Moss, A.J., Green, P., 1998. Sand and silt graints: predetermination of their
formation and properties by microfractures in quartz. Australian Journal of Earth
Mottershead, D.N., Pye, K., 1994. Tafoni on coastal slopes, South Devon, U.K.
Noble, S.K., Keller, L.P., 2006. Investigating the nanoscale complexity of lunar
space weathering. Geochimica Cosmochimica Acta Goldschmidt Conference
Abstracts, A448. http://dx.doi.org/10.1016/j.gca.2006.06.900.
Noguchi, T., Nakamura, T., Kimura, M., et al., 2011. Incipient space weathering
observed on the surface of Itokawa dust particles. Science 333, 11211125.
Paradise, T.R., 1995. Sandstone weathering thresholds in Petra, Jordan. Physical
Perelman, A.I., 1966. Landscape Geochemistry. Vysshaya Shkola, Moscow, 388 pp.
(Translation No. 676, Geological Survey of Canada, 1972).
Phillips, J.D., 2000. Signatures of divergence and self-organization in soils and
weathering profiles. Journal of Geology 108, 91102.
68
Pope, G., Dorn, R.I., Dixon, J., 1995. A new conceptual model for understanding
geographical variations in weathering. Annals of the Association of American
Pope, G.A., 1995. Newly discovered submicron-scale weathering in quartz:
geographical implications. Professional Geographer 47, 375387.
Pope, G.A., Meierding, T.C., Paradise, T.R., 2002. Geomorphologys role in the
study of weathering of cultural stone. Geomorphology 47, 211225.
Potter, R.M., 1979. The tetravalent manganese oxides: clarification of their structural
variations and relationships and characterization of their occurrence in the
terrestrial weathering environment as desert varnish and other manganese
oxides. PhD Dissertation. California Institute of Technology, Pasadena.
Potter, R.M., Rossman, G.R., 1977. Desert varnish: the importance of clay minerals.
Science 196, 14461448.
Pye, K., 1987. Aeolian Dust and Dust Deposits. Academic Press, London.
Pye, K., 1989. Processes of Fine Particle Formation. Dust Source Regions, and
Climatic Change. Kluwer Academic Publishers, Dordrecht, 330.
Reith, F., 2011. Life in the deep subsurface. Geology 39, 287288.
Rimer, J.D., Trofymluk, O., Lobo, R.F., Navrotsky, A., Vlachos, D.G., 2008.
Thermodynamics of silica nanoparticle self-assembly in basic solution of
monovalent cations. Journal of Physical Chemistry C 112, 1475414761.
Rimer, J.D., Trofymluk, O., Navrotsky, A., Lobo, R.F., Vlachos, D.G., 2007. Kinetic
and thermodynamic studies of silica nanoparticle dissolution. Chemistry of
Materials 19, 41894197.
Robinson, D.A., Williams, R.B.G., 1992. Sandstone weathering in the High Atlas,
Morocco. Zeitschrift fur Geomorphologie 36, 413429.
Schiffbauer, J.D., Xiao, S., 2009. Novel application of focused ion beam electron
microscopy (FIB-EM) in preparation and analysis of microfossil ultrastructures: a
new view of complexity in early eukaryotic organisms. Palaios 24, 616626.
Schleicher, A.M., van der Pluijm, B.A., Warr, L.N., 2010. Nano-coatings of clay and
creep of the San Andreas Fault at Parkfield, California. Geology 38 (7),
667670.
Sharma, R.K., Gupta, H.O., 1993. Dust pollution at the Taj Mahal a case study.
In: Thiel, M.-J. (Ed.), Conservation of Stone and Other Materials. E & FN Spon,
London, pp. 1118.
Smalley, I.J., Krinsley, D.H., 1978. Loess deposits associated with deserts. Catena
5, 5366.
Smalley, I.J., Kumar, R., Dhand, K.O., Jefferson, I.F., Evans, R.D., 2005. The
formation of silt material for terrestrial sediments: particularly loess and dust.
Smith, B.J., McAlister, J.J., Baptista-Neto, J.A., Silva, M.A.M., 2007. Postdepositional modification of atmospheric dust on a granite building in central
Rio de Janeiro: implications for surface induration and subsequent stone decay.
Geological Society, London, Special Publications 271, 153166,.
Smith, P.H., Tamppari, L.K., Arvidson, R.E., et al., 2009. H2O at the Phoenix landing
site. Science 325, 5861.
Smits, M.M., Herrmann, A.M., Duane, M., Duckworth, O.W., Bonneville, S.,
Benning, L.G., Lundstrom, U., 2009. The fungalmineral interface: challenges
and considerations of micro-analytical developments. Fungal Biology Review 23,
122131.
Song, W., Ogawa, N., Oguichi, C.T., Hatta, T., Matsukura, Y., 2010. Laboratory
experiments on bacterial weathering of granite and its constituent minerals.
Geomorphologie, Relief, Processus, Environnement 4, 327336.
Stoppato, M.C., Bini, A., 2003. Deserts. Firely Books, Buffalo, 256 pp.
Swoboda-Colberg, N.G., Drever, J.I., 1993. Mineral dissolution rates in plot-scale
field and laboratory experiments. Chemical Geology 105, 5169.
Tang, R., Wang, L., Orme, C.A., Bonstein, T., Bush, P.J., Nancollas, G.H., 2004.
Dissolution at the nanoscale: self-preservation of biominerals. Angewandte
Chemie 43, 26972701.
Tratebas, A.M., Cerveny, N., Dorn, R.I., 2004. The effects of fire on rock art:
microscopic evidence reveals the importance of weathering rinds. Physical
Turkington, A.V., Paradise, T.R., 2005. Sandstone weathering: a century of research

and innovation. Geomorphology 67, 229253.
Turkington, A.V., Phillips, J.D., Campbell, S.W., 2005. Weathering and landscape
evolution. Geomorphology 67, 16.
Twidale, C.R., 2002. The two-stage concept of landform and landscape development
involving etching: origin, development and implications of an idea. Earth
Van der Giessen, E., Needleman, A., 2002. Micromechanics simulations of fracture.
Annual Review of Material Research 32, 141162.
Velbel, M.A., 1993. Constancy of silicate-mineral weathering-rate ratios between
natural and experimental weathering: implications for hydrologic control of
differences in absolute rates. Chemical Geology 105, 8999.
Wadsten, T., Moberg, R., 1985. Calcium oxalate hydrates on the surface of lichens.
Lichenologist 17, 239245.
Wang, J.Z., Kalinichev, A.G., Kirkpatrick, R.J., 2006. Effects of substrate structure
and composition on the structure, dynamics, and energetics of water at mineral
surfaces: a molecular dynamics modeling study. Geochimica Cosmochimica Acta
70, 562582.
Washburn, A.L., 1969. Case hardening. In: Washburn, A.L. (Ed.), Weathering, Frost
Action and Patterned Ground in the Mesters District, Northeast Greenland.
Reitzels, Copenhagen, 15 pp.
Wasklewicz, T., 1994. Importance of environment on the order of mineral weathering
in olivine basalts, Hawaii. Earth Surface Processes and Landforms 19, 715735.
Wasklewicz, T., Dorn, R.I., Clark, S., et al., 1993. Olivine does not necessarily
weather first. Journal of Tropical Geography 14, 7280.
Waychunas, G.A., Kim, C.S., Banfield, J.F., 2005. Nanoparticle iron oxide minerals
in soils and sediments: unique properties and contaminant scavenging
mechanisms. Journal of Nanoparticle Research 7, 409433.
Wei, H., Sparks, R.S.J., Liu, R., et al., 2003. Three active volcanoes in China and
their hazards. Journal of Asian Earth Sciences 21, 515526.
Whalley, W.B., Marshall, J.R., Smith, B.J., 1982. Origin of desert loess from some
experimental observations. Nature 300, 433435.
White, A.F., 2005. Natural Weathering Rates of Silicate Minerals. Elsevier,
Whiteway, J.A., Komguem, L., Dickinson, C., et al., 2009. Mars water-ice clouds
and precipitation. Science 325, 6870.
Wilhelmy, H., 1964. Cavernous rock surfaces in semi-arid and arid climates.
Pakistan Geographical Review 19(2), 813.
Wright, J., Smith, B., Whalley, W.B., 1998. Mechanisms of loess-sized quartz silt
production and their relative effectiveness: laboratory simulations.
Geomophology 23, 1534.
Xu, K., Cao, P., Heath, J.R., 2010. Graphene visualizes the first water adlayers on
mica at ambient conditions. Science 329, 11881191.
Yaalon, D.H., Ganor, E., 1973. The influence of dust on the soils during the
Quaternary. Soil Science 116, 146155.
Young, R., Young, A.R.M., 1992. Sandstone Landforms. Springer, Berlin.
Young, R.W., 1988. Quartz etching and sandstone karst: examples from the East
Kimberleys, Northwestern Australia. Zeitschrift fur Geomorphologie 32, 409423.
Zhang, H., Gilbert, B., Huang, F., Banfield, J.F., 2003. Water-driven structure
transformation in nanoparticles at room temperature. Nature 424, 10251029.
Zhou, B.G., Liu, T., Zhang, Y.M., 2000. Rock varnish microlaminations from
northern Tianshan, Xinjiang and their paleoclimatic implications. Chinese
Science Bulletin 45, 372376.
Zhu, C., Veblen, D.R., Blum, A.E., Chipera, S.J., 2006. Naturally weathered feldspar
surfaces in the Navajo Sandstone aquifer, Black Mesa, Arizona: electron
microscopic characterization. Geochimica Cosmochimica Acta 70, 46004616.
Zorin, Z.M., Churaev, N., Esipova, N., Sergeeva, I., Sobolev, V., Gasanov, E., 1992.
Influence of cationic surfactant on the surface charge of silica and on the
stability of aqueous wetting films. Journal of Colloid and Interface Science 152,
170182.
69
Biographical Sketch
Ronald I Dorn has been a professor of geography at Arizona State University, Tempe, Arizona since 1988. He
served previously on the faculty at Texas Tech University. He is co-coordinator of the Arizona Geographic Alliance,
a K-12 outreach program to promote geographic education in Arizona. He has been president and secretary/
tresurer of the Geomorphology Specialty Group and as chair of the Nystrom Committee of the Association of
American Geographers. He is a fellow of the Geological Society of America and the Arizona/Nevada Academy of
Science, a Guggenheim Fellow, and a member of the Association of American Geographers and the American Rock
Art Research Association.
Steven J Gordon is an associate professor of geosciences at the United States Air Force Academy. His specialties are
geomorphology, spatial variations in rock weathering, and geographic information systems (GIS). Steve graduated
from Arizona State University and has published extensively in the area of mineral weathering at the micron scale.
David Krinsley served in the Army during WWII. After that, he attended the University of Chicago where he
received his PhD in geology. Following that, he took up a postdoctoral position at Columbia University. He then
transferred to the Queens College, City University of New York, where he served as chairman of the Department of
Geology, dean of sciences and acting provost (acting dean of the faculty). Four years were spent at the University
of Cambridge, UK, where he was a visiting professor in the Department of Geology and an oversea fellow at
Churchill College, University of Cambridge, UK. Leaving Queens College, he assumed the Chair at Arizona State
University in Geology; upon retiring he moved to the Department of Geological Sciences at The University of
Oregon, where he has been doing research in sedimentation and nanotechnology.
Kurt Langworthy is the director of nanofabrication and electron imaging at the Center for Advanced Materials
Characterization in Oregon (CAMCOR). He has over 8 years of experience with transmission electron microscopy
(TEM), scanning electron microscopy (SEM), and focused ion beam (FIB) instrumentation. Kurt graduated in
2006 from the University of Oregon where he achieved a masters degree in chemistry.
4.5 Rock Coatings

RI Dorn, Arizona State University, Tempe, AZ, USA
4.5.1
4.5.2
4.5.2.1
4.5.2.2
4.5.2.3
4.5.2.4
4.5.2.5
4.5.3
4.5.4
References
Introduction to Rock Coatings

Interpreting Rock Coatings through a Landscape Geochemistry Approach
First-Order Control: Geomorphic Stability
Second-Order Control: Subaerial Exposure of Subsurface Coatings
Third-Order Control: Competition from Lithobionts
Fourth-Order Control: Transport Pathways
Fifth-Order Control: Barriers to Transport
Importance of Rock Coatings in Geomorphology
Conclusion
Glossary
Carbonate skin Coating composed primarily
of carbonate, usually CaCO3, but sometimes
MgCO3.
Case hardening Addition of cementing agent to rock
matrix material; the agent may be manganese, sulfate,
carbonate, silica, iron, oxalate, organisms, or
anthropogenic.
Dust film Light powder of clay- and silt-sized particles
attached to rough surfaces and in rock fractures.
Heavy metal skins Coatings of iron, manganese, copper,
zinc, nickel, mercury, lead, and other heavy metals on rocks
in natural and human-altered settings.
Iron film Coating composed primarily of iron oxides or
oxyhydroxides.
Lithobiontic coatings Organisms forming rock coatings,
for example, lichens, moss, fungi, cyanobacteria, and algae.
Nitrate crust Potassium and calcium nitrate coatings on
rocks, commonly in caves and rock shelters in limestone
areas.
70
71
73
73
78
81
84
88
90
94
Oxalate crust Mostly calcium oxalate and silica with

variable concentrations of magnesium, aluminum,
potassium, phosphorus, sulfur, barium, and manganese.
Commonly found forming near or with lichens.
Phosphate skin Various phosphate minerals (e.g., iron
phosphates or apatite) sometimes mixed with clays and
sometimes manganese.
Pigment Human-manufactured material placed on rock
surfaces by people.
Rock varnish Clay minerals, Mn and Fe oxides, and minor
and trace elements; color ranges from orange to black in
color produced by variable concentrations of different
manganese and iron oxides.
Salt crust Chloride precipitates formed on rock surfaces.
Silica glaze Coating that is generally clear white to orange
shiny luster, but can be darker in appearance, composed
primarily of amorphous silica and aluminum, but
commonly with iron.
Sulfate crust Sulfates (e.g., barite and gypsum) on rocks;
not gypsum crusts that are sedimentary deposits.
Abstract
Fourteen different types of coatings cover rock surfaces in every terrestrial weathering environment, altering the appearance of
the underlying landform. Some accretions interdigitate, whereas others blend together, creating a great number of variations.
Rock coatings are important in geomorphology because coatings: alter weathering rates; play a role in case-hardening surfaces;
offer clues to understanding environmental change; and can provide chronometric insight into the exposure of the underlying
rock surface. Following a landscape geochemistry paradigm, five general hierarchies of control explain the occurrence of
different types of rock coatings: first order geomorphic processes control the stability of bedrock surfaces on which coatings
form; second order coatings originating in rock fissures occur on subaerial surfaces when erosion of the overlying rock
occurs; third order the habitability of surfaces for fast-growing lithobionts such as lichens determines whether slowly
accreting coatings occur; fourth order the raw ingredients must have a transport pathway to the rock surface, and of course,
they must be present; fifth order physical, geochemical, or biological barriers to transport then result in the accretion of the
coating.
Dorn, R.I., 2013. Rock coatings. In: Shroder, J. (Editor in Chief), Pope, G.A.,
Weathering and Soils Geomorphology, pp. 7097.
70
http://dx.doi.org/10.1016/B978-0-12-374739-6.00066-X
Rock Coatings
4.5.1
Introduction to Rock Coatings
Bare rock surfaces rarely display a lithologys true appearance,

where rhyolite appears pink, gneiss displays dark and light
banding or basalt is black. As surfaces accrete rock coatings
(Table 1), appearances can change dramatically. Consider
three examples. The sandstone of Petra, Jordan, is generally
darkened by case hardening caused by the accumulation of
iron and manganese in the upper millimeter of sandstone
(upper row in Figure 1). The light-colored granodiorite at
Stone Mountain, Georgia, is streaked by several different types
of rock coatings, including calcium oxalate (middle row in
Figure 1). Black basalt lava flows on the rainshadow side of
Hualalai and Mauna Loa volcanoes, Hawaii, are gradually
lightened in color as silica glaze accretes on lava flow surfaces
(lower row in Figure 1). The geomorphic reality is that coated
surfaces are far more common than uncoated rocks. Because
these accretions influence the weathering of the underlying
rock, a full understanding of the terrestrial weathering environment must consider rock coatings.
This chapter organizes rock coatings through a paradigm of
landscape geochemistry, as developed by Soviet geography
(Polynov, 1937; Perelman, 1961, 1966; Glazovskaya, 1968,
1973) with some adoption outside of Russia (Fortescue,
1980). This theoretical framework provides a way by which
rock coatings can be interpreted in terms of element abundance, element migration, geochemical flows, geochemical
gradients, and geochemical barriers with the classification,
interpretation, and spatial laws pertaining to geochemical
71
landscapes. At its simplest, a landscape geochemical approach

interprets the occurrence of rock coatings as being caused by
physical, chemical, or biological barriers to the transport of
elements. This chapter interprets rock coatings from the perspective of five different landscape geochemistry hierarchies of
controls on rock-coating formation.
4.5.2
Interpreting Rock Coatings through a

Landscape Geochemistry Approach
What influences the formation of a rock coating? Phrased more

empirically, what processes generate silica glaze on one surface,
case hardening on another, and oxalate crusts on a third as
exemplified by the three different types of rock coatings seen in
Figure 1. The answer characteristically ends up being an encyclopedic approach of analyzing unique site-specific factors.
In the case of case-hardened rock at Petra (upper row in
Figure 1), rock varnish once formed on a sandstone surface.
Manganese and iron were leached out of the varnish and
reprecipitated inside the pore spaces of the sandstone.
Granular disintegration of the sandstone surface eroded much
of the varnish, but case hardening by heavy metals including
manganese stabilized the outer millimeter. Another coating is
the oxalate crust on Stone Mountain, Georgia (middle row in
Figure 1); this oxalate, composed primarily of hydrated calcium oxalate (CaC2O4 H2O), accreted about a meter downslope from a patch of lichens. Overland flow carried oxalate
from the lichens, and evaporation assisted in the accretion of
Table 1 Major types of rock coatings

Coating
Description
Related terms
Carbonate
skin
Case
hardening
Dust film
Composed primarily of carbonate, usually CaC03, but sometimes MgCO3
Calcrete, travertine
Addition of cementing agent to rock matrix material; the agent may be manganese,
sulfate, carbonate, silica, iron, oxalate, organisms, or anthropogenic.
Light powder of clay- and silt-sized particles attached to rough surfaces and in rock
fractures.
Coatings of iron, manganese, copper, zinc, nickel, mercury, lead and other heavy
metals on rocks in natural and human-altered settings.
Composed primarily of iron oxides or oxyhydroxides.
Organisms forming rock coatings, for example, lichens, moss, fungi, cyanobacteria,
algae.
Potassium and calcium nitrate coatings on rocks, often in caves and rock shelters in
limestone areas.
Mostly calcium oxalate and silica with variable concentrations of magnesium,
aluminum, potassium, phosphorus, sulfur, barium, and manganese. Often found
forming near or with lichens.
Various phosphate minerals (e.g., iron phosphates or apatite) sometimes mixed with
clays and sometimes manganese.
Human-manufactured material placed on rock surfaces by people.
Clay minerals, Mn and Fe oxides, and minor and trace elements; color ranges from
orange to black in color produced by variable concentrations of different
manganese and iron oxides.
Chloride precipitates formed on rock surfaces.
Usually clear white to orange shiny luster, but can be darker in appearance,
composed primarily of amorphous silica and aluminum, but often with iron.
Sometimes called a particular type of

rock coating
Clay skins, clay films, soiling
Heavy metal
skins
Iron film
Lithobiontic
coatings
Nitrate crust
Oxalate crust
Phosphate
skin
Pigment
Rock varnish
Salt crust
Silica glaze
Sulfate crust
Sulfates (e.g., barite, gypsum) on rocks, not gypsum crusts that are sedimentary
deposits.
Also described by chemical

composition
Ferric oxide, iron staining
Organic mat, biofilms, biotic crust
Saltpeter, niter, icing
Oxalate patina, lichen-produced crusts,
patina, scialbatura
Organophosphate film, epilithic biofilm
Pictograph, paint, graffiti
Desert varnish, patina, Wustenlack
Halite crust, efflorescence

Desert glaze, turtle-skin patina,
siliceous crusts, silicaalumina
coating, silica skins
Sulfate skin
72
Rock Coatings
Figure 1 Rock coatings alter the appearance of bare-rock

landforms. The left column is the rock coating as seen through backscattered electron (BSE) microscopy. The right column illustrates
how thin rock coatings alter the visual appearance of rock. Top row:
case hardening by heavy metals seen through darkens the
appearance of a pink sandstone at Petra, Jordan. Middle row:
calcium oxalate on quartz illustrates how even thin coatings create
dark streaks on granodiorite at Stone Mountain, Georgia, USA.
Bottom row: silica glaze lightens the appearance of basalt flows as
seen in an ASTER image of the rainshadow of Hualalai Volcano,
Hawaii, courtesy of NASA (NASA, 2010), where even the 1859 Mauna
Loa lava flow has a thin silica glaze coating.
the oxalate crust. Silica glaze that has formed on the 1859
Mauna Loa lava flow (lower row in Figure 1) likely started with
soluble AlSi complexes [Al(OSi(OH)3)2 ] that are ubiquitous on silicate mineral surfaces. Gentle wetting from dew or
even water vapor is enough to mobilize these AlSi complexes.
There is a transition between complete and partial wetting; this
transition rests at about 2070 nm. When this transition is
crossed, the metastable wetting film on the silica surface is then
ruptured and silica precipitates.
Taken separately, each site-specific study represents an
empirical exemplar of inductive science. Such cases dominate
the rock-coating literature, but do not contribute to the development of any broader theory explaining the geography of
rock coatings or why different types of rock coatings form in
specific places. Because the science of weathering in geomorphology is a search for broader explanatory patterns, case
studies do not typically advance the development of any
general theory that explains rock coatings.
The field of landscape geochemistry (Polynov, 1937; Perelman, 1961, 1966; Glazovskaya, 1968, 1973) offers one systematic way of analyzing the geography rock coatings, working
toward a model that can predict what types of rock coatings
would develop in different locations. Using this spatial approach
to understanding biogeochemistry, Dorn (1998) proposed five
general hierarchies of controls on the development of rock
coatings. No alternative theoretical approach has yet been posited, although a dialog on development of theory on rock
coatings would be a welcome development.
A hierarchical approach necessarily orders the way of
analyzing rock coatings. For example, first and foremost, bare
rock faces must exist exposed by geomorphic processes such
as landsliding or glaciation. It is an obvious statement that
without the exposure of bare rocks, we would not see subaerial
exposures of rock coatings. As exposures of bare rock are most
common in deserts and alpine settings, rock coatings are more
commonly seen in these environments.
Second-order controls are exhibited where erosion exposes
rock surfaces that already had rock coatings that formed in the
subsurface. The subaerial exposure of coatings that formed in
rock fractures is a second-order control, because a great many
now-exposed coatings originated at depth.
Third-order controls come into play when fast-growing
lithobionts colonize rock faces, and lichens, fungi, and algae
effectively outcompete slower-growing inorganic rock coatings. Lithobionts are a third-order control, because they can
dramatically alter the biogeochemistry of rock surfaces. This
chemical change can often dissolve inorganic coatings or can
prevent them from accreting.
Fourth and fifth orders become relevant only if bare rock
faces occur (first order), if rock coatings are not exposed by an
erosional event (second order), and if fast-growing lithobionts
do not grow over the rock face (third order). All three of these
conditions must be for the development of many of the subaerial rock coatings listed in Table 1. For example, the silica
glaze on Stone Mountain (Figure 1) could accrete, because a
bare rock surface was not already coated by lithobions or an
accretion formed originally in the subsurface.
The fourth and fifth orders of control focus on element
abundance, element migration, geochemical flows, geochemical gradients, and geochemical barriers associated with
different types of inorganic rock coatings. The fourth order
focuses on the issue of whether or not the elemental ingredients of a rock coating are present and have a transport
pathway to a site; for example, the iron and manganese present in rock varnish at Petra was mobilized down into sandstone pores (upper row in Figure 1). However, just because
the elements of rock coatings occur and are transported to a
site does not mean that these elements will accrete.
The fifth order of control involves barriers to transport that
fix the coatings constituents. The transport barrier might be
physical, such as electrostatic or physical forces that hold dust
particles together (Jordan, 1954; Bishop et al., 2002; Ganor et al.,
2009). Another barrier could be geochemical, such as a change
in pH/Eh oxidizing and fixing (NIAIST, 2005) iron in films. Still
Rock Coatings
73
First-order issue:
Exposure of rocks on the land surface
Landscapes
dominated by hillslope
and fluvial processes
Landscapes
dominated by
other geomorphic
processes
Transport-limited
landscape
Weathering-limited
landscape
Extensive soil cover

Rocks exposed
in streams and
by mass wasting
Minimal soil cover

Rock surfaces exposed
in many places
Glaciers
expose
abundant
bare rock
Periglacial
processes
expose
abundant
bar rock
Waves
expose
rock along
shoreline
Figure 2 Subaerial rock coatings can only occur where bare rock faces have been exposed by geomorphic processes.
another barrier could be biological, exemplified by bacteria

oxidizing manganese in rock varnish (Northup et al., 2010).
Barriers to transport resulting in rock coatings generally require
combinations of biological, chemical, and physical processes.
4.5.2.1
Granular erosion of the

newly exposed coatings
First-Order Control: Geomorphic Stability
The first-order control on whether rock coatings occur is

whether bare rock surfaces exist. Rock control is an important
concept in geomorphology (Suzuki, 2002), and it is especially
critical for rock coatings. After all, a cover of soil, vegetation, or
regolith means that rock coatings have no subaerial exposure.
Glaciers, waves, landslides, and overland flow in weatheringlimited landscapes all exemplify geomorphic processes that
expose rocks to the atmosphere the primary precondition for
the occurrence of subaerial rock coatings.
Thus, the first order of control rests with geomorphic
processes that generate bare rock surfaces for subaerial rock
coatings (Figure 2). A wide variety of geomorphic processes
expose bedrock surfaces (Ehlen and Wohl, 2002). Mass wasting, glacial, periglacial, volcanic, faulting, flooding, and coastal
processes can all bring rocks into the subaerial environment.
Bare exposures are most common in arid regions, because
deserts are weathering-limited landscapes (Gilbert, 1877).
4.5.2.2
Re-exposure of same type of

coating by more spalling
Second-Order Control: Subaerial Exposure of

Subsurface Coatings
Rock coatings do develop in the subsurface and are exposed at

the surface by erosion of the overlying soil or rock material.
For example, iron films develop on clasts in the B-horizon
of soils (Haberland, 1975; Hayden, 1976) and are then exposed at the surface through soil erosion (Hunt and Wu,
2004). Manganiferous rock varnish forms in mountain soils
(Ha-mung, 1968). Weathered rock fractures are lined with
MnFe coatings (Weaver, 1978; Kim et al., 2006), dendrites or
branching structures (Xu et al., 2010), and silica glazes (Milnes
et al., 1991; Frazier and Graham, 2000). Clay-coated surfaces
can be a common component of fractured bedrock (Frazier
and Graham, 2000) and weathered minerals in general
(Meunier et al., 2007). Fissures in less-weathered bedrock host
iron films, rock varnish, laminar calcrete, silica glaze, and
Second-order
processes:
Exposure of
subsurface
coatings by
erosion
Growth of
lithobionts
on coatings
Erosion of subsurface coating

and growth of different
subaerial coating
Preservation of subsurface
coating and deposition of
different coating on top
Preservation of subsurface
coating and deposition
of same coating on top
Figure 3 Possible pathways of subsurface-formed coatings upon
exposure in the subaerial environment. Further erosion may lead to
exposure of other subsurface rock coatings. The joint face might
erode by millimeter-scale flaking, removing the surficial coating.
Lichens and other lithobionts might form on top of inorganic rock
coatings, potentially leading to dissolution from the secretion of
organic acids. A former subsurface coating might be dissolved by
carbonic acid in rainwater, freeing up the surface for the growth of a
subaerial coating more in equilibrium with the surficial environment.
A new rock coating might form or, in some cases, the same rock
coating will continue to accrete.
heavy metals (Douglas, 1987; Dorn and Dragovich, 1990;

Robinson and Williams, 1992; Mottershead and Pye, 1994;
Villa et al., 1995). These coatings formed because transport
pathways brought the necessary constituents, and because
barriers to further transport fixed these constituents on fracture sides, in regolith, or on rock fragments in soils.
74
Rock Coatings
20 cm
Rock varnish
black band
Iron film
orange
Dust, eroded
when fissure
spalls
Calcrete
Figure 4 Rock-coating sequence occurring on the walls of desert rock fractures. This fissure once hosted dust and weathered rock fragments.
Carbonate leached from desert dust reprecipitates as laminar calcrete skins deeper in the rock fracture. An orange, iron-rich coating forms where
dust is in constant contact with fissure sides. A centimeter-scale band of black rock varnish rims the outer edge of the fracture where the dust
has been washed away.
10 m
10 m
(a)
(b)
Figure 5 Optical thin sections show how changes in landscape geochemistry alter rock coatings originally formed in desert fissures
(cf. Figure 4). (a) White calcrete first formed a coat over the gray quartz. Then, when the fissure opened wide enough to accumulate dust and
weathered fragments, this dust fostered the formation of the orange film on top of the calcrete. (b) First, an iron film formed in the crevice.
Then, the rock fracture opened wide enough to wash the accumulated dust away from the rock surface. This allowed the formation of a black
rock varnish.
Rock Coatings
75
Figure 6 Spalling of sandstone in Whoopup Canyon in Wyoming exposes the silica glaze that coats fractures. The change from subsurface
fracture to subaerial surface enables rock varnish to accrete on top of the silica glaze. The left BSE image shows case hardening by Fe (and
some Mn) leached from the varnish. The right BSE image presents Mn and Fe combining with silica glaze in pore spaces, forming a different
type of case hardening.
5.0 m
Figure 7 Cobbles in desert pavements can develop a very shiny appearance. Some of this sheen, seen in the lower image from the Sierra
Pinacate, Mexico, comes from silica glaze. In the upper BSE image, an iron film (brighter material) first formed on top of quartz on the sidewalls
of a fissure. Then, dust accumulated in the fissure (fragments of gray material), and the dust was trapped by formation of more iron film. Then,
the rock spalled and this former fissure was embedded in a desert pavement. A micron-thick layer of silica glaze then formed over the prior rock
coatings at the ground-line band of the desert pavement cobble. This silica glaze allows some of the color of the underlying iron film to show
through a very shiny appearance.
76
Rock Coatings
Figure 8 Carbonate crusts form on the underside of large boulders

in Bk horizons in desert and semi-arid soils. Construction of a
prehistoric rock cairn in the Panamint Valley, California (upper image)
and 2010 debris flow in metropolitan Phoenix, Arizona (lower image),
both expose pedogenically formed carbonate crusts (arrows).
The landscape geochemistry setting of a rock-coating

changes when mass wasting exposes a former fracture coating,
when a gully exposes regolith coatings, or when scouring
brings a soil clast to the surface. Subsurface coatings brought
into the subaerial environment through rock or soil erosion
are extremely common. A critique of many rock coatings articles is that they do not even consider a subsurface origin for
the coating under study. The unstated assumption of much
research is that just because a coating occurs exposed to the
atmosphere, it originated in that location.
Upon exposure in the subaerial environment, what were
once subsurface coatings can experience several possible futures, diagrammed in Figure 3. Consider, for example, rock
coatings that are ubiquitous in fractures occurring in warm
deserts (Figure 4). A laminar calcrete skin precipitates where
the fissures are most narrow; as the fissures open up, orange
iron films accrete; a centimeter-wide band of black rock varnish grows where the fissures are close enough to the surface
for precipitation to seep in and wash dust from fissure sides.
This colorful sequence of rock coatings found on the
sides of rock fissures in warm and dusty deserts (Figure 4)
is modified as the landscape geochemical environment changes. As the fissure gradually opens wider, dust is washed from
the walls of the fissure by precipitation, and this change allows
the formation of manganese-rich rock varnish (Figure 5(b)).
With spalling, carbonic acid dissolves the laminar calcrete
skins that are exposed, except where the iron film has already
formed a protective covering (Figure 5(a)). Spalling also
promotes the formation of black manganese rock varnish over
the orange surface, because the removal of the alkaline dust no
longer inhibits microbial enhancement and fixation of
manganese.
Joints or fractures in rocks in semi-arid environments
commonly accumulate silica glaze that helps start the process
Variations over single hillslope

750
Elevation (m)
745
740
735
Rock varnish
730
Iron films
No coating
725
720
0.0
50.0
100.0
150.0
200.0
Distance (m)
Figure 9 A rock coating catena (Palmer, 2002) exemplifying how three different types of rock coatings change down a basalt hillslope in the
Mojave Desert. The hillcrest has less soil erosion, and hence rock varnish dominates. More and more iron films occur further down the slope,
reflecting greater soil erosion exposing the iron films that originally formed in the subsurface.
Rock Coatings
of case-hardening joint faces (Dorn, 1998). With erosion exposing joint faces to the subaerial environment, changes
typically ensue. Figure 6 documents two types of postexposure
changes seen in the semi-arid western USA. Both changes involve the accumulation of rock varnish, followed by leaching
of manganese and iron dissolved from the varnish which
then infiltrates into the underlying pores. Sometimes, the iron
(with some manganese) infills pore spaces (left image in
Figure 6). In other cases, the manganese and iron combine
with the silica glaze in the pore spaces and further contribute
to case hardening the rock surface (right image in Figure 6).
Another example of the influence of erosion altering the
landscape geochemistry of subsurface-formed rock coatings
comes from the spalling of a cobble in a Sonoran Desert
pavement (Figure 7). An iron film originally formed inside a
rock fissure in a desert pavement cobble. Dust particles then
adhered to the iron film, and iron film helped cement dust
to the fissure side walls. Then, the rock split open, changing
77
the landscape geochemical setting from a fissure to exposure at

the soil line in a desert pavement. Desert pavement cobbles
often develop a very shiny line at the soil line called a groundline band (Engel and Sharp, 1958; Dorn and Oberlander,
1982). Ground-line bands are shiny, in part, because
they develop coatings of silica glaze. Figure 7 exemplifies that
even a micron-thick layer of silica glaze can impart
this sheen.
Erosional processes in deserts, both anthropogenic and
natural, can expose carbonate crusts that become temporary
subaerial coatings (Figure 8). Originally formed in the Bk
horizon of soils, the carbonate is slowly dissolved through
interaction with carbonic acid in precipitation. A landscape
geochemistry interpretation is that the geochemical barrier
present in the soil no longer exists in the subaerial environment. Changes in the position of pedogenic carbonate crusts
have been used to analyze prehistoric geoglyphs and other
earthen features (Cerveny et al., 2006).
Figure 10 Examples of lithobionts in different stages of covering subaerial rock surfaces. Upper left: hillslope of welded tuff in central Arizona
where light green lichens cover rock surfaces that have remained stable for a few hundred years. Lower left: lichens coating joint faces at a
spring near Yunta, South Australia. Upper right: dark-colored fungi colonize the surface of a former joint face, while lichens grow inside of
engravings, Kahoolawe Island, Hawaii. Lower right: Whoopup Canyon, Wyoming, where lichens completely coat surfaces where the dark-colored
heavy-metal case-hardened joint face has spalled away.
78
Rock Coatings
Desert cobbles and small boulders in undisturbed settings

host black manganese-rich rock varnish in a subaerial position,
with an orange iron film forming where the clast remains in
contact with the underlying alkaline soil (Cerveny et al., 2006).
Rock coatings can offer a visually distinctive clue that modern
soil erosion has been active, because seeing an abundance of
orange iron films mean that clasts with the black rock varnish
have been eroded. A study of variability in rock coatings along
a single hillslope (Palmer, 2002) found greater amounts of soil
erosion in the steepest portion and at the bottom of the disturbed hillslope in the Mojave Desert, resulting in the exposure
of orange iron films at the surface (Figure 9). Eventually, if
erosion ceases or slows tremendously, black rock varnish will
reform. However, Figure 9 illustrates that the second order of

control, erosion, plays a key role in determining the type of
rock coating that is seen in disturbed settings.
4.5.2.3
Third-Order Control: Competition from Lithobionts
Lithobiontic coatings are organisms that live on the surface

(epiliths), bore tubes into rocks (euendoliths), occupy fissures
in rocks (chasmoendoliths), or live within the pore spaces of
weathering rinds (cyptoendoliths) (Golubic et al., 1981).
Lithobiontic coatings thinner than a millimeter are classified
as biofilms; those between 1 and 5 mm are biorinds; and
100 m
(a)
100 m
(b)
Figure 11 Granodiorite inselberg surfaces at Garden Butte, Papago Park, central Arizona, are dominated by lichens rather than rock varnish.
Images (a) and (b) compare secondary electrons (a) to back-scattered (BSE) electrons (b). Note how the lichens seen with secondary electrons
(a) penetrate into the rock along mineral boundaries and appear to enhance the spacings between minerals, either mechanically or through
chemical weathering.
Rock Coatings
coatings greater than 5 mm are called biocrusts (Viles, 1995;

Gorbushina, 2007).
Lichens, fungi, and algae lithobiontic coatings grow much
faster than most inorganic rock coatings such as rock varnish,
iron films, or silica glaze (Friedmann and Galun, 1974;
Rundel, 1978; Golubic et al., 1981; Dragovich, 1987; Viles,
1995; Souza-Egipsy et al., 2004; Bhatnagar and Bhatnagar,
2005; Loso and Doak, 2006). As a result, these subaerial
organisms commonly dominate rock faces (Figure 10). The
third order of control, thus, involves conditions that control the
79
growth of fast-growing lithobionts, where biofilm communities

can grow in decades to hundreds of years (Viles, 2001).
One way that lithobionts take possession of a rock surface
is through enhancing the weathering and erosion of rock
material (Paradise, 1997; Lee and Parsons, 1999). Lichens
grow not only on the surface, but also in pore spaces between
mineral grains (Figure 11). Increased spacing between mineral
grains destabilizes the rock surface and spalling takes place.
Part of the process involves enhanced dissolution of minerals
under lichens (Aghamiri and Schwartzman, 2002; Gordon
15 m
Figure 12 Lithobionts commonly secrete acids that dissolve inorganic rock coatings. The left image shows basalt talus in the Mojave Desert,
California with inset photo of a petroglyph that was sampled for electron microscopy. The right image is a secondary electron microscope image
from this inset showing a euendolith (tube boring) microcolonial fungi that is dissolving rock varnish.
+
hs
On soft
t
row
eg
ilithics
rocks
s
rfa
am
u
bs
lia
tra
ard
so
nh
ut
So
film
n
itio
os
p
de
er t
ish
n
es
r
d
a
e
V
jav
Mo
us
hA
roc
k
Al
Bio
Weathering efficiency
pin
e
str
h
nis
Var ion
s
Ero
Growth of rock varnish
Su
er ep
?
and oth
Lichens
Subsurface growth on very hard rocks
c
ma
Ata
Dry
Wet
Figure 13 A landscape geochemical conceptualization of how lithobionts and rock varnish interact together. Moisture plays a vital role in the
growth of rock varnish and in the weathering efficiency of lithobionts. Secondary factors presented in this model are competition from lithobionts
for rock varnish and rock hardness for lithobionts. Adapted from Viles (1995) and Dorn and Oberlander (1982).
80
Rock Coatings
Figure 14 Varnish and biofilms of fungi and lichens grow on schist

(lower image, where flower width is 2 cm). At the present time,
biofilms appear to be dissolving varnish, as illustrated in the BSE
image (image width is B100 mm; the epoxy has separated from the
fungi and dissolved varnish). The dissolving varnish is then
reprecipitating inside fractures, promoting case hardening with iron
and manganese heavy metals. This example corresponds with the
location of varnish erosion in Figure 13, where biofilms are growing
on hard rocks.
Figure 15 Anthropogenic paint balls transported and applied to rock

surfaces at South Mountain Park, Arizona.
and Dorn, 2005a). Lichens are then able to recolonize these

spalled surfaces much faster than inorganic coatings such as
rock varnish or iron films. Another way that lithobionts replace other rock coatings is by biochemically dissolving
(Dragovich, 1987) the preexisting coating (Figure 12).
An example of the third order of control on rock coatings
compares the behavior of lithobionts and another rock coating to the external forcing of moisture. Viles (1995: 32) diagrammed lithobiont weathering as a function of moisture and
hardness of the host rock. Dorn and Oberlander (1982) argued that the growth of rock varnish is influenced by moisture
and competition from lithobionts. These two conceptualizations are combined in Figure 13, linked through a common focus on moisture. Lichens are most common in moist
settings, whereas much drier environments foster endoliths.
Rock varnish survives best and grows the slowest in drier environments, but as moisture increases lithobionts biochemically dissolve more rapidly forming varnish (e.g., Figure 12).
Moisture conditions as diagrammed in Figure 13, however,
are not static. A site moves to the right and left on the diagram
with microclimatic oscillations. A slight shift to wetter conditions can foster the colonization of lithobionts and result in
varnish erosion. Considered from the perspective of Figure 13,
a site near the peak of varnish growth can shift into varnish
erosion where acid-producing fungi have colonized rock
varnish. The effect of such a change is seen in Figure 14.
Although lithobionts do chemically dissolve and erode preexisting rock coatings and prevent the formation of many inorganic accretions, it is important to note that lithobionts can
also stabilize rock surfaces by holding weathered fragments in
place (Gehrmann et al., 1988; Kurtz and Netoff, 2001; Viles
and Goudie, 2004). Lithobionts can also generate protective
Figure 16 Anthropogenic pigments applied next to the DeBrung

Monastery, Tibet.
Rock Coatings
coatings of silica glaze (Lee and Parsons, 1999) and oxalate

crusts (Souza-Egipsy et al., 2004).
4.5.2.4
Fourth-Order Control: Transport Pathways
Rock coatings require a sufficient abundance of constituent

elements and those elements are transported to accretion sites.
81
Sometimes, the constituents are fairly ubiquitous such as

silica, clay minerals, and iron. For other rock coatings, the
concentrations transported to a coating site are quite low and
enhancement is required such as with the manganese in rock
varnish or iron in heavy metal skins. In other cases, although
the overall abundance of material such as oxalate minerals
might be low on a rock face, locally strong geochemical
Figure 17 Streak of silica glaze formed over a petroglyph panel at McKonkey Ranch, Utah. The white streak was sampled above the engraving
on a natural joint face. The BSE image shows three types of rock coatings, where the host rock is on the left side of the electron micrograph. On
top of gray quartz is a thin layer of rock varnish (bright material) formed. Then, on top of that varnish is dust film loosely cemented with iron
and manganese. On top of this rests the water-streak silica glaze. The source of the water-mobilized and transported silica appears to be a rock
spall that redirected overland flow. The photograph is about 2 m in length, and the BSE image width is 10 mm.
10 m
Figure 18 A sandstone face at Medicine Lodge Creek, Wyoming, has a fracture occupied by birds. Droppings are mobilized and drip down the
rock art panel. The back-scatter electron microscope image shows phosphate mixing with detrital rock fragments, perhaps transported by water
flow.
82
Rock Coatings
gradients might exist near a source of calcium oxalate such as

lichens. The fourth general control on the occurrence of rock
coatings is the requirement that the constituents of a coating
have a pathway of transport to a rock-surface site.
Some transport pathways are obvious. Pigments are
applied to rock surfaces (Li et al., 2001; Hortola, 2005;
Simionescu et al., 2009) through anthropogenic transport
(e.g., Figures 15 and 16). Streaks of oxalate crust flowing
down from oxalate-producing lichens at Stone Mountain (e.g.,
Figure 1) reveal a visual trace of a transport pathway. A white
streak formed over a Utah petroglyph panel corresponds with
a thin coating of silica glaze (Figure 17). Transport pathways
can be dozens of kilometers for the dust that composes dust
films and rock varnish, or very short in the case of iron and
manganese mobilized from rock vanish and transported into
pores to case harden the underlying rock (Figures 6 and 14).
Transport of raw mineral ingredients involves two general
preconditions. The constituents must be present and they
must migrate to the rock face. Bird droppings (Arocena and
Hall, 2003) or microorganisms (Konhauser et al., 1994), for
example, generate the requisite material for a phosphate
skin. Then, phosphates are mobilized and reprecipitated
(Figure 18). In another example, the formation of silica glazes
on granite-building stones in Rio de Janeiro first requires dust
deposition, then silica is mobilized from the dust to reprecipitate as silica glaze (Smith et al., 2007). Deposited dust, in
general, appears to be a key agent in the formation of pollution-related coatings in urban settings where the first general
step is dust deposition, followed by complex interactions that
result in the net migration of elements from the dust into a
rock coating (McAlister et al., 2006).
Salt crust formation exemplifies how multiple transport
steps can involve very different processes (Oguchi et al., 2002).
The salt crust seen on weathered surfaces such as Mushroom

Rock in Death Valley, for example, was first transported by
wind from the salt playa to the soil adjacent to the talus
boulder. Then, salt was dissolved in precipitation and moved
by capillary action up the side of the boulder to precipitate as
salt crusts (Figure 19).
Multiple transport pathways are key to the formation of
black crusts occurring on marble tombstones and limestone
buildings in humid regions undergoing anthropogenic pollution. Sulfate and oxalate crusts both occur on such surfaces
as shown in Figure 19. Sulfur from carbon fuel combustion
interacts with the host carbonate to produce gypsum
sulfate crusts (Potgieter-Vermaak et al., 2004). Then, this
Figure 20 Sulfate crust on marble tombstone in the Old Fellowship

Cemetery in Atlanta, Georgia, USA.
10 m
15 m
Figure 19 Barium sulfate, sodium chloride, calcium sulfate, and strontium sulfate weathers Mushroom Rock in Death Valley, California. The BSE
micrographs on the right show a barium sulfate crust as the brighter material both fragmenting the basalt and covering mineral surfaces. The
bright white in these images are barite and darker crystals are weathered basalt silicate minerals.
Rock Coatings
Moisture and solutes

W
e
t
t
i
n
g
100 m
A
t
m
o
s
p
h
e
r
e
f
r
o
n
t
83
Gypsum crust
Gypsum replacement of
eroding host limestone
Host limestone
Moisture and solutes
Original surface
Carbonaceous
soot (dots)
Gypsum (needles)
Host limestone
Figure 21 Pollution-generated gypsum crusts can replace the host limestone through flow of sulfate-rich solutions migrating inward from the
surface or outward from the rock, and they can exist as needles over a dissolving surface.
Case hardening of
quartz cemented by
manganiferous varnish
Quartz loosely
cemented by clay minerals
100 m
Figure 22 Waterflow streaks can sometimes be fungi, lichens, heavy metal skins, and sometimes rock varnish. In this case, streaks are heavy
metal skins that impregnate the sandstone of Sedona, at Schnebly Hill Road, Arizona.
84
Rock Coatings
gypsum is then transported by capillary water microns to

millimeters on the rock surface to contribute to the development of biofilms of cyanobacteria (Ortega-Calvo et al., 1994)
(Figure 20).
The deposition of sulfate crust occurs in tandem with dissolution of the limestone, but the style of the accretion can
vary (Figure 21). In some cases, a thicker crust of several
hundred micrometers forms as the gypsum replaces the
limestone (Verge`s-Belmin et al., 1993) through a process by
which sulfate-rich solutions migrate outward from the rock in
the early stages of surface desiccation (Smith, 1994). In other
cases, the sulfate crust manifests itself as a mix of gypsum
needles and soot over karren (Camuffo et al., 1983).
Sometimes, the lack of a transport pathway for a key ingredient can determine what type of rock coating accretes. This
is especially true for rock varnish. Black streaks across sandstone surfaces of the Colorado Plateau are almost always attributed to rock varnish, but this is generally incorrect. These
water-flow deposits commonly lack clay minerals. As clays are
vital to the formation of rock varnish (Potter and Rossman,
1977; Krinsley et al., 1995; Krinsley, 1998; Dorn, 2007), the
manganese and iron deposited without the clays results in a
heavy metal skin (Figure 22). Another example comes from
Hawaii. Silica glaze forms on the rainshadow side of the island
of Hawaii (Figure 1), in part because of the paucity of clay
mineral transport to basalt flow surfaces.
4.5.2.5
barrier to transport exists. For example, even though dust is

ubiquitous in warm deserts (Goudie, 1978) and coatings
occur where electrostatic or physical forces hold dust particles
together (Jordan, 1954; Bishop et al., 2002; Ganor et al.,
2009), the lack of bare rock surfaces (first order), the exposure
of a subsurface coating (second order), the growth of lithobionts (third order), or the transport of other materials by
water or wind (fourth order) can all interfere with the occurrence of a dust coating (Figure 23).
Physical and chemical barriers commonly work in tandem
to generate inorganic rock coatings, as it the case for silica
glaze. Dorn (1998) proposed that silica glaze formation starts
Fifth-Order Control: Barriers to Transport
Physical, chemical, and biological barriers halting the transport of elements results in the accretion of rock coatings. This
fifth order of control is not ranked higher, because the other
controls can prevent the occurrence of a coating, even if a
100 nm
Figure 24 HRTEM image of silica glaze spheroids in a sample from
the Ashikule Basin, Tibetan Plateau. The bright dots are artifacts.
Figure 23 High-resolution transmission electron microscope image of a dust film that has been preserved inside a partially opened rock
crevice. Its position protects the coating from being washed away by incident precipitation. The sample was collected from a granodiorite
surface, South Mountain, central Arizona. The nature of the clays in this dust film shows an orientation parallel to magnetite surface. A mixture
of illite and smectite (sm) occurs. The illite has a spacing of 1.0 nm. The darker portions of the smectite are too thick for spacing measurement,
but the thinner portions show spacings of 1.3 nm. There may also be organic matter (OM?) as a part of the dust film.
Rock Coatings
85
5 m
Physical barrier: van der Walls
force promotes dust
accumulation, providing
raw ingredients of clays
Biological barrier: Mn and Fe fixation in bacteria sheaths

Si
Al
1 m
Varnish
K Ca
Mn Fe
Mn
Si
Al
0.2 m
K Ca
Ti
Fe
Spot on bacteria
Dissolution from
cell wall (granular
fragments)
20 nm
20 nm
8 nm
and
Physiochemical
barrier: Fixation
in clay matrix
cementing
mixed-layered
clays
10 nm
10 nm
20 nm
20 nm
Figure 25 Different barriers to transport are involved in the accretion of rock varnish. The top BSE image illustrates dust that has accumulated
on top of rock varnish from the Ashikule Basin, Tibet. HRTEM and secondary electron microscopy images of cocci bacterial forms from Antarctic
varnish, and energy-dispersive electron microscopy shows that the bacteria are concentrating manganese and iron in their sheaths. HRTEM of
samples from Peru, Death Valley, and Antarctica illustrate instability and then barriers at the nanometer scale where oxides are dissolved from
bacterial sheath fragments, transported a few nanometers, and then fixed into mixed-layered illitemontmorillonite clays.
86
Rock Coatings
with soluble AlSi complexes [Al(OSi(OH)3)2 ] that are

common at the waterrock interface (Browne and Driscoll,
1992). Dew or frost deposition supplies sufficient moisture to
mobilize AlSi complexes, where the transition between
complete and partial wetting on silica surfaces is about
2070 nm (Zorin et al., 1992; Churaev, 2003). Crossing this
physical transition would result in the deposition of silica with
particles in the size range of 2070 nm. High-resolution
transmission electron microscopy (HRTEM) finds evidence of
spheroids in this size range in silica glaze (Figure 24), an
observation consistent with this model of silica glaze formation (Langworthy et al., 2010).
A sequence of physical, biological, and physiochemical barriers are needed for the formation of rock varnish (Figure 25).
Varnishing begins with physical barriers of electrostatic or
physical forces holding dust particles together on subaerial
surfaces (Jordan, 1954; Bishop et al., 2002; Ganor et al., 2009).
Then, the biological action of bacteria concentrate manganese
and iron (Dorn and Oberlander, 1981; Hungate et al., 1987;
Dorn, 2007; Northup et al., 2010). Some of the bacterial sheaths
become microfossils (Dorn and Meek, 1995; Dorn, 1998;
Krinsley, 1998) that then dissolve. Decay of the MnFe casts
mobilizes nanometer fragments. Physiochemical fixation takes
place only a few nanometers away when the oxides are fixed
Figure 26 Carbonate crusts formation in different environments. The two left images show tufa formed in a lacustrine environment at Pyramid
Lake, Nevada, generating centimeter-scale crusts over rock surfaces. The upper right image illustrates a biogenic carbonate crust of beachrock
tens of centimeters thick along the northeastern shore of the Sinai Peninsular. The lower right BSE image shows carbonate crust formed as a
result of sprinkler water evaporating on a tombstone in Phoenix, Arizona.
Rock Coatings
87
into mixed-layered clays (Potter, 1979). HRTEM imagery (e.g.,

Figure 25) shows Mn and Fe inserted into mixed-layered clays.
Potter (1979: 174175) hypothesized:
Deposition of the manganese and iron oxides within the clay
matrix might then cement the clay layerythe hexagonal arrangement of the oxygens in either the tetrahedral or octahedral layers of
the clay minerals could form a suitable template for crystallization
of the layered structures of birnessite. The average 0-0 distance of
the tetrahedral layer is 3.00 A in illite-montmorillonite mixedlayered clays, which differs only 3.4 percent from the 2.90 A distance of the hexagonally closed-packed oxygens in birnessitey
Rock varnish formation, then, appears to be the result of a

sequence of barriers and nanometer-scale transport: first,
physical barriers to the migration of first dust that supplies clay
minerals; second, biological barriers that enhance manganese
and iron; third, Mn and Fe are transported nanometers to adjacent clay minerals; and fourth, the physiochemical processes
cement clay minerals to the rock and to prior varnish. This is
the polygenetic model of rock varnish formation (Dorn, 1998).
Although carbonate crusts do not necessarily require a
complex sequence, they can form from physical, chemical, or
biological barriers to transport of carbonate (Figure 26). Tufas
and other types of carbonate crusts are known to form from
both biotic and chemical processes that create barriers to
further transport (Pentecost, 1985; Pedley, 1990; Viles and
Goudie, 1990; Benson, 1994; Arp et al., 1999; Carter et al.,
2003; Yoshikawa et al., 2006). Marine carbonate cementing
rock surfaces in coastal areas appear to be related to activity of
algae and cyanobacteria (Krumbein, 1979; Kendall et al.,
1994). The physical barrier of evaporation can also form carbonate crusts (Schlesinger, 1985).
Barriers to elemental migration can shift, resulting in very
different types of rock coatings, even over scales of tens of
nanometers (Figure 27). In a sample from the dusty and alkaline Ashikule Basin, Tibet, layered varnish rests on top of
and underneath silica glaze composed of spheroids. In this
HRTEM image, varnish rests distinctly on the underlying
rock and silica glaze spheroids makes a distinct contact with
the varnish both underneath and above the silica. Energydispersive X-ray spectroscopy analysis of the middle zone
reveals that the spheroids are composed of Si and O, similar
to Figure 24. The shifting geochemical barrier could have
been something as subtle deposition of alkaline dust changing
the pH from near-neutral conditions favoring rock varnish
formation to higher pH conditions favoring silica glaze formation, and then back again to near-neutral pH values.
Geochemical barriers can shift over short or long timescales, ranging from hours to thousands of years. Temporary
rock coatings of ice can melt within a day (Hetu et al., 1994).
Organic coatings on andesite volcanic blocks form during the
final stages of the solidification of a lava flow, resulting in the
coalification of plant material (Donoghue et al., 2009). Acidfog generates coatings that mix silica glaze and jarosite, formed
from evaporative processes (Shiffman et al., 2006). Iron films
transform from ferrihydrite to goethite (Raiswell et al., 2009)
under glaciers and ice sheets over periods of less than 100
years. Silica glazes can form within a few decades (Gordon and
Dorn, 2005b) and can alter lava flow appearances dramatically
within centuries to millennia (Figure 1). Rock varnishes
50 nm
Figure 27 HRTEM image of silica glaze interdigitating with rock

varnish. The dark material at the bottom of the image is the
underlying rock. Layered rock varnish forms a 50-nm thick deposit
on the underlying rock. The barrier to transport changed, resulting in
silica glaze spheroids deposited in a layer about 200-nm thick. Then,
layered rock varnish deposited on top of the silica spheroids.
accrete evidence of millennial-scale climatic change through

shifting geochemical barriers (Figure 28); time periods with
stronger manganese barriers record wet intervals (Liu and
Broecker, 2000, 2007; Broecker and Liu, 2001).
The creation of one geochemical barrier can generate another
barrier. Manganese and iron oxides enhanced in rock varnish
(Figure 28) scavenge heavy metals (Jenne, 1968; Thiagarajan
and Lee, 2004; Wayne et al., 2006). Modern aerosols are generally much higher in such heavy metals such as lead from the
use of leaded gasoline (Ganor et al., 2009). The upper micron in
rock varnish, thus is greatly enhanced in lead, because iron and
manganese scavenges this pollutant and fixes it (Figure 28).
Rock coatings are rarely stable for more than a few thousand years. In a few cases, the layering pattern of varnish
microlaminations reveals that stable rock surfaces can host
coatings for a few hundred thousand years (Liu and Broecker,
2008b; Liu, 2010). Along the coast of southern Peru, meters of
salt crust have coated bedrock of a marine isotope stage 5e
marine terrace over the last 105 years (Figure 29), and sulfate
crusts formed in Victoria Land, Antarctica may be as old as the
last time a locus was glaciated (Giorgetti and Baroni, 2007).
Burial can preserve rock coatings in Antarctica (Marchant
et al., 1996). In one locale, geochemical barriers remained
intact after burial, preserving rock varnishes for 107 years
(Figure 30) (Dorn and Dickinson, 1989).
An issue that should not be neglected is the role of nucleation sites as important barriers for migration. This is the
case for bacteria that concentrate manganese and iron in rock
varnish (Figure 25) (Dorn and Oberlander, 1981; Hungate
et al., 1987; Dorn, 2007; Northup et al., 2010). Nucleation
sites can also be important in the formation of some sulfate
crusts (Figure 31). Carbon particles are known to serve as sites
of gypsum precipitation in urban polluted contexts (Del
Monte and Sabbioni, 1984; Del Monte et al., 1984).
In summary, landscape geochemistry theory interprets the
presence of rock coatings as reflecting physical, biological, or
chemical barriers to the transport of elements. However, the
presence of inorganic rock coatings can only occur where there
is limited competition from lithobionts such as lichens that
88
Rock Coatings
Figure 28 Microlaminations in rock varnish reflect shifting geochemical barriers. Wetter environments enhance barriers to Mn migration and
result in the accretion of dark Mn-rich layers. These ultra-thin sections are of rock varnishes collected from the Ma Ha Tuak Range, Phoenix,
Arizona. Wetter microenvironments increase rates of varnishing and can preserve relatively fine paleoclimatic information. The upper section
(thickness B28 mm) shows all three wet phases of the Little Ice Age WH1 signal (Liu and Broecker, 2007): WH1a, WH1b, and WH1c. Also
annotated on this section are wavelength-dispersive electron microprobe analyses of PbO; these analyses show the typical pattern of lead
contamination of the uppermost microns in varnish from twentieth-century automobile pollution. The values are in PbO weight percent, and the
approximate distance between the probe spots are 2 mm. Note how PbO drops down close to or below minimum detection limits underneath this
twentieth-century varnish. The bottom image was collected from a drier microenvironment, and this drier setting slows the rate of varnishing to
the point where only the major wet Holocene (WH) periods (Liu and Broecker, 2007) are recorded (thickness B40 mm).
can grow much faster. These barriers to transport can be

widespread, leading to extensive coatings of rock varnish.
These barriers can be linear, resulting in streaks of silica glaze
or oxalate crusts. Barriers can also be discrete places, such as
carbonate crusts accumulating on the undersides of desert
boulders. The geographical expression of such barriers leads to
the amazing variety of rock coatings occurring in terrestrial
weathering environment.
4.5.3
Importance of Rock Coatings in

Geomorphology
Rock coatings influence landforms in a variety of ways. At the

most general level, the esthetic and dramatic bare rock landforms that motivate the general public, students, and professionals alike are colored by ubiquitous rock coatings. The
appearance of favored icons of geomorphology, such as Uluru
Rock Coatings
Figure 29 Marine Terrace, southern Peru, where up to 4 m of

sodium chloride salt crust covers bedrock.
being coated with iron films (Dorn and Dragovich, 1990) or

the spectacular alluvial fans of Death Valley darkened by rock
varnish (Dorn, 1988), cannot be separated from landforms.
Personal satisfaction associated with field work is, thus, inseparably linked with scenic aspects imposed by rock coatings
(e.g., Figure 1).
Rock coatings are also relevant to geomorphological studies, because they can help stabilize landform surfaces through
case hardening. Case hardening commonly derives from the
mobilization and reprecipitation of rock-coating materials
inside pore spaces in the weathering rind (Conca and Rossman, 1982). Calcite is a common mineral that case-hardens
surfaces (Mellor et al., 1997). Heavy metals also case harden
by infilling of pore spaces in a weathering rind, for example,
sandstone at Petra, Jordan (Figure 1) and Sedona, Arizona
(Figure 22). Figure 32 illustrates stabilization of delicate
weathering features through reprecipitation of rock varnish or
silica glaze. In addition to the migration of single types of
coating material into the weathering rinds, different types of
mobilized rock coatings are commonly mixed together inside
the weathering rind (Figure 6).
The rock-coating material reprecipitated into weathering
rinds does more than physically protect through case hardening. In general, chemical weathering rates decrease over
time in large part because of the role of clay-coated mineral
surfaces (Meunier et al., 2007). Some types of rock coatings,
such as silica glaze, can slow the rate of chemical weathering
(Gordon and Dorn, 2005b). In a study of basalt surfaces of
known age in Hawaii, chemical dissolution of plagiocase
under silica glaze was much lower than plagioclase not covered by silica glaze (Table 2).
Other types of rock coatings can, in contradistinction,
accelerate the physical weathering of rocks. The calcrete crust
and dust films that precipitate and accumulate inside rock
fissures (Figure 4) wedge open those fractures (CoudeGaussen et al., 1984; Villa et al., 1995; Cerveny et al., 2006;
Dorn et al., 2008). Analysis of 20 classic bedrock desert
landforms in the Southwestern USA reveals minimum rates of
rock spalling from dirt cracking to be between 0.3% and 1.5%
per 1000 years. Thus, the precipitation of calcrete crusts and
the expansion and contraction of dust films inside fissures can
89
complete resurface a desert landform within a timescale in the

range of 66 000400 000 years.
The study of rock coatings includes an extensive effort
to use characteristics of these coatings to determine
exposure age. In particular, rock varnish has seen decades of
study in order to determine if different characteristics could be
used to assess when varnish started to form. Knowing the
approximate starting time for varnishing would, therefore,
provide insight into the age of a glacial moraine, alluvial-fan
deposit, stream terrace, landslide, or other feature. Table 3
presents the various strategies used to estimate the start of
varnishing.
Out of these different dating methods, the most advanced
approach is the result of two decades of research by Tanzhuo
Liu of Columbia University (Dorn, 2009). Scholarly research
into rock coating initiated with the field and laboratory studies
of rock varnish by Alexander von Humboldt (von Humboldt,
1812). Rock-coating studies over the past two centuries have
typically focused on a handful of samples. By contrast, Dr. Liu
has analyzed more than 10 000 microstratigraphic sequences
of rock varnishes obtained from millimeter-scale rock depressions (e.g., Figure 33). His painstaking analysis of varnishes, based on the study of three orders of magnitude more
samples than analyzed in a typical publication, has led to a
revolution in our understanding of how climatic changes are
recorded by varnish microlaminations (Liu and Broecker,
2007, 2008a, 2008b; Liu, 2010). The geomorphological implications of Dr. Lius research are profound, offering desert
geomorphologists a tool to study in tandem chronometry and
the influence of climatic change (Dorn, 2009).
Rock coatings have significance in extraterrestrial contexts,
such as the ubiquitous dust coatings on lunar surfaces (Gaier,
2005). Although it is possible that rock varnish might exist on
Mars inside rock fractures (Krinsley et al., 2009) and perhaps
in other settings (DiGregorio, 2010), most research on Mars
rock coatings view accretionary coatings and weathering rinds
as occurring together, perhaps in some type of mixture
(McAdam et al., 2002; Haskin et al., 2005).
Dorn (1998) predicted that silica glazes should be found
on Mars, and silica glazes in terrestrial volcanic settings have
been investigated as potential analogs (Minitti et al., 2005,
2007; Chemtob et al., 2010). An analysis of chemicalweathering conditions on Mars (Kraft et al., 2004) suggests
that four steps might be involved in the formation of silica
glazes on Mars: (1) a fresh rock surface is exposed; (2) dust is
deposited on rock surfaces; (3) thin films of liquid water form
between dust grains along the dust-rock interface leading to
(4) silica glaze deposition (Kraft et al., 2004, p. 2).
There is another possibility to explain the possible occurrence of silica glaze on Mars and Earth that has not yet been
considered: water vapor. An unsolved problem on the geography of terrestrial rock coatings concerns why rock varnish is
so common in subtropical deserts, but silica glaze is ubiquitous on basalt surfaces in rainshadows of Hawaiian volcanoes
Haleakala, Mauna Loa, Mauna Kea, and Hualalai. One possible explanation is that the paucity of dust inhibits varnish
formation that is dependent on clay minerals (Figure 25).
However, the abundance of water vapor in this tropical context could not be ruled out. Thus, a 20-year laboratory experiment exposed basalt rock chips to 80% and 90% levels
90
Rock Coatings
1 km
on
ny
rre
Ba
Qal
25
50
Arizona
Qtg N
Ca
Ksa
10
Tucson
U D
60
Tpa
Ksa
30 35
10
10
Adobe
tank
Qtg
Quaternary terrace
10
Tsg
Neogene basin fill

of Sonoita valley
(Miocene-Pliocene)
n
so
vid
Da
11040
Tsg
Tpa
Pantano formation
(Oligocene-Miocene)
Ksa
Salero formation
(Upper Cretaceous)
10
10
83
Sonoita
Quaternary alluvium
30
50
Tsg
31 50
3150
Qal
n
yo
an
30
35
Figure 30 Rock varnish formed originally on colluvium on a Miocene hillslope of the Pantano Formation has been preserved at Davidson
Canyon, southern Arizona. Arrows indicate the varnished colluvial horizon that is overlain depositionally by sandy gravels part of Neogene basin
fill. The composition of this varnish is similar to modern semi-arid varnishes (Dorn and Dickinson, 1989).
of relative humidity. No liquid water was involved. This experiment found that water vapor alone can generate silica
glaze and may be an important factor in explaining why silica
glaze is the dominant rock coating in humid warm drylands
(Figure 34). Although water vapor concentrations on Mars are
extremely small, the length of time for mineralwater vapor
interactions on Mars make it possible that coatings amorphous silica occurring on Mars could be a product of billions of
years of nanoscale water vapor interaction.
4.5.4
Conclusion
Fourteen general types of coatings cover rocks in the terrestrial

weathering environment. Some coatings are always found
separately, whereas others blend together creating an
almost infinite variety. The only general theory that has been
proposed to understand rock coatings is the paradigm of
landscape geochemistry, originally developed by Soviet geographers. This framework focuses on the spatial aspects of
Rock Coatings
Backscattered electrons
91
Secondary electrons
Silica glaze
Mn-rock varnish
Silica glaze
1
0
0
Gypsum and Mg-sulphate

precipitation in rind
Vitrinite
Figure 31 A particle of vitrinite serves as a nucleation site for sulfate crust precipitation at Karolta, South Australia. The vitrinite is seen
topographically in secondary electrons, but its low atomic number makes it appear dark in back-scattered electrons. After the sulfate crust
precipitated around the carbonaceous particle, the landscape geochemistry environment oscillated between favoring silica glaze and rock varnish
formation.
30 m
10 m
Figure 32 BSE imagery showing that case hardening occurs from the remobilization of rock coatings into pores in the underlying rock.
Left: rock varnish has dissolved and reprecipitated in the underlying mica, as weathering has gradually opened the mineral (left), from the
Ashikule Basin, Tibet. Right: silica glaze reprecipitated into pore spaces in the weathering rind of basalt, Kahoolawe Island, Hawaii. In both cases,
the weathered fragments of the host rock are held together by the reprecipitated rock coatings.
biogeochemistry and interprets rock coatings in terms of

element abundance, element migration, geochemical flows,
geochemical gradients, and geochemical barriers. These components of understanding rock coatings are reorganized here
in a hierarchy of five orders of control to explain what types of
rock coatings develop:
First order. Bare rock faces must be exposed by erosional

processes for rock coatings to be seen.
Second order. Coatings originally formed in the
subsurface can be exposed by erosional processes and are
inherited from a different landscape geochemistry
environment.
Third order. Lithobionts such as lichens dominate rock

surfaces in conditions where they grow much faster than
inorganic rock coatings.
Fourth order. The elements of rock coatings must be transported to bare rock surfaces.
Fifth order. Barriers to the further transport of constituents
result in the accumulation of rock coatings.
Rock coatings are important to the broader field of

geomorphology in several different ways. They alter the
appearance of bedrock landforms. Coatings can promote
surface stability through case hardening; this first involves
the mobilization of constituents from rock coatings and
92
Rock Coatings
Table 2 Weathering of plagioclasea grains of Hawaiian basalt surfaces exposed for different lengths of time, underneath silica glaze and not
under silica glaze
Surface
Not under silica glaze

Grain areab
Mauna Ulu-a (1974 AD)

Mauna Ulu-b (1974 AD)
Mauna Ulu-c (1974 AD)
flow f7d h7.9 (B700 BP)
flow f5d c8.2 (B2000 BP)
flow f5d p3.5 (B3150 BP)
Mauna Kea Glacial Polish (B14 000 BP)
217
216
202
238
303
141
201
500
000
500
500
000
000
500
Under silica glaze

Porosityc
0.07170.018
0.05870.014
0.00770.004
0.9470.17
1.6370.15
2.9070.25
26.20710.33
Grain areab
160
216
247
183
160
185
154
500
500
000
500
500
500
000
Porosityc
0.00970.018
0.00870.011
0.01070.022
0.1370.19
0.3970.23
0.5470.33
15.7175.82
Representative composition in oxide weight percent of analyzed plagiocalse minerals: 3.22% Na2O, 0.22% MgO, 29.12% Al2O3, 48.22% SiO2, 18.10% CaO, 0.12% TiO2, 0.40%
MnO, and 0.54% FeO.
b
Grain area is measured in square micrometers as a total for all mineral grains analyzed.
c
Each value indicates the average and standard deviation of the porosity measurements.
Table 3 Different methods that have been used to assess rock varnish chronometry
Method
Synopsis of method
Accumulation of Mn
and Fe
As more varnish accumulates, the mass of manganese and iron gradually increases. Occasionally this old idea is
resurrected (Lytle et al., 2002), but it has long ago been demonstrated to yield inaccurate results in tests against
independent control (Bard, 1979; Dorn, 2001).
The appearance of a surface darkens over time as varnish thickens and increases in coverage. However, much of this
darkening has to do with exposure of inherited coatings, and with the nature of the underlying weathering rinds, that do
not permit accurate or precise assignment of ages based on visual appearance. There is no known method that yields
reliable results.
Rock varnish contains elements that are leached (washed out) rapidly (Dorn and Krinsley, 1991; Krinsley, 1998). Over time,
a ratio of leached to immobile elements decline over time (Dorn, 2001). If the correct type of varnish is used, the method
performs well in blind tests (Loendorf, 1991). This method also has seen use in places such as China (Zhang et al., 1990),
Israel (Patyk-Kara et al., 1997), South Africa (Whitley and Annegarn, 1994), Yemen (Harrington, 1986), and elsewhere.
Rock carvings made historically may have used steel. The presence of steel remains embedded in a carving would invalidate
claims of antiquity, whereas presence of such material as quartz would be consistent with prehistoric antiquity (Whitley
et al., 1999).
Twentieth-century lead and other metal pollution is recorded in rock varnish, because the iron and manganese in varnish
scavenges lead and other metals. This leads to a spike in the very surface micron from twentieth-century pollution.
Confidence is reasonably high, because the method (Dorn, 1998: 139) has been replicated (Fleisher et al., 1999;
Thiagarajan and Lee, 2004; Hodge et al., 2005) with no publications yet critical of the technique that can discriminate
twentieth century from pre-twentieth-century surfaces.
Organic carbon exists in an open system in the rock varnish that covers petroglyphs. This method compares the more
mobile carbon and the more stable carbon. The method is best used in soil settings (Harrison and Frink, 2000), but it has
been applied experimentally to rock varnish in desert pavements (Dorn et al., 2001).
Calcium carbonate sometimes forms over varnish, and can be radiocarbon dated, providing a minimum age for such
features as rock art. The method has been used in Australia (Dragovich, 1986) and eastern California (Smith and Turner,
1975; Cerveny et al., 2006).
The hope is that carbon trapped by coating provides minimum age for the petroglyph. First developed in 1986, two
independent investigators working in a blind test (Dorn, 1997; Watchman, 1997) both found organic carbon that pre- and
post-dates the exposure of the rock surface. The only person who still uses organic carbon of unknown residues in
radiocarbon dating (Watchman, 2000; Huyge et al., 2001), Watchman now admits that he has not tested results against
independent controls (Watchman, 2002; Whitley and Simon, 2002a, 2002b).
The inorganic mineral oxalate (e.g., whewellite: CaC2O4 H2O) sometimes deposits on top of or underneath rock varnish
(Watchman et al., 2000). Because this mineral contains datable carbon, the radiocarbon age can provide a minimum age
for the underlying or overlying varnish. The most reliable research on radiocarbon dating of oxalates in rock surface
contexts has been conducted in west Texas (Rowe, 2001; Spades and Russ, 2005) and in a rock art shelter (Watchman
et al., 2005).
As radionuclides are enhanced in varnish (Marshall, 1962), uranium-series isotopes show potential (Knauss and Ku, 1980).
Complications surround acquiring the necessary amount of material from the basal layers and concerns over accounting
for the abundant thorium that derives from clay detritus instead of radioactive decay.
Climate fluctuations change the pattern of varnish microlaminations (VML). The confidence level is high, because the
method (Liu, 2003; Liu and Broecker, 2006) has been replicated in a rigorous blind test (Marston, 2003), and the method
is based on analyses of over 10 000 rock microbasins.
Appearance
Cation-ratio dating
Foreign material
analysis
Lead profiles
Organic carbon ratio
14
C carbonate
14
C organic
14
C oxalate
Uranium-series dating
Varnish microlaminations (VML)
Rock Coatings
93
H0:
H1:
H2:
12 500
16 500
16 500
Folsom:
Clovis:
11 50013 000
13 00013 400
LU-1
Oldest Dryas
Clovis
IACP
Older Dryas
Allerd biling
35
Folsom
GISP2 Ice core record (18O)
30
H0
H2
H1
a
WPO
a: 11 750
b: 12 350
c: 13 150
WPO+
a: 14 150
b: 14 950
40
WP1
a
b
b
WP0+
d
c
WP0
a: 15 950
b: 16 550
c: 17 150
d: 17 750
WP1
LU-3
LU-4
LU-2
45
0
5000
10 000
15 000
20 000
25 000
Calendar years BP
Figure 33 A varnish ultra-thin section from Galena Canyon fan in Death Valley (with the upper layers irregularly polished off during thin section
production) exemplifies how different terminal Pleistocene varnish microlamination patterns. The age correlation presented here comes from
independent numerical age control and from the Greenland ice core record. For the sake of temporal recognition, the well-known Folsom and
Clovis lithic technologies are placed in this high-resolution sequence. The nomenclature of layering units (LU), Heinrich Events (e.g., H0, H1, H2),
wet periods in the late Pleistocene (WP) identified by black varnish layers follows previous research (Liu and Broecker, 2007, 2008a, 2008b; Liu,
2010).
then reprecipitation within the pore spaces of the underlying

weathering rind. By contrast, rock coatings formed in
rock fissures in dusty warm deserts can accelerate physicalweathering rates substantially through wedging rocks apart.
Rock coatings likely occur as silica glaze on Mars. One of the

most significant aspects of rock coatings in geomorphology
rests with the revolutionary research by Tanzhuo Liu,
where two decades of painstaking analysis of over 10 000
94
Rock Coatings
(a)
(b)
Figure 34 BSE images of silica glaze formed on basalt rock chips after exposure to only air with a relative humidity of (a) 90% and (b) 80% at
18 1C in a 20-year-long laboratory experiment. Arrows show contacts between the silica glaze and the underlying basalt. The white dots indicate
the positions of focused beam electron microprobe measurements. A typical water vapor-formed silica glaze composition is MgO 0.06%, Al2O3
0.57%, 80.22% SiO2, 0.04% K2O, 0.22% CaO, 0.08% TiO2, 0.06% Fe2O3 with abundant porosity.
microsedimentary basins exemplifies how rock coatings can

be used to analyze paleoclimatic changes and also provide
minimum ages to understand landform evolution.
References
Aghamiri, R., Schwartzman, D.W., 2002. Weathering rates of bedrock by lichens: a
mini watershed study. Chemical Geology 188, 249259.
Arocena, J.M., Hall, K., 2003. Calcium phosphate coatings on the Yalour Islands,
Antarctica: formation and geomorphic implication. Arctic Antarctic and Alpine
Research 35, 233241.
Arp, G., Thiel, V., Reimer, A., Michaelis, W., Reitner, J., 1999. Biofilm
expolymers control microbialite formation at thermal springs discharging
into the alkaline Pyramid Lake, Nevada, USA. Sedimentary Geology 126,
159176.
Bard, J.C., 1979. The development of a patination dating technique for Great Basin
petroglyphs utilizing neutron activation and X-ray fluorescence analyses. PhD
Thesis, University of California, Berkeley.
Benson, L., 1994. Carbonate deposition, Pyramid Lake Subbasin, Nevada: 1.
Sequence of formation and elevational distribution of carbonate
deposits (tufas). Palaeogeography, Palaeoecology, Palaeoclimatology 109, 5587.
Bhatnagar, A., Bhatnagar, M., 2005. Microbial diversity in desert ecosystems.
Current Science 89, 91100.
Bishop, J.L., Murchie, S.L., Pieters, C.M., Zent, A.P., 2002. A model for formation
of dust, soil, and rock coatings on Mars: physical and chemical processes on
the Martian surface. Journal of Geophysical Research 107(E11), http://dx.doi.org
/10.1029/2001JE001581.
Broecker, W.S., Liu, T., 2001. Rock varnish: recorder of desert wetness? GSA Today
11(8), 410.
Browne, B.A., Driscoll, C.T., 1992. Soluble aluminum silicates: stoichiometry, stability,
and implications for environmental geochemistry. Science 256, 16671669.
Camuffo, D., Del Monte, M., Sabbioni, C., 1983. Origin and growth mechanisms of
the sulfated crusts on urban limestone. Water, Air and Solution Pollution 19,
351359.
Carter, C.L., Dethier, D.P., Newton, R.L., 2003. Subglacial environment inferred from
bedrock-coating siltskins, Mendenhall Glacier, Alaska, USA. Journal of
Cerveny, N.V., Kaldenberg, R., Reed, J., Whitley, D.S., Simon, J., Dorn, R.I., 2006.
A new strategy for analyzing the chronometry of constructed rock features in
deserts. Geoarchaeology 21, 181203.
Chemtob, S.M., Jolliff, B.L., Rossman, G.R., Eiler, J.M., Arvidson, R.E., 2010. Silica
coatings in the Kau Desert, Hawaii, a Mars analog terrain: a
micromorphological, spectral, chemical and isotopic study. Journal of
Geophysical Research 115(E04001), http://dx.doi.org/10.1029/2009JE003473.
Churaev, N.V., 2003. Surface forces in wetting films. Advances in Colloid and
Interface Science 103, 197218.
Conca, J.L., Rossman, G.R., 1982. Case hardening of sandstone. Geology 10, 520525.
Coude-Gaussen, G., Rognon, P., Federoff, N., 1984. Piegeage de poussie`res
eoliennes dans des fissures de granitoides due Sinai oriental. Compte Rendus
de lAcademie des Sciences de Paris II, 369374.
Del Monte, M., Sabbioni, C., 1984. Gypsum crusts and fly ash particles on
carbonate outcrops. Archives for Meteorology, Geophysics and Bioclimatology
Series B 35, 105111.
Del Monte, M., Sabbioni, C., Ventura, A., Zappia, G., 1984. Crystal growth from
carbonaceous particles. Science of the Total Environment 36, 247254.
DiGregorio, B.E., 2010. Life on the rocks. New Scientist 2747, 4043.
Donoghue, E., Troll, V.R., Schwarzkopf, L.M., Clayton, G., Goodhue, R., 2009.
Organic block coatings in block-and-ash flow deposits at Merapi Volcano,
central Java. Geological Magazine 146, 113120.
Dorn, R.I., 1988. A rock varnish interpretation of alluvial-fan development in Death
Valley, California. National Geographic Research 4, 5673.
Dorn, R.I., 1997. Constraining the age of the Coa valley (Portugal) engravings with
radiocarbon dating. Antiquity 71, 105115.
Dorn, R.I., 1998. Rock Coatings. Elsevier, Amsterdam, 429 pp.
Dorn, R.I., 2001. Chronometric techniques: engravings. In: Whitley, D.S. (Ed.),
Handbook of Rock Art Research. Altamira Press, Walnut Creek, pp. 167189.
Dorn, R.I., 2007. Rock varnish. In: Nash, D.J., McLaren, S.J. (Eds.), Geochemical
Sediments and Landscapes. Blackwell, London, pp. 246297.
Dorn, R.I., 2009. The rock varnish revolution: New insights from microlaminations
and the contribution of Tanzhuo Liu. Geography Compass 3, 18041823.
Dorn, R.I., Dickinson, W.R., 1989. First paleoenvironmental interpretation of a
pre-Quaternary rock varnish site. Davidson Canyon, south Arizona. Geology 17,
10291031.
Dorn, R.I., Dragovich, D., 1990. Interpretation of rock varnish in Australia: case
studies from the Arid Zone. Australian Geographer 21, 1832.
Dorn, R.I., Krinsley, D.H., 1991. Cation-leaching sites in rock varnish. Geology 19,
10771080.
Rock Coatings
Dorn, R.I., Meek, N., 1995. Rapid formation of rock varnish and other rock coatings
on slag deposits near Fontana. Earth Surface Processes and Landforms 20,
547560.
Dorn, R.I., Oberlander, T.M., 1981. Microbial origin of desert varnish. Science 213,
12451247.
Dorn, R.I., Oberlander, T.M., 1982. Rock varnish. Progress in Physical Geography 6,
317367.
Dorn, R.I., Stasack, E., Stasack, D., Clarkson, P., 2001. Through the looking glass:
analyzing petroglyphs and geoglyphs with different perspectives. American
Indian Rock Art 27, 7796.
Dorn, R.I., Whitley, D.S., Cerveny, N.C., Gordon, S.J., Allen, C., Gutbrod, E., 2008.
The rock art stability index: a new strategy for maximizing the sustainability of
rock art as a heritage resource. Heritage Management 1, 3570.
Douglas, G.R., 1987. Manganese-rich rock coatings from Iceland. Earth Surface
Dragovich, D., 1986. Minimum age of some desert varnish near Broken Hill, New
South Wales. Search 17, 149151.
Dragovich, D., 1987. Weathering of desert varnish by lichens. In: Conacher, A.
(Ed.), Readings in Australian Geography. Proceedings of the 21st IAG
Conference, Perth, pp. 407412.
Ehlen, J., Wohl, E., 2002. Joints and landform evolution in bedrock canyons.
Transactions, Japanese Geomorphological Union 232, 237255.
Engel, C.G., Sharp, R.S., 1958. Chemical data on desert varnish. Geological Society
of America Bulletin 69, 487518.
Fleisher, M., Liu, T., Broecker, W., Moore, W., 1999. A clue regarding the origin of
rock varnish. Geophysical Research Letters 26(1), 103106.
Fortescue, J.A.C., 1980. Environmental Geochemistry. A Holistic Approach.
Springer, New York, NY, 347 pp.
Frazier, C.S., Graham, R.C., 2000. Pedogenic transformation of fractured granitic
bedrock, southern California. Soil Science Society of America Journal 64,
20572069.
Friedmann, E.I., Galun, M., 1974. Desert Algae, Lichen, and Fungi. Academic Press,
New York, NY, pp. 165212.
Gaier, J.R., 2005. The Effects of Lunar Dust on EVA systems During the Apollo
missions. NASA/TM 2005-213610. NASA Center for Aerospace Information,
Hanover, MD, 66 pp.
Ganor, E., Kronfeld, J., Feldman, H.R., Rosenfeld, A., Ilani, S., 2009. Environmental
dust: a tool to study the patina of ancient artifacts. Journal of Arid Environments
73, 11701176.
Gehrmann, C., Krumbein, W.E., Petersen, K., 1988. Lichen weathering activities on
mineral and rock surfaces. Studia Geobotanica 8, 3345.
Gilbert, G.K., 1877. Geology of the Henry Mountains. US Geological and
Geographical Survey, Washington, DC, 160 pp.
Giorgetti, G., Baroni, C., 2007. High-resolution analysis of silica and sulphate-rich
rock varnishes from Victoria Land (Antarctica). European Journal of Mineralogy
19, 381389.
Glazovskaya, M.A., 1968. Geochemical landscape and types of geochemical soil
sequences. Transactions of 9th International Congress Soil Science, Adelaide IV,
303311.
Glazovskaya, M.A., 1973. Technogiogeomes the intial physical geographic objects
of landscape geochemical prediction. Soviet Geography Review and Translation
14(4), 215228.
Golubic, S., Friedmann, E., Schneider, J., 1981. The lithobiontic ecological niche,
with special reference to microorganisms. Journal Sedimentary Petrology 51,
475478.
Gorbushina, A.A., 2007. Life on the rocks. Environmental Microbiology 9,
16131631.
Gordon, S.J., Dorn, R.I., 2005a. In situ weathering rind erosion. Geomorphology
67, 97113.
Gordon, S.J., Dorn, R.I., 2005b. Localized weathering: implications for theoretical
and applied studies. Professional Geographer 57, 2843.
Goudie, A.S., 1978. Dust storms and their geomorphological implications. Journal
of Arid Environments 1, 291310.
Ha-mung, T., 1968. The biological nature of iron-manganese crusts of soil-forming
rocks in Sakhalin mountain soils. Microbiology 36, 621624.
Haberland, W., 1975. Untersuchungen an Krusten, Wustenlacken und Polituren auf
Gesteinsoberflachen der nordlichen und mittlerent Saharan (Libyen und Tchad).
21, 177.
Harrington, C.D., 1986. Investigation of desert rock varnish on a cobble from a
stone burial mound at al-Farrah, al-Jubah Quadrangle, Yemen Arab Republic.
In: Overstreet, W.C.A.O. (Ed.), Geological and Archaeological Reconnaissance in
the Yemen Arab Republic, 1985, the Wadi Al-Jubah Archaeological Project.
American Foundation for the Study of Man, Washington, DC, vol. 4, pp. 4146.
95
Harrison, R., Frink, D.S., 2000. The OCR carbon dating procedure in Australia: new
dates from Wilinyjibari Rockshelter, southeastern Kimberley, Western Australia.
Australian Archaeology 51, 615.
Haskin, L.A., Wang, A., Jolliff, B.L., et al., 2005. Water alteration of rocks and soils
on Mars at the Spirit rover site in Gusev crater. Nature 436, 6669.
Hayden, J., 1976. Pre-altithermal archaeology in the Sierra Pinacate, Sonora,
Mexico. American Antiquity 41, 274289.
Hetu, B., Vansteijn, H., Vandelac, P., 1994. Flows of frost-coated clasts A recently
discovered scree slope grain-flow type. Geographie Physique et Quaternaire 48,
322.
Hodge, V.F., Farmer, D.E., Diaz, T.A., Orndorff, R.L., 2005. Prompt detection of
alpha particles from Po-210: another clue to the origin of rock varnish? Journal
of Environmental Radioactivity 78, 331342.
Hortola, P., 2005. SEM examination of human erythrocytes in uncoated bloodstains on
stone: use of conventional as environmental-like SEM in a soft biological tissue
(and hard inorganic material). Journal of Microscopy Oxford 218, 94103.
von Humboldt, A., 1812. Personal Narrative of Travels to the Equinoctial Regions of
America during the Years 17991804 V. II (translated and edited by T. Ross in
1907). George Bell and Sons, London, 521 pp.
Hungate, B., Danin, A., Pellerin, N.B., Stemmler, J., Kjellander, P., Adams, J.B.,
Staley, J.T., 1987. Characterization of manganese-oxidizing (MnIIMnIV) bacteria
from Negev Desert rock varnish: implications in desert varnish formation.
Canadian Journal Microbiology 33, 939943.
Hunt, A.G., Wu, J.Q., 2004. Climatic influences on Holocene variations in soil
erosion rates on a small hill in the Mojave Desert. Geomorphology 58,
263289.
Huyge, D., Watchman, A., De Dapper, M., Marchi, E., 2001. Dating Egypts oldest
art: AMS 14C age determinations of rock varnishes covering petroglyphs at
El-Hosh (Upper Egypt). Antiquity 75, 6872.
Jenne, E.A., 1968. Controls on Mn, Fe, Co, Ni, Cu and Zn concentrations in soils
and water: the significant role of hydrous Mn and Fe oxides. In: Gould, R.F.
(Ed.), Trace Inorganics in Water. American Chemical Society, Washington, DC,
pp. 337387.
Jordan, D.W., 1954. The adhesion of dust particles. British Journal of Applied
Physics 5, S194S198.
Kendall, C.G.S., Sadd, J.L., Alsharhan, A., 1994. Holocene marine cement coatings
on beach-rocks of the Abu-Dhabi coastline (UAE) Analogs for cement fabrics
in ancient limestones. Carbonates and Evaporites 9, 119131.
Kim, J.G., Lee, G.H., Lee, J., Chon, C., Kim, T.H., Ha, K., 2006. Infiltration pattern
in a regolith-fractured bedrock profile: field observations of a dye stain pattern.
Knauss, K.G., Ku, T.L., 1980. Desert varnish: potential for age dating via uraniumseries isotopes. Journal of Geology 88, 95100.
Konhauser, K.O., Fyfe, W.S., Schultze-Lam, S., Ferris, F.G., Beveridge, T.J., 1994.
Iron phosphate precipitation by epilithic microbial biofilms in Arctic Canada.
Canadian Journal Earth Science 31, 13201324.
Kraft, M.D., Michalski, J.R., Sharp, T.G., 2004. High-silica rock coatings: TES
surface-type 2 and chemical weathering on Mars. Lunar and Planetary Science
35, 1936.pdf.
Krinsley, D., 1998. Models of rock varnish formation constrained by high resolution
transmission electron microscopy. Sedimentology 45, 711725.
Krinsley, D., Dorn, R.I., DiGregorio, B.E., 2009. Astrobiological implications of rock
varnish in Tibet. Astrobiology 9, 551562.
Krinsley, D.H., Dorn, R.I., Tovey, N.K., 1995. Nanometer-scale layering in rock
varnish: implications for genesis and paleoenvironmental interpretation. Journal
of Geology 103, 106113.
Krumbein, W.E., 1979. Photolithotrophic and chemoorganotrophic activity of
bacteria and algae as related to beachrock formation and degradation (Gulf of
Aqaba, Sinai). Geomicrobiology Journal 1, 139203.
Kurtz, H.D., Netoff, D.I., 2001. Stabilization of friable sandstone surfaces in a
desiccating, wind-abraded environment of south-central Utah by rock surface
microorganisms. Journal of Arid Environments 48(1), 89100.
Langworthy, K., Krinsley, D., Dorn, R.I., 2010. High Resolution Transmission
Electron Microscopy evaluation of silica glaze reveals new textures. Earth
Lee, M.R., Parsons, I., 1999. Biomechanical and biochemical weathering of lichenencrusted granite: textural controls on organic-mineral interactions and
deposition of silica-rich layers. Chemical Geology 161, 385397.
Li, F.X., Margetts, S., Fowler, I., 2001. Use of chalk in rock climbing: sine quanon
or myth? Journal of Sports Sciences 19, 427432.
Liu, T., 2010. VML Dating Lab. http://www.vmldating.com/ (accessed March 2011).
Liu, T., Broecker, W.S., 2000. How fast does rock varnish grow? Geology 28,
183186.
96
Rock Coatings
Liu, T., Broecker, W.S., 2007. Holocene rock varnish microstratigraphy and its
chronometric application in drylands of western USA. Geomorphology 84, 121.
Liu, T., Broecker, W.S., 2008a. Rock varnish evidence for latest Pleistocene
millennial-scale wet events in the drylands of western United States. Geology
36, 403406.
Liu, T., Broecker, W.S., 2008b. Rock varnish microlamination dating of late
Quaternary geomorphic features in the drylands of the western USA.
Loendorf, L.L., 1991. Cation-ratio varnish dating and petroglyph chronology in
southeastern Colorado. Antiquity 65, 246255.
Loso, M.G., Doak, D.F., 2006. The biology behind lichenometric dating curves.
Oecologia 147, 229233.
Lytle, F.W., Pingitore, N.E., Lytle, N.W., Ferris-Rowley, D., Reheis, M.C., 2002. The
possibility of dating petroglyphs from the growth rate of desert varnish. Nevada
Archaeological Association Meetings Abstracts.
Marchant, D.R., Denton, G.H., Swisher, C.C., Potter, N., 1996. Late Cenozoic
Antarctic paleoclimate reconstructed from volcanic ashes in the Dry Valleys
region of southern Victoria Land. Geological Society of America Bulletin 108,
181194.
Marshall, R.R., 1962. Natural radioactivity and the origin of desert varnish.
Transactions of the American Geophysical Union 43, 446447.
Marston, R.A., 2003. Editorial note. Geomorphology 53, 197.
McAdam, A.C., Leshin, L.A., Harvey, R.P., 2002. Investigation of weathering
products of Martian meteorite analog materials on Earth and applications to
Martian dust analysis. Astrobiology 218, 577578.
McAlister, J.J., Smith, B.J., Torok, A., 2006. Element partitioning and potential
mobility within surface dusts on buildings in a polluted urban environment,
Budapest. Atmospheric Environment 40, 67806790.
Mellor, A., Short, J., Kirkby, S.J., 1997. Tafoni in the El Chorro area, Andalucia,
southern Spain. Earth Surface Processes and Landforms 22, 817833.
Meunier, A., Sardini, P., Robinet, J.C., Pret, D., 2007. The petrography of
weathering processes: facts and outlooks. Clay Minerals 42, 415435.
Milnes, A.R., Wright, M.J., Thiry, M., 1991. Silica accumulations in saprolites and
soils in South Australia. In: Nettleton, W.D. (Ed.), Occurrence, Characteristics
and Genesis of Carbonate, Gypsum and Silica Accumulations in Soils. Soil
Science Society of America Special Publication 26, Madison, pp. 121149.
Minitti, M.E., Weitz, C.M., Lane, M.D., Bishop, J.L., 2005. Rock coatings from
vulcano, a Martian analog environment. Lunar and Planetary Science 36, 1835.pdf.
Minitti, M.E., Weitz, C.M., Lane, M.D., Bishop, J.L., 2007. Morphology, chemistry,
and spectral properties of Hawaiian rock coatings and implications for Mars.
Journal of Geophysical Research 112(E05015), http://dx.doi.org/10.1029/
2006JE002839.
Mottershead, D.N., Pye, K., 1994. Tafoni on coastal slopes, South Devon, UK. Earth
NASA, 2010. HUalalai Satelllite Images Aster Volcano Archive. http://
ava.jpl.nasa.gov/volcano.asp?vnum 1302-04 (accessed March 2011).
NIAIST, N.I.o.A.I.S.a.T., 2005. Atlas of EhpH diagrams. Intercomparison of
thermodynamic data bases. Geological Survey of Japan Open File Report 419,
1287.
Northup, D.E., Snider, J.R., Spilde, M.N., et al., 2010. Diversity of rock varnish
bacterial communities from Black Canyon, New Mexico. Journal of Geophysical
Research 115(G02007), http://dx.doi.org/10.1029/2009JG001107.
Oguchi, C.T., Matsukura, Y., Kuchitsu, N., 2002. Environmental and seasonal
influences on the spatial distribution of salt efflorescence and weathering on brick
kiln walls. Transactions, Japanese Geomorphological Union 232, 335348.
Ortega-Calvo, J.J., Arino, X., Stal, L.J., Saiz-Jimenez, C., 1994. Cyanobacterial
sulfate accumulation from black crust of a historic building. Geomicrobiology
Journal 12, 1522.
Palmer, E., 2002. Feasibility and implications of a rock coating catena: analysis of a
desert hillslope. M.A. Thesis Arizona State University, Tempe.
Paradise, T.R., 1997. Sandstone weathering from lichen overgrowth on Red
Mountain, Arizona. Geografiska Annaler A 79, 177184.
Patyk-Kara, N.G., Gorelikova, N.V., Plakht, J., Nechelyustov, G.N., Chizhova, I.A.,
1997. Desert varnish as an indicator of the age of Quaternary formations
(Makhtesh Ramon Depression, Central Negev). Transactions (Doklady) of the
Russian Academy of Sciences/Earth Science Sections 353A, 348351.
Pedley, H.M., 1990. Classification and environmental models of cool freshwater
tufas. Sedimentary Geology 68, 143154.
Pentecost, A., 1985. Association of cyanobacteria with tufa deposits: identify,
enumeration and nature of the sheath material revealed by histochemistry.
Geomicrobiology Journal 4, 2398.
Perelman, A.I., 1961. Geochemical principles of landscape classification. Soviet
Geography Review and Translation 11(3), 6373.
Perelman, A.I., 1966. Landscape Geochemistry. (Translation No. 676, Geological

Survey of Canada, 1972). Vysshaya Shkola, Moscow, 388 pp.
Polynov, B.B., 1937. The Cycle of weathering. [Translated from the Russian by A.
Muir]. Nordemann Publishing, New York, 220 pp.
Potgieter-Vermaak, S.S., Godol, R., Van Grieken, R., Potgieter, J.H., Oujja, M.,
Catillejo, M., 2004. Micro-structural characterization of black crust and laser
cleaning of building stones by micro-Raman and SEM techniques.
Spectrochimica Acta Part A: Molecular and Biomolecular Spectroscopy 61,
24602467.
Potter, R.M., 1979. The tetravalent manganese oxides: clarification of their structural
variations and relationships and characterization of their occurrence in the
terrestrial weathering environment as desert varnish and other manganese
oxides. PhD Thesis, California Institute of Technology, Pasadena.
Potter, R.M., Rossman, G.R., 1977. Desert varnish: the importance of clay minerals.
Science 196, 14461448.
Raiswell, R., Benning, L.G., Davidson, L., Tranter, M., Tulaczyk, S., 2009.
Schwertmannite in wet, acid, and oxic microenvironments beneath polar and
polythermal glaciers. Geology 37, 431434.
Robinson, D.A., Williams, R.B.G., 1992. Sandstone weathering in the High Atlas,
Morocco. Zeitschrift fur Geomorphologie 36, 413429.
Rowe, M.W., 2001. Dating by AMS radiocarbon analysis. In: Whitley, D.S. (Ed.),
Handbook of Rock Art Research. Altamira Press, Walnut Creek, pp. 139166.
Rundel, P.W., 1978. Ecological relationships of desert zone lichens. Bryologist 81,
277293.
Schlesinger, W.H., 1985. The formation of caliche in soils of the Mojave Desert,
California. Geochimica et Cosmochimica Acta 49, 5766.
Shiffman, P., Zierenberg, R., Marks, N., Bishop, J.L., Dyar, M.D., 2006.
Acid-fog deposition at Kilauea volcano: a possible mechanism for the
formation of siliceous-sulfate rock coatings on Mars. Geology 34,
921924.
Simionescu, B., Alflori, M., Olaru, M., 2009. Protective coatings based on
silsesquioxane nanocomposite films for building limestones. Construction and
Building materials 23, 34263430.
Smith, B.J., 1994. Weathering processes and forms. In: Abrahams, A.D., Parsons,
A.J. (Eds.), Geomorphology of Desert Environments. Chapman and Hall,
London, pp. 3963.
Smith, B.J., McAlister, J.J., Baptista-Neto, J.A., Silva, M.A.M., 2007. Postdepositional modification of atmospheric dust on a granite building in central
Rio de Janeiro: implications for surface induration and subsequent stone decay.
Geological Society, London, Special Publications 271, 153166.
Smith, G.A., Turner, W.G., 1975. Indian Rock Art of Southern California with
Selected Petroglyph Catalog. San Bernardino County Museum Association,
Redlands, 150 pp.
Souza-Egipsy, V., Wierzchos, J., Sancho, C., Belmonte, A., Ascaso, C., 2004. Role
of biological soil crust cover in bioweathering and protection of sandstones in a
semi-arid landscape (Torrollones de Gabarda, Huesca, Spain). Earth Surface
Spades, S., Russ, J., 2005. GC-MS analysis of lipids in prehistoric rock paints and
associated oxalate coatings from the lower Pecos region. Texas. Archaeometry
45, 115126.
Suzuki, T., 2002. Rock control in geomorphological processes: research history in
Japan and perspective. Transactions, Japanese Geomorphological Union 232,
161199.
Thiagarajan, N., Lee, C.A., 2004. Trace-element evidence for the origin of desert
varnish by direct aqueous atmospheric deposition. Earth and Planetary Science
Letters 224, 131141.
Verge`s-Belmin, Pichot, C., Orial, G., 1993. Elimination de croutes noires sur
marbre et craie: a` quel niveau arreter le nettoyage? In: Thiel, M.-J. (Ed.),
Conservation of Stone and Other Materials. E and FN Spon, London,
pp. 535541.
Viles, H.A., 1995. Ecological perspectives on rock surface weathering: towards a
conceptual model. Geomorphology 13, 2135.
Viles, H.A., Goudie, A.S., 1990. Tufas, travertines and allied carbonate deposits.
Viles, H.A., Goudie, A.S., 2004. Biofilms and case hardening on sandstones
from Al-Quawayra, Jordan. Earth Surface Processes and Landforms 29,
14731485.
Villa, N., Dorn, R.I., Clark, J., 1995. Fine material in rock fractures: aeolian dust or
weathering? In: Tchakerian, V. (Ed.), Desert Aeolian Processes. Chapman and
Hall, London, pp. 219231.
Watchman, A., 1997. Differences of interpretation for Foz Coa dating results.
National Pictographic Society Newsletter 8(1), 7.
Rock Coatings
Watchman, A., 2000. A review of the history of dating rock varnishes. Earth-Science
Reviews 49, 261277.
Watchman, A., 2002. A reply to Whitley and Simon. INORA 34, 1112.
Watchman, A., David, B., McNiven, I., Flood, J., 2000. Microarchaeology of
engraved and painted rock surface crusts at Yiwaralarlay (The Lightning Brothers
site), Northern Territory, Australia. Journal of Archaeological Science 27,
315325.
Watchman, A., OConnor, S., Jones, R., 2005. Dating oxalate minerals 2045 ka.
Journal of Archaeological Science 32, 369374.
Wayne, D.M., Diaz, T.A., Fairhurst, R.J., Orndorff, R.L., Pete, D.V., 2006. Direct
major-and trace-element analyses of rock varnish by high resolution laser
ablation inductively-coupled plasma mass spectrometry (LA-ICPMS). Applied
Geochemistry 21, 14101431.
Weaver, C.E., 1978. Mn-Fe coatings on saprolite fracture surfaces. Journal of
Sedimentary Petrology 48, 595610.
Whitley, D.S., Annegarn, H.J., 1994. Cation-ratio dating of rock engravings from
Klipfontein, Northern Cape Province, South Africa. In: Dowson, T.A., LewisWilliams, J.D. (Eds.), Contested Images: Diversity in Southern African Rock Art
Research. University of Witwatersrand Press, Johannesburg, pp. 189197.
97
Whitley, D.S., Simon, J.M., 2002a. Recent AMS radiocarbon rock engraving dates.
INORA 32, 1116.
Whitley, D.S., Simon, J.M., 2002b. Reply to Huyge and Watchman. INORA 34,
1221.
Whitley, D.S., Simon, J.M., Dorn, R.I., 1999. Vision quest in the Coso Range.
American Indian Rock Art 25, 131.
Xu, H., Chen, T., Konishi, H., 2010. HRTEM investigation of trilling todorokite and
nano-phase Mn-oxides in manganese dendrites. American Mineralogist 95,
556562.
Yoshikawa, K., Okura, Y., Autier, V., Ishimaru, S., 2006. Secondary calcite
crystallization and oxidation processes of granite near the summit of Mt.
McKinley, Alaska. Geomorhpologie Relief Processus Enviroment 3, 197204.
Zhang, Y., Liu, T., Li, S., 1990. Establishment of a cation-leaching curve of rock
varnish and its application to the boundary region of Gansu and Xinjiang,
western China. Seismology and Geology (Beijing) 12, 251261.
Zorin, Z.M., Churaev, N., Esipova, N., Sergeeva, I., Sobolev, V., Gasanov, E., 1992.
Influence of cationic surfactant on the surface charge of silica and on the
stability of aqueous wetting films. Journal of Colloid and Interface Science 152,
170182.
Biographical Sketch
Ronald I Dorn has been a professor of geography at Arizona State University, Tempe, AZ since 1988. He served
previously on the faculty at Texas Tech University. He is coordinator of the Arizona Geographic Alliance, a K-12
outreach program to promote geographic education in Arizona. He has been president and secretary/treasurer of
the Geomorphology Specialty Group and chair of the Nystrom Committee of the Association of American
Geographers. He is a fellow of the Geological Society of America and the Arizona/Nevada Academy of Science, a
Guggenheim Fellow, and a member of the Association of American Geographers and the American Rock Art
Research Association. He has received the Kirk Bryan Award, the G.K. Gilbert Award, the Wiley Award, and the
Castleston Award for his research with coauthors on rock coatings.
4.6 Weathering Rinds: Formation Processes and Weathering Rates

CT Oguchi, Geosphere Research Institute, Saitama, Japan
4.6.1
4.6.2
4.6.3
4.6.4
4.6.5
4.6.6
References
Introduction
Previous Research on Weathering Rinds
Temporal Changes in Rock Properties
Formation Processes of Weathering Rinds
A Porosity Concerned Model of Weathering Rind Development
Conclusions
Glossary
L*a*b* chromatic values It is a color-opponent space
with dimension L* for lightness and a* and b* for the
color-opponent dimensions of red-green and yellow and
blue, respectively.
Vickers microhardness The Vickers hardness test was
developed in 1924 by Smith and Sandland at Vickers Ltd.
98
99
99
104
105
108
108
The hardness number, Vickers Pyramid Number, is

determined by the load over the surface area of the
indentation and not the area normal to the force.
Weathering rinds The outer layer of a pebble, boulder, or
other rock fragment that has formed as a result of chemical
weathering.
Abstract
Models of weathering rind development have been discussed by many researchers. To understand this phenomenon,
considering suitable rock properties is important because weathering rates differ from rock property types. This chapter
briefly reviews the modeling types such as logarithmic, power, and relaxation function of previous researches, and then,
describes rock property changes by giving an example of an andesite rock type. The properties are major 10 elements, L,
a, and b chromatic values, bulk density, and porosity determined by mercury porosimetry, and Vickers microhardness.
This chapter also considers a formation process model of weathering rind development and a growth model of weathering rind thickness. A conceptual model of weathering-rind formation suggests that the inner white band is produced
by dissolution of alkali/alkaline earth metals related to the inward and subsequent outward movement of water. The
brown band is probably formed by both leaching of these metals and oxidation of irons. A porosity concerned growth
model, consisting of two diffusion equations based on the two bands, was proposed. One is a normal diffusion equation
showing the development of strongly weathered brown bands due to both oxidation and leaching. The other is an
equation with the diffusion coefficient exponentially related to porosity of the host rock showing the growth of total
weathering rinds mainly due to leaching. The important properties to consider weathering-rind developments are contents of iron and alkali/alkaline earth metals as well as porosity of rocks. Environmental conditions are also important
factors to determine the degree of oxidation or dissolution of sub-bands of weathering-rinds, although these studies are
expected in the near future.
4.6.1
Introduction
The meaning of the term weathering rate is slightly different

among researchers. The term has commonly been used as a
synonym of the chemical denudation rate (e.g., Waylen, 1979;
High and Hanna, 1970; Trudgill, 1975, 1976; Trudgill et al.,
Oguchi, C.T., 2013. Weathering rinds: formation processes and weathering

rates. In: Shroder, J. (Editor in Chief), Pope, G.A. (Ed.), Treatise on
Geomorphology. Academic Press, San Diego, CA, vol. 4, Weathering and
Soils Geomorphology, pp. 98110.
98
1981; Crabtree and Trudgill, 1985; Hirose et al., 1994, 1995).

The term also represents the formation rates of weathering
products that include soils and clay minerals (e.g., Alexander,
1985; Wakatsuki and Rasyidin, 1992; Garrels and Mackenzie,
1967; Yoshioka, 1975; and Suzuki and Hachinohe, 1995),
dated materials such as volcanic ash (e.g., Hay, 1960; Leneuf
and Aubert, 1960; Trendall, 1962; Ruxton, 1968; Haantjens
and Bleeker, 1970; Menard, 1974; Nahon and Lappartient,
1977; Amit et al., 1993), and thin weathered zones of rocks
such as weathering rinds, rock varnish, and hydration layers of
obsidian artifacts (e.g., Friedman and Smith, 1960; Friedman
and Long 1976; Katsui and Kondo, 1965). Another usage of
http://dx.doi.org/10.1016/B978-0-12-374739-6.00067-1
Weathering Rinds: Formation Processes and Weathering Rates
the term is the changing rates of rock properties with weathering. Kimiya (1975a, b) and Crook and Gillespie (1986)
investigated fluvial-terrace gravel of different ages and concluded that rock strength decreases exponentially with increasing weathering time. Oguchi et al. (1999) also pointed
out that both compressive and tensile strengths declined
drastically in the early stage of weathering. Such changes in
rock strength are crucial to discuss the effects of weathering on
geomorphic processes. However, studies on the changing rates
of rock properties such as mineralogical, chemical, physical,
and mechanical properties have been limited in number.
Analyses of various properties of weathering products are
also important because weathering progresses with simultaneous changes in several rock properties. For example,
chemical changes characterstically result in the reduction of
rock strength (Matsukura et al., 1983; Oguchi and Matsukura,
1999a). In order to discuss their relationships, both chemical
and mechanical properties should be analyzed. Most of the
previous studies on weathering, however, dealt with only
changing chemical and mineralogical properties (e.g., Craig
and Loughnan, 1969; Singer, 1984; Chesworth et al. 1981). So
far a limited number of studies (Saito et al., 1974; Eggleton
et al., 1987; Waragai, 1993; Oguchi and Matsukura, 1999a)
have investigated chemical and mineralogical properties along
with mechanical and physical properties.
4.6.2
Previous Research on Weathering Rinds
Weathering rinds that have developed on the surface of rocks

are useful for the study of weathering because very detailed
investigation can be performed on the rinds with small areal
extent. Moreover, the rates of change in rock properties due to
long-term weathering can be estimated using the weathering
rinds of rocks in dated deposits.
Table 1 is the list of the previous studies. Most of the
studies have generally been confined to the estimation of the
ages of Quaternary deposits. Cernohouz and Solc (1966) first
proposed that the relationship between weathering-rind
thickness and formative time is expressed by a logarithmic
function:
d A log 1 Bt
1
where d is weathering rind thickness, t is time, and A and B are

constants. Their research was made using basaltic rocks and
they termed weathering rind as weathering crust. Thus, the
term weathering crust includes the meaning of a thin weathered layer made on the gravel surface.
Absolute-age functions can be determined if some of the
deposits with weathering rinds are dated using other methods
such as 14C measurements. If the function is given, ages of
landforms can be calculated by substituting measured weathering-rind thickness into the equation. This method has been
applied in many examples to glacial deposits in high mountains
(e.g., Birkeland, 1973; Porter, 1975; Burke and Birkeland, 1979;
Anderson and Anderson, 1981; Chinn, 1981; Colman, 1981,
1982a, 1982b; Colman and Pierce, 1981; Whitehouse et al.,
1986; Knuepfer, 1988; Shiraiwa and Watanabe, 1991; Koizumi
and Seki, 1992; Koizumi and Aoyagi, 1993; Aoki, 1994).
99
Several studies proposed equations of weathering rind

development. A logarithmic equation was proposed not only
by Cernohouz and Solc (1966) but also by Colman and Pierce
(1981) (Table 1). Chinn (1981), Knuepfer (1988) and Oguchi
(2004) proposed power functions. Whitehouse et al. (1988)
modeled the phenomena using relaxation functions. Furthermore, Sak et al. (2004) proposed equilibrium dissolution
controlled by an interface reaction.
Some studies have examined not only thickness but also
other properties of weathering rinds. The series of studies by
Colman and Pierce (1981) and Colman (1981, 1982a) investigated weathering rinds on andesitic and basaltic stones
developed on the gravels of glacial deposits at 150 sites and 17
different areas in the Western US. Colman and Pierce (1981)
studied them from the viewpoint of a Quaternary age indicator, whereas Colman (1982a) focused at chemical and
mineralogical alteration on weathering rinds. Kuchitsu (1991)
identified the minerals of weathering rinds formed on lithic
artifacts. Matsukura et al. (1994a, 1994b), Oguchi and
Matsukura (1999b), and Oguchi (2001, 2004) studied mineralogical, chemical, and mechanical properties as well as
colors of weathering rinds on andesite gravel from central and
south Japan. Furthermore, biological effects on weathering
rind development were explained by Etienne (2002), in
which microerosion was considered. Dixon et al. (2002)
pointed out the difference between weathering rind and rock
coatings. Gordon and Dorn (1983) pointed out in situ weathering rind erosion using cosmogenic nuclides. These studies
are important in the application for absolute dating using
weathering rinds.
Systematic studies based on concurrent investigation of
several rock properties had become necessary to establish a
model for the formation of weathering rinds. The reason is
that the difference of rock property types sometimes shows
different weathering degrees. The relationship between weathering-rind properties and weathering time remains to be
determined. Arguments on the environmental influences were
also important. Considering these outstanding problems,
there is need to confirm the consistency between visible
characteristics and several rock properties: (i) weathering rinds
generally consist of sub-zones with different colors, and (ii)
the zone with a certain color near the rock surface may not
strictly correspond to the zone subjected to weathering. Thus,
systematic investigations of weathering rinds and inner zones
are important not only for improved understanding of rock
weathering, but also for the re-evaluation of dating methods
using weathering-rind thickness.
4.6.3
Temporal Changes in Rock Properties
This case study explains weathering rinds on andesite gravel in

river terrace deposits in central Japan (Oguchi and Matsukura,
1999a; Oguchi, 2001, 2004), arriving at general observations
relevant to weathering rinds in general. The rock samples were
collected from coalescing alluvial fans in Nasuno-ga-hara
area (361340 371050 N, 1391500 1411100 E, 120560 m asl)
(Figure 1). This area belongs to the humid temperate climate
region and has a mean annual temperature of 11.1 1C
( 0.2 1C in the coldest (January), 23.3 1C in the warmest
100
Rock type
Study area
Deposited
condition
Rock properties
Min.
Chem.
Phys.
Mech.
Color
Rate
Equation type
Reference
Logarithmic
Cernohouz and Solc (1966)
Basalt
Bohemia
Granite
Basalt
Colorado
North Cascade
Moraine
Moraine

Granite
Sandstone (A)a
Sierra Nevada
U.S. (Utah)
Moraine
Moraine

Sandstone (G)b
Andesite, Basalt
Andesite
Basalt
Sandstone (G)b
Sandstone (G)b
Gneiss
New Zealand
Western U.S.
ditto
ditto
New Zealand
New Zealand
Nepal Himalaya
Moraine
Moraine
ditto
ditto
Mne
Terrace
Moraine

L
4.64log(1 0.01t)
(L:thickness, t:
time)
45 mm/10 000 yr
(esti. 0.30.7 mm/
14 000 yr)
3 mm/10 000 yr
(Estimated 5 mm/
6000 yr)
6 mm/9500 yr

3 mm/315 000 yr
1.6 mm/140 000 yr
8 mm/20 000 yr
7 mm/15 000 yr
4.5 mm/3300 yr
Hornfels
Granodiorite
Quartz porphyry
Granodiorite
Andesite
Andesite
Andesite
Japan (Kanagawa)
Cen. Japan Alps
North. Japan Alps
Cen. Japan Alps
South Japan
ditto
Cen. Japan
Lithic artifacts
Moraine
Moraine
Moraine
Exposure
ditto
River terrace

4 mm/2000 yr
5 mm/20 000 yr
8 mm/50 000 yrs
7.8 mm/3500 yr

5.6 mm/660 000 yr
Andesite
Andesite
Basalt
ditto
ditto
Costa Rika
ditto
ditto
River terrac

3 mm/1000 yr
Quartzarenite.
Greywacke.
Birkeland (1973)
Porter (1975)
Power function
Logarithmic
ditto
Relaxation
Power function
Power function
Equilibrium dissolution
controlled by an
interface reaction
Burke and Birkeland (1979)

Anderson and Anderson
(1981)
Chinn (1981)
Colman (1982a, b)
Colman and Pierce (1981)
ditto
Whitehouse et al. (1986)
Knuepfer (1988)
Shiraiwa and Watanabe
(1991)
Kuchitsu (1991)
Koizumi and Seki (1992)
Koizumi and Aoyagi (1993)
Aoki (1994)
Matsukura et al. (1994a)
Matsukura et al. (1994b)
Oguchi and Matsukura
(1999b)
Oguchi (2001)
Oguchi (2001)
Sak et al. (2004)
Table 1 Summary of the previous studies on weathering rinds
139 50
101
140 0
0
400 km
Sea of Japan
R. N
aka
Pacific ocean
37 0
37 0
Kuroiso
R. sab
1
2
Sekiya
R.
ma
Ku
4
5
Sabui
6
7
R.
Nishinasuno
ki
Ho
Ohtawara
9
Kurobane
Nozaki
36 50
36 50
139 50
N
Sarado
0 1 2 3 4 5 km
140 0
Figure 1 Sample site locations within Nasuno-ga-hara, central Japan. 1: River Bed (0-ka surface); 2: Lower Terrace II; 3: Lower Terrace I (20ka surface); 4: Middle Terrace; 5: Upper Terrace (130-ka surface); 6: Lower Hill (290-ka surface); 7: Upper Hill (660-ka surface); 8: Mountains; 9:
Sampling sites. Reproduced from Oguchi, C.T., 2004. A porosity-related diffusion model of weathering-rind development. Catena 58, 6575.
(August) season), and has a mean annual precipitation of

1298 mm, according to the data from the nearest meteorological station (371080 N, 1401130 E, 354 m asl) (Japan Meteorological Agency, 1991). The fan deposits consist
predominantly of andesite pebbles and cobbles supplied from
volcanoes in the upstream area. All fans consist of andesite

gravel in loam including volcanic ash soils and represent
Quaternary fluvial surfaces of different ages (Watanabe and
Sagehashi, 1960; Koike, 1961; Suzuki et al., 1998). The following five surfaces suitable for rock sampling were investigated
102
in this paper: the present River Floodplain; the Lower Terrace I

formed at 20 ka BP; the Upper Terrace formed at 130 ka BP;
the Lower Hill formed at 290 ka; and the Upper Hill formed
approximately 660 ka BP. These ages are based on tephra
stratigraphy on the terrace deposits (Suzuki et al., 1998).
It is assumed that only fresh rocks were deposited on
floodplains and that weathering started soon after the terrace
formation. Thus, the time between the age of each terrace and
the present can be assumed to be the weathering period. The
studied geomorphic surfaces and rocks are, hereafter, designated by age as 0 ka, 20 ka, 130 ka, 290 ka, and 660 ka, respectively. The weathering of the gravel in the terrace deposits
may also have been affected by past changes in the weathering
environment. Takahashi and Hayakawa (1995) implied that
the air temperature in central Japan gently fluctuated between
0 and 7 1C during the Quaternary. This difference in temperature is small enough to cause only minor changes in
chemical reaction rates based on Arrhenius equation. Hence,
in the present study, it is assumed that weathering conditions
have been approximately constant over the past several hundreds of thousands of years.
Figure 2 shows rock samples with weathering rinds of
different ages. All the 0-ka rocks have fresh textures through
the rock surface to the interior. No brown weathered layers can
be observed. The 20-ka rocks have no distinctive weathered
layers, although these samples have very thin layers of
alteration that can be observed by the naked eye. The 130-ka,

290-ka, and 660-ka rock samples have both the outermost
weathered layers and inner fresh or relatively fresh parts. The
former layers with black or brown color under microscope
correspond to brown bands by naked-eye observation of cut
rocks, whereas the latter parts with pale colors are related to
the interior. The boundary between the brown bands and the
interior is sharp.
Chromatic, mineralogical, chemical, physical, and mechanical properties of the weathering rinds were determined.
Quantitative color measurements using a visible microspectrometer provide reproducible color data as well as information about the existence of ferric oxides and hydroxides
whose identification is otherwise difficult (Nakashima et al.,
1992). The color measurement yields three values of color
spaces denoted conveniently by L, a, and b, which can be
calculated from the basic spectral data. The L a b color
system is part of a standard methodology to describe colors in
a quantitative way, because they have uniform color space
(Hunt, 1980). L shows lightness, in which L 0 corresponds
to black, whereas L 100 corresponds to white. Both color
indicators a and b show chroma. A positive value of a
expresses red and a negative one indicates green, whereas a
positive value of b shows yellow and a negative one
Weathering rind
White band
80
Brown band
70
Rock interior
60
L*
0 ka
50
40
30
20
20 ka
a*
10
5
0
130 ka
290 ka
b*
5
40
30
20
10
0
10
10
12
14
Distance from rock surface (mm)

660 ka
Figure 2 Studied rock sample with weathering rinds.
Figure 3 Color changes with depth from the rock surface. The
values of L, a, and b represents lightness, redness and
yellowness, respectively. Modified from Oguchi, C.T., 2001. Formation
of weathering rinds on andesite. Earth Surface Processes and

represents blue. Figure 3 shows the L, a, and b values
plotted against the depth from the rock surface. The sample
shows that L values for the rock interior (more than 10 mm
depth) increase with decreasing depth, but a and b values do
not change. The values of a and b for the white band (from
10 mm to 4 mm depth) are ca. 0 and ca. 10, respectively. The
L values for the white band are constant with ca. 50, showing
whitish gray color. The L, a, and b values for the brown
band (from 4 mm depth to the rock surface) are from 55 to
65, from 0 to 6, and from 10 to 40, respectively.
Minerals were identified by X-ray diffraction analysis
(XRD). The fine powdered samples were analyzed to identify
both original rock-forming minerals and clay minerals. Most
samples from the brown layers consist mainly of kaolin
minerals, feldspar, pyroxene, and quartz. The main weathering
products are kaolin minerals derived from weathering of
feldspar and pyroxene.
Chemical composition of bulk samples was determined by
X-ray fluorescence analysis (XRF). Samples were collected
from the brown layers of 130-ka, 290-ka, and 660-ka rocks as
well as the interior of 0-ka, 20-ka, 130-ka, 290-ka, and 660-ka
rocks. The brown layers of 0-ka and 20-ka rocks could not be
collected because they were very thin. The collected samples
were ground into powders fine enough for analysis. Ten major
oxides were analyzed (SiO2, TiO2, Al2O3, FeO Fe2O3, MnO,
MgO, CaO, Na2O, K2O, and P2O5) and the relationship between chemical compositions for the brown layers and those
for the interior is shown in Figure 4. The proportion of the
contents of Na2O, CaO, and MgO for the brown bands to
those for the interior is 0.03:1 at minimum. On the other
hand, the proportion of the contents of ignition loss is 10
times higher than that of brown bands. The proportion of the
FeO+Fe2O3 content is 5 times higher that of brown bands.
There is little difference in total- and SiO2-values between the
brown layers and the interior. Electron-probe microanalysis
(EPMA) was conducted to examine element concentration of
the area (qualitative mapping) through rock surface to the
interior. Samples with thick and distinct weathering rinds

(660-A1) were analyzed using polished thin sections cut
normal to rock surfaces. Figure 5 shows the colored composition maps of nine elements: Si, Ti, Al, Fe, Ca, Mg, Na, K,
and Mn. The maps cover the zone from the interior to the
surface of the rocks. Color steps indicate the relative content of
each element: the white step corresponds to larger quantities
and the black step corresponds to pore space. The amounts of
most elements of the sample 660-A1 are also small in the
brown bands (the part between ca. 6.0-mm depth and the
rock surface) and large in the interior (Figure 5).
The contents of Si and Al are small within the brown bands,
but they are large at the outermost zone of 2-mm thickness.
The grains enriched in Si are thought to be secondary accumulated minerals. The contents of Ti are small in the brown
bands and large in the interior. The contents of Fe within the
groundmass of the brown bands are smaller than those of the
interior, but it is large within the outermost zone of ca. 3-mm
thickness. Other elements (K, Na, Mg, Ca, and Mn) for the
brown bands have amounts smaller than those for the interior,
and especially, Ca was much more leached out.
Physical properties such as bulk density, porosity, and
other physical properties are technically difficult to obtain
for weathering rinds. The common method uses cored rock
samples because weathering rinds are so thin. Mercury intrusion porosimetry (MIP), however, provides a method to investigate physical properties related to pores of rocks from the
measurement of pore-size distribution. This MIP measurement system also yields values of total pore volume (Vt) and
bulk density (rbulk). Thus, porosity (n) can be calculated by
multiplying these two values. Table 2 shows bulk density
and the calculated porosity. The interior of all the rocks have
the large bulk density of ca. 2.42.5 g cm3 on average. In
contrast, the bulk density for the brown bands of 130-ka,
290-ka, and 660-ka rocks are ca. 1.21.4 g cm3 on average.
Accordingly, porosity for the brown layers is larger than that
for the interior because of much larger pore volume. The
1000.00
10
10
Total
SiO2
AI2O3
TiO2
Fe2O3
MnO
Mg0
CaO
100.00
Interior (wt.%)
0.1
10.00
1.00
Na2O
K2O
P2O5
lg. loss
0.10
0.01
0.01
0.10
1.00
103
10.00
100.00
1000.00
Brown layer (wt.%)

Figure 4 Relationship between chemical composition for the brown bands and those for the interior. Reproduced from Oguchi, C.T., 2000.
Rates of rock property changes with weathering: andesite gravel in fluvial terrace deposits in Nasuno-ga-hara, Japan. Science Reports of the
Institute of Geoscience, University of Tsukuba, Section A 21, 5988.
104
Weathering rind
Interior
Si
Al
Fe
Ca
Count
7000
6000
Mg
5000
4000
3000
2000
Mn
1000
0
Figure 5 EPMA mapping of weathering rind. Modified from Oguchi, C.T., 2000. Rates of rock property changes with weathering: andesite gravel
in fluvial terrace deposits in Nasuno-ga-hara, Japan. Science Reports of the Institute of Geoscience, University of Tsukuba, Section A 21, 5988.
interior of all the rocks have a porosity of ca. 25% on average,

and the brown bands of 130-ka, 290-ka, and 660-ka rocks
have a porosity of ca. 3040% on average. As for mechanical
properties, the Vickers Hardness Number (VHN) (Smith and
Sandland, 1922) is a useful indicator. The results of the VHN
measurement are shown in Figure 6. Based on the changing
patterns of the VHN values (in gram-force per square micrometer, gf/mm2, units), three zones were identified as follows:
1. Brown bands have small VHN values from 10 gf mm2 to
80 gf mm2.
2. White bands have abrupt or gradual increase in VHN values with increasing depth, in which VHN values vary from
ca. 100 gf mm2 to ca. 500 gf mm2.
3. Rock interiors have large and constant VHN values of
ca. 500 gf mm2.
4.6.4
Formation Processes of Weathering Rinds
Figure 7 shows the changes in normalized values of the

four properties with depth. The maximum value of each parameter is set to 100% and the minimum to 0%. The changes
in values of CaO and FeO Fe2O3 are chosen because CaO
has been leached from the deepest point and the pattern of
change in FeO Fe2O3 is different from that of the other
elements. In the rock interior between depths of 15 and
10 mm, the CaO content decreases slightly with decreasing
depth, whereas L values slightly increase with decreasing
depth. In the white band, the contents of FeO p
Fe
2O3 and
L values increase with decreasing depth and a2 b2
values increase slightly between 10 and 7 mm depth, whereas

the contents of CaO and VHN values decrease with decreasing
depth and the contents of FeO Fe2O3 increase with
decreasing depth between 7 and 4 mm depth. In the outer
brown band (less than 4 mm depth), the contents of CaO
and VHN values decrease with decreasing depth, whereas
the contents of FeO Fep
2O
3 increase
with decreasing depth.

The values of L and a2 b2 first increase and then
decrease with decreasing depth. From these findings, rock
properties of the three parts can be summarized as follows.
The outer brown band is characterized by decreased alkali/
alkaline earth metal, FeO Fe2O3 contents, increased L, a,
and b values, and very low VHN values. The white band is
characterized by decreasing alkali/alkaline earth metals and
VHN values toward the surface and medium L values. The
rock interior is characterized by consistently high contents of
alkali/alkaline earth metals and large VHN values. The rock
interior also exhibits low contents of FeO Fe2O3 and small
L, a, and b values.
For the band described in this study, oxidized and dissolved
zones in the upper zone (the outer brown band) underwent not
only oxidation but also dissolution, and the formation of the
bands needs a two-way water movement. In the rocks comprising the fluvial-terrace deposits, weathering rinds tend to
occur parallel to the rock surface with nearly constant thickness.
Thus, the mechanism of weathering-rind formation is related to
smaller-scale water movement in and around the rocks, which
is different from the large-scale movement. Gravel in terrace
deposits is attacked by underground water percolating through
sediments. When rainfall is abundant, water is supplied to the
gravel. In this case, water movement is directed from the outside
105
Table 2 Physical properties of the interior and thickness of brown band (LI) and total weathering rind (LI II)
Samples
Total pore volume

Vt, mm3 kg 1
Bulk density
rbulk, 103 kg m 3
Porosity
n, %
Brown layer
LI, mm
Total weathering rind

LI II, mm
0-ka Rocks
0a
0b
0c
0d
0.018
0.009
0.011
0.016
2.40
2.44
2.55
2.55
4.30
2.19
2.79
3.98
20-ka Rocks
20 a
20 b
20 c
20 d
20 e
0.015
0.014
0.014
0.020
0.028
2.36
2.33
2.59
2.40
2.45
3.62
3.25
3.57
4.76
6.73
o0.10
o0.10
o0.10
o0.10
o0.10
2.88
3.03
1.88
2.00
3.00
130-ka Rocks
130 a
130 b
130 c
130 d
0.034
0.008
0.026
0.017
2.31
2.51
2.34
2.49
9.95
1.96
6.15
4.14
3.00
3.30
2.00
2.00
420.00
5.60
412.00
5.04
290-ka Rocks
290 a
290 b
290 c
290 d
290 e
290 f
290 g
0.008
0.003
0.010
0.018
0.036
0.027
0.015
2.61
2.68
2.61
2.54
2.42
2.40
2.55
2.20
0.85
2.48
4.49
8.59
6.54
3.91
2.42
1.87
3.00
3.25
3.03
3.10
3.16
2.79
4.26
6.50
410.20
421.00
415.00
8.42
660-ka Rocks
660 a
0.009
2.43
2.20
4.00
6.00
Vickers microhardness
number (gf m2)
Brown band
Rock interior
White band
600
400
200
0
5
10
Depth from rock surface (mm)
15
Figure 6 Vickers microhardness number versus depth from the

rock surface. Modified from Oguchi, C.T., 2001. Formation of
weathering rinds on andesite. Earth Surface Processes and
to the inside of the rock. In contrast, during periods of low

rainfall, water supply ceases, and water movement occurs from
inside to outside of a rock.
These different directions of water movement are responsible for the development of the two weathering bands. Based
on the analyses of the rock properties, a model to explain the
mechanism of the development of the two bands is proposed
(Figure 8). The growth of the white band is demonstrated as:
(1) when the water from rainfall percolates into the ground
and is supplied to the matrix around a cobble, water
movement toward the inside takes place; (2) alkali/alkaline

earth metals are dissolved from the rock into percolating
water;
(3) a bleached zone with depleted elements is formed near
the rock surface; and (4) this zone gradually thickens as the
dissolution front moves into the interior of the rock. Thus,
only the dissolution process formed this band. The growth of
the brown band is explained as: (1) when the groundwater
supply is reduced, the direction of water movement within the
rock is reversed, with movement of water occurring toward the
outside surface of the rock; (2) although most dissolved
metals move out of the rock with the water, Fe3 is precipitated near the rock surface because Fe2 is oxidized to
form Fe3 with low solubility in natural water (Ichikuni,
1972, p. 75); and (3) the outermost zone gradually turns
brown in color due to accumulation of ferric oxide/hydroxide
minerals and compounds such as Fe(OH)3, FeO(OH) (e.g.,
goethite and lepidochrocite), and Fe2O3 (e.g., hematite and
maghemite). Both dissolution and oxidation occur in the
brown band.
4.6.5
A Porosity Concerned Model of Weathering

Rind Development
The original definition of the weathering rind is an outer crust

or layer on a rock fragment formed by weathering (Gary et al.,
1972) or a hard and thin weathered layer with Si, Fe, or Mn
106
Outer brown
band
Inner white band
Rock interior
CaO
100
(%)
FeO + Fe2O3
0
100
(%)
L*
100
(%)
(a*2 + b*2)1/2
0
100
(%)
VHN
100
(%)
0
0
10
15
Depth from rock surface (mm)

Figure 7 Changes in rock properties with depth from the rock surface. Modified from Oguchi, C.T., 2001. Formation of weathering rinds on
andesite. Earth Surface Processes and Landforms 26, 847858.
enrichment at surfaces of rock blocks (Maruyama, 1981). In

recent years, the definition of the term weathering rind includes the formative process of oxidation with color alteration: oxidation phenomena which stain the parent rock
redyellow when exposed to air or near-surface groundwater for some time (Anderson and Anderson, 1981; Goudie
et al., 1985). Caine (1983) also called the weathering rinds
oxidationhydration rinds. However, the term weathering
rind has been used more frequently to represent a colored
weathered zone near the rock surface identified not by
chemical analyses but by eye observations (e.g., Chinn, 1981;
Watanabe, 1990; and Aoki, 1994). As noted before, the identification of such colored weathering rinds is generally difficult
because some rocks have more than one colored zone near the
surface. Andesite rocks investigated in the present study also
have both brown and white bands. Based on the measurements of rock properties, weathering rinds with two zones,
oxidation zone and dissolution zone, can be redefined.
Both the thicknesses of these zones are used to construct a
growth model of weathering rinds for the 0-ka, 20-ka, 320-ka,
450-ka, and 830-ka rocks. The relationships between these thicknesses and weathering period are shown in Figure 9. In general,
the thickness of a weathered zone has often been approximated
using a diffusion equation (e.g., Friedman and Long, 1976).
L D t1=2 D1=2 t 1=2
2
where L (mm) is the thickness of the weathered zone, t (year)

is the weathering period, and D (mm2/yr) is a diffusion coefficient. Using eqn [1], the diffusion coefficients for the
brown bands (DI) and the total weathering rinds (DI II) can
be calculated from the thickness data and the weathering time.
According to Drever (1997, p. 357), an effective diffusion
coefficient (Deff ) is related to an apparent diffusion coefficient
and rock porosity (n):
Deff D=n
D nDeff
3
In this study, the relationships between porosity (n) and

the diffusion coefficients DI and DI II cannot be expressed as
H2O
H2O
Fe2+
Brown band
White band
Ca2+
H2O
H2O
H2O
H2O
Mg2+
Ca2+
Rock surface
H2O
Fe3+
Fe2+
H2O
Na+
Fe3+ precipitation
Fe(OH)3, FeO(OH), Fe2O3
Na+
Fe2+
H2O
H2O
Ca2+
Mg2+
H2O
H2O
Rock surface
1. Alkali/alkaline earth metals

are dissolved due to the
effects of predominantly inward
moisture movement.
107
Rock surface
3. Precipitation of oxidized Fe3+
in the outermost layers gives
rise to the formation of the
brown band.
2. Mobilised elements in the

substrate are drawn out by the
predominant outward
movement of moisture creating
the bleached white band.
Dissolution zone
25
20
15
Oxidation
zone
10
5
0
0
200
400
600
800
1000
Time (ka)
Figure 9 Relationships between weathering time and thickness of
the oxidation zone and dissolution zone. Modified from Oguchi, C.T.,
2001. Formation of weathering rinds on andesite. Earth Surface
Diffusion coefficient (mm2 1000 yr1)
Oxidation and dissolution zones (mm)
Figure 8 Formation processes of white and brown bands. Reproduced from Oguchi, C.T., 2001. Formation of weathering rinds on andesite.

(dissolution zone)
y = 0.0431e0.4287x
R 2 = 0.7527
2
1
Brown band
(oxidation zone)
0
0
10
12
Porosity (n %)
the linear eqn (Figure 10). The DI values are low and almost
constant (DI 0.0283) irrespective of the n values, whereas the
DI II values increase exponentially with increasing n values.
The relationship between DI II and n is better expressed (see
Figure 11) as:
DIII 0:0431exp0:4287n
R2 0:7527
4
Combining eqns [1] and [3], the following equation is

obtained:
LIII Dt1=2 f0:0431exp0:4287n tg1=2
5
In the same way,

LI Dt1=2 0:0283 t1=2
6
Figure 10 Relationships between porosity and diffusion coefficient for

brown band (LI) and the weathering rind (LI II) (Oguchi, 2001). Open
circles (J) show total weathering rinds and solid circles (K) show
outer brown rinds. The data plots with arrows show the minimum
diffusion coefficients. Modified from Oguchi, C.T., 2001. Formation of
weathering rinds on andesite. Earth Surface Processes and Landforms
26, 847858, and Oguchi, C.T., 2004. A porosity-related diffusion
model of weathering-rind development. Catena 58, 6575.
Figure 10 shows the development of the total weathering

rind and brown band thicknesses based on the obtained diffusion equations. In the case of dense andesite (n 0), the
thickness of the total weathering rind is almost equal to that of
the brown band. Total weathering-rind thickness on porous
andesite, however, is much larger than that of the brown band.
It is revealed that not only weathering time but also rock
porosity controls the total weathering-rind thickness.
108

(dissolution zone)
Thicknesses of oxidation zone and

total weathering rind (mm)
20
n = 5.0 %
n = 4.0 %
15
n = 3.0 %
n = 2.0 %
10
n = 1.0 %
n=0%
5
Outer brown band
(oxidation zone)
0
0
200
600
400
800
1000
Time (ka)
Figure 11 Development of the thicknesses of the outer brown band (dotted curve) and weathering rind (solid curves) with different rock
porosity values (n) based on eqns [5] and [6]. Modified from Oguchi, C.T., 2004. A porosity-related diffusion model of weathering-rind
development. Catena 58, 6575.
4.6.6
Conclusions
In order to elucidate mechanisms of weathering-rind formation, the relationships between several rock properties were
examined specifically for weathering subbands, the brown
band and white band compared to the rock interior. The interior of the rock is composed of fresh rock-forming minerals
and is characterized by high contents of alkali/alkaline earth
metals, low L, a, and b color indicator values and high
Vickers microhardness numbers. The white band is depleted
in Ca, especially, relative to the rock interior of the rock. The
brown band is very brittle and contains clay minerals, which
are absent from the rock interior and white band. The brown
band also has much higher L, a, and b color indicator
values than the inner two parts, a higher content of
FeO Fe2O3, but a lower alkali/alkaline earth metal content.
A conceptual model of weathering-rind formation suggests
that the inner white band is produced by dissolution of highsolubility alkali/alkaline earth metals related to the inward and
subsequent outward movement of water. The brown band
is probably formed by: (1) dissolution of alkali/alkaline earth
metals and (2) oxidation of Fe2 to Fe3 , which has low
solubility and therefore forms mineral precipitates close to the
rock surface. This provides valuable information on the linkages
between weathering processes and rock properties. Based on
property analysis, a porosity-concerned growth model for
weathering rinds was also proposed. The model consists of two
diffusion equations based on the two types of weathering-rinds
identified. One is a normal diffusion equation showing the
development of strongly weathered brown bands due to both
oxidation and leaching. The other is an equation with the diffusion coefficient exponentially related to porosity of the host
rock showing the growth of total weathering rinds mainly due
to leaching. A significant implication of this result is that studies
which use weathering-rind thickness as a dating tool under a
humid temperate environment should also to take rindcharacterization and original rock porosity into consideration.
Weathering-rind developments are supposed to be different

from rock types. The important properties to consider them are
contents of iron as well as alkali and alkaline earth metals and
porosity of rocks. If the rock is a mafic type, the rinds are clearly
distinguished from the original rocks as its color turns to brown
due to oxidation of iron. Dissolution of calcium will occur as
well, because these rocks are rich in alkaline earth metals. If the
rock is a felsic type, the oxidation zone is expected to be unclear,
unless iron stains are input from outside of the rock or small
amounts of ferric minerals are decomposed within the rock.
Porosity will control thickness of the dissolution zones in most
cases. Degree of oxidation or dissolution should be determined
by temperature and precipitation as well. Therefore, it is also
important to consider rock properties and climate conditions
for weathering-rind studies.
References
Alexander, E.B., 1985. Rates of soil formation from bedrock or consolidated
sediments. Physical Geography 6, 2342.
Amit, R., Gerson, R., Yaalon, D.H., 1993. Stages and rate of the gravel shattering
process by salts in desert Reg soil. Geoderma 57, 295324.
Anderson, L.W., Anderson, D.S., 1981. Weathering rinds on quartzarenite clasts as
a relative-age indicator and the glacial chronology of Mount Timpanogos,
Wasatch range, Utah. Arctic and Alpine Research 13, 2531.
Aoki, T., 1994. Chronological Study of glacial advances based on the weathering-rind
thickness of morainic gravel in the northern part of the Central Japan Alps.
Geographical Review of Japan 64A, 601618 (in Japanese with English abstract).
Birkeland, P.W., 1973. Use of relative age-dating methods in a stratigraphic study of
rock glacier deposits, Mt. Sopris, Colorado. Arctic Alpine Res 5, 401416.
Burke, R.M., Birkeland, P.W., 1979. Reevaluation of multiparameter relative dating
techniques and their application to the glacial sequence along the eastern
escarpment of the Sierra Nevada, California. Quaternary Research 11, 2151.
Caine, N., 1983. The Mountains of Northern Tasmania. A. A. Balkema, Rotterdam,
200 pp.
Cernohouz, J., Solc, 1966. Use of sandstone wanes and weathered basaltic crust in
absolute chronology. Nature 212, 806807.
Chesworth, W., Dejou, J., Larroque, P., 1981. The weathering of basalt and relative
mobility of the major elements at Belbex France. Geochemi. Chosmohimi. Acta
45, 12351243.
Chinn, T.J.H., 1981. Use of rock weatheringrind thickness for Holocene absolute
age-dating in New Zealand. Arctic and Alpine Research 13, 3345.
Colman, S.M., 1981. Rock-weathering rates as functions of time. Quaternary
Colman, S.M., 1982a. Chemical weathering of basalts and andesites: evidence from
weathering rinds. U.S. Geological Survey Professional Paper. 1246.
Colman, S.M., 1982b. Clay mineralogy of weathering rinds and possible implications
concerning the sources of clay minerals in soils. Geology 10, 370375.
Colman, S.M., Pierce, K.L., 1981. Weathering rinds on andesitic and basaltic stones
as a Quaternary age indicator, western United States. U.S. Geological Survey
Professional Paper 1210, 156.
Crabtree, R.W., Trudgill, S.T., 1985. Chemical denudation on a magnesian limestone
hillslope, field evidence and implications for modeling. Earth Surface Processes
Craig, D.C., Loughnan, F.C., 1969. Chemical and mineralogical transformation
accompanying the weathering of basic volcanic rocks from New South Wales.
Australian Jour. Soil Res. 2, 218234.
Crook, Jr. R., Gillespie, A.R., 1986. Weathering rates in granitic boulders measured
by P-wave speeds. In: Colman, S.M., Dethier, D.P. (Eds.), Rates of Chemical
Weathering of Rocks and Minerals. Academic Press, Orlando, 603 pp.
Dixon, J.C., Thorn, C.E., Darmody, R.G., Cambell., S.W., 2002. Weathering rinds
Geological Society of American Bulletin 114, 226238.
Drever, J.I., 1997. The Geochemistry of Natural Waters: Surface and Groundwater
Environments-3rd ed. Prentice-Hall, Upper Saddle River, NJ, 436 pp.
Eggleton, R.A., Foudoulis, C., Varkevisser, D., 1987. Weathering of basalt: changes
in chemistry and mineralogy. Clay and Clay Miner 35, 161169.
Etienne, S., 2002. The role of biological weathering in perigratial areas: a study of
weathering rinds in South Iceland. Geomorphology 47, 7586.
Friedman, I., Long, W., 1976. Hydration rate of obsidian. Science 191, 347352.
Friedman, I., Smith, R.L., 1960. A new dating method using obsidian: part I, the
development of the method. American Artiq. 25, 476493.
Garrels, R.M., Mackenzie, F.T., 1967. Origin of the chemical compositions of some
springs and lakes. In: Gould, R.F. (Ed.), Equilibrium Concepts in Natural Water
Systems. Advances in Chemistry Series. American Chemical Society, Washing,
DC, vol. 67, pp. 222242.
Gary, M., McAfee, Jr. R., Wolf, D. (Eds.), 1972. Glossary of Geology. American
Geological Institute, Washington DC 805 pp.
Gordon, S.J., Dorn, R.I., 1983. In situ weathering rind erosion. Geomorphology 67,
97113.
Goudie, A., Atkinson, B.W., Gregory, K.J., Simmons, I.G., Stoddart, D.R.,
Sugden, D. (Eds.), 1985. The Encyclopedic Dictionary of Physical Geography.
Blackwell, Oxford, 611 pp.
Haantjens, H.A., Bleeker, P., 1970. Tropical weathering in the Territory of Pupua and
New Guinea. Australian Journal of Soil Research 8, 157177.
Hay, R.L., 1960. Rates of clay formation and mineral alteration in a 4000-year-old
volcanic ash soil on St. Vincent, B. W. I. American Journal of Sciences 258,
354368.
High, C., Hanna, F.K., 1970. A method for the direct measurement of erosion on rock
surfaces. British Geomorphological Research Group, Technical Bulletin 5, 25.
Hirose, T., Hatta, T., Matsukura, Y., 1995. Dissolution rates of several plutonic rocks
and limestone in closed- and batch-system experiments. Transactions of the
Japanese Geomorphical Union 16, 4351 (in Japanese with English abstract).
Hirose, T., Onda, Y., Matsukura, Y., 1994. Runoff and solute characteristics in four
small chatchments with different bedrocks in the Abukuma Mountains, Japan.
Transactions of the Japanese Geomorphical Union 15A, 3148.
Hunt, R.W.G., 1980. Colour term, symbols, and their usage. In: Grum, F., Bartleson,
C.J. (Eds.), Optical Radiation Measurement. Academic Press, New York 1131.
Ichikuni, M., 1972. Inorganic Geochemistry. Baihuukan, Tokyo, 148 pp. (in
Japanese).
Japan Meteorological Agency, 1991. Climatic Table of Japan, Volume 1. Monthly
Normals (19611990). Japan Meteorological Agency, Tokyo (in Japanese).
Katsui, Y., Kondo, Y., 1965. Dating of stone implements by using hydration layer of
obsidian. Japanese Journal of Geology and Geography 36, 4560.
Kimiya, K., 1975a. Tensile strength as a physical scale of weathering in granitic
rocks. Journal of Geology of Japan 81, 319364 (in Japanese with English
abstract).
Kimiya, K., 1975b. Rates of weathering of gravel of granitic rocks in Mikawa and
Tomikusa areas. Journal of Geology of Japan 81, 683696 (in Japanese with
English abstract).
Knuepfer, P.L.K., 1988. Estimating ages of late Quaternary stream terraces from
analysis of weathering rinds and soils. Geological Society of America Bulletin
100, 12241236.
109
Koike, K., 1961. Geomorphological developments of the drainage basin of the Naka
River. Geographical Review of Japan 34A, 498513 (in Japanese with English
abstract).
Koizumi, T., Aoyagi, S., 1993. Debris supply periods estimated from weatheringrind thickness in rubble on the west-facing slope of Mt. Yakushidake, the
Northern Japanese Alps. Geographical Review of Japan 66A, 269286 (in
Japanese with English abstract).
Koizumi, T., Seki, H., 1992. Use of rock weathering-rind thickness to redate fossil
periglacial slope in Mt. Kisokomagatake, the Central Japanese Alps. Quaternary
Journal of Geography 44, 245251 (in Japanese with English abstract).
Kuchitsu, N., 1991. Weathering of stone implements excavated from Sunadadai Site,
Kanagawa Prefecture Taphonomy for archeological properties. Quaternary
Research (Japan) 30, 4347 (in Japanese).
Leneuf, N., Aubert, G. 1960. Essai devaluation de la vitesse de ferrallitisation. Proc.
7th Int. Conf. Soil Sci., pp. 225228.
Maruyama, A., 1981. Weathering rinds. In: Machida et al., (Eds.), The Encyclopedia
of Geomorphology. Ninomiya-shoten, Tokyo, 767 pp (in Japanese).
Matsukura, Y., Kimata, M., Yokoyama, S., 1994a. Formation of weathering rinds on
andesite blocks under the influence of volcanic gasses around the active crater
of Aso Volcano, Japan. In: Robinson, D.A., Williams, R.B.G. (Eds.), Rock
Weathering and Landform Evolution, John Wiley & Sons, Chichester,
pp 8998.
Matsukura, Y., Maekado, A., Hatta, T., Yatsu, E., 1983. Vertical changes in
mineralogical, physical chemical and mechanical properties due to deep
weathering of Inada granitic rocks. Transactions of the Japanese Geomorphical
Union 4, 6580 (in Japanese with English abstract).
Matsukura, Y., Oguchi, C.T., Hatta, T., 1994b. Preliminary study on Vickers
microhardness of weathering rinds. Annual Report of the Institute of Geoscience,
the University of Tsukuba. No. 20, pp. 1517.
Menard, H.W., 1974. Geology, Resources and Society. W. H. Freeman, San
Francisco, 621 pp.
Nahon, D., Lappartient, J.R., 1977. Time factor and geochemistry in iron crust
genesis. Catena 4, 249254.
Nakashima, S., Miyagi, I., Nakata, E., et al., 1992 Colour measurement of some
natural and synthetic minerals. Report of Research Institute of Natural
Resources, Mining College, Akita University 57, pp. 5776.
Oguchi, C.T., 2001. Formation of weathering rinds on andesite. Earth Surface
Oguchi, C.T., 2004. A porosity-related diffusion model of weathering-rind
development. Catena 58, 6575.
Oguchi, C.T., Hatta, T., Matsukura, Y., 1999. Weathering rates over 40,000 years
based on changes in rock properties of porous rhyolite. Physical Chemistry of
the Earth (A) 24, 861870.
Oguchi, C.T., Matsukura, Y., 1999a. Microstructural influence on strength reduction
of porous rhyolite during weathering. Zeitschrift fur Geomorphologie. N.F.,
Supplement Band 119, 91103.
Oguchi, C.T., Matsukura, Y., 1999b. Effect of porosity on the increase in
weathering-rind thicknesses of andesite gravel. Engineering Geology 55,
7789.
Porter, S.C., 1975. Weathering rinds as a relative-age criterion: application to
subdivision of glacial deposits in the Cascade Range. Geology 3, 101104.
Ruxton, B.F., 1968. Measures of the degree of chemical weathering of rocks.
Journal of Geology 76, 518527.
Sak, P.B., Fisher, D.M., Gardner, T.W., Murphy, K., Brantley, S., 2004. Rates of
weathering rind formation on Costa Rican basalt. Geochimica et. Cosmochimica
Acta 68, 14531472.
Saito, T., Abe, M., Kunori, S., 1974. Study on weathering of volcanic rocks (I).
Butsuri-Tanko (Geophysical Exploration) 27, 315.
Shiraiwa, T., Watanabe, T., 1991. Late Quaternary glacial fluctuations in the
Langtang Valley, Nepal Himalaya, reconstructed by relative dating methods.
Arctic and Alpine Research 23, 404416.
Singer, A., 1984. Clay formation in saprolites of igneous rocks under semiarid to
arid conditions, Negev, Souther Israel. Soil Science 137, 332340.
Smith, R.L., Sandland, G.E., 1922. An accurate method of determining the hardness
of metals, with particular reference to those of a high degree of hardness.
Proceedings of the Institution of Mechanical Engineers 1, 623641.
Suzuki, T., Fujiwara, O., Danhara, T., 1998. Fission track ages of eleven
Quaternary tephras in north Kanto and south Tohoku regions, central Japan.
Quaterly Research (Daiyonki-Kenkyu) 37, 95106 (in Japanese with English
abstract).
Suzuki, T., Hachinohe, S., 1995. Weathering rates of bedrock forming marine
terraces in Boso peninsula, Japan. Trans Actions of the Japanese Geomorphical
Union 16, 93113.
110
Takahashi, M., Hayakawa, Y., 1995. Plant remains from the Nakanojo lake deposit in
Gunma Prefecture, central Japan. Science Reports of the Faculty of Education,
Gunma UniversityScience Reports of the Faculty of Education, vol. 43, 7186
(in Japanese).
Trendall, A.F., 1962. The formation of apparent peneplains by a process of
combined laterization and surface work. Zeitschrift fur Geomorphologie NF.
(Suppl. Bd.) 6, 183197.
Trudgill, S.T., 1975. Measurement of erosional weight-loss of rock tablets. British
Geomorphological Research Group, Technical Bulletin 17, 1319.
Trudgill, S.T., 1976. The subaerial and subsoil erosion of limestone on Aldabra
Atoll, Indian Ocean. Zeitschrift fur Geomorphologie NF (Suppl. Bd) 26,
201210.
Trudgill, S.T., High, C.J., Hanna, F.K., 1981. Improvements to the micro-erosion
meter. British Geomorphological Research Group, Technical Bulletin 29, 317.
Wakatsuki, T., Rasyidin, A., 1992. Rates of weathering and soil formation. Geoderma
52, 251263.
Waragai, T., 1993. Weathering process of andesite lava in Mt. Hachimen,
Yabakei, Oita prefecture. Proceeding of the Institute of Natural Sciences.
Nihon University, Tokyo, Japan, vol. 28, 1524 (in Japanese with English
abstract).
Watanabe, K., Sagehashi, N., 1960. Geologic history of the Nasu Plain, Tochigi
Prefecture (Hydrology of Nasuno-ga-hara No.3). Journal of Geology of Japan
66, 147156 (in Japanese with English abstract).
Watanabe, T., 1990. Relative dating methods mainly applied to glacial and
periglacial deposits. The Quaternary Research (Japan) 29, 4977 (in Japanese
with English abstract).
Waylen, M.J., 1979. Chemical weathering in a drainage basin underlain by Old Red
Sandstone. Earth Surface Processes 4, 167178.
Whitehouse, I.E., McSaveney, M.J., Knuepfer, P.L.K., Chinn, T.J.H., 1986. Growth of
weathering rinds on Torlesse sandstone, southern Alps, New Zealand.
In: Colman, S.M., Dethier, D.P. (Eds.), Rates of Chemical Weathering of Rocks
and Minerals. Academic Press, Orlando, 603 pp.
Yoshioka, R., 1975. Estimation of amounts of weathered products through chemical
composition of waters in the Kamenose landslide area. Bulletin of Disaster
Prevention Research InstItute, Kyoto University 25, 115.
Biographical Sketch
Chiaki T. Oguchi was born in Yokohama, Japan. She obtained her Bachelor Degree in Physical Geography in 1991
from Meiji University, Tokyo, Japan. She received her Master Degree in Science from the Institute of Geoscience,
University of Tsukuba, Japan in 1993. She was awarded Doctorate Degree in Science from the Institute of
Geoscience, University of Tsukuba, Japan in 1998. Her research title was Rates and Mechanisms of Development
of Weathering Rind on Andesite in the Dated Fluvial Terraces under the advice of Prof. Dr. Yukinori Matsukura.
She was appointed as Assistant Professor at the Institute of Tsukuba, University of Tsukuba, Japan in 1998. She
joined as a Research Fellow at Japan Science and Technology Corporation-Japan Society for the Promotion of
Science, Japan International Research Center for Agricultural Sciences in 2001. She joined as Associate Professor,
Geosphere Research Institute, Saitama University, Japan in 2004.
4.7 Tafoni and Other Rock Basins

TR Paradise, University of Arkansas, Fayetteville, AR, USA
4.7.1
4.7.1.1
4.7.1.2
4.7.1.3
4.7.2
4.7.2.1
4.7.3
4.7.3.1
4.7.4
4.7.5
4.7.5.1
4.7.5.2
4.7.5.3
4.7.5.4
4.7.5.5
4.7.6
References
Introduction
Tafoni
Gnamma
Climatic and Geographic Influences
Morphological Classification and Rate of Development
Tafoni
Stages of Tafone Development
Gnammas
Stages of Gnamma Progression
Processes of Development
Lithologic Influences
Environmental Influences and Salinity
Biotic Influences
Climate and Insolation
Feedback Cycles
Summary
Glossary
Alveolar weathering (singular alveolae,
alveolus) Describes tafoni occurring typically on vertical
surfaces and rarely larger than 35 cm in individual cavity
diameter. Alveolus come from the term little cavity
in Latin.
Feedback cycle Commonly used to describe tafone and
gnamma incipience and development, when the product
from a process impedes its own process as negative
feedback, or increases its own process as positive feedback.
Gnamma From an Aboriginal word that derives from the
Western Australia, gnammas are stone basins, cavities, or
rock-holes generally found in on nearly horizontal surfaces
(0151), commonly in sandstone or granite. These basins
are formed by weathering and are commonly narrow at the
opening and wider at the bottom, water generally collects
seasonally in these stone basins. Subcategories of gnammas
can include pits, pans, bowls, canoe, armchair, flask-shaped,
and paternoster gnammas. Also called tinajas, tinajitas,
cuencas (Spanish), Opferkessel, Baumverfallspingen,
Steingrube (German), kociolki (Polish), conche rocciose
(Italian), oric- angas (Portuguese), and caldeiraos, poc- os
(Brazilian Portuguese).
112
112
113
114
116
116
117
118
119
120
120
122
122
123
124
125
125
Honeycomb weathering A subcategory of tafoni that

represent small (o2 cm), regular and commonly patterned
cavities found in granular rock (i.e., granite, sandstone).
Stonelace A subcategory of tafoni that is commonly
synonymous with honeycomb weathering.
Tafoni (singular tafone) Small (o1 cm) to large (41 m)
cave-like features generally occurring in granular rock (i.e.,
granite, sandstone) with smooth concave cavities, and often
round rims and openings. Subcategories of tafoni can
include honeycomb, stonelace, alveolar (o2 cm), sidewall,
basal, nested, and relic. Also called nido dape roccioso
(Italian).
Weathering Is the breakdown of rocks, minerals, and soils
through contact with the climatic influences (i.e.,
insolation), technological influences (i.e., pollution) and/or
humans (i.e., touch). Weathering occurs in situ and should
not be confused with erosion, whereby the influences such
as water, ice, wind, and gravity degrade through/by
movement. Conventional classifications include: (1)
mechanical or physical weathering (breakdown through
direct contact with atmospheric conditions such as heat,
ice); and (2) chemical weathering (effect of biologicallyproduced and/or atmospheric chemicals).
Paradise, T.R., 2013. Tafoni and other rock basins. In: Shroder, J. (Editor in
Chief), Pope, G.A., (Ed.), Treatise on Geomorphology. Academic Press, San
http://dx.doi.org/10.1016/B978-0-12-374739-6.00068-3
111
112
Tafoni and Other Rock Basins
Abstract
Tafoni and gnamma are cavernous weathering phenomena that have been extensively surveyed, discussed, and studied.
From delicate, polygonal cells (24 mm) to huge circular pits (30 m ), they develop in a variety of rock substrates,
commonly in sandstone (tafoni) and granite (gnamma). Early research on their initiation and development was divisive,
but current studies indicate polygenetic, differential weathering from intrinsic factors including hydrolysis, hydration of
lithologic constituents, to extrinsic influences like moisture availability, insolation, and salinity. Increasingly, it is believed
that complex, feedback cycles are responsible for their sigmoidal rates of development.
4.7.1
Introduction
One of the most puzzling traits of honeycomb or panshaped weathering features is that they can develop in
seemingly homogenous rocks, in one portion or face, and not
another. So why does an intricate lacework pattern form on one
cliff face or boulder, and not on the one adjacent? Are the
processes responsible for their developed unknown, or are we
slowly beginning to understand their complex yet beautiful
nature? Studies have investigated the primary causes and influences for more than 100 years, but we are only now beginning to comprehend the integrated, coordinated, and feedback
systems involved with their initiation, development, and progression. Contemporary research emphasizes the role of
microclimatic conditions, salinity, and environmental factors in
tandem with feedback mechanisms on their development
(Hejl, 2005), but controversy still exists as to the polygenetic
nature of these generally unique features called tafoni, gnammas, armchair hollows, alveoli, solution pits, or honeycomb
weathering.
Tafoni and gnammas are lace-like, honeycomb, bowl, or
pan-shaped cavities occurring in a variety of rock types
and locations that show a commonly unique assemblage and
morphology. These oddly shaped, generally intricately created,
naturally occurring weathering and erosional features occur in a
wide range of lithologies including sandstone, limestone,
granite, greywacke, rhyolite, quartzite, conglomerate, and tuff.
They have been documented across diverse environments in
deserts and along coasts, and from assorted climatic zones that
include the mountainous, coastal, and arid regions of all continents, Antarctica, and even on Mars (Rodriguez-Navarro,
1998).
These recessional and erosional features come in various
forms and with assorted names. Tafoni, or singular tafone,
commonly occur on vertical faces, whereas gnammas are
weathering and erosional depressions that develop on relatively horizontal solid rock substrates, with a continuum of
weathering feature morphology from steep to flat settings.
Tafoni and gnammas are natural rock cavities that typically
take on an ellipsoidal, circular, oblate, or polygonal outer
edges, generally bowl, honeycomb, or pan-like in form. They
have been recorded as small as tiny pits, cells, or voids, to fistsize, bowl-size, or as large as an automobile or small house.
Blackwelder (1929) described the largest as shelter for horse
and rider. They form from as delicate lacey features measured
in millimeters, to the renown tafone cavities of the Remarkable Rocks of Kangaroo Island, Flinders Chase National Park
in Australia where these famous granite cavities measure up to
20 m across (Figure 1).
4.7.1.1
Tafoni
The etymology of the word tafoni is obscure; however, its use

across the Mediterranean implies that it may derive from the
Italian dialectic verb tafonare, to perforate. It has become a
colloquial term for windows across the western Mediterranean Basin, especially on Corsica (Boxerman, 2006), Sicily,
and Sardinia. However, the term tafone is commonly synonymous in Italy with nido dape roccioso (rocky honeycomb), whereas in English it can be synonymous with
a number of terms such as honeycomb, stone lace, alveoli,
fretting, and stone lattice.
Although the term tafoni has not been specifically assigned to features occurring only on vertical faces, similarly
appearing recessional forms on horizontal surfaces are most
commonly called gnamma, an Australian aboriginal word
that derives from the Western Desert languages (i.e., Nyungar)
which were/are spoken over a huge area of Western Australia.
The Western Desert people use gnamma to refer to rock-holes
occurring in sandstone or granite, sometimes containing
water. Described as basins formed by weathering that are
commonly narrow at the opening and wider at the bottom,
water generally collects seasonally in these stone basins
(Moore, 1842).
Honeycomb features were described by Strabo 2000 years
ago (22AD), and then commonly in the reports of nineteenth
century western explorers including Stephens (1837) and
Burton (1879). However, the term tafoni was first specifically
used by De Prado (1864) to describe the unusual weathered
Figure 1 Photograph of extraordinary tafoni development on

Kangaroo Island, Southern Australia. These Remarkable Rocks are
located in Flinders Chase National Park and have developed atop
granite corestones resting atop a small, partially exposed granite
dome. These boulders with tafoni have become a popular tourist
destination on the western side of the Island. The largest tafone in
this image measures 3 4 m, at an aspect of B310 1N.
Figure 2 Sketch by De Prado (1864) of tafone in granite near

Madrid. This is the first recorded reference to and illustration of
tafone/tafoni in western research.
forms occurring in Spanish granites (Figure 2), and later by

Reusch (1883) and Penck (1884) regarding Corsicas unusual
weathering features. Since the nineteenth century, more than
100 articles and reports alone have been published in the
western literature on these often ubiquitous, yet interesting
features (Hejl, 2005), and they continue to foment interest in
curious tourists and researchers alike.
In the western literature, the earliest reference to honeycomb weathering and their possible development is commonly attributed to Darwin (1839) with his discussion of
weathering features during his voyages on the HMS Beagle.
While sailing the coast of Western Australia, he described
features not unlike tafoni, which appeared to be carbonate
casts of tree roots.
After the global instability of the First World War (WWI),
exploration across the planets arid regions expanded,
prompting elaborate travel journals and new research directions and studies. Tafoni research increased and so did the
speculation as to their formation. Sir Charles Cottons
notable work on the geomorphology of New Zealand is one
of the first to document tafone development, in addition to
postulations on their initiation and morphological evolution.
He specifically attributed the formation of tafone and alveoli
to lithologic variation. Cotton (1922) cited the example of
limonite-rich seams that weathered at a slower rate than the
surrounding sandstone, which created polygonal cavities that
developed into tafoni.
In his work on the Giza Plateau of Egypt, Hume (1935)
speculated that sandstone deterioration and the formation of
tafone were products of salt mobilization from external
sources, and within the rock material itself. This was an innovative concept for the time and guides research to this day.
Bartrum (1936) followed investigating the role of carbonate
dissolution in less arid landscapes like New Zealand. Popoff
and Kvelberg (1937) would then hypothesize about the importance of microclimatic effects and moisture mobility on
the famous tafone of Corsica.
113
However, many researchers consider the early works

and related controversy of Bryan and Blackwelder as the fundamental watershed in tafoni research. Bryans examination
of the Chaco Canyon sandstones (1928) led to his hypotheses
on the crucial nature of moisture and rock permeability as
the primary influence on the development of tafone. However,
Blackwelder (1929) explained that the role of wind cannot
be discounted in their development, arguing that Bryans
work was site-specific and too narrow in his explanation
of honeycomb weathering. Although Bryan stressed the occurrence of these cavities near ephemeral waterfalls, weeps
and sapping indicated the role of moisture in dissolving
mineral matter from the sandstone cement, Blackwelder was
quick to explain that moisture and mineral alteration was
influential, whereas wind was the essential element needed to
remove the by-products of weathering. He emphasized the
cooperative effect of wind not as an abrasive agent, but as an
erosional force. This debate still haunts tafoni research today
is their development due to primary influences like wind or
lithology, or is it a complex and polygenetic interaction of
intrinsic and extrinsic controls on tafoni initiation and
development?
4.7.1.2
Gnamma
In the comprehensive Descriptive Vocabulary of Aboriginal

Words (Moore, 1842), the earliest reference is to gnamar, or a
hole or pool of water in a rock. Later Austin (1856) described
how his Aborigine guide led him to a water-hole in a parched
area of Western Australia. He stated that the guide depended
upon the precarious supply of rainwater accumulated in the
hollows of the rocks. Calvert (1897) and Carnegie (1898)
later coined the term namma-hole while touring and reporting on their journeys across Australias outback.
Gnammas are rock basins or cavities that generaly occur on
nearly horizontal surfaces (0151) in exposed granite landscapes where they can be common and well developed.
Gnammas can occur atop other substrates like sandstone or
greywacke, where the rock materials are relatively impermeable like granite. They can measure mere centimeters in width
and depth, to as large as 5 18 m and to a depth of 2 m.
Twidale and Corbin (1963) cited examples, however, of exceptionally deep gnammas at 6 m, whereas Netoff and Shroba
(2001) later discovered huge basins (tanks) in the sandstones
of the American Southwest that measured 16 m in depth and
up to 70 m in diameter.
Gnamma morphology is typically a round or oval depression atop a planar surface, generally wider than deeper.
They also follow a rough trend from small and shallow
to large and deep: circular and bowl-shaped at first termed
pits, to elliptical and flat-bottomed (generally sedimentfilled) as larger and older, called pans. As gnammas develop
on, or enlarge toward steeper slopes (415201), they enlarge
in dimension and often evolve into armchair hollows, where
the uphill depth is greatest, and the downslope depths decrease to essentially no rim, edge, or lip (Paradise and Yin,
1993).
Gnammas also act as reservoirs for water from precipitation so when it is full, the water overflows at the edge with
114
the lowest rim. This repeated drainage erodes a channel or

spillway often characteristic of advanced morphology. In fact,
spillways that connect gnamma to gnamma are distinctive of a
constellation representing mature gnammas, and gnamma
landscapes in advanced stages.
Gnamma enlargement has been attributed to the seasonal
filling of water; however, a pit with water is not necessarily a
gnamma. Their evolution and development morphologies
indicate that there is indeed an ordered progression sequence,
pointing toward forms and development that are not arbitrary.
This has been repeatedly corroborated through gnamma
morphometric analysis (GMA) where developmental relationships between width and depth indicate a coordinated
morphological progression (Dominguez-Villar, 2006).
Gnammas have been widely described in English and other
languages and include tinajas, tinajitas or cuencas (Spanish:
small tub, bowl, or basin), Opferkessel, Baumverfallspingen,
or Steingrube (German: sacrificial cauldron, stone pit or hollow), kociolki (Polish), conche rocciose (Italian: rock basins),
oric- angas (Portuguese), caldeiraos or poc- os (Brazilian Portuguese: cauldrons, rock basins), weathering pans, weathering
pits, solution pits, cups, or armchair hollows. Unlike tafoni
which develop on permeable substrates that enable moisture
mobility (i.e., sandstone, limestone), gnammas are generally
horizontal (or close) and develop on mostly impermeable
substrates (i.e., granite). In his notable work in Yosemite
National Park, California, Matthes (1930) coined the terms
weathering pit and weather pit terms that remain both in
the scholastic studies and travel books today.
In the western literature, research on gnamma morphology
and development in granite has been dominated by Australian
studies of Twidale and others (Twidale and Corbin, 1963;
Vidal Romani and Twidale, 1998); however, in the past century other geomorphologists have continued to investigate
these unusual rock pans, pits, and bowls in granite and other
rock substrates in Australia and abroad. In the Americas, research intensified Caldenius (1932) wrote extensively about
gnamma development on Patagonian granites and Matthes
(1930) wrote about their forms, distribution, and development on the exposed granite of the Sierra Nevada Mountains.
Friese (1938) speculated as to the many influences on the
progression of gnammas in Brazil. Investigations into their
morphology and development leapt in number and depth
after WWII with a number of important studies in the 1960s
(i.e., Dahl, 1966). These studies have led to works today that
link morphology, and morphology process.
Gnammas have also been used as relative dating tools. By
exploring pit widths, depths, and/or depths, research has
utilized their morphology and distribution to date surface
exposure after ice sheet retreat, or landscape change (i.e.,
Matthes, 1930; Dahl, 1966; Ives, 1978; Landvik, 1994). These
same researchers have also noted the dangers associated with
using gnammas in this manner; however, their use has led to
interesting speculation and continued research on glacial retreat, weathering rates, and landscape denudation.
This chapter will cover tafoni and gnammas, and their
morphology, development, substrates, and rate of development, as well as the range of influences. The discussion will
include salt-induced weathering, microclimatic effects, lithologic influences, positive and negative feedback cycles, and the
role of geographic controls in the development of some of

natures most intriguing weathering phenomena. However,
over the past century, an absence of consistent and standardized nomenclature has led to some confusion as what are
tafoni, gnammas, honeycomb, or alveolar weathering features? In field studies, are gnammas being discussed, tafoni
being studied or both, in the same study, or at the same site?
In both Bryans (1928) and Blackwelders (1929) seminal
works, although the terms niches and cavernous features
were used, descriptions and images indicate that both tafone
and gnamma morphologies were being addressed. In the work
of Sancho and Benito (1990) on sandstones in central Spain,
although the term tafoni was preferred, images show a range
of morphologies and locations of tafone, cavernous weathering, armchair hollows, and gnammas, developing as a
function of slope vertical slopes exhibited tafoni, whereas on
horizontal surfaces, gnammas were prevalent. This confusion
in nomenclature and overlap in terms will continue to haunt
this fascinating area in geomorphological and stone conservation research.
4.7.1.3
Climatic and Geographic Influences
It has been estimated that tafone development is so extensive

along the Earths coasts that they account for more than 10%
of all shoreline retreat (Gill et al., 1981; see Chapter 4.13). On
a global dimension and within human timescales, weathering
and erosional processes are responsible for the slow destruction of stone structures and monuments; however, most
of these processes are often invisible within each generation of
observation. Over time, nevertheless, tafone and gnamma
development and subsequent surface recession affect landscapes differently deserts at a slower rate than coasts. These
phenomena, however, are undeniably responsible for extensive surface recession and subaerial denudation to an enormous degree.
Although tafoni and gnammas have been recorded and
studied across the planet (and Mars), distinctive relationships
have been found between extrinsic factors including climate
and environmental conditions, and intrinsic controls like
lithologic structure, constituency, texture, and integrity (fresh
vs. weathered). However, their geographic distributions vary
widely and between themselves. They have been recorded at
sea level and at above 2000 m in the Mountains of Corsica,
and Patagonia, and above 2700 m in the Sierra Nevada of
California. However diverse are the landscapes of tafone and
gnamma occurrence, there is a broad consensus that their
development is more prevalent in temperate coastal environments and more hot and/or cold deserts.
Tafoni have been recorded and investigated in temperate
regions that include the coasts of Africa, America, Asia,
Australia, and Europe. The humid regions include Hong Kong,
the U.S. Midwest, and the Mediterranean Basin, arid, warm
landscapes in the American Southwest, North Africa and
Australia, and the cold deserts of Antarctica, Finland, Iceland,
China, and Mars. Similarly, gnammas have been recorded
across the globe and at varying latitudes and climatic influences, from sea level to the summits of mountains and
bornhardts. However influential is climate, the control of
climate has been found to be minimal when compared to

the substrate type, permeability, and surface conditions (i.e.,
varnish, glaze, crust, vegetative mats) (Figures 3 and 4).
Geographically, tafoni most commonly occur in coastal
environments and arid landscapes in permeable or quasipermeable rock substrates, whereas gnammas commonly
occur in more diverse climatic zones, atop or flanking crystalline rock substrates like granite or granitoid schists, although the occurrence of gnammas has been recorded in
115
igneous (extrusive, intrusive), sedimentary (clastic, chemical),

and metamorphic substrates. In fact, some of the largest
weathering pits were documented in the Entrada sandstones
of Utah (USA), measuring 370 m in diameter to 216 m in
depth (Netoff and Shroba, 2001).
Climate effects, like wind and environmental influences
(i.e., salinity), are generally crucial in the development of
tafoni. Although lithology and substrate composition commonly determine the location, distribution, and morphology
Figure 3 Map of recorded tafone locations from western research in English, French, German, and Italian languages. These sites have been
compiled from research dating from the 1840s. Some neighboring sites have been identified with one marker only.
Figure 4 Map of recorded gnamma locations from western research in English, French, German, and Italian languages. These sites have been
compiled from research dating from the 1840s. Some neighboring sites have been identified with one marker only.
116
of gnamma as they are generally found in outcrops of hard,

impermeable rock, particularly granite, and along the flanks
and tops of monadnocks and inselbergs (Twidale and Corbin,
1963). Matthes (1930) noted the size and volume of weather
pits increased toward the peaks of granite domes in Yosemite
National Park, causing speculation as to whether this was a
function of slope, exposure, temperature change, or another
environmental factor. Paradise and Yin (1993) noted that
their distribution was more a function of aspect, heating and
cooling cycles, and moisture availability. Tafoni and gnammas
both develop best in conditions where an ideal balance between wetting and drying exists. Too much water or too little
moisture can arrest their progression, which controls their
size, shape, and distribution.
Research has indicated that it is the availability of moisture,
and it is distribution that is paramount to the development
of tafoni and gnammas (Rodriguez-Navarro et al., 1999;
Huinink et al., 2004). Field observations and laboratory simulations point to the importance of the length of episodes between wetting and drying cycles in tafoni. In coastal
environments where moisture is more evenly distributed, tafoni
extents tend to be larger, whereas the individual cavities are
smaller (Boxerman, 2006). Although in arid regions where the
moisture is restricted to shaded alcoves and northern faces,
tafoni may develop into larger cavities, however occupying a
smaller extent. In near-shore areas where the rock is continually
wetted or moist (diurnal cycles), the substrate never or rarely
dehydrates completely; few salts or none can accumulate.
Hence, salt is evenly distributed across the surface as a function
of sea spray. This causes a relatively even development and
distribution of relatively uniformly sized cavities and ribs in
coastal settings where salt is more prevalent in comparison to
arid environments.
In arid regions like the U.S. Southwest or the Levant,
periods of moisture availability occur from seasonal events
more often than from diurnal variation. These cycles of seasonal precipitation are longer in duration, with lengthier
episodes of drying between wetting and/or saturation. These
dry periods are measured in weeks and months (arid) in
contrast to hours and days (coastal). These longer dry periods
in deserts allow for greater accumulations of salts inside the
tafone cells and cavities. This enables positive feedback cycles
to proliferate in arid landscapes where the sheltered cavities
accumulate salts, deteriorate faster to create larger cavities,
increasing the cavity volume. More salts accumulate in the
larger cavities, accelerating in-cavity weathering to propagate
the cycle.
In the broadest sense, near-shore settings facilitate smaller
and more uniform cells and ribs in tafoni, whereas hot and
cold arid environments sustain the development of larger cells
and ribs within a grouping of more irregularly size cavities.
Similarly, hyper-arid landscapes do not facilitate the growth of
gnammas, since moisture (and related heating and drying
cycles) are needed in the initiation and development of bowls
and pans. However, the occurrence of gnammas in humid
and/or tropical climates indicates that their progression is
more controlled by lithology than climate. Largely, it may be
said that tafoni growth is primarily influenced by extrinsic
factors, whereas gnamma development is affected by both
intrinsic and extrinsic influences.
4.7.2
Morphological Classification and Rate of

Development
Tafoni and gnammas evolve in distinct stages, from tiny cavities and depressions to large caves and pits over decades,
centuries, and millennia. It is this relatively long period of
development within human time scales that makes it difficult
for geomorphologists, stone conservators, geologists, and
geographers to effectively explain the rate of change, and the
occurrence of similarly shaped niches, bowls, and pans on
varied rock substrates, and in diverse environmental settings.
4.7.2.1
Tafoni
Tafoni, rock niches, or cavernous weathering can take on a

number of forms and configurations that have been discussed
and coined over the last 150 years. Some tafoni have been
grouped by locality (near-shore), position (cliff base), size
(huge, remarkable), shape (geometric, lacey), or having a
unique feature (human face, eagle). Various terms have been
collected from research conducted over the last 150 years:
Honeycomb is possibly the most common synonym or
alternate term for small-cavity or cell tafoni and is often
considered a subclass of tafoni. It has been interchanged with
alveolar weathering, implying a cell-like size, arrangement
(o2 cm), and configuration of cavities. Smaller honeycomb
features may be called lacework, stonelace, fretwork, or fretting. Some studies have identified honeycomb at centimeter
scales or smaller, and tafoni at centimeter scales or larger
(Viles, 2001).
Sidewall tafoni describe their occurrence on the steep
faces of boulders, outcrops, and cliffs, whereas basal tafoni
specifically explain their development below steep faces as a
function of moisture accumulation and wicking, environmental factors (i.e., sea spray, wind), and/or biotic instigation
(i.e., lichens, insects). Nested tafone is used to identify features that have developed within the cells, cavities, and walls
of other tafone. This nesting indicates that the tafoni are redeveloping within already developed cavities and represent a
later stage in tafone development. Nested tafoni may also be
relic tafoni where the existing cells, cavities, and walls are no
longer actively enlarging, receding, or weathering. The presence of lichens, mosses, cyanolichens, or other coatings indicates either a stable surface on which they may propagate, or a
surface with arrested weathering due to attachment and
growth of biotic coatings. Lichen attachment can both indurate and weather the rock substrate.
Iconic tafoni describe tafoni that have developed into
forms or shapes that resemble something else. These tafoni
can bear a resemblance to an animals head, face, mushrooms,
structures, and writing,
Tafoni evolve in five distinctive stages that are categorized
through changes in the cavity shapes, widthdepth ratios, and
heightdepth ratios. First, small pits and cavities develop.
Then the cavities and cells widen, deepen, and enlarge. The
cells begin to develop orthogonal walls and bases. Then the
cells enlarge, the walls thin, and the bases (backwalls) of
the cavities flatten. Finally, as the cavities enlarge, some of the
walls may collapse and/or breach, causing cavities to merge
creating larger cavities. Pitting may initiate in existing cavities
117
Stone
1
4,5
Figure 5 Diagram of rock profiles illustrating the evolution of tafone morphology. Development is affected by complex intrinsic and extrinsic
influences and negative and positive feedback cycles. (1, 2) Small depressions develop and cavities broaden and deepen into (3, 4) distinctive
honeycomb or lace-like features consisting of a cell-and-wall structure. Development continues as the cell bases flatten, walls narrow, and the
cavities deepen. (4, 5) Cavities merge as walls weather and/or break. Tafone can range in size from o1 cm to 410 m. Reproduced from Goudie,
A.S., Migon, P., 1997. Weathering pits in the Spitzkoppe area, Cental Namib Desert. Zeitschrift fur Geomorphologie 41, 417444, and Boxerman,
J. Z., 2006. The evolution of tafoni on coastal sandstones in northern California. Masters Thesis, Department of Geosciences, San Francisco
State University, unpublished.
to progress again. The tafoni evolve from cavity initiation and

development, to enlargement, to coalescence, and sometimes
back to cavity initiation (Figure 5).
4.7.3
Stages of Tafone Development
1. Small depressions and pits develop due to lithologic weakness or irregularity (i.e., permeability, porosity, bedding plane,
crack, or joint). Dimensions smaller than 1520 mm (cells
o2 cm, cavities 42 cm). Rough walls and ridges develop.
2. Pits and depressions enlarge and deepen, and cells and
cavities develop. Cavities round out in form (42030 mm).
Ribs and walls become defined and regular. Weathering byproducts may accumulate and/or be removed.
3. Cavity backwalls, rib and walls and wall intersections develop a more orthogonal and geometric configuration from
rounded form. Walls begin to thin to similar dimensions.
Coatings, skins, and surface rinds may develop on the
original surface. Backwalls and cavity roofs enlarge at a
faster rate than cell and cavity floors.
4. Walls between cavities and cells begin to breach and/or

collapse as walls near 25 mm in width, due to the
enlargement of each cavity. Rock coatings may develop into
case-hardening that promotes the growth of lips, flares,
hoods, and visors. Weathering by-products accumulate.
Overall deepening and enlargement continue until walls
are fully breached.
5. Walls erode into recessed ribs within the cavity voids. Many
cavities and cells coalesce to create relatively smooth, void
surfaces. Pitting may re-initiate to renew the development
progression. Deepening rates decrease, and breakdown rate
of wall remnants and portions increase. New depression
initiation may commence within existing cavities creating a
nested arrangement and orientation for tafoni.
Some researchers have speculated that tafone walls and ribs
regularly thin to 48 mm since that is the thinnest width in
which biotic overgrowth (i.e., lichen, algae, bacteria) is able to
stabilize the wall by connecting through the permeable substrate of the walls (Boxerman, 2006). This may represent the
thinnest width or depth at which the thalli or colonies on
118
opposite sides of the cavity wall can connect into single colonies to stabilize the wall, and arrest recession.
Tafoni develop at rates controlled by climate, lithology,
environmental factors, and anthropogenic factors. Climate
controls moisture availability, wettingdrying and freeze-thaw
cycles, and the intervals of desiccation between periods of
saturation. Lithology can affect variability in permeability and
porosity that controls moisture mobility and salt accumulation. Environmental influences can include tree shading or
removal, covering rockfalls or sand ramps, or groundwater
fluctuations (either natural or induced) affecting variable
wicking, and moisture transport. Prior conservation efforts,
pollution, and human-induced damage can affect rates of
development by modifying permeability or rock integrity.
However, climate has a significant effect on rates of tafone
development. It has been found that weathering increases
when evaporation rates are slow and the period of desiccation
is lengthy. In arid settings, since the drying episodes are longer
between periods of wetting and saturation, cavity and cell
deepening may be slower overall, although dry, hot climates
foment deterioration at rates up to two to four orders of
magnitude faster than in moister environments. This may be
attributed to the reason why tafone in aridhot settings can be
so much larger than in coastal environments. However, along
coasts, the constant wetting with seawater, cells and cavities
develop at some of the fastest rates known; rate of deterioration in millimeters per year are not uncommon (Huinink
et al., 2004; Boxerman, 2006).
Rates of tafone development are rarely arithmetic in nature,
but most commonly logarithmic, sigmoidal, and/or irregular.
Under ideal tafone-producing conditions, once the rock substrate is exposed, tafoni do not develop immediately but lag in
their pit initiation. Once the depression begins to enlarge, the
cavity deepens at a faster rate than it widens. Wall thinning
rates have been recorded as slower than back wall deepening.
The rate of deepening then decreases as the rate of widening
increases, until a uniform cell has developed (Sunamura,
1996). Cell and cavity wall breach and/or collapse commonly
indicate a decrease in the rate of widening and deepening,
representing a later stage in development.
Also, not all rock substrates weather, enlarge, widen, and
deepen at the same rates: tafoni tend to develop into the most
voluminous cavities in arid, hot settings, and containing the
most cells and cavities in near-shore environment. However,
part of the wonder and complexity of tafoni is that large
cavities occur along coasts, and intricate stonelace and
multiple-cell tafoni are common in deserts.
The fastest rates for tafone progression have been recorded
in coastal settings with up to 4.9 mm yr1 for cavity widening,
and 0.10.6 mm yr1 for cavity deepening, yet mean rates of
development for a 45 cm cavity have been recorded to take
100500 years (Mustoe, 1982; Sunamura, 1996; Pye and
Mottershead, 1995). Due to the nature of insolation-induced
drying and evaporation, lower walls in cells and cavities
weather and enlarge at slower rates than cavity roofs and
upper walls where evaporation rates are less. Consequently, for
a tafone cell to progress from stage 1 (cell) to stage 2 (cavity), a
century may pass for the initiation and development of a
1014 mm cell with the greatest dimension occurring in the
upper portions of each cell. Moreover, to extrapolate
Figure 6 Photograph of complex tafone development in the

Moenkopi Formation at Wupatki National Monument, Arizona. The
arrangement of ribs and cavities in the tafoni indicates the control of
horizontal lithologic structure (bedding planes), in addition to the
development stages of cavities-within-cavities-within-cavities
representing stages 45. The tafoni in this image measure 1.5 m
across, at an aspect of 255 1N.
weathering rates in near-shore marine settings, we can hypothesize about tafoni cycles as contributors to coastal retreat
(assuming no uplift or major climate change) at 100600 m of
retreat over 100 000 years (Figure 6).
4.7.3.1
Gnammas
Solution pits, weathering pits, pans, and gnammas are terms

that have been used to describe these pans and pits. Some
gnammas have been grouped by climate (in alpine, desert),
location (on bornhardts), topography (on summits, pediments), size (pans, pools), shape (oval, bowl), or having a
unique outline (animal shape, geometric). These are various
terms that have been collected from research conducted over
the last 150 years.
Gnamma is the term now widely used to describe an oval
or circular depression commonly containing water seasonally.
Pits, bowls are hemispherical in shape, whereas pans possess
flattened bottoms from increased lateral development over
vertical recession. Canoe is half of a pit developing against a
wall or joint, an elongated modification to a pit, whereas
armchairs or armchair hollows result from pan enlargement
on or toward a steeper slope (4201).
More unusual terms have been used for less common attributes of gnammas. Flask-shaped hollows exhibit distinctive
narrow openings to larger chambers. Paternoster gnamma or
pits is a rare but apt term that describes a series of pits and/or
pans visibly connected by spillways and channels. This term is
derived from their resemblance to rosary beads. Flares, lips,
hoods, and visors are somewhat synonymous terms that
have been used to describe the small ridges or rims that
overhang the pan or bowl. These generally occur in mature
tafone (stages 46) and may be due to case-hardening, coresoftening, and/or differential weathering.
Pot-hole is an incorrect use that has conventionally
applied to a circular depression as a result of stream channel
erosion.
Gnammas are affected by complex influences, and their
development by complicated feedback cycles. They evolve
through distinctive stages that are categorized through changes
in the pan and bowl shapes, widthdepth ratios, heightdepth
ratios, and sediment infilling. First minor depressions enlarge
and deepen, becoming circular or oval in outline through
progression. As bowls enlarge and bottoms flatten, spillways
119
Stone
Figure 7 Diagram of profiles and photographs illustrating the evolution of gnamma morphometry. Development is affected by complex
influences and feedback cycles but the cavities evolve from (1, 2) minor depressions and bowls, (3, 4) into broader and deeper hollows often
seasonally filling with water, (5, 6) until enough weathering by-products (i.e., grus, sand) accumulate to act as a plant substrate facilitating grass,
moss, sedge, and small plants to colonize the depression. Gnammas can range in size from o10 cm to 412 m. Reproduced from Twidale, C.R.,
Corbin, E.M., 1963. Gnammas. Revue de Geomorphologie dynamique 4, 120; Paradise, T.R., Yin, Z.Y., 1993. Weathering pit characteristics and
topography, Stone Mountain, Georgia. Physical Geography 14, 6881.
develop until channels linking adjacent gnammas develop.

Lateral growth increases as deepening decreases and erosion of
sediments decrease until they accumulate to support plant
grown. Once completely sediment-filled overall length, width
and depth weathering diminishes (Figure 7).
4.7.4
Stages of Gnamma Progression
1. Minor depressions and bowls develop from surface irregularity and/or lithologic weaknesses (i.e., jointing,
xenoliths, textural change).
2. Depressions enlarge and deepen, commonly seasonally
filling with water.
3. Bowls develop vertical walls resulting in pan-forms. Spillways are incipient. Seasonal precipitation removes accumulating silt, sand, and some pebbles.
4. Lateral growth rate increases and deepening decreases.
Bowls enlarge laterally and bottoms flatten to produce
broader pans. Surface skins and coatings may develop to
facilitate lipped and flask-like edges. Weathering byproducts accumulate.
5. Lateral growth rate decreases. Adjacent pits and pans may
coalesce into larger gnammas. Spillways enlarge with defined channels. Rock coatings may develop into case-
hardening that promotes the development of lips, visors,

and flask-shaped profiles. Sediment infilling begins with
weathering by-products (i.e., grus) or through aeolian
contributions of silt and sand.
6. Channel and spillways link adjacent gnammas. Sediment
accumulation supports plant growth (grass, moss, sedge,
small plants) to colonize the depression. Once sediment
and plant-filled, they act as sumps and/or as reservoirs for
adjacent downslope gnammas.
Gnammas, like tafoni, develop at rates controlled by climate and environmental factors. Climate controls moisture
availability both as atmospheric humidity and precipitation,
wettingdrying and freeze-thaw cycles, and through the
intervals of desiccation between episodes of bowl in-filling.
Lithology can affect gnamma initiation due from differential
weathering of the original surface (phenocrysts, xenoliths,
variability in constituent minerals, density, integrity),
and related disparities in permeability and porosity that
control water containment, and sediment accumulation. Environmental influences can include surrounding and ingnamma plant growth, spillway obstruction, and/or adjacent or
upslope tree growth or removal. Although anthropogenic
influences can affect development through aerial and
point-source pollution that may exacerbate weathering, and
human-induced damage (intentional or accidental) through
120
Irregular
depressions:
often multiple
Merging of
depressions
with
wth ng
o
r
l g peni
e
era
Lat ht de
lig
Large, shallow
irregular depressions
Flat-bottom,
shallow pits:
no sediment
Single pits:
sediment-filled with
connected channels
Ext
dee ensiv
pen e
ing
tom
Bot ing
Sumps
d
l
a
ra
gr
Late th
w
old
gro
esh
Thr nation
i
elim
De
epe
Integrated
nin
g
depressions
Deep isolated pits:

scarce sediments
Figure 8 Diagram illustrating the pathways of gnamma development in granite in the Namib Desert. Reproduced from Goudie, A.S., Migon, P.,
1997. Weathering pits in the Spitzkoppe area, Cental Namib Desert. Zeitschrift fur Geomorphologie 41, 417444.
spillway obstruction, trampling, and abrasion at popular

American tourist sites such as Stone Mountain (Georgia),
Yosemite National Park (California), and Enchanted Rock
(Texas), where visitors are free to wander and walk atop mature and incipient gnammas (Figure 8).
Gnammas are not believed to initiate, enlarge, widen, and
deepen at the same rates. Unlike tafoni that have environments that facilitate their maximal development such as arid
and coastal settings, gnammas occur across diverse landscapes
from marine to alpine and from humid to arid. However, it is
the substrate type and its permeability that appears to have the
greatest influence on their occurrence and progression.
The fastest rates for gnamma progression have been recorded on bare granite substrates such as the summits of
bornhardts, inselbergs, and mountaintops, yet their occurrence on steep slopes as armchair gnammas has been recorded
across the globe. However pervasive their occurrence on
granite and crystalline rock, other substrates including sandstone (arkosic), quartzite, greywacke, schists (granitoid), porphyry (quartz), siltstone, conglomerate, and limestone have
also been documented and studied.
Gnammas are believed to develop at rates much slower
than tafoni, and also at rates less able to be quantified due to
their much slower progression. In central Australia, gnammas
measuring 25 cm (1000 ) in diameter were believed to have
developed in a few thousands years at the most (Twidale and
Corbin, 1963), whereas Victorian-constructed seawalls exhibit
extensive tafone development over 150 years.
In cold climates like Patagonia, juvenile gnammas were
documented (stages 1 and 2) displaying no spillways, and
measuring 645 cm in length, 630 cm in width, and
1.17.5 cm in depth. Their volumes ranged from 4 l to less
than 0.1 l. Having initiated after glacial retreat between 600
and 1250 AD, it places their maximum recession for gnamma
length at 0.60.3 mm yr1, for width at 0.40.2 mm yr1 for
width, and 0.010.05 mm yr1 for depth (Dominguez-Villar,
2006). These represent rare measurements for gnamma
development, owing to their slow progression. Like all studies
of this nature, they cannot account for variations in rate over
time or nonlinear dimensional changes, so as lateral growth
increases over time, and deepening decreases, since these
gnamma are incipient, their maximum growth rates represent
only the earliest stages of morphological progression.
4.7.5
Processes of Development
Recent studies have determined the relative importance of

diverse and interacting influences on the formation of tafone
and gnamma. Increasingly research points toward the complex
and commonly nebulous polygenetic nature for their development where initiation and development is controlled by
many factors, both intrinsic (i.e., substrate composition,
structure) and extrinsic (i.e., climate, lichen overgrowth).
Here we will address the many influences that have been investigated and documented including the roles of lithology,
climate, environment, and plants and animals in their occurrence and growth.
4.7.5.1
Lithologic Influences
Gnamma and tafone initiate in zones of differential weathering on the rock surface, including variations in lithology,
structure, composition, and texture (Dragovich, 1969). Petrology and mineralogy of the substrate then play two primary
roles in the development of tafone and gnamma. Variability
in composition can influence differential weathering since
some lithologic constituents act as relative indurating agents
(i.e., iron, silica), or through the hydrolysis, dissolution, ionic
exchange, or hydration of weathering-susceptible minerals like
phyllosilicates (Mustoe, 1983), carbonates (Bartrum, 1936),
and feldspars (Sancho and Benito, 1990) which can weather
to fall out, enlarge to pry out neighboring constituents, and/or
weaken to erode from the substrate. Other components can
increase differential weatherability and subsequent cavity development through variations in integrity (i.e., jointing,
cleavage, texture). These variations may facilitate recession in
some portions, relative to little recession in others.
The other important role of lithology, and possibly the
more influential, lies in its structural and compositional effects
on porosity (percentage of void space to material) and permeability (percentage of active porosity or connected pore
space). In the ruined city of Petra, Jordan, Paradise (1995)
investigated the occurrence of tafone and cavernous weathering in sandstone. He found that minor constituent variation of iron oxides from 2 to 4% in the rock matrix,
significantly affected the rate of surface recession and tafone
development. Using architectural surfaces that have been
exposed since their construction (50 BC150 AD), analyses

showed that cavernous weathering features (tafoni, alveoli)
developed to depths of 200 mm when the matrix contained
only 2% iron oxides, whereas when the matrix contained 4%
iron levels, the stone exhibited nearly complete arrested recession. These high iron sandstone surfaces displayed original
Nabataean and Roman stonemason dressing marks. Although
iron was found to decrease weatherability, some components
(like carbonaceous matrices) showed accelerated recession
and increased development of tafoni.
Sunamura (1996) found that an increase in tafone development, distribution, and dimension was directly related to
the presence of weak bedding planes, jointing, and fractures in
the substrate rock. This was found to be especially important
in sheer cliff faces and single expanses of rock where variations
could be identified and related to tafone occurrence.
In Weston-Super-Mare, UK, limestone and sandstone
blocks were both used to build a seawall; the limestone remained relatively unweathered, whereas the sandstone exhibited obvious tafone development. Pye and Mottershead
(1995) found that this was due to the low permeability of
limestone and its decreased chemical susceptibility to saltinduced weathering, when compared to adjacent sandstone
blocks in the seawall. Lithologic constituents can facilitate the
initiation and development of tafoni and gnammas, and some
can nearly halt surface recession. Where salt is present it can
only facilitate tafone development when other factors (i.e.,
substrate type, permeability) are present.
In their research in central Spain, Sancho and Benito
(1990) confirmed that specific environmental factors influenced tafone (and gnamma) development in sandstone. They
found a positive correlation (r 92) between feldspar constituency in the rock (410%) and the distribution density of
tafoni a previously speculated influence (Mustoe, 1983). It
was explained that the relative susceptibility of feldspar to
weathering through hydrolysis was the culprit. It was clear that
the freed potassium cation (K ) mobilized to later recombine
and form potassium salts (i.e., sylvite) which would further
accelerate feldspar deterioration, releasing more potassium,
thus propagating a positive feedback cycle. They also found a
significant relationship between tafoni distribution and substrate conductivity (4150 mS cm1) in Spain a previously
unknown control. Increases in conductivity may be due to
lithologic constituents that can increase electron mobility, or
simply due to the increased presence of secondary salts known
to foment tafone and gnamma development.
Both in Mustoes (1983) seminal work on tafoni, and
Twidale and Corbins (1963) research on gnammas, the
decomposition of feldspars and phyllosilicates was cited as a
crucial influence on tafone development through differential
weathering. When all variables are normalized, then lithology
has been found to dramatically affect relative, differential
weathering.
However, it may be the role of lithologic structure that has
the greatest effect on cavernous weathering. The moment joint
patterns found in sandstones across the Colorado Plateau
have been shown to rapidly drain and dry these beds, allowing
for little throughflow, although these sandstone are some of
the more permeable rock formations on Earth. Yet when water
mobility is diminished such as in nonjointed areas, or at a
121
contact with a less permeable contact cavernous weathering

develops. This indicates that there is a balance required for
tafoni to develop between rock so permeable that water drains
out freely and quickly, and rock with so little permeability that
moisture is not mobilized within the strata. This may be
the reason why arid environments (hot and cold) exhibit a
broader diversity of cavernous features as a result of decreased
moisture, its movement through the substrate, and within its
ambient microclimate.
Secondary deposition of rock coatings and skins may also
influence permeability and subsequently the progression of
tafone and gnamma morphology. For example, since iron and
silica skins can diminish overall permeability, their initiation
can also decrease cavernous development. Spatial variation in
permeability due to lithologic changes or secondary skin development will cause variations in moisture mobility within
and across the substrate, which in turn produces a variability
in morphology. Coatings can also influence the rate at which
cavernous weathering proceeds through the precipitation of
secondary minerals or materials (i.e., iron oxide, calcite, kaolinite) into and among the gaps and pores of the substrate
surface. It has been found that rock subsurfaces may weather
faster than at the surface. In Baja California, it was found that
intrusive rock constituents (i.e., biotite, feldspar) beneath the
surface deteriorated to produce iron-rich solutions that mobilized outward to precipitated and indurate the surface. Accordingly, solutes produced from internal weathering were
precipitated at the surface to produce a coating, which facilitated cavernous weathering through a core-softening and/or
case-hardening (Conca and Rossman, 1985).
Since the earliest observations of gnammas across the
granite domes and outcrops of the Outback and Yosemite,
iron staining on the rims, lips, pans, and bowls has
been noted, measured, and studied. Prior work has detailed
that iron salts (oxides, sulfides, sulfates), amorphous silica,
silcretes, and/or calcretes can act as indurating and/or cementing agents that create coatings and skins. For instance, as
rocks and minerals deteriorate through hydration, hydrolysis,
and ionic exchange, weathering by-products (i.e., iron oxides
and clays) can combine to produce coating that can indurate
the surface (i.e., ferruginous cements). Measuring from 1 to
30 mm in thickness, although 15 mm skins are the most
prevalent (Goudie and Viles, 1997), these brown and orange
rinds have been recorded on tafoni and gnammas on each
continent and are considered by many researchers as prerequisites in the development of tafone and gnamma. Research continues to underline and examine the role of this
case-hardening from coatings in tafone and gnamma development, and may one day answer the question regarding their
complex nature.
With gnammas, lithology affects both their initiation and
development. Incipient gnammas commonly display an
irregularity in substrate granite or sandstone indicating that
their initiation was indeed a function of a variation in petrology (i.e., phenocryst size, xenoliths, orientation, shape,
texture). It was also found that on granite bornhardts like
Stone Mountain, Georgia, the gnammas occurred larger and in
greater frequency on spalled portions or leaves relative to
nonspalled substrates. It was hypothesized by Paradise and
Yin (1993) that in its exfoliation and subsequent expansion,
122
granites display less compact structures which permit greater

intercrystal water penetration which facilitates mineral hydration and deterioration. This would weaken the substrate,
accelerating granite breakdown, which aided in pit and pan
incipiency.
4.7.5.2
Environmental Influences and Salinity
Salt and its role in tafone development have been discussed

since the work of Bryan (1922) and Blackwelder (1929) in the
U.S. Southwest. Simply put, the common occurrence of tafoni
in saline environments like deserts and intertidal settings is a
sound indication that salt plays a crucial role in their development. However, the larger question looms why do tafoni
develop on some surfaces and not others, but in the same saltrich environments? Is salt essential, but only in conjunction
with other intrinsic and extrinsic influences?
Sodium chloride (halite) is the most prevalent form of salt
most commonly derived from seawater, however less common
salts occur, including sodium sulfate, calcium sulfate (gypsum), magnesium sulfate (epsomite), and potassium chloride
(sylvite) all representing common contributors to cavernous
weathering. These salts may be derived from seawater or
formed through the dissolution of outcrops of rock and
minerals, or water infiltration in and through strata containing
chlorides, carbonates, and sulfates. Moreover, salt can foment
both physical and/or chemical weathering. Dissolved salts
may be transported atmospherically or as crystalline aerosols
and deposited atop the rock substrate where they dissolve to
precipitate, or simply accumulate as salt crystals, expanding to
pry apart crystalline or grained rocks like granite or sandstone
(Rodriguez-Navarro et al., 1999). As salt crystals develop, they
can physically force apart crystal boundaries opening the rock
surface, causing disaggregation and surface recession, and the
subsequent deposition of its deterioration by-product, such as
grus, sand, or silt.
Salt also has an accelerating effect on the dissolution of
silica. Increasing the concentrations of salt ions such as chloride, sulfides, and/or sulfates increases the weathering rates of
silicate rocks (granite), minerals (quartz), and amorphous silicates (opal). In addition, the presence of salts can modify the
pH of water that, in turn, can accelerate weathering and tafone
and gnamma initiation and progression. The simple addition of
seawater (and sea salt) has been found to greatly increase the
solubility of quartz, whereas its presence can instigate the
weathering of pyrite and other sulfites to produce acidic solutions (pH 35) that also increase weathering. Salts can speed up
the deterioration of clasts and mineral components through
mechanical and chemical means, which then expands the
boundary interstices, permitting more salts or saline solutions
to enter, thus weathering more and creating a positive feedback
cycle of breakdown (Huinink et al., 2004).
It has also been found that the length of the wetting
and drying episodes and the rate of evaporation are crucial
parameters in the role of salt in tafone and gnamma development. When drying periods are short, the evaporation
boundary remains in contact with the substrate surface during
desiccation, depositing salts in proportion to the period of
drying. Consequently, salt deposition is greatest at the more
exposed portions of the cavities, pits, and pans. However,

when the drying episode is long, the evaporation boundary is
in little contact with the surface the greatest salt deposition
then occurs in the more sheltered parts of the cavities where
the drying rates are lower: tafone backwalls, and beneath
overhanging gnamma rims (Huinink et al., 2004). This disparity in salt deposition causes differential weathering which
then triggers marginal and subsurface areas to enlarge and
deepen. This may explain why tafone are commonly observed
expanding upward with floors weathering at a slower rate
than roofs, and gnammas expanding within the seasonal
waterline thus producing lips, rims, and flask-forms, from the
vertical pan walls.
Development of tafoni and gnammas can indeed accelerate
their own progression in saline environments. Many studies
emphasized the primary influence of salt on tafone and
gnamma development; however, it is increasingly clear that
these weathering and erosional features are polygenetic in
origin, attributed to complex tandem influences and feedback
cycles.
4.7.5.3
Biotic Influences
The role of plants and animals in stone weathering has been

discussed since Strabos observations across the Mediterranean
2000 years ago (22 AD). Like the effects of salinity, plants and
animals have been found to act as mechanical and chemical
weathering agents. The attachment of lichens on rock is
commonly considered a destructive agent on rock integrity
through the penetration of rhizines into the substrate and
prying and separating the substrate constituents. Also, lichen
attachment can roughen the substrate beneath the cortex,
abrading the surface, thus preparing it for differential weathering and the initiation of tafone or gnamma. In addition to
rhizinal penetration, the production of oxalic acid beneath the
lichen cortex can exacerbate the destructive capabilities of lichen overgrowth (Paradise, 1997). However, some lichen
species and subspecies (i.e., Lecanora sp) have been found to
indurate the rock beneath the cortex through the production
of an oxalate skin, whereby the sandstone, granite, or limestone actually weathers at a slower rate than areas adjacent
with no lichen overgrowth (Paradise, 2005). Additionally,
organic acids from humic decay have also been identified as
important weathering agents. Some organic compounds can
act as coatings on reactive surfaces, thus decreasing weatherability; however, in general, organic acids consistently accelerate weathering rates (Young et al., 2009). Amorphous and
crystalline silica solubility increases in the presence of organic
acids. In the Hawkesbury Sandstone near Sydney, Australia, it
was found that infiltrating acidic waters dissolved the iron
oxide matrix cement to release the clasts to produce cavernous
weathering (Young et al., 2009).
Commonly mistaken as tafoni and gnammas are the cavities and bowls produced by pholads (Pholas dactylus). These
small marine bivalve mollusks (26 cm) bore into wood, clay,
concrete, or soft rock for protection leaving a cavity not unlike
one or a group of tafone (Boxerman, 2006). Their appearance
is much like incipient tafoni; however, they are bowl-shaped
without thin ribbing and polygonal cell arrangements.
Although these boreholes can be mistaken as small gnammas

on horizontal surfaces, or tafoni on vertical faces, they were
created through organic processes. It must be noted however
that pholad boreholes can initiate the development of tafoni
in near-shore environments.
4.7.5.4
Climate and Insolation
The degree of differential weathering across a surface can be

further enhanced by differences in the environmental and
climatic factors like wind, humidity, and insolation. The rate
of salt weathering is known to accelerate as wetting and drying
cycles increase; however, the duration of the wetting and
drying episodes and the lapses between have been identified as
instrumental in the progression of tafoni and gnamma. Although singular influences on tafone and gnamma development have been researched and discussed (and emphasized),
most studies have also addressed an underlying effect(s)
controlled by environmental factors. Temperature variations,
insolation fluctuations, wind, and humidity variability have
been shown to affect changes in the rate of cavernous weathering development.
Since the late 1930s, insolation has been relatively abandoned as a primary factor in weathering due to the important
laboratory simulations conducted by Blackwelder (1933).
However rare, research since has examined the power of
sunlight and heat in tafone and gnamma development. Popoff
and Kvelberg (1937) found the highest frequency of gnamma
(although they are called tafoni) on granite surfaces exposed
to the greatest temperature fluctuations and refuted Blackwelders widely accepted findings.
Using aspect as a surrogate for insolation and temperature,
Paradise (2002) examined the frequency, arrangement, and
dimensions of tafoni occurring on architectural surfaces in
Petra, Jordan. By correlating tafone size and number to aspect,
significant relationships were revealed. In studies as these,
southern faces commonly exhibit the largest cavities; however,
in Petra, the widest and deepest tafoni were found between
2302701N and 1401701N. This bimodal distribution indicates a different role of sunlight beyond simple heating since
afternoon heating would have increased on western aspects,
123
and direct insolation would have facilitated growth on

southern exposures. Since the largest tafoni dominated on
both western and eastern, insolation was found to have the
greatest effect on the development of tafoni in arid settings,
when in conjunction with increased wettingdrying and/or
heatingcooling cycles. Temperature fluctuations and cycling
have been found to loosen sandstone clasts or weaken matrix
cements. (Figure 9)
Not only are thermal expansion and/or wettingdrying
cycles responsible for their growth, freezing and thawing fluctuations may be as important in frigid settings. In
Antarctica, French and Guglielmin (2000) explained that the
development of cavernous weathering on granitic and gneissic
metasedimentary rocks was related to the microfracturing
of quartz minerals in these cold environments. Quartz fractures
under cryogenic conditions, especially when salts as
by-products of weathering, lower the freezing temperature.
Their findings explain why cavernous weathering like gnamma
and tafoni may be so well developed in arctic and polar deserts.
Wind has also been emphasized as an overlooked influence; however, it may present a quandary common in all
geomorphological research. How important is wind in the
development of cavernous phenomena when it cannot be
separated from other effects like salt weathering, or hydrolysis
of mineral constituents? In his early work, Futterer (1899)
stressed the importance of wind in tafone and gnamma development in granite, whereas Blackwelder (1929) emphasized its role in deserts, explaining that it was an erosional
factor, and not an abrasive agent. However, in many arid settings, sand storms have been observed scouring cliff and
boulder faces in a single event, removing large expanses of
tafoni (Burton, 1879). So how does wind influence tafone and
gnamma development; does it scour the rock to produce
depressions, while somehow unaffecting its walls and rims?
As Blackwelder implied, wind may not act an as abrasional
element, but may be crucial in the removal of weathering
by-products within the cavities and niches, like sand and grus.
Rodriguez-Navarro et al. (1999) also corroborated the role of
salt in cavernous weathering (in oolitic limestone), but they
also found that wind was instrumental in accelerating surface
evaporation to produce the greatest relative disparities between
Tafoni dimensions and aspect at Anjar quarry, Petra, Jordan

200
(n = 540)
Feature dimension (cm)
Length
150
Width
Depth
100
50
0
(north)
90
(east)
180
(south)
270
(west)
360
(north)
Aspect (degrees)
Figure 9 Graph illustrating the relationship between aspect (0360 1N) and tafoni dimension in Petra, Jordan. Reproduced from Paradise, T.R.,
2002. Sandstone weathering and aspect in Petra, Jordan. Zeitscrift fur Geomorphologie 46, 117.
124
Dry rock
substrate
Developing
cavity
Enlarging
cavity
Rock
moisture
Moisture
boundary
Figure 10 Diagram of cavity development illustrating the importance of the moisture boundary and its relationship to tafoni or gnamma
morphology. The cavity evolves fastest (through deepening and/or widening) at the moisture boundary where wetting and drying cycles may be
diurnal or seasonal. Reproduced from Huinink, H.P., Pel, L., Kopinga, K., 2004. Simulating the growth of tafoni. Earth surface process.
Landforms 29, 12251233.
dry and wet portions. This leads to localized supersaturation

that promoted salt weathering. Moreover, it was discovered that
salt was not facilitating chemical weathering, but physical deterioration though constituent prying and heaving.
Since Bryan (1922) and Matthes (1930), it has been emphasized that moisture plays an important role, if not primary,
in the initiation and progression of rock cavities. Bryan mentioned visible moisture, dripping water, or sensed humidity in
the niches and cavities of Chaco Canyon, New Mexico, whereas
Matthes emphasized the need for channels from and between
gnammas that acted to remove grus and debris from the deteriorating bowl walls in Yosemite. This action then leaves a
fresh granite surface upon which weathering can progress. If
water cannot drain from the bowl or pan, then erosion of the
detritus is only possible through wind action a rare occurrence within these receding ground level cavities.
Finally, it was found that the role of moisture was explained
in terms of the boundaries that shift as ambient moisture regimes fluctuate. Huinink et al. (2004) established that cavities
like tafone grew as a function of mobilized moisture fronts
within the substrate. In essence it was found that the cavity
backside and interior develops within the dry rock, whereas the
walls and edges break down within moist or saturated areas of
the substrate. This brings new attention to the powerful influence
of wetting and drying within and on the rock surface (Figure 10).
4.7.5.5
Feedback Cycles
In nature and society, feedback occurs when the product from

a process impedes its own process (negative feedback), or
increases its own process (positive feedback). Such mechanisms and processes have been identified in cavernous weathering and represent the sequence of causes and effects that
accelerate the rate of development, or slow or stop the growth
of tafone and gnammas. Generally, weathering processes exploit lithologic variability so once deterioration begins, rocks
degrade at faster rates than neighboring unweathered rock,
and tafoni and gnammas develop. The bowls, cavities, pits,
and interiors then weather at rates faster than surface faces and
walls and a self-reinforcing loop develops. Weathering accelerates and the rate of change is exponential. However, as
cavities and bowls enlarge, they can instigate the growth of
epilithic coatings (like rock varnish or cyanolichens), which
decreases the rate of weathering, or altogether stops it. This
differential weathering can be attributed to intrinsic factors

like variations in lithology, and/or to the extrinsic influences
such as microclimatic variations between the substrate exterior
and its interior (i.e., salinity, humidity). Feedback loops can be
affected positively and/or negatively through these influences.
For example, the induration of surfaces can result in core
softening or case-hardening; crystalline rocks like granite and
quartzites tend to core soften, whereas clast-matrix rocks like
sandstone or arkose tend to develop case-hardening. As the
cavity enlarges, it create an environment that can accelerate its
own expansion or halt its development.
Moisture fluctuations, wettingdrying, and/or freezing
thawing can also have exacerbating effects on cavernous development. During episodes of saturation or high humidity,
water is retained in the sheltered cavities, and salt accumulation is diminished. However during dry periods, capillary
water is drawn to the cavity wall to evaporate, precipitating
interstitial salt crystallization. During both dry and moist
episodes, the environment created in the cells, bowls, and
cavities creates a positive feedback cycle that facilitates
weathering.
Tafone development, however, may pass through negative
and positive feedback cycles from incipience to maturity since
the sequence has been found to be sigmoidal in rate with a
nonlinear progression: slow during incipiency to accelerate
during its mature stages when cavities may deepen faster than
widen, to a slow rate of growth as the stages pass (Goudie and
Viles, 1997). In tafoni, the earliest stages of initiation and
development may be limited by low salinity, followed by enlargement and deepening dominated by increasing salt concentrations within a positive feedback cycle. Cavity
development then progresses until the cavity is too large to
effectively wick moisture from the substrate and/or maintain a
balance between salt mobilization, accumulation, and related
weathering. Hence, the cavity outgrows its own capacity to
enlarge and a negative cycle follows.
Gnammas also initiate and enlarge due to feedback influences. When water collects in the irregularities on rock
surfaces, then the water localizes physical (freeze-thaw) and
chemical (hydrolysis, hydration) weathering and granular
disintegration follows. Erosional forces like water and wind
eliminate the weathering by-products (i.e., grus), and the
depressions enlarge. This traps more water and the process
accelerates as a positive feedback cycle.
Negative feedback cycles in their development have also

been identified through the growth and/or accumulation
of secondary mineral precipitates, sediment accumulation, or
plant overgrowth. Mustoe (1982) and Paradise (2002) both
found that lichens decreased weathering by acting as protective biotic blankets. Although rhizinal penetration and oxalic
acid production in lichens has been found to accelerate
deterioration, adnate lichen overgrowth can diminish weathering and cavity enlargement in tafoni and gnammas. Secondary depositions like rock varnish may also develop,
indurating surfaces by filling mineral and clast pores and
boundaries. Such overgrowth and coatings act as negative
feedback mechanisms whereby enlargement facilitates endolithic overgrowth and coating buildup until weathering is
arrested and the tafoni and gnamma become relict, no longer
actively weathering.
4.7.6
Summary
Why tafoni and gnammas develop still puzzles geographers,

conservators, and geologists, although influences of salinity,
mineral solubility, lithology, and microclimatic influences are
still considered essential. Tafoni and gnammas develop over a
wide range of scales and environments, and many weathering
processes work in tandem, and in feedback cycles, to produce
these often delicate and symmetrical, or large, deep, and cavernous features. However, little is still understood as to how
these bowls, pans, honeycombs, alveoli, and armchairs initiate, develop, enlarge, and coalesce. Some develop within
decades and others over millennia depending on the complicated controls that may be intrinsic like rock type, or
extrinsic like microclimatic influences or broader shifts in
climate.
At smaller scales (centimeter to meter) it is now widely
accepted that salt mobility and evaporation is critical in cavity
development on vertical faces (tafoni, alveoli, stonelace,
niches), whereas differential weathering through hydrolysis,
hydration and/or ion exchange is crucial on horizontal to
vertical surfaces (gnammas, tafoni, niches). However important single agents may have been identified at specific sites, to
fully grasp the complicated and integrated processes responsible for tafoni and gnammas, we must begin to examine the
hierarchal associations between the many processes known to
affect their incipience and progression. Some sites may indeed
indicate that one process is principal, whereas another may
point to another influence as significant.
Since Blackwelder (1929) speculated on the polygenetic
origins of niches and cavernous weathering, we have identified
a number of processes that operate in tandem and/or individually. With contemporary research advocating feedback
sequences, the very nature of positive and negative feedback
cycles supports the idea that tafoni and gnammas develop
from small pits and hollow to niches large enough for a horse
and rider or basins big enough to bathe a family through
complicated and interconnected processes. Examples of such
an integrated process chain may include the following: (1)
Lithological disparity in the rock substrate initiates a depression. (2) Differential weathering deteriorates boundaries
of clasts and mineral constituents that instigate disaggregation
125
and cells and bowls develop. (3) As cavities enlarge, differential weathering (physical and chemical) continues. (4)
Erosion through water and wind remove the by-products of
weathering to support accelerated weathering and cavities
enlarge, or by-products remain to decrease rate of weathering
and cavity progression slows down or stops.
Rodriguez-Navarro (1998) speculated on the causes
of tafoni identified on Pathfinders photographs from Mars.
He argued that salt weathering was the primary agent, in
conjunction with chemical weathering and frost shattering.
Such conjecture may be valuable regarding our terrestrial
phenomena, as well as these fascinating cavities observed on
Martian boulders. As research continues on Earth (and other
planets), it is imperative that we continue to identify the influences on tafone and gnamma development and their
interconnected hierarchies of associated factors, how they relate to each site, at varying scales (spatial and temporal), on a
diversity of substrates, and within a varied range of environmental regimes.
References
Austin, R., 1856. Report by assistant surveyor Robert Austin of the expedition to
explore the interior of western Australia. Journal of the Royal Geographical
Society (London) 26, 235274.
Bartrum, J.A., 1936. Honeycomb weathering of rocks near the shoreline. The New
Zealand Journal of Science and Technology 18, 593600.
Blackwelder, E., 1929. Cavernous rock surfaces of the desert. American Journal of
Science 217, 393399.
Blackwelder, E., 1933. The insolation hypothesis of rock weathering. American
Boxerman, J.Z., 2006. The evolution of tafoni on coastal sandstones in northern
California. Masters Thesis, Department of Geosciences, San Francisco State
University, unpublished.
Bryan, K., 1922. Erosion and sedimentation in Papago country, Arizona. U.S.
Geological Survey Bulletin 730-B, 1990.
Bryan, K., 1928. Niches and other cavities in sandstone at Chaco Canyon,
New Mexico. Zeitschrift fur Geomorphologie 3, 125140.
Burton, R.F., 1879. The Land of Midian (vol. 1, 2). C. Kegan Paul and Company,
London, 665 pp.
Caldenius, C., 1932. Las glaciaciones cuaternarias en la Patagonia Y Tierra del
Fuego. Geografiska Annaler 14, 164.
Calvert, A.F., 1897. My Four Tours in Western Australia. Heinemann Publishers,
London.
Carnegie, D.W., 1898. Spinifex and Sand. Arthur Peason Printers, London.
Conca, J.L., Rossman, G.R., 1985. Core softening in cavernously weathered tonalite.
Journal of Geology 93-1, 5973.
Cotton, C.A., 1922. Geomorphology of New Zealand. Dominion Museum,
Wellington.
Dahl, R., 1966. Blockfields, weathering pits and tor-like forms. Geografiska Annaler
48, 5585.
Darwin, C.R., 1839. The Voyage of the Beagle. Collier & Son Publishers, New York.
De Prado, Casiano, 1864. Descripcion Fsica y Geologica de la Provincia de
Madrid. Junta General de Estadstica, Madrid, 232 pp.
Dominguez-Villar, D., 2006. Early formation of gnammas (weathering pits) in a
recently glaciated area of Torres del Paine, southern Patagonia (Chile).
Dragovich, D., 1969. The origin of cavernous surfaces (tafoni) in granitic
rocks of southern South Australia. Zeitschrift fur Geomorphologie 13(2),
163181.
French, H.M., Guglielmin, M., 2000. Cryogenic weathering of granite,
Northern Victoria Land, Antarctica. Permafrost and Periglacial Processes 11,
305314.
Friese, F.W., 1938. Inselberge und inselberglandschaften in granit- und gneissagebieten Brasilians. Zeitschrift fur Geomorphologie 10, 137168.
126
Futterer, K., 1899. Uber windkorrasion am Heidelberg Schlossturm. Mitteilunger

der grossherzoglichen badischen geologisches Landesanstalt 5, 471
495.
Gill, E.D., Segnit, E.R., McNeill, N.H., 1981. Rate of formation of honeycomb
weathering features on the Otway Coast, S.E. Australia. Proceedings of the Royal
Society of Victoria 92, 149154.
Goudie, A.S., Viles, H.A., 1997. Tafoni, Alveoles, Honeycombs, and Overhangs, in
Salt Weathering Hazards. Wiley, New York, 256 pp.
Hejl, E., 2005. A pictorial study of tafoni development from the 2nd millennium BC.
surface process. Landforms 29, 12251233.
Hume, W.F., 1935. Geology of Egypt. Government Press, Cairo.
Ives, J.D., 1978. The Maximum extent of the Laurentide Ice Sheet along the east
coast of North America during the last glaciation. Arctic 31, 2453.
Landvik, J.Y., 1994. The last glaciation of Germania land and adjacent areas,
Northeast Greenland. Journal of Quaternary Science 9, 8192.
Matthes, F.E., 1930. Geologic history of Yosemite Valley. USGS Professional Paper
160, 131 pp.
Moore, G.F., 1842. Diary of Ten Years of an Early Settler in Western Australia.
University of Western Australia Press (reprinted in 1978), Perth.
Mustoe, G.E., 1982. Origin of honeycomb weathering. Geologic Society of America
Mustoe, G.E., 1983. Cavernous weathering in the Capitol Reef Desert, Utah. Earth
Netoff, D.I., Shroba, R.R., 2001. Conical sandstone landforms cored with clastic
pipes in Glen Canyon National Recreation Area, southeastern Utah.
Paradise, T.R., 1995. Sandstone weathering thresholds in Petra, Jordan. Physical
Paradise, T.R., 1997. Sandstone weathering from Lichen Overgrowth, Red Mountain,
Arizona. Geografiska Annaler 79(3), 177184.
Paradise, T.R., 2002. Sandstone weathering and aspect in Petra, Jordan. Zeitscrift
fur Geomorphologie 46, 117.
Paradise, T.R., 2005. Weathering of sandstone architecture in Petra, Jordan:

influences and rates. In: Stone Decay in the Architectural Environment. GSA
Special Edition 390, pp. 3949.
Paradise, T.R., Yin, Z.Y., 1993. Weathering pit characteristics and topography, Stone
Mountain, Georgia. Physical Geography 14, 6881.
Penck, A, 1884. Geographische Wirkungen der Eiszeit. Deutschen Geographentages,
volume 13 of 15, S66 (Verhdgn. Des IV).
Popoff, B., Kvelberg, I., 1937. Die tafoni verwitterungserscheinung. Acta University
Latviensis 4(6), 129368.
Pye, K., Mottershead, D.N., 1995. Honeycomb weathering of carboniferous
sandstone in a seawall at Weston-super-Mare, UK. Quarterly Journal of
Reusch, H.H., 1883. Note sur la gelogie de la Corse. Societe Geologique de Paris
(Bulletin) 11, 5367.
Rodriguez-Navarro, C., 1998. Evidence of honeycomb weathering on Mars.
Rodriguez-Navarro, C., Doehne, E., Sebastian, E., 1999. Origins of honeycomb
weathering: the role of salts and wind. Geological Society of America Bulletin
1118, 12501255.
Sancho, C., Benito, G., 1990. Factors controlling tafoni weathering in the Ebro
Basin (NE Spain). Zeitschrift fur Geomorphologie 34, 165177.
Stephens, J.L., 1837. Incidents of Travel in Egypt, Arabia Petraea and the Holy
Land. Harper & Brothers, New York.
Sunamura, T., 1996. A physical model for the rate of coastal tafoni development.
Twidale, C.R., Corbin, E.M., 1963. Gnammas. Revue de Geomorphologie dynamique
4, 120.
Vidal Romani, J.R., Twidale, C.R., 1998. Formas y Paisajes Graniticos. Serie
Monografias 55. Servicio de Publications da Universidade da Coruna, A Coruna.
Young, R.W., Wray, R.A., Young, A.R., 2009. Sandstone Landforms. Cambridge
Biographical Sketch
Dr. Tom Paradise is a geography and geosciences professor at the University of Arkansas and past director of the
King Fahd Center for Middle East Studies. He comes from a diverse background in the geology, climatology,
material sciences, cartography, architecture, and Mediterranean geography. Having researched the unique,
decaying architecture of Petra, Jordan, since the 1980s, he has published more than 40 articles, chapters, and
books on the subject and continues to advise foreign agencies on cultural heritage management, stone architectural deterioration, and Middle Eastern and North African architecture. In addition, Paradise has published
four Atlases, including the recently released illustrated atlas of Arkansas. The Atlas of Hawaii is one of the most
popular books of its kind and won the Hawaii Book of the Year award (Ka Palapala Pookela). Prof. Paradise has
taught abroad at Universities in Rome, Venice, Amman, and Tunisia, as well as in the USA in Georgia, Hawaii,
Arizona, and California. He currently lives in Fayetteville, Arkansas, USA.
4.8 Weathering Mantles and Long-Term Landform Evolution

P Migon, University of Wroc"aw, Wroc"aw, Poland
4.8.1
4.8.2
4.8.3
4.8.4
4.8.5
4.8.5.1
4.8.5.2
4.8.5.3
4.8.5.4
4.8.6
References
Introduction
Weathering Mantles and How They Form
Deep Weathering Through Geological Time
Etching and Stripping
Geomorphological Signatures of Etchsurfaces
Inselbergs
Multiconvex Relief
Basins
Plains
Conclusions
127
128
131
133
137
137
139
141
142
142
142
Abstract
Weathering processes are not confined to surface conditions. Thick weathering mantles are ubiquitous around the world
and show various ages, from Mesozoic to Quaternary. Of considerable geomorphological importance is stripping of preweathered materials that exposes an etched surface. Etchsurfaces at different stages of evolution are present around the
globe, and current surface development in low latitudes is mainly through episodic etchplanation. Inselbergs, multi-convex
relief, and topographic basins are geomorphic signatures of landscape development through etching and stripping.
4.8.1
Introduction
Weathering, although generally defined as rock alteration and

breakdown at or near the topographic surface under the primary influence of atmospheric factors, is not a process restricted to the uppermost few meters of the lithosphere.
Nothing can be more incorrect than equating weathering with
the formation of superficial angular block fields, outcrop
flaking, and the origin of soils. In reality, the effects of weathering can be observed at much greater depths. Mineralogical
and textural changes driven by interactions with meteoric
waters are commonly traced down to a few tens of meters
(Figure 1), whereas weathered rocks with thicknesses exceeding 100 m are fairly common in ancient landmasses of
Australia or Africa (e.g., Taylor and Eggleton, 2001) and have
been recorded in other continents as well. Even in the formerly glaciated parts of Europe and North America, bare
bedrock surfaces and thick drift coexist with thick regolith,
inherited from the preglacial period, that survived glacial
erosion during the Quaternary (e.g., Lidmar-Bergstrom et al.,
1999). These occurrences are not accidents. They prove that
the significance of weathering extends through time and threedimensional space of the lithosphere and that well-developed
upper soil horizons and exfoliated boulders are all but tips of
the iceberg. To account for this great vertical extent and to
differentiate from surface breakdown, two related notions
Migon, P., 2013. Weathering mantles and long-term landform evolution. In:
Shroder, J. (Editor in Chief), Pope, G.A. (Ed.), Treatise on Geomorphology.
Academic Press, San Diego, CA, vol. 4, Weathering and Soils
have been introduced into the scientific vocabulary and are

widely used now. One of these is deep weathering. The key
message in the term is an emphasis on the great downward
extent of meteoric alteration and its effects, far beyond the
sphere of soil science interest. No numerical boundary value
can be easily provided for weathering to become deep, but
normally a thickness above 23 m would be expected. The
other one is weathering mantle, used to denote the presence
of material that has been produced through mineralogical and
textural alteration of parent rock, sometimes to the extent that
renders identification of this parent rock difficult, but which
still remains in situ. Saprolite is a term used as an equivalent
to weathered mantle, although in some cases it is restricted to
the upper part of the weathered mantle, where signs of alteration become pervasive (Figure 2). The weathering mantle
and any transported material that rests above it form regolith,
a term coined by Merrill (1897) to describe [this] entire
mantle of unconsolidated material, whatever its nature or
origin and used particularly in Australia.
The widespread occurrence of weathering mantles across
the globe and the efficacy of deep weathering cannot be
overlooked by geomorphologists. Some scenarios of longterm landscape evolution, whether classic and dated back to
the early 20th century or modern, based on sophisticated
numerical modeling might suggest otherwise. In fact, weathering is neither a mere preparatory process whose significance ends as soon as debris is liberated from parent rock, nor
is deep weathering an anomalous component in the overall
landscape evolution. In this chapter, evidence is provided to
demonstrate that weathered mantles are indeed extremely
widespread and diverse, have been formed, stripped, and
http://dx.doi.org/10.1016/B978-0-12-374739-6.00071-3
127
128
Weathering Mantles and Long-Term Landform Evolution
for using a general equation similar to the famous Jennys

equation for soils:
WM f R,C,G,t;
Figure 1 Deep weathering profile developed from granite,

Meghalaya Plateau, India. Faintly weathered boulders (right) and
in situ granites (bottom left) indicate proximity to the
weathering front.
Transported
material
Pedolith
Regolith
Saprolite
Regolith
Saprolite
ne
Saprock
to
es
W
ea
the
Co
ont
g fr
rin
Solid rock
Figure 2 Terminology used to describe weathering mantles. Note

that saprolite may mean different parts of the weathering mantle to
different researchers.
re-formed throughout geologic time, and that the deep weathering is a process fundamental for our understanding of
how landscapes evolve in the long-term, particularly in low- to
medium-relief regions.
4.8.2
Weathering Mantles and How They Form
Mantles of deeply weathered rock vary enormously in terms of

thickness, mineralogy, and physical properties. These differences result from a multitude of factors but can be accounted
where WM characteristics of weathered mantle, R rock

properties, C climatic factors, G geomorphology (local and
regional relief), and t time.
It is important to note that each of these factors, perhaps
except time, actually covers a range of parameters influencing
the extent and rate of weathering and that they may influence
the progress of weathering synergistically or antagonistically.
Rock properties and local morphology decide about susceptibility of material at a given location to weathering,
whereas climate and time are external factors that control rates
of weathering and the nature of the end product, respectively.
Among rock properties, mineralogy and discontinuities are the
most important. Mineralogical composition dictates the ease
with which particular rock-forming minerals can break down
in the presence of water, whereas discontinuities provide avenues for water movement through the rock mass and expose
more and more material to weathering. Discontinuities may
occur at different spatial scales, from intra-granular cracks
through joints of different origin, bedding planes in sedimentary rocks, to major fault lines commonly associated with
mechanical crushing of rock. Climatic factors influencing the
style and rate of deep weathering are many too, including
rainfall regimes (annual totals, seasonality, and intensity) and
hence water supply to the rock and temperature, but others
important for superficial weathering are largely irrelevant
such as very short-term changes in solar radiation. Table 1 lists
circumstances in which deep weathering may be expected to
be either most or least efficient.
Taken together, these three factors of space not only influence the rates with which the weathering front descends
into the rock and the mineralogical changes take place, but
decide about the likelihood of the resultant weathering mantle
to survive. The balance between the rate of weathering profile
lowering and the rate of regolith removal by erosion is critical
and will be discussed in more detail in the next section. The
factor of time needs to be considered separately. Taylor and
Eggleton (2001, p. 76) argued that same rocks in different
environments may weather to the same end product, but it
would just take longer if these environmental factors are less
favorable for deep weathering to occur. This is generally unlikely to happen as protracted periods of surface stability (no
erosion) would be required. With this constraint in mind,
kaolinite-rich weathered mantles derived from granite and
gneiss in mid- and high latitudes are generally presented as
products of efficient deep weathering in warm and humid
environments of the Palaeogene. However, Bird and Chivas
(1988) have shown the existence of low-temperature kaolinitic mantles in Australia, where the rates of long-term erosion
were very low indeed.
Given the multitude of factors influencing deep weathering, it is not surprising that there is such a variability between weathered mantles observed. As far as thickness is
concerned, figures quoted in the literature (e.g., Ollier, 1984;
Thomas, 1994; Migon and Lidmar-Bergstrom, 2001) give an
idea of how deep the surface weathering can reach. In the
129
Table 1 Factors enhancing deep weathering

Factor
Rock properties
Mineralogy and rock
chemistry
Fabric
Discontinuities
Conditions for enhanced deep weathering
Conditions for limited deep weathering
Rocks contain a large proportion of easily weatherable

minerals such as olivine, chain silicates or
carbonates
Rock composition involves less silica but more iron,
calcium, sodium and magnesium
Quartz-rich rocks are less prone to the formation of

thick weathered mantles
Fabric is weak and breakdown of rock/mineral

structure can take place easily (e.g. through
hydration of biotite)
Rock porosity (primary and secondary) increases
allowing for more efficient water percolation
Progress of weathering is slower in rocks with higher

silica content; in feldspar-rich rocks, potassium
feldspars are more difficult to break down than
Na- and Ca-feldspars
Interlocking crystals and grains, mineral swelling
insignificant
Primary porosity very low
Fracture density in rock is higher, allowing more to

infiltrate
Joints are discontinuous and form an irregular
network, slowing down water movement through the
rock
Massive rocks with widely spaced fractures offer few

avenues for water
Wide open and continuous joints allow for quick
passage of water through the rock, with limited
waterrock interaction
Water availability is higher in mid-slope and footslope

setting
Little surface erosion allows weathering products to
remain in place and the weathered mantle in increase
in thickness
Upper slopes and hilltops shed water and remain less

affected by weathering
In higher relief erosion is more efficient, limiting
preservation potential and residence time of
weathering products
Higher rainfall brings more water to the weathering

system, throughout a year or in specific seasons
Higher temperature impacts positively on the rates of
chemical reactions
Deep weathering limited in arid environments
Geomorphology
Local relief
Regional relief and longterm trends in landscape
evolution
Climate
Precipitation
Temperature
Figure 3 Thick tropical weathering profile on gneiss exposed in a

road cut in Serra do Mar, southeast Brazil. The uppermost 34 m is
colluvium, affected by soil creep and bioturbation.
tropical belt, weathering mantles of 2040 m thickness are

common (Figure 3), but in numerous places the actual
thickness is much higher, commonly in excess of 100 m. For
example, in metamorphic rocks building the Zaire copper belt
weathering profiles are at least 150 m deep (Thomas, 1994),
Low temperatures slow down chemical reactions;

ground freezing may temporarily shut the deep
weathering system (see Chapter 4.3 and
Chapter 4.15)
whereas Ollier (1984) referred to a number of thick examples

from Australia. Outside the tropics such considerable depths
to weathering front occur too. For example, slates in Rhenish
Shield, Germany, are weathered down to more than 100 m,
whereas in the Bohemian Massif kaolinization of granites and
arkosic sandstones may be in many places traced to nearly
100 m below their top surfaces (see review in Migon and
Lidmar-Bergstrom (2001)). In a few cases, figures as big as
4300 m are referred too, but these are generally suspect to
reflect weathering processes superimposed on the previous
phase (or phases) of hydrothermal alteration. For many years,
this was the preferred interpretation of thick commercially
exploited granite-derived kaolinitic clays in Dartmoor, SW
England, with emphasis on the earlier, hydrothermal phase.
More recently though, down-profile weathering has been
favored after Sheppard (1977) demonstrated that isotopic
signatures in kaolinite are consistent with its low-temperature
weathering origin.
Deep weathering is not necessarily uniform in space and
the weathering front is rarely flat and planar. The considerable
variability in depth can be seen even at an outcrop scale
(Figure 4), but it is most evident if dense networks of boreholes are compared. One of the early well-documented studies
was by Thomas (1966) from the Jos Plateau in northern
130
16 50
500 m
Thickness of weathering
mantle (in metres)
0
10
25
Outcrop of
unweathered
serpentinite
Figure 4 Deep weathering profiles are commonly highly complex

structurally. In this example, from Lausitz in east Germany, massive
corestones and less weathered quartz-rich veins occur side by side
with thoroughly disintegrated granite.
Nigeria, who showed that the thickness of the weathering

mantle varied from a few to more than 50 m over distances of
a few hundred meters. If local relief is higher, the variability
may be even more pronounced. An extensive database
from Hong Kong (Shaw, 1997), based on nearly 5000
boreholes covering an area of 22 km2, indicates that thickness
of weathered mantle over this granite and volcanic terrain
may exceed 90 m, but drops to a few meters only nearby,
with 2040 m being the most common values. Commonly,
pockets of deeply weathered rock occur within a mantle of
much smaller thickness, suggesting local lithological and/or
structural influence. In weathered serpentinite massif in
Szklary, SW Poland, the average thickness of saprolite is around
1020 m, but a few pockets of heavily decomposed rock reach
the depth exceeding 40 m, and locally as much as 70 m
(Figure 5).
Many attempts have been made to classify weathering
mantles and to establish their terminology. This can be done
using different criteria. One common approach is to use the
dominant grain-size of the weathering product as a criterion.
Consequently, weathering mantles can be arenaceous and
dominated by coarse fraction of sand (0.12 mm) and gravel
(42 mm), or argillaceous, with significant amount of silt and
clay. Since clays in rocks are mainly products of weathering,
brought about by mineralogical change, the amount and nature of secondary clays are general indicators of how advanced
weathering is. Hence, arenaceous mantles would be indicative
for rather early stages or rock weathering, whereas clay-rich
residual materials are expected if the history of weathering was
long and/or the intensity of weathering processes was high.
For weathered coarse crystalline rocks, particularly granite and
gneiss, the term grus (or gruss) is widely used to indicate the
granular nature of the saprolite. In rocks weathered to grus the
amount of mineralogical change is very limited, but disaggregation is fairly thorough and breakdown by hand can be
easily achieved (Migon and Thomas, 2002; Migon, 2006). In
many granite terrains, deep weathering is almost exclusively of
this type and the resultant profiles may be as much as 20 m
thick and more (Figure 6). One of advantages of the grain-size
based approach is that it allows for quick classification at the
stage of field work.
50 39
Figure 5 Variable thickness of deeply weathered serpentinites at

Szklary in southwest Poland. Modified from Niskiewicz, J., 2000.
Pokrywa zwietrzelinowa masywu Szklar i jej niklonosnosc [Engl.
summ.: The Szklary massif nickel-bearing weathering cover].
Geologia Sudetica 33(2), 107130.
A very different approach uses geochemistry of weathered

rock, allowing one to infer processes contributing to the origin
of the saprolite (Tardy, 1971). These processes are:
Allitisation, in which all basic cations and most of the silica

is removed. Hence the product is dominated by Fe and Al
hydroxides. The proportion of SiO2 to Al2O3 is below 2.
Monosiallitisation, in which silica is partially removed and
kaolinite is formed (SiO2:Al2O3 2).
Bisiallitisation occurs when most of the silica remains in
weathering products, 2:1 clay minerals originate and some
of the cations are retained. The proportion SiO2:Al2O3 is
above 2.
Corresponding to this distinction is the traditional division

of weathered mantles into kaolins (kaolinitic clays), formed
by monosiallitisation, and laterites, which are products of
allitisation. The latter in particular is very ambiguous and
various meanings of the word laterite can be found in literature. To avoid confusion, Taylor and Eggleton (2001) recommend the use of ferricrete to describe indurated, hard,
iron-rich upper zones of weathering profiles (see also
131
Figure 6 Thick granite-derived grus, rather uniformly weathered throughout, in the Karkonosze Mountains, Czech Republic.
Bourman and Ollier, 2002), but to speak about lateritic weathering profiles if a weathered mantle could be divided into an
uppermost ferruginous crust, followed in a downward direction by a mottled zone, pallid zone, and finally unweathered
bedrock. For some researchers this would be an ideal, complete weathering profile that has attained maturity.
Not every type of rock is likely to support a thick weathering mantle. Leaving surface erosion and its varying intensity through time aside, it is important to consider the
cohesion and strength of weathered material. The widespread
occurrence of thick saprolites in granite areas is in a few cases
taken as an indicator of particular susceptibility of granite to
weathering. This is not necessarily a correct interpretation
before we know residence times of saprolites. In fact, clayey
grus derived from granite can be fairly durable, with neoformed clays binding less weathered quartz and potassium
feldspar crystals. By contrast, some common rock types break
down into loose sand and silt that can be removed easily,
whether by overland flow, through piping, or by wind. Quartz
sandstones are a good example of a rock type that is hardly
associated with deep weathering profiles, even in humid tropics. At the extreme end of the scale are lithologies susceptible
to congruent dissolution such as limestone or gypsum. Insoluble residuals account for so little of the parent rock that
thin soils rather than thicker weathering mantles form
(see Chapters 4.3 and 4.15). Therefore, most deep weathering
profiles are known from basement areas, particularly from
granite, gneiss, schist, and serpentinite regions. In sedimentary
rocks, chances for the formation of a thick mantle increase if a
protective duricrust cap occurs at the surface.
For long, deep weathering tended to be associated with
conditions of low relief and visualized as a phenomenon accompanying advanced planation. In the same vein, it was
thought that thick saprolites in dissected terrains were inherited from a period preceding the onset of more vigorous
erosion. Such an interpretation cannot be sustained any
longer. Exploration and development of many hilly terrains,
especially in low latitudes (Brazil, Caribbean region, equatorial Africa, the Ghats, south-east Asia, tropical Pacific islands), has shown clearly that slopes are ubiquitously
underlain by a mantle of weathered rock, within which
unweathered compartments may or may not occur (Figure 7).
In specific locations the thickness of weathered rock in such
topographic setting may reach 5080 m (e.g., south-east Brazil). It was also observed that these weathered mantles are
commonly at a rather early or moderate stage of decomposition, being best described as arenaceous. Thomas
(1994) paid significant attention to the coexistence of
thick saprolites and dissected terrains and argued that this
combination may be reasonably expected if the following
characteristics are met: abundant rainfall, fractured bedrock,
and stepped relief with steep hydraulic gradient. Granitederived grus in particular, known from many mountainous
settings around the world, is proposed to represent a specific
response of the weathering system to the combination of
high relief, sufficient water availability, appropriate lithology
and structure, and efficacy of sediment transfer across
the topographic surface (Migon and Thomas, 2002). The
latter, by means of landsliding, gully erosion, soil creep, and
overland flow, truncates the top horizons of the saprolite
and inhibits more thorough geochemical and mineralogical
changes.
4.8.3
Deep Weathering Through Geological Time
Thick weathered mantles are widely present in the contemporary continental landscapes, except in the late Cenozoic
high mountain chains. The major role of deep weathering is
also evident in the geological record, in the form of either
relict saprolites, truncated by the current topographic surface
or buried under younger sediments, or detrital deposits whose
mineralogical characteristics point to deeply weathered lands
as source areas.
132
Figure 7 Deep weathering is not necessarily associated with tropical plains. This example comes from the western slope of Sierra Nevada in
California, in Kings Canyon National Park, and occurs within a slope which is more than 201 steep.
To fully explore geomorphological implications of deep

weathering and to use this record to constrain scenarios of
long-term landform evolution, ages of weathering should be
known. It is not easy to retrieve information. For a long time,
the only dating tool available was relation of saprolites to
geological materials (sediments and volcanics), whose ages
could have been independently established. Such an approach
worked well for weathered inter-basaltic horizons, but remained disappointingly inconclusive for most weathering
profiles on ancient basement rocks where saprolites occur at
the surface, or beneath very recent deposits. In addition, if
overlying strata are thin and permeable and the site is freely
drained, water may still have access to the saprolite below and
cause further weathering. Schmidt and Ollier (1988) have
described instances of Tertiary weathering mantles below Jurassic lava flows in Australia and documented such reverse ages
by palaeomagnetism, whereas Pavich and Obermeier (1985)
argued for late Cenozoic subsurface weathering under Cretaceous and Miocene strata in the Coastal Plain in eastern
United States. Locally, a morphostratigraphic approach can be
used. For example, Lidmar-Bergstrom (1995) examined relationships between landforms and cover rocks in southern
Sweden and identified distinctive landscape types associated
with specific saprolites that have evolved at different times
within the Phanerozoic. These associations have then been
used to infer ages of saprolites in other, not too distant areas.
Mineralogical characteristics and the nature of secondary
clays were routinely used as a proxy, but there were always
many uncertainties involved. The method is based on a sequence of assumptions, of which the critical one holds that
some secondary minerals, or their associations, are diagnostic
for certain types of climatic environments. The next step involves the identification of a period (or periods) within the
geological record, for which such climatic conditions have
been independently established. Major advances in this approach are associated with the French school of clay mineral
analysis in respect to climate (Tardy et al., 1973; Righi and
Meunier, 1995). However, the role of other factors such as
parent rock composition, site factors, including local geomorphology and hydrology, stage of weathering, and duration
of weathering has to be carefully considered as they may all
upset the suggested general relationships.
More recently, several novel methods and techniques of
dating ancient weathering products have been developed
(see review in Vasconcelos et al. (2008)). These can be divided
into indirect and direct dating. The former include the measurement of chemical remanent magnetization and oxygenisotope composition of pedogenic minerals, mainly kaolinite,
the latter uses KAr and 40Ar/39Ar methods to date several
neoformed manganese oxides (cryptomellane or hollandite)
and sulphates (alunite and jarosite). In the palaeomagnetic
method time is recorded when minerals formed through
weathering, usually hematite, acquire a stable chemical remanent magnetization. However, how it relates to the whole
history of weathering is imperfectly known. In ferruginized
profiles, most suitable for palaeomagnetic dating, the dates are
considered to represent the terminal lateritization, which is
not necessarily the main period of weathering. In the oxygen
isotope dating it is assumed that D/H and 18O/16O ratios in
minerals formed during weathering reflect the isotopic composition of groundwater in the presence of which they originated and the temperature of formation. The stable isotope
ratios in groundwater in turn reflect the average isotope ratio
of rainfall, itself dependent on the mean annual temperature
(Bird and Chivas, 1988). Hence, through the analysis of
stable isotope geochemistry of weathering products, one can
establish the likely temperature ranges of weathering environments and correlate them with the climatic history of a
given area. Using this method, Bird and Chivas (1988) were
able to identify different generations of residual kaolinitic
clays in Australia, spanning the period from the Permian to
late Cenozoic, whereas Gilg (2000) recognized two generations of kaolinitic weathered mantles in southern Germany,
one mid-Cretaceous and the other one mid-Cenozoic.
Indirect methods are useful in broadly constraining the age
of weathering profiles, but their resolution is rather low.
Vasconcelos et al. (2008) maintained that the required degree
of precision and accuracy can only be achieved through
radiogenic-isotope techniques. These are so far applied to
potassium-bearing minerals, and the 40Ar/39Ar method is
considered most promising as it enables dating single crystals,
overcoming difficulties in interpreting bulk samples with primary and weathering minerals intermixed. A good review of
the technique and its potential to answer geomorphological
questions has been provided by Vasconcelos (1999). Insofar,
direct dating of weathering products was attempted mainly in
Australia, South America and less commonly in Europe, but
overall, examples of this approach are few.
Applications of these different approaches leave no doubt
that deep weathering and formation of thick and laterally
extensive weathered mantles have been present through time
and space in the history of the Earth. Some old saprolites may
have been sealed by later deposition, but in many instances
products of palaeoweathering underlie the present-day topographic surface and bear directly on the contemporary geomorphic evolution. They are also important as sources of
information about the history of landscape evolution and how
it was accomplished.
Inevitably, the record of very ancient weathering is scarce,
although a generation of Permian kaolinitic saprolites could
have been identified in eastern Australia using oxygen isotopic
signatures preserved in kaolinites (Bird and Chivas, 1988).
However, Mesozoic saprolites are common and many deeply
weathered landmasses were sealed by Cretaceous marine
sediments formed at the time of late Cretaceous high sea-level
stands. With respect to western, central and northern Europe a
comprehensive review was provided by Migon and LidmarBergstrom (2001) and is summarized in Figure 8. PreCretaceous clayey saprolites occur in the Ardennes and the
Rhenish Slate Mountains (Belgium, Germany), in the Bohemian Massif (Czech Republic, Germany), in southern and
central Sweden (Figure 9), Finland, and Norway. At many
localities, the thickness of these saprolites exceeds 50 m. It is
likely that the onset of deep weathering and China Clay formation in Dartmoor, SW England, took place in the Mesozoic
too (Bristow, 1998). The record of deep weathering continues
in the Cenozoic. Precise dating is often not possible and it is
the overlying sediments used to constrain the minimum age of
the saprolites. Thus, in the Bohemian Massif thick kaolinitic
mantles derived from granite, gneiss and mica schist occur
beneath latest Oligocene/early Miocene clays, silts, and sands
in the down-faulted parts of the basement terrain. In the
French Massif Central, Simon-Coinc- on et al. (1997) documented several generations of deeply weathered palaeosurfaces, partially buried, which provide a record of deep
weathering covering most of the Mesozoic and the Cenozoic.
Many deep weathering profiles have been fossilized by products of volcanic eruptions which took place since the
133
Oligocene until the Quaternary. These include argillaceous

kaolinitic profiles of likely Palaeogene age, but also arenaceous mantles (grus) which are probably younger. In the
Massif Central granite-derived sandy saprolites occur beneath
lava flows, whose formation began 16 Ma ago, indicating that
deep disintegration has been present in the area since at least
the Early Miocene. Miocene lava flows are deeply weathered
and in specific localities, such as Vogelsberg in central
Germany, ferrallitic mantles up to 50 m have formed. Grus
saprolites are ubiquitous in basement areas throughout
Europe, from Ireland and Scotland through the Iberian Peninsula, southern Italy, central Europe to Fennoscandia (Migon
and Lidmar-Bergstrom, 2001; Migon and Thomas, 2002), although their ages generally remain unconstrained. It is important to note that grus mantles occur in different geomorphic
settings, from flat floors of intramontane basins and planar
upland surfaces to inselberg footslopes, moderately sloping
terrains (15251), and even shoulders of deeply incised valleys.
Such patterns of spatial distribution imply steady renewal of
grus profiles, with simultaneous truncation of the near-surface
horizons by hillslope processes and profile deepening at the
saprolite/bedrock interface. It was argued that the widespread
presence of grus and the absence of clay-rich profiles of evidently young (end-Cenozoic) age signifies a major change in
the style of weathering, induced by growing climatic and/or
tectonic instability (Migon and Thomas, 2002).
Long records of deep weathering, spanning the Mesozoic and
Cenozoic, are by no means confined to Europe. Bird and Chivas
(1988) in their oxygen isotope study of Australian saprolites
identified Cretaceous, Palaeogene, and Neogene generations, and
the antiquity of many Australian saprolites is now universally
acknowledged (Vasconcelos et al., 2008). Likewise, many instances of mid-Cenozoic saprolites have been recognized in ancient shields of South America through palaeomagnetic dating of
lateritic profiles (Theveniaut and Freyssinet, 2002), whereas
younger, late Miocene ages have been obtained by direct 40Ar/39Ar
dating for profiles in the Atlantic seabord of Brazil (Carmo and
Vasconcelos, 2006). In the Indian Peninsula, thick lateritic and
bauxitic weathering profiles developed upon latest Cretaceous/
Paleocene basalts but likely before late Eocene/Oligocene, indicating that the early Cenozoic was a period extremely conducive
to deep bedrock alteration (Widdowson and Gunnell, 1999). In
turn, Mesozoic deep weathering of crystalline basement may
be inferred from examination of sedimentary record offshore
(Gunnell, 1998). However, no radiometric dating is currently
available for Indian weathered mantles. A limited amount of data
for Africa suggest various Cenozoic ages for lateritic profiles
(Henocque et al., 1998), whereas in interior North America thick
(430 m) saprolites are known to occur beneath Cretaceous
sediments in Minnesota (Parham, 1970). The record of deep
weathering from the Appalachians and the Piedmont is younger
and considered Neogene to Quaternary in age (Pavich, 1989).
4.8.4
Etching and Stripping
Although the occurrence of thick weathered mantles struck

early explorers of the tropics from the turn of the twentieth
century, as attested by their writings (e.g., Branner (1896) in
Brazil or Falconer (1911) in Nigeria), it took a while until
134
Region
Triassic
Jurassic
Cretaceous
Paleogene
Neogene
Quaternary
British Isles
Northern Ireland
Ireland
Scotland
South Wales
SW England
Western Europe
Brittany
Massif Central
Ardennes
Rhenish Slate
Mountains
Harz
Central Europe
NW Bohemian Massif
Erzgebirge
Sudetes
SE Bohemian Massif
South Polish Uplands
Fennoscandia
Southern Sweden
North and central
Sweden
Bornholm
Norway
Finland
?
?
Kaolinitic weathering
Ferrallitic ( lateritic)
weathering
?
Sandy and grus weathering
(typically on coarse crystalline rocks)
Terra rosa and other limestone
residuals
Figure 8 Deep weathering and saprolites through time in western, northern and central Europe. Ages of most European saprolites are only
broadly constrained by overlying sediments or relationships to regional landforms, which is reflected by the length of arrows. In reality, the
relicts observed today may have formed over much shorter timescales. Modified from Migon, P., Lidmar-Bergstrom, K., 2001. Weathering
mantles and their significance for geomorphological evolution of central and northern Europe since the Mesozoic. Earth Science Reviews 56,
285324.
geomorphologists realized the significance of deep weathering

in long-term landform evolution and incorporated this process
into general conceptual models. Neither the geographical cycle
and peneplanation by Davis (1899), pediplanation concept by
Penck (1924), nor even Canons of landscape evolution by
King (1953) put any greater emphasis to the fact that rocks may
weather in situ to considerable depth and over long time before

erosion of the weathered material takes place. The concept of
etched plains developed by Wayland (1933) and Willis (1936)
in east Africa, as well as various miscellaneous observations
from south-east Asia and Australia (see Twidale, 2002) escaped
the attention of mainstream European and North American
135
Figure 9 Kaolinized granites at Ivon, south Sweden, have been revealed after overlying Cretaceous limestones were quarried away and are a
type locality of Mesozoic deep weathering in Europe. Note the very uneven weathering front. Thoroughly decomposed granite (left) passes along
a sharp subvertical boundary (a gully in the middle of the picture) into an unweathered massive dome (right). Here, complete stripping of the
weathering mantle reveals an inselberg-like hill.
geomorphology and it was the famous German geomorphologist Budel (1957) who secured the position of deep weathering among important processes contributing to landscape
evolution. As an outcome of later developments, both in terms
of concept and expanding field evidence, the etching and
stripping model of landscape evolution has become widely
recognized and is considered particularly applicable to areas of
low to moderate relief in at least seasonally humid low to midlatitudes (Thomas, 1994). Today, it is the only general model in
geomorphology that explicitly links weathering and landform
evolution over long timescales and across wide spatial scales.
However, the etching concept remains so far a largely conceptual model which is yet to be explored by numerical modeling of landscape development. The key problem here is that
the assumption of steady-state soil (saprolite) thickness, which
underlies many theoretical and mathematical models of slope
evolution, is invalid for deeply weathered lands, with inherited
saprolite cover (Phillips, 2010).
The word etching itself means corroding a surface by aggressive reagents. In geomorphology it is used to describe
progressive rock decomposition that occurs within deep weathering profiles, at the weathering front. In particular, it is
applied to situations where rocks differ in their resistance to
chemical decay and consequently thickness of a weathering
mantle becomes highly variable over short distances. Removal
of products of deep weathering generally referred to as
stripping will expose the bedrock surface, the topography of
which is the direct result of differential etching, thus it is an
etched surface. At the early stages of the development of
geomorphology science, until late 1950s, when focus on planation surfaces and peneplains was preeminent, etched surfaces were visualized as surfaces of low relief and thought of as
a special category of a peneplain, produced by subsurface rock
decay followed by stripping of weathered mantle. Accordingly,

the process of producing an etchplain through weathering
and stripping has later become known as etchplanation.
However, it was observed that a stripped surface, which lost its
weathered mantle, is rarely a plain. Instead, it tends to show
some relief, which reflects differential rock control on the
progress of etching. In granite areas in particular, stripped
surfaces are typified by inselbergs, domes, tors, basins, and
boulder piles, and to call them etchplains would be both
inappropriate and misleading. Therefore the term etchsurface
has been recommended for use wherever evacuation of weathering mantles reveals a varied topography.
Initially, etchplains were thought of as developing in two
major phases of unequal duration. A protracted period of deep
weathering was followed by a rather brief episode of stripping
weathering products. The reasons for change from weathering
to stripping regime are examined later, now implications for
landscapes will be considered. Stripping may have been
complete, down to weathering front and exposing bedrock, or
incomplete, restricted to the upper parts of the weathered
mantle or to certain locations in the landscape, for example,
the vicinity of major drainage lines. To account for the different proportions of original weathered mantle left in a
landscape and to name various field situations the following
classification was proposed (Thomas, 1994) (Figure 10):
Mantled etchplain weathering mantle is ubiquitous and

virtually no bedrock is exposed. Weathering progressively
attacks solid rock at the base of the mantle, shaping the
etched surface which is to be exposed later, but the mantle
can also be relict.
Partly stripped etchplain develops from mantled etchplain through selective removal of the weathering mantle
136
Mantled etchplain
Duricrust
Weathering mantle
Partly stripped etchplain
Stripped etchplain
Complex etchplain
Re-weathered
etchplain
Pedimented
etchplain
Pediments
Incised
etchplain
Buried etchplain
Younger sedimentary cover
Lava flow
Figure 10 Etchplain typology and evolution.
and exposure of bedrock surface, but part of the original

saprolite remains. The proportion of areas still underlain
by saprolite may vary from 10 to almost 100%.
Stripped etchplain most of the bedrock is exposed
from beneath a weathering mantle and only isolated patches of saprolite are left (o10% of the area). These characteristics conform with the original definition by Wayland
(1933).
Complex etchplain includes a few variants, in which
deeply incised valleys may be present (incised etchplain),
or removal of saprolite is accomplished by pedimentation
(pedimented etchplain), or a new generation of weathered
mantles begins to form (re-weathered etchplain).
Buried etchplain one which has been covered by younger
sediments or lava flows.
Exhumed etchplain one which has been reexposed after
burial.
This or similar classification schemes were used by various

workers to map the extent of etchsurfaces, both in low latitudes (Nigeria Thomas, 1969; Yilgarn block in southwestern
Australia Fairbridge and Finkl, 1980) and outside the tropics
(Finland Soderman, 1985), where deep weathering was
considered largely relict. Since a sequential development of
etchsurfaces is implied in the classification, their spatial distribution provides insights into spatial patterns of long-term
erosion and hence broader tectonic and climatic controls on
landscape development. In the formerly glaciated lands, it
proved useful in constraining models of varying efficacy of
glacial erosion.
The reasons of changing balance between weathering and
surface erosion, with the latter taking over the production of
new saprolite, may vary. One obvious control is tectonic. An
increase in thickness of the weathered mantle will be favored if
little erosion takes place, hence under conditions of low relief
and little change in base level. These should typify protracted
periods of crustal stability, particularly in continental interiors.
Therefore, etching was commonly associated with cratonic
settings and envisaged as a means through which morphology
of shields is shaped, for instance in Fennoscandia (LidmarBergstrom et al., 1999). Increase in available relief and shift to
stripping regime may be caused by different tectonic processes
such as thermally-driven domal uplift typical for Africa
(Summerfield, 1996), rifting and creation of escarpments, or
isostatic adjustment to lithospheric loading and unloading in
adjacent localities. Another plausible control is regional climatic environment and the extent to which climatic parameters are suitable for deep weathering to occur. Budel
(1957) was the champion of environmental change as a reason for enhanced stripping, and two principal scenarios emphasize transitions from humid to arid conditions and from
warm to cooler climates, with associated reduction in water
availability and temperature, respectively. In both cases, these
should cause the rates of weathering to decline. A shift toward
aridity in particular, resulting in ground exposure due to
vegetation change, would likely upset the balance and set the
conditions for vigorous erosion of weathered materials. Of
course, regional climate change may itself result from changing tectonic setting, for instance movement of a lithospheric
plate from low to high latitudes or vice versa, or appearance of
a mountain range at a convergent plate boundary. For example, the granite landscape of the Joshua Tree National Park
in the Mojave Desert, California, is an etchsurface stripped of
its weathered mantle. Uplift of the near-coast San Bernardino
Mountains turned the hinterland into a desert, limiting further
deep weathering and accelerating erosion (Oberlander, 1972).
Thomas (1994), in turn, drew attention to major environmental changes during the Quaternary alone, related to glacial/interglacial cycles and demonstrated their importance on
the example of the Koidu etchplain in Sierra Leone (Thomas
and Thorp, 1985).
The Koidu study, followed by many others in different areas
in the tropical and subtropical belt, has shown that the simple
bipartite division of timescales of etchplain formation into a
period of etching and a period of stripping does not portray the
reality adequately. Therefore, a concept of dynamic etchplanation was introduced to emphasize continuing landscape development through simultaneous etching and stripping. Key points
made are weathering and removal of its products going side by

side, lowering of both interfluves and valley floors, continuous
sediment transfer, redistribution and temporal storage of weathering products within drainage basins, and the importance of
minor environmental disturbances over short timescales
(104105 years). An important part played by episodic high-energy events is emphasized in the parallel term of episodic
etchplanation (Thomas and Thorp, 1985).
The greatest field of uncertainty in the etchplanation model
concerns relevant timescales which are characteristically
speculated about rather than demonstrated beyond doubt.
Fairbridge and Finkl (1980), working in cratonic settings,
suggested 107109 years necessary for the formation of a
weathered mantle and 105107 years for removal of saprolites.
This latter time span seems consistent with the situation in the
Mojave Desert, where stripping of pre-late Miocene saprolites
has been largely completed after etching was suppressed some
9 Ma ago (Oberlander, 1972). However, in areas of higher
relief such as certain passive margins (south-east Brazil, the
Ghats) or regions of domal uplift, where groundwater circulation is efficient because of higher hydraulic gradient, the
corresponding timescales may be shorter. In basement terrains, deep disintegration into grus is typical and it seems that
a grus mantle 520 m thick may form in just one million years
(Pavich, 1989), hence one to three orders of magnitude faster
than proposed by Fairbridge and Finkl (1980). There is little
scope for weathering profiles to advance beyond the grus/
are`ne stage because of surface denudation and progressive
truncation of the upper zones of weathering. Thus, etching
and stripping rates may be in long-term equilibrium, although
the former is continuous while the latter tends to be episodic.
In addition, it is important to distinguish between periods
necessary for weathered mantles to form a continuous blanket
over a landscape, to create a mantled etchplain, and their
subsequent residence times. The former may be much shorter
than the latter. For instance, in the early Paleocene volcanic
province of Antrim (Ireland) 15 m thick ferrallitic weathering
profiles have formed at the expense of basaltic lava flows in
less than one million years (Hill et al., 2000). Thomas (1994,
2006) reviewed some available data about rates of weathering
and concluded that in tropical environments they may be as
high as 2050 mm ka1. This means that in favorable circumstances, weathered mantles ca. 50 m thick may have
formed during the Quaternary alone. Yet they may subsequently persist for millions of years, especially if protected
from the top by duricrust, not uncommon in semi-humid low
latitudes. In the Mt. Isa region of Queensland, Australia,
ferricretes cap relict weathering profiles which acquired their
principal characteristics in early and middle Cenozoic
(Vasconcelos, 1998). Subsequent stripping has been accomplished mainly by scarp retreat and reduction of a primary
duricrusted plateau, gradually exposing lower surfaces and
initiating new cycles of weathering.
4.8.5
Geomorphological Signatures of Etchsurfaces
Landscape evolution through etching and stripping produces a

range of distinctive landforms. They are initiated at the weathering front reflecting structure-controlled selectivity of
137
etching, molded largely in the subsurface, and later exposed

after stripping takes place, undergoing further modifications in
subaerial conditions. They vary in shape, from upstanding to
concave, and in size. Minor landforms exposed from weathered mantles are discussed elsewhere in this volume
(see Chapter 4.10). Here the emphasis is on the larger landforms and their assemblages. Inselbergs, multi-convex relief,
topographic basins and plains are examined.
4.8.5.1
Inselbergs
Inselberg is a German term, which literally means an island

hill, but has found its way into English terminology in unchanged form. It was coined by a German naturalist and traveler from the turn of the twentieth century, Walter Bornhardt,
who was impressed by isolated rock-built hills rising steeply,
as if islands from the sea, from the savanna plains of East
Africa. There have been many attempts to define inselbergs.
Young (1972: p. 205) simply stated that they are steep-sided
isolated hills rising relatively abruptly above gently sloping
ground which sounds straightforward but nevertheless creates two problems which perplexed researchers, especially.
First, how steep ought inselbergs to be and how to differentiate them from gentle topographic rises. Second, how to define isolation which also bears on any attempts to draw
topographic limits of an inselberg. These considerations may
seem a futile academic exercise, but are important if quantitative studies are attempted. In one such study, the values of a
minimal distance of 0.8 km to the nearest neighbor, a minimum height of 15 m, and 25% as the minimum angle of the
steepest slope were adopted, but these were entirely arbitrary,
guided by the practical purposes of use of existing topographic
maps to identify inselbergs (Faniran, 1972). The height lower
limit should assist in differentiating inselbergs from smaller
residuals tors (see Chapter 4.10), but one can also argue that
they differ in landscape setting. Namely, a tor is a part of a
slope, whereas an inselberg possesses slopes itself. Thomas
(1967) noted great variability of form around a single, rather
low granite inselberg near Oyo in Nigeria and concluded that
advancing more restrictive criteria for inselbergs is not
realistic.
Inselbergs are not restricted to any single lithology, although they are particularly common in granites (Figure 11).
Ancient shields of Africa, South America and Australia, where
granites are abundant, host the most characteristic inselberg
landscapes (Thomas, 1978; Twidale, 1982; Twidale and VidalRomani, 2005; Migon, 2006). However, other lithologies may
support inselbergs as well, including gneiss, gabbro, quartzite,
conglomerate, and arkosic sandstone. The latter builds perhaps the most famous inselberg on Earth, Uluru in central
Australia. Inselbergs vary greatly in size and shape. In granite
areas, Twidale (1982) identified three principal morphological
types, namely domes, nubbins (or block- and boulder-strewn
inselbergs) and castle koppies (castellated inselbergs), but the
distinction between them in terms of visual appearance is not
always easy. Two parallel characteristics appear important,
namely the shape of constituent compartments and their
isolation from the rock core (Figure 12). Boulder inselbergs
are apparently chaotic heaps of detached rock compartments,
138
Figure 11 A granite inselberg on the Mojave Desert, California, exposed through stripping of the pre-late Miocene weathering mantle.
(a)
(b)
(c)
(d)
Figure 12 Different morphological types of granite inselbergs in the Mojave Desert, California. (a) Dome, (b) castellated inselberg, (c) boulder
inselberg (nubbin), and (d) individual corestones and their clusters excavated from a grus weathering mantle.
which generally mantle a more massive, bedrock-rooted core

or rest on a low-angle rock platform. Individual components
of a boulder inselberg can be either angular or rounded, or
these two can coexist. They measure from 12 m to more than
10 m across. Castellated inselbergs tend to have an angular

outline and are typically composed of a massive lower part,
which grades into an array of pillars, walls and clefts in the
upper part, occasionally termed as ruiniform relief. Visually,
they commonly show a stepped appearance, with vertical faces

separated by subhorizontal benches. The former relate to
vertical fractures, the latter to the horizontal ones. Typical
height of castellated inselbergs is of the order of tens of meters.
The most impressive inselbergs, however, are domes. Their
characteristic features include steep, bare and upward-convex
slopes, a sharp piedmont angle, and a mantle of talus derived
from joint-controlled degradation around at least a part of its
perimeter. The height of domed inselbergs is very variable.
Some domes are as low as 2030 m, others may well exceed
300400 m high, whereas the tallest examples such as Spitzkoppe in the Namib Desert achieve 600 m. However, low
domes 1530 m high, with slope angles as low as 10151 and
flat summit surfaces, may commonly rise above the plains.
The reasons for morphological variability of inselbergs relate
to different structural controls. Boulder inselbergs develop in
rather irregularly jointed rock masses, whereas castellated inselbergs form in orthogonally fractured bedrock, where partings are relatively few in number. Domes in turn are associated
with the most massive compartments, where the distance between fractures may be measured in tens of meters. Curved
sheeting surfaces may be present, influencing the rounded
appearance of an inselberg. However, the origin of sheeting is
still insufficiently known and competing explanations have
been offered in literature (Hencher et al., 2010). The relationships between the different morphological types of inselbergs are more extensively discussed in Migon (2006).
The origin of inselbergs has been a matter of long-standing,
often passionate debate. Two main hypotheses have been
proposed. One holds that inselbergs are specific by-products
of long-term scarp retreat and pediplanation, and is usually
presented in conjunction with the name of Lester King and his
Canons of landscape evolution (1953). According to the alternative model, A plane surface of granite and gneiss subjected to long-continued weathering at base level would be
decomposed to unequal depths, mainly according to the
composition and texture of the various rocks. When elevation
and erosion ensues, the weathered crust would be removed,
and an irregular surface would be produced from which the
most resistant rocks would project. (y) In this way would
arise the characteristic domes and turtlebacks which suffer
further denudation only through insolation and exfoliation
(Falconer, 1911: p. 246). This scenario, with its emphasis on
two separate stages in the evolution of inselberg landscapes,
was brought back to attention in the mid-1960s (Ollier, 1965;
Thomas, 1965) and linked with the etching/stripping model
of landscape development (Figure 13). Critics, including
Lester King, pointed out discrepancies between the height of
inselbergs and the known thicknesses of weathered mantles
from which the inselbergs would have been excavated. The
former, it was argued, are commonly much higher than the
depth to weathering front, rendering the two-stage model
unfeasible. To account for this situation, a model of episodic
exposure of inselbergs was introduced (Twidale and Bourne,
1975). In essence, the excessively high hills would have risen
through many repeated stages of deep weathering and stripping. Hence their height would have increased through time,
and ultimately surpassed the typical thickness of a weathering
mantle, with the inselbergs being multistage rather than twostage landforms. Solid evidence for the validity of the two
139
(multi)stage model was provided by discoveries of nascent

domes in weathered profiles (Figure 9), and further confirmation comes from geophysical studies (Beauvais et al., 2003).
Today, the model of inselberg development through etching
and stripping enjoys widespread support although it would be
wrong to dismiss the other scenarios completely. One such
area, where alternative concepts have been debated, is the
Namib Desert. Its central part is dotted by inselbergs of variable height, but the evidence for deep weathering is missing
and it is argued that the Namib was never more humid than
semidry during the last 80 million years or so. Selby (1982)
presented evidence that the inselbergs of the Namib owe their
form and position to lithology and structure and are perfectly
explainable in terms of ongoing differential denudation,
without the necessity to recourse to past deep weathering. In
some cases, the domical form was produced at the time of
intrusion and nowadays the forms are progressively exposed as
the overlying schist is wasted away. Other inselbergs, such as
Gross Spitzkoppe, Mirabib or Amichab have their form
strongly adjusted to the fracture patterns. Thus, although there
are good reasons to consider inselbergs as landforms produced by etching and stripping and geomorphic signatures of
efficient subsurface weathering, they remain examples of
geomorphological equifinality. The recognition of the history
of inselbergs in each specific area, and whether they emerged
from a weathering mantle, needs to be based on both an analysis of the residual form itself and wider consideration of
residual deposits, their sedimentary equivalents and the history of environmental change in the relevant period of
geological past.
4.8.5.2
Multiconvex Relief
Multiconvex landscapes have been introduced to literature

under this name by Thomas (1974) who attempted a morphographic classification of granite relief. Its major components are rather closely spaced and irregularly distributed
hilly compartments, typically a few hundred meters in diameter, separated by narrow valleys and channel-less elongated
depressions, with little intervening flat land. Individual hills
tend to be oval in plan and convex in shape, hence the name
for this type of relief. The relative relief within multi-convex
relief is typically small, rarely exceeding 100 m, with typical
slope inclination of 20251. Although a certain proportion of
hills may show an exposed rock core, the majority are weathered to a great depth, or even throughout (Figure 14).
Multiconvex landscapes are distributed across the intertropical
belt, with examples having been described from coastal and
inland Brazil, Guyana, equatorial Africa, Madagascar, Indian
peninsula, and south-east Asia (review in Migon, 2009), but
hardly from outside the tropics.
As the thorough decomposition of many convex compartments indicates, deep weathering is an important component of this type of relief. Multiconvex landscapes may be
thus considered as a transient landscape on the pathway from
a deeply weathered plain to a stripped etchsurface. They are
very dynamic, shaped by a variety of surface processes, including landsliding, gullying, and soil creep (Coelho Netto,
1999), which remove or redistribute products of weathering
140
Differential deep weathering of bedrock

of non-uniform resistance
Deep weathering continues
Two-phase evolution
Multiphase evolution
Complete stripping
Episodic stripping and exposure
Inselberg landscape
Figure 13 Model of inselberg evolution through etching and stripping. Reasons for location of particular inselbergs may be lithological (right) or
structural (left).
and reduce the weathered hills toward more subdued, gently

rolling topography. However, the preferred occurrence of
multiconvex landscapes in humid tropical regions, where deep
weathering is most efficient, suggests that their evolution is
more complex and is sustainable.
Two sets of antagonistic processes, acting concurrently, can
be identified in multiconvex terrains. Deep weathering continues to break down the bedrock and to increase the thickness
of the saprolitic mantle, whereas surface processes episodically
remove the weathered material. Since landslides are very potent
agents of denudation and can easily remove the whole thickness of weathering mantle, deep weathering has to be particularly efficient to compensate material losses in relatively short
periods. It is required that weathering operates efficiently in all
topographic settings, and dissection proceeds to sustain relative
relief. Otherwise the multi-convex landscape will be reduced to
a rolling plain, with occasional shield inselbergs remnants of
the unweathered cores of the long-gone hills, and widely distributed sedimentary veneer from past landslides. Referring to
general concepts in geomorphology of low latitudes (Thomas,
1994), it seems that multi-convex landscapes may survive under
the dynamic/episodic etchplanation regime, whereas their

occurrence indicates long-term landscape lowering and appears
to be dependent on both environmental and geotectonic
factors. Warm and humid conditions favor deep weathering,
which is able to keep pace with landsliding, which is the decisive surface process for this peculiar type of morphology.
However, long-term regional base-level lowering is necessary to
maintain the relief, hence a long-term relative surface uplift is
indispensable. Broad topographic swells and plateaux in nonorogenic settings provide the most appropriate locations for the
multi-convex landscape to develop. If the rate of uplift is too
high, balance between saprolite production and saprolite removal is lost and stripping increasingly takes over. Thomas
(1997) suggests the uplift rate of 0.5 mm year as the threshold
between saprolitic and nonsaprolitic weathering. The above
requirements match the distribution of the majority of multiconvex weathered terrains that are present near topographic
margins of elevated plateaus (Meghalaya, India), rift shoulders
(Nyika, Malawi), broad swells related to intraplate uplift (Sierra
Leone), or great escarpments along passive margins (south-west
Deccan, south-east Brazil). One of the challenges is to further
141
Figure 14 Multi-convex relief developed in deeply weathered gneisses and granites on the inland side of Serra do Mar, southeast Brazil.
quantify the relationships between weathering and surface denudation in multiconvex terrain, as those established so far
concern dissimilar geomorphic settings.
4.8.5.3
Basins
The examples reviewed so far focused on the origin of convex

landforms and an emphasis was placed upon structural and
lithological prerequisites that decide about an elevated resistance against rock alteration. However, it is the topographic
basins which are etched landforms per se and regionally these
may be as important components of morphology as inselbergs
are elsewhere.
Basins may be of different sizes, from much less than
1 km2 to 10100 km2, and nested basins may occur. Likewise,
some basins may stand in isolation and these are drained via
narrow breaches in their rocky rims, whereas others are
interconnected and form spatially complex patterns. In the
tropics elongated channel-less alluviated depressions surrounded by weathered interfluves are widespread and these
are known under a variety of local names, such as dambo in
central Africa or bolis in Sierra Leone. Their origin is complex
and involves various processes of sediment production,
transport and storage, but deep weathering plays an important
part, etching out an initial depression and contributing to its
later extension (Thomas, 1994). The evolution of dambos is
episodic in that sediment stored within their floors is occasionally flushed out by major floods, allowing for further
etching and concurrent lowering of interfluves and dambo
floors. The presence of dambo basins is thus a signature of
dynamic/episodic etchplanation in action.
Basins are common outside the tropics as well, although
they rarely occur within deeply weathered hilly lands. More
commonly, they are the only locations where thick weathering
mantles occur, possibly covered by younger sediments. If this

is the case, they are interpreted as places of incomplete stripping because of less elevated position, favoring the survival
of old saprolites. European uplands, from NE Scotland
throughout the French Massif Central to the Bohemian Massif
offer numerous examples of topographic basins retaining
pre-Quaternary weathering residual deposits. Other uplands,
as well as certain formerly glaciated regions, abound in
basins that do not have any weathered mantle present
within their limits. One such area in the coastal strip of SW
Sweden around Goteborg, where a multitude of saprolite-free
basins of different sizes occur in gneiss and granite terrain
(Johansson et al., 2001). Their geomorphic characteristics,
including a lattice-like distribution pattern, exclude fluvial or
glacial processes as main agents of basin origin. Instead, differential etching focused on fracture lines, and regional fault
lines is invoked to be the main contributor and the evolutionary history is traced back to the Mesozoic. The largest
and deepest basins occur at intersections of major fracture
lines and follow two parallel fault zones. Fracture-aligned
basins grade into another specific type of relief, named a
joint-valley landscape (Lidmar-Bergstrom, 1995). Various
concave landforms, from linear fluvial valleys to narrow clefts
and gorges, overprinted on major fractures zones typify this
morphological variant, recognized both inside and outside the
tropics.
Rock-cut basins are generally considered as products
of selective etching that may exploit fractures or petrological
and mineralogical differences (Thorp, 1967; Bremer, 1975)
and their adjustment to structure should be best visible at
an early stage of stripping the saprolite. However, with
time, they may extend into adjacent higher ground due to
enhanced scarp-foot weathering and the primary structural
control becomes obscure. Further development of concave
landforms, especially in granitic terrains, may be facilitated
142
by the contrasting behavior of granite in the presence and

absence of moisture. Once the relief differentiation into upland and basin compartments takes place, the former become
areas of runoff and moisture deficit, and hence are more
durable.
4.8.5.4
Plains
Plains and their relicts, which once so much preoccupied the

minds of geomorphologists, do not generate such an interest
today. Particularly, the significance of level rock-cut terrains in
reconstructing denudation chronologies has been seriously
doubted. Certainly, one of the reasons for this waning enthusiasm for relict peneplains has been the recognition of the
morphological signature of etching and stripping. Although
initially a stripped surface was envisaged as an etched plain,
hence an area of little relief, it was subsequently shown that
this is hardly compatible with the general principles of weathering, which does not act uniformly over bedrock but picks
up less resistant zones. These may be related to specific mineralogical composition, denser fracturing, antecedent hydrothermal alteration, different porosities, and the like. Scenarios
of long-term landform evolution through etching and stripping emphasize the origin of varied relief (e.g., Kroonenberg
and Melitz, 1983) or even an increase in relief amplitude over
time (Twidale, 1991). It is now assumed that etched planar
surfaces form in exceptional circumstances, within uniform
and susceptible bedrock such as schist and possibly close to
base level.
4.8.6
Conclusions
The examples discussed above leave no doubt that the phenomenon of deep weathering is not only one of the most
important aspects of weathering in general, but also its implications for the evolution of landforms are profound. Although deeply weathered rocks were recognized as early as in
the ninetieth century, it took many decades until this fact has
become adequately accounted for in theory and practice of
geomorphology. Deep weathering occurs in many different
settings, from plains to moderately high mountains, with the
probable threshold between saprolitic and non-saprolitic
weathering being placed at the uplift rate of 0.51 mm per
year (Thomas, 1997). Obviously, there will be differences in
the characteristics of weathered mantles between plains and
mountains. Complete profiles, with clay-rich weathering
zones near the top and possibly duricrusts typify many rolling
plains, whereas weathering grades encountered within dissected terrains are generally low and very thorough decomposition is rare. Interestingly, however, thicknesses may be
similar and depths to weathering front may be up to 100 m in
either setting, which indicates that rates of descent of the
weathering front do not necessarily correspond with weathering grade within an existing saprolite.
Weathered mantles observed nowadays need to be interpreted in two categories: formation and preservation. It is clear
that weathering is a pervasive, continuous process and never
stops (Taylor and Shirtliff, 2003). Yet it is equally clear from
examination of available geological record that ages of relict
weathered mantles are not uniformly distributed over the

geological timescale (e.g., Thiry et al., 1999; Migon and
Lidmar-Bergstrom, 2002), even if we allow for an uncertainty
what is actually meant by saprolite age. This apparent conflict
arises from the fact that weathered mantles, once formed, have
had different chances to survive. Early burial by sediments or
lava sheets sealed the record of weathering for tens or even
hundreds millions of years, whereas an exposed position
made saprolites vulnerable to erosion if climatic conditions or
tectonic controls changed.
The varying long-term histories of weathered mantles and
their implications for geomorphology are addressed in the
etching and stripping model of landscape evolution. Among
the major theoretical advances in geomorphology, it is the
only one that explicitly takes into account processes of deep
weathering and successfully explains the origin and nature of
many palaeosurfaces that can be found across the globe, both
in the stratigraphic record and as elements of contemporary
topography. Intriguing landforms such as inselbergs, jointaligned drainage-free corridors, and intramontane basins are
in many instances features of an exposed weathering front,
hence etched landforms.
The significance of deep weathering and relict saprolites is
not restricted to historical geomorphology. A thought that
weathered mantles are helpful to reveal patterns of geomorphic
evolution through time but not quite relevant to the challenges
faced by applied geomorphology of today cannot be more
wrong. Leaving aside the huge economic importance of saprolites (see Taylor and Eggleton, 2001), properties of saprolites
acquired from deep weathering are the key controls of hillslope
evolution, including landslide occurrence. This subject is explored in Chapter 4.10 in this volume.
References
Beauvais, A., Ritz, M., Parisot, J.-C., Bantsimba, C., 2003. Testing etching
hypothesis for the shaping of granite dome structures beneath lateritic
weathering landsurfaces using ERT method. Earth Surface Processes and
Landforms 28, 10711080.
Bird, M.I., Chivas, A.R., 1988. Oxygen-isotope dating of the Australian regolith.
Nature 331, 513516.
Bourman, R.P., Ollier, C.D., 2002. A critique of the Schellmann definition and
classification of laterite. Catena 47, 117131.
Branner, J.C., 1896. Decomposition of rocks in Brazil. Geological Society of
Bremer, H., 1975. Intramontane Ebenen, Prozesse der Flachenbildung. Zeitschrift fur
Geomorphologie N.F., Supplement-Band 23, 2648.
Bristow, C.R., 1998. China clay. In: Selwood, E.B., Durrance, E.M., Bristow, C.M.
(Eds.), The Geology of Cornwall. University of Exeter Press, Exeter, pp.
167178.
Budel, J., 1957. Die Doppelten Einebnungsflachen in den feuchetn Tropen.
Zeitschrift fur Geomorphologie N.F. 1, 201228.
Carmo, I., Vasconcelos, P., 2006. 40Ar/39Ar geochronology constraints on late
miocene weathering rates in Minas Gerais, Brazil. Earth and Planetary Science
Letters 241, 8094.
Coelho Netto, M.L., 1999. Catastrophic landscape evolution in a humid region (SE
Brasil): inheritances from tectonic, climatic and land use induced changes.
Supplementi di Geografia Fisica e Dinamica Quaternaria 3(3), 2148.
Davis, W.M., 1899. The geographical cycle. Geographical Journal 14, 481504.
Fairbridge, R.W., Finkl, Jr. C.W., 1980. Cratonic erosional unconformities and
peneplains. Journal of Geology 88, 6986.
Falconer, J.D., 1911. The Geology and Geography of Northern Nigeria. Macmillan,
London, 295 pp.
Faniran, A., 1972. Nearest-neighbour analysis of inter-inselberg distance: a case

study of inselbergs of south-western Nigeria. Zeitschrift fur Geomorphologie
N.F., Supplement-Band 20, 150167.
Gilg, H.A., 2000. DH evidence for the timing of kaolinization in northeast Bavaria,
Germany. Chemical Geology 170, 518.
Gunnell, Y., 1998. The interaction between geological structure and global tectonics
in multistoreyed landscape development: a denudation chronology of the South
Indian shield. Basin Research 10, 281310.
Hencher, S.R., Lee, S.G., Carter, T.G., Richards, L.R., 2010. Sheeting joints:
characterisation, shear strength and engineering. Rock Mechanics and Rock
Engineering. http://dx.doi.org/10.1007/s00603-010-0100-y.
Henocque, O., Ruffet, G., Collin, F., Feraud, G., 1998. 40Ar/39Ar dating of West African
lateritic cryptomelanes. Geochimica et Cosmochimica Acta 62, 27392756.
Hill, I.G., Worden, R.H., Meighan, I.G., 2000. Geochemical evolution of a
palaeolaterite: the interbasaltic formation, Northern Ireland. Chemical Geology
166, 6584.
Johansson, M., Migon, P., Olvmo, M., 2001. Development of joint-controlled rock
basins in Bohus granite, SW Sweden. Geomorphology 40, 145161.
King, L.C., 1953. Canons of landscape evolution. Geological Society of America,
Kroonenberg, S.C., Melitz, P.J., 1983. Summit levels, bedrock control and the
etchplain concept in the basement of Suriname. Geologie en Mijnbouw 62,
389399.
Lidmar-Bergstrom, K., 1995. Relief and saprolites through time on the Baltic Shield.
Lidmar-Bergstrom, K., Olsson, S., Roaldset, E., 1999. Relief features and
palaeoweathering remnants in formerly glaciated Scandinavian basement
areas. International Association of Sedimentologists, Special Publication 27,
275301.
Merrill, G.P., 1897. A Treatise on Rocks, Rock Weathering and Soils. Macmillan,
New York, 411 pp.
384 pp.
Migon, P., 2009. Are any granite landscapes distinctive of the humid tropics?
Reconsidering multiconvex topographies. Singapore Journal of Tropical
Migon, P., Lidmar-Bergstrom, K., 2001. Weathering mantles and their significance
for geomorphological evolution of central and northern Europe since the
Mesozoic. Earth Science Reviews 56, 285324.
Migon, P., Lidmar-Bergstrom, K., 2002. Deep weathering through time in central
and north-western Europe. Problems of dating and interpretation of geological
record. Catena 49, 2540.
Migon, P., Thomas, M.F., 2002. Grus weathering mantles problems of
interpretation. Catena 49, 524.
Oberlander, T., 1972. Morphogenesis of granitic boulder slopes in the Mojave
Desert, California. Journal of Geology 80, 120.
Ollier, C.D., 1965. Some features of granite weathering. Zeitschrift fur
Geomorphologie N.F. 9, 265284.
Ollier, C.D., 1984. Weathering. Longman, London, 270 pp.
Parham, W.E., 1970. Clay mineralogy and geology of Minnesotas kaolin clays.
Minnesota Geological Survey Special Publication Series SP-10, 142 pp.
Pavich, M.J., 1989. Regolith residence time and the concept of surface age of the
Piedmont peneplain. Geomorphology 2, 181196.
Pavich, M.J., Obermeier, S.F., 1985. Saprolite formation beneath Coastal Plain
sediments near Washington, DC. Geological Society of America, Bulletin 96,
886900.
Penck, W., 1924. Die Morphologische Analyse. Engelhorn, Stuttgart, 283 pp.
Phillips, J.D., 2010. The convenient fiction of steady-state soil thickness. Geoderma
156, 389398.
Righi, D., Meunier, A., 1995. Origin of clays by rock weathering and soil formation.
In: Velde, B. (Ed.), Origin and Mineralogy of Clays. Clays and the Environment.
Schmidt, P.W., Ollier, C.D., 1988. Palaeomagnetic dating of late cretaceous to early
tertiary weathering in New England, NSW, Australia. Earth-Science Reviews 25,
363371.
Selby, M.J., 1982. Form and origin of some bornhardts of the Namib Desert.
Shaw, R., 1997. Variations in sub-tropical deep weathering profiles over the
Kowloon Granite, Hong Kong. Journal of the Geological Society, London 154,
10771085.
Sheppard, S.M.F., 1977. The Cornubian batholith SW England: D/H and 18O/16O
studies of kaolinite and other alteration minerals. Journal of the Geological
Society, London 133, 573591.
143
Simon-Coinc- on, R., Thiry, M., Schmitt, J.-M., 1997. Variety and relationships of
weathering features along the early tertiary palaeosurface in the southwestern
French Massif Central and the nearby Aquitaine Basin. Palaeogeography,
Palaeoclimatology, Palaeoecology 129, 5179.
Soderman, G., 1985. Planation and weathering in eastern Fennoscandia. Fennia
163, 347352.
Summerfield, M.A., 1996. Tectonic, geology, and long-term landscape development.
In: Adams, W.M., Goudie, A.S., Orme, A.R. (Eds.), The Physical Geography of
Africa. Oxford University Press, Oxford, pp. 117.
Tardy, Y., 1971. Characterisation of the principal weathering types by the
geochemistry of waters from some European and African crystalline massifs.
Chemical Geology 7, 253271.
Tardy, Y., Bocquier, G., Paquet, H., Millot, G., 1973. Formation of clay from granite
and its distribution in relation to climate and topography. Geoderma 10,
271284.
Taylor, G., Eggleton, R.A., 2001. Regolith Geology and Geomorphology. Wiley,
Chichester, 375 pp.
Taylor, G., Shirtliff, G., 2003. Weathering: cyclical or continuous? An Australian
perspective. Australian Journal of Earth Sciences 50, 917.
Theveniaut, H., Freyssinet, P., 2002. Timing of lateritization on the Guiana Shield:
synthesis of paleomagnetic results from French Guiana and Suriname.
Thiry, M., Simon-Coinc- on, R., Schmitt, J.-M., 1999. Problems, progress and future
research concerning palaeoweathering and palaeosurfaces. International
Association of Sedimentologists, Special Publication 27, 317.
Thomas, M.F., 1965. Some aspects of the geomorphology of domes and tors in
Nigeria. Zeitschrift fur Geomorphologie N.F. 9, 6381.
Thomas, M.F., 1966. Some geomorphological implications of deep weathering
patterns in crystalline rocks in Nigeria. Institute of British Geographers,
Transactions 40, 173193.
Thomas, M.F., 1967. A bornhardt dome in the plains near Oyo, Western Nigeria.
Thomas, M.F., 1969. Geomorphology and land classification in tropical Africa.
In: Thomas, M.F., Whittington, G.W. (Eds.), Environment and Land Use in Africa.
Methuen, London, pp. 103145.
Thomas, M.F., 1974. Granite landforms: a review of some recurrent problems
of interpretation. Institute of British Geographers, Special Publication 7,
1337.
Thomas, M.F., 1978. The study of inselbergs. Zeitschrift fur Geomorphologie N.F.,
Supplement-Band 31, 141.
Thomas, M.F., 1994. Geomorphology in the Tropics. A Study of Weathering and
Denudation in Low Latitudes. Wiley, Chichester.
Thomas, M.F., 1997. Weathering and landslides in the humid tropics: a
geomorphological perspective. Journal of the Geological Society of China 40, 116.
Thomas, M.F., 2006. Lessons from the tropics for a global geomorphology.
Singapore Journal of Tropical Geography 27, 111127.
Thomas, M.F., Thorp, M.B., 1985. Environmental change and episodic etchplanation
in the humid tropics of Sierra Leone. In: Douglas, I., Spencer, T. (Eds.),
Environmental Change and Tropical Geomorphology. Allen and Unwin, London,
pp. 239267.
Thorp, M.B., 1967. Closed basins in Younger Granite Massifs, northern Nigeria.
Zeitschrift for Geomorphologie N.F. 11, 459480.
Twidale, C.R., 1982. Granite Landforms. Elsevier, Amsterdam, 372 pp.
Twidale, C.R., 1991. A model of landscape evolution involving increased
and increasing relief amplitude. Zeitschrift fur Geomorphologie N.F. 35,
85109.
Twidale, C.R., Bourne, J.A., 1975. Episodic exposure of inselbergs. Geological
Society of America Bulletin 86, 14731481.
Twidale, C.R., Vidal-Romani, J., 2005. Landforms and Geology of Granite Terrains.
Taylor and Francis, London, 352 pp.
Vasconcelos, P., 1998. Geochronology of weathering in the Mount Isa and Charters
Towers regions, northern Queensland. CRC LEME Open-file Report 139.
Vasconcelos, P.M., 1999. KAr and 40Ar/39Ar geochronology of weathering
processes. Annual Reviews of Earth and Planetary Science 27, 183229.
Vasconcelos, P.M., Knesel, K.M., Cohen, B.E., Heim, J.A., 2008. Geochronology
of the Australian Cenozoic: a history of tectonic and igneous activity,
weathering, erosion, and sedimentation. Australian Journal of Earth Sciences 55,
865914.
Wayland, E.J., 1933. Peneplains and some erosional landforms. Annual Report
Bulletin, Protectorate of Uganda, Geological Survey, Department of Mines 1,
144
7779 (reprinted in Adams, G.F. (Ed.) (1975). Planation Surfaces,

Benchmark Papers in Geology 22, Hutchinson and Ross, Stroudsburg,
pp. 355357).
Widdowson, M., Gunnell, Y., 1999. Lateritization, geomorphology and geodynamics
of a passive continental margin: the Konkan and Kanara coastal lowlands of
western peninsular India. International Association of Sedimentologists, Special

Publication 27, 245274.
Willis, B., 1936. East African Plateaus and Rift Valleys. Carnegie Institute
Publications, Washington, 470, 358 pp.
Young, A., 1972. Slopes. Oliver and Boyd, Edinburgh, 288 pp.
Biographical Sketch
Piotr Migon is Professor of Geography at the University of Wroc"aw, Poland, where he is Head of Department of
Geography and Regional Development. He was a Royal Society Fellow based at the University of Oxford, UK in
19951996, and a Fulbright Fellow based at the University of California, Santa Cruz in 2003. He served as the
Secretary of the International Association of Geomorphologists (IAG) in 19972001 and currently (20092013) is
one of its three Vice-Presidents. His principal subject of research is how geological structures influence landforms
and surface processes. Other explored issues include weathering, mass movement in mountain terrains, and longterm landform evolution. Piotr Migon is also involved in various geo-conservation activities and in promotion of
geomorphology. His publication record includes Granite Landscapes of the World (Oxford University Press,
2006), a textbook on geomorphology published in Poland (2006), co-editing of Encyclopedia of Geomorphology (Routledge, 2004) and numerous papers in international journals, including Geomorphology, Catena,
Zeitschrift fur Geomorphologie, and Earth-Science Reviews. His last publication is the edited volume
Geomorphological Landscapes of the World (Springer, 2010), which contains 36 chapters describing the most
fascinating geomorphic sceneries from all around the world.
4.9 Catenas and Soils

RJ Schaetzl, Michigan State University, East Lansing, MI, USA
4.9.1
4.9.2
4.9.3
4.9.3.1
4.9.3.2
4.9.3.3
4.9.3.4
4.9.3.5
4.9.3.6
4.9.4
4.9.5
4.9.6
4.9.7
References
Introduction
The Catena Concept
Elements and Characteristics of Catenas
Summits
Shoulders and Free Faces
Backslopes
Footslopes
Toeslopes
Catenary Variation as Affected by Sediments and Climate
Soil Variation on Catenas Why?
Soil Drainage Classes along Catenas
The Edge Effect
Summary
Glossary
Catena A sequence of soils along a slope, having different characteristics due to variation in relief, elevation, and drainage (depth to water table), as well as the
influence of slope processes on sediment removal and
delivery.
Cumulization The slow, upward growth of the soil surface
due to additions of sediment on top of the soil. The
additions, for example, alluvium, loess, and slopewash,
must occur slowly enough so that pedogenesis can
incorporate the sediment into the profiles horizons.
Debris flux The movement of sediment (organic and
inorganic) across a slope, usually on the surface but also
including subsurface transfers.
Drainage class (soil-drainage class) Under natural
conditions, not artificially drained, this term refers to a
group of soils defined as having a specific range in relative
wetness due to a water table (apparent or perched), in
conditions similar to those under which the soil developed.
Edge effect The condition whereby soils located near
sharp breaks in slope profile at slope edges are
markedly different from those upslope and downslope. Two
types of edge effects exist, at wet edges and dry edges.
Endosaturation The condition of saturation of a zone or
soil horizon by groundwater (not perched water).
Episaturation The condition in which the soil is saturated
with water in one or more layers but in which it also has
one or more unsaturated layers below. Episaturation is
usually synonymous with the condition of having a perched
water table.
146
146
148
148
149
149
149
149
150
150
154
155
156
156
Hydrosequence A sequence of related soils, usually along

a slope, that differ, one from the other primarily with regard
to wetness.
Moisture flux The redistribution of water and solutes on
and within soils on a slope.
Pedogenesis The natural processes involved in the
formation of soils.
Pedon A theoretical term that represents the smallest
volume that can be called a soil. Pedons are threedimensional bodies of soil with lateral dimensions large
enough to permit the study of horizon shapes and relations.
The area of a pedon typically ranges from 1 to 10 m2.
Redoximorphic (redox) processes Chemical processes
associated with wetting (saturation) and drying (aeration)
of soils. The term is an abbreviation of the chemical terms
reduction and oxidation.
Slope element A segment of a hillslope, as viewed in cross
section. Traditionally, slopes have five elements. From the
top of the slope, downward, these are: summit, shoulder,
backslope, footslope, and toeslope.
Solum (plural: sola) The upper and most weathered part
of the soil profile; the A and B horizons.
Time zero The moment at which soil formation begins.
The sudden draining of a lake, with subsequent exposure of
the bottom sediments to impact of climate and organisms,
illustrates how time zero may be introduced into an area.
Theoretically, each soil has had a time zero.
Toposequence A sequence of related soils on a slope that
differ, one from the other, primarily because of topography
as a soil-formation factor.
Schaetzl, R.J., 2013. Catenas and soils. In: Shroder, J. (Editor in Chief),
Pope, G.A. (Ed.), Treatise on Geomorphology. Academic Press, San Diego,
CA, vol. 4, Weathering and Soils Geomorphology, pp. 145158.
http://dx.doi.org/10.1016/B978-0-12-374739-6.00074-9
145
146
Catenas and Soils
Abstract
Soil development is intimately tied to the slopes on which soils form. Soils across slopes are connected, process-wise, like
links in a chain. This analogy has led to the concept of a catena a term for a series of soils on a slope. This chapter explores
the reasons for soil variation on catenas, focusing on (1) debris and moisture flux along the slope and (2) depth to the
water table. Fluxes of sediment, commonly facilitated by water, vary predictably as a function of position on the slope,
leading to soils that may be thinner or thicker than expected on steep slope segments where runoff is accentuated.
Conversely, soils on lower, flatter slope segments may be overthickened from many years of slow but episodic sediment
accumulations from upslope; when sediment accumulations are particularly fast or large, soils here can become buried. Soil
texture and infiltration capacities dramatically impact these processes; on slopes composed of coarser, more permeable
materials, catenary position is less important because there is less runoff, and thus, even on the steepest slope segments,
much of the water infiltrates vertically. Water tables, commonly deepest on the steepest slope segments, vary predictably as a
function of position on the slope. High water tables can dramatically affect internal soil processes, as well as weathering and
related phenomena.
4.9.1
Introduction
Soils form on surfaces, and, thus, they are affected by aboveground and below-ground slope processes and by their position on that slope. That soils vary as a function of slope
position has been known for decades, and a great amount of
research has been conducted on these interrelationships. The
field of soil geomorphology has at its core these fundamental
soilslope linkages. Most of this sort of research has focused
on documenting, and then explaining, the changes manifested
in soils as one traverses the various slope positions. This type
of research operates within the paradigm that soils on slopes
change regularly, but ultimately, in a predictable manner
across them and that each soil individual is genetically connected to the one next to it.
The examination of soils along a catena is one of the best
ways to discern the interrelationships between soils and
topography. A catena is a transect of soils from the top to the
base of a slope, generally interpreted as a transect perpendicular (or nearly so) to the slope contour. Its name comes from
the Latin catenaria, or chain. Soils in a catena can be visualized
as interconnected chain links; solid materials, fluids, and gases
move through and between the soils (links) that are connected
on the slope, making each soil (pedon) like a link in a chain.
Visualizing this type of chain catenation is what first resulted
in the formation of the catena concept.
Variation along and within a soil catena is a manifestation
of Jennys (1941) relief soil-forming factor. Although relief
itself is passive, it functions by providing potential and kinetic
energy to the soil system, through its impact on the flows of
matter and energy within the soil-landscape system.
4.9.2
The Catena Concept
Some of the earliest field (and theoretical) work on catenas

was done in the 1930s by Geoffrey Milne (Brown et al., 2004).
Milne and his colleagues were trying to explain some complex
soillandscape associations in Uganda. He noted that, over
large areas with regular (or rhythmic) topographic variability,
the soils also had repeatable patterns and that these patterns
would have to be interpreted as indicating the occurrence of
not a single soil but of a sequence of soilsy (Milne, 1932: 5).
This sequence of soils was Milnes catena a concept focused
on soils that have repeated patterns on the landscape, due

mainly to the effects of topography and slope (Milne, 1935a,
1935b). The term has stood the test of time and remains in
widespread use today. Later, Milne would note that soils along
catenas also vary because of: (1) differences in drainage conditions; and (2) the effects of erosion and sedimentation on
the slope. Holliday (2006) noted that Milnes unique contribution was actually in his linking of soilcatenary patterns
to specific slope-related processes: wetness, solute transport,
and erosion/deposition.
Milnes catena concept was found to be useful for soil
mapping in the United States and elsewhere, and hence it was
incorporated into one of the US Department of Agriculture
(USDA)s earliest soil-classification systems (Baldwin et al.,
1938). Mapping soils along catenary sequences of otherwise
similar parent materials helped to develop the notion of soildrainage classes (see below), as the overall wetness and depth
to the water table of these soils tended to also vary predictably
along slopes. Soils along catenas that differed only in drainage
class, as manifested in their profile characteristics, came to
form one of the smallest and tightly defined taxonomic classes
of soil the soil series (Soil Survey Staff, 1951). Holliday
(2006) noted that the early US soil-survey efforts utilized a
catena concept that focused on soils that vary along slopes
mainly due to the effects of topography and drainage. Together, these incremental advances in our knowledge of soils
within their various catenary settings led to fundamental
progress in the fields of soil genesis, soil geomorphology, and
soil classification.
The notion of parent-material uniformity along a catena,
although utilized by many in the US, including the Soil Survey, was not a part of Milnes original concept. However, it
soon would be. In a classic paper, Bushnell (1942) expanded
on the catena concept and offered additional theoretical suggestions and refinements. Bushnell formally suggested that the
term catena be limited to slopes where the soils have all
formed in one parent material, that is, all soil-forming factors
except topography (drainage was the word he used) are
held constant. Taken together, these developments devalued
slope-influenced transport/depositional processes, previously
so important to the explanation of soil development along
slopes.
Today, the catena concept is used primarily for soils along a
slope, but particularly where they are generally all formed in a
Catenas and Soils
similar parent material. Thus, it is assumed that their genetic

and morphologic differences have accrued due mainly to
the influence of topography. Knowing that such pure catenas
are more often the exception than the rule, two other
terms eventually came into use for the less than perfect
catenas. If the topographic influence is mainly manifested by
(or from) differences in soils wetness (or depth to the water
table), the term hydrosequence is sometimes used, as in
Cremeens and Mokma (1986), Mokma and Sprecher (1994),
and Smeck et al. (2002). The term toposequence is now
commonly used for any sequence of soils along a slope,
regardless of the uniformity or lack thereof of the other
soil-forming factors, as in Alvarado and Buol (1975), Evans
and Franzmeier (1986), and Bravard and Righi (1991).
Hydrosequences and toposequences commonly focus on
morphologic changes due to differing wetness conditions
(Hall, 1983), whereas soils along a true catena may differ
because of drainage condition changes and fluxes of sediment.
For ease of reading, in this chapter the term catena will be
used for any sequence of soils along a slope, essentially mirroring the term toposequence.
By definition, the term catena implies a two-dimensional
(2D) transect along a hillslope. However, by nature, of course,
flows of matter and energy within the geomorphic system are
inherently 3D. Flowlines down slopes can converge or diverge,
rendering the base of the slope drier or wetter, respectively,
than would be expected if the flowlines were simpler straight
down the slope (Figure 1). Soils on nose-slope catenas, for
example, are drier and generally thinner than might be
expected because nose slopes are water- and sedimentdiffusing slopes. These areas are called divergent slopes. On
147
head slopes, where concave contour/plan curvature exists,

flowlines will converge onto the lower parts of the slope; these
areas are called convergent slopes. Soils on convergent slopes
tend to be wetter and runoff here is initiated onto (and from)
these sites more rapidly, other things being equal. On the
lower ends of convergent slopes, sediment will preferentially
accumulate, and soils tend to be thicker (King et al., 1983).
Contrast that to areas at the base of nose slopes; these areas
tend not to accumulate as much sediment and run-on as do
other types of slopes, and soils are thinner here.
Obviously, slope curvature and complexity can dramatically affect soil development, and it occurs in more than
just the simple, catena-based, 2D manner. For this reason,
recent research has commonly focused on soils within a 3-D
system, whether it be via terrain/geographic information systems (GIS) modeling or empirical research (Huggett, 1975;
Moore et al., 1991, 1993; Western et al., 1999; Yang et al.,
2007; Pei et al., 2010; Zhao et al., 2010). The focus in this
chapter, however, is on the 2D, that is, the traditional, soil
catena.
Lastly, it is important to note that catenas fall into two
main categories: closed and open. Closed catenas are formed
in internally draining depressions, that is, little or no sediment
can leave the system (Richardson et al., 1994). In closed
catenas, soil burial is much more likely in lower slope positions because sediment continually accumulates there via
slopewash and other processes but cannot leave the system
(Walker and Ruhe, 1968; Weitkamp et al., 1996). Most catenas
on geomorphologically mature landscapes form as part of an
open system, where the lower slope segments connect to some
sort of integrated stream or drainage system, which can
Flowlines down slopes, as impacted by slope curvature

Contour
(plan)
Slope
Linear
Convex
Concave
Linear
Convex
Concave
Figure 1 Flowlines down slopes of various curvatures. Reproduced from Ruhe, R.V., 1975. Geomorphology. Houghton Mifflin Co., Boston, 246
pp and Huggett, R.J., 1975. Soil landscape systems: a model of soil genesis. Geoderma 13, 122.
148
Catenas and Soils
remove sediment from the system (Ruhe and Walker, 1968).

Thus, soil burial in lower slope segments is less common but
can be accomplished via flooding of streams.
4.9.3
Elements and Characteristics of Catenas
Slopes, like catenas, can be simple or complex, convex, concave or straight, short or long, and steep or gentle. There are
no rules. Perhaps the easiest way to comprehend this complexity is to partition the slope into smaller and, hence, more
easily handled units, or elements (Figure 2(b)).
Most simply, all slopes can be viewed as having an erosional component nearer the top, a transportational component in the mid-slope position, and a depositional component
nearer the base (De Alba et al., 2004). This three-part slope
categorization fits well with most slopes/catenas, especially the
steeper ones. Soils respond to these general slope-development processes, in that some may thicken or even get buried at
the bases of the slope, others may erode on the upper slope
components, while others may maintain a balance between
pedogenic and slope development. Thus, along a given slope
or catena, the evolutionary stories of soils change; some soils
may be undergoing long-term, gradual erosion, whereas others are becoming progressively buried. In short, catena slopes
can generally be considered time transgressive the ages of the
surfaces change progressively along the slope, in this case,
induced by erosion and burial processes. The introduction of
differing soil ages along the slope complicates the catena

concept, but is important and commonplace nonetheless.
These points illustrate that soils and slopes codevelop.
A related way to examine slopes is to consider that they have
three main components: a rounded upper edge or waxing
slope, a constant slope (of varying length), and a waning slope
where sediment and debris accumulation is most pronounced
(Wood, 1942; Ahnert, 1970). Waxing slopes are generally
concave downward; waning slopes are concave upward. Most
slopes also have a flat, horizontal (or nearly so) component in
the middle and at their base, even if these segments are short.
On simple slopes, the pattern (from the top) is convex, straight,
and concave. Many slopes are more complex or compound, but
each segment still retains the convexstraightconcave pattern,
in a downslope direction, over at least some part of the slope.
Taking slope description one step further, Ruhe (1960)
defined five elements of slope, and hence, of soil catenas.
From the top to the base, they are: summit, shoulder, backslope, footslope, and toeslope. Most slopes and catenas contain all of these elements, and soil properties vary consistently
and predictably as a function of these landscape positions
(Figure 2(b)).
4.9.3.1
Summits
Summits are generally stable surfaces, especially if flat, with

minimal amounts of erosion or accretion. Thus, summits are
commonly dominated more by chemical weathering than by
Types of slopes and slope elements

Divide
Si
slo de
pe
Head
slope
Interfluve
(summit)
Alluvial fill
Nose
slope
Flow of stream
(a)
Su
Sh
Sh
Bs
Su
Bs
Fs
Fs
Ts
Ts
Alluvium
Su - summit
Bs - backslope
Fs - footslope
Ts - toeslope
(b)
Figure 2 Three-dimensional slope types (a) and the various slope elements commonly used in soil geomorphology (b). Reproduced from Ruhe,
R.V., 1975. Geomorphology. Houghton Mifflin Co., Boston, 246 pp.
Catenas and Soils
physical weathering and erosion (King, 1957). Exceptions

occur either where the summit is narrow or where runoff is
common because of wet climates and slowly permeable soils.
Wide summits may be the oldest and most stable of the five
slope elements, because water is unlikely to run off. Most of
the water that falls on summits infiltrates, leading to betterleached and developed soils than on slope elements that are
immediately downslope. For example, clay maxima may be
deeper here than on steeper slope elements because more
water infiltrates. However, on slowly permeable sediment or
soils, or on bedrock, the summit may be an area where water
perches either on the soil surface or within the profile, making
these upland soils much wetter than would otherwise be
expected. Indeed, on many low-relief landscapes formed on
slowly permeable materials, the flat uplands are some of the
very wettest sites. Because water and sediment fluxes across
summits tend to be minimal and spatially uniform, soils are
also minimally varying across them, with the exceptions being
sharp-crested and/or undulating summits.
4.9.3.2
Shoulders and Free Faces
On shoulder slopes, slope convexity is maximal. Runoff and

erosion are dominant processes (Pennock, 2003). The steepest
of shoulder slopes are called free faces. On steep shoulder
slopes, runoff dominates to the point that erosion can effectively outstrip soil formation (Walker and Ruhe, 1968). As a
result of the continual erosion on shoulders, surfaces and soils
there are usually the youngest and least stable of all the surfaces on a catena. Soils on shoulders are comparatively thin,
lower in organic matter, coarser textured, and drier. Surface
instability on shoulders can be initiated either by surface
runoff or by lateral flow of water in the subsurface, which is
especially common where bedrock is near the surface. Nonetheless, the shoulder is generally the driest slope position, and,
during most years, it undergoes the greatest amount of watertable fluctuation (Khan and Fenton, 1994). Sites farther
downslope are more uniformly wet with high water tables,
whereas flat summit positions are wetter due to minimal
amounts of runoff.
4.9.3.3
Backslopes
Backslopes are comparatively steep, transportational slopes.

They are commonly straight, that is, neither convex nor concave, being positioned at the junction between upslope areas
dominated by erosion and lower-slope areas of sedimentation.
Where backslopes are short, soils can change markedly from
those just upslope to those downslope (King et al., 1983).
Materials move through and across backslopes, depending in
part on slope gradient and length. Water may run off on the
surface as well as in throughflow on top of less permeable
layers in the subsurface (Huggett, 1976; Schlichting and
Schweikle, 1980). On long backslopes, sediments may
become increasingly sorted and finer-textured downslope. As
is the case with slope processes on backslopes vis-a`-vis other
slope elements, soils here tend to be intermediate in character.
It is difficult to generalize about soil attributes on backslopes,
149
as much depends on the relationships between, and rates of,

runoff (erosion) versus pedogenesis.
4.9.3.4
Footslopes
Footslopes are the most concave parts of a catena and, because

they are also on the lower slope, footslopes are the most
favored sediment- and water-receiving slope positions. Materials, carried in solution and suspension, whether above or
below the surface, are commonly transported onto footslopes.
The results, generally, are thicker A horizons and sola, and
even buried soils, in footslope positions. These slope positions
are commonly cooler than are upslope areas because of cold
air drainage and more complete shading in the valleys. Wetter
conditions also exist here, due to higher water tables. Thus,
conditions at the base of the slope may promote more plant
productivity and inhibit decomposition of the organic materials they produce, combining to make soils in the footslope
(and toeslope) positions much richer in organic matter than
on other slope elements (Kleiss, 1970).
Water and sediment impinge upon footslopes, especially
where flowlines converge. As a result, spring sapping is common here, which may in turn lead to gullying. As these gullies
erode upslope, into the backslope, more sediment impacts the
footslope, and small depositional fans may form, burying still
more soils.
4.9.3.5
Toeslopes
Toeslopes, a.k.a alluvial toeslopes, are at the outward limit of

footslopes. Like footslopes, toeslopes are constructional sites
(Vreeken, 1973). Sediment accumulates here from upslope
areas and also from streams that flood and deposit overbank
alluvial sediments. Sediment accumulating on footslopes and
toeslopes tends to get progressively finer, farther away from
the upslope-contributing areas because slopewash processes
tend to transport finer material farther (Walker and Ruhe,
1968; Malo et al., 1974; Nizeyimana and Bicki, 1992). Accumulation of slope-derived sediment at the bases of slopes is
especially important in basins of closed drainage, where few
mechanisms exist to remove it (Walker and Ruhe, 1968).
Soil development on toeslopes reflects the inherent wetness of the site (relative to the other slope elements), the
tendency for high plant productivity and organic-matter production on such sites and the intermittent accumulations of
surficial sediment. Thus, A horizons tend to be thicker than
anywhere else along the catena, and soil burial and overthickening are common (Gregorich and Anderson, 1985).
Overthickened soils develop because of sediment accumulations from upslope-contributing areas. If this sediment
accumulates so gradually that it can be pedogenically incorporated into the upper profile, overthickened A horizons and
sola thicken. This process is referred to as cumulization or
developmental upbuilding; it is driven not only by slopewash
sediment, but also aeolian and anthropogenic additions of
mineral particles to the soil surface of a soil (Riecken and
Poetsch, 1960; Johnson, 1985). More rapid rates of sediment
accumulation in footslope and toeslope locations will, instead,
lead to soil burial. Cumulization can be considered a type of
150
Catenas and Soils
4.9.4
progressive pedogenesis, if the new material is effectively

assimilated into the profile (Johnson et al., 1990).
4.9.3.6
Soil Variation on Catenas Why?
Soils on a catena generally vary only subtly at a pedon-topedon scale, but taken as a whole, soils across the full catena
may be drastically different from the top of the slope to its
base, for example, Wieder et al. (1985) and Weitkamp et al.
(1996). Regardless, the soils on the catena are all linked to one
another through lateral translocations of fluids, gases, and
materials, as well as, potentially, by their similar inheritance of
the various state factors parent materials, climate, biota,
surface age, and relief (Dan and Yaalon, 1964). That is, at time
zero, all or most of the soils (or non-soil sediments) on the
slope may have been essentially or nearly identical.
Over time, pedogenesis and slope processes act nonuniformly across the slope, causing soils there to become
increasingly dissimilar over time, as each develops along its
own, unique pedogenic pathway (Johnson and Watson-Stegner,
1987). Vertical and lateral translocations within the soils
can vary in intensity at different locations on the slope
(Figure 5(b)), and therefore, so do claymineral genesis and
weathering. These, and a suite of other pedogenic processes, act
in unison to place each pedon on its own unique pedogenic
pathway. As time passes, the pedogenic divergence across the
slope usually increases. Of course, all of this developmental
divergence occurs because some (or all) of the state factors,
which may have been similar across the slope at time zero, sort
themselves out over time and become increasingly more variable, with time, across the slope. These state factor differences,
for example, in vegetation or depth to water table, eventually
Catenary Variation as Affected by Sediments and

Climate
Figure 3 summarizes many of the points made above, by

showing the relative intensity of various surficial processes on a
typical catena. Areas of more concentrated erosion and thus,
thinner soil profiles, that is, shoulder slopes, contrast with
more stable landscape positions. This general model of
soilslopeprocess linkages would best fit a catena developed
on moderately permeable soils in a humid climate. In areas of
highly permeable soils and deep water tables, there may be
minimal runoff, no water-table influence, and thus, much less
soil morphologic variation along the catena (Figure 4). The
same can be said for some catenas in dry climates, where, again,
water tables may be very deep, and thus, lower slope soils may
be as dry as soils on upper catenary positions. Runoff may be
considerable, however, in some of these catenas because of
heavy (albeit infrequent) rainfall events. Thus, Figure 3 presents the mode of processes and soils for catenas, but variations on this theme are myriad, and one must always strive
to understand the relative intensities of the various process
drivers on the catena in question, for example, precipitation
and snowmelt, as well as the inherent properties of the
soilsediment system, for example, infiltration capacity, slope
steepness, and water-table depth, all of which affect runoff
intensity and, hence, soil variation across the slope (Figure 5).
Solum thickness
Run off/erosion potential
Run on potential
Geomorphic stability
Organic matter in a horizon
A horizon thickness
Likelihood of burried soils
Bar order legend
Summit
Sh
ou
lde
Ba
ck
slo
pe
Foo
tslo
pe
Toeslope
Figure 3 Estimates of relative variation in soil characteristics and slope-process intensity along a typical catena in a humid climate.
Catenas and Soils
Summit
Shoulder
Elevation (m)
362
361
5.0
Hillslope
profile
Backslope
360
Footslope
359
Toeslope
358
0
30
60
90
120
Distance from summit (m)
150
Organic carbon in A horizon (%)
363
Y = 6.78 + 0.012 D 2.74 log D

R = 0.967
4.5
4.0
3.5
3.0
2.5
30
0
Y = 22.63 0.18 D + 0.00069 D 3
R = 0.975
1.50
125
100
75
50
25
0
30
60
90
Distance (m)
120
150
22.5
Weighted mean gravel (> 2 mm dia) in solum (%)
Depth to <1% organic carbon (cm)
175
150
151
60
90
Distance (m)
120
150
120
150
Y = 1.41 [(1 105)(D 2)] +

[(2 1012 )(D250)5]
R = 0.952
1.25
1.00
0.75
0.50
0.25
0
0
30
60
90
Distance (m)
36
17.5
Content in A horizon (%)
Geometric mean particle

size of A horizon (microns)
20.0
15.0
12.5
Y = 19.82 0.00006 D2
R = 0.906
10.0
Y = 35.5 0.0 14(D)

0.00016(D150)2
32
Sand
28
24
Clay
20
R = 0.875
16
7.5
0
30
60
90
120
150
Distance (m)
R = 0.951
30
Y = 21.5 + 0.02(0)
+ 0.00032(D150)2
60
90
Distance (m)
120
150
Figure 4 Quantitative relationships between slope (catenary) position and various soil properties. Reproduced from Malo, D.D., Worcester, B.K.,
Cassel, D.K., Matzdorf, K.D., 1974. Soil-landscape relationships in a closed drainage system. Soil Science Society of America Proceedings 38, 813818.
force these pedogenic process changes to occur, and these

processes drive the morphological and chemical changes in the
soils across the catena. State factors may vary dramatically
across the slope, but such variation is minimal between
adjoining pedons; this fact explains why adjacent soils may be
only subtly different, whereas soils on the summit are dramatically different from those in lower slope positions. Thus, soils
vary across slopes according to how the state factors vary, and
these pedogenic differences may become increasingly manifested over time.
152
Catenas and Soils

Infiltration capacity < precipitation intensity
Infiltration capacity precipitation intensity
Precipitation
Runoff
Precipitation
Amounts of runoff and runon,

erosion, and burial,
dependent on slope position
All soils equally leached
Infiltration
Infiltration
Runon
(a)
(b)
Figure 5 Two end-member scenarios illustrating the importance of, and interactions between, precipitation amount and intensity vs. infiltration
capacity along a catena, and how these variables affect slope/soil processes. Reproduced from Schaetzl, R.J., Anderson, S.N., 2005. Soils:
Genesis and Geomorphology. Cambridge University Press, Cambridge, 832 pp.
This section also explores some of the main process reasons that soils vary across catenas. Materials, gases, and liquids
are translocated on (above) and through (within) the slope
materials, ultimately forming the differences in these soils on
the various slope positions (Hall, 1983). Groups of processes
translocations, transformations, additions, and removals
that were championed by Simonson (1959) as useful for
explaining intra-pedon, that is, horizon-related, differences,
also help explain inter-pedon differences across catenas.
Schaetzl and Anderson (2005) explained that soils vary
along catenas for two main reasons: (1) fluxes of water and
matter, generally but not always in the downslope direction
and (2) influence of the water table in the lower portions of
the catena, or as perched water on flat summits (Figure 6).
They pointed out that there are two main types of fluxes along
slopes: debris flux and moisture flux (ignoring gas fluxes as
very minor components of the story). Debris fluxes involve
both sediment and organic material (Malo et al., 1974).
Debris flux is so important that, on many slopes, soils more
strongly reflect the hillslope-sedimentation system than the
pedologic system (Kleiss, 1970). Sommer and Schlichting
(1997) presented an idealized scheme to represent the main
types of fluxes within catenas, emphasizing that fluxes can be
downslope, horizontal, and even upward, and are usually
driven by water (Figure 7).
Debris flux (again, usually driven by moisture fluxes)
involves an erosional component and a depositional component, usually on the same catena. Where overland flow and
rainsplash are the primary drivers of this flux, the term
slopewash is generally used. On summits, water may infiltrate
or run off, depending on the slope curvature and the balance
between precipitation and infiltration rates (Figure 5).
Because they are the steepest slopes on the landscape,
shoulders and upper backslopes have the most potential for
runoff and hence exhibit the most eroded, thinnest soil profiles. Indeed, shoulders and upper backslopes are the most
likely areas for rock outcrops (Figure 3). Because finer sediment is more easily eroded, coarser materials may remain on
eroding slope positions, making soils there more sandy or
Processes that drive soil variation along and within catenas
Fluxes
Gases
Water table influence
Matter Moisture
(debris)
Infiltration/percolation
Return flow
Across the Below

surface ground
Lateral
throughflow
Surface Surface
runoff
runon
Figure 6 The main clusters of processes that drive soil variation
along and within catenas.
gravelly and rendering the depositional areas (and soils)

downslope finer textured. Debris-flux processes tend to sort
materials during transit; thus, sediment texture tends to get
finer as one progresses from the shoulder and backslope
positions to lower positions along the catena. This trend is
especially well expressed in closed basins because, in open
basins, either the translocated sediment may be removed by
streams, or other sediment may be added to the lower slope by
fluvial processes. Backslopes are generally areas of transportation, with maximal debris and moisture flux. If interflow or
throughflow is present along a catena, it will be most pronounced here, perhaps accentuated by slowly permeable layers
in the subsoil (Figure 7). Farther downslope, debris fluxes
slow, and deposition becomes an increasingly dominant process, such that cumulization can occur on toe- and footslopes.
Moisture flux parallels debris flux; indeed, it is the driving force
behind it. Slope steepness and curvature directly affect water
flow, both within and above the soil (Huggett, 1975, 1976).
The amount of water entering a soil on a slope depends on a
number of factors, such as infiltration capacity, slope steepness
Catenas and Soils

Soil surface
153
Soil surface
Soil surface
1
Pedon A
Pedon A
Pedon A
Pedon B
Pedon B
Pedon B
5
Groundwater
4
Groundwater
1 - Overland flow (runoff)
Groundwater
2 - Lateral subsurface flow (throughflow) 3 - Vertical seepage (percolation)
5 - Return flow (saturation overland flow)
4 - Capillary rise
Figure 7 Idealized diagram of the various fluxes that can occur along a catena. Reproduced from Sommer, M., Schlichting, E., 1997.
Archetypes of catenas in respect to matter a concept for structuring and grouping catenas. Geoderma 76, 133.
0
25
50
Water table
Depth (cm)
75
100
125
150
175
Porosity
Volumetric
water content
200
Water table
225
250
0
10 20 30 40 50 60 70 80 90 100 0
(a)
10 20 30 40 50 60 70 80 90 100
(b)
Volumetric water content and porosity (%)

Figure 8 Relationship between volumetric soil-water content and porosity in two soils in Iowa: Clarion ((a) summit, well drained) and Canisteo
((b) toeslope, poorly drained). Reproduced from Khan, F.A., Fenton, T.E., 1994. Saturated zones and soil morphology in a Mollisol catena of
central Iowa. Soil Science Society of America Journal 58, 14571464.
and plan curvature, intensity/frequency of rainfall events, and

cover type. In theory, inputs of water from precipitation
are evenly distributed along the catena. If soil-infiltration
capacities exceed rates of precipitation input, most of the precipitation will immediately enter the soil and catenary differences due to moisture flux will be small (Figure 5). However,
this situation may hold only in the coarsest textured catenas.
Generally, inputs of precipitation are redistributed unevenly
across the surface, affected by the varying slope gradients and
infiltration capacities, such that the actual amount of water that
infiltrates individual soils may be vastly different across the
catena.
Soils in lower slope positions may be influenced by the
water table, or not, depending on a number of local conditions, as well as the regional climate (Figure 8). If water
table is high, soils there may classify as aquic or hydromorphic
soils (Vepraskas and Guertal, 1992; Bell and Richardson,
1997; Vepraskas et al., 2004). In most humid areas, where the
water table subtly mimics the surface topography, soils in the
lower slope areas receive discharge from the groundwater and
are continually wet, variously impacted by slowly moving or

stagnant groundwater (Figure 9). In drier climates, the water
table may be much deeper. Here, water that runs off the slope
impacts the soils on the lower slope positions as run-in and,
potentially, added amounts of infiltration. Soils in these
recharge areas are more variably wet, ranging from dry to
saturated, and, because of the downward percolation of water,
are more leached (Figure 9).
High water tables inhibit certain pedogenic processes,
especially those related to vertical translocation, for example,
lessivage or podzolization. As vertically percolating soil water
encounters the water table, it loses kinetic energy and diffuses
into the groundwater. Any dissolved or suspended substances
in the water tend to be deposited at or near the water table,
and little or no additional vertical translocation, below the
water table, occurs.
The presence of a water table in soils brings with it an
entirely new suite of pedogenic processes, generally associated
with redoximorphic processes, as well as vertical influx
(including capillary rise) and lateral influx of soluble materials
154
Catenas and Soils
WD
WD
WD
Recharge wetland
Ru
no
ff
WD
MWD
SPD PD
Water table
VPD
Drainage class legend
WD
WD
Ru
WD:
MWD:
SPO:
PD:
VPO:
no
ff
MWD
Water table
SPD
Well drained
Moderately well drained
Somewhat poorly drained
Poorly drained
Very poorly drained
Discharge wetland
PD
VPD
Figure 9 Two end-member scenarios for discharge vs. recharge wetlands and conditions on the lower slopes of catenas. Soil-drainage classes
are also shown. Reproduced from Schaetzl, R.J., Anderson, S.N., 2005. Soils: Genesis and Geomorphology. Cambridge University Press,
Cambridge, 832 pp.
(Vepraskas, 1999). Redox processes affect mainly Mn, Fe, and

S compounds. However, saturated conditions also impact
many solubility-based pedogenic processes, such as leaching,
salinization, and sodicity (White, 1964; Peterson, 1980; Dan
and Yaalon, 1982; Berger and Cooke, 1997). Organic-matter
decomposition is commonly inhibited in the lower slope
positions in humid climate soils because saturated conditions
reduce the amount of oxygen needed for complete decomposition. For this reason, the lower portions of many catenas
in humid climates, especially in recently glaciated landscapes
where the lower slopes are replete with high water tables, have
thick sequences of Histosols. Here, the wet, saturated conditions, coupled with the inherently cool nature of the climate,
leads to only partial decomposition of any raw organic matter
that may accumulate there.
Many soil-classification schemes now refer to soils impacted by a shallow water table as having a type of endogenic
water regime, or the condition of endosaturation; for example,
the Natural Resources Conservation Service (NRCS) classifies a
wet Spodosol with a shallow water table as an Endoaquod.
Alternatively, perched water tables disconnected from the
much deeper apparent water table by an unsaturated zone
are distinguished, taxonomically, by the term epi, for example, Epiaquod or the condition of episaturation (McDaniel
and Falen, 1994). Water may perch on any slowly permeable
layer, such as a clay-rich B horizon, a Ca-cemented horizon, or
even slowly permeable, dense till in a C horizon. Perched
water tables can occur at any catenary position but are most
common on sites of lowest slope, where soil water has the
most difficult time moving out of the profile laterally. Thus, in
such situations, the water will have the greatest potential to

perch on the slowly permeable layer.
4.9.5
Soil Drainage Classes along Catenas
Loosely tied to landscape position, and generally correlating

with depth to the water table, is the concept of natural soil
drainage classes (Soil Survey Division Staff, 1993; Schaetzl and
Anderson, 2005; Table 1). With the exception of perched
water tables, natural soil drainage classes vary systematically as
a function of topography (Mackintosh and van der Hulst,
1978; Figure 10). Soils with deep water tables are on uplands
and upper backslopes, whereas, at the lower slope positions,
the highest water tables and the wettest drainage classes are
likely to occur. Indeed, on many landscapes, the driest soils
those with the deepest water tables and the driest drainages
class are on shoulders and upper backslopes, not on the flat
summits.
Schaetzl et al. (2009) integrated the concept of natural soil
drainage classes with that of other soil characteristics that
influence long-term wetness, for example, texture, depth to
bedrock, and organic-matter content, to develop the Natural
Soil Drainage Index (DI). The DI is an ordinal index, ranging
from 0 for the driest soils to 99 (open water) and is derived
from the soils taxonomic classification, which it assumes is a
reflection of its long-term wetness. Although the DI has some
degree of subjectivity, it has nonetheless proven to be highly
useful for various types of landscape scale, soil wetness analysis and visualization. Particularly enlightening is the use of
Catenas and Soils
155
Table 1 Generalized features of the various natural soil drainage classes recognized by the NRCSa
Natural soil drainage class
Typical depth to water

table (cm)
Uppermost horizon(s) in which redox

features and gleying are typically
observed
Other characteristics
Excessively drained
Somewhat excessively drained
Well drained
4150
4150
100150
None
None
Upper C or lower BC
Moderately well drained
75125
Lower B horizon
Somewhat poorly drained
3075
Upper B (gleying in lower B)
Poorly drained
o3050
Throughout the profile (gleying in all

or most B horizons)
Very poorly drained
015 (but usually above

the mineral soil
surface)
Throughout the profile (gleying

throughout)
Very coarse-textured soils

Coarse-textured soils
Typical of upland soils with deep water
tables
Slightly wetter and farther downslope
than upland, well-drained soils
Water table impacts some part of the
profile in all seasons
Water near or at the surface yearround; overthickened A and O
horizons
Organic materials accumulate above
mineral soil surface, sometimes to
great thicknesses
From Soil Survey Division Staff., 1993. Soil Survey Manual. USDA Handbook No. 18. US Government Printing Office, Washington, DC, 437 pp. and Schaetzl, R.J., Anderson, S.N.,
2005. Soils: Genesis and Geomorphology. Cambridge University Press, Cambridge, 832 pp.
0
Knoke (Footslope)
Depth of water table (cm)
Knoke
(Footslope)
100
Clarion
(Shoulder)
200
Clarion
(Shoulder)
Canisteo (Toeslope)
Knoke (Footslope)
Nicollet (Backslope)
Clarion (Shoulder)
Clarion (Summit)
300
1984
J
1985
J
1986
J
1987
J
1988
J
1989
J
1990
J
1991
J
1992
1993
Figure 10 Relationship between water-table depth and various drainage classes and catenary positions, for some soils in Iowa. Reproduced
from Khan, F.A., Fenton, T.E., 1994. Saturated zones and soil morphology in a Mollisol catena of central Iowa. Soil Science Society of America
Journal 58, 14571464.
the DI in a geographic information system, where the DI

values are shown in various colors, representing wetness,
ranging from light orange (the driest soils) through yellow,
green, blue, and finally purple (the wettest soils). Figure 11
illustrates the wetness of a sample landscape, using the DI
color scheme outlined above.
4.9.6
The Edge Effect
As noted earlier, shoulders tend to be some of the driest sites

along a catena (Daniels et al., 1971a). Especially dry are
shoulder slopes on stepped landscapes, where the shoulder
represents a drop down onto a steep escarpment. Such a site
is referred to as a landscape edge (Daniels and Gamble,
1967). The case above, in which the escarpment drops off to

a drier (shoulder slope) landscape, is a dry edge. Short
escarpments that fall off into a wet depression are called
wet edges. The edge effect is an example of the close
linkages between soil development, the water table, and
geomorphology.
Daniels and Gamble (1967) studied dry edges and
observed that soils change the most, pedon to pedon, along
catenas, at these edges. Here, state factors are changing most
rapidly across short horizontal distances. Particularly noteworthy is the great change in depth to the water table along
such an edge (Daniels et al., 1971b). At a dry edge, water tables
are generally deeper than at sites upslope or downslope, and
changes in the depth to the water table are most extreme.
Redder soil hues, especially in the B horizons, show that
156
Catenas and Soils
750 1500
3000
Meters
several others. These factors all combine to influence soil

development differently along up and down and across a
slope, through fluxes of water and sediment, as well as by
water-table influences. The term catena has traditionally been
used to refer to these soil linkages and changes that occur
along slopes. It is a valuable concept that anyone who studies
soils within landscapes needs to have in their conceptual bag
of tricks.
References
Figure 11 Soil-wetness map for a small area in northeastern Otsego

County, Michigan, using the Natural Soil Drainage Index of Schaetzl
et al. (2009). (See: http://www.drainageindex.msu.edu/). This is an
area of predominantly sandy soils, with very poorly drained Histosols
(purple) in the broad, swampy lowlands. Note the areas of Histosols
in the small bogs on the otherwise flat upland. The well-drained soils
(green) on the flat upland contrast with the incised backslopes, which
have notably drier soils (light green) due to their steeper slopes. Areas
with yellow hues are excessively drained sandy soils. In a few places,
somewhat poorly drained soils (teal) on footslopes can be seen,
intermediate in wetness between the Histosols of the valley floors and
the well-drained soils of the backslopes.
conditions are drier and water tables are deeper, nearer the
edge. For this reason, Daniels and Gamble (1967) also referred to this area as the red edge. Increased translocation of clay
in edge soils occurs, perhaps because this area experiences
more wetdry cycles per unit time that do other parts of the
landscape.
At wet edges, soils also change predictably but in the
opposite way. The wet soils directly downslope from a wet
edge are gray, low in iron, and have lost much of their clay to
weathering and translocation (Daniels and Gamble, 1967).
4.9.7
Summary
Soils are best examined and explained where they are viewed
fully within the landscape that they have formed in. To simply
examine a soil exposed in a pit face literally with blinders
on is to ignore perhaps the most important factor in that
soils development the landscape. In this chapter, the term
landscape includes all the components of slope: gradient,
curvature, permeability, and depth to water table, among
Ahnert, F., 1970. An approach towards a descriptive classification of slopes.

Zeitschrift fur Geomorphologie 9, 7184.
Alvarado, A., Buol, S.W., 1975. Toposequence relationships of Dystrandepts in
Costa Rica. Soil Science Society of America Proceedings 39, 932937.
Baldwin, M., Kellogg, C.E., Thorp, J., 1938. Soil classification. In: Kellogg, C.E.
(Ed.), Soils and Men. Yearbook of Agriculture, USDA, US Govt. Printing Office,
Washington, DC, pp. 9791001.
Bell, J.C., Richardson, J.L., 1997. Aquic conditions and hydric soil indicators for
Aquolls and Albolls. In: Aquic Conditions and Hydric Soils: The Problem Soils.
Soil Science Society of America Special Publication 50, 2340.
Berger, I.A., Cooke, R.U., 1997. The origin and distribution of salts on alluvial fans
in the Atacama Desert, Northern Chile. Earth Surfaces Processes and Landforms
22, 581600.
Bravard, S., Righi, D., 1991. Characterization of fulvic and humic acids from
an OxisolSpodosol toposequence of Amazonia, Brazil. Geoderma 48,
151162.
Brown, D.J., Clayton, M.K., McSweeney, K., 2004. Potential terrain controls on soil
color, texture contrast and grain-size deposition for the original catena landscape
in Uganda. Geoderma 122, 5172.
Bushnell, T.M., 1942. Some aspects of the soil catena concept. Soil Science
Cremeens, D.L., Mokma, D.L., 1986. Argillic horizon expression and classification
in the soils of two Michigan hydrosequences. Soil Science Society of America
Journal 50, 10021007.
Dan, J., Yaalon, D.H., 1964. The application of the catena concept in studies of
pedogenesis in Mediterranean and desert fringe regions. Transactions of the 8th
International Congress of Soil Science 83, 751758.
Dan, J., Yaalon, D.H., 1982. Automorphic saline soils in Israel. Catena Supplement
1, 103115.
Daniels, R.B., Gamble, E.E., 1967. The edge effect in some Ultisols in the North
Carolina coastal plain. Geoderma 1, 117124.
Daniels, R.B., Gamble, E.E., Cady, J.G., 1971a. The relation between
geomorphology and soil morphology and genesis. Advances in Agronomy 23,
5188.
Daniels, R.B., Gamble, E.E., Nelson, L.A., 1971b. Relations between soil
morphology and water-table levels on a dissected North Carolina coastal plain
surface. Soil Science Society of America Proceedings 35, 781784.
De Alba, S., Lindstrom, M., Schumacher, T.E., Malo, D.D., 2004. Soil landscape
evolution due to soil redistribution by tillage: a new conceptual model of soil
catena evolution in agricultural landscapes. Catena 58, 77100.
Evans, C.V., Franzmeier, D.P., 1986. Saturation, aeration, and color patterns in a
toposequence of soils in north-central Indiana. Soil Science Society of America
Journal 50, 975980.
Gregorich, E.G., Anderson, D.W., 1985. Effects of cultivation and erosion on soils
of four toposequences in the Canadian prairies. Geoderma 36, 343354.
Hall, G.F., 1983. Pedology and geomorphology. In: Wilding, L.P., Smeck, N.E., Hall,
G.F. (Eds.), Pedogenesis and Soil Taxonomy. I. Concepts and Interactions.
Holliday, V.T., 2006. A history of soil geomorphology in the United States.
In: Warkentin, B.P. (Ed.), Footprints in the Soil. People and Ideas in Soil
History. Elsevier Press, Amsterdam, pp. 187254.
Huggett, R.J., 1975. Soil landscape systems: a model of soil genesis. Geoderma
13, 122.
Huggett, R.J., 1976. Lateral translocations of soil plasma through a small valley
basin in the Northaw Great Wood, Hertfordshire. Earth Surface Processes and
Landforms 1, 99109.
Jenny, H., 1941. Factors of Soil Formation. McGraw-Hill, New York, NY, 281 pp.
Catenas and Soils
Johnson, D.L., 1985. Soil thickness processes. In: Jungerius, P.D. (Ed.), Soils and
Geomorphology Catena. Catena Verlag, Braunschweig, Supplement 6, pp. 2940.
Johnson, D.L., Keller, E.A., Rockwell, T.K., 1990. Dynamic pedogenesis: new views
on some key soil concepts, and a model for interpreting Quaternary soils.
Johnson, D.L., Watson-Stegner, D., 1987. Evolution model of pedogenesis. Soil
Science 143, 349366.
Khan, F.A., Fenton, T.E., 1994. Saturated zones and soil morphology in a
Mollisol catena of central Iowa. Soil Science Society of America Journal 58,
14571464.
King, G.J., Acton, St. D.F., Arnaud, R.J., 1983. Soillandscape analysis in relation
to soil distribution and mapping at a site within the Weyburn association.
Canadian Journal of Soil Science 63, 657670.
King, L.C., 1957. The uniformitarian nature of hillslopes. Transactions of the
Edinburgh Geological Society 17, 81102.
Kleiss, H.J., 1970. Hillslope sedimentation and soil formation in northeastern Iowa.
Soil Science Society of America Proceedings 34, 287290.
Mackintosh, E.E., van der Hulst, J., 1978. Soil drainage classes and soil water
table relations in medium and coarse textured soils in southern Ontario.
Canadian Journal of Soil Science 58, 287301.
Malo, D.D., Worcester, B.K., Cassel, D.K., Matzdorf, K.D., 1974. Soil-landscape
relationships in a closed drainage system. Soil Science Society of America
Proceedings 38, 813818.
McDaniel, P.A., Falen, A.L., 1994. Temporal and spatial patterns of episaturation in a
Fragixeralf landscape. Soil Science Society of America Journal 58, 14511457.
Milne, G., 1932. Conference of East African Soil Chemists Technical Conferences of
the East African Dependencies. Amani, Tanganyika Territory. Government Printer,
Nairobi.
Milne, G., 1935a. Some suggested units for classification and mapping, particularly
for East African soils. Soils Research. Supplements to the Proceedings of the
International Society of Soil Science (Berlin) 4, 183198.
Milne, G., 1935b. Composite units for the mapping of complex soil associations.
Transactions of the 3rd International Congress of Soil Science 1, 345347.
Mokma, D.L., Sprecher, S.W., 1994. Water table depths and color patterns in
Spodosols of two hydrosequences in northern Michigan, USA. Catena 22,
275286.
Moore, I.D., Gessler, P.E., Nielsen, G.A., Peterson, G.A., 1993. Soil attribute
prediction using terrain analysis. Soil Science Society of America Journal 57,
443452.
Moore, I.D., Grayson, R.B., Ladson, A.R., 1991. Digital terrain modeling a review
of hydrological, geomorphological, and biological applications. Hydrological
Processes 5, 330.
Nizeyimana, E., Bicki, T.J., 1992. Soil and soillandscape relationships in the north
central region of Rwanda, east-central Africa. Soil Science 153, 225236.
Pei, T., Qin, C.Z., Zhu, A.X., Yang, L., Luo, M., Li, B.L., Zhou, C.H., 2010. Mapping
soil organic matter using the topographic wetness index: a comparative study
based on different flow-direction algorithms and kriging methods. Ecological
Indicators 10, 610619.
Pennock, D.J., 2003. Terrain attributes, landform segmentation, and soil
redistribution. Soil Tillage Research 69, 1526.
Peterson, F.F., 1980. Holocene desert soil formation under sodium salt influence in
a playa-margin environment. Quaternary Research 13, 172186.
Richardson, J.L., Arndt, J.L., Freeland, J., 1994. Wetland soils of the prairie
potholes. Advances in Agronomy 52, 121171.
Riecken, F.F., Poetsch, E., 1960. Genesis and classification considerations of some
prairie-formed soil profiles from local alluvium in Adair County, Iowa.
Proceedings of the Iowa Academy of Science 67, 268276.
157
Ruhe, R.V., 1960. Elements of the soil landscape. Transactions of the 7th
International Congress of Soil Science, Madison, WI 4, 165170.
Ruhe, R.V., Walker, P.H., 1968. Hillslope models and soil formation. I. Open
systems. Transactions of the 9th International Congress of Soil Science 4,
551560.
Schaetzl, R.J., Anderson, S.N., 2005. Soils: Genesis and Geomorphology.
Schaetzl, R.J., Krist, Jr. F.J., Stanley, K.E., Hupy, C.M., 2009. The Natural Soil
Drainage Index: an ordinal estimate of long-term, soil wetness. Physical
Schlichting, E., Schweikle, V., 1980. Interpedon translocations and soil
classification. Soil Science 130, 200204.
Simonson, R.W., 1959. Outline of a generalized theory of soil genesis. Soil Science
Smeck, N.E., Bigham, J.M., Guertal, W.F., Hall, G.F., 2002. Spatial distribution of
lepidocrocite in a soil hydrosequence. Clay Minerals 37, 687697.
Soil Survey Manual. US Government Printing Office, Washington, DC, USDA
Handbook No. 18. 437 pp.
Soil Survey Staff, 1951. Soil Survey Manual. USDA Agriculture Handbook 18. US
Government Printing Office, Washington, DC.
Sommer, M., Schlichting, E., 1997. Archetypes of catenas in respect to matter a
concept for structuring and grouping catenas. Geoderma 76, 133.
Vepraskas, M.J., 1999. Redoximorphic Features for Identifying Aquic Conditions.
North Carolina Agricultural Research Service Technical Bulletin 301. North
Carolina State University, Raleigh, NC.
Vepraskas, M.J., Guertal, W.R., 1992. Morphological indicators of wetness.
In: Kimble, J.M. (Ed.), Proceedings of the Eighth International Soil Correlation
Meeting (VIII ISCOM): Characterization, Classification, and Utilization of Wet
Soils. USDA-SCS, National Soil Survey Center, Lincoln, NE, pp. 307312.
Vepraskas, M.J., He, X., Lindbo, D.L., Skaggs, R.W., 2004. Calibrating hydric soil
field indicators to long-term wetland hydrology. Soil Science Society of America
Journal 68, 14611469.
Vreeken, W.J., 1973. Soil variability in small loess watersheds: clay and organic
carbon content. Catena 1, 181196.
Walker, P.H., Ruhe, R.V., 1968. Hillslope models and soil formation. II. Closed
systems. Transactions of the 9th International Congress of Soil Science 4,
561568.
Weitkamp, W.A., Graham, R.C., Anderson, M.A., Amrhein, C., 1996. Pedogenesis of
a vernal pool EntisolAlfisolVertisol catena in southern California. Soil Science
Society of America Journal 60, 316323.
Western, A.W., Grayson, R.B., Bloschl, G., Willgoose, G.R., McMahon, T.A., 1999.
Observed spatial organization of soil moisture and its relation to terrain indices.
White, E.M., 1964. The morphologicalchemical problem in solodized soils. Soil
Science 98, 187191.
Wieder, M., Yair, A., Arzi, A., 1985. Catenary soil relationships on arid hillslopes.
Catena Supplement 6, 4157.
Wood, A., 1942. The development of hillside slopes. Proceedings of the Geological
Association 53, 128138.
Yang, X., Chapman, G.A., Gray, J.M., Young, M.A., 2007. Delineating soil landscape
facets from digital elevation models using compound topographic index in a
geographic information system. Australian Journal of Soil Research 45, 569576.
Zhao, Z.Y., Yang, Q., Benoy, G., Chow, T.L., Xing, Z.S., Rees, H.W., Meng, F.R.,
2010. Using artificial neural network models to produce soil organic carbon
content distribution maps across landscapes. Canadian Journal of Soil Science
90, 7587.
158
Catenas and Soils
Biographical Sketch
Randall J. Schaetzl is professor of geography at Michigan State University. His research interests span the fields of
soils and geomorphology, with particular emphasis on landscapes of the Great Lakes region. Quantification of soil
development and use of GIS in soil-landscape analysis has been a recent research focus. He is the lead author on a
research-grade soils textbook, Soils Genesis and Geomorphology, published by Cambridge University Press, and the
editor-in-chief of Michigan Geography and Geology, a Pearson Custom Publishers publication.
4.10 Weathering and Hillslope Development

P Migon, University of Wroc"aw, Wroc"aw, Poland
4.10.1
4.10.2
4.10.2.1
4.10.2.2
4.10.2.3
4.10.3
4.10.3.1
4.10.3.2
4.10.3.3
4.10.4
4.10.4.1
4.10.4.2
4.10.4.3
4.10.5
4.10.5.1
4.10.5.2
4.10.5.3
4.10.6
References
Introduction
Fundamentals
Weathering-Limited and Transport-Limited Slopes
Short-Term and Long-Term Controls and Feedback
Working Definitions
Weathering and Rock Slope Evolution
Strength of Weathered Rock Masses
Weathering-Induced Rock Slope Failures
Caprock Failures above Weathered Base
Deep Weathering and Landslides
Deep Weathering Profiles and their Properties
Landslides in Weathered Terrains
Geomorphic Signatures of Mass Movements in Weathered Materials
Weathering and Slope Landforms
Boulders and Boulder Fields
Tors
Flared Slopes
Conclusions
159
160
160
160
161
161
161
163
164
166
166
168
168
170
170
173
176
176
177
Abstract
Weathering plays an important role in the evolution of hillslopes. It decreases strength of a rock mass and hence contributes
to slope failures by fall, slide, or topple. Accelerated weathering of soft rock under hard caprock disturbs slope equilibrium
and results in long-term escarpment retreat. In deeply weathered terrains, patterns of mass movements are directly related to
the weathering grade. Minor mid-slope landforms such as boulders, tors, and flared slopes are in most instances products
of differential weathering and removal of weathering products. The significance of weathering is conceptualized in the
traditional distinction between weathering-limited and transport-limited slopes.
4.10.1
Introduction
Weathering plays a key role in hillslope evolution, and the

relationships between rock alteration processes, gravity-driven
processes, and slope morphology are manifold. First, weathering weakens the rock and decreases its strength, which
results in the lowering of slope gradient. Therefore, in weathered landscapes slope surfaces are generally less inclined and
topography more subdued than in areas built of the same kind
of rock, but little weathered. Second, weathering acts primarily
along discontinuities (joints, bedding planes, faults, cleavage
surfaces, etc.), loosening an otherwise tight rock mass. Depending on slope inclination, liberated discontinuity-bound
rock fragments may remain in place, contributing to the
development of a debris-mantled slope or may move downslope through fall or slide. In specific instances, advanced
alteration along a major discontinuity surface may set a stage
Migon, P., 2013. Weathering and hillslope development. In: Shroder, J.
Press, San Diego, CA, vol. 4, Weathering and Soils Geomorphology,
pp. 159178.
for a large-scale rockslide. Hence, weathering is an important

process to shape rockslope morphology. Third, deep weathering changes geomechanical characteristics throughout
rock mass and contributes to an increased susceptibility to
landsliding. Spatial and temporal patterns of landslides are
commonly related to the patterns of deep weathering. This is
why, with due caution, the latter can be used to predict the
former. Fourth, at the local scale, it is the slope morphology
that influences the patterns of weathering and soil formation.
Mid-slope hollows and footslopes focus water and are moist
microenvironments, within which the efficacy of weathering
processes is expected to be higher. Fifth, weathering and
evacuation of weathering products are directly responsible for
the origin of different hillslope landforms, among which solid
rock residuals tors are the most obvious example.
This chapter focuses on two tightly related subjects. On the
one hand, weathering control on hillslope processes will
be reviewed. An argument will be developed that an understanding of preconditioning of bedrock by weathering is crucial in any attempts to explain and predict the pathways of
slope evolution. On the other hand, landforms resulting from
in situ weathering and subsequent displacement of weathered
http://dx.doi.org/10.1016/B978-0-12-374739-6.00075-0
159
160
Weathering and Hillslope Development
material will be shown in the context of overall hillslope

morphology. It is also to be argued that, in retrospect, these
landforms can be used to infer the role of weathering in
shaping hillslopes in the long term.
4.10.2
4.10.2.1
Fundamentals
Weathering-Limited and Transport-Limited
Slopes
Classic models of hillslope evolution, such as those by W.M.

Davis or L. King, did include weathering as one of the controlling parameters, but were more preoccupied with the
general change of form through time rather than involved
detailed analysis of the effects of weathering on process and
form. It is fair to say that Davis (1899) envisaged a progressive
increase in the depth of waste cover and fining of its texture,
which in turn should have led to the reduction of slope gradient. However, as pointed out by Parsons (1988) and many
others, his scenario, as well as other general models of the
time, were based on intuition and reasoning rather than any
empirical data, particularly about the processes involved. The
major conceptual advance in considering weathering as a
factor in slope evolution was the distinction between weathering-limited hillslopes and transport-limited hillslopes,
formulated in full in the 1970s (Carson and Kirkby, 1972), but
traceable back to the pioneer work of G.K. Gilbert from the
late nineteenth century. Weathering-limited hillslopes are defined as those where the rate of weathering controls, through
the rate of supply of loose material available for transport, the
evolution of slope form through time (Figure 1(a)). In an
extreme case, if no material is made available by weathering,
no change takes place. Some bare slopes built of very massive
and resistant rock may be considered as examples. Transportlimited hillslopes, by contrast, are those where the efficacy
of transport processes controls the rate of change. One can
imagine a slope underlain by a thick mantle of weathered
material, yet low-angle and heavily vegetated. Under these
conditions, surface processes are severely suppressed and
no change may occur despite huge material availability.
However, transport limitations have to be considered with
respect to appropriate timescales. A mantled slope may remain
stable over short term, but fail through an episodic highmagnitude geomorphic event such as landslide (Figure 1(b)),
triggered by heavy rainfall or an earthquake.
Parsons (1988) criticized such a dichotomy and argued
that these are all but two end-members of a broad spectrum.
Instead, he suggested to consider detachment control and
offered a detachability continuum approach. The rate of slope
change is a function of both material availability and transport
capacity. Interestingly, an increased rate of weathering does
not necessarily increase material availability for transport. This
is because of pedogenic processes that result in the formation
of soil crusts and cohesive clays. Rates of weathering and the
thickness of regolith, critical to establish the status of a hillslope, whether weathering- or transport-limited, are the key
ingredients of diffusive mathematical models of hillslope
development (Carson and Kirkby, 1972), which again can be
traced back to the work of Gilbert (1909) and his concept of
(a)
(b)
Figure 1 Different types of slopes and different weathering

processes acting on them. (a) Bare rock slope of a granite inselberg
in the Namib Desert, with physical weathering dominant. (b) Mantled
and vegetated slope underlain by chemically altered saprolite, Serra
do Mar, Brazil. Rock slope failures are common in the first example,
whereas deep-seated slides occur on the latter.
steady state in landform evolution. However, although the

diffusion models work well in soil-mantled landscapes, their
application to slopes underlain by thick saprolitic covers
is more problematic because saprolite thickness is unlikely
to remain constant over time (Phillips, 2010). Although the
thickness of regolith can be measured relatively easily, to establish the rates of weathering front advance is much more
challenging. Measurements of cosmogenic radionuclides such
as 10Be in saprolites and bedrock (Dixon et al., 2009) should
be helpful in extending models of slope evolution to areas
with thick weathering mantles.
4.10.2.2
Short-Term and Long-Term Controls and

Feedback
It is important to distinguish between short-term and longterm relationships between form and process in hillslope
evolution. Many approaches to processresponse modeling
are underlain by an assumption that form is controlled by
process. In fact, although this stance is defensible for long
timescales, it is not necessarily correct in short timescales.
With respect to the latter, it is generally observed that occurrences, rates, and effects of surface processes are controlled by
the preexisting slope morphology, including factors such as

steepness, length, curvature, and others. Relationships between hillslope form and weathering patterns aptly illustrate
these feedbacks. Since water availability is critical to most
weathering processes, sites that collect more water may be
expected to undergo faster and more efficient weathering.
Conversely, surfaces that shed water will remain less affected.
In the context of hillslope morphology, two settings warrant
particular attention: footslopes and mid-slope hollows.
Enhanced footslope weathering is hypothesized for a range
of spatial scales, from individual rock outcrops (Twidale and
Campbell, 1992) to large intramontane basins (Bremer,
1975). Pedestal rocks, flared slopes of inselbergs, marginal
troughs around residual hills and specific relationships within
some basins, with some residual relief in the middle, are
offered as corroborating geomorphic evidence. In certain areas
it was possible to demonstrate that bedrock is indeed weathered to a greater depth at the slope/plain junction. For example, such distribution patterns have been documented for
deeply disintegrated granites grus (Migon and Thomas,
2002).
Slope hollows (or zero-order basins) generate attention
mainly because they commonly become loci of landslides
(Dietrich et al., 1986; Fernandes et al., 2004). They may vary
in origin. Some may be inherited from the Pleistocene, but
others owe their form to recent landslides in weathered bedrock. The resultant depressions influence subsequent water
and sediment movement on slopes. Not only is there gradual
filling of the hollows by material brought in by creep, wash,
and shallow sliding, but weathering continues under hollows
due to runoff diversion from the surrounding slopes. The next
landslide removes the fill and reestablished saprolite, deepening the hollow and producing scalloped hillslope topography. Geomorphology of slope hollows is discussed more
extensively in Section 4.10.4.
4.10.2.3
Working Definitions
Relationships between weathering and hillslope evolution

have to be considered separately for rock slopes, where weathering is nonsaprolitic, and mantled slopes, where thick
weathered mantles may form (Figure 1). These two basic
slope categories are defined as follows.
Rock slopes are inclined surfaces built of solid rock, with
no regolith and soil cover, except for some localized discontinuous patches. However, individual boulders may rarely
rest on solid rock pedestals if slope angle allows them to remain stable. Rock slopes may be vertical or even overhanging,
as is common along coastal cliffs or in glaciated high mountains, but they may have a very low angle (less than 101), too.
Likewise, their height may vary by a few orders of magnitude,
from a mere few meters to more than 500 m. Rock slopes are
typically associated with aprons of material derived from rock
face and accumulated at the base of the slope, termed scree or
talus. For a more extensive discussion of rock slopes, the
reader is directed to Selby (1993).
By contrast, mantled slopes are those within which
unweathered bedrock outcrops are either scarce or entirely
absent and the topographic surface is cut across unlithified
161
materials. These in turn may be derived straight from

in situ weathering, subject to limited downslope movement
(colluvium) or completely allochthonous with respect to
bedrock, brought in by aeolian or glacial processes. The
thickness of the mantle, also traditionally called a regolith,
may be very variable, from less than a meter to more than
100 m. Taylor and Eggleton (2001) discussed different aspects
of regolith geology, geochemistry, and geomorphology.
In the following sections, rock slopes and types of failures
associated with them are presented first, followed by a discussion of landslides in weathered materials.
4.10.3
4.10.3.1
Weathering and Rock Slope Evolution

Strength of Weathered Rock Masses
Weathering is an important factor in the evolution of rock

slopes, as they get set up for various types of slope failures.
Generally, failures occur when applied stress exceeds the
strength of the material and this may be achieved in two ways.
Strength may remain constant but stress increases, either instantaneously or gradually, leading to fall, slide, or topple
from a rock slope. Typical situations include earthquakes,
during which transient stresses induced by seismic waves travelling in the shallow subsurface are so high that previously
stable slopes fail (Keefer, 1984). Progressive fluvial undercutting of a rock slope at a meander bend may also result in an
ultimate crossing of tensile strength threshold.
Conversely, stress may remain approximately constant but
strength decreases until it falls beneath the threshold of stability. Reasons for decreasing material strength may vary.
Short-term or seasonal changes in porewater pressure, related
to rainfall distribution in time, impact on rock mass strength
and many slope failures occur either in association with heavy
rainfall episodes (e.g., arrival of typhoons) or during more
prolonged rainy seasons. Weathering is another factor contributing to diminishing rock strength over time and it does so
through a number of interactions with the rock:
weakening of fabric by boundary and intra-granular crack

development, repetitive volume changes arising from thermal expansion/contraction and hydration/dehydration;
dissolution of cement, most evident if cement is carbonate,
but siliceous and clayey cement are affected too, albeit at a
slower rate;
dissolution of soluble minerals such as calcite or gypsum;
preferential rock alteration along fractures, particularly
along fault lines, where fault gouges and breccia are particularly prone to weathering and clayey material may be
susceptible to swelling; and
joint opening and enlargement due to salt and freezethaw
weathering, both involving alien crystal growth (salt,
water/ice) in the voids and cracks.
Weathering operates at different scales (Viles, 2001) but it

is the scale of a rock mass that is most relevant to rock slope
failures. Characteristics of a rock mass include, in addition to
lithology (mineralogy, fabric, and arrangement of mineral
constituents) and resultant intact strength, the presence of
discontinuities of different origin and age, their geometric
162
properties, and the presence of water (Figure 2). To account

for all these features, different approaches have been proposed
to obtain a reasonably complete picture of rock mass strength.
One of these is the rock mass rating (RMR) developed by
Bieniawski (1989) and particularly relevant to geotechnical

applications (Table 1), where it is used widely. It is less useful
in geomorphological research, as it involves determination of
intact rock strength through geotechnical experimental tests
Regular joint spacing

Irregular joint spacing
Slickensided
joint
Joint surfaces may be

smooth or rough
Joints of variable
orientation
Discontinuous
joints
Presence of
groundwater
Infilled joint
Figure 2 Rock mass properties affecting rock mass strength.
Table 1 Rock mass rating system

Parameter
Range of values
Intact rock strength

(uniaxial compressive strength, in
MPa)
Rating
4250
100250
50100
2550
525
15
o1
15
12
Drill core quality

Rating
90100%
20
7590%
17
5075%
13
2550%
8
o25%
3
Discontinuity spacing
Rating
4200 cm
20
60200 cm
15
2060 cm
10
620 cm
8
o6 cm
5
Conditions of
discontinuities
Very rough
surfaces
Slightly rough
surfaces
Slightly rough
surfaces
Soft gouge 45 mm thick
Not continuous
Separation
o1 mm
Slightly weathered walls
Separation
o1 mm
Highly weathered walls
Slickensided
surfaces or
gouge
o5 mm thick
Separation
15 mm
Continuous
Continuous
Rating
Unweathered
wall rock
30
25
20
10
Groundwater
Rating
Completely dry
15
Damp
10
Wet
7
Dripping
4
Flowing
0
No separation
Source: Based on Bieniawski, Z.T., 1989. Engineering Rock Mass Classifications. Wiley, New York, 251 pp.
Separation 45 mm
163
Table 2 Rock mass strength classification

Parameter
Strength
Very strong rock
Strong rock
Moderately strong
rock
Weak rock
Very weak
rock
Intact rock strength (R-value

determined by N-type
Schmidt hammer)
Partial rating
10060
6050
5040
4035
3510
20
18
14
10
Degree of weathering
Unweathered
Slightly weathered
Moderately weathered
Partial rating
10
Highly
weathered
5
Completely
weathered
3
Joint spacing
Partial rating
43 m
30
13 m
28
0.31 m
21
0.050.3 m
15
o 0.05 m
8
Joint orientation
Moderate dip into the

slope
18
Horizontal or nearly
vertical (in hard
rock)
14
Moderate dip
out of the
slope
9
Steep dip out

of the slop
Partial rating
Steep dip into the

slope; cross joints
interlock
20
Joint width
Partial rating
o 0.1 mm
7
0.11 mm
6
15 mm
5
520 mm
4
4 20 mm
2
Joint continuity and infill
None continuous
Few continuous
Continuous, no infill
Partial rating
Continuous,
thin infill
4
Continuous,
thick infill
1
Groundwater outflow
None
Trace
Slight (o25 l per min

per 10 m2)
Partial rating
Moderate
(25125 l
per min
per 10 m2)
3
Great
(4125 l
per min
per 10 m2)
1
Total rating
91100
7190
5170
2650
o26
Source: Based on Selby, M.J., 1980. A rock-mass strength classification for geomorphic purposes: with tests from Antarctica and New Zealand. Zeitschrift fur Geomorphologie N.F.
24, 3151.
(uniaxial compressive strength or point-load strength) and

rating of drill core quality, assessed on the basis of the degree
of fracturing recorded in a drill core.
To overcome both inconveniencies, Selby (1980) modified
the RMR approach and suggested a rock mass strength (RMS)
approach instead, which can be applied without any expensive
equipment, except the Schmidt hammer (see Day and Goudie,
1977; Goudie, 2006). The RMS approach consists of in-parallel,
semiquantitative determination of seven parameters and their
subsequent rating to obtain a numerical value characterizing
the rock mass (Table 2). The final score is dimensionless but
allows for comparison of different rock masses. Among the
parameters taken into account, some have direct relationship
with the susceptibility of rock to weathering and the current
state of weathering. Intact strength reflects the fabric of the
rock and the presence of intergranular voids and hence, how
prone to weakening of the fabric a given rock type is. The
weathering parameter is assessed qualitatively; using characteristics of weathering grade (see Table 3) and describes the
current state of the near-surface portion of the rock. Joint
spacing and orientation are inherent features of a rock
mass, but both joint width and continuity/infill are related
to weathering. With the progress of weathering, fractures become more open, adjacent cracks may join, and products of
jointwall alteration accumulate along discontinuities. In

effect, cohesion and friction angle decrease and the shear
strength of joint surfaces diminishes markedly in comparison
to intact rock, especially if infilling clays are smectites, which
have high water-adsorption capacity. Altogether, the RMS rating for weathered rock may be significantly lower than that for
its unweathered counterpart, particularly if dealing with rather
weak, densely jointed rock masses.
4.10.3.2
Weathering-Induced Rock Slope Failures
Decrease of rock mass strength may lead to rock slope failures

and these may be of different types, broadly divided into two
groups. One involves detachment of individual rock fragments, one by one, and their subsequent movement downslope, by free fall, sliding, and/or bouncing, depending on
slope angle and available energy. The detachment itself usually
occurs as a result of expansion of salt, ice, or plant roots in
cracks and joints. Sorted scree aprons below rock faces build
up in this way. The second group includes mass movements
involving simultaneous transfer of large rock masses, from
a few cubic meters to thousands of cubic meters in volume
(e.g., Wieczorek, 2002). These large-scale rock slope failures
164
Table 3 Weathering grades for deeply weathered rock masses

Ruxton and Berry (1957)
Geological Society (1997)
Description
Formal name
Weathering
zone code
Weathering
zone code
Formal name
Description
Structureless sandy clay or

clayey sand, up to 30% of
clay, dominance of quartz
and kaolin
Few rounded corestones,
less than 10% of the
section area, less than 5%
of clay
Much of the original rock
structure preserved,
corestones occupy 10 to
50% of the section area
Residual debris
VI
Residual soil
Residual debris with

corestones
IIa
Completely
weathered
IIb
IV
Highly
weathered
All rock material converted to

soil; original structure and
fabric destroyed; large
change in volume
All rock decomposed and
disintegrated but original
structure and fabric still
largely intact
More than 50% of rock
decomposed and
disintegrated; fresh or
weakened rock locally
present, including as
corestones
Less than 50% of rock
decomposed and
disintegrated; rock
discolored and partly
weakened; corestones may
be common
Discoloration as the main
sign of change, mainly
along discontinuities; some
weakening of fabric
Slight discoloration along
major discontinuities
No visible signs of
weathering
Dominance of rectangular
corestones (5090%),
surrounded by residual
matrix
Corestones with
residual debris
III
III
Moderately
weathered
More than 90% is solid rock,

minor amounts of debris
along fractures, iron
staining may be present
No visible signs of rock
weathering
Partially weathered
rock
IV
II
Slightly
weathered
IB
Faintly
weathered
Fresh
Bedrock
IA
are classified as rock slides, rock-mass falls, rock avalanches,

and topples (Selby, 1993). Although the geomorphic impact
and resultant landforms may be similar chaotically arranged
piles of boulders mantling the lower slope and extending onto
valley floors or pediments, the contribution of weathering to
each type of movement is different.
Rock slides occur along major preexisting discontinuity
surfaces and involve sliding movement of the rock mass above
this boundary surface. Failure planes are generally structural
surfaces, such as schistosity planes in metamorphic rocks,
bedding planes in sedimentary rocks, or sheeting surfaces in
igneous rocks, inclined down the slope and broadly parallel to
the topographic surface. Shear strength along partings is a
critical parameter and rock slides are initiated when its value is
significantly reduced. Reduction of shear strength often results
from various weathering processes: secondary clay accumulation within joints, dissolution along joint surfaces, or disintegration of asperities.
Rock-mass falls are displacements of large volumes of
rock, which may be internally fractured themselves, but are
separated from the rest of the rock mass by a major failure
surface, commonly stepped, following many individual joint
surfaces. Reasons for such massive failures are generally sought
in processes other than weathering, such as seismic triggers,
elevated cleft-water pressures, and propagation of stress release
fractures parallel to cliff faces. However, weathering may be a
contributing factor, especially if it is concentrated at the base

of rock slope. This is common in arid areas, where slope/plain
junctions are the sites of enhanced salt weathering. In an inselberg terrain around Al-Quwayra in south Jordan, Goudie
et al. (2002) documented huge alcoves, up to 10 m high and
50 m wide, and showed their increased levels of salinity. At the
same time, rock slopes of the sandstone inselbergs bear evidence of multiple rock-mass falls, some involving detachment
from a free face as much as 80 m high. Young et al. (2009)
discussed at length reasons of rock slope failures in sandstones
and noticed substantial strength reduction with an increasing
content of water. By whatever reason, in poorly fractured rock
masses, with one joint set parallel to the rock face, undercutting by weathering may lead to massive slab failures.
4.10.3.3
Caprock Failures above Weathered Base
In the examples described above it is ongoing weathering of

rock mass building the rock slope that ultimately brings it to
cross the stability threshold. The failures just discussed occur
within lithologically homogeneous rock slopes or their segments, although failed masses may travel far beyond such
segments. A separate group concerns rock slopes in layered
structures, where a massive, resistant top rock layer (caprock)
overlies mechanically much weaker rock units below. This
165
Figure 3 Hard caprock and weaker, weathering-prone rock units below, partially mantled with products of caprock failures. Near Moab,
Utah, USA.
Figure 4 Enhanced weathering in silty layers beneath conglomerate caprock and unstable overhangs, Danxiashan, south China.
situation is typical in dissected plateaus developed on ancient

sedimentary basins, where massive well-cemented sandstones
or limestones occur above poorly lithified sandstones, shales,
or calcarenites, creating a considerable geomechanical contrast
(Figure 3). Similar conditions may occur in volcanic terrains,
where lava flows or ignimbrites (welded tuffs) overlie poorly
consolidated pyroclastic deposits or weak sedimentary layers.
Generally, caprock supports very steep, even vertical slopes,
which grade into much less inclined surfaces (15351) in
underlying units. In these cases, in addition to weathering of
caprock itself, it is alteration of the weaker units below which
influences caprock slope evolution. Weathering mechanisms

operating under caprock may be variable and site specific. The
presence of water is essential and a perched aquifer generally
exists beneath permeable, jointed caprock. Weak units are not
uncommonly clayey and much less permeable, hence forced
subsurface drainage toward escarpments and seepage at the
caprock/underlying rock boundary occurs, which may be very
efficient (Howard and Kochel, 1988; Oberlander, 1989). Repetitive wetting and drying, freezing and thawing, dissolution
and clay alteration disaggregate the subcaprock unit and rearrange its structure, weakening support for caprock itself. All
166
these processes cause slope recession in the weaker unit, resulting in the development of an overhang in caprock. This
will remain temporarily stable, but eventually tensile stresses
will exceed the critical strength and rock-mass fall will occur
(Figure 4).
Consequences for caprock are different if the weaker clayey
units beneath it deform plastically and show flow-like behaviour. They may do so under the weight of caprock and
because of an increase in water content and pore-water pressure. As the clays are squeezed out, jointed caprock compartments begin to separate from each other, change position, and
lose stability. The most common type of movement of caprock
blocks is forward toppling, but backward toppling (sagging)
may occur as well. In extreme examples, long-distance, kilometer-scale gliding of caprock blocks over deformable clays
and siltstone is possible. Instances of mass movements along
caprock escarpments affected by substratum weathering and
deformation are known from all around the world, including
the Colorado Plateau, the Sydney Basin, the Isle of Portland
in Great Britain, Ethiopian Highlands, Bohemian Massif, and
many other localities (see Oberlander, 1989; Young et al.,
2009).
4.10.4
4.10.4.1
Deep Weathering and Landslides

Deep Weathering Profiles and their Properties
Landsliding is a ubiquitous geomorphic process and occurs in

many different settings and environments, but its particular
intensity is recorded in deeply weathered terrains (see Chapter
4.8). Thick weathered mantles abound in humid low latitudes
and it is these tropical and subtropical hilly to moderately
mountainous regions that are notorious for landslides of different types, not uncommonly disastrous in their effects
(Figure 5). The best known examples come from heavily urbanized, high-relief settings such as Hong Kong or Rio de
Janeiro, but in fact extensive tracts of countryside in Mexico,
the Caribbean region, Venezuela, eastern Brazil, equatorial
Africa, India, and southeast Asia are molded by landslides
developed in thick saprolites (Thomas, 1994). Away from the
intertropical belt, landslides in weathered materials occur too,
especially in easily weatherable mudstones and deeply disaggregated granite (grus). The nature of mass movement in
weathered rock depends on many factors, but primarily on the
extent of deep weathering and properties of weathered mass.
Therefore, basic characteristics of deep weathering profiles
need to be considered first.
The notion of deep weathering refers to instances of
thorough in situ disintegration and decomposition of parent
rock that penetrates throughout the rock mass to a depth of
at least a few meters, but generally much more than this.
For example, granite-derived grus is typically 520 m thick
(Migon and Thomas, 2002), whereas more decomposed clayrich saprolites may be up to 3040 m thick (Thomas, 1994;
Taylor and Eggleton, 2001). In stable cratonic settings, subject
to deep weathering for millions of years, complete weathering
profiles are commonly in excess of 100 m (Ollier, 2010). In the
context of mass movements it is important to note that deep
weathering is widespread in hilly, low-latitude landscapes, not
just under extensive tropical plains.
In places such as Serra do Mar mountains in southeast
Brazil or Hong Kong and the adjacent Guangdong province of
China thicknesses of weathered mantles more than 50 m have
been widely reported (Thomas, 1995; Shaw, 1997; Lan et al.,
2003). Despite considerable variability in thickness it has long
Figure 5 Damage to a building in effect of a slide in deeply weathered granite, south China. Photo courtesy of Xu Hongen.
been observed that weathering mantles commonly show clear

vertical differentiation, particularly well displayed in granite.
They are recognizable in the field, through the variable percentage of little weathered rock compartments (corestones),
visibility of original structural features such as joints, and
color change, and these differences generally correspond with
geochemical and mineralogical characteristics, with the percentage of neoformed minerals growing upward. An important structural feature is the weathering front that separates
fresh and weathered rock and may be very sharp, although a
gradational transition is more often observed. These observations led Ruxton and Berry (1957) to propose a formal
subdivision of granite weathering profiles, which included five
principal classes (Table 3; Figure 6). Later work refined their
concept and a sixfold division was proposed (Dearman et al.,
1978), which has been widely adopted (Geological Society,
1997; Table 3). Although devised for granite and allied rocks,
the concept of zonation of weathering profiles is universal and
applicable to all types of rocks, even if some horizons or
structural features may be poorly or not at all represented in
specific field examples. It is important to note that weathering
profiles are not necessarily complete, in the sense that all
structural horizons are present at any given location. These
situations are explained in two ways. First, weathered mantles
observed nowadays may be all but relicts of much thicker
profiles existing in the past, which have been partly destroyed
(beheaded) due to accelerated surface erosion. If this is the
case, we speak about truncation of a profile and the material
IIa
eroded can be found in adjacent sedimentary basins in the

form of clay-rich deposits. Second, weathering profiles may
have never reached the stage of an ideal, complete sequence
and they failed to do so because of two reasons. Either
environmental conditions were unsuitable for mineralogical
alteration to advance beyond a certain threshold (e.g., in cold
climate), or steepness of terrain and efficient surface erosion
prevent attaining higher level of mineralogical change. The
latter possibility appears to explain numerous instances
of apparently incomplete weathering profiles on igneous
rocks in steep tropical and subtropical terrains (e.g., Migon
and Alcantara-Ayala, 2008).
Not only structural differentiation and mineralogical
change is observed within deep weathering profiles. It has
been demonstrated that physical properties of saprolites
change too, bearing on the susceptibility of saprolitic materials to landsliding. Upper parts of weathering profiles are
typically more fine-grained, with the percentage of silt and clay
increasing. In mature saprolites clay-size material may account
for 7080% of the total by weight, with kaolinite present in
major quantities. In less advanced profiles, or in those subject
to truncation, the percentage of clay may be less and the
dominant clays may be from the illite and smectite families.
Further properties that change due to weathering are porosity
and permeability, crucial for the ability to transmit and hold
water within the saprolite. In originally massive rocks such as
granite, porosity may increase from 1% to 2% in fresh rock to
more than 25% in the completely decomposed one, whereas
in granular sedimentary rocks the respective change may
be from 510% to 3040%. Simultaneously, bulk density
decreases. Physical changes in turn influence geomechanical
properties of saprolites. The decrease of strength may be
demonstrated in the field, applying different crude tests for
material strength (using penetrometer, Schmidt hammer, and
the like), but is most evident if compressive strength is
measured experimentally (Figure 7). However, with respect to
saprolites, intact strength has to be distinguished from joint
strength. Weathered joint planes are not only structural
Max value
III
Uniaxial compressive strength (Mpa)
160
IIb
167
140
Mean value
120
Min value
100
80
60
40
20
IV
0
I
Figure 6 Zoning of weathering profile following coding system by
Ruxton and Berry (1957). Further explanation in the text and Table 3.
Based on Migon (2010).
II
III
IV
VI
Weathering grade
Figure 7 Decrease of rock intact strength due to weathering, based
on data from Arel and Tugrul (2001).
168
discontinuities within saprolites; but they are also often infilled by kaolinitic clays and the infill may be from less than a
millimeter to a few centimeters thick. For moderately decomposed granite and volcanic rocks, joint strength is only
0.50.8 of intact strength but both parameters equal for highly
decomposed rocks (Au, 1996). Significant decrease of strength
along relict joints points to structural discontinuities as
potential slip surfaces in saprolitic materials. In addition, relict
fractures are not uncommonly indurated or slickensided,
acting thereby as efficient barriers to groundwater circulation.
4.10.4.2
Landslides in Weathered Terrains
Reduction in strength of saprolitic materials with respect to

that of fresh rock makes the former susceptible for landsliding,
even if fresh rock is hardly ever affected by landslides, as is the
case of granites and massive volcanic rocks. However, the style
of landsliding varies and shows distinct relationship with the
weathering grade and properties of weathered rock masses
(Figure 8). Early stages of weathering, typified by alteration
focused along discontinuity planes and opening of major
fractures due to tensile stress (grades III, see Figure 6), are
associated with rock slides and rock falls, depending on local
slope gradient. In either case, shear strength of the rock along a
parting is significantly reduced (Au, 1996). Secondary clay
accumulations within dilation joints, which may be very thin
but laterally very extensive, also contribute to slope failures in
an otherwise massive rock. Catastrophic slope collapses on the
hillsides of granite or gneissic domes are typical gravitational
phenomena causally related to incipient weathering. At more
advanced stages of deep weathering (grade III), with a mantle
of grus (are`ne) or stony loam is present under the slope surface, shallow debris slides are preferentially generated. In terrains where the mantle of grus is thin (a few meters), and the
weathering front separating grus from solid rock is sharp,
slides may occur at this interface because the weathering front
provides a barrier to groundwater movement, resulting in the
development of ephemeral perched water-tables during rainstorms. Hence, the rather thin grus above the interface
Rotational
slides
Earthflows
IIa
Debris
slides
IIb
III
Debris flows
Corestones
Boulder
falls
Rock slope
failures
IV
becomes saturated and the pore-water pressure rises above the

limit of material stability. Numerous examples of slope instability in granite-derived grus have been described from
mountainous regions of Japan. Steep slopes and efficient
erosion do not allow for thick weathered mantles to develop
and grus is widespread instead. This material is fairly
stable when dry, but conditions drastically change during
heavy rainfall episodes brought in by seasonal tropical storms.
Grus becomes quickly saturated and begins to slide down
along available slip surfaces. Slope-parallel microsheeting
facilitates mass movement. In metamorphic rocks (gneiss,
schist), relict schistosity planes are ready slip planes, especially
if oriented parallel to the slope. Chigira (2001) observed that
in many instances landsliding was widespread in grus and less
extensive in more weathered material, which appears counterintuitive, given progressive decrease of rock mass strength
with weathering. However, two factors play a role here. First,
zones of localized weakness (relict joints, rock/grus interfaces), hence potential sliding surfaces, are common in grus
but poorly represented in more homogeneous clayey saprolites. Second, disintegrated granite, by virtue of its higher mass
strength, can support slopes that are steeper than those developed in fine-grained saprolites. These steep slopes are more
prone to abrupt yielding induced by an extreme precipitation
event, whereas gentler slopes underlain by heavily weathered
granites release stress through creep and shallow soil flow,
which occur in a steadier manner.
The style of landsliding changes again if weathered rock
reaches grades IV and V. These rock masses, with few corestones
left and significant percentages of secondary clays, are typically
affected by rotational slides (slumps). The development of
curved failure surfaces is facilitated by the homogenization of
the saprolite and disappearance of joint-related discontinuities.
The relationships between the type of landslide and weathering grade are further influenced by local relief. If a hillside
affected by mass movement, whether a debris slide or a slump,
is long and steep enough, and topographic and hydrological
conditions are conducive, then an initial slide may transform
into a flow-type phenomenon. Geomorphological settings
favoring the transformation are hillslope hollows and valley
floors occupied by streams present at the time of failure. After a
slide reaches a local watercourse, the content of water in the
sediment increases, the pore-water pressure builds up and the
sliding mass liquefies. Devastating debris flows that frequently
affect built-up footslope areas in Rio de Janeiro (Coelho Netto,
1999; Fernandes et al., 2004), Hong Kong (Ng et al., 2002), or
coastal Venezuela (Larsen and Wieczorek, 2006), are the result
of transformation of initial slides in weathered material and
their confinement within narrow, steep gradient valleys. In more
weathered terrains, colluvium from rotational slides may liquefy
and continue moving as earthflows, spreading wide over the
available footslope space. The wide depositional surfaces below
the landslide scars in multiconvex terrain are the product of this
type of transformation (Coelho Netto, 1999; Migon, 2009).
joint
traces
Figure 8 Generalized relationships between types of mass

movement and weathering grade in deeply weathered rocks. Based
on Migon (2010).
4.10.4.3
Geomorphic Signatures of Mass Movements in

Weathered Materials
The relationships between deep weathering patterns, mass

movement, and hillslope morphology are complex and
involve a range of feedback. The preexisting topography dictates the spatial pattern of mass movements, including the
location of slope failures, the travel path, and the run-out
distance. However, landslides and other types of mass movement play a major role themselves in shaping the topography
of weathered terrain, creating new landforms, and providing
new avenues for sediment transfer. In addition, they leave
geomorphic signatures, which may persist over long time,
helping to identify ancient landslides and hence to evaluate
landslide hazard. Thus, mapping the geomorphological record
of past mass movement is a valuable contribution to wider
hazard and risk assessment.
Geomorphological consequences of landslides in the
weathered terrain may be considered in terms of both hillslope form and process. Mass movement effects are essentially
erosional in the upper/middle section of a slope and depositional in the lower slope and in the valleys below, but will
differ depending on the type of mass movement taking place.
If weathered mantles are thin, shallow landslides are capable
of complete stripping of regolith and the former weathering
front is revealed as a rock slope facet (Figure 9). Landslide
scars formed in this way are not necessarily very deep but may
be laterally extensive, coalescing into large rock slope units,
and long-lived since reestablishment of vegetation and soil on
a bare steep rock surface may take tens, and perhaps hundreds
of years. Thus, slides within a saprolitic cover are important
contributors in episodic dome excavation from the weathered
mantle, consistent with the two-phase evolution of residual
hills (see Chapter 4.8). This model of dome evolution was
advocated by Hurault (1963) using evidence from Guyana
and further examples were offered from Serra do Mar in Brazil
(De Ploey and Cruz, 1979). Debris slides in the upper slope
may transform into debris flows that shape the morphology of
169
lower slopes and adjacent valleys floors. Landform inventories

produced by debris flows in deeply weathered terrain do not
differ from those produced by debris flows in other landscape
settings, for instance in high-mountain terrain. However, the
structure of associated levees reflects the nature of the saprolite
subject to stripping. If corestones were abundant in the weathered mantle, then boulder-rich levees would easily form,
with densely packed boulder ridges built by individual rounded blocks up to 45 m long as in the Veladero granite massif,
Mexico (Lugo et al., 2002).
In highly weathered hilly areas, where deeper rotational
slumps prevail over near-surface debris slides, hemispherical
slope hollows become the dominant landform. The density of
such hollows may be considerable, pointing to the significance
of landslides in landform evolution. Shroder (1976) mapped
landslides in the deeply weathered and moderately dissected
Nyika Plateau in Malawi, Africa, and recorded 235 individual
landslides over an area of 277 km2. One classic area to study
the origin and evolution of landslide-produced slope hollows
is the inland side of Serra do Mar in south-east Brazil, particularly around the town of Bananal (Coelho Netto, 1999;
Figure 10). Hollows vary in size, but are generally less than
100 m across and their main scarps are 1015 m high. Likewise, their shapes are different, from broad and smooth to
deeper and abruptly narrowing at the lower end. Many are
structurally controlled and their elongation follows the predominant joint direction.
The position of hollows within the slope varies. Some
are located close to the footslope and their floors connect
with depositional surfaces around the hills, whereas others
are disconnected from the footslope and may be described
as hanging. The latter, however, commonly grade into funnelshaped elongated depressions that provide a link to the
Figure 9 Bare rock slopes exposed by landslides. Note blocky talus, partly overgrown, at the footslope. Near Petropolis, south-east Brazil.
170
Figure 10 A landslide-related rampa complex in the vicinity of Bananal, south-east Brazil.
footslope. Footslopes below slope hollows are typically developed as low-angle (o51) fans and aprons built of landslide-derived colluvial material may further connect with
alluviated valley floors, local relief permitting. The thickness of
colluvial units may exceed 10 m, whereas the extent of these
depositional units may reach a few hundred meters, indicating
a long and complex erosional and depositional history, with
recurrent landslides, probably extending back well into the
Pleistocene (De Meis and Monteiro, 1979; Modenesi, 1988).
The association of landforms described above is so inseparable
that locally it is described using a single name. In south-east
China a term benggang is in use and literally means a collapsing hill (Xu, 1996), whereas in south-east Brazil such
hollow-and-footslope fan pairs are referred to as rampa
complexes, whereas the depositional surfaces themselves are
called rampas (De Meis and Monteiro, 1979).
The occurrence of landslides sets the stage for further
geomorphic processes, effective because of high erodibility of
weathered material if compared with solid rock. Among them,
gully erosion is worth particular attention as it works side-byside with landsliding. Generally, funnel-shaped transitional
zones used by earthflows and footslope fans built of poorly
consolidated material favor the subsequent development of
gully networks (Coelho Netto, 1999). Gullied slope hollows
and valley floors are a common sight in deeply weathered
terrains, such as south-east Brazil or southeast Asia. Gully
extension, in turn, promotes seepage, tunnel erosion, and
slope undermining, triggering new landslides, albeit on a rather local scale. Finally, the preweathered hillslope material
detached by landslides and carried away by earthflows and
sediment-laden ephemeral streams reaches valley floors,
causing widespread alluviation and valley filling. It is thus
evident that the geomorphic role of landsliding in the development of weathered hillslopes extends far beyond the spatial
scale of individual slope units affected by movement.
4.10.5
Weathering and Slope Landforms
Weathering dictates patterns of mass movement, the main

through rather indirect contribution of weathering to hillslope
evolution (see Chapter 4.17). Notwithstanding this fact, there
are different minor landforms present within the slopes,
which may be directly attributable to rock breakdown and
alteration. Among them, boulder fields, tors, and flared slopes
are the most conspicuous and have the widest distribution.
Moreover, they may have more than one origin.
4.10.5.1
Boulders and Boulder Fields
Boulder fields are defined as slope units, in which the topographic surface is made of closely spaced large rock fragments,
with virtually no finer material in between (Figure 11). Jointed
bedrock in situ may not be visible, being mantled by a boulder
blanket a few meters thick, but in other instances boulders
may be strewn over a rock platform. No universally agreed
threshold of minimum boulder size exists, but the length of
0.5 m is a reasonable limit and most papers report about
boulders of this size or larger. A distinction is sometimes made
between boulder fields and block fields. In the former, individual rock fragments are rounded, whereas angular shapes
typify the latter. Boulder fields occur on slopes of different
inclination, from as little as a few degrees to nearly 451. They
may also cover nearly flat summit surfaces and the phrase
mountain-top detritus is then occasionally used (Ballantyne
and Harris, 1994).
The size of individual boulders is controlled by the
spacing of discontinuities in the parent rock. The largest are
known from poorly jointed granites and these may exceed
10 m in length, although the length of 12 m is much more
common. Other rock types producing large boulders include
171
Figure 11 Quartzite block field produced by predominant physical weathering under cold-climate conditions, Jested, Czech Republic.
granodiorites, gabbro, rhyolites, quartzites, and massive

sandstones. By contrast, densely jointed and fissile rocks, such
as schist or mudstone, rarely disintegrate into boulders. The
dominant size of boulders determines the structure of an entire boulder blanket. If boulders are large, a few meters long,
and little finer material is available, then spaces between the
boulders cannot be easily filled and hence remain empty. In
certain instances, these voids may be so big and interconnected that their penetration by humans is possible. Such
shallow underground labyrinths are commonly referred to as
caves of nonkarstic origin. In Sweden, boulder caves in massive granites are known to exceed 1 km long in total and
similar lengths are reported from some boulder caves in
sandstone terrains in the Czech Republic. However, if the size
of individual fragments hardly exceeds 1 m, boulder blankets
are more densely packed.
There appears to be a relationship between boulder angularity and lithology. Fine-grained igneous and metamorphic
rocks preferably weather into sharp-edged fragments, whereas
coarse igneous and sedimentary rocks tend to produce rounded shapes, best known from granite areas. This is attributed
to more efficient grain-by-grain breakdown in coarse and
granular materials.
Boulder-mantled slopes and block fields are produced by
different surface processes. Some are definitely unrelated to
weathering such as talus slopes derived from rock fall or slide,
with subsequent disintegration of the failed mass on impact.
Others may be of glacial origin (boulder trails). Rapid cooling
of lava and resultant thermally induced rock breakdown may
also result in the formation of extensive boulder-mantled
slopes. Finally, some block fields are suspected to originate
from particularly severe seismic events accompanying deglaciation, as hypothesized with respect to certain localities in
Fennoscandia (Morner, 2003).
However, the majority of boulder-mantled slopes appear to
be a result of in situ weathering, although three evolutionary
pathways are possible. One scenario involves combination

of differential subsurface weathering and selective erosion
and is particularly applicable to regularly jointed rock masses.
Weathering is focused along water-bearing fractures and,
over time, differentiation of the rock mass into an orthogonal
lattice of thoroughly disintegrated zones and little weathered
compartments in between occurs (Figure 12). The latter
are named corestones and have been noted in excavations
as early as the late eighteenth century (Twidale, 1978).
Corestone-rich weathered mantles may be many meters
thick (see Chapter 4.8), with corestones present
at different depths within a profile. As a general guidance,
one may expect large, cubic corestones near the weathering
front, somewhat smaller but more rounded corestones in
the middle part of a profile, and least voluminous but
nearly perfectly spherical compartments in the top part
(Ruxton and Berry, 1957). Such a distribution would be
consistent with the assumption that time is the key factor
controlling the progress and effects of deep weathering.
However, rock inhomogeneities, whether related to mineral
composition or fracture density, interfere and may upset
simple relationships between duration of weathering and the
degree of alteration. Hence, corestones may concentrate in
particular horizons within weathering profiles, not necessarily
at the base, close to the weathering front. In the scenario
discussed, differential weathering leading to the origin of
corestones is generally considered as the first phase. Boulder
fields are the result of the second phase, which involves the
removal of fine material by running water from between corestones, leaving the latter in place but now exposed at the
surface. It may be hypothesized that in the course of longterm weathering and stripping, exposed boulders will passively settle down concurrently with general landscape lowering, to eventually form a continuous boulder blanket. This
model of boulder field origin bears close affinities to models
of two-stage origin of inselbergs (see Chapter 4.8) and tors (see
172
Figure 12 Corestone-rich weathering profile in granite, Mi"kow, south-west Poland.
(a)
(b)
Figure 13 Differential weathering of syenite around Aksum, Ethiopia: (a) general view of a footslope boulder field and (b) close up of structural
predisposition of boulder production by weathering.
below).
Reasons
for
acceleration
of
surface
erosion that outpaces deep weathering may be different,
or even site-specific. In the Mojave Desert of southwest
US, a shift from more humid conditions toward aridity in
the late Miocene was offered as an explanation (Oberlander,
1972), but in the granite areas of the Meghalaya plateau
in northwest India rapid stripping and boulder excavation can
be attributed to recent human impact, widespread deforestation, and surface wash during torrential rains (Prokop,
2007).
However, boulder fields in deserts are not necessarily the

evidence of past humid conditions and major environmental
change. In the central Namib Desert, around Gobabeb, the
granite landscape is dominated by spherical and half-spherical
boulders, as well as low bedrock convexities, rising from an
essentially rock-cut plain, with no deep weathering present.
Here, structural predisposition of granite to produce rounded
shapes is observed, further enhanced by flaking and exfoliation
after exposure. Figure 13(a) shows another boulder field, developed on syenite, likely formed in this way. Coexistence of

4.10.5.2
massive and foliated compartments (Figure 13(b)) indicates

primary differentiation within the rock itself, subsequently exploited by near-surface processes of weathering and erosion.
The essence of the third model is superficial rock breakdown in situ, with no contribution from either deep weathering or preexposure structural differentiation. The resultant
boulder fields are composed of angular rather than rounded
material and are particularly widespread on plateau surfaces
and gently inclined slopes in high latitudes and high altitudes,
hence considered as typical periglacial landforms. A variant
of the above is a block slope, whose gradient may be up to
30401 and bedrock is completely hidden beneath the stone
blanket (Figure 11). Various processes of physical weathering
are generally called for to explain cold-climate boulder fields,
mainly freezethaw and thermal weathering, with the debate
continuing about the actual role of each (see Chapter 4.15).
The structure of periglacial block fields is further modified by
frost heaving and sorting, as well as downslope movement in
the presence of cementation ice.
173
Tors
Tors belong to the most intriguing landforms within slopes

(see Chapter 4.8). The word is of local Cornish origin but
derives from the Old Welsh twr or twrr, meaning a heap or
pile. It was long used to describe castellated granite outcrops
rising from the moorlands of Dartmoor and Bodmin Moor in
south-west England and formally defined by Linton (1955:
470) as solid rock outcrops as big as a house rising abruptly
from the smooth and gentle slopes of a rounded summit or
broadly convex ridge. The definition, origin, and significance
of tors, as well as their relation to inselbergs, were particularly
debated in the late 1950s and 1960s (see Migon, 2006: 8587,
102105 for review) and a consensus finally emerged that tors
are outcrops of solid rock rising from a regolith-veneered
surface or rock platform, whether flat or sloping, which are
composed of more than one individual compartment (boulder),
but
are
too
small
to
be
considered
as separate hills. Tors have global distribution and are not
8
9
6
7
1
12
21
10 11
20
13
22
23
19
5
14
15
16
17
24
18
25
26
(a)
(b)
(c)
(d)
Figure 14 (a) Worldwide tor distribution and selected examples of granite tors; (b) castellated tor in the Karkonosze Mts, south-west Poland;
(c) Haytor, Dartmoor, south-west England; (d) massive domical tor in Serra da Estrela, Portugal. Tor localities: 1 Sierra Nevada, 2 Idaho,
3 Laramie Range, 4 Mojave Desert, 5 Acapulco, 6 Cairngorms, 7 Dartmoor and Bodmin Moor, 8 Aurivaara, northern Sweden,
9 southern Finland, 10 Harz, 11 Bohemian Massif, 12 Massif Central, 13 Iberian Peninsula, 14 Nigeria, 15 Uganda, 16 Tanzania,
17 Zimbabwe, 18 Swaziland, 19 Aravalli Range, 20 Khangai Mountains, 21 south-east Mongolia, 22 Korean Peninsula, 23 Japanese
Alps, 24 Devils Marbles, central Australia, 25 Bega Basin, 26 South Island of New Zealand.
174
Fractured rock mass

Weathering mantle
Figure 15 Model of tor evolution due to selective weathering and stripping.
Figure 16 Bouldery tor exposed in a quarry, after loose weathering products have been removed for road building purposes, Mylliem,
Meghalaya, north-east India. Photo courtesy of Agneiszka Latocha.
confined to any specific climatic zone (Twidale, 1982; Migon,

2006). Nor are they endemic to granite terrains although
classic examples have been indeed reported from granite uplands of Europe (Dartmoor and Cairngorms Great Britain,
Massif Central France, Harz Germany, Karkonosze Poland/Czech Republic, Serra da Estrela Portugal (Figure 14).
The origin of tors is generally considered a two-phase
process, hence similarly to the origin of certain boulder fields
as reviewed above. This evolutionary scenario is linked with
the name of David Linton, but the concept can actually be
traced much further back in time (Twidale, 1978). As Linton
(1955: 476) put it: A tor is a residual mass of bedrock produced below the surface level by a phase of profound rock
rotting effected by groundwater and guided by joint systems,

followed by a phase of mechanical stripping of the incoherent
products of chemical action (Figure 15). The key evidence
validating the theory is the presence of tor-like solid rock
masses surrounded by friable products of in situ disintegration, revealed in many grus excavations, road cuts, and
quarries all over the world (Figure 16). This scenario is easily
acceptable for low residuals, but needs refinement if very big
tors, 410 m high, are considered. First, the examples of yet-tobe-exposed tors show these features as a few to ten meters high
at most, whereas many tors are imposing residuals exceeding
20 m high. Second, the sides of certain tors exhibit features
typical for rock face/soil cover junction such as flared slopes
(see below), now standing high above the surrounding

ground surface. Third, elevated parts of tall tors bear traces of
protracted subaerial weathering and show a rich assemblage of
microforms (weathering pits, karren, tafoni), which are absent
or much less developed in their lower parts. These observations
suggest that the process of tor exposure may be episodic and
proceeds through a few cycles of deep weathering and stripping, with the top part having been exposed much earlier.
Not all tors have necessarily evolved according to the twophase model. Palmer and Neilson (1962) came up with an
alternative explanation, later known as a one-phase model.
They conceded that molding of tors would have occurred
entirely at the topographic surface and no preceding subsurface weathering (phase one) was involved. However, in
detail, periglacial tors would have been shaped in three
distinct phases. First, stripping of any preexisting regolith
by solifluction and exposure of solid granite occurred. Later
granite is subject to frost action and breaks down along
partings to form block fields. Finally, downhill movement of
released blocks by solifluction would expose a castellated tor.
The evidence in favor of this proposal includes the presence of
angular boulders around many tors, angular shapes of the tors
themselves, and widespread occurrence of solifluction deposits on adjacent slope surfaces. Paucity or absence of deep
weathering in the surrounding has also been cited but this is
not decisive. In fact, the lack of thick saprolites is consistent
with the two-phase model of tor origin. Today it is argued that
many tors, especially in Europe, have had complex history. For
example, Gerrard (1988) suggested that the tors of Dartmoor
have their roots in the pre-Quaternary period, when domes
and valleys formed and deep weathering operated, but details
of their morphology have been sculpted in the periglacial
environment of the Pleistocene. In other areas, for example, in
the Bohemian Massif of Central Europe, tors of different origin coexist, pointing to the polygenesis of the present-day
geomorphic landscape (Demek, 1964).
175
Thus, although the appearance of a tor is a poor and rather

unreliable clue to its geomorphic history, it is beyond doubt
that weathering processes are fundamental in shaping the
tors at every stage of their history. They continue to be active
on exposed tors and work toward their destruction. Selective
Figure 18 Flared slope of a granite boulder, Serra da Estrela,

Portugal.
Figure 17 A mushroom-like sandstone tor due to differential weathering of bedded and massive layers, Sto"owe Mts., south-west Poland.
176
weathering of tor surfaces leads to the origin of curious

landforms such as perched blocks (balanced rocks) and pedestal rocks, understood as residuals consisting of two parts:
a wide cap resting on a narrow pillar. The latter are most
common in bedded sandstones and conglomerates, where
weathering exploits differences between adjacent units
(Figure 17). Minor weathering landforms such as pits and
pans, tafoni, and karren (see Chapter 4.7) commonly develop
on tor surfaces, in some places rising to imposing dimensions.
Dzulynski and Kotarba (1979) presented a conceptual model
of how the evolution of weathering pans may reduce an upstanding tor to a low rock platform.
Phase 1
Runoff
Footslope deposits
Phase 2 - Accelerated weathering in footslope setting;

downward and inward migration
of weathering front (I > IV)
Runoff
I II
III
IV
Phase 3 - Removal of weathering products

and exposure of flared slope
Runoff
4.10.5.3
Flared Slopes
Flared slopes are smooth concavities in the scarp-foot zone of

hill-slopes or boulders, particularly widespread in granite
terrains but present in other rock types as well (Figure 18).
These basal concavities show a wide range of dimensions, even
along a single outcrop. They may be as low as 0.5 m, but
spectacular overhangs more than 10 m high are also known,
such as the 14-m high and nearly 100-m long Wave Rock
at Hyden Hill in Western Australia. They can be of a much
localized occurrence, and may also be present virtually all
along the perimeter of a hill, for example, at the Pildappa Hill,
Eyre Peninsula, Australia, where they make up 95% of the
lower slopes (Twidale, 1962). Many pedestal rocks have their
lower slopes flared and gradations are observed from flared
slopes to overhangs of basal tafoni and scarp-foot caves on
the one hand, and to gently sloping rock slope surfaces on
the other.
Although basal concavities are comparably minor surface
features, they are regarded as important indicators of geomorphic evolution of an area. This is because their origin is
generally ascribed to enhanced subsurface weathering at the
rock slope/footslope junction, the latter being either sediment-mantled or developed across saprolite (Figure 19).
Runoff from exposed rock slopes supplies water to sediment/
saprolite, creating a local microenvironment of aggressive
weathering, leaving the bare rock surface above little affected
by weathering. With the passage of time, this contrasting behavior would cause significant steepening of the rockregolith
boundary surface around an outcrop. Subsequent stripping of
loose material reveals the concavity. Subsurface, two-phase
origin of flared slopes implies that landforms within which
they occur have developed through at least one cycle of
differential weathering followed by evacuation of regolith.
Multiple flared slopes, at different heights above the base, have
been used as the evidence of long-term surface lowering
and episodic exposure of rock compartments (Twidale and
Bourne, 1975).
Flared slopes may continue to develop after exposure, as
attested by widespread flaking and exfoliation within the
concavities. Indeed, Ollier and Bourman (2002) argued that
flared slopes should not be uncritically assumed to be of
subsurface initiation and may develop solely by subaerial
processes, through slope retreat and preferential weathering at
the slope base. Whichever the model, though, it is evident that
these curious and sometimes truly spectacular features owe
their origin and appearance mainly to weathering rather than
any other geomorphic process.
The examples discussed above do not exhaust the range of
hillslope geomorphic features, which result from differential
weathering. These include various types of caverns (tafoni,
bedding caves), joint-guided clefts, mid-slope benches, rock
arches, and hoodoos (see Chapter 4.7).
Flared slope
4.10.6
Figure 19 Model of flared slope development as envisaged by

Twidale (1962).
Conclusions
This chapter demonstrated the direct bearing of rock weathering processes on patterns and pathways of hillslope evolution. The relationship between weathering and surface
processes has been formally recognized in the concept of

weathering-limited slopes and transport-limited slopes, contrasted with each other by means of relative efficacies of processes involving disaggregation of material and movement
of this material downslope. Although such a dichotomy is
an oversimplification, it remains a useful starting point to
examine how weathering influences the evolution of slope
morphology in different timescales.
Weathering controls mass movement processes operating
within a hillslope, dictating their nature and rate with which
they occur. If bedrock alteration proceeds at a fast rate and
reaches deeply beneath the surface, rate of change of slope
morphology increases. Hillslope processes influence hillslope
form, which may become characteristic for certain environments. For example, if weathered mantle is thin and surface
creep is the dominant process of mass transfer, convex
hilltops may form. However, if weathered mantle is thick and
landslides are common, scalloped hillslope topography with
channelless hollows will develop. In the short-term though,
processform relationships may also be reverse. It is an inherited hillslope form that controls patterns of weathering and
eventually, mass wasting.
Weatheringtransportform linkages may be examined for
both rock slopes, where physical weathering plays the major
role, and mantled slopes, where chemical weathering operates
efficiently. In either case, the role of weathering consists in
preparation, through various mechanisms of disintegration
and decomposition, rock material for transport. An important
factor here is the rock itself and characterization of rock mass
properties, whether little or thoroughly weathered, is essential
if pathways of hillslope evolution are to be understood and
predicted.
It has been shown that various categories of rock fall and
rock slide operate within rock slopes, whereas landslides,
earthflows, debris flows, and gully erosion mold slopes
underlain by weathered materials. Each of these mass-wasting
processes leaves behind erosional and depositional landforms,
which add to the complexity of hillslope morphology. However, careful analysis of these landforms, coupled with examination of footslope deposits, may allow for back-analysis and
recognition of the process, even if that has not been observed
directly. Here, geomorphology has a major role to play in
hazard assessment and in extension of observational record
into the more distant past, often essential to derive meaningful magnitudefrequency relationships.
Several classic landforms are products of slope evolution
in weathered materials, including famous tors and boulders.
Although other origins are possible, the majority of these inspiring landforms is produced by removal of weathered rock
from slope surfaces.
References
Arel, E., Tugrul, A., 2001. Weathering and its relation to geomechanical properties
of Cavusbasi granitic rocks in northwestern Turkey. Bulletin of Engineering
Geology and Environment 60, 123133.
Au, S.W.C., 1996. The influence of joint-planes on the mass strength of Hong Kong
saprolitic soils. Quarterly Journal of Engineering Geology 29, 199204.
Ballantyne, C.K., Harris, C., 1994. The Periglaciation of Great Britain. Cambridge
177
Bieniawski, Z.T., 1989. Engineering Rock Mass Classifications. Wiley, New York,
251 pp.
Bremer, H., 1975. Intramontane Ebenen, Prozesse der Flachenbildung. Zeitschrift fur
Geomorphologie N.F., Supplement-Band 23, 2648.
Chigira, M., 2001. Micro-sheeting of granite and its relationship with landsliding
specifically after the heavy rainstorm in June 1999, Hiroshima Prefecture, Japan.
Coelho Netto, M.L., 1999. Catastrophic landscape evolution in a humid region
(SE Brasil): Inheritances from tectonic, climatic and land use induced
changes. Supplementi di Geografia Fisica e Dinamica Quaternaria 3(3),
2148.
Davis, W.M., 1899. The geographical cycle. Geographical Journal 14,
481504.
Day, M.J., Goudie, A.S., 1977. Field assessment of rock hardness using the
Schmidt test hammer. British Geomorphological Research Group, Technical
Bulletin 18, 1929.
Dearman, W.R., Baynes, F.J., Irfan, T.Y., 1978. Engineering grading of weathered
granite. Engineering Geology 12, 345374.
De Meis, M.R.M., Monteiro, A.M.F., 1979. Upper Quaternary rampas: Doce river
valley, Southeastern Brazilian plateau. Zeitschrift fur Geomorphologie N.F. 23,
132151.
De Ploey, J., Cruz, O., 1979. Landslides in the Serra do Mar, Brazil. Catena 6,
111122.
Demek, J., 1964. Castle koppies and tors in the Bohemian Highland
(Czechoslovakia). Biuletyn Peryglacjalny 14, 195216.
Dietrich, W.E., Wilson, C.J., Reneau, S.L., 1986. Hollows, colluvium, and landslides
in soil-mantled landscapes. In: Abrahams, A.D. (Ed.), Hillslope Processes. Allen
and Unwin, London, pp. 361388.
Dixon, J.L., Heimsath, A.M., Amundson, R., 2009. The critical role of climate and
saprolite weathering in landscape evolution. Earth Surface Processes and
Landforms 34, 15071521.
Dzulynski, S., Kotarba, A., 1979. Solution pans and their bearing on the
development of pediments and tors in granite. Zeitschrift fur Geomorphologie
N.F. 23, 172191.
Fernandes, N.F., Guimaraes, R.F., Gomez, R.A.T., Vieira, B.C., Montgomery, D.R.,
Greenberg, H., 2004. Topographic controls of landslides in Rio de Janeiro: field
evidence and modeling. Catena 55, 163181.
Geological Society, 1997. Fookes, P.G. (Ed.), Tropical Residual Soils. A Geological
Society Engineering Group Working Party Revised Report.
Gerrard, J., 1988. Periglacial modification of the Cox Tor Staple Tors area of
western Dartmoor, England. Physical Geography 9, 280300.
Gilbert, G.K., 1909. The convexity of hilltops. Journal of Geology 17, 344350.
Goudie, A., Migon, P., Allison, R.J., Rosser, N., 2002. Sandstone geomorphology of
the Al Quwayra area of south Jordan. Zeitschrift fur Geomorphologie, N.F. 46,
365390.
Goudie, A.S., 2006. The Schmidt hammer in geomorphological research. Progress
in Physical Geography 30, 703718.
Howard, A.D., Kochel, R.C., 1988. Introduction to cuesta landforms and sapping
processes on the Colorado Plateau. In: Howard, A.D., Kochel, R.C., Holt, H.R.
(Eds.), Sapping Features of the Colorado Plateau. NASA SP-491, NASA,
Hurault, J., 1963. Recherches sur les inselbergs granitiques nus en Guyane Francaise. Revue de Geomorphologie Dynamique 14, 4961.
Keefer, D.K., 1984. Landslides caused by earthquakes. Geological Society of
America, Bulletin 95, 406421.
Lan, H.X., Hu, R.L., Yue, Z.Q., Lee, C.F., Wang, S.J., 2003. Engineering and
geological characteristics of granite weathering profiles in South China. Journal
of Asian Earth Sciences 21, 353364.
Larsen, M.C., Wieczorek, G.F., 2006. Geomorphic effects of large debris flows and
flash floods, northern Venezuela, 1999. Zeitschrift fur Geomorphologie N.F.,
Supplement-Band 145, 147175.
Linton, D.L., 1955. The problem of tors. Geographical Journal 121, 470487.
Lugo, J., Vazquez Conde, M.T., Zamorano Orozco, J.J., Matas Ramrez, G., Gomez
Arizmendi, A., 2002. El huracan Pauline en Acapulco, octubre de 1997.
In: Lugo, J., Inbar, M. (Eds.), Desastres Naturales en America Latina. Fondo de
Cultura Economica, Mexico, pp. 267288.
384 pp.
Migon, P., 2009. Are any granite landscapes distinctive of the humid tropics?
178
Migon, P., 2010. Mass movement and landscape evolution in weathered granite and
gneiss terrains. In: Calcaterra, D., Parise, M. (Eds.), Weathering as a
Predisposing Factor to Slope Movements. Engineering Geology Special
Publications, Geological Society, London, vol. 23, pp. 3345.
Migon, P., Alcantara-Ayala, I., 2008. Weathering and landform development in a
subtropical mountainous terrain, Veladero massif, Mexico. Zeitschrift fur
Geomorphologie, N.F. 52, 116.
Migon, P., Thomas, M.F., 2002. Grus weathering mantles problems of
interpretation. Catena 49, 524.
Modenesi, M.C., 1988. Quaternary mass movements in a tropical plateau (Campos
do Jordao, Sao Paulo, Brazil). Zeitschrift fur Geomorphologie, N.F. 32,
425440.
Morner, N.A., 2003. Paleoseismicity of Sweden. A Novel Paradigm. Stockholm,
320 pp.
Ng, K.C., Parry, S., King, J.P., Franks, C.A.M., Shaw, R., 2002. Guidelines for
natural terrain hazard studies. GEO Report no. 138, Geotechnical Engineering
Office, Hong Kong, 138 pp.
Oberlander, T., 1972. Morphogenesis of granitic boulder slopes in the Mojave
Desert, California. Journal of Geology 80, 120.
Oberlander, T., 1989. Slope and pediment systems. In: Thomas, D.S.G. (Ed.), Arid
Zone Geomorphology. Halsted Press, London, pp. 5686.
Ollier, C.D., 1965. Some features of granite weathering. Zeitschrift fur
Geomorphologie, N.F. 9, 265284.
Ollier, C.D., 2010. Very deep weathering and related landslides. In: Calcaterra, D.,
Parise, M. (Eds.), Weathering as a Predisposing Factor to Slope Movements.
Engineering Geology Special Publication, vol. 23, pp. 514.
Ollier, C.D., Bourman, R.P., 2002. Flared slopes, footslopes and the retreat of
overhanging slopes: examples of convergent landform development. Physical
Palmer, J., Neilson, R.A., 1962. The origin of granite tors on Dartmoor, Devonshire.
Proceedings of the Yorkshire Geological Society 33, 315340.
Parsons, A.J., 1988. Hillslope Form. Routledge, London and New York, 212 pp.
Phillips, J.D., 2010. The convenient fiction of steady-state soil thickness. Geoderma
156, 389398.
Prokop, P., 2007. Land degradation of the southern slope of the Meghalaya plateau,
India. Prace Geograficzne 210, 1162IGiPZ PAN, Warszawa.
Ruxton, B.P., Berry, L., 1957. Weathering of granite and associated erosional
features in Hong Kong. Geological Society of America, Bulletin 68, 12631282.
Selby, M.J., 1980. A rock-mass strength classification for geomorphic purposes:
with tests from Antarctica and New Zealand. Zeitschrift fur Geomorphologie N.F.
24, 3151.
Selby, M.J., 1993. Hillslope Materials and Processes. Oxford University Press,
Oxford, 451 pp.
Shaw, R., 1997. Variations in sub-tropical deep weathering profiles over the
Kowloon Granite, Hong Kong. Journal of the Geological Society, London 154,
10771085.
Shroder, Jr. J.F., 1976. Mass movements on the Nyika Plateau, Malawi. Zeitschrift
fur Geomorphologie N.F. 20, 5677.
Chichester, 375 pp.
Thomas, M.F., 1994. Geomorphology in the tropics. Weathering and Denudation in
the Low Latitudes. Wiley, Chichester, 460 pp.
Thomas, M.F., 1995. Models for landform development on passive margins. Some
implications for relief development in glaciated areas. Geomorphology 12, 315.
Twidale, C.R., 1962. Steepened margins of inselbergs from north-western Eyre
Peninsula, South Australia. Zeitschrift fur Geomorphologie N.F. 6, 5169.
Twidale, C.R., 1978. Early explanations of granite boulders. Revue de
Geomorphologie dynamique 27, 133142.
Twidale, C.R., Bourne, J.A., 1975. Episodic exposure of inselbergs. Geological
Twidale, C.R., Campbell, E.M., 1992. On the origin of pedestal rocks. Zeitschrift fur
Geomorphologie N.F 36, 113.
Wieczorek, G.F., 2002. Catastrophic rockfalls and rockslides in the Sierra Nevada,
USA. Geological Society of America, Reviews in Engineering Geology 15,
165190.
Xu, J., 1996. Benggang erosion: the influencing factors. Catena 27, 249263.
Young, R.W., Wray, R.A.L., Young, A.R.M., 2009. Sandstone Landforms. Cambridge
Biographical Sketch
Piotr Migon is professor of geography at the University of Wroc"aw, Poland, where he is Head of Department of
Geography and Regional Development. He was a Royal Society Fellow based at the University of Oxford, UK, in
19951996, and a Fulbright Fellow based at the University of California, Santa Cruz, in 2003. He served as the
Secretary of the International Association of Geomorphologists (IAG) in 19972001 and currently (20092013) is
one of its three vice-presidents. His principal subject of research is how geological structures influence landforms
and surface processes. Other explored issues include weathering, mass movement in mountain terrains, and longterm landform evolution. Piotr Migon is also involved in various geo-conservation activities and in promotion of
geomorphology. His publication record includes Granite Landscapes of the World (Oxford University Press,
2006), a textbook on geomorphology published in Poland (2006), coediting of Encyclopedia of Geomorphology (Routledge, 2004) and numerous papers in international journals, including Geomorphology, Catena,
Zeitschrift fur Geomorphologie, and Earth-Science Reviews. His last publication is the edited volume
Geomorphological Landscapes of the World (Springer, 2010), which contains 36 chapters describing the most
fascinating geomorphic sceneries from all around the world.
4.11 Weathering in the Tropics, and Related Extratropical Processes

4.11.1
4.11.1.1
4.11.1.2
4.11.2
4.11.2.1
4.11.2.2
4.11.2.3
4.11.2.4
4.11.2.5
4.11.3
4.11.3.1
4.11.3.2
4.11.3.3
4.11.4
References
Overview
Heritage
The Tropical Geomorphic Region: Defining Tropical in Geography and Time
Weathering Processes and Their Relation to Tropical Conditions
Factors
The Processes
End Products of the Weathering Process
Rates of Weathering
Weathering Maxima Outside the Tropics
Weathering-Related Landforms of the Tropics
Weathering Voids: Solutional Landforms
Deep Weathering Mantles
Conclusion
Glossary
Allitization Total loss of silica and alkaline elements,
production of gibbsite (the aluminum oxide residual),
ferric hydrates, and 1:1 clays (such as kaolin), as
told by Pedro (1968), to be centered in the core
tropics.
Bisiallitization Moderate loss of silica, formation of 2:1
clays such as smectite and vermiculite, some retention of
alkaline cations, as told by Pedro (1968) to be typical of
temperate climates.
Bornhartdt A dome-shaped rock inselberg, named after
West African explorer Wilhelm Bornhardt.
Chelation Biochemical weathering process, the
preferential extraction in minerals of metal ions by organic
compounds.
Dissolution Multiple-step chemical weathering process in
the presence of acidic agents, sometimes referred to as
incongruent solution or hydrolysis.
Duricrust An illuvial soil hardpan formed of secondary
precipitates, such as iron or silica.
Etchplain A low-relief exhumed surface, in which deep
weathered rock material has been removed.
Ferrallitization Extensive leaching, creating end-product
soils (including ferrisols and ferricrete) in a transition zone
beyond the wettest rainforest to the seasonally wet savannas.
Ferricrete An iron-rich duricrust or hardpan in the soil.
Ferrisols Iron-rich aluminum-silicate end product soil
common in the tropics.
180
180
183
184
184
185
186
189
189
190
190
191
192
193
193
Gibbsite End product aluminum hydroxide mineral, a

component of bauxite, common in but not limited to the
tropics.
Hydration Chemical weathering process involving
the absorption of a hydroxide molecule into the crystal
matrix.
Inselberg Literally, island mountain, a resistant hill or
peak derived from deep weathering; may be a single dome
or bouldery.
Kaolinite A 1:1 clay mineral, hydrated aluminum silicate,
lacking base cations. Kaolinization is the genesis of kaolinite,
and end-stage weathering product.
Laterite An iron- and aluminum-rich, autochthonous
(originating in-place) weathering product, variously
defined in common and academic use. Laterization is the
genesis of laterite, common but not limited to tropical
regions.
Monosiallitization Partial loss of silica, total loss of
alkaline elements, production of 1:1 clays (kaolinite) and
ferric hydrates, as told by Pedro (1968) to be associated
with subhumid tropics.
Multiconvex topography Also known as half-orange or
meias laranjas hills; low, irregularly spaced dome-shaped
hills known in tropically weathered locations.
Neoformation Refers to new minerals formed out of ions
in solution, the product of weathering.
Oxidation Chemical weathering process by which oxygen
molecules incorporate into the mineral lattice.
Pope, G.A., 2013. Weathering in the tropics, and related extratropical

processes. In: Shroder, J. (Editor in Chief), Pope, G.A. (Ed.), Treatise on
Geomorphology. Academic Press, San Diego, CA, vol. 4, Weathering and
Soils Geomorphology, pp. 179196.
http://dx.doi.org/10.1016/B978-0-12-374739-6.00059-2
179
180
Weathering in the Tropics, and Related Extratropical Processes
Oxisol Deeply weathered soils, with prominent reddish

tint brought about by iron oxidation, also known as
ferralsol.
Pedogenic Referring to the process of soil formation.
Pressure unloading A type of mechanical weathering, in
which rock sheets or segments separate along preferred
weakness zones subparallel with the rock surface, due to the
relaxation of pressure above the surface (such as overburden
or glacial ice).
Regolith Accumulation of unconsolidated or poorly
consolidated sediment and weathered rock, above bedrock,
at the surface.
Rubification Reddening of soil color, by oxidation of
iron.
Saprolite Weathered bedrock, in situ.
Silcrete A silica-rich duricrust or hardpan in the soil,
formed of secondary deposition.
Silica karst A class of solution landforms, parallel to the
more common carbonate karst, formed in silicate rocks
such as sandstone or quartzite.
Solution Simple, single-stage dissociation o f elements in
the presence of water or acids; also referred to as congruent
solution.
Tor Residual outcrop or pile of corestone boulders, the

remains of exhumed deep weathering; a small boulder
inselberg.
Tower karst Tall pinnacle of resistant rock remaining after
karst solution, usually applied to limestone landscapes,
though some instances of silicate rock tower karst exist.
Ultisol Well-developed soil with low base saturation,
found in humid climates, but not leached to the extent of
oxisols.
Weathering Rock and mineral decay by surface chemical
and mechanical agents, the precursor to erosion and
sediment generation and a source of dissolved elements in
surface and near-surface waters.
Weathering potential A measure of the degree of
weathering that can take place, expressed as a ratio of
alkaline oxides in the rock to all oxides; high weathering
potential would have a high alkaline oxide ratio, being less
depleted of alkaline oxides.
Weathering product A measure of the degree of
weathering that has occurred, expressed as a ratio of silica
oxides over the combination of resistant oxides (silica,
titanium, aluminum, and iron); rock more depleted in silica
would have a lower weathering product.
Abstract
Weathering processes are partially responsible for a characteristic geomorphology that occurs in the tropics and subtropics.
Resistant landforms such as inselbergs, extreme solution processes such as silica karst, and deep weathering profiles with
end stage weathering products such as laterite and kaolin are common features of tropical weathering. Many of these
features also occur outside the tropics. In part, climate change and paleotectonics were responsible for tropical conditions
in areas not now tropical. But, some processes assumed to require tropic conditions that are not so limited, sufficient
moisture, and time sufficing for their development. This chapter reviews the weathering processes and distinctive landforms
of the broadly defined tropics, and explores the debates over weathering factors as they pertain to tropic and extratropical
environment.
4.11.1
Overview
Those who study weathering often look at optimal scenarios,

in which conditions are most conducive to weathering. The
literature gravitates toward the extremes: the hottest, coldest,
driest, wettest, and most associated chemically active or
physically excessive environments that bring about observed
weathering. Indeed, much of the science of geomorphology is
the same, and we commonly recognize today that the cataclysmic and extreme are as important if not more important
than the average or moderate in shaping the landscape. The
preoccupation with extremes quickly identified the tropics as
an incubator of intense chemical weathering and associated
landforms and soils, given the abundance of precipitation,
higher temperatures, and omnipresence of organic acids. Yet,
these assumed ingredients are not always necessary at once for
the genesis of tropical-like weathering and geomorphology,
and tropic-like geomorphology can be observed beyond the
tropics. This chapter explores the processes of weathering
relevant to the tropics, leading to characteristic landforms
and regolith. Convergently, evolutionary landforms and

regolith beyond the tropics are also discussed, revealing a
debatable controversy that is ongoing but not always recognized beyond the introductory literature.
4.11.1.1
Heritage
The processes of weathering provided seemingly sensible affirmation for the overarching theories of climatic geomorphology, initiated by Branner (1896) and Falconer (1911), and
advanced in the mid- and later-twentieth century, the works of
Budel (1948, 1977), Derbyshire (1973), and Tricart and
Cailleux (1972) being most prominent to the movement.
Deep laterites and etchplains of the tropics seemed as obvious
as sand dunes in the deserts and glaciers in the Arctic, and
weathering (particularly in the tropics) received due attention.
Three key figures emerged as seminal to the concepts of climate-controlled weathering in general: Louis Peltier, Nikolai
Strakhov, and Georges Pedro. Their models influenced theory,
though with increasing criticism as weathering factors are

rationalized.
Peltiers (1950) study, first among the three, focused on
periglacial geomorphology and the numerous processes involved, including weathering. Despite the cold-climate focus,
his Figures 13 (Peltier, 1950:219) of that paper presented a
broad and appealing graphic model of weathering types across
all climate zones (Figure 1(a)), precipitation and temperature
being the primary controls of weathering in this view. This set
of figures is perhaps the most widespread diagramatic model
of weathering factors seen, still common in introductory and
more specialized textbooks (including this one) 60 years
hence. In Peltiers model, as temperature and moisture increase, the efficacy (and dominance) of chemical weathering
becomes stronger, whereas mechanical weathering diminishes
to weak to absent or insignificant at the positive extremes of
temperature and moisture. Strong chemical weathering thus
dominates in the tropics. Peltier, however, was not aware of
the ubiquity of chemical weathering in cold and arid climates.
Though not necessarily as vigorous as in wetter regions,
chemical weathering is in fact present and perhaps even
dominant across most climatic regions (Pope et al., 1995),
whereas mechanical weathering is also present and important
in the humid tropics.
Strakhovs (1967) work was intended as an examination of
sedimentary processes and the genesis of sedimentary rocks,
but the chapter on weathering likewise offered a diagramatic
model that lived on well beyond the expected lifetime of a
somewhat obscure text. The most cited diagram (Figure 1(b),
from Strakhov, 1967: 6) expressed temperature and moisture
controls similar to Peltier that added the influence of biotic
agents in weathering (in the form of plant litter fall), translating to weathering efficacy evident in depth of weathering
and regolith zonation across a latitudinal transect. An accompanying world map, seldom cited, generalized these
weathering climate zones. Like Peltiers, the model was conceptual: Although actual values for latitude, temperature,
moisture, and biomass were used, depths of weathering zones
were approximate (but still attempted to suggest the vertical
and spatial irregularity of weathering fronts, at least illustratively). The interesting feature of the diagram was the zonation
(Al-oxide, Fe Al oxides, kaolinite, illite-montmorillonite,
and incipient weathering), the primary intensity peak of the
tropics, a secondary intensity peak of the subpolar and middle
latitudes (Strakhov used Taiga zone, with the increase of
acidic conifer needle mulch), and minimal weathering profiles
in warm and cold drylands. The diagram illustrated a cumulative effect of abundant moisture (as both weathering agent
and remover of solutes), plentiful decomposed vegetation,
and high temperature to produce deep weathering profiles in
the tropics. Strakhovs tropic zone included the regions
dominated by ferrisols and ferrallitic soils, the product of
rapid weathering and leaching. The diagram, however, was a
broad generalization, not a data-based model, and with
notable exceptions to the presumed weathering zonation (see
Chorley et al., 1984; Ollier and Pain, 1996; Pope et al., 1994).
For one thing, the assumed curves for the bioclimatic weathering factors do not closely approximate actual measurements now available (Figure 1(b)). As it turns out, leaf litter
fall is actually greater in midlatitudes than in the tropics
181
(Potter et al., 1993), and this corresponds to higher soil carbon in middle latitudes (Post et al., 1985), providing the
midlatitudes with potentially greater biochemical weathering
agents than in the tropics.
Pedros (1968) study was parallel to Strakhovs, developed
independently though using similar concepts. Pedros work
directly addressed weathering, particularly surface chemical
weathering, translating globally across different climatic zones,
resulting in different forms of weathering and associated
soils. The geographic extents of certain phenomena (for instance, soil rubification in oxisols) was one example of weathering defining the region of tropical geomorphology. Broad
pedogenic processes allitisation (laterite, aluminum oxides),
monosiallitisation (kaolinite), bisiallitisation (vermiculitemontmorillonite), and podsolization were distinguished
using ratios of silicon oxide to aluminosilicate and silicon
oxide to bases (an important distinction from the Peltier and
Strakhov models, which were not actually quantitative). These
in turn corresponded directly with climatic zones (tropical,
temperate, and cool) and rates of decomposition (rapid to
slow). The resulting map (Figure 1(c), from Pedro, 1968: 463)
of weathering pedogenesis somewhat matched the distrubution of major soil orders (FAO, 1985; Thomas, 1974, 1994).
Recent authors built on the groundwork of these initial
tropical weathering paradigms. Foremost among these are
Ollier (1969, 1984), Ollier and Pain (1996), Twidale (1982,
2002), and Thomas (1974, 1996). Both Ollier and Twidale,
while working with tropical geomorphology, importantly
noted that many of their subjects were not by necessity tropical and were in fact controlled in many cases by nonclimatic
factors. Thomas, in comparison, did emphasize the importance of tropic climates. Although recognizing with his studies
the continuum of weathering and pedogenic processes beyond
the tropics, he emphasized the unique qualities of the tropics
in his definitive texts on tropical geomorphology, which relied
heavily on the foundation of weathering.
Trudgill (1976: 89) rightly pointed out y the unqualified
tenet of climatic geomorphology y states simply that different
landforms occur in different climatic zones. Climatic geomorphology proved unsatisfying as an over-arching theory, as
climate was not the only or even primary factor in many geomorphic systems (see Holzner and Weaver, 1965; Stoddart,
1969; Selby, 1985; Twidale and Lageat, 1994). Although conceptually and diagrammatically appealing, the climatic weathering models are problematic in their simplicity. Chorley et al.
(1984) pointed out the absence of maritime temperate climates
from the Strakhov concept. Huggett (2007) noted Peltiers
omission of mechanical weathering not caused by ice (mentioning the importance of mechanisms involving heat and salt,
but not going far enough by also mentioning the roles of roots,
expansive clays, and rock stress relaxation). Thomas (1994)
differed with the relative proportions of tropical rainforest and
savanna climates. Pope et al. (1995) drew issue with both the
Strakhov and Peltier models as too simplistic. The models, and
a widespread repetition of them throughout the research and
academic literature, ignored frequent evidence that contradicted
the assumptions of the models:
1. Chemical weathering is not weak in cool or dry climates,
nor is mechanical weathering absent in warm-wet climates;
Mean Ann. Rainfall

80
70 60
50 40 30
20
10 Inches
Mod. mech.
10 F
strong mechanical
Mean Ann. Temp.
20
Slight
weathering
mech. wea
mod.
30
Mean Ann. Temp.
Mean Ann. Rainfall

80 70 60 50 40 30 20 10
10 F
20
Moderate
chemical
weathering
ak
We
50
60
Absent or insignificant
80
Frost weathering
Mean Ann. Rainfall
80 70 60 50 40 30 20 10
Inches
10F
Very
slight
weathering
80
Weathering regions
Mean Ann. Temp.
Strong
chemical
weathering
ng
ro
te
era
d
Mo
40
50
60
Inches
St
30
70
mech
Mod. chem. wea

wea with frost
action
40
70
Weak
20
30
40
Moderate
50
60
70
Strong
80
(a)
Chemical weathering
30
5
4
3
2
1
Mean temperature C
12
20
10
8
10
6
0
4
10
Little chemical alteration

Illite-montmorillonite
Kaolinite
Al2O3
Fe2O3 + Al2O3
(b)
Tectonic
area
Tectonic
area
Tectonic
area
Tectonic areas
Weathering zones
Allitization
Hyperarid
Ice cover
Kaolinization
Podzolization
Smectization
(c)
Tectonic
area
Tectonic
area
Total plant litter (Pg of carbon)
Mean precipitation (mm day1)
182
2. Deep weathering profiles occur commonly outside of the

tropics (and taiga zones);
3. In terms of denudation rates (a function of both weathering and erosion), there is significant overlap in the
range of values between climatic regions.
These revised beliefs are realized by a growing number of
weathering geomorphologists reflected in this book.
Apart from the morphoclimatic debate, the summary
message is that weathering processes in the tropics were
identified by early climatic geomorphologists as important to
the overall geomorphic landscape of the tropics, and key
works, namely by Peltier (1950), Strakhov (1967), and Pedro
(1968), provided conceptual models for weathering processes
that persist to this day. The oft-repeated tenet is difficult to
refute: The high temperatures and high moisture of the tropics
provide an intense chemical weathering environment. Add to
this the presence, in some areas, of abundant organic decomposition, providing organic weathering agents. Even when
considered at the microscale boundary layer at which actual
weathering takes place (Pope et al., 1995), large-scale bioclimatic factors should filter in to smaller scales to be relevant.
Absent from the tropical weathering concept is the mention of
mechanical weathering processes, though these too can be
added. What requires further explanation is the observed
landforms beyond the tropics that appear to be, and are
claimed to be, derived from tropical processes. Are they? Can
nontropical environments produce similar landforms? Or, do
Figure 2 Chongra bornhardt face of part of Mount Mlange massif in

eastern Malawi, Africa. The erosional residuals (inselbergs,
bornhardts) of this part of the African erosion surface in East Africa
are quartz- and orthoclase feldspar-rich and more resistant to
chemical weathering than the mafic-rich country rock of the
foreground which has weathered and been erosional stripped to a
lower altitude (Shroder, 1973). Photograph courtesy of J. Shroder.
183
these outlying examples provide proof of previous tropical

conditions?
4.11.1.2
The Tropical Geomorphic Region: Defining

Tropical in Geography and Time
The common notion of the tropics consistent heat,

humidity, and abundant precipitation seasonally if not
continuously is complicated in the perspective of geomorphology. Naturally, it would incorporate regions of tropical rainforest and tropical savannah. With the help
of ocean currents and regional prevailing wind variations,
tropical climates can extend from the equator to beyond
25301 N or S latitude (for instance, southeast Africa, northern India, and northwest Mexico). Given the long intervals of
geomorphic evolution along with climate change, neighboring
regions to the core tropical ones also experienced tropical
conditions worth including. In contrast, highland areas (such
as the South American Altiplano, East Africa, mountainous
Southeast Asia, and Indonesia) within the tropics would not
experience the same elevated temperatures. Areas under subtropic high pressure or in mountain rain shadows could be
significantly drier, qualifying as deserts.
Tropic can be precisely defined, but depends on the context.
The classic climate region classification of Koppen and Geiger
specified tropical or equatorial (A-type climates) for areas of
the world warmer than 18 1C (Kottek et al., 2006). One popular
introductory textbook (Gabler et al., 2007) defines the tropical
climatic region simply as warm all year, which probably suffices
short of defining temperature thresholds for chemical or physical processes. Subsets to the tropical definition are based on the
amount and seasonality of precipitation, which are of importance to geomorphic processes. At present, tropical climates cover
B22% of the continental surface (Kalvova et al., 2003). From an
ecological perspective, tropical also varies by temperature and
precipitation, but defined in ways coinciding with vegetation
types (also relevant to the types of weathering). Wolfe (1979)
distinguished divisions in forest types between tropical
(425 1C mean annual temperature, MAT), paratropical
(2025 1C MAT), and subtropical (1320 1C MAT). The term
paratropical was defined in an earlier paper by Wolfe (1969),
referring to a fossil Paleogene flora occurring in the present Gulf
of Alaska (revealing a crucial point: in the time spans relevant to
weathering landscapes, extents of climatic regimes vary greatly.)
Regions peripheral to the true tropics may also be relevant to
the weathering geomorphology discussion, as will be discussed
later in this chapter. These include the hot deserts (in the
context of climate change and potentially more humid conditions in the past) as well as portions of mesothermal humid
continental regions (Koppen-Geiger Cfa and Csa classification).
Figure 1 Graphical models of weathering factors with implications for tropical weathering. (a) From Peltier (1950), weathering type by
precipitation and temperature. Chemical weathering is now recognized to occur in all climate types. Reproduced from Peltier, L.C., 1950. The
geographical cycle in periglacial regions as it is related to climatic geomorphology. Annals of the Association of American Geographers 40(3),
214236, with permission from Taylor & Francis. (b) Weathering depth and factors, by latitudinal transect, derived from Strakhov (1967).
Strakhovs estimation of temperature (red dotted line), precipitation (blue dashed line) and litter fall (green dash-dot line) do not correspond with
realistic and up-to-date measurement of latitudinal mean temperature (solid red line) and mean precipitation (heavy blue line) (both from Kalnay
et al. (1996)) and plant litter (Potter et al., 1993). (c) From Thomas (1974), based on Pedro (1968), soil types and weathering based on climatic
region.
184
+40
0
Montmorillonite
40
Kaolinite,
halloysite,
anauxite
80
Cuban laterite
ite
120
ux
/ba
e
t
i
ter trend
La
160
Arkansas bauxite
20
40
Potential index [bases (bases + SiO2 + R2O3)]
Igneous rocks
200
60
80
Product index [SiO2 (SiO2 + TiO2 + R2O3)]

Figure 3 Weathering product and weathering potential indexes, Arkansas Bauxite and Cuban Laterite are examples described in Reiches text.
Colmam (1982) utilized the same graphic in for the Western United States, but extending only to the kaolinite field. Modified from Reiche,
P.,1950. A Survey of Weathering Processes and Products. University of New Mexico Publications in Geology, No. 3. Albuquerque: The University
of New Mexico Press, 91 pp.
In the familiar climatic geomorphology classification, the

tropic morphoclimatic region can be defined in terms of
weathering and soil development. Ferrisols and fersiallitic
soils are roughly congruent to Pedros (1968) world weathering zones: the zone of alltisation (kaolinite gibbsite)
in the core lowland tropics; and the zone of kaolinization
(kaolinite) in surrounding areas, extending as far south as
Uruguay, South Africa, and southeast Australia, and as far
north as India, South China, and the southeast US, in other
words, covering the classic tropics, and extending into subtropic to even middle latitude regions. These various classifications follow the foundation of morphogenetic regions,
summed well in Chorley et al. (1984), but including the
seminal works (overgeneralized though they may be) of Budel
(1948), Peltier (1950), Strakhov (1967), Pedro (1968), Tricart
and Cailleaux (1972), and Derbyshire (1973).
Defining climate regions applies to current environments, though past environments are equally if not more important. Ollier and Pain (1996: 80) pointed out that assuming
present-day climate is responsible for observed regolith features
is nave and misleading. Weathering can be among the slowest
of geomorphic processes. Significant weathering landforms
may have evolved over millions or even tens of millions of
years, a time span in which entire continents shift into different
climatic latitudes or portions of landmass thrust upward into
cooler elevations, not to mention large swings in climate cycles
determined by solar input and ocean currents. Even smaller
weathering forms may be relict of different climatic regimes due
to the cycles of climate change. Thus, studies of tropical weathering in mid latitude and even subarctic regions are possible
(Cunningham, 1969; Dury, 1971; Derbyshire, 1972; Pavich,

1986; Yapp, 2008; and Solbakk et al., 2010), though there is
sometimes legitimate debate as to whether the weathering
observed is actually tropical in genesis.
4.11.2
Weathering Processes and Their Relation to

Tropical Conditions
The weathering processes of the tropics are similar to those

elsewhere, with just a few exceptions. Apart from a few
mechanical processes, all are relevant, but with special emphasis related to the climatic conditions and environment,
enumerated here.
4.11.2.1
Factors
Weathering is a syngergistic system, the whole of which may be

greater than the sum of its parts (see Chapter 4.2; Pope et al.,
1995). Within this synergy, what factors contribute to rapid
weathering, and are these factors prevalent in the tropics? Curtis
(1976: 50) answered this question summarized as follows:
1. A warm climate translates to sustained high temperatures,
influencing chemical reaction rates, following the Arrhenius equation:
k A expEA =RT
where k is the reaction rate, A is the rate constant for the
reaction (a function of molecular collision, related to
efficacy of the weathering agent in this case), EA is the

activation energy of the reaction, R is the gas constant, and
T is the absolute temperature. From the perspective of the
chemical reaction, temperature is a primary environmental
variable.
2. High precipitation (or at least where precipitation is significantly greater than evapotransipiration, cf. Ollier and
Pain, 1996) provides a long term presence of water as a
medium for reactions, a supply of reacting agents (including H2O, CO2, and O2), and removal of solutes
without reaching saturation. Variability in precipitation
influences the type of chemical reaction, discussed shortly.
Large seasonal variations in precipitation and humidity,
also present in some tropical areas, would be pertinent to
several mechanical weathering processes.
3. Climate in turn fosters high organic productivity, which
also supplies key reacting agents such as acids and chelates.
Bluth and Kump (1994) reinforced Curtiss observations in
terms of chemical denudation, but also stressed the role of
evacuating weathering products from the system:
y dissolved yield of a given drainage basin is determined by a
balance between physical and chemical weathering; thus, a warm,
wet climate, or the presence of abundant vegetation cannot guarantee high rates of chemical denudation unless accompanied by
high rates of physical removal.
Nonclimatic factors are important as well. Fresh parent

materials would have greater weathering potential (WPoI), as
would rocks of high surface exposure (by way of macro- or
microporosity). The tropical region has examples of recent
tectonism or volcanism that would easily expose fresh rock,
but also examples of long term tectonic stability (such as
Australia, South Africa, and Brazil). Tectonisms and erosion
act to provide significant topographic relief, also relevant in
providing good drainage, influencing the presence of water as
reaction medium and solute remover. Time is a factor not
mentioned above, completely divorced from climatic environment. Although process rates may be accelerated in the
tropics, given enough time and stability, an equifinality of
weathering extremes may be possible regardless of climate. In
sum, although various weathering factors are aided by tropical
environments, other factors occur regardless of climate.
4.11.2.2
The Processes
Chemical processes are strong in the tropics, or at least obvious,

but mechanical processes are present and important. Mechanical
processes go together with chemical, it is seldom that one does
not exist without the other, and rather positively reinforce each
other. The processes will be reviewed here one by one, though in
reality processes work together in a synergistic fashion (see also
Chapter 4.2).
Of the mechanical processes, ice is unlikely to be an agent
in the classically defined tropics, as well as stress from subfreezing temperature excursions, apart from climate cycles that
could be relevant at higher elevations or at higher latitudes.
There is some debate as to whether thermal shock at
high temperatures is relevant (see Bland and Rolls, 1998;
Eppes et al., 2010). Even if the tropics do not attain the high
185
air temperatures of the deserts (though some may come close),

rock surface temperatures may well exceed 70 1C, particularly
on dark colored rocks (Thomas, 1994). High temperature itself may not be sufficient to create brittle fracture without large
temperature extremes, but the subject has not been well researched in the tropics. Fires, outside of the rainforest during
the dry seasons and in droughts, are known to exert extreme
temperatures capable of brittle rock fracture (Goudie et al.,
1992; Dorn, 2003). Crystal growth within confined pores or
fractures may be causes of mechanical weathering in the tropics. Normally, rapidly growing minerals such as salts, calcite,
and gypsum are easily dissolved and flushed away by rain.
However, in the aggressive chemical environment, rapid release of elements such as sodium, calcium, and potassium
from rock forming minerals ensures a supply for new mineral
growth, given a chance. That chance may take place during dry
seasons which can assume suddenly and salts have the
opportunity to accumulate within voids, fractures, and grain
boundaries. Salt weathering plays a role in the granular disintegration and cavernous weathering of coarse crystalline
rocks observed in wetdry tropics as well as arid regions
(Young, 1987; Turkington and Paradise, 2005). Seasonal
wetdry tropics are capable of sustaining pedogenic gypsum in
soils over carbonate rocks (Luzzadder-Beach and Beach, 2008),
another possible source of crystal expansion by means of hydrating calcite. Expansive clays and neoformed iron oxides may
also exert pressure (Nahon and Merino, 1997). Silica reprecipitation after dissolution can be responsible for further
opening grain boundaries and fractures at the micron scale and
lattices and crystal faults at the nanometer scale (Chapter 4.4).
Pressure unloading, sometimes known as dilation or
sheeting, is the relief of overburden stress that causes expansion and then brittle fracture of formerly buried rocks. Resistant rock bodies, by way of differing petrology or structure,
survive weathering and erosion to become exposed as domeshaped remnants (bornhardts, inselbergs, tors, or other related
terms). The exposed outer surfaces are thus vulnerable to
pressure release, fracturing parallel to the rock surface and
normal to the surface to release slabs. Twidale (1973) offered
an opposing opinion that dome-shaped jointing preexists exposure by way of compression (not extension), such that
domed inselbergs are so because of their fractures, not that the
fractures are so because the rock is domed. Regardless, although the phenomenon is commonly observed in doming
rocks of various lithology in the tropics (Figure 2, see also
Shroder, 1973), the process is not limited to the tropics.
It is important to note that the mechanical weathering
processes, except for crystal growth of neoformed minerals, are
restricted to and determined by surface conditions. Because
the weathering profiles may be many meters in thickness,
these surface conditions and processes are but a fraction of the
total weathering system (Ahnert, 1976).
The combination of abundant weathering agents and
higher temperatures ensures the potential for an active
chemical weathering environment in the tropics. That said,
weathering end-products the kaolinite, gibbsite, and iron
oxides common in tropical soils and regolith also indicate
an eventual chemical stability, explaining the dearth of nutrients available in some tropical soils. Details of chemical
weathering are best explained in Yatsu (1988), Nahon (1991),
186
and Taylor and Eggleton (2001), but summarized here with

emphasis on the tropical relevance.
Solution and dissolution are most prominent among the
chemical weathering reactions, with widely recognized results
in the tropics. Solution is the simpler of the two, occurring in a
single-step process, also known as congruent. The solution of
calcium carbonate is commonly cited as a good example.
Quartz, although resistant (Goldich, 1938), also dissolves
congruently in water:
SiO2 2H2O H4SiO4
The resultant silicic acid, H4SiO4, can be transported out
in surface water or groundwater, but also has the ability to
dissociate and reprecipitate silica as neoformed quartz or
amorphous silica, relevant in the process of cementing sediments, creating duricrusts in regolith, or in case of hardening
of boulders (Conca and Rossman, 1982). Silica solution is
generally seen as a minor process compared to the dissolution
weathering of other silicate minerals, and slow. However,
studies by Schulz and White (1998) and Murphy et al. (1998)
show that chemical weathering of quartz in a tropical
envrionment generates 2575% of the dissolved silica in
regolith pore water (over all other silicate minerals). Solution
also generates smaller particles (see Chapter 4.17; Pye (1983))
attributed tropical humid weathering of Pleistocene sand
dunes to the formation of silt-sized quartz, which accumulated to 10% of the bulk sediment in the B and C horizons of
the soil. Quartz solution is also the process that is responsible
for the generation of silica karst (see Section 4.11.3.1).
Most aluminosilicate minerals undergo dissolution, also
known as incongruent solution or hydrolysis, a multistep and
parallel process involving acids. The generalized process involves the attack by water and acid to produce a clay, possible
other neoformed minerals, cations in solution, and silicic
acid. Water itself is a weak H proton donor, but acids are
much more efficient. Carbonic acid is the default and ubiquitous acidic weathering agent, via rain water charged with
atmospheric CO2, or soil water charged with CO2 from the
soil air (concentrated more than two orders of magnitude
higher, when compared to the atmosphere, Ugolini and
Sletten, 1991). Organic acids, derived from organic decay as
well as biotic functions (such as plant roots), are also important (Ugolini and Sletten, 1991), and possibly even dominant in some instances (Wasklewicz, 1994).
The dissolution process of the feldspar mineral albite in the
presence of water and carbonic acid (implied with the inclusion of CO2) is a good example:
2NaAlSi3 O8 3H2 OCO2 -Al2 Si2 O5 OH4 4SiO2 2Na 2HCO3
albite
kaolinite
quartz ions in solution
Further, kaolinite can dissolve to gibbsite (typical of

bauxitic laterite, a weathering residual) and silicic acid (carried
away in aqueous solution):
Al2 Si2 O5 OH4 105H2 O-AlOH3 42H4 SiO4
kaolinite
gibbsite
silicic acid
What distinguishes solution from dissolution depends on
the parent material (mineral), but also the supply of water as a
weathering agent or weathering agent medium, hence responsive to different variations of tropical moisture. Taylor and
Eggleton (2001) explain that during incongruent dissolution,
there are intermediate stages of dynamic equilibrium. Saturation and mineral neoformation would take place during periods of water limitation, a temporary chemical equilibrium.
Addition of new water rejuvenates the system, establishes
chemical disequilibrium, and the remaining primary minerals
along with neoformed minerals are subject to attack.
The process of oxidation is essentially inseparable from the
dissolution process. Oxidation is relevant to iron-bearing, and to
a lesser extent manganese-, titanium-, and sulfate-bearing minerals. Several of the primary rock-forming minerals are ironbearing: biotite, olivine, amphiboles, and pyroxenes. Oxidation
alters the crystal structure which in turn leads to a weakened rock
fabric, which in turn allows further penetration of other weathering agents (Taylor and Eggleton, 2001). At the same time,
oxidation is responsible for fixing stable iron oxides, and parallel
to hydrolysis, also creates some dissolved silica. Olivine, an ironbearing aluminosilicate in many igneous rocks, provides a good
example of an oxidation reaction in the presence of water:
2Fe2 SiO4 H2 O O2 -FeO OH dissolved silica
olivine
goethite
Further, goethite dehydrates to form hematite. Iron oxides
such as goethite and hematite are stable and residual in the
soil and weathering profile. These oxidized minerals impart
the vivid yellow (goethite), orange, and red (hematite) colors
to tropical soils.
Hydration is a process similar to oxidation, in which hydroxide (OH) ions, rather than oxygen, are incorporated into
the mineral matrix. Phyllosilicates, including clays, are most
notable for hydration, where hydroxide ions are incorporated
between silicate layers. Yatsu (1988) considered hydration to
be a mechanical rather than a chemical process, an argument
parallel to that presented in Chapter 4.4.
Biochemical processes are now recognized as important to
weathering (Krumbein and Dyer, 1985; Reith et al., 2008),
and involve a suite of reactions including those mentioned
above as well as chelation, a uniquely biochemical process.
Ollier and Pain (1996) explained that oxidation is involved in
a plants uptake of iron and other nutrients by way of the
roots. Silica depletion is said to be enhanced by bacterial action (Ollier and Pain, 1996). McFarlane (1987) demonstrated
the importance of microorganisms in the evolution of bauxite.
Chelation is the process by which metals are preferentially
extracted by organic molecules, derived from decomposing
vegetation. It is presumed, but not well researched, that rapid
organic decomposition in rainforest soils could produce an
abundance of chelating weathering agents. Tropical soils do
harbor an immense diversity of microbes, concommitant with
the above-ground biodiversity (Borneman and Triplett, 1997).
4.11.2.3
End Products of the Weathering Process
The totality of weathering processes evolves continuously to

eventual stable, low-potential-energy weathering products.
Reiches (1950) graphic representation of weathering potential
versus weathering product best illustrates this evolutionary
187
Hematite, Goethite
100%
100
90
10
Ferrite
20
Alu
fer min
rite ous
40
s
ou
i ce
Sil rite
fer
30
80
70
60
50
50
40
60
u
er o
rrif
Fe olin
ka
Laterite
30
20
80
Fe
ba rrifer
ux ou
ite
s
70
90
10
Bauxite
100
0
10
20
Bauxititious
kaolin
Siliceous
bauxite
30
40
100%
Gibbsite + Boehmite
50
60
70
Kaolin
80
90
0
100
100% Kaolinite and

other clay minerals
Figure 4 Weathering end product ternary diagram. Modified from Bardossy, J., Aleva, G.J.J., 1990. Lateritic bauxites. Developments in
Economic Geology 27, 624.
path (Figure 3). The weathering product index (WPrI) is the

ratio of silica to combined silica titanium sesquioxides
(e.g., iron and aluminum oxides), in molar oxide form, and
decreases in value as silica leaches out with respect to less
mobile titanium and sesquioxides
WPrI
molesSiO2
molesSiO2 TiO2 Al2 O3 R2 O3
The WPol is the ratio of the alkaline earths to the total of

all common elements (also in molar oxide form). The potential index decreases as alkaline earths preferentially leach
A fresh igneous rock, for instance, has a high weathering
potential and high product ratio (high proportion of silica to
aluminum and iron oxides). End-stage bauxite and laterite
have very low weathering potential and low product ratio
(greater dominance of the sesquioxides). Intermediate mineral
phases such as clays fall between the extremes.
The tropics are well known for end-stage weathering
products such as iron, silica duricrusts, and residual alumina.
Based on end-stage weathering products, Strakhov (1967) and
Pedro (1968, 1983) assigned regions of dominant weathering
processes (see also Figures 1(b), (c)) according to present
climatic conditions:
WPol
Allitization total loss of silica and alkaline elements, production of gibbsite (the aluminum oxide residual), ferric
hydrates, and 1:1 clays (such as kaolin); centered in the core
tropics;
Monosiallitization partial loss of silica, total loss of alkaline elements, production of 1:1 clays (kaolinite) and ferric
hydrates; in the paratropics and subtropics (with a secondary frequency in the subpolar taiga zone, according to
Strakhov, 1967);
Ferrallitization the production of ferrisols and ferricrete in
a transition zone beyond the wettest rainforest to the seasonally wet savannas.
(definitions from Thomas, 1994; Pedro, 1983; and Schaetzl
and Anderson, 2005)
Laterite and laterization fall in within these definitions,
though the terms are complicated, as are correlating terms and
regions described by Pedro and Strakhov. Bourman (1993)
and Ollier and Pain (1996) contended that the use of the term
laterite was too diverse, Ollier and Pain preferring to fold it
into the rubric of ferricrete or iron-based duricrusts. Thomas
(1994), however, described the range of laterite, bauxite, and
similar duricrusts by way of a Fe-kaolinite-gibbsite ternary
diagram (Figure 4) after Bardossy and Aleva (1990). Widdowson (2007) generalized laterite to be an iron-rich,
molesCaO Na2 O MgO K2 O H2 O

molesSiO2 TiO2 Al2 O3 Fe2 O3 Cr2 O3 CaO Na2 O MgO K2 O
188
autochthonous weathering product of tropical or subtropical

conditions, and distinct from allochthonous ferricretes,
though both are end members of the same spectrum of ironbased weathering residua.
As per Pedro and Strakhov, it was assumed that lateritic
processes required tropical conditions to form, given the
obvious frequency of laterite in the tropics. Still, lateritic
end products are not unique to the tropics, a point known
for quite some time. In his chapter on tropical weathering,
Reiche (1950) noted the presence of bauxite in Arkansas
(USA) and laterite in the Mediterranean, and cited Jennys
(1941) mention of laterization in the Appalachian (USA)
Piedmont as well as Goldschmidts (1928) observation of
aluminum hydrates in Norway. Dury (1971) assigned duricrust formation at several midlatitude locations to humid
tropical conditions at the Eocene optimum; therefore the
current exposures are relict. With the ability now to date or
calibrate weathering profiles, several researchers pin ages to
weathering formations and their paleoenvironments. Cecil
et al. (2006) used (U-Th)/He thermochronology to date the
exhumation history of the northern Sierra Nevada, California,
and thereby equate a period of tectonic stability with a lateritic
paleosol, the Ione formation, of Eocene age. Although this
does not conclusively determine that laterite followed warmwet conditions, Yapp (2008) established a paleotemperature
5 1C warmer than present, and wetter, at the same location
and time period, using oxygen isotope ratios derived from the
paleosol. Similarly, Retallack (2007) concluded that wet/warm
conditions existed in northwest and west-central North
America based on the weathering of Eocene, Miocene, and
Pliocene paleosols.
The questions arise, then, were much larger regions subject
to tropical processes, did the weathering take place when a
landmass was at a different paleolatitude, or can the formation
of tropic-like regolith proceed without tropical conditions?
Are these nontropical examples exceptions to the rule, or
within the spectrum? Conversely, why should end product
FeAlSi weathering residua appear so commonly in the tropics? There are several answers. It is easy to assume that the
weathering derives from the distant tectonic past when currently nontropical land areas were at one time situated in or
near the tropics, and have tectonically drifted over time out of
the tropics. Some demonstrably old weathering profiles may
fit this category, and it is possible to verify the age of these
examples (Pillans, 2008) in order to correlate with their
paleotectonic geography.
However, weathering-then-tectonic drift does not necessarily explain all tropical profiles in the nontropics. A
growing number of authors now recognize that genesis of
these end products do not require tropical conditions (cf.
Paton and Williams, 1972; Bird and Chivas, 1988; Ollier,
1988; Taylor et al., 1992; Bourman, 1993, 1995). Abundant
moisture and time appear to be key factors, as well as exposure
to groundwater in the weathering profile. Recognition of ferrous end-product saprolite in some present day drylands (see
Figure 5(a)) may derive from wetter periods in the geologic
past, or simply very long and stable exposures. Certainly the
end-product regolith of coastal California (Burke et al., 2007),
the California Sierra Nevada (Cecil et al., 2006; Yapp, 2008),
the Rocky Mountains (Wanty et al., 1992; Figure 5(b)), the
(a)
(b)
Figure 5 Examples of deep weathering. (a) Oxidized end-product

saprolite exposed on the side of a granitic inselberg, near Phoenix,
Arizona (USA). Photo courtesy of R. Dorn. (b) Deep weathered grus
with corestones, near Pikes Peak, Colorado (USA). Presumed to be
the result of warmer climates of the Eocene.
Appalachian Piedmont (Pavich, 1986), and Bohemian Massif

(Vitek, 1983) benefit from abundant precipitation.
Second, long exposure and stable tectonics tend to favor
preservation of deep and highly evolved weathering residua.
Thus, deep tropical weathering mantles survive. It so happens
that the Brazilian Shield, Australia, and the tropical African
Plateau satisfy these conditions. Though Thomas (1994: 19)
downplayed this chance relationship to deep weathering,
several subcontinental areas have been placed in an equatorial
position and relatively unscathed by major orogenic events
for more than 20 million years (more than 100 million years
in some cases). Table 1 compares a rough estimate of the
proportion of gross geomorphic surfaces between tropical
and mid-latitude belts, showing that exposed Precambrian
shields are more than three times more prevalent in tropical
zones than they are in midlatitudes. In other locations of the
world with suitable weathering conditions and tectonic stability, there are also deep weathering mantles. Surprisingly,
these can survive major glaciations; remnant saprolites, tors,
and inselbergs remain in Scotland (Hall and Mellor, 1988),
on the Fennoscandian Shield (Lidmar-Bergstrom, 1995;
189
Table 1 Comparison of landform categories between Tropical and Mid-Latitude zones. Landform classification is based on that of Murphy
(1968)
Landform/region
Recent mountains of the Alpine

Orogeny, newly rifted areasa
Caledonian/Hercynian/Appalachian
mountain remnantsb
Exposed Gondwana or Laurasian Shieldc
Sediment-covered lowlandsd
Tropics and subtropics (301 N to 301 S)e
Mid-latitudes (311 N to 601 N 311 S to 601e
% of latitude zone
% of earths land
area
% of latitude zone
% of earths land
area
29
17
30
12
15
43
27
25
15
13
42
5
17
The post-Jurassic Alpine orogenic belt includes the Cordillera of the Western Hemisphere and Circum-Pacific subduction chain, and the Eurasian-Himalayan System, mountains or
widely-spaced mountains. Newly rifted areas include the African-Red Sea rift system, the Flinders region of Australia, and the Baikal Rift System.
b
The Caledonian, Hercynian, and Appalachian range mountain remnants are from Paleozoic to early Mesozoic orogenies and have had no significant rejuvenation.
c
Exposed surfaces of the Gondwana and Laurasian shields are primarily Tertiary erosion surfaces on mainly Precambrian parent material, and include tablelands, plains, and some
mountains.
d
Sediment-covered lowlands include plains and low hills, mostly covering shield areas with Paleozoic to Recent sediments.
e
Latitudinal belts Include areas of tropical desert or mid-latitude desert.
Percentages are calculated based on combined area of 51 51 grid cells, in which each cell is assigned a single dominant landform class.
Ebert and Hattestrand, 2010) and on the Canadian Shield

(Bouchard and Jolicoeur, 2000). Where erosion has become
more efficient, weathering residua have been removed, exposing etch surfaces.
4.11.2.4
Rates of Weathering
Of the myriad factors that influence weathering processes

(Pope et al., 2005), time is probably the dominant factor in
the evolution of major weathering landforms. Simply, the
more time evolved, regardless of other factors including climate, the greater the degree of weathering. Landscapes of great
antiquity usually exhibit the most extensive weathering landforms. Likewise, antiquity allows for inheritance of landforms
from previous environments (Thomas, 1994).
Weathering rates in tropical regions are generally accelerated, following the expectations of the Arrhenius equation
temperature kinetics (increasing rate with increasing temperature), demonstrated by numerous studies (cf. Haantjens
and Bleeker, 1970; Dorn and Brady, 1995; Navarre-Sitchler
and Brantley, 2007). White et al. (1998) described a regolith
propagation rate of 58 m Ma1 (or 58 B, where B Bubnoff
units 1 mm per 1000 years) in Puerto Rico. They considered
this rate of chemical dissolution of silicate rock (granodiorite
in this case) to be the fastest on Earth, and happens to be
equivalent to Thomass (1994) presumed maximum rate of
tropical weathering, and an order of magnitude greater than
the global mean of B6 B. Saunders and Young (1983), in their
review of denudation studies, reported a range of 2 to 15 B for
chemical denudation or weathering of silicate rocks in regions
spanned by the tropics (subtropic wetdry to tropical rainforest); and 11 to 500 B for carbonate rocks. This compared to
a range of 0126 B for silicate rocks and 13210 B for carbonate rocks of the temperate midlatitudes. Rates of denudation overlap for several reasons. Apart from temperature, other
factors account for variation in weathering rates. Moisture or
precipitation (and subsequently stream runoff discharge) is a
prominent factor in weathering rates (cf. Langbein and

Dawdy, 1964; Haantjens and Bleeker, 1970; Dunn, 1978;
Bluth and Kump, 1994; Li et al., 2011). Organic chemistry
(Viers et al., 1997) and presence of organic weathering agents
(Dorn and Brady, 1995; Kelly et al., 1998), lithology (Bluth
and Kump, 1994), strength of weathering agents, including
variations in pH (Casey and Sposito, 1992; Cama et al., 2002),
soil and weathering profile depth (White et al., 1998), seasonal hydroclimatic changes (Li et al., 2011), and topographic
or geomorphic position (Stallard, 1992) are all influential in
determining weathering rates.
There are some cases where tropical weathering rates are
not as great as expected. In Sri Lanka, von Blanckenburg et al.
(2004) reported a low weathering and denudation rate, despite high temperatures and precipitation. In this case, an already-weathered plateau (low on Reiches (1950) Weathering
Potential and Weathering Product indexes) was the source of
chemical denudation. Chemical denudation rates in Taiwan
were also lower, but for the opposite reason, according to
Selveraj and Chen (2006). In steep terrain, immature sediments had only a brief residence time within the weathering
system; the topographic factor dominated the mountain environment. In this system, Selvaraj and Chen claimed that
physical weathering dominates, a direct contradiction of
Peltiers (1950) prediction that chemical weathering should
dominate in warm, wet regions.
4.11.2.5
Weathering Maxima Outside the Tropics
Several studies, from disparate sources, suggest weathering

process maxima unrelated to tropical characteristics, but possibly responsible for tropical-like landforms. Strakhov (1967)
is frequently referenced. Apart from the major overwhelming
tropical factors, he suggested a midlatitude/subpolar submaximum of weathering depth (Figure 1(b)). Like comparable works of the period, there was ample simplification, and
most modern weathering process studies focus not on the
190
topdown factors (regional climatic) but rather the bottom up

(microscale dominance, see Pope et al., 1995). A novel reversal of this was undertaken by Scull (2010), a study of soil
forming factors in the continental US, but still relevant in that
outcomes such as profile development and clay genesis are
also weathering events. Sculls spatial model, utilizing finescale environmental data based on temperatures and precipitation, was more black-box, with future work required to
elucidate the reasons for relationships. What is interesting at
this stage is the variegated spatial correlation between environmental factors and soil (weathering) factors. The study
showed strong positive correlation between total clay and
precipitation (increasing precipitation - increasing clay) in
New England and portions of the Midwest (regions with
moderate to abundant precipitation), but not in the South or
Pacific Northwest, equally if not more humid but quite different in temperatures. The assumption of the Arrhenius
temperature function associated with weathering rate was not
apparent. When temperature, in turn, was related to total clay,
areas of strong positive correlation (increasing temperature increasing clay) appeared along the Pacific Coast and in a
variegated pattern in the Central US from the Gulf Coast to the
Canada, but poor or negative correlation in warm locations
such as the Deep South and Desert Southwest. Again, tropical
temperature and moisture factors appeared to show no logical
spatial gradient.
Another detailed attempt at topdown weathering factors
was made by Fowler and Petersen (2004), applying Peltiers
(1950) parameters (Figure 1(a)) to predict theoretical weathering regions in the continental US using fine-scale climatic
data. Independently, Pope and Pobanz (2011) used coarser
climate data but a combined chemical mechanical weathering potential in one index. Both studies predicted strong
chemical weathering in the Gulf Coast states expected as the
outer fringe of the subtropics and also mountainous locations
in the Southern Appalachians and Northeast States, and a wide
continuum of strong chemical weathering in the Sierra Nevada,
Cascade Range, and Coast Ranges of the Pacific Coast states. It is
interesting that abundant moisture and related biomass were
seen as sufficient for aggressive chemical weathering despite
cooler temperatures, particularly in the Sierra Nevada case
where oxisols were shown to be formed in warmer, wetter climatic conditions (Yapp, 2008). Not only does this partially
match Sculls (2010) predictions for deeper soil profiles or
higher clay genesis, it also closely corresponds with the occurrence of ultisols (of any suborder) in the US. Further work on
these geospatial models would use proxy weathering data for
validation, for instance, small stream solute loads or depth of
weathering profile. A more appropriate estimate of temperature
would be an integration of temperatures over the presumed
lifetime of the weathering profile, similar to Wehmillers (1982)
Equivalent Quaternary Temperature curve used for amino acid
racemization geochronology.
4.11.3
Weathering-Related Landforms of the

Tropics
Legitimate questions exist as to whether landforms said to

be tropical are truly dependent on tropical conditions of
weathering and erosion. The discussion of weatheringrelated landforms here is classified by means of generalized
morphology introduced in Chapter 4.1 of this Treatise: weathering voids, weathering resistant landforms, and weathering
residua. This classification has an inherent scalar and temporal
organization. At increasing spatial and temporal scales, specific weathering processes diminish in importance replaced by
the works of the entire weathering system.
4.11.3.1
Weathering Voids: Solutional Landforms
Karst geomorphology is, of course, a product of solutional

weathering. The acidic properties of groundwater act on sedimentary rock (generally, but not limited to, carbonate rocks) to
produce caves, karst landscapes, and microscale solution features. Karst geomorphology is included along with weathering,
lumped into the same chapter, in most introductory textbooks,
but the uniqueness and variety of karst geomorphology justifies
complete and separate treatment (Frumkin, 2013). This chapter
makes brief mention of solutional landforms relevant to weathering in tropical regions.
Carbonate karst refers to solutional features primarily not
only in limestone but also in marble, dolomite, and some
carbonate-cemented sandstones. As much as these rock types
are quite common across the planet, karst landscapes are also
widely distributed. Dramatic karst landscapes of southeast
Asia and Indonesia, Central America, and the Caribbean attest
to the active elements of solutional weathering in the tropics:
abundant rainfall and high CO2 content via high biomass,
combined with fast chemical reaction due to warmer temperatures (Monroe, 1976). One aspect of the greater relief
(and more common representation) of karst development in
the tropics is the lack of interfering geomorphic processes
(such as glaciation and periglaciation). Still, extensive karst is
also seen in temperate regions, in areas such as southeast
Europe (the type locality of karst), England, the Ozark and
Applachian highlands of the US, and southern Australia. Extensive cave systems perhaps evolved over several million years
(Granger et al., 2001), though many sizeable systems are
Quaternary in age, and surface karst formations seen in England, Ireland, and Germany are certainly postglacial. The
combination of precipitation and CO2 saturation in groundwater is sufficient for karst developent in these areas. Major
karst systems are thought to be influenced by warmer and
wetter climates; Ford (2010), Twidale (2002), and Maslyn
(1977) suggested that the significant tower karst evident in
mid- and high-latitudes are exhumed weathering relicts from
the warm-humid past (either by paleogeography or climate
change).
Not as common as carbonate solution (though more
common than previously thought), silica karst solution
landforms in sandstone, quartzite, and siliceous igneous rocks
are recognized in numerous locations. Wray (1997) see also
Chapter 6.36 and Young et al. (2009) provide the most
complete review of silica solutional landforms. Wray (2003)
argued that any rock-solutional feature, including those in
silicate rocks, is true karst (as opposed to pseudokarst). Because of the higher activation energies of dissolving quartz,
warmer temperatures would be most condusive to silica karst,
and indeed silica karst was first recognized in the tropics

(Martini, 1979; Twidale, 1984; Young, 1986). In tropical environments, the constant supply of water establishes a continuous chemical disequilbrium, pertinent to the congruent
solultion process. That said, there are silica karst regions outside of the tropics. Netoff et al. (1995) and Netoff and Chan
(2009) reported large doline-like pits in the Entrada and
Navajo sandstones of arid southern Utah (USA). They attributed the formation of these pits to mainly mechanical weathering processes (such as clay and salt growth) and some
solution of calcium cements in the sandstone matrix, with
debris winnowed out of the pit bottoms by wind. But, although the region is presently arid, wetter conditions were
possible at times over Quaternary period when chemical
weathering may have been more efficient. May and Warne
(1999) theorized that the so-called Carolina Bays (elliptical,
oriented basins, and ponds found in coastal sediments along
the US Atlantic and Gulf coasts) are silica-karst features, and
include alteration of kaolinite to gibbsite and comcomitant
loss of volume, hence sinkholes. This is but one of many
different theories used to explain Carolina Bays. Vitek (1983)
and Demek and Kopecky (1994) recognized pseudokarst
forms, including tower karst, in the sandstones of the Bohemian Massif. Vitek (1983) suggested that their development
occurred in recent mild humid conditions. Other rock forest
or rock city formations may qualify as silicate karst. Cammeraat and Seijmonsbergen (2010) reclassified the Bosques
de Rocca area in the Peruvian Andes, an ignimbrite formation
widely thought to be wind-eroded, as silica karst. The vitreous
nature of ignimbrite would have a lower solution threshold
than quartz, compensating for the cooler temperatures and
slower rates at high altitudes. This author noted similar pinnacle formations in ignimbrite and tuff in the San Juan
Mountains of Colorado (USA) and the famous Cappadocia
region of Turkey. City of Rocks, in southwestern New Mexico
(USA), may be another example. Described by Mueller and
Twidale (1988) as a joint-controlled, exhumed etch surface
formed in a warmer-wetter subsurface environment, which may
well be correct, the difference between solutional karst and
dissolutional subsurface etching is probably fuzzy.
4.11.3.2
Weathering-resistant landforms include positive-relief features

that are first more durable to weathering attack, and then
secondly more resistant to erosion. These include the inselbergs (boulder or domes) and plateaus, and are resistant
mostly for structural or lithologic reasons (Twidale, 1982,
2002; Migon, 2009) or heterogeneous groundwater distribution, less a factor of subaerial weathering agents or regional
or microclimates. Still, many introductory textbooks include
a picture of a bornhardt or boulder inselberg as an example
of resistant landforms in the tropics. Indeed, weatheringresistant landforms have been a focus of research in tropical
geomorphology, even if weathering resistance is not a function
of climate. Thomas (1994:343) noted the widespread occurrence of boulder inselbergs (including tors) and domed inselbergs regardless of latitude in the tropics and subtropics,
suggesting that their distribution appears not to be controlled
by climate. However, tropical weathering provides ideal
191
conditions for their development: periods of aggressive deep

weathering.
Classic domed inselbergs (bornhardts) abound in the tropics, including Brazil, Australia, South, West, and East Africa,
(Figure 2) and at the tropical fringes as far as Texas and
Georgia (cf. Shroder, 1973; Twidale, 1982; Thomas, 1978,
1994). They are massive, not necessarily of a single lithology
(Petersen, 1988), but distinctly less jointed and fractured
compared to surrounding terrain (except for the near-surface
parallel joints that develop out of expansion to create unloading slabs). Thomas (1994) believed that deep weathering
and saprolite development takes place preferentially along the
well-jointed rock. Later, this easily eroded weathered rock is
stripped, leaving the resistant bornhardt. The domed morphology is a result of pressure release due to exposure, or as
Twidale (1973) argued, preexisting compression stress due to
pluton emplacement.
The formation of boulder inselbergs is discussed thoroughly by several authors (Linton, 1955; Thomas, 1978;
Twidale, 1982; Ollier, 1984). Included here are the conical,
boulder-mantled inselberg hills, as well as classic tors. Boulder
inselbergs are formed in a similar manner to bornhardts, except that the preexisting rock is more jointed, allowing for
stacks of segmented, in situ, spheroidally weathered boulders.
The size of these boulders depends on the degree of weathering and the spacing of preexisting joints. Tors occur on all
continents including the present extremes of temperature and
moisture. Where they are seen outside of the present tropics, it
is sometimes assumed that previous tropical conditions were
responsible for the first part of their formation, the deep
weathering that isolates resistant blocks and corestones (cf.
Linton, 1955; Cunningham, 1969). A completely different
explanation uses frost weathering and periglacial slope action
to expose the resistant corestones of tors (Palmer and Neilson,
1962). Yet, chemical weathering is valid even in cold regions,
and Derbyshire (1972) attributed chemical weathering processes responsible for the formation of tors in favorable
microclimates within the prevailing polar desert conditions of
Southern Victoria Land, Antarctica. Mechanical weathering by
ice and temperature extremes is likely in this case, though the
granular disintegration and rounding of edges is typical of
chemical weathering.
The evidence of tors in tropical and nontropical locations as
diverse as West Africa, Dartmoor, the central Rocky Mountains,
the Bohemian Massif, Portugal, the Mojave Desert, and even
Antarctica is one to suggest convergent evolution (cf. Campbell
and Twidale, 1995; Twidale, 1984: 333334). Weathering
landforms are often seen as good examples of convergent
evolution, or the alternate term, equifinality. Harrison
(2009:359), for instance, exemplified tors as developing in either periglacial action or by deep chemical weathering. Convergence or equifinality presume that different processes in
different climates are responsible for similar forms, an idea
parallel to the biologists convergent evolution (Dendy, 1916).
This may not be the case, in the strictest sense: It is possible that
geomorphic processes themselves (in this case, chemical weathering) are identical or at least very similar. The only difference
might be the rates of change over time (through concomitant
climatic change) and the sequence of events. Antarctic and
Tropical Savanna tors would be convergent if formed
192
differently, if chemical weathering was responsible for the

granular disintegration and block rounding in the warmer
regions, whereas ice weathering was responsible for the same
in cold regions. It is more likely that chemical weathering
is active in both areas, and that tors were primarily the result
of chemical weathering, thus not convergent but in fact
identical.
Both boulder and domed inselbergs have an implication
of climate change in their genesis (Thomas, 1994). Deep
weathering would progress under one regime (tropical conditions, for instance), whereas exhumation and stripping would
require either seasonal or drier conditions (swinging, for instance, toward the weathering-limited direction of the
spectrum).
Although Migon (2006), Twidale (1982), and others
questioned the requirement of tropical conditions for many
bedrock landforms, Migon (2009) revived the notion of
granitic multiconvex topography as occurring in the tropics
and not reported in any other climatic region. Multiconvex
topography falls within a spectrum of morphologies of weathered granitic terrain extending from plains to all-slopes
topography of considerable relief and steep slopes (Migon,
2006). Multiconvex topography, also mentioned by Thomas
(1994), with alternate names demi orange or meias laranjas
hills was described as having developed out of deep weathering, consisting of closely-spaced, irregularly distributed
low hills resembling half-cut oranges. These hills retain weathered mantles, and some may be weathered throughout, establishing a sort of equilibrium in form between continued
weathering mantle development and erosion. Unlike other
bedrock weathering landforms, it is thought that multiconvex
topography and the weathering-stripping equilibrium cannot
survive major climate changes, and thus exists entirely within
the tropical regime (in areas that have retained consistent
tropical characteristics). Rates of formation and times of persistence of these and other resistant landforms is worthy of
further investigation.
A class of weathering-resistant landform involves duricrusts,
which protect surfaces. Duricrusts are the product of weathering, cemented by secondary mineralization of weathering
products. Laterite is a duricrust (Widdowson, 2007), as is silcrete (Nash and Ullyott, 2007). Silcrete, like the iron duricrusts,
occurs in many type of environments, not just the tropics, and
may be the result of pedogenic, weathering, or groundwater
processes. Laterite, ferricrete, and silcrete are capable of indurating surface regolith, and like resistant caprock, can form
plateaus by protecting softer underlying regolith. Where silcrete
has been deposited in association with stream courses, sinuous
paleochannels can be protected and remain in positive relief as
the remaining landscape erodes down.
4.11.3.3
Deep Weathering Mantles
Deep weathering mantles are fairly common and familiar in

the tropics but certainly not limited to the tropics
(Figure 5(b), see also Chapter 4.8) Deep weathering is naturally associated with the humid tropics, with plentiful
moisture and biomass. Thomas (1994: 80) outlined ten optimal conditions for deep weathering, occuring in any combination of conditions:
1. wet equatorial or monsoon climates, rainfall

41500 mm yr1, together with rainforest vegetation;
2. predominantly warm and humid conditions during substantial periods (B106 years) over B108 years of paleoclimatic history;
3. cratonic terranes in continental interiors or passive
margins;
4. domed and plateau uplifts with strong tensional stress
fracture patterns;
5. shear zones and intersections of dense fracture patterns;
6. hydrothermal altered rock (diagenesis, not weathering);
7. fissile metamorphic rocks or igneous rocks with dense
microcrack systems;
8. old (pre-Neogene) land surfaces at moderate to high altitude, even in suboptimal climates, where long-term denudation rates have been low;
9. proximity to structural depressions promoting a strong
groundwater gradient;
10. free-draining sites beneath interfluves and hillslopes
o201, particularly if protected by a duricrust cap.
Note that only the first two factors concern climate, and
therefore relevance to tropical conditions. All others are rooted
in climatically-azonal lithologic, tectonic, topographic, or
historical conditions. Chapter 4.3, discusses processes in
which chemical weathering in cool environments would be
responsible for extensive soil development and deep
weathering.
The classic transport-limited (versus weathering limited)
landscape relationship is relevant here, keeping in mind other
factors also responsible for erosion or lack of erosion of
regolith (including vegetation cover, slope, and precipitation
intensity). Where terrain is stable, and not liable to rapid
erosion, regolith greater than 100 m thickness is possible
(Thomas, 1994; Ollier and Pain, 1996). Deep weathering
profiles are condusive to further weathering because they are
commonly in contact with groundwater, and tend to be moist
all the time above groundwater (Ollier, 1988). Ahnert (1976)
expressed a zone of optimal chemical weathering somewhat
below the surface, decreasing as overburden cover increased.
Recent evidence from Burke et al. (2007) refined this model
(but in granite terrain of coastal California) to verify that soil
production rates, chemical weathering indexes, and acidity
decreased with increasing soil thickness, but immediately
from the soil surface and not, as Ahnert suggested, from an
optimum subsurface point.
The age of these weathering profiles is a matter of contention, and therefore the process of their creation would vary.
Climatic extremes may be important. A frequent response in
geomorphic studies is to assume that deep weathering is a
product of tropical conditions (cf. Pavich and Obermeier,
1985; Pavich, 1986; Kabata-Pendias and Ryka, 1989; Cecil
et al., 2006; Solback et al., 2009). These cases may well be the
result of tropical paleogeography or climate change, but development of deep weathering is possible in temperate to cool
climates (therefore in the midlatitudes during colder phases,
or seasonally). Molina Ballesteros and Canto Martin (2002)
questioned the need to invoke tropical conditions for deep
saprolites in Iberia, when time under climate conditions
analogous to todays may be sufficient. Deep weathering
profiles recognized in the Canadian Shield, Scotland, and the

Baltic Shield are attributed to preglacial or periglacial conditions, and are also perhaps quite ancient (Goodfellow, 2007;
Bouchard and Jolicoeur, 2000; Lidmar-Bergstrom, 1995; Hall
and Mellor, 1988).
Saprolites on Appalachian ridgetops and summits provide
an example of the debate between temperate or periglacial
weathering and tropical weathering. Cryoplanation, the periglacial process of weathering and freeze-thaw heave, is suggested for the creation of deeply weathered profiles in
highland areas of the eastern and central US (Clark and
Ciolkosz, 1988). Similar formations are reported in the Rocky
Mountains (Munroe, 2006). In these once-colder areas, summit lowering and weathered regolith to depths of several
meters are possible within a Pliocene to Pleistocene time
frame. Marsh (1999: 6163), however, argued for deep weathering that easily predates any periglacial climate. On the flat
ridge tops of the very resistant Tuscarora quartzite, there exists
an in situ weathered sandy soil approximately two meters
deep, underlain by eight to ten meters of weathered quartzite
saprolite. Further, subsurface profiles indicated an irregular
regolith boundary, similar to that produced by weathering and
not by periglacial process. Small, low quartzite tors crop out at
the summit. Marsh believed that Pennsylvania mountaintop
weathering profiles were much older, tens of millions of years
(and has been argued, perhaps even dating to late Paleozoic),
and not cryplanated in Pleistocene times. This would associate
the deep weathering to much warmer climates, both globally
and latitudinally for North America.
4.11.4
Conclusion
Weathering processes form an intergral part of tropical geomorphology, perhaps more so than any other environment on
earth. Areas dominated by the existence of deep and long-term
weathering are known for characteristic landforms and soils:
duricrusts, saprolite, resistant bedrock landforms, and solutional formations on both carbonate and silicate rocks. These
in turn influence the other geomorphic processes of this environment, including groundwater and surface water flow,
sediment transport and deposition by rivers, and slope
movements.
Weathering processes of the tropics are not unlike those
elsewhere on the planet, but owing to the availability of water
and enhanced temperature, rates of chemical weathering are
accelerated and more aggressive. Pervasive conceptual models
of weathering and weathering factors distinguish weathering
efficacy based on climate, though these generalizations tend to
obscure or marginalize observations that do not fit the models. Despite the obviousness of weathering landforms in the
tropics, similar landforms occur elsewhere, outside of the
present tropics. These outer examples initiate a debate, whether other processes are responsible for similar landforms
convergent evolution or whether climatic conditions change
over weathering-process time scales such that tropical conditions existed in greater areas at different invervals of the
geologic past. In fact, both sides of the debate can be true.
Climates have changed, and land masses that existed within
tropical regions have drifted out. As well, so-called tropical
193
landforms can exist without the temperatures of the tropics,

simply given abundant moisture (and weathering agents) and
increased time.
Recent research recognizes the spectrum of weathering
landforms, and accepts factors for weathering types and rates
relevant to specific situations, which may include climate but
may also find over-riding factors. The frontiers of weathering
geomorphology continue to work on integrating observations
across different environment. New techniques in surface and
regolith dating methods, geographic information systems, and
remote sensing, biogeochemical process modeling, as well as
continued refinement of the understanding of environmental
change at global as well as regional scales, will afford rapid
expansion of new ideas and integration with old ideas in order
to answer persistent questions.
References
Ahnert, F., 1976. Description of a comprehensive three-dimensional model of
landform development. Zeitschrift fur Geomorphologie Supplementbande 25,
2949.
Bardossy, J., Aleva, G.J.J., 1990. Lateritic bauxites. Developments in Economic
Geology 27, 624.
Bird, M.I., Chivas, A.R., 1988. Stable-isotope evidence for low-temperature kaolinitic
weathering and post-formational hydrogen-isotope exchange in Permian
kaolinites. Chemical Geology: Isotope Geoscience Section 72(3), 249265. http:/
/dx.doi.org/10.1016/0168-9622(88)90028-0.
von Blanckenburg, F., Hewawasam, T., Kubik, P.W., 2004. Cosmogenic nuclide
evidence for low weathering and denudation in the wet, tropical highlands of Sri
Lanka. Journal of Geophysical Research 109(F03008), 22. http://dx.doi.org/
10.1029/2003JF000049.
Bland, W., Rolls, D., 1998. Weathering: An Introduction to the Scientific Principles.
Bluth, G.J.S., Kump, L.R., 1994. Lithologic and climatological controls of river
chemistry. Geochemica et Cosmochimica Acta 58(10), 23412359.
Borneman, J., Triplett, E.W., 1997. Molecular microbial diversity in soils from
eastern Amazonia: evidence for unusual microorganisms and microbial
population shifts associated with deforestation. Applied and Environmental
Microbiology 63(7), 26472653.
Bouchard, M., Jolicoeur, S., 2000. Chemical weathering studies in relation to
geomorophological research in southeastern Canada. Geomorphology 32(34),
213238.
Bourman, R.P., 1993. Perennial problems in the study of laterite: a review.
Bourman, R.P., 1995. A review of laterite studies in southern South Australia.
Transactions of the Royal Society of South Australia 119(1), 128.
Branner, J.C., 1896. Decomposition of rocks in Brazil. Geological Society of
Budel, J., 1948. Das system der klimatischen geomorphologie. Verhandlungen
Deutscher Geographentag Munchen 27(4), 65100.
Budel, J., 1977. Klima-Geomorphologie. Gerbtruder Borntraeger, Berlin, pp.
Burke, B.C., Heimsath, A.M., White, A.F., 2007. Coupling chemical weathering with
soil production across soil-mantled landscapes. Earth Surface Processes and
Landforms 32(6), 853873. http://dx.doi.org/10:1002/esp.1443.
Cama, J., Metz, V., Ganor, J., 2002. The effect of pH and temperature on kaolinite
dissolution rate under acidic conditions. Geochimica et Cosmochimica Acta
66(22), 39133926.
Cammeraat, E., Seijmonsbergen, H., 2010. Significance of silica karst in ignimbrite
landscapes in the Peruvian Andes and their vulnerability. Geophysical Research
Abstracts, (European Geosciences Union General Assembly, May 27, 2010,
Vienna Austria) 12, EGU2010-13703, accessed at (http://
meetingorganizer.copernicus.org/EGU2010/poster_programme/2075) .
Campbell, E.M., Twidale, C.R., 1995. Lithologic and climatic convergence in
granite morphology. Cadernos do Laboratorio Xeoloxico le Laxe Coruna 20,
381403.
Casey, W.H., Sposito, G., 1992. On the temperature dependence of mineral
dissolution rates. Geochimica et Cosmochimica Acta 56, 38253830.
194
Cecil, M.R., Ducea, M.N., Reiners, P.W., Chase, C.G., 2006. Cenozoic exhumation
of the northern Sierra Nevada, California, from (U-Th)/He thermochronology.
Geological Society of America Bulletin 118(1112), 14511488.
Chorley, R.M., Schumm, S.A., Sugden, D.E., 1984. Geomorphology. Methuen,
London, 605 pp.
Clark, G.M., Ciolkosz, E.J., 1988. Periglacial geomorphology of the Appalachian
Highlands and Interior Highlands south of the glacial border a review.
Colmam, S., 1982. Clay mineralogy of weathering rinds and possible implications
concerning the sources of clay minerals in soils. Geology 10, 370375.
Conca, J.L., Rossman, G.R., 1982. Case-hardening of sandstone. Geology 10,
520523.
Cunningham, F.F., 1969. The Crow Tors, Laramie Mountains, Wyoming, USA.
Zeitschrift fur Geomorphologie 13, 5674.
Curtis, C.D., 1976. Chemistry of rock weathering: fundamental reactions and
controls. In: Derbyshire, E. (Ed.), Geomorphology and Climate. Wiley, London,
pp. 2557.
Demek, J., Kopecky, J., 1994. Geomorphological processes and landforms in the
southern part of the Policka Vrchovina Highland (Czech Republic). GeoJournal
32(3), 231240.
Dendy, A., 1916. Outlines of Evolutionary Biology, Second ed. D. Appleton and
Company, New York, 457 pp.
Derbyshire, E., 1972. Tors, rock weathering and climate in Southern Victoria Land,
Antarctica. In: Boulton, G.S., Price, R.J., Sugden, D.E. (Eds.), Polar Geomorphology.
Institute Of British Geographers Special Publication, vol. 4, pp. 9395.
Derbyshire, E. (Ed.), 1973. Climatic Geomorphology. Barnes and Noble Books, New
York, 296 pp.
Dorn, R.I., 2003. Boulder weathering and erosion associated with a wildfire, Sierra
Ancha Mountains, Arizona. In: Butler, D.R., Walsh, S.J., Malanson, G.P. (Eds.),
Mountain Geomorphology Integrating Earth Systems; Proceedings of the 32nd
Annual Binghamton Geomorphology Symposium Chapel Hill, NC, USA, October
2001. (Geomorphology 5(14), 155171.)
Dorn, R.I., Brady, P.V., 1995. Rock-based measurement of temperature-dependent
plagioclase weathering. Geochimica et Cosmochimica Acta 59(13), 28472852.
Dunn, T., 1978. Rates of chemical denudation of silicate rocks in tropical
catchments. Nature 274, 244246.
Dury, G.H., 1971. Relict deep weathering and duricrusts in relation to the
paleoenvironments of the middle latitudes. The Geographical Journal 137(4),
511522.
Ebert, K, Hattestrand, C, 2010. The impact of Quaternary glaciations on inselbergs
in northern Sweden. Geomorphology 115(12), 5666.
Eppes, M.C., McFadden, L.D., Wegmann, K.W., Scuderi, L.A., 2010. Cracks in
desert pavement rocks: further insight into mechanical weathering by direct
insolation. Geomorphology 123(12), 97108.1016/j.geomorph.2010.07.003.
Falconer, J.D., 1911. The Geology and Geography of Northern Nigeria. Macmillan,
London.
Ford, D., 2010. The South Nahanni: high-latitude limestone landscapes. In: Migon,
P. (Ed.), Geomorphological Landscapes of the World. Springer, New York, pp.
1320.
Fowler, R., Petersen, J., 2004. A spatial representation of Louis Peltiers weathering,
erosion, and climatic graphs using geographic information systems. 2004 ESRI
User Conference Proceedings, August 913, 2004, obtained at (http://
proceedings.esri.com/library/userconf/proc04/abstracts/a1752.html).
Frumkin, A., 2013. New developments of karst geomorphology concepts. In:
Shroder, J. (Editor in Chief), Frumkin, A. (Ed.), Treatise on Geomorphology.
Academic Press, San Diego, CA, vol. 6, pp. 113.
Gabler, R.E., Peterson, J.F., Trapasso, L.M., 2007. Essentials of Physical Geography.
Thomson Brooks/Cole, Belmont, California, 658 pp.
Goldich, S.S., 1938. A study of rock weathering. Journal of Geology 46, 1758.
Goldschmidt, V.M., 1928. The formation of laterite as a weathering product of
Labradorite Rock. Festskrift til H. Sorlie, pp. 2124.
Goodfellow, B.W., 2007. Relict non-glacial surfaces in formerly glaciated
landscapes. Earth-Science Reviews 80(12), 4773.
Goudie, A.M., Allison, S.J., McLaren, R.J., 1992. The relationship between
modulus of elasticity and temperature in the context of experimental simulation of
rock weathering by fire. Earth Surface Processes and Landforms 17(6), 605615.
Granger, D.E., Fabel, D., Palmer, A.N., 2001. Pliocene-Pleistocene incision of the
Green River, Kentucky, determined from radioactive decay of cosmogenic 26Al
and 10Be in Mammoth Cave sediments. Geological Society of America Bulletin
113(7), 825836.
Hall, A.M., Mellor, R., 1988. The characteristics and significance of deep weathering
in the Gaick area, Grampian Highlands, Scotland. Geografiska Annaler 70A(4),
309314.
Haantjens, H.A., Bleeker, P., 1970. Tropical weathering in the territory of Papua and
New Guinea. Australian Journal of Soil Research 8(2), 157177.
Harrison, S., 2009. Environmental systems: philosophy and applications in
physical geography. In: Clifford, N.J., Holloway, S.L., Rice, S.P., Valentine, G.
(Eds.), Key Concepts in Geography, Second ed. SAGE Publications, London, pp.
251264.
Holzner, L., Weaver, G.D., 1965. Geographic evaluation of climatic and climatogenetic geomorphology. Annals of the Association of American Geographers
55(4), 592602.
Huggett, R.J., 2007. Fundamentals of Geomorphology, Third ed. Routledge, Oxon,
UK, 516 pp.
Jenny, H., 1941. Factors of Soil Formation. McGraw Hill, New York, 281 pp.
Kabata-Pendias, A., Ryka, W., 1989. Precambrian Weathered Crusts in North-Eastern
Poland. Weathering: Its Products and Deposits. Products Deposits
Geotechnics. Theophrastus Publications, Athens, vol. II, pp. 241255.
Kalnay, E., Kanamitsu, M., Kistler, R., et al., 1996. The NCEP/NCAR reanalysis 40year project. Bulletin of the American Meteorological Society 77, 437471.
Kalvova, J., Halenka, T., Bezpalcova, K., Nemesova, I., 2003. Koppen climate types
in observed and simulated climates. Studia Geophysica et Geodaetica 47(1),
185202.
Kelly, E.F., Chadwick, O.A., Hilinski, T.E., 1998. The effect of plants on mineral
weathering. Biogeochemistry 42, 2153.
Kottek, M, Grieser, J., Beck, C., Rudolf, B., Rubel, F., 2006. World map of the
Koppen-Geiger climate classification updated. Meteorologische Zeitschrift 15(3),
259263.
Krumbein, W.E., Dyer, B.D., 1985. This planet is alive: weathering and biology, a
multi-faceted problem. In: Drever, J.I. (Ed.), Chemistry of Weathering. D. Reidel
Publishing Company, Dordrecht, 324.
Langbein, W.B., Dawdy, D.R., 1964. Occurrence of dissolved solids in surface
waters in the United States. Geological Survey Research 1964, United States
Geological Survey Professional Paper 501-D, pp. D115D117.
Li, S., Lu, X.X., He, M., Zhou, Y., Bei, R., Li, L., Ziegler, A.D., 2011. Major element
chemistry in the upper Yangtze River: a case study of the Longchuanjiang River.
Lidmar-Bergstrom, K., 1995. Relief and saprolites through time on the Baltic Shield.
Linton, D.L., 1955. The problem of tors. Geographical Journal 121(4), 270287.
Luzzadder-Beach, S., Beach, T., 2008. Arising from the wetlands: mechanisms and
chronology of landscape aggradation in the Northern Coastal Plain of Belize.
Annals of the Association of American Geographers 99(1), 126.
Marsh, Ben, 1999. Periglacial Landscapes of Central Pennsylvania. Sixty-Second
Annual Reunion, Northeast Friends of the Pleistocene, May 2223, 1999. Field
Trip Guidebook. Bucknell University, Lewisburg, Pennsylvania, 69 pp.
Martini, J.E.J., 1979. Karst in Black Reef Quartzite near Kaapsehoop, Eastern
Transvaal. Annals of the South African Geological Survey 13, 115128.
Maslyn, R.M., 1977. Fossil tower karst near Molas Lake, Colorado. The Mountain
Geologist, Rocky Mountain Association of Geologists 14(1), 1745.
May, J.H., Warne, A.G., 1999. Hydrogeologic and geochemical factors required
for the development of Carolina Bays along the Atlantic and Gulf of
Mexico, coastal plain, USA. Environmental and Engineering Geoscience 5(3),
261270.
Migon, P., 2006. Granite Landscapes of the World. Oxford University Press, Oxford.
Migon, P., 2009. Are granite landscapes distinctive of the humid tropics?
Migon, P., 2013. Weathering mantles and long term landform evolution. In:
Shroder, J. (Editor in Chief), Pope, G.A. (Ed.), Treatise on Geomorphology,
Academic Press, San Diego, CA, vol. 4, Weathering and Soils Geomorphology,
pp. 127144.
Molina Ballesteros, E., Canto Martin, M., 2002. Study of weathering processes
developed on old piedmont surfaces in Western Spain: new contributions to the
interpretation of the Rana profiles. Geomorphology 42, 279292.
Monroe, W.H., 1976. The karst landforms of Puerto Rico. USGS Professional Paper
899, 69 p.
Mueller, J.E., Twidale, C.R., 1988. Geomorphic development of City of Rocks, Grant
County, New Mexico. New Mexico Geology 10(4), 7379.
Munroe, J.S., 2006. Investigating the spatial distribution of summit flats in the
Uinta Mountains of northeastern Utah, USA. Geomorphology 75, 437449.
Murphy, R.E., 1968. Annals map supplement number 9, landforms of the world.
Annals of the Association of American Geographers 58(1), 198200.
Nahon, D., Merino, E., 1997. Pseudomorphic replacement in tropical weathering:
evidence, geochemical consequences, and kinetic-rheological origin. American
Journal of Science 297, 393417. http://dx.doi.org/10.2475/ajs.297.4.393.
Nash, D.J., Ullyott, J.S., 2007. Silcrete. In: Nash, D.J., McLaren, S.J. (Eds.),
Geochemical Sediments and Landscapes. Blackwell Publishing, Ltd., Malden,
Massachusetts, pp. 95143.
Netoff, D.I., Chan, M.A., 2009. Aeolian activity at a giant sandstone weathering pit
in arid south-central Utah. Earth Surface Processes and Landforms 34, 99108.
http://dx.doi.org/10.1002/esp.1697.
Netoff, D.I., Cooper, B.J. , Shroba, R.R., 1995. Giant sandstone weathering pits near
Cookie Jar Butte, southeastern Utah. In: van Riper III, C. (Ed.), Proceedings of
the 2nd Biennial Conference on Research Colorado Plateau National Parks,
2528 October, 1993. National Park Service Transactions and Proceedings
Series, NPS/NRNAU/NRTP-95- 11, pp. 2553.
Ollier, C.D., 1969. Weathering. American Elsevier, New York, 304 pp.
Ollier, C.D., 1984. Weathering, 2nd edn. Longman, London, 270 pp.
Ollier, C.D., 1988. Deep weathering, groundwater, and climate. Geografiska Annaler
70A(4), 285290.
Ollier, C.D., Pain, C., 1996. Regolith, Soils, and Landforms. Wiley, Chichester, UK,
316 pp.
Palmer, J., Neilson, R.A., 1962. The origin of granite tors on Dartmoor. Proceedings
of the Yorkshire Geological Society 33, 315340.
Paton, T.R., Williams, M.A.J., 1971. The concept of laterite. Annals of the
Association of American Geographers 62(1), 4256.
Pavich, M.J., 1986. Process and rates of saprolite production and erosion on a
foliated granitic rock of the Virginia Piedmont. In: Colman, S.M., Dethier, D.P.
(Eds.), Rates of Chemical Weathering of Rocks and Minerals. Academic Press,
New York, pp, pp. 552590.
Pavich, M.J., Obermeier, S.F., 1985. Saprolite formation beneath Coastal Plain
sediments near Washington, D.C. Geological Society of America Bulletin 96,
886900.
Pedro, G., 1968. Distribution des principauz types dalteration chimique a la surface
du globe. Revue de Geographie Physique et de Geologie Dynamique 10(5),
457470.
Pedro, G., 1983. Structuring of some basic pedological processes. Geoderma 31(4),
289299.
Peltier, L.C., 1950. The geographical cycle in periglacial regions as it is related to
climatic geomorphology. Annals of the Association of American Geographers
40(3), 214236.
Petersen, J. F., 1988. Influence of porphyritic texture on granite weathering features;
Enchanted Rock, Texas. In Aplin, G. (compiler), 26th congress of the
International Geographical Union; abstracts, Technical Program Abstracts
International Geographical Congress, 26, 448.
Pillans, B, 2008. Regolith Dating Methods: A Guide to Numerical Dating Techniques
in Australia. Cooperative Research Centre for Landscape Environments and
Mineral Exploration (CRC LEME), Bently, Australia, 52 pp.
understanding geographical variations in weathering. Annals of the Association
of American Geographers 86(1), 3864.
Pope, G.A., and Pobanz, S., 2011. Weathering geomorphologys black box:
integrating region-scale factors into weathering models with GIS. Abstracts,
Association of American Geographers Annual Meeting, April 1216, 2011,
Seattle, Washington, available at www.aag.org.
Potter, C.S., Anderson, J.T., Field, C.B., Matson, P.A., Vitousek, P.M., Mooney, H.A.,
Klooster, S.A., 1993. Terrestrial ecosystem production: a process model based
on global satellite and surface data. Global Biogeochemical Cycles 7(4),
811841.
Pye, K., 1983. Formation of quartz silt during humid tropical weathering of dune
sands. Sedimentary Geology 34, 267282.
Reiche, P., 1950. A Survey of Weathering Processes and Products. University of
New Mexico Publications in Geology. The University of New Mexico Press,
Albuquerque, vol. 3, 91 pp.
Reith, F., Durr, M., Welch, S., Rogers, S.L., 2008. Geomicrobiology of the regolith.
In: Scott, K.M., Pain, C.F. (Eds.), Regolith cience. Springer Science Business
Media B.V., Dordrecht, and Collingwood, Australia: CSIRO Publishing, pp.
127157.
Retallack, Gregory J., 2007. Cenozoic paleoclimate on land in North America. The
Schaetzl, R., Anderson, S., 2005. Soils: Genesis and Geomorphology. Cambridge
Scull, P., 2010. A top-down approach to the state-factor paradigm for use in
macro-scale soil analysis. Annals of the Association of American Geographers
100(1), 112.
Selby, M.J., 1985. Earths Changing Surface. Clarendon Press, Oxford.
195
Selveraj, K., Chen, C.-T.A., 2006. Moderate chemical weathering of subtropical

Taiwan: constraints from solid-phase geochemistry of sediments and
sedimentary rocks. The Journal of Geology 114, 101116.
Shroder, J.F., 1973. Erosion remnants near the Rift edge in Malawi. Zeitschrift fur
Geomorphologie Supplementbande 18, 121143.
Solbakk, T., Bering, D., Bronner, M., et al., 2010. Tropical weathering in Norway,
the NGU-NPD Twin Project. 29th Nordic Geological Winter Meeting, Oslo,
January 11-13, 2010, poster, available online (http://www.npd.no/en/
Publications/Presentations/TWIN-Tropical-Weathering-in-Norway/)
Stoddart, D.R., 1969. Climatic geomorphology: review and reassessment. Progress
in Geography 1, 159222.
Strakhov, N.M., 1967. Principles of Lithogenesis. Oliver and Boyd, Edinburgh.
Schulz, M., White, A.F., 1998. Chemical weathering in a tropical watershed,
Luquillo Mountains, Puerto Rico III: quartz dissolution rates. Geochimica et
Cosmochimica Acta 63(3/4), 337350 and Sons, 375 pp.
Stallard, R.F., 1992. Tectonic processes, continental freeboard, and the ratecontrolling step for continental denudation. In: Butcher, S.S., Charlson, R.J.,
Orians, G.H., Wolfe, G.V. (Eds.), Global Biogeochemical Cycles. Academic Press,
London, pp. 93121.
Taylor, G., Eggleton, R.A., Holzhauer, C.C., Maconachie, L.A., Gordon, M., Brown,
M.C., McQueen, K.G., 1992. Cool climate lateritic and bauxitic weathering. The
Chichester, 375 pp.
Thomas, M.F., 1974. Tropical Geomorphology. MacMillan, London.
Thomas, M.F., 1978. The study of inselbergs. Zeitschrift fur Geomorphologie,
Supplementband 31, 243264.
Thomas, M.F., 1994. Geomorphology of the Tropics. Wiley, Chichester, 460pp.
Tricart, J., Cailleaux, A., 1972. Introduction to Climatic Geomorphology. Longman,
London, pp.
Trudgill, S.T., 1976. Rock weathering and climate: quantitative and experimental
aspects. In: Derbyshire, E. (Ed.), Geomorphology and Climate. Wiley, New York,
pp.
and innovation. Geomorphology 25, 229253.
Twidale, C.R., 1973. On the origin of sheet jointing. Rock Mechanics and Rock
Engineering 5(3), 163187.
Twidale, C.R., 1984. So-called pseudokarst in granite. Boletin de la Sociedad
Venezuela de Espeleologia 21, 312.
involving etching: origin, development, and implications of an idea. EarthScience Reviews 57(12), 3774.
Twidale, C.R., Lageat, Y., 1994. Climatic geomorphology: a critique. Progress in
Physical Geography 18(3), 319334.
Ugolini, F.C., Sletten, R.S., 1991. The role of proton donors in pedogenesis as
revealed by soil solution studies. Soil Science 151, 5975.
Viers, J., Dupre, B., Polve, M., Schott, J., Dandurand, J.-L., Braun, J.-J., 1997.
Chemical weathering in the drainage basin of a tropical watershed (Nsimi-Zoetele
site, Cameroon): comparison between organic-poor and organic-rich waters.
Chemical Geology 140, 181206.
Vitek, J., 1983. Classification of pseudokarst forms in Czechoslovakia. International
Journal of Speleology 13, 118.
Wanty, R.B., Folger, P.F., Frishman, D., Briggs, P.H., Day, W.C., and Poeter, E.,
1992, Weathering of Pikes Peak Granite: field, experimental, and modeling
observations. In Kharaka, Y.K., and Maest, A.S., Proceedings of the 7th
International Symposium on Water-Rock Interactions. A.A. Balkema, Rotterdam,
pp. 599602.
Wasklewicz, T.A., 1994. Importance of environment on the order of mineral
weathering in olivine basalts, Hawaii. Earth Surface Processes and Landforms
19(8), 715734.
Wehmiller, J.F., 1982. A review of amino acid racemization studies in
Quaternary mollusks: srategraphic and chronology applications in coastal
and interglacial sites, Pacific and Atlantic coasts, United States, United Kingdom,
Baffin Island, and tropical islands. Quaternary Science Reviews 1,
83120.
White, A.F., Blum, A., Schulz, M.S., et al., 1998. Chemical weathering in a
tropical watershed, Luquillo Mountains, Puerto Rico: I. Long-term versus
short-term weathering fluxes. Geochimica et Cosmochimica Acta 62(2),
209226.
Widdowson, M, 2007. Laterite and ferricrete. In: Nash, D.J., McLaren, S.J. (Eds.),
Geochemical Sediments and Landscapes. Malden, Massachusetts. Blackwell
Publishing, Ltd., Malden, Massachusetts, pp. 4694.
196
Wolfe, J.A., 1969. Paleogene flora from the Gulf of Alaska Region. Government
Printing Office, Washington, DC, 114 p. United States Geological Survey Open
File Report.
Wolfe, J.A., 1979. Temperature parameters of humid to mesic forests of Eastern
Asia and relation to forests of other regions of the Northern Hemisphere and
Australasia. United States Geological Survey Professional Paper 1106.
Government Printing Office, Washington, 42 pp.
Wray, R.A.L., 1997. A global review of solutional weathering forms on quartz
sandstones. Earth Science Reviews 42(3), 137160.
Wray, R.A.L., 2003. Quartzite dissolution: karst or pseudokarst? Speleogenesis and
Evolution of Karst Aquifers 2, 19.
Yapp, C.J., 2008. 18O/16O and D/H in goethite from a North American oxisol of the
early Eocene climatic optimum. Geochimica et Cosmochimica Acta 72,
58385851. http://dx.doi.org/10:1016/j.gca.2008.09.003
Yatsu, E., 1988. The Nature of Weathering: An Introduction. SOZOSHA, Tokyo, 624 pp.
Young, A.R.M., 1987. Salt as an agent in the development of cavernous weathering.
Geology 15(10), 962966. http://dx.doi.org/10.1130/0091-7613(1987)15o962:
SAAAIT42.0.CO;2.
Young, R.W., 1986. Tower karst in sandstone: bungle Bungle massif, northwestern
Australia. Zeitschrift fur Geomorphologie 30, 189202.
Biographical Sketch
Gregory Pope is an associate Professor in Earth and Environmental Studies at Montclair State University in New
geoarchaeology, working in China, Latin America, Western Europe, and the Western and Northeastern United
States. He is active in both the Geological Society of America and Association of American Geographers, and
served as chair of the Geomorphology Specialty Group and a Regional Councilor for the AAG.
4.12 Weathering in Arid Regions

PA Warke, Queens University Belfast, Belfast, UK
4.12.1
4.12.2
4.12.2.1
4.12.2.1.1
4.12.2.1.2
4.12.2.2
4.12.2.2.1
4.12.2.2.2
4.12.2.2.3
4.12.3
4.12.3.1
4.12.3.2
4.12.3.2.1
4.12.3.2.2
4.12.3.2.3
4.12.3.2.4
4.12.3.3
4.12.3.4
4.12.3.4.1
4.12.3.4.2
4.12.3.5
4.12.3.5.1
4.12.3.5.2
4.12.4
4.12.4.1
4.12.4.1.1
4.12.4.1.2
4.12.4.1.3
4.12.4.1.4
4.12.4.2
4.12.4.2.1
4.12.4.2.2
4.12.4.2.3
4.12.4.2.4
4.12.4.3
4.12.5
4.12.6
References
Introduction
Climate and Weathering Presumed Connections and Observed Disparities
Temperature
Air temperature
Rock and sediment temperature
Moisture Availability
Rainfall
Dewfall and fog
Groundwater
Nature and Complexity of Weathering Processes
Insolation Weathering (Thermoclastis)
Salt Weathering
Crystallization
Hydration/dehydration phase change
Thermal expansion/contraction
Chemical dissolution effects
Frost (FreezeThaw) Weathering
Chemical Weathering
Mobilization and removal of elements
Mobilization and precipitation of elements
Biochemical effects
Biophysical effects
The Desert Weathering System
System Components
Materials
Processes
Form/morphology
Environment
Desert Weathering A Nonlinear Dynamic System?
Feedback mechanisms
Magnitude and frequency
Form convergence (equifinality)
Sensitivity and system components
Scale Issues (Spatial and Temporal)
Inheritance and the Concept of Palimpsest
Conclusion
Glossary
Albedo The proportion of incident light or solar radiation
reflected from a surface. Light-colored rock or sediment surfaces
are described as having a high albedo and hence reflect back
more incident light in comparison with dark surfaces.
198
200
200
201
201
203
203
203
204
204
205
207
209
210
212
212
212
213
214
214
215
216
217
218
218
218
218
218
218
218
219
219
219
221
221
222
222
224
Alveolae (Honeycomb Weathering Features) Small

hollows that are generally less than 50 cm in
diameter and are commonly found in association
with larger cavernous weathering features called
tafoni.
Warke, P.A., 2013. Weathering in arid regions. In: Shroder, J. (Editor in

Chief), Pope, G.A. (Ed.), Treatise on Geomorphology. Academic Press, San
http://dx.doi.org/10.1016/B978-0-12-374739-6.00060-9
197
198
Weathering in Arid Regions
Bornhardts (Inselbergs) Steep-sided, dome-like hills of

bare rock, usually found in areas of granite or gneiss geology.
Boundary Layer Conditions Temperature and humidity
conditions in the thin layer of air that passes over and is in
immediate contact with rock surfaces.
Case hardening Formation of a resilient surface layer on
rock outcrops and boulders by infilling of pore spaces by
natural cements.
Cryptoendolithic Algae and microbial organisms that
colonize preexisting fractures or spaces within rock.
Dolines Circular or near-circular depressions that form by
the dissolution and/or the collapse of limestone.
Holistic (Holism) Refers to all the parts of a system and
recognition that the system is more than the sum of its
individual parts (see Synergy).
Intergranular Describes the boundaries between
individual mineral grains.
Karst Collective term for the assemblage of distinct
landforms associated with regions of limestone geology in
which the dominant weathering process is dissolution.
Microenvironmental Temperature and humidity
conditions at the rock/air interface.
Microfractures These are submillimeter-scale breakages in
rock or individual mineral grains.
Nonlinear Dynamical Systems Systems in which it is
difficult to predict behavior because of the complex nature
of variables involved and the interactions between them.
Palimpsest A term used by geomorphologists to describe

a landscape (or landform) in which traces of past events or
environmental conditions remain visible.
pH Scale A measure of acidity or alkalinity, with values
greater than pH 7 being described as alkaline and those less
than pH 7 being described as acidic.
Playa A common feature in arid regions manifest as an
enclosed depression periodically flooded after extreme
rainfall events and forming the endpoint of a desert
drainage system.
Rock Varnish (Desert Varnish) A very thin rock coating
composed of clay minerals cemented by iron and
manganese oxides and commonly found in, but not
exclusive to, hot deserts.
Subaerial Describes conditions at the ground surface.
Synergy Describes the interaction between two or more
factors or variables that produces a result that is not
achievable by variables acting in isolation.
Tafoni Large cavernous weathering features generally
more than 50 cm in diameter and typically occurring on
vertical rock faces.
Thermal Conductivity The ability of a substance to
conduct heat energy from its surface to its core. Materials
with a high thermal conductivity transfer heat energy
more quickly than materials with a low thermal
conductivity.
Abstract
Many misconceptions exist regarding weathering in arid regions. Chief among these are assumptions that physical processes
dominate and are not very effective because of a perceived lack of moisture. This chapter explores the factors that combine
to make weathering in arid regions spatially and temporally complex, reflecting the range of surface microenvironmental
conditions. Because of desert landscape complexity, attempts at interpreting weathered features must take into account the
long-term history of rock outcrops and debris that mantle them, as most desert landscapes contain legacies of weathering
forms and products, which were developed when moisture was more readily available in the past.
4.12.1
Introduction
Weathering in hot arid regions, like weathering in any other

climatic zone, reflects the fact that rock is no longer in a state
of balance or equilibrium with prevailing conditions of temperature, pressure, and moisture availability compared with
those under which it was formed. Rock typically forms under
conditions of high temperature and/or pressure and once it
approaches the subaerial or near subaerial conditions of the
Earths surface, it starts to deteriorate both structurally and
mineralogically:
Conditions of temperature and moisture availability initiate chemical and physical reactions that alter previously
stable minerals, creating new phases that are more stable in
the subaerial environment (Banfield et al., 1999).
The decrease or complete removal of overburden pressures
can result in the formation of differential expansive stresses
in rock that contribute to the development of a host of joints
and fractures that occur at all scales from submicron to

kilometer scale (Figure 1). In terms of rock weathering, all of
these joints and fractures are critical because they increase
the surface area upon which weathering processes can act
and provide routes into the fabric of rock for weathering
agents such as moisture and salt (Whalley and Warke, 2005).
Structural weaknesses are particularly important in arid
zone weathering because of the general absence of a buffering
layer of soil and vegetation that tends to shield rock from
direct interaction with components of the subaerial environment. Consequently, strong structural controls are commonly
evident in some arid zone weathering forms, creating many
distinctive weathering features such as bornhardts and tafoni,
although it is suggested that some of these features may owe
their origins to former more humid climatic conditions (see
Chapter 4.8).
Historically, within arid zone weathering studies, there has
been a tendency to concentrate investigation on distinctive
199
(a)
Figure 1 Large-scale sheeting in sandstone, Arches National Park,

Utah. Such features are typically attributed to pressure release
through erosion and removal of overburden. These features often
have smaller scaling features superimposed upon them and both are
significant because they increase the potential reactive rock surface
area in direct contact with subaerial weathering processes.
and often visually prominent weathering features, with a resultant emphasis in the literature on the unusual at the expense of the mundane (Figures 2(a) and 2(b)). However, the
visually prominent and exceptional weathering forms are not
truly representative of surface conditions as a whole in the
majority of deserts, which typically tend to be dominated by
extensive debris mantles and aeolian deposits (Thomas, 1997)
(Figures 3(a)3(d)). Greater understanding of the origins
and subsequent weathering of this material can only lead
to a much better appreciation of the complexity and potential
dynamism of many desert surfaces and recognition of the
role of present-day processes in long-term landscape
development.
Several misconceptions exist regarding rock weathering in
arid regions. Chief among these are the assumptions that
weathering is dominated by physical processes and that these
processes lack dynamism especially when assessed within the
short-term context of the typical three-year research project
and when compared with other more humid regions. Consequently, this has resulted in research into rock weathering in
arid regions assuming a somewhat subservient role to the
more obviously dynamic land-surface modification associated
with aeolian and fluvial processes.
Desert weathering has been previously characterized by its
selectivity and superficiality (Cooke and Warren, 1973),
characteristics that reflect the shallowness of moisture penetration and the strong geological controls made parti-
(b)
Figure 2 (a) Sculpted sandstone pedestal in southern Jordan. Note

the alveolar weathering features and preferential weathering of
bedding planes that are inherent lines of structural weakness.
(b) Sandstone arch, Arches National Park, Utah.
cularly evident in deserts because of the general lack of

developed soil and vegetation cover. However, this characterization belies the complexity of a situation whereby weathering processes operate cumulatively and sequentially with
variable spatial and temporal intensity over time to create a
variety of weathering features and to provide the principal
means of sediment release for subsequent erosion, transport,
and deposition by aeolian, fluvial, and gravity-related slope
processes. Together, these processes contribute to landscapes
underpinned by complex weathering systems, the effectiveness
of which cannot be assessed within the confines of a shortterm research project or even within the human lifespan.
Despite the difficulties associated with weathering research
in arid regions, difficulties that reflect the inaccessibility and
extreme conditions experienced in these regions, the latter half
of the twentieth century has seen the identification of several
major conceptual issues that have become increasingly important in guiding the focus of current research and informing
better understanding of the nature and rate of rock weathering
in arid regions. These issues provide a framework for the rest
of this chapter and include the following:
The role of climate in determining the nature and intensity

of weathering in which there are many presumed connections that are challenged by observed disparities.
200
(a)
(b)
(c)
(d)
Figure 3 (a) Sand plain, Wadi Rum, southern Jordan; (b) debris mantled desert plain, Amargosa Valley, Nevada; (c) desert pavement, Panamint
Valley, California; (d) climbing dune and debris mantled surface, southern Jordan.
The nature and complexity of individual weathering

processes and the potential for synergistic reactions
between them.
The desert weathering system itself, which exhibits properties typical of a nonlinear dynamical system in which
complexity and unpredictability are key characteristics reflecting the effect of complex feedback mechanisms.
Desert landscapes typically exhibit multi-cycle characteristics (Twidale, 2007) with traces of past weathering under
different climatic conditions that persist within the landscape and continue to influence present-day processes. This
can lead to the creation of a landscape palimpsest in which
a complex assemblage of inherited and contemporary
weathering forms bear witness to the role of water despite
prevailing conditions of aridity.
4.12.2
Climate and Weathering Presumed

Connections and Observed Disparities
Traditional climate-based classifications of morpho-climatic

weathering zones such as the work by Peltier (1950) were used
to indicate dominant weathering processes and the depth of
weathering profiles (e.g., Strakhov, 1967). These classification
systems, together with the experiences in hot deserts of early
researchers, supported the interpretation of rock breakdown
in these regions as being dominated by physical weathering
processes in landscapes characterized by high temperatures
and an obvious lack of moisture.
However, it is now widely accepted that such overly simplistic classifications, although possessing some validity at the
regional scale, belie the complexity of arid zone weathering on
the ground (Pope et al., 1995). On desert surfaces, spatial and
temporal variability in temperature conditions and moisture
supply gives rise to a weathering system in which physical,
chemical, and biological processes all contribute to rock
weathering and breakdown to varying degrees recognition of
this complexity is central to understanding the dynamic nature
of arid zone weathering.
More extensive field travel and technological advances that
allow longer-term remote data collection have highlighted
the importance of microenvironmental controls within the
atmospheric boundary layer, with significant differences
recorded between large-scale regional conditions and smallscale local conditions at the rock/air interface (Pope et al.,
1995; Warke, 2000).
The two key environmental components that drive the arid
zone weathering system are temperature and moisture availability. An understanding of their spatial and temporal variability and the factors that influence these is central to gaining
an understanding of the complexity of the weathering system
in arid regions and the inaccuracy of oversimplifying this
system by ascribing dominance to one group of weathering
processes.
4.12.2.1
Temperature
High daytime temperatures typically define hot, arid environments, and anecdotes abound whereby rock surfaces are
reported as being hot enough to fry eggs. Such anecdotes

demonstrate the popular fixation with extreme conditions and
the many factors that combine to create these, with air temperature being only one component of the thermal characteristics of arid regions. The other comprises the temperature
characteristics of desert surfaces that are often much more
extreme. Differences between the two arise from emission of
long-wave terrestrial radiation and the influence of thermal
properties of the materials (rock and sediment) that comprise
desert surfaces and that can significantly increase rock surface
temperatures over and well above that of local air temperature.
4.12.2.1.1
Air temperature
Historically, the emphasis was on the identification of extremes of air temperature usually recorded in the middle of
the day at the height of summer. Such data were commonly
used to characterize desert conditions and to justify the use of
extreme experimental conditions in laboratory weathering
simulations. There are many examples of these data cited in
the literature and used to illustrate not just the severity of rock
surface conditions but also the wide range of temperatures
encountered, thereby reinforcing the view that while high rock
surface temperatures frequently occur, they do not present a
complete picture of typical conditions (Goudie, 1997; Smith,
2009). Consequently, in recent decades, the emphasis has
shifted toward data collection that reflects the annual, seasonal, diurnal (night and day), and shorter-term air temperature fluctuations typical of most arid environments (e.g.,
Warke and Smith, 1994; Warke, 2000; Viles and Goudie,
2003), with specific attention now being paid to the effects on
rock weathering of the rapid temperature and humidity fluctuations associated with the transition of day to night and vice
versa when conditions of microenvironmental change are
more likely to facilitate weathering activity.
Air temperature in deserts is both temporally and spatially
variable. In terms of temporal variability, desert temperature
regimes are best described as comprising a hierarchy of conditions whereby seasonal fluctuations provide the foundation
on which diurnal and shorter-term fluctuations, operating
over hours and minutes associated with the passage of cloud,
gusts of wind, etc., are imposed (Figure 4). The scale of
seasonal temperature gradients varies in different deserts, reflecting the effects of altitude, latitude, and continentality but,
in general, they follow a common pattern of marked temperature differences between the extremes of day-time heating
and night-time cooling, whereby radiative heat loss under
clear night skies can be rapid and extreme.
In terms of spatial variability, significant differences normally exist between large-scale regional conditions and smallscale low-level boundary layer conditions at the rock/air
interface. This reflects different controlling factors, with regional temperatures reflecting latitude, altitude, and large-scale
meteorological conditions. Small-scale temperatures in the
near-surface layer of air are typically higher than air several
meters above the ground surface because of the lack of vertical
and reduced horizontal mixing associated with the frictional
drag created when air flows over the ground surface. This
creates marked horizontal and vertical temperature differences
during the daytime in the layer of air overlying the ground
surface where air is heated by the passage of solar radiation,
201
energy reflected back from the surface and outgoing long-wave

terrestrial radiation. Hence, conditions immediately above a
rock surface may differ significantly from measurements made
more than 1 m above the ground surface. For example, air
temperature recorded 5 cm above ground level in Death Valley
at 14.55 on 3 August 1992 was 65.8 1C but some 20 1C cooler
at 45 1C just 1 m above this (Warke, 2000).
The morphology of the rock/air interface also has an impact on air temperatures and this is best demonstrated by
the temperature conditions within weathered hollows such
as alveolae and tafoni, where, because of the effects of shade,
air temperatures are typically less extreme than on rock
surfaces exposed to direct insolation (Evans, 1970; Mustoe,
1982, 1983; Smith and McAlister, 1986; Turkington, 1998;
Turkington and Paradise, 2005). Together, this complex suite
of conditions gives rise to a dynamic and potentially extreme
temperature backdrop against which a range of weathering
processes can operate (see Chapter 4.7).
4.12.2.1.2
Rock and sediment temperature
Temperatures experienced by rock surfaces are the other part

of the desert temperature equation and these, like air temperatures, are both spatially and temporally variable. Spatial
variability arises from differences in the thermal properties of
materials across a desert surface. For example, the albedo and
thermal conductivity properties of rock will determine the
intensity of surface heating and the efficiency with which this
heat energy is conducted away from the rock surface into
subsurface layers. In deserts, the albedo (color) of rock surfaces can be changed by the development of rock varnish,
a commonly occurring iron- and manganese-enriched rock
surface alteration/depositional feature (Figures 5(a) and
5(b)). The presence of a well-developed rock varnish can
lower albedo and lead to the development of surface temperatures much higher than would normally be experienced
by an unvarnished surface (see Chapter 4.5).
Further complexity is introduced with heterogeneous rock
types such as granite, where different minerals have different
thermal properties, leading to differential intergranular heating. Consequently, a desert surface comprising a mix of intact
rock outcrops, boulders, gravel, and a range of finer sediment
types can lead to the development of a complex patchwork of
surface temperature responses and depth of heat transfers.
Temperature gradients established in fine playa sediments are
shown in Figure 6, where over a 24-h recording period, playa
sediments show a lag between changes in air temperature and
sediment response and marked differences between sediment
temperatures some 100 mm below the surface and overlying
air temperature with up to 10 1C of difference between the
two. This is particularly notable at night, when there is a reversal in temperature conditions, with subsurface sediment
being warmer than surface and air temperature conditions.
Temporal variability of temperatures in desert surface materials reflects the hierarchy of air temperature fluctuations but
tends to lag behind these primarily because rock and sediment
take time to respond to changes in solar radiation through
either the loss or the absorption of energy. Figure 7 clearly
shows the influence of rock thermal properties on temperature
response, whereby a combination of low albedo and low
thermal conductivity in a dark gray basalt rock produces
Average monthly temperature (C)
202
40
30
20
10
Average monthly temperature (C)
0
1995 1996 1997 1998 1999 2000 2001 2002 2003 2004 2005 2006
(a)
Recording period (years)
40
30
20
10
0
Jan Feb Mar Apr May Jun Jul Aug Sep Oct Nov Dec
Recording period (months)
(b)
70
Temperature (C)
60
50
40
30
0
10:00
13:00
16:00
19:00 22:00 01:00 04:00

Recording period (24 h)
13:30
14:00
Recording period (2 h)
(c)
07:00
10:00
Temperature (C)
70
65
60
55
0
13:00
(d)
14:30
15:00
Figure 4 Air temperature variability at different temporal scales, Death Valley, California. Graphs (a) and (b) are based on average monthly
temperature data, while graphs (c) and (d) reflect temperature data recorded at intervals of one minute, some 5 mm above the ground surface.
203
62
58
Temperature (C)
54
(a)
50
46
42
38
34
Site sandy sediment on playa margin
Time of year August
30
0
11:00
15:00
19:00
23:00
03:00
07:00
11:00
Time of day (hundred hours)

Air temperature recorded approximately 30mm
above the ground surface
Ground surface temperature
Ground temperature 50 mm below the surface
Ground temperature 100 mm below the surface
(b)
Figure 5 (a) Varnished rock debris that forms a well-developed

desert pavement in Panamint Valley; (b) the underside of clast X
(shown in (a)) is much lighter in color than its surface, which has
been darkened by a well-developed manganese-rich varnish.
surface temperatures that exceed air temperature while a

comparatively higher albedo and coefficient of thermal conductivity results in rock surface temperatures in the limestone
that are well below air temperature even though both lithologies are exposed to the same air temperature conditions
(Warke and Smith, 1998).
4.12.2.2
Moisture Availability
Deserts are defined on the basis of their aridity and this,

combined with high temperature conditions, has traditionally
underpinned the perceived dominance of physical weathering
processes. However, moisture availability within deserts is now
recognized as being much more complex, with considerable
spatial and temporal variability.
4.12.2.2.1
Rainfall
Given the spatial and temporal variability of rainfall events,

some parts of a desert catchment may not receive any rainfall
in a given year while other parts receive the total annual
rainfall input in one storm-event. The way rainfall is delivered
is also important especially with regard to its infiltration and
retention in the soil and groundwater store. Rainfall events can
be of variable intensity, with high-intensity delivery providing
significant energy for reshaping desert surfaces through sediment mobilization while low-intensity events are less energetic
but provide sufficient moisture to facilitate small-scale rock
weathering. There is evidence that high-intensity events
generally associated with arid regions are less frequent than
Figure 6 Summer temperature characteristics of Death Valley playa

sediments and those of the overlying air recorded at intervals of one
minute over a 24-h period.
assumed, with moisture input arriving more often in the form

of low-intensity rainfall (Peel, 1974) rainfall that is more
likely to be absorbed by rock and fine sediment and therefore
of more significance to weathering activity. Pope et al. (1995)
contend that spatial variability in rock weathering reflects
small-scale processes, which in turn are controlled by moisture
availability at the microscopic scale.
4.12.2.2.2
Dewfall and fog
Unlike more humid environments, rainfall may not necessarily account for the majority of moisture input, with occult
forms of precipitation such as dewfall and fog often contributing significantly to the total moisture budget in some desert
regions (Evanari et al., 1982; Zangvil, 1996; Ebner et al., 2011).
For example in the Negev Desert (Israel), dewfall makes a
significant contribution to the average annual moisture input
and similarly, in the coastal Namib Desert (Namibia), the
input of moisture from fog often exceeds that derived from
rainfall events (Pietruszka and Seely, 1985), with the average
annual rainfall contributing 24.5 mm to the total precipitation
budget while input from fog accounted for 31.7 mm (Goudie,
1997). Because of the long dry periods between rainfall events
these occult forms of moisture along with moisture in
groundwater stores and ambient atmospheric humidity come
to assume great significance in facilitating the operation of
rock weathering processes (Goudie et al., 1997; Goudie and
Parker, 1998; Viles, 2005; Viles and Goudie, 2007). The role of
dew in stabilizing desert surface sediments by facilitating the
development of microbiotic crusts is widely recognized and is
another illustration of how very small amounts of moisture
204
72.65 C
73
Limestone
Basalt
65.80 C
69
71.15 C
Temperature (C)
65
61
57.65 C
65.80 C
57
57.50 C
53
49
Thermal conductivity 1.53 W m1 K1

Time of year August
45
Thermal conductivity 0.96 W m1 K1

Time of year August
0
11:00
12:00
13:00
14:00
15:00
11:00
12:00
13:00
14:00
15:00
Time of day (hundred hours)

Air temperature recorded approximately 50 mm above ground level
Rock surface temperature
Rock subsurface temperature (25 mm below surface)
Figure 7 Graph showing the influence of thermal conductivity and albedo on temperature response characteristics of basalt and limestone rock
samples exposed to summer heating in Death Valley, California. Reproduced from Figure 2 in Warke, P.A., 2000. Micro-environmental conditions
and rock weathering in hot arid regions. Zeitschrift fur Geomorphologie 120, 8395, with permission from Schweizerbart.
can play a role of local and regional geomorphological significance by restricting sediment mobilization (Kidron et al.,
2000; Kidron et al., 2002).
4.12.2.2.3
Groundwater
In addition to direct surface inputs from precipitation, groundwater acts as a major source of moisture even in the most hyperarid deserts. Groundwater is significant in terms of both largescale landscape development and small-scale rock weathering.
With regard to landscape development, groundwater can
be of major significance in valley development through the
effects of spring sapping, whereby slope failure and retreat is
associated with the outflow of groundwater at the base of a
slope, which weakens the rock or sediment through which it
flows, leading to the undermining and eventual collapse of
overlying strata (Nash, 1997). The proximity of the groundwater store can also be of major significance to playa development by determining their hydrological status, with
discharge playas developing where the water-table and hence
the groundwater store is close to the surface, while recharge
playas are associated with a water-table and associated capillary fringe that lies well below the ground surface.
In terms of arid zone weathering, groundwater provides an
important source of moisture that is of special significance to
salt weathering. High temperatures draw moisture to the surface through a process called evaporative pumping (Cooke
et al., 1982) and where rock is in contact with the capillary
fringe above the water table, moisture will also be drawn up
through the rock fabric by capillary rise, a process referred to

as the wick effect (Goudie, 1986). The significance of this
moisture is that it generally brings with it salts in solution,
which then crystallize out at the ground surface or above the
ground surface within or on rock debris and outcrops.
Groundwater thus provides both a mechanism for mobilizing
salt by moving it from subsurface to subaerial environments
and provides the necessary moisture to drive the mechanisms
of salt weathering in particular.
The role of moisture in shaping desert landscapes is an
anomalous one (Smith, 2009) because these regions have
historically been classified as arid, reflecting the lack of
moisture in comparison with other climatic zones. However,
many landforms in contemporary regions owe their existence
to the work of water and although moisture may have been
more abundant in the past when climates in some regions
were more humid, it is now accepted that moisture is more
readily available than previously thought, an availability reflected in widespread evidence of weathering activity.
4.12.3
Nature and Complexity of Weathering

Processes
Following exposure at the earths surface, rock in an arid environmental setting can be exposed to the cumulative and sequential actions of a variety of weathering processes. These
processes can be grouped into one of two categories those that
cause a weakening of rock fabric without the addition, chemical
alteration, or removal of material and those that result in the

addition of externally derived material and/or the chemical alteration and removal of constituents of the rock fabric. The
former describes insolation weathering while the latter include
salt, frost, chemical, and biological weathering processes.
Traditionally, there has been an emphasis on the comparative significance of physical and chemical weathering in
arid regions but desert weathering systems include chemical
and biological processes, with the resultant weathering forms
created being the products of complex interactions between
these processes. Consequently, although each is examined
individually in the following sections, it is important to remember that rarely, if ever, do they operate in isolation, more
often interacting synergistically where the action of one enhances the effectiveness of another.
4.12.3.1
Insolation Weathering (Thermoclastis)
Desert environments tend to be characterized by high daytime

temperatures and commonly large diurnal temperature ranges
with rapid loss of heat energy at night as a result of radiative
losses. Description of these regions by early explorers of landscapes that were uniformly dry and that baked during the
daytime under relentless cloud-free skies created a deeply rooted perception of regions in which the effects of great heating
were associated with rock breakdown. Because of this emphasis
on temperature, insolation weathering has traditionally been
cited as a possible mechanism of rock breakdown, although its
efficacy has never been conclusively proven or disproven and its
status, as an effective weathering process, is still hotly contested.
The rationale for insolation weathering is based on the
development of differential volumetric expansion of individual mineral grains and/or surface and near-surface rock
layers in response to repeated diurnal heating and cooling
(Whalley and Warke, 2005). It is suggested that the stresses
imposed on rock fabric by repeated expansion and contraction
over a long period of time are sufficient to gradually weaken
intergranular bonds and/or the bonds between surface and
near-surface layers of rock, thereby contributing to breakdown
through a fatigue effect (Goudie and Viles, 2000).
Thermal fatigue has long been recognized as a possible
explanation for rock breakdown, although one that probably
operates over several hundreds of years (Hockman and
Kessler, 1950). It has been suggested that the development of a
thermal fatigue effect may be facilitated in mineralogically
heterogeneous rocks (e.g., granite) by differences in the
thermo-elastic behavior of mineral grains across their
boundaries (Yatsu, 1988; Gomez-Heras et al., 2006) and, even
in mineralogically homogeneous rocks, differences in crystallographic alignment between adjacent grains may give rise to
considerable intergranular stress during heating and cooling
(Hall and Andre, 2003). For example, experimental evidence
shows that when a calcite crystal is heated, it expands parallel
to its C or long axis and actually contracts perpendicular to
the C axis, thereby making the crystal structure longer and
narrower (Winkler, 1973).
Failure of early dry thermal cycling laboratory-based experiments to produce rock breakdown raised the original
doubts about insolation weathering as being an effective
agent of rock breakdown (Blackwelder, 1925; Griggs, 1936).
205
However, advances in technology indicate that thermal cycling

in the absence of moisture can be sufficient to initiate surface
microfracturing at temperatures around 70 1C (Richter and
Simmons, 1974; Aires-Barros et al., 1975; Aires-Barros, 1977;
Yong and Wang, 1980; Goudie and Viles, 2000). Consequently,
although debate persists, insolation weathering cannot be
dismissed as a component of the desert weathering system.
It is now recognized that fluctuations in rock surface temperatures can be complex, reflecting the interaction between
aspect-related differences in heating, diurnal temperature
cycles, and shorter-term temperature change occurring over
several minutes and related to interruptions in the receipt of
insolation and changes in windspeed (Jenkins and Smith,
1990; Hall and Hall, 1991; Warke, 2000; Hall and Andre,
2001; McFadden et al., 2005; Eppes et al., 2010; Molaro and
McKay, 2010). Receipt of insolation may be interrupted by the
passage of cloud or when rock surfaces suddenly emerge from
or enter shade thrown by another landform (Smith, 1977;
Warke and Smith, 1994; Smith and Warke, 1997; Smith,
2009). Such interruptions may cause particularly extreme
temperature fluctuations under conditions where ambient air
temperatures are low during cooler winter months. The work
by Hall has shown that any interruption of incident insolation
under subzero temperature conditions can lead to rapid surface cooling with the potential to create conditions of thermal
shock (Hall and Hall, 1991). In addition, sudden decreases in
rock surface temperatures may reverse surface/subsurface
temperature gradients, creating complex stress patterns within
surface and near-surface layers of rock (Smith et al., 2011).
Along with stress created as a result of differential thermal
expansion between mineral grains and between surface and
subsurface rock fabric, the rate of temperature change may
also play an important role in thermally driven rock breakdown. For example, it is suggested that rates of temperature
change greater than 2 1C min1 may be sufficient to cause
micro-fracturing along grain boundaries (Richter and Simmons, 1974; Yatsu, 1988; Hall, 1999; Hall et al., 2008; McKay
et al., 2009).
Rates of temperature change are certainly significant as
shown by the destructive effects of bushfires, which result in
widespread splitting and spalling of rock (Bierman and
Gillespie, 1991; Dragovich, 1993; Dorn, 2003), although these
are extreme events that are not representative of typical daily
temperature regimes and nor can such events be described as
insolation weathering. However, with regard to rock weakening and breakdown, these rare but high-magnitude events
can accomplish a great deal of geomorphic work and leave
structurally and mineralogically degraded rock and debris
surfaces that may be more susceptible to subsequent weathering (Dorn, 2003). Their significance in truly arid areas is
debatable but they may have a greater role to play in the
semiarid margins, where sufficient scrub and grassy vegetation
occurs to fuel bushfires. Although such events may be rare, the
geomorphic work achieved by fire in environments where bare
rock and debris surfaces are for the most part unprotected by
soil cover is out of all proportion to their frequency.
The significance of the thermal properties of rock as a
whole and its mineral constituents has been briefly discussed
in a previous section but it is important to consider the relevance of these, particularly albedo and conductivity, in the
206
context of insolation weathering. When low levels of albedo

and thermal conductivity are combined, high surface temperatures and steep surface/subsurface temperature gradients
may be initiated even during relatively low ambient temperature conditions (McGreevy, 1985a; Warke and Smith,
1994, 1998). Consequently, in rock with these thermal characteristics, the thermal stresses incurred as a result of temperature fluctuations may be sufficiently severe to cause a
gradual weakening of intergranular bonds that may lead directly to surface breakdown or act to increase the susceptibility
of the rock surface to other exploitative weathering processes.
Split stones are commonly occurring and widely reported
features in arid regions (Yaalon, 1970 Israel; Mabbutt, 1977
Algeria; Goudie and Day, 1980 Death Valley; Ollier, 1984
Australia; Smith, 1988 Morocco; McFadden et al., 2005
southwestern USA and New Mexico) and present as either
irregular splitting, where the boulder or cobble has randomly
cracked into several pieces, or parallel splitting, where the
cleavage of the clast occurs along two or more subparallel
fractures roughly normal to the ground surface (Figures
8(a)8(d)). Much debate persists about the causes of stone
splitting but it is accepted that it is a two-stage process beginning with the initiation of splitting, followed by the subsequent widening of the fractures and separation of the split
fragments (Cooke and Warren, 1973). The absence of
chemical alteration on the split surfaces supports a model for

splitting based on purely physical weathering processes and
there are two schools of thought on the nature of the processes
involved, with one supporting salt weathering (Yaalon, 1970;
Goudie and Day, 1980; Goudie and Watson, 1984) and the
other supporting insolation weathering (Evans, 1971; Ollier,
1963, 1965, 1984; McFadden et al., 2005; Eppes et al., 2010).
Supporters of salt weathering attribute splitting to salt
crystal growth within microfractures and along preexisting
lines of weakness such as bedding or cleavage planes. An increased incidence of stone splitting with increased proximity
to the Death Valley saltpan was statistically demonstrated by
Goudie and Day (1980) but association should not be confused with causation, with uncertainty persisting as to whether
salt weathering mechanisms actually initiate splitting or
merely exploit preexisting structural weaknesses. Insolation
weathering has been invoked to explain both parallel and irregular stone splitting, with anecdotal evidence supporting the
role of thermal shock when a boulder surface warmed by insolation split shortly after a rainstorm, which caused rapid
surface cooling (Whitaker, 1974).
A feature common to split stones in many deserts is their
partial burial in underlying sediment where temperature
conditions differ from those above ground. Temperature gradients within the rock may be greatly enhanced if it is partially
(a)
(b)
(c)
(d)
Figure 8 (a), (b) Small cobble-sized clasts that exhibit a regular form of in-situ parallel/subparallel splitting that cuts across geological
structures such as bedding planes; (c) an irregularly split boulder with a well-developed rock varnish; (d) irregular splitting of the debris shown
has resulted in complete disintegration that will remain stored on the ground surface until further breakdown produces sediment fine enough to
be blown away or a flood event mobilizes the larger particles.
buried in cooler sediments, with expansive stresses initiated

through the heating of upper exposed rock surfaces opposed
by compressive stresses in the cooler confined portion of the
rock (Figure 9). This situation of conflicting thermal stress can
be further complicated by aspect-related differences in the
heating of rock, with Eppes et al. (2010) citing this as playing a
key role in the comminution of boulders and cobbles comprising desert pavement through the initiation and propagation of meridianally aligned fractures.
Separation of the rock fragments following the initiation of
splitting may reflect the properties of the finer sediment in
which the debris is partially buried (Figure 10). If the sediments contain salts, repeated episodes of crystallization and/
or hydration/dehydration may be sufficient to gradually push
the fragments apart (Yaalon, 1970), whereas the presence and
repeated wetting and drying of swelling clay minerals in
sediment has also been cited as a possible mechanism for
fragment separation (Smith, 1988).
Traditional emphasis on the pursuit of extreme temperature conditions and the search for associated dramatic rock
fracturing has occurred at the expense of investigation of the
more subtle effects of insolation weathering. Although work in
recent decades has sought to address this and much progress
has been made in understanding the potentially disruptive
effects of repeated and differential surface heating of rock
Warm
Cool
Sediment
Rock
Day
Warm
207
(see Smith et al., 2011), it is important to remember that insolation weathering probably operates over prolonged periods
of time in combination with other forms of weathering such
as the thermal expansion and contraction of salt crystals deposited within pore spaces, micro-fractures, and naturally occurring structural weaknesses.
Debate over the role of insolation weathering and rock
breakdown will undoubtedly continue because of the difficulty of replicating its effects definitively under controlled
laboratory conditions over the relatively short time-scales of
most simulation studies. However, weathering mechanisms
such as insolation do not operate at a set rate over a fixed
period of time but will fluctuate in effectiveness in response to
changing environmental conditions, the action of other weathering processes and changes in rock characteristics over time.
A much greater understanding of the complexity of rock
temperature conditions and the factors that control them
now exists and, therefore, the effects of repeated heating and
cooling cannot be discounted, especially given the potential
for synergistic interactions with other weathering processes.
4.12.3.2
Salt Weathering
Unlike insolation weathering, the effectiveness of salt weathering in rock breakdown is not contested. In fact, in many
instances, the very presence of salt is used to infer its causal
association with a range of weathering forms, but as will be
shown, salt weathering is extremely complex and often the
forensic trail of evidence linking it directly to rock breakdown
is not clear (Smith and Warke, 1997).
Deserts are characterized by an abundance of salt, which
reflects two key factors. First, the combination of low rainfall
amounts and high levels of evaporation means that there is a
predominant upward movement of subsurface moisture
bringing soluble salts to the surface, where they concentrate in
the sediment column over time. Second, salts are derived from
geological sources, as byproducts of rock weathering, atmospheric fallout (especially in coastal deserts), and from salinized
Cool
Cool
Warmer
Sediment
Night
Rock
Cool
Warmer
Figure 9 Diagram showing the theoretical effects of differential

heating of rock surface and subsurface layers within a rock and how
this notionally varies in response to diurnal patterns of heating and
cooling.
Figure 10 Parallel split rock (above key fob) with separation of

fragments. In this example, the salt-rich characteristics of the fine
sediment in which the clast is embedded may have contributed to
this, through repeated dissolution and crystallization of salts gradually
pushing the fragments apart. Note the smooth and polished
appearance of some of the surrounding clasts caused by long-term
abrasion by windblown fine sediment.
208
groundwater either locally derived or draining an external region. The degree of salt concentration reflects input so that salt
deposits described as saltpans result from the evaporation of
salt-rich ephemeral lake waters (e.g., Death Valley saltpan) that
leave behind typically thick (41 m) salt crusts (Figure 11) that
are further enhanced by the addition of salts drawn up from
saline groundwater after the lake has dried out. These extremely salt-rich environments are characteristically end-points
in a large drainage system that received input during the geological past from an active system of lakes and rivers, with the
contemporary salt deposits providing a record of past more
humid conditions (Goudie and Wells, 1995).
Salt concentration will generally tend to be lower and less
visually evident in topographic lows or playas that are occupied by ephemeral lakes only very rarely, but rely instead on
the upward migration of salt from groundwater and occasional inputs from flashflood events (Figure 12). In more
humid environments, perennial rivers and abundant rainfall
ensure that soluble salts are removed from the sediment column and prevented from accumulating within the landscape.
But in deserts characterized by endoreic drainage patterns,
salts accumulate within the landscape, although distribution
tends to be spatially variable. Salts, although typically concentrated in drainage basins, can be mobilized during
aeolian activity when surface sediments dry out and are widely
transported to contribute to rock weathering elsewhere
(Cooke et al., 1993).
Because of the logistical difficulties associated with working
in characteristically remote and climatically challenging desert
regions, most of our understanding about salt weathering has
been provided by laboratory experimentation and it is
Figure 11 Death Valley saltpan showing well-developed evaporite

deposits of salt with patterned ground phenomena evident in the
form of salt polygons with clay pinnacles.
arguable that some of the output from early studies more

closely reflected experimental design rather than the actual
conditions encountered in arid regions because of the use of
unrealistic temperature regimes and salt concentrations
(McGreevy and Smith, 1982; Smith et al., 2005). Advances in
technology that have enabled the remote collection of detailed
and long-term environmental data have greatly improved
understanding of the complexity of microenvironmental
conditions in deserts (Viles, 2005), which in turn has resulted
in significant refinement of laboratory simulation experiments
to reflect this, with the resultant production of data that
are much more representative of real-world conditions. What
follows is an overview of the key components of salt weathering but it is important to note that under real-world
conditions, these mechanisms rarely if ever operate in isolation but are closely interlinked.
There are three key aspects of salt weathering that need to
be considered. First, salt is an exploitative agent of rock weathering requiring the preexistence of structural weaknesses
within rock that enable its ingress into rock fabric. Second,
moisture is an essential component of effective salt weathering
and it may be derived from direct rainfall or from more occult
sources such as fog, dewfall, and groundwater. Finally, under
natural conditions, salts do not occur singly or in pure forms
and hence their behavior may differ significantly from that
exhibited under controlled laboratory conditions, where salts
are typically investigated in isolation and in pure solutions.
For example, laboratory experimentation has shown that
different salts vary in their weathering effectiveness, primarily
reflecting their crystallization and hydration properties
(Sperling and Cooke, 1985; Cooke et al., 1993; Goudie and
Viles, 1997), with sodium sulfate and magnesium sulfate
being the most effective, although it is always important to
remember that such effectiveness may be a reflection of the
experimental conditions. In the natural environment, salts
such as calcium sulfate, sodium chloride, and calcium carbonate are the most commonly occurring salts (Goudie and
Viles, 1997) most commonly occurring in association with
active weathering features but these salts do not perform so
effectively under laboratory conditions, indicating that the
effectiveness of salts under real-world conditions is as much a
result of synergistic interactions with other salts and minerals
as it is a reflection of the efficacy of the salt alone.
Figure 12 Panamint Valley playa, California.
209
Salt weathering involves both physical and chemical

mechanisms. The former includes crystallization pressures,
volumetric increase arising from phase change related to
hydration/dehydration and thermal expansion/contraction,
while the latter involves the creation of alkaline conditions in
the presence of salts under which normally durable elements
such as silica can be destabilized and subject to chemical
dissolution. The effectiveness of each of these is difficult to
assess because, under natural conditions, they may operate
sequentially or in combination, the action of one enhancing
or detracting from the effective operation of another (Smith
et al., 1992).
4.12.3.2.1
Crystallization
Laboratory simulation studies suggest that salt crystallization

is the most effective in terms of rock breakdown (Sperling and
Cooke, 1985) and involves the disruptive expansive pressures
exerted by salts in solution crystallizing out within pore
spaces, along cleavage planes in minerals, and along microfractures and joints (Figures 13(a) and 13(b)). Salts such as
thenardite (Na2SO4) and halite (NaCl) can exert as much as
340 and 650 atmospheric pressure, respectively, when oversaturated by a factor of two at 50 1C (Winkler and Singer,
1972). The significance of these pressures lies in the fact that
typically the tensile strength of rock generally falls in a range
of 14300 atmospheric pressure (Goudie, 1985), with the
result that salt crystallization within pore spaces and microfractures may be sufficient to disrupt intergranular bonds, resulting in granular disintegration and release.
Salt crystallization is a complex mechanism, with its
effectiveness determined by factors such as fluctuating temperature and humidity conditions and the solubility and
crystallographic properties of different salts. In the main,
strong evaporation and high temperatures favor effective
crystallization, although some salts such as sodium sulfate and
magnesium sulfate exhibit a significant decrease in solubility
when temperatures decrease. Therefore, when sufficient
moisture is available for salt to go into solution and enter the
fabric of rock depending on the salts or combination of salts
present, crystallization can potentially occur in response to
both evaporative moisture loss with temperature increase
and also to a decrease in temperature associated with nocturnal cooling (Goudie, 1985). Consequently, when different
salts are present, crystallization may have the potential to
operate effectively over a wide range of microenvironmental
conditions.
Some salts are more soluble than others and therefore have
a greater potential for penetration of rock fabric. Sodium
chloride for example exhibits consistently high solubility values across a wide range of temperatures in sharp contrast to
calcium sulfate, which, in its pure form, has poor solubility
characteristics. However, the solubility characteristics of individual salts as determined under laboratory conditions may
not accurately reflect actual conditions in the real world because of the role of synergistic interactions between different
salts. For example, the solubility of calcium sulfate increases in
the presence of sodium chloride and other salts such as
magnesium sulfate (Goudie and Viles, 1997; Wollmann and
Voigt, 2008), thereby demonstrating the difficulty of providing
(a)
(b)
Figure 13 Scanning electron microscopy images showing

(a) sodium chloride salt crystals between the cleavage planes of
mica and (b) sodium chloride crystals developed along the line of
a microfracture within feldspar note the cubic habit of the
sodium chloride crystals. Image width for (a) is 70 mm and for
(b) is 120 mm.
a definitive measure of the weathering effectiveness of

individual salts.
The crystallographic characteristics of salts are also important in determining their efficacy. For example, salts with
an acicular (needle-like) crystal nature (e.g., calcium sulfate,
sodium sulfate) and a preferred crystal orientation appear to
be the most effective in producing rock breakdown primarily
because of the focused crystal growth pressures exerted on
confining pore walls and within microfractures (Figure 14).
However, although crystallographic properties are important,
they are secondary to the solubility characteristics of a particular salt because, first and foremost, for a salt to achieve
rock breakdown, it must be able to enter the fabric of the rock
in sufficient quantities.
Aside from the properties of different salts, the effectiveness
of salt crystallization in terms of rock breakdown reflects the
rate of crystallization and the location and degree of salt
concentration within the fabric of the rock. Crystallization
rates in turn reflect such factors as air and rock temperature
and the extent of air movement over rock surfaces that can
increase evaporation rates (Rodriguez-Navarro and Doehne,
1999). The greater the crystallization rate, the greater the
pressures exerted on confining minerals and hence the greater
210
subsequent breakdown and surface retreat through multiple

scaling or flaking (Warke and Smith, 2000) (Figures 16(a)
and 16(b)).
4.12.3.2.2
Figure 14 Scanning electron microscopy image showing the

acicular (needle like) crystal habit of calcium sulfate deposited within
the cleavage planes of mica Image width is 260 mm.
Figure 15 Scaling features in sandstone; scales are typically 12 cm

in depth and 10 cm or more in diameter. Flaking is more often
associated with fine-grained rock types while scaling can develop in
both fine- and coarse-grained lithologies.
the potential for disruption of intergranular bonds. The

location and concentration of crystallizing salt within rock
will have an impact on the nature of rock breakdown. Salts
commonly occur in association with the loss of surface rock
layers through scaling and flaking (Dragovich, 1967; Rodriguez-Navarro et al., 1999; Warke and Smith, 2000), with the
depth of material released reflecting wetting depth (Dragovich, 1967), which describes the extent to which moisture and
hence salt in solution can penetrate rock before being drawn
back to the surface by evaporation (Figure 15). The mechanics
of salt accumulation are complex, reflecting both the solubility
characteristics of the salt, the porosity and permeability characteristics of the rock and the external microenvironmental
conditions such as air temperature and air movement
(Camuffo, 1984). Further complexity arises when more than
one salt is present as differences in solubility and hence mobility characteristics may result in crystallization occurring
at different depths within stone, which may accelerate
Hydration/dehydration phase change
Some salts exhibit significant volumetric change associated

with phase change as they alter from their dehydrated to
hydrated forms and vice versa. The volumetric increase or
decrease associated with such phase changes can be considerable, with salts such as sodium sulfate and sodium carbonate exhibiting the greatest volume changes of more than
300% during hydration (Goudie, 1985; Goudie and Viles,
1997). However, although such volumetric expansion measured in pure samples of salt under controlled laboratory
conditions is undoubtedly significant in terms of its potential
for rock weathering and breakdown, laboratory evidence tends
to suggest that it is a less effective mechanism when compared
with the disruptive effects of salt crystallization. In reality, it is
impossible to separate out the relative disruptive effects of salt
crystallization and volumetric change arising from phase
change because a single salt may be involved in both the
mechanisms of initial crystallization and subsequent hydration/dehydration. Further complexity is introduced when salts
occur in combination as this can alter the temperature at
which phase changes occur. For example, the phase transition
temperature for a pure form of sodium sulfate decreases from
32.4 to 17.9 1C where sodium chloride is present (Goudie,
1985; Obika et al., 1989; Goudie and Viles, 1997), bringing it
well within the range of temperature conditions prevalent in
most desert environments, meaning that such phase changes
could occur in the natural environment with much greater
frequency than indicated by laboratory experiments.
Not all salts exhibit phase changes. Sodium chloride, for
example has no stable hydrated form at temperatures above
0 1C (Johnston, 1973) and is therefore ineffective with regard
to rock breakdown through the formation of hydration pressures alone. However, sodium chloride is a deliquescent salt,
which means that it is strongly hygroscopic, readily absorbing
moisture from the atmosphere. Deliquescence differs from
phase change in that the former results in salt absorbing sufficient moisture to go into solution, while the latter describes
changes in the salts solid crystalline structure as it absorbs
atmospheric moisture and alters from its dehydrated to its
hydrated form (or vice versa) without the salt going into
solution.
The significance of deliquescence, particularly in a salt-rich
environment with fluctuating atmospheric humidity conditions, is that it provides a mechanism for the wetdeposition, surface accumulation, and gradual penetration of
salts into the fabric of the stone (Price, 1993; Zezza and Macr,
1995; Goudie and Viles, 1997; Camuffo, 1998). Deliquescence forms part of a complex process whereby
ycondensation nuclei begin to absorb a number of water molecules, and then grow in size becoming first deliquescent, then a
very concentrated hygroscopic solution and later a dilute solutiony(Camuffo, 1998: 141142).
Different hygroscopic salts exhibit particular threshold

relative humidity values above which they start to deliquesce.
211
Surface
Flake 1
Scrabo sandstone
Fracture 1
Epoxy resin infilling

pore spaces
Subparallel fractures filled
with epoxy resin
Flake 2
ZONE 1
Zone of clearly defined multiple,
subparallel fracture development
affecting the outer 4mm of stone
Fracture 2
Flake 3
Fracture 3
10 mm
ZONE 2
Subsurface zone of increased
porosity reflected in penetration
of epoxy resin. SEM examination
of this zone shows extensive interand intra-granular microfracturing
with widening and coalescence of
many of these microfractures
ZONE 3
Zone of gradually decreasing
porosity with increasing
distance from surface layers.
This decrease in porosity and
increased granular cohesion is
reflected in reduced penetration
of epoxy resin into pore spaces
200 mm
Location of sample section

on sandstone block B
See above
for detail
Zone of salt
enrichment
(a)
(b)
Figure 16 (a) Multiple flaking in sandstone related to salt deposition in subsurface layers. Reproduced from Figure 3 in Warke, P.A., Smith,
B.J., 2000. Salt distribution in clay-rich weathered sandstone. Earth Surface Processes and Landforms 25, 13331342. (b) Cross-sectional view
of the upper 10 mm of sandstone compiled using a composite of scanning electron microscopy images (magnification 40). This diagram
shows the different structural components associated with the development of surface flaking phenomena.
212
These values are called the equilibrium relative humidity

threshold values. Laboratory experimentation on pure salt
samples has shown the value of this for halite (NaCl) to be
between 75.1 and 75.5% over a temperature range between 0
and 30 1C (Arnold and Zehnder, 1990; Price, 1993). Above the
equilibrium humidity value, hygroscopic salts will exhibit
deliquescence but if the ambient humidity decreases below
this, the salt solution will become saturated and with a
continued humidity decline salt crystallization will begin.
Under real-world conditions where impurities abound, these
threshold values of equilibrium relative humidity may be
quite different but they give an indication of the type of
conditions under which deliquescence operates.
4.12.3.2.3
Thermal expansion/contraction
As with insolation weathering, greater theoretical knowledge

exists regarding the potential role of volumetric expansion and
contraction of salt crystals associated with heating and cooling
than knowledge based on actual data from field and laboratory study. Cooke and Smalley (1968) and Winkler (1973)
demonstrated that the theoretical possibility exists for differences in coefficients of thermal expansion between certain
salts and common rock minerals to create conditions of
intergranular stress within rock fabric. For example, sodium
chloride has a thermal expansion coefficient that is greater
than that of granite, expanding by 0.5% when heated through
temperatures of 060 1C, with a temperature increase of
100 1C needed to produce a 1% volumetric increase (Cooke
and Smalley, 1968). These temperature conditions are not
unrealistic but occur frequently on rock surfaces in arid regions, especially when a low rock surface albedo is combined
with poor thermal conductivity (Warke and Smith, 1994;
Smith and Warke, 1997). Although thermal expansion of salt
has been cited as contributing to rock breakdown (Johanessen
et al., 1982; Mottershead, 1989; Smith et al., 2005), it is
almost impossible to extricate the disruptive weathering effects
directly attributable to salt crystallization in the first instance
from those resulting from subsequent thermal expansion and
contraction. As Smith et al. (2005) suggest, the actual disruptive effect of repeated expansion and contraction of salt in
pore spaces within surface and near-surface layers of rock may
be minor in comparison with the effects of the other salt
weathering mechanisms and may operate over long time
periods but it should not be discounted because of this.
4.12.3.2.4
Chemical dissolution effects
The chemically disruptive effect of salt is commonly overlooked

in a general review of salt weathering, with the emphasis placed
on the physically disruptive effects described in the previous
sections. However, the combination of salt and moisture creates alkaline conditions that have the potential to destabilize
normally stable and durable elements such as silica and the
minerals that it forms a significant part of (e.g., aluminosilicates). This is particularly true for quartz, which, although a
normally durable mineral, can become more susceptible to
chemical corrosion under the alkaline conditions associated
with an abundance of salt with disruption of intergranular
bonds and crystal lattices and dissolution and removal of silica
(Young, 1987; Magee et al., 1988; Young, 1988; Goudie, 1997).
As quartz is a major component of many common rock
types such as sandstone and granite, any mechanism that

contributes to its deterioration will have significant rock weathering implications, with release of material through granular
disintegration creating openings for subsequent moisture ingress. Persistent alkaline conditions may also contribute to the
disruption of alumino-silicate minerals (e.g., feldspars and
mica) through degradation of the silica component within
crystal lattices and eventual disintegration or dissolution of the
mineral grain.
Chemical dissolution of specific elements such as silica
under alkaline conditions is not primarily dependent on
fluctuating subaerial conditions but may operate equally
well beneath the ground surface where temperature and
humidity conditions are comparatively stable. However, salt
crystallization, thermal expansion and contraction, and phase
changes are dependent on the temperature and humidity
fluctuations in the subaerial environment for their effective
operation. The significance of these different process domains
is evident in the rapid post-excavation deterioration of previously buried archeological masonry that, although buried in
salt-rich sediments, retained its structural integrity until exposed to the subaerial environment (Goudie, 1977; Selwitz,
1990; Maekawa and Agnew, 1996). This effect is also demonstrated by previously unpublished field data where quarry
fresh blocks of limestone were partially embedded in fine saltrich sediments of the Death Valley playa for 18 months. As
Figure 17 shows, the upper portion of the limestone block
exposed to fluctuating temperatures and humidity underwent
significant deterioration as salts drawn up in solution from
the playa sediments disrupted structural integrity through the
combined effects of crystallization, thermal expansion, and
phase changes. In comparison, the lower section of the block
embedded in the playa sediments for 18 months and protected from the fluctuating subaerial environmental conditions retained its structural integrity despite being loaded
with salts kept in solution by the damp playa sands.
Salt weathering is an important component of rock breakdown in arid regions primarily because of the availability of
salt. However, it is important to reiterate that there are two
prerequisites to the effective operation of the mechanisms
comprising salt weathering: first, the presence of existing
lines of weakness such as intergranular joints, micro-fractures,
cleavage, and bedding planes and second, the presence of
moisture, which enables mobilization of salts in solution and
their entry into the fabric of the rock. Without these factors
being present, salt remains a relatively inert component of the
rock weathering system in desert regions.
4.12.3.3
Frost (FreezeThaw) Weathering
Moisture is essential for the operation of frost weathering

and therefore it is of limited effectiveness with regard to
rock breakdown in the majority of contemporary deserts.
However, in high altitude and some mid-latitude deserts (e.g.,
Patagonia Argentina; Nevada USA), its significance may be
greater and therefore warrants inclusion here (see Chapter
4.15).
Traditionally, frost weathering was attributed to the 9%
volumetric expansion that accompanies the conversion of
water to ice within rock fractures and joints (French, 1976;
Portland limestone
cm
cm
(a)
disruption of rock by frost has been gained from laboratory

simulation experiments that take no account of the effect of
salts and other impurities in the water as well as the dynamic
nature of temperature fluctuations that occur in the natural
environment and that may greatly alter the actual conditions
required for water to freeze (McGreevy, 1981). The efficacy of
frost weathering will also be determined by structural properties of the rock such as the number and size of natural
planes of weakness and porosity characteristics.
Frost weathering is typically associated with the production
of angular shattered rock debris, but in desert regions, such
shattering of rock is also associated with insolation weathering, where it is postulated that long-term exposure to repeated cycles of heating and cooling lead to the physical
disintegration of rock. Because moisture is essential for the
effective operation of frost weathering, its role under contemporary conditions in the majority of deserts is probably of
limited significance. However, on occasions, conditions of
moisture availability and subzero temperatures may occur and
during such times the amount of geomorphological work
triggered by frost weathering exploiting structural weakness
created by other weathering processes may be considerable.
4.12.3.4
Untreated
cm
cm
(b)
Figure 17 Limestone block, (a) before and (b) after 18 months of

exposure to conditions on the salt-rich playa surface in Death Valley,
California. The lower quarter of the block that was buried in surface
sediments has retained its structural integrity despite being loaded
with salt while the upper portions of the block exposed to fluctuating
subaerial conditions have undergone significant deterioration. X-ray
diffraction identified sodium chloride and calcium carbonate as being
the main salts associated with the deterioration observed.
Ollier, 1984). However, an alternative theoretical model based

on the migration of water through freezing rock was proposed
by Walder and Hallet (1985), whereby crack or fracture
growth within rock occurs in a quasi-steady manner at temperatures well below freezing as water migrates toward freezing centers and exerts considerable hydraulic pressure at the
migration front, which gradually extends the crack or fracture.
The pressures developed by the freezing of water depend to a
large extent on the width and the depth of the fissure in which
the ice forms and the overall effectiveness of this process is
thought to be related to the frequency of freezing events and
the rate of freezing. Lautridou and Ozouf (1982) suggest that
rapid freezing is much more damaging to rock because it reduces the amount of water lost by evaporation and cryosuction and therefore increases the potential energy available for
frost weathering. Consequently, if freezing occurs sufficiently
rapidly and the water/ice cannot escape from rock pores,
pressure will result from the 9% volumetric expansion as the
water changes to ice. As with salt weathering, much of our
understanding of the actual mechanisms involved in the
213
Chemical Weathering
The traditional assumption of a relatively simple relationship

between weathering and climate (Peltier, 1950; Strakhov,
1967) resulted in the perception that chemical weathering
played no significant role in rock weathering in arid regions.
Although this perception remains, it is now acknowledged
that such simplistic generalizations do not represent the
complexity of conditions on the ground (Cooke and Doornkamp, 1990) and that even under extreme conditions,
chemical weathering can occur, albeit in a spatially and temporally limited form (Balke et al., 1991).
The extent to which rock in deserts is susceptible to
chemical weathering reflects interactions between intrinsic
and extrinsic factors. The former includes the mineralogical
characteristics of the rock whereas the latter includes environmental factors such as moisture availability, temperature,
slope, aspect, drainage conditions, the presence of salts, and
biotic activity. As previously discussed, weathering in general
has been characterized by its superficiality and selectivity
(Cooke et al., 1993) and this description is especially applicable to chemical weathering, where superficiality reflects the
typically shallow zone of moisture penetration while selectivity arises from the susceptibility of certain rock types and
minerals to chemical weathering. Chemical weathering, like
salt weathering, comprises several different mechanisms and
in arid regions the most important of these include:
Solution dissolution of minerals under both acid and

alkaline conditions.
Hydration addition of moisture to a mineral and is important with regard to the weathering effects of clay minerals and salts.
Oxidation addition of oxygen to a mineral.
Chelation chelating agents are produced by plants and
lithobionts and they extract ions from minerals, thereby
214
increasing the weathering rate over and above that which

occurs in the absence of plants.
Solution is probably the most significant of these reactions and
unlike other environments arid regions can give rise to both
alkaline and acidic conditions. The former reflects the salt-rich
nature of these regions and creates the potential for weathering of normally durable minerals such as quartz because
of the susceptibility of the element silica to dissolution under
strongly alkaline conditions (Magee et al., 1988; Bland and
Rolls, 1998), a fact that will also have implications for any of
the alumino-silicates such as feldspar and mica in which silica
forms a major constituent of the crystal lattice and, thus, the
rock types in which they occur (Figure 18). At the other end of
the pH scale, dissolution under acidic conditions normally
reflects the impact of mildly acidic precipitation produced by
the interaction between atmospheric carbon dioxide and water
to form a weak carbonic acid, which reacts with minerals such
as calcite, and hence has implications for the stability of rock
types such as limestone, dolomite, and calcareous sandstones.
Dissolution under acidic conditions may also arise from
the effects of biological weathering by lithobionts (e.g., algae,
cyanobacteria, fungi, and lichen) that all secrete organic acids
as a part of their normal metabolic functions. These are discussed in more detail in the following section on biological
weathering.
Moisture availability is a fundamental requirement for
chemical weathering to occur and even under the most arid
climatic conditions, moisture is available in a variety of forms
including episodic rainfall, dew, fog (mostly coastal deserts),
and groundwater. The spatial and temporal variability of these
sources of moisture has been previously discussed and is reflected in the spatial and temporal variability of chemical
weathering. In arid regions, chemical weathering may be a
slow progressive process that takes place over many thousands
of years under particularly severe environmental conditions
but interspersed with episodes of accelerated activity during
periods of greater moisture availability. Despite its perceived
limitations, chemical weathering has twofold significance for
rock breakdown: first, because it provides a contemporary
mechanism for creation of weaknesses within rock through
the chemical and physical alteration and/or dissolution of
minerals with modification of microstructural properties

such as porosity changes that provide entry points for other
exploitative agents of weathering such as salt and second,
because chemical weathering may act to strengthen rock under
certain circumstances through the mobilization and precipitation of certain elements within rock to create a hardened
outer surface. This is referred to as case hardening and is
commonly associated with the development of tafoni particularly in sandstone.
4.12.3.4.1
4.12.3.4.2
10 cm
Figure 18 Pseudo-microkarst development in a quartzite clast from
Death Valley, California, with preferential dissolution-related
weathering along lines of bedding.
Mobilization and removal of elements
Weathering features normally associated with chemical dissolution and removal of material are described as karst and are
typically, although not exclusively, associated with calcareous
rock types. It is argued that because of the paucity of moisture,
vegetation, and as a consequence, soil carbon dioxide, karst
development should be very poor in arid regions (Jennings,
1983). However, as always, reality is often more complex than
the theories suggest and this is demonstrated by the presence
in many deserts of a range of karstic weathering features from
large (4100 m) to small scale (o10 mm). Large-scale features
include dolines, caves, and a range of karren and although
they can be found in many contemporary deserts, their initiation and development is usually attributed to former, wetter
climatic conditions, with only superficial modification taking
place under present-day aridity (Smith, 2009).
Rillensteine are one of the most frequently encountered
examples of small-scale or microkarstic features. These are
small sinuous rills typically o12 mm in depth and width,
occurring in association with calcareous rocks such as limestone and dolomite, and formed by the action of mildly acidic
dew (Figures 19(a) and 19(b)). Rillensteine can become
elongated in locations where prevailing winds consistently
blow dew across the rock surface and, once formed, they may
be self-sustaining as dew collects preferentially in the troughs
formed by previous dissolution, causing the progressive removal of calcite and surface lowering. Rillensteine have been
observed in many desert regions (Dunkerly, 1979; Sweeting
and Lancaster, 1982; Smith, 1987, 1988) and it is suggested
that these features are distinctive arid zone microkarst because
they develop slowly in response to limited moisture availability
and will be removed by more generalized dissolution under
wetter climatic conditions (Laudermilk and Woodford, 1932).
Although spatially restricted and limited to specific rock
types, the presence of rillensteine in even hyper-arid deserts
serves as a reminder of the availability and weathering
potential of even meager amounts of moisture.
Mobilization and precipitation of elements
The formation of features such as weathering rinds and casehardened surfaces in arid regions is attributed to the mobilization, precipitation, and gradual accumulation of specific
elements on or within the near-surface layers of rock, creating
a layer that is comparatively harder and thus more durable
than subsurface material (see Chapter 4.6). Iron and calcium
are probably the two most frequently cited constituent elements of case hardening and may be derived from both intrinsic
and extrinsic sources. Conca and Rossman (1982) identify an
aeolian source for calcium in case-hardened layers developed
on sandstone in Nevada with calcium-bearing windblown
215
(a)
Figure 20 Breaching of a case-hardened surface layer on sandstone

with preferential weathering of substrate material to create hollows as
substrate material is lost. Petra, southern Jordan. Image height is
approximately 2 m.
(b)
Figure 19 (a) Rillensteine developed by the action of dew on

calcareous rock with dissolution, creating a convoluted pattern of rills
that can become elongated when wind blows frequently from a
consistent direction, and (b) scanning electron microscope image
showing detail of the morphology of rillensteine with peaks and
troughs that may eventually become self-sustaining as they focus the
collection of moisture.
dust deposited in rock surface pore spaces and discontinuities,

where it is subsequently mobilized in solution during episodes of moisture availability, bonding with the host rock
during subsequent crystallization.
Case hardening may also result from the outward migration of elements such as calcium carbonate, which may cause
the substrate to be weakened as surface layers become
strengthened and consequently more susceptible to weathering if the case-hardened layer was ever breached (Conca
and Rossman, 1985). In many instances, development of a
case-hardened layer may reflect the action of several weathering mechanisms. For example, Goudie et al. (2002) attributed the case hardening of sandstone at Wadi Rum in
southern Jordan to the precipitation of iron by rainwash, with
subsequent more detailed analysis of substrate material
showing the presence of a concentration of other elements
such as calcium in association with crypto-endolithic organisms that may have played some part in the case-hardening
development through the formation of a calcareous subsurface cement (Viles and Goudie, 2004) (Figure 20).
The significance of case hardening for rock weathering in
arid regions is twofold in that it can act to temporarily retard
rock weathering or conversely contribute to the formation of
distinctive cavernous weathering features. With regard to the
former, while the case-hardened layer remains intact, it will

afford some protection against the effects of physical weathering processes. However, where the case-hardening constituents are derived from within the rock, strengthening of
surface layers will occur at the expense of an increasingly
weakened substrate. Once the case-hardened layer is breached,
weathering of the underlying substrate may proceed relatively
rapidly to form cavernous weathering features such as pits,
hollows, and eventually tafoni. Once established, Martini
(1978) suggests that these forms become self-perpetuating
because of the microenvironmental conditions within them
where absolute temperatures and temperature ranges tend to
be lower than on exposed rock surfaces, with humidity typically maintained at a higher level than outside the hollow
(Dragovich, 1981; Rogner, 1987; Smith and Warke, 1997).
Evidence indicates that the microenvironmental conditions
established are particularly conducive to salt weathering, with
many examples of where the back wall retreat within these
hollows proceeds through the combined effects of granular
disintegration, flaking, and scaling in the presence of salts
(Dragovich, 1967; Mustoe, 1983; McGreevy, 1985b; Sancho
and Benito, 1990).
As many of the earths desert regions show evidence of
Holocene climate change with a gradual shift from more humid
to the present-day arid conditions, it could be postulated that
chemical weathering under past climatic conditions played a
more significant role in rock weathering than under present-day
climatic conditions. However, it is important to remember that
chemical reactions are not stopped by low average annual
precipitation and desiccating heat and that the ubiquity and
efficacy of chemical weathering processes is evident at various
scales in contemporary arid regions (Pope et al., 1995).
4.12.3.5
Biological weathering is an umbrella term for a complex group

of processes that include the actions of a wide range of
organisms. In arid regions, higher plant assemblages tend,
for the most part, to be sparsely distributed or absent
altogether from areas that experience particularly extreme
environmental conditions. Consequently, in the context of
216
biological weathering in arid regions, the emphasis is on

the more primitive but nonetheless ancient organisms that
comprise the lithobionts a group that includes cyanobacteria, algae, fungi, and lichen that inhabit rock surfaces.
Biofilms are another part of this group and describe a mixed
community of microorganisms that form a thin layer and are
typically, but not exclusively, epilithic (Koestler et al., 1997).
Traditionally, only scant attention has been paid to the role of
biological weathering in arid regions primarily because of the
perceived absence of plants and the assumption that less
complex forms such as algae and lichen would play no significant role in rock weathering and breakdown. However,
although debate persists about the importance of biological
weathering processes per se, greater field investigation and
advances in microscopy and associated analytical techniques
have supported a growing recognition of the reality of biological weathering in arid regions (Viles, 1995).
The type of organisms that colonize rock surfaces reflects
the combined effects of the mineralogical and structural
characteristics of the rock its bioreceptivity (Guillitte and
Dreesen, 1995; Hutchens, 2009) and the limitations imposed
by the prevailing environmental conditions.
Bioreceptivity is mainly dependent on structural properties
such as rock surface roughness and porosity and the dominant
mineralogical characteristics. Large pore spaces facilitate surface colonization by algae, fungi, and lichen through the creation of sheltered anchor points for these organisms to fix
themselves to the rock surface (Guillette and Dreesen, 1995)
and sites where any incident moisture would tend to linger.
The chemistry of the mineralogical constituents of rock has
significant implications for the type of species able to colonize,
with, for example, species adapted to alkaline conditions
being most successful on calcareous rock such as limestone,
while those with a preference for more acidic conditions
preferring rock types rich in silica such as sandstone (Lisci
et al., 2003). Because of the severity of environmental conditions in arid regions, many organisms are endolithic in
habit, carrying out their lifecycle a few millimeters beneath the
rock surface or in microfractures and cracks primarily for
protection from the effects of prolonged desiccation. Consequently, the light transmissivity properties of rock and its
mineral constituents are extremely important because many of
these organisms are reliant on the process of photosynthesis
(Hall et al., 2008).
Viles (1995) identified the optimum environmental conditions for biological growth and weathering as occurring in
regions with abundant moisture; thus, in arid regions, biological activity will necessarily be more restricted because of
the stress imposed by aridity. However, with regard to colonization of rock, organisms can occupy either epilithic or
endolithic niches (Viles and Goudie, 2004). The former describes organisms that live on rock surfaces, while the latter
includes organisms that have evolved to cope with extreme
environmental conditions by living below the rock surface in
preexisting microfractures and cracks or within pore spaces
and boreholes. Although epilithic organisms can occur in
arid regions, endolithic forms become increasingly dominant
as moisture availability declines (Viles, 1995) and hence tend
to be the most significant in the context of rock weathering by
breaking down individual minerals and intergranular bonds,
thereby creating subsurface weaknesses that may be subsequently exploited by salts in solution.
Biological weathering describes rock breakdown arising
from biochemical and/or biophysical mechanisms and although these are described separately here, they are closely
interlinked with the action of one facilitating the action of
the other.
4.12.3.5.1
Biochemical effects
The iron-chelating ability of lichens and fungi is well documented (e.g., Danin and Garty, 1983; Seaward, 1997, 2004;
Banfield et al., 1999; Lisci et al., 2003; Allsopp et al., 2004) but
it is a mechanism that is more commonly associated with
environmental conditions that are more humid than those
typically associated with contemporary arid regions. It is important to remember that the complexity of microenvironmental conditions across desert surfaces creates microscale
ecological niches in which organisms can comfortably live and
where biochemical mechanisms can actively weather rock
(Friedmann and Galun, 1974). Biochemical weathering is also
typically associated with the secretion of organic acids during
normal metabolic functions. For example, a byproduct of
cellular respiration by algae, fungi, and lichens is carbon dioxide, which combines with organic moisture to form carbonic acid that in turn can contribute to localized small-scale
dissolution of susceptible minerals (Chen et al., 2000). Calcareous rock types are particularly susceptible because of the
high solubility of elements such as calcium (Smith et al.,
2000), although the significance in terms of rock weathering is
only evident at the submillimeter scale but comprising a
hierarchy of microdissolution forms associated with isolated
microbial organisms as well as those created by larger colonies
(Figures 21(a)21(c)). Another commonly secreted organic
acid usually associated with the metabolic functions of lichen
is oxalic acid. Oxalic acid is a chelating agent and reacts with
various minerals to form a range of different oxalates, with
some such as calcium oxalate being relatively insoluble so that
once formed it can play a bioprotective role on rock surfaces.
The complexity of biological weathering is highlighted by the
potential for bioprotection as opposed to biodeterioration
arising from the effects of biochemical mechanisms operating in
certain conditions. Although this remains a poorly understood
and widely debated aspect of biological weathering, examples
from arid regions and the significance for rock weathering have
been reported in the literature (e.g., Kurtz and Netoff, 2001;
Viles and Goudie, 2004). Viles and Goudie (2004) described the
case-hardened sandstone at Wadi Rum in southern Jordan as
occurring in association with cryptoendolithic biofilms containing cyanobacteria and fungi, which they suggested may have
facilitated the development of hardened surface layers of rock
through cementation of mineral grains, and that the breaching
of this layer led to the formation of caverns and hollows by the
preferential weathering and loss of weakened substrate material.
The complexity of biochemical weathering mechanisms in
arid regions is much better understood than it used to be but
much remains to be discovered about the factors that determine whether there is a bio-protective or a bio-destructive
outcome and how this might change over time in response to
changing microenvironmental conditions. Despite the obvious limitations of biological weathering in arid regions, the
217
lichens primarily because of the catalytic effect of organic acids

and enzymes that create a more aggressive weathering environment. Although biochemical mechanisms operate at the
submillimeter scale, there may be an up-scaling effect,
whereby if these mechanisms are sufficiently widespread, their
small-scale actions may effect the development of meso-scale
features such as caverns and hollows as noted in more humid
climatic environments (McIlroy de la Rosa et al., 2011).
4.12.3.5.2
(a)
NONE
SEI 10.0kV
X950
10m WD 9.2mm
(b)
(c)
Figure 21 (a) Hierarchy of micro-scale pits formed by the in-situ

biochemical dissolution of limestone in Tunisia by microbial
organisms. The morphology of these features is attributed to the
excretion of organic acids associated with the normal metabolic
functions of epilithic microbial organisms, with the size of the pit
reflecting the action of single or colonies of organisms; (b) scanning
electron image of a cluster of epilithic microorganisms established on
the surface of sandstone covered in a manganese rich rock varnish
shown in (c); (c) remnant of rock varnish-covered sandstone (Utah)
that shows some evidence of a secondary varnish development but
the surrounding sandstone is extremely friable and unstable.
significance of biochemical mechanisms in the context of rock

weathering is that rates of weathering may be greater than
chemical weathering in the absence of algae, fungi, and
Biophysical effects
Biophysical weathering effects consist of those that are directly

associated with rock breakdown and those that play an indirect role. Direct physical weathering effects are most efficiently demonstrated by lichen where stresses arising from the
periodic wetting and drying of lichen thalli are sufficient to
pluck fragments and individual mineral grains from rock
surfaces (Cooks and Otto, 1990; Moses and Smith, 1993;
Seaward, 2004). The penetration of fungal hyphae from beneath lichen into the fabric of rock may also facilitate physical
breakdown by providing a mechanism for moisture and salt
ingress and by the gradual separation of mineral grains or
parts of grains as the number of hyphae and the extent of their
penetration increase. At a smaller scale, algae also contribute
to the direct physical disruption of rock through the expansion
and contraction of the algal mucilage associated with repeated
wetting and drying of individual algal bodies and colonies
located beneath rock surfaces in pore spaces and/or microfractures (Hall, 1990).
Indirect biophysical weathering effects include changes to
rock surface temperatures caused by the presence of lichens
that can alter albedo characteristics of the rock surface. Darkcolored lichens have been shown to significantly increase
heating of underlying rock and it is suggested that these
lichens make an indirect contribution to rock breakdown
through enhancement of thermal stress between surface and
subsurface rock layers, leading eventually to fatigue-related
failure (Schwartzman et al., 1997; Carter and Viles, 2004; Hall
et al., 2005). The higher temperatures established may also
enhance the effectiveness of any biochemical weathering
mechanisms associated with acids secreted during lichen metabolism. Any physical deterioration of rock surfaces or subsurface material arising from biophysical actions will leave
points of weakness that may subsequently be exploited by
other agents of weathering such as salt and moisture.
The connection between biochemical and biophysical
mechanisms is highlighted by Paradise (1997), who reported the
operation of both beneath lichen on sandstone in Arizona,
where biochemical weathering through the effects of acid secretion and mineral dissolution weakened the microstructure of the
sandstone, which in turn aided subsequent hyphal penetration
and related biophysical weathering as the lichen grew and the
thallus expanded outward to continue the process on fresh rock.
In the last few decades, technological advances in microscopy have enabled better understanding of the micro-scale
changes caused by a range of organisms from lichens to algae
and has demonstrated their presence in even the harshest of
environmental settings. Field evidence indicates that these
organisms are well adapted to thrive under extreme environmental conditions, entering a state of minimal metabolic activity during periods of desiccation and becoming more active
218
in response to episodes of moisture availability. Their role in

the weathering of rock in arid regions may appear to be
somewhat restrictive and dependent on the availability of
moisture but nevertheless biological weathering is another
component of the desert weathering system and, as such,
warrants serious consideration.
4.12.4
The Desert Weathering System
Comment has been made several times in this chapter and

elsewhere concerning the superficial and selective nature of rock
weathering in contemporary arid regions (Mabutt, 1977; Cooke
et al., 1993), with much effort expended on laboratory investigation of the action and efficacy of individual weathering
processes. However, despite advances in understanding of the
factors controlling these processes, the link between data gained
in the laboratory and the actual real-world operation of processes and their association with weathering features is still not
strong. As Smith (2009: 69) rightly observes y.weathering
studies [in deserts] continue to be characterised more by uncertainties and gaps in knowledgey.., with concentration on
individual processes producing a poor understanding of how to
translate data collected from short-term studies into a robust
and thorough appreciation of the cumulative effects of smallscale weathering processes on large-scale and long-term landform and landscape development.
show evidence of micro-solutional weathering features while

noncalcareous rock types do not, despite being exposed to the
same conditions.
4.12.4.1.3
4.12.4.1.4
4.12.4.1
System Components
These gaps and uncertainties in knowledge reflect an incomplete understanding of all the components of the desert
weathering system and the extent to which they are interconnected. As described by Smith (1996), rock weathering
and breakdown in any environmental setting reflects the
interactions between the principal components of this system,
with change in any one of these components having a significant impact on the others.
4.12.4.1.1
Materials
Material characteristics include factors such as porosity, permeability, thermal properties, and mineralogy and these have
a major controlling impact on the susceptibility of a particular
rock type to weathering. Some material characteristics can
change over time. For example, weathering and the gradual
removal of silica over time under alkaline environmental
conditions may lead to breakdown of primary minerals such
as feldspar and mica and an overall reduction in the cohesive
strength of the rock, with a resultant increase in its weathering
susceptibility despite no apparent change in the nature and
intensity of weathering processes acting upon it.
4.12.4.1.2
Processes
Process characteristics reflect the prevailing microenvironmental conditions but their effectiveness is closely linked to
material characteristics. For example, the efficacy of chemical
dissolution under mildly acidic conditions increases in the
presence of soluble minerals such as calcite. Consequently,
even under conditions of hyper-aridity, calcareous rock may
Form/morphology
Form characteristics include the surface morphology of the

rock outcrop or debris. For example, the presence of certain
points or lines of weakness on rock surfaces may provide the
focus for weathering and the development of features such as
pits, hollows, and alveolae (Figures 22(a) and 22(b)). However, form is not only related to material characteristics but
may also influence the nature of weathering processes active at
the rock/air interface. Some rock weathering features such as
alveolae and rillensteine, once established, may be self-perpetuating. Conditions in shaded hollows and alveolae tend to
have higher levels of atmospheric humidity and remain moist
for longer periods than exposed rock surfaces, with the result
that processes such as salt weathering can remain active for
longer periods and hence achieve more rock breakdown, thus
further enhancing the development of the alveolae. Similarly,
once the microsolution rock surface weathering features called
rillensteine have developed, they tend to preferentially channel subsequent moisture flow over the rock surface, thereby
reinforcing the development of the existing rillensteine.
Environment
Environmental characteristics have a major influence on

interactions between materials, form, and processes. A distinction may be made between macro and microenvironmental conditions the former referring to regional
conditions of temperature and moisture while the latter describes conditions at the rock/air interface. Changes in macroenvironmental inputs affect the nature and effectiveness of
weathering and transport processes and, although microenvironmental conditions are primarily determined by macroenvironmental factors, they are also strongly influenced by
the form or morphology of rock surfaces. For example, where
the rock surface is relatively smooth, temperature and humidity conditions at the rock/air interface will remain relatively uniform across the surface but an uneven surface with
pits and shaded hollows will give rise to spatial variability of
temperature and humidity conditions conditions that may
be conducive to enhanced weathering activity.
4.12.4.2
Desert Weathering A Nonlinear Dynamic

System?
In addition to understanding the key components of the

typical desert weathering system, it is important to remember
that this system displays the classic characteristics of a nonlinear dynamical system typified by unpredictable responses to
energy inputs (Phillips, 2005). This unpredictability reflects
the complexity of the system as a whole and the interactions
between variables operating within the system. The impact of
complexity and the unpredictability of weathering pathways
are explained by the factors outlined below (see Chapter 4.2).
219
may create microenvironmental conditions conducive to more

intense weathering activity and hence perpetuate preferential
surface breakdown and retreat.
4.12.4.2.2
(a)
(b)
Figure 22 (a) Alveolae weathering features developed in sandstone;

(b) exploitation of cross-bedding structures in sandstone that form
inherent lines of weakness where porosity and permeability properties
may facilitate the ingress of moisture and salt (southern Jordan).
Image height is approximately 1.5 m.
4.12.4.2.1
Feedback mechanisms
Complexity and associated unpredictability within arid zone

weathering arises from the operation of positive and negative
feedback mechanisms. Positive feedback causes a change in
the system state through the acceleration of weathering activity
while negative feedback acts to retard weathering activity. The
switch from positive to negative feedback conditions or vice
versa may be triggered by seemingly insignificant changes in
intrinsic or extrinsic system components sufficient to destabilize the system and initiate an increase or decrease in
weathering activity. Changes in intrinsic factors such as the
ability of the rock fabric to resolve imposed stresses arising, for
example, from repeated salt crystallization, may result in
breakdown without any increase in the stress applied. With
regard to extrinsic factors, an increase in atmospheric humidity associated with a drop in air temperature at night may
be sufficient to initiate deliquescence in certain susceptible
salts such as sodium chloride and trigger a period of active salt
mobilization and penetration of rock fabric that may facilitate
subsequent disruption of rock associated with crystallization
pressures as air temperature increases and humidity levels
decrease at daybreak. Feedback mechanisms are also invoked
to explain the development and preferential weathering
within alveolae and tafoni, whereby air temperature and humidity conditions within a hollow on a rock surface or in an
area shaded for longer periods of time than exposed surfaces
Magnitude and frequency
Sudden, high-magnitude increases in the stress burden on

rock through the operation of weathering processes may result
in rapid or catastrophic breakdown of rock fabric. Breakdown
may not have occurred, or occurred less dramatically, if the
stress had been applied more gradually and this is related to
the threshold concept, whereby, if the critical threshold of
material strength is greater than the stress applied, no apparent change will occur. However, the absence of obvious visual
damage in rock can be misleading as it implies that the material is unaltered even though microscopic external and internal changes may have occurred, the accumulation of which
may eventually reduce cohesive strength and lead to fatigue
failure. Similarly, the efficacy of extreme heating and high rates
of surface temperature change (thermal shock) are clearly
demonstrated in the natural environment during bushfires,
when widespread splitting and spalling of natural rock outcrops as critical thresholds of strength are exceeded by highmagnitude externally derived stress operating over very short
periods of time (minutes) as the fire-front passes.
Depending on the nature of the environmental stress,
prolonged and continuous stressing of rock may be less destructive or not so obviously destructive as repeated highmagnitude short-term events. Although the magnitude and
frequency characteristics of stress events are important, material characteristics must also be considered. For example, the
impermeability of some rock types may leave them relatively
unaffected by high-magnitude weathering events, primarily
because weathering agents such as moisture and salt cannot
penetrate the rock fabric for lack of microfractures and pore
spaces. Other rock types, exposed to the same environmental
conditions, may be more susceptible because of their structural and mineralogical characteristics, which facilitate the
ingress of exploitative weathering agents. Consequently, in a
landscape where different lithologies are present, it is probable
that the nature and extent of weathering response will also be
different, creating a complex pattern of weathering activity and
hence sediment release.
4.12.4.2.3
Form convergence (equifinality)
The surfaces of many desert regions are mantled with angular

shattered debris similar to that observed in many cold highlatitude and/or high-altitude environments but the production of this debris in both hot and cold regions probably
reflects the action of quite different processes exploiting
structural weaknesses within the rock but resulting in a similar
debris endform. For example, in hot deserts thermal fatigue
and salt weathering may have a significant role to play in the
generation of angular shattered debris. It is important to recognize that different processes can give rise to similar weathering forms and because arid environments are widely
recognized as being rich in salts due to high rates of evaporation and a lack of fluvial outflow, there has been a tendency
to associate salt weathering with the widespread breakdown
and release of debris. Too often, weathering processes that are
immediately associated with observed rock weathering forms
220
Rock debris form adjustment to pulses of change
Lag
time
Relaxation
time
Characteristic form
time
Lag
time
(a)
Sensitive areas
<10 years
Geomorphological change
<10 years
Lag time
Relaxation time
Impulse of change
(b)
Insensitive areas
>100 years
Characteristic form
time
(c)
Time
Figure 23 Landscape sensitivity and the characteristics of different energy pathways. (a) Shows the basic stages involved in a system reaction
to an input of energy, with the initial impulse of change providing the driving force but with a lag time between that and obvious response in the
system, followed by a period of relaxation during which the system adjusts to the new state and assumes a characteristic form appropriate to
this new state. (b, c) Show the differences between these stages in sensitive and insensitive systems, whereby impulses of change may be
more frequent and lag times may be shorter in the former because of the proximity to energy pathways. In contrast, the latter shows a system
dominated by conditions of stability. Reproduced with permission from Figure 8 in Brunsden, D., 1980. Applicable models of long-term landform
evolution. Zeitschrift fur Geomorphologie 36, 1626.
221
are judged to be solely responsible for their formation. But

this assumption of guilt by association can obscure the role
of both past and present processes. For example, repeated
heating and cooling of rock surfaces over many years, while
leaving no obvious physical trace, may have contributed
to microscopic alteration through a thermal fatigue effect,
whereby intergranular bonds are gradually weakened, facilitating the ingress of moisture and contributing to the apparent
efficacy of salts (Smith and Warke, 1997).
4.12.4.2.4
Sensitivity and system components
Further complexity arises from the coexistence of sensitive and

insensitive components of the landscape. Within the context
of landscape change, Brunsden (1980) identified components
that are either sensitive or insensitive to impulses of change,
with the former being fast responding and attaining new states
easily, whereas the latter are slow to respond, thereby taking
much longer to achieve new forms. Although Brunsdens
model was originally applied to landscape change, it is equally
relevant to explanation of the complexity of weathering response (Figure 23). In most desert regions, sensitive and insensitive areas can coexist, sometimes in relatively close
proximity, with the former located close to impulse pathways
such as ephemeral stream channels, the base of cliff faces, and
around playa margins, where an abundance of salt and sporadic recharging of subsurface and surface moisture can result in
active weathering and debris breakdown and change.
In such areas, various rock weathering forms can occur
but other depositional features such as rock varnish will be
absent because surface changes and deterioration of rock
occur more rapidly than the time required for varnish
formation (Figure 24). In contrast, insensitive areas may
undergo many decades or even hundreds of years between
impulse events because these areas are located away from
impulse pathways functioning as isolated surfaces. Such areas
can include rock outcrops and debris on hill-slopes and
abandoned alluvial fan surfaces, where weathering and
breakdown is very slow. Between endpoints of the sensitivity
spectrum lies a continuum along which sensitivity varies and
it is worth noting that the sensitivity status of an area may
change through time in response to re-routing of ephemeral
stream channels or environmental change associated with
normal seasonal changes in rainfall receipt or the long-term
effects of climate change.
4.12.4.3
Figure 24 Complex assemblage of hollows (alveolae) and caverns

(tafoni) developed in a sandstone outcrop (Utah), with the spatial
distribution of these features reflecting both the microenvironmental
and the groundwater conditions at the site. Groundwater rise and
higher levels of atmospheric humidity in the air at the base of the
outcrop create conditions conducive to rock weathering, with the
intensity reflected in the extent of material loss. Higher up the
outcrop above the level of the alveolae, the presence of a welldeveloped rock varnish indicates surface stability because of the time
needed for its formation. Image height is approximately 3 m.
Scale Issues (Spatial and Temporal)
Considerable knowledge has been gained from laboratory

experiments on the modus operandi of individual weathering
mechanisms operating at a small scale over relatively short
time periods. However, despite this knowledge, a disconnect
exists between these small-scale studies of process and the
means by which they translate into the actual initiation and
development of common weathering forms such as alveoli
and tafoni, split stones, scales, and flakes. Such uncertainty
may reflect the time-bounded nature of doctoral and research
projects and the consequent lack of attempts at up-scaling
from small-scale, short-term single process-centered studies to
larger-scale, longer-term multiprocess conditions that are
Figure 25 Principal landscape components of basin-range-type

desert with mountains fronted by alluvial fans that spill onto the
valley floor, where they integrate with playa sediments.
222
undoubtedly necessary for the development of weathering

forms commonly encountered in many desert environments.
All of the above contribute to the complexity of contemporary arid zone weathering and the variety of landforms
and associated weathering features typically occurring in most
deserts. In addition to these issues is the fact that many contemporary arid regions have only acquired this aridity during
the Holocene, with considerable climatic fluctuations and
periods of much wetter conditions during the Pleistocene,
which have left traces on present-day landscapes.
4.12.5
Inheritance and the Concept of Palimpsest
Weathering studies have often sought to explain the features

observed in terms of a prevailing set of environmental conditions. However, it is clear that as rock weathers over many
hundreds or thousands of years, it is exposed to a range of
weathering environments and processes. These different conditions can arise in two ways: first, from the effects of longterm climate change acting on rock outcrops and debris that
remain relatively static within the landscape and, second, from
the spatial relocation of debris from high to low altitude
as a result of flooding events and/or gravity-related mass
movements.
With regard to the former, most contemporary hot desert
environments have undergone climate change, with many
landforms and weathering features inherited from periods
when moisture was more readily available (Budel, 1982;
Twidale, 2007). These features may form large-scale (e.g., large
karstic cave systems) or small-scale (e.g., corroded cobbles
covered with rock varnish) components of the contemporary
landscape but they provide critical evidence of past conditions
in which chemical weathering processes may have played a
more dominant role when moisture was more abundant
within the weathering system.
Consequently, many desert landforms, despite forming an
integral part of the present-day landscape, owe their existence
to quite different morphogenetic conditions unrelated to
current environmental parameters. This recognition that many
characteristic desert landforms are inherited features has not
been so widely acknowledged in the study of smaller-scale
weathering forms developed on rock outcrops and released
debris. Yet, scale differences should not preclude consideration
that many smaller scale debris weathering forms may not be
the direct product of the environmental conditions in which
they are currently found (Butler and Mount, 1986) and that
past conditions may have a profound effect on the weathering
response of debris to contemporary conditions.
In addition to long-term climate change acting on
relatively static debris, complex multistage debris weathering
can arise from the spatial relocation of debris from high to
low altitude through intermittent fluvial transport under
comparatively stable regional climatic conditions. For example, in the basin-range deserts of the southwest USA,
where desiccation has become established only within the last
few thousands of years, the proximity of high mountain ranges and deep valley troughs creates a range of weathering
environments (Figure 25). Material initially released in a highaltitude subalpine environment is transported downslope to
low-altitude salt-rich playa margins, where comparatively

rapid disintegration can occur disintegration that is typically
attributed directly to the action of salt. Most of this material is
not transported directly from high to low altitude but experiences often, prolonged periods of storage within the
landscape system. During different episodes of storage, it is
exposed to weathering activity associated with site-specific
microenvironmental conditions. By the time debris eventually
reaches alluvial fan/playa margins, it may possess a complex
weathering history reflecting the cumulative effects of exposure to multiple weathering processes (Figure 26) (Warke,
2007).
Because weathering conditions within the natural environment are so complex, and because each fragment of rock
can carry a unique weathering-related legacy of damage and
alteration, the weathering response of debris to changes in
environmental conditions through spatial relocation may
widely differ, with some debris readjusting to new conditions
through accelerated breakdown while other material, often of
the same lithology, remains relatively stable and apparently
unchanging (Warke, 2007). Any change in the rate of debris
breakdown may have significant implications for the amount
of fine sediment available for subsequent aeolian and/or fluvial mobilization.
Because of the complexity of desert landscapes, our attempts at interpreting what we see must necessarily take into
account the long-term history of the landscape itself and the
rock outcrops and debris that mantle it. The absence for the
most part of a protective layer of soil and vegetation leaves a
surface that is directly susceptible to the effects of subaerial
environmental conditions. This means abandoning oversimplification with regard to the weathering system and accepting that it is an inherently complex system in which past
weathering conditions contribute to the effectiveness of contemporary processes.
4.12.6
Conclusion
Our understanding of rock weathering in arid regions has

come a long way from early observations about the importance of heat and wind in desert weathering that underpinned
many misguided inferences regarding the formation of common weathering features. It is no longer sufficient to categorize arid regions as being dominated by physical
weathering processes on the basis of their aridity. The work of
recent decades in collecting extensive field data that have been
used to inform the design of more realistic laboratory experimentation has highlighted the complexity of surface
conditions and the interactions between different weathering
processes operating sequentially and/or cumulatively at a
number of spatial and temporal scales.
Weathering in arid regions should more accurately be described as a continuum with physical weathering processes at
one end and chemical weathering at the other and where
different areas within an individual arid region may be at
different locations on this continuum, reflecting their specific
microenvironmental conditions conditions that can change
across a range of time-scales. We now have a much better
understanding of the long-term environmental complexity of
223
BEDROCK
Debris release
Mountain
slopes
Mountain slopes and ephemeral stream channels
Joint
blocks
Fractured
blocks
Fine
debris
Gravity related mass movement / slopewash

Ephemeral
stream
channels
Mixed alluvium
Rapid
Delayed
Subsurface
Surface
Debris comminution
Remobilisation
Debris comminution
Alluvial fans
Alluvial
fans
Mixed alluvium
Rapid
Delayed
Surface
Subsurface
Debris comminution
Remobilisation
Debris comminution
Playa
margins
Playa margins and playa
Mixed alluvium
Surface
Subsurface
Debris comminution
Playa
Predominantly fine sediment
Surface
Subsurface
Decreasing altitude, debris size and moisture availability Increasing time, temperature salinity and
stress inheritance / weathering history
Key to weathering processes and transportation
Frost action
Chemical alteration & Biological
weathering
Salt weathering, Chemical alteration
& Biological weathering
Deflation
Cementation
Fluvial transport and comminnution.

Includes chemical dissolution, and wetting/drying
Rock varnish formation
Salt weathering, Chemical alteration, Biological

weathering & Insolation weathering
Links
Pathway
Debris type
Process group
Sediment store
Figure 26 Conceptual model of potential downslope debris transport pathways and weathering environments encountered in basin-range-type
desert environments. Adapted from Figure 2 in Warke, P.A., 2007. Complex weathering in drylands: implications for rock breakdown and
sediment release. Geomorphology 85, 3048.
224
arid regions and a greater recognition of the many weathering

forms that, although undergoing modification under contemporary conditions, owe their initiation and early development to past more humid climatic conditions.
Despite advances in our knowledge of arid zone weathering, many significant gaps in knowledge remain. Chief
among these is the need for a much clearer understanding of
the actual mechanics of synergistic interactions between weathering processes as these can have significant implications for
both the effectiveness of individual processes and the breadth
of environmental conditions over which particular weathering
processes can operate.
The long-held view that desert landscapes are for the most
part inactive where, yaridity mummifies a landscape by
slowing down the rates of change just as the dry sand and desiccating air preserved Egyptians and Peruvians for the museums
of the twentieth century (Walton, 1969) is now actively challenged by evidence from the more extensive field data collected
in the last couple of decades. These data have shown us that
weathering in arid regions amounts to more than just the sum of
its individual parts and it is clear to anyone who has worked in
these environments that in terms of weathering activity, arid
regions possess an inherent dynamism that make them exciting
and continually evolving landscapes.
References
Aires-Barros, L., 1977. Experiments on thermal fatigue of non-igneous rocks.
Aires-Barros, L., Grac- a, R.C., Velez, A., 1975. Dry and wet laboratory tests and
thermal fatigue of rocks. Engineering Geology 9, 249265.
Allsopp, D., Seal, K., Gaylarde, C., 2004. Introduction to Biodeterioration.
Cambridge University Press, Cambridge.
Arnold, A., Zehnder, K., 1990. Salt weathering on monuments. In: Zezza, F. (Ed.),
The Conservation of Monuments in the Mediterranean Basin. Brescia, Grafo, pp.
3158.
Balke, J., Haendel, D., Kruger, W., 1991. Contribution to the weathering-controlled
removal of chemical elements from the active debris layer of the Schirmacher
Oasis, East Antarctica. Zeitschrift fur Geologische Wissenschaften 19(2),
153158.
Banfield, J.F., Barker, W.W., Welch, S.A., Taunton, A., 1999. Biological impact on
mineral dissolution: application of the lichen model to understanding mineral
weathering in the rhizosphere. Proceedings of the National Academy of Sciences
96, 34043411.
Bierman, P., Gillespie, A., 1991. Range fires: a significant factor in exposure-age
determination and geomorphic surface evolution. Geology 19, 641644.
Blackwelder, E., 1925. Exfoliation as a phase of rock weathering. Journal of Geology
33, 793806.
Bland, W., Rolls, D., 1998. Weathering, An Introduction to the Scientific Principles.
Arnold, London.
Brunsden, D., 1980. Applicable models of long-term landform evolution. Zeitschrift
Budel, J., 1982. Climatic Geomorphology. Princetown University Press, New Jersey.
Butler, P.R., Mount, J.F., 1986. Corroded cobbles in southern Death Valley: their
relationship to honeycomb weathering and lake shorelines. Earth Surface
Camuffo, D., 1984. Condensation-evaporation cycles in pore and capillary systems
according to the Kelvin model. Water, Air and Soil Pollution 21, 151159.
Camuffo, D., 1998. Microclimate for Cultural Heritage. Developments in
Atmospheric Science 23. Elsevier, Amsterdam.
Carter, N.E.A., Viles, H.A., 2004. Lichen hotspots: raised temperature beneath
Verrucaria nigrescens on limestone. Geomorphology 62, 116.
Chen, J., Blume, H.P., Beyer, L., 2000. Weathering of rocks induced by lichen
colonisation a review. Catena 39, 121146.
Conca, J.L., Rossman, G.R., 1982. Case hardening of sandstone. Geology 10,
520523.
Conca, J.L., Rossman, G.R., 1985. Core softening in cavernous weathered tonalite.
Cooke, R.U., Brunsden, D., Doornkamp, J.C., Jones, D.K.C., 1982. Urban
Geomorphology in Drylands. Oxford University Press, Oxford.
Cooke, R.U., Doornkamp, J.C., 1990. Geomorphology in Environmental
Management. Oxford University Press, Oxford.
Cooke, R.U., Smalley, I.J., 1968. Salt weathering in deserts. Nature 220, 12261227.
Cooke, R.U., Warren, A., 1973. Geomorphology in Deserts. Batsford, London.
Cooke, R.U., Warren, A., Goudie, A.S., 1993. Desert Geomorphology. UCL Press,
London.
Cooks, J., Otto, E., 1990. The weathering effects of the lichen Lecidea Aff.
Sarcogynoides (Koerb.) on Magaliesberg Quartzite. Earth Surface Processes and
Danin, A., Garty, J., 1983. Distribution of cyanobacteria and lichens on hillsides of
the Negev highlands and their impact on biogenic weathering. Zeitschrift fur
Dorn, R., 2003. Boulder weathering and erosion associated with a wildfire, Sierra
Ancha Mountains, Arizona. Geomorphology 55, 155171.
Dragovich, D., 1967. Flaking, a weathering process operating on cavernous rock
surfaces. Bulletin of the Geological Society of America 78, 801804.
Dragovich, D., 1981. Cavern microclimates in relation to preservation of rock art.
Studies in Conservation 26, 143149.
Dragovich, D., 1993. Fire-accelerated boulder weathering in the Pilbara, western
Australia. Zeitschrift fur Geomorphologie N.F 37(3), 295307.
Dunkerly, D.L., 1979. The morphology and development of rillenkarren. Zeitschrift
Ebner, M., Miranda, T., Roth-Nebelsick, A., 2011. Efficient fog harvesting by
Stipagrostis sabulicola (Namib dune bushman grass). Journal of Arid
Environments 75, 524531.
Eppes, M.C., McFadden, L.D., Wegmann, K.W., Scuderi, L.A., 2010. Cracks in
desert pavement rocks: further insights into mechanical weathering by directional
insolation. Geomorphology 123, 97108.
Evanari, M., Shanan, L., Tadmor, N., 1982. The Negev, The Challenge of a Desert.
Harvard University Press, Cambridge MA.
Evans, I.S., 1970. Salt crystallisation and rock weathering: a review. Revue de
Geomorphologie Dynamique 19(4), 153177.
Evans, I.S., 1971. Physical weathering in hot deserts: the processes of haloclasty
and thermoclasty. Proceedings of the British Geomorphological Research Group
Conference, Durham (25th September 1971).
French, H.M., 1976. The Periglacial Environment. Longman, London.
Friedmann, E.I., Galun, M., 1974. Desert algae, lichens and fungi. In: Brown, G.W.
(Ed.), Desert Biology. Academic Press, London, vol. 2, pp. 165212.
Gomez-Heras, M., Smith, B.J., Fort, R., 2006. Surface temperature differences
between minerals in crystalline rocks: implications for granular disaggregation of
granites through thermal fatigue. Geomorphology 78, 236249.
Goudie, A.S., 1977. Sodium sulphate weathering and the disintegration of MohenjoDaro, Pakistan. Earth Surface Processes and Landforms 2, 7586.
Goudie, A.S. 1985. Salt Weathering. Oxford University School of Geography
Research Paper 8.
Goudie, A.S., 1986. Laboratory simulation of the wick effect in salt weathering of
rock. Earth Surface Processes and Landforms 11, 275285.
Goudie, A.S., 1997. Weathering processes. In: Thomas, D.S.G. (Ed.), Arid Zone
Geomorphology, Second ed. Wiley, Chichester, pp. 2539.
Goudie, A.S., Day, M.J., 1980. Disintegration of fan sediments in Death valley,
California by salt weathering. Physical Geography 1, 126137.
Goudie, A.S., Parker, A.G., 1998. Experimental simulation of rapid rock block
disintegration by sodium chloride in a foggy coastal desert. Journal of Arid
Goudie, A.S., Viles, H.A., 1997. Salt Weathering. Wiley, Chichester.
Goudie, A.S., Viles, H.A., 2000. The thermal degradation of marble. Acta
Universitatis Carolinae, Geographica XXXV, 716.
Goudie, A.S., Viles, H.A., Parker, A.G., 1997. Monitoring of rapid salt weathering in
the central Namib Desert using limestone blocks. Journal of Arid Environments
37, 581598.
Goudie, A.S., Watson, A., 1984. Rock block monitoring of rapid salt weathering in
southern Tunisia. Earth Surface Processes and Landforms 9, 9598.
Goudie, A.S., Wells, G.L., 1995. The nature, distribution and formation of pans in
arid zones. Earth Science Reviews 38, 169.
Goudie, A.S., Wright, E., Viles, H.A., 2002. The roles of salt (sodium nitrate) and
fog in weathering: a laboratory simulation of conditions in the northern Atacama
Desert, Chile. Catena 48, 255266.
Griggs, D.T., 1936. The factor of fatigue in rock exfoliation. Journal of Geology 74,
733796.
Guillitte, O., Dreesen, R., 1995. Laboratory chamber studies and petrographical
analysis as bioreceptivity assessment tools of building materials. The Science of
the Total Environment 167, 365374.
Hall, K., 1990. A note on biological weathering of nunataks of the Juneau Icefield,
Alaska. Permafrost and Periglacial Processes 1, 189196.
Hall, K., 1999. The role of thermal stress fatigue in the breakdown of rock in cold
regions. Geomorphology 31, 4763.
Hall, K., Andre, M.-F., 2001. New insights into rock weathering from high-frequency
rock temperature data: an Antarctic study of weathering by thermal stress.
Hall, K., Andre, M.-F., 2003. Rock thermal data at the grain scale: applicability to
granular disintegration in cold environments. Earth Surface Processes and
Hall, K., Arocena, J.M., Boelhouwers, J., Liping, Z., 2005. The influence of aspect
on the biological weathering of granites: observations from the Kunlun
Mountains, China. Geomorphology 67, 171188.
Hall, K., Guglielmin, M., Strini, A., 2008. Weathering of granite in Antarctica: 1.
Light penetration into rock and implications for rock weathering and endolithic
communities. Earth Surface Processes and Landforms 33, 295307.
Hall, K., Hall, A., 1991. Thermal gradients and rock weathering at low temperatures:
some simulation data. Permafrost and Periglacial Processes 2, 103112.
Hockman, A., Kessler, D.W., 1950. Thermal and moisture expansion studies of
some domestic granites. Journal of the Research of the National Bureau of
Standards, U.S. Department of Commerce Research Paper 2087(44), 395410.
Hutchens, E., 2009. Microbial selectivity on mineral surfaces: possible implications
for weathering studies. Fungal Biology Reviews 23(4), 115121.
Jenkins, K.A., Smith, B.J., 1990. Daytime rock surface rock surface temperature
variability and its implications for mechanical rock weathering: Tenerife, Canary
Islands. Catena 17, 449459.
Jennings, J.N., 1983. The disregarded karst of the arid and semi-arid domain.
Karstologia 1, 6173.
Johanessen, C.L., Feiereisen, J.J., Wells, A.N., 1982. Weathering of ocean cliffs by salt
expansion in a mid-latitude coastal environment. Shore and Beach 50, 2634.
Johnston, J.H., 1973. Salt weathering processes in the McMurdo Dry Valley region
of South Victoria Land, Antarctica. New Zealand Journal of Geology and
Kidron, G.J., Barzilay, E., Sachs, E., 2000. Microclimate control upon sand
microbiotic crusts, western Negev Desert, Israel. Geomorphology 36, 118.
Kidron, G.J., Herrnstadt, I., Barzilay, E., 2002. The role of dew as a moisture source
for sand microbiotic crusts in the Negev Desert, Israel. Journal of Arid
Koestler, R.J., Warscheid, T., Nieto, F., 1997. Biodeterioration: risk factors and their
management. In: Baer, N.S., Snethlage, R. (Eds.), Saving Our Architectural
Heritage: the Conservation of Historic Stone Structures. Wiley, Chichester, pp.
2536.
Kurtz, H.D., Netoff, D.I., 2001. Stabilisation of friable sandstone surfaces in a
desiccating, wind-abraded environment of south-central Utah by rock surface
micro-organisms. Journal of Arid Environments 48, 89100.
Laudermilk, J.D., Woodford, A.O., 1932. Concerning rillensteine. American Journal
of Science 23, 135154.
Lautridou, J.P., Ozouf, J.C., 1982. Experimental frost shattering: 15 years of
research at the Centre de Geomorphologie du C.N.R.S. Progress in Physical
Lisci, M., Michela, M., Pacini, E., 2003. Lichens and higher plants on stone: a
review. International Biodeterioration and Biodegradation 51, 117.
Mabbutt, J.A., 1977. Desert Landforms. MIT Press, Cambridge MA.
Maekawa, S., Agnew, N.H. 1996. Investigation of environmentally driven
deterioration of the Great Sphinx and concepts for protection. In: Roy, A., Smith,
P. (Eds.), Proceedings of the Copenhagen Congress, Archaeological
Conservation and its Consequences, August 1996, pp. 116120.
Magee, A.W., Bull, P.A., Goudie, A.S., 1988. Chemical textures on quartz grains. An
experimental approach using salts. Earth Surface Processes and Landforms 13,
665676.
Martini, I.P., 1978. Tafoni weathering with examples from Tuscany, Italy. Zeitschrift
McFadden, L.D., Eppes, M.C., Gillespie, A.R., Hallet, B., 2005. Physical weathering
in arid landscapes due to diurnal variation in the direction of solar heating.
McGreevy, J.P., 1981. Some perspectives on frost shattering. Progress in Physical
Geography 5, 5678.
McGreevy, J.P., 1985a. Thermal properties as controls on rock surface temperature
maxima and possible implications for rock weathering. Earth Surface Processes
225
McGreevy, J.P., 1985b. A preliminary scanning electron microscope study of

honeycomb weathering of sandstone in a coastal environment. Earth Surface
McGreevy, J.P., Smith, B.J., 1982. Salt weathering in hot deserts: observations on
the design of simulation experiments. Geografiska Annaler 64A, 161170.
McIlroy de la Rosa, J.P., Warke, P.A., Smith, B.J. 2011. Microscale biopitting by the
endolithic lichen Verrucaria baldensis and its proposed role in mesoscale
solution basin development on limestone. Earth Surface Processes and
Landforms, published on-line: 10.1002/esp.2244.
McKay, C.P., Molaro, J.L., Marinova, M.M., 2009. High-frequency rock temperature
data from hyper-arid desert environments in the Atacama and the Antarctic Dry
Valleys and implications for rock weathering. Geomorphology 110, 182187.
Molaro, J.L., McKay, C.P., 2010. Processes controlling rapid temperature variations
on rock surfaces. Earth Surface Processes and Landforms 35, 501507.
Moses, C.A., Smith, B.J., 1993. A note on the role of the lichen Collema auriforma
in solution basin development on a carboniferous limestone substrate. Earth
Mottershead, D.N., 1989. Rates and patterns of bedrock denudation by coastal salt
spray weathering: a seven year record. Earth Surface Processes and Landforms
14, 383398.
Mustoe, G.E., 1982. The origin of honeycomb weathering. Geological Society of
Mustoe, G.E., 1983. Cavernous weathering in the Capitol reef Desert, Utah. Earth
Nash, D., 1997. Groundwater as a geomorphological agent in drylands. In: Thomas,
D.S.G. (Ed.), Arid Zone Geomorphology, Second ed. Wiley, Chichester, pp. 319348.
Obika, B., Freer-Hewish, R.J., Fookes, P.G., 1989. Soluble salt damage to thin
bituminous road and runway surfaces. Quarterly Journal of Engineering Geology
22, 5973.
Ollier, C.D., 1963. Insolation weathering: examples from Central Australia. American
Ollier, C.D., 1965. Dirt cracking a type of insolation weathering. Australian
Ollier, C.D., 1984. Weathering, Second ed. Oliver & Boyd, Edinburgh.
Paradise, T.R., 1997. Disparate sandstone weathering beneath lichens, Red
Mountain, Arizona. Geografiska Annaler 79(3), 177184.
Peel, R.F., 1974. Insolation weathering: some measurements of diurnal temperature
changes in exposed rocks in the Tibesti region, Central Sahara. Zeitschrift fur
Geomorphologie Supplement Band 21, 1928.
Peltier, L., 1950. The geographic cycle in periglacial regions as it is related to
climatic geomorphology. Annals of the Association of American Geographers 49,
214236.
67, 255272.
Pietruszka, R.D., Seely, M.K., 1985. Predictability of two moisture sources in the
Namib desert. South Africa Journal of Science 81, 682685.
Price, C.A., 1993. Preventive conservation of salt-contaminated masonry in the
Wakefield Tower, HM Tower of London. Institute of Archaeology Bulletin 30,
121133.
Richter, D., Simmons, G., 1974. Thermal expansion behaviour of igneous rocks.
International Journal of Mechanics and Mineral Science 11, 403411.
Rodriguez-Navarro, C., Doehne, E., 1999. Salt weathering: influence of evaporation
rate, supersaturation and crystallisation pattern. Earth Surface Processes and
111, 12501255.
Rogner, K., 1987. Temperature measurements of rock surfaces in hot deserts
(Negev, Israel). In: Gardiner, V. (Ed.), International Geomorphology. Wiley,
Sancho, C., Benito, G., 1990. Factors controlling tafoni weathering in the Ebro
Basin (N.E. Spain). Zeitschrift fur Geomorphologie N.F 34(2), 165177.
Schwartzman, D., Aghamiri, R., Bailey, S., 1997. Lichen weathering rates from miniwatershed studies. Abstracts with Programs, Geological Society of America 29,
361 pp.
Seaward, M.R.D., 1997. Major impacts made by lichens in biodeterioration
processes. International Biodeterioration and Biodegradation 40(2), 269273.
Seaward, M.R.D., 2004. Lichens as subversive agents of biodeterioration. In: St.
Clair, L., Seaward, M. (Eds.), Biodeterioration of Stone Surfaces. Kluwer
Academic Publishers, The Netherlands, pp. 918.
226
Selwitz, C., 1990. Deterioration of the Great Sphinx: an assessment of the literature.
Antiquity 64, 853859.
Smith, B.J., 1977. Rock temperature measurements from the northwest Sahara and
their implications for rock weathering. Catena 41, 4163.
Smith, B.J., 1987. An integrated approach to the weathering of limestone in an arid
area and its role in landscape evolution: a case study from southeast Morocco.
In: Gardiner, V. (Ed.), International Geomorphology. Wiley, Chichester, pp. 637657.
Smith, B.J., 1988. Weathering of superficial limestone debris in a hot desert
environment. Geomorphology 1, 355367.
Smith, B.J., 1996. Scale problems in the interpretation of urban stone decay.
In: Smith, B.J., Warke, P.A. (Eds.), Processes of Urban Stone Decay. Donhead,
Smith, B.J. 2009. Weathering processes and forms. In: Parsons, A.J., Abrahams,
A.D. (Eds.), Geomorphology of Desert Environments, Second ed., Springer
Science and Business, New York, pp. 69100.
Smith, B.J., McAlister, J.J., 1986. Observations on the occurrence and origins of
salt weathering phenomena near Lake Magadi, southern Kenya. Zeitschrift fur
Smith, B.J., Srinivasan, S., Gomez-Heras, M., Basheer, P.A.M., Viles, H.A., 2011.
Near-surface temperature cycling of stone and its implications for scales of
surface deterioration. Geomorphology 130, 7682.
Smith, B.J., Warke, P.A., 1997. Controls and uncertainties in the weathering
environment. In: Thomas, D.S.G. (Ed.), Arid Zone Geomorphology, Second ed.
Wiley, Chichester, pp. 4154.
Smith, B.J., Warke, P.A., McGreevy, J.P., Kane, H.L., 2005. Salt weathering
simulations under hot desert conditions: agents of enlightenment or
perpetuations of preconceptions? Geomorphology 67, 211227.
Smith, B.J., Warke, P.A., Moses, C.A., 2000. Limestone weathering in contemporary
arid environments: a case study from southern Tunisia. Earth Surface Processes
Smith, B.J., Whalley, W.B., Magee, R., 1992. Assessment of building stone decay: a
geomorphological approach. In: Webster, R.G.M. (Ed.), Stone Cleaning and the
Nature, Soiling and Decay Mechanisms of Stone. Donhead, Shaftesbury, pp.
249257.
Sperling, C.H.B., Cooke, R.U., 1985. Laboratory simulation of rock weathering by
salt crystallisation and hydration processes in hot arid environments. Earth
Strakhov, N.M., 1967. Principles of Lithogenesis (Translation: J.P. Fitzsimmons).
Oliver & Boyd, Edinburgh.
Sweeting, M.M., Lancaster, N., 1982. Solutional and wind erosion forms on
limestone in the central Namib Desert. Zeitschrift fur Geomorphologie 26,
197207.
Thomas, D.S.G., 1997. Arid environments: their nature and extent. In: Thomas,
D.S.G. (Ed.), Arid Zone Weathering, Second ed. Wiley, Chichester, pp. 312.
Turkington, A.V., 1998. Cavernous weathering in sandstones: lessons to be learned
from natural exposure. Quarterly Journal of the Geological Society 31, 375383.
and innovation. Geomorphology 67(12), 229253.
Twidale, R.C., 2007. Bornhardts and associated fracture patterns. Revista de la
Asociacion Geologica Argentina 62(1), 139153.
Viles, H.A., 1995. Ecological perspectives on rock surface weathering: towards a
conceptual model. Geomorphology 13, 2135.
Viles, H.A., 2005. Microclimate and weathering in the central Namib Desert,
Viles, H.A., Goudie, A.S., 2003. Interannual, decadal and multidecadal scale climatic
variability and geomorphology. Earth Science Reviews 61, 105131.
Viles, H.A., Goudie, A.S., 2004. Biofilms and case hardening on sandstones from
Al-Quwayra, Jordan. Earth Surface Processes and Landforms 29, 14731485.
Viles, H.A., Goudie, A.S., 2007. Rapid salt weathering in the coastal Namib desert:
implications for landscape development. Geomorphology 85, 4962.
Walder, J.S., Hallet, B., 1985. A theoretical model of the fracture of rock during
freezing. Geological Society of America Bulletin 96, 336346.
Walton, K., 1969. The Arid Zones. Aldine Transaction, USA.
Warke, P.A., 2000. Micro-environmental conditions and rock weathering in hot arid
regions. Zeitschrift fur Geomorphologie 120, 8395.
Warke, P.A., 2007. Complex weathering in drylands: implications for rock
breakdown and sediment release. Geomorphology 85, 3048.
Warke, P.A., Smith, B.J., 1994. Short-term rock temperature fluctuations under
simulated hot desert conditions: some preliminary data. In: Robinson, D.,
Williams, R.B.G. (Eds.), Rock Weathering and Landform Evolution. Wiley,
Warke, P.A., Smith, B.J., 1998. Effects of direct and indirect heating on the validity
of rock weathering simulation studies and durability tests. Geomorphology
22(34), 347357.
Warke, P.A., Smith, B.J., 2000. Salt distribution in clay-rich weathered sandstone.
Whalley, W.B., Warke, P.A., 2005. Weathering. In: Selley, R.C., Cocks, L.R.M.,
Plimer, I.R. (Eds.), Encyclopaedia of Geology. Elsevier, Oxford, Volume 5, pp.
581590.
Whitaker, C.R., 1974. Split boulder. Australian Geographer 12, 562563.
Winkler, E.M., 1973. Stone: Properties, Durability in Mans Environment. SpringerVerlag, New York.
Winkler, E.M., Singer, P.C., 1972. Crystallisation pressure of salts in stone and
concrete. Geological Society of America Bulletin 83, 35093514.
Wollmann, G., Voigt, W., 2008. Solubility of Gypsum in MSO4 Solutions (M Mg,
Mn, Co, Ni, Cu, Zn) at 298.15 K and 313.15 K. Journal of Chemical
Engineering Data 53, 13751380.
Yaalon, D.H., 1970. Parallel stone cracking, a weathering process on desert
surfaces. Geological Institute of Bucharest Technical Economic Bulletin
(Pedology) 18, 107111.
Yatsu, E., 1988. The Nature of Weathering. Sozosha, Tokyo.
Yong, C., Wang, C.Y., 1980. Thermally induced acoustic emission in Westerly
granite. Geophysical Research Letters 7, 10891092.
Geology 15, 962966.
Young, R.W., 1988. Quartz etching and sandstone karst: examples from the East
Kimberleys, northwestern Australia. Zeitschrift fur Geomorphologie N.F 32(4),
409423.
Zangvil, A., 1996. Six years of dew observations in the Negev Desert, Israel.
Journal of Arid Environment 32, 361371.
Zezza, F., Macr, F., 1995. Marine aerosol and stone decay. The Science of the Total
227
Biographical Sketch
Patricia Warke was appointed as a lecturer in 1999 in the School of Geography, Archaeology, and Palaeoecology at
Queens University Belfast. A geomorphologist by training, her PhD research involved investigation of rock
weathering in contemporary hot arid environments with particular emphasis on the role of inheritance effects in
development of present-day weathering phenomena. Since then she has worked in the deserts of southwestern
USA, North Africa, and the Middle East, and while weathering in the natural environment underpins her work she
has also been involved in the investigation of the general decay dynamics of stone in built structures, both
archaeological and historical. This latter research strand has stimulated a focus on development of condition
classification systems for buildings and ancient monuments based on the Staging System approach used in
medicine. As a member of the Weathering Research Group at Queens University, she is involved in a wide variety
of laboratory and field-based weathering research projects with a current focus on strengthening links between
laboratory and field studies through the up-scaling of small-scale data to help explain the development of mesoscale weathering forms.
4.13 Coastal Weathering

D Mottershead, University of Portsmouth, Portsmouth, UK
4.13.1
4.13.1.1
4.13.1.2
4.13.2
4.13.2.1
4.13.2.2
4.13.2.3
4.13.2.4
4.13.3
4.13.3.1
4.13.3.1.1
4.13.3.1.2
4.13.3.1.3
4.13.3.2
4.13.3.2.1
4.13.3.3
4.13.3.3.1
4.13.3.3.2
4.13.3.4
4.13.4
4.13.4.1
4.13.4.2
4.13.4.3
4.13.5
4.13.5.1
4.13.5.2
4.13.5.3
References
Introduction
The Coastal Weathering Environment: Definition
Conditions in the Coastal Weathering Environment
Marine Salt in the Coastal Environment
The Nature of Marine Salts
Salts in the Atmosphere: Uptake and Transfer
Onshore Salt Deposition
Salts in Coastal Rocks
Weathering Processes Facilitated by the Coastal Environment
Mechanical Weathering Processes
Thermal variation
Slaking
Phase changes
Chemical Weathering Processes
Dissolution
Biological Weathering Processes
Macrobiological processes
Microbiological processes
Complexity and Rates of Rock Weathering by Salts
Coastal Landforms Associated with Weathering
Shore Platforms
Tafoni (Singular Tafone)
Honeycomb Weathering
Conclusion
Environmental Factors
Lithological Factors
Process Factors
228
228
229
230
230
230
231
232
233
233
233
233
234
235
235
235
235
235
236
237
238
238
239
241
241
241
241
242
Abstract
Coastal weathering operates on rocky coasts. Marine salts ejected from ocean and coastal waves to the atmosphere are
transferred to the coastal zone where they impregnate surface geomaterials. Coastal rocks are subject to many weathering
processes, some of which are enhanced in this environment, particularly those dependent on salt, water, and direct
exposure to atmospheric temperature changes. Weathering contributes to the development of shore platforms, tafoni and
honeycombed rock. It is carried out by a combination of individual mechanisms that may be difficult to identify, and leads
to very high rates of rock erosion around high tide level (HTL).
4.13.1
Introduction
4.13.1.1
The length of the coastline on planet Earth amounts to approximately 440 000 km (Pethick, 1984). Throughout their
length, these coastlines display many variations in landforming materials, climatic regimes, oceanic regimes, and
consequent geomorphic process regimes.
Mottershead, D., 2013. Coastal weathering. In: Shroder, J. (Editor in Chief),

228
The Coastal Weathering Environment: Definition
Geomorphologically, the coast is a high-energy environment

in which erosion processes frequently proceed very actively.
Weathering, however, operates but slowly in the spectrum of
geomorphological processes, and in order for it to occur,
geomaterials need to be capable of withstanding erosion processes for a sufficient time. In effect, therefore, the
active coastal weathering environment is limited to those
coasts formed by consolidated rock, where erosion is weathering-limited (Gilbert, 1877). It is suggested (Trenhaile,
1987; Naylor et al., 2010) that up to 80% of the worlds
oceanic coastlines may be deemed rocky, and may include
http://dx.doi.org/?10.1016/B978-0-12-374739-6.00064-6?
Coastal Weathering
coastal buildings constructed of stone that, indeed, have facilitated significant contributions to the study of coastal
weathering.
The spatial extent of the coastal weathering environment
comprises both longshore and orthogonal components. The
shoreline itself defines the former, but the lateral extent ranges
from low tide level at its seaward limit to a landward
boundary that identifies a meaningful limit to the coastal effect. Marine salt, a signature influence in the coastal weathering environment, may be transported by atmospheric
processes many hundreds of kilometers inland in declining
concentrations (Eriksson, 1960), to create an imprecisely defined coastal hinterland of gradually diminishing marine influence. Field studies have shown that salt deposition, at high
levels within the coastal environment, is at a maximum on
surfaces opposed to the maritime winds delivering the salt
onshore. A reasonable approach, therefore, would be to define
the coastal weathering environment to include coastal slopes
facing the sea, directly opposing onshore winds and in position to intercept incoming aerosol salts. This definition would
include weathering on sea cliffs, coastal slopes bearing rock
outcrops, and structures built with stone. This may be more
formally stated as follows: the coastal weathering environment
exists along rocky coasts between low water mark and the crest
of any coastal slope facing the sea.
The coastal zone thus defined may range in width from
perhaps 20 m in the case of a microtidal cliffed coast to some
hundreds of meters for a coast with an extensive backing slope.
It encompasses the intertidal zone in which weathering takes
place during those phases of the tidal cycle when rock happens
to be exposed to the atmosphere. It also includes the supratidal
zone, above high water mark, where the shore zone is exposed
to wave splash as seawater is thrown up onshore by breaking
waves. Further onshore sea spray, a wet aerosol, may deposit salt
on land surfaces, whereas deposition of sea salt in dry aerosol
form may occur under suitably windy conditions. It is clear that
any shore-normal zonation defined by such conditions is
highly variable through time, according to the independently
varying fluctuations of tides, and weather conditions.
4.13.1.2
Conditions in the Coastal Weathering

Environment
Conditions in the coastal weathering environment are a response to environmental factors that include physical factors
of the sea and physical factors of the land (Davies, 1972).
These include oceanic factors such as wave and tidal regimes,
climatic factors, local relief, and geomorphological regimes.
These various factors, which can be used to characterize a
coastline at the macroscale, and the permutations of interplay
between them, will determine the particular nature of the
weathering environment at an individual site.
Coastal weathering can be viewed as a system (Figure 1),
an organized framework in which materials and energy are
brought together, interacting through a range of mechanical
and chemical processes through which the landscape is
modified by the alteration and movement of geomaterials
(White et al., 1992). An alternative approach is to consider the
individual weathering site, where the weathering processes
themselves operate, and the various controls come into play.
Here, we are dealing with processes operating at the scale of a
pore or a micropore in the case of mechanical processes or the
molecule or ion in the case of chemical processes. The conceptual framework of Pope et al. (1995) offers a relevant approach to conditions and controls at the weathering boundary
layer, where mutually incompatible materials may come
Atmosphere
Wind
Mist
Solar
radiation
Rainfall
Wind
Seeds
and spores
Onshore
wind
Salt
Microbes
Biosphere
Waves
Mist
Splash
Spray
Weathering processes:
Salt weathering
Crystallization pressure
Thermoclasty
Thermal shock
Slaking
Frost weathering
Hydration
Dissolution
Macrobiological weathering
Microbial weathering
Windblow
Rainwash
Swash
Detachment
Tides
Gravitational
energy
229
Rocks
minerals
Gravitational
energy
Geosphere
Figure 1 The coastal weathering system showing the relationships between energy and material flows and weathering processes.
230
Coastal Weathering
together to create a weathering reaction. In this framework,

weathering rate (Wr) is expressed as a function of 16 (or more)
variables that determine the nature of the action at the actual
locus of weathering.
The distinctive characteristic of the coastal weathering environment is that bare rock is directly exposed at the land
surface to weathering conditions and weathering agents of both
atmospheric and marine origin. In response to solar radiation,
coastal rocks experience a wide range of temperatures, including
subzero temperatures, and rapid temperature changes. They
also receive direct inputs of rainwater and especially saline
seawater, with consequent wetting and drying according to atmospheric temperature, windspeed, and humidity. The most
distinctive characteristic, and of particular significance, is the
presence of marine salt, which differentiates coastal weathering
environments from others with similar climatic conditions. The
biotic influence in this weathering environment is provided
largely by lichens, algae, and microbes.
The combination of these factors creates a weathering environment whose individual nature renders it worthy of separate definition and study. What makes the coast unique is the
balance in significance of the rock weathering controls in this
environment, by which particular weathering mechanisms are
either enhanced or suppressed in comparison with otherwise
similar noncoastal environments.
4.13.2
4.13.2.1
Marine Salt in the Coastal Environment

The Nature of Marine Salts
Normal seawater is a dilute solution of a large number of

different salts. It is a six-component NaKMgSO4ClH2O
system with a density at 25 1C of 1.03 g cm1 and a salinity of
35 7 2%. It contains a substantial range of dissolved ions
(Table 1), with more than 40 trace elements at an abundance
of o1 ppm.
The more abundant cations combine to form salts, which
crystallize out on evaporation or freezing (Table 2), creating
potentially powerful agents of weathering. The most abundant
elements both in terms of mass and salinity are sodium and
chloride, which combine on evaporation to form halite,
comprising B78% of the total crystallate. Significant amounts
of magnesium, potassium, and calcium salts also form a range
of minerals in various hydration states.
Table 1 Major ions (40.01% of total salinity) in seawater
Ion
% mass
Na
Mg2
Ca2
K
Sr2
30.62
3.68
1.18
1.10
0.02
10.56
1.27
0.40
0.38
0.01
Cl
SO4
HCO3
Br
H2BO3
55.07
7.72
0.40
0.19
0.01
18.90
2.65
0.14
0.06
0.03
Source: Reproduced with permission from Anikouchine, W.A., Sternberg, R.W., 1973.
The Worlds Ocean; An Introduction to Oceanography. Prentice Hall, New Jersey, 339 pp.
Episodic wetting by and drying of seawater are common

events in coastal rocks. The behavior of seawater brine during
evaporation is therefore crucial to an understanding of the
consequences of these processes for rock weathering. There is a
consistent order of crystallization of salts during the progressive evaporation of seawater as the removal of water causes
the density and salinity of the remaining brine to increase
(Warren, 1999; Eugster et al., 1980; Harvie et al., 1980;
McCaffrey et al., 1987). Minerals then begin to form by crystallization, selectively removing elements from the solution.
The first to crystallize is CaCO3, usually as aragonite, followed
by the onset of gypsum/anhydrite at 5% evaporation, halite at
11%, magnesium sulfate (epsomite/kieserite) at B70%, and
potassium salts (sylvite, carnallite, kainite, polyhalite) at
B80%. This selective removal of elements alters the chemical
composition of the brine during the evaporation process.
Thus, the brine is always dominated by chloride, initially
along with sodium up to ~40% evaporation, and at higher
concentrations with sulfate and finally magnesium. During
the final stages of evaporation, the brine attains a density of
41.2, becoming a very sticky fluid (Table 3).
Several of the salts crystallizing out of seawater exist in both
anhydrous and hydrated forms. These are sodium chloride,
calcium sulfate, magnesium sulfate, magnesium chloride, and
polyhalite. Sodium chloride hydrates to NaCl 2H2O at temperatures of o0.1 1C, a factor of potential significance in
environments where subzero temperatures occur. Calcium
sulfate exists in both anhydrous (anhydrite) and dihydrate
(gypsum) forms, magnesium chloride up to the hexahydrate
form (bischofite), and magnesium sulfate up to heptahydrate
(epsomite). Together, these hydrating salts comprise B22% of
the total crystallate mass. Complex double and triple chlorides
and sulfates formed during evaporation of supersaline brine
may also form hydrates (kainite, polyhalite, and carnallite).
Many of these phase transitions occur within the normal range
of climatic conditions characteristic of coastal environments.
The significance of these processes in the coastal weathering environment lies in their effects within rock pores and
fissures, as wetting and drying of exposed rocks takes place.
4.13.2.2
Salts in the Atmosphere: Uptake and Transfer
The transfer of marine salt from the ocean to land surfaces in

the coastal zone comprises three stages, namely release from
ocean into the atmosphere, lateral transfer through the atmosphere, and deposition onto a land surface.
Table 2 Major salts crystallizing out of seawater during evaporation
Salt
Mineral/salt
Yield g kg1 seawater
NaCl
MgCl2
MgSO4
CaSO4
K2SO4
CaCO3
MgBr2
Total
Halite
Magnesium chloride
Epsomite, kieserite
Gypsum, anhydrite
Polyhalite
Calcite, aragonite
Magnesium bromide
27.21
3.81
1.66
1.26
0.86
0.12
0.08
35.00
Source: After Kuenen, P.H., 1950. Marine Geology. Wiley, New York, 568 pp.
Coastal Weathering
231
Table 3 Mineral crystallization and changing brine conditions during the evaporation of seawater
Brine stage
Mineral precipitate
Salinity (% )
Degree of concentration
Water loss (%)
Density (g cm3)
Normal marine
Mesohaline
Penesaline
Penesaline
Supersaline
Supersaline
Seawater
Calcite/aragonite
Gypsum/anhydrite
Gypsum / halite
Halite
KMg salts
35
35140
140250
250350
4350
Variable
1x
14x
47x
711x
411x
460x
0
075
7585
8590
490
B99
1.04
1.041.10
1.1001.126
1.1261.214
41.214
41.214
Source: After Warren, J., 1999. Evaporites Their Evolution and Economics. Blackwell, Oxford, 438 pp.
The major source of sea salt transfer to the atmosphere is

bubbles from breaking waves (Berner and Berner, 1996; Beer,
1997; Wells, 1997; Bigg, 2003). This occurs on the open
ocean, where whitecaps are an abundant source, forming
across wide areas under storm conditions, and also in the
coastal zone where waves break on entering shallow water.
Waves begin to break and form whitecaps at windspeeds of
B34 ms1 (Blanchard and Woodcock, 1957; ODowd and
de Leeuw, 2007), equivalent to Beaufort Wind Scale Force 3,
when wind shear causes the wave to topple forward, the water
mass plunging back into the ocean surface. The plunging mass
contains bubbles of atmospheric gases, whose lower density
lifts them back toward the water surface, where they burst,
ejecting water droplets into the atmosphere. At a windspeed of
16 ms1, Mason (2001) stated that breaking waves occupy
some 5% of the ocean surface, ejecting air bubbles in foam
patches at a rate of 40 cm2 s1. The larger particles tend to
fall back into the sea, whereas the smaller ones tend to
evaporate, releasing salt particles into the lower atmosphere. A
1 mm radius bubble is likely to produce droplets which, on
evaporation, will yield up to 0.2 mg of salt, equivalent to dry
particles in the range 0.825 mm (Brimblecombe, 1986).
Windspeed plays a crucial role in atmospheric salt production from the ocean surface. A positive correlation has
been shown between windspeed and salt concentration, with
winds of 4 ms1 generating 7 mg m3 of salt, increasing to
20 mg m3 at a windspeed of 9 ms1 (Woodcock, 1953), and
up to 1000 mg m3 at windspeeds of 1520 ms1 (Lewis and
Schwartz, 2004). Turbulence and convection then lift salt
particles further into the atmosphere. The smaller particles
remain aloft for longer, and are available for a sufficient length
of time to be transported for distances of many hundreds of
kilometers, carrying them across continental coasts and beyond into the interior.
Davies (1972) produced maps, based on data from the
US Navy Marine Climatic Atlas of the World, showing the
global distribution of the frequency of oceanic winds of
Beaufort Scale 4 and above, including zones in which they
intercept continental coastlines. The distributions of these
windspeeds would appear to be indicative of those zones in
which, globally, coastal salt deposition attains its highest
values. Erickson and Duce (1988) estimate that global deposition of marine salt, over both ocean and land, amounts to
15004500 Tg a1.
Marine salt is delivered in both wet and dry form to the
coastal zone by sedimentation and rainout processes (Cadle,
1973; Hidy, 1973). This atmospheric fallout will occur ubiquitously on land surfaces and be intercepted (scavenged) by
vegetation above high water mark in the supratidal zone and

episodically on surfaces within the intertidal zone.
In the intertidal zone, the daily tidal cycle will cause the
saline seawater to wash directly over subjacent rocks, where
the tidal head will tend to force saline water into any bedrock
submerged during the tidal cycle. Above high water mark,
seawater in the form of fluid or foam may be splashed up onto
rocks, depending on the energy of the breaking waves.
4.13.2.3
Onshore Salt Deposition
Estimates have been made of the rates of salt deposition in the

coastal zone at national and local scales. Because of the constancy of composition of seawater, an individual ion can be
used as a measure of total marine salt. Erickson and Duce
(1988) presented a map of the US of chloride concentrations in
atmospheric precipitation, showing the highest values in a
narrow coastal zone, with 48 mg l1 Cl in rainfall along the
North Atlantic coast, and hotspots of 420 mg l1 in the Pacific
northwest and the west of the Gulf of Mexico, falling rapidly
inland to values of o1 mg l1. Hingston and Gailitis (1976)
presented a map of Cl deposition across Western Australia
showing declining values inland from the coast (Figure 2).
At the local scale studies, in particular by plant ecologists,
have revealed rates of coastal salt deposition and instructive
patterns, both spatial and temporal, in its concentration.
Randall (1970) measured the deposition of sea salt on the
coast of Barbados, exposed to north east trade winds through
most of the year with a mean velocity of 16.8 km h1. Wave
amplitudes on the exposed east coast are commonly four to
eight times greater than the sheltered west coast. Aerial salt
concentration was found to diminish by two orders of magnitude over the first 150 m inland (Figure 3). The accumulated seasalt concentrations in soil are much higher on the
exposed east coast than the west (reaching values of 440 mg
salt per gram of soil). Salt deposition is also increased by
ground surface irregularities; higher deposition occurs on both
windward and leeward sides of upstanding terrain obstacles.
Randall (1974) sampled airborne salt on a stormy Atlantic
coast environment on the island of North Uist, UK. At this
location, 44% of days experience gale force winds of 467
km h1. Sampling during summer only, values were obtained
of up to 670 mg m2 day1 of aerial seasalt equivalent. A linear
transect inland from the Atlantic coast shows a reduction in
salt collection of 470% within 100 m of the coast. Randall
demonstrates a threshold wind velocity of deposition of
15 km h1, with increasing salt concentrations at higher
232
Coastal Weathering
200 000
100 000
Intermittent lake
20
Cl precipitated 1973 (kg
ha1)
2
3
20
Salt in Micromhos per cm
10
Medium wind
High Sea
5
10 000
6
4
7
8
9
5
6
10
7
8
1000
Medium wind
Medium Sea
50
150
50
Geraldton
100
0
10
100
Perth
20
50
100
0
Albany
500 km
Figure 2 Annual chloride (a surrogate for sea salt) precipitation on

the western coast of Australia. Deposition rates are higher with the
higher rainfall of the southwest. All decline markedly inland but still
maintain significant levels at over 200 km from the coast. Redrawn
with permission from Hingston, F.J., Gailitis, V., 1976. The
geographic variation of salt precipitated over Western Australia.
Australian Journal of Soil Research 14, 319335.
windspeeds. There is a pattern of greater salt concentration on

slopes facing the coast and a reduction on leeward slopes.
Malloch (1972) presented data on cation deposition from
marine aerosol on an exposed coastal site at Lizard, Cornwall,
UK. He estimated total annual salt production ranging from a
maximum value of 18.3 kg ha1 some 20 m from the coastline, falling rapidly over 100 m to 3.14 kg ha1. He observed
that salt deposition is associated with strong onshore winds,
noting a sharp increase in deposition at a threshold windspeed velocity of 14 ms1.
Vale (1991) monitored Cl deposition on the island of
Flatholm in the Bristol Channel, UK, calculating the annual
total of salt deposition at 170 kg ha1. The author suggested
that the total catch may be underestimated by 30%, implying
a more realistic estimate of 4200 kg ha1. Some 70% of
deposition took place within a single sampling interval,
100
200
Distance from Sea (m)

Figure 3 Ground surface salt deposition in the immediate coastal
zone, showing the influence of wave amplitude (sea state) and
coastal proximity. Under comparable wind conditions, higher seas
yield more salt deposition. The rates of deposition decline rapidly
inland over the first 150 m. Redrawn with permission from Randall,
R.E., 1970. Salt measurement on the coast of Barbados, West Indies.
Oikos 21(1), 6570.
coincident with strong Atlantic onshore winds. There is also a

strong correlation between rainfall and element deposition,
suggesting that rainout is a major deposition mechanism, and
also indicating strongly episodic deposition.
Mustoe (1982) observed soluble cations of marine aerosol
origin in weathering debris from the coast near Bellingham,
WA, USA, and showed strong inverse correlations of concentration with elevation above sea level over a range of
0.52.4 m asl, showing similar abundance in contrasting
periods of both storm and calm conditions. The author inferred that the salt input is therefore derived from the more
continuous wave splash rather than airborne aerosol.
It is evident from this brief review that the rates of coastal
salt deposition are controlled by a number of factors. Temporally, episodes of high winds, high surf, and precipitation
deliver salt to coastal sites. Spatially, the magnitude of salt
deposition diminishes with distance from and elevation above
the shoreline, with aspect and shelter varying locally to create
a mosaic of salt capture at the land surface.
4.13.2.4
Salts in Coastal Rocks
Although marine salts within coastal rock samples have frequently been observed microscopically, the presence of marine
Coastal Weathering
salts in coastal rocks has been assessed quantitatively in only a

limited number of geomorphological studies. Laboratory experiments by Tingstad (2008) showed seawater to penetrate a hard
limestone to a depth of 0.3 mm in 28 days whereas in Davisons
(1986) work, seawater penetrated mica schist samples to depths
of between 5 and 18 mm in only 50 diurnal cycles, showing that
salt penetration into rock may take place very rapidly.
At Weston super Mare, UK, sea salt concentration in a
sandstone capping of a seawall, mostly adjacent to HTL, shows
values of 1.737.72 mg g1 of rock (Mottershead, 1994). The
role of wind is again emphasized by the fact that on a wall of
circular planform, the maximum salt concentration/weathering is aligned with the azimuth of the prevailing wind,
rather than proximity to the shoreline.
Mottershead and Pye (1994) presented data for chloride
content from a core taken from a greenschist outcrop on a
coastal slope at Prawle, UK, at approximately 30 m above sea
level. The maximum value of seasalt content within the rock
(from Cl determinations) was calculated at 1.47 mg g1 of
rock. Seasalt content declined inward to a value of 0.5 mg g1
at the base of the core, implying a total penetration of approximately 0.2 m, and quite possibly representing a long-term
equilibrium. It is not known how representative these limited
observations are or how the salt content of coastal rocks varies
in different lithologies or with varying degrees of exposure, and
this appears to represent a significant gap in current knowledge.
4.13.3
Weathering Processes Facilitated by the

Coastal Environment
The most distinctive conditions of the coastal weathering environment arise from the facts that rocks experience direct
inputs of marine salt and are directly exposed the atmosphere.
Lacking the moderating effects of an overburden or soil cover,
they are exposed to ambient atmospheric conditions and
processes, with widely varying inputs of solar radiation and
precipitation, and periods of wetting and drying. Weathering
processes likely to be facilitated in this environment will now
be reviewed in respect of their potential contribution to the
weathering of coastal rocks.
4.13.3.1
Mechanical Weathering Processes
Several mechanical weathering mechanisms involve the

growth of a solid substance within the confining space of a
pore, whereby a force is exerted on the confining wall, especially if the pore is completely filled. This force creates a tensile
stress within the rock, which, if exceeding rock tensile
strength, will create fractures (microcracks) in which the stress
concentrates at the tip. Further stress causes propagation of the
crack at the tip, leading to splitting and disintegration of the
rock. Pore expansion also occurs during low-level stressing of
rock by slaking, a cumulative effect that leads toward rock
fatigue, as the rock becomes progressively weakened by pore
expansion and contraction.
4.13.3.1.1
Thermal variation
Rocks exposed directly to solar radiation commonly heat up to

temperatures well in excess of the ambient air temperature
(Peel, 1974; Sumner et al., 2007), extending the daily
233
temperature range and increasing temporal thermal variation

in rocks (see Chapters 4.12 and 4.15). According to their
properties of albedo, thermal expansion, and transmissivity,
minerals and rocks will expand and contract differentially,
setting up internal stresses. Thermal stresses are also created by
the temperature gradient between the exposed surface and the
rock mass interior. All such processes have the capacity to
create mechanical stresses within the rock during temperature
change (Yatsu, 1988). The cumulative effect of this repeated
low-level stressing leads progressively to rock fatigue, as the
rock becomes weakened by continuing expansion and contraction of fissures and pores.
Salt crystal expansion/contraction: This effect has been postulated in respect of halite, the dominant coastal salt. Halite
has a substantially higher coefficient of thermal expansion
between 0 and 60 1C than quartz, calcite, and granites
(Cooke and Smalley, 1968). Differential thermal expansion
of halite crystals in pores within less expansive minerals will
create tensile stresses. The ubiquity of halite at the coast
means that this process is likely to be effective, especially on
coasts with more extreme temperature variations.
Thermal shock: Recent insights into this phenomenon have
been considerably aided by the acquisition of highfrequency rock temperature data (Hall, 1999; Hall and Andre,
2001, 2003; McKay et al., 2009; Molaro and McKay, 2010).
Incoming solar radiation causes the rock surface to heat up
to temperatures greater than the adjacent atmosphere. When
the radiation input ceases, outward radiation occurs, cooling
the rock surface. The cessation may be quite abrupt with, for
example, the passage of a cloud or a shadow extending over
the rock surface. If a substantial temperature difference occurs
between rock and the adjacent air, then cooling is very rapid
indeed, with consequent rapid contraction of the surface rock.
Rates of thermal change (dT/dt) of 42 1C min1 require rocks
to adjust more rapidly than their capacity for plastic deformation allows, thereby causing thermal shock. Such values of
dT/dt may occur at any temperature, but may be particularly
frequent during rapid cooling in arctic environments under
clear sky conditions and low ambient temperatures.
The thermal conductivity of rock means that it is unlikely
that such brief effects are propagated more than a few millimeters beneath the rock surface, but they are nevertheless
capable of affecting the rock surface over time at this scale,
encouraging granular disintegration and flaking of the surface
skin to a depth of millimeters (Hall et al., 2008). Although not
yet apparently investigated in specifically coastal rocks, under
suitable climatic conditions, they are exposed to this
phenomenon.
4.13.3.1.2
Slaking
Slaking is the process of alternate wetting and drying. Rocks,

especially those containing clays, tend to swell on wetting,
with subsequent contraction on drying. When water enters the
pores of a rock, the rock dilates, creating tensile stresses and
generating tension cracks (Yatsu, 1988). The consequent development of microcracks and their propagation further increase porosity.
Wetting/drying sequences necessarily occur in the coastal
zone, notably on shore platforms during tidal cycles, but also
234
Coastal Weathering
above high water mark through episodic wetting by wave

splash during heavy seas and by rainfall. Such repeated expansion/contraction sequences are likely to impose stresses on
coastal rocks. Experiments indicate that slaking is capable of
producing rock breakdown, and also show that saltwater may
be more or less effective than pure water in slaking, depending
on the rock type.
Laboratory studies of slaking have been carried out by Hall
and Hall (1996) and Kanyaya and Trenhaile (2005), using
deionized water to eliminate the effect of salt. The former
showed slaking to cause mass loss of the rock, and that water
uptake increased during the procedure, interpreted as indicating pore enlargement; the latter showed that maximum
water content can be achieved very rapidly and, importantly,
that the effectiveness of slaking is governed by the length of
the drying cycle and whether it is allowed to run its full course
before a subsequent wetting phase.
4.13.3.1.3
Phase changes
These occur when a substance within rock pores undergoes a

phase change from liquid to solid (see Chapters 4.12 and
4.15). In the context of rock weathering, phase changes occur
when water freezes to form ice, and salts crystallize either from
saline fluid or by hydration. The force of crystallization then
acts against the confining walls, creating a tensile stress.
Freezing
(of water): Water freezing within rock expands by

9% with the theoretical potential under ideal conditions to
generate a force of 207 MPa, substantially greater than the
tensile strength of most rocks (normally 110 MPa). These
ideal conditions include a water content of 8090% of the
saturation level (McGreevy and Whalley, 1985), and are
most likely to be met within the outer few centimeters of an
already saturated rock (Matsuoka and Murton, 2008).
The freezing (and thawing) of ice within rock has recently
been re-evaluated by Hall (2004), with internal temperature
data for bricks exposed to freezing, and showed that rock
surface temperatures vary according to aspect and slope, thus
creating a spatial mosaic of temperature variation across and
through the rock surface.
Crystallization pressure: Correns (1949) developed a saltspecific theoretical basis for calculating the pressure generated by a growing crystal, a function primarily of the
supersaturation of the fluid from which it is crystallizing.
Steiger (2006) updated the Correns model, and found that
with respect to NaCl Correns underestimated the forces
generated by a factor of B3. Since NaCl does not readily
supersaturate, it is therefore likely to be achieved only under
kinetic conditions, during rapid change such as evaporation
or cooling.
Rodriguez-Navarro et al. (1999) and Doehne (2002) presented benchmark reviews of salt weathering mechanisms.
They showed that rock breakdown depends not only on the
supersaturation ratio but also the location of crystallization
relative to the rock surface. If this occurs within the rock,
blocking pores and inhibiting fluid migration, then nondamaging subflorescence forms. Crystallization as efflorescence
at the rock surface is also non-damaging. Crystallization close
beneath the rock surface, however, can cause significant
damage. The behavior of the salt solution is critical in this and

its movements are strongly influenced by its own physical
properties such as viscosity, surface tension, and vapor pressure, as well as evaporation and consequent flow rates. With
respect to NaCl it appears that rapid evaporation, encouraged
by conditions of high temperature, high wind flow, and low
relative humidity, is required to induce subflorescent precipitation. Given the prevalence of humid conditions at the coast,
however, it would appear that temperature and windflow are
critical factors in salt weathering in this domain.
Hydration, in which anhydrous minerals adsorb water
and consequently swell, also creates crystallization pressure
(Mortensen, 1933). Some common salts readily hydrate and
dehydrate under daily temperature and humidity variations,
creating calculable stresses during expansion (Winkler and
Wilhelm, 1970). Steiger (2006), however, showed that the hydrate becomes unstable with increasing pressure, thereby causing the loaded crystal faces to dehydrate, reducing crystal
volume, and limiting the crystallizing force of the hydrating salt.
Seawater yields five salts that may exist in hydrated form.
Of these, NaCl is easily the most dominant by volume
(B78%), but exists in hydrate form only below 0.1 1C and is,
therefore, zonally limited in its effects. Other hydrating salts in
seawater include anhydrite and polyhalite, but have as yet
unknown although probably limited potential to contribute
significantly to rock weathering by crystallization.
Despite a substantial body of literature on experimental
weathering of rocks by crystallization from saline solutions,
only a small number have used seawater. Mottershead (1982a)
presented a diagnostic study of four environmental conditions
in discriminating between potential weathering processes
acting on greenschist. Davison (1986) investigated a range of
temperature regimes in experimental weathering by sea salt.
Tingstad (2008), in an experiment of 28 diurnal cycles, used
SEM imaging to identify the penetration of seawater into a
limestone in comparison with sodium chloride and sodium
sulfate, showing differing penetration behaviors by the various
fluids.
A significant recent development would appear to be that
the locus of salt crystallization is influenced by the heating
regime used in experimentation. Gomez-Heras and Fort (2007)
found that by using radiative rather than convective heating,
NaCl crystallized in the more damaging subsurface form rather
than as relatively harmless subflorescence or efflorescence. This
may be one cause of the earlier difficulty in demonstrating rock
breakdown by this salt (e.g., Goudie, 1974).
Frost
and salt combination: Frost and salt crystallization are

each individually powerful rock weathering mechanisms.
Several laboratory studies of the combined effects of
freezing salt solutions have been undertaken (e.g., Williams
and Robinson, 1981; McGreevy, 1982; Williams and
Robinson, 2001), with sometimes contradictory results.
Jerwood et al. (1990a, b) demonstrated the complexity of
rock responses to frost plus salt in the case of chalk. The
potential mechanisms in play with salt and frost in combination have been fully reviewed by Williams and Robinson (1991), who suggested that they vary from site to site
according to the freezing regime, the salts present and their
concentrations, and the physical properties of the rock.
Coastal Weathering
Trenhaile and Rudakas (1981) studied the effect of seawater on frost cycling in rocks experimentally and were
successful in creating rock breakdown. It was found that
saline water at half the concentration of seawater was more
effective than both normal and higher seawater concentrations, and deionized water.
A field test of the hypothesis that frost enhances salt weathering is provided by Robinson and Jerwood (1987a, b).
Chalk shore platforms in Sussex, UK, are constantly subject to
salt weathering, and in the harsh winter of 1985, were additionally subjected to frequent frosts. A substantial increase in
rock weathering and the creation of weathered debris occurred
during that period, implying that the freezing effect augmented
that of salt in considerably enhancing weathering rates.
Salt mixtures: Seawater is a mixture of salts, which raises the
question of how its overall weathering effects compare with
its individual component salts. McGreevy (1982) suggested
that salt combinations may be more effective than single
salts. Robinson and Williams (2000) investigated this experimentally with a range of salts, finding different and
unpredictable behaviors in respect of enhancement or inhibition of rock breakdown. They did, however, find a
consistent tendency for a halite/gypsum combination to be
more effective than either of the two salts individually, as
observed previously by Goudie (1986).
A further perspective on this question is provided by
Mottershead (1982a, b), facilitating a comparison of complete seawater with each of its major component salts at
equivalent concentrations, and showing that NaCl singly
and alone equaled the capacity of seawater in weathering
greenschist in a wetting and drying regime, and implying
that the effects of the other salts are negligible. Robinson and
Williams (2000) suggested that the proportions of salts in
combinations are important, and it may well be that the
dominance by volume of NaCl in seawater accounts for its
apparently dominant behavior in rock weathering.
Clay mineral swelling: Clay minerals may undergo volume
changes on wetting and drying, an effect that may be enhanced in the presence of salt. Rock with a clay mineral
component will undergo disruptive internal stresses as a
consequence (Pye and Mottershead, 1995).
4.13.3.2
4.13.3.2.1

Dissolution
Most rock-forming minerals are slightly soluble in water,

whereas a small number such as calcite, gypsum, and halite
are highly soluble. Mineral solubility in seawater, however,
differs. Seawater, for example, is saturated with respect to
calcium carbonate and is not a natural solvent for that particular mineral. Joly (1904) conducted carefully controlled
experiments on the comparative solvent effects of seawater
and distilled water on a small selection of rocks (basalt, obsidian) and minerals (hornblende, orthoclase). The total rockforming elements found in solution following the seawater
treatment substantially exceeded those in the equivalent distilled water treatment, by a factor of not less than 3 in the case
of basalt, not less than 8 for hornblende, more than 4 for
obsidian, and more than 14 for orthoclase.
235
Joly used seawater at its natural concentration. During

evaporation, however, the composition of seawater brine
changes and it becomes more concentrated. It appears highly
possible that its power as a solvent of rock-forming minerals
may increase in the latter stages of evaporation when in a
hyperconcentrated state, a consequence that may be of significance in coastal rock weathering. Corroded crystal margins
have been reported in coastally exposed rocks (McGreevy,
1985; Young, 1987; Mottershead and Pye, 1994), and the
consequent weakening of intercrystal bonds may be the consequence of just such a dissolution process.
4.13.3.3
4.13.3.3.1
Biological Weathering Processes

Macrobiological processes
Sachs (1865) first demonstrated the effects of higher plants on

rock weathering. In only a few days, seeds had germinated,
sending down roots whose exudates had caused their traces to
become etched by corrosion into polished rock surfaces (Mottershead and Viles, 2004). A range of higher plants are pioneer
species in the supratidal zone, although they do not appear to
have been the focus of specific weathering studies.
4.13.3.3.2
Microbiological processes
In contrast, interest in the weathering effects of microorganisms has developed rapidly in recent years (Viles, 1995;
Ehrlich, 1998; Wakefield and Jones, 1998). Several groups of
microorganisms are involved in rock weathering, including
bacteria, fungi (photosynthetic organisms including algae and
cyanobacteria), and lichens. These organisms themselves
produce a range of potential stone deteriogens, including acids
and chelating compounds, which are capable of etching
mineral surfaces and leaching Fe, Mg, and silicate minerals
(Wakefield and Jones, 1998).
Fungi possess hyphae that physically penetrate rock and
mineral structures, mechanically disrupting mineral surfaces.
Bacteria and algae create mucilage, an extracellular amorphous
matrix that is hygroscopic and absorbs moisture from the atmosphere, expanding and shrinking in the process and creating disruptive tensile stresses on rock surfaces. The existence
of these organic bodies on a rock surface, in creating their own
habitat, also mediates the microenvironmental conditions in
respect of temperature and moisture regimes of the subjacent
rock. It is thought that the overall effects of microbial action
may contribute 2030% of the total rock weathering.
Brehm et al. (2005) reviewed the ways in which organic
biofilms break down quartz in a variety of forms crystalline,
subcrystalline, and granular, by modifying the rock surface
environment. This biofilm modifies the temperature and
moisture characteristics of the rock surface and can shift the pH
into the alkaline range, where quartz becomes soluble. Song
et al. (2007) experimentally demonstrated bacterial weathering
of plagioclase and quartz surfaces infected with Bacillus subtilis
and showing substantial new pitting in only 30 days.
In a coastal environment, Mottershead et al. (2003)
showed substantial differences between north and south aspects in weathering rates in relation to biofilm coverage. It was
concluded that biofilm protection from or exposure to
236
Coastal Weathering
Fort Perch Rock north face

(a)
Fort Perch Rock south face

(b)
30 m
100 m
Figure 4 Contrasting biofilm cover on the shaded north (a) and

sunny south (b) walls of Fort Perch Rock, Merseyside, UK. On the
former, halite crystals form on the biofilm that protects the rock
surface beneath; on the latter, the halite is in direct contact with the
rock surface.
weathering agents is a major controlling factor on consequent

weathering rates (Figure 4).
4.13.3.4
Complexity and Rates of Rock Weathering by

Salts
It is clear from this review that many weathering processes

have the potential to operate in the coastal zone, raising the
problem of diagnosing the operative mechanism at a particular site. It is, for example, only too easy to conclude in
favor of salt weathering on the basis of salt presence alone; this
propensity to infer mechanisms without further supporting

observation has been described as guilt by association
(McGreevy, 1985). It does not, however, identify which of
several potential salt-related mechanisms may be operative as
apparently conflicting evidence may come to light over time,
as exemplified by studies of coastal weathering of greenschist
in South Devon, UK.
Evidence from field monitoring of greenschist in the immediate supratidal zone revealed an annual weathering rate of
0.6 mm a1, in the presence of granular debris and salt crystals
(Mottershead, 1982a). A four-way controlled laboratory experiment was undertaken using combinations of seawater and
distilled water, under both wetting and drying and continuously humid conditions, in order to identify the effects of
these various environmental conditions. The four combinations in this black box approach allowed inferences regarding
the different weathering mechanisms permitted in each treatment. The outcome, in which wetting and drying in seawater
was the sole treatment condition to cause rock weathering
within six months, and exactly simulating the field Wr, was
interpreted as implying that salt crystallization was the effective mechanism. In particular, it was notable that six months
of continuous immersion in a seawater solution without
drying produced no observable rock weathering as either
weight loss or debris, implying that no significant dissolution
took place over that timescale, in contrast to the rapid mass
loss under wetting and drying. A further experiment simulating wetting and drying with individual salts showed that
NaCl alone was effective and sufficient to produce the weathering rate observed in the field (Mottershead, 1982b).
Continued field monitoring, however, revealed a consistent
seasonal variation in the Wr, with a clear summer maximum,
clearly indicating a climatic control (Mottershead, 1989).
Given the range of potential weathering mechanisms available
in this context, it could be interpreted either that more frequent drying of the rock was causing more frequent crystallization or that more complete evaporation was exposing the
rock to aggressive late-stage brine, or that the higher temperatures were directly creating greater thermal stress or greater
differential expansion of halite in the rock pores.
A further study of tafoni (see Section 4.13.4.2) in nearby
cliffs within the same rock, showing at least one order of
magnitude lower weathering rate yielded further conflicting
evidence (Mottershead and Pye, 1994). Microprobe analysis
revealed the presence of chlorine throughout the rock, although no NaCl crystallization was visible. Crystal margins
showed signs of corrosion. It was concluded that Cl was finely
distributed throughout rock pore surfaces and had caused the
weakening of intercrystal bonds by corrosion. The same study
showed that the rock was substantially reduced in strength
throughout the top 27 mm from the weathering surface.
In this case, continuing study in both field and laboratory revealed the potential for the existence of a range of
salt-related mechanisms, and offers the possibility of reconciling apparently conflicting evidence. It now appears likely that
mineral dissolution takes place at grain boundaries, perhaps on
a long-term basis, leading to weakening of the surface layer. The
subsequent detachment of weathered material appears to be
driven by mechanical responses to short-term environmental
events such as salt crystallization or temperature changes.
Coastal Weathering
237
30 m
Fort Perch Rock south face
The mere presence of salt crystals alone on a rock surface

does not necessarily imply that crystallization weathering is
occurring (Figure 5), nor does an abundance of halite crystals
within rock pores necessarily mean that the force of crystallization is exerting a critical tensile stress against cavity margins within the rock (Figure 6). A range of field and laboratory
evidence requires careful analysis before a specific mechanism
may be safely identified rather than simply inferred.
Despite the less than complete understanding of how salts
facilitate rock breakdown, it is, nevertheless, evident from
field studies that coastal salt weathering may be a very aggressive process. Field measurements of surface lowering of
41 mm a1 have been observed, up to a maximum of 5.25
mm a1 over 20 years (Takahashi et al., 1994), which are exceptional values for indurated and insoluble rocks. These very
high weathering and erosion rates occur around HTL and are
influenced also by aspect, embracing exposure to solar radiation and storm waves (Figure 7). Field measurements of
surface lowering 41 mm a1 have been reported with a value
of 2.2 mm a1 over 38 years reported by Aoki and Matsukura
(2007) (Figure 7). These are exceptional erosion figures for
insoluble and indurated rocks, and are associated usually with
tafoni or honeycomb formation close to a high water level.
W
2
Altitude (m)
Figure 5 Coastal brickwork showing abundant salt efflorescence.

Doehne (2002) showed that efflorescence is not in itself damaging,
and further subsurface investigation would be required in this case to
identify the mechanism of material breakdown. Interestingly, the brick
matrix is less resistant to coastal weathering than the intervening
mortar, forming miniature tafoni bounded by the more resistant
mortar.
1
MHWS
MTL
0
10
Depth of erosion
15
11
10
9
8
7
6
5
4
3
2
1
Stone courses
Figure 6 Halite crystals on sandstone from a coastal environment.

Although relatively abundant, none of the crystals is evidently
growing against the margin of a rock cavity. There is, however,
evidence of both small and larger etch pits, indicative of corrosion of
the foreground crystal face.
20 cm
Figure 7 Erosion depth of sandstone in a coastal structure after 38

years of exposure, showing maximum values around the mean high
water spring tide level. The relationship between erosion and aspect,
with the highest rate on the exposed south face and the lowest on
the shaded north face, implies that solar radiation has a major
influence on the weathering rate. Redrawn with permission from
Takahashi, K., Suzuki, T., Matsukura, Y., 1994. Erosion rates of a
sandstone used for a masonry bridge pier in the coastal spray zone.
In: Robinson, D.A., Williams, R.B.G. (Eds.), Rock Weathering and
Landform Evolution. Wiley International, Chichester, pp. 175192.
4.13.4
Coastal Landforms Associated with

Weathering
Some characteristic coastal landforms, such as shore platforms, can be attributed at least in part to weathering action.
Others, tafoni (see Section 4.13.4.2) and honeycombed rocks,
238
Coastal Weathering
develop rapidly by rock weathering under coastal conditions

but are not exclusively coastal, forming in other environments
also characterized by the presence of abundant salts.
1.80 mm a1
Lowest high tide
4.13.4.1
0.91 mm a1
0.98 mm a1
Extreme high tide

Mean high spring tide
Shore Platforms
Shore platforms form in the intertidal zone as a consequence

of the erosion of cliffs bounding a landmass. The platform
surface is then exposed to subaerial weathering for a considerable part of the tidal cycle.
The relative contributions of subaerial rock weathering and
mechanical erosion by incoming waves to platform development are currently contested (Dasgupta, 2011). Stephenson and
Kirk (2000a, b) argued that the energy of incoming waves at
Kaikoura, New Zealand, especially storm waves, is dissipated in
shallowing waters offshore and at the seaward margin of the
platform, and that minimal energy then remains for platform
erosion by waves. They maintained, on the basis of the mechanics of breaking waves, that wave contribution to platform
lowering is negligible. In contrast, they observed a reduction of
up to 50% in the mechanical strength of rocks on the platform,
and abundant physical evidence of weathering, with rock particles becoming readily available for removal. Trenhaile and
Kanyaya (2007) subsequently demonstrated that the role of
wave energy varied according to tidal range, being more effective
on macrotidal coasts where waves are able to reach the coastline, riding across and eroding the platforms before breaking.
The data of Porter et al. (2010a; their Table 4) facilitated
comparison of laboratory experimental weathering rates and
field platform lowering rates. They show good agreement in
both sandstone and basalt platforms between field and
simulated rates above lowest high tide (LHT) level, implying
that weathering processes alone are sufficient to cause the
observed loss. Below LHT weathering appears on average to
account for B35% of lowering on sandstone and B23% on
basalt platforms, implying that weathering is inhibited and
that wave mechanical action plays a more significant role in
the mid and lower parts of the platform.
The relative effectiveness of the different weathering processes likely to be in play on a shore platform has been assessed experimentally in the laboratory, demonstrating that
slaking, salt weathering, and freezing and thawing all produce
weathering effects under simulated intertidal weathering
conditions (Kanyaya and Trenhaile, 2005; Moses et al., 2006;
Porter et al., 2010a, b).
Field measurements of shore platform lowering by a variety
of authors (Foote et al., 2006; Stephenson and Kirk, 2000b,
Porter et al., 2010a, b) showed that maximum lowering tends
to occur in the upper intertidal zone above the LHT level.
This shows that more rapid lowering occurs in locations
with a low inundation duration, the highest frequency of
slaking cycles (Stephenson and Kirk, 2000b), and where the
drying phase of the slaking cycle is longest. These conditions
favor slaking, salt weathering, frost weathering (where
suitable temperatures apply), and biological activity, a cumulative circumstance in which a wide range of weathering
mechanisms may flourish.
Overall, and current contention notwithstanding, it is becoming clear that weathering processes contribute a significant
component of the material available for erosion. This issue will
Mid-tide
0.61 mm a1
Highest
0.76 mm a1
low tide
0.024 to
0.027 mm a1
Extreme
low tide
Figure 8 Relationship between shore platform downwearing and
elevation shown by laboratory simulation studies. The highest
lowering rates occur in the upper intertidal range, between the lowest
high tide and the mean high spring tide levels. Redrawn with
permission from Porter, N.J., Trenhaile, A.S., Prestanski, K.J.,
Kanyaya, J.I., 2010a. Shore platform downwearing in eastern Canada:
the mega-tidal Bay of Fundy. Geomorphology 118(12), 112.
perhaps become clarified when similarly intensive datasets are

available for a wider range of coastal environments.
The field measurements of Porter et al. (2010a) are consistent with the observations of Aoki and Matsukura (2007) and
Mustoe (2010) that revealed maximum weathering rates around
high tide level (Figure 8). This is consistently shown to be the
locus of the most aggressive weathering, and confirms the
qualitative model (Figure 9) of Johannessen et al. (1982).
The longer term implications of these process considerations
are the subject of modeling of shore platform development
(Trenhaile, 2000, 2004, 2008). Dasgupta (2011) pointed to future avenues for further research on shore platforms.
4.13.4.2
Tafoni (Singular Tafone)
Tafoni are cavernous features, at a scale of centimeters

to meters, eroded into rock outcrops at elevations from close
to sea level up to 50 m or more (see Chapter 4.7). They are
characterized
by freshly
weathered
interior
faces
and commonly partially enclosed by lateral walls and an
overhanging visor. Such tafoni are commonly associated with
rocks possessing an outer skin of hardened rock, indurated by
mineral accumulation, which, when breached, opens out to a
cavernous interior (Figure 10).
Tafoni exist in a wide range of climatic conditions and in
particular, in coastal and desert environments, in both of
which salts are abundant. They may develop both on natural
rock surfaces and also on built structures (McGreevy, 1985;
Takahashi et al., 1994).
Recent studies of coastal tafoni include Matsukura and
Matsuoka (1991) and Mottershead and Pye (1994). The latter
show that tafoni may develop from a small puncture through
a case hardened surface, which then opens out internally as
less resistant rock behind becomes exposed to weathering
agents. Adjacent small tafoni may coalesce ultimately to form
recessed freshly weathering rock faces several meters in lateral
and vertical dimensions.
Internal walls may be covered with flaking or scaling surfaces, disaggregating crystals/grains, and the tafone floor may
239
So
il
Coastal Weathering
Highest storm wave attack

Fresh water protected zone
limited salt spray,
washed by rain and soil drainage
Honeycomb
surface
23 m
Fault
Maximum normal spray zone
West
East
as
si
ve
sa
n
ds
to
ne
1
1.
Salt spray zone

Max. exposure to salt deposition
by spray: max. heating by sun
Salt water
protected zone
Washed with
every tide
Low tide
(swash limit)
High tide swash zone

Honeycomb
surface
Figure 9 Coastal rock weathering and erosion in relation to sea level, Cape Arago, Oregon. Redrawn with permission from Johannessen, C.L.,
Feiereisen, J.J., Wells, A.N., 1982. Weathering of ocean cliffs by salt expansion in a mid-latitude coastal environment. Shore and Beach 50,
2634.
be covered with a powdery sediment (rock meal) that has

become dislodged from the walls above. Earlier authors observed the association of tafoni with salts and inferred that salt
crystallization (including hydration) was responsible for rock
breakdown. Takahashi et al. (1994) showed that weathering
rates are positively correlated with received solar radiation.
Whatever the precise mechanism, it is clear that the conditions
for salt weathering are met, and involve the ingress of coastal
salts into the rock, and their movement in solution as controlled by wetting and drying processes. These will depend on
the microclimatic conditions within the tafone itself.
Theoretical models of tafone formation were developed by
Matsukura and Matsuoka (1991), Takahashi et al. (1994),
Sunamura (1996), and Huinink et al. (2004). All show an
exponentially diminishing rate of deepening over time. This is
interpreted as a feedback function, whereby the deepening of
the tafone modifies its internal conditions with respect to
temperature, shelter, or salt supply.
4.13.4.3
Figure 10 Tafone formed in greenschist on a coastal slope. The

weathering face is overhung by the visor, indicative of the depth of tafone
surface recession. Rock meal accumulates at its base. South Devon, UK.
Honeycomb Weathering
Honeycomb weathering is included by some authors under

tafoni (see Chapter 4.7). The two weathering forms differ,
however, in distribution, pattern, and scale and are therefore
treated separately here. Honeycomb, alternatively termed alveolar weathering or maladie alveolaire (Pauly, 1976), is a
swarm of cavities/alveoli at a scale of centimeters penetrating a
rock surface in a regular or an irregular pattern. Commonly,
they range from 10 to 50 mm in both diameter and depth,
240
Coastal Weathering
although diameter increases rapidly where they coalesce

(Figure 11). There appears to exist a developmental sequence
from small isolated alveoli to a dense mutually adjacent pattern. Where fully developed, they are separated by a thin wall
of remnant rock up to the initial surface, and just a few
millimeters wide, which may then decay, leading to coalescence of the hollows and ultimately a planar surface from
which the entire honeycombed layer has been stripped
(Mottershead, 1994).
Honeycomb weathering may develop on horizontal, vertical, sloping, or overhanging surfaces. It appears to develop
preferentially on sandstone, arkose, schist, and tuff. Pye and
Mottershead (1995) reported a highly honeycombed sandstone
in close juxtaposition to an indurated limestone from which
they are absent. This suggests that rock characteristics such as
mineralogy, texture, porosity, and strength may predispose
particular rock types to honeycomb development.
Honeycombed surfaces are commonly concentrated in a
zone either side of HTL and extending up to B3 m above HTL
(Mustoe, 2010). Marked gradients exist, with the intensity of
development decreasing with increased elevation above sea
level, distance from sea level, and direct exposure to the
marine influence (Mottershead, 1994). They also intensify on
the shoreward rather than the landward aspect, with a concentrated focus on the prevailing wind azimuth. These distributional characteristics point clearly to the marine origin of
the weathering agencies responsible.
Quayle (1992) suggested that wind eddies passing across
the rock surface may be responsible for honeycomb formation. Rodriguez-Navarro et al. (1999) demonstrated this
experimentally by passing an airflow across a rock surface
under controlled conditions and developing honeycomb formation in only 38 days. Once initiated, the cavities promoted
turbulence in the airflow, creating suction and increased
evaporation at the cavity surface, a positive feedback that increased the salt supply to the cavities. The weathering mechanism was inferred to be salt weathering by crystallization
promoted by rapid evaporation, as the saline fluid was preferentially drawn toward incipient hollows in the blasted surface. The balance between saline fluid input by capillary rise,
and rapid surface loss by evaporation caused damaging
Figure 11 Honeycomb well developed in coastal sandstone. The

walls have decayed in some places, causing the alveoli to coalesce.
Lens cap is 51 mm in diameter. Weston super Mare, UK.
subflorescences to develop in the cavities, thus deepening

them at the expense of the walls, which became isolated from
the supply of solution. The rapidity of honeycomb formation
suggests that breakdown by crystallization was responsible,
rather than solution.
Honeycomb cavities generally contain rock debris in the
form of rock meal, indicating rock breakdown by granular
disintegration; many of the ions in seawater are also present.
Mustoe (1982) noted the influence of microorganisms in
honeycomb development, later stressing their significance
(Mustoe, 2010). A thin film of blue green algae colonizes the
internal walls although not the floor, thereby inhibiting seawater penetration into the rock walls. In this way, the cavities
are able to deepen, whereas the protected walls thus prevent
decay of the honeycomb pattern into coalescence.
Honeycombs appear to develop very rapidly, shown by their
appearance on dated structures adjacent to the sea. Observations by Grisez (1960), Mustoe (1982), McGreevy (1985),
and Mottershead (1994) imply weathering rates of B0.5 mm
a1, with maxima of 41 mm a1 (Figure 12) (Figure 13).
Figure 12 Evidence of rapid coastal weathering (1). The oil has

protected the subjacent rock from weathering, and a pedestal
B10 mm high has emerged as the surrounding rock was weathered
and subsequently eroded. Prawle, UK.
Figure 13 Evidence of rapid coastal weathering (2). A layer of

honeycomb weathering has formed and subsequently been stripped
from this sandstone coping stone since its emplacement a century
earlier, indicating a weathering rate of 1 mm a1. Weston super
Mare, UK.
Coastal Weathering
4.13.5
Conclusion
This brief survey of coastal weathering reveals that considerable knowledge has accumulated of weathering activity within
this domain. In conclusion, some perspectives emerge on
prospective future avenues for the investigation of potential
processes and their controls.
4.13.5.1
Environmental Factors
Field studies of coastal rock weathering have been undertaken

in a range of environments globally, notably on UK, North
American, Australian and New Zealand, Japanese, and Mediterranean coasts. Some environmental conditions are, however, notably underrepresented, including tropical, hot desert,
and polar coastlines. It is apparent that a significant number of
accelerating factors, such as thermal shock, frost and salt
combinations, frost action, and freezing brine with associated
NaCl dihydrate formation, are prevalent in the polar domain;
scant observations of coastal weathering outcomes, however,
have been reported from that zone. This uneven spread limits
the currently available perspective of major zonal controls on
coastal weathering, as insufficient field observations are reported from contrasting domains for globally valid conclusions to be drawn.
Environmental controls also act at the site scale (Pope
et al., 1995). The presence of a coastline moderates climate
locally, creating the specific microclimate of the shoreline and
the backshore zone. This implies the existence of an environmental gradient normal to the shoreline, along which
climatic and saline parameters vary landward, observable at a
range of scales meters, tens, and hundreds of meters, which
may be expected to control weathering processes and
outcomes.
It is the case that recent studies of surface microclimates
have been revolutionized with the acquisition of highly
detailed environmental datasets via the application of
microprobes and electronic datalogging. The magnitude and
frequency of climatic inputs to, and their potential as controls
on, the coastal weathering system (Goudie and Viles, 1999)
are yet to be fully explored. Although wave data and standard
climatic variables such as temperature, wind, and precipitation
are readily measured, or available, those of splash, spray,
mist, and salt deposition processes are scarcely represented,
and yet are crucial to weathering processes. There has as yet
been a marked scarcity of such observations within the coastal
zone.
4.13.5.2
Lithological Factors
Laboratory and field observations have revealed the behavior

of a wide range of rocks in the coastal environment, both real
and simulated. It has been shown that rocks respond to seawater in differing and sometimes contradictory ways (e.g.,
Matsukura and Yatsu, 1985; Pye and Mottershead, 1995;
Porter et al., 2010a). The reasons for these varying responses
seem worthy of investigation, for they have the potential to
identify the material properties acting as controls on the rock
response and further illuminate the processes responsible. A
241
systematic analysis of even existing data in this respect would

surely yield interesting insights into lithological responses to
the coastal environment.
There has been limited utilization in weathering studies of
the mechanical properties of rocks, particularly tensile
strength, in respect of interpreting weathering processes
(e.g., Mottershead, 1982b) or weathering responses (e.g.,
Mottershead, 1989). The mechanical weakening of coastal
rocks by weathering before erosion is an aspect that merits
greater attention, particularly with respect to rock fatigue.
The demonstration of effective weathering is considerably
strengthened by measurements of mechanical strength of fresh
and subsequently weathered rock.
4.13.5.3
Process Factors
The crux of weathering processes lies in events at the pore scale,

and despite a wide range of research approaches in recent decades, identifiable gaps in knowledge remain. For example, the
chemical effects of seawater on rock minerals remain little
known. The work of Joly (1904) appears to represent the most
recent investigation of this topic, and may merit further study in
relation to the dissolutional capability of seawater. Furthermore,
the increasing concentration and changing composition of saline
brine during evaporation (or NaCl hydration) has hitherto been
little considered in the context of rock weathering. The potential
effects of the later minerals to crystallize has not, apparently,
been investigated. Given suggestions in more recent work of
chemical weathering, the presence of Cl and Na ions alone as
corrosive agents (Young, 1987) may also merit further investigation. There are, therefore, aspects of pore-scale chemical activity yet to be explored.
Much of the current knowledge of salt/rock interaction is
drawn from laboratory experiments in which the salts, their
concentrations, temperature and humidity conditions, and the
mode of moisture application are controlled. Such experiments, commonly in black box form, rest on assumptions
about the conditions and processes within the box and have
frequently produced conflicting conclusions. There is a need
for careful critical consideration of the assumptions of the
experiment design and the subsequent inferences made regarding the processes within the box.
Whereas rapid field weathering is perhaps intuitively more
easily interpreted in terms of mechanical processes, it remains
the case that there appears to be no published evidence of a
salt crystal actively prying apart a rock pore or microfracture.
The case for crystallization pressure weathering is not yet
proven, other than in the prying away of already loosened
particles.
Microbiological weathering processes have been little
studied to date, and their relative contribution to the overall
weathering budget is far from understood.
The coastal zone, particularly at the shoreline, is shown to
be a highly aggressive weathering environment, with contemporary weathering rates around HTL in the range of
101100 mm a1. Haslett and Curr (1998) have postulated a
weathering rate of 0.25 mm a1 throughout the Holocene for
coastal granites in Brittany. A nearby noncoastal site shows
rates at least one order of magnitude lower on similar rocks
242
Coastal Weathering
(Sellier, 1997), providing a rough initial measure of the

coastal effect
The central activity of coastal rock weathering takes place
in rock pores. Yet firm knowledge of processes at this scale
remains elusive and is based on inference and post hoc observation to a significant extent. Indeed, Johannessen et al.
(1982) construct an interpretation of entire coastal landscape
development on just such an inference of pore-scale activity.
The development of new techniques, for example ESEM, may
be expected to lead to better means of observing events at the
pore and, indeed, micropore scale, which are the key to a
deeper understanding of these processes.
References
Aoki, H., Matsukura, Y., 2007. Effects of rock strength and location heights on
growth rates of tafoni-like depressions at sandstone blocks used for a masonry
bridge pier in the coastal spray zone. Zeitschrift fur Geomorphologie NF
51(Suppl. 1), 115132.
Anikouchine, W.A., Sternberg, R.W., 1973. The World Ocean; An Introduction to
Oceanography. Prentice Hall, New Jersey, 339 pp.
Beer, T., 1997. Environmental oceanography 2E. CRC Press, Boca Raton, FL, 367 pp.
Berner, E.K, Berner, R.A., 1996. Global Environment: Water Air and Geochemical
Cycles. Prentice-Hall, New Jersey, 376 pp.
Bigg, G.R., 2003. The Oceans and Climate 2E. Cambridge University Press,
Cambridge, 288 pp.
Blanchard, D.C., Woodcock, A.H., 1957. Bubble formation and modification in the
sea and its meteorological significance. Tellus 9(2), 145158.
Brehm, U., Gorbushina, A., Mottershead, D., 2005. The role of microorganisms and
biofilms in the breakdown and dissolution of quartz and glass. In: Noffke, N.
(Ed.), Geobiology: Objectives, Concepts, Perspectives. Palaeogeography,
Palaeoclimatology, Palaeoecology 219, (12). Elsevier, Oxford, pp. 117129.
Brimblecombe, P., 1986. Air Composition and Chemistry. Cambridge University
Press, Cambridge, 268 pp.
Cadle, R.D., 1973. Particulate material in the lower atmosphere. In: Rasool, S.I.
(Ed.), Chemistry of the Lower Atmosphere. Plenum Press, London, pp. 69120.
Cooke, R.U., Smalley, I.J., 1968. Salt weathering in deserts. Nature 220, 12261227.
Correns, C.W., 1949. Growth and dissolution of crystals under linear pressure.
Discussions of the Faraday Society 5, 267271.
Dasgupta, R., 2011. Whither shore platforms? Progress in Physical Geography
35(2), 183209.
Davies, J.L., 1972. Geographical Variation in Coastal Development. Oliver & Boyd,
Edinburgh, 204 pp.
Davison, A.P., 1986. An investigation into the relationship between salt weathering
debris production and temperature. Earth Surface Processes and Landforms
11(3), 335341.
Doehne, E., 2002. Salt weathering: a selective review. In: Siegesmund, S, Weiss, T,
Vollbrecht, A (Eds.), Natural Stone, Weathering Phenomena, Conservation
Strategies and Case Studies. Geological Society of London Special Publications
205. The Geological Society, London, pp. 5164.
Ehrlich, H.L., 1998. Geomicrobiology: its significance for geology. Earth Science
Reviews 45(12), 4560.
Erickson, III. D.J., Duce, R.A., 1988. On the global flux of atmospheric sea salt.
Journal of Geophysical Research 93(C11), 1407914088.
Eriksson, E., 1960. The yearly circulation of chloride and sulfur in nature:
meteorological, geochemical and pedological implications. Part II. Tellus 12(1),
63109.
Eugster, H.P., Harvie, C.E., Weare, J.H., 1980. Mineral equilibria in a six-compound
seawater system, NaKMgCaSO4ClH2O at 25 1C. Geochimica
Cosmochimica Acta 44(9), 13351347.
Foote, Y., Plessis, E., Robinson, D.A., Henaff, A., Costa, S., 2006. Rates and
patterns of downwearing of chalk shore platforms of the Channel: comparisons
between France and England. Zeitschrift fur Geomorphologie N.F. 144, 93115.
Gilbert, G.K., 1877. The Geology of the Henry Mountains. United States
Geographical and Geological Survey, Washington.
Gomez-Heras, M., Fort, D., 2007. Patterns of halite (NaCl) crystallisation in building
stone conditioned by laboratory heating regimes. Environmental Geology 52(2),
259267.
Goudie, A.S., 1974. Further experimental investigation of rock weathering by salt

and other mechanical processes. In: Schick, A.P., Yaalon, D.H., Yair, A. (Eds.),
Geomorphic Processes in Arid Environments, Zeitschrift fur Geomorphologie
N.F. Suppl. 2. Schweizerbart, Stuttgart, pp. 112.
Goudie, A.S., 1986. Laboratory simulation of the wick effect in salt weathering of
rock. Earth Surface Processes and Landforms 11(3), 275285.
Goudie, A.S., Viles, H.A., 1999. The frequency and magnitude concept in relation to
rock weathering. In: Crozier, M., Mausbacher, A. (Eds.), Magnitude and
Frequency in Geomorphology, Zeitschrift fur Geomorphologie Suppl. 115.
Schweizerbart, Stuttgart, pp. 175189.
Grisez, L., 1960. Alveolisation littorale de schistes metamorphiques. Revue de
Geomorphologie Dynamique 11, 164167.
regions. Geomorphology 31(14), 4763.
Hall, K., 2004. Evidence for freeze-thaw events and their implications for rock
weathering in Northern Canada. Earth Surface Processes and Landforms 29(1),
4357.
Hall, K., Andre, M.-F., 2001. New insights into rock weathering from high-frequency
rock temperature data: an Antarctic study of weathering by thermal stress.
Landforms 28(8), 823836.
Hall, K., Guglielmin, M., Strini, A., 2008. Weathering of granite in Antarctica: II.
Thermal stress at the grain scale. Earth Surface Processes & Landforms 33(3),
475493.
Hall, K., Hall, A., 1996. Weathering by wetting and drying: some experimental
results. Earth Surface Processes and Landforms 21(4), 365376.
Harvie, C.E., Weare, J.H., Hardie, L.A., Eugster, H.P., 1980. Evaporation of seawater:
calculated mineral sequences. Science 208, 498500.
Haslett, S.K., Curr, R.J.F., 1998. Coastal rock platforms and Quaternary sea-levels in
the Baie dAudierne, Brittany, France. Zeitschrift fur Geomorphologie NF 42(4),
507515.
Hidy, G.M., 1973. Removal processes of gaseous and particulate pollutants.
In: Rasool, S.I. (Ed.), Chemistry of the Lower Atmosphere. Plenum press,
London, pp. 121176.
Hingston, F.J., Gailitis, V., 1976. The geographic variation of salt precipitated over
Western Australia. Australian Journal of Soil Research 14(3), 319335.
Surface Processes and Landforms 29(10), 12251233.
Jerwood, L.C., Robinson, D.A., Williams, R.B.G., 1990a. Experimental frost and
salt weathering of chalk I. Earth Surface Processes and Landforms 15(7),
611624.
Jerwood, L.C., Robinson, D.A., Williams, R.B.G., 1990b. Experimental frost and
salt weathering of chalk II. Earth Surface Processes and Landforms 15(8),
699708.
Johannessen, C.L., Feiereisen, J.J., Wells, A.N., 1982. Weathering of ocean cliffs by
salt expansion in a mid-latitude coastal environment. Shore and Beach 50,
2634.
Joly, J., 1904. Some experiments on denudation by solution in fresh and salt water.
Proceedings, Royal Irish Academy Section A 24, 2133.
Kanyaya, J.I., Trenhaile, A.S., 2005. Tidal wetting and drying on shore platforms: an
experimental assessment. Geomorphology 70(12), 129146.
Lewis, E.R., Schwartz, S.E., 2004. Sea salt aerosol production. Mechanisms,
methods, measurements and models: a critical review. Geophysical Monograph
152. American Geophysical Union, Washington, DC.
Malloch, A.J.C., 1972. Salt-spray deposition on the maritime cliffs of the Lizard
peninsula. Journal of Ecology 60(1), 103112.
Mason, B.J., 2001. The role of sea-salt particles as cloud condensation nuclei over
the remote oceans. Quarterly Journal Royal Meteorological Society 127(576),
20232032.
Matsukura, Y., Matsuoka, N., 1991. Rates of tafoni weathering on uplifted shore
platforms in Nojima-Zaki, Boso peninsula, Japan. Earth Surface Processes and
Landforms 16(1), 5156.
Matsukura, Y., Yatsu, E., 1985. Influence of salt water on slaking rate of Tertiary
shale and tuff. Transactions Japanese Geomorphological Union 6(2), 163167.
Matsuoka, N., Murton, J., 2008. Frost weathering: recent advances and future
directions. Permafrost and Periglacial Processes 19(2), 195210.
McCaffrey, M.A., Lazar, B., Holland, H.D., 1987. The evaporate path of seawater and
the coprecipitation of Br and K with halite. Journal of Sedimentary Research
57(5), 928937.
McGreevy, J.P., 1982. Frost and salt weathering: further experimental results. Earth
Surface Processes and Landforms 7(5), 475488.
Coastal Weathering
McGreevy, J.P., 1985. A preliminary scanning electron microscope study of

honeycomb weathering of sandstone in a coastal environment. Earth Surface
McGreevy, J.P., Whalley, W.B., 1985. Rock moisture content and frost weathering
under natural and experimental conditions: a comparative discussion. Arctic and
Alpine Research 17(3), 337346.
data from hyper-arid desert environments in the Atacama and the Antarctic Dry
Valleys and implications for rock weathering. Geomorphology 110(34), 182187.
Molaro, J.L., McKay, C.P., 2010. Processes controlling rapid temperature
variations on rock surfaces. Earth Surface Processes and Landforms 35(5),
501507.
Mortensen, H., 1933. Die Salzsprengung und hire Bedeutung fur die regional
klimatische Gliederung der Wusten. Petermanns Geographische Mitteilungen 79,
130135.
Moses, C.A., Robinson, D.A., Williams, R.B.G., Marques, F.M.F.S., 2006. Predicting
rates of shore platform downwearing from rock geotechnical properties and
laboratory simulation of weathering and erosion. In: Robinson, D.A., Lageat, Y.
(Eds.), European Shore Platform Dynamics, Zeitschrift fur Geomorphologie NF
Suppl. 144. Schweizerbart, Stuttgart, pp. 1937.
Mottershead, D.N., 1982a. Coastal spray weathering of bedrock in the supratidal
zone at East Prawle, South Devon. Field Studies 5, 663684.
Mottershead, D.N., 1982b. Rapid weathering of greenschist by coastal salt spray,
East Prawle, south Devon; a preliminary report. Proceedings of the Ussher
Society 5, 347353.
Mottershead, D.N., 1989. Rates and patterns of bedrock denudation by coastal salt
spray weathering: a seven year record. Earth Surface Processes & Landforms
14(5), 383398.
Mottershead, D.N., 1994. Spatial variations in intensity of alveolar weathering of a
dated sandstone structure in a coastal environment, Weston super Mare, UK,
151174. In: Robinson, D.A., Williams., R.B.G. (Eds.), Rock Weathering and
Landform Evolution. Wiley International, Chichester, 519.
Mottershead, D.N., Gorbushina, A., Lucas, G., Wright, J.S., 2003. The influence of
aspect, marine salts and microbes in the weathering of sandstone in two historic
structures. Building and Environment 38(910), 11931204.
Mottershead, D.N., Pye, K., 1994. Tafoni on coastal slopes, South Devon, UK. Earth
Surface Processes & Landforms 19(6), 543563.
Mottershead, D.N., Viles, H.A., 2004. Experimental studies of rock weathering
by plant roots: updating the work of Julius Sachs (18321897). In: Mitchell, D.J,
Searle, D.E. (Eds.), Stone Deterioration in Polluted Urban Environments, Land
Reconstruction and Management 3. Science Publishers, Enfield, NH, pp. 6172.
Mustoe, G.E., 1982. The origin of honeycomb weathering. Bulletin of the Geological
Society of America 93(2), 108115.
Mustoe, G.E., 2010. Biogenic origin of coastal honeycomb weathering. Earth
Surface Processes & Landforms 35(4), 424434.
Naylor, N.A., Stephenson, W.J., Trenhaile, A.S., 2010. Rock coast geomorphology: recent advances and future research directions. Geomorphology
114(12), 311.
ODowd, C.D., de Leeuw, G., 2007. Marine aerosol production: a review of the
current knowledge. Philosophical Transactions, Royal Society A 365(1856),
17531774.
Pauly, J.-P., 1976. Maladie alveolaire: conditions de formation et devolution.
In: Rossi-Manresi, R (Ed.), The Conservation of Stone. Centro per la
conservazione delle sculture all aperto, Bologna, pp. 5580.
Peel, R.F., 1974. Insolation weathering: some measurements of diurnal temperature changes in exposed rocks in the Tibesti region, central Sahara. In:
Schick, A.P., Yaalon, D.H., Yair, A. (Eds.), Geomorphic Processes in Arid
Environments. Geomorphologie N.F. Suppl. 21. Schweizerbart, Stuttgart, pp.
1928.
Pethick, J., 1984. An Introduction to Coastal Geomorphology. Edward Arnold,
London, 272 pp.
understanding geographical variation in weathering. Annals Association
Porter, N.J., Trenhaile, A.S., Prestanski, K., Kanyaya, J.I., 2010. Patterns of surface
downwearing on shore platforms in eastern Canada. Earth Surface Processes
and Landforms 35(15), 17931810.
Porter, N.J., Trenhaile, A.S., Prestanski, K.J., Kanyaya, J.I., 2010a. Shore platform
downwearing in eastern Canada: the mega-tidal Bay of Fundy. Geomorphology
118(12), 112.
Porter, N.J., Trenhaile, A.S., Prestanski, K.J., Kanyaya, J.I., 2010b. Shore platform
downwearing in eastern Canada: micro-tidal Gaspe, Quebec. Geomorphology
116(12), 7786.
243
Pye, K., Mottershead, D.N., 1995. Honeycomb weathering of Carboniferous

sandstone in a sea wall at Weston-super-Mare, U.K. Quarterly Journal
Quayle, N.J.T., 1992. Alveolar decay in stone its possible origins. In: Rodrigues,
J.D., Henriques, F., Jeremias, F.T. (Eds.), Proceedings of the 7th International
Congress on Deterioration and Conservation of Stone. Laboratorio Nacional de
Engenharia Civil, Lisbon, Vol. 1, pp. 109118.
Randall, R.E., 1970. Salt measurement on the coast of Barbados, West Indies. Oikos
21(1), 6570.
Randall, R.E., 1974. Airborne salt deposition and its effects upon coastal plant
distribution: the Monach Isles National Nature Reserve, Outer Hebrides.
Transactions Botanical Society Edinburgh 42, 153162.
Robinson, D.A., Jerwood, L.C., 1987a. Frost and salt weathering of chalk shore
platforms near Brighton, Sussex, UK. Transactions Institute British Geographers
N.S. 12(2), 217226.
Robinson, D.A., Jerwood, L.C., 1987b. Subaerial weathering of chalk shore
platforms during harsh winters in Southeast England. Marine Geology 77(12),
114.
Robinson, D.A., Williams, R.B.G., 2000. Experimental weathering of sandstone by
combinations of salts. Earth Surface Processes and Landforms 25(12), 1309
1315.
Rodriguez-Navarro, C., Doehne, E., 1999. Salt weathering: influence of evaporation
rate, supersaturation and crystallization pattern. Earth Surface Processes and
Landforms 24(3), 191209.
111(8), 12501255.
Sachs, J., 1865. Handbuch der Experimental-Physiologie den Pflanzen. Wilhelm
Engelmann, Leipzig, 514 pp.
Sellier, D., 1997. Utilisation des megaliths commes marqueurs de la vitesse de
lerosion des granites en milieu tempere: enseignements apportes par les
alignements de Carnac (Morbihan). Zeitschrift fur Geomorphologie N.F 41(3),
319356.
Song, W., Ogawa, N., Oguchi, C.T., Hatta, T., Matsukura, Y., 2007. Effect of
Bacillus subtilis on granite weathering: a laboratory experiment. Catena 70(3),
275281.
Steiger, M., 2006. Crystal growth in porous materials: influence of supersaturation
and crystal size. In: Fort, R., Alvarez de Buergo, M., Gomez-Heras, M.,
Vasquez-Calvo, C. (Eds.), Heritage Weathering and Conservation, Proceedings of
the International Conference on Heritage, Weathering and Conservation, HWC2006, 2124 June 2006, Madrid, Spain. Taylor & Francis, London, Vol. 1, pp.
245251.
Stephenson, W.J., Kirk, R.M., 2000a. Development of shore platforms on Kaikoura
Peninsula, South Island, New Zealand: Part One: The role of waves.
Stephenson, W.J., Kirk, R.M., 2000b. Development of shore platforms on Kaikoura
Peninsula, South Island, New Zealand: II: The role of subaerial weathering.
Sumner, P.D., Hedding, D.W., Meiklejohn, K.I., 2007. Rock surface temperatures in
southern Namibia and implications for thermally driven physical weathering.
Sunamura, T., 1996. A physical model for the rate of coastal tafoni development.
Journal of Geology 104(6), 741748.
Takahashi, K., Suzuki, T., Matsukura, Y., 1994. Erosion rates of a sandstone used
for a masonry bridge pier in the coastal spray zone. In: Robinson, D.A.,
Williams, R.B.G. (Eds.), Rock Weathering and Landform Evolution. Wiley
International, Chichester, pp. 175192.
Tingstad, A., 2008. Simulation of salt weathering in a closely replicated coastal
environment. Geografiska Annaler Series A 90(2), 165171.
Trenhaile, A.S., 1987. The Geomorphology of Rock Coasts. Clarendon Press,
Oxford, 384 pp.
Trenhaile, A.S., 2000. Modeling the development of wave-cut shore platforms.
Marine Geology 166(14), 163178.
Trenhaile, A.S., 2004. Modeling the effect of tidal wetting and drying on shore
platform development. Journal of Coastal Research 20(4), 10491060.
Trenhaile, A.S., 2008. Modeling the role of weathering in shore platform
development. Geomorphology 94(12), 2439.
Trenhaile, A.S., Kanyaya, J.I., 2007. The role of wave erosion on sloping and
horizontal shore platforms in macro- and mesotidal environments. Journal of
Coastal Research 23(2), 298309.
Trenhaile, A.S., Rudakas, P.A., 1981. Freeze thaw and shore platform development in Gaspe Quebec. Geographie Physique et Quaternaire 35(2),
171181.
244
Coastal Weathering
Vale, J.A., 1991. Atmospheric deposition of heavy metals to the Severn Estuary.
Ph.D. thesis, University Stirling, unpublished, 345 pp.
Viles, H., 1995. Ecological perspectives on rock surface weathering: towards a
conceptual model. Geomorphology 13(14), 2135.
Wakefield, R.D., Jones, M.S., 1998. An introduction to stone colonizing microorganisms and biodeterioration of building stone. Quarterly Journal Engineering
Geology and Hydrology 31(4), 301313.
Warren, J., 1999. Evaporites Their Evolution and Economics. Blackwell, Oxford,
438 pp.
Wells, N., 1997. The Atmosphere and Ocean, 2E. John Wiley, Chichester, 394 pp.
White, I.D., Mottershead, D.N., Harrison, S.J., 1992. Environmental Systems: An
Introductory Text 2E. Chapman & Hall, London, 616 pp.
Williams, R.B.G., Robinson, D.A., 1981. Weathering of sandstone by the combined
action of frost and salt. Earth Surface Processes and Landforms 6(1), 19.
of salts a review. Permafrost and Periglacial Processes 2(4), 347353.
Williams, R.B.G., Robinson, D.A., 2001. Experimental frost weathering of sandstone
by various combinations of salts. Earth Surface Processes and Landforms 26(8),
811818.
Winkler, E.M., Wilhelm, E.J., 1970. Salt burst by hydration pressures in
architectural stone in urban atmosphere. Geological Society of America Bulletin
81(2), 567572.
Woodcock, A.H., 1953. Salt nuclei in marine air as a function of altitude and wind
force. Journal Meteorology 10, 362371.
Yatsu, E., 1988. The Nature of Weathering: An Introduction. Sozosha, Tokyo, 624 pp.
Geology 15(10), 962966.
Biographical Sketch
Derek Mottershead obtained a PhD in geomorphology from University of London Kings College in 1972. He
became interested in rock weathering studies with a coastal focus in the late 1970s and has published a significant
number of papers in this field. He has observed landforms in a range of environments in Norway, Spain, Malta,
Mallorca, New Zealand, Canada, and USA, in addition to his home field area of southwest England. He is also a
coauthor of a systems-based undergraduate text on physical geography. He has been a longstanding committee
member of the British Geomorphological Research Group (now British Society for Geomorphology). He is
currently Visiting Research Fellow at University of Portsmouth, UK.
4.14 Chemical Weathering in Cold Climates

JC Dixon, University of Arkansas, Fayetteville, AR, USA
4.14.1
4.14.2
4.14.2.1
4.14.2.2
4.14.2.3
4.14.2.4
4.14.2.5
4.14.2.6
4.14.2.7
4.14.2.8
4.14.3
4.14.4
4.14.5
4.14.6
4.14.7
4.14.8
References
Introduction
Solution
Hydrolysis
Hydration
Ion Exchange
Oxidation/Reduction
Carbonation
Chelation
Controls on Rates of Chemical Weathering
Bedrock Weathering
Rock Coatings
Soil Development in Cold Climates
Chemical Weathering in Glacial and Proglacial Environments
Chemical Denudation in Arctic and Alpine Environments
Conclusions
Glossary
Alpine Mountain zone above the limit of tree growth.
Aluminum glaze The concentration of aluminum at the
Earths surface in the form of thin rock surface coatings.
Arctic Latitudinal climatic zone beyond 66.51 N.
Commonly, but not always, beyond the limit of northern
tree growth.
Carbonation The reaction of carbon dioxide dissolved in
water with rocks and minerals of the Earths crust. Especially
affects carbonate minerals and Ca-rich feldspars.
Chelation Reaction between largely immobile metal ions
and organic complexing agents.
Chemical weathering Chemical and mineralogical
alteration of rocks and minerals of the Earths crust
to secondary minerals more stable at the Earths
surface.
Denudation Vertical lowering of the Earths surface by the
combined effects of physical and chemical erosion.
Hydration Incorporation of water into the crystal lattice
of minerals with no chemical reaction.
246
246
246
246
247
247
247
247
247
247
248
249
250
253
253
254
254
Hydrolysis Reaction between minerals and hydrogen ions

split from water molecules.
Ion exchange The substitution of ions in solution for
those on mineral grains.
Iron films Thin rock coatings composed of iron oxides
and oxyhydroxides. Differ from desert varnish in containing
little clay.
Oxidation/reduction Reactions involving the transfer of
electrons from one surface to another as a result of change
in charge. Oxidation is the loss of electrons and reduction is
the gain of electrons.
Periglacial processes Frost dominated processes
operating in cold nonglacial climatic regimes.
Silica glaze Clear white to orange glaze with a shiny luster
composed principally of amorphous silica and aluminum
with variable amounts of iron.
Solution The dissociation of molecules into their
constituent anions and cations.
Sulfate crust Rock coating composed of a diversity of
sulfates including gypsum and barite.
Abstract
Chemical weathering processes in cold climates have traditionally been viewed as subordinate in kind, intensity, and rate
compared to mechanical weathering processes. Empirical work by Jackli (1956) and Rapp (1960), however, demonstrated
that in fact chemical denudation in arctic and alpine environments exceeded mechanical denudation. Since that time a
growing body of research has shown that in fact chemical weathering processes operating in cold climates are the same as
Dixon, J.C., 2013. Chemical weathering in cold climates. In: Shroder, J.

Press, San Diego, CA, vol. 4, Weathering and Soils Geomorphology, pp.
245257.
http://dx.doi.org/10.1016/B978-0-12-374739-6.00063-4
245
246
Chemical Weathering in Cold Climates

those operating in milder climatic regimes and that the intensity of weathering as reflected in weathering products and soil
formation is also as great as in milder climates. Chemical denudation rates approximate those of milder climates and
frequently exceed those of mechanical denudation.
4.14.1
Introduction
Chemical weathering has long been believed to be a relatively

insignificant geomorphic process in cold climates due to the
combination of low temperatures, limited liquid moisture, and
associated lack of vegetation. Chemical reaction rates have
largely been viewed to be temperature driven. This view has
persisted despite the fact that Tamm (1924) clearly established
that chemical reactions at prevailing Earth surface temperatures
are not inhibited in the coldest climates. This view however has
persisted in geomorphology as a consequence of the emphasis
placed on freeze-thaw weathering by ozinski (1909, 1912) in
his definition of the periglacial realm and is emphasized in the
classic work of Peltier (1950) who deductively reasoned that as
temperature decreases chemical weathering in cold climates
becomes progressively less significant.
Beginning in the 1960s geomorphologists and geochemists
began to realize that there was considerable evidence for the
operation of chemical weathering in arctic and alpine environments. This change of perspective was based on empirical
measurements and the realization that in fact temperature is
only one, and a subordinate one at that, control on rates of
chemical weathering. In fact more significant is the availability
of moisture (specifically H ion availability) in driving chemical reactions. Second, there is an abundance of organic acids
available in association with arctic and alpine plants that facilitate the migration of even the most resistant cations as
chelates. Amongst the earliest geomorphologal studies to
demonstrate the important role of chemical weathering in
landscape denudation in alpine environments was Jackli (1956)
who demonstrated that solute transport accounted for greater
sediment transport than all other hillslope processes and glacial transport. Similar results were obtained by Rapp (1960)
from his detailed study of hillslope processes in the arctic alpine
environment of northern Sweden. This study was seminal in
that it showed that with respect to hillslope denudation, solute
transport was not only substantial but dominant. Since the
work of Rapp there has been an ever increasing body of literature that has shown the dominance of chemical weathering over
physical weathering in cold climate landscape denudation.
Following the work of Jackli and Rapp empirical studies by
Reynolds (1971) and Reynolds and Johnson (1972) clearly
established that primary minerals weather into secondary clay
products in alpine environments and that meltwater streams
draining different lithologic environments had distinctive
compositions controlled by bedrock composition rather than
by precipitation, aeolian infall, or evaporation. Importantly,
stream waters draining different watersheds had total ionic
concentrations that were significantly greater than that of incoming precipitation.
These four seminal papers established that chemical weathering in cold climates was a reality and that this set of
processes collectively could be responsible for substantial and
significant landscape denudation. The most important aspect
of cold climate chemical weathering is that associated solutes
are transported in a highly efficient manner from the drainage

basin and runoff waters from snowmelt and glacial ablation
transport exceptionally large quantities of dissolved mass.
This chapter considers the nature of chemical weathering
processes operating in cold climates, controls on chemical
weathering, bedrock weathering in cold climates, glacial
foreland weathering, arctic/alpine soil formation, and geochemical landscape denudation.
4.14.2
4.14.2.1

Solution
Solution, or more precisely dissolution, refers to the dissolving

of minerals by water and the removal of constituent atoms
from the mineral surface (Dixon, 2004; Olson, 2004; Yadav
and Chakrapani, 2006; Anderson and Anderson, 2010). Dissolution may be either congruent or incongruent. Congruent
dissolution occurs when a mineral is separated into its constituent ions as a result of its reaction with water, whereas
incongruent dissolution involves the formation of secondary
mineral products (Berner and Berner, 1997; Blum and Stillings, 1995; Yadav and Chakrapani, 2006; Anderson and
Anderson, 2010). The most common examples of congruent
dissolution are the weathering of calcite into aqueous calcium
and bicarbonate and quartz into aqueous silicic acid. The
weathering of olivine is also an example of pure congruent
dissolution. Mineral dissolution has been widely reported
from cold climates. Predominantly congruent dissolution, in
which mineral surfaces simply dissolve and weathering pits or
porosity develop, is widely reported from cold climates. Dixon
et al. (2002a, b) reported extensive dissolution pit development on a diversity of mineral surfaces as well as along mineral grain boundaries from arctic Sweden and Norway,
respectively. Similarly extensive grain dissolution was reported
by Dixon (1983) from soils in the Front Range of the Colorado Rocky Mountains and by Dixon et al. (1984) from the
Juneau Icefield of Alaska. The resulting increase in mineral
grain porosity facilitates a feedback effect which leads to increased weathering rates and ultimately rock disintegration.
4.14.2.2
Hydrolysis
Hydrolysis, or incongruent dissolution, involves the reaction of

minerals with water to produce a weak acid or base. When the
water reacts with the mineral surface it is split into hydrogen
ions and hydroxyl anions. H replaces other cations in the
mineral structure and disrupts structural bonds thus weakening
the mineral structure (Olson, 2004). The displaced cations
react with hydroxyl ions in adhered water on the mineral
surface and form secondary clay minerals (Blum and Stillings,
1995). Common hydrolysis weathering reactions include the
formation of kaolinite or illite from orthoclase feldspar, kaolinite or smectite from albite, and kaolinite from anorthite.
These reactions and accompanying clay mineral formation are

widely reported from both arctic and alpine environments.
Bockheim (1979) and Locke (1979) both reported clay mineral formation in soils from arctic Canada. The formation of
iron-rich montmorillonite from the weathering of dolerite has
been reported by Claridge and Campbell (1984), as well as the
formation of vermiculite from the weathering of mica
(Campbell and Claridge, 1987) in Antarctica. Ellis (1980) reported smectite formation in soils from Arctic-Alpine Norway
whereas Darmody et al. (2005) reported clay mineral formation from micas in alpine Norway. Extensive investigations
of clay mineral formation have been undertaken recently in the
Italian Alps by Egli and coworkers who reported widespread
development of soil smectites (Egli et al., 2003a, 2004, 2007).
4.14.2.3
Hydration
Hydration results in the addition of water molecules to a

mineral structure but without accompanying dissociation occurring in hydrolysis. In short there is no reaction between
water and mineral, rather water is simply added to the structure of the mineral. A simple example is the formation of
gypsum as a result of the addition of water to anhydrite. In
addition, some clay minerals are especially susceptible to hydration as water molecules enter and leave interlayer locations
such as montmorillonite (Olson, 2004). Although principally
an expansive rather than chemical process hydration does
produce secondary minerals such as hydrobiotite which
commonly occur in arctic (Isherwood, 1975; Watts, 1981) and
alpine (Dixon, 1983; Dixon et al., 1984) soils.
4.14.2.4
Ion Exchange
Clay minerals are also especially affected by ion exchange as

they have a strong capacity for exchange within interlayers and
with surface ions. During exchange of ions, the basic structure
of the clay mineral is unchanged, but the interlayer spacing
changes depending on the size of the ions entering interlayer
positions. As a result it is possible for one clay mineral to
transform to another clay mineral such as in the case of illite
transforming to smectite (Olson, 2004). In cold climates one
of the most widespread ion exchange reactions involves the
weathering of biotite to vermiculite or montmorillonite. This
reaction involves the loss of interlayer K and adsorption of Na
and Ca. Vermiculite formation has been widely reported from
both arctic and alpine environments.
4.14.2.5
247
oxidation of Fe-rich minerals has been widely reported from

the Canadian Arctic by Isherwood (1975), Locke (1979),
Watts (1981) and from Antarctica by Kelly and Zumberg
(1961), Ugolini (1986), Campbell and Claridge (1992).
Widespread iron oxidation, especially of pyrite, is reported
from Arctic-Scandinavia (Dixon et al. 2002a, b); Darmody
et al., 2007) and from Alaska (Anderson et al., 2003).
4.14.2.6
Carbonation
Carbonation is a chemical weathering process involving the

reaction of carbon dioxide dissolved in water with minerals. It
involves the acid hydrolysis of both carbonates and silicates
due to CO2 dissociation. In subglacial settings the process is
restricted due to the limited availability of atmospheric gases
in subglacial waters. This process is integral in the weathering
of carbonate rocks, especially limestone and related rocks. The
reaction is fundamentally one of carbonic acid and calcium.
Similarly carbonation is fundamental initial process in the
weathering of Ca-rich aluminosilicates such as plagioclase
feldspar, a process widely reported from glacial and proglacial
environments (Tranter, 2005). Carbonation of limestone and
dolomite has been widely reported from cold climates (Ford,
1979, 1996; Anderson et al., 2003; Egli et al., 2008) where
low temperatures substantially enhance carbonate dissolution
rates.
4.14.2.7
Chelation
Chelation is the reaction between a metal ion and an organic

complexing agent resulting in the formation of a ring structure
that encompasses the metal ion and removes it (Olson, 2004).
Dissolved organic molecules bind to metals in solution or on
the surface of minerals. A molecule on an organic compound
that can form these associations is referred to as a ligand and
the ligandcation or ligandmetal association is called a coordination complex (Anderson and Anderson, 2010). The formation of coordination complexes on mineral surfaces weakens
the bond of the metal or cation to the crystal lattice and facilitates dissolution (Brantley, 2008). The important role played
by acid chelation in soil development in Arctic Canada has been
examined by Lev and King (1999) on Devon Island, Nunavut,
Canada. Extensive spodosol development is reported from
arctic-alpine environments of Scandinavia where chelation is a
dominant process in the leaching of Fe and Al to form strongly
developed E horizons (Ellis, 1980; Darmody et al., 2000).
Oxidation/Reduction
Oxidation and reduction reactions involve the transfer of

electrons between atoms. Oxidation involves loss of electrons
whereas reduction involves gain of electrons. Chemical
elements must have at least two viable oxidation states to be
involved in redox reactions. Fundamentally oxidation is the
reaction of minerals with oxygen in the presence of water. The
most common redox reactions in weathering in the natural
environment involve iron and sulfur. Ferrous iron is oxidized
to ferric iron and the oxidation of sulfide to sulfate (Dixon,
2004; Olson, 2004, Anderson and Anderson, 2010). Extensive
4.14.2.8
Controls on Rates of Chemical Weathering
There are several major controls on the rate at which minerals

weather and release ions into the natural environment. These
controls are temperature, pH, biological, and surface morphology. The relationship between weathering rate and temperature is captured in the Arrhenius equation. In essence this
equation establishes that higher the temperature, or lower the
activation energy (energy required for a chemical reaction to
occur), faster the reaction occurs. For instance an increase in
temperature from 01 to 25 1C results in an increase in the
248
weathering rate by a factor of about 10 (Anderson and

Anderson, 2010).
Most silicate minerals dissolve at greater rates at higher and
lower pH with a region of relatively low solubility in the neutral
pH range. At lower pH, due to proton (H )-promoted dissolution the rate increases exponentially with increasing hydrogen ion concentration. Above some transition pH, in the range
of 4 or 5, dissolution rates are generally independent of pH. At
higher pH, generally above 8, hydroxyl-promoted dissolution
occurs and rates go up with increasing pH. The total dissolution
rate is expressed as the sum of the individual rates (Yadav and
Chakrapani, 2006; Brantley, 2008; Anderson and Anderson,
2010). Protons and hydroxyls adsorb to cations on mineral
surfaces, which weakens the bonds between the cation and the
crystal lattice. This accelerates the release of cations into solution
at a rate that is proportional to the concentration of H or OH
in a nonlinear fashion (Anderson and Anderson, 2010).
Soil microorganisms and plants produce a variety of acids
including carbonic, sulfuric, and humic acids. These acids
provide hydrogen ions that replace cations on mineral surfaces
and lead to the mineral disintegration. As outlined previously,
organic ligand adsorption accelerates the dissolution rates of
silicate minerals by forming organometallic complexes on
mineral surfaces, especially in association with Al and Fe,
thereby weakening bonds between the metal and the solid.
The ligand-promoted rate is proportional to the concentration
of adsorbed ligand, which is a function of the activity in the
solution (Yadav and Chakrapani, 2006).
The rate of dissolution of silicate minerals is also controlled by the surface morphology of the crystal (Hochella and
Banfield, 1995; Yadav and Chakrapani, 2006). Crystal surfaces
are seldom uniform but rather display a variety of imperfections called dislocations in which one part of the crystal is
offset from another. Dislocations are rows of atoms that are
slightly displaced and are therefore highly energetic. The rate
of dissolution is regulated by the type and density of dislocations on the crystal surface. The two principal types of
dislocations are kinks and steps. Kinks have three sides exposed and steps have two, whereas planar crystal surfaces have
only one. During dissolution, atoms with highest energy detach most rapidly from the surface, so kinks are favored location for both focused dissolution as well as the adsorption
of species from solution. As a result of selective dissolution,
etch pits form on the mineral surfaces, which reflect the crystal
structure of the underlying mineral and therefore are regular
in shape and aligned in specific directions Yadav and Chakrapani (2006).
4.14.3
Bedrock Weathering
Bedrock weathering, as reflected in the development of weathering rinds (Colman, 1982; Oguchi, 2001), is widely expressed in alpine environments (Figure 1). Weathering rind
thicknesses on regolith boulders in a snowpatch setting in the
Colorado Front Range have been reported with thicknesses up
to 3 mm compared to 0.5 mm on adjacent control sites
(Thorn, 1975, 1976). To the southwest of Denver in the alpine
zone of Mt. Sopris, Birkeland (1973) reported rind thicknesses
of up to 45 mm. The weathered rock displays strong oxidation
Figure 1 Weathering rind developed on schist from Karkevagge,

Swedish Lapland. The rind is dominated by iron accumulation with a
thin bleached zone at the weathered/fresh rock boundary.
of ferromagnesian minerals with enhanced iron III contents

compared to that of fresh material (Dixon and Thorn, 2005).
Weathering rind formation with thicknesses of up to 3 mm
within 100 years was reported from Iceland by Etienne (2002).
Geochemical investigations of weathering rinds on bedrock debris from the alpine environment of Karkevagge,
Swedish Lapland by Dixon et al. (2002a) reveal extensive iron
oxidation as well as absolute losses of alkalis, alkali earths,
and silicon. Mineral dissolution results in the development of
extensive porosity that stands in marked contrast to that of the
underlying fresh rock (Figure 2). Significantly, the weathering
rinds lack the formation of secondary clay minerals suggesting
effective flushing of dissolved weathering products.
Investigations of the nature of chemical weathering associated with the development of rinds on cobbles of ultramafic
rocks in the Jotunheimen of Norway reveals rapid onset of
weathering following deglaciation. Significant porosity development is observed in buried cobbles after 70 years of landscape exposure. After 190 years surface lichen covered rocks
display significant porosity development compared to lichenfree surfaces on exposed cobbles (Thorn et al., 2007). This
study was based on analysis of plagioclase grains in which
porosity is associated with strongly oriented voids controlled
by crystal structure. Clearly, availability of moisture (in the
case of buried cobbles) and organic acids (in the case of lichen-covered surfaces) are significant controls on chemical
weathering in these environments.
Similar observations were made from incipient weathering
rind development on granite disks buried in soil profiles in
Karkevagge Swedish Lapland. Thin weathering rinds on the
249
Norway was indicated previously. In Iceland, Etienne (2002)

has demonstrated the substantial role that bacteria play in the
rapid development of weathering rinds. Dixon et al. (2002a)
also noted the presence of substantial quantities of organic
material incorporated in weathering rinds but were unable to
determine its exact nature.
4.14.4
Figure 2 Backscatter scanning electron micrograph of weathering

rind from Karkevagge, Swedish Lapland. The weathering rind is
characterized by an extensive development of porous structure. The
overall porosity of the weathering rind is 4.4% whereas the fresh
rock interior is 0.6%.
outer edges of the granite discs develop over a period of time as

short as 4 years. The initial stage of rind development involves
the development of porous structure consisting of a combination of pits and fractures that have been solutionally enlarged
and modified. Solution is preferentially concentrated on the
surface of both feldspars and quartz. Strong orientation of the
solutionally enlarged voids suggests strong structural control by
feldspar crystallography. Additionally, iron oxides are present
along grain boundaries and mineral interiors and are interpreted
to have been derived from the oxidation of ferromagnesian
minerals. Spatial variability in weathering rind development is
strongly driven by moisture availability (Dixon et al., 2006).
A 10-year study of weathering rates of potential weathering
of limestone, dolomite, and granite disks in the same location,
as in the above study, reveal mass losses of up to 25% for
limestone, 14% for dolomite, and 0.3% for granite in the
wettest sites (Thorn et al., 2006a, b). Mass loss rates were
greater in the second 5-year period when temperature and
moisture were both greater than in the first 5-year period.
Significantly, when the rates of limestone disk mass loss are
compared to mass loss rates for the same rock material in a
variety of other climatic settings (Gams, 1985; Thorn et al.,
2006a, b). The rates for the wettest sites are an order of
magnitude greater than for any other climatic environment.
For the driest sites mass loss rates are of a comparable magnitude to those in tropical, arid/semiarid, and Mediterranean
climates. A companion study of surface pebble weathering
rates of dolomite and granite revealed similar results to the
disk study with dolomite consistently weathering faster than
granite. Over the 10-year period dolomite experienced a
1.910% mass loss and granite experienced a 0.74.02% loss.
These results compare similarly with those obtained by Caine
(1979) from the San Juan Mountains of Colorado. As with the
disks the mass loss was greater in the second 5-year period
compared to the first.
Biological processes have also been identified as playing a
significant role in bedrock weathering in cold climates. The
apparent importance of lichens in bedrock weathering in
Rock Coatings
Coatings of diverse chemical composition are widespread on

rock surfaces in cold climates and represent the accumulation of
transported solutes under disequilibrium conditions. Amongst
the most widespread coatings are those dominated by Fe, Si, Al,
Ca, and K (Figure 3) giving rise to iron films, silica glaze, aluminum glazes, and sulfate crusts, respectively (Dorn, 1998;
Dixon et al. 2002a, b). These coatings, resulting from both inorganic and organically mediated weathering processes, occur
extensively in both periglacial and glacial settings.
In Swedish Lapland, Dixon et al. (2002a) suggested that
iron films are primarily the result of the oxidation of pyrite
and short distance transport of iron which is reprecipitated as
goethite. They have also suggested that bacterial-facilitated
precipitation may also occur (Dixon et al., 1995), which
subsequent investigations have confirmed. Iron films have
been widely reported in association with subglacial meltwaters
(Andersen and Sollid, 1971; Drewry, 1986; Whalley et al.,
1990). These films are also widely reported from the polar
desert environment of Antarctica (Dorn, 1998).
Silica glazes are widely reported from a variety of cold
climate settings including glacier margin and subglacial locations (Hallet, 1975; Peterson and Morseby, 1979). They are
also reported from a variety of contexts in Antarctica (Dorn,
1998). Dixon et al. (2002a, b) reported the occurrence of silica
glazes on isolated boulders in Karkevagge, Swedish Lapland.
The source of the silica is principally from silicate mineral
dissolution in local bedrock (Tranter, 2005) or in the case of
Antarctica from the dissolution of airborne quartz and clay
(Dorn, 1998).
A preliminary study of rock coatings in Karkevagge,
Swedish Lapland (Dixon et al., 2002a, b), revealed the presence of aluminum glazes in association with iron films. A
more focused study of rock coatings in Karkevagge by Darmody et al. (2007) revealed valley wall coatings to be overwhelmingly dominated by aluminum. The coatings are
predominantly an amorphous aluminum oxyhydroxide sulfate (basaluminite) derived from aluminisilicate dissolution
by waters associated with pyrite oxidation.
In Karkevagge, where rock coatings are protected from
seasonal snowmelt and summer rain, the dominant rock
coatings are gypsum and jarosite. The presence of these minerals further attests to the significant role played by waters
associated with pyrite oxidation in the weathering of Ca- and
K-rich minerals in the valley walls. Ca-rich coatings on gravels
have been widely reported from subglacial environments
(Hallet, 1976; Watts, 1985), and from alpine environments in
the Canadian Arctic (Bunting and Christensen, 1978). Extensive gypsum coatings in Antarctica have been attributed to
redeposition of Ca derived from weathering of carbonate cement in the local bedrock (Meiklejohn and Hall, 1997).
250
(a)
(b)
(c)
(d)
Figure 3 Backscatter scanning electron micrographs of chemical rock coatings, Karkevagge, Swedish Lapland. (a) Smooth surface of iron
coating, (b) Al-dominated basaluminite coating, (c) gypsum coating, and (d) gypsum crystals on jarosite coating.
4.14.5
Soil Development in Cold Climates
Soil formation is one of the strongest expressions of the effectiveness of chemical weathering of bedrock and regolith. In
cold climates soil formation is characteristically most strongly
expressed on regolith derived from glacial, periglacial, and
colluvial processes. Arctic soils have typically been characterized as being dominated by cryogenic processes with limited
chemical weathering (Rieger, 1974; Tedrow et al., 1958; Hofl
et al., 1998) and alpine soils by limited or no profile development, frozen subsoils, little chemical weathering, and limited organic matter (Kubiena, 1953). More recently, with
increasing interest in the impact of climate change in arctic
and alpine environments, there has been renewed interest and
investigation of cold climate soils.
Alpine soil formation is highly variable due to topographic, climatic, organic, and parent material changes over
short distances (Retzer, 1965; Burns, 1990; Legros, 1992).
Detailed studies of alpine soils from the Alps, the Rockies, the
Jotunheimen, and the Scandes all point to relatively rapid and
strong chemical weathering under favorable conditions. Soil
formation encompasses the full spectrum of development
from young recently formed soils such as inceptisols and
entisols to more complex soil orders such as spodosols
(podsols) and mollisols (Figure 4).
Investigations in the Front Range of the Rocky Mountains of

Colorado (Birkeland et al., 1987; Dixon, 1983; Burns and
Tonkin 1982) reveal strong soil development in glacial and
periglacial deposits over extremely short periods of geologic
time. Soils developed on deposits younger than 12 000 years
BP. With increasing age, soils consistently show increasing
amounts of organic carbon, total nitrogen, organic bound
phosphorus, clay, extractable iron and aluminum, and higher
cation exchange capacities (Birkeland et al., 1987). These trends
are reflected in the development of progressively more complex
soil profiles with the youngest soils displaying A/Cox profiles
and the oldest soils displaying A/Bw (Bt)/Cox horizons.
In addition to the overall trend of increasing clay in progressively older soils there are clear patterns of alteration of the
clay minerals indicating strong chemical weathering. Dixon
(1983, 1986) found kaolinite to be the dominant mineral in
the soils which decreased in abundance with increasing soil
development. He identified the alteration of biotite to
hydrobiotite, vermiculte, and smectite (Figure 5). Mahaney
(1974) identified the alteration of illite to chlorite, montmorillonite, and mixed layer illitemontmorillonite. Shroba
and Birkeland (1983) similarly identified the alteration of
mica to mixed layer clays, but subsequently (1987) and correctly pointed out the challenge of separating weathering from
aeolian addition.
Figure 4 Spodosol soil profile, Karkevagge, Swedish Lapland

showing strong development of leached E horizon and accumulation
of iron at depth.
Dixon (1983, 1986) reported extensive loss of silicon, alkalis, and alkali earths relative to Fe and Al as well as the
oxidation of FeII to FeIII compared to parent material (Cn)
horizons in soils developed on glacial and periglacial deposits
in the Front Range of the Colorado Rocky Mountains. He
identified the extensive dissolution of Ca/Na plagioclase,
quartz, biotite, and K-feldspar. These trends are consistent with
the progressive increase in CEC (cation exchange capacity) of
soils with increasing age.
A study of soils in the Alps by Legros (1992) similarly
revealed strong evidence of chemical weathering. Soils developed on calcshale (mica, chlorite, and calcite) with some
FeS2 display calcite dissolution, sulfide oxidation, and phyllosilicate formation. Weathering profiles are typically greater
than 23 m thick. Dissolution of the calcite results in the
production of a phyllosilicate and quartz residuum. In limestone terrains there is strong limestone dissolution and associated formation of rendzinas and calcic brown soils. Soils
developed on crystalline rocks, particularly granites, gneiss,
and mica schist are typically entisols and spodosols. Weathering is extensively expressed in the grussification of
the crystalline rocks with feldspars showing evidence of dissolution and clay formation. The common clay formation
pathways are chlorites and micas transforming to interstratified minerals, vermiculite, and smectite and kaolinite
forming from feldspars (Legros, 1992). Soils developed on
volcanic rocks display strong altitudinal zonation. At highest
251
elevations andosols with strong podsolic characteristics (typic

haplohumods) develop. At intermediate elevations strongly
developed andosols are developed. At low elevations, andic
properties are less strongly developed and the soils are dominated by andic haplumbrepts. At the foot of volcanoes the
typical soil order is inceptisols and vertisols. Soils developed
on ultramafic rocks are predominantly inceptisols or where
aeolian additions are abundant spodosols have developed.
Investigation of spodosols in the Italian and Swiss Alps over
the past decade similarly demonstrate rapid and intense pedogenesis (Mirabella et al., 2002; Egli et al., 2003a, b, 2004,
2007, 2008). Although variations in the intensity of weathering
and soil formation were noted with respect to elevation, aspect,
parent material, and vegetation cover several recurring trends
are noteworthy. All soils demonstrated varying losses of Ca, Mg,
K, Na, Si, and Mn relative to parent material contents but
were generally greater at higher altitudes. Si losses are particularly high suggesting intense weathering, especially in the
subalpinealpine transition. Fe and Al eluviation occurred in all
soils, but was again uniformly high. The dominant clay mineral
was smectite derived from the weathering of chlorite and mica.
Intermediate transformation minerals include interstratified
mica-vermiculite and vermiculitesmectite. Minor amounts of
kaolinite occurred and were attributed to the weathering of
feldspars. Significantly, the strong spodosol development and
associated weathering summarized in this body of research
occurred in Holocene age soils.
Soil development is also strongly expressed in surficial
deposits in the Scandes Mountains of northern Sweden and in
the Jotunhemen of southern Norway. In Karkevagge, Swedish
Lapland, six different soil orders were identified depending on
their location in the valley with respect to landscape stability,
elevation, and vegetation cover. At low elevations on
stable landscape surfaces with covers of birch and empetrum
heath, haplocryods occur. These soils display strong base and
Al and Fe leaching. Entisols dominate the valley bottom and
side slopes. Inceptisols are the dominant residual soil at high
ridge-top locations. These soils display some evidence of Ca
loss. Mollisols develop in thick colluvial deposits on steep
slopes. Histosols occur in isolated bedrock depressions in the
valley bottom. The rapid formation of soils on Holocene deposits is facilitated by a combination of coarse textured parent
material, abundant snowmelt water in the spring, and especially the presence of pyrite in the soil parent materials
(Darmody et al., 2000). The soils display a wide variety of clay
minerals. The dominant clay-size mineral is muscovite that is
derived from the local bedrock, with subordinate amounts of
chlorite. Secondary clay minerals associated with the weathering of these primary clay minerals include a variety of
interstratified minerals including mica-smectite, chloritesmectite, and corrensite (Allen et al., 2001).
Soil development in the Jotunheimen of southern Norway is
overall generally slow with the soils being predominantly
cryorthents (Darmody et al., 2005). However, over a period of
some 200 years through an elevation range of some 300 m
discernible trends in soil development are observed. Generally,
extractable Fe and Al, cation exchange capacity, organic carbon,
and the C/N ratio increase with time and pH, base saturation,
and extractable cations decrease with time. Extractable cations
tend to increase with elevation. Secondary clay minerals,
252
m
m
c
c
i
c
b
c
d
(a) Soil A6A
(b) Soil SM22C
12
15
12
15
m
m
a
a
c
c
c
b
c
(c) Soil DH2C2
(d) Soil B12BC
12
15
12
15
Figure 5 XRD spectra showing secondary clay minerals resulting from the weathering of mica. C, chlorite; I, interstratified mica-smectite; m,
mica. Pretreatments (left side of spectra) (a) heated to 550 1C, (b) ethylene glycol solvated, (c) K-saturated, (d) Na-saturated and glycerol
solvated.
especially hydrobiotite increase with both age and elevation

(Darmody et al., 2005). Overall, however, secondary clay mineral formation is greatest in wet sites (Darmody et al., 1987).
Dominant clay minerals include vermiculite, kaolinite, and
smectite and are derived from the weathering of the pyroxenegranulite gneiss. Earlier soil research in the region had identified
the weathering of biotite to hydrobiotie on older landscape

surfaces (Mellor, 1985, 1987) and the tendency for soil profile
thickening, increasing organic carbon content and cation exchange capacity, and decreasing pH with soil age (Messer, 1988).
Soil formation and associated chemical weathering in
Antarctica have received considerable attention since the
twentieth century, evolving strongly in the 1960s (Tedrow and

Ugolini, 1966) and continuing until the present time. In their
classic study of chemical weathering Kelly and Zumberg
(1961) concluded that because of the exceptionally low temperatures chemical weathering is essentially nonexistent except
for limited iron oxidation. This view was subsequently challenged by the work of Ugolini and Anderson (1973), Ugolini
and Jackson (1982), Campbell and Claridge (1987) who have
continued to demonstrate that despite low annual temperatures and spatially and temporally restricted water availability
soil formation and associated chemical weathering is widespread (Bockheim, 2002; Bockheim and Wilson, 1992).
In favorable edaphic settings soils develop deep profiles
commonly approaching 2 m. Chemical weathering processes
are dominated by iron oxidation, however analysis of soil
solutes also reveals that some constituents are derived from
rock weathering. Clay mineral formation is also observed.
Common clay minerals include mica and illite, but under
favorable moisture conditions species including hydrous
micas, vermiculite, and hydroxyl interlayered vermiculite
(Campbell and Claridge, 1992).
4.14.6
Chemical Weathering in Glacial and

Proglacial Environments
Extensive investigations of glaciated terranes reveals that

chemical weathering rates in such environments is often
equivalent to or greater than continental averages (Sharp et al.,
1995; Anderson et al., 1997, 2000; Wadham et al., 1997;
Hodson et al., 2000; Tranter, 2005). This counterintuitive
observation results from the fact that glaciated catchments are
characterized by very high runoff, high concentrations of fresh
rock and mineral faces on small silt-sized rock flour, and
general lack inorganic and organic rock surface precipitates
that inhibit weathering (Tranter, 2005). The total flux of glacial solutes is generally associated with high discharges and
dilute waters (Anderson et al., 2000; Tranter, 2005). Glacial
runoff is generally a dilute Ca, HCO3, SO4 solution with
variable amounts of Na and Cl (Brown, 2002; Tranter, 2005).
The concentration of major ions in glacial runoff is in the
range of 103500 microequivalents (Tranter, 2005).
The chemical composition of glacial meltwaters results
from a series of chemical reactions controlled by kinetics and
microbial activity (Tranter, 2005). The initial reaction that
occurs when the dilute melt interacts with glacial flour derived
from diverse rock types is carbonate and silicate hydrolysis.
This raises pH, lowers the partial pressure of CO2 and maximizes the waters potential to absorb CO2 (Tranter et al., 1997,
2002). Melt water in contact with glacial flour facilitates the
exchange of divalent cations from solution for monovalent
cations on rock surfaces. There is ion exchange of Ca and Mg
for Na and K (Tranter, 2005).
The high pH of melt waters resulting from hydrolysis of Carich rock increases the dissolution of aluminosilicate minerals
as both Al and Si increase in solubility at pH 4 9. Because of
the increased absorption of CO2 by the strongly alkaline
waters, water acidity increases. In addition to acidification by
CO2 adsorption, sulfide oxidation increases acidity. Consequently, subglacial meltwaters are commonly undersaturated
253
with respect to calcium and as a result carbonate dissolution

has a disproportionately large impact on meltwater chemistry
despite its minor abundance (Tranter, 2005).
Sulfide oxidation is the dominant chemical reaction in
subglacial environments. After hydrolysis, this is the most
significant reaction responsible for lowering pH, decreasing
the saturation index of carbonates, and allowing more carbonate dissolution. Sulfide oxidation occurs predominantly in
debris-rich environments where ground-up rock is first in
contact with water. The oxidation has been found to be
microbially mediated and preferentially dissolves carbonates
because this rate is orders of greater magnitude (Tranter et al.,
2002; Tranter, 2005).
Chemical weathering has been shown to be enhanced in
the proglacial zone over subglacial environments (Anderson
et al., 2000; Wadham et al., 2001; Tranter, 2005; Anderson,
2007). Enhanced weathering is to be expected in the proglacial
environment given abundant fragmental debris, potential for
reworking, colonization by vegetation, potential for perturbation of groundwaters, ingress of gases, and potential for
long-term oxic conditions (Tranter, 2005). Several field and
laboratory studies confirm these expectations. Studies of
weathering in forelands in Switzerland and Svalbard found
that solute concentrations and fluxes increased downstream
by 3047%. The increase in flux is due to increased solute
concentration as discharge did not increase the proglacial
reaches. In both of these cases the increase in solute flux is due
to carbonate dissolution and pyrite oxidation. In a detailed
study of chemical weathering on the foreland of the Bench
Glacier in southern Alaska, Anderson et al. (2000) demonstrated that commonly patterns of chemical denudation may
be complex. These workers found that the carbonate dissolution and pyrite oxidation account for 90% of the solute
flux in the youngest sediments, but once these constituents are
leached from the soil total cation denudation decreases rather
than increases. In contrast however, silica denudation rates
increase as silicate weathering increases in the oldest sediments due to progressively increasing vegetation. The increase
in the case of the Swiss study is attributed to dissolution of
suspended sediments in the stream whereas in the case of the
Svalbard study it is attributed to highly mineralized waters
shallow groundwater from the glacial foreland.
4.14.7
Chemical Denudation in Arctic and Alpine

Environments
The importance of chemical denudation in contemporary cold

climate landscape evolution was initially recognized by Jackli
(1956) and Rapp (1960). Since that time these early findings
have been corroborated by subsequent workers on several
continents. Chemical erosion is responsible for not only the
transportation of considerable mass, but also for the removal
of considerable quantities of dissolved material from entire
arctic and alpine drainage basins. Essentially chemical denudation represents complete removal of mass from a drainage basin (excluding vegetative uptake and pedogenesis). The
landscape model of Ahnert (1977) distinguished total removal
as opposed to point-to-point movement as the single most
important discriminator amongst slope processes. In the case
254
of solute removal extremely strong coupling with the drainage

system facilitates rapid and effective removal of dissolved mass
from the drainage basin (Dixon and Thorn, 2005).
Caine and Thurman (1990) and Caine (1992, 1995, 2001)
reported results of a multi-year study from the relatively dry
environment of the Front Range of the Rocky Mountains in
which they found dissolved solid concentrations to be relatively low, but associated rates of geochemical denudation to
be an order of magnitude greater than those of sediment removal from the drainage basin. Although rates of geochemical
denudation were high compared to sediment removal, they
were low compared to other high mountain catchments such
as those reported by Reynolds and Johnson (1972), Eyles et al.
(1982), Collins (1983), and Souchez and Lemmens (1987).
Rates of chemical denudation from numerous catchments
in the Pacific Northwest of the U.S. range from 19100 t
km2 yr1. With a mean of approximately 34 t km2 yr1. In
the Bohemian massif in the Czech Republic, Paces (1986) reported denudation rates of 945 t km2 yr1. In both of these
examples the variability strongly reflects contrasts in lithologies.
Rapps (1960) estimates of 27 t km2 yr1 chemical denudation from Karkevagge, Swedish Lapland were based on few
analyses and a lack of discharge determinations. Subsequent
determinations by Campbell et al. (2002) based on thriceweekly measurements of discharge-normalized TDS at the
drainage basin outlet revealed a chemical denudation rate over
an 82 day melt year of 46 t km2 yr1. In the same general area,
Beylich et al. (2004a, b) and Beylich (2008) reported chemical
denudation rates of 5.4 t km2 yr1 and mechanical erosion
rates of 2.3 t km2 yr1. Although both these rates are lower
than for Karkevagge, especially with respect to chemical denudation, it is important to note that the chemical erosion rate
is double that of the mechanical denudation rate. In contrast,
long-term studies by Beylich (2009) from subarctic Iceland and
Finland reveal chemical denudation rates that are an order of
magnitude greater than those reported from Lapland.
The rates of chemical erosion from these alpine sites, with
the exception of the study by Beylich, are all comparable to the
global average for all lithologies of 45 t km2 yr1 (Meybeck,
1987) and substantially exceed the global average for crystalline rocks in general of 19 t km2 yr1 (Meybeck, 1987). Significantly, these values are also very similar to the largest
catchments in the world such as the Mississippi and the
Amazon at 3545 t km2 yr1.
Chemical denudation rates for glacierized drainage basins
are generally higher than those of nonglacierized (Anderson
et al., 2000). In arctic environments, rates of chemical denudation are strongly tied to the thermal characteristics of the
glaciers. Although catchments dominated by warm-based
glaciers are likely to yield relatively high denudation rates,
those dominated by cold based glaciers are likely to be characterized by low denudation rates (Anderson et al., 2000;
Hodgkins et al., 1997; Wadham et al., 1997).
4.14.8
Conclusions
Despite continuing misconceptions, chemical weathering processes have long been established as important and generally
significant components of the contemporary geomorphic and
pedogenic process suite in cold climates. The processes operating have been established to be fundamentally the same as
those that operate in milder climatic environments. Although
the rates at which they operate may be slowed under the most
extreme conditions of temperature and moisture availability, it
has been established by field experimentation that they can also
operate at extremely rapid rates.
Chemical weathering processes operating in cold environments are of considerable geomorphic significance. They
represent processes that result in solutional mass that is largely
completely removed from the drainage basin and as such result in highly efficient landscape denudation. Significantly,
geochemical denudation generally exceeds mechanical denudation, thus establishing the ongoing need to more fully
understand the complexity of processes limited by moisture,
organic acids, and extended periods of extreme cold.
References
Ahnert, F., 1977. Some comments on the quantitative formulation of
geomorphological processes in a theoretical model. Earth Surface Processes 2,
191201.
Allen, C.E., Darmody, R.G., Thorn, C.E., Dixon, J.C., Schlyter, P., 2001. Clay
mineralogy, chemical weathering, and landscape evolution in Arctic-alpine
Sweden. Geoderma 99, 277294.
Andersen, J.L., Sollid, J.L., 1971. Glacial chronology and glacial geomorphology in
the marginal zones of glaciers, Midtdahlsbreen and Nigardsbreen, south Norway.
Norsk Geografisk Tidsskrift 25, 135.
Chemistry of Landscapes. Cambridge University Press, Cambridge, 637p.
Anderson, S.P., 2007. Biogeochemistry of glacial landscape systems. Annual Review
of Earth and Planetary Sciences 35, 375399.
Anderson, S.P., Drever, J.I., Frost, C.D., Holden, P., 2000. Chemical weathering in
the foreland of a retreating glacier. Geochimica, et Cosmochimica Acta 64,
11731189.
Anderson, S.P., Drever, J.I., Humphry, N.F, 1997. Chemical weathering in glacial
environments. Geology 25, 399402.
Anderson, S.P., Longacre, S.A., Kraal, E.R., 2003. Patterns of water chemistry and
discharge in the glacier-fed Kennicott River, Alaska: evidence for subglacial water
storage cycles. Chemical Geology 202, 297312.
Berner, R.A, Berner, E.K., 1997. Silicate weathering and climate. In: Ruddiman, W.F.
(Ed.), Tectonic Uplift and Climate Change. Plenum Press, New York, pp.
353365.
Beylich, A.A., 2008. Sediment fluxes and sediment budget in Latnjavagge and the
potential for applying unified methods for integrating investigations on sediment
fluxes and budgets in cold environment catchments. Geological Survey of
Norway, Special Publication 11, 111130.
Beylich, A.A., 2009. Chemical and mechanical fluvial denudation in cold
environments: comparison of denudation rates from three catchments in subArctic eastern Iceland, sub-Arctic Finnish Lapland and Arctic Swedish Lapland.
Jokulll 59, 1932.
Beylich, A.A., Kolstrup, E., Thyrsted, T, Gintz, D., 2004b. Water chemistry and its
diversity in relation to local factors in the Latnjavagge drainage basin, arcticoceanic Swedish Lapland. Geomorphology 58, 125143.
Beylich, A.A., Kolstrup, E., Thyrsted, T., Linde, N., Pedersen, L.B., Dynesius, L.,
2004a. Chemical denudation in arctic-alpine Latnjavagge (Swedish Lapland) in
relation to regolith as assessed by radio magnetotelluric-geophysical profiles.
Birkeland, P.W., 1973. Use of relative age-dating methods in a stratigraphic study of
rock glacier deposits Mt. Sopris, Colorado. Arctic and Alpine Research 5,
401416.
Birkeland, P.W., Burke, R.M., Shroba, R.R., 1987. Holocene alpine soils in gneissic
cirque deposits, Colorado Front Range. U.S. Geological Survey Bulletin 1590E,
21 pp.
Blum, A.E., Stillings, L.L., 1995. Feldspar dissolution kinetics. In: White, A.F.,
Brantley, S.L. (Eds.), Chemical Weathering Rates of Silicate Minerals. Reviews in
Mineralogy, vol. 31, pp. 291351.
Bockheim, J.G., 1979. Properties and relative age of soils of southwestern

Cumberland Peninsula, Baffin Island, N.W.T., Canada. Arctic and Alpine
Bockheim, J.G., 2002. Landform and soil development in the McMurdo Dry Valleys:
a regional synthesis. Arctic Antarctic and Alpine Research 34, 308317.
Bockheim, J.G., Wilson, S.C., 1992. Soil forming rates and processes in cold
desert soils of Antarctica. In: Gilichinski, D.A. (Ed.), Proceedings of the First
International Symposium on Cryopedology. Russian Academy of Science,
Puschino, pp. 4256.
Brantley, S.L., 2008. Kinetics of mineral dissolution. In: Brantley, S.L., Kubicki, J.D.,
White, A.F. (Eds.), Kinetics of Water-Rock Interaction. Springer-Verlag,
Heidelberg, pp. 151210.
Brown, G.H., 2002. Glacier meltwater hydrochemistry. Applied Geochemistry 17,
855883.
Bunting, B.T., Christensen, L., 1978. Micromorphology of calcareous crusts from
the Canadian High Arctic. Geologiska Foereningan I Stockholm Foerhandlingar
100, 361367.
Burns, S.F., 1990. Alpine soil distribution and development, Indian Peaks, Colorado
front Range. Ph.D. dissertation, University of Colorado, Boulder.
Burns, S.F., Tonkin, P.J., 1982. Soil-geomorphic models and the spatial distribution
and development of alpine soils. In: Thorn, C.E. (Ed.), Space and Time in
Caine, N., 1979. Rock weathering rates at the soil surface in an alpine environment.
Catena 6, 131141.
Caine, N., 1992. Spatial patterns of geochemical denudation in a Colorado alpine
environment. In: Dixon, J.C., Abrahams, A.D. (Eds.), Periglacial Geomorphology.
Caine, N., 1995. Temporal trends in the quality of stream water in an alpine
environment: Green Lakes Valley. Colorado Front Range, U.S.A. Geografiska
Annaler 77A, 207220.
Caine, N., 2001. Geomorphic systems of Green Lakes Valley. In: Bowman, W.D.,
Seastedt, T.R. (Eds.), Structure and Function of an Alpine Ecosystem. Niwot
Ridge, CO, Oxford University Press, Oxford, pp. 4574.
Caine, N., Thurman, E.M., 1990. Temporal and spatial variations in the solute
content of an alpine stream, Colrado Front Range. Geomorphology 4, 5572.
Campbell, I.B, Claridge, G.G.C., 1987. Antarctica: Soils Weathering Processes and
Environment. Elsevier, Amsterdam, 368 pp.
Campbell, I.B., Claridge, G.G.C., 1992. Soils of cold climate regions. In: Martini,
I.P., Chesworth, W. (Eds.), Weathering, Soils, and Paleosols. Elsevier,
Campbell, S.W., Dixon, J.C., Thorn, C.E., Darmody, R.G., 2002. Chemical denudation
rates in Karkevagge, Swedish Lapland. Geografiska Annaler 84A, 179185.
Claridge, G.G.C., Campbell, I.B., 1984. Mineral transformation during the
weathering of dolerite under cold arid conditions in Antarctica. New Zealand
Journal of Geology and Geophysics 27, 537545.
Collins, D.N., 1983. Solute yield from a glacierized high mountain basin. Dissolved
Load of Rivers and Surface WaterQuantity/Quality Relationships. International
Association of Scientific Hydrology, Publication 141, pp. 4149.
Colman, S.M., 1982. Chemical weathering of basalts and andesites: evidence from
weathering rinds. U.S. Geological Survey Professional Paper 1246. 51pp.
Darmody, R.G., Allen, C.E., Thorn, C.E., 2005. Soil topo-sequences at
Storbreen, Jotunheimen, Norway. Soil Science Society of America Journal 69,
12751287.
Darmody, R.G., Thorn, C.E., Dixon, J.C., 2007. Pyrite-enhanced chemical
weathering in Karkevagge, Swedish Lapland. Geological Society of America
Bulletin 119, 14771485.
Darmody, R.G., Thorn, C.E., Dixon, J.C., Schlyter, P., 2000. Soils and landscapes of
Karkevagge, Swedish Lapland. Soil Science Society of America Journal 64,
14551466.
Darmody, R.G., Thorn, C.E., Rissing, J.K., 1987. Chemical weathering of fine debris
from a series of Holocene moraines: Storbreen, Jotunheimen, southern Norway.
Geografiska Annaler 69A, 405413.
Dixon, J.C., 1983. Chemical weathering of late Quaternary cirque deposits in the
Colorado Front Range. Ph.D. dissertation, University of Colorado, Boulder. 174 pp.
Dixon, J.C., 1986. Solute movement on hillslopes in the alpine environment of the
Colorado Front Range. In: Abrahams, A.D. (Ed.), Hillslope Processes. Allen and
Unwin, London, pp. 139159.
Dixon, J.C., 2004. Weathering. In: Goudie, A.S. (Ed.), Encyclopedia of
Dixon, J.C., Campbell, S.W., Thorn, C.E., Darmody, R.G., 2006. Incipient weathering
rind development on introduced machine-polished granite discs in an Arctic
alpine environment, northern Scandinavia. Earth Surface Processes and
255
mid-latitude alpine environments. Geomorphology 67, 127145.
Dixon, J.C., Thorn, C.E., Darmody, R.G., 1984. Chemical weathering processes on
the Vantage Peak nunatak, Juneau Icefield, southern Alaska. Physical Geography
5, 111131.
Dixon, J.C., Thorn, C.E., Darmody, R.G., 1995. Preliminary investigation of
geochemical process reponses to potential environmental change in Karkevagge,
northern Scandinavia. Geografiska Annaler 77A, 259267.
Dixon, J.C., Thorn, C.E., Darmody, R.G., Campbell, S.W., 2002a. Weathering rinds
Dixon, J.C., Thorn, C.E., Darmody, R.G., Campbell, S.W., 2002b. Postglacial rock
weathering processes on a roche moutonee in the Riksgransen area (681 N),
northern Norway. Norsk Geografisk Tidsskrift 56, 257261.
Dorn, R., 1998. Rock Coatings. Alsevier, Amsterdam, 429 p.
Drewry, D.J., 1986. Glacial Geological Processes. Arnold, London, 276 pp.
Egli, M., Merkli, C., Sartori, G., Mirabella, A., Plotze, M., 2008. Weathering,
mineralogical evolution and soil organic matter along a Holocene soil
toposequence developed on carbonate-rich materials. Geomorphology 97,
675696.
Egli, M., Mirabella, A., Fitz, P., 2003a. Formation rates of smectites derived from
two Holocene chronosequences in the Swiss Alps. Geoderma 117, 8198.
Egli, M., Mirabella, A., Mancabelli, A., Sartori, G., 2004. Weathering of soils in
alpine areas influenced by climate and parent material. Clays and Clay Minerals
52, 287303.
Egli, M., Mirabella, A., Sartori, G., Fitz, P., 2003b. Weathering rates as a function of
climate: results from a climosequence of the Val Genova (Trentino Alps).
Egli, M., Mirabella, A., Sartori, G., Giaccai, D., Zanelli, R., Plotze, M., 2007. Effect
of slope aspect on transformation of clay minerals in Alpine soils. Clay Minerals
42, 373398.
Ellis, S., 1980. An investigation of weathering in some Arctic-alpine soils on the
northeast flank of Oksskolten, north Norway. Journal of Soil Science 31,
371385.
Etienne, S., 2002. The role of biological weathering in periglacial areas: a study of
weathering rinds in south Iceland. Geomorphology 47, 7586.
Eyles, N., Sassaville, D.R., Slat, R.M., Rogerson, R.J., 1982. Geochemical
denudation rates and solute transport mechanisms in a marine temperate glacier
basin. Canadian Journal of Earth Sciences 18, 4149.
Ford, D.C., 1979. A review of alpine karst in the southern Rocky Mountains of
Canada. Bulletin of the National Speleological Society 41, 5365.
Ford, D.C., 1996. Karst in a cold climate. In: McCann, S.B., Ford, D.C. (Eds.),
Geomorphology sans Frontie`res. Wiley, Chichester, pp. 153179.
Gams, I., 1985. International comparative measurements of surface solution
by means of standard limestone tablets. Razprave IV Razreda Sazu 26,
361386.
Hallet, B., 1975. Subglacial silica deposits. Nature 254, 682685.
Hallet, B., 1976. Deposits formed by subglacial precipitation of CaCO3. Geological
Hochella, Jr. M.F., Banfield, J.F., 1995. Chemical weathering of silicates in nature: a
microscopic perspective with theoretical considerations. In: White, A.F., Brantley,
S.L. (Eds.), Chemical Weathering Rates of Silicate Minerals. Reviews in
Mineralogy, vol. 31, pp. 353406.
Hodgkins, R., Tranter, M., Dowdeswell, J., 1997. Solute provenance, transport and
denudation in a high Arctic glacierized catchment. Hydrological Processes 11,
18131832.
Hodson, A., Tranter, M., Vatne, G., 2000. Contemporary rates of chemical
denudation and atmospheric CO2 sequestration in glacialbasins an Arctic
perspective. Earth Surface processes and Landforms 25, 14471471.
Hofl, C., Ping, C.L., Kimble, J., 1998. Properties of permafrost soils on the
northern Seward Peninsula, northwest Alaska. Soil Science Society of America
Journal 62, 16291639.
Isherwood, D., 1975. Soil geochemistry and rock weathering in an Arctic
environment. Ph.D. dissertation, Department of Geological Sciences, University
of Colorado, Boulder.
Jackli, H., 1956. Gegenwartsgeologie des bundnerischen Rheingebietes-ein Beitrag
zur exogenen Dyanmik alpine Gebirgslandschaften. Beitrage zur Geologie der
Schweiz, Geotechnical Series 36, 126p.
Kelly, W.C., Zumberg, J.H., 1961. Weathering of a quartz diorite at marble Point,
McMurdo Sound, Antarctica. Journal of Geology 69, 433446.
Kubiena, W.L., 1953. The Soils of Europe. T. Murby, London.
Legros, J.P., 1992. Soils of Alpine mountains. In: Martini, I.P., Chesworth, W.
(Eds.), Weathering, Soils and Paleosols. Elsevier, Amsterdam, pp. 155181.
256
Lev, A., King, R.H., 1999. Spatial variation in soil development in a high Arctic soil
landscape: Truelove Lowland, Devon Island, Nunavut, Canada. Permafrost and
Periglacial Processes 10, 289307.
Locke, W.W., 1979. Etching of hornblende grains in Arctic soils: an indicator of
relative age and paleoclimate. Quaternary Research 11, 197212.
ozinski, W., 1909. Uber die mechanische Verwitterung der sandsteine im
gemassigten Klima. Bulletin International de lAcademie des Sciences de
Cracovie class des Sciences Mathematique et Naturelles 1, 125.
ozinski, W., 1912. Die periglaziale fazies der mechanischen Verwitterung. 11th
International Geological Congress, Stockholm 1910, Comptes Rendus,
10391053.
Mahaney, W.C., 1974. Soil stratigraphy and genesis of Neoglacial deposits in the
Arapaho and Henderson Cirques, Central Colorado Front Range. In: Mahaney,
W.C. (Ed.), Quaternary Environments. Atkinson College/York University
Monograph, vol. 5, pp. 197240.
Meiklejohn, I., Hall, K., 1997. Aqueous geochemistry as an indicator of chemical
weathering on southeastern Alexander Island, Antarctica. Polar Geography 21,
101112.
Mellor, A., 1985. Soil chronosequences on Neoglacial moraine ridges,
Jostedalsbreen and Jotunheimen, southern Norway: a quantitative pedogenic
approach. In: Richards, K.S., Arnett, R.R., Ellis, S. (Eds.), Geomorphology and
Soils. Allen and Unwin, London, pp. 289308.
Mellor, A., 1987. A pedogenic investigation of some soil chronsequences on
Neoglacial moraine ridges, southern Norway: examination of soil chemical data
using principal components analysis. Catena 14, 369381.
Messer, A.C., 1988. Regional variations in rates of pedogenesis and the influence of
climatic factors on moraine chronosequences, southern Norway. Arctic and
Alpine Research 20, 3139.
Meybeck, M., 1987. 1987. Global chemical weathering of surficial rocks estimated
from river dissolved loads. American Journal of Science 287, 401428.
Mirabella, A., Egli, M., Carnicelli, S., Sartori, G., 2002. Influence of parent material
on clay mineral formation in Podsols of Trentino, Italy. Clay Minerals 37,
699707.
Oguchi, C., 2001. Formation of weathering rinds on andesite. Earth Surface
Olson, C., 2004. Chemical weathering. In: Goudie, A.S. (Ed.), Encyclopedia of
Geomorphology. Routledge, London.
Paces, T., 1986. Rates of weathering and erosion derived from mass balance in
small drainage basins. In: Colman, S.M., Dethier, D.P. (Eds.), Rates of Chemical
Weathering of Rocks and Minerals. Academic Press, Orlando, pp. 531550.
Peltier, L., 1950. Geographic cycle in periglacial regions as it is related to climatic
geomorphology. Annals of the Association of American Geographers 40,
214236.
Peterson, J., Morseby, J., 1979. Subglacial travertine and associated deposits in the
Carstensz area Irian Jaya, republic of Indonesia. Zeitschrift fur Gletscherkund
und Glaziologie 15, 2326.
Rapp, A., 1960. Recent development of mountain slopes in Karkevagge and
surroundings, northern Scandinavia. Geografiska Annaler 42A, 65200.
Retzer, J.L., 1965. Present soil forming factors and processes in Arctic and alpine
regions. Soil Science 99, 3844.
Reynolds, Jr. R.C., 1971. Clay mineral formation in an alpine environment. Clays
and Clay Minerals 19, 361374.
Reynolds, Jr. R.C., Johnson, N.M., 1972. Chemical weathering in the temperate
glacial environment of the northern Cascade Mountains. Geochimica
Cosmochimica Acta 36, 537554.
Rieger, S., 1974. Arctic soils. In: Ives, J.D., Barry, R.G. (Eds.), Arctic and Alpine
Environments. Methuen, London, pp.
Sharp, M., Tranter, M., Brown, G.H., Skidmore, M., 1995. Rates of chemical
denudation and CO2 drawdown in a glacier covered alpine catchment.
Geology 23, 6164.
Shroba, R.R., Birkeland, P.W., 1983. Trends in late Quaternary soil development in
the Rocky Mountains and Sierra Nevada western United States. In: Porter, S.C.
(Ed.), The Late Pleistocene, Late Quaternary Environments of the United States.
University of Minesota Press, Minneapolis, MN, vol. 1, pp. 145156.
Souchez, R.A., Lemmens, M.M., 1987. Solutes. In: Curnall, A.M, Clark, M.J. (Eds.),
Glacio-fluvial Sediment Transfer. Wiley, Chichester, pp. 285303.
Tamm, O., 1924. Experimental Studies on Chemical Processes in the Formation of
Glacial Clay. Sveriges Geologiska Undersoekning, Arsbok. The Swedish
Geological Survey, Sweden, vol. 18, 20 pp.
Thorn, C.E., 1975. Influence of late-lying snow on rock-weathering rinds. Arctic and
Thorn, C.E., 1976. Quantitative evaluation of nivation in the Colorado Front Range.
Thorn, C.E., Darmody, R.G., Campbell, S.W., Allen, C.E., Dixon, J.C., 2007.
Microvariability in the early stages of cobble weathering by microenvironment
on a glacier foreland, Storbreen, Jotunheimen, Norway. Earth Surface Processes
Thorn, C.E., Dixon, J.C., Darmody, R.G., Allen, C.E., 2006a. Ten years (19942004)
of potential weathering in Karkevagge, Swedish Lapland. Earth Surface
Thorn, C.E., Dixon, J.C., Darmody, R.G., Allen, C.E., 2006b. A 10 year record of
the weathering rates of surficial pebbles in Karkevagge, Swedish Lapland.
Catena 65, 272278.
Tedrow, J.C.F., Drew, J.V., Hill, D.E., Douglas, L.A., 1958. Major genetic soils of the
arctic slope of Alaska. Journal of Soil Science 9, 3345.
Tedrow, J.C.F., Ugolini, F.C., 1966. Antarctic soils. In: Tedrow, J.C.F. (Ed.), Antarctic
Soils and Soil Forming Processes. Antarctic Research Series. American
Geophysical Union, Washington, DC, vol. 8, pp. 161177.
Tranter, M., 2005. Geochemical weathering in glacial and proglacial environments.
In: Drever, J.I. (Ed.), Surface and Ground Water, Weathering, and Soils. Treatise
on Geochemistry, Elsevier, Amsterdam, vol. 5, pp. 189205.
Tranter, M., Sharp, M.J., Brown, G.H., et al., 1997. Variability in the chemical
composition of in situ subglacial meltwaters. Hydrological Processes 11, 5977.
Tranter, M., Sharp, M.J., Lamb, H.R., Brown, G.H., Hubbard, B.P., Willis, I.C., 2002.
Geochemical weathering at the bed of Haut Glacier dArolla, Switzerland-a new
model. Hydrological Processes 16, 959993.
Ugolini, F.C., 1986. Processes and rates of weathering in cold and polar desert
environments. In: Colman, S.M., Dethier, D.P. (Eds.), Rates of Chemical
Ugolini, F.C., Anderson, D.M., 1973. Ionic migration and weathering in frozen
Antarctic soils. Soil Science 115, 461470.
Ugolini, F.C., Jackson, M.L., 1982. Weathering and mineral synthesis in Antarctic
Soils. In: Cradock, C. (Ed.), Antarctic Geoscience. University of Wisconsin,
Madison, WI, pp. 11011111.
Wadham, J.L., Cooper, R.J., Tranter, M., Hodgkins, R., 2001. Enhancement of
glacial solute fluxes in the proglacial zone of a polythermal glacier. Journal of
Wadham, J.L., Hodson, A.J., Tranter, M., Dowdeswell, J.A., 1997. The rate of
chemical weathering beneath a quiescent, surge-type polythermal based glacier,
southern Spitsbergen. Annals of Glaciology 24, 2731.
Watts, S.H., 1981. Bedrock weathering features in a portion of eastern High Arctic
Canada: their nature and significance. Annals of Glaciology 2, 170175.
Watts, S.H., 1985. A scanning electron microscope study of bedrock microfractures
in granites under high Arctic conditions. Earth Surface Processes and
Whalley, W.B., Gellatly, A.F., Gordon, J.E., Hansom, J.D., 1990. Ferromanganese
rock varnish in northern Norway: a subglacial origin. Earth Surface Processes
Yadav, S.K., Chakrapani, G.J., 2006. Dissolution kinetics of rock-water interactions
and its implications. Current Science 90, 932937.
257
Biographical Sketch
John C Dixon is a professor in the Department of Geosciences at the University of Arkansas, Fayetteville, AR, USA.
His research interests lie primarily in the area of geochemical processes and landscape evolution in cold climates.
He has conducted research in Swedish Lapland over the past 15 years, as well as in the Jotenheimen of Norway,
the Juneau Icefield of Alaska, and the Rocky Mountains of Colorado. He and his research colleagues have
published more than 70 papers in the major geomorphology journals and in specialized monographic volumes.
4.15 Mechanical Weathering in Cold Regions

K Hall, University of Northern British Columbia, Prince George, BC, Canada
4.15.1
4.15.1.1
4.15.1.2
4.15.1.3
4.15.2
4.15.2.1
4.15.2.2
4.15.2.2.1
4.15.2.2.2
4.15.2.2.3
4.15.2.2.4
4.15.2.2.5
4.15.2.3
4.15.2.4
4.15.2.4.1
4.15.2.5
4.15.2.6
4.15.2.7
4.15.3
4.15.3.1
4.15.4
References
Introduction
The Literature
Weathering Processes
Landforms Associated with Weathering
Weathering Processes in Cold Regions
Introduction
Frost Weathering
Background to the freezethaw concept
Freezethaw in cold regions: Laboratory studies
Thermal and moisture conditions
Rock properties
From the past to the present
Weathering by Wetting and Drying
Thermal Stress Fatigue/Thermal Shock
Thermal stress within the cold region context
Salt Weathering
Dilatation/Pressure Release
Where Are We at Where Do We Go?
Landforms
Landform Associations
Where are We at and Where are We Going?
Glossary
Chemical weathering Changes to the structure/chemistry
of a rock through purely chemical mechanisms, which, in
some cases, may be reversible.
Dilatation (pressure release) The expansive movement of
rock, leading to failure planes parallel to the relieving
surface, due to the loss of overburden through erosive
action.
Exotherm A term used to describe the release of latent
heat when water freezes within rock. The exotherm is the
recorded sudden increase in temperature, due to that latent
heat, during a time when the air temperature continues to
decrease.
Freezethaw weathering A complex process, with
many different identified mechanisms, that is said to
cause the (mechanical) weakening and failure of rock at
different scales (see microgelivation/macrogelivation)
through the stresses developed by the freezing of water
within the rock.
Macrogelivation A term conflating process and product
that refers to the breakdown of rock by the freezing of water
associated with larger scale lines of weakness (e.g., joints)
that is said to result in pebble-size and larger products
(particularly blocks).
259
259
260
260
260
260
261
261
261
262
263
263
264
265
266
266
267
268
269
269
272
272
Mechanical weathering Irreversible breakdown of

rock by mechanisms that do not involve any chemical
action.
Microgelivation A term conflating process and product
that refers to the breakdown of rock by the freezing of water
at (roughly) the grain scale and is said to cause granular
disintegration or flaking.
Pseudokarren Karren refers to elongate, solutional
furrows on slopes developed in karst areas while
pseudokarren is a term used to describe features that have
the same form but that are not developed in calcareous
rocks; they are often found in granitic rocks.
Salt weathering A mechanical weathering mechanism
causing stresses and failure in rock due to the expansive
nature of salt crystals within the rock.
Tafoni (alveolar or honeycomb weathering) Hollows
developed on (sub)vertical walls in various lithologies.
A combination of weathering and debris removal creates
the hollows; the terms alveolar or honeycomb weathering
are sometimes applied to smaller forms.
Thermal stress fatigue/shock Mechanisms by which
differential stresses induced within the rock by thermally
induced expansion and contraction of the component
minerals can lead to fatigue and failure of that rock,
Hall, K., 2013. Mechanical weathering in cold regions. In: Shroder, J. (Editor
in Chief), Pope, G.A. (Ed.), Treatise on Geomorphology. Academic Press,
San Diego, CA, vol. 4, Weathering and Soils Geomorphology, pp. 258276.
258
http://dx.doi.org/10.1016/B978-0-12-374739-6.00062-2
Mechanical Weathering in Cold Regions
characteristically at the granular scale. When the stresses

exceed the ability of the minerals/rock to deform in
response to the thermally induced changes, instant failure
can occur: thermal shock.
Weathering pits Hollows found on (sub)horizontal
surfaces in many lithologies that are presumed to have
259
developed by a combination of weathering and transport of

the weathering products.
Wetting and drying (slaking) A mechanical weathering
mechanism by which stresses, possibly leading to failure,
are generated by the expansive nature of minerals when
they take up water (without chemical change).
Abstract
The title of this chapter is really an oxymoron insofar as there are no processes unique to such regions and the landforms
observed may not be the product of the present climate. Indeed, the zonal approach may be one that is disingenuous and,
ultimately, misleading, and has hindered our process and landform understanding. The linkage of weathering to landforms is also frequently ambiguous (e.g., weathering pit) for there are few landforms solely due to weathering; there must
also be the equally important removal of weathering products. Here, a number of foundational issues related to the notion
of both weathering per se and of weathering in cold regions are discussed. With regard to most weathering processes, it is
noted that it is moisture availability that is the primary limiting factor rather than temperature. The exception is thermal
stresses where, although an azonal process, it may be enhanced in cold environments when large values of DT/Dt occur
during a decline in temperature. Ultimately, the cold region is argued not to be a dictator of process but rather one of
enhancing or diminishing process efficacy or duration. It is further argued that it is the rock properties, rather than climate,
that play the major role in determining process. An attempt is made to outline the literature for both processes and
landforms, to note our current understanding of processes, and to offer some possible scale-related approaches that may
advance our understanding of both weathering and landforms.
4.15.1
Introduction
Mechanical weathering produces a change in the physical attributes of rock (e.g., fractures, pores, etc.), without chemical
action, and it generally leads to a decrease in strength. Although considered to act in situ, without movement, Chorley
et al. (1984, p. 205) recognized that there must be some
movement involved and thus it is impossible to distinguish
between the mechanisms of weathering and those of mass
movement. Simplistically, the majority of mechanical weathering processes are driven by the weather the specific
(spatial and temporal) attributes of an area interact with the
rock and its properties to determine the process(es). Here,
weathering is considered within the framework of a cold
climate. Weathering due to mechanical processes is irreversible
and will, ultimately, lead to the breakdown of the rock into
its component materials and is hence a producer of clastic
sediments. Although discussed separately (see Chapter 4.2),
mechanical processes frequently, in some synergistic fashion,
operate with chemical weathering. Ultimately, both constructive (e.g., screes) and destructive (e.g., tafoni) landform
evolution depends indirectly, and to varying degrees, on
mechanical weathering processes.
The indirect role of weathering in landform production
is emphasized as mechanical weathering processes, on their
own, cannot generally produce landforms but rather they
prepare the material for gravity-based transportation/erosional
processes, as well as the potential for chemical or biological
weathering, the total interaction of which creates the resultant
landforms. Thus, sensu stricto, it is incorrect, and certainly
misleading, to talk of weathering landforms and to apply
names to landforms (e.g., weathering pits) that suggest this
weathering origin the pit being due to the removal of the
weathering products. However, this is not to say that mechanical weathering does not play a major role within the
development of landforms. As is noted below, the primary
mechanical weathering processes normally considered are
weathering due to freezethaw, wetting and drying, salt,
thermal stresses, and the action of dilatation; only the last is
a weathering process not directly associated with weather but
can be associated with deglaciation and hence may be significant within cold regions.
4.15.1.1
The Literature
A number of texts deal specifically with cold region geomorphology, for example, Tricart (1963), Davies (1969),
Washburn (1973), Embleton and King (1975), Yershov (1998),
and French (2007). A number of very topic-specific cold region
texts also exist: for example, on landforms (Barsch, 1996), on a
process (Seppala, 2004), as well as region-specific texts (Ballantyne and Harris, 1994 or Godard and Andre 1999), and
collected works by Clark (1988), Dixon and Abrahams (1992),
and Evans (1994).
A number of papers provide reviews of literature and
thoughts, as well as some consideration of our gaps in
knowledge. McGreevy (1981) provided a review of frost
shattering, which was followed by a discussion regarding the
role of rock temperature in cold region weathering (McGreevy
and Whalley, 1982) and rock moisture content in frost
weathering (McGreevy and Whalley, 1985). Whalley and
McGreevy (1983, 1985, 1987) went on to provide progress
reports on weathering that continue to be a superb resource.
Douglas et al. (1983) also provided a review regarding the
interaction of temperature and moisture in causing frost
shattering, which Ritchie and Davison (1968) had previously
260
undertaken in an applied study focused on masonry materials.

Ugolini (1986) detailed valuable information on the mechanical, chemical, and biological processes in cold and polar
desert environments, whereas Hall and Boelhouwers (2003) is
a more recent literature resource. Ives (1973) raised a number
of questions pertaining both to freezethaw weathering and to
an associated landform nivation hollows (see Thorn and
Hall, 1980, 2002). Ives (1973) questioned the efficacy of frost
weathering and notes (Ives, 1973, p. 3):
I am not arguing that the freezethaw cycle is in no way involved
y but most certainly the concept of numerous diurnal or shortterm fluctuations of temperature across the freezing point does not
provide an adequate explanation.
This constitutes a set of concerns that have gone essentially

unheeded and uninvestigated, wherein the simple counting
of crossings of some (unknown) freezing point (e.g., Fortin,
2010) has continued to bedevil cold region weathering and
landform concepts.
4.15.1.2
Weathering Processes
Historically, weathering within cold regions has been seen

as climatic zone specific (e.g., ozinski, 1909). However,
all mechanical weathering processes are azonal in operation;
even freezethaw weathering occurs within temperate
climatic regions and such hot areas as the Namib Desert. As a
corollary, processes generally more associated with hot environments (e.g., thermal stresses) are being recognized (e.g.,
Leask and Wilson, 2003; McKay et al., 2009) as significant
within cold regions. In many ways, much of the material here
could legitimately appear within other weathering sections.
Specifically, high polar or high-altitude dry cold regions are
really no different from any other Arid Region (aridity being
the driver not temperature), and many low-elevation or lowlatitude cold regions do not differ significantly, in terms of
weathering, from that of Temperate Regions (save perhaps in
rate). Furthermore, processes operate, temporally and spatially, in a synergistic fashion not only among the mechanical processes alone but also in multiple combinations with
both their chemical and biologically induced counterparts.
Indeed, here, it is argued that it is this very climatic foundation, with its emphasis on cold, that has been a significant
hindrance to our understanding of landform development.
Furthermore, we have, through books perhaps more so
than in research papers, a situation wherein, either because it
is believed or, following Ritter et al. (2002), because it is a
matter of convenience, weathering processes set out as discrete
entities. Such a presentation promotes the (unhealthy) notion
that the processes are indeed discrete and that they (somehow) operate alone. For example, in the case of freezethaw
weathering, given the presence of water, it is assumed that
wetting and drying, chemical weathering, thermal stresses,
and potentially salt weathering must also be taking place.
Thus, as such, there really is no such thing as freezethaw
sensu stricto on our planet and perhaps our approach, both
conceptually and experimentally, is misleading. Indeed, Smith
(1994, p. 40) makes the very same point: the reductionist
tendency has fostered a segregationist approach wherein the
specific mechanisms such as insolation weathering, salt crystallization, or freezethaw are examined in isolation. Further,
the impression is clearly given, if only by default, that these
identified processes are both the only ones existing when it
seems incredibly naive to think we have found all that may
exist (e.g., Hall, 1989) and that these processes are real (i.e.,
that they actually exist in the manner that we portray them).
The emphatic presentation of processes by rote lends credence
to the unspoken (mis)conception that these are, in the guise
presented, the only weathering processes.
4.15.1.3
Landforms Associated with Weathering
As noted above, there really cannot be, by the very definition

of weathering, any landforms that result from weathering
alone (thermal contraction cracks in rock perhaps being the
exception that proves the rule). However, it is clear that certain processes, or more commonly process suites, are linked to
cold region landforms. There are strong associations of weathering with nivation or cryoplanation and yet, even here,
questions regarding the assumption of a cold climate origin
have arisen (e.g., Guilcher, 1950). Although some landforms
are commonly linked with cold region weathering (e.g., tafoni:
Calkin and Cailleux, 1962), they are actually azonal features
(see Strini et al., 2008); the question of the origin versus
preservation of other features under the presently exhibited
cold climate remains an enigma (e.g., pseudokarren in Antarctica: Bardin, 1962). Equally, there are landforms such as
boulder fields commonly associated with cold regions that
may, or may not, have anything to do with a cold region
origin. Boulder field characteristics are not much different
from those of a boulder pavement in a hot desert (Cooke,
1970). Indeed, the stone streams of the Falkland Islands, the
sites for the origin of the periglacial solifluction concept
(Andersson, 1906), were recently argued, on the basis of
SEM and XRD analyses, to be of tropical origin (Andre et al.,
2008). Indeed, Andre (2003) discussed whether periglacial
landscapes even evolve under periglacial conditions. Thus,
the linkage of weathering with landforms that have a determined cold region origin is perhaps more tenuous (imagined?) than real.
4.15.2
4.15.2.1
Weathering Processes in Cold Regions

Introduction
Any general text on weathering in periglacial regions is likely

to state something to the effect that: Freezethaw action is
undoubtedly the most important process of rock weathering
in the periglacial zone; it is the primary agent responsible for
such features as talus accumulations and blockfields, and the
breakdown of debris into particles fine enough to be handled
by running water and the wind (Embleton and King, 1975,
p. 4). Many texts justify this argument via the features that are
(directly or indirectly) said to result from weathering as well as
by recourse to the presence of highly angular debris that
generally occurs in periglacial regions (see McCarroll et al.,
1995; Hall and Thorn, 2011). This is generally accompanied by
statements that suggest chemical weathering to be weak in
cold regions and that weakness can then devolve into the
dismissal of chemical weathering and the overemphasis of
mechanical processes (e.g., Selby, 1985). The reality, in a range
of cold environments, may be that processes other than
freezethaw dominate and that biological or chemical
weathering may be more prevalent than is usually assumed
(e.g., McCarroll and Viles, 1995). Even the most recent cold
region texts (e.g., French, 2007, p. 60) cannot divest themselves of the historical importance of frost weathering, maintaining the formprocess connotation of angular rocks as
equating to frost action. However, as Thorn (1992, p. 11)
stated: what periglacial geomorphologists need more than
any other single item is a way to determine in the field whether or not bedrock fragments have been frost weathered!
The body of literature, both direct and indirect, on weathering in cold regions (see Hall et al., 2002) is simply enormous and is accompanied by another voluminous literature
that is relevant but not climate specific such that a whole
volume would be needed to deal with the topic. Thus, any
omissions that the reader notice herein are due to space
limitations and my own shortcomings and should not be
viewed as deliberate.
4.15.2.2
Frost Weathering
As a mechanism recognized since at least the 1500s

(Thompson, 1959), the role of freezethaw weathering and its
general acceptance as dominating periglacial regions have best
been discussed by Thorn (1988, 1992):
From its inception, periglacial geomorphology has been dominated
by the concept of frost wedging, a synonym for weathering by
freezing and thawing. The story is one of casual empiricism gathering respectability by repetition until it attained the stature of an
article of faith. (Thorn, 1992, p. 10)
Further,
that have been voiced over the last hundred years. Luquer
(1895), in consideration of frost tests on building stone, noted
that all used saturated conditions and that the number
of freeze cycles to effect damage was quite varied. He stated
(Luquer, 1895, p. 237) As in many cases, it is almost impossible, or very inconvenient, to get the real action of frost.
Although Merrill (1906, p. 177) stated Provided, then, sufficient water be contained within the pores of a stone, it is easy
to understand that the results of freezing must be disastrous,
he also noted that although very porous stone will readily take
up water it will also give it up very easily, and when frozen, the
water has space in which to expand and the ice can be extruded more easily without exerting any force on the rock,
what was referred to as freezeout (Luquer, 1895, p. 256).
Further, Branner (1896, p. 282) noted that frozen water on
rock may be protective rather than destructive as it prevents
the penetration of water.
The physicist Grawe (1936) called attention to some of the
points outlined above plus added physical constraints upon
the frost mechanism and, in so doing, cautioned against its
casual acceptance or presumption. Grawe (1936) pointed out
that the theoretical power of freezing water (2115 kg cm2)
can only be attained if the following criteria are met: (1) the
system comprises only water, (2) that the temperature is
22 1C and declining, (3) it is pure water, (4) the ice is
confined in a closed system, and (5) the constraining material
is actually strong enough to resist the calculated pressure,
something no rock can do. However, in moving toward that
possible maximum pressure, so rock breakdown could occur.
Grawes point was that it is meaningless to cite this maximum
theoretical value as it will not occur in nature. Reiche (1950,
p. 12), in his survey of weathering and weathering products,
reiterated the following point: Thus y the figures often cited
for the maximum disruptive force of freezing are quite irrelevant and yet it continues to be cited as if it has any validity
(e.g., French, 2007, p. 63).
4.15.2.2.2
At present there is no adequate field criterion to establish that
bedrock weathering or further comminution of rock fragments has
been dominated by freezethaw weathering. However, the most
common argument to substantiate the importance of freezethaw
weathering is both circumstantial and circular. The concept that
freezethaw weathering dominates cold regions gained respectability long before there was the ability to test it in the field.
As angular fragments are common in cold environments they
were assumed to be the product of the dominant process, namely
freezethaw weathering. (Thorn, 1992, p. 11)
The problem is that the argument for frost action is both

circular and circumstantial, and is used as a foundation for
secondary concepts such as nivation or cryoplanation. None
of this is to say that freezethaw does not take place but rather
it may be less ubiquitous than we assume or other processes
can play a role, even a dominant one (see Andre, 2003). As, for
cold regions, frost weathering is generally considered the
prime weathering agent and the key driver behind landforms;
consequently, it will receive preferential treatment.
4.15.2.2.1
Background to the freezethaw concept
To begin with, it is worth noting some of the criticisms regarding the efficacy and prevalence of freezethaw weathering
261
Freezethaw in cold regions: Laboratory

studies
Much of the freezethaw problem is founded on laboratory

experiments thought to replicate field situations, which are
then used to validate/exemplify the operation/efficacy of the
process. However, as Warren (1914a, p. 547) stated experiments cannot easily reproduce natural conditions and if they
do not adequately reproduce natural conditions they cannot
directly show the product of those conditions. Warren
(1914b, p. 413) further made the point that it would be unsound to assume that the result of a certain experiment must
also be reproduced by natural agencies, without evidence that
similar conditions exist in Nature to those employed in the
experiment. Thus, any simulation is only as good as the data
used to structure it. In many instances, the factors of rock
properties, moisture, and thermal conditions have been inadequately quantified and so, in reality, the findings are severely flawed. Further, most simulations attempt to consider
but a single process and ignore both inherent and longer term
process synergies, as well as the spatial variability of the given
process(es). It is believed that these issues weaken many of the
laboratory undertakings attempting to simulate freezethaw
weathering cited below.
262
Outside of engineering/building applied approaches (e.g.,

Butterworth, 1964; Davison and Sereda, 1978; Davidson and
Nye, 1985), early key studies, using a variety of rock types,
were those of Tricart (1956), Wiman (1963), Martini (1967),
Potts (1970), Keeble (1971), and Brockie (1972). All these
undertakings generally utilized two main kinds of freeze
thaw cycles: an Icelandic cycle (rapid, low-amplitude fluctuations) or a Siberian cycle (with slow, large-amplitude
freezes). The findings from these studies varied considerably
(Tricart found the Siberian regime to produce the most weathering but both Wiman and Potts suggested the Icelandic to
be the more efficacious), the reality being that the results were
a product of the experiments. Even the two cycles were a
function of the perceived climate and had nothing to do with
what the rock actually experienced, the more so as samples
were usually immersed in water during the wetting phase.
In Japan, workers at the Institute of Low Temperature
Science carried out a range of experiments (e.g., Fukada, 1972;
Fukuda and Matsuoka, 1982; Fukuda et al., 1983; Matsuoka,
1988, 1990) that used, to great effect, techniques such as
porosimetry, ultrasonic velocity, and electrical resistance to
monitor freezing and water migration. Indeed, works by
Fukuda and Matsuoka (1982) predate the theoretical model of
Hallet (1983), where they argued for (p. 223) a frost heave
mechanism as the cause of frost shattering. The main problem, for the English-speaking world, is that most of the text is
in Japanese and so these works have not received the recognition they are due. To some extent, the same is the case for
the studies from France and elsewhere (e.g., Van Steijn, 1979),
which are both extensive and detailed and yet not so often
recognized.
In France, notable early works are by Masseport (1959) and
Godard and Hovel-Gangloff (1965). Frost experiments at
Caen University are many, but Lautridou and Ozouf (1982)
provided a 15-year summary of the work. Later works from
Caen include Lautridou (1984), Lautridou and Seppala
(1986), Letavernier and Ozouf (1987), Mante et al. (1983),
and Ozouf (1987); elsewhere in France, a notable study is
that of Le Ber and Oter-Duthoit (1987). Many of the above
experiments used two temperature regimes essentially the
Icelandic and Siberian and samples were commonly located within trays of water. In replicating observed local
moisture conditions, the experiments have some validity, but
as an imitation of many other situations the moisture conditions are excessive. Significant in the study of Le Ber and
Oter-Duthoit (1987, Figure 8) is the observation of an exotherm (at c. 3 1C) indicative of the moisture to ice phase
change within the sample. This finding is in accord with the
laboratory studies of Hall (e.g., Hall, 1986a, 1986b, 1988a,
1988b; Hall et al., 1989), who also recorded exotherms (e.g.,
Hall 1988b, Figure 3) on subjecting large rock samples to unidirectional freezing. Hall (1988b) evaluated rock weathering
in the maritime Antarctic, by mechanical processes, based on
replicating both thermal and moisture conditions (Hall et al,
1989) recorded at Antarctic sites. Ultrasonic evaluation of rock
samples during freezethaw events showed a change in pulse
velocity simultaneous with recorded exotherms, thereby validating the thermal signal as being a clear indicator of waterto-ice phase change. Simulation data also showed that, despite
the suggestions by Walder and Hallet (1985, 1986), the
cooling rate of the rock was controlled by the fixed final

temperature and that the slow rates of decline required by
their model were not occurring at the simulated study sites;
equally slow rates of decline of temperature applied to the
rock produced sudden, massive freezes that also conflict with
the model of Walder and Hallet (1985, 1986). In addition,
attempts were made to monitor acoustic emissions, reported
by Hodder (1976) to occur during freezing of supercooled
water, but none were recorded at the time of the freeze
event although some were recorded several hours after the
initial freeze; their significance is unknown but may relate to
rock fracture subsequent to initial freezing. Overall, while
answering some questions, and attempting to directly relate
simulations to field sites, the results from the studies of Hall
tended to lead to other questions.
4.15.2.2.3
Thermal and moisture conditions
McGreevy and Whalley (1982, p. 161) presented a discussion

on rock temperatures and their role in frost weathering with
the aim of fueling the current controversy concerning the role
of frost shattering in rock breakdown. Based on their detailed
analysis of available data, they show that 3 1C is the probable thermal threshold for frost shattering; noting that
Matsuoka (1990) used a decline below 2 1C, followed by
an increase above 2 1C, McGreevy and Whalley (1982,
p. 161) pointed out that although we are acquiring some data
regarding rock thermal conditions it, nevertheless, becomes
increasingly necessary to gather information on the availability of moisture at sites where frost shattering is thought to
be active. Without it, evaluation of the geomorphic significance of temperature changes in rocks remains speculative.
See comments below (Hall, 2004, 2007) regarding thermal
events failing to show any water-to-ice transformation because
there was no water present. Speculation regarding moisture
content is dealt with in a separate discussion (McGreevy and
Whalley, 1985), the key finding of which was that the results
of laboratory experiments may not reflect what is taking place
in nature. They find that many, if not all, laboratory simulations probably exaggerate the role of freezethaw as the
water conditions used in experiments far exceed those found
in the field. The problem is summed up when they state Only
by monitoring actual rock temperature and moisture regimes
in cold climates y can their geomorphic effects be evaluated
(McGreevy and Whalley, 1985, p. 345). Schneebeli et al.
(1995) provided a detailed discussion regarding techniques
(notably time domain reflectometry (TDM) and thermocouple psychrometers (TP)) for monitoring water content in
unsaturated crystalline rock, whereas Sass (1998, 2004, 2005)
has made major advances regarding rock moisture content
using these techniques and in applying them to frost
weathering.
These discussions by McGreevy and Whalley (1982, 1985)
bring us full circle to the opening questions and concerns
raised so eloquently by Thorn (1988, 1992). Thus, there is a
century of researchers who, from different perspectives, have
raised questions regarding the efficacy and role of freezethaw
weathering. Their points still stand today, very few of the
questions they raised (regarding field data) actually having
been answered, and when tackled frequently confound the
universal application of freezethaw weathering.

4.15.2.2.4
Rock properties
Although there is a great deal of information regarding rock

properties from engineering studies, as McGreevy and Whalley
(1984) pointed out, geomorphologists have generally been
slow or reticent to use this extensive database. Standard tests
and equipment exist for the analysis of many properties,
whereas attributes such as albedo and thermal properties of
the constituent minerals can be obtained from various books
of constants (e.g., Clark, 1966). Unfortunately, McGreevy
and Whalleys (1984, p. 555) desire that their discourse and
extensive references would encourage more frequent recourse
to this literature and range of techniques appears not to have
reached as wide an audience as they would have wished. In
fairness, however, it is evident that certain techniques for assessing rock strength variations due to weathering, such as the
use of a Schmidt hammer (e.g., McCarroll et al., 1995; Hall,
1987; Day, 1981) and ultrasonic testing (e.g., Fahey and
Gowan, 1979; Hall, 1988b; Allison, 1990), have been utilized
increasingly in weathering studies.
4.15.2.2.5
From the past to the present
With respect to the historical perception of freezethaw in the

field, the literature is simply enormous and cannot be dealt
with here. However, two things appear clear. First, there is
substantial repetition regarding the assumption/occurrence of
freezethaw weathering. Second, it is clear that, in many cases,
early perceptions were constrained by the concepts of climatic
geomorphology (e.g., Tricart, 1956); see Andre (2009) for a
discussion regarding the recent shift from climatic approaches.
Roles for processes other than freezethaw are simply not
entertained as a source of rock breakdown in cold regions
climatic geomorphology approaches dictating that, due to the
cold, chemical weathering is not possible. The problem is the
cold related to the air temperatures and not those of the rock,
and, in the absence of data on the latter, the notion of cold,
and its weathering limitations, prevailed. This very issue is
dealt with in detail by Sumner et al. (2004), where they show
that (p. 90) the ranges and fluctuations of air and rock
temperature on a diurnal basis and in total over the monitoring periods highlight a remarkable similarity in thermal
oscillations experienced by rock in both (a hot arid desert and
a wet alpine situation) settings. They continue Temperature
fluctuations at high latitudes can be of the same order of
magnitude as those found in a hot desert and thus if a
degree of azonality applies to rock thermal conditions, this
would highlight the importance of rock moisture in determining process operation and effectiveness under different
settings. In much the same way (Hall, 2004, 2007), showed
that in northern Canada, given the presence of water, chemical
weathering could occur throughout the winter despite air
temperatures down to 40 1C on some slope aspects and
inclinations. The key is the presence of water. This is important
as it is the freezing and thawing of water that is the foundation
of freezethaw weathering and, just as Hall (2004, 2007)
showed proxy evidence for freeze events through the release
of latent heat, that same evidence indicated the presence of
water when the material temperature was as high as 27 1C and
thus could facilitate chemical weathering; this was also noted
by Etzelmuller and Sollid (1991), Talkington et al. (1982), and
Balke et al. (1991). However, the absence of this same proxy
263
evidence showed the many times when thermal conditions

were conducive but no freeze pressure occurred simply because there was no water present to freeze. To further confound the common, simplistic approach of simply counting
thermal crossings of some threshold and assuming they
equate to freezethaw events (e.g., Ritchie and Davison,
1968; Coutard and Francou, 1989), the exotherm data clearly
showed that, for the same point, freezing could occur through
a range of temperatures from c. 1.9 1C to c. 9.0 1C. This is
important, for it has been extremely common to use some
thermal threshold and count freezethaw events and, finding
large numbers of these events, to use this as a validation for
the aggressive, efficient nature of freezethaw weathering.
Never is there any evidence that water was actually present,
that it actually froze and caused any damage!
This recognition of the problems regarding freezethaw
is not new. Ives (1973, p. 1), in a review of the topics in
most need of study within periglacial geomorphology, identified The efficacy of freezethaw process in the role of
bedrock disintegration, an issue reiterated by White (1976).
These two authors once again brought forth the point that
we really know very little regarding the action and role of
freezethaw weathering. White (1976) discussed the background to the theory of freezethaw weathering, noting the
questions raised by Grawe (1936) and showing how they were
still not being heeded by researchers. As White (1976, p. 2)
cynically states, researchers preferred to repeat the old
familiar phrases, and tacitly assumed the process of splitting
open coherent rock by frost had been demonstrated
by someone; as also noted by Thorn (1992) nearly 20 years
later! White makes the very important distinction, not
made by many authors, between the breaking of previously
split rock and the breaking of solid rock (see Hall, 2006a
regarding the concept of solid rock). White then goes on to
make a very good case for much of what is generally thought
to be frost weathering being, in fact, the result of hydration
shattering. It is, however, the Questions for Reflection that
White raises at the end of his paper that are highly pertinent.
Apropos the problem regarding rock moisture content, White
(1976, p. 5) asks:
In how many mountain ranges or on how many arctic plains will
bedrock fortuitously ever become 450% water-saturated from
melting snow and rain and then undergo rapid freezing to crack the
rock? What are the chances that the restrictions of 450% watersaturation followed by sudden deep freeze overcoming the increasing strength of the rock as temperature decreases are always
satisfied, or satisfied only half the time, or are never satisfied?
One aspect of freezethaw weathering that has received

recent attention is that of two-sided freezing within areas
of permafrost. Work by Murton (Murton, 1996; Murton et al.,
2001, 2006; Matsuoka and Murton, 2008) provides convincing evidence for the production of segregation ice with
both downward and upward freezing and that this results in
the brecciation of the rock, which may itself be a good indicator of previous frost action (Murton, 1996). In addition to
raising a number of questions pertaining to the nature of this
type of frost action, they (Matsuoka and Murton, 2008) also
provide a useful summary regarding freezethaw weathering
processes, controls, and effects. Of all the proposals regarding
264
the nature and effects of frost action in rock, these studies by

Murton appear to offer, for permafrost areas with permeable
rock and adequate moisture, the most significant (and rigorously tested) advances yet. However, one contentious
issue must remain the notion of microgelivation and macrogelivation as cited by them, and particularly by Matsuoka
(2001), as an extension of the original work by Tricart (1956).
This issue has been dealt with from a variety of perspectives by
Hall (2006a) and Hall and Thorn (2011):
the fusion of climate geomorphology and process geomorphology
in the on-going development of the freezethaw concept has resulted in the confusion of product with process whereby microgelivation (producing small products) and macrogelivation
(producing large products) are seen, due to the product dichotomy,
as distinctly different processes.
Overall, the concept of Tricart (1956) is riddled with

what should be recognized today as misunderstandings
(Hall, 2006a) and yet we continue to promote the idea
without, seemingly, grasping the difference between process
and product, such that The inherent logical flaws within the
definitions of microgelivation and macrogelivation at the very
least hamper research pursued within their fuzzy frameworks
(Hall and Thorn, 2011). Ultimately, at the very least, we
should be asking the fundamental question, If we accept this
dichotomy, are there no other scales of operation as well?
Hall (2006a) and Hall and Thorn (2011) deal with these issues
at length.
Thus, the application of the freezethaw concept in geomorphology has all the attributes of a classic geomyth
(Dickinson, 2003, p. 856), wherein: The style of myth posits
dominance of a single perception to the exclusion of others.
This approach makes geomyths difficult to falsify y (and
they) y tend to be more durable but also less specific than
leading or ruling hypotheses. Only in recent years (see Andre,
2009, p. 12) has there begun a move away from the
freezethaw dogma to the recognition of other weathering
processes within cold regions.
4.15.2.3
Weathering by Wetting and Drying
Water is well known to exert an influence on the hardness,

strength, and compressibility of rock (Michalopoulos and
Triandafilidis, 1976). Further, wetting and drying of rock is
known to cause swelling and shrinkage to a whole range of
rocks such as basalt and flint (Nepper-Christensen, 1965),
schist, marble, obsidian, granite, slate, and limestone (Nishioka and Harada, 1958). Nepper-Christensen (1965) examined effects on basalt and flint and found that basalt shrank
0.015 to 0.020% from saturated to equilibrium at 65% relative
humidity a value comparable to the 0.020% expansion
identified by Nishioka and Harada (1958). Importantly, both
sets of experiments found that the greater part of expansion
(and contraction) took place in the 90% to 100% saturation
range. As a comparison, the contraction upon drying of
granite was in the order of 0.04% while for a temperature
range of 50 1C the contraction is in the order of 0.01%
(Nepper-Christensen, 1965, p. 554). Hudec and Sitar (1975),
working with carbonate rocks, suggested that their expansion
upon wetting was comparable to a thermal expansion of a

78 1C temperature range. The important factor is that, as
Nepper-Christensen (1965, p. 555) stated, total shrinkage or
swelling of rocks is frequently of the same order of magnitude
as the thermal changes in volume that accompany heating
and cooling at the earths surface. Further, there is a hysteresis
effect for, as Nishioka and Harada (1958, p. 66) show,
an andesite has a 0.000 07% expansion at 15% wetting, which
increases to a maximum of 0.000 36% at saturation but
reverts to 0.000 03% elongation at 15% wetting during
the drying phase. Hames et al. (1987) also examined the
expansion of various rock types (mostly limestones) and
found that the samples showed an elongation from 1 to
18 mm cm1 and that some columns did not revert to their
original length upon drying. Yatsu (1988) stated that the actual cracking occurs during this wetting phase (rather than
during the drying) as a result of the expansion strains. Pissart
and Lautridou (1984) used columns of Bathonian limestone
with a porosity of 25 to 28%, which, when immersed, expanded rapidly but then with time the expansion became
slower. They argued that the disintegration may well be caused
by changes in the humidity of the air surrounding a rock
surface, and showed that the changes in length of the rock
were relatively large despite only slight changes in humidity.
Ultimately, the net effect of this swelling and shrinking is to
cause fatigue and failure in much the same manner as thermal
stress; the process is azonal.
A difficult question to answer is what then, if anything,
does this azonal process have to do with cold regions?
The very nature of a cold environment necessitates that, for
significant periods, any available water will be frozen. Conversely, although spatially constrained, in nivation sites (see
Thorn and Hall, 2002) water continues to be released from a
melting snowpatch when the surrounding areas are snow free.
Here, wetting and drying may be enhanced, as indirectly
suggested by Berrisford (1991), and mechanical weathering
may be potentially influenced. Paradoxically, for cold regions,
it will be the warm periods that are conducive to this process
and the only relationship with cold will be in the provision
of snow that will later melt. Thus, episodic melting of snow
during the summer may facilitate spatially limited (see
the furrows of French and Guglielmin, 2002) weathering
due to wetting and drying. However, the role of this process in
cold regions has rarely been considered and needs greater
examination.
Importantly, wetting and drying may play a synergistic role
with other weathering processes. Pissart and Lautridou (1984)
saw wetting and drying as synergistic with freezethaw weathering, as too did Moss et al. (1981), who found that the
combination led to increased breakdown. Fahey (1983) considered the connectivity of frost action with hydration weathering within a laboratory setting and showed them to be
synergistic. It was found that (p. 535) hydration effects
were not as effective as frost action in causing particle size
reduction, as (p. 544) frost action y was y three to four
times more effective than hydration. Regarding these synergistic attributes, Mugridge and Young (1983, p. 568) investigated the interaction of wetting and drying with frost
action, and found that the combination of freezethaw with
wetting and drying is more rapid than disintegration by
wetting and drying only. Moss et al. (1981) also observed that
granite detritus broke down due to the combined action
of ice and adsorbed water, while Hall (1988c) found that there
was a synergistic interaction between the freezethaw stresses
and a hysteresis effect related to strength reduction due
to repeated wetting and drying. Hall (1988c) has argued
that freezethaw cannot operate without a wetting and drying
component. If this is correct, consideration of this process
must then be germane to weathering in cold regions.
Further, Hall (1991, 1993), from field studies in Alaska and
Antarctica, argued that wetting and drying frequently occurred
in a subsurface zone because the rocks were not always completely dry before being supplied with moisture again. Thus,
weathering by wetting and drying is taking place below the
rock surface and so (a) may not be recognized, or (b) may
facilitate other processes that ultimately break the rock.
Studies in both Alaska and Antarctica (Hall, 1991, 1993) also
clearly indicated significant aspect controls on the role of
wetting and drying.
Briefly, it is worth noting the findings of Hall and Hall
(1996), who examined the weathering effects resulting from
the manner in which water is provided to samples in weathering experiments as this relates to both weathering by
wetting and drying and the findings of freezethaw simulations. In many weathering simulations, rocks are supplied
with water by complete immersion, partial immersion, or
by pouring or spraying water over the rock (McGreevy and
Whalley, 1985). Using these three methods, samples of
sandstone and a dolerite were subjected to daily wetting and
drying. Results are not fully comprehended, but a number of
observations are relevant. First, material was weathered free,
with the covered and half-covered samples showing more
mass loss than the sprayed ones. However, the weathering
products from immersed and half-immersed samples were
splinters, whereas from the sprayed samples, there were
only granular products. Sumner and Loubser (2008), in a
laboratory study using sandstones, also found that the method of water application influenced the nature of weathering.
Internally, all samples showed an increase in percentage
saturation with time (reflecting an increase in the size of
pores or the number of pores) but covered 4 half-covered 4
sprayed. There was also a change in the time it took to
saturate samples: at the start of the experiment, it took
two days to fully saturate but by the end, it took five days.
4.15.2.4
Thermal Stress Fatigue/Thermal Shock
A number of early researchers (e.g., Bartlett, 1832) considered

the role of thermal stress fatigue or thermal shock (a.k.a.
insolation weathering) to be important, and Merrill (1906,
p. 159) observed flaking due to, what he termed, dry cold and
that In cold climates y the results of such flaking are
sometimes strikingly manifest, an apparent indirect reference
to thermal stresses. However, the laboratory experiments of
Blackwelder (1933) and Griggs (1936) failed to find any indication of weathering due to thermal stresses and so frequently was this evidence cited that it (Ollier, 1984, p. 20) led
to an almost complete abandonment of insolation for many
years. For example, Reiche (1950, p. 10) commented that
265
Griggs had published an account of a thoroughly convincing

series of laboratory experiments. However, there were several
researchers who were still aware of the validity of thermal
stress fatigue (e.g., Rice, 1976; Winkler, 1977; Garner; 1976),
and, as Ollier (1984, p. 20) notes, the experimental procedures of both Blackwelder and Griggs were flawed. Despite
results that clearly indicate the possibility of thermal stresses
(e.g., Hockman and Kessler, 1950; Richter and Simmons,
1974; Rodriguez-Rey and Montoto, 1978; Bahr et al., 1986;
Williams, 1986), geomorphologists, in the main, have ignored
the process as well as aspects such as buttressing that can
exacerbate thermal stresses (e.g., Winkler, 1984; Folk et al.,
1982).
Beyond geomorphology, many other disciplines continued
to study the role of thermal stresses. For example, Johnson and
Parsons (1944) studied 123 specimens of aggregate material as
well as crystals of calcite, quartz, and feldspar over the temperature range 20 1C to 60 1C at a heating rate of 0.5 1C
min1. Of significance, they found the orientation of crystals
to have an influence on expansivity the DL per degree Celsius
being positive for calcite parallel to the c-axis and negative
normal to the c-axis. For quartz, DL normal to the c-axis is
greater than parallel to the c-axis per degree C. Interestingly,
for cold regions, Johnson and Parsons (1944, p. 121) found
that y the materials expand on heating until temperatures
between 0 1C and 10 1C y with further heating they either
expand at a lesser rate y or contract until temperatures between 15 1C and 30 1C y when expansion is resumed.
Hockman and Kessler (1950, p. 409) carried out a comparable
series of studies using the same temperature range and concluded: It is probable that the repeated thermal expansion
and contraction of granite resulting from natural weather
conditions, is one of the many factors contributing to the
eventual disintegration of the stone. Kingery (1955, p. 3)
points out that stresses will arise due to difference in expansion between crystals and these may lead to serious
weakening or fracture. Further, stresses associated with heating
and cooling are complex as both tensile and compressive
stresses occur at the same time: On cooling, the maximum
stress is the tensile stress on the surface and the center is in
compression. On heating, the maximum stress is the compressive stress on the surface, and the center is in tension
(Kingery, 1955, p. 4). Importantly, within the context of cold
region studies, on cooling, primary stresses are tensile (Kingery, 1955), a point also made by Marovelli et al. (1966), and
rocks have a relatively low tensile strength. Further, Marovelli
et al. (1966) and Bahr et al. (1986) showed diagrams of
fracture patterns for basalt for DT values of 72 1F (22.2 1C)
through to 450 1F (232.2 1C) and these are comparable to the
fracture patterns seen on rocks in cold, dry regions that experience large thermal variations (Figure 1). As Bahr et al.
(1986) noted, crack density increases with the severity of
the thermal shock and initial flaws in the material that are
induced to grow into cracks by the thermal changes become
smaller as DT/Dt increases. There is, at least in the experiments
conducted by Bahr et al. (1986, p. 2719), a hierarchical order
of cracks that stands out clearly against the random background: something very noticeable in some crack patterns
occurring in the field on rocks considered to have been broken
by thermal shock.
266
Figure 1 Fracture patterns on a large block in the high Andes

(Argentina). Note that the fractures do not follow the lines of
weakness within the rock and that fractures are normal to each other
and show a hierarchy of sizes.
4.15.2.4.1
Thermal stress within the cold region context
Weathering by thermal stress has begun to receive increasing

recognition in cold environments (Andre, 2003), where it may
play a significant role (Hall and Andre, 2003). The cold region
context is that of cold air conditions but with rocks warmed by
solar radiation (Hall, 2006b) and the stresses that can result as
the warming influence is removed (Hall and Hall, 1991). Hall,
et al. (2008a, b), following Gomez-Heras et al. (2006),
showed differences in temperatures between grains in an
Antarctic situation and extended the argument of Andre et al.
(2004) regarding the possible influence of light transmissive
minerals on the rocks near-surface thermal regime. Thermal
stresses resulting from both mineral properties and translucent
minerals were explored further by Hall et al. (2010) and Hall
(2011). Regarding the cold environment of the planet Mars,
Leask and Wilson (2003, p. 1320) found that Abrupt illumination, or to a lesser extent shadowingy.as the sun moves
behind topography can induce thermal stresses large enough
to cause exfoliation or granular disintegration as long as the
changes in illumination last longer than B1 min. This work
further validates the finding of Hall (1999, 2004, 2006a), Hall
and Andre (2001), and Hall et al. (2008a, b), who showed that
with high-frequency thermal data, rock surfaces in cold environments can be shown to experience large-amplitude,
short-duration thermal fluctuations, often 420 1C min1. The
argument was made that such fluctuations, coupled with the
different thermal responses of the minerals in terms of expansion and contractions (e.g., Siegesmund et al., 2000),
could lead to grain-to-grain stresses and ultimately granular
disintegration. Recent studies by McKay et al. (2009) and
Molaro and McKay (2010) extended the work of Hall (1999,
2003, 2004) and investigated the nature of thermal changes at
1 s and 0.375 s intervals. With the faster record interval, they
obtained DT/Dt values as high as 137 1C min1 and found
that these high DT/Dt rates were correlated with wind speed.
Overall, as high as these rates are, their impact is scale constrained but they may be significant in granular disintegration.
McFadden et al. (2005) discussed the fracture patterns
generated on rocks in areas of extreme solar radiation receipt

and show the impact of the (p. 161) seasonally varying,
latitude-dependent solar elevation angle. Moores et al. (2008,
p. 472) showed that cracks in clasts (in the Southwest U.S.A.)
have a preferred orientation independent of rock fabric and
shape and that crack growth is by hydration or thermal
weathering and is primarily a function of local water at the
crack tip. Pointedly, it is the crack tips that experience less
solar energy receipt, and hence retain more moisture, and
that weather faster than those that receive more solar radiation. The findings of McFadden et al. (2005) and Moores
et al. (2008) have yet to be considered within a cold environment but there is no reason to believe that they would
not be important. Deng et al. (2009, p. 328) also note the
complexity regarding thermal south (i.e., the aspect receiving
the greatest thermal input), wherein A deviation angle may
therefore exist between the insolation (e.g., geographic) south,
and the thermal south the warmest slope azimuth. In the
same vein, Hall (2006b) reports the varying seasonal influence
of aspect and slope angle on rock temperatures whereas,
at a different scale, there is also the influence of the rocks
surface roughness on thermal receipt through the day (Hall
et al., 2008b).
4.15.2.5
Salt Weathering
Salt weathering is an azonal process whose operation

within cold regions is primarily linked with either an evaporative environment or one, as along ice-free coasts, from
which there is a direct input of salts to the land. In the case of
the cold arid environment (e.g., the Dry Valleys of Antarctica),
the role of salt is identical to that within a hot arid environment (see Goudie, 1999) it being in the evaporative
environment that is the primary reason for the process taking
place. That the salt, once in the rock, can undergo thermal
expansion and contraction or hydration/dehydration is not,
in itself, a cold region attribute, rather the cold environment
simply determines the timing and extent of these drivers.
Regarding salt weathering per se, good sources include Evans
(1970) or Doornkamp and Ibrahim (1990) for general discussions, Ugolini (1986) for cold environments, and Goudie
(1997) or Smith (1994) for arid/desert environments.
The specific climatic regime plays an important role in determining process timing and extent, temperature and humidity
cycling playing critical roles. Given the range of cold climates
(from wet, temperate (e.g., Scotland) to extremely cold
and arid (e.g., Antarctic Dry Valleys)), salt weathering can play
a varying role. Coastal attributes of salt weathering are
also azonal, the climatic situation being, again, one that inhibits, limits, or enhances the activity spatially and temporally.
In some cold region coastal zones, the impact of temperature
in causing water to freeze may actually inhibit salt action
by protecting the rock from penetration by saline solutions;
seasonally, high-latitude locations may also experience coastal
protection due to sea ice. In some specific geographic
sites (e.g., the Vestfold Hills, Schirmacher Oasis, and McMurdo
Dry Valleys of Antarctica), the presence of saline lakes provides another source of salts along the lake margins (e.g.,
Mcleod, 1964).
In the context of cold environments, the origin of salts can

be quite important as it determines which salts are present,
approximately 40 different secondary salt minerals having
been identified in Antarctica (Richter and Bormann, 1995).
Keys and Williams (1981) discussed the marine origin of salts
in the McMurdo region (Antarctica) and showed that sulfate,
chloride, sodium, and calcium salts are the most common,
but with the sodium and chloride more dominant nearer
the coast. Inland, Tasch and Angino (1968) identified sulfate
and carbonate salts as predominating. Importantly, for this region, biological, volcanic, and hydrothermal processes were
also seen as contributing locally. Recognition of which salts
are present is important for, as Robinson and Williams (2000)
have shown, weathering can be enhanced or diminished depending on the salt combination(s). Penguin rookeries provide
phosphate salts as well as a source of uric acid that can play
a role in chemical weathering and hence the formation of other
salts (Keys and Williams, 1981). A corollary to this is the biological (penguin) production of calcium phosphate minerals
that can coat rocks, thereby inhibiting the penetration of
moisture (Arocena and Hall, 2003). Many references (e.g.,
Gibson, 1962; Goudie, 1984; Richter and Bormann, 1995) deal
with salt efflorescence or encrustations that may cover the rock
surface but not necessarily cause any weathering. Richter and
Bormann (1995, p. 201) observed, at the Schirmacher Oasis,
that secondary salts form by Meltwaters containing salts of
mainly marine-aerosol origin enter the substratum and are later
transported back to the surface as a result of intense evaporation and can contain ions derived from chemical weathering.
Gypsum has also been found to be active in salt weathering in
Antarctica (Gibson, 1962; Hayashi and Miura, 1989; Richter
and Bormann, 1995) and in the Karakoram Mountains (Goudie, 1984), while Miotke (1983) identified calcite, mirabilite,
thenardite, and halite within the Darwin Mountains of Antarctica. Tedrow (1966) and Watts (1983) identified thenardite
and other salts in the high Canadian Arctic on Prince Patrick
Island and Ellesmere Island, respectively.
Fahey (1985, p. 110), noting that salts are not uncommon
in polar or alpine regions, used an experimental approach to
studying the role of salt weathering for rock breakdown in
cold climates. Although he found that the presence of salts did
not augment frost weathering (as had been supposed), it was
found to play an important role in causing particle size reduction. Salt weathering was determined to be active primarily
in the spring to fall period. Fahey (1985), in showing the
interaction of salt, wettingdrying, and frost processes, indicated the complexity of linking landforms with a specific
weathering process. Fitzner and Kalde (1991) also conducted a
laboratory simulation of frost and salt weathering and concluded that, combined, there was a greater weathering effectiveness. Miotke (1983) referred to the interaction of salt
fretting with both chemical weathering and frost shattering in
the Darwin Mountains and the Dry Valleys of Antarctica, as
was also recorded by Johnstone (1973). Williams and
Robinson (1991) and McGreevy (1982) also investigated the
synergy of frost and salt weathering, with the former finding
that some salts can enhance rock breakdown by frost and the
latter finding that certain salts in solution can inhibit frost
damage. McGreevy (1982, p. 485) perhaps best summed up
the issues where he stated.
267
There is a real danger that patterns of rock breakdown produced in

experiments will be more reflections of experimental design and
procedures rather than environmental conditions which rocks
might experience in nature.
The key problems in this regard are the manner in which

the specimens are frozen and the amount of salts provided.
The validity of these, and any experiments, once more
returns to:
Experiments tend to be performed under conditions favorable for
breakdown, but the question as to whether or not such conditions
persist in nature requires some thought. (McGreevy, 1982, p. 486)
Ultimately, the reality is the combination of frost shattering and salt weathering is a much more complex phenomenon than either of these two processes operating
independently (McGreevy, 1982, p. 487) and this begs the
question as to whether any of these processes ever operate
independently. Thus, there is still a need for more realitybased experimentation before any valid conclusions can be
drawn regarding synergistic weathering activities. Williams
and Robinson (1991) provided an excellent review of the
interaction of frost and salt weathering, outlining the process
combination(s) that may enhance or actually inhibit each
other. Holmer (1998, p. 52) discussed how thermal stresses
link with salt weathering, but considers the solar influence as
primarily important in drying the saline solution and then
causing expansion of the salt crystals. In the coastal context,
Cook (1952) detailed the interaction of salt with frost action
in marine locations and Holmer (1998) for the weathering
along the Swedish west coast.
4.15.2.6
Dilatation/Pressure Release
As a totally azonal process, dilatation (unloading, stress release: see Yatsu, 1988, p. 140) has no cold region-specific attribute. However, pressure release is commonly associated with
deglaciation (e.g., Harland, 1957) and thus may be of indirect
significance in cold regions that experience(d) glaciation;
Harland (1957) links exploitation of the dilatation fractures by
subsequent frost action to an increase in resulting rock failure.
This notion is an extension of earlier work by Lewis (1954), the
focus of these studies being the development of alpine terrain
and glacial valleys in particular. However, Twidale and Bourne
(2009, p. 152) argued that neither erosional nor glacial unloading is now considered a viable explanation for spalling
(dilatation fractures) but, at the same time (p. 158) Yet
pressure release cannot be discarded as ineffective, for it appears to be a significant contributory cause of expansion and
rupture. Simplistically, since early works (e.g., Twidale, 1973),
the argument has been that the land surface is consequent on
structure and not the converse (Yatsu, 1988, p. 145). However,
following Yatsu (1988, p. 150), the shape of the land surface
produces the structure at times and hence some degree of
parallel-to-relieving face fractures may be possible in some rock
types and under some circumstances. Indeed, Ishimaru and
Yoshikawa (2000, p. 54), working in Antarctica, observed rock
sheeting on nunataks that has grown since the exposure of the
nunataks, and argued that the causes of the sheeting are a result
268
of erosional unloading, gravitational body force (internal stress

of the rock body resulting from the overburden mass), tectonic
residual stress, and differential weathering. Thus, although
such direct dilatation-with-deglaciation associations as given
by Small (1972) are perhaps no longer valid, it would seem
that erosional unloading in glacial environments can still play
a role (Hambrey, 1994). Clearly, this is a complex problem
with multiple possible mechanisms: realignment of stress
trajectories (Twidale and Bourne, 2009, p. 158), gravitational
body force (Augustinus and Selby, 1990), or stress relief
from overburden removal. Whatever the case, fracturing (and
microfracturing) can result and it can be driven by glacial
erosion and unloading.
Thus, pressure release and the resultant dilatation cracking,
including microdilatation (see Crook and Gillespie, 1986),
requires a driver, but once existent, such fractures can be extensively exploited by other weathering processes. Less understood but perhaps significant to weathering is the formation of
microcracks. Given that microcracks are those planar discontinuities that are too small to be seen in a hand specimen
(Engleder, 1987, p. 31) and that the fracture relieves strain
normal to their surfaces (Engleder, 1987, p. 64), the question
arises as to the development of microfractures parallel to the
relieving face as surface material is lost. Thus, the question
arises, as yet untested within the context of weathering per se, as
to whether, given the residual stresses with a material, microfractures can be created in rock as weathering and gravity
continues to remove surface materials. If so, then this adds yet
another scale dimension to weathering and to the possible
interaction of this microdilatation with other weathering processes that can then exploit these microfractures. In relation to
all of this, the arguments of Twidale and Bourne (2009) are
recognized but with the caveat that more detailed studies are
still needed to accept or refute the nature of cracking at the
micro (nano-?) scale (Butenuth, 2001; Hall, 2006a).
4.15.2.7
Where Are We at Where Do We Go?
Where are we? In truth, we are still largely working within the
(possibly) false assumption that: Frost weathering is a fundamental geomorphic process operating from polar regions
to tropical high mountains, wherever rocks experience a
temperature fluctuation across freezing point and moisture is
present (Matsuoka and Murton, 2008, p. 195). Frost weathering may be the reality, but it is far from proven and
our approach tends towards the fulfillment of a geomyth
(Dickinson, 2003) rather than an objective evaluation of
weathering in cold regions. Questions regarding proof of
freezethaw operation and its unequivocal effect(s) in the
landscape continue to be avoided. Further, the simplistic
dichotomy of microgelivation and macrogelivation is, at
best, nave, and is likely more confusing than enlightening
(Hall and Thorn, 2011). Basically, with respect to freezethaw
weathering: It would seem that geomorphologists can better
go forward by forgetting the past and adopting an approach
more related to materials science coupled with holistic field
monitoring with data acquisition appropriate to a host of
scales and processes (Hall and Thorn, 2011). Using a terrible metaphor, it would indeed seem that we are slowly
weathering our historical bonds with freezethaw as the primary process and landform producer for cold regions and that
cracks are slowly starting to appear in the edifice cracks that
are often being actively covered over, but there nevertheless. In
some ways, given our paucity of data and the questions raised
for over 100 years, it is more amazing that freezethaw still
holds such dominance!
We are at a stage where appropriate data are finally
superseding untested assumptions and opening the door to
more holistic consideration of weathering and weathering not
necessarily predicated upon cold. Indeed, recent data indicate
just how hot rock temperatures can be, what thermal fluctuations occur, and when and if water is present, plus the local
impacts of aspect, slope, and roughness on the resulting
weathering regimes. Much is confounding both freezethaw
weathering and how we approach it, and much strongly suggests the importance of other weathering processes. Further, as
weathering starts to be considered over long time frames,
process variation and synergy is leading to new questions
regarding landform development. Overall, we are perhaps at
the cusp of actually going forward rather than continuing to
mark time, but we are still carrying a great deal of historical
baggage with us that we are slow to shed. With all of our
processes being azonal, rock properties are once more gaining
the importance they should have in evaluating weathering,
wherein the climate is but a control on the timing and efficacy
of the processes facilitated by the rock itself.
Where do we go? The reality is we can only advance if we
are prepared to challenge our core ideas and to consider that,
perhaps, there is the operation of (as of yet) unknown processes. Although deductive approaches have been the most
common, there is perhaps a need for a move to critical realism, wherein reality is viewed as both stratified and differentiated. Reality (being) made up of entities that interact
according to underlying laws related to underlying structures;
it is differentiated (Inkpen, 2005, p. 30), and with this approach, the level of our observation is not the only level at
which explanation can or should occur. Ultimately, perhaps
the most contentious, difficult, and problematic question we
should be asking is Have we really found all of the weathering processes? coupled with the necessity for revisiting
our perception/approach to currently known processes. For
example, Hall (1989) proposed the stress exerted by extraneous rock material forced into cracks by wind as a new weathering process; how many others might there be that we do
not see simply because we do not look? Perhaps more pertinent is a reconsideration of the conceptual frame within
which we look at our presently known weathering processes.
Even though process synergy is accepted, it is accepted within
our current individual process recognition and thus primarily
as to how (e.g.) frost action works with salt weathering. Perhaps it is not two named processes working together but rather a single process that is the combination of freezing and
thawing and salt weathering in some (unknown) manner (not
constrained or evaluated by our conceptual confines of the
two processes). Recognizing that the processes (any processes)
occur within a range of climates, the climate must exert its
specific influence on the nature (e.g., combination mixture, as
well as timing, duration, efficacy, etc.) of that weathering,
thereby validating any climate prefix (as in cold region
weathering). Accepting this, reconsideration of process combinations (or however we better term these but preferably
without carrying any of the old conceptual baggage in the
name) as weathering entities in their own right, the variability
within these process combinations being a derivative of the
climate, is perhaps the way forward for rethinking, monitoring, and applying weathering. With such an approach, the
misleading climatic prefixes we ascribe to weathering environments now become meaningful, and our approach to
linking weathering to landforms becomes less climate constrained as a landform is seen as more the product of changes
in climatic drivers and less of specific weathering processes.
Conceptually, this is difficult to articulate, and clearly more
difficult to undertake, laden as we are with such a strong
history of considering processes as individual entities (e.g.,
Bland and Rolls, 1998; French, 2007), even when they are
closely intertwined. However, once we can grasp the notion,
and perhaps younger researchers less burdened by the old
sacred cows (Thorn, 1988) have more capacity to facilitate
this way of thinking, we may start to evaluate weathering in
cold regions (and everywhere) in a more realistic fashion.
4.15.3
Landforms
Although a number of landforms closely connected with

weathering appear to be associated with, or linked to, cold
regions (e.g., Section 6 of Robinson and Williams (1994),
Chapter 9 of Ballantyne and Harris (1994), or Chapter 13 of
French (2007)), the reality is that: (1) weathering is but part of
the process synergy required to create the landform; and (2)
both the process(es) and the landform are ultimately azonal
to some degree. Indeed, within the cold region milieu, there is
almost as much connection of weathering with glacial landforms as there is with periglacial (see Chapters 26 and 27
of Robinson and Williams, 1994). Even within the context
of freezethaw process studies, Battle (see Whalley and
McGreevy, 1991 for a detailed review) conducted his studies in
relation to glaciers and the role of the bergschrund in cirque
development. Circular arguments, based on the presence of
angular clasts and angular clasts being the solely product of
freezethaw weathering, are a common foundation for explaining the origin of many landforms (e.g., the nivation
hollows of Hanvey and Lewis, 1991). Frequently, the identification of landforms within present (or recently) cold climates appears to require their origin to be through cold-based
weathering processes, particularly freezethaw, (e.g., altiplanation terraces (Boch and Krasnov, 1943) or mountain-top
detritus (Ballantyne and Harris, 1994)). This is not to say that
processes such as freezethaw may not have been involved but
rather that this is often the assumed starting point. In recent
years, there has been a slow change in perspective. For example, Andre et al. (2008) suggested that the stone runs of
the Falklands may be tropical in origin rather the product of
freezethaw. Although still invoking frost processes for the
formation of an alpine blockfield, Goodfellow et al. (2008)
also suggested that a mixture of other mechanical processes
play a role and, more importantly (p. 18), that the blockfield
developed during the Quaternary interglacial and interstadial periods and hence under changing climates. Similarly,
269
Andre (1995) found postglacial frost weathering on glacial

roches moutonnes to be ineffective.
4.15.3.1
Landform Associations
Mainstream thinking regarding cold region landforms and

weathering is well presented by French (2007). Here, the aim
is rather to try and reconsider our perceptions of weatheringrelated landforms to open new avenues of thought. For
example, although landforms such as nivation hollows or
cryoplanation terraces are, within the bulk of the literature
(see French, 2007), firmly linked to freezethaw weathering,
this remains more assumed than proven (Thorn and Hall,
2002). Indeed, where relevant data have become available,
more questions regarding the nature and timing of the weathering have been raised, rather than assumptions validated.
Although Berrisford (1991) observed that there was enhanced
mechanical weathering associated with snowpatches, this
varied spatially within a snowpatch site largely as a result of
moisture availability. Significantly, evidence did not suggest
that it was diurnal (short-term) freezethaw but rather that it
related to the annual freeze coupled with moisture availability
and to hydration-based weathering. If the findings of Berrisford have a larger geographical acceptance then this would be
an example of a specific weathering enhancement within a
cold-based environment but clearly not what is commonly
associated with nivation.
Although tafoni are clearly azonal, there has been extensive
discussion regarding their formation within cold climates (see
Strini et al., 2008); differences in terminology (e.g., honeycomb or alveolar forms) are more related to scale than process
(Roman et al., 2006, p. 812). Again, the formative processes,
wherein those causing the removal of the weathered material
seem less considered than those effecting breakdown (e.g.,
French, 2007), are completely azonal and the only
present cold connection is with respect to the timing and
magnitude of process(es). Indeed, Andre and Hall (2005) observed that current weathering of tafoni on Alexander Island
(Antarctica) appeared limited by the present cold, dry climate,
but there was evidence to suggest that it was more active under
milder conditions. French (2007, p. 73) argued that In spite of
all these uncertainties and ambiguities, the pervasive impression remains that salt is important in the development of
cavernous weathering, and certainly the occurrence of tafoni in
coastal and deserts worldwide makes for a good argument in
favor of salts. Here, a problem still remains: whether there is
any significant cold context? Another question linked to tafoni
is why is their form so rounded when (cold region) mechanical processes are deemed to produce characteristic angularity?
At the heart of the issue, and prevalent throughout cold region
landform consideration, is the unfounded assumption that
mechanical processes produce angular material (see Hall and
Thorn, 2011). Although many cracks are linear in form, they can
be curvilinear (Figure 2) as a result of internal rock structures,
and an untested suggestion regarding tafoni is that their form
reflects exploitation of curvilinear lines of weakness. For example, in the granites studied by Strini et al. (2008), there is no
reason to not consider macroscopic flow structures y folding,
layering, crystal alignment, concentric structures as part of a
270
Figure 2 Curvilinear weathering within a basalt (near Mt

Washington, Antarctica). Weathering in the hollow is along curvilinear
planes rather than straight (angular) lines.
Figure 3 Furrows and hollows in granites on Mt Keinath

(Antarctica). There is some evidence to suggest that, through time,
the furrows become ridges and the developing tubes become the
furrows in an on-going sequence of topographic reversal. Actual
weathering processes (as well as the age and the impact of climatic
change) are unknown.
granite body (Roman et al., 2006). Other studies by Alvarez

and Roman (2006a, b) further extend the linkage of rock
structure with resulting weathering forms in granite. This is a
theme in need of further consideration.
Also linked (at least spatially) with tafoni are grooves or
furrows observed in some rocks (particularly granites) in
Antarctica (Figure 3). French and Guglielmin (2002) referred
to these as cryogenic grooves due to meltwater associated
with repeated freezing and thawing. This, though, is misleading as there is (as of yet) no unequivocal evidence of a
cryogenic origin, the more so as (for example) Twidale and
Bourne (2003) referred to comparable forms (in granite) in
Australia, and Tobisch et al (1997) in California, and provide
noncryogenic origins. The subsurface origin suggested by
Twidale and Bourne (2003) for the origin of flutes in Australia
needs consideration as personal observation of the flutings
referred to by French and Guglielmin (2002) shows that they
are far more complex in form (suggesting an on-going sequence of growth and inversion) than the cryogenic hypothesis can explain.
Mountain-top detritus (see Ballantyne and Harris (1994)

or Boelhouwers (2004) for reviews) is strongly linked to
freezethaw weathering and a periglacial origin. In consideration of mountain-top autochthonous blockfields (rather than
allochthonous blockfields), descriptions are not unlike those
for stone pavements in deserts (see Cooke, 1970). The question is are they really different? Our perceptions are perhaps
colored by trying to compare mountain tops with extensive
desert surfaces. However, Cooke (1970, p. 560) identified the
desert pavement as an intricate mosaic of coarse angular or
rounded particles, usually one or two stones thick, set in deposits of sand, silt or clay, and that the pavements can occur
(p. 562) on weathered residue on bedrock surfaces, an absence of surface fines possibly being due to deflation by wind.
Although evidence indicates blockfields to have a noticeably
deeper weathering zone than stone pavements; Ballantyne and
Harris (1994, p. 173) referred to mountain-top blockfields
as including the presence of fines below the surface and a
tendency for coarse debris to decrease in size with depth and
that there is an absence of surface fines. The aim is not to
overstate or belabor the convergence of form, but two points
arise. First, once the notion of cold is removed, it is hard to
see, in terms of weathering associations, how they are not
comparable. Second, the operative weathering processes, including freezethaw (see Cooke, 1970, p. 573), may differ
little and thus the question must arise as to what then is different (regarding weathering attributes)? Indeed, this is the
very point made by Boelhouwers (2004, p. 133), where he
stated Mechanisms of block production increasingly appear
to be the result of azonal weathering processes such as
chemical weathering and thermal stress fatigue, with bedrock
fractures a prerequisite rather than an outcome of frost action.
Further, Boelhouwers noted (p. 133) that blockfields appear
to be the result of complex interactions and climatic histories
that defy straightforward palaeoclimatic interpretations. Further papers by Whalley et al. (2004), from the Arctic, and
Sumner et al. (2004), from the sub-Antarctic, reiterate many of
the points made by Boelhouwers. Whalley et al. (2004, p. 93)
made the point that laboratory investigations show no
freezethaw weathering products and, echoing Boelhouwers,
that this is due to the paucity of microcracks available to be
exploited. Further, they argued (p. 93) that the presence of
clays and the depth of the blockfield points to pre-Quaternary
weathering. Significantly within the context of the general
assumption of the role of freezethaw weathering in blockfield origin, the work by Whalley et al. (2004, p. 109) found
that It is difficult to explain the breakdown of bedrock into
fine material as seen in the blockfield by any mechanical
means usually attributed to freezethaw processes.
Landforms such as frost heaves of bedrock (e.g., French,
1993) may be relevant in this present discussion, and are
perhaps linked to two-sided freezing in permafrost areas (see
Matsuoka and Murton, 2008) in association with A-tents
(Jennings, 1978). A-tents, which may or may not (Twidale and
Bourne, 2009) be linked to pressure release from glacial unloading, consist of (Jennings, 1978, p. 34): pairs of roughly
rectangular slabs roofing a shallow cavity in the manner of a
wall-less A-tent. Frost heave of blocks, whereby ice grows in
fracture planes in association with two-sided freezing due to
the presence of permafrost and lifts those blocks is really a
form of weathering. If fractures are related to pressure release

and are then exploited by ice growth that moves the rock, then
that, if it causes (further) fracturing, is freezethaw weathering.
However, this idea hides within it the very questions raised
by Ives (1973) and White (1976) insofar as, if it is movement
of already separated blocks it is transport rather than
weathering and the same is the case in any other situation
where there is ice growth that cannot be proven to have caused
fracture (Hall and Thorn, 2011). Thus, if frost heaving of
bedrock in permafrost does involve fracture then it is really a
landform related to freezethaw weathering.
Thermal contraction cracks in bedrock (Figure 4) can cause
polygonal forms. Technically, there is nothing zonal about this
process and so, again, the cold-based relationship may be
spurious (e.g., polygonal fracturing of sandstone in Utah:
Howard and Selby, 1994). However, considering that thermal
stresses are greater during a decline in temperature, cold environments may be more conducive to this action. It remains
an azonal process, but one that is perhaps enhanced within
cold environments. That being so, bedrock contraction polygons may be more prevalent in cold environments, but are
certainly not indicative of former cold regions unless there is
evidence of ice having filled the cracks. Much is the same
argument for tors landforms occurring in (e.g., Derbyshire,
1972) and historically linked to (e.g., Palmer and Radley,
1961) cold regions, but that are azonal in occurrence (e.g.,
Gibbons, 1981). Although tors appear frequently within the
periglacial literature (see Ballantyne and Harris, 1994) and
some authors (e.g., French, 2007, p. 330) consider them to be
regarded as indicative of previous cold-climate conditions,
the reality is that their ability to be indicative is solely based
on foreknowledge of that previous cold climate and, even
then, they probably have a polygenetic origin (Ballantyne and
Harris, 1994, p. 181).
If this perceived indicative capacity of landforms (associated with weathering for cold regions) is examined, there is
really very little substance. In truth, the greater bulk of the
evidence is based on either the landform currently being
Figure 4 Thermal contraction cracks in bedrock, Alexander Island

(Antarctica). These mudstones show polygonal cracking due to
thermal stresses. The surface of the rock is weathered and shows no
signs of glacial debris. It is not known whether the cracks have a
filling of ice or sand.
271
found within a cold region or in one known to have been

recently cold. Given this knowledge, attempts are made to
provide supporting information generally, in the context of
weathering, based on observed angular clasts being the assumed product of freezethaw weathering and this is then
used as validation for the original cold premise (e.g., Hanvey
and Lewis, 1991). To have validity, there needs to be information that can, without foreknowledge of climate, exclusively indicate a cold environment origin. This also needs
to extend beyond climate by association (e.g., stratified
slope deposits), where the arguments remain circular and still
founded on freezethaw weathering: Intense freezethaw
action, together with slopewash activity is generally believed
to result in rhythmically stratified slope-waste deposits
(French, 2007, p. 330). A significant fallacy, common
throughout and yet used as an indicator of the climate
through the perceived weathering association, is the assumption (see Hall and Thorn, 2011) that clast angularity is indicative of frost action. Given an angular clast, what in that
form indicates frost action? If, with clast in hand (of unknown
provenance), one cannot unequivocally show why it has that
form due solely to frost action then the entire argument fails,
as does any landform construct utilizing that assumption. If,
for example, Al-Farraj and Harvey (2000) can discuss the use
of clast angularity for the classification of hot desert pavements, terraces and fans, then what meaning does the same
form index have in designating cold region landforms and
sediments? This is a topic worthy of greater discussion but
cannot be included here due to space constraints. However,
the many studies by Smith (e.g., Smith et al., 2005) regarding
hot deserts should be viewed from the perspective of: what, if
any differences are there from the landform origin and weathering questions raised for cold climates?
Smith and Warke (1997, p. 48 and Table 4.1) emphasized
how microscale structural discontinuities within rocks facilitates weathering, a point also forcibly made by Power et al.
(1988), the key being that our obsession with the weathering
processes has all but ignored the role of the material (the
rock) on which they work, and that, in turn, influence what
processes may, or may not, take place irrespective of climate,
recognizing that all the processes are azonal (Hall, 2006a;
Hall and Thorn, 2011). Hall (2006a) and Hall and Thorn
(2011) referred to this for cold regions, whereas (among others) Smith (1994) and Smith et al. (2005) dealt with exactly
the same for hot environments. Equally, rather than climate,
the focus should be on microclimate that which, at the
appropriate scale, affects the rock face or individual minerals
(Viles, 2005). The whole issue of climate measurement being
at a scale commensurate with the problem has been dealt with
(e.g., Thorn, 2003; Hall, 2006a) but there still needs to be
more rigorous application. Ground climates are the controlling weather phenomenon and not regional climates (Thorn,
personal communication, 2010)!
Thus, in summary, are there really any true cold-region
weathering landforms: ones that can be unequivocally ascribed
to a solely cold climate origin? Equally, and linked to the previous question, one needs ask, are there any unequivocal indicators of a cold climate process? Once the rhetoric, the
assumptions, the circular arguments, and, most importantly,
foreknowledge of the climate, are removed, I suspect the answer
272
to both questions is (at the moment) a resounding No! and

yet it appears, Nothing is so firmly believed as that which is
least known (de Montaigne, 1580). Indeed, the baggage we
carry, which constrains and limits our thinking, is that very
prefix cold, coupled with, as discussed by Yatsu (1992, p. 115),
authoritarianism, heroism and idolatry (so) deeply buried in
the mental substrata of geomorphologists that it appears to
maintain and perpetuate our cold region geomyths.
4.15.4
Where are We at and Where are We Going?
Very slowly, there seems to be a shift from the freezethaw

obsessed, cold-attributes fixation to a more holistic consideration of weathering in high-latitude/high-elevation locations.
At the same time, there has been an increase in actual data
collection pertaining to all aspects of weathering, as well as
recognition of the synergistic complexity of weathering processes. The more we obtain, the more complex process(es) are
found to be and the climaticgeomorphology constraints are
broken (notably by inclusion of chemical and biological processes), thereby adding to this overall complexity. Further,
more realistic attributes of scale are being entertained, particularly of micro (nano-) scale processes being the originators
of large-scale products. Not only are micro/nano-scale processes now seriously considered, but also scale attributes are
now being studied within proper scale boundaries, rather than
the previous somewhat arbitrary qualitative manner. Importantly, the dichotomy between process scale and product scale,
whereby processes can operate at a scale much smaller than the
ensuing product(s), has also been recognized. That said, such
studies are now only just beginning and there is a need for
much more research in this direction with experiments, primarily in the field, structured to clearly evaluate processes, and
temporal process variation/synergy, at appropriate scales.
Concurrent with the process evaluations there needs to be
greater recognition of the role that rock properties play in
controlling/determining those processes. Although some
authors have overtly identified this and urged us to take greater
recognition of rock properties, nevertheless, it has remained
subservient to the investigation of process per se. Simultaneous
with the consideration of rock properties there needs to be a
better application of rock fracture mechanics, thereby facilitating how process and rock properties interact to cause failure.
With such an approach, we will move away from our current
simplistic, assumptive notions of weathering in cold regions.
Compounding the above with ever-increasing recognition
of many landforms being extremely old (rather than just postglacial or even Quaternary in age), the ramifications of quite
different climates come into play and pave the way to the very
real possibility of many of our long-cherished cold region
landforms as having an origin under quite different environments, perhaps only being preserved during cold phases.
Our obsession with angular clasts equating to frost action is
slowly being overcome, thus destroying the many simplistic
underpinnings of landform evolution arguments. If we can
overcome our obsession with the climatic prefix to our weathering zone as the determinant of process, then it is possible
that the landform may owe a great deal of its origin to alternate climates or processes. By extension, many of the climate
prefixes ascribed to weathering and landforms may fall away

and we will find increasingly less fits into such preconceived
climatic boxes. Given this, the role of climate, which clearly
cannot be ignored, is now valid as a determinant, not of
process, but of process timing, efficacy, and extent, and is
interactive with spatial attributes (aspect, slope). In this
manner, climate becomes useful to our understanding rather
than the hindrance it currently is.
To conclude, perhaps I can return to the very opening sentences where there was an attempt to define weathering (at least
as it is used here), as this too underpins some of our problems
and possible future directions. A common definition, as cited
by Ollier (1984, p. 1), mirroring Reiche (1950), would be the
breakdown and alteration of materials near the Earths surface
to products that are more in equilibrium with newly imposed
physico-chemical conditions. However, we have perhaps come
to a juncture where our definition of weathering itself is no
longer functional and is blurring at its boundaries Surprisingly,
many texts, on both weathering and cold region processes, use
the term weathering but provide no definition as to what is
meant it appears we all understand and are in agreement. But
as Thorn (2009, personal communication) suggests regarding
the meaning of weathering: For those not actively researching
the topic, there is a widely accepted, but hazy, meaning; while
for those actively researching the topic, it is pervaded by burgeoning complexity as it is pursued at ever-more reductionist
scales; hence, the result is that terminology continues to be
used with old, outdated connotations, on the one hand while
simultaneously being used by some with new connotations
and in a worst case scenario the two groups talk past one
another when they think they are talking to one another. Thus,
perhaps, we are not as much in accord as we may assume.
Perhaps, as much as we need to reconsider weathering processes per se, so, simultaneously, we might want to reconstruct
what the conceptual framework that we are defining them
within actually is.
References
Al-Farraj, A., Harvey, A.M., 2000. Desert pavement characteristics on wadi terrace and
alluvial fan surfaces: Wadi Al-Bih, U.A.E. and Oman. Geomorphology 35, 279297.
Allison, R.J., 1990. Developments in a non-destructive method of determining rock
strength. Earth Surface Processes and Landforms 15, 571577.
Alvarez, E., lRoman, J.R.V., 2006a. Estructura y forma: studio experimental en
cacholas (tafoni) de Galicia (NW Peninsular). Cursos E Congresos Da
Universidade De Santiago De Compostela 171, 785793.
Alvarez, E., l Roman, J.R.V., 2006b. Trayectorias morphologicas de las pas
(vasques, ganammas). Cursos E Congresos Da Universidade De Santiago De
Compostela 171, 795802.
Andersson, J.G., 1906. Solifluction; a component of subaerial denudation. Journal
of Geology 14, 91112.
Andre, M.-F., 1995. Postglacial microweathering of granite roches moutonnees in
northern Scandanavia (Riksgransen area, 681 N). In: Slaymaker, O. (Ed.),
Steepland Geomorphology. John Wiley, Chichester, pp. pp. 103127.
Andre, M.-F., 2003. Do periglacial landscape evolve under periglacial conditions?
Andre, M.-F., 2009. From climatic to global change geomorphology: contemporary
shifts in periglacial geomorphology. In: Knight, J., Harrison, S. (Eds.),
Periglacial and Paraglacial processes and Environments. The Geological Society,
London, Special Publication, 320, pp. 528.
Andre, M.-F., Hall, K., 2005. Honeycomb development on Alexander Island, glacial
history of George VI Sound and palaeoclimatic implications (Two Steps Cliffs/
Mars Oasis, W. Antarctica). Geomorphology 65, 117138.
Andre, M.-F., Hall, K., Bertran, P., Arocena, J., 2008. Stone runs in the Falklands:
periglacial or tropical? Geomorphology 95, 524543.
Andre, M.-F., Hall, K., Comte, V., 2004. Optical rock properties of rocks and
weathering processes in polar environments (with special reference to
Antarctica). Polar Geography 27, 384403.
Arocena, L., Hall, K., 2003. Calcium phosphate accumulation on the Yalour Islands
(Antarctica): surface coatings and geomorphic implications. Arctic, Antarctic and
Augustinus, P.C., Selby, M.J., 1990. Rock slope development in McMurdo Oasis,
Antarctica, and implications for interpretation of glacial history. Geografiska
Annaler 72A, 5562.
Bahr, H.-A., Fischer, G., Weiss, H.-J., 1986. Thermal-shock crack patterns explained
by single and multiple crack propagation. Journal of Materials Science 21,
136140.
Balke, J., Haendal, D., Kruger, W., 1991. Contribution to the weathering-controlled removal of chemical elements from the active debris layer of the
Schirmacher Oasis, East Antarctica. Zeitschrift fur Geologische Wissenschaftern
19, 153158.
Ballantyne, CK, Harris, C., 1994. The Periglaciation of Great Britain. Cambridge
Bardin, V., 1962. Borozdy na skalakh. Information Bulletin, Sovetskaja
Antarkticeskafa Expedicisa 35, 6364.
Barsch, D., 1996. Rock-glaciers. Springer, Berlin, 331 pp.
Bartlett, W.H.C., 1832. Experiments on the expansion and contraction of
building stones, by variations of temperature. American Journal of Science 22,
136140.
Berrisford, M.S., 1991. Evidence for enhanced mechanical weathering associated
with seasonally late-lying and perennial snow patches, Jotunheimen, Norway.
Permafrost and Periglacial Processes 2, 331340.
Blackwelder, E., 1933. The insolation hypothesis of rock weathering. American
Bland, W., Rolls, D., 1998. Weathering: an Introduction to the Scientific Principles.
Boch, S.G., Krasnov, I.I., 1943. O nagornykh terraskh I drevnikh poverkhnostyakh
vyravnivaniya na Urale I svyazannykh s nimi problemakh. Vsesoyuznogo
Geograficheskogo obshchestva Izvestiya 75, 1425.
Boelhouwers, J., 2004. New perspectives on autochthonous blockfield development.
Polar Geography 28, 133146.
Branner, J.C., 1896. Decomposition of rocks in Brazil. Bulletin of the Geological
Society of America 7, 255314.
Brockie, W.J., 1972. Experimental frost shattering. Proceedings of the 5th New
Zealand Geographical Conference, pp. 177185.
Butenuth, C., 2001. Strength and Weathering of Rock as Boundary Layer Problems.
Imperial College Press, London, 270 pp.
Butterworth, B., 1964. The frost resistance of concrete and tiles: a review. Journal
of the British Ceramic Society 1, 203223.
Calkin, P., Cailleux, A., 1962. A quantitative study of cavernous weathering (tafoni)
and its application to glacial chronology in Victoria Valley, Antarctica. Zeitschrift
Chorley, R.J., Schumm, S.A., Sugden, D.E., 1984. Geomorphology. University
Clark, M.J. (Ed.), 1988. Advances in Periglacial Geomorphology. John Wiley,
Chichester, 481 pp.
Clark, S.P., 1966. Thermal conductivity. Geological Society of America Memoir 97,
461482.
Cook, H.K., 1952. Experimental exposure of concrete to natural weathering in
marine locations. American Society for the Testing of Materials 52, 11691180.
Cooke, R.U., 1970. Stone pavements in deserts. Annals of the Association of
Coutard, J.-P., Francou, B., 1989. Rock temperature measurements in two alpine
environments: implications for frost shattering. Arctic and Alpine Research 21,
399416.
Crook, Jr. R., Gillespie, A.R., 1986. Weathering rates in granite boulders measured
by P-wave speeds. In: Colman, S.M., Dethier, D.P. (Eds.), Rates of Chemical
Weathering of Rocks and Minerals. Academic Press, London, pp. 395417.
Davidson, G.P., Nye, J.F., 1985. A photoelastic study of ice pressure in rock cracks.
Cold Regions Science and Technology 11, 141153.
Davies, J.L., 1969. Landforms of Cold Climates. M.I.T Press, Cambridge, 200 pp.
Davison, J.I., Sereda, P.J., 1978. Measurement of linear expansion in bricks due to
freezing. Journal of Testing and Evaluation 6, 144147.
Day, M.J., 1981. Rock hardness and landform development in the Gunong Mulu
National Park, Sarawak, east Malaysia. Earth Surface Processes and Landforms
6, 165172.
273
De Montaigne, M., 1580. Essays. Book 1, Chapter 32, Gallimard, Bibliothe`que de la

Pleiade, Paris.
Deng, Y., Goodchild, M.F., Chen, X., 2009. Using NDVI to define thermal south in
several mountainous landscape of California. Computers and Geosciences 35,
327336.
Derbyshire, E., 1972. Tors, rock weathering and climate in southern Victoria Land,
Antarctica. In: Price, R.J., Sugden, D.E. (Eds.), Polar Geomorphology. Institute of
British Geographers, Special Publication, 42, pp. 93105.
Dickinson, W.R., 2003. The place and power of myth in geoscience: an associate
editors perspective. American Journal of Science 303, 856864.
Dixon, J.C., Abrahams, A.D. (Eds.), 1992. Periglacial Geomorphology. John Wiley,
Chichester, 354 pp.,
Doornkamp, J.C., Ibrahim, H.A.M., 1990. Salt weathering. Progress in Physical
Douglas, G.R., McGreevy, J.P., Whalley, W.B., 1983. Rock weathering by frostshattering processes. In: Proceedings of the 4th International Conference on
Permafrost. National Academy Press, Washington, DC, pp. 244248
Embleton, C., King, C.A.M., 1975. Periglacial Geomorphology. Edward Arnold,
London, 203 pp.
Engleder, T., 1987. Fracture Mechanics of Rocks. Academic Press, London.
Etzelmuller, B., Sollid, J.L., 1991. The role of weathering and pedological processes
for the development of sorted circles on Kvadehuksletta, Svalbard a short
report. Polar Reasearch 9, 181191.
Evans, D.J. (Ed.), 1994. Cold Climate Landforms. Wiley, Chichester, 526 pp.
Evans, I.S., 1970. Salt crystallization and rock weathering: a review. Revue de
Geomorphologie dynamique 19, 153177.
Fahey, B.D., 1983. Frost action an hydration as rock weathering mechanisms
on schist: a laboratory study. Earth Surface Processes and Landforms 8,
535545.
Fahey, B.D., 1985. Salt weathering as a mechanism of rock breakup in cold
climates: an experimental approach. Zeitschrift fur Geomorphologie 29, 99111.
Fahey, B.D., Gowan, R.J., 1979. Application of the sonic test to experimental freezethaw studies in geomorphic research. Arctic and Alpine Research 11, 253260.
Fitzner, B., Kalde, M., 1991. Simulation of frost-thaw cycle and salt weathering
nature adapted tests. Colloque International de la Deterioration des Materiaux de
Construction, Universite de Poitiers, La Rochelle, France, pp. 103114.
Folk, R.L., Patton, A., Patton, E.B., 1982. Buttressed expansion of granite and
development of grus in Central Texas. Zeitschrift fur Geomorphologie 26, 1732.
Fortin, G., 2010. Variabilite et frequence des cycles de gel-degel dans la region de
Quebec, 19772006. The Canadian Geographer 54, 196208.
French, H.M., 1993. Cold-climate processes and landforms. In: French, H.M.,
Slaymaker, O. (Eds.), Canadas Cold Environments. McGill-Queens University
Press, Montreal & Kingston, pp. 143167.
French, H.M., 2007. The Periglacial Environment. Wiley, Chichester, 458 pp.
French, H.M, Guglielmin, M., 2002. Observations on granite weathering
phenomena, Mount Keinath, Northern Victoria Land, Antarctica. Permafrost and
Fukada, M., 1972. Freezing-thawing process of water in pore spaces of rocks II.
Low Temperature Science, Series A 30, 183189.
Fukuda, M., Matsuoka, N., 1982. Pore water pressure profile in freezing porous
rocks. Low Temperature Science, Series A 41, 217224.
Fukuda, M., Miura, S., Nishiura, Matsuoka, N., 1983. The breaking of stone
remains caused by freezing and its protection by treatment with synthetic resins.
Science for Conservation 22, 114.
Garner, J., 1976. Insolation warmed over: comment. Geology 4, 264.
Gibbons, C.L.M.H., 1981. Tors in Swaziland. The Geographical Journal 147,
7278.
Gibson, G.W., 1962. Geological investigations in southern Victoria Land, Antarctica.
New Zealand Journal of Geology and Geophysics 5, 361374.
Godard, A., Andre, M.-F., 1999. Les milieux polaires. Armand Colin, Paris, 451 pp.
Godard, A., Hovel-Gangloff, F, 1965. Essais de gelifraction artificielle pratiques sur
des calcaires et des gres lorrains. Revue Geographique de lest 2, 125139.
Gomez-Heras, M., Smith, B.J., Fort, R., 2006. Surface temperature differences
between minerals in crystalline rocks: implications for granular disintegration of
granites through thermal fatigue. Geomorphology 78, 236249.
Goodfellow, B.W., Fredin, O., Derron, M.-C., Stroeven, A., 2008. Weathering
processes and Quaternary origin of an alpine blockfield in Arctic Sweden.
Boreas 10, 120.
Goudie, A.S., 1984. Salt efflorescences and salt weathering in the Hunza valley,
Karakorum Mountains, Pakistan. In: Miller, K.J. (Ed.), International Karakorum
Project. Cambridge University Press, Cambridge, pp. 607615.
Goudie, A.S., 1997. Weathering processes. In: Thomas, D.S.G (Ed.), Arid Zone
Geomorphology. John Wiley, Chichester, pp. 2539.
274
Goudie, A.S., 1999. Experimental salt weathering of limestones in relation to rock

properties. Earth Surface Processes and Landforms 24, 715724.
Grawe, O.R., 1936. Ice as ana agent of rock weathering: a discussion. Journal of
Geology 44, 173182.
Griggs, D.T., 1936. The factor of fatigue in rock exfoliation. Journal of Geology 44,
781796.
Guilcher, A., 1950. Nivation, cryoplanation et solifluction quaternaires dans les
colins de Bretagne Occidentale et du Nord de Devonshire. Revue
Geomorphologie dynamique, 5378.
Hall, K., 1986a. Rock moisture content in the field and the laboratory and its
relationship to mechanical weathering studies. Earth Surface Processes and
Hall, K., 1986b. Freeze-thaw simulations on quartz-micaschist and their implications
for weathering studies on Signy Island, Antarctica. British Antarctic Survey
Bulletin 73, 1930.
Hall, K., 1987. The physical properties of quartz-micaschist and their relationship to
mechanical weathering studies. Earth Surface Processes and Landforms 12,
137149.
Hall, K., 1988a. Daily monitoring of a block of indigenous rock at a Maritime
Antarctic site: moisture and weathering results. British Antarctic Survey Bulletin
79, 1725.
Hall, K., 1988b. A laboratory simulation of rock breakdown due to freeze-thaw in a
maritime Antarctic environment. Earth Surface Processes and Landforms 13,
369382.
Hall, K., 1988c. The interconnection of wetting and drying with freeze-thaw : some
new data. Zeitschrift fur Geomorphologie, N.F. Suppl. Bd. 71, 111.
Hall, K., 1989. Wind blown particles as weathering agents? An Antarctic example.
Hall, K., 1991. Rock moisture data from the Juneau Icefield (Alaska), and its
significance for mechanical weathering studies. Permafrost and Periglacial
Hall, K., 1993. Rock moisture data from Livingston Island (Maritime Antarctic) and
implications for weathering studies. Permafrost and Periglacial Processes 4,
245253.
regions. Geomorphology 31, 4763.
Hall, K., 2003. Micro-transducers and high-frequency rock temperature data:
changing our perspectives on rock weathering in cold regions. In: Phillips, M.,
Springman, S.M., Arenson, L.U. (Eds.), Proceedings of the 8th International
Conference on Permafrost. Balkema, Lisse, pp. 349354.
weathering in northern Canada. Earth Surface Processes and Landforms 29,
4357.
Hall, K., 2006a. Perceptions of rock weathering: some thoughts on attributes of
scale. Geomorphologie 3, 187196.
Hall, K., 2006b. Monitoring of thermal conditions in building stone with particular
reference to freeze-thaw events. In: Kourkoulis, S.K. (Ed.), Fracture and Failure
of Natural Building Stones: applications in the Restoration of Ancient
Monuments. Springer, Dordrecht, pp. 373394.
weathering in northern Canada, II: the temperature at which water freezes in
rock. Earth Surface Processes and Landforms 32, 249259.
Hall, K., 2011. Natural building stone composed of light-transmissive lithologies:
impacts on thermal gradients, weathering and microbial colonization.
Environmental Earth Sciences 62, 289297.
Hall, K., Andre, M.-F., 2001. New insights into rock weathering as deduced from highfrequency rock temperature data: an Antarctic study. Geomorphology 41, 2335.
Hall, K., Boelhouwers, J., 2003. Bibliography on Southern hemisphere permafrost
and periglacial landforms and processes. Circumpolar Active-Layer Permafrost
System (National Snow and Ice Data Center CD-ROM) 3, 97.
Hall, K., Cullis, A., Morewood, C., 1989. Antarctic rock weathering simulations:
simulator design, application and use. Antarctic Science 1, 19.
Hall, K., Hall, A., 1991. Thermal gradients and rates of change of temperature in
rock at low temperature: new data and significance for weathering. Permafrost
and Periglacial Processes 2, 103112.
Hall, K., Hall, A., 1996. Weathering by wetting and drying: some experimental
results. Earth Surface Processes and Landforms 21, 365376.
Hall, K., Thorn, C., 2011. The historical legacy of spatial scales in cold region
weathering: misrepresentation and resulting misdirection. Geomorphology. 130,
8390.
Hall, K., Thorn, C., Matsuoka, N., Prick, A., 2002. Weathering in cold
regions: some thoughts and perspectives. Progress in Physical Geography 4,
576602.
Hall, K., Guglielmin, M., Strini, A., 2008a. Weathering of granite in Antarctica I:
light penetration in to rock and implications for rock weathering and endolithic
communities. Earth Surface Processes and Landforms 33, 295307.
Hall, K., Guglielmin, M., Strini, A., 2008b. Weathering of granite in Antarctica II:
thermal data at the grain scale. Earth Surface Processes and Landforms 33,
475493.
Hall, K., Meiklejohn, I., Sumner, P., Arocena, J., 2010. Light penetration into
Clarens sandstone and implications for deterioration of San rock art.
Geoarchaeology 25, 122136.
Hallet, B., 1983. The breakdown of rock due to freezing: a theoretical model. In:
Proceedings of the 4th International Conference on Permafrost. National
Academy Press, Washington, DC, pp. 433438.
Hambrey, M. 1994. Glacial Environments. UCL Press, London, 296 pp.
Hames, V., Lautridou, J.-P., Ozer, A., Pissart, A., 1987. Variations dilatometriques de
roches soumises a` des cycles humidification-sechage. Geographie physique et
Quaternaire 41, 345354.
Hanvey, P.M., Lewis, C.A., 1991. Sedimentology and genesis of slope deposits at
Sonskyn, Eastern cape Drakensberg, South Africa. Permafrost and Periglacial
Processes 2, 3138.
Harland, W.B., 1957. Exfoliation joints and ice action. Journal of Glaciology 3(21),
810.
Hayashi, M., Miura, K., 1989. Glacial landforms and weathering processes in the
Balchenfjella region, eastern part of the Sr Rondane Mountains. East Antarctica.
Proceedings of the NIPR Symposium on Antarctic Geoscience 3, 6580.
Hockman, A., Kessler, D.W., 1950. Thermal and moisture expansion studies of
some domestic granites. Research of the U.S. National Bureau of Standards 44,
433438.
Hodder, A.P.W., 1976. Cavitation-induced nucleation of ice: a possible mechanism
for frost-cracking in rocks. New Zealand Journal of Geology and Geophysics 19,
821826.
Holmer, B., 1998. Flaking by insolation drying and salt weathering on the Swedish
west coast. Zeitschrift fur Geomorphologie 42, 3955.
Howard, A.D., Selby, M.J., 1994. Rock slopes. In: Abrahams, A.D., Parsons, A.J.
(Eds.), Geomorphology of Desert Environments. Chapman & Hall, London,
pp. 123172.
Hudec, P.P., Sitar, N., 1975. Effect of water sorption on carbonate rock expansivity.
Canadian Geotechnical Journal 12, 179186.
Inkpen, R., 2005. Science, Philosophy and Physical Geography. Routledge,
Abingdon, 164 pp.
Ishimaru, S., Yoshikawa, K., 2000. The weathering of granodiorite porphyry in the
Thiel Mountains, inland Antarctica. Geografiska Annaler 82A, 4557.
Ives, J.D., 1973. Arctic and alpine geomorphology a review of current outlook
and notable gaps in our knowledge. Research in Polar and Alpine
Geomorphology. GeoAbstracts, Norwich, pp. 110.
Jennings, J.N., 1978. Genetic variety in A-tents and related features. Australian
Geographer 14, 3438.
Johnson, W.H., Parsons, W.H., 1944. Thermal expansion of concrete aggregate
materials. Journal of Research of the U.S. National Bureau of Standards 32,
101126.
Johnstone, J.H., 1973. Salt weathering processes in the McMurdo Dry Valley
regions of South Victoria Land, Antarctica. New Zealand Journal of Geology and
Geophysics 16, 221224.
Keeble, A.B., 1971. Freeze-thaw cycles and rock weathering in Alberta. The Albertan
Geographer 7, 3442.
Keys, J.R., Williams, K., 1981. Origin of crystalline, cold desert salts in the McMurdo
region, Antarctica. Geochimica et Cosmochemica Acta 45, 22992309.
Kingery, W.D., 1955. Factors affecting thermal stress resistance of ceramic
materials. Journal of the American Ceramic Society 38, 317.
Lautridou, J.-P., 1984. Les experimentations dans le domaine de la cryoclastie.
Centre de Geomorphologie, Bulletin 27, 1538.
Lautridou, J.-P., Ozouf, J.-C., 1982. Experimental frost shattering. Progress in
Lautridou, J.-P., Seppala, M., 1986. Experimental frost shattering of some
Precambrian rocks, Finland. Geografiska Annaler 68A, 89100.
Leask, H.J., Wilson, L., 2003. Heating and cooling of rocks on Mars: consequences
for weathering. 34th Annual Lunar and Planetary Science Conference, Texas,
Abstract # 1320.
Le Ber, M., Oter-Duthoit, L., 1987. Dsagregation des parois rocheuses et climat:
approche thermique et thermodynamique. Bulletin de lassociation francaise pour
letude du Quaternaire 31, 147159.
Letavernier, G., Ozouf, J.-C., 1987. Resultats des recherches recentes sur la
gelifraction des calcaires. Inter-Nord 18, 5365.
Lewis, W.V., 1954. Pressure release and glacial erosion. Journal of Glaciology
2(16), 417422.
ozinski, M.W., 1909. Uber die mechanische Verwitterung der sandsteine im
gemassigten Klima. Adadmie sci. cracovie Bull. Internat., cl. Sci. math. Et
naturelles 1, 125.
Luquer, L.McI., 1895. The relative effects of frost and the sulphate of soda
efflorescence tests on building stones. American Society of Civil Engineers 33,
235256.
Mante, Cl., Ozouf, J.-C., Coutard, J.-P., Fraugeres, L., Hiolle, Ph., 1983. Deux
applications de LAnalyse des donnees en geomorphologie. Centre de
Geomorphologie, 115.
Marovelli, R.L., Chen, T.S., Veith, K.F., 1966. Thermal fragmentation of rock.
American Institute of Mining and Metallurical and Petroleum Engineering 235,
115.
Martini, A., 1967. Preliminary studies of frost weathering of certain rock types from
the west Sudetes. Biuletyn Pergglacjany 16, 147194.
Masseport, J., 1959. Premiers resultats dexpereriences de laboratoire sur les
roches. Revue de Geographie Alpine 47, 531537.
Matsuoka, N., 1988. Laboratory experiments on frost shattering of rocks.
Science Report of the Institute of Geosciences, University of Tsukuba, Section A
9, 136.
Matsuoka, N., 1990. The rate of bedrock weathering by frost action: field
measurement and a predictive model. Earth Surface Processes and Landforms
15, 7390.
Matsuoka, N., 2001. Microgelivation versus macrogelivation: towards bridging the
gap between laboratory and field frost weathering. Permafrost and Periglacial
Matsuoka, N., Murton, J., 2008. Frost weathering: recent advances and future
decisions. Permafrost and Periglacial Processes 19, 117.
McCarroll, D., Ballantyne, C.K., Nesje, A., Dahl, S.-O., 1995. Nunataks of the last
ice sheet in northwest Scotland. Boreas 24, 305323.
McCarroll, D., Viles, H., 1995. Rock-weathering by the lichen Lecidea auriculata in
an arctic alpine environment. Earth Surface processes and Landforms 20,
199206.
McFadden, L.D., Eppes, M.C., Gillespie, A.R., Hallet, B., 2005. Physical weathering
in arid landscapes due to diurnal variation in solar heating. Geological Society
McGreevy, J.P., 1981. Some perspectives on frost shattering. Progress in Physical
Geography 5, 5675.
McGreevy, J.P., 1982. Frost and salt weathering: further experimental results. Earth
McGreevy, J.P., Whalley, W.B., 1982. The geomorphic significance of rock
temperature variations in cold environments: a discussion. Arctic and Alpine
McGreevy, J.P., Whalley, W.B., 1984. Weathering. Progress in Physical Geography
8, 543569.
McGreevy, J.P., Whalley, W.B., 1985. Rock moisture content and frost weathering
under natural and experimental conditions: a comparative discussion. Arctic and
data from hyper-arid desert environments in the Atacama and Antarctic Dry
Valleys and implications for weathering. Geomorphology 110, 182187.
Mcleod, I.R, 1964. The saline lakes of the Vestfold Hills, Princess Elizabeth Land.
In: Adie, R (Ed.), Antarctic Geology. North Holland Publ., Amsterdam, pp.
6572.
Merrill, G.P., 1906. A treatise on rocks. Rock-Weathering and Soils. Macmillan,
London, 400 pp.
Michalopoulos, A.P., Triandafilidis, G.E., 1976. Influence of water on hardness,
strength, and compressibility of rock. Bulletin of the Association of Engineering
Geologists 13, 122.
Miotke, F.-D., 1983. Salt fretting and chemical weathering in the Darwin Mountains
and the Dry Valleys, Victoria Land, Antarctica. Polar Geography and Geology 7,
83122.
Molaro, J.L., McKay, C.P., 2010. Processes controlling rapid temperature
variations on rock surfaces. Earth Surface Processes and Landforms 35,
501507.
Moores, J.E., Pelletier, J.D., Smith, P.H., 2008. Crack propagation by differential
insolation on desert surface clasts. Geomorphology 102, 472481.
Moss, A.J., Green, P., Hutka, J., 1981. Static breakage of granitic detritus by ice
and water in comparison with breakage by flowing water. Sedimentology 28,
261272.
275
Mugridge, S-J, Young, H.R., 1983. Disintegration of shale by cyclic wetting

and drying and frost action. Canadian Journal of Earth Science 20,
568576.
Murton, J.B., 1996. Near-surface bercciation of chalk, Isle of Thanet, south-east
England: a comparison with ice-rich brecciated bedrocks in Canada and
Spitsbergen. Permafrost and Periglacial Processes 7, 153164.
Murton, J.B., Coutard, J.-P., Lautridou, Ozouf, J.-C., Robinson, D.A., Williams,
R.B.G., 2001. Physical modeling of bedrock brecciation by ice segregation in
permafrost. Permafrost and Periglacial Processes 12, 255266.
Murton, J.B., Peterson, R., Ozouf, J.-C., 2006. Bedrock fracture by ice segregation
in cold regions. Science 314, 11271129.
Nepper-Christensen, P., 1965. Shrinkage and swelling of rocks due to moisture
movements. Meddelser fra Dansk Geologisk Forening 15, 548555.
Nishioka, S., Harada, T., 1958. Elongation of stones due to absorption of water. Japan
Cement Engineering Association, Review of the 12th Meeting, Tokyo, pp. 6667.
Ollier, C., 1984. Weathering. Longman, Harlow, 270 pp.
Ozouf, J.-C., 1987. Etude experimentale sur la gelifraction des calcaires et silex
signification climatique. In: Miskovsky, J.-C. (Ed.), Geologie de la Prehistoire.
Association pour lEtude de lEnvironment Geologique de la Prehistoire. Paris,
pp. 403412.
Palmer, J, Radley, J., 1961. Gritstone tors of the English Pennines. Zeitschrift fur
Pissart, A., Lautridou, J.-P., 1984. Variations de longueur de cylinders de Pierre de
Caen (calcaire bathonien) sous leffect de secharge et dhumidification. Zeitschrift
fur Geomorphologie, Supplement Bund 49, 111116.
Potts, A.S., 1970. Frost action in rocks: some experimental data. Transactions of the
Institute of British Geographers 49, 109124.
Power, T., Smith, B.J., Whalley, W.B., 1988. Fracture patterns and grain release in
physically weathered granitic rocks. In: Soil Micromorphology and Applied
Science. Elsevier, Amsterdam, pp. 545550.
Reiche, P., 1950. A Survey of Weathering Processes and Products University of
New Mexico Albuquerque, Geological Series, vol. 3, 95 pp.
Rice, A., 1976. Insolation warmed over. Geology 5, 61190.
Richter, D., Simmons, G, 1974. Thermal expansion behaviour of igneous rocks.
International Journal of Rock mechanics, Mining Science and Geomechanics
Abstracts 11, 403411.
Richter, W., Bormann, P., 1995. Geomorphology. In: Bormann, P., Fritzsche, D
(Eds.), The Schirmacher Oasis, Queen Maud Land, East Antarctica, and its
Surroundings. Thomas Muntzer, Bad Langensalza, pp. 171206.
Ritchie, T., Davison, J.I., 1968. Moisture content and freeze-thaw cycles on masonry
materials. Journal of Materials 3, 658671.
Ritter, D.F., Kochel, R.C., Miller, J.R., 2002. Process Geomorphology. McGraw Hill,
New York, 560 pp.
Robinson, D.A., Williams, R.B.G. (Eds.), 1994. Rock Weathering and Landform
Evolution. John Wiley, Chichester, 519 pp.
Robinson, D.A., Williams, R.B.G., 2000. Experimental weathering of sandstone by
combination of salts. Symposium on the Deterioration and Protection of Stone
Monuments, Paris, pp. 114.
Rodriguez-Rey, A., Montoto, M., 1978. Insolation weathering phenomena in
crystalline rocks. Alteration et protection des monuments en pierre. Deterioration
and protection of stone monuments, Colloque International. International
Symposium, Paris, 59 June, pp. 18.
Roman, J.R.V., Alvarez, E., Mosquera, D.F., Sanchez, J.S., 2006. Proposal of
Nomenclature for granitic forms. Cursos E Congresos Da Universidade De
Santiago De Compostela 171, 803814.
Sass, O., 1998. Die Steuerung von Steinschlagmenge und -verteilung durch
Mikroklima, Gesteinsfeuchte und Gesteinseigenschaften im westlichen
Karwendelbirge (Bayerische Alpen). Munchener Geographische Abhandlungen
Reihe B B29, 175.
Sass, O., 2004. Rock moisture fluctuations during freeze-thaw cycles: preliminary
results derived from electrical resistivity measurements. Polar Geography 28, 1331.
Sass, O., 2005. Rock moisture measurements: techniques, results, and implications for weathering. Earth Surface Processes and Landforms 30,
359374.
Schneebeli, M., Fluhler, H., Gimmi, T., Wydler, H., Laser, H.-P., Baer, T., 1995.
Measurements of water potential and water content in unsaturated crystalline
rock. Water Resources Research 31, 18371843.
Selby, M.J., 1985. Earths Changing Surface. Clarendon Press, Oxford, 607 pp.
Seppala, M., 2004. Wind as a Geomorphic Agent in Cold Climates. Cambridge
Siegesmund, S., Ullemeyer, K., Weiss, T., Tschegg, E., 2000. Physical weathering of
marbles caused by thermal anisotropic expansion. International Journal of Earth
276
Small, R.J., 1972. The Study of Landforms. Cambridge University Press,

Cambridge, 486 pp.
Smith, B., Warke, P., 1997. Controls and uncertainties in the weathering
environment. In: Thomas, D.S.G. (Ed.), Arid Zone Geomorphology. John Wiley,
Smith, B.J., 1994. Weathering processes and landforms. In: Abrahams, A.D.,
Parsons, A.J. (Eds.), Geomorphology of Desert Environments. Chapman and
Smith, B.J., Warke, P.A., McGreevy, J.P., Kane, H.L., 2005. Salt-weathering
simulations under hot desert conditions: agents of enlightenment or perpetuators
of preconceptions? Geomorphology 67, 211227.
Van Steijn, H., 1979. Experimentele gelifractie, voorlopige resultaten van enkele
proeven. Geografisch Instituut Rijksununiversiteit, Berichten Fysisch Geografisch
12, 1119.
Strini, A., Guglielmin, M., Hall, K., 2008. Tafoni development in a cryotic
environment: an example from Northern Victoria Land. Earth Surface Processes
Sumner, P.D., Loubser, M.J., 2008. Experimental sandstone weathering using
different wetting and drying moisture amplitudes. Earth Surface Processes and
Sumner, P.D., Meiklejohn, K.I., Nel, W., Hedding, D.W., 2004. Thermal attributes of
rock weathering: zonal or azonal? A comparison of rock temperatures in
deifferent environments. Polar Geography 28, 7992.
Talkington, R.W., Mayewski, P.A., Gaudette, H.E., 1982. Geochemical and glaciogeomorphic implications of basalt weathering in the Queen maud Mountains,
Antarctica. Geological Magazine 119, 553566.
Tasch, P., Angino, E.E., 1968. Sulphate and carbonate efflorescences from the
Antarctic interior. Antarctic Journal of the United States 3, 239240.
Tedrow, J.C.F., 1966. Polar desert soils. Soil Science Society of America Journal
30, 381387.
Thompson, H.R., 1959. Bernard Palissy on frost shattering and river ice. The
Thorn, C.E., 1988. Nivation: a geomorphic chimera. In: Clark, M.J. (Ed.), Advances
in Periglacial geomorphology. John Wiley, Chichester, pp. 331.
Thorn, C.E., 1992. Periglacial geomorphology. what, where, when? In: Dixon, J.C.,
Abrahams, A.D. (Eds.), Periglacial Geomorphology. John Wiley and Sons,
Chichester, UK, pp. 130.
Thorn, C.E., 2003. Making the most of new instrumentation. Permafrost and
Thorn, C., Hall, K.J., 1980. Nivation : an Arctic-Alpine comparison and reappraisal.
Thorn, C., Hall, K., 2002. Nivation and cryoplanation: the case for scrutiny and
integration. Progress in Physical Geography 4, 533550.
Tobisch, O.T., McNulty, B.A., Vernon, R.H., 1997. Microgranitoid enclave swarms in
granitic plutons, central Sierra Nevada, California. Lithos 40, 321339.
Tricart, J., 1956. Etude experimentale du probleme de la gelivation. Biuletyn
Peryglacjalny 4, 285318.
Tricart, J., 1963. Geomorphologie des regions froides. Presses Universitaires de
France, Paris, 289 pp.
Twidale, C.R., 1973. On the origin of sheet jointing. Rock Mechanics 5, 163187.
Twidale, C.R., Bourne, J.A., 2003. Origin and inversion of fluting in granitic rocks.
Twidale, C.R., Bourne, J.A., 2009. On the origin of A-tents (pop-ups), sheet
structures, and associated forms. Progress in Physical Geography 33, 147162.
Ugolini, F.C., 1986. Processes and rates of weathering in cold and polar desert
environments. In: Colman, S.M., Dethier, D.P. (Eds.), Rates of Chemical
Viles, H.A., 2005. Microclimate and weathering in the central Namib Desert,
Walder, J., Hallet, B., 1985. A theoretical model of the fracture of rock during
freezing. Geological Society of America Bulletin 96, 336346.
Walder, J.S., Hallet, B., 1986. The physical basis of frost weathering: toward a more
fundamental and unified perspective. Arctic and Alpine Research 18, 2732.
Warren, S.H., 1914a. Some points on the Eolithic controversy. Geological magazine
6, 546548.
Warren, S.H., 1914b. The experimental investigation of flint fracture and its
application to problems of human implements. Journal of the Royal
Anthropological Institute 44, 412450.
Washburn, A.L., 1973. Periglacial processes and Environments. Arnold, London,
320 pp.
Watts, S.H., 1983. Weathering processes and products under arid arctic conditions.
Geografiska Annaler 65A, 8598.
Whalley, W.B., McGreevy, J.P., 1983. Weathering. Progress in Physical Geography
7, 559586.
9, 559581.
11, 358370.
Whalley, W.B., McGreevy, J.P., 1991. The contribution of W.R.B. Battle to
mechanical weathering studies. Permafrost and Periglacial Processes 2,
341346.
Whalley, W.B., Rea, B.R., Rainey, M.M., 2004. Weathering, blockfields and fracture
systems and the implications for long-term landscape formation: some evidence
from Lyngen and ksfjordjkelen areas in northern Norway. Polar Geography
28, 93119.
White, S.E., 1976. Is frost action really only hydration shattering? A review. Arctic
and Alpine Research 8, 16.
Williams, R.E., 1986. The thermal spallation drilling process. Geothermics 15,
1722.
of salts a review. Permafrost and Periglacial Processes 2, 347354.
Wiman, S., 1963. A preliminary study of frost weathering. Geografiska Annaler 45,
113121.
Winkler, E.M., 1977. Insolation warmed over: comment and reply. Geology 5,
188190.
Winkler, E.M., 1984. Buttressed expansion of granite and development of grus in
central Texas: discussion. Zietschrift fur Geomorphologie 28, 383384.
Yatsu, E., 1988. The Nature of Weathering: an Introduction. Sozosha, Tokyo, 624 pp.
Yatsu, E., 1992. To make geomorphology more scientific. Transactions, Japanese
Geomorphological Union 132, 87124.
Yershov, E.D., 1998. General Geocryology. Cambridge University Press, Cambridge,
580 pp.
Biographical Sketch
Kevin has worked for over 40 years on mechanical weathering processes in the laboratory as well as in both the
Antarctic and Arctic, in Tibet, the Andes, the Canadian Rockies, and the Drakensberg mountains (plus the Namib,
Sahara, Kalahari, Mojave, and Gobi deserts). His goal has always been to find ways to obtain suitable data in these
harsh environments and to use these to test (rather than just accept) old theory and to create new ideas. For him,
observations (and data) suggest just how small the difference is between the Namib and the Antarctic in terms of
both process and landforms, and the overwhelming conclusion he comes to is just how little data we have and
how little we really know. Having had a few lucky insights, the bulk of his work owes much to colleagues, notably
Colin Thorn, David Walton, Ian Meiklejohn, Marie-Franc- oise Andre, Brian Whalley, Augusto Perez-Alberti, Bernie
Smith, and Paul Sumner. They often provided the ideas, he simply had the opportunities.
4.16 Soil Chronosequences

GL Schmid, Minnesota State University, Mankato, MN, USA
4.16.1
4.16.2
4.16.2.1
4.16.2.2
4.16.2.3
4.16.3
4.16.4
4.16.4.1
4.16.4.2
4.16.4.3
4.16.4.4
4.16.4.5
4.16.5
References
Introduction
Soil Characteristics Supporting Chronosequence Development
Organic Enrichment
Leaching of Soluble Salts
Translocation of Clay Minerals
Issues Complicating the Development and Use of Chronosequences
Chronosequence Applications
Marine and Fluvial Terraces
Glacial and Periglacial Chronosequences
Vegetation Change and Management Issues
Fire Disturbance
Climate Change and Carbon Sequestration
Summary and Conclusion
Glossary
Aggradation To build up a surface or soil through the
addition of organic and inorganic materials such as flood
sediments, glacial deposits, or mass wasting deposits.
Horizonation The formation of virtually horizontal layers
in a soil profile. These layers develop over time through the
gravity-driven relocation of various soil components such as
salts, clays, and organic matter.
Illuviation The movement and accumulation of soil
components into a soil horizon. Salts and clay minerals are
some of the more common illuviated soil materials.
277
278
278
278
278
279
279
279
279
280
281
281
282
282
Pedogenesis The process of soil development to create a

three-dimensional soil unit. Pedogenesis results from a
combination of environmental factors interacting over time.
These environmental factors include the original parent
material, the climate, and the impact of plants and animals.
Rubification The reddening of a soils color resulting
from the accumulation (illuviation) of clay minerals.
Translocation The movement of various materials
through a soil. The downward displacement is generally
driven by gravity and can include the movement of salts,
clay minerals, and organic matter in solution.
Abstract
A soil chronosequence is a group of soils that have formed with all soil development variables being constant, except time.
Geomorphic settings for the development of soil chronosequences include marine terraces, fluvial terraces, glacial moraines, glacier forelands, and volcanic flows. Anthropogenic disturbances such as fire, mining, and cultivation generate
chronosequences through the interruption of soil development. Natural hazards including floods, landslides, debris flows,
and mudflows also reset soil time. The task of holding the majority of soil variables constant and studying the response of
the remaining variables to ecosystem changes results in an extremely complex, but a powerful research tool.
4.16.1
Introduction
The term chronosequence derives from the Greek chronos

meaning time and Latin sequentia meaning to follow. Soil
chronosequences are, therefore, a series of distinct soils that
develop in succession over a period of time. Looking at the
characteristics of the individual soils in a chronosequence provides insight into how time affects soil development. Soil
Schmid, G.L., 2013. Soil chronosequences. In: Shroder, J. (Editor in Chief),
chronosequences and the development of soils over time is a

concept useful in many scientific disciplines including, but by
no means limited to, soil science, agronomy, biology, ecology,
forestry, climatology, geography, and Earth science.
The process of soil development is called pedogenesis and
involves a number of interrelated variables. Despite the diversity of variables involved in soil development, Jenny (1941)
succinctly summarized the five main factors influencing soil
formation: climate, organisms, relief, parent material, and
time. The concept of a soil chronosequence essentially holds
four of these variables constant, with the only difference between soils being the time of soil formation.
http://dx.doi.org/10.1016/B978-0-12-374739-6.00076-2
277
278
4.16.2.1
Pygmy forest
Hardpan
Sand dunes
Beach deposits
Bedrock
5
4
3
2
1
Age 100 ka 200 ka
50 m
Eleva- 30 m
tion
300 ka
90 m
400 ka
130 m
500 ka
200 m
Figure 1 Idealized emergent marine terrace chronosequence

modeled after Jenny 1980. Courtesy of F.D. Wilkerson.
Examples of landforms developing over time that naturally

support soil chronosequence development include sequential
lava flows, uplifted marine terraces, and recessional glacier
forelands. If all other major soil-forming factors (climate, organisms, relief, and parent material) remain relatively constant, a soil chronosequence becomes a useful tool for
investigating the influence of time on soil development
(Jenny, 1980). In chronosequence settings where there has
been some type of disturbance to the soil (e.g., cultivation,
fire, mining, and climate change), the variation in soil characteristics observed following the time of disturbance allows
an investigation into how that particular disturbance affects
soil development (Figure 1).
4.16.2
Soil Characteristics Supporting

Chronosequence Development
Soils develop over time through a variety of interrelated processes. All soils start with some type of parent material (bedrock, alluvium, glacial till, etc.) that interacts with the
surrounding environment over time. Through time, several
distinct and identifiable physical and chemical soil characteristics develop, with some of the key traits being horizonation,
organic enrichment, movement of salts, clay accumulation,
and changes in pH. The development of discrete layers
(horizons) is accomplished through the accumulation and
translocation of mineral and chemical components. The
movement of salts and clay minerals, and the addition of
organic matter from vegetation and microbes all contribute to
the enhancement of soil horizonation over time. Individual
soil characteristics related to chronosequence development are
numerous, but all have that underlying driving variable of
time, assuming all other soil forming factors are held constant.
Extensive descriptions of soil formation processes can be
found in numerous introductory soil science textbooks, dictionaries, and encyclopedias (e.g., Singer and Munns, 2006;
Brady and Weil, 2008; Soil Science Glossary Terms Committee,
2008; Schmid, 2010).
Organic Enrichment
The establishment of vegetation on the parent material initiates the addition of organic matter to the surface. This organic
material becomes incorporated into the mineral components
of the underlying soil, and an organic enriched surface horizon continues to form over time. There are numerous field
and laboratory ways to characterize the relative amounts of
organic enrichment over time, including consideration of soil
color, depth of enrichment, organic carbon content, organic
nitrogen, phosphorus, and changes in soil structure and texture (e.g., Weaver et al., 1994; Sparks, 1996; Dane and Topp,
2002).
4.16.2.2
Leaching of Soluble Salts
Over time, water moves through the soil profile and redistributes various mineral and chemical components. Given
adequate water and time, some of these soil components will
be completely removed from the profile (e.g., salts), whereas
others move more slowly and tend to concentrate in subsurface horizons (e.g., clay minerals, iron, and aluminum
oxides). Salts are the most readily leached, with sodium being
the most soluble, followed by potassium and calcium compounds. The depth and degree of salt leaching are indicators
of both climate and time, with the accumulation of carbonates
(generally CaCO3) the most common example. In arid to
semiarid environments, carbonates are translocated through
the soil profile to a depth that is indicative of the mean annual
precipitation. Birkeland (1999) included extensive discussions
of soluble salts and carbonates, including field descriptions of
different stages in the development of carbonate morphologies within a soil profile. Laboratory analyses for quantification of soluble salts have been presented in Sparks (1996)
and Ulery and Drees (2008).
4.16.2.3
Translocation of Clay Minerals
After soluble salts have been leached out of the upper soil
horizons, mineral solids can be translocated and illuviated
into subsurface horizons. Clay-sized soil particles are the
smallest mineral solid portion (o0.002 mm) of a soil and are
dominated by clay minerals. Clay minerals can physically
weather from the parent material, or can form as secondary
minerals from chemical weathering within the soil profile.
Regardless of their source, clay-sized particles are susceptible
to translocation through the soil profile. Accumulations of
illuviated clay tend to cause soil color to redden in a process
called rubification, and the illuviated clay commonly results in
a blocky soil structure (Soil Survey Staff, 1993). As with salts
and other translocated particles, the movement of clays proceeds downward with time, so a comparison of clay content
(and mineralogy) between unaltered parent material and clay
content at depth is essential (Birkeland, 1999).
Alterations of the soil parent material including weathering, mineralization, translocation, and subsurface accumulation of organic matter, soluble salts, silicate clays, iron
and aluminum oxides all evolve with time. Many of these soil
variables progress through time in a linear manner, but that
linear development can be disturbed or interrupted. The
279
characteristics described above are the most obvious soil

characteristics supporting the development of a soil chronosequence. More unique and scale-dependent characteristics
will be introduced in the discussion of applied chronosequence research (Section 4.16.3). For more basics on pedogenesis and chronosequence theory, see Bockheim (1980),
Harden (1982), Hugget (1998), Phillips (1999), and Phillips
and Lorz (2008).
A review of the professional literature published during any

time period will reveal numerous research studies utilizing soil
chronosequences. The studies presented here were selected to
reinforce the role of well-established chronosequence variables while illustrating some of the more unique applications.
These applications also serve to demonstrate commonly used
chronosequence methodology along with newer, innovative
methods.
4.16.3
4.16.4.1
Issues Complicating the Development and

Use of Chronosequences
Any type of disturbance that interrupts soil development can

disrupt the linear progression of such characteristics as horizonation, organic enrichment, and clay accumulation. When
these disturbances occur in relation to the development of a
soil chronosequence, the interpretation of that chronosequence becomes more complicated (e.g., Schumm, 1991;
Hugget, 1998; Birkeland, 1999; Phillips, 1999).
A disturbance essentially resets the soil development time
clock, and may also change the interactions between other
soil-forming factors such as climate, organisms, and relief.
Severe disturbances can introduce a new parent material by
burying the original soil profile, which then becomes a buried
paleosol. Natural disruptions to the development of a soil
chronosequence can include climate change, periglacial activity, glacial advances, landslides, debris flows, mudflows,
floods, and fire. Anthropogenic disturbances include fire, cultivation, logging, and mining. Many of these disturbances may
not occur across the entire chronosequence suite of landforms,
but may impact on a microscale, affecting only some of the
soil profiles in the sequence.
Disturbances to soil chronosequence development can,
however, open new windows of investigation. A disturbance
that changes the relationship between any one of the other four
soil-forming factors (climate, organisms, relief, parent material)
and time can make it possible to investigate the role of the
disturbed factors after the point that the original time sequence
was interrupted. General examples include investigating the
role of vegetation change following fire, cultivation, logging, or
mining disturbances. Chronosequences of disturbances can be
used to explore the effects of those specific disturbances on soil
development over time. More specific examples are provided in
the discussion of chronosequence applications.
4.16.4
Chronosequence Applications
The application of soil chronosequences to investigate soil and

landscape relationships takes on a variety of landform and soil
variable approaches. This discussion of chronosequence applications starts with the classic chronosequence landform
applications on marine and fluvial terraces, glacial forelands,
and associated periglacial areas. Vegetation change applications cover a variety of landforms and climates, focusing on
management issues. Fire chronosequences and associated
vegetation changes are included as well. The discussion ends
with applications associated with the contemporary issues of
climate change and carbon sequestration.
Marine and Fluvial Terraces
Rates of tectonic uplift associated with marine terrace chronosequences are expected to be correlated with rates of soil development. The investigation by Huang et al. (2010) on marine terraces in eastern Taiwan used profile development
indices to show a relationship between soil development rates
and tectonic uplift. Their chronosequence results also reinforced the role of climate in controlling weathering and soil
development rates.
Grasslands on a southern California marine terrace chronosequence allowed the investigation of microbial communities
across time and between surface and subsurface relationships
(Moore et al., 2010). With age and soil development, microbial
populations shifted from biogeochemical cycles at depths
o1 m to geochemical cycles in the subsurface depths 41 m in
depth. This shift had previously been documented in more
tropical climates, but not under temperate grasslands.
In alluvial settings, climate plays a major role in soil development rates, but can also act as a disturbance and override
time in controlling soil development. Eppes et al. (2008) used
chronosequence data to establish that the period of soil development on alluvial terraces in the Reno River Valley, Italy,
crossed a climatic threshold between drier periods of CaCO3
accumulation and wetter periods associated with translocation
of clay minerals. From the soil evidence documenting the
change in climate and depositional environments, Eppes et al.
(2008) could establish that the rivers geomorphic responses
to climate change were responsible for the periods of aggradation in the northern Apennines, rather than accelerated
anthropogenic erosion as previously suspected.
As discussed earlier, soil development over time will leach
carbonates and illuviate translocated clay in subsurface horizons. These processes will facilitate a decrease in pH from the
original parent material and rubification in clay-rich horizons.
Through field and standard laboratory analyses, Calero et al.
(2009) found these and other pedogenic processes to hold
true in a fluvial sequence in southern Spain. They reinforced
these findings through the use of scanning electron microscopy to analyze the fabric of the interior of soil peds (soil
structural units) from the soil chronosequence profiles.
4.16.4.2
Glacial and Periglacial Chronosequences
Glacial foreland areas set up natural soil chronosequences as

the glacier retreats, with soil development factors forming a
steep environmental gradient across the landscape. Noll and
Wellinger (2008) focused on bacterial communities across the
chronosequence time gradient. They found typical soil responses across the gradient, with changes in soil organic
280
carbon, pH, soil water content correlating with soil age. Previous research had suggested that soil age and pH were two of
the most important factors in controlling bacterial communities in soil. Noll and Wellinger (2008) separated the influence of soil age and soil water, and found that pH was not a
statistically significant factor to bacterial communities, but soil
age and soil water were the only significant factors, despite the
variety of other soil development variables present. Noll and
Wellinger (2008) felt that their results would be helpful predicting ecosystem responses to climate change.
Periglacial environments are near-glacial environments
dominated by freezethaw processes. Haugland (2004) argued
that periglacial patterned ground can be short-lived landforms
when compared with glacier forelands, and their structure and
sorting processes do affect soil development. He also found
that overall soil development in the foreland periglacial areas
followed expected horizonation over longer-term deglaciation.
However, the interior of fine-scale sorted patterns, where active
cryogenic processes dominate, showed a definite lag in soil
development rates. Haugland (2004) recognized the disturbance role of the periglacial processes and their role in establishing a threshold between pattern genesis and stabilization,
and the role of periglacial processes in creating landscape
heterogeneity.
Haugland continued this periglacial research by investigating whether soil profile development indices could be applied to these very young, disturbed periglacial soils
(Haugland and Owen Haugland, 2008). Although the young
soils exhibited less-developed soils than adjacent areas, they
could still be characterized using a soil profile development
index. Haugland and Owen Haugland (2008) could see the
utility of being able to use soil profile development indices as
gauges of disturbance for other fine-scaled pedogenic processes like bioturbation and tree throw.
4.16.4.3
Vegetation Change and Management Issues
Loss of protective plant cover and soil erosion are both recognized as detrimental to soil quality. What is not well established is how quickly a soil can recover from disturbance,
and how quickly soil can respond to revegetation. An et al.
(2008, 2009) investigated the response of degraded soils on
the Loess Plateau in northwestern China. These studies looked
at soil erosion and the associated decline in soil quality on
land that was deforested and then cultivated (An et al., 2008),
and the recovery of desertified grasslands allowed to revegetate
under natural succession (An et al., 2009).
Erosion on the cultivated land resulted in high loss of soil
aggregates during the first 12 years following deforestation (An
et al., 2008). During the next 12 years, degradation continued,
but rates slowed as the soil reached a quasi-equilibrium with
the cultivated environment. Excessive erosion from the deforestation and cultivation also resulted in loss of soil nutrients, total soil N, and organic C.
The overgrazed, desertified grasslands were fenced to keep
out grazing animals, and were allowed to revegetate by natural
succession (An et al., 2009). Soil microbial populations were
one of the factors measured as an indicator of soil quality. Soil
recovery was exceptionally rapid during the first 2023 years
of grazing exclosure, with organic soil C doubling along with

microbial activity. These plots were allowed to revegetate for
another 2023 years, during which time the recovery rates
slowed down dramatically.
Through the use of soil chronosequences in both these
studies (An et al., 2008, 2009), the authors concluded that soil
reclamation on the Loess Plateau cannot be put in the context
of short-term goals, but rather must be viewed as a long-term
project. The Chinese recognize the importance of soil sustainability and want to develop realistic management goals,
and these chronosequence applications are contributing to the
development of those goals.
A chronosequence study of soil recovery after vegetation
disturbance was conducted over 30 years on a tallgrass prairie
in Kansas (Jangid et al., 2010). Four grassland treatments were
used: conventionally tilled cropland, restored grasslands, annual burns on native tallgrass prairie, and unburned native
prairie. Soil bacterial communities were used to compare the
disturbed prairie treatments to the undisturbed. Jangid et al.
(2010) found transitional bacterial communities present on
the prairie during restoration. These restoration transitional
communities were not found in soils associated with any of
the other three treatments (tilled, burned, and unburned).
Although the transitional communities might appear to be a
sign of adaption to restoration, the disturbed tallgrass plots
did not reached normal levels of the typical tallgrass prairie
bacterial community during the 30-year study period (Jangid
et al., 2010).
A tallgrass prairie study in the US state of Iowa created a
13-year chronosequence of conversion of row-crop cultivation
back to tallgrass prairie (Schilling and Jacobson, 2010).
Groundwater quality was the focus of this soil chronosequence study; and after 13 years there were significant declines in groundwater nitrate and chloride levels under the
older prairie restoration plantings. Schilling and Jacobson
(2010) felt that these relatively short-term water quality recovery times will apply to other tallgrass prairie restorations in
similar geomorphic environments.
Agroforestry plantations need to be managed for sustainable nutrient availability and optimum production. Sharma
et al. (2009) looked at nutrient levels across a 40-year
chronosequence under an aldercardamom plantation in
northeast India. Plots were established at 5-, 10-, 15-, 20-, 30-,
and 40-year intervals, where the N2-fixing alders act as shade
trees for the fruit producing cardamom understory. Nutrient
levels were lowest in the youngest and oldest plots, and
peaked at 1520 years. Sharma et al. (2009) concluded that,
given the nutrient demands of the cardamom, the
aldercardamom plantations should be managed on a 20-year
replantation cycle for the best economic and environmental
sustainability.
The anthropogenic impact of mining acts as a disturbance
to soil development and soil chronosequences can be used to
evaluate soil reclamation responses to revegetation. Banning
et al. (2008) looked at 26-year chronosequence to evaluate the
recovery of a eucalyptus forest from deforestation due to
bauxite mining in western Australia. Soil organic matter levels
and microbial biomass and activity levels returned to nonmined levels in the 26-year period. Levels of nutrient mobility
did not reach previous nonmining levels. The lower levels of
available N are a concern if the lower nutrient mobility levels

affect longer-term forest growth. In addition, Banning et al.
(2008) were not sure about the impact of other management
practices like thinning and fire.
4.16.4.4
Fire Disturbance
Forest fire ash contributes nutrients to the soil immediately

following the fire, but these nutrient effects are thought to be
short lived. Duran et al. (2008) looked at an 18-year fire
chronosequence in pine forests on the Canary Islands in order
to investigate the long-term effects of fire on N and P availability. Fire increased the available N, but only for 1 year; and
by 5 years postburn, N levels were similar to those on unburned plots, with no long-term postfire reduction in available N (Duran et al., 2008). Available P did not increase after
fire disturbance, instead it actually decreased significantly for
the first year postfire. Available P levels did gradually increase
on the burned plots over time, and lowest P levels were found
on the unburned plots (Duran et al., 2008).
Wildfires in natural areas can affect both animal and plant
populations. Vamstad and Rotenberry (2010) looked at the
impact of increased wildfire size and frequency in the fire
tolerant Joshua tree woodland ecosystem of the Mojave Desert. A 65-year fire sequence was used to evaluate post-fire rodent diversity and plant species composition. There were no
significant differences in rodent populations over time between the burned and unburned areas. A directional change in
species composition did occur on the burned sites, with no
apparent convergence back to preburn conditions. In simpler
terms, species that were present before the burn and were removed by the fire did not return in the years following the fire.
Vamstad and Rotenberry (2010) attributed this to the slow reestablishment rates of some plant species. Soil chronosequences will come more directly into play with future
management considerations. Questions of the feasibility and
sustainability of fire suppression, along with the presence of
exotic grass species are some of the more immediate management concerns. The authors stated their concern over
continued environmental pressures with potential climate
change scenarios and the threat of more invasive species, and
that despite the remoteness of this location, that it is impacted
by anthropogenic disturbance (Vamstad and Rotenberry,
2010).
4.16.4.5
Climate Change and Carbon Sequestration
Growing concerns over global climate change and the role of

atmospheric CO2 lead to discussions of the sequestration of C
in various ecosystem components. Many of the previously
discussed studies have climate-change components, and the
few that follow provide just a small additional sampling.
The stability of soil organic matter is important to retain C
in the soil. Egli et al. (2010) used a 150-year soil chronosequence of proglacial soil development in the Swiss Alps to
investigate the time frame for formation and stability of soil
organic matter. They were able to determine that even at the
outset of soil development, a large portion of the soil organic
matter was in a stable form (6% of total soil organic carbon
281
and 10% of total soil N were in stable organic matter). The

authors recognized this as a strong organic-mineral association at the onset of proglacial soil formation, and realized the
significance of this finding relative to potential new soil areas
exposed with glacial retreat and climate change (Egli et al.,
2010).
Volcanic soils have a high C accumulation potential, especially when forested. Volcanic chronosequences also
provide a very long-term soil development platform.
Pena-Ramrez et al. (2009) used an 183530 500-year BP
chronosequence under pine and fir forests to evaluate soil C
storage in the Chichinautzin Volcanic Field in Mexico. Their
results showed a linear increase in the accumulation of C in
the soils from 1835 to 10 000 years BP, but after 10 000 years
the C accumulation decreased by 30%. Tree sampling showed
that in the first 8000 years site quality improves with soil
development and tree establishment. As with Egli et al. (2010)
in the Swiss Alps, C accumulation in the early years (first
10 000 years in Chichinautzin) is accomplished in organicmineral complexes. After 10 000 years, Pena-Ramrez et al.
(2009) attribute C stabilization to less efficient rates associated
with clay minerals and Al oxides.
The question of how much C a degraded soil can recover,
and how quickly it can recover is a common research
question. Preger et al. (2010) addressed this question using a
31-year soil sequence in South Africa. The soil sequence contained primary grassland, degraded cropland, and secondary
pasture reclaimed from cultivated land. Over the chronosequence time span, the secondary pastures had 3094% more
soil organic C than the cultivated cropland, but the pastures
never regained the same C levels as the original grasslands. In
addition, all the soil organic recovery occurred in the surface
horizon, so it is still very susceptible to soil erosion and degradation (Preger et al., 2010).
In China, grazing exclusion on semiarid grasslands is a
management option. Grazing exclusion is known to increase
soil C storage, but it is unclear what happens with long-term
grazing exclusion. He et al. (2009) used a 28-year soil
chronosequence ranging from free grazing to grazing exclusion to investigate the long-term soil impacts. Their results
show significant increases in C storage with grazing exclusion,
with a logarithmic increase for at least two decades. However,
after two decades, the potential for further C sequestration
declines, and is strongly related to N inputs.
Cropping practices in the Brazilian savanna have shifted
from traditional tillage and rotations to a 2-year no-tillage
mulch-based system that alternates soybeans or maize with
cereal crops. Neto et al. (2010) used a 23-year soil chronosequence to evaluate soil C levels. After 23 years of cultivation,
the traditionally tilled soils had 30% less soil organic C than
the natural savanna. The no-tillage mulch soils had soil organic carbon levels equal to the natural savanna after only
12 years of management. Neto et al. (2010) have clearly
documented an important role of tropical soils in storing
organic C.
Carbon sequestration in some reclaimed mine soils is
difficult to measure due to the presence of large amounts of
coal and inorganic C. Shrestha and Lal (2010) used a 25-year
reclaimed mine soil and biomass chronosequence in Ohio to
investigate soil C and N pools under forest and pasture
282
revegetation. Sequestration of ecosystem organic C peaked in

6 years for the pasture and 14 years for the forest. Pasture
ecosystems always sequestered less overall organic C, but had a
higher percentage of soil C. The forest ecosystems have a larger
amount of organic C, and their highest sequestration percentages are in biomass storage. Despite these differences,
Shrestha and Lal (2010) noted that both N and C sequestration rates peaked in the forest and pasture ecosystems after
1015 years of reclamation. This short-term peak shows the
relative efficiency of these ecosystems for sequestration of C,
and they should be included in management plans for offsetting carbon emissions (Shrestha and Lal, 2010).
4.16.5
Summary and Conclusion
Soil chronosequences can be established on a variety of

landforms over a wide range of time spans. Specific landforms
such as marine and fluvial terraces, glacial moraines and
forelands, and progressive volcanic deposits can support
classic chronosequences that range from decades to hundreds
of thousands of years. Revegetation chronosequences usually
cover shorter decadal time sequences, but in some regions can
be established over hundreds of years. Regardless of the time
period covered by a sequence, soil chronosequences allow us
to investigate the role of time as dominant soil forming factor.
But perhaps the most powerful contribution of soil chronosequences is facilitating the connections between time and
other soil development factors such as climate, vegetation, and
microorganisms.
References
An, S., Huang, Y., Zheng, F., 2009. Evaluation of soil microbial indices along a
revegetation chronosequence in grassland soils on the Loess Plateau, Northwest
China. Applied Soil Ecology 41(3), 286292.
An, S., Zheng, F., Zhang, F., Van Pelt, S., Hamer, U., Makeschin, F., 2008. Soil
quality degradation processes along a deforestation chronosequence in the
Ziwuling area, China. Catena 75(3), 248256.
Banning, N.C., Grant, C.D., Jones, D.L., Murphy, D.V., 2008. Recovery of soil
organic matter, organic matter turnover and nitrogen cycling in a post-mining
forest rehabilitation chronosequence. Soil Biology and Biochemistry 40(8),
20212031.
New York, 430 pp.
Bockheim, J.G., 1980. Solution and use of chronofunctions in studying soil
development. Geoderma 24(1), 7185.
Brady, N.C., Weil, R., 2008. The Nature and Property of Soils, 14th ed. Pearson
Higher Education, Saddle River, NJ, 992 pp.
Calero, J., Delgado, R., Delgado, G., Martn-Garca, J.M., 2009. SEM image
analysis in the study of a soil chronosequence on fluvial terraces of the middle
Guadalquivir (southern Spain). European Journal of Soil Science 60(3),
465480.
Dane, J.H., Topp, G.C. (Eds.), 2002. Methods of Soil Analysis. Part 4. Physical
Methods. SSSA Book Series No. 5. Soil Science Society of America, Madison,
WI, 1692 pp.
Duran, J., Rodrguez, A., Fernandez-Palacios, J.M., Gallardo, A., 2008. Changes in
soil N and P availability in a Pinus canariensis fire chronosequence. Forest
Ecology and Management 256(3), 384387.
Egli, M., Mavris, C., Mirabella, A., Giaccai, D., 2010. Soil organic matter formation
along a chronosequence in the Morteratsch proglacial area (Upper Engadine,
Switzerland). Catena 82(2), 6169.
Eppes, M.C., Bierma, R., Vinson, D., Pazzaglia, F., 2008. A soil chronosequence
study of the Reno valley, Italy: insights into the relative role of climate versus
anthropogenic forcing on hillslope processes during the mid-Holocene.

Geoderma 147(34), 97107.
Harden, J.W., 1982. A quantitative index of soil development from field
descriptions: examples from a chronosequence in central California. Geoderma
28(1), 128.
Haugland, J.E., 2004. Formation of patterned ground and fine-scale soil
development within two late Holocene glacial chronosequences: Jotunheimen,
Norway. Geomorphology 61(34), 287301.
Haugland, J.E., Owen Haugland, B.S., 2008. Cryogenic disturbance and pedogenic
lag effects as determined by the profile developmental index: the
Styggedalsbreen glacier chronosequence, Norway. Geomorphology 96(12),
212230.
He, N., Wu, L., Wang, Y., Han, X., 2009. Changes in carbon and nitrogen in soil
particle-size fractions along a grassland restoration chronosequence in northern
China. Geoderma 150(34), 302308.
Huang, W.-S., Tsai, H., Tsai, C.-C., Hseu, C.-C., Hseu, Z.-Y., Chen, Z.-S., 2010.
Subtropical soil chronosequence on Holocene marine terraces in eastern Taiwan.
Soil Science Society of America Journal 74(4), 12711283.
Huggett, R.J., 1998. Soil chronosequences, soil development, and soil evolution: a
critical review. Catena 32(34), 155172.
Jangid, K., Williams, M.A., Franzluebbers, A.J., Blair, J.M., Coleman, D.C.,
Whitman, W.B., 2010. Development of soil microbial communities during
tallgrass prairie restoration. Soil Biology and Biochemistry 42(2), 302312.
Jenny, H., 1941. Factors of Soil Formation: a System of Quantitative Pedology.
McGraw Hill Book Company, Inc., New York, NY, (Reprinted 1994 by
Ronald Amundson, Dover Publications Inc., 281 pp).
Jenny, H., 1980. The soil resource. Origin and Behavior. Ecological Studies 37.
Springer-Verlag, New York, 377 pp.
Moore, J., Macalady, J.L., Schulz, M.S., White, A.F., Brantley, S.L., 2010. Shifting
microbial community structure across a marine terrace grassland
chronosequence, Santa Cruz, California. Soil Biology and Biochemistry 42(1),
2131.
Neto, M.S., Scopel, E., Corbeels, M., Cardoso, A.N., Douzet, J.M., Feller, C.,
Piccolo, M.C., Cerri, C.C., Bernoux, M., 2010. Soil carbon stocks
under no-tillage mulch-based cropping systems in the Brazilian Cerrado:
an on-farm synchronic assessment. Soil and Tillage Research 110(1),
187195.
Noll, M., Wellinger, M., 2008. Changes of the soil ecosystem along a receding
glacier: testing the correlation between environmental factors and bacterial
community structure. Soil Biology and Biochemistry 40(10), 26112619.
Pena-Ramrez, V.M., Vazquez-Selem, L., Siebe, C., 2009. Soil organic carbon
stocks and forest productivity in volcanic ash soils of different age
(183530,500 years B.P.) in Mexico. Geoderma 149(34), 224234.
Phillips, J.D., 1999. Earth surface systems. Complexity, Order, and Scale. Blackwell
Publishers, Malden, MA, 180 pp.
Phillips, J.D., Lorz, C., 2008. Origins and implications of soil layering. EarthScience Reviews 89(34), 144155.
Preger, A.C., Kosters, R., Brodowski, S., Amelung, W., DuPreez, C.C., 2010. Carbon
sequestration in secondary pasture soils: a chronosequence study in the South
African Highveld. European Journal of Soil Science 61(4), 551562.
Schilling, K.E., Jacobson, P., 2010. Groundwater conditions under a reconstructed
prairie chronosequence. Agriculture, Ecosystems and Environment 135(12),
8189.
Schmid, G.L., 2010. Soils. Encyclopedia of Geography. Sage Publications,
Thousand Oaks, CA, 8 pp.
Schumm, S.A., 1991. To interpret the Earth: ten ways to be wrong. Cambridge
University Press, Cambridge, UK, 133 pp.
Sharma, G., Sharma, R., Sharma, E., 2009. Impact of stand age on soil C, N and P
dynamics in a 40-year chronosequence of aldercardamom agroforestry stands
of the Sikkim Himalaya. Pedobiologia 52(6), 401414.
Shrestha, R.K., Lal, R., 2010. Carbon and nitrogen pools in reclaimed land
under forest and pasture ecosystems in Ohio, USA. Geoderma 157(34),
196205.
Singer, M.J., Munns, D.N., 2006. Soils: an Introduction, Sixth ed. Prentice Hall,
Saddle River, NJ, 464 pp.
Soil Science Glossary Terms Committee, 2008. Glossary of Soil Science Terms
2008. Soil Science Society of America, Madison, WI, 92 pp.
Soil Survey Staff, 1993. Soil Survey Manual, USDA Handbook vol. 18. United
States Department of Agriculture. Natural Resources Conservation Service. U.S.
Government Printing Office, Washington, DC.
Sparks, D.L. (Ed.), 1996. Methods of Soil Analysis. Part 3: Chemical Methods.
SSSA Book Series No. 5. Soil Science Society of America, Inc., American
Society of Agronomy, Inc., Madison, WI, 1390 pp.
Ulery, A.L., Drees, R. (Eds.), 2008. Methods of Soil Analysis. Part 5. Mineralogical
Methods. SSSA Book Series No. 5. Soil Science Society of America, Madison,
WI, 521 pp.
Vamstad, M.S., Rotenberry, J.T., 2010. Effects of fire on vegetation and small
mammal communities in a Mojave Desert Joshua tree woodland. Journal of Arid
Environments 74(10), 13091318.
283
Weaver, R.W., Angle, S., Bottomley, P., Bezdiecek, D. (Eds.), 1994. Methods of Soil
Analysis. Part 2. Microbiological and Biochemical Properties. SSSA Book Series
No. 5. Soil Science Society of America, Madison, WI, 1121 pp.
Biographical Sketch
Ginger L. Schmid holds degrees in earth science, soils, and environmental geography. The two most important
components of Gingers professional career are her students and her field research. She teaches numerous
undergraduate and graduate level courses including geomorphology and seminars on soils. Her field research is
centered on soils geomorphology in wild land areas including the role of soils in mass wasting, periglacial
environments, and climate change. Her main research areas include Glacier National Park and the White
Mountains of California.
4.17 Weathering and Sediment Genesis

4.17.1
4.17.2
4.17.3
4.17.3.1
4.17.3.2
4.17.3.3
4.17.4
4.17.5
4.17.5.1
4.17.5.2
4.17.5.3
4.17.5.4
4.17.6
4.17.7
References
Weathering, Sediments, and the Rock Cycle

Processes: Disintegration and Chemical Alteration
Factors of Weathering Relevant to Sediment Production
Parent Material
Climate
Drainage and Topographic Relief
Sediment Maturity and Weathering in Transport
Types of Sediment
Scree (or talus) and Other Rock Fragments
Sand
Silt
Clay
The Role of Weathering in Cementing Sediment
Summary
Glossary
Arenization A process in granitic rock (but possibly also
relevant to any coarse crystalline rock) by which chemical
decomposition results in saprolite and the generation of
coarse sand or gravel-sized fragments.
Bioturbation A process in soil and sediment by which
plants or animals turn and displace material, thereby
mixing the soil or sediment in a vertical sense.
Bog iron Iron deposit in wetlands formed of precipitated
iron ions derived from weathering of other material.
Calcrete or caliche A pedogenic (soil formed) and
indurated deposit of calcium carbonate, derived from the
precipitation of calcium ions, usually found in arid and
semiarid environments.
Case hardening Induration of the outer shell of rock by
means of cementation by secondary minerals derived from
weathering.
Chelation Chemical weathering process by which
complex organic molecules (some proteins and amino
acids, for instance) selectively extract metal ions from
minerals, weakening the mineral.
Chemical weathering (chemical alteration) Breakup of
rocks and minerals by way of chemical agents, removing
material by molecules into solution; caused by any of
several chemically reacting agents.
Clastic Term referring to particles of rock, of any size.
Clay 1) One of the phyllosilicate (layered) minerals
usually formed from weathering of primary minerals or
precipitated from solutes after weathering; 2) A class of
285
285
287
287
287
287
288
288
288
288
288
289
291
291
291
particle, in the grain size scale, of any mineralogy, defined as

being less than 4 mm in diameter.
Comminution Mechanical crushing and abrasion of
sediments in transport.
Cosmogenic isotope dating Surface exposure dating
technique that uses accumulation of unique and rare
isotopes (such as Be10 or He3), generated through
interaction with high-energy cosmic rays.
Diagenesis Chemical or physical alteration to a sediment
following deposition, leading to compaction and
lithification. Usually excludes surface weathering, though
transitional gradation to groundwater diagenesis is ill
defined.
Disaggregation In situ breakup of rock particles to form
smaller particles, distinct from chemical alteration but may
be aided by chemical processes.
Fissuresols Fine grained sediments, weathering or aeolian
in origin, accumulating in rock fractures.
Fragmentation Creation of rock pieces, influenced
existing fractures or rock weaknesses, largely mechanical
processes though aided by chemical processes. Predecessor
to disaggregation.
Hydration Chemical weathering process by which a
hydroxide ion is incorporated into the crystal structure,
weakening the lattice. Noted in mica and clay minerals.
Laterite A residual weathering product concentrated in
aluminum and iron oxides and depleted of silica and most
other elements, often indurated.
Loess Windblown sediment of mostly silt-sized particles.
Pope, G.A., 2013. Weathering and sediment genesis. In: Shroder, J. (Editor
in Chief), Pope, G.A. (Ed.), Treatise on Geomorphology. Academic Press,
San Diego, CA, vol. 4, Weathering and Soils Geomorphology, pp. 284293.
284
http://dx.doi.org/10.1016/B978-0-12-374739-6.00402-4
Weathering and Sediment Genesis
Luminescence dating Surface exposure dating technique

that measures accumulated radiation in minerals following
most recent exposure to sunlight or intense heat.
Maturity In sediments, the composition and grain
morphology relevant to the age and duration of transport;
more mature sediments are of one mineralogy (usually
quartz), and with well-rounded and well-sorted grains.
Mechanical weathering Breakup of rocks and minerals
without chemical alteration, created by physical forces of
pressure, strain, or shear, caused any of several agents:
crystal growth (including ice), expansion or contraction due
to changes in temperature, relaxation due to removal of
pressure (unloading), and pressure from plant roots.
Neoformed minerals see secondary minerals.
Oxidation Chemical weathering process by which oxygen
molecule is incorporated into the crystal structure,
weakening the lattice. Oxidation of iron imparts reddish to
yellowish colors in rocks, minerals, and soils.
Periglacial Describes processes related to nonglacial ice,
and regions dominated by these processes (some of which
may have been previously glaciated, others never glaciated
but still cold).
Proglacial Area immediately down-gradient from or in
front of a glacier.
285
Regolith Encompassing term for unconsolidated material

at the Earths surface, including soil, sediments, and
weathered rock.
Sand In the grain size scale, a class of particles ranging
from 62.5 mm to 2 mm in diameter.
Saprolite Any weathered bedrock that remains in situ.
Scree Rock fragments, usually angular, caused by
weathering, that accumulate on slopes. Sometimes
erroneously referred to as talus, which is more properly the
landform of a scree deposit.
Secondary minerals Minerals created from the solute
products of weathering or altered in situ by weathering (as
opposed to primary minerals which derive from magma).
Sometimes referred to as neoformed minerals.
Sediment Particles or dissolved material, derived from
rock and mineral weathering, that is carried by surface
processes of water, ice, wind, or gravity.
Silt In the grain size scale, a class of particles ranging from
4 mm to 62.5 mm in diameter.
Structure In rock, the inherent physical properties of the
rock, including jointing, faulting, folding, bedding, or
foliation.
Abstract
The genesis of detrital sediments could not take place without weathering. Both chemical and mechanical weathering
processes work toward the fragmentation, disaggregation, and chemical alteration of rocks to form clastic particles (clay to
boulder size) and solutes. Clastic particles are entrained into the sediment system by various geomorphic processes. Solutes
indurate soils and near surface sediments as well as cementing buried sediments. The weathering genesis of particles and
weathering of sediments becomes relevant to concerns within geomorphology and sedimentology, including the recognition of fine-grained particles derived from weathering, and the implications of sediment weathering on environmental
factors and dating methods.
4.17.1
Weathering, Sediments, and the Rock Cycle
Sediments are a significant concern in the science of geomorphology. The overwhelming majority of these sediments
are generated by weathering processes: no weathering, no
sediment, nothing to transport or deposit. Weathering is a
critical node in the rock cycle published in many forms in
innumerable introductory textbooks (Figure 1), most of which
quickly mention the process that can attack any rock to start the
sedimentary rock segment. Not usually included in this diagram is the feedback of direct weathering products (included in
Figure 1), the near-surface, neoformed, in situ minerals that in
many cases cement or otherwise solidify soil and regolith, prior
to entering the sediment system. These byproducts, too, can
weather, an additional step on the way to the sediment system.
Weathering is equal in importance to magma production in the
rock cycle process; both are responsible for recycling of earth
materials and genesis of new materials.
White et al. (1992) also in an introductory textbook, provided a more directly focused diagram of the weathering system (Figure 2), the specific segment within the rock cycle. In
this cascading system, rock breakdown produces rock and

mineral fragments and mobile and residual chemical products, all of which may be temporarily stored in the regolith,
recycled, or leave the local system to enter the detrital sediment and dissolved solute budget. A feedback loop is suggested in the weathering of end product and new material
created by the weathering process. The diagram was intended
as a simplified representation, in reality probably more
interaction and parallel development of chemical and
mechanical weathering agents occurs at all stages. Chemical
weathering helps the mechanical fragmentation and disaggregation process, whereas mechanical forces continue to
work during the chemical alteration process.
4.17.2
Processes: Disintegration and Chemical

Alteration
The processes by which rock is broken down in the sedimentary

system involve both disaggregation and chemical alteration.
Details of these processes are covered by other chapters in this
286
Igneous
rocks
Solidifying
Weathering
Magma
Weathering
products
Melting
Sediments
Metamorphic
rocks
Erosion, transport
and deposition
Sedimentary
rocks
Metamorphism
Figure 1 The rock cycle.
Unaltered
rock
Agencies activating
fragmentation
Agencies activating
disaggregation
Clastic rock
fragments
Agencies activating
chemical alteration
Disaggregated
mineral grains
Mobile
products
Agencies activating
resynthesis
Residual
materials
Mobile
products
New
material
Residual
material
Regolith
Figure 2 Weathering system, generating particles and chemical solutes. Reproduced from White, I.D., Mottershead, D.N., Harris, S.J., 1992.
Environmental Systems: An Introductory Text, Second edn. Chapman and Hall.
book, respective to different geomorphologies, but can be

briefly outlined here. Disaggregation is the in situ breakup
of rock particles to form smaller particles. Fragmentation begins
this process by creating rock pieces, largely controlled by
preexisting fractures or other rock weaknesses (such as bedding
or foliation). Such fragments range in size from millimeters
to many meters in diameter. Disaggregation continues with
the release of individual crystalline grains, crystal fragments, clusters of crystals, or, in the case of existing sedimentary rocks, individual detrital particles (e.g., sand, silt, or
clay).
Mechanical weathering processes are important in the
fragmentation and disaggregation. Ice, salt, and neoformed
minerals (such as iron and quartz) are similar in their ability
to grow crystals within confined pore spaces or fractures,
exerting enough pressure to cause brittle fracture. Plant roots
likewise are able to expand within fractures and microfractures. Anisotropic response to heating and cooling (in
rocks or individual crystals) causes thermal shock over extreme events or rock fabric fatigue over long periods (see
Chapters 4.15 and 4.12). Drying and wetting cycles can be
particularly relevant to shrinkswell clay minerals such as
vermiculite and smectite, a mechanism for the mechanical
release of clay-size particles. Finally, dilation or pressure unloading follows the release of overburden, after rapid erosion
or melting of large glacial mass. These mechanical processes
do not operate devoid of chemical weathering; chemical
weathering helps to open existing weaknesses to further attack,
including mechanical weathering. The White et al. (1992)
weathering system (Figure 2) would be most accurate to
consider chemical weathering happening in parallel to
mechanical weathering, even at the onset of exposure of
unaltered rock.
Chemical alteration includes any of the chemical weathering
processes. Dissolution (hydrolysis) would be most important
for the silicate minerals, whereas solution would be most
relevant specifically to rocks containing quartz and calcite.
Other chemical weathering processes include oxidation, hydration (addition of oxygen or hydroxide molecules, respectively, to the mineral matrix), and chelation (selective removal of
metallic ions by organic agents). Some of these processes target
specific minerals. Biotite, for instance, is vulnerable to oxidation
and chelation. Loss of the specific minerals would in turn
weaken the rock matrix, a factor in the arenization of some
granitic rocks. Not only the loss of entire minerals, but also the
ionic loss from chemical weathering cause intramineral voids,
ultimately decreasing the unit weight (bulk density) of the rock
(McNally, 1992) and leading to more rapid disintegration as
well as collapse of weathering profiles.
4.17.3.1
Factors of Weathering Relevant to Sediment

Production
Prothero and Schwab (1996) outlined several factors

controlling the processes of sediment generation by weathering. These factors are similar to those promoted by
Johnsson (1993), Hill and Rosenbaum (1998), and Pope et al.
(1995).
Parent Material
The parent material rock has a composition that encompasses

the mineralogy, texture, and rock structure of the host rock.
Mineralogy renders an inherent weathering susceptibility,
owing to the elemental content and crystal structure. This familiar relationship introduced by Goldich (1938) is sometimes overwhelmed by environmental factors (Wasklewicz,
1994). Rocks composed of different mineralogies (for instance, granite) will undergo differential weathering, the more
weatherable minerals are the first to fail, but this also compromises the integrity of the entire rock, in that more resistant
mineral grains (such as quartz) can more easily enter the
sediment train.
Structure and texture are related in their control of access to
weathering agents and importance to the strength of the rock
fabric. In general, fine-grained rocks (such as volcanic, mudstone, and slate) are more susceptible to chemical weathering
due to their increased surface area per unit volume (sometimes referred to as the wetted surface area). Rock structure
includes fissuring and jointing in the rock, foliation (in
metamorphic rocks), and bedding (in sedimentary rocks). In
general, increased joint density and finer foliation or bedding
promote increased weathering. Bedding, foliation, and jointing interrupt the rock fabric and provide conduits for moisture
(aiding chemical weathering and providing access for mechanical weathering such as salt crystallization or ice). At large
scales, the spatial distribution of resistant rock landforms is
determined by joint density (Young et al., 2009; Twidale and
Romani, 2005; Campbell, 1997).
4.17.3.2
Climate
Climate influences the amount and seasonality of precipitation, the mean and variation in temperature, and in turn the
type of ecosystem with associated biota. All of these elements
are important to weathering, though the relationships are not
always as simple as presumed (see Chapters 4.3, 4.14, 13.11,
4.15, 4.12, 4.1, and 4.13). In general, abundant precipitation
and organic acids promote chemical weathering; higher temperatures increase chemical weathering rates; temperature extremes promote physical stress on the rock; and cold
temperatures are necessary for the formation of ice and,
therefore, physical stress within pores and fractures. Regional
climatic regimes as well as variations from these norms by way
of microclimates would influence conditions at the mineralscale boundary layer. Weathering environments are susceptible
to climate change, as weathering is generally a slow and longterm process.
4.17.3.3
4.17.3
287
Drainage and Topographic Relief
Both drainage and topographic relief relate to the influx of

weathering agents and the removal of solute and solid weathering products. Within the rock or soil, the resident time of
reactive chemical agents influences the rate of chemical weathering. Well-drained weathering profiles carry solutes away
quickly, but the weathering agents have short contact time. In
the same sense, slopes promote the downward movement of
weathering products. Steep slopes allow rapid removal of
288
material; gentle relief would allow accumulation of weathered

material and retention of weathering agents within the colluviums, promoting additional chemical weathering.
4.17.4
Sediment Maturity and Weathering in

Transport
Maturity in sediments refers to characteristics of composition,

grain size and sorting, and grain morphology, which evolve in
a down-path direction from source to end (most often pertaining to fluvial sediments, but equally relevant to aeolian,
marine, and glacial sediments). A mature sediment is more
exclusively quartz, well sorted, and with greater particle
rounding. Weathering, both mechanical and chemical, is a
part of this maturation. The works of Krinsley and Mahaney
provided extensive discussion of in-transport alterations of
sediment grains (Krinsley and Doornkamp, 1973; Krinsley,
1998; Mahaney, 1995, 2002; Mahaney et al., 1996). Although
not specifically mentioning weathering or sediment maturity,
the comprehensive grain morphology rubric presented by
Blott and Pye (2008) permits an organized comparison of
factors (such as sphericity, irregularity, and rounding).
It is sometimes assumed for methodological reasons (especially dating methods) that weathering does not take place
during sediment transport (cf., Cockburn and Summefield,
2004; Bierman and Steig, 1996). Weathering may be negligible
for the purposes of cosmogenic nuclide dating methods,
though the impacts are not well researched (cf., McFarlane
et al., 2005; Zreda and Phillips, 2000); weathering during
temporary storage has been shown to be a significant complicating factor in luminescence dating (Jeong et al., 2007).
Weathering can and does occur with sediments in transport,
including during temporary storage.
Comminution, a term used to describe mechanical
crushing of sediments in transport, is arguably a weathering
process (the break-down of rock or mineraly), but is usually not considered as such, senso stricto, in that it takes place
during transport and not in situ. Comminution does produce
morphological changes (scars and rounding) and smaller
particles (Peterknecht and Tietz, 2011; Wright and Smith,
1993). For sediment (in active transport or in temporary
storage), researchers report various means of weathering
(Johnsson et al., 1991; Johnsson, 1993; Pye and Mazzullo,
1994; Nesbitt et al., 1996).
4.17.5
Types of Sediment
Huggett (2011) gave a broad classification of weathering

debris, the products of weathering: solids (detrital particles,
including clay, silt, sand, and boulders), solutes (ions in solution), and colloids (clumps of ions combine with organic
matter B1100 mm). Solutes reappear in precipitated secondary minerals within the weathering system, in cements in
sedimentary rocks, or as precipitated sediments (for instance,
evaporites or calcium carbonate sediments in oceans and
lakes). Solids and colloids become the clastic detrital sediment
system, discussed in detail below.
4.17.5.1
Scree (or talus) and Other Rock Fragments
Scree properly refers to rock fragments that accumulate at the

bottom of a slope (Stratham, 1973), the accumulating depositional slope is known as a talus (Bloom, 1998). (Some
authors also refer to talus and scree interchangeably as the
particles, and further differentiate scree as the smaller of the
particle range.) A wide range of particle sizes occurs, from gravel
to large boulders. Ritter et al. (1995: 129) described scree as
controlled by weathering along mechanical discontinuities in
the cliff rocky. Therefore, the size of the blocks is largely inherited from the structural characteristics of the parent bedrock.
Even so, chemical weathering probably plays a role in rock
fabric weakening. Thornburry (1966) recognized the possible
importance of mineral hydration in the process of spalling.
Granular or gravel-sized scree is probably reliant more on
chemical weathering than mechanical, where deep weathering
loosens the rock fabric along crystal boundaries (Figure 3).
Smaller scree particles are more readily entrained into the
sedimentary system, whereas larger blocks would remain
stable at footslopes, subject to further weathering (or larger
forces of transport, such as glaciers). Boulder- and gravel-bedded
streams in steep terrain, as well as the much of the coarse debris
entrained in lateral moraines of glaciers are derived from weathering-supplied rock fragments from the surrounding slopes.
Large rock fragments also develop in situ in the soil and
weathering profile during chemical weathering. Philips et al.
(2005) described the prevalence of rock fragments in soil and
the importance of lithology, sandstones surviving as fragments
preferentially to more weatherable shale. Both down-slope
colluvial processes and tree-throw bioturbation are important
in the lateral and vertical distribution of rock fragments in the
soil. Phillips (2004) also explained particle-size differentiation
in soil and weathering profiles. These vertical textural contrasts, zoned segregation in particle size (coarse vs. fine) that
vary by depth, are caused by erosional winnowing or downward translocation of fines from the surface, surface and
subsurface inputs of coarse or fine material (by weathering, for
instance), and bioturbation.
4.17.5.2
Sand
The dominant mineralogy of most sand is quartz, which is

relatively resistant in the weathering environment (though not
completely). Sand-sized particles of other mineralogy are also
possible, but tend to diminish with sediment maturity (such
as feldspars, see above). Sand is derived from crystalline source
rocks, or from previously deposited sandstones. In the former,
grains are typically coarse, commonly beyond sand-sized,
and further weathering or comminution diminishes the size to
sand-sized fraction or smaller (Pope 1995a; see also Figure 4).
Saprolites are an important source of detrital quartz
(Figure 3). Arenization refers to this process of sand genesis
in saprolite, the term literally derived from the Latin (and
French, Portuguese, and Spanish) word for sand: arena
(Sequeira Braga et al., 2002).
4.17.5.3
Silt
A large volume of research literature on production of weathering sediment concerns silt-sized particles. Silt is common
(a)
(b)
Figure 3 (a) Gravel scree. Gravel scree, Gold Camp Road, near
Pikes Peak, Colorado. Deeply weathered granite freely disintegrates
on an eroded cliff face, accumulating as gravel-sized debris cones at
the bottom of the cliff. Photograph by author. (b) Larger scree slope.
Larger scree slope, with angular clasts dominated by preexisting joint
and bedding. Glacier National Park, Montana, Firebrand Pass area.
Photo source, P. Carrara, USGS Photographic library. Reproduced
from US Geological Survey photographic library, photograph by P.
Carrara, ID. Carrara, P. 317 ct. Image file: http://
libraryphoto.cr.usgs.gov/htmllib/batch41/batch41j/batch41z/batch41/
car00317.jpg
289
in river sediments, and because it is easily windborne, silt is

predominant in loess. The long-standing assumption was that
silt occurring in loess was derived from proglacial and periglacial areas during glacial episodes, supported by the occurrence of loess downwind of the glacial limit in Eurasia and
North America. Comminution, abrasion (Wright and Smith,
1993), and crushing (Smalley, 1966, 1995; Riezebos and Van
der Waals, 1974) do produce silt-sized fragments in quartz,
the former evident in the glacial flour that imparts a milky
color to proglacial rivers and lakes. Tectonic crushing (Smalley,
1995) was also suggested as a source for silt-sized quartz
outside of glacial regions. However, weathering is now recognized as an important source for silt-sized sediments. Wright
et al. (1998) compared the various mechanisms possible with
laboratory simulations, whereas Wright (2007) provided a
comprehensive review of the evidence for weathering-generated silt. In the latter study, Wright pointed out that weatherings role in silt production has been considerably
underestimated, and illustrated a number of examples of
weathering-produced silts in glaciated and proglacial environments that predate glaciation. Further evidence of weathering sources include Mazzullo et al. (1988), who identified
on the North American continental shelf, preglacial Cenozoic
and Precambrian/Paleozoic/Mesozoic silt sediments derived
from weathering, in addition to angular silt-sized quartz derived from Quaternary glaciations; and Pope (1995a), who
illustrated with electron microscopy a variety of silt-sized
particles generated within quartz sand grains.
Dry playas and alluvial plains were assumed to provide the
bulk of silt-sized dust in arid and semiarid regions. Again, although silt is plentiful and easily entrained in the wind in
drylands, weathering sources are also relevant. Smith et al.
(1987) demonstrated the production of silt from quartz sandstone in laboratory simulations of desert conditions. Salt and
temperature stress were tested as mechanisms, thus mechanical
weathering processes were shown to be important. In the natural environment, Goudie et al. (1979) described the production of silt-sized quartz from salt weathering of dune sands.
Dune sands also produce quartz silt by chemical weathering
during periods of humid climate (Pye, 1983). In another arid
example, Villa et al. (1995) suggested both aeolian and in situ
weathering origins for fine-grained sediments or fissuresols,
occurring in rock fractures, depending on the climatic conditions. Chemical weathering would be possible in such fissures
during wetter climatic periods. Other examples of chemical
weathering production of silt-sized quartz in tropical regions
were described by Eswaran and Stoops (1979), Johnsson et al.
(1991), and Pye and Mazzullo (1994). Suspended sediment
load, both silt and clay derived from chemical weathering, accounts for the pale whitewater characteristic of tropical rivers
such as the Amazon and its tributaries (Sioli, 1984).
4.17.5.4
Clay
Clay is a particle size and a class of minerals, and in both cases

a possible weathering product. Clay in the sedimentary system
is derived from three sources (Eberle, 1984):
1. Inherited from parent material (such as shale or mudstone). Mechanical weathering or abrasion would release
290
(a)
(c)
(b)
(d)
Figure 4 Backscatter electron (BSE) micrographs of sand grains seen in cross section. BSE shading is proportional to the chemistry of the
mineral, brighter backscatter corresponds to heavier elements. All samples except (d) are derived from weathered granitic saprolite exposed on
marine terraces on the California (USA) coast. Scale differs for each micrograph; scale bar (in micrometers) appears in the lower right in each
micrograph. (a) A coarse sand quartz grain, exposed near Montara Beach. The quartz grain is internally sound, with no fractures or grain
boundaries, but the edges are disintegrating into smaller particles. (b) A fine sand quartz grain, exposed near Montara Beach. The grain laced
with internal fractures or crystal boundaries and is shattered into angular silt- and clay-sized fragments. (c) A coarse sand quartz grain, exposed
near Granite Canyon, Carmel, California (USA). The quartz grain is relatively intact except for minor fracturing, but an amalgam of clay- and siltsize particles, possibly a weathering rind, adheres to the upper rim of the grain. (d) Sample from beach sand near Carmel, California, derived
from proximal weathered granite. The sediment is immature, evident by the mix of minerals (quartz darkest gray, sodium and potassium
feldspars lighter gray) and angularity of the cross section. Weathering is essentially nonexistent on the quartz; the feldspars show minor etching
of grain boundaries. All micrographs by author.
clay from parent material to the sediment stream. Inherited

clays exhibit little, if any further, alteration and are not
usually susceptible to weathering in transport.
2. In situ alteration product of weathering (or deep diagenesis) of silicate minerals, such as micas or feldspars
(common in igneous and metamorphic rocks).
3. Neoformation, clay minerals precipitated from solution or
incorporating amorphous silica. This is also a weathering
product, dissolution or solution being an intermediate
step. Amorphous silica may come from weathering or
directly from volcanic sources.
Clay as a mineral is at or near the end-stage of chemical

weathering, and relatively stable (Reiche, 1950). Clay is also
relatively stable in terms of erosion, resistant to entrainment,
easily carried once transported, and last to drop out of quiet
water (Hjustrom, 1935). The type of clay formed out of
weathering is partially dependent on climatic functions such
as temperature and precipitation, though time and parent
material are critical and sometimes overriding. Birkeland
(1999) provided an overview of climate and clay genesis, but
also pointed out where these straightforward relationships do
not always follow, for instance, voluminous clay genesis in
291
arid and semiarid regions. Clay species can change or modify

over climate change. Weathering during extreme events may
also alter clay mineralogy: Reynard-Callanan et al. (2010)
found that intense wildfires can alter clay minerals in nearsurface soil horizons, and that these alterations can persist
over several years.
Clay-sized quartz, unlike actual clay phyllosilicate minerals, is similar in nature to fragmented silt-sized quartz, discussed above, though less commonly studied. Leschak and
Ferrell (1988) mentioned both pedogenic weathering and
glacial crushing origins for clay-sized quartz in suspended
sediment of the Mississippi River. Popes (1995b) nanoscale
study of quartz weathering revealed clay-sized (and smaller)
fragments of quartz, probably the result of hydration in the
crystal lattice.
4.17.6
The Role of Weathering in Cementing

Sediment
The processes of chemical weathering liberate ions that, if not

carried away in runoff or groundwater, remain in the weathering profile as recrystallized secondary minerals. These
weathering products act as indurating cements to detrital
particles at or near the surface. Bog iron (cemented by goethite), laterite (cemented by aluminum oxides), and calcrete
or caliche (cemented by calcium carbonate) are examples of
soils and surface sediments hardened by ions previously in
solution. Secondary minerals from precipitated ions are also
important at small scales, accounting for case hardening (Viles
and Goudie, 2005) as well as infills of microfractures within
and surrounding crystal grains (Figure 4, and see Chapters
4.10 and 4.5). Secondary minerals also play a role in cementing buried sediments. Silica, calcite, iron and aluminum
oxides, and clays are all known as cementing agents in sedimentary rocks. The secondary minerals may be derived from
surface chemical weathering and work into the groundwater
system, or may be derived from diagenetic processes (such as
hydrothermal alteration). Figure 5 illustrates differential
weathering due to secondary cementing in arkose, which has
indurated specific layers that resist weathering.
Cementing processes may also create aggregate particles.
Eswaran and Bin (1977) reported silt- and sand-sized aggregates of gibbsite, halloysite, and kaolinite in deep saprolite in
Malaysia, creating anomalously higher proportions of those
size fractions.
4.17.7
Summary
Weathering processes are an integral component of the rock

cycle, the system of earth materials. Primarily, weathering is
responsible for generating clastic sediment and soluble products that cement sediments, and in this process acts as a recycler as important as melting in creating and regenerating
rocks. An understanding of weathering processes is important
to research in sedimentology, including genesis of sediments,
alteration of sediments in transport or temporary storage, and
cementing of sediments in the near-surface environment as
well as those buried in sedimentary accumulations.
Figure 5 Pinnacles and mushrooms of Dawson Arkose protected

by iron-cemented caps, near Colorado Springs, Colorado (USA), in
this vintage image from 1914. The arkose itself is an immature
sediment product of nearby weathered Pikes Peak granite. Later,
weathering of the arkose resulted in kaolinitic decomposition of the
constituent feldspars as well as liberation of iron oxides which in turn
cemented certain layers, resulting in the pinnacle caps and
midpinnacle indurated layers. Photo source: N.H. Darton, US
Geological Survey, Plate 12, Folio 203. Reproduced from US
Geological Survey photographic library, photograph by N.H. Darton,
ID 1055 dnh01055, Plate 12, in El Paso County, Colorado, Folio (Folio
203). Image file: http://libraryphoto.cr.usgs.gov/htmllib/btch329/
btch329j/btch329z/btch329/dnh01055.jpg
Both mechanical and chemical weathering processes are

active, acting in tandem to weaken crystals and rocks, following paths of least resistance: fractures and joints, bedding
and foliation, grain or crystal boundaries, weaker minerals,
and internal defects within crystals. All ranges of particle sizes
are relevant to the weathering sediment system, from the
submicron scale of clay to the meter scale of boulders. Recent
research efforts revealed the importance of weathering in the
generation of silt- and clay-sized particles that become part of
the sediment transport system.
References
Bierman, P., Steig, E.J., 1996. Estimating rates of denudation using cosmogenic
isotope abundance in sediment. Earth Surface Processes and Landforms 21,
125139.
Oxford, 430 pp.
Bloom, A.L., 1998. Geomorphology: A Systematic Analysis of Late Cenozoic
Landforms, Third ed Waveland Press, Long Grove, Illinois, 482 pp.
Blott, S.J., Pye, K., 2008. Particle shape: A review and new methods of
characterization and classification. Sedimentology 55(1), 3163.
Campbell, E.M., 1997. Granite landforms. Journal of the Royal Society of Western
Australia 80, 101112.
292
Cockburn, H.A.P., Summefield, M.A., 2004. Geomorphological applications

of cosmogenic isotope analysis. Progress in Physical Geography 28(11),
142.
Eberle, D.D., 1984. Clay mineral formation and transformation in rocks and soils.
Philosophical Transactions of the Royal Society of London, Series A 311,
241257.
Eswaran, H., Bin, W.C., 1977. A study of a deep weathering profile on granite in
peninsular Malaysia: II. Mineralogy of the clay, silt, and sand fractions. Soil
Science Society of America Journal 42(1), 149153.
Eswaran, H., Stoops, G., 1979. Surface textures of quartz in tropical soils. Soil
Science Society of America Journal 13, 420424.
Goldich, S.S., 1938. A study in rock-weathering. Journal of Geology 46, 1758.
Goudie, A.S., Cooke, R.U., Doornkamp, J.C., 1979. The formation of silt from
quartz dune sand by salt-weathering processes in deserts. Journal of Arid
Hill, S.E., Rosenbaum, M.S., 1998. Assessing the significant factors in a rock
weathering system. Quarterly Journal of Engineering Geology 31, 8594.
Hjustrom, F., 1935. Studies of the morphological activity of rivers as illustrated by
the River Fyris. Bulletin of the Geological Institute, University of Uppsala 25,
221527.
Huggett, R.J., 2011. Fundamentals of Geomorphology, Third edn. Routledge,
London, 516 pp.
Jeong, G.Y., Cheong, C.-S., Choi, J.-H., 2007. The effect of weathering on
optically stimulated luminescence dating. Quaternary Geochronology 2(14),
117122.
Johnsson, M.J., 1993. The system controlling the composition of clastic sediments.
In: Johnsson, M.J., Basu., A. (Eds.), Processes Controlling the Composition of
Clastic Sediments. Geological Society of America, Boulder, Colorado, Geological
Society of America Special Paper 284, pp. 119.
Johnsson, M.J., Stallard, R.F., Lundberg, N., 1991. Controls on the composition of
fluvial sands from a tropical weathering environment: sands of the Orinoco River
drainage basin, Venezuela and Colombia. Geological Society of America Bulletin
103(12), 622647.
Krinsley, D.H., 1998. Backscattered Scanning Electron Microscopy and Image
Analysis of Sediments and Sedimentary Rocks. Cambridge University Press,
Cambridge, 193 pp.
Krinsley, D.H., Doornkamp, J.C., 1973. Atlas of Quartz Sand Surface Textures.
Leschak, P., Ferrell, Jr. R.E., 1988. Morphologies of suspended clay-sized quartz
particles in the Mississippi River and their relation to sedimentary sources.
Geology 16(4), 334336.
Mahaney, W.C., 1995. Glacial crushing, weathering and diagenetic histories of
quartz grains inferred from scanning electron microscopy. In: Menzies, J. (Ed.),
Glacial Environments. Vol. 1: Modern Glacial Environments Processes,
Dynamics and Sediments. Butterworth-Heineman, Oxford, UK, pp. 487506.
Mahaney, W.C., 2002. Atlas of Sand Grain Surface Textures and Applications.
Oxford University Press, Oxford, 237 pp.
Mahaney, W.C., Claridge, G., Campbell, L., 1996. Microtextures on quartz grains in
tills from Antarctica. Paleogeography, Paleoclimatology, Palaeoecology 121(1-2),
89103.
Mazzullo, J., Leschak, P., Prusak, D., 1988. Sources and distribution of late
Quaternary silt in the surficial sediment of the Northeastern Continental Shelf of
the United States. Marine Geology 78, 241254.
McFarlane, M.J., Eckardt, F.D., Ringrose, S., Coetzee, S.H., Kuhn, J.R., 2005.
Degradation of linear dunes in Northwest Ngamiland, Botswana and the
implication for luminescence dating of periods of aridity. Quaternary International
135(1), 8390.
McNally, G.H., 1992. Engineering geology of duricrusts. In: Branagan, D.F.,
Williams, K.F. (Eds.), Papers of the Fifth Edgeworth David Day Symposium. The
Edgeworth David Society, University of Sydney, Sydney, pp. 8996, September,
1992.
Nesbitt, H.W., Young, G.M., McLennan, S.M., Keays, R.R., 1996. Effects of
chemical weathering and sorting on the petrogenesis of siiclastic sediments,
with implications for provenance studies. The Journal of Geology 104,
525542.
Peterknecht, K.M., Tietz, Gerd F., 2011. Chattermark trails: surface features on
detrital quartz grains indicative of a tropical climate. Journal of Sedimentary
Research 81(2), 153158.
Phillips, J.D., 2004. Geogenesis, pedogenesis, and multiple causality in the
formation of texture-contrast soils. Catena 58, 275295.
Philips, J.D., Luckow, K., Marion, D.A., Adams, K.R., 2005. Rock fragment
distributions and regolith evolution in the Ouachita Mountains, Arkansas, USA.
Pope, G.A., 1995a. Internal weathering in quartz grains. Physical Geography 16,
315338.
Pope, G.A., 1995b. Newly discovered submicron-scale weathering in quartz:
geographical implications. The Professional Geographer 47(4), 375387.
Pope, G.A., Dorn., R.I., Dixon, J.C., 1995. A new conceptual model for
Prothero, D.R., Schwab, F., 1996. Sedimentary Geology: Introduction to Sedimentary
Rocks and Stratigraphy. W.H. Freeman, New York, 575 pp.
Pye, K., 1983. Formation of quartz silt during humid tropical weathering of dune
sands. Sedimentary Geology 34, 267282.
Pye, K., Mazzullo, J., 1994. Effects of tropical weathering on quartz grain shape; an
example from northeastern Australia. Journal of Sedimentary Research 64(3a),
500507.
Reiche, P., 1950. A Survey of Weathering Processes and Products. University of
New Mexico Publications in Geology. The University of New Mexico Press,
Albuquerque, No. 3, 91 pp.
Reynard-Callanan, J., Pope, G.A., Gorring, M.L., Feng, H., 2010. Effects of highintensity forest fires on soil clay mineralogy. Physical Geography 31(5),
407422.
Riezebos, P.A., Van der Waals, L., 1974. Silt-sized quartz particles: a proposed
source. Sedimentary Geology 12(4), 279285.
Ritter, D.F., Kochel, R.C., Miller, J.M., 1995. Process Geomorphology, Third ed
William C. Brown Publishers, Dubuque, Iowa, 546 pp.
Sequeira Braga, M.A., Paquet, H., Begonha, A., 2002. Weathering of granites in a
temperate climate (NW Portugal): granitic saprolites and arenization. Catena 49,
4156.
Sioli, H., 1984. The Amazon and its main affluents: hydrography, morphology of
the river courses, and river types. In: Sioli, H. (Ed.), The Amazon. Limnology
and Landscape Ecology of a Mighty Tropical River and its Basin. Dr. W. Junk,
The Hague, pp. 127165.
Smalley, I.J., 1966. The properties of glacial loess and the formation of loess
deposits. Journal of Sedimentary Research 36(3), 669676.
Smalley, I.J., 1995. Making the material: the formation of silt sized primary
mineral particles for loess deposits. Quaternary Science Reviews 14(7-8),
645651.
Smith, B.J., McGreevy, J.P., Whalley, W.B., 1987. Silt production by weathering of a
sandstone under hot arid conditions: an experimental study. Journal of Arid
Stratham, I., 1973. Scree slope development under conditions of surface particle
movement. Transactions of the Institute of British Geographers 59, 4153.
Thornburry, W.D., 1966. Principles of Geomorphology. John Wiley and Sons, New
York, 618 pp.
Twidale, C.R., Romani, J.R.V., 2005. Landforms and Geology of Granite Terrains.
Taylor and Francis Group, London, 359 pp.
Viles, H.A., Goudie, A.S., 2005. Biofilms and case hardening on sandstones from
Al-Quwayra, Jordan. Earth Surface Processes and Landforms 29, 14731485.
Villa, N., Dorn, R.I., Clark, J., 1995. Fine material in rock fractures: Aeolian dust or
weathering. In: Tchakerian, V.P. (Ed.), Desert Aeolian Processes. Chapman and
Wasklewicz, T., 1994. Importance of environment on the order of mineral weathering
in olivine basalt, Hawaii. Earth Surface Processes and Landforms 19, 715734.
White, I.D., Mottershead, D.N., Harris, S.J., 1992. Environmental Systems: An
Introductory Text, Second edn. Chapman and Hall.
Wright, J., Smith, B., 1993. Fluvial comminution and the production of loess-sized
quartz silt. Geografiska Annaler 75A(12), 2534.
Wright, J., Smith, B., Whalley, B., 1998. Mechanisms of loess-sized quartz silt
production and their relative effectiveness: laboratory simulations.
Wright, J.S., 2007. An overview of the role of weathering in the production of
quartz silt. Sedimentary Geology 202, 337351.
Zreda, M.G., Phillips, F.M., 2000. Cosmogenic nuclide buildup in surficial
materials. In: Noller, J.S., Sowers, J.M., Lettis, W.R. (Eds.), Quaternary
Geochronology: Methods and Applications. American Geophysical Union,
293
Biographical Sketch
Gregory Pope is an Associate Professor in Earth and Environmental Studies at Montclair State University in New
geoarchaeology, working in China, Latin America, Western Europe, and the Western and Northeastern US. He is
active in both the Geological Society of America and the Association of American Geographers, and served as
Chair of the Geomorphology Specialty Group and a Regional Councilor for the AAG.
5.1 Dedication to Dr. Kurt Lang Frankel

LA Owen, University of Cincinnati, Cincinnati, OH, USA
References
March 21, 1978July 2, 2011

This volume of the Treatise of Geomorphology is dedicated to
the memory of Dr. Kurt Lang Frankel.
Kurt was an Assistant Professor in the School of Earth and
Atmospheric Sciences at the Georgia Institute of Technology in
Atlanta, Georgia, USA. Throughout his short career, Kurts research and teaching was focused on tectonic geomorphology,
in particular along active plate margins where his work furthered our understanding of the nature and dynamics of
landscape evolution. Kurt applied innovative field-based
methods, light detection and ranging (LiDAR), and cosmogenic
nuclides to study active faults and fluvial systems. Although his
professional career was just beginning to bloom, it was clear to
Kurts colleagues and collaborators, and others in our community that he was already becoming one of the leaders in his
field and a role model for his students and peers alike.
Kurts doctoral work at the University of Southern
California focused on defining geomorphic rates of displacement along the Death ValleyFish Lake Valley fault to help
quantify partitioning of deformation across the Eastern California Shear Zone (Frankel, 2007). His love for Death Valley
and the surrounding areas first developed from his undergraduate days at the University of North Carolina at Chapel
Hill when he undertook a geomorphic study of Holocene fault
scarps near Badwater (Frankel et al., 2001). Kurt continued to
work on tectonics and landscape development in Death Valley
and the adjacent regions after completing his doctorate in
2007, and he published numerous benchmark papers on the
region (Frankel and Dolan, 2007; Frankel et al., 2007a, b,
2011; Ganev et al., 2010, 2012; Hoeft and Frankel, 2010; Foy et
al., 2012). In addition to this research, Kurt also worked on
active faulting and landscape evolution in other regions of the
world, including Mongolia (Frankel et al., 2010c), Italy, and

Turkey.
Kurts research for his Masters degree at Lehigh University
began his interest in tectonics, erosion, and topographic development (Frankel, 2002). This interest developed into research on mountain fronts, drainage systems, and knickpoints,
resulting in important contributions that include Frankel and
Pazzaglia (2005, 2006), Frankel et al. (2007c), Stock et al.
(2009) and Gallen et al. (2011).
Kurts interest in cosmogenic nuclide geochronology included both research to improve methodologies and applications to geomorphic studies (Frankel et al., 2010b; Owen
et al., 2011). Not only did Kurt apply these methods vigorously
and whole heartily in the field, but he was also a gifted laboratory scientist and he established a start-of-the-art preparation laboratory for cosmogenic nuclides at Georgia Institute
of Technology.
Kurt thrived in the field. His enthusiasm for field geology
was very contagious. Kurts field research took him to farreaching parts of the world, extending from the Great Basin to
the Appalachians, and to Turkey, Mongolia, and Italy. Kurt
organized many memorable field trips for students and colleagues, editing several impressive field guidebooks (Frankel
et al., 2008, 2010c).
At Georgia Tech, Kurt was a force to be reckoned with.
Coming straight from Graduate School at the University of
Southern California, Los Angeles, USA, he hit the ground
running, taking on a large and energetic group of students,
developing a state-of-the-art laboratory, and winning over
undergraduates with his teaching, which he won awards for.
Quick to laugh, and diligent with his research and teaching,
his time at Georgia Tech will be remembered fondly.
Kurts contribution to the professional community was
exemplary. He chaired numerous meeting topical sessions,
served on a National Science Foundation panel, and was the
chair of the steering committee for the National Center for
Airborne Laser Mapping.
Kurt was clearly not only an outstanding young scientist
and academic, but he was also a wonderful and generous individual. He will be immensely remembered for his profound
friendship and devotion to his friends, students and colleagues; and for his sincerity, modesty, and humor. Kurt leaves
behind his wife, Stephanie Briggs, who he was devoted to
since they met as undergraduates at the University of North
Carolina. Kurt will be greatly missed by all his family, friends,
and the academic community.
References
Owen, L.A., 2013. Dedication to Dr. Kurt Lang Frankel. In: Shroder, J.
(Editor in Chief), Owen, L.A. (Ed.), Treatise on Geomorphology. Academic
Press, San Diego, CA, vol. 5, Tectonic Geomorphology, pp. 12.
doi:10.1016/B978-0-12-374739-6.09006-0
Foy, T.A., Frankel, K.L., Lifton, Z.M., Johnson, C.W., Caffee, M.W., 2012. Distributed
extensional deformation in a zone of right-lateral shear: Implications for geodetic
Dedication to Dr Kurt Lang Frankel
versus geologic rates of deformation in the eastern California shear zone-Walker

Lane. Tectonics 31, TC4008.
Frankel, K.L., 2002. Quantitative topographic differences between erosionally
exhumed and tectonically active mountain fronts: Implications for late Cenozoic
evolution of the southern Rocky Mountains. M.S. thesis. Lehigh University,
Bethlehem, 138 pp.
Frankel, K.L., 2007. Fault slip rates, constancy of seismic strain release, and
landscape evolution in the eastern California shear zone. Ph.D. dissertation.
University of Southern California, Los Angeles, 179 pp.
Frankel, K.L., Brantley, K.S., Dolan, J.F., et al., 2007a. Cosmogenic 10Be and 36Cl
geochronology of offset alluvial fans along the northern Death Valley fault zone:
implications for transient strain in the eastern California shear zone. Journal of
Geophysical Research Solid Earth, 112. doi:10.1029/2006JB004350.
Frankel, K.L., Dolan, J.F., 2007. Characterizing arid-region alluvial fan surface
of Geophysical Research Earth Surface, 112. doi:10.1029/2006JF000644.
Frankel, K.L., Dolan, J.F., Owen, L.A., Finkel, R.C., Hoeft, J.S., 2007b. Spatial
variations in slip rate along the Death Valley-Fish Lake Valley fault system from
LiDAR topographic data and cosmogenic 10Be geochronology. Geophysical
Research Letters, 34. doi:10.1029/2007GL030549.
Frankel, K.L., Dolan, J.F., Owen, L.A., Ganev, P.N., Finkel, R.C., 2011. Spatial and
temporal constancy of seismic strain release along and evolving segment of the
Pacific-North America plate boundary. Earth and Planetary Science Letters 304,
565576. doi:10.1016/j.epsl.2011.02.034.
Frankel, K.L., Finkel, R.C., Owen, L.A., 2010b. Terrestrial cosmogenic nuclide data
reporting standards needed. Eos 91, 3132. doi:10.1029/2009ES002782.
Frankel, K.L., Glazner, A.F., Kirby, E., et al., 2008. Active tectonics of the Eastern
California shear zone. In: Duebendorfer, E.M., Smith, E.I. (Eds.), Field Guide to
Plutons, Volcanoes, Reefs, Dinosaurs, and Possible Glaciation in Selected Areas
of Arizona, California, and Nevada. Geological Society of America, Denver, CO,
Field Guide 11, pp. 4381. doi:10.1130/2008.fld011(03).
Frankel, K.L., Jayko, A.S., Glazner, A.F., 2001. Characteristics of Holocene fault
scarp morphology, southern part of the Black Mountains fault zone, Death
Valley. In: Machette, M.N., Johnson, M.L. (Eds.), Field Trip Guidebook for the
Pacific Cell of the Friends of the Pleistocene. USGS, Death Valley, CA, Open
File Report 01-51, pp. M205M216.
Frankel, K.L., Lee, J., Bishop, K., Dawers, N., Ganev, P., Unruh, J., Owen, L.,
2010c. Miocene Quaternary tectonic evolution of the northern eastern
California shear zone. In: Clifton, H.E., Ingersoll, R.V. (Eds.), Geologic
Excursions in California and Nevada: Tectonics, Stratigraphy, and Hydrogeology.

Pacific Section, SEPM (Society for Sedimentary Geology), Tulsa, OK, Book 108,
pp. 173231.
Frankel, K.L., Pazzaglia, F.J., 2005. Tectonic geomorphology, drainage basin
metrics, and active mountain fronts. Geografia Fisica e Dinamica Quaternaria 28,
721.
Frankel, K.L., Pazzaglia, F.J., 2006. Mountain fronts, base level fall, and landscape
evolution: insights from the southern Rocky Mountains. In: Willett, S., Hovius,
N., Brandon, M., Fisher, D. (Eds.), Tectonics, Climate, and Landscape Evolution.
Geological Society of America, Denver, CO, Special Paper 398, pp. 419434.
Frankel, K.L., Pazzaglia, F.J., Vaughn, J.D., 2007c. Knickpoint evolution in a
vertically bedded substrate, upstream dipping terraces, and Atlantic slope
bedrock channels. Geological Society of America Bulletin 119. doi:10.1130/
B25965.1.
Frankel, K.L., Wegmann, K.W., Bayasgalan, A., et al., 2010c. Late Pleistocene slip
rate of the Hoh Serh-Tsagaan Salaa fault zone, Mongolian Altai and
intracontinental deformation in central Asia. Geophysical Journal International
183, 11341150. doi:0.1111/j.1365-246X.2010.04826.x.
Gallen, S.F., Wegmann, K.W., Frankel, K.L., et al., 2011. Hillslope response to
knickpoint migration in the southern Appalachians: implications for the evolution
of post-orogenic landscapes. Earth Surface Processes and Landforms 36,
12541267.
Ganev, P.N., Dolan, J.F., Frankel, K.L., Finkel, R.C., 2010. Rates of extension along
the Fish Lake Valley fault and transtensional deformation in the eastern
California shear zone-Walker Lane. Lithosphere 2, 3349.
Ganev, P.N., Dolan, J.F., McGill, S.F., Frankel, K.L., 2012. Constancy of geologic
slip rate along the central Garlock fault: implications for strain accumulation and
release in southern California. Geophysical Journal International 190, 745760.
Hoeft, J.S., Frankel, K.L., 2010. Temporal variations in extension rate along the
Lone Mountain fault and strain partitioning in the eastern California shear zoneWalker Lane. Geosphere 6, 917936. doi:10.1130/GE00603.1.
Owen, L.A., Frankel, K.L., Knott, J.R., Reynhout, S., Finkel, R.C., Dolan, J.F.,
Lee, J., 2011. Beryllium-10 terrestrial cosmogenic nuclide surface exposure
dating of Quaternary landforms in Death Valley. Geomorphology 125, 541557.
doi:10.1016/j.geomorph.2010.10.024.
Stock, G.M., Frankel, K.L., Ehlers, T.A., Schaller, M., Briggs, S.M., Finkel, R.C.,
2009. Spatial and temporal variations in denudation of the Wasatch Mountains,
Utah, USA. Lithosphere 1, 3440. doi:10.1130/L15.1.
5.2 Tectonic Geomorphology: A Perspective

5.2.1
Introduction
5.2.2
Development of Tectonic Geomorphology and Advances Related to the Discipline
5.2.3
Recent Research Foci (Subdisciplines)
5.2.3.1
Landscape and Tectonic Evolution of Active Plate Margins
5.2.3.2
Mountain Building
5.2.3.3
Development of Fault and Fold Systems
5.2.3.4
Evolution of Passive Margins, Continental Interiors, and Plateau Uplift
5.2.3.5
Volcanic Geomorphology
5.2.3.6
Paleoseismology and Seismic Hazard Assessment
5.2.3.7
Tectonics, Climate Change and Erosion, and Polygenetic Landscapes
5.2.4
Future Advances
Acknowledgments
References
Glossary
Base level Lowest point that a drainage can flow to and is
generally sea level or the level of a internally drained lake.
Continental drift The movement of Earths continents
relative to each other.
Davisian cycle of erosion A model for landscape
development proposed by William Morris Davis in which
landscape change from youthful to mature by processes of
erosion.
Delamination Loss/detachment and sinking of part of the
lowermost lithosphere.
Dendrochronology Method of dating using tree rings.
Digital elevation models (DEMs) Computer model or
three-dimensional representation of a terrains surface.
Fault scarp degradation Reduction of the original form of
a scarp produced by a fault by processes of erosion,
transportation, and deposition of the scarp material.
Geochronology Measurement of time intervals or dating
of events on the geological timescale.
Geodesy Study of the shape and size of Earth by survey
and mathematical means.
Geodetic Pertaining to the study of shape and size of
landforms, landscapes, and/or Earth.
Geomorphic indices Quantitative method/descriptor for
a landform and/or landscape.
Glacioisostasy Crustal subsidence or uplift related to
loading or unloading of the crust due to glacier/ice sheets
growth or melting.
Global positioning systems (GPS) A space-based global
navigation satellite system (GNSS) that provides a location
on Earth.
Ground-penetrating radar (GPR) Geophysical method
that uses radar to image the subsurface.
4
4
7
7
7
8
8
8
9
9
9
10
10
Hot-spots Volcanic center, 100200 km across, that

persists for a few tens of millions of years.
Lahars Debris flow/mudflow comprising volcaniclastic
material.
Magnetic polarity timescale A record of the succession of
magnetic reversals or anomalies of Earths magnetic field
determined from the magnetic properties of\rock or
sediment successions that can be used to determine the
geologic age of a rock or sediment.
Magnetometer Geophysical equipment used to measure
Earths magnetic field or the magnetic characteristics of rock
and sediment.
Mantle upwelling Movement of mantle material from
lower depths within the mantle toward the crust.
Mountain front sinuosity Geomorphic indices that
quantifies the straightness of a mountain, which is a function
of the tectonically activity along the mountain front.
Optically stimulated luminescence (OSL) dating A method
used to determine the age of sediment using the acquired
luminescence signal induced in minerals by ionizing radiation
in the ground and from cosmic ray particles.
Orogen A mountain belt.
Orogenic Pertaining to mountains.
Paleoelevation Former elevations/altitudes.
Paleoseismology Study of past earthquakes.
Passive margin The transition between oceanic and
continental crust where there is little/no movement
between the plates.
Radiocarbon (14C) dating Method used to determine
the age of material containing carbon (usually organic)
by measuring the decay of the 14C isotope within the
material.
Owen, L.A., 2013. Tectonic geomorphology: a perspective. In: Shroder, J.

http://dx.doi.org/10.1016/B978-0-12-374739-6.00077-4
Tectonic Geomorphology: A Perspective
Recurrence interval The characteristic time between two

similar events such as an earthquake or volcanic eruption.
Stream-gradient indices Geomorphic indices that
describe changes in stream gradient with respect to the
position along the length of the stream.
Terrestrial cosmogenic nuclides (TCN) Nuclides
produced by the interaction of cosmic ray particles with
target atoms (generally in rock or sediment) at Earths
surface, including such nuclides as 10Be, 14C, 26Al, and 36Cl.
TCNs are used to date surfaces or buried sediment.
Thermochronology Study of the thermal evolution of a

region of a planet.
Tomography Three-dimensional imaging using
sections derived from the use of any kind of penetrating
wave.
Transform plate boundary A lithospheric plate margin
where the relative motion of the two plates is horizontal, so
that they essentially slide past each other.
Abstract
Tectonic geomorphology is a relatively young subdiscipline of geomorphology. During the past decades, tectonic geomorphology has developed into several main areas of emphasis including: landscape evolution of active plate margins;
mountain building; development of active fault and fold systems; the evolution of passive margins, continental interiors
and plateau uplift; volcanic geomorphology; paleoseismology and seismic hazard assessment; and interaction of tectonics,
climate change, erosion, and polygenetic landscapes and hazard mitigation. These areas of focus reflect the growth of new
studies in tectonics, climate and, Earth surfaces processes, and technological advances such as remote sensing, global
positioning systems (GPSs), computers, geochronology, shallow geophysics and geochemistry.
5.2.1
Introduction
Tectonic geomorphology, as its name suggests, is the merging

of the study of tectonics and geomorphology. But it is far more
than simply an overlap between these two immense fields.
Tectonic geomorphology also incorporates the study of geodesy, geophysics, geochemistry, geochronology, archaeology,
and paleoclimatology among others. This chapter examines
the development of the relatively young discipline of tectonic
geomorphology and highlights current areas of research emphasis and possible future directions of study. In writing this
perspective, the author has tried to be as objective as possible
and to credit the major contributions to tectonic geomorphology. However, he should like to sincerely apologize at the
outset if anyone or any major study has not been credited.
Tectonic geomorphology is growing into a vast discipline and
it is becoming increasingly challenging to keep track of all the
exciting new developments that are emerging from so many
different fields. The chapters in the treatise (Owen, 2013) that
follow this perspective will expand in considerable depth and
detail many of the subjects mentioned in a very cursory way in
the text below.
5.2.2
Development of Tectonic Geomorphology and

Advances Related to the Discipline
Tectonic geomorphology is one of the youngest branches of

geomorphology. Yet the study of the relationship between
Earths internal (tectonic/endogenetic) and external (surface/
exogenetic) processes in relation to the evolution of landscapes has a long history. Tectonic geomorphology, however,
introduced to geomorphology a far deeper awareness of the
importance of tectonics in helping to drive landscape development and influence surface processes.
Tectonics is essentially the study of structures and processes

that are associated with the deformation of Earths crust as a
result of deep-seated processes. The Oxford English Dictionary
defines tectonic as Belonging to the actual structure of the
Earths crust, or to general changes affecting it. Also with
reference to other planets.
The use of the term tectonic is generally considered to be
relatively new, relating to the study of plate tectonics that
substantially developed during the late 1960s and 1970s. But
the use of the term dates back to at least 1894, when Dawkins
(1894) and Lapworth and Watts (1894) used it in relation to
geologic structures in southern Britain. Tectonics has become a
vital part of the Earth science lexicon with whole branches of
the discipline devoted to its study, including journals with
such names as Tectonics and Tectonophysics.
There is no formal definition of tectonic geomorphology,
but it is broadly defined as the study of landforms produced or
affected by tectonic processes such as faulting, tilting, folding,
uplift or subsidence, and the application of geomorphology to
the study of tectonic problems. Moreover, many studies in
tectonic geomorphology commonly include a climate component and attempt to analyze how the evolution of landforms, tectonic processes, and climate are linked. Combining
the use of the terms tectonic with geomorphology to distinguish the field of tectonic geomorphology really did not
come into vogue until the 1980s.
Probably the first most significant publication on tectonic
geomorphology was the proceedings of the fifteenth Annual
Binghamton Geomorphology Symposia in 1984 for which the
theme was tectonic geomorphology (Morisawa and Hack,
1985). Morisawa and Hacks (1985) proceedings volume
comprised 16 contributions ranging from continental- to
landform-scale studies. The first papers cited in ScienceDirect
that appear with tectonic geomorphology in their titles were
published in 1988 (Alexander, 1988; Wells et al., 1988). But it
was not until 1998 that tectonic geomorphology was used in

the keywords of articles cited in ScienceDirect, which include
the papers of Demoulin (1998) and Li et al. (1998). Tectonic
geomorphology, for example, does not even appear as a topic
in the Macmillan Encyclopedia of Earth Sciences (Dasch,
1996) and the Encyclopedia of Geology (Seller et al., 2005).
Further, tectonic geomorphology, senso stricto, receives less
than a page entry in Goudies (2004) Encyclopedia of Geomorphology. Moreover, during 2009 and 2010 only 14 papers
cited tectonic geomorphology in their keyword lists for
papers listed in ScienceDirect. Yet, without doubt tectonic
geomorphology is now one of the most rapidly growing
branches of geomorphology and as such a whole volume of
the Treatise of Geomorphology is devoted to this subject,
which includes this perspective (Shroder, 2013; Owen, 2013).
The development of tectonic geomorphology as a discipline has a long history and has been ingrained into geomorphology for well over a century. Yet, of course, the field
was not called tectonic geomorphology until the 1980s. Tectonic geomorphology was an intimate part of such concepts as
the Davisian cycle of erosion and various models of landscape evolution, but, of course, researchers talked more of
phases of uplift and subsidence, and then related geomorphology to the current paradigms of the time. Beckinsale and
Chorley (1991), Burt et al. (2008), and Grapes et al. (2008)
provided comprehensive and fascinating reviews of the history
of the development of the study of landforms from 1890 to
1965, which highlighted some of the progress of studies of the
relationship between crustal processes and geomorphology
before plate tectonics was established as a theory and the key
framework for geology. Kamp and Owen (2013) also described some of these developments with reference to polygenetic landscapes.
The first step toward the development of tectonic
geomorphology was the continental drift hypothesis of
Wegener (1915). With the development of the magnetic
polarity timescale, which was an essential part of the
VineMatthewsMorley hypothesis connecting patterns of
magnetic reversals on the ocean floors with sea-floor spreading, Wegeners view was confirmed and ultimately the theory
of plate tectonics was established (Vine and Matthews, 1963).
Marvin (1973), Hallam (1973), Glen (1982), Schwarzbach
(1986), and Stewart (1990) presented good reviews of these
developments of the science of plate tectonics. Plate tectonics
clearly set the background for tectonic geomorphology.
With plate tectonics established, geomorphologists had a
modern framework to examine landscapes and their development on all scales, but most significantly on global and
continental scales. Of particular note during the early years of
the development of plate tectonics was the pioneering work
done by Hack (1973) who developed a stream-gradient index,
which crudely correlates to stream power and reflects the
extent of surface uplift, and Bull (1977) who analyzed landscape development along active mountain fronts. Bull and
McFadden (1977), Keller (1986), and Bull (1991) continued
the development of geomorphic indices to quantify and
compare how tectonic activity, notably including mountain
front sinuosity and ratio of valley-floor width to valley height.
The study of fault scarp degradation was an area of great
focus during this period (Wallace, 1977; Bucknam and
Anderson, 1979; McCalpin, 1982). Much attention was also

given to understand the origin of marine terraces and how
they could be used to reconstruct former sea level stands (e.g.,
Chappell, 1974), rates of tectonic uplift (Bull and Cooper,
1986; Muhs et al., 1990), and rates of glacioisostatic uplift
(Morner, 1990).
Probably two of the most significant events in helping to
formulate tectonic geomorphology as a discipline were the
Geological Society of America Penrose Conference held in
1983 at Winnemucca, Nevada and the fifteenth Annual Binghamton Geomorphology Symposium, both on tectonic
geomorphology.
The Penrose Conference set the stage for what would be
the focus of much work for the next decade or so. The
56 scientists from six countries who attended discussed the
following topics: (1) tectonic processes, forms, and rates;
(2) geomorphic processes, rates, and landforms; (3) interactions of tectonic movements and fluvial systems; (4)
marine terraces and landforms of tectonically active coastlines;
(5) effects of Quaternary climate change on fluvial systems
in areas of differing rates of vertical and horizontal
tectonic deformation; (6) chronosequences of soils and
dating of displaced landforms; (7) faults, slip rates, and recurrence intervals; and (8) modeling of the effects of base-level
fall on elements of fluvial landscapes (Bull and Wallace,
1985).
At the Penrose Conference, many new exciting developments were presented and discussed, including: (1) the recognition of landforms that are particularly sensitive in
recording vertical or horizontal deformation; (2) devising new
dating techniques (and extension of old techniques) in assessing the ages of stream terraces, glacial deposits, and marine
terraces; and (3) development of processresponse models to
explore the relative importance of tectonism in the operation
of geomorphic systems and to estimate times of discrete tectonic events.
The Annual Binghamton Geomorphology Symposium that
followed a year later helped crystallize the current thinking
and the conference proceedings, Morisawa and Hack (1985),
became a reference source for many of the young geomorphologists who began to take an interest in tectonics.
As tectonic geomorphology was beginning to take on
meaning, there was also growing interest in neotectonics and
active tectonics. Neotectonics, a subdiscipline of tectonics, is
essentially synonymous with active tectonics, and as such, is
the study of contemporary motions and deformations of
Earths crust including those of the very recent geologic past.
The term neotectonics is attributed to Obruchev (1948).
Methods used in the study of neotectonics can include remote
sensing, structural geology, seismology, regional geophysics,
engineering geology, and geomorphology (Hancock and
Williams, 1986).
Much of what was being done under the auspice of
neotectonics included studies in tectonic geomorphology
(Doornkamp, 1986). However, neotectonics and active
tectonics are generally considered a broader range of study
than what many geomorphologists were studying in the
latter decades of the twentieth century. Nevertheless, the
overlap between neotectonics and tectonic geomorphology is
considerable.
Interest in neotectonics grew in the 1980s with the establishment in 1981 of the Commission on Neotectonics of the
International Union for Quaternary Research (INQUA), which
was presided over by Nils-Axel Morner. The commission ran
numerous workshops and conference symposia.
In 1983, the American Geophysical Union ran a symposium entitled Active Tectonics which was directed by
Robert E Wallace. It brought together geomorphologists,
structural geologists, Quaternary geochronologists, and geophysicists (including geodesists) mainly with a focus on establishing earthquake geology as a discipline. The meeting
resulted in the publication of the book, Active Tectonics,
under the auspice of the Geophysics Study Committee, Geophysics Research Forum, Commission on Physical Sciences,
Mathematics, and Resources, and the National Research
Council, which included studies of tectonic geomorphology.
The book was edited by Wallace (1986a) and its first chapter,
also written by Wallace (1986b), providing an overview and
recommendations, was particularly important as it showed the
significance of active tectonics to society. Wallace is not credited in the publication, which was probably a reflection of
Wallaces sincere modesty (Yeats, 2002). The conference and
publication were instrumental in bringing the use of the term
active tectonics into vogue (somewhat replacing the use of
term neotectonics).
Yet notable conferences continued to be held under the
title of neotectonics, and neotectonics still persists as a distinct
discipline. Important conferences on neotectonics include a
Tectonics Study Group (TSG) meeting of the Geological
Society of London (GSL) and the International Conference on
Neotectonics hosted jointly by the Neotectonics Commission
of INQUA, the Quaternary Research Association and the TSG
at Burlington House in London during December 1984 and
January 1992, respectively (Hancock and Williams, 1986;
Stewart, 1993; Stewart et al., 1993; Owen et al., 1993).
Vita-Finzis (1986) book, Recent Earth Movements provided a
great introduction to the modern study of neotectonics, alluding much to the geomorphology aspects of the study of
active tectonics.
Keller and Pinters (1996) book, Active Tectonics provided
a wealth of information on tectonic geomorphology, including
a chapter called Landforms, Tectonic Geomorphology, and
Quaternary Chronology, and other chapters on geomorphic indices, rivers, coastlines, and mountain building, all pertinent to
tectonic geomorphology. The second edition, published in 2002
(Keller and Pinter, 2002), retains the same structure, but provides
more updated examples. Yeats et al. (1997) devoted a whole
chapter to tectonic geomorphology in their seminal book, The
Geology of Earthquakes.
By the turn of the twenty-first century, tectonic geomorphology had truly become a discipline in its own right.
The publication of Burbank and Andersons (2001) textbook,
Tectonic Geomorphology, solidified the discipline and has
been a main reference source for tectonic geomorphology over
the last decade and is used by many universities as a textbook
for teaching tectonic geomorphology courses. The second
edition of Burbank and Anderson (2012) updates much of this
seminal volume.
From the 1970s through the 1990s, much research focused
on the role of climate in geomorphology, and the discipline of
climatic geomorphology became well established (Derbyshire,

1973, 1976; Budel, 1982; Bull, 1991). As detailed paleoclimate
records evolved through ice core studies in the 1990s (e.g.,
Dansgaard et al., 1993), it became evident just how rapidly
climate could change and that the magnitude of that change
was large. Increasing recognition was given to how landscapes
likely adjust after times of climatic transition (Ballantyne,
2002). As such, the role of climate became integrated into
tectonic geomorphic studies.
As new technologies developed throughout the 1990s and
2000s, tectonic geomorphology acquired additional boosts.
These included advances in remote sensing, geochronology,
geophysical exploration methods, numerical modeling,
and enhanced computing, particularly developments in
microcomputing.
Remote sensing now involves numerous methods, including simple photography from balloons and airplanes to
spaceborne techniques ranging from multispectral imaging to
space-geodetic methods. As the National Research Council
(2008) highlighted in their report on the first 50 years of scientific achievements in Earth Observation from Space, the
transformative nature of modern remote sensing is demonstrated by the fact that a mere 50 years ago, a traveler might
not know his/her position on Earth to better than 500 m, even
after expending considerable effort in tedious reduction of
geodetic observations, yet, today, a person can determine their
position to within a meters precision in real time, anywhere
on the planet, using an inexpensive Global Positioning System
(GPS) receiver.
Since GPS technology measures a highly accurate threedimensional position on Earths surface at a precise time
relative to the known positions of several satellites, it is possible to determine rates of tectonic motion and develop highly
precise maps. Moreover, the International Earth Reference
System is accurate to better than 1 cm in all components, including the time-dependent position of the geocenter, to
provide us with a robust reference frame (National Research
Council, 2008). Even more impressive is that we now have the
ability to determine relative positioning to a millimeter anywhere on the surface of the planet, or in orbit. Developments
in light detection and ranging (Lidar), which use a laser beam
mounted on an aircraft, satellite, truck, ship, or hand-held
device to measure accurate distances to a target point on land
or underwater, has allowed us to use this method for precise
surveying in numerous tectonic geomorphic studies (National
Research Council, 2010). In addition, the development of
interferometric synthetic aperture radar (InSAR) which uses
radar mounted on an airplane or satellite to measure the
strength and round-trip travel time of a microwave signal to
gather elevation data, has allowed us to determine and map
deformation experienced during an earthquake over large
areas (102103 km2) with centimeter accuracy (Massonnet
et al., 1993; Zebker et al., 2000).
These positioning data can be used to produce a digital
surface model (DSM) which can then be processed to yield
digital elevation models (DEMs), digital terrain models, contours, and three-dimensional feature data (National Research
Council, 2008, 2010). The topographic models that are developed can be used to in conjunction with computer modeling for analysis of landscape evolution.
The National Research Council (2010) highlighted new

developments in geochronology in their report on new horizons for research on Earths surface that have helped advance
the study of Earth surface processes and landforms. In particular, new geochronologic methods have accelerated the
study of tectonic processes on geomorphic timescales
(100105 years). It should be noted that it was only in 1949
that Willard Frank Libby and his colleagues at the University
of Chicago developed radiocarbon dating (Arnold and Libby,
1949), probably the most widely used numerical dating
technique in tectonic geomorphology. So it really has only
been during the last 4050 years that radiocarbon dating has
been providing ages on landforms and sediments in geomorphology. The method essentially revolutionized much of
what we can do in tectonic geomorphology and other branches of geomorphology.
The development of terrestrial cosmogenic nuclide (TCN)
surface exposure dating methods since about the late 1980s has
helped accelerate interest in tectonic geomorphology because
now many landforms that could not be previously dated with
radiocarbon methods, due to the lack of organic material, can
today be relatively easily dated. Moreover, the TCN method can
allow landforms from a few decades to several million years old
to be dated, far beyond the radiocarbon range of about
3050 ka. Common TCNs used in tectonic geomorphic studies
include 10Be, 26Al, 36Cl, and 10Ne (Frankel et al., 2007).
Advances in luminescence dating, specifically optically stimulated luminescence (OSL) dating, have paralleled those of TCN
dating. In particular, OSL methods have now been very successfully applied to dating sediments in fault trenches for
paleoseismic studies and other tectonic studies (Tsukamoto
et al., 2009). Recently, OSL methods have also been applied as a
thermochronological tool (Herman et al., 2010).
Another major area of growth has been the use of lowtemperature thermochronometer techniques to provide information on hundred thousand- to million-year rates of
erosion and topographic evolution. Common systems used
include U-Th-Sm/He and fission track methods in apatite
and zircon, and 40Ar/39Ar age-spectrum dating of a variety of
silicate minerals (Lee et al., 2009; Adams et al., 2009). In
addition, sidereal methods such as lichenometry, dendrochronology, and varve chronology have contributed significantly to defining ages on landforms for tectonic geomorphic
studies (Bull, 2007).
Advances in geophysical prospecting technologies have
also contributed greatly. These include advanced seismic
methods that use two- and three-dimensional imaging with
acoustic waves to resolve fossilized surface features to resolutions of 1 m or better at shallow depths and up to several
meters to kilometers below the surface (National Research
Council, 2010). Other advances include electromagnetic resistivity surveys, tomography, and magnetometers that measure electrical or magnetic fields, to yield data on groundwater
flow and subsurface geometry and structures. Groundpenetrating radar can now yield two- and three-dimensional
subsurface images at imaging depths between 1 and 50 m.
These instruments can be hand-held or mounted on vehicles.
Computing has advanced so much now that sophisticated
programs can be run on desktop computers. Of particular note
in examining the relationships among tectonics, climate, and
surface processes is the modeling work pioneered Beaumont

et al. (1999, 2001, 2004).
5.2.3
Recent Research Foci (Subdisciplines)
Since its recent establishment as a distinct branch of geomorphology, tectonic geomorphology has developed several
main areas of emphasis. These reflect the growth of tectonics,
climate studies, and Earth surfaces processes over the past few
decades, particularly in the light of the new technological
advances discussed in the previous section. Each one of these
areas of emphasis is outlined briefly below and is reflected in
the chapters in this treatise (Owen, 2013) that follow this
perspective. A common theme in each of these areas of study is
the interrelationship among tectonics, climate, and Earth
surface processes. Probably the most exciting and rewarding
aspect of tectonic geomorphology is the realization of the
complex links and feedbacks between exogenic and endogenic
processes in the evolution of landscapes and the need for
multidisciplinary, collaborative research. Fully understanding
these links and feedbacks is essential for environmental
management and sustainable development.
5.2.3.1
Landscape and Tectonic Evolution of Active Plate

Margins
With the development of plate tectonics a need has arisen to

quantify rates of vertical and horizontal crustal displacement
to help understand how plate margins evolve over time. In
particular, recently there has been growing interest in comparing rates over geologic (100 to 107 years), geomorphic (100
to 105 years), and geodetic (101 to 102 years) timescales.
Knowledge of the rates of displacement is essential in helping
to formulate tectonic and landscape models.
The HimalayanTibetan collisional orogen and the Gulf of
California-San Andreas transform plate boundary are probably the most studied in terms of tectonic geomorphology.
Owen (2010) described the development of the study
of landscape evolution for the HimalayanTibetan orogen,
specifically in terms of the advancement of tectonic
climatelandscape concepts and models. Some of concepts
and models are reviewed and expanded on in more detail in
Hodges and Adams (2013), Koons et al. (2013) and Spotila
(2013); and Schoebohm (2013) provide an overview of
our understanding of the geomorphology of continental
continental collision. Bull (2007, 2009) and Frankel and
Owen (2013) provided overviews and summaries of the tectonic geomorphology of the Gulf of CaliforniaSan Andreas
transform plate boundary.
It is probably an understatement to say that tectonic geomorphic studies have been undertaken on nearly every active
plate margin. These studies are now providing important insights into how landscapes develop in varied tectonic and
climatic settings and are building frameworks for applied
studies such as seismic hazard mitigation.
5.2.3.2
Mountain Building
Tectonic geomorphology has been used extensively in the

study of mountain building. As Koons et al. (2013)
emphasize, some of the most spectacular topography on the

planet arises from the product of localized interactions among
tectonics, climate, and erosion. Schoenbohm (2013) highlights that one of the key topics is whether or not steady state
can be achieved, and how it can be recognized. Moreover,
Schoenbohm (2013) highlights that one of the main topics
is determining the importance of the relative controls of
tectonics and climate on the width and asymmetry of
mountain belts.
Tectonic geomorphology is beginning to be used to identify the lateral transport of crust, either brittlely along strikeslip faults, or ductility through lower crustal flow. In particular,
Schoenbohm (2013) shows how drainage reorganization is a
particularly powerful tool for inferring uplift patterns. Hodges
and Adams (2013) show how the Tibetan Plateau serves as a
model for understanding the possible influence of middle and
lower crustal flow on the geomorphic and tectonic evolution
of continental plateaus. They present arguments in favor of a
weak middle (and, in some areas, lower) crust beneath the
plateau, and explore the development and features of popular
models of the HimalayanTibetan orogenic system that incorporate lateral flow of this weak crust under the influence of
gravity. Hodges and Adams (2013) argue that many of the
long- and short-wavelength features of Tibetan and Himalayan
landscapes are reasonably explained as a consequence of
crustal flow.
Schoenbohm (2013) points out that study of tectonic
geomorphology can also be important in helping to explain
the persistence of topography over long-dead orogens.
5.2.3.3
Development of Fault and Fold Systems
The development of fault and fold systems has attracted much

attention, not only for understanding landscape development,
but also for seismic hazard mitigation. In particular, there is
growing recognition that large earthquakes can occur within
fold belts, commonly on blind thrusts, such as in the case of
the North Ridge earthquake in Southern California in 1994
(Davis and Namson, 1994). This recognition has stimulated
much research on the development of fold and thrust belts
for earthquake hazard assessment. Keller and DeVecchio
(2013) highlight that the questions being addressed at present
in the field of tectonic geomorphology of fold and thrust
belt systems mainly relate to the lateral propagation of folds
and the development of traverse drainages across fold belts.
Recent studies are concerned with examining the development
of transverse drainage across folds to relate them to uplift
and fold growth. Much effort, therefore, has been focused
on developing adequate chronology to quantify fold
growth and erosion by determining rates of lateral propagation and incision by streams that cross-cut growing
folds.
Recent studies on fault systems have addressed a range of
scales. These vary from continental-scale transform and strikeslip faults, to mountain/basin-scale bounding faults, to individual fault scarps (Frankel and Owen, 2013; Nash, 2013).
Much attention has been focused on the PacificNorth
American plate boundary incorporating the San Andreas
fault and the Eastern California Shear Zone, and the
PacificAustralian plate boundary, the CaribbeanNorth

American plate boundary, and intracontinental strike-slip
systems with the HimalayanTibetan orogen to help develop
tectonic and landscape evolution models, but also for seismic
hazard mitigation in the most densely populated areas
(Frankel and Owen, 2013; Kondo and Owen, 2013).
5.2.3.4
Evolution of Passive Margins, Continental

Interiors, and Plateau Uplift
Explaining continental-scale landforms such as great escarpments and plateaus along passive margins has attracted much
attention in recent years. Blenkinsop and Moore (2013)
explain how passive margins and their associated landscapes
and landforms are created when extension related to rifting
forms oceanic crust. Blenkinsop and Moore (2013) point out
that the diversity of passive margins reflects variations
amongst thermal, isostatic, flexural, buckling, and dynamic
(mantle) sources of stress that affect margins. Flexural isostasy
plays a significant role in defining the shape of passive margins
as a consequence of erosion. Blenkinsop and Moore (2013)
present three current models for escarpment evolution which
are: (1) backwearing from an initial abrupt increase in topography; (2) downwearing to an inland drainage divide; and
(3) downwarping. Recent research is beginning to focus on
constraining these models using thermochronology, numerical modeling, and geomorphic observations. Blenkinsop and
Moores (2013) point out that the elevation of the continental
interior might be due partially to mantle upwelling (dynamic
topography), however, drainage divides within the continent,
at least in southern Africa, are more likely the result of plate
boundary stresses.
Studies of the evolution of continental interiors have focused on the very long-term geomorphic evolution of shield
areas. However, with the development of climatic geomorphology and the recognition of the complex behavior of
continental ice sheets, there has been much recent focus on
the Late Quaternary glacial and periglacial development of
continental interiors.
One of the great challenges in tectonic geomorphology has
been to define rates of plateau uplift, specifically quantifying
paleoelevation. Babault and Van Den Driessche (2013) provide a detailed review of the many aspects of the tectonic
geomorphology of plateau uplift. New techniques have included the use of oxygen, hydrogen, and carbon isotopes
within paleosols and the size of vesicles in volcanic rocks
(Garzione et al., 2000; Sahagian et al., 2002; Quade et al.,
2007; Sahagian and Proussevitch, 2007; Ghosh et al. , 2006).
Various mechanisms for plateau uplift have been proposed,
including denudational unloading, uplift associated with
hotspots, and delamination (Blenkinsop and Moore, 2013).
A further area of research related to plateau development
concerns the feedback between plateau uplift and global climate change (Raymo and Ruddiman, 1992; Ruddiman, 1997;
Clift et al., 2010).
5.2.3.5
Volcanic Geomorphology
The tectonic geomorphology of volcanoes has focused much

on hazard mitigation (Huff and Owen, 2013). This commonly
includes the study of volcanic landslides/lahars, many of

which have had catastrophic consequences. The Nevada del
Ruiz eruption in Columbia in 1985 resulted in the death of
423 000 people when Armero was buried by a lahar (Barberi
et al., 1990).
A clear area of interest in tectonic geomorphic studies of
volcanoes is the evolution of volcanic islands, which varies
from the formation of new, ephemeral islands such as Kavachi
and Surtsey, to large islands over hotspots such as Hawaii and
Iceland, and large archipelagoes along subduction zones, such
as Japan and the Philippines (Villeneuve and Bachelery, 2013;
Huff and Owen, 2013).
Active volcanoes are continuously and rapidly evolving and
this makes them one of the most exciting topics of tectonic
geomorphology. Real-time monitoring by such research
groups and organizations as the US Cascade Volcano Observatory has yielded important new insights into how volcanoes
grow and erupt. Clearly, these studies are important for volcanic hazard mitigation and monitoring will continue in the
coming years, especially in areas near large populations. These
studies are adding greatly to our understanding of tectonic
geomorphology.
5.2.3.6
Paleoseismology and Seismic Hazard Assessment
The seminal work of Wallace (1975) and Sieh (1978) that

included studies of Late Quaternary history of faults has led to
a mini-revolution in earthquake geology. Studies first concentrated on trying to define earthquake recurrence intervals
for important faults such as the San Andreas, Himalayan
Frontal Thrust, and then determining potential seismic hazards (Sieh and Jahns, 1984; Kumar et al., 2010). Most of this
work involved fault trenching and dating sediment layers
using radiocarbon methods. As paleoseisomology developed,
earthquake geologists began to recognize that earthquakes
may cluster in time and so other dating methods such as OSL
were more readily applied (Rockwell et al., 2000). As TCN
methods became more utilized, much work focused on defining partitioning of deformation across plate boundaries
such as the San AndreaGulf of California transform plate
boundary and along individual faults for seismic hazard assessment (Blisniuk et al., 2010; Frankel et al., 2011). Paleoseimological methods and studies are described in much
detail in Kondo and Owen (2013).
Another major focus of interest has been earthquaketriggered landslides. Keefer (2013) highlights the many
different types of seismically generated landslides and
emphasizes that they occur in virtually every geologic environment. Keefer (2013) stresses that even small earthquakes
(M 4) can trigger landslides, and of course, the number and
geographic distribution of landslides increases substantially
for earthquakes of larger magnitude. Earthquake-triggered
landslides can be extremely devastating as highlighted by the
Kashmir earthquake in 2005 (Owen et al., 2008). During the
largest earthquakes, tens of thousands of landslides can occur
throughout regions ranging from tens to hundreds of thousands of square kilometers, moving millions of cubic meters
of debris, (Keefer 2013). Much emphasis has been placed on
explaining the distribution of earthquake-triggered landslides
for hazard mitigation (Kamp et al., 2008, 2010).
5.2.3.7
Tectonics, Climate Change and Erosion, and

Polygenetic Landscapes
Probably the overarching theme that has enhanced the study

of tectonic geomorphology over the past few decades has been
the realization that landscapes are continuously evolving and
are the result of complex interactions between endogenetic
and exogenetic processes. Intimate links exist between tectonics, climate and erosion, and there has been a growing
realization that landscapes are polygenetic in origin. The history and an overview of the study of these links are provided
by Kamp and Owen (2013).
Of particular note is the development of the glacial buzz-saw
model in which glacial erosion generates local relief in the cores
of mountain ranges, driving rock uplift because of erosional
unloading of the crust. Spotila (2013) describes how this
mechanism has the potential for significant positive feedbacks
since rock uplift raises peaks, which in turn results in larger
glaciers. The tectonic aneurysm model argues that localized
uplift caused by enhanced glacial and fluvial erosion can
weaken the lithosphere through increased crustal heat flux and
thereby enhancing bedrock uplift (Koons et al., 2013). Koons
et al. (2013) point out that mountain uplift can lead to orographic effects that cause inhomogeneous distribution of precipitation that then becomes further concentrated by rivers and
glaciers into spatially restricted zones with a high potential of
erosion. However, as Spotila (2013) highlights, there is significant debate over the relative importance of mountain glaciers,
fluvial erosion, and mass movement in shaping mountain
landscapes. Koons et al. (2013) show that petrologic, geomorphic, and geodynamic investigations of areas of anomalous
high uplift and topography suggest a nonlinear interaction but
reinforcing actions between tectonics and surface processes,
which are likely universal, but are most clearly manifested in
those regions where high process rates produce a high signal-tonoise ratio. These models and hypotheses linking tectonics,
processes of surface erosion, climate and their interactions have
generated much research in recent times and are among the
most exciting branches of tectonic geomorphology.
5.2.4
Future Advances
Tectonic geomorphology is clearly an exciting and expanding

branch of geomorphology, which has seen many significant
advances in the past few decades. Predicting its future development is difficult, but the areas of concentration highlighted
in the past section will clearly continue to expand, and attract
much interest and research.
As the discipline develops, we are likely to see increasing
interest in examining the complex variability between regions
and on different timescales (ranging from 100 to 106 years).
Increasing attention is also likely to be paid to quantifying
the timing, rates, and magnitude of landscape evolution,
particularly since we have so many new tools now at their
disposal, including satellite remote sensing and global positioning systems, plus new developing analytical methods,
specifically those for numerical dating.
As with many branches of science, we are increasingly
asked to provide justification in terms of broader impacts.
10
As Hough and Bilham (2005) eloquently highlight in their

book After the Earth Quakes: Elastic Rebound on an Urban
Planet human population is increasing dramatically in tectonically active regions, and most of the worlds megacities are
located in areas of high seismic risk. Never has there been a
more important and urgent time in Earths history to understand and study tectonic geomorphology to aid in hazard
mitigation. Exciting times are ahead for tectonic geomorphology and the discipline is likely to continue to expand and
develop in the coming years.
Acknowledgments
Drs. C Dietsch and R Jayangondaperumal for their constructive reviews on this perspective.
References
Adams, B., Dietsch, C., Owen, L.A., Caffee, M., Spotila, J., Haneberg, B., 2009.
Exhumation and incision history of the Lahul Himalaya, northern India, based
on (U-Th)/He thermochronometry and terrestrial cosmogenic nuclide dating
techniques. Geomorphology 107, 285299.
Alexander, D., 1988. A review of the physical geography of Malta and its
significance for tectonic geomorphology. Quaternary Science Reviews 7, 4153.
Arnold, J.R., Libby, W.F., 1949. Age determination by radiocarbon content: checks
with samples of known age. Science 110, 678680.
Babault, J., Van Den Driessche, J., 2013. Plateau uplift, regional warping and
subsidence. In: Owen, L.A. (Ed.), Treatise on Geomorphology. Academic Press,
San Diego, vol. 5, pp. 93128.
Ballantyne, C.K., 2002. Paraglacial geomorphology. Quaternary Science Reviews 21,
19352017.
Barberi, F., Martini, M., Rosi, M.F., 1990. Nevado del Ruiz volcano (Columbia);
preeruption observations and the 13 November 1985 catastrophic event. Journal
of Volcanology and Geothermal Research 42, 112.
Beaumont, C., Jamieson, R.A., Nguyen, M.H., Lee, B., 2001. Himalayan tectonics
explained by extrusion of a low-viscosity crustal channel coupled to focus
surface denudation. Nature 414, 738742.
Beaumont, C., Jamieson, R.A., Nguyen, M.H., Medvedev, S., 2004. Crustal channel
flows: 1. Numerical models with applications to the tectonics of the HimalayanTibetan orogen. Journal of Geophysical Research 109, B06406. http://dx.doi.org/
10.1029/2003JB002809.
Beaumont, C., Kooi, H., Willett, S., 1999. Coupled tectonic-surface process models
with applications to rifted margins and collisional orogens. In: Summerfield,
M.A. (Ed.), Geomorphology and Global Tectonics. John Wiley and Sons Ltd,
Beckinsale, R.P., Chorley, R.J., 1991. The history of the study of landforms or the
development of geomorphology. Vol. 3: In: Historical and Regional
Geomorphology 18901950. Routledge, London, 528 pp.
Blenkinsop, T., Moore, A., 2013. Tectonic geomorphology of passive margins and
continental hinterlands. In: Owen, L.A. (Ed.), Treatise on Geomorphology.
Academic Press, San Diego, vol. 5, pp. 7192.
Blisniuk, K., Rockwell, T., Owen, L.A., et al., 2010. Late Quaternary slip rates
gradient defined using high-resolution topography and 10Be dating of offset
landforms on the southern San Jacinto Fault zone, California. Journal of
Geophysical Research 115, B08401. http://dx.doi.org/10.1029/2009JB006346.
Bucknam, R.C., Anderson, R.E., 1979. Estimation of fault-scarp ages from scarpheightslope-angle relationship. Geology 7, 1114.
Budel, J., 1982. Climatic Geomorphology. Princeton University Press, Princeton,
444 pp.
Bull, W.B., 1977. The alluvial fan environment. Progress in Physical Geography 1,
222270.
Bull, W.B., 1991. Geomorphic Response to Climate Change. Oxford University
Press, New York, 352 pp.
Bull, W.B., 2007. Tectonic Geomorphology of Mountains: A New Approach to
Paleoseismology. Blackwell, Oxford, 328 pp.
Bull, W.B., 2009. Tectonically active landscapes. Wiley-Blackwell, Chichester,

320 pp.
Bull, W.B., Cooper, A.F., 1986. Uplifted marine terraces along the Alpine fault,
New Zealand. Science 234, 12251228.
Bull, W.B., McFadden, L.D., 1977. Tectonic geomorphology north and south of the
Garlock Fault, California. In: Dochring, D.O. (Ed.), Geomorphology in Arid
Regions. George Allen & Unwin, London, pp. 115138.
Bull, W.B., Wallace, R.E., 1985. Penrose Conference report: tectonic geomorphology.
Geology 13, 216.
Burbank, D.W., Anderson, R.S., 2001. Tectonic Geomorphology. Blackwell Science,
Oxford, 274 pp.
Burbank D.W., Anderson, R.S., 2012. Tectonic Geomorphology. Wiley-Blackwell,
Chichester, (2nd edition), 454 pp.
Burt, T.P., Chorley, R.J., Brunsden, D., Cox, N.J., Goudie, A.S. (Eds.), 2008.
Volume 4: Quaternary and Recent Processes and Forms (18901965) and
the Mid-Century Revolutions. The Geological Society of London, London,
1056 pp.
Chappell, J., 1974. Geology of Coral Terraces, Huon Peninsula, New Guinea:
A study of quaternary tectonic movements and sea-level changes. Geological
Clift, P.D., Tada, R., Zheng, H. (Eds.), 2010. Monsoon Evolution and TectonicClimate Linkage in Asia. Special Publication of the Geological Society, London,
342, 312 pp.
Dansgaard, W., Johnsen, S.J., Clausen, H.B., et al., 1993. Evidence for general
instability of past climate from a 250-kyr ice-core record. Nature 364, 218220.
Dasch, E.J. (Ed.), 1996. Macmillan Encyclopedia of Earth Sciences. Macmillan,
New York, 1273 and 800 pp.
Davis, T.L., Namson, J.S., 1994. A balanced cross-section of the 1994 Northridge
earthquake, southern California. Nature 372, 167169.
Dawkins, W.B., 1894. The relation existing between the tectonic anticlines and
synclines in the districts of South Wales, Gloucester, and the West of England.
Geological Magazine 459.
Demoulin, A., 1998. Testing the tectonic significance of some parameters of
longitudinal river profiles: the case of the Ardenne (Belgium, NW Europe).
Derbyshire, E., 1973. Climatic Geomorphology. Macmillan, New York, 296 pp.
Derbyshire, E., 1976. Geomorphology and Climate. Wiley, Chichester, 514 pp.
Doornkamp, J.C., 1986. Geomorphological approaches to the study of neotectonics.
Journal of the Geological Society, London 143, 335342.
Frankel, K.L., Brantley, K.S., Dolan, J.F., et al., 2007. Cosmogenic 10Be and 36Cl
Implications for transient strain in the eastern California shear zone. Journal of
Geophysical Research Solid Earth 112, B06407.
Frankel, K.L., Dolan, J.F., Owen, L.A., Ganev, P., Finkel, R.C., 2011. Spatial and
temporal constancy of seismic strain release along an evolving segment of the
Pacific- North America plate boundary. Earth and Planetary Science Letters 304,
565576.
Frankel, K.L., Owen, L.A., 2013. Transform plate margins and strike-slip fault
systems. In: Owen, L.A. (Ed.), Treatise on Geomorphology. Academic Press, San
Diego, vol. 5, pp. 3770.
Garzione, C.N., Quade, J., DeCelles, P.G., English, N.B., 2000. Predicting
paleoelevation of Tibet and the Himalaya from d18O vs. altitude gradients in
meteoric water across the Nepal Himalaya. Earth and Planetary Science Letters
183, 215229.
Ghosh, P., Garzione, C.N., Eiler, J.M., 2006. Rapid uplift of the Altiplano revealed
through 13C18O bonds in paleosol carbonates. Science 311, 511515.
Glen, W., 1982. The Road to Jaramillo: Critical years of the revolution in Earth
sciences. Stanford University Press, Stanford, 480 pp.
Goudie, A.S. (Ed.), 2004. Encyclopedia of Geomorphology. Routledge, London,
1200 pp.
Grapes, R.H., Oldroyd, D., Grigelis, A. (Eds.), 2008. History of geomorphology
and Quaternary geology. Geological Society Special Publication, London, 301,
344 pp.
Hack, J.T., 1973. Stream-profile analysis and stream-gradient index. US Geological
Survey Journal of Research 1, 421429.
Hallam, A., 1973. A revolution in the Earth Sciences: From Continental Drift to
Plate Tectonics. Clarendon Press, Oxford, 138 pp.
Hancock, P.L., Williams, G.D., 1986. Neotectonic. Journal of the Geological Society
143, 325326.
Herman, F., Rhodes, E.J., Braun, J., Heiniger, L., 2010. Uniform erosion rates and
relief amplitude during glacial cycles in the Southern Alps of New Zealand, as
revealed from OSL-thermochronology. Earth and Planetary Science Letters 297,
183189.
Hodges, K.V., Adams, B.A., 2013. The influence of middle and lower crustal flow
on the landscape evolution of orogenic plateaus: insights from the Himalaya and
Tibet. In: Owen, L.A. (Ed.), Treatise on Geomorphology. Academic Press, San
Diego, vol. 5, pp. 350369.
Hough, S.E., Bilham, R.G., 2005. After the Earth Quakes: Elastic Rebound on an
Urban Planet. Oxford University Press, Oxford, 336 pp.
Huff, H., Owen, L.A., 2013. Volcanic landformation and hazards. In: Owen, L.A. (Ed.),
Treatise on Geomorphology. Academic Press, San Diego, vol. 5, pp. 148191.
susceptibility mapping for the 2005 Kashmir Earthquake Region. Geomorphology
101, 631642.
Kamp, U., Owen, L.A., 2013. Polygenetic landscapes. In: Owen, L.A. (Ed.), Treatise
on Geomorphology. Academic Press, San Diego, vol. 5, pp. 148192.
Kamp, U., Owen, L.A., Growley, B.J., Khattak, G., 2010. Back analysis of
landslide susceptibility zonation mapping for the 2005 Kashmir earthquake: an
assessment of the reliability of susceptibility zoning maps. Natural Hazards 54,
125.
Keefer, D.K., 2013. Landslides generated by earthquakes: immediate and long-term
effects. In: Owen, L.A. (Ed.), Treatise on Geomorphology. Academic Press, San
Diego, vol. 5, pp. 250266.
Keller, E.K., 1986. Investigation of active tectonics: use of surficial earth processes.
In: Wallace, R.E. (Ed.), Active Tectonics. National Academy Press, Washington,
D.C., pp. 136147.
Keller, E.A., DeVecchio, D., 2013. Tectonic Geomorphology of Active folding and
development of transverse drainages. In: Owen, L.A. (Ed.), Treatise on
Geomorphology. Academic Press, San Diego, vol. 5, pp. 129147.
Keller, E.A., Pinter, N.P., 1996. Active Tectonics: Earthquakes, uplift and landscape,
1st Edition Pentice Hall, Upper Saddle River, NJ, 338 pp.
Keller, E.A., Pinter, N.P., 2002. Active Tectonics: Earthquakes, uplift and landscape.
Second Ed. Pentice Hall, Upper Saddle River, NJ, 326 pp.
Kondo, H., Owen, L.A., 2013. Paleoseismology. In: Owen, L.A. (Ed.), Treatise on
Geomorphology. Academic Press, San Diego, vol. 5, pp. 267299.
Koons, P.O., Zeitler, P.K., Hallet, B., 2013. Tectonic aneurysms and mountain
building. In: Owen, L.A. (Ed.), Treatise on Geomorphology. Academic Press, San
Diego, vol. 5, pp. 318349.
Kumar, S., Wesnousky, S.G., Jayangondaperumal, R., Nakata, T., Kumahara, Y.,
Singh, V., 2010. Paleoseismological evidence of surface faulting along the
northeastern Himalayan front, India; timing, size, and spatial extent of great
earthquakes. Journal of Geophysical Research 115, B12422. http://dx.doi.org/
10.1029/2009JB006789.
Lapworth, C., Watts, W.W., 1894. The geology of Southern Shropshire: with special
reference to the district to be visited during the long excursion. Proceedings of
the Geologists Association 13, 297355.
Lee, J., Stockli, D.F., Owen, L.A., Finkel, R.C., Kislitsyn, R., 2009. Exhumation of
the Inyo Mountains, California: Implications for the timing of extension along
the western boundary of the Basin Range Province and distribution of dextral
fault slip rates across the Eastern California Shear Zone. Tectonics 28, TC1001.
http://dx.doi.org/10.1029/2008TC002295.
Li, Y., Yang, J., Xia, Z., Mo, D., 1998. Tectonic geomorphology in the Shanxi
Graben System, northern China. Geomorphology 21, 7789.
Marvin, U.B., 1973. Continental Drift: The Evolution of a Concept. Smithsonian
Institution Press, Washington, DC, 256 pp.
Massonnet, D., Rossi, M., Carmona, C., et al., 1993. The displacement field of the
Landers earthquake mapped by radar interferometry. Nature 364, 138142.
McCalpin, J.P., 1982. Quaternary geology and neotectonics of the western flank of
the northern Sangre de Cristo Mountains, south-central Colorado. Colordao
School of Mines Quarterly 77, 97.
Morisawa, M., Hack, J.T. (Eds.), 1985. Tectonic Geomorphology. Proceedings of the
15th Annual Binghamton Geomorphology Symposium, September 1984. Allen
and Unwin, Boston, 390 pp.
Morner, N.-A., 1990. Glacial isostsay and long-term crustal movements in
Fennoscandia with respect to lithopsheric and asthenospheric processes and
properties. Tectonophysics 176, 1324.
Muhs, D., Kelsey, H., Miller, G., Kennedy, G., Whelan, J., McInelly, G., 1990. Age
Estimates and Uplift Rates for Late Pleistocene Marine Terraces Southern
Oregon Portion of the Cascadia Forearc. Journal or Geophysical Research 95,
66856698.
Nash, D., 2013. Tectonic geomorphology of normal fault scarps. In: Owen, L.A.
(Ed.), Treatise on Geomorphology. Academic Press, San Diego, vol. 5, pp.
234249.
National Research Council, 2008. Earth Observations from Space: The First 50
Years of Scientific Achievements. The National Academies Press, Washington
DC, 142 pp.
11
National Research Council, 2010. Landscapes on the Edge: New horizons for
Research on Earths Surface. The National Academies Press, Washington DC,
180 pp.
Obruchev, V.A., 1948. Osnovnye cherty kinetiki i plastiki neotektonik. Izv. Akad.
Nauk. Ser. Geol. 5, 1324.
Owen, L.A., 2010. Landscape development of the HimalayanTibetan orogen: a
review. Special Publication of the Geological Society of London 338, 389407.
Owen, L.A. (Ed.), 2013. Treatise on Geomorphology. Academic Press, San Diego,
vol. 5, pp. 1393.
Owen, L.A., Kamp, U., Khattak, G.A., Harp, E.L., Keefer, D.K., Bauer, M.A., 2008.
Landslides triggered by the October 8, 2005, Kashmir Earthquake.
Owen, L.A., Stewart, I., Vita-Finzi, C. (Eds.), 1993. Neotectonics-recent advances.
Quaternary Proceedings. City Cartographic & Desktop Publishing Unit, London
Guild-hall University, Lodon, 3, 122 pp.
Quade, J., Garzione, C., Eiler, J., 2007. Paleoelevation reconstruction using
pedogenic carbonates. Reviews in Mineralogy and Geochemistry 66, 5387.
Raymo, M.E., Ruddiman, W.F., 1992. Tectonic forcing and late Cenozoic climate.
Nature 359, 117122.
Rockwell, T., Lindvall, S., Herzberg, M., Murbach, D., Dawson, T., Berger, G., 2000.
Paleoseismology of the Johnson Valley, Kickapoo and Homestead Valley Faults:
clustering of Earthquakes in the Eastern California Shear Zone. Bulletin of the
Seismological Society of America 90, 12001236.
Ruddiman, W.F. (Ed.), 1997. Tectonic uplift and climate change. Plenum Press,
New York, 535 pp.
Sahagian, D., Proussevitch, A., 2007. Paleoelevation measurement on the basis of
vesicular basalts. Reviews in Mineralogy and Geochemistry 66, 195213.
Sahagian, D., Proussevitch, A., Carlson, W., 2002. Timing of Colorado Plateau
uplift: initial constraints from vesicular basalt-derived paleoelevations. Geology
30, 807810.
Schoenbohm, L.A., 2013. Continentalcontinental collision zone. In: Owen, L.A.
(Ed.), Treatise on Geomorphology. Academic Press, San Diego, vol. 5, pp.
1336.
Schwarzbach, M., 1986. Alfred Wegener: The Father of Continental Drift, translated
from the German. Science Tech., Madison, WI, 241 pp.
Seller, R.C., Cocks, L.R.M., Plimer, I.R. (Eds.), 2005. Encyclopedia of Geology
Elsevier, Oxford, (five volumes), 2750 pp.
Shroder, J., 2013. Treatise of Geomorphology 12 volumes. Academic Press: San
Diego.
Sieh, K., 1978. Prehistoric large earthquakes produced by slip on the San Andreas
Fault at Pallett Creek, California. Journal of Geophysical Research 83,
39073939.
Sieh, K.E., Jahns, R.H., 1984. Holocene activity of the San Andreas fault at Wallace
Creek, California. Geological Society of America, Bulletin 95, 883896.
Spotila, J.A., 2013. Glacially influenced tectonic geomorphology: the impact of the
glacial buzzsaw on topography and orogenic systems. In: Owen, L.A. (Ed.),
Treatise on Geomorphology. Academic Press, San Diego, vol. 5,
pp. 300317.
Stewart, I., Owen, L.A., Vita-Finzi, C. (Eds.), 1993. Neotectonics and Active Faulting.
Zeitschrift fur Geomorphologie, Suppl. Bd, 94, 328 pp.
Stewart, I.S., 1993. Neotectonics: current themes and future prospects. Conference
Report. Journal of the Geological Society 150, 785787.
Stewart, J.A., 1990. Drifting Continents and Colliding Paradigms: Perspectives on
the Geoscience Revolution. Indiana University Press, Bloomington, 304 pp.
Tsukamoto, S., Duller, G., Murray, A., Choi, J.-H., 2009. Luminescence: analysis of
Quaternary tectonic movements and environmental change. Quaternary
Villeneuve, N., Bachelery, P., 2013. Hot spots and Large Igneous Provinces.
In: Owen, L.A. (Ed.), Treatise on Geomorphology. Academic Press: SanDiego,
vol. 5, pp. 193233.
Vine, F.J., Matthews, D.H., 1963. Magnetic anomalies over oceanic ridges. Nature
4897, 947949.
Vita-Finzi, C., 1986. Recent Earth Movements: An Introduction to Neotectonics.
Academic Press, London, 226 pp.
Wegener, A., 1915 Die Entstehung der Kontinente und Ozeane Bravnschweig,
Druckund verlag von Friedr. Vieweg & Sohn.
Wallace, R.E., 1975. In: Crowell, J.C. (Ed.), San Andreas Fault in Southern
California. California Division of Mines and Geology, Special Report, 118,
243 pp.
Wallace, R.E., 1977. Profiles and ages of young fault scarps, north-central Nevada.
Wallace, R.E. (Ed.), 1986a. In: (Ed: anonymous) Active Tectonics. National Academy
Press, Washington, DC, 148 pp.
12
Wallace, R.E., 1986b. Overview and recommendations. In: Wallace, R.E. (Ed.), Active
Tectonics. National Academy Press, Washington, DC, pp. 319.
Wells, S.G., Bullard, T.F., Menges, C.M., et al., 1988. Regional variations in tectonic
geomorphology along a segmented convergent plate boundary pacific coast of
Costa Rica. Geomorphology 1, 239265.
Yeats, R., 2002. Citation by Robert Yeats: Structural Geology and Tectonics Career
Contribution Award presented to Robert E. Wallace. http://rock.geosociety.org/
sgt/2002_Career_Award_Wallace.pdf
Yeats, R.S., Sieh, K., Allen, C.R., 1997. The Geology of Earthquakes. Oxford
University Press, Oxford, 568 pp.
Zebker, H.A., Amelung, F., Jonsson., S., 2000. Remote sensing of volcano surface and
internal processes using radar interferometry. In: Mouginis-Mark, P.J., Crisp, J.A.,
Fink, J.H. (Eds.), Remote Sensing of Active Volcanism. AGU Monograph No. 116.
American Geophysical Union, Washington, DC, pp. 179205.
Biographical Sketch
Prof. Lewis Owen received his PhD in Geomorphology from the University of Leicester, U.K., in 1988. Before
joining the University of Cincinnati in 2004, he held positions at the Hong Kong Baptist University, Royal
Holloway University of London, and the University of California Riverside. Prof. Owens research focuses on
the Quaternary geology and geomorphology of tectonically active mountain belts and their forelands, particularly
in the Himalayan-Tibetan orogen and the Cordilleras of North and South America. He has also undertaken
research in other tectonically active regions, including the Red Sea margin in Yemen and the Atlas and Anti-Atlas
Mountains of Morocco. In 2011, Prof. Owen was awarded the Busk Medal from the Royal Geographic Society for
his field research in Quaternary history and geomorphology in tectonically active areas.
5.3 ContinentalContinental Collision Zone

LM Schoenbohm, University of Toronto Mississauga, ON, Canada
5.3.1
5.3.2
5.3.2.1
5.3.2.2
5.3.2.3
5.3.3
5.3.3.1
5.3.3.1.1
5.3.3.1.2
5.3.3.2
5.3.3.2.1
5.3.3.2.2
5.3.3.2.3
5.3.4
5.3.4.1
5.3.4.2
5.3.5
5.3.5.1
5.3.5.1.1
5.3.5.1.2
5.3.5.1.3
5.3.5.2
5.3.5.2.1
5.3.5.2.2
5.3.5.2.3
5.3.5.2.4
5.3.6
5.3.7
References
Introduction
Southern Alps of New Zealand
Overview of Geology and Geomorphology
Steady State
Orography, Subduction Polarity, and Range Symmetry
AfricaEurope Collision
Pyrenees
Overview of geology and geomorphology
Origin and significance of low-relief surfaces
Alps
Cause of recent uplift of the Alps
Range width and climate change
ArabiaEurasia Collision
Mantle Lithosphere Delamination, Uplift, and Drainage Reorganization
IndiaAsia Collision
Himalaya
Monsoon precipitation and climatetectonic coupling
Channel flow
Syntaxes tectonic aneurysms
Tibet
Paleoaltimetry
Lower crustal flow
Drainage reorganization, uplift, and crustal shear
Landslides and glacial dams
Ancient Orogens
Conclusion
Glossary
Anatexis The differential, or partial melting of rocks.
Anatectic granites are typically associated with local, in situ
partial decompression melting of rock triggered by rapid,
tectonic unroofing.
Coulomb critical wedge A geometric conception of a
mountain range as a one- or a two-sided wedge, in which
the geometry (slope of the upper topographic surface and of
the lower detachment surface) depends on the influx of
accreting material, the efflux of material through erosion,
the frictional resistance along the base, and the compressive
strength of the wedge material.
ELA Equilibrium line altitude, or the average elevation of
the balance point on a glacier between net ablation and net
accumulation, averaged over time.
14
17
17
17
19
19
19
19
20
21
21
22
23
23
23
24
26
26
26
26
27
28
28
28
29
30
30
31
31
Erosion surface A relatively low-relief geomorphic surface

produced through erosion. Such surfaces can form at low
elevation (generally through fluvial processes) or high
elevation (generally through glacial, periglacial, or hillslope
processes).
GPS Global Positioning System, a network of
satellites that allow the precise measurement of location
on Earth. Continuously recording GPS stations or
repeated measurements of the same monument can
determine the elastic and inelastic deformation of tectonic
plates.
LGM Last Glacial Maximum, the time of the maximum
advance of ice sheets and mountain glaciers during the last
glacial period.
Schoenbohm, L.M., 2013. Continentalcontinental collision zone. In:

Shroder, J. (Editor in Chief), Owen, L.A. (Ed.), Treatise on Geomorphology.
Academic Press, San Diego, CA, vol. 5, Tectonic Geomorphology, pp.
1336.
http://dx.doi.org/10.1016/B978-0-12-374739-6.00081-6
13
14
ContinentalContinental Collision Zone
Moho Short term for the Mohorovicic discontinuity,

which is the boundary between the Earths crust and the
underlying mantle lithosphere.
Orogeny The result of the collision of tectonic plates,
which leads to deformation of the Earths crust and the
production of mountain ranges.
Orographic precipitation Adiabatic cooling,
condensation, and precipitation caused when air masses are
forced over topographic features.
Peneplain A low-relief surface (an erosion surface) that is
interpreted to represent the final stage of fluvial erosion
after a long period of tectonic quiescence. The original
concept was developed by geomorphologists William Davis
and Walther Penck in the early 1900s.
Relict landscape A relatively low-relief geomorphic
surface graded to a different, transient base-level compared
with the rest of the landscape. In contrast to the term
erosion surface, a relict landscape implies that a variety of
geomorphic processes may have been active on the
landscape (deposition as well as erosion), the original relief
structure may be complex, and that the age of formation
may vary across the landscape.
Saprolite Highly weathered, thoroughly decomposed,
clay-rich rock formed by in situ chemical weathering.
Steady state A condition and which various aspects of a
mountain range do not change with time, generally as a
result of constructive (tectonic) and destructive (erosional)
forces. Types of steady states include: (1) flux steady state, in
which the amount of material accreted to a range during
continental collision is exactly balanced by the amount
removed by erosion; (2) topographic (or geomorphic)

steady state, in which the general form of topography (e.g.,
the mean or the maximum elevation of the orogen or the
topographic wavelength) is constant with time; (3) thermal
steady state, in which the temperature field of an orogen is
time invariant; and (4) exhumational steady state, in which
the spatial pattern of exhumation is constant with time and
therefore the pattern of reset thermochronometer ages is
constant and nested, or organized according to closure
temperature, with the highest closure temperature mineral
showing an age minima in the core of the range, with
increasingly wider zones of decreasing closure-temperature
minerals around it.
Syntaxis A syntaxis is a sharp bend in an orogenic belt,
accompanied by a convergence of mountain belts toward
the bend. Syntaxes typically occur at the corners of
indentors, such as at the western and eastern ends of the
Indian plate, where it indents Asia.
Thermochronology The study of the thermal evolution of
a region in order to interpret its tectonic and erosional
history, based on the use of the specific closure temperature
of different isotopic systems in different minerals. Common
low-temperature thermochronometers include (UTh)/He
dating in apatite and zircon, with closure temperatures of
approximately 70 and 180 1C, respectively, fission-track
dating in apatite and zircon, with closure temperatures of
115 and 230 1C, respectively, and 39Ar/40Ar in biotite and
muscovite, with closure temperatures of 350 and 380 1C,
respectively.
Abstract
The major zone of continentalcontinental collision on Earth today is along the AlpineHimalayan Chain, which is home
to the Pyrenees, European Alps, Turkish-Iranian Plateau, the Himalaya, and the Tibetan Plateau. Continentalcontinental
collision also occurs in the Southern Alps of New Zealand and has produced a number of ancient orogens, including the
Appalachians, Urals, and Caledonides. The extensive, long-lived deformation, and particularly the crustal thickening, and
therefore weakening, associated with continental collisions results in unique tectonic geomorphology. An important topic
in understanding the tectonic geomorphology of continentalcontinental collision zones is whether or not flux,
topographic, thermal, or exhumational steady state can be achieved and how each can be recognized. Another subject of
study is the relative controls of tectonics and climate on range width and asymmetry. Tectonic geomorphology has also
been used to identify the lateral transport of crust, either brittlely, along strikeslip faults, or ductilely, through lower crustal
flow. Drainage reorganization is a particularly powerful tool for inferring uplift patterns. Geomorphologic studies also yield
insights into lithospheric foundering. Finally, tectonic geomorphology can help explain the persistence of topography over
long-innactive orogens. In this contribution these topics are explored on an orogen-by-orogen basis.
5.3.1
Introduction
Of the many plate tectonic settings, continentalcontinental

collision inarguably produces the most extensive crustal
deformation, the most spectacular mountain uplift, and has
the most profound influence on climate dynamics on a
planetary scale. Such collisions begin with the subduction of
oceanic crust, which causes the development of a volcanic
arc on the overriding plate. As subduction proceeds, smaller
terranes may be accreted. The process ends as the opposing
continent finally reaches the subduction zone. Largely
unable to subduct because of their relatively low density, the

continents instead collide. The properties of each piece of
the collision control how the orogeny develops; in some
cases, these properties may be similar, leading to more
symmetric deformation, whereas in other cases the overriding plate has been softened by arc magmatism and deformation during terrane accretion, leading to a highly
asymmetric distribution of deformation. Also considered is
the case of intracontinental collision, in which two continents or subcontinents converge to close a relatively small
ocean or continental extensional basin.
15
ns
hia
c
la
pa
Ap
Ural
s
Norwegian
egian Caledonides
aledonides
Alps
lps
Pyrenees
enees
Tibet
et
Hima
layas
Iranian
anian Plateau
Pl eau
Anatolian
olian Plateau
Pl eau
Southern
outhern Alps
lps
Figure 1 World topographic map on SRTM3 base with GTOPO30 north of 601 N and south of 601 S showing the location of the major
mountain ranges of the AlpineHimalayan Chain (Pyrenees, Alps, Anatolian Plateau, Iranian Plateau, Himalayas, and Tibet), and other active
(Southern Alps) and ancient (Appalachians, Urals, and Norwegian Caledonides) continental-continental collisions.
The major zone of continentalcontinental collision occurring on Earth today is in the AlpineHimalayan chain,
stretching from the Mediterranean to the South China Sea
(Figure 1). The collision between a variety of microplates
trapped between the converging African and European plates
has produced, among others, the Alps and the intracontinental Pyrenees and Carpathian Mountains. Further east along
the chain, the collision of the Arabian Peninsula and the
Iranian Plateau with Asia has caused the formation of distinct
mountain ranges, such as the Zagros fold-thrust belt and the
Alborz Mountains, and has led to broad deformation and
uplift of the TurkishIranian Plateau. Finally, the impressive
convergence between India and Asia is responsible for forming
the highest mountain range on Earth, the Himalaya, and for
significantly contributing to the uplift of an area greater than
2.5 million km2, the Tibetan Plateau. Continental collision
also occurs, albeit on a smaller scale, within the Southern Alps
of New Zealand (Figure 1). The planet is also scarred by ancient continental collisions, including the Variscan and Caledonian orogens in Europe, the related Taconic, Acadian, and
Alleghenian orogens in North America and the Urals of
western Russia and Kazakhstan in the Paleozoic, the Grenville
orogen in North America in the Mesoproterozoic, and the
Neoproterozoic Pan-African orogeny. Many of the Paleozoic
mountain ranges are still topographically high today.
Given the extreme disruption of the landscape and the
profound changes that occur within the crust as a collision
develops, it is not surprising that the tectonic geomorphology
of continentalcontinental collisions is a rich and deep field
of study. Although several authors have made advances in
understanding the specifics of uplift, shortening, deformation,
landscape evolution, and climatetectonic interactions within
the various mountain ranges outlined above, these sorts
of studies are well covered by other chapters in this volume
(e.g., Chapters 5.13 and 5.14). It is also beyond the scope of

this contribution to address the topic of continental
collision and global climate change. Although mountain
building of any type can interact with global climate,
continentalcontinental collisions tend to both alter atmospheric circulation patterns by producing the highest and most
widespread topography and change ocean circulation patterns
by closing seaways. For example, uplift of Tibet and the collision of Arabia with Eurasia are implicated in both long-term
climatic deterioration and cooling spikes in the Cenozoic
(Raymo et al., 1988; Allen and Armstrong, 2008). Plateau
uplift associated with continental collision also appears to
have strengthened the Indian Summer Monsoon (e.g., Molnar
et al., 1993; Clift et al., 2008). However, studies associated
with this topic are generally oceanographic, geochemical, or
geodynamic.
In this contribution, the focus is instead on larger questions addressed by tectonic geomorphology directly that are
unique to continentalcontinental collision. These questions
include:
1. Steady state: Whether or not a mountain range ever
achieves a constant form, in terms of the pattern of exhumation, the deeper structure of the range, or the topography, is a topic of much debate. Although the question of
steady-state achievement is not limited to continental collisions, much of the active research does focus on two
prominent examples, the European Alps and the Southern
Alps of New Zealand, and thus is included here. The different types of steady-state behavior (Figure 2) include:
(1) flux steady state, in which the amount of material accreted to a range during continental collision is exactly
balanced by the amount removed by erosion; (2) topographic (or geomorphic) steady state, in which the general
16
Topographic steady state

Divide position
6
Topographic Range
mean elevation
4
2
0
(a) km
Flux steady state: FA = FE1 + FE2

FE1
FE2
Pro-
Retro-
FA
Eroded material
Fixed
Plate
Deforming orogen
(b)
Undeforming plate
Thermal steady state: T= 0
Isotherms
(c)
Exhumational steady state

Pre-orogenic age
ab
c d
Age
Reset age
Nested
ages
Distance
Ta
Tb
Tc
Td
Closure temperatures
Particle paths
(d)
Figure 2 Different forms of steady state (Willet and Brandon (2002)).

(a) Topographic steady state, in which topographic changes occur, but
remain within a narrow range over time. (b) Flux steady state, in which
material accreted to the orogen (FA) equals material eroded from the
orogen (FE). (c) Thermal steady state, in which the temperature within
the orogen remains constant with time. (d) Exhumational steady state,
in which the temperature and particle paths are constant with time,
resulting in a nested pattern of thermochronometers within the range.
form of topography (e.g., the mean or the maximum elevation of the orogen or the topographic wavelength) is
constant with time; (3) thermal steady state, in which the
temperature field of an orogen is time invariant; and
(4) exhumational steady state, in which the spatial pattern

of exhumation is constant with time and therefore the
pattern of reset thermochronometer ages is constant and
nested, or organized according to closure temperature, with
the highest closure temperature mineral showing an age
minima in the core of the range, with increasingly wider
zones of decreasing closure-temperature minerals around it
(Willet and Brandon, 2002). Common low-temperature
thermochronometers include (UTh)/He dating in apatite
and zircon, with closure temperatures of approximately
70 and 180 1C, respectively, fission-track dating in apatite
and zircon, with closure temperatures of 115 and 230 1C,
respectively, and 39Ar/40Ar in biotite and muscovite, with
closure temperatures of 350 and 380 1C, respectively (see
Reiners and Brandon, 2006, for an overview of the application of thermochronology to orogenesis).
2. Range width and symmetry: a related topic is how convergence geometry, rheology of the colliding crust, and
climate affect both the width of a mountain range and its
symmetry. According to Coulomb critical wedge models,
range width and taper (slope) are controlled by, among
other factors, accretionary influx and erosional efflux (e.g.,
Dahlen, 1990; Whipple and Meade, 2004; Roe et al.,
2006). Erosion acts to narrow and steepen a range, whereas
increased flux (either in the form of an increased convergence rate or increased off-scraping of material and
incorporation into the wedge) acts to widen a range.
Processes in the European Alps are discussed, where
evidence from the pattern of deformation, erosion, and
deposition has been used to argue for (and against) climate-driven changes in orogen width. The geometry of the
collision zone most importantly, which is the overriding
and which is the subducting plate control range geometry
as well, for example, in the Pyrenees and in the Southern
Alps. Rheology can also play a role, for example, in the
IndiaAsia collision, in which India acts as a rigid indentor
into the relatively hot, weak Tibetan crust. Finally, orographic precipitation may play a role in shaping range
asymmetry.
3. Orogen-parallel transport of crust: tectonic geomorphology
is an invaluable tool in evaluating the lateral transport of
crust, bridging the gap between estimates of fault slip rates
from short-term geodetic data and longer term geologic
data, for example, in the TurkishIranian and Tibetan
Plateaus. It is also uniquely positioned to address questions
of lower crustal flow, which is thought to produce broad
uplift and crustal thickening in the absence of evidence for
deformation at the surface. Tectonic geomorphology has
been used, for example, to infer lower crustal flow on both
the southeast and the northeast margins of the Tibetan
Plateau. The related phenomenon of channel flow, in
which the extrusion of midcrustal material in the Himalaya
is coupled to intense erosion at the surface, is addressed in
other chapters in this volume (Chapters 5.13, 5.14 and
5.15), and is therefore not discussed in detail here.
4. Lithospheric foundering: awareness is growing on the importance of lithospheric removal events, either through
delamination, in which the mantle lithosphere peels away
from the overlying, thickened, crust, or through convective
removal, in which mantle lithosphere and potentially
lower crusts are removed. Although geophysical data can

provide a snapshot of the state of the lithosphere and
upper mantle today, tectonic geomorphologic studies can
provide an insight into the nature and timing of these removal events, as the associated uplift and extensional
faulting cause drainage reorganization and incision.
Lithospheric foundering is thought to play a prominent
role in the development of the TurkishIranian plateau as
well as the Alps, Pyrenees, and Tibet.
5. Persistence of topography: inactive orogens can maintain
significant height tens to even hundreds of millions of
years after convergence ceases. Tectonic geomorphology
has yielded an insight into the preservations and even resurgence of topography in ranges such as the Pyrenees, the
Appalachians, and the Norwegian Caledonides.
In this contribution, focus is on these five topics by exploring
past and present continentalcontinental collisions. This discussion begins with the Southern Alps of New Zealand, in the
earliest stages of collision, then works along the Alpine
Himalayan chain from west to east, from the smallest ranges
in Europe and North Africa, to the largest collision zone on Earth
in the Himalaya and Tibet. Finally, there is a brief discussion of
older orogens.
5.3.2
5.3.2.1
Southern Alps of New Zealand

The Southern Alps of New Zealand (Figure 3) are the product

of the oblique collision between continental fragments imbedded along the edges of larger plates: the Chatham Rise
and Campbell Plateau in the Pacific plate to the east and
the Challenger Plateau in the Australian plate to the west
(Walcott, 1978). Uplift began around 9.8 Ma when
PacificAustralian plate motion changed (Stock and Molnar,
1982), and the present tectonic regime was established
by 5 Ma (Sutherland, 1996; Walcott, 1998; Chamberlain
et al., 1995; Batt and Braun, 1999; Batt et al., 2000). Since
that time, oblique convergence has been partitioned
into strikeslip displacement of B38 mm yr1, mostly accommodated along the Alpine fault, and shortening of
B13 mm yr1, accommodated both on the Alpine fault and
distributed through the range (Norris et al., 1990; Norris and
Cooper, 1997). The eastern side of the Southern Alps is
dominated by the Permo-Triassic Torlesse graywacke, whereas
the metamorphic equivalent, the Alpine schist, crops out in a
narrow belt adjacent to the Alpine fault, and reaches amphibolite facies (Figure 3(a) and Tippett and Kamp, 1993).
The Otago schist, for which the Torlesse graywacke is also a
protolith, outcrops to the south of the range (Figure 3(a) and
Tippett and Kamp, 1993).
The range is highly asymmetric (Figure 3(b)), with the
steepest slopes and most rapid exhumation, up to B10 mm
yr1, occurring on the western, upwind side of the range
(Tippett and Kamp, 1993; Little et al., 2005). Precipitation on
the western flank (Figure 3(b)), produced by orographic forcing of moist air masses flowing eastward from the Tasman
Sea, is as much as 5 m yr1 (Griffiths and McSaveney, 1983;
17
Willet, 1999). The drier, downwind, eastern side receives less

than 2 m rainfall per year (Griffiths and McSaveney, 1983;
Willet, 1999). This is reflected in the steep, deeply incised
high-gradient rivers and abundant evidence for mass wasting
on the wet, west side of the range compared with the subdued
relief and longer, lower gradient rivers on the east (Whitehouse, 1987; Willet, 1999). Although fluvial erosion and
landsliding are dominant erosional forces in the Southern
Alps, most of the western side of the orogen is glacially influenced as well (Herman and Braun, 2006), and the range
was covered by an icecap during the Last Glacial Maximum
(Denton and Hughes, 1980). The strong precipitation gradient
is also reflected in lower modern and paleo-equilbirum line
altitudes (ELAs) on the western side of the range as compared
with the east (Soons, 1979). Finally, erosion rates vary markedly from west to east, being 3100 times higher on the west
side of the range (Adams, 1980).
5.3.2.2
Steady State
Feedback between climate and tectonics is argued to drive

mountain ranges toward a steady-state condition, in which, if
the external conditions remain the same, the range maintains
a constant topographic, dynamic, and thermal form (Figure 2;
Willet and Brandon, 2002). However, how long it takes to
achieve steady state under specific climatic and tectonic conditions, how quickly a system responds to changes, and
whether or not a true steady state is even possible, given the
continually fluctuating Quaternary climate, remain controversial. This vigorous debate has been focused on the Southern
Alps of New Zealand.
Whether or not a range is in a thermal or an exhumational
steady state (Figure 2) is particularly important for interpreting thermochronometric cooling ages in terms of the
exhumation rate. Although thermokinematic models can account for the dynamic advection of heat in an orogen (e.g.,
Batt and Braun, 1999; Herman et al., 2007), many studies
make the simplifying assumption of steady-state conditions in
calculating the exhumation rate from cooling ages. The thermochronometric record in the Southern Alps indicates that
the modern tectonic regime was established only 5 million
years ago (Sutherland, 1996; Walcott, 1998; Chamberlain
et al., 1995; Batt and Braun, 1999; Batt et al., 2000). Batt
(2011) suggests that the pattern of thermochronometric age
reflects the transition to a steady state: it has been achieved in
the central Southern Alps, where ages are consistently less than
2 Ma, but the range is still in transition to the south, marked
by a sharp transition in KAr mica ages at B5 Ma near the Fox
Glacier. Willet and Brandon (2002) cited the nested nature of
different thermochronometers (Figure 2) as an indication that
an exhumational steady state has been achieved in the
Southern Alps: reset apatite fission-track samples are restricted
to within 40 km to the east of the Alpine fault, zircon fission
track samples to within 1520 km, and muscovite KAr
samples to less than 10 km (Figure 3). This pattern is a primary prediction of the model of an exhumational steady state,
with the asymmetry reflecting the polarity of accretion, which,
in this case, is from the east (Willet and Brandon, 2002).
18
1000 km
Challenger
Plateau
Norrth Island
No
North
m/y
38 m
Australian
plate
Chatham Rise
South Island
South
-2000 m
Pacific
plate
b
Campbell
Plateau
A
Fox Glacier
ult
Fa
e
in
ult
Alp
Hope Fa
Alpine
Schist
gu
rT
re
n
ch
Torlesse
Greywacke
Metamorphic Grade
Otago Schist
Pu
ys
e
h
nc
Tre
i
g
ran
iku
Garnet-Oligoclase
Biotite
Chlorite
200 km
Reset Ages
Zircon Fission Track
Apatite Fission Track
Wind
Depth (km)
basin
10
20
30
Main Divide
basin
A 0
pa
rti
cle
mylonite
zone
40 Australian
Plate
50
greywacke
pa
th
s
lower crust
schist
amphibolite
13 mm/yr
3
0
Elevation (km)
Alpine Fault
nt
decolleme
Pacific Plate
Figure 3 Southern Alps of New Zealand (Beaumont et al., 1996; Little et al., 2005). (a) Overview map, showing oceanic plateaus (shown by
area below 2000 m in depth in gray), tectonic plate configuration, and rate of convergence. (b) Simplified geologic map indicating metamorphic
rocks exhumed in the Cenozoic. The Southern Alps lie directly east of the Alpine fault. (c) Simplified cross section through the crust along
section line AA0 shown in (b), indicating the convergence rate normal to the Alpine Fault and the location of orographic precipitation. Nested
zones of reset zircon and apatite fission track thermochronometers are indicated above the figure.
Although the Southern Alps may be at or transitioning to

an exhumational steady state, a number of authors argue that
a topographic steady state (Figure 2) may be impossible to
achieve. Modeling and geomorphic analysis suggests that although the location of the divide is relatively steady, features
may be tectonically advected from the eastern side of the range
to the west, disrupting the topographic form of the landscape
(Willet et al., 2001; Herman and Braun, 2006). Further,
modeling indicates that significant changes to the topography

occur during glacial and interglacial periods, and thus the
landscape can never achieve a steady state (Whipple, 2001;
Herman and Braun, 2006). Tomkin (2003) confirms that although a topographic steady state may be achieved in fluvially
dominated orogens, glacial erosion, and in particular the
variability of ice thickness, produce chaotic behavior that
resists the establishment of steady-state conditions. A flux
steady state (Figure 2) is likely to have been achieved in the

Southern Alps, given the high rates of erosion that correlate
with exhumational patterns (Adams, 1980).
5.3.2.3
Orography, Subduction Polarity, and Range

Symmetry
A related topic is the origin of the extreme asymmetry of the

Southern Alps, in terms of topography, exhumation, and
precipitation (Figure 3(b)). Koons (1990) and Koons
et al. (2003) emphasized the role of erosion on the windward
side of the range in concentrating deformation along the Alpine fault. This fault accommodates both transcurrent and
compressional displacement; concentrated erosion and associated thermal weakening is key to allowing these opposing
strains to be concentrated on a single structure (Koons et al.,
2003). Other models explore the coevolution of topography
and orographic precipitation. Batt and Braun (1999) argue
that the asymmetry of the range is a result of higher orographic rainfall on the western flanks. Further, they found that
there is a lag between the onset of deformation and the development of topography high enough to produce a significant orographic effect, and link this to a B2 Myr delay
between the onset of rock uplift and surface uplift in the
Southern Alps east of the divide (Batt and Braun, 1999). Finally, recent work suggests that erosion could play an important role in the geodynamic evolution of the mantle
lithosphere as well; efficient removal of crust through surface
erosion promotes a stable, subduction-like geometry for the
mantle lithosphere, whereas the absence of erosion inhibits
subduction, promoting slab steepening and detachment
(Pysklywec, 2006). Interestingly, these observations suggest a
paradox: the windward side of a range should undergo greater
erosion, valley lowering, and headward channel incision,
which in turn would lengthen and lower the slope over time
(Bonnet, 2009). In the Southern Alps, in fact, the opposite
symmetry is observed, with the long, gentle slope on the leeward side of the range (Willet et al., 2001). Thus, in the
Southern Alps, although erosion may influence the shape of
the orogen by focusing deformation, it is not sufficient to
override tectonic forces and push the divide toward the dry
side of the orogen.
5.3.3
AfricaEurope Collision
The Mediterranean region provides an excellent example of

just how complex a continental collision can be. Africa and
Europe are converging at about 5 mm yr1 in the west and up
7 mm yr1 in the east (Nocquet and Calais, 2004), slowly
wiping out the last remnants of the Tethyan Ocean. However,
the intervening region is composed of a number of microplates, with local subduction and back-arc extension (Dewey
et al., 1973). The Apennines in Italy, for example, are the result
of shortening in the overriding plate as the largely oceanic
crust of the Ionian Sea subducts beneath it from the east
(Robertson and Grasso, 1995; Gueguen et al., 1998). Trench
roll-back pulls the overriding plate along with it, causing backarc extension in the Tyrrhenian Sea to the west. Only one part
19
of this collision zone could be considered continentalcontinental collision, in its earlier stages: subduction of the Apulian
Platform in the southern Apennines is choking off the subduction zone there (Gueguen et al., 1998). Similarly complex
relationships characterize the Betics, Maghrebides, Dinarides,
Hellenides, Taurides, and Carpathians. Significant continental
collision characterizes only two ranges within the Africa
Europe collision zone today: the European Alps and the
Pyrenees. Each is discussed in turn, focusing on specific tectonic geomorphic observations and controversies.
5.3.3.1
5.3.3.1.1
Pyrenees
The Pyrenees Mountains (Figure 4), straddling the border

between France and Spain, are the product of Late Cretaceous
to early Miocene intracontinental convergence between the
Iberian and the European plates (Roest and Srivastava, 1991),
and accommodated at most 165 km of upper crustal shortening (Beaumont et al., 2000). Present-day convergence across
the range is negligible. The doubly vergent orogen is cored by
Hercynian basement and is flanked by fold-thrust belts in
Mesozoic and syn-deformational Tertiary rocks (Sinclair et al.,
2005). Although exhumation from the late-Paleoecene to the
mid-Eocene is focused in the core of the range, significant
asymmetry developed in the Oligocene and early Miocene as a
result of underplating and rapid erosion of the pro-wedge,
antiformal stack (Fitzgerald et al., 1999; Sinclair et al., 2005).
Apatite fission-track ages are as young as 2036 Ma on the
southern, pro-wedge side of the orogen, where Iberia subducted beneath Europe (Sinclair et al., 2005). As expected,
shortening is much greater in the southern, pro-wedge Ebro
basin, at B128 km, than in the northern, retro-wedge
Aquitaine basin, where it is only B37 km (Beaumont et al.,
2000; Naylor and Sinclair, 2008). The asymmetry of shortening and exhumation within the range can be adequately
explained by the geometry of subduction and preexisting weak
detachments within the southern part of the wedge (Fitzgerald
et al., 1999; Beaumont et al., 2000). Asymmetry is thus a result
of the geometry and rate of movement of the converging
plates; orographic precipitation, heavier on the northern side
of the range, does not appear to have played a role in either
shaping the exhumation pattern or rejuvenating uplift on the
pro-wedge side of the range in the Oligocene or Miocene.
The central Pyrenees reach an average elevation of about
2000 m above sea level (asl), with peaks up to 3404 m high.
The southern flank is slightly lower in slope than the northern
flank, reflecting the larger area of the pro-wedge side of the
range and the fold-thrust belt in the Ebro basin, with the
northern flank being slightly steeper, reflecting the steeply
dipping thrust slices there (Babault et al., 2005). The most
striking geomorphological feature of the Pyrenees is the
presence of low-relief, high-elevation erosion surfaces, first
noted in the literature by Penck (1894). These surfaces are at
about 2000 m elevation and range from tens to hundreds of
square kilometers in aerial extent (Babault et al., 2005). They
are absent in the western Pyrenees and may reflect glacial
erosion in the central Pyrenees (Penck, 1894; Gunnell and
Calvet, 2006), but their origin in the eastern Pyrenees is
20
France
W3
W2
W1
E1
Aquitaine Basin
Aquitaine
Bilbao
Spain
A
N43
Pyre
nees
N42
N
Ebro
Ebr
Ebro Basin
A
Barcelona
100 km
N41
Neogene to Pliocene
Late Cretaceous/Paleogene
Upper Eocene to Miocene
Mesozoic
Lower Eocene to Oligocene
Paleozoic
(a)
Pro-wedge
Ebro Basin
South Pyrenean
Thrust Belt
Retro-wedge
Axial Zone
North
Pyrenean
Zone
Aquitaine
Basin
0
10
20
30
Lower crust
Lower crust
40
50
km
50 km
(b)
Figure 4 Pyrenees Mountains (Beaumont et al., 2000; Sinclair et al., 2005). (a) Simplified geological map of the Pyrenees. (b) Structural cross
section along line AA0 shown in (a), based on the ECORS seismic profile.
more enigmatic. The lower age of the surfaces in the eastern

Pyrenees is constrained by the overlying Late Miocene strata
(Roca, 1996). Precipitation rates today are higher on the
northern side of the range (Meteo-France, 1987), but the
southern Pyrenees and Ebro basin was warm and humid
during the late Eocene as well (Lopez Blanco et al., 2000),
suggesting a more symmetric paleoclimate.
5.3.3.1.2
Origin and significance of low-relief surfaces
Although the Pyrenean erosion surfaces are discussed in

more detail in another contribution to this volume (Chapter
5.6), the different models for their origin are discussed here
because of the importance of the debate for a broader
understanding of the development of low-relief surfaces in
convergent orogens. Similar surfaces have been recognized in
other orogens, including the U.S. Rocky Mountains, Central
Andes, Carpathians, and Sierra Nevada (Ollier and Pain,
2000), and the southeast margin of the Tibetan Plateau (Clark
et al., 2006).
The first set of models envisions these eastern Pyrenean
surfaces forming at a low altitude through peneplanation
(Figure 5(a)), and then undergoing subsequent uplift during
either slab detachment (Gunnell et al., 2008) or thermal
erosion (Gunnell et al., 2009). The standard mechanism of
isostatic uplift produced through crustal thickening is unrealistic, as the process of thickening the crust through compressional faulting would disrupt the surface. An alternative
therefore is removal of the dense mantle lithosphere root: this
would cause uplift without any deformation or thickening at
the surface. Gunnell et al. (2009) argued that the eastern
Pyrenees were rejuvenated at B12 Ma as a result of thermal
erosion of the subcrustal lithosphere, coincident with minor
volcanism and extension. This would also explain the relatively thin crustal thickness (2535 km) of the eastern
Pyrenees, compared with other collisional orogens (Gunnell
et al., 2008, 2009).
Other researchers favor altiplantion (Figure 5(b)), in
which the surfaces are formed at their current elevation and
have experienced minimum uplift or subsidence since that
time. Babault et al. (2005) suggested that piedmont aggradation essentially buried the range and the high-elevation surfaces were graded in the Miocene. Subsequent erosion then
reexhumed the range, stranding the surfaces. There is unequivocal evidence for significant sediment aggradation on the
southern side of the range only (e.g., Beamud et al., 2011), and
therefore, this explanation is only applicable to the southern
surfaces. Fission track data can be interpreted to support burial
and up to 23 km postorogenic exhumation of the southern
Peneplanation
Altiplanation
In situ formation
High local relief
High local relief
High local relief
Sea Level
Sea Level
Sea Level
Uplift
Uplift
Uplift
Piedmont
Aggradation
Low-altitude Peneplain Formation
21
High-altitude Erosion
Glacial/Hillslope Erosion
Piedmont
Aggradation
(c)
Erosion
Erosion
Renewed Uplift
(a)
(b)
Figure 5 Models for the formation of Pyrenean low-relief, high-elevation surfaces. (a) Peneplanation, in which the mountain range is eroded to
a low-elevation, low-relief peneplain, and is then subsequently uplifted, possibly as a result of lithospheric foundering. (b) Altiplanation, in which
a high-elevation, low-relief piedmont is formed continuous with an aggraded surface. Subsequent erosion strands the high-elevation, low-relief
surface (Babault et al., 2005). (c) In situ formation of low-relief surfaces at high altitude through glacial, periglacial, or hillslope processes.
flank of the range (Fitzgerald et al., 1999). However, lower

temperature thermochronometers (Apatite (U-Th)/He and
fission-track) demonstrate only slow exhumation of
0.10.2 mm yr1 after B20 Ma (Gibson et al., 2007). Further,
critics point to the absence of other remnant surfaces at lower
elevations that might define the carapace of this hypothesized
Miocene graded surface, and also suggest that the graded
surface would have been steeper, at B1.251, than typical
braided alluvial fans, which generally support slopes of
40.10.51 (Gunnell and Calvet, 2006). However, Quaternary
piedmont slopes of up to 2.851 have been described in
Tibet (Meyer et al., 1998; Meriaux et al., 2005). Reincision
could have been caused by Quaternary climate change
(Babault et al., 2005) or base-level lowering from the capture
of the Ebro River to the Mediterranean (Coney et al., 1996;
Garcia-Castellanos et al., 2003).
A final option envisioned is the formation of low-relief
surfaces in situ (Figure 5(c)), not requiring peneplanation and
subsequent uplift or altiplanation and subsequent incision.
Advocates for this model downplay the extent of the surfaces
(Calvet, 1994), suggest that some of the surface may be glacial
cirque floors (Gunnell and Calvet, 2006), or point to lithologic controls (Sinclair et al., 2009). Similar surfaces in the
Rocky Mountains have been shown to be parabolic hillslopes
decoupled from the surrounding deeply glacially carved
landscape (Small et al., 1999; Anderson, 2002). In the absence
of direct, reliable paleoelevation data, resolution of this debate
may have to wait.
5.3.3.2
5.3.3.2.1
Alps
Geologists have been studying the geology and geomorphology of the European Alps (Figure 6) for longer and in more
detail than any other collisional orogen on Earth. Despite the
wealth of data available for the Alps, fundamental questions
remain unanswered. The orogen is the result of the collision of
the Adriatic promontory of Africa and Europe. Convergence

began as early as the Cretaceous, but the main phase of deformation has occurred mainly since the Oligocene, as Europe
subducted beneath the Adriatic margin (Schmid et al., 2004).
The Alps are a bivergent wedge, cored by a crystalline European basement and flanked by the Jura fold-thrust belt and the
Molasse Basin to the north and a series of south-vergent
basement nappes and the Po Basin to the south. The Insubric
line is interpreted to be the EuropeAfrica suture.
The Alps began to expand laterally in the early Miocene, with
the major phase of lateral growth around 1820 Ma (Schmid
et al., 1996). Deformation seems to have shifted back into the
interior of the range by the end of the Miocene (Willet et al.,
2006). Sediment accumulation rate has increased two- to
threefold since 35 Ma (Kuhlemann et al., 2002), although
this observed increase could also be explained by the incorporation of longer hiatuses in deposition is longer compared with shorter records, which will produce the apparent
effect of increased sedimentation rate toward the present
(Schumer and Jerolmack, 2009). Some thermochronologic
data are interpreted to indicate that the Alps have had relatively constant long-term exhumation rates and are therefore
at an exhumational steady state (Bernet et al., 2001, 2009).
However, most studies suggest a complex and time-variable
exhumation history (e.g., Carrapa et al., 2003; Kuhlemann
et al., 2006; Carrapa, 2009), including an increase in the exhumation rate since the Pliocene (Vernon et al., 2008), consistent with sedimentologic studies (Kuhlemann et al., 2002).
Convergence today, driven by the counterclockwise rotation of
the Adriatic promontory, implies a complicated pattern of
dextral shear, divergence, and convergence in the Alps, but all
at rates of less than 12 mm yr1 (Nocquet and Calais, 2004).
Leveling data, collected since B1903, indicate up to
1.3 mm yr1 vertical uplift of the core of the range (Champagnac et al., 2009).
The Alps have been heavily influenced by glacial erosion
in the Quaternary (Kelly et al., 2004; van der Beek and
Bourbon, 2008), although the erosion rate over the short term
22
E5
E10
100 km
N
Zurich
N47
Bern
Lepontine
Dome
K LINE
INSUBRIC
Como
A
N45
North Alpine
Foreland Basin
Jura
Eastern Alps
Central Alps
Molasse Basin
Austroalpine Nappes
Jura Mountains
Penninic Nappes
Ophiolite
Helvetic Nappes
Tertiary Granitoids
Southern Alpine Nappes
Western
Alps
Po Basin
Ap
p
en
nin
es
Pre-Triassic crystalline basement
(a)
A
Austroalpine
Molasse Basin
Po Basin
Helvetic
ust
lower cr
Adriatic
Euro
pe
50
km
Insubric Line
an lo
wer c
ru
50 km
st
Mesozoic-Cenozoic Sediments
Continental Upper Crust
Continental Lower Crust
Oceanic Crust
(b)
Figure 6 European Alps (Schmid and Kissling, 2000; Pfiffner et al., 2002). (a) Simplified tectonic map of the Alpine orogen. (b) Cross section
along line AA0 shown in (a) based on the ECORS-CROP seismic profile.
(B1000 years) as measured by cosmogenic 10Be inventory

in detrital sediments matches quite well with longer term,
geodetically derived rock uplift rates (Whittmann et al., 2007).
Precipitation, dominated by the Atlantic Gulf Stream, is relatively symmetric across the Alps, reaching up to B3.8 m yr1
on the northern and southern flanks of the range (Frei and
Schar, 1998; Schwarb et al., 2001).
5.3.3.2.2
Cause of recent uplift of the Alps
The origin of relatively rapid uplift in the Alps poses a geologic

puzzle. Current shortening rates across the range can only
account for o0.2 mm yr1 uplift, far short of the observed
value (Champagnac et al., 2009). Additionally, seismicity does
not correlate with present-day uplift or long-term exhumation
patterns, supporting a decoupling between shortening and
uplift in the Alps (Vernon et al., 2009). A number of theories
have been put forward to explain this phenomenon. As a
purely tectonic explanation, recent foundering of the lithosphere is one possible explanation for the recent uplift of the
Alps. Relatively low modern elevation considering crustal
thickness and geophysical tomographic images suggests the
presence of a complex, partially delaminated lithosphere beneath the Western and Central Alps (Lyon-Caen and Molnar,
1989; Lippitsch et al., 2003). However, the longer wavelength
uplift associated with deep mantle lithospheric removal does
not adequately explain small-scale variations in rock uplift
(Champagnac et al., 2009).
A number of researchers have explored the role of glacial
and fluvial erosional denudation in driving recent uplift. The
strong correlation between short-term indicators of rock-uplift
rate (geodesy and stream power) and erosion rate (modern
sediment yield) and long-term indicators of the same (thermochronology, sedimentation rate) supports this view (e.g.,
Schlunegger and Hinderer, 2001; Kuhlemann et al., 2002;
Cederbom et al., 2004; Whittmann et al., 2007; Champagnac

et al., 2009; Korup and Schlunegger, 2009; Vernon et al.,
2009). Kuhni and Pfiffner (2001) and Korup and Schlunegger
(2009) pointed out the importance of rock type in the Alps in
determining how fast rocks erode, but also supported the
correlation of the long-term uplift rate and geomorphic signifiers of erosion rate. The fact that erosion is dominated by
glaciers in the Alps seems to be important: glacial conditioning as a result of glacialinterglacial fluctuation prevents
the landscape from reaching an equilibrium form, maintaining high rates of erosion over the long term (Norton et al.,
2010). This erosion-driven rebound may have only been established in the last few million years. Climate change after
the B4.6 Ma closure of the Isthmus of Panama intensified the
Atlantic Gulf Stream, increasing precipitation over much of
Europe (Haug and Tiedemann, 1998). Cederbom et al. (2004)
presented apatite fission-track data from the northern Alpine
foreland basin that imply 13 km of erosion after B5 Ma.
Linking this with erosion rate data for the Swiss and Western
Alps, Kuhlemann et al. (2001) argued that B4.6 Ma climate
change intensified erosion across the range, driving rebound
(Cederbom et al., 2004). Finally, a few studies have pointed to
the very recent rebound from simply removing ice from the
landscape since the Last Glacial Maximum (Gudmundsson,
1994), but the effects are not thought to be sufficient to explain
more than a B1 mm yr1 of the observed uplift (Persaud and
Pfiffner, 2004; Stocchi et al., 2005; Barletta et al., 2006).
Although questions of course remain, climate-driven denudation appears to be an important driving mechanism of
the recent uplift in the Alps. Elastic modeling convincingly
demonstrates that most of the uplift can be explained by
lithospheric bending associated with mass redistribution on
the northern flank of the Alps; to the south, a B30% deficit
could be explained by mantle delamination (Champagnac
et al., 2009).
5.3.3.2.3
Range width and climate change
Late Miocene climate change has also been implicated for

changes in the width of the deforming zone of the Alps over
time. Deformation in the Jura basin and along the Insubric line
at 2018 Ma testifies to expansion of the orogen (Bourquin
et al., 1946; Schmid et al., 1996). A drier, continental climate
prevailed at that time (Berger, 1992), which would have
promoted lower erosion rates. Additionally, the exposure of
crystalline basement rocks in the core of the Central Alps, on
the basis of 39Ar/40Ar cooling ages in detrital white mica at
B18 Ma (von Eynatten et al., 1999), would have lowered
erosional efficiency. Schlunegger and Simpson (2002) argued
that lower erosion rates, meaning less removal of mass from
the orogen, would promote lateral growth of the range.
However, Rosenberg and Berger (2009) questioned the correlation of increased erosion rates with lateral orogen growth,
emphasizing instead the role of lower crustal configuration.
The Alps seems to have undergone a major contraction in
the latest Miocene and the early Pliocene. Deformation to the
north of the range ceased by 3.4 Ma (Bolliger et al., 1993;
Becker, 2000) and the Molasse Basin is today being eroded
(Cederbom et al., 2004), both consistent with deformation
stepping into the interior of the orogen by the Pliocene (Willet
et al., 2006). To the south, the timing of deformation is well
23
constrained by seismic reflection data; it ceased there by

5.3 Ma. The exhumation rate appears to have increased
(Vernon et al., 2008) and sediment yield may have increased
at about this time (Kuhlemann, 2000; Kuhlemann et al.,
2002), although the latter may reflect instead the incompleteness of the sedimentary record (Schumer and
Jerolmack, 2009). One explanation for this shrinkage is that
plate convergence rates have slowed from the Miocene to the
present; if erosion rates are maintained in the face of decreased
accumulation rates, the orogen will decrease in width
(Rosenberg and Berger, 2009). However, an increase in erosion
could also alter the dynamics of the doubly vergent wedge; to
maintain the slope of the wedge in the face of increased erosion, deformation would need to step back into the range
(Whipple and Meade, 2004). The increased erosion could be
driven by closure of the Isthmus of Panama and the
strengthening of the Atlantic Jet Stream at around 4.6 Ma
(Cederbom et al., 2004). However, the transition occurred
earlier in the Southern Alps (by 5.3 Ma) and is coincident with
the Messinian Salinity crisis. This period of drying of the
Mediterranean would have significantly lowered the base level
(Hsu et al., 1977) for the southern parts of the range, thus
providing an alternative explanation for the onset of erosion
and narrowing of the orogen (Willet et al., 2006). However,
Rosenberg and Berger (2009) questioned the correlation of
erosion and range-width reduction as well, citing uncertainties
in the termination of tectonic activity in the Alpine foreland.
5.3.4
5.3.4.1
ArabiaEurasia Collision
Along the AlpineHimalayan chain to the east, the Arabian

plate (rifted from the African margin) is colliding with Eurasia,
causing widespread deformation, extrusion of continental
crust along strikeslip faults, and the formation of the Anatolian and Central Iranian plateaus (Figure 7). After a series of
small accretion events, final closure of the Neo-Tethys and
collision between Arabia and Eurasia began in the Middle to
Late Miocene (1012 Ma; Dewey et al., 1986; McQuarrie et al.,
2003), although other authors have argued for closure as early
as the Late Cretaceous (Agard et al., 2005). Collision initiated
shortening, uplift, and mountain building in the Zagros
Mountains adjacent to the Persian Gulf and in the Caucasus,
Talash, Alborz, and Kopet-Dag mountains wrapping around
the Black and Caspian seas (Figure 7; e.g., Stocklin, 1968;
Falcon, 1974; Berberian and King, 1981; Berberian et al.,
1982; Berberian, 1983, 1995; S- engor et al., 1985; Alavi, 1994).
Modern and historical seismicity and GPS data suggest
that the intervening region, including the Turkish and
Central Iranian plateaus, largely behaves as a series of rigid
blocks, with relatively minor internal deformation (e.g.,
Jackson and McKenzie, 1984; Vernant et al., 2004). Although
shortening is active along the flanks of the Central Iranian
Plateau (the Bitlis-Zagros fold-thrust belt to the south and
the Caucasus Mountains to the north), the plateau itself
undergoes almost exclusively strikeslip and normal faulting
(Copley and Jackson, 2006). The region experienced a transition from collision to escape with the initiation of the
20 E
30 E
40 E
50 E
60 E
45 N
Eurasia
NAF
Anatolian
Anatolian
Plateau
Plateau
Taurus
Taurus
35 N
Kop
et-D
ag
Talash
Talash
EAF
35 N
40 N
Gr. C
auca Caspian
sus
Sea
Black Sea
40 N
45 N
24
Albo
Alborz
lborrz
MRF
Above 1000 m
Za
gr
os
C
entral
Central
IIranian
ranian
P
lateau
Plateau
40 N
40 N
Mediterranean Sea
25 N
30 E
(a)
Zagros
fold-thrust belt
25 N
Arabia
Persian
Gulf
Red
R
ed
S
ea
Sea
40 E
50 E
Central Iran
Alborz
South Caspian Basin
60 E
MOHO
200 km
no vertical exaggeration
Sedimentary Cover
Oceanic-type basement
(b)
Figure 7 TurkishIranian Plateau region (Allen et al., 2004). (a) Simplified tectonic map of the ArabiaEurasia collision zone. Area above 1000 m
elevation shown by light shading. NAF, North Anatolian fault; EAF, East Anatolian fault; MRF, Main Recent Fault. (b) Crustal cross section along
AA0 shown in (a).
East and North Anatolian strikeslip faults (Figure 7) around

57 Ma, which accommodate westward extrusion of the
Anatolian Plate (e.g., S- engor, 1979; Westaway, 1994, 2003;
Armijo et al., 1999). The Main Recent fault (Figure 7) to the
east of the Anatolian plateau in Iran accommodates significant
strikeslip movement as well. Strikeslip and normal faults on
the TurkishIranian plateau appear to have originated at
around the same time (572 Ma; Copley and Jackson, 2006).
The modern convergence rate between Arabia and Eurasia is
2272 mm yr1 (Vernant et al., 2004). The Anatolian plateau
has been volcanically active since 68 Ma (Innocenti et al.,
1976a, 1976b; Pearce et al., 1990). The crust beneath
the Anatolian plateau is a relatively thin 3850 km, based on
receiver function data (Zor et al., 2003), and seismic
data indicate that the mantle lithosphere is entirely absent
(Gok et al., 2007).
The Anatolian and the Central Iranian plateaus (Figure 7)
are arid and partly internally drained, with an average elevation of B2000 m. Geographically, these plateaus are not as
striking as their counterparts in Tibet or the central Andes, but
they are of relatively low relief, consistent elevation, and are
partially internally drained, particularly in Turkey. The central

part of the Anatolian plateau is characterized by a number of
large basins, including Lake Van, and the Erzurum, Pasinler,
Mus, and Ararat basins and active volcanism in the Kars volcanic plateau (e.g., Dhont and Chorowicz, 2006). Although
rare, glaciers were present during the Last Glacial Maximum in
some parts of the region (Erol, 1991).
5.3.4.2
Mantle Lithosphere Delamination, Uplift, and

Drainage Reorganization
Evidence in the ArabiaEurasia collision seems to point to some

form of delamination of the subcontinental mantle lithosphere
beneath the Anatolian plateau, in which the lithosphere peels
away from the overlying crust, detaches, and descends into the
asthenosphere (Copley and Jackson, 2006; Faccenna et al.,
2006). The influx of hot asthenosphere would buoy the plateau
and produce extension and volcanism within it (Bird, 1979;
S- engor et al., 2003; Keskin, 2003). Recent modeling suggests
that delamination could produce a 2-km-high plateau, but
interestingly, also points to a pulse of migrating subsidence over

the hinge of the detaching lithosphere; the interior of the
plateau is also predicted to be in extension, whereas compression continues on the margins (Gogu-s and Pysklywec,
2008). Evidence for delamination beneath the Anatolian
plateau includes thin crust and absent mantle lithosphere (Zor
et al., 2003; Gok et al., 2007), extensional basins within the
plateau (Dhont and Chorowicz, 2006), and recent, widespread volcanism (Innocenti et al., 1976a, 1976b; Pearce et al.,
1990). Delamination or the convective removal of mantle
lithosphere through RayleighTaylor instability has been
suggested for plateaus in other collisional (Tibet; e.g., Molnar
et al., 1993) and noncollisional (Central Andes; e.g., Kay and
Kay, 1993) settings.
Lithospheric delamination is reflected in the geomorphology of the Anatolian plateau, as the crust undergoes shortening or extension and changes in surface elevation, resulting
in a complex history of basin filling and reincision (Ocakoglu,
2002; Jaffey and Robertson, 2005) and of drainage reorganization and incision (e.g., Monod et al., 2006) on the southern
plateau margin. A number of studies have examined the uplift
history of the plateau region, the Taurus Mountains, and the
history of drainage reorganization, and have linked their
70 E
72 E
74 E
Nanga Parbat
Parbat
Neogene-Quaternary Sediments
Sub-Himalaya
Tethyan Himalayan Sequence
Lesser Himalayan Sequence
Greater Himalayan Complex
Cenozoic Plutons and Leucogranties
Asian crust
36 N
78 E
34 N
80 E
32 N
70 E
observations to evidence for lithospheric delamination. Jaffey

and Robertson (2005) used the history of several small sedimentary basins in southeastern Turkey to infer a complex
history of uplift and extension there. The region was low and
topographically subdued enough to allow a marine incursion
in the Oligocene-early Miocene, although paleocurrent and
provenance data indicate that the Taurus Mountains were
uplifting, probably related to subduction of the Neotethys
(Jaffey and Robertson, 2005). The region underwent significant uplift in the Plio-Quaternary, driving river incision in the
Taurus Mountains, although the authors link this to underplating rather than to delamination (Jaffey and Robertson,
2005). Demir et al. (2008) focused instead on river incision as
a proxy for surface uplift. Regionally, in the central part of the
plateau, the Euphrates River records B600 m incision since
B9 Ma. Incision and deformation are recorded along the
Tigris River as well, which drains the eastern Taurides (Nicoll,
2010). The Taurus Mountains, with their relatively flat-lying,
carbonate, stacked thrust sheets, have also yielded a number of
important insights into the uplift history of the region. For
example, Monod et al. (2006) identified a relict surface at
15002200 m asl in the western Taurides (Figure 7), along
with a paleo-drainage network, which has been incised by the
76 E
.
sR
du
In
34 N
82 E
72 E
AT
74 E
30 N
84 E
san
gpo
R.
76 E
86 E
88 E
90 E
92 E
94 E
30N
Namche Barwa
Bar
Barwa
S
South
outh T
Tibetan
ibetan Detachment
Detachme
etachmen
nt
Mt.
M
t. E
Everest
verest
80 E
82 E
26 N
84 E
ra
ut
ap
m
ah
Br
Main Central Thrust
Shilong Plateau
Plateau
86 E
(a)
88 E
90 E
92 E
94 E
MCT
STD
MBT
MC
ST
T
GCT
10
10
40
50
28N
26N
96 E
km
10
30
R.
20
96 E
A
28 N
78 E
km
10
25
20
Indian basement
30
channel
40
50
(b)
Figure 8 Himalaya (Yin, 2006). (a) Simplified tectonic map of the Himalayan orogen with major rivers shown. (b) Cross section along line AA0
shown in (a). MBT, Main Boundary thrust; MCT, Main Central thrust; STD, South Tibetan detachment; GCT, Great Counter thrust.
26
modern river network to form gorges more than 1000 m deep.

The relict landscape is Miocene in age, suggesting more than
1500 m uplift since that time, most likely in the PlioceneQuaternary (Monod et al., 2006). Although much of the research in this region lacks a coherent theme, it is apparent that
surface uplift has disrupted and rerouted the regional drainage
system, either as a result of local deformation along faults or
more regional deformation related to delamination of the
mantle lithosphere.
5.3.5
IndiaAsia Collision
The culmination of the AlpineHimalayan chain occurs in the

east, where the convergence of India and Asia has created the
highest mountains and the largest continental plateau on Earth
(Figures 8 and 9). Collision occurred around 55 Ma (e.g.,
Garzanti and van Haver, 1988; Rowley, 1996; Zhu et al., 2005;
Green et al., 2008; Dupont-Nivet et al., 2010). Although the
rate of convergence has declined since then, it remains high, at
3444 mm yr1 (Molnar and Stock, 2009). A long history
occurs of accretion of terranes to the Asian margin before final
collision, which may have thermally weakened and caused
early thickening of the crust that today underlies the Tibetan
Plateau (e.g., Kapp et al., 2005a, 2005b). Deformation penetrates deeply into the interior of the Asian continent (Figure 9),
causing, for example, rejuvenation of the Kyrgyz Tian Shan,
and uplift in Mongolia, nearly 2000 km from the suture zone
(Tapponnier and Molnar, 1979). The Tibetan crust today is the
thickest on Earth (more than 70 km in places; Holt and
Wallace, 1990). The plateau may host a partially molten channel
in its lower part (Nelson et al., 1996) that has been
hypothesized to flow toward the Himalaya, drawn in part
by intense surface erosion there (Beaumont et al., 2001), or
toward the southeast margin of the Tibetan plateau (Royden
et al., 1997; Clark et al., 2000). Although crustal thickening is
an important first-order manifestation of the IndiaAsia
collision, a number of major strikeslip faults slice the
plateau and adjacent regions (Figure 9; Molnar and Tapponnier,
1975). They may have accommodated significant extrusion of
material away from the indentor (Tapponnier et al., 1982); the
Ailao Shan shear zone in southwest China, for example,
accommodated up to 1000 km of eastward movement of the
Indochina block (Tapponnier et al., 1990; Leloup et al., 1995),
and the Altyn Tagh Fault and Kunlun faults may have accompanied extrusion as well (Zhou and Graham, 1996). However,
the importance of these faults in accommodating the major
deformation of the plateau is much debated (e.g., England and
Houseman, 1988; Wright et al., 2004).
Most of the precipitation in the Himalaya falls during the
summer months during the Indian Summer Monsoon. Orographic forcing sets up a strong precipitation gradient across
the Himalaya and precipitation decreases along the strike of
the range to the west as well (Bookhagen and Burbank, 2006).
The Himalayas support countless glaciers, fed by the summer
precipitation. To the north, the B5000 m plateau is largely
arid and the western half is internally drained. The eastern half
of the plateau is of low relief, but is incised by a few major
rivers, including the Yellow, Yangtze, and Salween rivers. The
northern margin of the plateau is coincident with the Altyn
Tagh fault and is relatively steep, and a portion of the eastern

margin is steep as well adjacent to the Sichuan basin. The
southeast and northeast margins, however, are more diffuse
(e.g., Clark and Royden, 2000). The Himalaya is discussed
briefly and then a number of tectonic geomorphic problems
associated with the Tibetan plateau in more detail.
5.3.5.1
Himalaya
The tectonic geomorphology of the Himalaya and particularly

the interesting ideas of channel flow and tectonic aneurysm
are discussed in more detail in other contributions to this
volume (Chapters 5.13, 5.14, and 5.15), but because of its
importance in considering tectonics, climate, landscape evolution, and geomorphology, a brief overview of the Himalaya
is given here.
5.3.5.1.1
Monsoon precipitation and climatetectonic

coupling
A number of studies have pointed out the correspondence

between maximum precipitation and exhumation rate or a
related correspondence between major, erosive rivers, and
rapid exhumation (e.g., Lave and Avouac, 2001; Hodges et al.,
2004; Bojar et al., 2005; Thiede et al., 2005; Wobus et al.,
2005). Other studies point to time correspondence. For example, Wobus et al. (2008) documented an increase in the
exhumation rate in the central Nepalese Himalaya at B10 Ma
that could be linked to a strengthened Indian monsoon,
whereas Huntington et al. (2006) observed an increase in the
Pliocene that they linked to global climate change. In an
interesting study, Grujic et al. (2006) investigated both spatial
and temporal climatetectonic coupling. The rise of the Shilong Plateau (Figure 8(a)) in the Indian foreland in the late
Miocene blocked moisture to the Himalayas, creating a downwind rain shadow. The erosion rate decreased around this
time in the Himalaya, and the erosion rate along strike is
significantly lower in this zone (Grujic et al., 2006). Despite
these compelling correlations, a number of studies have argued against climatetectonic coupling in the Himalaya. In a
study combining locally measured weather data with thermochronology, Burbank et al. (2003) concluded that spatial
exhumation rates are invariant despite large gradients in precipitation. Interestingly, although exhumation and precipitation have been shown to correlate in some studies, Thiede
et al. (2009) did not find a correlation between a regionally
compiled thermochronology data set and present-day rainfall,
local relief, or specific stream power, further calling into
question the nature of climatetectonic coupling in the
Himalaya. Finally, many studies that suggest a Pliocene increase in the erosion rate may not adequately account for the
advection of isotherms or for the ramp-flat geometry of the
Himalayan wedge (Figure 8(b)), in which exhumation over a
ramp will produce a signal of accelerating exhumation as the
rock approaches the surface, in the absence of a change in
climatic variables or the tectonic convergence rate.
5.3.5.1.2
Channel flow
An outgrowth of the idea of climatetectonic coupling in

the Himalaya is the channel flow model, in which it is
80 E
85 E
90 E
95 E
40 N
105 E
an
Sha
Qai
n
dam
Bas
in
.
gh F
n Ta
y
t
l
A
Ka
m
ra
ko
ra
Yan
gtz
eR
M
.
ek
o
Salw
lween ng
R. R
.
Tibet
ibet
T
110 E
40 N
Qili
Tarim Basin
Tarim
35 N
100 E
27
Kun
lun F
.
35 N
F.
Min
M
in Shan
en
m
ng an
Lo Sh
30 N
30 N
Tsan
Tsangpo
sang
gpo R.
A
25 N
o
la
Ai
25 N
Above 2000 m
Above 4000 m
Indochina
Indochina
an
Sh
India
Re
d
1000 km
20 N
Ri
ve
r
Bay
B
ay of
Bengal
Bengal
80 E
85 E
90 E
20 N
South
South
ea
China S
Sea
95 E
100 E
105 E
110 E
(a)
A
HIMALAYA
MBT
0
70
150
MCT STD
GCT
MOHO
LHASA
BLOCK
QIANGTANG
BLOCK
Kunlun
Fault
Qaidam
Basin QILIAN
SHAN
A
Hexi
Corridor
MOHO
MOHO
km
(b)
Figure 9 Tibet. (a) Simplified tectonic map of Tibet (Tapponnier et al., 2001). Area above 2000 and 4000 m elevation shown by shading. Major
rivers shown in gray. (b) Cross section (Yin and Harrison, 2000) along line AA0 shown in (a). MBT, Main Boundary thrust; MCT, Main Central
thrust; STD, South Tibetan detachment; GCT, Great Counter thrust.
hypothesized that ductile extrusion of hot, weak crust from

beneath Tibet is linked to intense erosion in the Himalaya
(Beaumont et al., 2001). This model explains many of the firstorder observations of the Himalaya, including the presence of
a zone of high-grade gneisses and young leucogranites bounded by the Main Central thrust below and the extensional
South Tibetan detachment above (Figure 8(b)), and the inverted metamorphic gradient, and the PT paths both within
and adjacent to the inferred channel (Beaumont et al., 2001,
2004; Grujic, 2006). Since the publication of this model, a
number of field-based studies have provided support to this
(e.g., Searle and Szulc, 2004; King et al., 2007). However, other
studies point out critical inconsistencies with the channel flow
model. For example, Robinson et al. (2006) argued for normal
propagation of the thrust belt in western Nepal since the early
Miocene, with no evidence for disruption by channel flow;
exhumation of Greater Himalayan rocks is the result of
stacking of thrust sheets and erosion. In Bhutan, Long and
McQuarrie (2010) argued for less than 20 km displacement
along the South Tibetan Detachment, thus limiting the extent

of possible channel flow there. In a study focused on PTt
data, Kohn (2008) also demonstrated that many of the firstorder features of the Himalaya can be successfully explained
by critical taper wedge models, and are inconsistent with the
channel flow model. Thus, the degree and structural importance of climatetectonic coupling in driving exhumation of
the zone between the Main Central thrust and the South
Tibetan detachment remain controversial.
5.3.5.1.3
Syntaxes tectonic aneurysms
An extreme form of tectonicclimate coupling has been proposed to explain enigmatic features in the syntaxial corners of
the orogen at the peaks of Nanga Parbat in the west and
Namche Barwa in the east (Figure 8(a)): the tectonic aneurysm model. It is hypothesized that large-magnitude river
incision removes the strong upper crust, thus advecting deeper, hotter rocks toward the surface; this increase in the nearsurface geothermal gradient locally weakens the crust.
28
Deformation subsequently becomes focused into this weak,

topographically steep zone, leading to a positive feedback
cycle (Zeitler et al., 2001; Koons et al., 2002). In the western
syntaxis, the IndiaAsia suture is warped into a NS-trending
antiform that is host to two high peaks (Nanga Parbat at
8126 m asl and Haramosh at 7397 m asl) that lie south of the
main trend of the high Himalaya. The antiform is bisected by
the Indus River, which flows at about 1000 m asl elevation.
The combination of high peaks and the deeply incised river
has created some of the greatest relative relief on the planet.
The entire antiform, but Nanga Parbat in particular, is
underlain by young metamorphic rocks cut by 13 Ma anatectic granites, an extremely steep metamorphic gradient
(60 1C km1), and a shallow brittleductile transition
(B3 km, marked by the lower cutoff in microseismicity), all of
which are interpreted to indicate rapid exhumation and the
presence of hot, weak rocks near the surface of the Earth
(Zeitler et al., 2001). Similar relationships are observed in the
eastern syntaxis, where the Tsangpo River wraps around the
Namche Barwa peak (Zeitler et al., 2001), and a similar NStrending antiform has been exhuming rapidly since 4 Ma
(Seward and Burg, 2008). Finnegan et al. (2008) found a
strong correspondence between incision, relief, and young
cooling ages, arguing for a sustained balance between rock
uplift and denudation. They argued that the impressive
knickpoint in the TsangpoBrahmaputra River has been pinned in place for at least the last 1 m yr, protecting the landscape upstream in Tibet from incision. Thus, although the
nature and extent of tectonicclimate coupling remains
debated in the Himalaya, there is strong field- and modelingbased evidence for a relationship between erosion and
tectonics both along the length of the range and within the
syntaxial corners.
5.3.5.2
Tibet
An important tectonicgeomorphic question in Tibet

(Figure 9) relates to its elevation during the history of its
evolution: how did it grow upward and outward with time
and how is this reflected in the geology, regional geomorphology, and drainage pattern. Although a comprehensive review
of paleoaltimetry is beyond the scope of this contribution, it is
discussed here briefly in the context of the uplift history of the
Tibetan plateau and is expanded on in Chapter 5.6. Focus is
also on recent research on the northeast, east, and southeast
plateau margins into how the plateau grew in these directions,
and the associated implications for the underlying tectonic
controls.
5.3.5.2.1
Paleoaltimetry
Most estimates for the uplift history of Tibet are based on

stable isotope paleoaltimetry, in which the fractionation of
18
O and 2H (D) in water vapor with increasing elevation is
exploited to determine approximate elevation from carbonate
rocks is soils and lakes (e.g., Rowley and Garzione, 2007).
Although the determination of altitude is complicated by
issues of moisture transport pathway, calibration to modern
precipitation, and carbonate diagenesis, the results appear to
be consistent in Tibet, and data indicate that much of central
Tibet has been at its current elevation since at least 2640 Ma

(Rowley and Currie, 2006; DeCelles et al., 2007). In northcentral Tibet, data indicate that the plateau was lower than
2 km at 3540 Ma (Cyr et al., 2005). These data indicate a
steady north and eastward march of the high-elevation margin
of the plateau since B40 Ma (Rowley and Garzione, 2007;
Kent-Corson et al., 2009). Paleoaltimetry data are thus inconsistent with models that call for the recent, overall uplift
of the Tibetan Plateau as a result of lithospheric foundering
(e.g., Molnar et al., 1993; Wang et al., 2006), favoring instead
models of gradual outward plateau growth (Tapponnier et al.,
2001), potentially through the flow of lower crust (e.g.,
Royden et al., 1997).
5.3.5.2.2
Lower crustal flow
One of the more controversial proposals for the development

of the Tibetan Plateau is that the margin has propagated
outward over time as a result of flow of lower crust (e.g.,
Royden et al., 1997). There is evidence for partially molten or
fluid-rich material at a depth of greater than 15 km beneath
the southern plateau (Nelson et al., 1996). It is hypothesized
that the pressure gradient from the thick crust and high
topography of Tibet compared with the thinner crust and
lower elevation of the foreland drives lower crustal flow in a
channel that may be on the order of B15 km thick (Figure 10;
Royden et al., 1997; Clark and Royden, 2000). Most attention
has been focused on the southeast plateau margin. There is
some support for this model in terms of geology, geophysics,
and geodynamic modeling, but it is mixed. For example, there
is a gradient in crustal thickness from the plateau southeast
toward the South China Sea (Li and Mooney, 1998), but no
apparent evidence for upper crustal shortening that could
account for this thickening (Wang and Burchfiel, 1997; Wang
et al., 1998), leading to the inference that thickening occurred
through influx of material in the mid to lower crust instead
(Figure 10(b)). Some geophysical studies of shear wave
splitting suggest decoupling of movement in the upper crust
relative to the lithospheric mantle, possibly because of the
presence of the weak, channelized, flowing crust (Lev et al.,
2006). However, other studies have found a correspondence
between surface and mantle shear wave splitting data in Tibet,
successfully modeling surface displacement using a mantledriven approach (Flesch et al., 2005), precluding large-scale
lower crustal flow. Finally, although many modeling studies
support the channel flow hypothesis (e.g., Royden et al., 1997;
Clark and Royden, 2000; Beaumont et al., 2001, 2004; Cook
and Royden, 2008), others find that material cools too quickly
within the channel, preventing it from intruding more than
B150 km from the plateau margin (Rey et al., 2010), far less
than is hypothesized for the southeast margin of the Tibetan
plateau.
In the context of these competing geophysical and modeling studies, key evidence for channel flow models therefore
comes from the tectonic geomorphology. A low-relief, relict
landscape characterizes much of the southeast margin of the
plateau, forming a carapace that slopes smoothly from Tibet to
the South China Sea (Figure 10(a); Clark et al., 2006). This
surface is deeply incised by several major rivers, including the
Mekong, Salween, Red, and Yangtze (Figures 9 and 10; Clark
and Royden, 2000; Clark et al., 2006). It is geomorphologically
low-elevation cratonic areas

Internally drained plateau
erosion surface
(relict landscape)
major rivers
flux
29
A. weak foreland
normal
crust
added crust
channel
added crust
B. strong foreland
normal crust
channel
flux
added
crust
(b)
(a)
Figure 10 (a) Geomorphic map of the Tibetan Plateau (Clark et al., 2006) illustrating the location of the low-relief, relict landscape in eastern
Tibet and on the northeast and southeast plateau margins. Major drainages are also shown. (b) Schematic illustration (Clark and Royden, 2000)
of the viscous flow model, in which material flows from Tibet into the margins in a narrow crustal channel. Excess material is accreted to the top
and bottom of the channel to produce crustal thickening, which in turns results in isostatic crustal uplift. Areas with weak forelands form long,
low-gradient slopes similar to the southeast margin of the plateau; areas with strong forelands form steep margins, similar to the Sichuan basin.
Approximate locations of schematic cross sections are shown in (a).
striking: for example, the Ailao Shan Mountains (Figure 9) in

Yunnan are beveled and deeply saprolitized, and contrast with
steep, landslide-prone, incised tributaries to the Red River
(Schoenbohm et al., 2004). River profiles reflect the presence
of the transient, relict landscape as well (Schoenbohm et al.,
2004). Evidence from thermochronology, paleontology, and
geomorphology indicates that incision of at least 1500 m
began in the northwest around 913 Ma (Clark et al., 2005b),
and propagated to the southeast at 50200 mm yr1, reaching
the Red River by the Pliocene (Schoenbohm et al., 2006).
Incision in these studies is interpreted to reflect uplift of a
regional, low-relief, and originally low-elevation landscape.
Additional support for this uplift pattern is seen in major
drainage reorganization, discussed in the next section.
The strong crust of the Sichuan basin is hypothesized to
have impeded the flow of lower crust along the eastern margin
of the Tibetan plateau (Figure 10(b); Royden et al., 1997; Clark
and Royden, 2000; Cook and Royden, 2008), causing the development of welts of anomalously high topography adjacent
to the basin (Clark et al., 2005a). Displacement on thrust faults
in the Min Shan and Longmen Shan (Figure 9), which define
the border of the plateau, is minimal (Wang and Burchfiel,
1997; Kirby et al., 2000), despite evidence for rapid uplift from
tilting of alluvial surfaces and thermochronology (Kirby et al.,
2000). River profile analysis demonstrates uplift focused in the
mountains immediately adjacent to the plateau (Kirby et al.,
2003), and large landslides are frequent (Ouimet et al., 2007).
The situation is more complicated on the northeast margin
of the plateau, although lower crustal flow is hypothesized
there as well (Clark and Royden, 2000). For example,
although river incision is thought to lag only slightly behind
uplift on the southeast margin of the plateau, Craddock et al.
(2010) demonstrated that incision along the Yellow River lags

uplift by nearly 10 Ma, implying that climate change may exert
the primary control on river incision and integration in Tibet.
5.3.5.2.3
Drainage reorganization, uplift, and crustal

shear
Uplift of the plateau has reorganized the drainage system of

central Asia. Much of western Tibet is internally drained, and
only a few other major rivers tap the Plateau (Figure 10(a)).
The watershed shape of these rivers is unique: they are widely
spaced and low gradient in Tibet, are narrow and closely
spaced as they cross the Himalaya or traverse the southeast
plateau margin, and then spread apart again, terminating in
widely separated seas (Figures 9 and 10(a); Brookfield, 1998;
Clark et al., 2004). One hypothesis to explain this unusual
morphology is that a single major river with a dendritic pattern once drained the eastern part of the plateau, traveling
through the modern Red River and reaching the South China
Sea, but that as a result of uplift on the southeast plateau
margin, a series of river captures diverted flow away from the
uplifting region (Brookfield, 1998; Clark et al., 2004). For
example, Clark et al. (2004) proposed that the modern
Yangtze (Figures 9 and 10) is a result of a series of capture and
drainage reversal events, culminating in the capture of the
upper Yangtze away from the Red River Drainage. The capture
of the Tsangpo by the Brahmaputra River in the eastern syntaxis is proposed as well (Figures 7 and 8; Clark et al., 2004).
Alternatively, Hallet and Molnar (2001) argued that the major
rivers of the southeast plateau margin have always been separate, and that their close modern juxtaposition reflects
shortening and crustal-scale shear: in this scenario, the rivers
act as passive strain markers. Although some such shear can be
30
accommodated by models for lower crustal flow, it must be

relatively minimal, because otherwise the low-relief relict
landscape would be disrupted. Similarly, Seward and Burg
(2008) proposed that the present geometry of the
TsangpoBrahmaputra River reflects the advection of the river
during the formation of the Namche Barwa anticline rather
than a capture event. Interestingly, modeling indicates that
both uplift-driven river capture and crustal strain could play a
role in shaping the large-scale drainage pattern of eastern Tibet
(Stuwe et al., 2008). Drainage reorganization in the western
Himalaya is suggested as well. Initial diversion of the Indus to
its current position around 18 Ma is inferred from sediment
provenance (Najman et al., 2003), although it could have
been younger at B12 Ma (Shroder and Bishop, 2000). A
change in the abundance and isotopic composition of sediments in the Indus fan indicates expansion of the Indus
drainage by capture around 5 Ma (Clift and Blusztajn, 2005).
5.3.5.2.4
Landslides and glacial dams
Although landslides occur in many both tectonically active and

inactive regions around the world, and are heavily influenced
by climate, they may play a role in Tibet and the Himalaya in
protecting the plateau from erosion, and thus deserve special
mention here. Paleoaltimetry and structural data indicate that
the plateau has been high since as early as 40 Ma. Preservation
in the face of active incision by the aggressive rivers of the
Himalaya therefore presents a conundrum. Although differential rock uplift may account for much of this stalling of river
incision, Korup and Montgomery (2008) and Korup et al.
(2010a, 2010b) suggested that glacial damming, at least since
the Quaternary, can impede incision through impounding
sediment upstream of the dam and allowing aggradation during interglacial periods. Their mapping indicates that over 80%
of certain channel networks would have been affected by glacial
dams during Holocene advances (Korup and Montgomery,
2008). Similar aggradation is produced by large landslides on
the steep margin of the plateau adjacent to the Sichuan Basin
(Ouimet et al., 2007), where they are thought to hinder the
adjustment of river channels in response to plateau uplift. In
the Himalaya, landslides appear to reflect climate variability
and times of increased penetration of monsoon precipitation.
For example, along the Sutlej River in the western Himalaya,
late Pleistocene- and Holocene-intensified monsoon phases are
linked to a fivefold increase in the sediment transport rate, as
reflected in landslide-dammed lakes (Bookhagen et al., 2005)
and the formation of strath terraces and alluvial fans (Dortch
et al., 2011). The landslides themselves are likely triggered by
the intensified monsoon phases (Dortch et al., 2009). Finally,
cosmogenic data indicate that even the most deeply incised of
Himalayan Rivers can become filled with hundreds of meters of
sediments during times of strengthened monsoon (Pratt et al.,
2002).
5.3.6
Ancient Orogens
Despite being tectonically inactive for up to hundreds of millions of years, a number of ancient orogens retain their topographic signature. Examples include the Paleozoic ranges
around the margins of the rifted Atlantic basin, such as the
Appalachians in North America and Scandinavian Caledonides

in Norway, the Lachlan fold belt of southeast Australia, and the
Ural Mountains (Figure 1). The question in each case is whether the modern expression of these orogens is a remnant of the
ancient topography that has persisted because of mechanisms
that slow the erosion rate over time, preserving topography, or
whether the ranges have been tectonically rejuvenated, perhaps
by processes operating in the mantle lithosphere or the mantle.
Two examples are discussed below that have been the subject of
much debate in the recent literature.
The Appalachians are the result of the collision of fragments of Gondwana and then Gondwana itself (modern
Africa) with Laurentia (North America) in the Ordovician to
Permian, forming the supercontinent Pangaea. The last tectonic event to affect the region was the rifting of the North
Atlantic in the Triassic. Topography in the Appalachians
reaches up to 1500 m asl. River profiles are smooth and concave up (Hack, 1960). Offshore stratigraphy may indicate a
significant increase in erosional flux in the late Cenozoic
(Pazzaglia and Brandon, 1996), suggesting rejuvenation of the
range, although this could also reflect bias in the sedimentary
record (e.g., Schumer and Jerolmack, 2009). One nontectonic
possibility is that recent uplift could be a response to climatedriven erosion of the Appalachians, enhanced in the late
Cenozoic by global cooling and climate deterioration; erosional mass removal in the mountains and loading in the
adjacent ocean basin would lead to flexural uplift (Pazzaglia
and Gardner; 1994; Pazzaglia and Brandon, 1996). Matmon
et al. (2003), however, disputed the recent increase in erosion.
They found that the erosion rates measured over the long term
using fission-track and sediment budget are in agreement with
short-term sediment yield and cosmogenic basin-wide erosion
rate data, suggesting steady erosion at B30 m Myr1 since the
Mesozoic, precluding a recent increase in the erosion rate.
They instead ascribed the persistent topography in the Appalachians to slow erosion of a deep crustal root, and the associated, but diminishing, uplift (Matmon et al., 2003). A
related, nontectonic explanation for the persistent topography
calls on the dynamics of erosion and fluvial channels. Fluvial
modeling by Baldwin et al. (2003) indicated that isostatic
rebound of a thick crustal root may increase the longevity of a
range, but including a threshold value for critical shear stress
on the channel bed inhibits erosion significantly, and that
erosion rates decrease even further in weathering-limited, lowrelief environments. They find that inherent geomorphic factors can explain the persistence of residual topography for
hundreds of millions of years (Baldwin et al., 2003).
Another prominent example of persistent topography over an
ancient mountain range is in the Caldeonides of western Scandanavia. The orogen is the result of the Silurian collision of
Laurentia and Baltica (Soper et al., 1992; Cocks and Torsvik,
2002), followed by rifting and collapse in the late Paleozoic-early
Mesozoic (Dunlap and Fossen, 1998; Fossen and Dunlap, 1999;
Andersen et al., 1999; Mosar, 2003). However, peaks today reach
above 2000 m asl in both northern and southern Norway. Evidence for Cenozoic uplift of the Scandanavian Caledonides,
summarized by Nielsen et al. (2009), includes: (1) coarsening of
sedimentation in offshore basins; (2) AFT data that suggest
rapid exhumation starting at 30 Ma (Rohrman et al., 1995);
(3) PliocenePleistocene tilting of offshore strata away from the
mainland (Clausen et al., 1999; Faleide et al., 2002; Gabrielsen

et al., 2010) and overburial of sedimentary strata along the coast;
and (4) the presence of low-relief, concordant surfaces at high
elevation in the landscape, interpreted as an extensive surface
originally graded to sea level. Uplift in the Paleogene may be
linked to passage of the Icelandic Plume (Nadin et al., 1997;
Saunders et al., 1997; Clift et al., 1998; Dore et al., 2008). Evidence for active normal faulting also exists in the range, in jumps
in apatite fission-track ages across faults, and INSAR and earthquake data (Osmundsen et al., 2010). Overall, the long wavelength of the uplifted area however suggests mantle processes at
work (Cloetingh et al., 2005), such as asthenospheric diapirism
(Rohrman and van der Beek, 1996). However, Nielsen et al.
(2009) argued against any tectonic uplift in the Cenozoic. According to their isostacy-climate-erosion (ICE) hypothesis
(Nielsen et al., 2009), topography in the Scandinavian Caledonides is sustained by the deep crustal root given that a reduction of regional elevation by 1 km requires 56 km of
erosion, topography could be maintained for hundreds of millions of years if erosion rates are very low (on the order of
30 m Myr1, Nielsen et al., 2010). In this hypothesis, changes in
offshore sedimentation reflect climate change, and remnant
surfaces at high elevation do not reflect uplift of a peneplain,
but rather were formed in situ through glacial and periglacial
processes. Critics of this hypothesis point out the lack of correlation between Moho depth and topography, a mismatch between climate and sedimentation records (Chalmers et al., 2010),
and the presence of recent, deeply incised valleys and reexposed
relief (Lidmar-Bergstrom and Bonow (2009)). If the volume of
publication on this topic in recent years is any indication, this
matter is far from resolved.
5.3.7
Conclusion
Continentalcontinental collisions around the globe, active

and ancient, generate intense debate about tectonicclimate
coupling, tectonic geomorphology, and landscape evolution.
Younger or smaller collisions, such as in the Southern Alps of
New Zealand, the European Alps, or the Pyrenees, have been
described as behaving as frictional wedges. One important
topic that has been studied in terms of such orogens is the
concept of a steady state, whether in terms of topography or
exhumation. The Southern Alps appear to be close to or at an
exhumational steady state, but glacial erosion, and particularly
glacialinterglacial cycling, likely precludes the achievement of
a topographic steady state. The question remains unresolved in
the European Alps. Asymmetry in some of these ranges is
mostly tectonically controlled, particularly in the Pyrenees and
European Alps, but climate may play a role in the Southern
Alps. Range width, however, does seem to be sensitively tuned
to climate and erosion.
Larger, more extensive collisions occur further east along
the AlpineHimalayan chain, forming the TurkishIranian
Plateau, and Tibet and the Himalaya. These larger, hotter
orogens with thicker crust are associated with lateral and
vertical movements driven by lower crustal flow and processes
in the mantle lithosphere. In Turkey and Iran, collision of
Arabia and Eurasia has produced a moderately elevated, partially internally drained plateau characterized by large-scale
31
drainage reorganization. Lateral crustal movement is accommodated by strikeslip faults. Vertical uplift and extension of
the plateau region may result from delamination of the mantle
lithosphere. In Tibet, uplift of the plateau margins, possibly
the result of lower crustal flow, has also resulted in large-scale
drainage reorganization. In the Himalaya, climatetectonic
coupling appears to reach extremes; intense monsoon-related
erosion along the Himalayan front may drive exhumation of
channelized, lower crustal material from beneath Tibet, and
focused coupling in the eastern and western syntaxes may have
resulted in the development of tectonic aneurysms, in which
focused erosion drives the advection of hotter, and therefore
weaker, crust toward the surface in a narrow region.
In ancient orogens, topography may persist for hundreds
of millions of years after active tectonics cease in part because
of isostatic uplift as the mountain root is slowly eroded or
perhaps because of threshold conditions in fluvial erosion that
slow the erosion rate over time.
References
Adams, J., 1980. Contemporary uplift and erosion of the Southern Alps, New
Zealand: summary. GSA Bulletin Part I 91, 24.
Agard, P., Omrani, J., Jolivet, L., Mouthereau, F., 2005. Convergence history across
Zagros (Iran): constraints from collisional and earlier deformation. International
Journal of Earth Science 94, 401419.
Allen, M., Jackson, J., Walker, R., 2004. Late Cenozoic reorganization of the
ArabiaEurasia collision and the comparison of short-term and long-term
deformation rates. Tectonics 23, TC2008.
Allen, M.B., Armstrong, H.A., 2008. ArabiaEurasia collision and the forcing of
mid-Cenozoic global cooling. Palaeogeography, Palaeoclimatology,
Palaeoecology 265, 5258.
Alavi, M., 1994. Tectonics of the Zagros orogenic belt of Iran: new data and
interpretation. Tectonics 229, 211238.
Anderson, R.S., 2002. Modeling the tor-dotted crests, bedrock edges, and parabolic
profiles of high alpine surfaces of the Wind River Range, Wyoming.
Andersen, T.B., Torsvik, T.H., Eide, E.A., Osmundsen, P.T., Faleide, J.I., 1999.
Permian and Mesozoic extensional faulting within the Caledonides of Central
south Norway. Journal of the Geological Society 156, 10731080.
Armijo, R., Meyers, B., Hubert, A., Barka, A., 1999. Propagation of the North
Anatolian fault into the northern Aegean: timing and kinematics. Geology 27,
1153.
Babault, J., Van Den Driessche, J., Bonnet, S., 2005. Origin of the highly elevated
Pyrenean peneplain. Tectonics 24, TC2010.
Baldwin, J.A., Whipple, K.X., Tucker, G.E., 2003. Implications of the shear stress
river incision model for the timescale of postorogenic decay of topography.
Barletta, V.R., Ferrari, C., Diolaiuti, G., Carnielli, T., Sabadini, R., Smiraglia, C.,
2006. Glacier shrinkage and modeled uplift of the Alps. Geophysical Research
Letters 33, L14307.
Batt, G.E., 2011. The approach to steady-state thermochronological distribution
following orogenic development in the Southern Alps of New Zealand. American
Batt, G.E., Braun, J., 1999. The tectonic evolution of the Southern Alps,
New Zealand: Insights from fully thermally coupled dynamical modeling.
Geophysical Journal International 136, 403420.
Batt, G.E., Braun, J., Kohn, B.P., McDougall, I., 2000. Thermochronological analysis
of the dynamics of the Southern Alps, New Zealand. Geological Society of
Beamud, E., Munoz, J.A., Fitzgerald, P.G., Baldwin, S.L., Garces, M., Cabrera, L.,
Metcalf, J.R., 2011. Magnetostratigraphy and detrital apatite fission track
thermochronology in syntectonic conglomerates: constraints on the exhumation
of the South-Central Pyrenees. Basin Research 23, 309331.
explained by extrusion of a low-viscosity crustal channel coupled to focused
32
flows: 1. Numerical models with applications to the tectonics of the HimalayanTibetan orogen. Journal of Geophysical Research 109, B06406.
Beaumont, C., Kamp, P.J.J., Hamilton, J., Fullsack, P., 1996. The continental
collision zone, South Island, New Zealand: comparison of geodynamical models
and observations. Journal of Geophysical Research 101, 33333359.
Beaumont, C., Munoz, J.A., Hamilton, J., Fullsack, P., 2000. Factors controlling the
Alpine evolution of the central Pyrenees inferred from a comparison of
observations and geodynamical models. Journal of Geophysical Research 105,
81218145.
Becker, A., 2000. The Jura Mountains An active foreland fold-and-thrust belt?
Berberian, F., Muir, I.D., Pankhurst, R.J., Berberian, M., 1982. Late Cretaceous and
early Miocene Andean-type plutonic activity in northern Makran and Central
Iran. Journal of the Geological Society of London 139, 605614.
Berberian, M., 1983. The southern Caspian: a compressional depression floored by
a trapped, modified oceanic crust. Canadian Journal of Earth Science 20,
163183.
Berberian, M., 1995. Master blind thrust faults hidden under the Zagros folds:
active basement tectonics and surface morphotectonics. Tectonophysics 241,
193224.
Berberian, M., King, G.C.P, 1981. Towards a paleogeography and tectonic evolution
of Iran. Canadian Journal of Earth Sciences 18, 210285.
Berger, J.P., 1992. Paleontologie de la Molasse de Suisse Occidentale, Taxonomie,
Biostratigraphi, Paleoecologie, Paleogeographie et Paleoclimatologie [Habilitation
thesis]. Universite Fribourg, Fribourg, 405 p.
Bernet, M., Brandon, M., Garver, J., Balestieri, M.L., Venura, B., Zattin, M., 2009.
Exhuming the Alps through time: clues from detrital zircon fission-track
thermochronology. Basin Research 21, 781798.
Bernet, M., Zattin, M., Garber, J.I., Brandon, M.T., Vance, J.A., 2001. Steady-state
exhumation of the European Alps. Geology 29, 3538.
Bird, P., 1979. Continental delamination and the Colorado Plateau. Journal of
Bojar, A.-V., Fritz, H., Nicolescu, S., Bregar, M., Gupta, R.P., 2005. Timing and
mechanisms of Central Himalayan exhumation: discriminating between tectonic
and erosion processes. Terra Nova 17, 427433.
Bolliger, T., Engesser, B., Weidmann, M., 1993. Premiere decouverte de mammiferes
pliocenes dans le Jura neuchateois. Eclogae Geologicae Helvetiae 86,
10311068.
Bonnet, S., 2009. Shrinking and splitting of drainage basins in orogenic
landscapes from the migration of the main drainage divide. Nature Geoscience
2, 766771.
Bookhagen, B., Burbank, D.W., 2006. Topography, relief, and TRMM-derived rainfall
variations along the Himalaya. Geophysical Research Letters 33, L08405.
Bookhagen, B., Thiede, R.C., Strecker, M.R., 2005. Late Quaternary intensified
monsoon phases control landscape evolution in the northwest Himalaya.
Geology 33, 149152.
Bourquin, P., Suter, H., and Fallot, P., 1946. Geologischer Atlas der Schweiz,
Altasblatt 15, Blatter 114 Biaufond, 116 La Ferriere, 115 Les Bois, 117 St.
Imier: Wabern (Bern), Geologische Kommission der Schweizerischen
Naturforschenden Gesellschafter, Landesgeshydrologie und geologie.
Brookfield, M.E., 1998. The evolution of the great river systems of southern Asia
during the Cenozoic India-Asia collision: rivers draining southwards.
Burbank, D.W., Blythe, A.E., Putkonen, J., et al., 2003. Decopuling of erosion and
precipitation in the Himalaya. Nature 426, 652655.
Calvet, M., 1994. Morphogenese dune Montagne Mediterraneenne. Thesis,
Universite de Paris, Paris, 177 pp.
Carrapa, B., 2009. Tracing exhumation and orogenic wedge dynamics in the
European Alps with detrital thermochronology. Geology 37, 11271130.
Carrapa, B., Wijbrans, J., Bertotti, G., 2003. Episodic exhumation in the Western
Alps. Geology 31, 601604.
Cederbom, C.E., Sinclair, H.D., Schlunegger, F., Rahn, M.K., 2004. Climate-induced
rebound and exhumation of the European Alps. Geology 32, 709712.
Chalmers, J.A., Green, P., Japsen, P., Rasmussen, E.S., 2010. The Scandinavian
mountains have not persisted since the Caledonian orogeny. A comment on
Nielsen et al. (2009a). Journal of Geodynamics 50, 94101.
Chamberlain, C.P., Zeitler, P.K., Cooper, A.F., 1995. Geochronologic constrains of
the uplift and metamorphism along the Alpine Fault, South Island, New Zealand.
New Zealand Journal of Geology and Geophysics 38, 515523.
Champagnac, J.-D., Schlunegger, F., Norton, K., von Blanckenburg, F., Abbuhl,
L.M., Schwab, M., 2009. Erosion-driven uplift of the modern Central Alps.
Clark, M.K., Bush, J.W.M., Royden, L.H., 2005a. Dynamic topography produced
by lower crustal flow against rheological strength heterogeneities bordering
the Tibetan Plateau. Geophysical Journal International 162, 575590.
Clark, M.K., House, M.A., Royden, L.H., Whipple, K.X., Burchfiel, B.C., Zhang, X.,
Tang, W., 2005b. Late Cenozoic uplift of southeastern Tibet. Geology 33,
525528.
Clark, M.K, Royden, L.H., 2000. Topographic ooze: building the eastern margin of
Tibet by lower crustal flow. Geology 28, 703706.
Clark, M.K., Royden, L.H., Whipple, K.X., Burchfield, B.C., Zhang, X., Tang, W.,
2006. Use of a regional, relict landscape to measure vertical deformation of the
eastern Tibetan Plateau. Journal of Geophysical Research 111, F03002.
Clark, M.K., Schoenbohm, L.M., Royden, L.H., et al., 2004. Surface uplift, tectonics,
and erosion of eastern Tibet from large-scale drainage patterns. Tectonics 23,
TC1006.
Clausen, J.A., Gabrielsen, R.H., Reksnes, P.A., Nysaether, E., 1999. Development of
intraformational (Oligocene-Miocene) faults in the northern North Sea: influence
from remote stresses and doming of Fennoscandia. Journal of Structural
Geology 21, 14571475.
Clift, P.D., Blusztajn, J., 2005. Reorganization of the western Himalaya river system
after five million years ago. Nature 438, 10011003.
Clift, P.D., Carter, A., Hurford, A.J., 1998. The erosional and uplift history of NE
Atlantic passive margins: constraints on a passive plume. Journal of the
Clift, P.D., Hodges, K.V., Heslop, D., Hannigan, R., Long, H.V., Calves, G., 2008.
Correlation of Himalayan exhumation rates and Asian monsoon intensity. Nature
Geoscience 1, 875880.
Cloetingh, S., Ziegler, P.A., Beekman, F., et al., 2005. Lithospheric member, state of
stress and rheology: neotectonic controls on Europes intraplate continental
topography. Quaternary Science Reviews 24, 241304.
Cocks, L.R.M., Torsvik, T.H., 2002. Earth geography from 500 to 400 million years
ago: a faunal and palaeomagnetic review. Journal of the Geological Society of
London 159, 631644.
Coney, P.J., Munoz, J.A., McClay, K.R, Evenchick, C.A., 1996. Syntectonic burial
and post-tectonic exhumation of the southern Pyrenees foreland fold-thrust belt.
Journal of the Geological Society of London 153, 916.
Cook, K.L., Royden, L.H., 2008. The role of crustal strength variations in shaping
orogenic plateaus, with application to Tibet. Journal of Geophysical Research
113, B08407.
Copley, A., Jackson, J., 2006. Active Tectonics of the Turkish-Iranian Plateau.
Tectonics 25, TC6006.
Craddock, W.H., Kirby, E., Harkins, N.W., Zhang, H.P., Shi, X.H., Liu, J.H., 2010.
Rapid fluvial incision along the Yellow River during headward basin integration.
Nature Geoscience 3, 209213.
Cyr, A., Currie, B.S., Rowley, D.B., 2005. Geochemical and stable isotopic
evaluation of the Fenghuoshan Group lacustrine carbonates, north-central Tibet:
implications for the paleoaltimetry of Late Eocene Tibetan Plateau. The Journal
of Geology 113, 517533.
Dahlen, F.A., 1990. Critical taper model of fold- and -thrust belts and
accretionary wedges. Annual Reviews of Earth and Planetary Science Letters 18,
5599.
DeCelles, P.G., Quade, J., Kapp, P., Fan, M.J., Dettman, D.L., Ding, L., 2007. High
and dry in central Tibet during the Late Oligocene. Earth and Planetary Science
Letters 253, 389401.
Demir, T., Seyrek, A., Westaway, R., Bridgland, B., Beck, A., 2008. Late Cenozoic
surface uplift revealed by incision by the River Euphrates at Birecik, southeast
Turkey. Quaternary International 186, 132163.
Denton, G.H., Hughes, T.J., 1980. The Last Great Ice Sheets. John Wiley, Hoboken,
NJ, 484 pp.
Dewey, J.F., Hempton, M.R., Kidd, W.S.F, Saroglu, F., S-engor, A.M.C., 1986.
Shortening of continental lithosphere: the tectonics of eastern Anatolia a
young collisional zone. Geological Society Special Publication 19, 336.
Dewey, J.F., Pitmann, III W.C., Ryan, W.B.F., Bonnin, J., 1973. Plate tectonics and
the evolution of the Alpine system. Geological Society of America Bulletin 84,
31373180.
Dhont, D., Chorowicz, J., 2006. Review of the neotectonics of the Eastern TurkishArmenian Plateau by geomorphic analysis of digital elevation model imagery.
International Journal of Earth Science (Geol Rundsch) 95, 3449.
Dore, A.G., Lundin, E.R., Kusznir, N.J., Pascal, C., 2008. Potential mechanisms for
the genesis of Cenozoic domal structures on the NE Atlantic margin: pros, cons
and some new ideas. In: Johnson, H., Dore, A.G., Gatliff, R.W., Holdsworth, R.,
Lundin, E.R., Ritchie, J.D. (Eds.), The Nature and Origin of Compression in
Passive Margins. Geological Society of London Special Publication, London,
306, pp. 16.
Dortch, J.M., Dietsch, C., Owen, L.A., Caffee, M.W., Ruppert, K., 2011. Episodic
fluvial incision of rivers and rock uplift in the Himalaya and Transhimalaya.
Journal of the Geological Society, London 168, 783804.
Dortch, J.M., Owen, L.A., Haneberg, W.C., Caffee, M.W., Dietsch, C., Kamp, U.,
2009. Nature and timing of large landslides in the Himalaya and Transhimalaya
of northern India. Quaternary Science Reviews 28, 10371054.
Dunlap, W.J., Fossen, H., 1998. Early Palaeozoic orogenic collapse, tectonic
stability, and late Palaeozoic continental rifting revealed through
thermochronology of K-feldspars, southern Norway. Tectonics 17, 604620.
Dupont-Nivet, G., Lippert, P.C., van Hinsbergen, D.J.J., Meijers, M.J.M., Kapp, P.,
2010. Palaeolatitude and age of the Indo-Asia collision: palaeomagnetic
constraints. Geophysical Journal International 182, 11891198.
England, P.C., Houseman, G.A., 1988. The mechanics of the Tibetan Plateau.
Philosophical Transactions of the Royal Society of London A 326, 301320.
Erol, O, 1991. Geomorphological evolution of the Taurus Mountains, Turkey.
Zeitschrift fur Geomorphologie, Neue Folge, . Supplement 82), 99109.
Faccenna, C., Bellier, O., Martinod, J., Piromallo, C., Regard, V., 2006. Slab
detachment beneath eastern Anatolia: a possible cause for the formation of the
North Anatolian Fault. Earth and Planetary Science Letters 242, 8597.
Falcon, N.L., 1974. Southern Iran: Zagros Mountains. In: Spencer, A.M. (Ed.),
MesozoicCenzoic Orogenic Belts, Data for Orogenic Studies. Geological Society
of London Special Publications, London, pp. 199211.
Faleide, J.I., Kyrkjeb, R., Knennerud, T., Gabrielsen, R.H., Jordt, H., Fanavoll, S.,
Bjerke, M.D., 2002. Tectonic impact on sedimentary processes during Cenozoic
evolution of the northern North Sea and surrounding areas. In: Dore, A.G.,
Catwright, J.A., Stoker, M.S., Turner, J.P., White, N. (Eds.), Exhumation of the
North Atlantic Margin: Mechanisms and Implications for Petroleum Exploration.
Geological Society of London Special Publication 196, London, pp. 235269.
Finnegan, N.J., Hallet, B., Montgomery, D.R., Zeitler, P.K., Stone, J.O., Anders, A.M.,
Liu, Y.P., 2008. Coupling of rock uplift and river incision in the Namche BarwaGyala Peri massif, Tibet. Geological Society of America Bulletin 120, 142155.
Fitzgerald, P.G., Munoz, J.A., Coney, P.J., Baldwin, S.L., 1999. Asymmetric
exhumation across the Pyrenean orogen: implications for the tectonic
evolution of a collisional orogen. Earth and Planetary Science Letters 173,
157170.
Flesch, L.M., Holt, W.E., Silver, P.G., Stephenson, M., Wang, C.-Y., Chan, W.W.,
2005. Constraining the extent of crust-mantle coupling in central Asia using
GPS, geologic, and shear wave splitting data. Earth and Planetary Science
Letters 238, 248268.
Fossen, H., Dunlap, J., 1999. On the age and tectonic significance of PermoTriassic dikes in the Bergen-Sunnhordland region, southwestern Norway. Norsk
Geologisk Tidsskrift 79, 169177.
Frei, C., Schar, C., 1998. A precipitation climatology of the Alps from highresolution rain-gauge observations. International Journal of Climatology 18,
873900.
Gabrielsen, R.H., Faleide, J.I., Pascal, C., Braathen, A., Nystuen, J.P., Etzelmuller, B.,
ODonnell, S., 2010. Latest Caledonian to Present tectonomorphological
development of southern Norway. Marine and Petroleum Geology 27, 709723.
Garcia-Castellanos, D., Verges, J., Gaspar-Escribano, J., Cloetingh, S., 2003.
Interplay between tectonics, climate, and fluvial transport during the Cenozoic
evolution of the Ebro Basin (NE Iberia). Journal of Geophysical Research 108,
2347.
Garzanti, E., van Haver, T., 1988. The Indus clastics: Forearc basin sedimentation in
the Ladakh Himalaya (India). Sedimentary Geology 59, 237249.
Gibson, M., Sinclair, H.D., Lynn, G.J., Stuart, F.M., 2007. Late- to post-orogenic
exhumation of the Central Pyrenees revealed through combined
thermochronological data and modeling. Basin Research 19, 323334.
Gogu-s , O.H., Pysklywec, R.N., 2008. Mantle lithospheric delamination driving
plateau uplift and synconvergent extension in eastern Anatolia. Geology 36,
723726.
Gok, R., Pasyanos, M., Zor, E., 2007. Lithospheric structure of the continentcontinent collision zone: eastern Turkey. Geophysical Journal International 169,
10791088.
Green, O.R., Searle, M.P., Corfield, R.I., Corfield, R.M., 2008. Cretaceous-Tertiary
carbonate platform evolution and the age of the India-Asia collision along the
Ladakh Himalaya (northwest India). Journal of Geology 116, 331353.
Griffiths, G.A., McSaveney, M.J., 1983. Distribution of mean annual precipitation
across some steepland regions of New Zealand. New Zealand Journal of Science
26, 197208.
Grujic, D., 2006. Channel flow and continental collision tectonics: an overview.
In: Law, R.D., Searle, M.P., Godin, L. (Eds.), Channel Flow, Ductile Extrusion
and exhumation in Continental Collision Zones. Geological Society of London
Special Publications 268, London, pp. 2537.
33
Grujic, D., Coutand, I., Bookhagen, B., Bonnet, S., Blyth, A., Duncan, C., 2006.
Climatic forcing of erosion, landscape, and tectonics in the Bhutan Himalayas.
Geology 34, 801804.
Gudmundsson, G., 1994. An order-of-magnitude estimate of the current uplift-rates
in Switzerland caused by the Wurm Alpine deglaciation. Eclogae Geologicae
Hevetiae 87, 545557.
Gueguen, E., Doglioni, C., Fernandez, M., 1998. On the post-25 Ma geodynamic
evolution of the western Mediterranean. Tectonophysics 298, 259269.
Gunnell, Y., Calvet, M., 2006. Comment on Origin of the highly elevated Pyrenean
peneplain by Julien Babault, Jean Van Den Driessche, and Stephane Bonnet,
Sebastien Castelltort, and Alain Crave. Tectonics 25, TC3003.
Gunnell, Y., Calvet, M., Brichau, S., Carter, A., Aguilar, J.-P., Zeyen, H., 2009. Low
long-term erosion rates in high-energy mountain belts: insights from thermoand biochronology in the Eastern Pyrenees. Earth and Planetary Science Letters
278, 208218.
Gunnell, Y., Zeyen, H., Calvet, M., 2008. Geophysical evidence of a missing
lithospheric root beneath the Eastern Pyrenees: consequences for post-orogenic
uplift and associated geomorphic signatures. Earth and Planetary Science Letters
276, 302313.
Hack, T.J., 1960. Interpretation of erosional topography in humid temperate regions.
American Journal of Science 258-A, 8097.
Hallet, B., Molnar, P., 2001. Distorted drainage basins as markers of crustal strain
east of the Himalaya. Journal of Geophysical Research 106, 1369713709.
Haug, G.H., Tiedemann, R., 1998. Effect of the formation of the Isthmus of Panama
on Atlantic Ocean thermohaline circulation. Nature 393, 673676.
Herman, F., Braun, J., 2006. Fluvial response to horizontal shortening and
glaciations: a study of the Southern Alps of New Zealand. Journal of
Geophysical Research 111, F01008.
Herman, F., Braun, J., Dunlap, W.J., 2007. Tectonomorphic scenarios in the
Southern Alps of New Zealand. Journal of Geophysical Research 112, B040201.
Hodges, K.V., Wobus, C., Ruhl, K., Schildgen, T., Whipple, K., 2004. Quaternary
deformation, river steepening, and heavy precipitation at the front of the Higher
Himalayan ranges. Earth and Planetary Science Letters 7012, 111.
Holt, W.E., Wallace, T.C., 1990. Crustal thickness and upper mantle velocities in the
Tibetan plateau region from inversion of regional Pnl waveforms: evidence for a
thick upper mantle lid beneath southern Tibet. Journal of Geophysical Research
95, 1249912525.
Hsu, K.J., Montadert, L., Bernoulli, D., et al., 1977. History of Mediterranean
salinity crisis. Nature 267, 399403.
Huntington, K.W., Blythe, A.E., Hodges, K.V., 2006. Climate change and late
Pliocene acceleration of erosion in the Himalaya. Earth and Planetary Science
Letters 252, 107118.
Innocenti, F., Mazzuoli, R., Pasquare, G., Radicati di Brozolo, F., Villari, L., 1976a.
Evolution of volcanism in the area of interactions between the Arabian, Anatolian
and Iranian plates (Lake Van, Eastern Turkey). Journal of Volcanology and
Innocenti, F., Mazzuoli, R., Pasquare, G., Radicati di Brozolo, F., Villari, L., 1976b.
Tertiary and Quaternary volcanism in the Erzerum-Kars area (eastern Turkey):
geochronological data and geodynamic evolution. Journal of Volcanology and
Jackson, J., McKenzie, D., 1984. Active tectonics of the Alpine-Himalayan Belt
between western Turkey and Pakistan. Geophysics Journal Royal Astronomical
Society 77, 185264.
Jaffey, N., Robertson, A., 2005. Non-marine sedimentation associated with
Oligocene-Recent exhumation and uplift of the Central Taurus Mountains, S
Turkey. Sedimentary Geology 173, 5389.
Kapp, J.L.D., Harrison, T.M., Kapp, P., Grove, M., Lovera, O.M., Lin, D., 2005a.
Nyainqentanglha Shan: a window into the tectonic, thermal, and geochemical
evolution of the Lhasa block, southern Tibet. Journal of Geophysical Research
110, B08413.
Kapp, P., Yin, A., Harrison, T.M., Ding, L., 2005b. Cretaceous-Tertiary shortening,
basin development, and volcanism in central Tibet. Geological Society of
American Bulletin 117, 865878.
Kay, R.W., Kay, S.M., 1993. Delamination and delamination magmatism.
Kelly, M.A., Buoncristiani, J.F., Schluchter, C., 2004. A reconstruction of the last
glacial maximum (LGM) ice-surface geometry in the western Swiss Alps and
contiguous Alpine regions in Italy and France. Eclogae Geologicae Helvetiae 97,
5775.
Kent-Corson, M.L., Ritts, B.D., Zhuang, G.S., Bovet, P.M., Graham, S.A.,
Chamberlain, C.P., 2009. Stable isotopic constraints on the tectonic,
topographic, and climatic evolution of the northern margin of the Tibetan
Plateau. Earth and Planetary Science Letters 282, 158166.
34
Keskin, M., 2003. Magma generation by slab steepening and break off beneath
a subduction accretion complex: an alternative model for collisionrelated volcanism in Eastern Anatolia, Turkey. Geophysical Research Letters 30,
8046.
King, J., Harris, N., Argles, T., Parrish, R., Charlier, B., Sherlock, S., Zhang, H.F.,
2007. First field evidence of southward ductile flow of Asian crust beneath
southern Tibet. Geology 35, 727730.
Kirby, E., Whipple, K.X., Burchfiel, B.C., Tang, W.Q., Berger, G., Sun, Z.M., Chen,
Z.L., 2000. Neotectonics of the Min Shan, China: implications for mechanisms
driving Quaternary deformation along the eastern margin of the Tibetan Plateau.
Kirby, E., Whipple, K.X., Tang, W.Q., Chen, Z.L., 2003. Distribution of active
rock uplift along the eastern margin of the Tibetan Plateau: inferences from
bedrock channel longitudinal profiles. Journal of Geophysical Research 108,
2217.
Kohn, M.J., 2008. PTt data from central Nepal support critical taper and
repudiate large-scale channel flow of the Greater Himalayan Sequence.
Koons, P.O., 1990. Two-sided orogen: collision and erosion from the sandbox to
the Southern Alps, New Zealand. Geology 18, 679682.
Koons, P.O., Norris, R.J., Craw, D., Cooper, A.F., 2003. Influence of exhumation on
the structural evolution of transpressional plate boundaries: an example from the
Southern Alps, New Zealand. Geology 31, 36.
Korup, O., Densmore, A.L., Schlunegger, F., 2010a. The role of landslides in
mountain range evolution. Geomorphology 120, 7790.
Korup, O., Montgomery, D.R., 2008. Tibetan plateau river incision inhibited by
glacial stabilization of the Tsangpo gorge. Nature 455, 786790.
Korup, O., Montgomery, D.R., Hewitt, K., 2010b. Glacier and landslide feedbacks to
topographic relief in the Himalayan syntaxes. Proceedings of the National
Academy of Sciences 107, 53175322.
Korup, O., Schlunegger, F., 2009. Rock-type control on erosion-induced uplift,
eastern Swiss Alps. Earth and Planetary Science Letters 278, 278285.
Kuhlemann, J., 2000. Post-collisional sediment budget of circum-Alpine basins
(Central Europe). Memorie degli Istituti de Geologia e Mineralogia dell
Universia di Padova 52, 191.
Kuhlemann, J., Dunkl, I., Brugel, A., Spiegel, C., Frisch, W., 2006. From source
terrains of the Eastern Alps to the Molasse Basin: detrital record of non-steady
state exhumation. Tectonophysics 413, 301316.
Kuhlemann, J., Frisch, W., Dunkl, I., Szekely, B., 2001. Quantifying tectonic versus
erosive denudation by the sediment budget: the Miocene core complexes of the
Alps. Tectonophysics 330, 123.
Kuhlemann, J., Frisch, W., Szekley, B., Dunkl, I., Kazmer, M., 2002. Postcollisional
sediment budget history of the Alps: tectonic versus climatic control.
International Journal of Earth Sciences 91, 818837.
Kuhni, A., Pfiffner, O.A., 2001. The relief of the Swiss Alps and adjacent areas and
its relation to lithology and structure: topographic analysis from a 250-m DEM.
Lave, J., Avouac, J.P., 2001. Fluvial incision and tectonic uplift across the
Himalayas of central Nepal. Journal of Geophysical Research 106,
2656126591.
Leloup, P.H., Lacassin, R., Tapponnier, P., et al., 1995. The Ailao Shan-Red River
shear zone (Yunnan, China), Tertiary transform boundary of Indochina.
Lev, E., Long, M.D., van der Hilst, R.D., 2006. Seismic anisotropy in Eastern Tibet
from shear wave splitting reveals changes in lithospheric deformation. Earth and
Li, S., Mooney, W.d., 1998. Crustal structure of China from deep seismic sounding
profiles. Tectonophysics 288, 105113.
Lidmar-Bergstrom, K., Bonow, J.M., 2009. Hypotheses and observations on the
origin of the landscape of southern Norway a comment regarding the
Isostacy-climate-erosion hypothesis by Nielsen et al., 2008. Journal of
Geodynamics 48, 95100.
Lippitsch, R., Kissling, E., Ansorge, J., 2003. Upper mantle structure beneath the
Alpine orogen from high-resolution teleseismic tomography. Journal of
Little, T.A., Cox, S., Vry, J.K., Batt, G.E., 2005. Variations in exhumation level and
uplift rate along the oblique-slip Alpine Fault, central Southern Alps, New
Zealand. Geological Society of America Bulletin 117, 707723.
Long, S., McQuarrie, N., 2010. Placing limits on channel flow: insights from the
Bhutan Himalaya. Earth and Planetary Science Letters 290, 375390.
Lopez Blanco, M., Marzo, M., Burbank, D., Verges, J., Roca, E., Anadon, P., Pina,
J., 2000. Tectonic and climatic controls on the development of large, foreland
fan deltas: Montserrat and Sant Llorenc del Munt systems (middle Eocene, Ebro
basin, NE Spain). Sedimentary Geology 138, 1739.
Lyon-Caen, H., Molnar, P., 1989. Constraints on the deep structure and dynamic
processes beneath the Alps and adjacent regions from an analysis of gravity
anomalies. Geophysical Journal International 99, 1932.
Matmon, A., Bierman, P.R., Larsen, J., Southworth, S., Pavich, M., Caffee, M.,
2003. Temporally and spatially uniform rates of erosion in the southern
Appalachian Great Smoky Mountains. Geology 31, 155158.
McQuarrie, N., Stock, J.M., Verdel, C., Wernicke, B.P., 2003. Cenozoic evolution of
Neotethys and implications for the causes of plate motions. Geophysical
Meriaux, A.-S., Tapponnier, P., Ryerson, F.J., et al., 2005. The Aksay segment of the
northern Altyn Tagh fault: tectonic geomorphology, landscape evolution, and
Holocene slip rate. Journal of Geophysical Research 110, B04404.
Meteo-France, 1987. Atlas Climatique de la France. Paris, 30 pp.
Meyer, B., Tapponnier, P., Bourjot, L., et al., 1998. Crustal thickening in GansuQinghai, lithospheric mantle subduction and oblique, strike-slip controlled
growth of the Tibet Plateau. Geophysical Journal International 135, 147.
Mosar, J., 2003. Scandinavias North Atlantic passive margin. Journal of
Molnar, P., England, P., Martinod, J., 1993. Mantle dynamics, uplift of the Tibetan
Plateau, and the Indian monsoon. Reviews of Geophysics 31, 357396.
Molnar, P., Stock, J.M., 2009. Slowing of Indias convergence with Eurasia since
20 Ma and its implications for Tibetan mantle dynamics. Tectonics 28, TC3001.
Molnar, P., Tapponnier, P., 1975. Cenozoic tectonic of Asia: effects of a continental
collision. Science 189, 419426.
Monod, O., Kuzucuoglu, C., Okay, A.I., 2006. A Miocene palaeovalley network in
the Western Taurus (Turkey). Turkish Journal of Earth Sciences 15, 123.
Nadin, P.A., Kusznir, N.J., Cheadle, M.J., 1997. Early Tertiary plume uplift of the
North Sea and Faeroe-Shetland basins. Earth and Planetary Science Letters 148,
109127.
Najman, Y., Garzanti, E., Pringle, M., Bickle, M., Stix, J., Khan, I., 2003. EarlyMiddle Miocene paleodrainage and tectonics in the Pakistan Himalaya.
Naylor, M., Sinclair, H.D., 2008. Pro- vs. retro-foreland basins. Basin Research 20,
285303.
Nelson, K.D., Zhao, W., Brown, L.D., et al., 1996. Partially molten middle crust
beneath southern Tibet; synthesis of Project INDEPTH results. Science 274,
16841688.
Nicoll, K., 2010. Landscape development within a young collision zone:
implications for post-Tethyan evolution of the Upper Tigris River system in
southeastern Turkey. International Geology Review 52, 404422.
Nielsen, S.B., Clausen, O.R., Jacobsen, B.H., Thomsen, E., Huuse, M., Gallagher,
K., Balling, N., Egholm, D., 2010. The ICE hypothesis stands: How the dogma
of late Cenozoic tectonic uplift can no longer be sustained in the light of data
and physical laws. Journal of Geodynamics 50, 102111.
Nielsen, S.B., Gallagher, K., Leighton, C., et al., 2009. The evolution of western
Scaninavian topography: a review of Neogene uplift versus the ICE (Isostacyclimate- erosion) hypothesis. Journal of Geodynamics 47, 7295.
Nocquet, J.-M., Calais, E., 2004. Geodetic measurements of crustal deformation in
the Western Mediterranean and Europe. Pure and Applied Geophysics 161,
661681.
Norris, R.J., Cooper, A.F., 1997. Erosional control on the structural evolution of a
transpressional thrust complex on the Alpine fault, New Zealand. Journal of
Structural Geology 19, 13231342.
Norris, R.J., Koons, P.O., Cooper, A.F., 1990. The obliquely-convergent plate
boundary in the South Island of New Zealand: implications for ancient collision
zones. Journal of Structural Geology 12, 715725.
Norton, K.P., Abbuhl, L.M., Schlunegger, F., 2010. Glacial conditioning as an
erosional driving force in the Central Alps. Geology 38, 655658.
Ocakoglu, F., 2002. Palaeoenvironmental analysis of a Miocene basin in the high
Taurus Mountains (southern Turkey) and its palaeogeographical and structural
significance. Geological Magazine 139, 473487.
Ollier, C., Pain, C., 2000. The Origin of Mountains. Routledge, London, 345 p.
Osmundsen, P.T., Redfield, T.F., Hendriks, B.H.W., et al., 2010. Fault-controlled
alpine topography in Norway. Journal of the Geological Society of London 167,
8398.
Ouimet, W.B., Whipple, K.X., Royden, L.H., Sun, Z.M., Chen, Z.L., 2007. The
influence of large landslides on river incision in a transient landscape: eastern
margin of the Tibetan Plateau (Sichuan, China). Geological Society of America
Bulletin 119, 14621476.
Pazzaglia, F.J., Brandon, M.T., 1996. Macrogeomorphic evolution of the postTriassic Appalachian mountains determined by deconvolution of the offshore
basin sedimentary records. Basin Research 8, 255278.
Pazzaglia, F.J., Gardner, T.W., 1994. Late Cenozoic flexural deformation of the
middle U.S. Atlantic passive margin. Journal of Geophysical Research 99,
12,14312,157.
Pearce, J.A., Bender, J.F., De Long, S.E., et al., 1990. Genesis of collision
volcanism in eastern Anatolia, Turkey. Journal of Volcanology and Geothermal
Penck, A., 1894. Studien uber das Klima Nordspaniens wahrend der jungeren
Terciarperiod und der diluvialperiode. Zeitschfrift Ges. Erdkunde Berlin 29,
109141.
Persaud, M., Pfiffner, O.A., 2004. Active deformation in the eastern Swiss
Alps: post-glacial faults, seismicity and surface uplift. Tectonophysics 385,
5984.
Pfiffner, O.A., Schlunegger, F., Buiter, S.J.H., 2002. The Swiss Alps and their
peripheral foreland basin: stratigraphic response to deep crustal processes.
Tectonics 21, 1009.
Pratt, B., Burbank, D.W., Heimsath, A., Ojha, T., 2002. Impulsive alluviation during
early Holocene strengthened monsoons, central Nepal Himalaya. Geology 30,
911914.
Pysklywec, R.N., 2006. Surface erosion control on the evolution of the deep
lithosphere. Geology 34, 225228.
Raymo, M.E., Ruddiman, W.F., Froelich, P.N., 1988. Influence of late Cenozoic
mountain building on ocean geochemical cycles. Geology 16, 649653.
Reiners, P.W., Brandon, M.T., 2006. Using thermochronology to understand
orogenic erosion. Annual Reviews of Earth and Planetary Sciences 34,
419466.
Rey, P.F., Teyssier, C., Whitney, D.L., 2010. Limit of channel flow in orogenic
plateau. Lithosphere 2, 328332.
Robertson, A.H.F., Grasso, M., 1995. Overview of the Late Tertiary-Recent tectonic
and palaeo-environmental development of the Mediterranean region. Terra Nova
7, 114127.
Robinson, D.M., DeCelles, P.G., Copeland, P., 2006. Tectonic evolution of the
Himalayan thrust belt in western Nepal: implications for channel flow models.
Roca, E., 1996. The Neogene Cerdanya and Seu dUrgell intramontane basins
(eastern Pyrenees). In: Friend, P.F., Dabrio, C.J. (Eds.), Tertiary Basins of Spain:
The Stratigraphic Record of Crustal Kinematics. Cambridge University Press,
Roe, G.H., Stolar, D.B., Willett, S.D., 2006. Response of a Steady-State Critical
Wedge Orogen to Changes in Climate and Tectonic Forcing. Geological Society
of America Special Paper 398, pp. 227239.
Roest, W.R., Srivastava, S.P., 1991. Kinematics of the plate boundaries between
Eurasia, Iberia and Africa in the North Atlantic from the late Cretaceous to the
present. Geology 19, 613616.
Rohrman, M., van der Beek, P., 1996. Cenozoic postrift domal uplift of North
Atlantic margins: an asthenospheric diapirism model. Geology 24, 901904.
Rohrman, M., van der Beek, P.A., Andriessen, P.A.M., Cloetingh, S., 1995. MesoCenozoic morphotectonics evolution of southern Norway: neogene domal uplift
inferred from apatite fission track thermochronology. Tectonics 14, 704718.
Rosenberg, C.L., Berger, A., 2009. On the causes and modes of exhumation and
lateral growth of the Alps. Tectonics 28, TC6001.
Rowley, D.B., 1996. Age of initiation of collision between India and Asia: a review
of stratigraphic data. Earth and Planetary Science Letters 145, 113.
Rowley, D.B., Currie, B.S., 2006. Palaeo-altimetry of the late Eocene to Miocene
Lunpola basin, central Tibet. Nature 439, 677681.
Rowley, D.B., Garzione, C.N., 2007. Stable isotope-based paleoaltimetry. Annual
Royden, L.H., Burchfiel, B.C., King, R.W., Wang, E., Chen, Z., Shen, F., Liu, Y.,
1997. Surface deformation and lower crustal flow in eastern Tibet. Science 276,
788790.
Saunders, A.D., Fitton, J.G., Kerr, A.C., Norry, M.J., Kent, R.W., 1997. The north
Atlantic Igneous Province. Geophysical Monograph 100, 4593.
Schlunegger, F., Hinderer, M., 2001. Crustal uplift in the Alps: why the drainage
pattern matters. Terra Nova 13, 425432.
Schlunegger, F., Simpson, G., 2002. Possible erosional control on lateral growth of
the European Central Alps. Geology 30, 907910.
Schmid, S.M., Fugenschuh, B., Kissling, E., Schuster, R., 2004. Tectonic map and
overall architecture of the Alpine orogen. Eclogae Geologicae Helvetiae. 87,
93117.
Schmid, S.M., Kissling, E., 2000. The arc of the western Alps in the light of
geophysical data on deep crustal structure. Tectonics 19, 6285.
35
Schmid, S.M., Pfiffner, O.A., Froitzheim, N., Schonborn, G., Kissling, E., 1996.
Geophsycialgeological transect and tectonic evolution of the Swiss-Italian Alps.
Tectonics 15, 10361064.
Schoenbohm, L.M., Burchfiel, B.C., Chen, L.Z., 2006. Propagation of surface uplift,
lower crustal flow, and Cenozoic tectonics of the southeast margin of the
Tibetan Plateau. Geology 34, 813816.
Schoenbohm, L.M., Whipple, K.X., Burchfiel, B.C., Chen, L., 2004. Geomorphic
constraints on surface uplift, exhumation, and plateau growth in the Red River
region, Yunnan Province, China. Geological Society of America Bulletin 116,
895909.
Schumer, R., Jerolmack, D.J., 2009. Real and apparent changes in sediment
deposition rates through time. Journal of Geophysical Research 114, F00A06.
Schwarb, M., Daly, C., Frei, C., Schar, C., 2001. Mean Annual and Seasonal
Precipitation in the European Alps 19711990. Institue of Geography of the
University of Berne, Hydrological Atlas of Switzerland: Berne, Switzerland.
Searle, M.P., Szulc, A.G., 2004. Channel flow and ductile extrusion of the high
Himalayan slab the Kangchenjunga-Darjeeling profile, Sikkim Himalaya.
Journal of Asian Earth Sciences 25, 173185.
S- engor, A.M.C., 1979. The North Anatolian transform fault: its age, offset and
tectonic significance. Journal of the Geological Society of London 136,
269282.
S- engor, A.M.C., Gurur, N., Saroglu, F., 1985. Strike slip faulting and related basin
formation in zones of tectonic escape: Turkey as a case study. In: Biddle, K.T.,
Christie-Blick, N. (Eds.), Strike Slip Faulting and Basin Formation. Spec. Publ.
Econ. Paleontol. Mineral., vol. 37, pp. 227264.
S- engor, A.M.C., Ozeren, S., Zor, E., Genc, T., 2003. East Anatolian high plateau as
a mantle-supported, N-S shortened domal structure. Geophysical Research
Letters 30, 8045.
Seward, D., Burg, J.-P., 2008. Growth of the Namche Barwa Syntaxis and
associated evolution of the Tsangpo Gorge: constraints from structural and
thermochronological data. Tectonophysics 451, 282289.
Shroder, J.F., Bishop, M.P., 2000. Unroofing the Nanga Parbat Himalaya. In: Asif
Khan, M., Treloar, P.J., Searle, M.P., Qasim Jan, M. (Eds.), Tectonics of the
Western Himalaya and Karakoram. Geological Society of London, Special
Publication 170, pp. 163179.
Sinclair, H.D., Gibson, M., Lynn, G., Stuart, F., 2009. The evidence for a
Pyrenean Resurrection: a response to Babault et al., 2008. Basin Research 21,
143145.
Sinclair, H.D., Gibson, M., Naylor, M., Morris, R.G., 2005. Asymmetric growth of
the Pyrenees revealed through measurement and modeling of orogenic fluxes.
American Journal of Science 305, 369406.
Small, E.E., Anderson, R.S., Hancock, G.S., Finkel, R.C., 1999. Estimates of regolith
production from 0Be and 26Al: Al: evidence for steady state alpine hillslopes.
Soons, J.M., 1979. Late Quaternary environments in the central South Island of
New Zealand. New Zealand Geography 25, 1623.
Soper, N.J., Strachan, R.A., Holdsworth, R.E., Gayer, R.A., Greiling, R.O., 1992.
Sinistral transpression and the Silurian closure of Iapetus. Journal of the
Stocchi, P., Spada, G., Gianetti, S., 2005. Isostatic rebound following the Alpine
deglaciation: impact on the sea level variations and vertical movements in the
Mediterranean region. Geophysical Journal International 162, 137147.
Stock, J., Molnar, P., 1982. Uncertainties in the relative positions of the Australia,
Antarctica, Lord Howe, and Pacific plates since the Late Cretaceous. Journal of
Stocklin, J., 1968. Structural history and tectonics of Iran: a review. American
Association of Petroleum Geologists Bulletin 52, 12291258.
Stuwe, K., Robl, J., Hergarten, S., Evans, L., 2008. Modeling the influence of
horizontal advection, deformation and late uplift on the drainage development in
the India-Asia collision zone. Tectonics 27, TC6011.
Sutherland, R., 1996. Transpressional development of the Australia-Pacific boundary
through southern South Island, New Zealand: constrains from Miocene-Pliocene
sediments, Waiho-1 borehole, South Westland. New Zealand Journal of Geology
and Geophysics 39, 251264.
Tapponnier, P., Lacassin, R., Leloup, P.H., et al., 1990. The Ailao Shan/Red River
metamorphic belt: tertiary left-lateral shear between Indochina and south China.
Nature 343, 431437.
Tapponnier, P., Molnar, P., 1979. Active faulting and Cenozoic tectonics of the Tien
Shan, Mongolian and Baykal regions. Journal of Geophysical Research 84,
34253459.
Tapponnier, P., Peltzer, G., Armijo, R., Le Dain, A.-Y., Cobbold, P., 1982.
Propagating extrusion tectonics in Asia: new insights from simple experiments
with plasticine. Geology 10, 611616.
36
Tapponnier, P., Xu, Z.-Q., Roger, F., Meyer, B., Arnaud, N., Wittlinger, G.,
Yang, J.-S., 2001. Oblique stepwise rise and growth of the Tibetan plateau.
Science 294, 16711677.
Thiede, R.C., Arrowsmith, R.J., Bookhagen, B., McWilliams, M.O., Sobel, E.R.,
Strecker, M.R., 2005. From tectonically to erosionally controlled development of
the Himalayan orogen. Geology 33, 689692.
Thiede, R.C., Ehlers, T.A., Bookhagen, B., Strecker, M.R., 2009. Erosional variability
along the northwest Himalaya. Journal of Geophysical Research 114, F01015.
Tippett, J.M., Kamp, P.J.J., 1993. Fission track analysis of the Late Cenozoic
vertical kinematics of continental Pacific crust, South Island, New Zealand.
Tomkin, J.H., 2003. Erosional feedbacks and the oscillation of ice masses. Journal
of Geophysical Research 108, 2488.
van der Beek, P., Bourbon, P., 2008. A quantification of the glacial imprint on relief
development in the French western Alps. Geomorphology 97, 5275.
Vernant, Ph., Nilforoushan, F., Hatzfeld, D., et al., 2004. Present-day crustal
deformation and plate kinematics in the Middle East constrained by GPS
measurements in Iran and northern Oman. Geophysical Journal International
157, 381398.
Vernon, A.J., van der Beek, P.A., Sinclair, H.D., 2009. Spatial correlation between
long-term exhumation rates and present-day forcing parameters in the western
European Alps. Geology 37, 859862.
Vernon, A.J., van der Beek, P.A., Sinclair, H.D., Rahn, M.K., 2008. Increase in
late Neogene denudation of the European Alps confirmed by analysis of fissiontrack thermochronology database. Earth and Planetary Science Letters 270,
3163293.
von Eynatten, H., Schlunegger, F., Gaupp, R., Wijbrans, J.R., 1999. The exhumation
of the Lepontine: evidence from 39Ar/40Ar laserprobe dating of detrital white
micas from the Swiss Molasse Basin. Terra Nova 11, 284289.
Walcott, R.I., 1978. Present tectonics and late Cenozoic evolution of New Zealand.
Geophysical Journal Royal Astronomical Society 52, 137164.
Walcott, R.I., 1998. Modes of oblique compression: late Cenozoic tectonics of the
South Island of New Zealand. Reviews of Geophysics 36, 126.
Wang, E., Burchfiel, B.C., 1997. Interpretation of Cenozoic tectonics in the rightlateral accommodation zone between the Ailao Shan shear zone and the Eastern
Himalayan Syntaxis. International Geology Review 39, 191219.
Wang, E., Burchfiel, B.C., Royden, L.H., Chen, L., Chen, J., Li, W., Chen, Z., 1998.
Late Cenozoic Xianshuihe-Xiaojiang, Red River and Dali fault systems of
southwestern Sichuan and central Yunnan, China. Geological Society of America
Special Paper, Boulder 327, 108 pp.
Wang, Y., Deng, T., Biassatti, D., 2006. Ancient diets indicate significant uplift of
southern Tibet after ca. 7 Ma. Geology 34, 309312.
Westaway, R., 1994. Present-day kinematics of the Middle East and eastern
Mediterranean. Journal of Geophysical Research 99, 1207112090.
Westaway, R., 2003. Kinematics of the Middle East and Eastern Mediterranean
updated. Turkish Journal of Earth Sciences 12, 546.
Whipple, K.X., 2001. Fluvial landscape response time: how plausible is steady-state
denudation? American Journal of Science 301, 313325.
Whipple, K.X., Meade, B.J., 2004. Controls on the strength of coupling among
climate, erosion and deformation in two-sided, frictional orogenic wedges at
steady state. Journal of Geophysical Research 109, F01011.
Whitehouse, I.E., 1987. Geomorphology of a compressional plate boundary,
Southern Alps, New Zealand. In: Gardiner, V. (Ed.), International Geomorphology.
Wiley, Chichester, UK, pp. 897924.
Whittmann, H., von Blanckenburg, F., Kruesmann, T., Norton, K.P., Kubik, P.W.,
2007. The relation between rock uplift and denudation from cosmogenic
nuclides in river sediment in the Central Alps of Switzerland. Journal of
Geophysical Research 112, F04010.
Willet, S.D., 1999. Orogeny and orography: the effects of erosion on the structure
of mountain belts. Journal of Geophysical Research 104, 2895728981.
Willet, S.D., Brandon, M.T., 2002. On steady state in mountain belts. Geology 30,
175178.
Willet, S.D., Schlunegger, F., Picotti, V., 2006. Messinian climate change and
erosional destruction of the central European Alps. Geology 34, 513616.
Willet, S.D., Slingerland, R., Hovius, N., 2001. Uplift, shortening, and steady sate
topography in active mountain belts. American Journal of Science 201,
455485.
Wobus, C., Heimsath, A., Whipple, K., Hodges, K., 2005. Active out-of-sequence
thrust faulting in the central Nepalese Himalaya. Nature 434, 10081011.
Wobus, C., Pringle, M., Whipple, K., Hodges, K., 2008. A Late Miocene
acceleration of exhumation in the Himalayan crystalline core. Earth and Planetary
Wright, T.J., Parsons, B., England, P.C., Fielding, E.J., 2004. InSAR observations of
low slip rates on the major faults of western Tibet. Science 305, 236239.
Yin, A., 2006. Cenozoic tectonic evolution of the Himalayan orogen as constrained
by along-strike variation of structural geometry, exhumation history, and foreland
sedimentation. Earth Science Reviews 76, 1131.
Yin, A., Harrison, T.M., 2000. Geologic evolution of the Himalayan-Tibetan orogen.
Annual Reviews of Earth and Planetary Science 28, 211280.
Zeitler, P.K., Meltzer, A.S., Koons, P.O., et al., 2001. Erosion, Himalayan
geodynamics and the geomorphology of metamorphism. GSA Today 11, 49.
Zhou, D., Graham, S.A., 1996. Extrusion of the Altyn Tagh wedge: a kinematic
model for the Altyn Tagh fault and palinspastic reconstruction of northern China.
Geology 24, 427430.
Zhu, B., Kidd, W.S.F., Rowley, D.B., Currie, B.S., Safique, N., 2005. Age of initiation
of the India-Asia collision in the east-central Himalaya. Journal of Geology 113,
265285.
Zor, E., Sandvol, E., Gurbuz, C., Turkelli, N., Seber, D., Barazangi, M., 2003. The
crustal structure of the East Anatolian plateau from receiver functions.
Biographical Sketch
Lindsay Schoenbohm earned an undergraduate degree in Geology from Carleton College in Northfield, Minnesota, in 1997. She completed her PhD from MIT in Geology in 2004 on the tectonics and geomorphology of
the southeast margin of the Tibetan Plateau. She went on to a postdoc at the University of Potsdam in Germany,
and then an assistant professorship at the Ohio State University in 2006, until moving to the University of
Toronto in 2009.
5.4 Transform Plate Margins and Strikeslip Fault Systems

KL Frankelw, Georgia Institute of Technology, Atlanta, GA, USA
5.4.1
Introduction
5.4.2
General Tectonic Setting
5.4.3
Advances in Studying Continental Transform Systems
5.4.3.1
Geochronology and Thermochronology
5.4.3.2
Remote Sensing
5.4.3.3
Tectonic Geodesy
5.4.4
Major Continental Transform Plate Boundaries and Strikeslip Fault Systems
5.4.4.1
Pacific-North America Plate Boundary
5.4.4.1.1
San Andreas Fault
5.4.4.1.2
Eastern California shear zone-Walker Lane: An Evolving Plate Boundary
5.4.4.1.3
Garlock Fault: An Intracontinental Transform
5.4.4.2
Pacific-Australian Plate Boundary
5.4.4.3
Caribbean-North American Plate Boundary
5.4.4.4
Himalayan-Tibetan Orogen Strikeslip Systems
5.4.4.5
Alaska
5.4.4.6
North and Eastern Anatolian-Dead Sea Fault System
5.4.4.7
Miscellaneous
5.4.5
Important Questions and Future Directions
5.4.5.1
Constancy of Seismic Strain Accumulation and Release
5.4.5.2
Quantifying Deformation Rates
5.4.5.3
Calibrating and Cross-Checking Geochronometers
5.4.6
Conclusions
Acknowledgments
References
Glossary
Dextral motion Right lateral displacement.
Digital elevation model A three-dimensional
representation of a terrains surface.
Exhumation The uncovering or exposure of rock through
erosion.
Fault trenching Excavated trench across a fault used to
help determining the timing of past earthquakes.
Fluvial (river) terraces Remnants of a former floodplain
and/or riverbed that are preserved above the present river
level as a consequence of uplift and/or incision.
Geochronology Measurement of time intervals or dating
of events on the geological time scale.
Global Positioning System (GPS) A space-based global
on Earth.
38
39
43
44
45
46
47
47
48
50
55
55
55
55
57
57
57
57
58
59
63
65
66
66
Holocene Geologic epoch that began at 11 700 years ago

and continues to the present.
Interferometric synthetic aperture radar (InSAR) A
satellite radar technique that uses two or more synthetic
aperture radars (SAR) to produce maps of surface
deformation or digital elevation by comparing differences
in the phase of waves returning to the satellite.
Intracontinental Pertaining to the interior of a continent.
Light Detection And Ranging (LiDAR or lidar) An
optical remote sensing technology, ground based or
airborne, that can measure distance, which is commonly
used to produce digital elevation models.
Marine terrace Former beach or shoreline that is above
present sea level formed by uplift and/or changes in global
sea level.
Moment magnitude (Mw) A measurement of the energy
released during an earthquake that is proportional to the
slip on a fault times the area of the fault surface that has
Frankel, K.L., Owen, L.A., 2013. Transform plate margins and strikeslip
fault systems. In: Shroder, J. (Editor in Chief), Owen, L.A. (Ed.), Treatise on
Geomorphology. Academic Press, San Diego, CA, vol. 5, Tectonic
w
Deceased.
http://dx.doi.org/10.1016/B978-0-12-374739-6.00082-8
37
38
Transform Plate Margins and Strikeslip Fault Systems
slipped. The scale is logarithmic, such that for every increase

of 1 Mw is equivalent to a total energy increases of 32 times
the smaller magnitude.
Neotectonics Study of the present motion and
deformation of Earths crust and over recent geologic time
(usually last few million years).
Optically stimulated luminescence (OSL) dating A
method used to determine the age of sediment using the
acquired luminescence signal induced in minerals by ionizing
radiation in the ground and from cosmic ray particles.
Orogen A mountain belt.
Paleoearthquakes Past earthquakes.
Paleoseismic Pertaining to past earthquakes.
Paleoseismology Study of past earthquakes.
Pleistocene Geological period that spans approximately
the last 2.6 million years to approximately 11 600 years
before present.
Pressure ridge A linear topographic feature produced by
uplift due to active faulting.
Probability density functions (PDF) A function that
describes the relative likelihood for a random variable to
occur at a given point.
Radiocarbon (14C) dating Method used to determine the
age of material containing carbon (usually organic)
by measuring the decay of the 14C isotope within the

material.
Sag ponds Small pond or lake produced by subsidence
related to extension along a strike-slip fault.
Shutter-ridge An isolated ridge formed by displacement
of a spur by strike-slip faulting.
Strain Deformation of material resulting from an applied
force.
Terrestrial cosmogenic nuclides (TCN) Nuclides
produced by the interaction of cosmic ray particles with
target atoms (usually in rock or sediment) at Earths surface,
which include such nuclides as 10Be, 14C, 26Al and 36Cl.
TCNs are used to date surfaces or buried sediment.
Thermochronology Study of the thermal evolution of a
region of a planet.
Transpression Pertaining to compression due to a bend
in the length of a strike-slip fault.
Transtension Pertaining to extension due to a bend in the
length of a strike-slip fault.
Triple junction The location where three plate boundary
meet.
U-series geochronology Dating of biogenic and
inorganic carbon that involves three U-series decay chains.
Abstract
Tectonic landforms associated with transform plate margins and strikeslip faults vary from the plate boundary scale to the
regional and local scale, and they help provide an understanding of the dynamics and the nature of structural and
landscape development. Classic examples of these plate margins and fault systems include: the Pacific-North American plate
boundary incorporating the San Andreas fault and the eastern California shear zone-walker Lane; and the Pacific-Australian
plate boundary; the Caribbean-North American plate boundary; and intracontinental strikeslip systems of the HimalayanTibetan orogen. New advances in remote sensing, geochronology, and tectonic geodesy are helping to better resolve
patterns of deformation and landscape development along these fault systems.
5.4.1
Introduction
Transform plate boundaries, also known as conservative plate

boundaries, occur where lithospheric plates slide past each
other and where the crust is neither destroyed nor formed.
These plate boundaries are commonly associated with, and
evolve as a consequence of, the accommodation of displacement along spreading ridges. The motion along a transform
plate boundary typically occurs along major transform faults,
which on continents are commonly referred to as strikeslip
faults. These strikeslip faults range in size from a few tens of
meters to over a thousand kilometers in length (Figure 1).
Continental transform faults play a critical role in accommodating strain along major tectonic plate boundaries. These
strikeslip faults and their associated structures are defined by
linkage of mid-ocean spreading centers and subduction zones,
and are exposed onshore as they traverse the continental crust
between mid-ocean ridges and other plate boundary segments
(Figure 2). Often times, due to the predominantly horizontal
motion along transform faults, the geomorphic expression of
these structures is subtle as there is generally little or no surface or rock uplift driving base level fall and river incision.
However, continental transform faults generally have segments

that are well-expressed geomorphically and it is these regions
that tend to offer the greatest insight into transform plate
boundary deformation.
This chapter reviews current knowledge on the tectonic
geomorphology of transform plate boundaries and major
strikeslip faults. The tectonic landforms associated with these
tectonic settings are examined from the plate boundary scale
to the regional and local scale to provide an understanding of
the dynamics and the nature of landscape development along
transform plate boundaries and strikestrike faults. Major
tectonic regimes in a variety of geographic regions are examined to illustrate the variety and complexity of tectonic processes, landforms, and landscape development in these
settings. In addition, new advances in remote sensing, geochronology, and tectonic geodesy are discussed as they pertain
to better resolving patterns of deformation and landscape
development along transform plate boundaries and strikeslip
faults. We conclude the chapter with a discussion by outlining
important unresolved questions and future directions of tectonogeomorphic research along transform plate boundaries
and strikeslip fault systems.
39
BATHURST F.
MACDONALD F.
CARPATHIANS
Dzungarian
North
ALPS
F.
Anatolian
Bolnai F.
Fergano
Kamchatka F.
Altun
F.
Kansu
Herat F. Tagh F.
F.
Median
Kopet
Tectonic
Levant Dagh F.
Kang
ANTALYA
Line
Karakorum F.
F.Z.
Ting
COMPLEX
Longitudinal F.
ARAKAPAS F.
Chaman F.
Red River F.
Gulf
Owen
Philippine F.
of Guinea
F.Z.
Sorong F.Z.
Afar
Sangaing
New
Hebrides
Andaman Sea
Bogota
Semangko
penninsula
system
New
Caledonia
Sillon
Alpine F.
AGULHAS
F.Z.
COSTAL
COMPLEX
Chugach
Fairweather
Queen Charlotte
Island system
San Andreas
system
MAGDALEN
BASIN
KINGSKAWEATH
Middle
America
Trench
Gibbs F.Z.
CANTABRIANS
Oceanographer F.Z.
AzoreAtlantis F.Z.
Gilbraltar F.Z.
Northern
Caribbean
Kane
Zone
Oca F. Verna
F.Z.
St. Pauls
F.Z.
El Pilar
system
Romanche F.Z.
DoloresAscension
Guayaquil F.
F.Z.
and Bocono F.
ATACAMA F.
Hikurangi
Margin
Falklands
F.Z.
Menard F.Z.
Eltanin F.Z.
Normal
Oblique
Ridge-transform systems
Other divergent boundaries
} Trenches
Boundary transforms
Other active strike-slip
faults (e.g. Kansu F.)
Ancient strike-slip faults

(e.g. ARKAPAS F.)
Figure 1 Location of major transform plate boundaries and strikeslip fault systems. Adapted from Woodcock, N., 1986. The role of strikeslip
fault systems at plate boundaries. Philosophical Transactions of the Royal Society 317, 1329, and Molnar, P., Dayem, K.E., 2010. Major
intracontinental strikeslip faults and contrasts in lithospheric strength. Geosphere 6, 444467.
Trench
Indent-linked
Strike-slip fault
Back-arc basin
Trench-linked
Strike-slip fault
Continental
crust
Arc
Boundary
transform
Ridge
Oceanic crust
Ridge-transform
Figure 2 Tectonic setting of major classes of transform faults. Adapted from Woodcock, N., 1986. The role of strikeslip fault systems at plate
boundaries. Philosophical Transactions of the Royal Society 317, 1329.
5.4.2
General Tectonic Setting
Earths major transform plate boundaries and strikeslip faults

are shown in Figure 1. Each of these locations links mid-ocean
ridge or subduction zone segments as is shown in Figure 2,

which summarizes the plate-tectonic settings for the major
classes of transform faults. The structures associated with these
transform plate boundaries and intracontinental strikeslip
40
faults produce a unique geomorphic expression on the

landscape.
Regions where streams flow orthogonal to the plate boundary generally provide the richest record of deformation along
strikeslip faults. In addition, places where there are bends in
the transform systems result in vertical motion expressed as
reverse faulting, folding, rock and surface uplift or subsidence,
and extension (Figure 3). These areas also commonly preserve
a unique record of tectonic activity in the form of pressure
ridges, sag ponds, shutter ridges, and marine and fluvial terraces. Looking even deeper into the landscape can also provide
geomorphic clues to the deformational history along continental transform faults as various topographic metrics derived
from streams and drainage basins can be useful indicators of
the rates and patterns of continental transform activity.
The geomorphic expression of deformation along continental transform faults can be expressed at scales ranging from a
few meters to hundreds of kilometers (Figures 4 and 5). In
many cases, lineaments, such as the San Andreas and Altyn
Tagh faults, can be observed from space stretching many
hundreds of kilometers (Figures 4(a) and 5). Individual
ground-rupturing earthquakes can leave relatively subtle imprints on the landscape in the form of fault scarps or offset
streams and landforms (Figure 4(b)(e)). Repeated seismic
activity over Holocene and late Pleistocene timescales generally leads to landscapes that preserve a record of 10s to 100s
of meters of offset (Figure 4(b)(d)). On the million-year
timescales over which plate boundary transform faults usually
persist, it is not uncommon for the landscape to record many
kilometers of strikeslip displacement and/or rock uplift.
The trends of strikeslip faults commonly curve along their
length, resulting in zones of transpression and transtension,
which in turn produce their own characteristic landforms
(Christie-Blick and Biddle, 1985; Sylvester, 1988; Figure 3).

The San Andreas fault provides good examples of variation of
transtension and transpression along its length (Figure 6).
Steep hillslopes and high mountains are common in zones of
transpression. Classic examples include the Transverse Ranges
in Southern California that form along the San Andreas faults
and the Gobi Altai Mountains in Mongolia that have formed
along the North Gobi Altai and Gobi-Tien Shan fault systems
(Meisling and Weldon, 1989; Cunningham et al., 1997;
Spotila et al., 2007a, b). In contrast, basins are common
within zones of transtension. For example, the Death Valley,
Dead Sea, and Turfan basins are located among the lowest
continental elevations along releasing bends in the Death
Valley, Dead Sea, and Gobi Tien Shan fault systems, respectively (e.g., Burchfiel and Stewart, 1966; Garfunkel et al., 1981;
Rotstein et al., 1991; ten Brink et al., 1999; Cunningham et al.,
1997).
The geomorphic expression of continental transform plate
boundaries was recognized early on in the study of these
structures (Vedder and Wallace, 1970; Wallace, 1975). Figure 7
illustrates the main mesoscale landforms that are present
along most strikeslip fault systems, reflecting both zones of
transpression and transtension. Transtensional features typically include sag ponds, linear valleys, and troughs, whereas
transpression is generally expressed by linear ridges and
benches. Along straight segments of strikeslip faults, valley
spurs and ridges are commonly offset to form shutter ridges
and streams are characteristically deflected across fault traces
or tectonic landforms. Figure 8 demonstrates the modification
of streams flowing orthogonal to strikeslip faults. Fault scarps
are also common along strikeslip systems (e.g., Figure 4) and
may form a series of composite benches representing multipleslip events (i.e., ground-rupturing earthquakes).
Horsetail splay
Pull-apart
basin
Release bend
Principal
displacement
zone
(PDZ)
Restraining
bend
Normal seperation fault

Reversel seperation fault
Fault with sense of
strike-slip
Fold
Overturned fold
Areas of subsidence and
sediment accumulation
Pull-apart
basin
Rotated
folds
Thrusting
and uplift
Figure 3 Schematic diagram showing the geometry of strikeslip faults within transform settings and the associated landforms. Adapted from
Burbank, D.W., Anderson, R.S., 2001. Tectonic Geomorphology. Blackwell, Oxford, 274 pp., and Christie-Blick, N., Biddle, K.T., 1985. Deformation
and basin formation along strike-slip faults. In: Biddle, K.T., Christine-Blick, N. (Eds.), Strike-slip Deformation, Basin Formation, and
Sedimentation. Society of Economic Paleontologists and Mineralogists, Tulsa, OK, Special Publication 37, pp. 134.
41
1830' N
7225' W
(a)
5 km
(c)
(b)
(d)
(e)
Figure 4 Landforms and displacement along continental transform faults at differing scales. (a) Oblique high-altitude view of the San AndreasGulf of California transform plate boundary from the Space Shuttle. Image STS095704922 courtesy of the Image Science & Analysis
Laboratory, NASA Johnson Space Center (http://eol.jsc.nasa.gov). (b) Image at fifteen-meter-resolution of false-color Advanced Spaceborne
Thermal Emission and Reflection Radiometer (ASTER) of the Enriquillo-Plantain Garden fault in Haiti from 21 January 2010. The image was taken
9 days after a MW 7.0 earthquake along the fault. Note the large plate boundary-scale left-lateral stream offsets. Red color indicates vegetation.
White dots are bare bedrock exposed as a result of landslides from the 12 January 2010 earthquake. The city of Port-au-Prince is the gray
region in the northeast corner. Image courtesy of NASA/GSFC/METI/ERSDAC/JAROS, and US/Japan ASTER Science Team. (c) Oblique aerial
photograph of northern Death Valley fault at Red Wall Canyon, showing B29779 m of offset of alluvial fan channels (photograph courtesy of
Jeffrey Knott). (d) Field photograph looking to the NW across the dextrally offset Wallace Creek along the central San Andreas fault. Wallace
Creek is displaced 128 m as a result of multiple ground-rupturing earthquakes. The 1857 Fort Tejon earthquake resulted in 9 m of displacement
at this location (Sieh and Jahns, 1984) (US Geological Survey photograph by Ken Hudnut). (e) Single-event right-lateral oblique channel offsets
of 12 m from the 4 April 2010 MW 7.2 El Mayor-Cucape earthquake in Baja Mexico. View looking to the NW (US Geological Survey photograph
by Ken Hudnut).
42
Figure 5 Oblique high-altitude photographs of the central Altyn Tagh fault. North is to the top in the photograph. STS08870594 (Altyn Tagh
fault) are courtesy of the Image Science & Analysis Laboratory, NASA Johnson Space Center (http://eol.jsc.nasa.gov).
Transpression
(+)
20
0
Transtension
()
0
200
400
(a)
34 N
600
Along-strike distance (km)
38 N
n
Be
San Andreas fault
100 km
()
20
10
N
0
118 W
(b)
1000
30
Big
800
121 W
(+)
Transtension
<30
Transpression
0 2
20
% Fault length
Obliquity ()
40
>30
30 25 20 15 10 5 0 0 5 10 15 20 25 30
Obliquity () ()
(c)
Figure 6 Transpression and transtension along the San Andreas. Adapted from Spotila, J.A. Niemi, N., Brady, R., House, M., Buscher, J., Oskin,
M., 2007a. Long-term continental deformation associated with transpressive plate motion: the San Andreas fault. Geology 35, 967970, and
Spotila, J.A., House, M.A., Niemi, N.A., Brady, R.C., Oskin, M., Buscher, J.T., 2007b. Mechanisms of transpression along the San Andreas fault,
based on patterns of bedrock uplift. In: Till, A.B., Roeske, S.M., Sample, J.C., Foster, D.A. (Eds.), Exhumation Associated with Continental
StrikeSlip Systems: Geological Society of America Special Paper 434. (a) The angle of obliquity between relative plate motion and local
orientation of San Andreas fault, plotted for the entire plate boundary. (b) Map of major fault traces associated with the San Andreas fault
system, showing relative plate motion. (c) Histogram of obliquity along the San Andreas fault. Bars represent percent of fault length within the
range of obliquity shown, in 51 bins. The middle (unfilled) bar represents percent of fault that is oriented within 721 of local plate motion.
Offset drainage
channel
Offset drainage
channel
Linear valley
Linear ridge
Scarp
Spring
Bench
43
Linear valley
Shutter or ridge
ridge
Sag pond
Scarp
Fault
Shear zone
Figure 7 Block diagram of landforms associated with strikeslip faults. Adapted from Vedder, J.G., Wallace, R.E., 1970. Map showing recently
active breaks along the San Andreas and related faults between Cholame Valley and Tejon Pass, California: U.S. Geological Survey Miscellaneous
Geologic Investigations Map 1-574, scale 1:24 000.
Relatively crude maps and topographic surveys gave early

workers clues as to the topographic expression resulting from
ongoing tectonic activity along these strikeslip faults. However,
recent technological advances are helping researchers better
detect, date, and more fully understand the geomorphic expression of continental transform faults. Specifically, highresolution imagery from airborne lidar and terrestrial laser
scanning are providing submeter resolution, and in some cases,
centimeter-scale, topographic data for tectonic and geomorphic
investigations (e.g., Frankel et al., 2007a,b, 2011; Oldow and
Singleton, 2008). Figure 9 compares several different types of
remotely sensed imagery for part of Death Valley and highlights
the fine detail expressed in airborne lidar.
Furthermore, recent advances in Quaternary dating techniques such as radiocarbon (14C), optically stimulated luminescence (OSL), terrestrial cosmogenic nuclides (TCN), and
uranium-series (U-series) dating are helping refine the timing
of landform deposition and rates of deformation and denudation. Combined with short-term rates of deformation
determined by global positioning system (GPS) surveys and
other geodetic techniques (e.g., interferometric synthetic
aperture radar (InSAR)), these advances are improving the
evaluation of the role that continental transform faults play in
accommodating plate boundary deformation over a wide
range of spatial and temporal scales.
A recent area of major focus is comparing geologic/geomorphic slip rates (B102106 years) with geodetic slip rates
(100101 years) to assess whether geodetic observations characterize the nature of strikeslip faulting. Comparisons of
short-, intermediate-, and long-term rates of deformation have
important implications for seismic hazard assessment and
mitigation (e.g., Field et al., 2009) and for developing numerical models for the interactions and feedbacks between
tectonics and landscape evolution. Figure 10 summarizes
comparisons of geologic versus geodetic rates of displacement
along strikeslip fault systems. These data show strong discordance between the two sets of data for many of the
strikeslip systems. The reason for this is not fully understood
and is discussed in more detail below.
5.4.3
Advances in Studying Continental Transform

Systems
Over the past few decades, major advances in geochronology,

remote sensing, and tectonic geodesy have led to a vast improvement of our understanding of continental transform
fault systems. The advances in each of these fields are directly
responsible for furthering detailed investigations of deformed
landscapes along continental transform and strikeslip faults
and allowing geologists and geophysicists to tease out additional information about fault system behavior in these important plate boundary settings.
44
Beheaded
Beheaded
(a)
Fault
(b)
(c)
Fault
Shutter ridge
(d)
(e)
(f)
Fault
(g)
(h)
Meters
400
Approximate scale
Stream channel
Ridge or spur
Adandoned channel
Fault trace
Figure 8 Diagrammatic representation of patterns of fault-related stream channels occurring along the San Andrea Fault in the Carrizo Plain
area of California. Offset channels are shown in (a) and (b). (a) The misalignment of single channels directly related to the amount of fault
displacement and the age of the channel there is no ridge on the downslope side of the fault and beheading is common. (b) Misaligned paired
stream channels. Patterns resulting from a combination of offset and deflection are represented in (c)(f). (c) Compound offsets of ridge spurs
and both left and right offsetting and deflection of channels. (d) Trellis drainage produced by multiple fault strands and shutter ridges. (e)
Indicates how a combination of offset and deflection by shutter ridges can give rise to exaggerated or reversed apparent offset. (f) Shows how
capture by an adjacent channel, followed by right-lateral slip may produce a Z pattern. Two types of false offset are represented by (g) and (h).
(g) Shows how differential uplift may deflect streams to produce false offsets. Adapted from Wallace, R.E., 1975. The San Andreas fault in the
Carrizo PlainTemblor Range region, California. In: Crowell, J.C., (Ed.), The San Andreas Fault in Southern California. California Division of Mines
and Geology, Special Report 118, pp. 241250, and Summerfield, M., 1991. Global Geomorphology, Longman, Harlow, pp. 537.
(h) Illustrates en echelon fractures over a fault zone, followed by streams producing false offsets.
5.4.3.1
Geochronology and Thermochronology
Improvements and advances in Quaternary geochronometers

are particularly beneficial for understanding the tectonic geomorphology of continental transform plate margins, among
other regions of deformation (e.g., Noller et al., 2000).
Methods such as TCN geochronology (e.g., Gosse and Phillips,
2001; Dunai, 2010), OSL (e.g., Tsukamoto et al., 2009;
Rhodes, 2011), U-series geochronology (e.g., Bourdon et al.,
2003; van Calsteren and Thomas, 2006), 14C (e.g., Weldon
et al., 2004, 2005), and various forms of thermochronology
(e.g., Reiners and Ehlers, 2005) are allowing researchers the
opportunity to unravel the timing and rates of deformation
with unprecedented accuracy and precision.
The ability to accurately and precisely date landforms that
have been offset and deformed by faults is critical to assessing
the rates and styles of deformation. In most cases, geochronometers are well suited for specific timescales of deformation.
Typically, TCN geochronology can be used to determine the
timing and rates of deformation over timescales ranging from
103 years to, under certain favorable circumstances, 106 years
(e.g., van der Woerd et al., 1998, 2000, 2002; Frankel et al.,
2007a,b, 2010a, b; Granger, 2006). Although highly precise,
radiocarbon is generally limited to the past 40 ka (e.g., Pigati
et al., 2007). U-series dating is extremely precise, and is confined to timescales of a few hundred to about 500 000 years,
as is OSL. Thermochronometers, however, are generally useful
for deciphering deformation rates over timescales ranging
from 106 to 109 years (e.g., Reiners and Shuster, 2009).
In most cases, 14C, TCN, OSL, and U-series methods are used
to determine the timing of deposition of deformed alluvial and
fluvial landforms. TCNs can also be used to quantify the rates of
150m
(a)
45
150m
(b)
150m
(c)
150m
(d)
Figure 9 Comparison of airborne lidar topography data with aerial photography and other commonly available digital topographic data. Adapted
from Frankel, K.L., 2007. Fault slip rates, constancy of seismic strain release, and landscape evolution in the eastern California shear zone. Ph.D.
dissertation, Los Angeles, University of Southern California, 179 pp. Images are map views of Badwater, the lowest point in the western
hemisphere, in Death Valley, California. (a) Thirty-meter horizontal resolution, NASA shuttle radar topography mission (SRTM), digital topographic
data. (b) Ten-meter horizontal resolution, USGS national elevation data (NED), digital topography. (c) Black and white 1:12 000 scale aerial
photograph. (d) One-meter horizontal resolution airborne lidar image.
denudation and the timing and duration of sediment burial

(e.g., Granger et al., 1996; Granger, 2006; Dortch et al., 2011).
Thermochronology is the main source of data related to the
exhumational history of large-scale topography, which, along
continental transform faults, is typically formed along
large bends (e.g., the Big Bend of the San Andreas fault) or as a
result of oblique convergent strain regimes (e.g., Spotila et al.,
2007a, b).
In all cases, the power to unravel the tectonogeomorphic
history of a continental transform system comes from combining high-fidelity age control with high-resolution, remotely
sensed imagery. Only by combining these two data sets can the
Quaternary deformation history of a continental transform
fault be fully understood.
5.4.3.2
Remote Sensing
Any successful study of tectonics and topography must begin

with a detailed analysis of the landscape. A key component to
many recent investigations of plate boundary fault systems is
the acquisition of both ground-based and airborne lidar
digital topographic data. These data facilitate the efficient
identification, mapping, and analysis of deformed landforms
in unprecedented detail (e.g., Carter et al., 2007; Hudnut et al.,
2002; Arrowsmith and Zielke, 2009; Frankel et al., 2007a,b;
Oskin et al., 2007). Lidar data have several great advantages
over previously available forms of remotely sensed and digital
topographic data (Figure 9). In particular, surveying of deformed landforms (e.g., offset channels and alluvial fans) and
46
40
Long-term geologic rate (mm/yr)
35
gy
30
olo
e
=g
GP
25
20
15
10
5
0
0
10
15
20
25
30
35
40
Short-term GPS rate (mm/yr)

Figure 10 Plots of geologic versus geodetic rates of displacement along strikeslip fault systems. Modified from Thatcher, W., 2009. How the
continents deform: the evidence from tectonic geodesy. Annual Review of Earth and Planetary Sciences 37, 237262, with additional data.
the construction of high-precision topographic maps, which

would take days to weeks with traditional surveying methods,
can be accomplished in minutes with lidar data.
Moreover, the high-resolution topographic data can be
digitally manipulated to enhance and reveal subtle topographic
features, something not possible with aerial photographs, satellite imagery, or lower-resolution digital elevation models
(Figure 9). For example, the landscape can be artificially illuminated from any angle to highlight previously unrecognized
features of the landscape and surface slope and slope aspect
maps can also be easily constructed to reveal and retrodeform
extremely subtle topographic features (Frankel, 2007; Frankel
et al., 2007a,b; Oskin et al., 2007; Arrowsmith and Zielke, 2009;
Zielke et al., 2010). Furthermore, this technology holds the
ability to see through canopies in vegetated regions, thus revealing an image of the bare earth beneath the trees and illuminating landscapes that could not previously be investigated
in detail with aerial photographs or other forms of remotely
sensed images (e.g., Carter et al., 2007; Prentice et al., 2009).
These data can also be used to objectively map and quantify the
evolution of alluvial landforms through time by developing
various algorithms to measure surface roughness characteristics
(e.g., Frankel and Dolan, 2007). Combining detailed landform
analyses from lidar or similar data with age control from improved Quaternary geochronometers is leading to an unprecedented insight into the rates and patterns of fault activity
and landscape evolution (e.g., Frankel et al., 2011).
5.4.3.3
Tectonic Geodesy
Although not directly related to the study of landforms, tectonic

geodesy is shedding new light on rates and styles of deformation
along continental transform and other forms of plate boundaries as well as continental plate interiors (e.g., Thatcher, 2009;
Calais et al., 2002, 2010). GPS and InSAR are providing both
near real-time views and snapshots of the present-day strain

field, and helping elucidate patterns of active faulting and deformation (e.g., Gourmelen et al., 2011; Chen and Khan, 2010).
Critical to understanding how deformation is accommodated
through the lithosphere is an understanding of how constant or
variable rates of deformation are in time and space.
Tectonogeomorphic studies integrate rates and patterns of
deformation over hundreds to millions of years, whereas
geodetic studies commonly contain only a few years to 10s of
years of information about lithospheric deformation. However, integrating short-term data from geodetic measurements
with intermediate- and long-term data from geomorphology
and geochronology helps provide a fully integrated view of the
behavior of plate boundary fault systems. Indeed, combining
short- and long-term studies of deformation can help provide
important information on whether temporal and spatial
changes in the rates of deformation are characteristic of tectonic plate boundaries or tied more directly to structurally
complex regions (Figure 10).
For example, discrepancies in short-term and long-term
rate data have been reported along parts of major continental
transform boundaries (e.g., Bennett et al., 2003; Dolan et al.,
2007; Frankel et al., 2007b; Oskin et al., 2008) and other
strikeslip fault systems (e.g., Chavelier et al., 2005; Ganev
et al., 2012). These mismatches in rates could reflect real
temporal changes in tectonic activity, or alternatively, could be
the result of uncertainties associated with individual techniques and/or off-fault deformation that may not be captured by
geomorphic data sets focused solely on discrete structures.
Moreover, integrating geodetic and geomorphic data sets
yields an important insight into whether strain is accommodated primarily along individual structures separating
continental blocks or as a continuum of deformation spread
out across broad regions of the lithosphere (e.g., Molnar and
Tapponnier, 1975; Thatcher, 2009).
5.4.4
Major Continental Transform Plate Boundaries

and Strikeslip Fault Systems
motion along the plate boundary is that of the right-lateral

San Andreas fault, which accommodates the majority
(B75%) of Pacific-North America plate boundary
motion (King and Savage, 1983; DeMets and Dixon, 1999;
Bennett et al., 1996; Fialko, 2006). However, the plate
boundary is also characterized by large regions of transpression near the northern termination at the Mendocino triple
junction and the Big Bend of the San Andreas fault at the
Transverse Ranges. In addition, to the east of the San Andreas
fault exists a belt of transtension known as the eastern California shear zone-Walker Lane, which eventually gives way to
the purely extensional Basin and Range province. The plate
boundary can essentially be thought of as a 200800 km-wide
transition zone from extension to strikeslip motion (e.g.,
Bull, 2007, 2009).
The major continental transform plate boundaries and

strikeslip fault systems are briefly summarized in this section.
More details are provided for the Pacific-North American
Boundary because it is generally considered the archetype for
transform plate boundaries and strikeslip fault systems, and
as such, has arguably been studied more than the other plate
boundaries.
5.4.4.1
Pacific-North America Plate Boundary
The Pacific-North America plate boundary is formed in a

highly diverse tectonic setting (Figure 11). The dominant
125 W
120 W
115 W
IDAHO
40 N
OREGON
NEVADA
Ba
sin
alk
dR
La
ne
rra
ge
da
va
Ne
NORTH
AMERICAN
PLATE
IA
N
OR
LIF
F
SA
EC
GF
SZ
35 N
an
an
er
Sie
CA
PACIFIC
PLATE
47
N
200 km
Figure 11 Quaternary faults and topography of the Pacific-North America Plate Boundary. Faults plotted from the USGS Quaternary faults and
fold database and 90-m resolution SRTM dataset. ECSZ, eastern California shear zone; GF, Garlock fault; SAF, San Andreas fault.
48
5.4.4.1.1
San Andreas Fault
Perhaps the best present-day example of a continental transform system is the San Andreas fault, which is responsible for
accommodating the majority of deformation along the PacificNorth America plate boundary. The San Andreas fault extends
some 1300 km in length from the spreading center at the
Mendocino triple junction in the north to the midocean ridges
responsible for rifting in the Gulf of California in the south
(Figure 11).
The San Andreas fault is one of the worlds most active
fault systems. Weldon et al. (2004) reconstructed a 6000-yearlong record of earthquakes and offsets by undertaking paleoseismic investigations across the San Andreas fault near the
town of Wrightwood, California. This included excavating 45
trenches over 18 years. At least 30 prehistoric earthquakes were
characterized and Weldon et al. (2004) showed that for the
past 1500 years, the mean recurrence interval is 105 years
(31165 years for individual intervals) and the mean slip
per event was 3.2 m (0.77 m per event) (Figure 12). The
series was slightly more ordered than random and had a
notable cluster of events, during which stress was released at
three times the long-term average rate. Generally, short

intervals tend to coincide with large displacements and
long intervals with small displacements. The most significant
correlation that Weldon et al. (2004) found was that earthquakes were more frequent following periods of net strain
accumulation spanning multiple seismic cycles. The extent
of paleoearthquake ruptures was inferred by correlating
event ages between different trenching sites along the San
Andreas fault.
Pallet and Wallace Creeks are arguably the most famous
paleoseismic sites along the San Andreas fault. Sieh (1984)
and Sieh and Jahns (1984) first examined these locations.
The classic right-laterally offset drainages present at Wallace
Creek are illustrated in Figures 4(d) and 13. Offset drainages
are abundant along other stretches of the San Andreas fault;
however, until recent advances in fault zone imaging with
lidar, it was generally difficult to determine the exact offset.
Figure 14 shows an example of a channel that was offset during
the 1857 Fort Tejon earthquake. This channel offset, along
with 148 others, were measured by airborne lidar-derived
digital elevation models along the 60-km-long Carrizo segment
50
45
Cumulative slip (m)
40
35
2.4 cm/yr
R2 = 0.94
30
25
8.9 cm/yr
R2 = 0.99
3.1 cm/yr
R2 = 0.95
20
15
10
5
0
400
600
800
1000
1200
1400
1600
1800
2000
Calendar year (A.D.)
(a)
Cumulative strain (m)

departure from mean
4
2
0
2
4
6
8
10
500
(b)
750
1000
1250
1500
Calendar years (A.D.)
1750
2000
Figure 12 Paleoseismic data for part of the San Andreas fault at Wrightwood. Adapted from Weldon, R.J., Scharer, K.M., Fumal, T.E., Biasi,
G.P., 2004. Wrightwood and the earthquake cycle: what a long recurrence record tells us about how faults work. GSA Today 14, 410.
of the San Andreas by Zielke et al. (2010). They used a

statistical, goodness-of-fit measurement to determine fault offset during the 1857 Fort Tejon earthquake and determined
that surface displacement during this event was 5.371.4 m,
significantly lower than the previous estimate of 9.072.0 m
(Sieh, 1978a; Wallace, 1968; Sieh and Jahns, 1984; Liu-Zeng
et al., 2006). Additional well-defined channel offsets indicate
that the characteristic displacement during ruptures along the
Carrizo segment of the San Andreas fault may be B5 m.
The topographic expression of the San Andreas fault is
impressive, ranging from deflected drainages, sag ponds,
pressure ridges, and marine terraces to transpressional
mountains that rise to 43500 m above sea level (asl). The
35 1615 N
119 4945 W
200 m
Figure 13 Lidar image of a segment of the San Andreas fault at

Wallace Creek showing right-laterally offset drainages.
49
evolution of these landforms is being studied in increasing

detail. One recent study used a space-for-time substitution
method of applying high-resolution lidar-derived topography,
geomorphic mapping of active surface processes, and geologic
mapping, to examine the topographic and erosional response
of small drainage basins to rock uplift along the Dragons
Back pressure ridge in the Carrizo Plain segment of the San
Andreas fault (Hilley and Arrowsmith, 2008; Figure 15).
Hilley and Arrowsmith (2008) were able to show that progressive deformation and rock uplift is accompanied by increases in channel steepness and basin relief. They argued that
as uplift ceases, channel concavity rapidly increases, causing
channels to undercut hillslopes and produce landslides,
which causes basin relief to be greatest just after rock uplift
has stopped. The analysis showed that channels in the pressure ridge respond to changes in rock uplift rates over thousands of years. In contrast, hillslope processes may take more
than an order of magnitude longer to adjust to changes in
rock uplift rates.
The Mendocino triple junction marks the northern end of
San Andreas fault. Impressive flights of marine terraces are
present at, and just south of, the Mendocino triple junction; as
many as 14 flights of marine terraces are present in this region.
These marine terraces may be the result of regional surface
uplift caused by the growth of a slab window (Merritts and
Bull, 1989). Terraces date from 3 to 18 ka and temporal
changes in uplift rates vary from o1 to 35 mm yr 1 and
allow reconstructions of the tectonic regime and indicate
temporally and spatially variable rates of coastal uplift that
track with the northward migration of the Mendocino triple
junction (Figure 16). Snyder et al. (2003) suggest that the
incision rates that Merritts and Bull (1989) inferred as related
Elevation (m)
616
615
0
20
40
60
80
616
351303N
614
0
20
40
60
Distance along profile (m)
Fault trace
X-sections
Channel trend
1/SDelevation
(b)
50 meters
(a)
1194618W
15
min
80
max
5
3
5
7
Displacement (m)
(c)
Figure 14 (a) Lidar-derived hillshade map of a channel offset during the 1857 Ft. Tejon earthquake along the central San Andreas fault. (b)
Channel topographic cross-sectional profiles along red and blue lines projected onto the fault plane, based on channel
orientation relative to the
P
fault in (a). The blue profile is the best-fit estimate of the back-slipped offset. (c) Measure of goodness of fit (1/ D elevation) calculated for the
back-slip model in (b). Modified from Zielke, O., Arrowsmith, J.R., Ludwig, L.G., Akc- iz, S.O., 2010. Slip in the 1857 and Earlier Large
Earthquakes Along the Carrizo Plain, San Andreas Fault. Science 327, 11191122.
50
elevation
1 km
588
meters
731
3 Rock uplift rate (mm/yr)

2
1
0
4
(e)
150 Cumulative rock uplift (m)

100
(a)
50
Qal
Qoa
Qls
QTPp
??
(f)
60 Local relief (m)
40
20
0
QTP
San Andreas Fau
500 m
QTPt
lt
QTPg
600 m
(g)
1 km
(b)
rock uplift
N
0
meters
Hillslope gradient (m/m)
0.5
80
(h)
10 Normalized channel steepness (mean concavity = 0.68)

5
1 km
0
(c)
N
rock uplift rate

0
mm/yr
1.7
0.5
(i)
Basin area (km3)
0
0
0
1 km
(j)
1
2
3
4
30
60
90
120
Distance NW along Dragons Back pressure ridge (km)
time (ka)
(d)
Figure 15 Uplift zone of the Dragons Back pressure ridge along the central San Andreas fault. (a) Airborne lidar-derived digital elevation model
(DEM). Scale at the bottom corresponds to x-axis values in (e)(j). (b) Geologic map overlain on lidar DEM. QTP undifferentiated Paso Robles
Formation; QTPt, QTPp, and QPTg are tan, pink, and gold members of the Paso Robles Formation; Qoa Pleistocene alluvim; Qls landslide
deposits; Qya young alluvium. Dashed contours indicate inferred subsurface location of the San Andreas fault. (c) Along-strike total rock uplift.
(d) Along-strike rock uplift rate assuming a right-lateral slip rate of 33 mm yr 1 on the San Andreas fault. (e) Mean rock uplift rate along the
Dragons Back pressure ridge. (f) Total rock uplift. (g) Local relief (the solid line is the mean value and the shaded area is 95% bounds). (h)
Hillslope gradient (the solid line is the mean value and the shaded area represents the 95% bounds). (i) Channel steepness (the solid line is the
500 m running average). (j) Drainage basin area (the solid line is the 500 m running average). Rock uplift, rock uplift rate, and various
topographic metrics are plotted as a function of distance from the southeast extent of the Dragons Back pressure ridge. Modified from Hilley,
G.E., Arrowsmith, J.R., 2008. Geomorphic response to uplift along the Dragons Back pressure ridge, Carrizo Plain, California. Geology 36,
367370.
to uplift might be influenced by drainage basin size and discharge, which suggests that other factors besides uplift are
influencing incision in this area.
Spotila et al. (2007a, b) used the San Andreas fault to
provide a synthesis for transpressive mountain building, as
recorded by rock uplift and topography along the entire length
of the fault (Figure 17). Their study revealed a complex crustal
response to oblique plate motion. Moreover, Spotila et al.
(2007a, b) showed that convergent deformation increases
toward the fault but does not correlate with the angle of platemotion obliquity. The shortening estimated from rock uplift
was insufficient to account for the fault-normal motion based
on relative plate velocity. Spotila et al. (2007a, b) suggested
that near-field convergence is influenced by local structural
complexity and is not purely driven by regional transpression,
and that the fault-normal component of plate motion is partly
accommodated elsewhere.
South of the Cajon Pass, where the San Andreas fault traverses the San Gabriel and San Bernardino mountains of the
Transverse Ranges, the importance of the San Andreas fault in
accommodating Pacific-North American relative plate motion
may be less, with a significant amount of motion taken up on
the San Jacinto fault. Using high-resolution airborne laserswath topography and selected offset landforms that were
dated using TCN 10Be, Blisniuk et al. (2010) suggested that
since B3050 ka, the slip rate along the southern San Jacinto
fault zone has been equivalent to or slightly less than the rate
along the southernmost San Andreas fault (Figure 18).
5.4.4.1.2
Eastern California shear zone-Walker Lane: An

Evolving Plate Boundary
The eastern California shear zone-Walker Lane is an evolving

component of the Pacific-North America plate boundary system (e.g., Faulds et al., 2005; Wesnousky, 2005; Frankel et al.,
2011). This region of predominantly right-lateral strikeslip
faults is thought to accommodate B2025% of the total
relative motion between the Pacific and the North American
plates (Bennett et al., 2003; Dixon et al., 2000, 2003; Dokka
and Travis, 1990; Hearn and Humphreys, 1998; Humphreys
and Weldon, 1994; McClusky et al., 2001; Thatcher et al.,
1999). The area of active deformation extends northward from
the eastern end of the Big Bend of the San Andreas fault near
Palm Springs, California, for B500 km through the Mojave
Surveyed profile
of marine terraces
Altitude (m)
150
Altitude from
7.5 map
Sea-level fluctuation curve

100
50
Sea Level
0
0
500
1000
1500
?
50
distance (m)
?
?
100
Terrace preserved and present

Terrace submerged or reoccupied
Terrace missing in profile
150
80
(a)
240
320
Inferred uplift (m)
140
120
100
80
60
40
20
00
(b)
y = 6.2 + 0.4x R2 = 0.99

100
200
300
400
Time before present (ka)
Figure 16 Diagrams illustrating uplift rates at the northern termination of the San Andreas fault near the Mendocino triple junction based on marine terraces. (a) Correlation of uplifted marine terrace
present-day altitudes with a eustatic sea-level curve. (b) Plot of correlated marine terrace age versus surface uplift of marine terraces. Linear regression through the data yields an uplift rate of
0.4 m ka 1. Modified from Merritts, D., Bull, W. B., 1989. Interpreting quaternary uplift rates at the mendocino triple junction, northern California, from uplifted marine terraces. Geology 17, 10201024.
160
Time before present (ka)
51
(mm yr1)
52
10
0.1
(a)
(m)
1500
1000
500
(b)
0
(deg.)
400
600
800
1000
N
Elevation
200
400
600
800
1000
Slope
10
0
0.2
(m yr1)
200
0
20
(c)
Rock uplift rate
200
400
600
800
1000
Precipitation
0.1
0.0
200
400
(d)
600
800
1000
Distance (nm)
Figure 17 The geomorphology along the San Andreas fault, showing the variance of: (a) envelope of maximum rock uplift along fault; (b) mean
elevation in an 80-km-wide swath along fault, (c) mean slope along a 20-km swath along, and (d) mean annual precipitation along fault. Adapted
from Spotila, J.A. Niemi, N., Brady, R., House, M., Buscher, J., Oskin, M., 2007a. Long-term continental deformation associated with
transpressive plate motion: The San Andreas fault. Geology, 35, 967970, and Spotila, J.A., House, M.A., Niemi, N.A., Brady, R.C., Oskin, M.,
Buscher, J.T., 2007b. Mechanisms of transpression along the San Andreas fault, based on patterns of bedrock uplift. In: Till, A.B., Roeske, S.M.,
Sample, J.C., Foster, D.A. (Eds.), Exhumation Associated with Continental StrikeSlip Systems: Geological Society of America Special Paper, 434.
Desert and along the western edge of the Basin and Range east
of the Sierra Nevada (Figure 11).
In the Mojave Desert, south of the left-lateral Garlock fault,
the eastern California shear zone comprises a 100-km-wide
network of NNW-trending right-lateral faults. Geodetic data
indicate that elastic strain is accumulating across this zone at a
rate of 1272 mm yr 1 (Savage et al., 1990; Gan et al., 2000;
McClusky et al., 2001; Miller et al., 2001; Peltzer et al., 2001).
Seismological and paleoseismological data also indicate that
this part of the eastern California shear zone releases strain at
a relatively rapid rate; portions of several of these faults ruptured during the 1992 moment magnitude (Mw) 7.3 Landers
and 1999 Mw 7.1 Hector Mine earthquakes. Moreover,
paleoseismologic data indicate that these two earthquakes
are part of an ongoing, B1000-year-long seismic cluster
(Rockwell et al., 2001).
However, recent work in the Mojave section of the eastern
California shear zone indicates that the total long-term slip
rate across this fault system is on the order of 57 mm yr 1 or
about half of the current rate of strain accumulation determined from space-based geodesy (Oskin and Iriondo, 2004;
Oskin et al., 2007, 2008). These observations suggest a pronounced strain rate transient across the Mojave section of the
eastern California shear zone. The Garlock fault bisects
the eastern California shear zone, forming a major geologic
and physiographic boundary between the Mojave Desert
and the western Basin and Range (Figure 1; Davis and
Burchfiel, 1973).
Displacement from the southern part of the eastern California shear zone in the Mojave Desert is funneled northward
across the Garlock fault onto four main fault systems: the
Owens Valley, Panamint Valley-Hunter Mountain-Saline Valley, Death Valley-Fish Lake Valley, and Stateline fault zones.
The only fault in the northern eastern California shear zone to
experience significant historic seismic activity is the Owens
Valley fault, which last ruptured in an estimated Mw 7.5
earthquake near the town of Lone Pine in 1872 (Beanland and
Clark, 1994; Hough and Hutton, 2008). A series of down-tothe-NW normal faults transfer slip between the Owens Valley,
Panamint ValleyHunter MountainSaline Valley, and Death
ValleyFish Lake Valley faults (Figure 19; Dixon et al., 1995;
Lee et al., 2001; Reheis and Dixon, 1996; Frankel et al., 2011).
In contrast to the strain transient south of the Garlock, the
region-wide rate of dextral shear across these four faults appears to have remained constant at 910 mm yr 1 over late
Pleistocene to recent timescales (Bacon and Pezzopane, 2007;
Bennett et al., 2003; Frankel et al., 2007a; Guest et al., 2007;
Lee et al., 2001, 2009; Oswald and Wesnousky, 2002).
Farther north, dextral motion between the Sierra Nevada
block and North America is focused on two faults bounding the
east and west sides of the White Mountains: the White Mountains fault zone to the west and the Fish Lake Valley fault zone
to the east. Of these two, GPS data suggest that the Fish Lake
Valley fault system is storing at least 75% of the elastic strain
accumulating in this region (Dixon et al., 2000). However, recent late Pleistocene slip rate studies on these two faults reveal a
53
Figure 18 Tectonic geomorphology of part of the San Jacinto fault, southern California. Adapted from Blisniuk, K., Rockwell, T., Owen, L.A.,
Oskin, M., Lippincott, C., Caffee, M.W., Dortch J., 2010. Late Quaternary slip rate gradient defined using high-resolution topography and 10Be
dating of offset landforms on the southern San Jacinto Fault zone, California. Journal of Geophysical Research 115, B08401. (a) ALSM image
showing the fault trace and displaced channels. (b) Location of beheaded and deflected channels at Rockhouse Canyon: (left) the present-day
configuration of Channels 1, 2, and 3; (middle) the reconstruction for Channel 2; and (right) the reconstruction for Channel 1. The blue solid
lines indicate the maximum and minimum displacements from the source drainage(s) for offset channels. The yellow dots indicate the location of
boulder samples collected for 10Be exposure dating.
discrepancy between geologic and geodetic slip rates similar to

that in the Mojave, whereby the White Mountains and Fish
Lake Valley faults account for less than half of the region-wide
rate of shear determined from GPS data (Bennett et al., 2003;
Frankel et al., 2007b, 2011; Kirby et al., 2006).
Across this same region, extension is accommodated on the
Round Valley fault in the west and across the Volcanic Tableland, White Mountains fault, and Fish Lake Valley fault to the
east (Ganev et al., 2010; Kirby et al., 2006). Unlike the dextral
component of shear in this region, when compared with
short-term rates of extension, late-Pleistocene extension rates
appear to agree well with geodetic data (e.g., Bennett et al.,
2003; Ganev et al., 2010; Wesnousky, 2005). The mismatch
between short- and long-term rates of right lateral deformation and the good temporal correlation among rates of extension have important implications for lithospheric
dynamics, the distribution of deformation, and strain transfer
along this important component of the Pacific-North America
plate boundary (e.g., Frankel et al., 2007b, 2011; Oldow,
2003). Although most of the structures in the northern eastern
California shear zone are relatively well characterized, there
are some regions where debate still exists as to the exact origin
and significance of landforms. For example, the Poverty Hills,

which are located along the dominantly right-lateral Owens
Valley fault, have been explained as both a thrust-bound popup structure (e.g., Taylor and Dilek, 2001; Taylor, 2002) and as
a large landslide (e.g., Bishop and Clements, 2006; Bishop in
Frankel et al., 2010a). Critical to understanding the role of the
eastern California shear zone in plate boundary deformation
is unraveling the origin and importance of enigmatic features
such as the Poverty Hills.
Using lidar topographic data and TCN 10Be dating of
offset alluvial fans, Ganev et al. (2010) showed that the oblique-normal-dextral Fish Lake Valley fault accommodates the
majority of Pacific-North America plate-boundary extensional
deformation east of the San Andreas fault in the northern part
of the eastern California shear zone (Figure 19). Comparison
of these rates with geodetic measurements of extension across
the northern Eastern California shear zone indicates that the
Fish Lake Valley fault accommodates approximately half of the
current rate of regional extension.
In one of the most detailed studies of the spatial and temporal constancy of seismic strain release along a strikeslip fault,
Frankel et al. (2011) showed that the late Quaternary slip
373442N
54
1180028W
1175932W
373338N
Qfi
N
500 m
1180028W
Late Pleistocene right-lateral slip rate (mm/yr)
Figure 19 Hill-shaded, LiDAR-derived digital elevation model showing normal faults within a transtensional zone in Fish Lake Valley, at the
CaliforniaNevada border. Adapted from Ganev, P.N., Dolan, J.F., Frankel, K.L., Finkel, R.C., 2010. Rates of extension along the Fish Lake Valley
fault and transtensional deformation in the eastern California shear zone-Walker Lane belt. Lithosphere 2, 3349. Solid lines are fault scarps and
the arrows show sense of motion on strikeslip faults, tick marks are on the hanging wall of normal faults. The surface highlighted in brown is
approximately 90 ka. Note the shutter ridge in the lower central area.
NW
SE
Cucumongo
Canyon
6
5
Red Wall
Canyon
4
3
2
Indian
Creek
Furnace
Creek
South Mud
Canyon
1
0
175
150
125
100
75
50
25
Along-strike distance (km)

Figure 20 The along-strike distribution of late Pleistocene slip rates on the Death Valley-Fish Lake Valley fault system. Adapted from Frankel,
K.L., Dolan, J.F., Owen, L.A., Ganev, P., Finkel, R.C., 2011. Spatial and temporal constancy of seismic strain release along an evolving segment
of the Pacific-North America plate boundary. Earth and Planetary Science Letters 304, 565576.
rates decreases northward and southward from the central part

of the Death Valley-Fish Lake Valley fault system (Figure 20).
Frankel et al. (2011) argued that slip is transferred onto northtrending zones of distributed normal faulting toward the
northeast and southwest of the central zone of a rapid slip rate.
This complex pattern of strain accommodation likely reflects
structural evolution toward a straighter, structurally simpler
zone of dextral shear that locally utilizes well-established dextral
faults that are linked where necessary by nascent zones of deformation. Summing of the rates of all four major faults in the
eastern California shear zone at the 371 N latitude in northern
Death Valley, Frankel et al. (2007a) showed that the cumulative
geologic rate of B8.510.0 mm yr 1 is indistinguishable from
the B9 mm yr 1 geodetic rate. Frankel et al. (2011) argued that
although the cumulative rate on the major faults of the ECSZ is
slower to the north and south, this probably reflects more distributed deformation in these areas, rather than transient strain
accumulation. Their results demonstrate the importance of obtaining multiple-slip rates to effectively document the behavior
of any fault system, in studies of seismic hazard assessment and
comparisons of geologic and geodetic rate data.
5.4.4.1.3
Garlock Fault: An Intracontinental Transform
The Garlock fault is part of, but oriented perpendicular to, the
prevailing deformation along the Pacific-North America plate
boundary. In a classic paper on western North America tectonics, Davis and Burchfiel (1973) referred to the Garlock fault
as an intracontinental transform. Although the fault does not
connect oceanic spreading centers, it does separate the highly
extended Basin and Range province to its north from the relatively undeformed Mojave block to its south. The Garlock fault
is somewhat enigmatic in that nowhere does it appear to offset,
or be offset by, NW-trending right-lateral eastern California
shear zone faults (Figure 11). A major remaining question in
eastern California shear zone tectonics is how Pacific-North
America plate boundary strain is transferred through, or around,
the Garlock fault. Since the late Pleistocene, there has been at
least 18 km of offset along the fault, and yet little to no significant earthquake activity has occurred over at least the past
B300 years (Carter, 1980; Dawson et al., 2003; McGill and
Sieh, 1991). Although no historical ground-rupturing earthquakes have occurred on the fault, numerous Holocene events
have been recorded in paleoseismic studies (e.g., McGill and
Sieh, 1991; Dawson et al., 2003). In addition, the geomorphic
expression of the Garlock fault is nothing short of spectacular,
with numerous left-laterally offset landforms. The long-term slip
rates on the Garlock fault are at significant odds with short-term
geodetic data, which record essentially no present-day motion
across the structure (Meade and Hager, 2005; McGill et al.,
2009; Ganev et al., in press).
McGill and Sieh (1991) showed that landforms offset along
the central and eastern Garlock fault record the amount of
surface slip associated with prehistoric earthquakes. In conjunction with available slip rates for the Garlock fault, the
geomorphic offsets suggest that average recurrence intervals are
probably within the range of 6001200 years south of the El
Paso Mountains, about 200750 years in Searles Valley, about
2001300 years in Pilot Knob Valley, and about 2003000 years
near Leach Lake and in the Avawatz Mountains. Dawson
et al. (2003) examined the paleoseimic record at the El
55
Paso Peaks site and have refined and extended the record of
paleoearthquakes for the central Garlock fault, identifying
six well-resolved earthquakes during the last B7000 years
(Figure 21). The event ages indicate that earthquake recurrence
is highly irregular at the El Paso Peaks site, with individual
preferred intervals ranging from as little as 215 years to as many
as 3300 years. Dawson et al. (2003) argued that the irregular
recurrence of surface-rupturing earthquakes along the central
Garlock fault might be related to temporal and spatial clustering
of earthquakes proposed for faults in the eastern California
shear zone (e.g., Rockwell et al., 2001; Dolan et al., 2007).
McGill et al. (2009) suggested that the high rate of motion
on the western Garlock fault is most consistent with a model
in which the western Garlock fault acts as a conjugate shear to
the San Andreas fault. Other mechanisms, involving extension
north of the Garlock fault (e.g., Davis and Burchfiel, 1973)
and block rotation at the eastern end of the fault (e.g., Guest
et al., 2003), may be relevant to the central and eastern sections of the fault, but they cannot explain the high rate of slip
on the western Garlock fault.
5.4.4.2
Pacific-Australian Plate Boundary
The Alpine fault, which traverses the South Island of New Zealand, accommodates most of the slip along the transform plate
boundary of the Pacific and Australian plates (Figure 22), which
equates to B40 mm yr 1 parallel to the fault (Molnar and
Dayem, 2010). Spectacular tectonic landforms are present along
its trend as illustrated in Figure 23 (Norris and Cooper, 2001;
Sutherland et al., 2006; Barnes, 2009). The evolution of the Alpine fault and the associated Southern Alps is complex, involving
feedback between uplift, climate, and erosion (Beaumont et al.,
1996; Waschbusch et al., 1998; Pysklywec et al., 2002).
5.4.4.3
Caribbean-North American Plate Boundary
The Caribbean-North American plate boundary is complex.

The Cayman Trough marks the northernwestern limit of the
Caribbean plate (Lyon-Caen et al., 2006) and behaves as a
transform fault separating oceanic lithosphere of the Caribbean and North American plates. The Motagua fault system
continues westward, splitting into several splays of the
Polochic and Motagua fault systems. To the east of the Cayman Trough, several major strikeslip faults traverse the region
as shown in Figure 24 (Lyon-Caen et al., 2006; Rogers and
Mann, 2007). Of particular significance is the EnriquilloPlantain Garden fault in Haiti, which is the likely structure
responsible for the infamous Mw 7.0, 12 January, 2010 Haiti
earthquake (USGS, 2011; Figures 25 and 26). The east-trending El Pilar fault marks the boundary of the Caribbean and
South American plates, which splays into the Bocono and
Oca Faults at its western end. Transpression along the Bocono
has resulted in the formation of the Venezuelan Andes
(Figure 27), which rise to 44980 m asl (Wesnousky et al.,
2012).
5.4.4.4
Himalayan-Tibetan Orogen Strikeslip Systems
The Himalayan-Tibetan orogen is a consequence of the collision of the Indian and Eurasian continental lithospheres
56
1173330W
70
Number of offset landforms
352840N
100 m
60
50
40
30
20
10
0
(a)
Seismic moment release (dyne-cm x 1027 yr-1)
(b)
0.1
8 10 12 14 16 18 20 22 24 26 28 30
Displacement (m)
Event U
Event Q
0.09 Event W
Event F
Event K
0.08
Event R
0.07
0.06
0.05
Cluster 3
0.04
0.03
Cluster 2
0.02
Cluster 1
0.01
Cluster 4 & 5?
0
0
10
(c)
12
14
16
18
20
22
24
Age (ka)
Figure 21 (a) Lidar-derived hillshade image showing left-lateral stream displacements along the east-central Garlock fault. (b) Histogram
illustrating measured mean left-lateral displacements of 223 landforms along the Garlock fault from prehistoric earthquakes. Data are from McGill
and Sieh (1991). (c) Probability distribution plots of earthquake frequency for the Garlock fault (black lines) and faults in the Mojave (gray lines)
for the last 24 ka. Reproduced from Dawson, T.E., McGill, S.F., Rockwell, T.E., 2003. Irregular recurrence of paleoearthquakes along the Central
Garlock fault near El Paso Peaks, California. Journal of Geophysical Research 108(B7), 2356.
(Yin and Harrison, 2000). Underthrusting of the Indian plate

beneath the Himalaya accommodates around half of its
B45 mm yr 1 northward movement, but much of the remaining movement is accommodated by lateral extrusion of
the Tibetan Plateau mostly along crustal-scale strikeslip faults
(Tapponnier and Molnar, 1976, 1977; Molnar and Tapponnier, 1978; Armijo et al., 1986; Tapponnier, et al., 2001;
Kreemer et al., 2003; Kirby et al., 2007; Molnar and Stock,
2009; Taylor and Peltzer, 2006; Meade, 2007). Across Tibet,
the major strikeslip fault systems include the Karakoram,
Kunlun, and Altyn Tagh, while the Red River and Chaman
faults are the major systems along the margins of the Himalayan-Tibetan orogen (Figure 28). However, deformation
persists significantly beyond Tibet into the interior of Central
Asia the Gobi, where continental scale strikeslip faults such
as the Gobi Altai-Tien Shan, North Gobi, and Mongolian Altai
fault systems are actively producing mountains in zones of

transpression, which have average elevations of 20003000 m
asl and some exceeding 3500 m asl (Cunningham et al., 1997;
Figure 29).
The evolution and rates of displacement along these continental scale strikeslip faults are hotly debated. Of particular
note is the Karakoram fault (Figure 30; Robinson 2009a, b).
Robinson (2009a) pointed out that there are two different
models to interpret the Karakoram fault: (1) a long-lived,
stable feature analogous to a plate boundary structure or (2)
an evolving fault system with limited displacement
(o200 km), either acting as a transfer structure linking trust
belts in the Pamir and western Tibet and/or accommodating
outward radial growth of the Himalayan arc. Slip rates along
the Karakoram fault are highly contested, both within geologic/geomorphic studies and between geologic/geomorphic
42 S
168 E
57
172 E
Australian plate
ine
Alp
lt
fau
d)
an
ut
So
(
nd
sl
hI
Pacific plate
46 S
ala
ew
Ze
N
200 km
Figure 22 Digital elevation model showing the Australia-Pacific Plate Boundary along the South Island of New Zealand and the trace of the
Alpine fault.
and geodetic studies (e.g., Chevalier et al., 2005; Brown et al.,

2006). The Altyn Tagh fault is another major strikeslip fault
that has attracted much attention and controversy (Washburn
et al., 2001, 2003; Meriaux et al., 2004, 2005; Cowgill, 2007;
Cowgill et al., 2009; Gold et al., 2001). Cowgill (2007)
pointed out that based on geomorphic studies, the slip rate
can vary by a factor of 1.25 for the same site on the Altyn
Tagh fault, depending on whether the reconstruction uses the
age of the upper or lower terrace for the initiation of riser
offset. Figure 31 shows Cowgills (2007) geomorphic indices
for evaluating the suitability of upper- versus lower-terrace
reconstruction. Moreover, Cowgill et al. (2009) have shown,
using displaced terraces rises, that at least for the last 46 ka,
the Altyn Tagh faults has slipped at only 149 mm yr 1, which
is consistent with the rates derived from geodetic, paleoseismic, and geologic measurements.
5.4.4.5
Alaska
Four major strikeslip faults are present in Alaska: Queen

Charlotte Islands; Fairweather; Denali; and Totschunda
faults (Figure 32(a)). These accommodate strain due to Pacific-North America plate motion in southeast Alaska, which is
essentially parallel to the Queen Charlotte Islands strikeslip
fault (Molnar and Dayem, 2010). Slip on the Denali and
Totschunda faults accommodates much of the relative plate
motion that is not absorbed by subduction in southern
Alaska. Displaced moraines along the Denali fault help
define the offset, which ranges from about 6 to 12 mm yr 1
(Matmon et al., 2006; Meriaux et al., 2009; Figure 32(b)).
5.4.4.6
North and Eastern Anatolian-Dead Sea Fault

System
Impressive strikeslip faults, the North and Eastern Anatolian

faults, traverse Turkey. The Dead Sea fault system essentially
transfers slip south of the Eastern Anatolian fault (Garfunkel
et al., 1981; Rotstein et al., 1991; ten Brink et al., 1999).
Figure 33 shows some of the characteristics of the earthquake
geology along the North Anatolian fault.
5.4.4.7
Miscellaneous
Other notable strikeslip faults include the Sumatra fault, the

Longitudinal Valley fault of Taiwan, the Philippine Fault, and
numerous faults that traverse Iran. The evolution and history
of these faults is not well defined, but good summaries are
provided in Molnar and Dayem (2010).
5.4.5
Important Questions and Future Directions
Although numerous studies have addressed continental transform plate boundary deformation from tectonic geomorphology, among other perspectives, there is much work still to be
done in order to fully understand how these fault systems behave and evolve through time and space. As new technology
improves our ability to detect geomorphic signatures of deformation in the landscape and the determination of ages and
rates, so will our ability to further decipher the tectonic and
geomorphic processes acting along transform and other plate
boundaries. Below is a discussion of some of the key questions
facing modern studies of active continental deformation.
5.4.5.1
Constancy of Seismic Strain Accumulation and

Release
One of the key questions in modern active tectonics and tectonic geomorphology is the degree to which the rates of deformation are constant or nonconstant in time and space. This
is of particular interest along major plate boundaries, where
the majority of strain is accommodated in the lithosphere.
72260 W
18290N
58
200 m
200 m
18290N
(a)
Figure 23 Oblique aerial photograph of geomorphic features along

the Alpine fault near the Cascade River (view is looking N from near
Woodhen Creek toward the lower Cascade River). Sag ponds, a
pressure ridge, and dextrally offset streams are clearly visible along
the fault trace. White arrows denote the trace of the fault. This photo
is from the southernmost onshore Alpine fault, where the fault is
primarily strikeslip but with a slight NW-side up offset (in contrast
to further north, where it is dextral-reverse SE-side up). Photograph
courtesy of Nicolas C. Barth.
(b)
72260 W
Figure 25 Lidar images of a stretch of the Enriquillo-Plantain

Garden fault in Haiti, trending eastwest in the lower third of the
figure. (a) Vegetation cover preserved in the processed data and (b)
vegetation removed after data processing.
80 W
70 W
NORTH AMERICAN PLATE

N
Bahamas bank
200 km
20 N
CUBA
n troug
Cayma
GONAVE MICROPLATE
HISPANOLA
a
Be
at
CARIBBEAN PLATE
Muertos Trough
Ri
dg
Jamaica
Figure 24 Digital elevation model showing the Caribbean-North America Plate Boundary and the traces of major transform faults.
Comparisons of short-term (decadal) geodetic data and longterm (103106 years) geologic plate motion data indicate that
the rates of strain storage and release are relatively constant
along most plate boundaries over a wide range of timescales
(e.g., Sella et al., 2002; Argus and Gordon, 2001). On parts of
plate boundary fault zones where both detailed geologic and
(a)
(b)
Figure 26 Oblique aerial photographs showing the geomorphic

expression of the Enriquillo-Plantain Garden fault in Haiti (US
Geological Survey photographs by Ken Hudnut). (a) Shutter ridges.
(b) Sag pond. These classic fault-related landforms are characteristic
of many continental transform and strikeslip faults. White arrows
delineate fault trace.
59
geodetic rate data are available, such as the central San Andreas and parts of the eastern California shear zone-Walker
Lane belt, the rates of strain release appear to be more or less
constant over Pleistocene to Recent timescales (Sieh and
Jahns, 1984; Argus and Gordon, 2001; Bennett et al., 2003;
Frankel et al., 2007a; Lee et al., 2009). However, along other
parts of fault systems, such as the Mojave section of the eastern
California shear zone (Oskin et al., 2008) and the central
Walker Lane (Frankel et al., 2007b; Figure 10) rates of strain
release vary over the late Quaternary.
Observations suggesting discrepancies between long- and
short-term rates of deformation raise some basic questions about
how strain is distributed throughout the lithosphere along plate
boundaries, including: (1) how temporally and spatially constant rates of strain accumulation are? (2) are geologic slip rates
averaged over thousands to millions of years compatible with
short-term geodetic rates, or are secular variations in rates of
deformation common? (3) if mismatches between short- and
long-term rate data do occur, over what temporal and spatial
scales do the strikeslip operate? (4) are strain transients localized features that characterize regions of relative structural
complexity or are they characteristic of plate boundaries as a
whole? or (5) are rate discrepancies simply a function of the time
interval over which these processes are measured (e.g., Gardner
et al., 1987; Dixon et al., 2003; Friedrich et al., 2004; Bennett
et al., 2004; Wernicke et al., 2000, 2004, 2005; Dolan et al.,
2007; Nicol et al., 2009)?
Data on the manner in which strain is distributed across
plate boundaries in time and space are necessary to understand the complex behavior of plate boundary fault systems
and the lithospheric deformation that they accommodate.
Such analyses require a comparison of deformation rate data
over a wide range of temporal and spatial scales using both
short-term rate data from GPS measurements and longer-term
tectonogeomorphic data to determine an integrated strain
budget for study regions. Integrating geophysical, geologic,
and geomorphic data in this way will lead to a better understanding of plate boundary tectonic processes over a wide
range of temporal and spatial scales.
5.4.5.2
Quantifying Deformation Rates
Although by and large, it is now known what kinds of measurements are necessary to quantify both short- and long-term
Figure 27 Digital elevation model of the Venezuela Andes showing the trace of the Bocono fault (pink line). The red line highlights the extent of
Pleistocene glaciation.
Alta
iM
TF
13 7
JHF
NTSF
ou
T
M
M
nt
0
02
ai
ns
Ind
o-
M
A
Ga
ng
200 km
ITS
MB
eti
10
m Ba
Kun L
u
sin
TIBET
15
53
Tan
ggu
la S
JS
han
157 A
6
SR
Nyainqentanglha
R
Y A
11
HF
KLF
n Sha
XF
BNS
KJF
I
18
Qiada
Kun
NQ
S
a
Na
nS
nS
han
ha
n
Qili
n
ATFLun ShaAKMS
ATF
20
KS
untains
21
KT
ra
ko
ra
m
ltai Mo
NGF
GTFS
30
Tarim Basin
Ka
ST
32
in
SF
s
Ku
KF
du
10
Pamirs
Gobi A
tain
10 5
Tien Shan
22 6
oun
AF
Junggar
Basin
LS
F
60
<5
Shan
MCT
cP
lai
21.5 1.5
Figure 28 Digital elevation model of the Himalayan-Tibetan orogen showing the major strikeslip faults and other structures. Adapted from
Owen, L.A., 2004. Cenozoic evolution of global mountain systems. In: Owens, P.N., Slaymaker, O. (Eds.), Mountain Geomorphology. Edward
Arnold (Publishers) Ltd., London, pp. 3358. Estimates of late Quaternary strikeslip, convergence, and extension rates are shown in mm yr 1.
AKMS Ayimaqin-Kunlun-Mutztagh Suture; ASRR Ailao Shan-Red River Shear Zone; AF Altai Fault; ATF Altyn Tagh fault; BNS Bangong
Nujiang Suture; GTFS Gobi-Tien Shan Fault System; HF Haiyuan fault; ITS Indus Tsangpo Suture; JHF Junggar Hegen Fault; JS Jinsha
Suture; KF Karakoram Fault; KJFZ Karakoram Jiali Fault Zone; KLF Kunlun fault; KS Kudi Suture; LSF Longmen Shan Fault; MBT Main
Boundary Thrust; MCT Main Central Thrust; MKT Main Karakoram Thrust (Shyok Suture Zone); MMT Main Mantle Thrust; NGF North
Gobi fault; NQS North Qilian Suture; NTSF North Tien Shan fault; STSF South Tien Shan fault; TFF Talus-Fergana Fault;
XF Xianshuihe Fault.
rates of deformation in order to make meaningful comparisons between the two, uncertainties in rate determinations
need to be accurately characterized and reported. The key to
effectively characterizing these rates of deformation along
faults lies in accurately measuring displacements and their
associated uncertainties with new technologies such as
lidar and objectively combining these measurements with
geochronologically determined dates and associated errors
(e.g., Bird, 2007; Zechar and Frankel, 2009; McGill et al., 2009;
Gold and Cowgill, 2011).
The high resolution of lidar data is allowing even subtle
fault offsets to be measured with incredible accuracy (e.g.,
Oskin et al., 2007; Frankel et al., 2007a,b; Arrowsmith and
Zielke, 2009). However, accurate assessment of the uncertainty
in these measurements remains a problem. Commonly, metrics such as the width of offset channels or the standard deviation of the multiple offset piercing point are used to report
displacement uncertainty (e.g., Frankel et al., 2007b). Yet,
these sorts of measurements are only semiquantitative, characteristically subjective, and somewhat unsatisfying. Improvements in the objective measurement and quantification
of horizontal offsets along transform plate boundaries are
being made, however (e.g., Zielke, 2011). Zielke (2011) developed a Matlab code (LaDiCaoz), which uses a statistical
goodness-of-fit algorithm to match the morphology of channels offset horizontally across strikeslip faults, and report the
best-fit, maximum, and minimum displacement values. Although still in the early stages of development, freely available
tools such as the LaDiCaoz are critical to improving displacement histories along active faults.
Furthermore, the accurate and objective calculation and
reporting of uncertainties in geologically and geomorphically
determined slip rates is critical to making meaningful rate
comparisons both between faults and across various temporal
and spatial scales. Quantitative slip rate estimates are essential
to understanding crustal deformation processes and assessing
seismic hazard. The two fundamental ingredients for computing slip rates estimates of the age of the offset landform
or deposit and displacement of the fault of interest contain
uncertainty, and therefore, slip rates are inherently uncertain.
Because standards by which to compute and report slip rates
have been lacking, slip rate data have been presented inconsistently and are frequently ambiguous. In particular, slip
rate uncertainty is commonly insufficiently characterized or
recorded. To date, two efforts have been aimed at objectively
quantifying slip rates and objectively calculating uncertainties
in these rates. The first uses a rigorous, probabilistic approach
to computing and reporting fault slip rates that involves
combining probability density functions (PDF) of ages and
offsets to produce a slip rate PDF containing error estimates
(e.g., Zechar and Frankel, 2009). The second approach uses
a Monte Carlo simulation to determine geologically
80
100
61
120
RUSSIA
L. BAIKAL
BF
GAF MONGOLIA
F
HS FF
50
TIEN SHAN
ATF
TB
RUSSIA
QS
/NS
PAMIRS
SHAM 4.1
TIBET
MH
T
30
CHINA
RR
F
4 cm/yr
ORLK 0.1
UZER 0.6
INDIA
LIST 1.3
10
ERZN 1.0
50
KIAT 3.5
SA
BF
1905 Mw 8.2
ULGI 4.4
FF
KNF
HS
MA
JF
OF
TN
CA
ULAA 2.9
1905 BOL2
4.2
Mw 8.3
TSET 3.8
HD
AH
HOVD
4.4
MDL
ULAB 2.2
ALTA 4.7
1931
Mw 8.0
BH
1957
Mw 8.3
JB
GAF
GA
MONGOLIA
URUM 9.0
DALA 5.3
TS
CHINA
40
TB
500 km
ATF
90
100
110
Figure 29 Topography, main active faults, GPS vectors in Mongolia and adjacent China plotted on shuttle radar topography mission (SRTM)
topography and focal mechanism plots (beachballs) for historical earthquakes. Adapted from Frankel, K.L., Wegmann, K.W., Bayasgalan, A., et al.,
2010b. Late Pleistocene slip rate of the Hoh Serh-Tsagaan Salaa fault zone, Mongolian Altai and intracontinental deformation in central Asia.
Geophysical Journal International 183, 11341150. Arrows represent GPS vectors (velocities in mm yr 1 are numbers adjacent) relative to stable
Eurasia with 95% confidence ellipses. Digital topography depicted in the inset map of Asia is from the ETOPO2 2-minute dataset. Faults: AH Ar
Hotol fault; ATF Altan Tagh fault; BF Bulnay fault; BH Batatar Hyarhan fault; FF Fu-yun fault; GAF Gobi Altai fault; HSF Hoh Serh fault;
JF Jid fault; KNF Kha-us-Nuur fault; MHT Main Himalayan thrust; OF Olgiy fault; QS/NS Qilian Shan/Nan Shan fault; RRF Red River
fault; TNT Tolbo Nuur thrust. Mountain Ranges: CA Chinese Altai; HD Hangay Dome; GA Gobi Altai; MA Mongolian Altai; SA Siberian
Altai; TS Tien Shan. Basins: JB Junggar Basin; MDL Mongolian Depression of Lakes; TB Tarim Basin.
acceptable slip rates that can fit the range of displacement and
age data from offset landforms (e.g., Gold and Cowgill, 2011).
Both of these methods provide robust treatments of age and
displacement data to determine rates and their uncertainties.
Continued efforts to develop statistical techniques that more

accurately portray deformation rates and their associated uncertainties are critical to further assessing the temporal and
spatial patterns of deformation.
62
(a)
(b)
(d)
(c)
(e)
Figure 30 (a) Simplified tectonic map of the Karakoram Fault and associated faults and sutures at the western end of the HimalayanTibetan
orogen. Offset exposures of the Aghil formation from Robinson, A.C., 2009b. Geologic offsets across the northern karakoram fault: implications
for its role and terrane correlations in the western Himalayan-Tibet orogen. Earth and Planetary Science Letters 286, 158170. IYS IndusYalu
suture; BNS BangongNujiang suture; JS Jinsha suture; KS Kunlun suture; KuS Kudi Suture; SS Shyok Suture; RPZ Rushan Pshart
Zone; TS Tanymas Suture. Views of (b) the Nubra valley and (c) Pangong Tso in northern India that follows the traces of Karakoram Fault; and
(d) Kongur Shan and (e) Gurla Mandata at the northern and southern terminus of the Karakoram Fault, respectively.
Offset of primary feature vs. riser:
Riser offset greater than

width of inset channel:
Inset channel
width
Riser offset
Upper-terrace reconstruction:
Lower-terrace reconstruction:
T2
T1
T2
63
T1
T1
(a) Upper terrace reconstruction
T2
(b)
Scarp morphology:
Diachronous tread abandonment:
( = topographic curvature)
Upper-terrace reconstruction:
base < crest

base > crest
base = crest
Zone of deposition
of riser base
(c)
Zone of erosion
of riser crest
or
(e)
Parallel slip vectors link surface to date:
Deflected riser:
Fault scarp
Upper terrace
(d)
Upper terrace reconstruction:

= Eroded
= True slip direction
Lower terrace
(f)
= Parallel slip vectors
= Divergent vectors
Figure 31 Geomorphic indices for evaluating suitability of upper- versus lower-terrace reconstruction. Adapted from Cowgill, E., 2007. Impact of
riser reconstructions on estimation of secular variation in rates of strikeslip faulting: revisiting the Cherchen River site along the Altyn Tagh
fault, NW China. Earth and Planetary Science Letters 254, 239255. (a) Magnitude of riser offset relative to width of channel inset into lower
tread. (b) Magnitude of riser offset relative to displacement of primary features (e.g., relict channels) on tread surface. (c) Scarp morphologies.
The bottom profile shows a symmetric scarp resulting from diffusion, the middle profile shows an asymmetric scarp with a low curvature crest
and high curvature base resulting from removal of material from the toe of the scarp during partial riser refreshment; the top profile shows an
asymmetric scarp with a high curvature crest and a low curvature base resulting from addition of loess and/or the presence of a cohesive unit in
the upper terrace. (d) Abrupt truncation versus curved riser at intersection with fault. (e) Isochronous versus diachronous tread ages. (f)
Parallelism of vectors linking riser crests and bases.
5.4.5.3
Calibrating and Cross-Checking Geochronometers
Another area that is becoming an important research focus is in

the improvement of geochronology methods. Because many
Quaternary geochronometers rely on relatively new techniques,
continual refinement of the methods is necessary. For example,
although now used widely in tectonic and geomorphic studies,
the production rates and even decay constants of TCNs are
continually being revised (e.g., Nishiizumi et al., 2007; Balco
et al., 2008). There is currently no consensus on how best to
incorporate time-variable TCN production as a function of the
changing geomagnetic field into age and denudation rate calculations (e.g., Lal, 1991; Stone, 2000; Balco et al., 2008;
Desilets and Zreda, 2003; Desilets et al., 2006; Dunai, 2001;
Lifton et al., 2005; Pigati and Lifton, 2004; Staiger et al., 2007).
In addition, recent research from a variety of geomorphic and
tectonic settings suggests that concentrations of TCNs in different grain size fractions contain important information for
interpreting exposure histories of alluvium (Niemi et al., 2005;
Belmont et al., 2007; Yanites et al., 2009; Frankel, 2010). As a
consequence, it has become increasingly clear that there is a
critical need in the geochronology, geomorphology, and tectonics communities for an improved understanding of what
variations in TCN concentrations among different alluvial grain
size fractions can tell us about the provenance and transport
history of material being sampled. This is equally important for
the interpretation and comparison of landform ages derived
from TCN concentrations in a variety of grain sizes (i.e., sand
versus pebbles versus cobbles).
Because of the relative youth of many of the techniques
used to date offset and/or deformed landforms, studies focused on the intercomparison of two or more Quaternary
geochronometers are important for technique validation and
calibration. Although rarely done (mainly due to the time and
cost involved), these sorts of comparison are crucial to better
understanding the systematics, advantages, and limitations of
various geochronologic techniques (e.g., Owen et al., 2007,
2011; DeLong and Arnold, 2007; Frankel et al., 2007a; Gold
et al., 2009; Behr et al., 2010; Fletcher et al., 2010). Furthermore, these intercomparisons can help with the interpretation
of the geomorphic context of the deposit of interest. For example, Owen et al. (2011) dated an offset alluvial fan
64
150
165 W
135
7.59.4 9.311.7
12.0 1.8
10.913.3
61
81
12.0 2.1
7.510.5
6.7 1.2
60
65 N
CANADA
ALASLKA
65 N
120 W
DENALI
FAULT
60
3.8 1.4
TOTSCHUNDA
FAULT
4.85.8
6.0 1.2
FAIRWEATHER
FAULT
55 N
45.6 2.0
55 N
u
Al e
tia
Tr
en
ch
QUEEN CHARLOTTE
ISLAND FAULT
150 W
(a)
135 W
11.2 1.2
12.4 1.3 12.1 1.3
11.2 1.2
12.2 1.3
12.1 1.3
11.9 1.5
11.5 1.2
12.8 1.4
12.8 1.4
11.9 1.3
(b)
Figure 32 (a) Map of Alaska showing the traces of the Queen Charlotte Islands, Fairweather, Denali, and Totschunda faults. Adapted from
Molnar, P., Dayem, K.E., 2010. Major intracontinental strikeslip faults and contrasts in lithospheric strength. Geosphere 6, 444467. Rates in
italics are based on geodetic measurements and those in normal fonts are based on dating of Quaternary offsets. (b) Offset moraine along the
Denali fault showing the locations where boulder and sediment samples were collected from both sides of the Fault to determine a slip rate.
Adapted from Matmon, A., Schwartz, D.P., Haeussler, P.J., Finkel, R., Lienkaemper, J.J., Stenner, H.D., Dawson, T.E., 2006. Denali fault slip rates
and Holocenelate Pleistocene kinematics of central Alaska. Geology 34, 645648. White open circles sample locations, ages in thousands of
years; red line fault trace; black line offset moraine crest; blue line offset channel (offset appears foreshortened because of photo
perspective). White arrow points to 2002 earthquake offset of B5 m. Sediment ages are in red ovals beside the moraine.
in central Death Valley, CA, using TCN and OSL methods.

Although both methods are commonly used successfully in
arid regions, the two data sets returned disparate results
(Figure 34). Berryllium-10 concentrations in quartzite clasts
from the fan surface yielded an age of B70 ka, whereas OSL
results suggested an age of 17 ka. In this case, understanding
the geomorphic context of the fan deposit (i.e., fan
morphology, soil development, etc.) was critical in assessing

which age more accurately reflected the timing of fan deposition and it became clear that OSL dates were more reasonable; the 10Be dates clearly reflected a prior exposure
history whereby the clasts were exposed on bedrock hillslopes
before erosion and transport from the source area to depocenter, or resided in older fan deposits or fill terraces for an
Black
Cumulative right-lateral
slip (m)
Sea
1939
6
TURKEY
4
1942
SYRIA
(a)
1971
1943
1957 1944
1992
1951
1949
1966
1967
0
500
Distance north
of 41 N latitude (km)
65
(b)
400
100 Istanbul
300
200
100
1967 Hendek Dzce
50 Iznik Geyve
Sapanca
1957
Yalova
100
500
400
300
100
Ruputered
Unruputered
BLACK
Karamrsel
Izmit
50
TH
NOR
200
300
SEA
1943
Esenay
400
Northeast
Anatolian
1942
Erzincan
1951 Kursunlu
ANKARA
200
100
Ezine
1939
Alaca
100
1949
1965
Yadisu
1971
East
Anatolian
Ovacik
Goy..
nuk
(c)
200
Distance east of 35 E longitude (km)
600
1992
A N A TO L I A N
Merziton
1944
500
300
400
500
600
Figure 33 Characteristics of some of the earthquake geology of the North Anatolian fault. Adapted from Stein, R.S., Barka, A.A., Dieterich, J.H.,
1997. Progressive failure on the North Anatolian fault since 1939 by earthquake stress triggering. Geophysical Journal International 128,
594604. (a) Active faults in Turkey with the North Anatolian fault in bold. GPS observations establish a 2471 mm yr 1 (1s confidence) deep
slip rate on the North Anatolian fault and reveal that the right-lateral transform describes a small circle about a (33.470.51 E, 31.171.31 N)
Anatolia-Eurasia rotation pole, which lies at the intersection of the dashed red lines. (b) Cumulative right-lateral slip associated with MZ6.7
earthquakes; the sequence ruptured from warm to cool colors. Slip in the 1949, 1966, and 1971 shocks is approximate, and the reverse-slip
component for the Ezine and Kursualu faults is not shown. (c) The region inscribed by the solid red line in (a) is projected relative to the
Anatolia-Eurasia rotation pole, so that a transform fault would strike due eastwest; the North Anatolian fault is seen to deviate less than 40 km
from being a simple right-lateral transform.
5.4.6
10Be
ages (n = 18)
Relative probability
OSL ages (n = 7)
50
100
150
Age (ka)
200
250
Figure 34 Probability density functions for 10Be dates for surface

cobbles on, and OSL dates for sediments in latest Pleistocene alluvial
fans in the Mud Canyon area of Death Valley, highlighting the
uncertainty associated with the different data sets. Adapted from
Frankel, K.L., Dolan, J.F., Owen, L.A., Ganev, P., Finkel, R.C., 2011.
Spatial and temporal constancy of seismic strain release along an
evolving segment of the Pacific-North America plate boundary. Earth
and Planetary Science Letters 304, 565576.
extended period of time during transport. In any case, this

example underscores the importance of using more than one
Quaternary geochronological methods and a full understanding of the geomorphic and geologic context of the
landform being dated is important.
Conclusions
Tectonic geomorphology is a critical component to studies of

continental transform plate boundaries. Analysis of geomorphic signatures in the landscape provides a unique opportunity to investigate rates of deformation over timescales
ranging from the late Holocene to early Pleistocene, or even
late Pliocene. Using measurements of landscape features such
as offset stream channels, marine and fluvial terraces, and displaced alluvial fans can help reconstruct the deformational
history of continental transform plate boundary structures over
these time periods. Combining measurements of deformation
with geochronologically determined ages of offset landforms
allows for rates and patterns of deformation to be deciphered
over intermediate- and long-term timescales. Comparison of
these geologically and geomorphically determined rates with
short-term rate data provided by tectonic geodesy allows for an
integrated view of the rates and patterns of plate boundary
strain. Advances in remotely sensed data and geochronology, as
well as geodetic methods and the quantification of uncertainties in deformation-rate data, are helping to determine rates of
deformation both more accurately and precisely. However, a
number of key active tectonic issues remain unresolved, such as
the degree to which strain accumulation and release is constant
in both time and space along continental transform (and other)
plate boundaries and how the patterns and rates of deformation change through time, both on individual structures as
well as across entire plate boundary systems as plate boundaries
evolve. Ultimately, geomorphic data from continental transform plate boundaries will provide the critical clues to
66
understanding the distribution of strain in the lithosphere over

multiple temporal and spatial scales.
Acknowledgments
A number of figures and photographs were graciously provided by Ken Hudnut, George Hilley, Olaf Zielke, Peter Koons,
Nicolas Barth, and Plamen Ganev. We thank Tim Phillips for
help with drafting many of the figures. Lidar data for the
western United States were collected by the National Center
for Airborne Laser Mapping (NCALM) and are based on services provided by the Plate Boundary Observatory operated by
UNAVCO for EarthScope (http://www.earthscope.org) and
supported by the US National Science Foundation (EAR0350028 and EAR-0732947). The Haiti lidar data collection
was funded by the World Bank and the Global Facility for
Disaster Recovery. All lidar data used in this paper are publicly
available from the OpenTopography Facility (http://www.
opentopgraphy.org), which is supported by the National Science Foundation (award numbers 0930731 and 0930643).
Zach Lifton and Jeffrey Knott provided insightful reviews,
which helped clarify and improve the paper.
References
Argus, D.F., Gordon, R.G., 2001. Present tectonic motion across the Coast Ranges
and San Andreas fault system in central California. Geological Society of
Armijo, R., Tapponnier, P., Mercier, J.L., Hanz, T.-L., 1986. Quaternary extension in
southern Tibet: field observations and tectonic implications. Journal of
Arrowsmith, J.R., Zielke, O., 2009. Tectonic geomorphology of the San Andreas
fault zone from high resolution topography: an example from the Cholame
segment. Geomorphology 113, 7081.
Bacon, S., Pezzopane, S.K., 2007. A 25 000-year record of earthquakes on the
Owens Valley fault near Lone Pine, California: implications for recurrence
intervals, slip rates, and segmentation models. Geological Society of America
Bulletin 119, 823847.
Balco, G., Stone, J.O., Lifton, N.A., Dunai, T.J., 2008. A complete and easily
accessible means of calculating surface exposure ages or erosion rates from
10
Be and 26Al measurements. Quaternary Geochronology 3, 174195.
Barnes, P.M., 2009. Postglacial (after 20 ka) dextral slip rate of the offshore Alpine
fault, New Zealand. Geology 37, 36.
Beanland, S., Clark, M.M., 1994. The Owens Valley fault zone, eastern California,
and surface rupture associated with the 1872 earthquake. United States
Geological Survey Bulletin 1982, 29 pp.
Beaumont, C., Kamp, P.J.J., Hamilton, J., Fullsack, P., 1996. The continental
collision zone, South Island, New Zealand; comparison of geodynamical models
and observations. Journal Geophysical Research 101, 33333359.
Behr, W.M., Rood, D.H., Fletcher, K.E., et al., 2010. Uncertainties in slip rate
estimates for the Mission Creek strand of the southern San Andreas fault at
Biskra Palms Oasis, southern California. Geological Society of America Bulletin
122, 13601377. http://dx.doi.org/10.1130/B30020.1.
Belmont, P., Pazzaglia, F.J., Gosse, J.C., 2007. Cosmogenic 10Be as a tracer for
hillslope and channel sediment dynamics in the Clearwater River, western
Washington State. Earth and Planetary Science Letters 264, 123135.
Bennett, R.A., Friedrich, A.M., Furlong, K.P., 2004. Codependent histories of the
San Andreas and San Jacinto fault zones from inversion of fault displacement
rates. Geology 32, 961964.
Bennett, R.A., Rodi, W., Reilinger, R.E., 1996. Global Positioning System constraints
on fault slip rates in southern California and northern Baja, Mexico. Journal of
Geophysical Research 101(21), 21,94321,960.
Bennett, R.A., Wernicke, B.P., Niemi, N.A., Friedrich, A.M., Davis, J.L., 2003.
Contemporary strain rates in the northern Basin and Range province from GPS
data. Tectonics, 22. http://dx.doi.org/10.1029/2001TC001355.
Bird, P., 2007. Uncertainties in long-term geologic offset rates of faults: General
principles illustrated with data from California and other western states.
Geosphere 3, 577595.
Bishop, K.M., Clements, S., 2006. Poverty Hills, Owens Valley,
CaliforniaTranspressional uplift or ancient landslide deposits. Environmental
and Engineering Geoscience 12, 301314.
Blisniuk, K., Rockwell, T., Owen, L.A., Oskin, M., Lippincott, C., Caffee, M.W.,
Dortch, J., 2010. Late Quaternary slip rate gradient defined using highresolution topography and 10Be dating of offset landforms on the southern San
Jacinto Fault zone, California. Journal of Geophysical Research 115, B08401.
Bourdon, B., Turner, S., Henderson, G.M., Lundstrom, C.C., 2003. Introduction to
U-series geochemistry. In: Bourdon, B., Henderson, G., Lundstrom, C., Turner,
S. (Eds.), Uranium-series geochemistry. Reviews in Mineralogy and
Geochemistry, 52, pp. 121.
Brown, E.T., Molnar, P., Bourles, D.L., 2006. Comment on Slip-rate measurements
on the Karakorum Fault may imply secular variations in fault motion. Science
309, 1326b.
Bull, W.B., 2007. Tectonic Geomorphology of Mountains. Blackwell, Oxford, 316 pp.
Bull, W.B., 2009. Tectonically Active Landscapes. Blackwell, Oxford, 320 pp.
Burchfiel, B.C., Stewart, J.H., 1966. Pull-apart origin of the central segment of
Death Valley, California. Geological Society of America Bulletin 77, 439442.
Calais, E., Freed, A., Mattioli, G., et al., 2010. Transpressional rupture of an
unmapped fault during the 2010 Haiti Earthquake. Nature Geoscience 3,
794799.
Calais, E., Mazabraud, Y., Mercier de Lepinay, B., Mann, P., Mattioli, G., Jansma, P.,
2002. Strain partitioning and fault slip rates in the northeastern Caribbean from
GPS measurements. Geophysical Research Letters, 29. http://dx.doi.org/10.1029/
2002GL015397.
Carter, B.A., 1980. Quaternary displacement on the Garlock fault, California. In: Fife,
D.L., Brown, A.R. (Eds.), Geology and Mineral Wealth of the California Desert.
South Coast Geological Society, Santa Ana, CA, pp. 457466.
Carter, W.E., Shrestha, R.L., Slatton, K.C., 2007. Geodetic laser scanning. Physics
Today 60, 4147.
Chavelier, M.-L., Ryerson, R.J., Tapponnier, P., Finkel, R.C., Van der Woerd, J.,
Haibing, L., Qing, L., 2005. Slip-rate measurements on the Karakorum fault may
imply secular variations in fault motion. Science 307, 411414.
Chen, L., Khan, S.D., 2010. InSAR observation of the strikeslip faults in the
northwest Himalayan frontal thrust system. Geosphere 6, 731736.
Chevalier, M.-L., Ryerson, F.J., Tapponnier, P., Finkel, R.C., Van Der Woerd, J.,
Haibing, L., Qing, L., 2005. Slip-rate measurements on the Karakoram Fault may
Cowgill, E., 2007. Impact of riser reconstructions on estimation of secular variation
in rates of strikeFslip faulting: revisiting the Cherchen River site along the
Altyn Tagh fault, NW China. Earth and Planetary Science Letters 254, 239255.
Cowgill, E., Gold, R.D., Chen, X., Wang, X.-F., Arrowsmith, J.R., Southon, J.R.,
2009. Low Quaternary slip rate reconciles geodetic and geologic rates along the
Altyn Tagh fault, northwestern Tibet. Geology 37, 647650.
Christie-Blick, N., Biddle, K.T., 1985. Deformation and basin formation along strikeslip faults. In: Biddle, K.T., Christine-Blick, N. (Eds.), Strike-slip Deformation,
Basin Formation, and Sedimentation. Society of Economic Paleontologists and
Mineralogists, Tulsa, OK, Special Publication 37, pp. 134.
Cunningham, W.D., Windley, B.F., Owen, L.A., Barry, T., Dorjnamjaa, D.,
Badamgarav, G., 1997. Geometry and style of partitioned deformation within a
Late Cenozoic transpressional zone in the eastern Gobi Altai Mountains,
Mongolia. Tectonophysics 277, 285306.
Davis, G.A., Burchfiel, B.C., 1973. Garlock faulk: an inter-continental transform
structure, southern California. Geological Society of America Bulletin 84,
14071422.
Dawson, T.E., McGill, S.F., Rockwell, T.E., 2003. Irregular recurrence of
paleoearthquakes along the Central Garlock fault near El Paso Peaks, California.
Journal of Geophysical Research 108(B7), 2356.
DeLong, S.B., Arnold, L.J., 2007. Dating alluvial deposits with optically stimulated
luminescence, AMS 14C and cosmogenic techniques, western Transverse
Ranges, California, USA. Quaternary Geochronology 2, 129136.
DeMets, C., Dixon, T.H., 1999. New kinematic models for PacificNorth America
motion from 3 Ma to present, I: evidence for steady state motion and biases in
the NUVEL-1A model. Geophysical Research Letters 26, 19211924.
Desilets, D., Zreda, M., 2003. Spatial and temporal distribution of secondary
cosmic-ray nucleon intensities and applications to in-situ cosmogenic dating.
Earth and Planetary Science Letters 206, 2142.
Desilets, D., Zreda, M., Prabu, T., 2006. Extended scaling factors for in situ
cosmogenic nuclides: new measurements at low latitude. Earth and Planetary
Dixon, T.H., Miller, M., Farina, F., Wang, H., Johnson, D., 2000. Present-day
motion of the Sierra Nevada block and some tectonic implications for the Basin
and Range province, North America Cordillera. Tectonics 19, 124.
Dixon, T.H., Norabuena, E., Hotaling, L., 2003. Paleoseismology and global
positioning system: earthquake-cycle effects and geodetic versus geologic fault
slip rates in the eastern California shear zone. Geology 31, 5558.
Dixon, T.H., Robaudo, S., Lee, J., Reheis, M.C., 1995. Constraints on presentday Basin and Range deformation from space geodesy. Tectonics 14, 755772.
Dokka, R.K., Travis, C.J., 1990. The role of the Eastern California shear zone in
accommodating PacificNorthAmerican plate motion. Geophysical Research
Letters 17, 13231326.
Dolan, J.F., Bowman, D.D., Sammis, C.G., 2007. Long-range and long-term fault
interactions in southern California. Geology 35, 855858.
Dortch, J.M., Owen, L.A., Schoenbohm, L.M., Caffee, M.W., 2011. Asymmetrical
erosion and morphological development of the Ladakh Range, northern India.
Dunai, T., 2001. Influence of secular variation of the magnetic field on production
rates of in situ produced cosmogenic nuclides. Earth and Planetary Science
Letters 193, 197212.
Dunai, T.J., 2010. Cosmogenic Nuclides: Principles, Concepts and Applications in
the Earth Surface Sciences. Cambridge University Press, Cambridge, 187 pp.
Faulds, J.E., Henry, C.D., Hinz, N.H., 2005. Kinematics of the northern Walker Lane:
an incipient transform fault along the Pacific-North America plate boundary.
Geology 33, 505508.
Fialko, Y., 2006. Interseismic strain accumulation and the earthquake potential on
the southern San Andreas fault system. Nature 441, 968971.
Field, E.H., Dawson, T.E., Felzer, K.R., et al., 2009. Uniform California earthquake
rupture forecast, version 2 (UCERF 2). Bulletin of the Seismological Society of
America 99, 20532107.
Fletcher, K.E.K., Sharp, W.D., Kendrick, K.J., Behr, W.M., Hudnut, K.W., Hanks, T.C.,
2010. 230Th/U dating of a late Pleistocene alluvial fan along the southern San
Andreas fault. Geological Society of America Bulletin 122, 13471359.
Frankel, K.L., 2007. Fault slip rates, constancy of seismic strain release, and
landscape evolution in the eastern California shear zone. Ph.D. dissertation, Los
Angeles, University of Southern California, 179 pp.
Frankel, K.L., 2010. Establishing rates of deformation along an evolving segment of
the Pacific-North America plate boundary with cosmogenic nuclide and optically
stimulated luminescence geochronology. Geological Society of America Annual
Meeting Abstracts with Programs, Denver, Abstract.
Frankel, K.L., Brantley, K.S., Dolan, J.F., et al., 2007b. Cosmogenic 10Be and 36Cl
implications for transient strain in the eastern California shear zone. Journal of
Geophysical Research, 112. http://dx.doi.org/10.1029/2006JB004350.
Frankel, K.L., Dolan, J.F., 2007. Characterizing arid region alluvial fan surface
of Geophysical Research, 112. http://dx.doi.org/10.1029/2006JF000644.
Frankel, K.L., Dolan, J.F., Owen, L.A., Finkel, R.C., Hoeft, J.S., 2007a. Spatial
variations in slip rate along the Death Valley-Fish Lake Valley fault system from
LiDAR topographic data and cosmogenic 10Be geochronology. Geophysical
Research Letters, 34. http://dx.doi.org/10.1029/2007GL030549.
Frankel, K.L., Dolan, J.F., Owen, L.A., Ganev, P., Finkel, R.C., 2011. Spatial and
temporal constancy of seismic strain release along an evolving segment of the
Pacific-North America plate boundary. Earth and Planetary Science Letters 304,
565576.
Frankel, K.L., Lee, J., Bishop, K., Dawers, N., Ganev, P., Unruh, J., Owen, L.A.,
2010a. Miocene Quaternary tectonic evolution of the northern eastern
California shear zone. In: Clifton, H.E., Ingersoll, R.V. (Eds.), Geologic
excursions in California and Nevada: Tectonics, stratigraphy and hydrogeology.
Pacific Section. SEPM (Society for Sedimentary Geology) Book, vol. 108, pp.
173231.
Frankel, K.L., Wegmann, K.W., Bayasgalan, A., et al., 2010b. Late Pleistocene slip
rate of the Hoh Serh-Tsagaan Salaa fault zone, Mongolian Altai and
intracontinental deformation in central Asia. Geophysical Journal International
183, 11341150.
Friedrich, A.M., Lee, J., Wernicke, B.P., Sieh, K., 2004. Geologic context of geodetic
data across a Basin and Range normal fault, Crescent Valley, Nevada. Tectonics,
23. http://dx.doi.org/10.1029/2003TC001528.
Gan, W., Svarc, J.L., Savage, J.C., Prescott, W.H., 2000. Strain accumulation across
the eastern California shear zone at latitude 36130 N. Journal of Geophysical
Research 105, 1622916236.
Ganev, P.N., Dolan, J.F., Frankel, K.L., Finkel, R.C., 2010. Rates of extension along
the Fish Lake Valley fault and transtensional deformation in the eastern
California shear zone-Walker Lane belt. Lithosphere 2, 3349.
67
Ganev, P.N., Dolan, J.F., McGill, S.F., Frankel, K.L., 2012. Constancy of geologic
slip rate along the central Garlock fault: implications for strain accumulation
and release in southern California. Geophysical Journal International 190,
745760.
Gardner, T.W., Jorgensen, D.W., Shuman, C., Lemieux, C., 1987. Geomorphic and
tectonic process rates: effects of measured time interval. Geology 15, 259261.
Garfunkel, Z., Zak, I., Freund, R., 1981. Active faulting in the Dead Rift.
Gold, R.D., Cowgill, E., 2011. Derivation of fault-slip histories from dated and
displaced landforms to test for secular variation in slip. Earth and Planetary
Gold, R.D., Cowgill, E., Arrowsmith, J.R., Gosse, J., Chen, X., Wang, X.-F., 2009.
Riser diachroneity, lateral erosion, and uncertainty in rates of strikeslip faulting:
a case study from Tuzidun along the Altyn Tagh fault, NW China. Journal of
Gold, R.D., Cowgill, E., Arrowsmith, J.R., Chen, X., Sharp, W.D., Cooper, K.M.,
Wang, X.-F., 2001. Faulted terrace risers place new constraints on the late
Quaternary slip rate for the central Altyn Tagh fault, northwest Tibet. GSA
Bulletin 123, 958978.
Gosse, J.C., Phillips, F.M., 2001. Terrestrial in situ cosmogenic nuclides: theory
and application. Quaternary Science Reviews 20, 14751560.
Gourmelen, N., Dixon, T.H., Amelung, F., Schmalzle, G., 2011. Acceleration and
evolution of faults: an example from the Hunter Mountain-Panamint Valley fault
zone, eastern California. Earth and Planetary Science Letters 301, 337344.
Granger, D.E., 2006. A review of burial dating methods using 26Al and 10Be, In:
Siame, L.L., Bourles, D.L., Brown, E.T., (Eds.), In-Situ-Produced Cosmogenic
Nuclides and Quantification of Geological Processes. Geological Society of
America Special Paper 415, pp. 116.
Granger, D.E., Kirchner, J.W., Finkel, R., 1996. Spatially averaged long-term erosion
rates measured from in situ-produced cosmogenic nuclide in alluvial sediment.
Guest, B., Niemi, N., Wernicke, B., 2007. Stateline fault system: a new component
of the MioceneQuaternary eastern California shear zone. Geological Society of
Guest, B., Pavlis, T.L., Golding, H., Serpa, L., 2003. Chasing the Garlock: a study
of tectonic response to vertical axis rotation. Geology 21, 553556.
Hearn, E.H., Humphreys, E.D., 1998. Kinematics of the southern Walker Lane Belt
and motion of the Sierra Nevada block, California. Journal of Geophysical
Research 103, 2703327049.
Hilley, G.E., Arrowsmith, J.R., 2008. Geomorphic response to uplift along the
Dragons Back pressure ridge, Carrizo Plain, California. Geology 36, 367370.
Hough, S.E., Hutton, K., 2008. Revisiting the 1872 Owens Valley, California,
earthquake. Bulletin of the Seismological Society of America 98, 913949.
Hudnut, K.W., Borsa, A., Glennie, C., Minster, J.-B., 2002. High-resolution
topography along surface rupture of the 16 October 1999 Hector Mine,
Calfornia, earthquake (MW 7.1) from airborne laser swath mapping. Bulletin of
the Seismological Society of America 92, 15701576.
Humphreys, E.D., Weldon, R.J., 1994. Deformation across the western United
States: a local estimate of Pacific-North America transform deformation. Journal
King, N.E., Savage, J.C., 1983. Strainrate profile across the Elsinore, San Jacinto,
and San Andreas faults near Palm Springs, California, 19731981. Geophysical
Kirby, E., Burbank, D.W., Reheis, M., Phillips, F., 2006. Temporal variations in slip
rate of the White Mountain Fault Zone, eastern California. Earth and Planetary
Kirby, E., Harkins, N., Wang, E., Shi, X., Fan, C., Burbank, D., 2007. Slip rate
gradients along the eastern Kunlun fault. Tectonics 26, TC2010. http://dx.doi.org
/10.1029/2006TC002033.
Kreemer, C., Holt, W.E., Haines, A.J., 2003. An integrated global model of presentday motions and plate boundary deformation. Geophyicsal Journal International
154, 834.
Lal, D., 1991. Cosmic ray labeling of erosion surfaces: in situ nuclide production
rates and erosion models. Earth and Planetary Science Letters 104, 424439.
Lee, J., Rubin, C.M., Calvert, A., 2001. Quaternary faulting history along the Deep
Springs fault, California. Geological Society of America Bulletin 113, 855869.
the Inyo Mountains, California: implications for the timing of extension along
the western boundary of the Basin and Range Province and distribution of
dextral fault slip rates across the eastern California shear zone. Tectonics 28. htt
p://dx.doi.org/10.1029/2008TC002295.
Lifton, N., Bieber, J., Clem, J., Duldig, M., Evenson, P., Humble, J., Pyle, R., 2005.
Addressing solar modulation and long-term uncertainties in scaling secondary
68
cosmic rays for in situcosmogenic nuclide applications. Earth and Planetary

Liu-Zeng, J., Klinger, Y., Sieh, K., Rubin, C., Seitz, G., 2006. Serial ruptures of the
San Andreas fault, Carrizo Plain, California, revealed by three-dimensional
excavations. Journal of Geophysical Research 111, B02306.
Lyon-Caen, H., Barrier, E., Lasserre, C., et al., 2006. Kinematics of the North
AmericanCaribbeanCocos plates in Central America from new GPS
measurements across the Polochic-Motagua fault system. Geophysical Research
Letters 33, L19309.
Matmon, A., Schwartz, D.P., Haeussler, P.J., Finkel, R., Lienkaemper, J.J., Stenner,
H.D., Dawson, T.E., 2006. Denali fault slip rates and Holocenelate Pleistocene
kinematics of central Alaska. Geology 34, 645648.
McClusky, S.C., Bjornstad, S.C., Hager, B.H., et al., 2001. Present-day kinematics of
the Eastern California shear zone from a geodetically constrained block model.
McGill, S.F., Sieh, K., 1991. Surficial offsets on the central and eastern Garlock
fault associated with prehistoric earthquakes. Journal of Geophysical Research
96, 21,59721,621.
McGill, S.F., Wells, S.G., Fortner, S.K., Kuzma, H.A., McGill, J.D., 2009. Slip rate of
the western Garlock fault, at Clark Wash, near Lone Tree Canyon, Mojave
Desert, California. Geological Society of America Bulletin 121, 536554.
Meade, B.J., 2007. Present-day kinematics at the IndiaAsia collision zone. Geology
35, 8184.
Meade, B.J., Hager, B.H., 2005. Block models of crustal motion in southern
California constrained by GPS measurements. Journal of Geophysical Research
110. http://dx.doi.org/10.1029/2004JB003209.
Meisling, K.E., Weldon, R.J., 1989. Late Cenozoic tectonics of the northwestern San
Bernardino Mountains, southern California. Geological Society of America
Bulletin 101, 106128.
Merritts, D., Bull, W.B., 1989. Interpreting Quaternary uplift rates at the Mendocino
triple junction, northern California, from uplifted marine terraces. Geology 17,
10201024.
Meriaux, A.-S., Ryerson, F.J., Tapponnier, P., et al., 2004. Rapid slip along the
central Altyn Tagh fault: morphochronologic evidence from Cherchen He and
Sulamu Tagh. Journal of Geophysical Research 109, B06401.
Meriaux, A.-S., Tapponnier, P., Ryerson, F.J., et al., 2005. The Aksay segment of the
northern Altyn Tagh fault: tectonic geomorphology, landscape evolution, and
Holocene slip rate. Journal of Geophysical Research 110, B04404.
Meriaux, A.-S., Sieh, K., Finkel, R.C., Rubin, C.M., Taylor, M.H., Meltzner, A.J.,
Ryerson, F.J., 2009. Kinematic behavior of southern Alaska constrained by
westward decreasing postglacial slip rates on the Denali fault, Alaska. Journal of
Geophysical Research 114, B03404.
Miller, M.M., Johnson, D.J., Dixon, T.H., Dokka, R.K., 2001. Refined kinematics of
the eastern California shear zone from GPS observations, 19931994. Journal
Molnar, P., Dayem, K.E., 2010. Major intracontinental strikeslip faults and
contrasts in lithospheric strength. Geosphere 6, 444467.
20 Ma and its implications for Tibetan mantle dynamics. Tectonics 28. TC3001,
TC3001.http://dx.doi.org/10.1029/2008TC002271
Molnar, P., Tapponnier, P., 1975. Cenozoic tectonics of Asia: effects of a continental
Molnar, P., Tapponnier, P., 1978. Active tectonics of Tibet. Journal of Geophysical
Research 83, 53615375.
Nicol, A., Walsh, J., Mouslopoulou, V., Villamor, P., 2009. Earthquake histories and
Holocene acceleration of fault displacement rates. Geology 37, 911914.
Niemi, N.A., Oskin, M., Burbank, D.W., Heimsath, A.M., Gabet, E.J., 2005. Effects of
bedrock landslides on cosmogenically determined erosion rates. Earth and
Nishiizumi, K., Imamura, M., Caffee, M.W., Southon, J.R., Finkel, R.C., McAninch,
J., 2007. Absolute calibration of 10Be AMS standards. Nuclear Instruments and
Methods in Physics Research, B 258, 403413.
Noller, J.S., Sowers, J.M., Lettis, W.R., 2000. Quaternary Geochronology: Methods
and Applications. American Geophysical Union, Washington, 582 pp.
Norris, R.J., Cooper, A.F., 2001. Late Quaternary slip rates and slip partitioning
on the Alpine fault, New Zealand. Journal of Structural Geology 23, 507520.
Oldow, J.S., 2003. Active transtensional boundary zone between the western Great
Basin and Sierra Nevada block, western U.S. Cordillera. Geology 31,
10331036.
Oldow, J.S., Singleton, E.S., 2008. Application of terrestrial laser scanning in
determining the pattern of late Pleistocene and Holocene fault displacement from
the offset of pluvial lake shorelines in the Alvord extensional basin, northern
Great Basin, USA. Geosphere 4, 536563.
Oskin, M., Iriondo, A., 2004. Large-magnitude transient strain accumulation on the
Blackwater fault, eastern California shear zone. Geology 32, 313316.
Oskin, M., Perg, L., Blumentritt, D., Mukhopadhyay, S., Iriondo, A., 2007. Slip rate
of the Calico fault: implications for geologic versus geodetic rate discrepancy in
the eastern California shear zone. Journal of Geophysical Research 112,
B03402.
Oskin, M., Perg, L., Shelef, E., Strane, M., Gurney, E., Singer, B., Zhang, X., 2008.
Elevated shear zone loading rate during an earthquake cluster in eastern
California. Geology 36, 507510.
Oswald, J.A., Wesnousky, S.A., 2002. Neotectonics and Quaternary geology of the
Hunter Mountain fault zone and Saline Valley region, southeastern California.
Owen, L.A., Bright, J., Finkel, R.C., et al., 2007. Numerical dating of a Late
Quaternary spit-shoreline complex at the northern end of Silver Lake playa,
Mojave Desert, California: a comparison of the applicability of radiocarbon,
luminescence, terrestrial cosmogenic nuclide, electron spin resonance,
U-series and amino acid racemization methods. Quaternary International 166,
87100.
Owen, L.A., Frankel, K.L., Knott, J.R., Reynhout, S., Finkel, R.C., Dolan, J.F., Lee,
J., 2011. Beryllium-10 terrestrial cosmogenic nuclide surface exposure dating of
Quaternary landforms in Death Valley. Geomorphology 125, 541557.
Peltzer, G., Crampe, F., Hensley, S., Rosen, P., 2001. Transient strain accumulation
and fault interaction in the eastern California shear zone. Geology 29, 975978.
Pigati, J.S., Lifton, N.A., 2004. Geomagnetic effects on time-integrated cosmogenic
nuclide production with emphasis on in situ 14C and 10Be. Earth and Planetary
Pigati, J.S., Quade, J., Wilson, J., Jull, A.J.T., Lifton, N.A., 2007. Development of
low-background vacuum extraction and graphitization systems for 14C dating of
old (4060 ka) samples. Quaternary International 166, 414.
Prentice, C.S., Crosby, C.J., Whitehill, C.S., Arrowsmith, J.R., Furlong, K.P.,
Phillips, D.A., 2009. GeoEarthscope LiDAR illuminates northern Californias
active faults. Eos, Transactions of the American Geophysical Union 90, 55.
Pysklywec, R.N., Beaumont, C., Fullsack, P., 2002. Lithospheric deformation during
the early stages of continental collision: numerical experiments and comparison
with South Island, New Zealand. Journal of Geophysical Research 107, 213. http
://dx.doi.org/10.1029/2001JB000252.
Reheis, M.C., Dixon, T.H., 1996. Kinematics of the eastern California shear zone:
evidence for slip transfer from Owens and Saline Valley fault zones to Fish Lake
Valley fault zone. Geology 24, 339342.
Reiners, P.W., Ehlers, T.A., 2005. Low-temperature thermochronology: techniques,
interpretations, and applications. Reviews in Mineralogy & Geochemistry 58, 622.
Reiners, P.W., Shuster, D.L., 2009. Thermochronology and landscape evolution.
Physics Today, X, 3136.
Rhodes, E.J., 2011. Optically stimulated luminescence dating of sediments over the
past 200,000 years. Annual Reviews of Earth and Planetary Sciences 39,
461488.
Robinson, A.C., 2009a. Evidence against Quaternary slip on the northern Karakorum
Fault suggests kinematic reorganization at the western end of the HimalayanTibetan orogen. Earth and Planetary Science Letters 286, 158170.
Robinson, A.C., 2009b. Geologic offsets across the northern Karakorum fault:
implications for its role and terrane correlations in the western HimalayanTibetan orogen. Earth and Planetary Science Letters 286, 158170.
Rockwell, T.K., Lindvall, S., Herzberg, M., Murbach, D., Dawson, T., Berger, G.,
2001. Paleoseismology of the Johnson Valley, Kickapoo, and Homestead Valley
faults: clustering of earthquakes in the eastern California shear zone. Bulletin of
the Seismological Society of America 90, 12001236. http://dx.doi.org/10.1785/
0119990023.
Rogers, R.D., Mann, P., 2007. Transtensional deformation of the western
CaribbeanNorth America plate boundary zone. In: Mann, P. (Ed.), Geologic and
Tectonic Development of the Caribbean Plate Boundary in Northern Central
America. Geological Society of America Special, 428, pp. 3764.
Rotstein, Y., Bartov, Y., Hofstetter, A., 1991. Active compressional tectonics in the
Jericho area, the Dead Sea Rift. Tectonophysics 198, 239259.
Savage, J.C., Lisowski, M., Prescott, W.H., 1990. An apparent shear zone trending
northnorthwest across the Mojave Desert into Owens Valley, California.
Sella, G.F., Dixon, T.H., Mao, A., 2002. REVEL: a model for Recent plate velocities
from space geodesy. Journal of Geophysical Research 107. doi:10.1029/
2000JB000033.
Sieh, K.E., 1978a. Slip along the San Anreas fault associated with the great 1857
earthquake. Bulletin of the Seismological Society of America 68, 14211448.
Sieh, K., 1984. Lateral offsets and revised dates of large earthquakes at Pallett
Creek, California. Journal of Geophysical Research 89, 76417670.
Sieh, K.E., Jahns, R.H., 1984. Holocene activity of the San Andreas fault at Wallace
Creek, California. Geological Society of America Bulletin 95, 883896.
Snyder, N.P., Whipple, K.X., Tucker, G.E., Merritts, D.J., 2003. Channel response to
tectonic forcing: field analysis of stream morphology and hydrology in the
Mendocino triple junction region, northern California. Geomorphology 53, 97127.
Spotila, J.A., House, M.A., Niemi, N.A., Brady, R.C., Oskin, M., Buscher, J.T.,
2007b. Mechanisms of transpression along the San Andreas fault, based on
patterns of bedrock uplift. In: Till, A.B., Roeske, S.M., Sample, J.C., Foster, D.A.
(Eds.), Exhumation Associated with Continental StrikeSlip Systems: Geological
Society of America Special Paper, 434.
Spotila, J.A., Niemi, N., Brady, R., House, M., Buscher, J., Oskin, M., 2007a.
Long-term continental deformation associated with transpressive plate motion:
the San Andreas fault. Geology 35, 967970.
Staiger, J., Gosse, J., Toracinta, R., Oglesby, B., Fastook, J., Johnson, J.V., 2007.
Atmospheric scaling of cosmogenic nuclide production; climate effect. Journal of
Stone, J.O., 2000. Air pressure and cosmogenic isotope production. Journal of
Summerfield, M., 1991. Global Geomorphology. Longman, Harlow, pp. 537.
Sutherland, R., Berryman, K., Norris, R., 2006. Quaternary slip rate and
geomorphology of the Alpine fault: implications for kinematics and seismic
hazard in southwest New Zealand. Geological Society of America Bulletin 118,
464474.
Sylvester, A.G., 1988. Strikeslip faults. Geological Society of America Bulletin 100,
16661703.
Tapponnier, P., Molnar, P., 1976. Slip-line theory and large-scale continental
tectonics. Nature 264, 319324.
Tapponnier, P., Molnar, P., 1977. Active faulting and tectonics in China. Journal of
Tapponnier, P., Xu, Z., Roger, F., Meyer, B., Arnaud, N., Wittlinger, G., Yang, J., 2001.
Oblique stepwise rise and growth of the Tibet Plateau. Science 294, 16711677.
Taylor, M., Peltzer, G., 2006. Current slip rates on conjugate strikeslip faults in
central Tibet using synthetic aperture radar interferometry. Journal of
Taylor, T.R., 2002. Origin and structure of the Poverty Hills, Owens Valley fault zone,
Owens Valley, California. M.S. thesis, Miami University, Miami, OH, 93 pp.
Taylor, T.R., Dilek, Y., 2001. Origin of the Poverty Hills, Owens Valley, California,
and implications for the tectonic evolution of the Owens Valley Basin. Geological
Society Abstracts with Programs 33, A-395.
ten Brink, U., Rybakov, M., Al-Zoubi, A.S., et al., 1999. Anatomy of the Dead Sea
transform. Geology 27, 887890.
Thatcher, W., 2009. How the continents deform: the evidence from tectonic geodesy.
Thatcher, W., Foulger, G.R., Julian, B.R., Svarc, J., Quilty, E., Bawden, G.W., 1999.
Present-day deformation across the Basin and Range province, western United
States. Science 283, 17141718.
Tsukamoto, S., Duller, G., Murray, A., Choi, J.-H., 2009. Application of
luminescence dating to Quaternary tectonic movements and environmental
change. Quaternary International 199, 12.
USGS, 2011. Magnitude 7.0 Haiti Region Summary. USGS Earthquake Hazards
Program. http://earthquake.usgs.gov/earthquakes/recenteqsww/Quakes/
us2010rja6.php#summary
van Calsteren, P., Thomas, L., 2006. Uranium-series dating application in natural
environmental science. Earth Science Reviews 75, 155175.
van der Woerd, J., Ryerson, F.J., Tapponnier, P., et al., 1998. Holocene left-slip rate
determined by cosmogenic surface dating on the Xidatan segment of the Kunlun
fault (Qinghai, China). Geology 26, 695698.
van der Woerd, J., Ryerson, F.J., Tapponnier, P., et al., 2000. Uniform slip-rate
along the Kunlun fault: implications for seismic behaviour and large-scale
tectonics. Geophysical Research Letters 27, 23532356.
69
van der Woerd, J., Ryerson, F.J., Tapponnier, P., et al., 2002. Uniform post-glacial
slip-rate along the central 600 km of the Kunlun fault (Tibet), from 26Al, 10Be
and 14C dating of riser offsets, and climatic origin of the regional morphology.
Vedder, J.G., Wallace, R.E., 1970. Map showing recently active breaks along the
San Andreas and related faults between Cholame Valley and Tejon Pass,
California: US, Geological Survey Miscellaneous Geologic Investigations Map 1574, scale 1:24 000.
Wallace, R.E., 1968. In: Dickinson, W.R., Grantz, A. (Eds.), Proceedings of
Conference On Geologic Problems of the San Andreas Fault System, Stanford
University publications in the Geological Sciences, vol. 11.
Wallace, R.E., 1975. The San Andreas fault in the Carrizo PlainTemblor Range region,
California. In: Crowell, J.C., (Ed.), The San Andreas Fault in Southern California.
California Division of Mines and Geology, Special Report 118, pp. 241250.
Washburn, Z., Arrowsmith, J.R., Forman, S.L., Cowgill, E., Wang, X., Zhang, Y.,
Chen, Z., 2001. Late Holocene earthquake history of the central Altyn Tagh fault,
China. Geology 29, 10511054.
Washburn, Z., Arrowsmith, J.R., Dupont-Nivet, G., Wang, X.-F., Zhang, Y.-Q., Chen,
Z., 2003. Paleoseismology of the Xorxol segment of the Central Altyn Tagh fault,
Xinjiang, China. Annals of Geophysics 5, 10151034.
Waschbusch, P., Batt, G., Beaumont, C., 1998. Subduction zones retreat and recent
tectonics of the south island of New Zealand. Tectonics 17, 267284.
Weldon, R., Scharer, K., Fumal, T., Biasi, G., 2004. Wrightwood and the earthquake
cycle: what a long recurrence record tells us about how faults work. GSA Today
14, 410.
Weldon, R.J., Scharer, K.M., Fumal, T.E., Biasi, G.P., 2004. Wrightwood and the
earthquake cycle: what a long recurrence record tells us about how faults work.
GSA Today 14, 410.
Weldon, R.J., Funal, T.E., Biasi, G.P., Scharer, K.M., 2005. Past and future
earthquakes on the San Andreas fault. Science 308, 966967.
Wernicke, B., Davis, J.L., Bennett, R.A., Normandeau, J.E., Friedrich, A.M., Niemi,
N.A., 2004. Tectonic implications of a dense continuous GPS velocity field at
Yucca Mountain, Nevada. Journal of Geophysical Research 109, B12404.
Wernicke, B., Friedrich, A.M., Niemi, N.A., Bennett, R.A., Davis, J.L., 2000.
Dynamics of plate boundary fault systems from Basin and Range Geodetic
Network (BARGEN) and geologic data. GSA Today 10, 17.
Wernicke, B., Niemi, N., Davis, J.L., Bisnath, S., Elosegui, P., 2005. Tectonic
megaoscillators: implications for large-scale stress transfer in the lithosphere.
Geological Society of America Abstracts with Programs 36.Abstract 55-13.
Wesnousky, S.G., 2005. The San Andreas and Walker Lane fault systems, western
North America: transpression, transtension, cumulative slip and the structural
evolution of a major transform plate boundary. Journal of Structural Geology 27,
15051512.
Wesnousky, S.G., Aranguren, R., Rengifo, M., Owen, L.A., Caffee, M.W., Murari,
M.K., Perez, O.J. 2012. Towards quantifying geomorphic rates of crustal
displacement, landscape development, and the age of glaciation in the
Venezuelan Andes. Geomorphology, in preparation.
Yanites, B.J., Tucker, G.E., Anderson, R.S., 2009. Numerical and analytical models
of cosmogenic radionuclide dynamics in landslide-dominated drainage basins.
Journal of Geophysical Research, 114. http://dx.doi.org/10.1029/2008JF001088.
Yin, A., Harrison, T.M., 2000. Geologic evolution of the HimalayanTibetan orogen.
Annual Reviews of Earth and Planetary Sciences 28, 211280.
Zechar, J.D., Frankel, K.L., 2009. Incorporating and reporting uncertainties in fault
slip rates. Journal of Geophysical Research Solid Earth, 114. http://dx.doi.org/
10.1029/2009JB006325.
Zielke, O. 2011. LaDiCaoz. http://www.public.asu.edu/Bozielke/programs_tools/
index.html
Zielke, O., Arrowsmith, J.R., Ludwig, L.G., Akc- iz, S.O., 2010. Slip in the 1857 and
Earlier Large Earthquakes Along the Carrizo Plain, San Andreas Fault. Science
327, 11191122.
70
Biographical Sketch
Dr. Frankel received his PhD from the University of Southern California in 2007 and, before his most untimely
death, he was an Assistant Professor in the School of Earth and Atmospheric Science at Georgia Institute of
Technology, Atlanta. Dr. Frankels research focused on the interaction between active tectonics and geomorphology, using a multidisciplinary approach that included field-based, process-oriented research in subject areas
ranging from examining the constancy of seismic strain accumulation and release along major plate-boundary
fault systems to investigating tectonic and climatic controls on range-scale erosion and exhumation.
Professor Owen received his PhD in Geomorphology from the University of Leicester, UK, in 1988. Before joining
the University of Cincinnati in 2004, he held positions at the Hong Kong Baptist University, Royal Holloway
University of London, and the University of California Riverside. Prof. Owens research focuses on the Quaternary geology and geomorphology of tectonically active mountain belts and their forelands, particularly in the
Himalayan-Tibetan orogen and the Cordilleras of North and South America. He has also undertaken research in
other tectonically active regions, including the Red Sea margin in Yemen and the Atlas and Anti-Atlas Mountains
of Morocco. In 2011, Professor Owen was awarded the Busk Medal from the Royal Geographic Society for his field
research in Quaternary history and geomorphology in tectonically active areas.
5.5 Tectonic Geomorphology of Passive Margins and Continental Hinterlands

T Blenkinsop, School of Earth and Environmental Sciences James Cook University, Townsville, QLD, Australia
A Moore, African Queen Mines Ltd., Botswana, South Africa, and Rhodes University, Grahamstown, South Africa
5.5.1
Introduction
5.5.2
Igneous and Tectonic Processes Associated with Rifting
5.5.3
Prerifting Continental Topography and Elevation
5.5.4
Postrifting Evolution of Marginal Escarpments
5.5.4.1
Kings Scarp Retreat Model Backwearing
5.5.4.2
Pinned Drainage Divide Downwearing
5.5.4.3
Downwarping of the Continental Margin
5.5.4.4
Isostasy and Flexure
5.5.4.5
Perspectives from Integrating Low-Temperature Geochronology and Numerical Modeling
5.5.4.6
Sinuosity of Escarpments
5.5.4.7
Low-Relief Passive Margins without a Marginal Escarpment
5.5.5
Evolution of Continental Hinterlands
5.5.5.1
Cyclic Erosion
5.5.5.2
Dynamic Topography
5.5.5.3
Plate Boundary Stresses and Lithospheric Buckling
5.5.5.4
Implications for Continent-Wide Erosion Cycles and the Origin of Uplifts
5.5.6
Concluding Remarks
Acknowledgments
References
71
76
78
79
79
79
80
81
83
83
83
85
85
86
87
88
89
89
89
Abstract
Passive margins, created at the margins of rifted continents, are affected by thermal, isostatic, flexural, buckling (plate
tectonic), and dynamic (mantle) stresses. Variations among these give rise to diverse topographic expressions. The geometry
of rifting also has a major effect on topography. Thus, many low-relief margins lacking a fringing escarpment occur at the
failed arms of triple junctions. Variations in lithospheric flexural rigidity influence the response of passive margins following rifting. Postrifting evolution of the continental hinterland is more readily explained by stresses related to plate
tectonic processes than by the dynamic uplift over plumes.
5.5.1
Introduction
Passive margins form where continental rifting results in plate

divergence and the formation of new oceanic crust. They
constitute B50% of continental margins today (Figure 1;
Gallagher and Brown, 1999), and they are highly significant
to humankind, both as dwelling places for a large proportion
of the worlds population and as repositories of vital resources
most notably hydrocarbons. Passive margins separate
oceanic crust from a continental hinterland, which has a
history that is closely linked to that of the adjacent margin
(e.g., Summerfield, 2000). Plate tectonics provides a first-order
framework for understanding the formation of these features
(Bishop, 2007), and yet, the evolution of their tectonic geomorphology can only be linked indirectly to plate tectonic
processes because they are commonly remote from plate
boundaries.
Some characteristic features of passive margins associated
with extrusive igneous rocks are illustrated in Figure 2(a). The
Blenkinsopand, T., Moore A., 2013. Tectonic geomorphology of passive
margins and continental hinterlands. In: Shroder, J. (Editor in Chief),
Owen, L.A. (Ed.), Treatise on Geomorphology. Academic Press, San Diego,
CA, vol. 5, Tectonic Geomorphology, pp. 7192.
shelf break, or slope, approximately at the continentocean

boundary, appears to mark the maximum Neogene sea lowstand linked to PlioPleistocene glacial advances, and joins
the continental shelf to the abyssal depths of the ocean. The
continentocean boundary is not always very well defined
(Brown et al., 2003), but at least conceptually, it corresponds
to the point at which continuous vertical dikes of oceanic crust
are joined to some continental crust. The shelf break is separated from the exposed continent by the continental shelf,
which consists of a sedimentary pile of variable thickness
underlain by continental crust. Ocean-dipping normal faults
have syn-tectonic sediments in their hanging walls; the tops of
tilted fault blocks may be eroded. The faults may be listric, and
rotate the sediments in their hanging walls (Figure 2). Postrift
sediments can overlie these features, and these sequences may
themselves be tilted and eroded. Continental margin sediments form seaward-dipping seismic reflectors characterized
by low P-wave velocities (o7 km s1) (Figure 2). Sedimentary
features of the offshore passive margin consist of fans, channels, canyons, and carbonate mounds, with evidence of
slumps, debris flows, and regional unconformities (Bagguley
and Prosser, 1999; OGrady et al., 2000; Stoker et al., 2010).
Rifted passive margins develop where disrupted continental fragments diverge essentially perpendicular to the
http://dx.doi.org/10.1016/B978-0-12-374739-6.00083-X
71
72
Tectonic Geomorphology of Passive Margins and Continental Hinterlands
GCS_WGS_1984
Datum: D_WGS_1984
Projection: Robinson
Prepared by Darin Pinto
11/30/2007
Volcanic and Non volcanic Margins
Added by Tejeev Patel
Sorced from:
Melluso et al. 2001
Geoffroy, 2005
Watkeys, 2002
Leroy, 2008
Global distribution of passive margins
Legend
Continents
Ocean
2 000 000 4 000 000
8 000 000 Meters
Passive margin
Volcanic passive margin
Non volcanic passive margin
Uncertain non volcanic passive margin
Uncertain volcanic passive margin
N
A passive margin is the transition between oceanic and continental crust which is not an active
plate margin. It is constructed by sedimentation above an ancient rift. Continental rifting creates
new ocean basins. Eventually the continental rift forms a mid oceanic ridge. The transition
between the continental and oceanic crust that is created by the rift is known as a passive margin.
Figure 1 Global distribution and classification of passive margins. http://commons.wikimedia.org/wiki/File:Globald.png
coastline, as for example, the Atlantic coasts of Africa and

South America. Sheared passive margins form where separation is along strikeslip faults, as in the case of the separation
of the Agulhas Bank from the Falkland plateau along a dextral
strikeslip fault (Figure 3(a)). These types of margins have
high continental shelves and deep sedimentary basins (Brown
et al., 2003). Passive margins can also be divided into Nonvolcanic and Volcanic margins (Figure 1). The former, also
referred to as sedimentary passive margins, develop where
rifting and associated crustal stretching and thinning is not
accompanied by volcanic activity; they are generally broader
than volcanic passive margins (Geoffroy, 2005). Passive margins can indeed be classified by their width: Narrow passive
margins are less than 100 km wide, compared with Wide
margins with a broad continental shelf (e.g., Brown et al.,
2003). These types are well illustrated by contrasts around the
Australian continent, where the southeast and northwest
Australian margins are typical examples of the narrow margin
and the wide margin, respectively.
The majority of classified rifted passive margins are volcanic (Figure 1), for example, the margins of southern and
eastern Africa and Greenland, and the western margin of
India. These passive margins are generally linked to large
volumes of basaltic lavas, sometimes with associated acid
extrusives, which form Large Igneous Provinces (LIPs). The

volcanism may precede continental rifting, as in the case of the
Drakensberg in southern Africa (Marsh et al., 1997), or may
be broadly coeval, as exemplified by the Deccan Traps (de Wit,
2003), and volcanic activity may persist after breakup. Seaward-dipping reflectors in some volcanic passive margins are
now considered to be due to rollover above listric normal
faults that dip toward the continent (e.g., Geoffroy, 2005;
Figure 2(b)). The lower crust of these margins typically has an
anomalously high seismic P-wave velocity (7.17.8 km s1),
which is interpreted to reflect basic igneous rocks termed
Lower Crustal Bodies (LCBs), linked to the volcanic activity,
which have been welded to the lower crust by a process
termed underplating (McKenzie, 1984). There is evidence
that volcanic rifted margins are divided along strike into
5070-km-long segments, which may have originally represented separate magma chambers (Geoffroy, 2005).
Onshore, passive margins are characterized by diverse
topographic expressions. Thus, extensive low-relief coastal
plains (Low Elevation passive margins; e.g., Gilchrist and
Summerfield, 1990) are a feature of central south Australia,
the Argentinean coast of South America, and the west coast of
central Africa (Fullard and Darby, 1974). Such topography
contrasts markedly with High Elevation passive margins
10
100
200
Prerift to synrift transition

Coast
Inner SDRS
Flood basalts
(km)
300
Shelf
400
73
500
Rise
Slope
Outer SDRS
0
Sealevel (km)
LIP
SDRS
Upper crustal
magma chambers
10
20
30
Oceanic crust
Transitional
igneous
crust
Transitional
continental
crust
HVLC
Lower crustal
Magma chambers
Lithospheric mantle
Moho
40
Vertical exaggeration 5
(a)
Extensional volcanic
passive margin
5080 km
No crustal extension coeval

with plume activity
Eventual external highs
Eventual pre-plume sedimentary basin

Pre-breakup traps
Post-breakup sediments
Onshore geology
Oceanic SDR
Sills
SDR
int
SDR
Thick oceanic crust

Continental crust
one
Moho
1230 km
ext
ty z
loci
h-ve
Hig
Moho
(b)
Figure 2 Features of volcanic rifted margins. (a) The SDRS may be due to ocean-dipping listric faults. Reproduced with permission from Menzies,
M.A., Klemperer, S.L., Ebinger, C.J., Baker, J., 2002. Characteristics of volcanic rifted margins. In: Menzies, M.A., Klemperer, S.L., Ebinger, C.J.,
Baker, J. (Eds.), Volcanic Rifted Margins. Geological Society of America (GSA), Boulder, CO, Special Paper 362, pp. 114. HVLC is the high-velocity
lower crust, LIP is Large Igneous Province. (b) Alternatively, the SDRS is due to tilting of volcanic rocks and sediments in the hanging walls of
continent-dipping listric faults. Reproduced from Geoffroy, L., 2005. Volcanic passive margins. Comptes Rendus Geoscience 337, 13951408.
Abbreviations: SDRint, internal seaward-dipping reflectors; SDRext, external seaward-dipping reflectors; SDRS, seaward-dipping reflectors series.
with dramatic, commonly continuous, seaward-facing scarps,

which are located some 60200 km inland of the coast. Examples of the latter are India (Ollier and Powar, 1985), the
east coast of Australia (Pain, 1985; Ollier, 2004), and the remarkable horseshoe-shaped escarpment that girdles virtually
all of southern Africa (Ollier and Marker, 1985) (Figure 4),
except where breached by the Orange and Limpopo rivers.
It is important to emphasize however, that the nature of
the southern African escarpment varies along its length. Thus,
the Drakensberg section is characterized by dramatic seawardfacing cliffs (Figure 4(b)), whereas elsewhere it may be represented by a marked increase in gradient separating the
relatively low-relief coastal plain from the interior plateau
(Figure 4(a)). In some sections of the escarpment, there is an
apparent accordance of summits, which have been inferred to
reflect the dissected relics of a Gondwana peneplain, extant

before rifting (King, 1976; Lister, 1987). However, this view
has been challenged by Partridge and Maud (1987) and
Fleming et al. (1999).
The morphology of escarpments can be classified into two
groups: Shoulder Type (e.g., the southern African escarpments), in which the escarpment forms a drainage divide
between a coastal plain and the continental hinterland, and
Arch Type (e.g., southeast Australia), in which the drainage
divide is inland of the escarpment (Figure 5; Ollier, 1984).
The classification of escarpments by Pazzaglia and Gardner
(2000), supplemented by Bishop (2007), distinguishes Type 1
escarpments as those etched into an upland associated with
rifting (e.g., southern Africa), Type 2, which are etched into a
postrift flexural bulge and are dominated by postrift offshore
60 W
40 W
20 W
20 E
30 S
30 S
74
Africa
tlantic
40 S
Ridge
69
AFT
AFF
69
60
60
50 S
FI
FP
FB
AFFZ
MEB
40
AS
60 W
40 W
20 W
40
50 S
20
20
0
40 S
20 E
(a)
NW
SE
Agulhas bank and Outeniqua basin
Sedimentary basin
6
8
Strongly compacted or sligthly

metamorphosed sedimentary rocks
Pre-break-up metasediments ?
12
Agulhas passage
Multiple faults
Sedimentary wedge
ry basin
10
Depth (km)
AFFZ
Sedimenta
2
4
DMR
0
2
Sediments
6
Continentoceantransition
Strongly compacted
sedimentary rocks
or metasediments
Oceanic crust
10
Upper continental crust
12
Mantle
14
14
16
Mid and lower continental crust

Interbedded volcanic
flows and sediments
18
20
18
22
24
24
Mantle
26
26
30
30
(b)
16
20
Sheared margin
22
32
240
32
280
320
360
400
440
480
520
560
600
640
Distance (km)
Figure 3 Rifted and Sheared passive margins. (a) The facing margins of South America (north of 47 1S) and Africa (north of 35 1S) are rifted. The margins to the north of the Falkland Plateau (FP) and
along the SE coast of Africa are Sheared margins. Reproduced with permission from Parsiegla, N., Gohl, K., Uenzelmann-Neben, G., 2007. Deep crustal structure of the sheared South African continental
margin: first results of the Agulhas-Karoo Geoscience Transect. South African Journal of Geology 110, 393406. FB Falkland Basin, FIFalkland Islands, MEBMaurice Ewing Bank, AFTAgulhasFalkland Transform, MT Mallory Trough segment, DRDiaz Ridge segment, ELEast London segment, DDurban segment. (b) Cross-section along the bold gray line (between O and B of the
Outeniqua Basin in Figure 3(a)) to show the crustal structure of a sheared margin.
D
EL
D
8 R
M 0
80 80 T
80
AP
Mid-A
South
America
75
KB
Bi
Plateau
KB
Zambezi
KB
20
KB
KB
KB
Orange
Limpopo
Moambique
Plain
ra
nt k e n s
ain
burg
s
Khomas
Highlands
-D
uli o u
h
t
M
Ma
30
Legend
High : 3473
500 km
Low : 0
20
30
40
(a)
(b)
Figure 4 (a) SRTM digital image for southern Africa, illustrating how the low-relief coastal plain and continental interior are separated either by a
dramatic escarpment cliff (see Figure 4(b)) or by a marked increase in slope. Prominent breaks in the escarpment (labeled) are associated with the
Limpopo, Zambezi drainage basins. There is a less pronounced break in the western escarpment, where it is incised by the Orange River (with the
major north bank Fish River tributary). (b) Prominent escarpment of the Drakensberg-Maluti mountains (also known as the Mathuli-Drakensberg
mountains). Reproduced from African Imagery.
loading (e.g., U.S. Atlantic), and Type 3, an escarpment

characterized by embayments and gorges caused by rivers that
rise at a preexisting continental divide inboard of the escarpment (e.g., parts of southeast Australia; Nott et al., 1996). This
genetic classification embodies some of the central controversies about escarpment evolution. Inland of these dramatic
escarpments, the continental hinterland commonly defines a
broad, low-relief basin (Ollier, 1984).
The tectonic geomorphology of passive margins and continental hinterlands is primarily concerned with vertical
movements of the crust and associated landscape evolution,
including epeirogenic (long wavelength) elevation changes.
Some of the factors proposed to cause vertical movements are
shown in Figure 6.
The aim of this study is to develop a tectonic and geomorphic framework to understand the processes responsible
76
for sculpting the diverse landforms of passive continental

margins and continental hinterlands. The study focuses on the
origin of the highlands that form major sea-facing escarpments on some margins and how this has impacted landscape
evolution. Many examples are taken from southern Africa,
which has archetypal status in the study of passive margin and
continental interior landscapes.
It is convenient to discuss passive margin and continental
hinterland tectonic geomorphology in two separate time
frames: in the first, the igneous and tectonic processes associated with rifting that precede continental breakup and generate
the prerift topography are discussed. The second time frame
deals with postrifting processes. The study of continental hinterlands is most naturally treated after the latter. Despite the
above observation that passive margin and continental interior
tectonics, dominated by vertical movements, can only be linked
indirectly to plate tectonics, in which horizontal displacements
are prevalent, it is argued that passive margin and continental
interior evolution have to be viewed in the context of underlying plate tectonic controls.
5.5.2
Igneous and Tectonic Processes Associated

with Rifting
Modern rifts, such as the East African Rift System (EARS) and
Red SeaGulf of Suez, are bounded by uplifted marginal
Drainage divide
Arched
Shoulder
0.53 km
100s km
Figure 5 Arch vs. Shoulder Type margins. Vertical line with double
arrows shows the drainage divide.
shoulders (Steckler and Omar, 1994), and by analogy, similar

topography is anticipated to have formed during the early
development of other rifted passive margins. A variety of
mechanisms, not mutually exclusive, have been proposed for
such elevated rift shoulders. Initial rifting may involve upwelling of hot, buoyant asthenosphere, leading to thermal
expansion and crustal updoming mainly due to dynamic effects (e.g., White and Mckenzie, 1989). This asthenosphericrelated uplift may be preceded or postdated by extension
(active vs. passive rifting: e.g., Sengor and Burke, 1978; Turcotte and Emerman, 1983; Huismans et al., 2001). If volcanism accompanied these processes, penetrative magmatism
associated with the intrusion of dikes and sills would provide
further advective heating (Gilchrist and Summerfield, 1994).
Plume acolytes and plume skeptics would join battle over the
role of putative mantle plumes in contributing to lithospheric
heating and uplift. Uplift of rift flanks can be predicted from
models that allow nonuniform extension in the crust and
mantle, whether the extension is continuous or discontinuous
(Royden and Keen, 1980; Rowley and Sahagian, 1986).
However, these effects would be transient, and are envisaged
to decay over a timespan of B60 Ma, so that they cannot
account for the existence of escarpments on rifted margins of
greater vintage.
Magmatic underplating of the lower crust by basic intrusive
igneous rocks, less dense than the underlying ultrabasic
mantle (McKenzie, 1984; White and Mckenzie, 1989), could
in principle lead to a buoyancy of a rifted margin, thus
permanently sustaining an initial thermal uplift (Figure 6).
However, a number of considerations suggest that underplating may not play a major role in maintaining uplifted
margins. Cox (1993) demonstrated that in a tholeitic LIP, the
ratio of cumulates to extruded magma would be approximately 0.5, which would imply that a lava thickness of 2000 m
(appropriate for the Drakensberg in South Africa) would
be associated with B1000 m of underplating, which would
produce a very modest uplift. Far greater amounts of underplating would be implied in the case of the eruption of the
estimated 7.5-km-thick succession of rhyolites interbedded
Flexure
Mantle
ascent
Sedimentation
Elastic
Cooling
Crustal
thinning
Erosion
Underplating
Thermal
Underplating
Isostatic
2
km
Buckling
2
4
Dynamic
Figure 6 Factors that cause vertical movements at passive margins and continental interiors. The factors can be grouped into elastic, thermal,
isostatic, buckling, and dynamic effects. Isostatic responses are obtained from sedimentation, crustal thinning, erosion, and underplating of
intrusions. Opposite elastic responses occur on either side of normal faults. Thermal effects may include heating in the early stages of rifting,
followed by cooling. Dynamic topography is created by mantle flow. The diagram includes different factors that operate at different times during
passive margin evolution, so that it does not represent a single time in passive margin evolution.
10
77
km
F1
(a)
D1
D2
(b)
Figure 7 The Geology of the Lebombo monocline. (a) Geology: pink color indicates rhyolites, blue color indicates basalts, and black lines
indicate bedding. (b) Model of the crustal structure, showing how extension occurs during intrusion of dykes in phases D1 and D2, and
monoclinal downwarping is accommodated by domino block faulting with transverse fault zones (F1). Reproduced with permission from Klausen,
M.B., 2009. The Lebombo monoclene and associated feeder dyke swarm: diagnostic of a successfully and highly volcanic margin.
with basalts (Cox, 1993) of the low-lying (450550 m high)

Lebombo mountain range on the southern African east coast
(Figure 7).
These volcanic rocks were supplied by a dense local network
of feeder dikes, which show that the Lebombo was an area of
major crustal extension at the time of eruption (du Toit, 1930).
The rhyolites have Sr isotopic signatures identical to the associated basalts, suggesting that they were either derived from the
latter by fractional crystallization or by partial melting of source

rocks similar to the basalts. Either mechanism requires that the
rhyolites represent B1116% of the original basaltic source
rocks, which thus would have to be 4575 km thick (Cox,
1993). A lower, but still substantial amount of underplating
would be implied on the basis of a thinner (3.5 km) sequence
of rhyolites inferred by Klausen (2009). Underplating of this
magnitude would be expected to produce substantial
78
The interior of southern Africa has an anomalously high elevation (generally 41000 m asl) compared with average elevations of 400500 m asl that characterize cratonic areas in
80 E
84 E
Ga
ng
24 N
Maiwa
20 N
lf
Gu
of
Narmada
Tap
ti
y
ba
m
Ca
Arabian
sea
16 N
Bay of
Bengal
ats
Prerifting Continental Topography and

Elevation
76 E
Gh
5.5.3
72 E
68 E
Western
permanent uplift of the margin, whereas the Lebombo rises

only slightly above the low veld of the surrounding low-lying
(o300 m) east coast plain of southern Africa.
The Lebombo volcanic belt has further important implications for understanding the processes associated with rifted
margins. It has a monoclinal structure, dipping eastwards
toward the coast, where it is covered beneath younger Cretaceous sedimentary rocks (Dingle and Scrutton, 1974). The
lavas are cut by dikes with compositions that can be matched
with those of the volcanic rocks, indicating that they represent
the feeder system. The highest dyke density is associated with
the monoclinal axis and the displacement resulting from these
intrusions indicates a crustal extension of at least 15% (du
Toit, 1930). The oldest dikes dip to the west (Figure 7),
showing that they have been tilted by the monoclinal warping,
whereas the youngest intrusive rocks are close to vertical.
These direct field observations led du Toit to conclude that the
maximum period of flexure must have been during the
interval of volcanic eruption of the Lebombo belt a conclusion recently endorsed by Klausen (2009). Cox (1993)
noted that had the monocline been below sea level, it would
certainly have been identified as a sequence of seawarddipping reflectors by marine geophysicists.
The Lebombo belt shows a sharp change in strike at the
Limpopo river from almost north-south to trending north
northeast. This point of inflection marks the eastern end
of a major dyke swarm that trends west northwest through
southern Zimbabwe and across Botswana for a distance of
approximately 1200 km. Reeves (1978) suggested that the
dykes exploited the failed arm of a triple junction, with the
trends of the Lebombo volcanic rocks marking the line of
failure associated with opening of the Indian Ocean.
The Deccan Traps flood basalts in India are associated with
the rifted passive margin that forms the west coast of the
country. They occur on the northern end of the major
seaward-facing escarpment known as the Western Ghats
(Figure 8). In the extreme north of the outcrop, the lavas show
a seaward-dipping monoclinal flexure, which dies out to the
south (Ollier and Powar, 1985). The northern margin of the
Traps is associated with a rift, which intersects the west coast at
the Gulf of Cambay. The Tapti and Narmada Rivers exploit this
rift to flow to the west coast, and thus in the opposite direction
to the major drainages further to the south, which rise off the
western Ghats and flow eastwards to the Bay of Bengal. Cox
(1989) interpreted the rift, which was exploited by the two
major west-flowing rivers, to represent the failed arm of a
triple junction. This points to a common tectonic setting at a
triple junction for the monoclinal flexures of the Lebombo
and the northern extremity of the Western Ghats (Figure 8).
Both are characterized by a low-relief interior that contrasts
with the well-defined escarpment elsewhere on the southern
African and West Indian passive margins.
12 N
200
400
Kilometres
Figure 8 Drainage patterns in India and the Deccan Traps. Most of

peninsular India is characterized by dome flank drainage, but the
Narmada and Tapti drainages, flowing in the opposite direction, are
examples of rift-related drainage. The extent of the inferred plume is
shown by the stipple and the extent of the Deccan traps is shown in
green. Western Ghats is the line of the escarpment. Diagram partly
based on Cox (1989).
other continents (Lithgow-Bertelloni and Silver, 1998; Gurnis

et al., 2000). A critical issue in accounting for such differences
is the elevation and nature of the topography of the prerift
land surface. At some rifted margins, there are very tight
constraints, as in the opening of the Red SeaGulf of Suez,
where normal faults related to rifting, dated at 2119 Ma, cut
across a sedimentary sequence capped by shallow water Eocene marine carbonates (Steckler and Omar, 1994). However,
by 17 Ma coarse conglomerates deposited into the rifted
marine basin provide evidence for the existence of high marginal relief, reflecting rift flank uplift, which at the present time
attains altitudes in excess of 2000 m along the Red SeaGulf of
the Suez margin (Steckler and Omar, 1994). However, Japson
et al. (2006, 2009) used landscape analysis and interpretation
of apatite fission track (AFT) data to show that following
rifting at B65 Ma, the Greenland west coast was low lying,
and experienced Palaeogene and Miocene marine transgressions. The modern relatively high elevation (up to 2000 m asl)
and rugged topography of this margin was inferred to have
developed relatively recently following two phases of Neogene
uplift (1110 and 72 Ma). Japson et al. (2006, 2009) suggested that several other elevated plateaux in continental
hinterlands may also postdate breakup, and may be relatively

recent.
In southern Africa, the Lower Triassic Beaufort sediments
are interpreted to have been deposited in a terrestrial low-relief
environment of marshes crossed by intermittent streams
(Truswell, 1970). Triassic uplift of the Cape Fold Belt resulted
in highlands in the south of the region, but absolute elevations are not at all constrained. Little geological evidence
exists for the configuration of the drainage system in the Jurassic, which was marked by the eruption of the Karoo flood
basalts in the early part of this period (B182 Ma; Marsh et al.,
1997). However, by the early Cretaceous, following opening
of both the Indian and the Atlantic Oceans, the configuration
of continental drainage of south-central Africa changed drastically from the Triassic river system that preceded the Karoo
volcanism (Moore and Blenkinsop, 2002).
Cox (1989) ascribed the development of the post-Gondwana
drainage network in southern Africa to updoming over the
Karoo and Parana plumes, centered on the east and west coasts,
respectively. This concept was extended by Moore and Blenkinsop (2002), who noted that initial uplift, related to the Karoo
volcanic event in the east, would have produced a westwarddraining network, but that the subsequent Parana-Etendeka
volcanism, centered on the Atlantic margin of southern
Africa, would have reversed the postrift paleo-slopes, initiating
the major eastward-flowing rivers that dominate the modern
drainage system. It should be stressed, however, that these
drainage reconstructions are independent of whether Gondwana
disruption was initiated by the putative Karoo and Parana
mantle plumes. Advective heating by penetrative volcanism
related to the extensive dike swarms associated with the Karoo
and Parana-Etendeka volcanism, possibly coupled with rift-flank
uplift, could also explain the doming invoked by Cox (1989).
The magnitude of the prerift topography in southern Africa
remains a highly contentious issue. Partridge and Maud
(1987) envisaged a relatively high (though not specified) altitude for southern Africa before rifting. This was supported by
Gilchrist and Summerfield (1991), who calculated the prerifting continental elevation to be B1200 m asl by reloading
the mean continental denudation of B1800 m estimated
by Rust and Summerfield (1990) for the Atlantic-draining
catchment and making an adjustment for Airy isostasy. In
contrast, Burke and Gunnell (2008) argued that the continent
was low lying, and characterized by low relief immediately
before continental breakup. On the basis of topographic
modeling, Doucoure and de Wit (2003) concluded that the
modern bimodal topography of Africa had been established
by the Cretaceous, with elevated areas centered on east and
southern Africa, a conclusion that was also reached by interpretation of AFT data (Kounov et al., 2008, 2009; Tinker et al.,
2008). However, Doucoure and de Wit (2003) envisaged that
both areas were at considerably lower average elevations
(200400 m) than the modern topography (41000 m average
elevations asl). The difference between their estimate and that
of Gilchrist and Summerfield (1991) serves to underline the
uncertainties inherent in the assumptions underlying geophysical modeling.
The arguments for a low elevation of Africa immediately
before, and for an extensive period following, continental
breakup must also be seen against the firm geological evidence
79
that in the Red SeaGulf of the Suez passive margin, rift flank
uplift, erosion, and consequent isostatic rebound had produced a strong inland relief in less than 3 Ma following rifting,
and elevations in excess of 2000 m asl at the present time,
some 21 Ma after rifting (Steckler and Omar, 1994). The
topography around the time of rifting of regions with anomalously high present-day elevations, as in the case of east and
southern Africa, remains strongly debated, and may vary
considerably in different geological/tectonic settings (e.g.,
Greenland vs. Red SeaGulf of Suez).
5.5.4
Postrifting Evolution of Marginal Escarpments
One of the most controversial issues in postrifting landscape

evolution has been the evolution of the dramatic seawardfacing escarpments that occur on some passive margins. This
debate must naturally be extended to explain why they are
absent on other margins and why some pairs of margins
have apparently complementary geometries, such as the welldeveloped elevated escarpment of Namibia in southern Africa
and the extensive low-relief plain on the margin of the
Argentinean seaboard.
5.5.4.1
Kings Scarp Retreat Model Backwearing
The elevated mountainland formed by the Drakensberg Karoo

basalts (Figure 4), which represents the highest point of the
almost continuous coast-parallel escarpment that bounds
southern Africa, has long held central stage in the debate regarding the evolution of passive margins. King (1955, 1963)
envisaged that the coastal plain seaward of the Drakensberg
mountainland had been planed by a process of parallel scarp
retreat (backwearing), analogous to that originally proposed
by Penck (1924) (Figures 9(b) and 9(c)). King considered
that the original scarp had formed at the coast at the time of
disruption of Gondwana and had advanced inland over a
period of 120140 Ma. The coastal plain would accordingly be
regarded as a pediplain.
5.5.4.2
Pinned Drainage Divide Downwearing
Kooi and Beaumont (1994, 1996) used numerical landscape

modeling to show that if an escarpment formed at the coast at
the time of continental breakup, it would very rapidly degrade
if there was an antecedent drainage system with an inland
drainage divide (Figure 9(d)). They demonstrated that a new
escarpment could then rapidly be initiated at the location of
the original drainage divide, followed by very slow rates of
inland retreat. This model for location of the escarpment at a
pinned drainage divide is strongly dependent on a number of
poorly known variables, and is inapplicable to the Eastern
Australian coastal escarpment and other arched-type escarpments, where the modern divide is located well inland of the
scarp. Nevertheless, it provides a potential explanation for
the broad coincidence of the escarpment and modern coastal
drainage divide in southern Africa and the Western Ghats
of India.
Fleming et al. (1999) first proposed that the pinned drainage divide concept was more appropriate to the evolution of
80
(a) Flexural response to different Te
Te = 7.5 km
1 km
15 km
22.5 km
(b) Backwearing with flexural isostacy
(c) Backwearing with flexural isostacy
(d) Downwearing with pinned drainage divide
(e) Downwarping
200 km
Figure 9 Escarpment evolution at passive margins. (a) Modeled riftflank uplift associated with flexural isostasy for various values of
elastic thickness Te. The typical profile across the western margin of
southern Africa is shown as a dashed line (Gilchrist and
Summerfield, 1990). (b) Models of scarp retreat and rift-flank uplift
associated with flexural isostasy (Gilchrist and Summerfield, 1990).
Colors in all diagrams represent successive stages. (c) Models of
scarp retreat with flexural isostasy (Van der Beek et al., 2002). (d)
Plateau degradation model, pinned drainage divide (Van der Beek
et al., 2002). (e) Model of downwarping of the Continental margin
(Ollier and Pain, 1997). All diagrams redrawn from original sources
and rescaled to approximately the same scale.
the Drakensberg escarpment than Kings scarp retreat model

on the basis of cosmogenic isotope studies, also carried out by
Kounov et al. (2007) with similar implications. These indicated very slow modern retreat rates of this major scarp, incompatible with inland migration from an original coastal
position at a constant rate since continental breakup. This
view was subsequently supported by Brown et al. (2002) and
Van der Beek et al. (2002) on the basis of AFT studies and
numerical landscape modeling, which also assumed a constant rate of scarp retreat subsequent to breakup.
The models for downwearing of the Drakensberg escarpment to a pinned drainage divide were criticized by Moore
and Blenkinsop (2006). They noted that a major shortcoming
was the dependence of the models on variables that were
difficult to constrain and the fact that the models failed to
account for the observation that the Drakensberg is bounded
by major scarps on three sides. Further, the modern scarp is
capped by very thick, resistant basalt flows, and the presentday basalt outcrop is surrounded by a dense boxwork of dykes,
4060 km wide, that was the source of the lavas, suggesting a
major local focus of volcanic activity. Moore and Blenkinsop
(2006) argued that massive lava flows would be more likely
in close proximity to the boxwork of feeder dykes and that
erosion rates would be dramatically slowed once the lavas
were exposed. Erosion rates would further be influenced by
climatic factors and the late Neogene evolution of the C4
grasses. Modern erosion rates, determined on the basis of
cosmogenic nuclides, may have no bearing on historic rates,
and cannot therefore be used as a basis to reject Kings scarp
retreat model.
The importance of lithology in controlling the inland migration of an initial escarpment formed in response to rifting
is further illustrated in the classic investigation of the Red
SeaGulf of Suez passive margin of Steckler and Omar (1994).
In this area, rifting has incised a sequence of massive, resistant
Eocene limestones that cap the friable and readily eroded
Nubian sandstone. The entire sequence thins from the north
to the south, and as a result, erosion has resulted in the initial
removal of the resistant Eocene limestone capping to expose
the Nubian Sandstone in the south. Erosion of the latter
formation has caused rapid undercutting of the overlying
limestone cap and the greatest inland advance of the scarp in
the south.
5.5.4.3
Downwarping of the Continental Margin
Ollier drew attention to the presence of a major continental

divide inland of the escarpment that overlooks the coastal
plain of much of eastern Australia (Ollier, 1982; Ollier and
Pain, 1997; Ollier and Stevens, 1989). These authors suggested
that marginal downwarping of eastern Australia at the time of
continental breakup had occurred along an axis corresponding to the inland divide. Thereafter, headward river erosion into the downwarped shoulder of the continent resulted
in the development of the coastal escarpment (Figure 9(e)).
Although Ollier (1982) envisaged that scarp retreat continues
at the present time, a Pliocene basalt that has flowed down
one of the escarpment valleys shows relatively little evidence
of erosion. The implied slow recent rates of erosion of the
eastern Australian escarpment might reflect the drastic vegetation change linked to the late Neogene evolution of widespread C4 grasses (Moore and Blenkinsop, 2006). This is
broadly consistent with AFT data that suggest rapid erosion
linked to Cretaceous breakup (e.g., OSullivan et al., 1995),
followed by slow erosion through the Cenozoic.
Olliers model has been criticized by Nott and Horton
(2000), who presented field evidence that along the northern
sector of the eastern Australian escarpment, the inland drainage
divide was already extant in the early Jurassic and thus predates
the Late-Jurassic to early Cretaceous formation of the Queensland and Townsville Basins near the modern coastline (Struckmeyer and Symonds, 1997). Nevertheless, the Lebombo
monocline in the east of southern Africa formed synchronously
with the associated lavas, erupted at B182 Ma (Du Toit, 1930;
Klausen, 2009), and thus significantly before 155 Ma, when
opening of the Indian Ocean is generally inferred to have
commenced (Reeves and de Wit, 2000; McMillan, 2003). If this
sequence of events is correct, the existence of a drainage divide in
eastern Australia before continental breakup in this area would
not be incompatible with a marginal downwarp that has remained stationary and existed for longer than generally assumed
(cf. Young, 1989). It should be noted, however, that Klausen
(2009) has proposed that continental rifting in the east of
southern Africa to initiate the Indian Ocean commenced immediately after eruption of the Lebombo volcanics. A monoclinal geometry is also suggested by Ghebreab and Talbot (2000)
for the Red Sea margin in Eritrea. Here, extension began at
40 Ma, followed by a second event at 23 Ma, which formed the
monoclinal flexure, uplifting 30 Ma flood basalts in its hanging
wall, at about the time that the Arabian and Nubian shields
separated. Flood basalts, monoclinal flexure, and ocean opening
therefore appear to have occurred simultaneously here, similar
to the scenario proposed for southern Africa by Klausen (2009).
5.5.4.4
Isostasy and Flexure
King (1955) proposed that erosional isostatic rebound of

the coastal plain would occur once a critical threshold of
B480 km of inland erosion had occurred. Thereafter, uplift
would take place, initiating a new cycle of scarp retreat and
pediplanation, giving rise to the development of cyclic land
surfaces. Kings (1955) envisaged multicycle erosion surfaces
have been strongly criticized (e.g., Summerfield, 1988) because geophysical models show that the isostatic response to
erosional unloading would be immediate and continuous,
rather than delayed and episodic. Furthermore, King (1955)
did not account for the flexural response of the lithosphere.
Nevertheless, as elaborated in a later section, objections to the
mechanism proposed by King (1955) cannot be used to
counter the field evidence for polycyclic erosion in southern
Africa. The concept of flexural isostacy was applied to escarpments by Gilchrist and Summerfield (1990, 1991, 1994),
who pointed out that one of the consequences of rifting
would be higher local relief and a marked lowering of the
erosional base level on the coastal margin of a rift flank,
leading to higher denudation rates than in the relatively lowrelief interior. Numerical modeling showed that if the lithosphere was treated as an elastic plate, this differential
81
denudation would result in landward erosion, coupled with

progressive elevation of the escarpment separating the coastal
plain from the interior plateau (Figures 9(a) and (b)).
One of the main uncertainties in applying the flexural isostacy model to passive margins is the elasticity of the lithosphere, encapsulated by the flexural rigidity Te. Estimates of Te
for continental margins are quite variable, even along the same
margin. For example, Tassara et al. (2007) showed estimates
of o10 to 30 km for the passive margins of South America.
Pazzaglia and Gardner (1994) suggested an average value of
40 km for the U.S. east coast, within the range of 2060 km for
previous estimates, and much larger than the 15 km value for
the SW African margin (Gilchrist and Summerfield, 1991). Although Te estimates are subject to some uncertainty, there appears to be a real variation between different passive margins
and variation along the length of the same margin.
The flexural isostacy model was applied to the escarpment
along the west coast of South Africa (Gilchrist and Summerfield, 1991), where up to 5 km of postrift denudation
has occurred on the coastal plain (Brown et al., 2002; Kounov
et al., 2009). Gilchrist and Summerfield (1991) concluded
that the best correspondence between modeled topography
and that observed for the coastal plain, marginal escarpment,
and inland plateau was obtained when Te 15.0 km
(Figure 9(a)). A better fit was obtained for the escarpment
and interior plateau for a lower elastic thickness (7.5 km),
but this also predicted a major crustal warp on the coastal
plain, which contrasts with the very subdued upwards
convex profile of the west coast of South Africa. In contrast, it
is noted that scarp retreat initiated by the Red SeaGulf of Suez
rift has been associated with deep erosion (25 km) and
major uplift on the coastal plain (Steckler and Omar, 1994),
similar to that predicted by the Gilchrist and Summerfield
(1991) for their model assuming a low elastic thickness for the
continental margin. The implication is that the observed differences in Te for different continental margins could to lead
to differences in the isostatic flexural response to differential
denudation associated with the development of a marginal
escarpment.
An important prediction of the model put forward by
Gilchrist and Summerfield (1991) is that inland sediment dips
would be expected in the vicinity of the escarpment. This is
indeed the case for the Karoo sequence on the coastal plain
below the Drakensberg escarpment in the east of South Africa,
where a minimum of 4.5 km of denudation (Brown et al.,
2002) is estimated to have occurred in the B150 Ma since
initiating of rifting (Figure 10; de Wit, 2003). Differential
denudation and flexural isostasy thus appear to have played a
major role in the evolution of this portion of the southern
African horseshoe escarpment.
Gunnell and Fleitout (2000) demonstrated that a lithospheric isostatic flexural response to the differential rates of
erosion of the inland plateau and coastal plain could in
principle account for the major escarpment formed by the
Western Ghats (Figure 8). The best fit between the actual and
the modeled topography was obtained when the lithosphere
was treated as a broken plate, with a high flexural rigidity
(Te 70 km) inland of the rifted margin and a low rigidity
(Te 1.3 km) on the seaward side of the break. However,
Gunnell and Fleitout (2000) stressed that this was a
82
8
Constrained
Kilometres
4
Total (1300 Ma)
2
0
(a)
8
Constrained
Kilometres
Total
Extrapolation through
independent estimates
13065 Ma
65
0M
(b)
0
400
300
200
100
Distance (km)
(b)
Figure 10 Interpolation between the denudation estimates at the two borehole sites constrained across the Lesotho Highlands by independent
estimates derived from zeolite distribution and in situ-produced cosmogenic 36Cl concentrations. (a) Total denudation, shown relative to presentday surface. (b) Denudation estimates for Late Cretaceous and Tertiary periods. Reproduced with permission from Brown, R.W., Summerfield,
M.A., Gleadow, A.J.W., 2002. Denudational history along a transect across the Drakensberg Escarpment of southern Africa derived from apatite
fission track thermochronology. Journal of Geophysical Research 107, 2350, doi:10.1029/2001JB000745.
50
160
320
335
270
225
520
Erosion depth (m)
1750 1400 1150
2000
Butte
Altitude (m)
1000
Sea level
0
1000
2000
NNW
3000
0
MF
50
100
SSE
150
Distance (km)
Figure 11 Strike-perpendicular cross section of Jabal Samhan, Arabian peninsula. MF master fault. Marginal downwarping is attributed to
sediment loading. Reproduced with permission from Gunnell, Y., Carter, A. Petit, C., Fournier, M., 2007. Post-rift seaward downwarping at
passive margins: new insights from southern Oman using stratigraphy to constrain apatite fission-track and (U-Th)/He dating. Geology 35,
647650.
nonunique solution in view of the poorly known variables

involved in their models.
A study by Gunnell et al. (2007) has potentially important
implications for the roles of both the flexural isostasy and the
marginal downwarping of the continent in controlling escarpment evolution. These authors noted that the sediments
forming the escarpment on the Oman coast (Arabian Peninsula) dip inland, as predicted by the modeled flexural response to differential erosion of the interior plateau and
coastal plain. However, the unconformity at the base of the
escarpment sedimentary sequence is preserved at a lower level

closer to the coast, and the elevation differential cannot be
explained by faulting. It was therefore inferred that marginal
downwarping of the continent had occurred along an axis of
flexure on the coastal side of the escarpment (Figure 11) and
that 1.75 km of denudation had occurred on the coastal plain.
On the basis of numerical modeling and AFT constraints, the
seaward flexure was inferred to be the result of sediment
loading on the continental shelf. This contrasts with the
model for early continental flexure linked directly to
83
continental rifting proposed by Ollier (2004), which is supported by the synchronicity of the Lebombo Karoo age volcanics and coastwards monoclinal flexure of the eastern
southern African coastal plain. An implication of the modeling carried out by Gunnell et al. (2007) is that coastal flexures
may have more than one origin, and by extension, different
temporal links to continental rifting.
The passive margins discussed above have all experienced
deep (several km) erosion extending well inland of the present
escarpment (Brown et al., 1999). A common feature of these
margins and their hinterland was an originally thick cover of
readily eroded sediments. In contrast, cover sediments were
either thin or absent over the eastern third of Madagascar
(Emmel et al., 2004) before rifting in the late Cretaceous
(De Wit, 2003). AFT studies across the eastern Madagascar
(Emmel et al., 2004; Seaward et al., 2004) indicate that there
has been minimal erosion inland of the escarpment since the
time of rifting. This suggests that the limited depth of postrifting erosion along this margin reflects the absence of a thick
readily eroded sedimentary cover overlying the more resistant
granitic basement.
The studies discussed above indicate that flexural isostasy
has played a major role in the formation and growth of escarpments on passive margins. The possibility of variable
flexural properties of passive margins along strike suggests the
potential for along-strike variations in modes of landscape
development and rates of erosion. Incision rates may vary by
an order of magnitude, and early landscapes may be very
different from one point to another, as documented for the
southeastern Australian margin by Van der Beek et al. (2001).
concepts. The spatial pattern of low-temperature geochronological results is sensitive to a range of variables, including the
geotherm and erosion rates, in addition to the mode of passive
margin evolution, which requires evaluation by numerical
modeling. The modeling suggests that the low-temperature
geochronology of transects from the coast to the escarpment
cannot generally discriminate between backwearing and
downwearing (Figure 12): this has been confirmed for (U-Th)/
He data from the Southeast Australian escarpment (Persano
et al., 2002). To constrain the mode of escarpment development, transects are needed parallel as well as perpendicular to
the escarpment (Braun and Van der Beek, 2004).
5.5.4.5
The controversy surrounding the origin of escarpments inland

of many passive margins is mirrored by the question of why
they are absent along others. This issue is well illustrated by the
margin of Namibia, with its well-developed escarpment and
high interior, and the conjugate low-lying Argentinean coast
line, where the Saldo rift basin intersects the coast at right angles
along the low-relief Argentinean portion of the South American
margin. The Saldo basin was associated with strikeslip faulting
that accommodated crustal reorganization during the early
phase of separation of Africa and South America (Nurnberg and
Muller, 1991) and represents a failed rift system. The major
Parana-Plate river, with headwaters inland of the escarpment
located to the north, enters the Atlantic via the Saldo Rift.
In southern Africa, there is a prominent break in the escarpment associated with the Lebombo triple junction
(Figure 4(a)), which is exploited by the Limpopo river, with
headwaters well inland of the coastal escarpment. In India, the
major Tapti and Narmada Rivers empty into the Arabian Sea at
the major inflection in the coastline forming the Gulf of
Cambay, at the northern end of the escarpment formed by the
western Ghats (Figure 8). Cox (1989) inferred that this was a
triple junction and that these two rivers have exploited a failed
rift arm.
There are thus several examples of low-relief passive margins lacking an inland escarpment that are associated with
failed rift arms linked to the disruption of Gondwana. These
are invariably exploited by major rivers, with headwaters well
inland. The geomorphic expression of some passive margins is
Perspectives from Integrating Low-Temperature

Geochronology and Numerical Modeling
Ollier and Pain (1997) and Gallagher et al. (1998) suggested

that AFT analyses offer a useful way to distinguish downwarping from flexural isostasy models for passive margin
evolution. Because the base of escarpments in a downwarped
margin will be exhuming deeper stratigraphic levels, AFT ages
should decrease from the coast toward the escarpment base.
However, the rapid and deep exhumation of scarps near the
coast in the flexural models should lead to AFT ages that
increase inland toward the escarpment (Gallagher et al.,
1998). A compilation of AFT data from Southeast Australia,
Southeast Brazil, Southwest Africa, Yemen, and West India
shows a general increase in age inland, and none of the data
sets show the predicted pattern for downwarping. However,
this is not the pattern observed in Scandinavia, where the
youngest AFT ages occur at the heads of the fiords (Rohrman
et al., 1995). Ollier and Pain (1997) interpret these data to
confirm the downwarping model.
These conclusions need to be reevaluated in the light of
recent numerical modeling. Braun and Van der Beek (2004)
stated that A retreating escarpment would progressively reset
the thermochronological ages as it migrates from the coastline
to its present-day position and therefore lead to a clear decrease in thermochronological ages from the coastline toward
the escarpment. The modeling of Braun and van der Beek
(2004) revealed the reason for this contradiction with previous
5.5.4.6
Sinuosity of Escarpments
The sinuosity of escarpments appears to evolve with age and to

depend on the type of margin. Matmon et al. (2002) showed
that escarpments at passive margins generally had higher
sinuosity than those of rifts, implying an evolution toward
more complex margins through time. Arch-Type margins have
higher sinuosity than Shoulder Type. Both these aspects of escarpment geomorphology are consistent with higher measured
rates of retreat at knick points compared with interfluves, which
would lead to greater sinuosity over time. The greater frequency
and size of drainages that cut Arch Type escarpments are also
compatible with this model for escarpment evolution.
5.5.4.7
Low-Relief Passive Margins without a Marginal

Escarpment
84
160
140
(U-Th)/He age (Myr)
120
100
80
60
40
20
Escarpment retreat scenario

Plateau degradation scenario
0
0
20
40
Distance from coast (km)
60
80
Figure 12 Distribution of apatite (UTh)/He ages as a function of distance from the coastline as computed from Pecube temperature predictions
under the two end-member scenarios. Reproduced with permission from Braun J, van der Beek P., 2004. Evolution of passive margin
escarpments: what can we learn from low-temperature thermochronology? Journal of Geophysical Research 109, F04009. doi: 10.1029/
2004JF000147
Lower plate passive margin

Detachment fault
1
Sea floor
spreading
Upper plate passive margin

Passive margin
Anorogenic
mountains
granite
Upper crust
Upper crust
Lower crust
Lower crust
Moho
Moho
Stretched
lithosphere
Basaltic
underplating
Asthenosphere
Figure 13 Conjugate margins differentiated as the Upper plate (hanging wall of detachment) with high elevations and Lower plate (footwall)
with low elevations. Redrawn from Lister, G.S., Etheridge M.A,. Symonds P.A., 1991. Detachment models for the formation of passive continental
margins. Tectonics 10, 10381064.
closely linked to structural controls, in particular the location

of rifts associated with triple junctions.
An entirely different perspective on low-relief passive margins is offered by the asymmetric rifting model of Lister et al.
(1986, 1991; Figure 13). Based on the premise that asymmetry
created by detachment faults is an inherent feature of crustal
extension, Lister et al. (1986) proposed that passive margins
can be considered in terms of an upper plate (the hanging wall
to the detachment) and a lower plate (the footwall). The lower
plate margin is envisaged to consist of rotational normal faults

and tilt blocks, compared with the upper plate with higher
angle normal faults. The upper plate is subject to uplift due to
exposure to rising asthenosphere in the footwall of the detachment fault and thus becomes a High-Elevation passive
margin, whereas in the lower plate, substitution of mantle for
lower crustal material results in subsidence to form a LowElevation passive margin. An interesting result from numerical
modeling of uplift in these models is that the only mechanism
85
to ensure long-lived elevation of passive margins is underplating (Lister et al., 1991).

Typically, then, passive margins should consist of opposed,
conjugate High- and Low-Elevation margins. Examples cited
by Lister et al. (1991) include the Antarctic and South Australia, the Blake Plateau and the Guinea basin on the North
Atlantic, and Southwest Africa and the Argentinean margin in
the South Atlantic. Along-strike changes in polarity of the
asymmetric extension may occur at transform faults, in a
manner that is very similar to the alternating half-graben
geometry of continental rifts (e.g., Rosendahl, 1987). Although the asymmetric rifting model is capable of explaining
the diversity of passive margin expressions, one problem is the
difficulty of finding strong observational evidence for its
diagnostic feature: large displacement detachment zones.
5.5.5
Evolution of Continental Hinterlands
Figure 14 The African erosion surface (arrow) at Grahamstown,

Eastern Cape, South Africa, with silcrete capping kaolinite.
Southern Africa will be used as the backdrop to explore the

nature of the controversies about landscape evolution inland of
the escarpment following continental breakup. These center
around two broad, but interrelated issues. The first is the concept of cyclical erosion episodes, which was championed for
southern Africa by King (1949, 1955, 1963), and the second is
the anomalous average elevation (41000 m asl) of the interior
plateau of the region. The latter problem raises the issue of
dynamic topography, which is currently an area of great interest
in tectonic geomorphology (e.g., Moucha et al., 2008).
5.5.5.1
Cyclic Erosion
King (1955, 1963) inferred after careful observation that landscape evolution involved successive phases of planation by
scarp retreat, to produce a low-relief pediplain. He postulated
that continental uplift was responsible for initiating a new cycle
of erosion, which would incise the surface formed in the preceding cycle, and that six surfaces of differing ages could be
identified in southern Africa. The oldest of these, termed the
Gondwana Surface, was considered to have been planed before
the disruption of Gondwana. Remnants of this surface were
recognized in the highlands of the Drakensberg, and subsequently also the eastern highlands of Zimbabwe (Lister,
1987). The most widespread surface, termed the African Surface,
formed following continental breakup, and was located over
wide areas of the continent. Much of the early criticism of Kings
(1955) model centered on the problem of identifying planation
surfaces and particularly their correlation over extensive distances. As discussed earlier, a more recent criticism has been
leveled at the mechanism of punctuated isostasy that King
(1955) proposed to account for episodic continental uplift.
In a major review, and arguably one of the most extensively
cited references to landscape evolution in Africa, Partridge and
Maud (1987) expressed strong support for cyclical erosion in
southern Africa, but proposed a simplified model, involving
three major cycles. They argued that Gondwana age surfaces
are nowhere preserved in African, except where they were exhumed from beneath younger cover rocks. This view was
subsequently endorsed by Fleming et al. (1999) on the basis
Figure 15 Granite inselbergs characteristic of the Post-African I

planation surface, Matopos Hills from Rhodes Grave, Zimbabwe.
of erosion rates in the Drakensberg highlands, determined

using cosmogenic nuclides, which implied substantial surface
lowering since continental breakup.
Partridge and Maud (1987) proposed that there was a
lengthy period of erosion following the disruption of Gondwana, with the early stages during the Cretaceous involving
deep weathering under tropical conditions. This resulted in a
senile surface with a deeply kaolinized mantle, capped by silcrete and ferricrete duricrusts, which was equated to Kings
African Surface (Figure 14). It was regarded as multicyclic in
origin, but Partridge and Maud (1987) considered that there
was insufficient evidence to confidently identify subcycles. The
characteristic African Surface weathering profile was taken as a
reference point for identifying younger surfaces. Partridge and
Maud (1987) inferred that the African cycle of erosion was
terminated by Miocene uplift, initiating a relatively short-lived,
and therefore less extensive, Post-African I planation surface.
Development of this surface primarily involved stripping of
the weathered African Surface carapace, and could therefore be
considered an etch surface (Moore et al., 2009). Granite inselbergs (Figure 15), which are commonly characteristic of this
surface, have been interpreted to represent more resistant areas
86
associated with lower joint densities, resulting in a shallower

base to the African weathering profile (Twidale, 2002). The
Post-African I planation surface was in turn envisaged to have
been interrupted by major Plio-Pleistocene uplift of up to
900 m, focused on a line termed the Ciskei-Swaziland Axis,
located on the eastern seaboard of southern Africa.
The broad synthesis proposed by Partridge and Maud
(1987) has resulted in an interesting dichotomy of geological
viewpoints. It is widely accepted in southern Africa that the
diagnostic African Surface weathering mantle provides a reliable
datum for long-range correlations and thus for recognizing
younger cycles of erosion. However, as with criticisms leveled at
Kings (1955) original proposals for cyclic erosion surfaces, a
major objection continues to be the absence of a mechanism to
account for episodic uplift of the continent. Thus, Gilchrist and
Summerfield (1991) suggested that uplift of a marginal scarp
should be continuous rather than episodic. However, it should
be emphasized once again that the criticism of the mechanism
proposed by King (1955) does not negate the clear field evidence for multiple erosion cycles. A potential tectonic trigger to
account for their initiation is presented in a later section.
Kings (1955) model has been widely applied. For example
in the Northern Territory, Australia, four distinct erosion surfaces have been identified and associated with cyclic erosion
(e.g., Hays, 1968). However, this work has been criticized on
the basis that the surfaces are structurally controlled, cannot
necessarily be extrapolated over large distances, and have inappropriate ages, and therefore, the surfaces do not constitute
evidence for cyclic erosion (Nott, 1994, 1995).
5.5.5.2
Dynamic Topography
Dynamic topography is caused by flow in the mantle that

generates vertical stresses to support deformation of the
Earths surface (Hager, 1985; Lithgow-Bertelloni and Silver,
1998). There is currently debate about the possible magnitude
of dynamic topographic effects, which have been predicted to
provide over a kilometer of elevation (e.g., Lithgow-Bertelloni
and Gurnis, 1997; Daradich et al., 2003; Conrad et al., 2004)
or to have effects on the order of hundreds of meters (e.g.,
Wheeler and White, 2000). In either case, dynamic topography could therefore play an important role in passive
margin and continental hinterland evolution.
Evidence for the role of relatively subdued dynamic topography has been presented for the continental interior of Australia by Sandiford (2007) and Sandiford and Quigley (2009).
These studies point out that Australia is a particularly favorable
location to examine dynamic topographic effects because of the
exceptionally fast movement of the Australian plate over
mantle anomalies, and because the aridity and low relief of the
continent preserve some epeirogenic features very well.
The Australian continent is moving north toward an area of
low dynamic topography at the convergent boundary between
the AustralianIndian plate and the Eurasian plate. An area of
low dynamic topography also occurs to the south of the continent (the AustraliaAntarctic Discordance). The combination
of motion toward low dynamic topography to the north and
away from low dynamic topography to the south may have
tilted the continent at a rate of 20 m Ma1 over the last 15 Ma,
and is responsible for the low elevation of the northern edge of

the Australian plate, reflected in its broad continental shelf and
the dearth of Cenozoic marine sediments in the northern part
of the continent (Figure 16; Sandiford and Quigley, 2009). By
contrast, Cenozoic sediments are found at elevations of 300 m
up to 400 km inland from the southern coast.
The tilting of the Australian continent appears to be due
mostly to dynamic topographic effects, but some 10% may
also be attributed to movement of the continent across
gradients in the geoid (Sandiford, 2007). There is a major
geoid high over the convergent plate boundaries of the
IndonesiaNew Guinea region, which causes geoid height
variations of 90 m over the continent. The anomaly is attributed to the deep mantle, caused by the subduction history of
the AustralianIndian plate (Sandiford, 2007).
A variety of proposals to explain interior elevation of
southern Africa by more dramatic dynamic topography related to plumes have been put forward. Burke (1996) and
Burke and Gunnell (2008) argued that lithospheric heating
and thermal uplift will be at a maximum when plate movement is slow relative to an underlying mantle plume. They
suggested that the northeast extreme of the African continent
came to rest over the Afar Plume at approximately 30 Ma. It
was envisaged that the resultant lithospheric heating triggered
the Afar volcanism, coupled with uplift that was responsible
for the elevated topography of south and eastern Africa.
This model poses more problems than those it attempts to
address. Firstly, plume skeptics would question whether the
uplift associated with the Afar volcanism is a reflection of a
deep mantle plume or rather the result of advective heating
associated with penetrative magmatic processes, including the
intrusion of dykes and sills linked to the Afar volcanism. It is
further difficult to envisage the putative Afar plume being responsible for a broad region of high ground in southern Africa, at the southern extreme of the continent, when it is
surrounded by much more proximal lower elevations immediately to the east and west in north Africa, although the
channeled plume explanation was offered for this phenomenon by Ebinger and Sleep (1998). Moreover, there is no
evidence for a major volcanic episode in southern Africa at
30 Ma, and in this region, post-Gondwana alkaline volcanic
pipe clusters show a systematic age pattern, younging from
the hinterland toward the coastal margins (Moore et al.,
2008). The youngest volcanic events (Early Palaeocene and
EoceneOligocene) would not therefore relate to the high
interior plateau.
Lithgow-Bertelloni and Silver (1998) and Gurnis et al.
(2000) have suggested that the anomalous elevation of
southern Africa reflects dynamic uplift related to a putative
superplume. However, seismic tomography, utilizing an array
of receiver stations, extending from the southern tip of the
subcontinent B2000 km to the northwest, failed to indicate a
major thermal anomaly that could be ascribed to such a
plume (Fouch et al., 2004). Further, Moore et al. (2009) point
out that the geophysical plume models proposed imply domal
uplift of southern Africa, which would be anticipated to result
in a radial drainage pattern. In contrast, the most elevated
ground is associated with the marginal escarpment and the
continental interior is the relatively lower-lying site of the
Kalahari basin (Figure 4a). Further, in contrast to the radial
120 E
130 E
140 E
150 E
10 S
10 S
110 E
87
Gulf of
Carpentaria
Barkly
tablelands
20 S
20 S
Approximate
tilt axis
Cape
range
Nullarbor plain
Eucla basin
Otway
basin
Early Miocene shoreline

(adapted from Veevers, 2000)
110 E
120 E
Murray
basin
130 E
140 E
Gippsland
basin
40 S
40 S
30 S
30 S
Eyre
basin
150 E
Figure 16 The tilting continent. Geological evidence, such as the position of the early Miocene shoreline and the asymmetry of the continental
shelf, suggests that Australia is tilting northwards due to dynamic topographic effects. Reproduced with permission from Sandiford, M., 2007.
The tilting continent: a new constraint on the dynamic topographic field from Australia. Earth and Planetary Science Letters. doi:10.1016/
j.epsl.2007.06.023.
drainage predicted by the dynamic plume model, southern

Africa is characterized by a remarkable pattern of three broadly
concentric drainage divides, roughly parallel to the coastline,
and also the mid-ocean ridges surrounding Africa (Figure 17).
These three divides have been interpreted as flexure axes that
reflect three episodes of epeirogenic uplift of southern Africa
(du Toit, 1933; Moore, 1999), younging from the coast to the
interior.
5.5.5.3
Plate Boundary Stresses and Lithospheric Buckling
Buckling of the whole lithosphere due to plate boundary

stresses was first suggested for oceanic lithosphere (e.g.,
McAdoo and Sandwell, 1985), and has also been proposed for
continental lithosphere on 100 km wavelengths that may be
determined by the mechanical strength of the mantle (Tikoff
and Maxson, 2001).
The ages inferred for the three flexure axes described above
in southern Africa are broadly coeval with major changes in
plate motion at the spreading ridges surrounding southern Africa. Thus, the marginal Escarpment Axis formed at the time of
initial Gondwana breakup (B125 Ma) and would have migrated inland as an escarpment flexure, as envisaged by Gilchrist
and Summerfield (1990). The age of the central Etosha-Griqualand-Transvaal (EGT) Axis correlates with a major shift in
the pole of rotation of the Atlantic at B84 Ma (Nurnberg and
Muller, 1991). The interior Ovambo-Kalahari-Zimbabwe (OKZ)
Axis formed during the late Palaeogene, broadly coeval with
both a reorganization of spreading regime in the Indian Ocean
(Reeves and de Wit, 2000) and a marked increase in spreading
rate in the mid-Atlantic (Nurnberg and Muller, 1991). The formation of each one of the flexure axes is marked in the offshore
sedimentary record of Southern Africa by erosion and unconformities (McMillan, 2003), which also correspond with clusters of volcanic ages (Moore et al., 2008, 2009).
These correlations suggest that the flexure axes, and thus
topography in southern Africa, could be primarily linked
to deformation events associated with plate reorganizations.
Although the exact mechanisms are speculative and require
investigation, major episodes of erosion in the British Isles
have also been correlated with plate boundary tectonic episodes (Hillis et al., 2008; Holford et al., 2009), which can also
be recognized in distinct stratigaphic sequences separated by
i
ez
mb
Za river
Axi
s
88
K Z
20
Axis
MM
po
po
Lim river
M
T
e
Orang
river
30
500 km
Escar
10
pmen
20
Axis
30
40
Figure 17 Concentric pattern of epeirogenic axes in Southern Africa. Colors denote stream order: purple - 1, blue - 2, green - 3, orange - 4,
red - 5. M Molopo River, NNossob River, MM Mahura Muhtla. The major river divides are interpreted to reflect epeirogenic uplift Axes. EGT
Axis Etosha-Griqualand-Transvaal Axis; OKZ Axis Ovambo-Kalahari-Zimbabwe Axis. Data from USGS EROS, http://eros.usgs.gov/products/
elevation/gtopo30/hydro/af_streams.html, and Moore, A.E., Blenkinsop, T.G., Cotterill, F.P.D., 2009. Southern Topography and erosion history:
plumes or plate tectonics. Terra Nova 21, 310315.
unconformities (Stoker et al., 2010). The projection of compressional stresses over distances 41000 km from oceanic
ridges into the interiors of continental plates can account for
a broad spectrum of shortening-related intraplate deformation
styles which vary in scale from upper crustal folding y to
whole lithosphere buckling (Holford et al., 2009).
On the scale of 1001000 km, Australia preserves a record
of Cenozoic uplift and sedimentary basin formation (e.g.,
Flinders ranges, Torrens and Frome basins) that shows topographic undulations of several hundred meters. These have
been attributed to lithospheric buckling by Celerier et al.
(2005) because of their correlation to Bouguer gravity
anomalies. The state of stress in the Australian continent is
largely compressional (The World Stress Map web site), and
the orientation of the Flinders ranges (northsouth) can be
matched with the measured and modeled compressional
(eastwest) stresses prevailing from collisional boundary
forces in the Himalaya, the Indonesian margin, and the
New Zealand margin (Reynolds et al., 2002). In this example,
there also appears to be a clear link between continental
hinterland deformation and plate boundary stresses.
5.5.5.4
Implications for Continent-Wide Erosion Cycles

and the Origin of Uplifts
A major criticism of the concept of cyclic erosion surfaces has

always been the question of the mechanisms responsible for
their initiation. A solution to this problem is hinted at by

successive uplift along axes described above for southern
Africa located progressively inland from the coast, initiated by
intraplate stresses associated with spreading reorganizations at
the ocean ridges. Each episode of epeirogenic tectonism would
have rejuvenated the drainage network of southern Africa,
thus providing a series of triggers to initiate new cycles of
erosion in the continental hinterland (King, 1963; Lister,
1987; Partridge and Maud, 1987). The close correlation of the
ages of these flexure axes with the major unconformities in the
Congo Basin lends support to the development of contemporaneous erosion surfaces over wide areas of Africa, as
postulated by King (1963). A reevaluation of erosion surfaces
in southern African in relation to the three flexure axes offers
a potential framework for refining understanding of their
chronology and interrelationships.
The ages of the three flexure axes also correspond to continent-wide episodes of alkaline volcanism recognized by Bailey
(1993), underlining the link between epeirogenesis and associated lithospheric stresses, and alkaline volcanism (Bailey, 1993;
Moore et al., 2008). The broad upwarps represented by the
flexure axes would be associated with relative tensional stresses
in the upper surface of the plate. In contrast, the lower plate
surface would experience relative tension beneath the basins
surrounding the axes. This link between the uplift axes and the
distribution of lithospheric stresses could explain the coastward
younging of alkaline volcanism in southern Africa, which contrasts with the inland age progression of the three axes.
In summary, the observed topography of Southern Africa

does not correlate well with dynamic topography predicted by
plume models. Rather, the major features of the topography
are determined by concentric flexural uplift axes, each coeval
with an episode of plate boundary reorganization. This suggests that the dominant influence on the modern topography
in southern Africa reflects stresses associated with plate
kinetics, rather than mantle plumes.
5.5.6
Concluding Remarks
Passive margins are subject to a variety of sources of stress,

mediated through several different processes, as summarized
in Figure 6. The diversity of passive margins may reflect a
corresponding variety in the relative importance of the processes shown in the figure. The clear temporal links between
passive margin/continental hinterland evolution and plate
tectonic reorganization demonstrated for southern Africa, and
the directional links between basins and ranges and in situ
stress in Australia, are suggestive evidence that the processes
responsible for vertical movements are directly linked to firstorder plate tectonic considerations. This link can also be extended, although with less confidence, to the earlier geological
record of passive margins.
Responses to rifting vary along the length of passive margins and from one margin to another. Thus, marginal escarpments are typically absent in the vicinity of rifts associated
with triple junctions. The local lithospheric flexural thickness
will also influence the isostatic response and geomorphic
evolution of passive margins. This is exemplified in the contrast between the marked flexural uplift on the Red Sea coastal
plain and the more subdued response along the margin of
southern Africa. Where rifted margins have a thick sedimentary cover, rapid deep erosion will follow, but this is not necessarily the case where basement rocks are exposed or where
there is only a thin sedimentary capping.
Escarpments are perhaps the most spectacular and puzzling
landforms associated with passive margins. There are three
main current models for escarpment formation and evolution:
flexural isostasy with backwearing, a pinned drainage divide
with downwearing, and downwarping. Despite the application of newer techniques such as low-temperature thermochronology and numerical modeling, it is not yet clear which
of these models is generally applicable, or whether more than
one model, or combinations of them, may be appropriate for
different passive margins. This remains an area for interesting
future research.
This article has focused on MesozoicHolocene passive
margins because their record is so much more visible, but
passive margins can be traced in the geological record back to
at least the late Archean, although their abundance and their
duration seem to have changed through time (Bradley, 2008).
The abundance of passive margins appears to be closely
related to the assembly and breakup of Pangea, but not clearly
to earlier hypothesized supercontinent cycles. The greater
duration of passive margins in the Precambrian was interpreted by Bradley (2008) to indicate that, paradoxically, plate
motion was slower in the past.
89
This article suggests that several fundamental issues still

remain to be resolved about passive margin and continental
hinterland tectonic geomorphology, which to some extent
invoke larger tectonic questions. These issues can be summarized in the form of several questions:
1. Do mantle plumes play an important role, or indeed any
role, in rifting and continental breakup?
2. How do escarpments form and evolve?
3. Is the high elevation of continental hinterland sustained
over the long term, and if so, how?
4. Do detachments exist under passive margins?
The answers to these elusive questions are clearly critical in
understanding the links between plate tectonics processes and
intraplate responses.
Acknowledgments
We are grateful to Robert Dutton for assistance with the
bibliography and to Jon Nott for valuable comments on an
early draft. Excellent reviews by Paul Green and David Nash
brought a necessary degree of circumspection to the first
submission, and Lewis Owen did an excellent editorial job.
References
Bagguley, J., Prosser, S., 1999. The interpretation of passive margin depositional
processes using seismic stratigraphy; examples from offshore Namibia.
Geological Society Special Publications 153, 321344.
Bailey, D.K., 1993. Petrogenic implications of the timing of alkaline, carbonatite and
kimberlite igneous activity in Africa. South African Journal of Geology 96, 6774.
Beaumont, C., Keen, C.E., Boutilier, R., 1982. On the evolution of rifted continental
margins: comparison of models and observations for the Nova Scotia Margin.
Royal Astronomical Society Geophysical Journal 70, 667715.
Bishop, P., 2007. Long-term landscape evolution: linking tectonics and surface
processes. Earth Surface Processes and Landforms 32, 329365.
Bradley, D.W., 2008. Passive margins through Earth history. Earth-Science Reviews
91, 126.
Braun, J., van der Beek, P., 2004. Evolution of passive margin escarpments: what
can we learn from low-temperature thermochronology? Journal of Geophysical
Research 109, F04009. http://dx.doi.org/10.1029/2004JF000147.
Brown, R.W., Gallagher, K., Gleadow, A.J.W., Summerfield, M.A., 1999.
Morphotectonic evolution of the south Atlantic margins of Africa and South
America. In: Summerfield, M.A. (Ed.), Geomorphology and Global Tectonics.
Brown, R.W., Summerfield, M.A., Gleadow, A.J.W., 2002. Denudational history along
a transect across the Drakensberg Escarpment of southern Africa derived from
apatite fission track thermochronology. Journal of Geophysical Research 107,
2350. http://dx.doi.org/10.1029/2001JB000745.
Brown, R.W., Muller, R.D., Gaina, C., Struckmeyer, H.I.M., Stagg, H.M.J., Symonds,
P.A., 2003. Formation and evolution of Australian passive margins: implications
for locating the boundary between the continental and oceanic crust. Special
Publication of the Geological Society of Australia 22, 223243.
Burke, K., 1996. The African plate. South African Journal of Geology 99, 339409.
Burke, K., Gunnell, Y., 2008. The African Erosion Surface: a continental-scale
synthesis of geomorphology, tectonics, and environmental change over the past
180 million years. Geological Society of America Memoir 201, 166.
Celerier, J., Sandiford, M., Hansen, D.L., Quigley, M., 2005. Modes of active
intraplate deformation, Flinders Ranges, Australia. Tectonics 24, http://dx.doi.org/
10.1029/2004&C001679.
Conrad, C.P., Lithgow-Bertelloni, C., Louden, K.E., 2004. Iceland, the Farallon slab,
and dynamic topography of the North Atlantic. Geology 32, 177180.
Cox, K.G., 1989. The role of mantle plumes in the development of continental
drainage patterns. Nature 342, 873877.
90
Cox, K.G., 1993. Continental magmatic underplating. Philosophical Transactions of

the Royal Society of London 342, 155166.
Daradich, A., Mitrovica, J.X., Pysklywec, R.N., Willet, S.D., Forte, A.M., 2003.
Mantle flow, dynamic topography, and rift-flank uplift of Arabia. Geology 31,
901904.
De Wit, M.J., 2003. Madagascar: heads its a continent, tails its an island. Annual
Reviews of Earth Sciences 31, 213248.
Dingle, R.V., Scrutton, R.A., 1974. Continental break-up and the development of
post-Paleozoic sedimentary basins around Africa. Bulletin of the Geological
Society of America 85, 11671175.
Doucoure, C.M., de Wit, M.J., 2003. Old inherited origin for the present near
bimodal topography of Africa. Journal of African Earth Sciences 36, 371
388.
Du Toit, A.L., 1930. The volcanic belt of the Lebombo: a region of tension.
Transactions of the Royal Society of South Africa 18, 189217.
Du Toit, A.L., 1933. Crustal movements as a factor in the geographical evolution of
South Africa. South African Geographical Journal 16, 133.
Ebinger, C.J., Sleep, N.H., 1998. Cenozoic magmatism throughout east Africa
resulting from the impact of a single plume. Nature 395, 788791.
Emmel, B., Jakobs, J., Razakamanana, T., 2004. Titanite and apatite fission track
analyses on basement rocks of central-southern Madagascar: constraints on
exhumation and denudation rates along the eastern rift shoulder of the
Morondava basin. Journal of African Earth Sciences 38, 343361.
Fleming, A., Summerfield, M.A., Stone, J.O., Fifield, L.K., Cresswell, R.G., 1999.
Denudation rates for the southern Drakensberg escarpment, SE Africa, derived
from in-situ-produced cosmogenic 36Cl: initial results. Journal of the Geological
Society of London 156, 209212.
Fouch, M.J., James, D.E., Vandecar, J.C., Van Der Lee, S.The Kaapvaal Seismic
Group, 2004. Mantle seismic structure beneath the Kaapvaal and Zimbabwe
Cratons. South African Journal of Geology 107, 3344.
Fullard, H., Darby, H.C., 1974. The University Atlas (Sixteenth ed.). George Philip &
Son. Ltd., London, 176 pp.
Gallagher, K., Brown, R., 1999. Denudation and uplift at passive margins: the
record on the Atlantic margin of southern Africa. Philosophical Transactions of
the Royal Society London A357, 835859.
Gallagher, K., Brown, R.W., Johnson, C., 1998. Fission track analysis and its
application to geological problems. Annual Review of Earth and Planetary
Science 26, 519572.
Geoffroy, L., 2005. Volcanic passive margins. Comptes Rendus Geoscience 337,
13951408.
Ghebreab, W., Talbot, C.J., 2000. Red Sea extension influenced by PanAfrican tectonic grain in eastern Eritrea. Journal of Structural Geology 22,
931946.
Gilchrist, A.R., Summerfield, M.A., 1990. Differential denudation and flexural
isostacy in formation of rift-margin upwarps. Nature 346, 739742.
Gilchrist, A.R., Summerfield, M.A., 1991. Denudation, isostacy and landscape
evolution. Earth Surfaces Processes and Landforms 16, 555562.
Gilchrist, A.R., Summerfield, M.A., 1994. Tectonic models of passive margin
evolution and their implication for theories of long-term landscape development.
In: Kirby, M.J. (Ed.), Process Models and Theoretical Geomorphology. Wiley,
Gunnell, Y., Carter, A., Petit, C., Fournier, M., 2007. Post-rift seaward downwarping at passive margins: new insights from southern Oman using
stratigraphy to constrain apatite fission-track and (U-Th)/He dating. Geology 35,
647650.
Gunnell, Y., Fleitout, I., 2000. Morphotectonic evolution of the Western Ghats, India.
In: Summerfield, M.A. (Ed.), Geomorphology and Global Tectonics. Wiley,
Gurnis, M., Mitrovica, J.X., Ritsema, J., van Heijst, H.-J., 2000. Constraining
mantle density structure using geological evidence of surface uplift rates: the
case of the African superplume. Geochemistry, Geophysics, Geosystems 1,
144.
Hays, J., 1968. Notes on geomorphology. In: Walpole, B.P., Dunn, P.R., Randal,
M.A. (Eds.), Geology of the Katherine-Darwin region, Northern Territory.
Bureau of Mineral Resources, Geology and Geophysics Bulletin 82, Canberra,
pp. 161170.
Hillis, R.R., Holford, S.P., Green, P.F., et al., 2008. Cenozoic exhumation of the
southern British Isles. Geology 36, 371374.
Holford, S.P., Green, P.F., Duddy, I.R., Turner, J.P., Hillis, R.J., Stoker, M.S., 2009.
Regional intraplate exhumation episodes related to plate boundary deformations.
Huismans, R.S., Podladchikov, Y.Y., Cloetingh, S.A.P.L., 2001. Transition from
passive to active rifting; relative importance of asthenospheric doming and
passive extension of the lithosphere. Journal of Geophysical Research 106(B6),

11,27111,291.
Japson, P., Bonow, J.M., Green, P.F., Chalmers, J.A., Lidmar-Bergstrom, K., 2006.
Elevated passive continental margins: long term highs or Neogene uplifts?
New evidence from Greenland. Earth and Planetary Science Letters 248,
315324.
Japson, P., Bonow, J.M., Green, P.F., Chalmers, J.A., Lidmar-Bergstrom, K., 2009.
Formation, uplift and dissection of planation surfaces at passive continental
margins a new approach. Earth Surface Processes and Landforms 34,
683699.
King, L.C., 1949. On the ages of African land-surfaces. Quarterly Journal of the
Geological Society 104, 439459.
King, L.C., 1955. Pedeplanation and isostacy: an example from South Africa.
Quarterly Journal of the Geological Society 111, 353359.
King, L.C., 1963. South African Scenery. Oliver and Boyd, Edinburgh, 308 pp.
King, L.C., 1976. Planation remnants upon high lands. Zeitschrift fur
Klausen, M.B., 2009. The Lebombo monocline and associated feeder dyke swarm:
diagnostic of a successful and highly volcanic rifted margin. Tectonophysics
468, 4262.
Kooi, H., Beaumont, C., 1996. Large-scale geomorphology: classical concepts
reconciled and integrated with contemporary ideas via a surface processes
model. Journal of Geophysical Research 101, 33613386.
Kooi, H., Beaumont, G., 1994. Escarpment evolution on high-elevation rifted
margins: insights derived from a surface process model that combines
diffusion, advection and reaction. Journal of Geophysical Research 99,
1219112209.
Kounov, A., Niedermann, S., Viola, G., de Wit, M., Andreoli, G., 2007. Present
denudation rates at selected sections of the South African escarpment and the
elevated continental interior based on cosmogenic 3He and 21Ne. South African
Kounov, A., Viola, G., de Wit, M., Andreoli, G., 2008. A Mid Cretaceous paleoKaroo River valley across the Knersvlakte plain (northwestern coast of South
Africa): Evidence from apatite fission-track analysis. South African Journal of
Geology 111, 409420.
Kounov, A., Viola, G., de Wit, M., Andreoli, G., 2009. Denudation along the Atlantic
passive margin: new insights from apatite fission-track analysis on the western
coast of South Africa. Geological Society of London Special Publication 324,
287306.
Lister, G.S., Etheridge, M.A., Symonds, P.A., 1986. Detachment faulting and the
evolution of passive margins. Geology 14, 246250.
Lister, G.S., Etheridge, M.A., Symonds, P.A., 1991. Detachment models for the
formation of passive continental margins. Tectonics 10, 10381064.
Lister, L.A., 1987. The erosion surfaces of Zimbabwe. Bulletin of the Zimbabwe
Geological Survey, Harare 90, 163 pp.
Lithgow-Bertelloni, C., Gurnis, M., 1997. Cenozoic subsidence and uplift
of continents from time-varying dynamic topography. Geology 25,
735738.
Lithgow-Bertelloni, C., Silver, P., 1998. Dynamic topography, plate driving forces
and the African superswell. Nature 395, 269272.
Marsh, J.S., Hooper, P.R., Rehacek, J., Duncan, R.A., Duncan, A.R., 1997.
Stratigraphy and Age of Karoo Basalts of Lesotho and Implications for
correlations within the Karoo Igneous Province. American Geophysical Union
Geophysical Monograph 100, 247272.
Matmon, A., Bierman, P., Enzel, Y., 2002. Pattern and tempo of great escarpment
erosion. Geology 30, 11351138.
McAdoo, D., Sandwell, D.T., 1985. Folding of oceanic lithosphere. Journal of
Geophysical Research 90(B10), 85638569.
McKenzie, D.P., 1984. A possible mechanism for epeirogenic uplift. Nature 307,
616618.
McMillan, I.K., 2003. Foraminiferally defined biostratigraphic episodes and
sedimentary patterns of the Cretaceous drift succession (Early Barremian to Late
Maastrichtian in seven basins on the South African and southern Namibian
continental margin. South African Journal of Science 99, 537576.
Menzies, M.A., Klemperer, S.L., Ebinger, C.J., Baker, J., 2002. Characteristics of
volcanic rifted margins. In: Menzies, M.A., Klemperer, S.L., Ebinger, C.J., Baker,
J. (Eds.), Volcanic Rifted Margins. Geological Society of America (GSA),
Boulder, CO, Special Paper 362, pp. 114.
Moore, A.E., 1999. A reappraisal of epeirogenic flexure axes in southern Africa.
South African Journal of Geology 102, 363376.
Moore, A.E., Blenkinsop, T.G., 2006. Scarp retreat versus pinned drainage divide in
the formation of the Drakensberg escarpment, southern Africa. South African
Moore, A.E., Blenkinsop, T.G., 2002. The role of mantle plumes in the development
of continental-scale drainage patters: the southern African example revisited.
Moore, A.E., Blenkinsop, T.G., Cotterill, F.P.D., 2008. Controls of post-Gondwana
alkaline volcanism in southern Africa. Earth and Planetary Science Letters 268,
151164.
Moore, A.E., Blenkinsop, T.G., Cotterill, F.P.D., 2009. Southern African topography
and erosion history: plumes or plate tectonics. Terra Nova 21, 310315.
Moore, A.E., Cotterill, F.P.D., Broderick, T.G., Plowes, D., 2009. Landscape evolution
in Zimbabwe from the Permian to present, with implications for kimberlite
prospecting. South African Journal of Geology 112, 6586.
Moucha, R., Forte, A.M., Mitrovica, J.X., Rowley, D.B., Quere, S., Simmons, N.A.,
Grand, S.P., 2008. Dynamic topography and long-term sea-level variations; there
is no such thing as a stable continental platform. Earth and Planetary Science
Letters 271, 101108.
Nott, J., 1994. Long-term landscape evolution in the Darwin region and its
implications for the origin of landsurfaces in the north of the Northern Territory.
Nott, J., Horton, S., 2000. 180Ma continental drainage divide in northeastern
Australia: role of passive margin tectonics. Geology 28, 763766.
Nott, J.F., 1995. The antiquity of landscapes on the North Australian Craton and the
implications for theories of long-term landscape evolution. Journal of Geology
103, 1932.
Nott, J.F., Young, R.W., McDougall, I., 1996. Back-wearing versus down-wearing
versus gorge extension in the long-term denudation of a highland mass:
quantitative evidence from the Shoalhaven catchment, southeastern Australia.
Nurnberg, D., Muller, R.D., 1991. The tectonic evolution of the south Atlantic from
Late Jurassic to present. Tectonophysics 191, 2753.
OGrady, D.B., Syvitski, J.P.M., Pratson, L.F., Sarg, J.F., 2000. Categorizing the
morphologic variability of siliciclastic passive continental margins. Geology 28,
207210.
OSullivan, P.B., Kohn, B.P., Foster, D.A., Gleadow, A.J.W., 1995. Fission track data
from the Bathurst Batholith: evidence for rapid mid-Cretaceous uplift and
erosion within the eastern highlands of Australia. Australian Journal of Earth
Sciences 42(6), 597607.
Ollier, C.D., 1982. The Great Escarpment of eastern Australia: tectonic and
geomorphic significance. Journal of the Geological Society of Australia 29,
1323.
Ollier, C.D., 1984. Morphotectonics of continental margins with great escarpments.
In: Morisawa, M., Hack, J.T. (Eds.), Tectonic Geomorphology. Allen & Unwin,
Boston, pp. 324.
Ollier, C.D., 2004. The evolution of mountains on passive continental margins.
In: Owens, P.N., Slaymaker, O. (Eds.), Mountain geomorphology. Edward Arnold,
London, UK, pp. 5988.
Ollier, C.D., Marker, M.E., 1985. The Great Escarpment of southern Africa. Zeitung
Ollier, C.D., Pain, C.F., 1997. Equating the basal unconformity with the palaeoplain:
a model for passive margins. Geomorphology 19, 115.
Ollier, C.D., Powar, K.B., 1985. The Western Ghats and the morphotectonics of
peninsular India. Zeitschrift fur Geomorphologie 54, 5769.
Ollier, C.D., Stevens, N.C., 1989. The Great Escarpment of Queensland. In: Maitre,
Le (Ed.), Pathways in Geology Essays in Honour of Edwin Sherbon Hills.
Blackwells Scientific, Melbourne, pp. 140152.
Pain, C.F., 1985. Morphotectonics of the continental margin of Australia. Zeitung fur
Partridge, T.C., Maud, R.R., 1987. Geomorphic evolution of southern Africa since
the Mesozoic. South African Journal of Geology 90, 179208.
Pazzaglia, F.J., Gardner, T.W., 1994. Late Cenozoic flexural deformation of the
middle U.S. Atlantic passive margin. Journal of Geophysical Research 99(B6),
12,14312,157.
Pazzaglia, F.J., Gardner, T.W., 2000. Late Cenozoic landscape evolution of the US
Atlantic passive margin: insights into a North American Great Escarpment.
In: Summerfield, M.A. (Ed.), Geomorphology and Global Tectonics. Wiley,
Penck, W., 1924. Morphological Analysis of Landforms, English translation by
Czeck, H. and Boswell, K.C., London 1953, 423 pp.
Persano, C., Stuart, F.M., Bishop, P., Barford, D.N., 2002. Apatite (U-Th)/He
age constraints on the development of the Great Escarpment on the southeastern Australian passive margin. Earth Planetary Science Letters 200, 114.
Reeves, C., de Wit, M.J., 2000. Making ends meet in Gondwana: retracing the
transforms of the Indian Ocean and reconnecting the continental shear zones.
Terra Nova 12, 272280.
91
Reeves, C.V., 1978. A failed Gondwana spreading axis in southern Africa. Nature
273, 222223.
Reynolds, S.D., Coblentz, D.D., Hillis, R.R., 2002. Tectonic forces controlling the
regional intraplate stress field in continental Australia: results from new finiteelement modelling. Journal of Geophysical Research 107, http://dx.doi.org/
10.1029/2001JB000408.
Rohrman, M., Van der Beek, P., Andriessen, P., Cloetingh, S., 1995. MesoCenozoic morphotectonic evolution of southern Norway: neogene domal
uplift inferred from apatite fission track thermochronology. Tectonics 14,
704718.
Rosendahl, B.R., 1987. Architecture of continental rifts with special reference to East
Africa. Annual Review of Earth and Planetary Sciences 15, 445503.
Rowley, D.B., Sahagian, D., 1986. Depth-dependent stretching: a different approach.
Geology 14, 3235.
Royden, L., Keen, C.E., 1980. Rifting processes and the thermal evolution of the
continental margin of eastern Canada determined from subsidence curves. Earth
Rust, D.J., Summerfield, M.A., 1990. Isopach and borehole data as indicators of
rifted margin evolution in southwestern Africa. Marine and Petroleum Geology 7,
277287.
Sandiford, M., 2007. The tilting continent: a new constraint on the dynamic
topographic field from Australia. Earth and Planetary Science Letters 261,
152163. http://dx.doi.org/10.1016/j.epsl.2007.06.023.
Sandiford, M., Quigley, M., 2009. TOPO-OZ: insights into the various modes
of intraplate deformation in the Australian continent. Tectonophysics 474,
405416.
Seaward, D., Grujie, D., Schreurs, G., 2004. An insight into the breakup of
Gondwana: identifying events through low-temperature thermochronology from
the basement rocks of Madagascar. Tectonophysics 23, TC3007. http://dx.doi.org
/10.1029/2003TC001556,2004.
Sengor, A.M.C., Burke, K., 1978. Relative timing of rifting and volcanism on
Earth and its tectonic implications. Geophysical Research Letters 5,
419421.
Steckler, M.S., Omar, G.I., 1994. Controls on erosional retreat of the uplifted flanks
at the Gulf of Suez and northern Red Sea. Journal of Geophysical Research
99(B6), 1215912173.
Stoker, M.S., Holfrod, S.P., Hillis, R.R., Green, P.F., Duddy, I.R., 2010. Cenozoic
post-rift sedimentation off northwest Britain: recording the detritus of episodic
uplift on a passive continental margin. Geology 38, 595598.
Struckmeyer, H.I.M., Symonds, P.A., 1997. Tectonostratigraphic evolution of the
Townsville Basin, Townsville Trough, offshore northeastern Australia. Australian
Journal of Earth Sciences 44, 799817.
Summerfield, M.A., 1988. Global tectonics and landscape development. Progress in
Summerfield, M.A., 1991. Global Geomorphology: An Introduction to the Study of
Landforms. Longman Scientific and Technical, London, 537 pp.
Summerfield, M.A. (Ed.), 2000. Geomorphology and Global Tectonics. Wiley, Chichester.
Tassara, A., Swain, C., Hackney, R., Kirby, J., 2007. Elastic thickness structure of
South America estimated using wavelets and satellite-derived gravity data. Earth
Tikoff, B., Maxson, J., 2001. Lithospheric buckling of the Laramide foreland during
Late Cretaceous and Paleogene, western United States Rocky Mountain. Geology
36, 1315.
Tinker, J., de Wit, M., Brown, R., 2008. Mesozoic exhumation of the southern Cape,
South Africa, quantified using apatite fission track thermochronology.
Truswell, J.F., 1970. An Introduction to the Historical Geology of South Africa.
Purnell, Cape Town, 167 pp.
Tucker, G.E., Slingerland, R.L., 1994. Erosional dynamics, flexural isostacy, and
long-lived escarpments: a numerical study. Journal of geophysical Research 99,
1222912243.
Turcotte, D.L., Emerman, S.H., 1983. Mechanisms of active and passive rifting.
Van der Beek, P., Pulford, A., Braun, J., 2001. Cenozoic development in the Blue
Mountains (southeast Australia): lithological and tectonic control on rifted
margin morphology. The Journal of Geology 109, 3556.
Van der Beek, P., Summerfield, M.A., Braun, J., Brown, R.W., Fleming, A., 2002.
Modelling post breakup landscape development and denudational history
across the southeastern African (Drakensberg Escarpment) margin. Journal
of Geophysical Research 107, B12. http://dx.doi.org/10.1029/2001JB000744.
92
Wheeler, P., White, N., 2000. Quest for dynamic topography: observations from
Southeast Asia. Geology 28, 963966.
White, R.S., Mckenzie, D.P., 1989. Magmatism at rift zones: the generation of
volcanic continental margins and flood basalts. Journal of Geophysical Research
94, 76857730.
Young, R.W., 1989. Crustal constraints on the evolution of the continental divide of
eastern Australia. Geology 17, 528530.
Relevant Website
http://dc-app3-14.gfz-potsdam.de/pub/introduction/introduction_frame.html
World Stress map.
Biographical Sketch
Tom Blenkinsop is Professor and Head of Discipline in the School of Earth and Environmental Sciences, James
Cook University, Australia, where he has been working since 2002 after teaching and researching at the University
of Zimbabwe in Harare for 13 years. His main interests are in structural geology and the areas of overlap between
this subdiscipline, tectonics, and landscape evolution. His interest and experience in passive margins and continental interiors derive from studies in southern Africa and Australia, which make several fascinating comparisons. As a structural geologist, he is particularly interested in faulting and fluid flow in the upper crust, and the
role of structures in the formation of mineral deposits. These themes are linked by considerations of the past and
present states of stress in the Earths crust.
Andy Moore is Vice President (Exploration, Diamonds) of African Queen Mines a Canadian-listed mineral
exploration company, and also an honorary Research Associate of the Geology Department of Rhodes University,
South Africa. He has been involved in diamond exploration in Botswana, Zimbabwe, Namibia, Madagascar, and
South Africa since 1980. He initiated and managed private diamond exploration companies in Botswana an
undertaking that proved to be a far more interesting and character-building way to part with most of ones capital
than the gaming tables. Then, concerned about possible character deficiencies, he commenced diamond prospecting in Zimbabwe. Initial research interests were focused on kimberlite mineralogy and geochemistry and
their application to the economic evaluation of kimberlites. Kimberlite prospecting led to an interest in drainage
evolution in southern Africa and its application to interpreting the dispersion of kimberlite pathfinder minerals
away from the primary source rocks. An unexpected spin-off was exploring the linkage between landscape evolution and the evolution and dispersion of plant fish and mammal species. A more recent curiosity has been the
temporal links between continental tectonics that modified drainage evolution, and both episodes of alkaline
volcanism, and the spreading history of the ocean ridges surrounding southern Africa. In essence, an archetypal
kimberlite prospector jack of all trades and master of none.
5.6 Plateau Uplift, Regional Warping, and Subsidence

J Babault, Universitat Auto`noma de Barcelona, Barcelona, Spain
J Van Den Driessche, Universite Rennes 1, Rennes, France
5.6.1
5.6.1.1
5.6.1.2
5.6.1.3
5.6.1.3.1
5.6.1.3.2
5.6.1.3.3
5.6.1.3.4
5.6.1.3.5
5.6.1.3.6
5.6.1.4
5.6.1.4.1
5.6.1.4.2
5.6.1.4.3
5.6.1.5
5.6.1.6
5.6.2
5.6.2.1
5.6.2.1.1
5.6.2.1.2
5.6.2.1.3
5.6.2.1.4
5.6.2.2
5.6.2.2.1
5.6.2.2.2
5.6.2.2.3
5.6.2.2.4
5.6.2.3
5.6.2.4
5.6.2.4.1
An Introduction to Surface and Deep Features of High Plateaus

Components and Scales of Landscape Dynamics
Definition and Types of Plateaus at Earths Surface
The Main High Plateaus
The Tibetan plateau
The AltiplanoPuna plateau
The Colorado Plateau
The Eastern Anatolian plateau
The East African and Ethiopian plateaus
The southern African plateau
Deep Structures of the Main High Plateaus
Continentcontinent collision plateaus (Tibet and Anatolian plateaus)
Oceancontinent collision plateau (The Andes)
Intraplate plateaus (Colorado and African plateaus)
High Plateaus: Uplifted Peneplains, Growing Plateau or Applanation at High Elevation?
On the Existence of Past High Plateaus in the Earth History
Evidence for Plateau Uplift, Regional Warping, and Subsidence
Geomorphic Markers
Low-relief, high-elevation erosional surfaces
Drainage network development and reorganization on a plateau
River longitudinal profiles: Steepness indices
Longitudinal paleoprofile reconstruction of rivers
Paleoaltimetry from Sedimentology
Paleoaltimetry from marine sediments
Paleoaltimetry from paleohorizontality of lacustrine sediments
Paleoaltimetry from paleoslopes: Large-scale patterns of deposition
Grain size distribution in piedmont sedimentation
Paleoaltimetry Data Based on Paleobotany
Paleoaltimetry Data Based on Stable Isotopes
Paleoaltimetry data based on the stable isotopic records (d18O and d2H) of carbonates derived from meteoric and
surface waters
5.6.2.4.2
Paleoaltimetry data based on the abundance of 18O13C16O (D47): The carbonate clumped isotope paleothermometer
5.6.2.5
Paleoaltimetry Data Based on Paleoatmospheric Pressure from Basalt Vesicularity
5.6.2.6
Paleoaltimetry Data Based on Cosmogenic Nuclides
5.6.2.7
Cooling-History and Erosion Rates as a Proxy for Rock or Surface Uplift
5.6.2.7.1
Vertical profiles of thermochronological data combined with other lines of evidence as a proxy for rock and surface
uplift
5.6.2.7.2
Horizontal profiles of thermochronological data combined with other lines of evidence as a proxy for rock/surface
uplift
5.6.2.7.3
Incision rates obtained from thermochronological data combined with other lines of evidence as a proxy for uplift
5.6.3
Tectonic Mechanisms and Associated Surface Uplift Rates for Plateau Uplift, Regional Warping, and
Subsidence
5.6.4
Plateau Uplift and Global Climate Change
5.6.5
Conclusion
Acknowledgments
References
94
94
95
97
97
98
98
98
98
98
99
99
101
102
104
106
107
107
107
108
108
109
110
110
110
110
111
111
113
113
114
115
115
116
116
117
117
118
119
120
120
120
Babault, J., Van Den Driessche, J., 2013. Plateau uplift, regional warping,
and subsidence. In: Shroder, J. (Editor in Chief), Owen, L.A. (Ed.), Treatise
on Geomorphology. Academic Press, San Diego, CA, vol. 5, Tectonic
http://dx.doi.org/10.1016/B978-0-12-374739-6.00087-7
93
94
Plateau Uplift, Regional Warping, and Subsidence
Glossary
d2H Indicates the hydrogen isotopic composition of a
sample. It is the difference between the isotope ratio of 2H
to 1H in a sample and that in a standard, divided by the
ratio in the standard, and expressed as parts per thousand
per million.
d18O Indicates the oxygen isotopic composition of a
sample. It is the difference between the isotope ratio of 18O
to 16O in a sample and that in a standard, divided by the
ratio in the standard, and expressed as parts per thousand
per million.
Convective instability Also potential instability, a
condition in which a stable fluid forced to rise or fall
becomes unstable thereby triggering its later vertical
motion.
Cosmogenic nuclides Rare isotopes within minerals
induced by nuclear reactions induced by cosmic rays
(mostly neutrons and muons). The length of surface
exposure, or alternatively, the rate of surface erosion, is
computed from the concentration of cosmogenic nuclides
in a landform.
Delamination Separation of the dense lower part of the
continental or oceanic lithosphere replaced by warmer, less
dense mantle rock.
Endorheic basin Said of a basin characterized by internal
drainage, a condition in which no surface drainage reaches
the ocean.
Hypsometry The distribution of topographic surface area
with respect to altitude.
Knickpoints (Nickpoint) refers to a substantially
steepened reach of a stream long profile.
Laramide orogeny Also Columbian orogeny, was an

oceancontinent collision orogen like the Andes now. The
subducting plate, the Farallon oceanic plate caused
thrusting and orogen building from what are now the
south-western USA to northern South America.
Local relief Generally calculated in 5 or 10 km, a moving
window, depending on the scale at which the topography
has to be analyzed. It is the difference between the
maximum elevation and the minimum elevation. It can
also be calculated along a swath profile.
Magnetotelluric sounding The use of changing
components of the geomagnetic field to study the electrical
conductivity of the rocks within the Earth. It is generally
applied to studies of the lower crust and upper mantle.
Mean elevation The critical variable that measures the
force balance in orogens and reflects the isostatic balance of
the lithospheric column. It is generally calculated in a 30
km moving window on a digital elevation model (DEM, a
raster of altitudes), that is, at the scale of the crust thickness.
It can also be calculated along a swath profile of at least
30 km wide.
Paleosurface A relict landscape more or less preserved, often
a low-relief landscape deeply incised by the current drainage
network. Its local characteristics may vary as a function of
bedrock lithology, deformation, elevation, and latitude.
Stable isotope paleoaltimetry Determination of past
elevations based on naturally occurring stable isotopes of
an element.
Stream power The rate at which a stream can work for
erosion and transport of its load, and measured over a
specific length.
Abstract
The authors present the physiognomic and deep characteristics of the seven largest plateaus on Earth. In the second
part, the evidence used by the Earth sciences community to infer rock and/or surface uplifts of plateaus are
reviewed. Finally, the authors briefly present the debate between the feedbacks between plateaus uplift and global climate
change. Some of the results contained in this chapter are still under debate and ongoing research is still needed to confirm
them because different studies invoke for a same object contradictory processes, in that they result in conflicting uplift
history.
5.6.1
An Introduction to Surface and Deep Features

of High Plateaus
5.6.1.1
Components and Scales of Landscape Dynamics
Earths surface results from the competition between deep

processes induced by the tectonic system that deforms, raises,
or lowers the topography, and, from surface processes controlled by the erosion-transport system that destroy the highs
and fill the lows. The former deep processes are a combination
of the motion of lithospheric plates and/or mantle convection, the vertical component of which moves rocks with
respect to the geoid, the sea level, or a reference ellipsoid. This

vertical movement is generally referred to as rock uplift
following England and Molnars (1990) definition. Surface
processes rely on erosion, transport capacity, and sedimentation, processes that are at least dependant on potential energy, hence, on vertical motions.
Taking into account the duration of tectonics and geomorphic events, it is evident that the interactions between
deep-Earth and surface-Earth systems depend on the relative
rates and response times of one system with respect to the
other. For that reason the relative rates and response times act
on the resulting topographic evolution. For example, erosion
tends to compensate rock uplift, but by definition a delay

exists between both rock uplift and removal of rocks, otherwise no mountains would form.
Large-scale tectonic processes act on the whole thickness
of the crust (B30 km) or the entire lithosphere (B100 km
thick). Large-scale tectonics breaks, folds, tilts, raises, and
lowers the landscape over a very large expanse of land, by
more than 103 104 km2 (England and Molnar, 1990), defining in this way, the regional-scale of tectonic geomorphology. Consequently, the study of plateau uplift, regional
warping and subsidence focus on extensive landmasses. More,
in the understanding of plateau generation, rock uplift data
only may not be a pertinent measure of plateau uplift since
the surface uplift equals the rock uplift minus the erosion (or
tectonic denudation) (England and Molnar, 1990; Molnar and
England, 1990). Evidence of plateau uplift will be discussed in
Section 5.6.2 of this chapter.
Driving mechanisms that produce plateaus uplift have
received considerable attention in the literature over the past
four decades, since the broad acceptance of plate tectonics
are still under debate. Timing and rates of vertical movements
are critical to understand which processes are the causes of the
vertical forces acting on topography. Low rates of surface uplift
(o0.5 mm yr1) can be explained by crustal shortening,
thickening and consequent isostatic compensation, whereas
high rates of surface uplift (40.5 mm yr1) during one or
several millions of years have their origin in mantle dynamics
(Garzione et al., 2006). Mantle processes responsible for high
rates of surface uplift are (among others): mantle lithosphere
delamination following intracontinental subduction (Bird,
1978); convective instability of cold, dense thickened-lithospheric mantle in orogens; and, its replacement by relatively
hotter and lighter mantle asthenosphere (England and
Houseman, 1989; Houseman et al., 1981). This point will be
150 W
120 W
90 W
60 W
30 W
elaborated in Section 5.6.3 of this chapter. The elevation history of plateaus, and how long high elevations of plateaus last,
is also of great interest in order to understand the interaction
between asthenospheric, lithospheric and climatic processes.
Since paleoelevations can only be obtained using proxies, the
way the inferred estimates of vertical motions occur are greatly
interpreted and debated (e.g., Molnar, 2005).
Paleoelevation data are, in general, sparse in continental
regions; generally they cover only a very small part of a plateau, preventing them being used as representative of the
whole plateau (e.g., Molnar et al., 2006). Recently, new tools
for paleoelevations determination based on stable isotope
paleoaltimetry (13C-18O bonds) provide new evidence for the
dynamics of surface uplift and the nature of its driving
mechanism in the Andes (Garzione et al., 2008; Ghosh et al.,
2006b; Hoke et al., 2009) and in Tibet (Garzione, 2008).
However, doubts have been shed on the reliability of this new
method (Poulsen et al., 2010).
5.6.1.2
Definition and Types of Plateaus at Earths Surface
As defined in the Glossary of Geology (Neuendorf et al.,

2005), a plateau is, broadly, any comparatively flat area of
great extent and elevation, specifically an extensive land region
considerably elevated (4150 m in altitude) above the adjacent country or above sea level. It is commonly limited on
at least one side by an abrupt descent; has a flat or nearly
smooth surface but is often dissected by deep valleys and
surmounted by high hills or mountains; and has a large part
of its total surface at or near the summit level. Although the
term tableland also applies to a broad, elevated region with a
nearly level or undulating surface of considerable extent
(Glossary of Geology, Neuendorf et al., 2005), it might be
30 E
60 E
90 E
120 E
150 E
60 N
EAn
T
30 N
A
0
EAfr
Elevation (m)
30 S
60 S
0 150
150 500
500 1000
1000 2000
2000 5000
5000 8300
95
AP
SA
Figure 1 Location of the main studied uplifted-plateaus (black outlined areas) on the ETOPO1 DEM. T: Tibetan Plateau, A-P: AltiplanoPuna
plateau, C: Colorado Plateau, EAn: Eastern Anatolian, A: Afar (or Ethiopian) Plateau, SA: southern African plateau, EAfr: East African plateau.
Green, Red, and Blue curves correspond to divergent, transform-fault and convergent plate boundaries, respectively.
Normalized cumulative area (Grey)

0.00
0.25
0.50
0.75
1.00
7000
7000
5000
4000
3000
5000
4000
3000
2000
2000
1000
1000
0
0.00
(a)
0.01
0.02
Normalized area (Black)
0.00
(b)

0.00
7000
0.75
0.00
1.00
4000
3000
2000
1000
1000
0
0.01
0.02
0.04
0.00
(d)
0.01
0.00

0.25
0.50
0.75
1.00
7000
Afar plateau
Altitude (m)
6000
5000
4000
3000
0.02

0.2
0.4
0.6 0.8
1.0
East Anatolian plateau
6000
0.75
5000
2000
7000
0.50
Colorado plateau
6000
3000
0.0
0.25
7000
4000
(c)
0.01
0.02
1.00
Altitude (m)
Altitude (m)
0.50
5000
0
0.00
Altitude (m)
0.25
AltiplanoPuna
6000
5000
4000
3000
2000
2000
1000
1000
0
0.00
(e)
0.01
0.02
0.04
0.00
(f)
0.01
0.02

0.00
0.25
0.50
0.75
1.00

0.00
0.25
0.50
0.75
1.00
7000
7000
Southern African plateau
6000
4000
3000
East African plateau
6000
Altitude (m)
5000
2000
5000
4000
3000
2000
1000
1000
0
0.00
(g)
Tibetan plateau
6000
Altitude (m)
Altitude (m)

0.00
0.25
0.50
0.75
1.00
Landmasses
6000
Altitude (m)
96
0.01
0.02
0.00
(h)
0.01
0.02
0.07
used for smaller plateau-like features (Press et al., 2004) such

as the Cheyenne Tablelands, a preserved interfluve between
the North and South Platte Rivers located in western Nebraska
and southeastern Wyoming, USA (e.g., Leonard, 2002).
From the theory of plate tectonics, the authors know that
continental collision produces orogens and, in some cases,
high plateaus such as the Tibetan Plateau in central Asia or the
AltiplanoPuna plateau in southern America, for the largest
and highest plateaus on Earth. However, plateaus can also
develop in plate-interior regions, such as the Colorado Plateau
in the western US, the Ordos block in China or the southern
African plateau. Plateaus also exist in the oceans and they have
been extensively studied since the onset of oceanographic
campaigns and, sediments and rocks sampling of the seafloor
since 1968 when the Deep Sea Drilling Project started, then
during the Ocean Drilling Program and now within the Integrated Ocean Drilling Program. The authors center this
chapter on landmass plateaus and leave out any synthesis of
studies concerned with the oceanic plateaus. Broadly, oceanic
plateaus are believed to be formed by extensive basaltic floods
induced by rising mantle plumes. The thickening of the
oceanic crust and the replacement of cold mantle lithosphere
by a hot mantle plume produces significant surface uplift
(Olson and Nam, 1986) and generates a plateau that is usually, at least in part, subaerial. After the onset of magmatic
activity, the relative plate motion from the plume source forces
the plateau to subside below sea level (Coffin, 1992; Detrick
et al., 1977). Although the origin of oceanic plateaus is still
under debate as it is the case for the Ontong-Java plateau, the
largest basaltic plateau on Earth (e.g., Roberge et al., 2005),
most studies agree that mantle convection affects subsidence
and bathymetry of oceanic sea floor (e.g., Sutherland et al.,
2010).
Continental high plateaus stand on almost every continent
(Figure 1), and if not, they may have formed in the past and
they may have then subsided as has occurred in Australia (e.g.,
Gurnis et al., 1998), or collapsed and eroded as in the Eastern
US (Coney and Harms, 1984). The mean elevation of most of
the studied high-plateaus lies above 1000 m above sea level
(asl) contrasting with the 600 m mean elevation of the con-
97
tinental areas (Figure 2). This chapter will summarize the

main geomorphologic characteristics of the most carefully
studied high plateaus.
5.6.1.3
The Main High Plateaus
To compare the topography of the Tibetan Plateau, the

AltiplanoPuna plateau, the Colorado Plateau, the Eastern
Anatolian plateau, the East African and Ethiopian plateaus,
and the southern African plateau (Figure 1) the authors performed a new analysis of elevation and slope frequencies
(Figures 2 and 3) using the digital elevation model (DEM,
SRTM90v4) published by Jarvis et al. (2008).
5.6.1.3.1
The Tibetan plateau
The internally drained area of Tibet (Figure 1) is situated on

the average at 5023 m (7200 m) asl, it is characterized by
mean slopes of 51 over 250 m windows (Fielding et al., 1994).
At longer wavelengths of 100-km-wide swaths crossing Tibet at
its greatest width and length, Fielding et al. (1994) obtained
relief (maximum minus minimum elevations) estimates of
B1 km or less for most of Tibet. The authors performed an
analysis using the SRTM90v4 DEM published by Jarvis et al.
(2008) and they obtained a similar mean elevation (5028 m,
Figure 2) to Fielding et al. (1994) for the internally drained
Tibetan Plateau (B700 000 km2). Tibet is an almost flat and
high area of great extent (Figure 4).
Almost 60% of the local slopes are less than 51 (Figure 3).
Recent and active tectonic structures such as compressional
mountain ranges, strikeslip fault-related uplifts and normal
fault-related escarpments within the plateau (e.g., Liu-Zeng
et al., 2008; Molnar and Tapponnier, 1978) explain the other
40% of slopes that are greater than 51. The narrow Kunlun
Mountains define the northern edge of the Tibetan Plateau, its
steepest boundary. To the south it is bounded by the 2500-kmlong Himalaya which, mostly in its central part, defines a steep
boundary (Duncan et al., 2003). To the east the Tibetan
Plateau slopes gently over 1000 km at a mean elevation of
3500 m asl up to the abrupt Longmen Shan topographic front.
Figure 2 Elevation frequencies (black lines) and hypsometries (heavy grey lines) computed at 10 m bins for the main-studied uplifted-plateaus
of the world (see Figure 1 for locations). (a) Histogram of the global landmasses used as a reference plot. Elevation values are derived from the
SRTMv4 DEM (Jarvis, A., Reuter, H.I., Nelson, A., Guevara, E., 2008. Hole-filled SRTM for the globe Version 4. Available from the CGIAR-CSI
SRTM 90 m: http://srtm.csi.cgiar.org) excepted for the global landmasses analysis (a) for which elevations have been extracted from ETOPO1
bedrock (Amante, C., Eakins, B.W., 2009. ETOPO1 1 Arc-minute global relief model: procedures, data sources and analysis. In: Memorandum,
N.T. (Ed.), National Geophysical Data Center, Marine Geology and Geophysics Division, Boulder, CO, p. 19). Hypsometries of Tibet (b) and
AltiplanoPuna (c) are computed for the internal drainage areas. The boundaries of the Colorado Plateau (d) used in this study are that used in
Roy, M., Jordan, T.H., Pederson, J., 2009. Colorado Plateau magmatism and uplift by warming of heterogeneous lithosphere. Nature (London)
459, 978982. The boundaries of the Eastern Anatolian Plateau (e) are from that defined by S- engor, A.M.C., Ozeren, M.S., Keskin, M., Sakinc,
M., Ozbakir, A.D., Kayan, I., 2008. Eastern Turkish high plateau as a small turkic-type orogen; implications for postcollisional crust-forming
processes in turkic-type orogens. Earth-Science Reviews 90, 148. The boundaries of the Afar plateau (f) and the East African plateau (h) are
from the boundaries published by Ebinger, C.J., Bechtel, T.D., Forsyth, D.W., Bowin, C.O., 1989. Effective elastic plate thickness beneath the East
African and Afar plateaus and dynamic compensation of the uplifts. Journal of Geophysical Research 94, 28832901; the narrow rift valleys of
the Main Ethiopian Rift, the Kenyan and Western rift systems, which correspond to flat and depressed areas superimposed on the plateaus, have
been removed for the topographic analysis. Elevations for the southern African plateau (g) correspond to the landmasses south of 201 S and in
the hinterland of southern Africa behind the Great Escarpment as defined by Ollier, C.D., Marker, M.E., 1985. The great escarpment of southern
Africa. Zeitschrift fur Geomorphologie. Supplementband 54, 3756. Reproduced from Gurnis, M., Mitrovica, J.X., Ritsema, J., van Heijst, H.-J.,
2000. Constraining mantle density structure using geological evidence of surface uplift rates; the case of the African Superplume. Geochemistry,
Geophysics, Geosystems 1(7), 1020. doi:10.1029/1999GC000035.
98
To the west, the Tibetan Plateau merges with the jagged relief
of the Karakoram and Pamir ranges.
5.6.1.3.2
The AltiplanoPuna plateau
The AltiplanoPuna plateau is the second highest on Earth

(B4000 m mean elevation, Isacks, 1988) with much of the
plateau (a 250300 km wide area) being internally drained
(Vandervoort et al., 1995) (Figure 5). Isacks (1988) defined
the plateau as the area of landmasses lying above the 3-km
elevation contour which is considerably broader than the
internally drained areas of the Altiplano and Puna. The
authors analyzed the distribution of elevations and slopes
within the internally drained areas, using the SRTM90v4
DEM published by Jarvis et al. (2008), in order to discard the
deep incisions of the Andean plateau margins and to get a
mean elevation closer to the definition of surface uplift following the definition of England and Molnar (1990). They
obtained from the SRTM90v4 (Jarvis et al., 2008) a 4020 m asl
mean elevation (Figure 2(c)) for the internally drained areas
(B330 000 km2, Figure 1), a value a little higher than the
3660 m asl mode value (Figure 2(c)) that corresponds to the
Salar de Uyuni, in Bolivia. Other spikes in the histogram
(Figure 2(c)) correspond to other major endorheic basins like
the Salar de Atacama basin at 2300 m asl in northern Chile
(Figure 2(c)). On the AltiplanoPuna plateau, the mean elevation is higher than the mode value and, than all the spikes
in the histogram (Figure 2(c)) that are due to the mountains
that surmount the main sedimentary basins. The mean elevation of the Puna is approximately 1000 m higher than that
of the Altiplano (Allmendinger et al., 1997; Isacks, 1988;
Wdowinski and Bock, 1994; Whitman et al., 1996). The cumulative hypsometric curve (Figure 2(c)) shows that 80% of
the internally drained Andes lie between 2270 m and 4400 m
asl. The AltiplanoPuna plateau is one of the flattest plateaus
on Earth, 84% of local slopes are less than 51 (Figure 2) and
B70% are less than 0.61 (1%). The great number and extent of
sedimentary basins, combined with the lack of major active
tectonics in the AltiplanoPuna plateau, explains the smooth
Andean landscape. The AltiplanoPuna plateau is flanked to
the West by the Western Cordillera, a smooth edge sloping
gently toward the Pacific trench (Faras et al., 2005; Isacks,
1988; Lamb et al., 1997), whereas the Eastern flank, the
Eastern Cordillera, consist of a rough topography (Horton,
1999; Isacks, 1988; Lamb and Hoke, 1997; Lamb et al., 1997;
Masek et al., 1994) containing smooth, highly elevated,
topographic remnants of low-relief paleosurfaces (Kennan,
2000; Kennan et al., 1997; Servant et al., 1989).
5.6.1.3.3
The Colorado Plateau
The Colorado Plateau (Figure 1) in the southern United States

of America has a mean elevation of 1941 m asl (Figure 2). This
is the third highest plateau on Earth. The physiographic extent
of the Colorado Plateau used here is the same as defined by
Hunt (1956) and also used recently by Pederson et al. (2002)
and by Roy et al. (2009). The plateau is surrounded to the
northwest and to the south by the Basin and Range Province,
to the northeast by the Rocky Mountains, to the east by the
Rio Grande Rift. Mostly in its southwestern part, the plateau is
deeply dissected by the Colorado River and its tributaries
(Rigby, 1977). The western and central parts consist of high
plateaus separated from each other by faults or more or less

incised by canyons (Rigby, 1977). The southeastern part of the
plateau is extensively covered by volcanic rocks whereas the
northern part consists of the Uinta Basin (Rigby, 1977). Only
63% of the local slopes are less than 51 (Figure 3). The relative
roughness of the Colorado Plateau is mostly a consequence of
its deep dissection by the Colorado River and its tributaries
(Pederson et al., 2002).
5.6.1.3.4
The Eastern Anatolian plateau
The East Anatolian plateau in Turkey (Figure 1) is bounded to

the north by the Caucasus and to the southeast by the Zagros
Mountains, and it extends from longitudes 411 E to 451 E (e.g.,
S- engor et al., 2008) and has a mean elevation of 1947 m asl
(Figure 2). Two depressions at elevations above 1500 m asl
compose the plateau: the Erzurum-Kars plateau in the north
(Atalay and Koc- man, 1979) and the Murat region in the
south. They are separated by the Central Range with crest
above 3000 m asl (e.g., S- engor et al., 2008). The Murat region
is an internally drained region of 12 000 km2 restricted to the
surrounding of the Saline, 450-m-deep and 4406 km2 of extent, Lake Van, perched at 1643 m (Spike in Figure 2(e)).
Elsewhere, the Murat plateaus are dissected by permanent
powerful streams including: the Euphrates; the Tigris; the
Greater Zap; the Araxes; and their tributaries (e.g., S- engor
et al., 2008). Only 38% of the local slopes are less than 51
(Figure 3), making the East Anatolian plateau the most
rugged of all the plateaus described in this chapter. Recent
deformation and deep incisions easily explain such a rough
topography.
5.6.1.3.5
The East African and Ethiopian plateaus
A striking feature of the almost 3000 km of the East African

Rift System is the presence of high plateaus surrounding the
axial rift valley (Figure 1). These highly elevated topographies
are distributed in two broad domes: the 1500-km-wide East
African (or Kenyan) dome, or plateau, and the 1000-km-wide
Ethiopian (or Afar) dome, or plateau. Such plateaus are
characterized by elevations mainly above 800 m asl (Ebinger
et al., 1989). Analysis of the topography of the high plateaus
only, using the SRTM90v4 DEM, yields a mean elevation
1572 m asl and 1274 m asl for the Afar and East African
plateaus, respectively (Figure 2). The frequency of local slopes
(Figure 3) highlights the highest degree of dissection of the
Afar plateau, by the Blue Nile among other streams (Ayalew
and Yamagishi, 2004; Gani et al., 2007; Pik et al., 2003), with
only 55% of slopes less than 51 compared to 77% for the
less incised East African plateau. The spike in the histogram of
elevations (Figure 2, Afar plateau) is approximately 2000 m
asl, corresponding to the Tana basin, which resulted from
the junction of three grabens during the late Cenozoic
(Chorowicz et al., 1998).
5.6.1.3.6
The southern African plateau
The authors used the SRTM90v4 DEM published by Jarvis

et al. (2008) to analyze the region south of latitude 201 S and
enclosed by the Great Escarpment that bound the plateau on
the coast of South Africa. The South African plateau has a
mean elevation of 1166 m asl (Figure 2), the lowest value of
the seven plateaus we analyzed. Local slope frequency reveals
Frequency
0.6
0.4
0.03
0.02
Frequency
Area of plateau with slope < 5

Southern African plateau 92%
AltiplanoPuna 84%
East African plateau 77%
Colorado plateau 63%
Tibetan plateau 59%
Afar plateau 55%
East Anatolian plateau 38%
99
0.01
0.2
0.0
0
0.00
0
10
Local slope ()
15
20
25
30
35
Local slope ()
Figure 3 Relative occurrence (frequency) of local slopes binned at 0.01 (m m 1); (left) lower than 51, and (right) higher than 51. Data are
derived from the SRTMv4 DEM. Reproduced from Jarvis, A., Reuter, H.I., Nelson, A., Guevara, E., 2008. Hole-filled SRTM for the globe Version
4. Available from the CGIAR-CSI SRTM 90m: http://srtm.csi.cgiar.org.
Figure 4 View of the Tibetan Plateau between the Tanggula Shan

and the Nyainqentangulha Shan. In the foreground the elevation is
B4500 m asl, and the summits in the background are at
50005200 m asl. Photo taken by Lewis Owen on the Golmud and
Lhasa highway.
that the southern African plateau is also the flattest with 92%
of its area with slopes lower than 51 (Figure 3). The southern
African plateau, the East African plateau and the southeastern
Atlantic Ocean basin to the southwest form together a region
of anomalous topography and bathymetry called the African
Superswell (Nyblade and Robinson, 1994).
5.6.1.4
Deep Structures of the Main High Plateaus
The crust and upper mantle structures for these high plateaus
are important constraints for evaluating possible surface uplift
mechanisms. The following is an up-to-date summary of our
knowledge of their physical and thermal structures; most of
these results and interpretations are still under debate and will
probably be improved in the near future.
Figure 5 View to the south-east of the Puna plateau. The Salar in

the background (white area close to the center of the picture) is at
4200 m asl and the highest summits in the right part of the picture
are the Cordon de Puntas Negras, a chain of volcanoes reaching
5850 m asl. Photo taken by Dominique Savanier at 5400 m asl, from
the Lascar volcano located on the eastern margin of the Salar de
Atacama (Chile) and west of the Aguas Calientes Volcano.
5.6.1.4.1
Continentcontinent collision plateaus (Tibet

and Anatolian plateaus)
5.6.1.4.1.1 The Tibetan Plateau

Located within a broad zone of NS continentcontinent collision (Figure 1), compression is active on the
borders of the Tibetan Plateau, whereas extension is active
in its center and orientated EW (e.g., Armijo et al.,
1986; Molnar and Tapponnier, 1978; Chapters 5.3 and
5.15).
The Tibetan Plateau is underlain by 60-up to 80-km-thick
crust (Figure 6), with the maximum thickness being just beneath the Zangbo (Tsangpo) suture at B301 N (Hirn et al.,
1984a, b; Holt and Wallace, 1990; Li and Mooney, 1998).
A widespread low-velocity zone is evidenced from seismic
studies in the midcrust and lower crust, and is interpreted as
Depth (km)
32
Tibet
33
34
10
5
Qiangtang block
Lhasa block
Altitude (km)
Ga
n
Himalaya
31
0
20
40
60
80
100
70
40
(a)
100
200
300
Himalayan orogenic
MHT
prism
Ind. upper cr.

Ind. lower cr.
Depth (km)
30
Moho depth (km)
29
28
BNS
Nepal
ge
M sB
FT .
M
BT
27
YTS
M
C
STT
D
100
400
500
31N boundary
7.0 km s1
5.0
600
700
Shallow LVZs
Tibetan crust
50
Mantle
100
India
ia
Mantle
ras
Eu
150
200
0
(b)
100
200
300
400
500
600
700
Songpan-Ganz
terrane
Qaidam
basin
Distance from MFT (km)

19911992 Passcal Tibetan plateau experiment
28
31
SANG
35
34
33
AMDO WNDO
lun
un
N.K fault
30
LHSA
Qiantang
terrane
a
sh
Jin ture
su
STD
MBT
MCT
Lhasa terrane 32
29
us
Ind ngpo
a
Ts ture
su
an
thy a
Te alay
Him
26
ng
go
ng e
Ba utur
s
ya
ala
im
ya
hH
ala
Hig
Him
er
ss
Le
Indepth
27
36
37
ERDO BUDO
38
TUNL
MHT RZT GT
Moho
40
Moho
Inefficient Sn-low Pn- zone
Lower crust
Indian mantle
Lithosphere
Anisotrophy developed by
lateral flow
vp = 6.37.5 km s1
200
240
200 km
vp = 7.27.5 km s1
Lower crust
Tanggula
shan
Himalayas
Zangbo Sut.
(Early tertiary)
India
80
Asian
120
continental
lithosphere 160
an
thy
Te eanic e?
oc pher
os
lith
Low S-wave zone
SW
40
N
Indian cru
st
(c)
39
~ 2510 My
Bangong Sut.
(Jurassic)
~ 12040 My
v ~ 5515 My
K
Kunlun
shan
Jinsha Sut.
(Triassic)
+ v ~5020 My
NE
Qilian
Shan
Kunlun Sut.
(Paleozoic)
Qilian Sut.
(Lower Paleozoic)
+ v ~200 My
E-N1
GOBI
N2 -Q
70
150
?
300
450
km 0
(d)
600 km
Eocene
to
present
Eocene
to
miocene
Miocene
to
quaternary
Pliocene
to
present
Figure 6 Deep structure of the Himalaya-Tibetan Plateau. (a) Receiver function image along the main profile showing the principal contrasts within the
lithosphere (red and blue colors represent interfaces with increasing and decreasing impedance with depth, respectively. Reproduced from Nabelek, J.,
Hetenyi, G., Vergne, J., et al., 2009. Underplating in the Himalaya-Tibet Collision Zone Revealed by the Hi-CLIMB Experiment. Science 325, 13711374.
(b) Interpretative cross section of the IndiaEuria collision zone. Reproduced from Nabelek, J., Hetenyi, G., Vergne, J., et al., 2009. Underplating in the
Himalaya-Tibet Collision Zone Revealed by the Hi-CLIMB Experiment. Science 325, 13711374. (c) Representative cross-section of the Tibetan Plateau
modified from the work of Owens, T.J., Zandt, G., 1997. Implications of crustal property variations for models of Tibetan plateau evolution. Nature 387,
3743; Beaumont, C., Jamieson, R.A., Nguyen, M.H., Medvedev, S., 2004. Crustal channel flows: 1. Numerical models with applications to the tectonics
of the Himalayan-Tibetan orogen. Journal of Geophysical Research 109, B06406; using information compiled by DeCelles, P.G., Robinson, D.M., Zandt,
G., 2002. Implications of shortening in the Himalayan fold-thrust belt for uplift of the Tibetan Plateau. Tectonics 21, 1062; Johnson, M.R.W., 2002.
Shortening budgets and the role of continental subduction during the IndiaAsia collision. Earth-Science Reviews 59, 101123; Tilmann, F.,
Ni, J., Team, I.I.S., 2003. Seismic imaging of the downwelling Indian lithosphere beneath Central Tibet. Science 300, 14241427; Haines, S.S.,
Klemperer, S.L., Brown, L., et al. , 2003. INDEPTH III seismic data: from surface observations to deep crustal processes in Tibet. Tectonics 22, 1001.
(d) Representative cross-section of the Tibetan Plateau by Tapponnier, P., Zhiqin, X., Roger, F., Meyer, B., Arnaud, N., Wittlinger, G., Jingsui, Y., 2001.
Oblique stepwise rise and growth of the Tibet Plateau. Science 294, 16711677, in which the Asian continental lithospheric manlte is preserved,
decoupled from the Asian thickened crust and subducted. Note that cross-sections (a) and (b) correspond only to the left hand half part of the crosssections (c) and (d) that share the same horizontal scale.
partly molten crust or a layer of aqueous fluids (Xu et al.,

2007; Yao et al., 2008). Magnetotelluric data show that this
low-velocity zone is consistent with a weak crust at depth of
2040 km distributed in two main zones that extend from the
Tibetan Plateau as far as 800 km into southwest China (Bai
et al., 2010).
At lithospheric scale, some seismological studies of the
central Tibetan Plateau (at 921 E to 931 E) argue for the Indian
plate underthrusting southern Tibet (Figures 6(a) and 6(b))
at least up to the south Lhasa block, B291 N (Hirn et al.,
1984b; Li et al., 2008; Nelson et al., 1996; Schulte-Pelkum
et al., 2005) but with a northern boundary uncertain. Beneath
northern Tibet, north of B341 N (McNamara et al., 1995,
1997; Ni and Barazangi, 1983), and south of the Kunlun
range, south of B361 N (Wittlinger et al., 1996), studies argue
for a low-velocity body in the mantle (Figure 6(c)). Together
with the mechanical properties of the crust of northern Tibet
(Owens and Zandt, 1997) and the volcanic activity of northern Tibet (Turner et al., 1993), such seismic properties have
been interpreted as a portion of crust and mantle of higher
temperature, that is, a thinned lithosphere and a hot and
partially molten crust (Molnar et al., 1993; Owens and Zandt,
1997; Zhu et al., 1995). Such a 250300 km dome-like
structure of hot mantle lies between two cold lithospheres
characterized by relatively high seismic velocities (Wittlinger
et al., 1996). These deep physical and thermal structures have
been interpreted as the evidence for the underthrusting of the
cold Indian lithospheric mantle beneath the Asian crust up to
the BangongNujiang suture at B331 N (Figure 6(c)) (Owens
and Zandt, 1997). Later, Kosarev et al. (1999) reexamined
these data and suggested that the Indian lithosphere has
underthrust the entire Asian lithosphere (not only the crust)
from the base of the Moho (80 km depth), 50 km north of
B29.51 N, where the IndusTsangpo suture is located, as far
as the BangongNujiang suture at B331 N, up to a depth of
200 km where it is inferred to be cut by a mostly vertical
subducting structure. From 331 N to 36.51 N, Kosarev et al.
(1999) also showed a discontinuous Asian lithospheric mantle made of fragments interpreted as a process of destruction
and subduction of the sinking Asian lithosphere. Alternatively,
Tapponnier et al. (2001) proposed the origin of the low velocity zone between the Tanggula and the Kunlun ranges (between 341 N and 361 N) to lie in the thick crust of the plateau.
Their interpretation is based on the seismic work of Griot et al.
(1998) and it implies that the Asian lithosphere has not been
thinned (Figure 6(d)). Recently, Nabelek et al. (2009) have
shown with a continuous seismic image that the Indian
lithosphere has underthrust the Asian crust only, and that the
Indian lower crust extends horizontally beneath the Tibetan
Plateau up to approximately 311 N (Figures 6(a) and 6(b)), as
it is present in eastern and western Tibet (Kind et al., 2002;
Wittlinger et al., 2004). The Asian and Indian crust are shown
to be separated by the main Himalayan thrust (MHT) from
Nepal (B27.51 N) into a midcrustal low velocity zone observed across the southern Lhasa Block (B29.51 N311 N).
Based on the differences in the seismic characteristics of the
lower crust and uppermost mantle, Nabelek et al. (2009)
proposed that the zone located around 311 N321 N, and
south of the BangongNujiang suture (B331 N), is a boundary between the Eurasian plate and the Indian plate that is
101
underthrusting the Tibetan upper crust. The Nabelek et al.

(2009) interpretation shifts to the south the plate boundary
inferred to be at B331 N by Kosarev et al. (1999). South and
north of 311 N321 N Nabelek et al.s (2009) boundary,
mantle lineations dip north and south, respectively, suggesting
a mantle downwelling along this plate boundary (Figures
6(a) and 6(b)), a continental subduction also put in evidence
east of Nabelek et al.s profile (Shi et al., 2004; Tilmann et al.,
2003). Such continental subduction of the Indian and Eurasian plates beneath and south of the BongongNujiang suture (southern Tibet) mainly agrees with the lithospheric-scale
cross-section proposed by Tapponnier et al. (2001)
(Figure 6(d)) and based on compiled surface and subsurface
data (see references in Tapponnier et al., 2001). One of the
main differences between the two models of the deep structure
of the Tibetan Plateau is the preservation, or not, of the Asian
lithosphere (Figures 6(c) and 6(d)). In Tapponnier et al.s
(2001) model, the crust thickened since the collision of Indian
and Asian plates, whereas the mantle did not. Such a model
implies that the lithospheric mantle is decoupled from the
weak lower crust, and subducted into the hot asthenospheric
mantle.
5.6.1.4.1.2 The Eastern Anatolian and Iranian plateaus
Like the Tibetan Plateau, the Eastern Anatolian plateau
corresponds to a zone of northsouth continental collision
(Figure 1). Folding is confined to the southwestern part of
the plateau (as defined by S- engor et al., 2008) and thrusting
is confined to eastwest depressions in the southwestern,
central, and northeastern parts of the plateau (for a compilation of neotectonic data see S-engor et al., 2008). However,
the plateau is dominated by strikeslip faults: a sinistral
northeastsouthwest orientated set and a dextral northwest
southeast striking set (S- engor et al., 2008 and references herein).
Seismic data show a mean crustal thickness of 45 km beneath the eastern Anatolian plateau (Figure 7) with 42 km
near the southern part of the Biltis suture zone (the southern
edge of the plateau) and 50 km along the northern edge of the
Anatolian plateau, whereas in the southeastern part of the
plateau the crust is thinner than 40 km (Zor et al., 2003).
Furthermore, seismic data argue for a thinned or totally
removed lithospheric mantle beneath the plateau (Al-Lazki
et al., 2004; Gok et al., 2003, 2007) which is consistent
with the high heat flow and the late Cenozoic volcanic
activity (Keskin, 2003; Pearce et al., 1990; Yilmaz et al., 1998).
Tomography indicates that it is not only the Turkish plateau
but all the TurkishIranian plateau, the Middle East region
north of the BitlisZagros suture, that is affected by a lithospheric mantle thinning and its replacement by a warm or
perhaps partially molten uppermost mantle (Maggi and
Priestley, 2005; Tabatabai Mir et al., 2008) to a depth of
B200 km (Zor, 2008). The region also lacks subcrustal
earthquakes, which is evidence for the absence of the subducting Arabian Plate beneath the Anatolian plateau (Turkelli
et al., 2003).
5.6.1.4.2
Oceancontinent collision plateau (The Andes)
The Andes result from oceanic-continent convergent margin

processes. The Nazca oceanic plate is subducted on a 301-eastdipping direction beneath the AltiplanoPuna plateau,
102
Crust
PML
Crust
BitlisPoturge massif
Crust
Sublithospheric mantle
AML
PML = Pontide mantle lithosphere

AML = Arabian mantle lithosphere
100 km
Figure 7 Lithospheric structure beneath eastern Anatolia from Gogus, O.H., Pysklywec, R.N., 2008. Mantle lithosphere delamination driving
plateau uplift and synconvergent extension in eastern Anatolia. Geology 36, 723726, originally modified from S- engor, A.M.C., Ozeren, S., Genc- ,
T., Zor, E., 2003. East Anatolian high plateau as a mantle-supported, north-south shortened domal structure. Geophysical Research Letters 30,
8045; Dhont, D., Chorowicz, J., 2006. Review of the neotectonics of the Eastern TurkishArmenian Plateau by geomorphic analysis of digital
elevation model imagery. International Journal of Earth Sciences 95, 3449; Gok, R., Michael, E.P., Ekrem, Z., 2007. Lithospheric structure of the
continentcontinent collision zone: eastern Turkey. Geophysical Journal International 169, 10791088, and Keskin, M., 2003. Magma generation
by slab steepening and breakoff beneath a subduction-accretion complex: an alternative model for collision-related volcanism in Eastern Anatolia,
Turkey. Geophysical Research Letters 30, 8046.
whereas in the narrower northern and southern Andes the

plate shallows and becomes nearly flat (Cahill and Isacks,
1992, and references herein). Gephart (1994) showed that
the central Andean orogen as well as the underlying
WadatiBenioff zone exhibit a high symmetry with the axis of
symmetry for the two surfaces nearly coincident. More,
Gephart (1994) showed that these axes were coincident with
the Nazca/South America finite relative rotation pole for the
period 3620 Ma, that is, coincident with plate kinematics,
which has been stable since the mid-Tertiary. Such geometric
and kinematics evidence suggest that the tectonic forces that
built the Andean topography are strongly related to the subduction process, or plate tectonics, in this noncollisional
orogen (Gephart, 1994) whatever the significant along-strike
geological variations.
Crustal thicknesses in the Andes vary strongly on the region
considered, and, also depending on the kind of methods used
to estimate it (see Allmendinger et al., 1997). From north to
south, seismic studies show that they range from 59 to 70 km
in southern Peru (Figure 8), from 4960 to 80 km beneath
the Central Andes and from 4050 to 6070 km above the
Pampean flat slab of the Southern Andes (Beck et al., 1996;
Fromm et al., 2004; Heit et al., 2008; McGlashan et al., 2008;
Yuan et al., 2002). Crust maximum thickness is beneath the
Western and Eastern Cordilleras where it reaches B70 km
(Beck and Zandt, 2002; Beck et al., 1996; Swenson et al.,
2000). Now considering the AltiplanoPuna plateau alone,
depending on different studies, crustal thicknesses vary beneath the PunaAltiplano from 60 to 65 km (Beck and Zandt,
2002; Beck et al., 1996) or they range between 57 and 82 km
beneath the Altiplano, and they thin to 4045 km beneath the
1000 m higher Puna (McGlashan et al., 2008; Yuan et al.,
2002). Alternatively, the crust under the Altiplano and Puna
plateau has been deduced to be 60 and 67 km thick, respectively, in correlation with the higher 1000 m asl mean elevation
of the Puna (Baumont et al., 2002). Gravity studies also yield
the conclusion that the crust beneath the Altiplano and Puna
is less than 70-km thick (Fukao et al., 1989; Gotze et al., 1994;
Kono et al., 1989). Beneath the Altiplano, a low-velocity zone
has been inferred for the midcrust to lower crust to be felsic
(quartz-rich) composition and it may correspond to a weak
layer decoupled from the brittle upper crust (Beck and Zandt,
2002; Koulakov et al., 2006; Yuan et al., 2000). Mafic lower
crust is inferred by seismic velocities to be in the eclogite
facies or to be absent in the Central Andes (Beck and Zandt,

2002).
Lithospheric thickness has been estimated to reach 150 km
beneath the Altiplano, to thin under the Western Cordillera
and to thin by 50 km beneath the Puna plateau (Whitman
et al., 1992; Whitman et al., 1996). Seismic velocities under
the AltiplanoEastern Cordillera transition also suggest that
the asthenosphere reaches near the base of the crust (Dorbath
and Granet, 1996; Koulakov et al., 2006; Myers et al., 1998;
Wigger et al., 1994). Geochemistry of hydrothermal fluids and
gases of the Altiplano and the Eastern Cordillera also supports
the views that the lithosphere has been thinned (Hoke et al.,
1994). The upper part of the subducted Nazca plate is topped
by a low-velocity zone interpretated as magma generation
zone induced by dehydratation of subducted material dragged
along the slab (Baumont et al., 2002). To the East of the Altiplano, numerous studies suggest that the mechanically
strong Brazilian lithosphere has underthrust the Sub Andes
and the Eastern Cordillera but not the Altiplano (Beck and
Zandt, 2002, and references herein).
5.6.1.4.3
Intraplate plateaus (Colorado and African

plateaus)
5.6.1.4.3.1 The Colorado Plateau

A Mesozoic oceancontinent collision started at 150 Ma in the
western coast of North America. Consequently, the east dipping Farallon plate subduction, and the associated compression triggered compressional deformation and the
formation of the Sevier orogen. During the late Cretaceous
and early Tertiary, the shallowing of the Farralon plate subduction (e.g., Dickinson and Snyder, 1978) propagated the
deformation into the interior of the North American plate
leading to the formation of the Laramide orogen. These orogenies lead to an overthickened crust up to 60 km thick
(Coney and Harms, 1984). The Colorado Plateau was located
close to the eastern front of the Laramide orogen but it experienced only a very small amount of crustal shortening, less
than a few percent (Spencer, 1996). The Colorado Plateau is
now bordered by the Basin and Range Province to the west
and south and the Rio Grande Rift to the east; both are
thinned regions (e.g., Bird, 1988; Sonder and Jones, 1999;
Wernicke et al., 1987) where extension is still active. However,
no significant extension is reported on the Colorado Plateau.
103
Depth (km)
Elevation (km)
Edge of the Brazilian

lithosphere
Local compensation
weak lower crust
Lithospheric flexure
strong lower crust
6
4
2
18
Westren
cordillera
Altiplano
Eastern cordi
100
NS fast
EW fast
Brazilian litho
zc
as
lab
300
69
EW fast
20
Chaco
sphere
Na
200
SubAndes
llera
68
Removal of
lithosphere
67
66
65
64
63
62
61
60 W
Change in shear wave

splitting direction
Figure 8 Schematic cross-section from Beck, S.L., Zandt, G., 2002. The nature of orogenic crust in the central Andes. Journal of Geophysical
Research 107, 2230 showing their interpretation of the lithospheric structure of the central Andes from geophysical and geological studies. Red
and blue indicate upper mantle P wave velocities that are slower and faster, respectively, than the reference IASPEI-91 model. Reproduced from
Myers, S.C., Beck, S., Zandt, G., Wallace, T., 1998. Lithospheric-scale structure across the Bolivian Andes from tomographic images of velocity
and attenuation for P and S wavesJournal of Geophysical Research 103, 2123321252.
Colorado
plateau
Present-Day
Great Basin
Rio grande
rift
Great plains
SE
NW
0 km
100
km
?
100 km
200
km
Sinking plate or
lithosphere
Crust (topography exaggerated)

Paleoproterozoic mantle lithosphere
Altered Paleoproterozoic mantle lithosphere
Altered Phanerozoic or juvenile mantle
Subducted plate or delaminated lithosphere
Figure 9 Deep structure beneath the Colorado Plateau and the Basin and Ranges, Rio Grande Rift and Great Plains neighboring regions, from
Sine, C.R., Wilson, D., Gao, W., Grand, S.P., Aster, R., Ni, J., Baldridge, W.S., 2008. Mantle structure beneath the western edge of the Colorado
Plateau. Geophysical Research Letters 35, L10303.
The crust beneath the Colorado Plateau is inferred

by seismic studies to be 44- to 47-km-thick (Figure 9)
(Das and Nolet, 1998; Wilson et al., 2005). Crustal thickness
is almost similar, 4246 km, in the Great Plains, whereas the
crust underwent thinning to 35 km centered on the Rio
Grande Rift (Wilson et al., 2005). The Basin and Range
Province underwent thinning of the crust up to 2030 km
thick (Coney and Harms, 1984), seismic studies suggest that
the crust thins from 30 to 25 km eastward (Das and Nolet,

1998, and references therein). The crust of the Rocky Mountains to the north is inferred to be 50-km-thick (Das and
Nolet, 1998, and references therein).
Seismic data argue for a 120- to 150-km-thick lithosphere
beneath the Colorado Plateau and B200 km beneath the
Great Plains to the east (West et al., 2004). In between,
thinned lithosphere, 45- to 55-km-thick, is inferred to exist
104
beneath the Rio Grande Rift (West et al., 2004). To the west,
the lithosphere is also thin at 60-km-thick beneath the Basin
and Range (Zandt et al., 1995).
A thick low-velocity zone is inferred beneath the Colorado
Plateau lithosphere (West et al., 2004) from a depth of 150 to
300 km. Based on seismic studies Wang et al. (2008) and Sine
et al. (2008) inferred mantle regional flow at the southeastern
edge and at the western edge of the Colorado Plateau.
5.6.1.4.3.2 The East African (Kenyan) and Ethiopian plateaus
The East African and Ethiopian plateaus are located in the
African plate along the East African Rift System. The Ethiopian
plateau is covered by extensive basalts dated at 3129 Ma
(e.g., Hofmann et al., 1997). Sedimentary and volcanic sequences are modeled to reach 2 to 5 km of thickness on the
Ethiopian plateau (Mackenzie et al., 2005) with 1 to 2 km of
flood basalts and volcanoes on top of the western Ethiopian
plateau (Pik et al., 2003). Depending on the studies carried out,
the crust is inferred to be 42- to 50-km-thick west of the Ethiopian rift and 37- to 40-km-thick to the east (Mackenzie et al.,
2005) or is inferred to be 33- to 44-km-thick on either side of
the Main Ethiopian rift (Dugda et al., 2005). Beneath the
western Ethiopian plateau, a 15-km-thick layer composing the
base of the lower crust is characterized by seismic high-velocities. Crustal thicknesses vary from 37 to 38 km west and
from 39 and 42 km east of the Kenyan rift (Dugda et al., 2005).
These studies indicate that crust away from the rifts has not
been significantly modified by the Cenozoic rifting and magmatism, and the Cenozoic flood basalts on the Ethiopian
plateau do not appear to be thick enough to alter the depth of
the base of the crust (Dugda et al., 2005). Only a slight thinning
of the crust is observed below the Main Ethiopian Rift (Mackenzie et al., 2005) and crustal thickness is inferred to vary from
27 to 38 km (Dugda et al., 2005).
Beneath the Ethiopian plateau, the lithosphere-asthenosphere boundary is imaged at 90 km depth (Ayele et al.,
2004). Almost similar values are obtained by Dugda et al.
(2007) who inferred the base of the lithosphere at 7080 km
beneath the Ethiopian plateau and at 50 km beneath the Main
Ethiopian Rift and the Afar. By contrast, the lithosphere of the
Tanzanian craton beneath the East African plateau reaches
deeper levels up to approximately 170200 km (Ritsema and
van Heijst, 2000) (Figure 10(b)), though 5080 km shallower
than the lithosphere of the thicker Congo craton to the west.
In the Afar and surrounding regions a deep low-velocity
anomaly in the upper mantle (Debayle et al., 2001; Ritsema
et al., 1999; Ritsema and van Heijst, 2000) is interpreted as
hot mantle extending to at least 660-km-depth (e.g., Debayle
et al., 2001). Beneath the East African plateau a low-velocity
anomaly is present in the upper mantle that is B500 km wide
and extends from the eastern edge of the Main Ethiopian Rift
westward (Bastow et al., 2008). A low velocity zone also
extends up to a depth of B400 km and to at least 300 km
beneath the eastern branch in Tanzania and in Kenya, respectively (Park and Nyblade, 2006; Ritsema et al., 1998;
Weeraratne et al., 2003). The top and bottom of the transition
zone (410660 km) beneath the Eastern Plateau is depressed
and it could be related to a thermal anomaly in the
lower mantle (Figure 10(a)) (Huerta et al., 2009) (see
Section 5.6.1.4.3.3).
Following the pioneer work of Dziewonski (1984),

Ritsema et al. (1999) presented a global tomographic model
inferring (like Dziewonski) a broad low-velocity zone in the
lower mantle and initiating at the core mantle boundary beneath southern Africa and propagating up to the upper mantle
beneath the Afar Plateau (Figure 10(a)). Such anomalous low
seismic velocity is interpreted as hot and less dense lower
mantle material combined with a chemical anomaly (Ritsema
et al., 1999; Simmons et al., 2007) and referred to as the deepmantle African Superplume.
5.6.1.4.3.3 The Southern African plateau
The southern African plateau is located within a stable continent and bordered by Mesozoic passive margins (Figure 1).
The southern African plateau is bounded on the coast by rift
shoulders, the Great Escarpment and the Drakensburg
Mountains, characterized by a warping down of the topography from 1000 m asl toward the margin (Ollier, 1985;
Ollier and Marker, 1985; ten Brink and Stern, 1992).
Crustal thickness beneath the southern African plateau
varies between 33 and 45 km beneath the Archean cratons and
45 and 50 km beneath the Proterozoic orogenic belts (Nguuri
et al., 2001).
A 250 to 300 km thick lithosphere has been inferred by
Ritsema and van Heijst (2000) and James and Fouch (2002),
respectively. High lithospheric thickness and low heat flow,
o45 mW m2 (Nyblade et al., 1990), are considered to mark
a strong upper mantle which is corroborated by high estimates
of effective lithospheric thicknesses (Stark et al., 2003).
Beneath southern Africa and the southern Atlantic Ocean,
global tomography suggests that there is a low-velocity zone in
the lower mantle and initiating at the core mantle boundary
(Ritsema et al., 1999) (Figure 10(a)).
5.6.1.5
High Plateaus: Uplifted Peneplains, Growing

Plateau or Applanation at High Elevation?
Davis (1899) interpreted elevated and smooth landscape as a

step in a sequence or a cycle in which ongoing erosion of a
previous landscape eventually leads to the complete destruction of the relief and the formation of a low-relief erosional surface at almost sea level: the peneplain. Daviss cycle
of erosion (1899) was based on the idea that after a rapid
surface uplift of the peneplain, the increase in potential energy
enhances erosion that in turn increases significantly the local
relief and local slopes. A wave of fluvial and hillslope erosion
then propagates inside the plateau from the borders. Ongoing
erosion eventually leads to the complete destruction of
the topography and the formation of another lowrelief erosional surface at almost sea level. A new uplift of
such peneplain means a wave of erosion starts afresh: a
new cycle of erosion began (see Chapter 5.16 for more
discussion).
In contrast, several models have been proposed for the non
instantaneous development of topography as it has been the
case for the Tibetan Plateau. They argue for the progressive
surface uplift of the whole plateau (Fielding et al., 1994),
or for progressive surface uplift of the whole followed by
rapid surface uplift events (England and Houseman, 1989;
fr. P. East
South A
Afr.
105
P.
Af
ar
P.
670
km
00
29
Ma
ntl
e
re
Co
(a)
km
East African plateau

(~4.5 S)
Westren
branch
Eastern
Labait
branch
volcano
Kibaran belt
Tanzania craton
Mozambique
belt
Moho
3000 m
S.L
Moho
?
Mantle
lithosphere
?
?
100 km
Thermally modified
mantle lithosphere
200 km
300 km
-olivine to -phase
*410* discontinuity
400 km
500 km
600 km
Majorite-out
discontinuity
Heat
flux
28
(b)
30
32
34
36
Longitude E
700 km
38
40
Figure 10 (a) Cross-sections through the South and eastern African plateaus, and across the Afar plateau, obtained from a tomographic model
extending up to the core-mantle boundary along a 1401 wide great circle arcs (inset). Relatively high velocity and low velocity regions are indicated by
blue and red colors, respectively. The dashed line represents the 670-km seismic discontinuity. Reproduced from Ritsema, J., Heijst, H.J.V.,
Woodhouse, J.H., 1999. Complex shear wave velocity structure imaged beneath Africa and Iceland. Science 286, 19251928. (b) Schematic crosssection of the East African plateau at B4.51 S showing receiver function stacks of mantle transition zone and cartoon of the associated thermal
upwelling from Huerta, A.D., Nyblade, A.A., Reusch, A.M., 2009. Mantle transition zone structure beneath Kenya and Tanzania: more evidence for a
deep-seated thermal upwelling in the mantle. Geophysical Journal International 177, 12491255, which is an updated version from Nyblade, A.A.,
Owens, T.J., Gurrola, H., Ritsema, J., Langston, C.A., 2000. Seismic evidence for a deep upper mantle thermal anomaly beneath east Africa. Geology
28, 599602.
106
Harrison et al., 1992; Molnar et al., 1993) or for the successive

surface uplift of external parts of the plateau incorporated into
the plateau to produce its present shape (Tapponnier et al.,
2001).
Plateaus in compressional settings, such as Tibet and the
Andes, are associated with erosion, transport and sedimentation in now highly elevated endorheic sedimentary
basins. Forward propagation of the deformation has been
proposed as a mechanism for the progressive incorporation
of endorheic basins in Tibet (Tapponnier et al., 2001) and in
the AltiplanoPuna plateau (Sobel et al., 2003; Strecker et al.,
2007). Based on the study of the Pyrenean landscape evolution and experimental modeling of surface processes interaction between rock uplift, erosion, and sedimentation it has
been shown that the accumulation of sediments at the foot of
mountain ranges produces a rise of the apex of sedimentation,
that is, the effective base level for erosion, which in turn inhibits the erosion at high elevation (Babault et al., 2005a, b,
2007). Such processes, combined with the progressive incorporation of internally drained basins, allowed the formation of a low-relief topography in the Tibetan Plateau and
in the AltiplanoPuna plateau during and after their rise
(Babault et al., 2005a; Liu-Zeng et al., 2008).
5.6.1.6
On the Existence of Past High Plateaus in the

Earth History
The demonstration of the existence of past high plateaus requires determining both the paleoelevation and the flatness of
large continental areas. This is less easy for past high plateaus
that have, by definition, disappeared. Paradoxically, the way
they have been destroyed suggests their past existence. Indeed,
currently very large areas of continental extension such as the
Basin and Range Province are viewed as resulting from the
collapse of overthickened crustal welts resembling high
plateaus (Coney and Harms, 1984). The intense bimodal
magmatism that accompanies extension in such areas is
considered as indicative of the partial melting of the thermally
relaxed crust and the underlying upper mantle (e.g., Buck,
1991; Dewey and Burke, 1973; Sonder and England, 1989),
attesting to crustal thickening and therefore past high
elevation of Earths surface in those areas. But the main
reason why this problem is so difficult to resolve relates
to the problems of deciphering the paleoelevation of
continents.
Paleoelevation is commonly deduced from reconstruction
of crustal thickness assuming isostatic compensation, or from
paleotemperature estimates through paleoenvironment and
stable isotopes studies (see Section 5.6.2). Methods to determine paleoatmospheric pressures allowing the determination
of paleoelevation, such as the use of basalt vesicularity are rare
(e.g., Sahagian and Proussevitch, 2007; Sahagian et al., 2002a,
see Section 5.6.2.5 of this chapter). Whatever the methods
used, the results are subjected to debate, mostly because of
the great margins of error associated with the methods used in
providing paleoelevation estimates.
Coney and Harms (1984) estimated the amount of Tertiary
extension in the Basin and Range to be between 40 and 75%.
They concluded that the initial crustal thickness could have
reached 60 km over a several hundred kilometers width and

they suggested that, during the Late MesozoicEarly Cenozoic
times, this area resembled the present-day Tibetan Plateau.
However, Coney and Harms (1984) did not provide any information regarding the flatness of this area so that they could
not definitively argue the existence of a high plateau. Best et al.
(2009) used the compositions of contemporaneous calc-alkaline lava flows as well as the configurations of the ignimbrite
sheets to show that the Basin and Range Province during the
middle Cenozoic was a relatively smooth plateau underlain by
unusually thick crust. In particular, Best et al. (2009) observed
that the generally small between-site variations in the
palaeomagnetic directions of individual sheets lend further
support for a relatively smooth landscape over which the
sheets were draped and they conclude that during the middle
Cenozoic this area was a relatively flat plateau.
Wolfe et al. (1998) compared the physiognomy of leaves
from modern vegetation of known climates with that of Eocene and Oligocene fossil leaf assemblages from middle latitudes of western North America and they concluded that most
of the western US and southern Canada was higher in elevation in early Cenozoic time than today. Moreover, where
normal faulting has been intense, paleoelevations commonly
suggest a subsidence of the region.
Menard and Molnar (1988) have suggested that the Variscan belt of Europe was a Hercynian Tibetan plateau that
collapsed into a late Paleozoic European Basin and Range
Province. Although most studies on the Variscan belt have
concluded that the crust was significantly thickened during the
Late Paleozoic, Menard and Molnar (1988) observed that
Late Paleozoic sedimentary basins are presently juxtaposed
with metamorphic basement rocks. Erosion cannot therefore
explain the present-day crust thickness of 30 km. As no pervasive extensional tectonics occurred during the Mesozoic and
the Cenozoic, Menard and Molnar (1988) concluded that the
thickened crust returned to normal thickness by collapse
during the Late Paleozoic, resulting in the exhumation of the
deep crust and the concomitant development of sedimentary
basins at the surface in a similar way to that of the North
American Cordillera, in the western US, during the Cenozoic.
Becq-Giraudon and Van Den Driessche (1994) and BecqGiraudon et al. (1996) have tentatively argued the paleoaltitude of this Variscan belt and its morphology resembled that
of a high plateau by looking at the characteristics of the late
Carboniferous-early Permian continental sedimentation. In
addition, Becq-Giraudon et al. (1996) remarked that the
presence of many sedimentary futures such as stratified slope
deposits, diamictites, water escape craters and ghosts of ice
crystals are related to periglacial effects. The clay mineral assemblage, dominated by the association illite-chlorite-interlayed clays and the absence of significant weathering of
feldspathic clasts and granitic pebbles favor the prevalence of a
cold to temperate climate during sedimentation. The flora and
entomofauna also indicate a seasonally varying climate with
cold winters. Furthermore, the extreme homogeneity of the
flora composition from one basin to another indicate a lack of
appreciable altitude zoning and a plateau-like morphology
(Becq-Giraudon et al., 1996). In view of the paleoequatorial
position of the Variscan belt during the late Caboniferousearly Permian, Becq-Giraudon et al. (1996) concluded that the
morphology of the chain was that of a near 4500 m asl high

plateau.
5.6.2
Evidence for Plateau Uplift, Regional

Warping, and Subsidence
5.6.2.1
5.6.2.1.1
Geomorphic Markers
Low-relief, high-elevation erosional surfaces
A striking feature of high plateaus margins is the occurrence of

deeply dissected, almost flat topographic remnants at high
elevation. As mentioned above (Section 5.6.1.5) and following Daviss cycle of erosion (1899), a smooth landscape
generates from the destruction of a previous smooth landscape uplifted and then progressively more dissected and
characterized by steeper slopes. This is followed by successive
stages, with a decrease in slope gradients associated with the
decrease of the elevations near to sea level. Based on these
views, rock uplift is usually deduced from the elevation of
more or less localized changes in the local slopes that separate
a subdued relief above, also called a relict landscape (for a
summary see Widdowson, 1997), from a steeper one below
(e.g., Clark et al., 2006; Epis and Chapin, 1975; Gregory and
Chase, 1994; Kennan et al., 1997).
Following Powell (1875), a stream cannot erode below its
base level, which is the lower limit in the landscape and is
ultimately represented by sea level. If the process of formation
of low-relief erosional surfaces follows Davis model, the key
point to infer rock uplift from highly elevated remnants of
low-relief erosional surfaces lies in the assumption that they
formed close to sea level. Such assumption is however almost
never demonstrated in orogens and plateaus because of the
destructive nature of erosion or the lack of marine sedimentation on the low-relief erosional surfaces. Consequently,
where it cannot be demonstrated that low-relief erosional
surfaces developed close to sea level, it cannot be ruled out
that a smooth landscape may have developed well above
sea level. Within continents, if an erosional smooth landscape
is 1000 km away from its base level and that the formational
regional slope is 0.00571 (104) or 0.0571 (103) then a lowrelief erosional surface can develop 100 to 1000 m above its
base level.
An alternative model of planation is that of parallel scarp
retreat by backwearing of slopes (King, 1953). As that of
Davis, in Kings model a discontinuous surface uplift is responsible for the high elevation of a smooth landscape that
starts to develop near the sea level. The plateau successively
expands just above sea level by river incision, scarp retreat,
and the formation of pediments at its base. Ongoing erosion
creates pediments that coalesce to form a pediplain. A next
surface uplift will promote a new planation surface near to sea
level. However, ongoing lateral erosion of the first pediplain
occurs above sea level and is disconnected from the younger
pediplain by an escarpment. Pediments are broad, concave-up
and sloping at an angle of 671 at the base of the retreating
escarpment, whereas pediplains are inclined gently (commonly thought to be less than 11 although never demonstrated) toward the sea level. As a consequence, low-relief
erosional surfaces developed in this way form near to sea level
107
only close to the coast line. Such considerations imply that

rock uplift cannot be deduced from their elevation alone, a
correcting factor must take into account for the regional slope
of erosion down to the coast.
Drainages of mountains and plateaus are not always open
to the sea. Where streams flow into closed basins (endorheic
basins), their base level is considered to be the basin floor or
the terminal lake. In the Andes or in Tibet, for example, highly
elevated endorheic basins cover extensive areas setting the base
level for their streams at high elevation. But long-term landscape evolution of mountain belts argues for the apex of
sedimentation in foreland basins to be the effective base level
for erosion below which no erosion occurs on the long-term
(Babault et al., 2005a, b; Carretier and Lucazeau, 2005), instead of a coastline or a lakes shoreline where drainages are
opened or closed, respectively. Babault et al. (2005a, b) and
Carretier and Lucazeau (2005) considered that it is not so
much the ultimate base level of stream waters but the transport
capacity of rivers, in controlling the long-term limit between
the erosion and deposition, that sets the effective base level for
erosion. In such a case, slopes and concavity of the rivers, of the
alluvial and of the fluvial fans, control how high the elevation
of the apex of sedimentation stands above the coastline or the
terminal lake. Gradually over time, the aggradation of the
products of erosion raises the base level when a sedimentary
basin is overfilled, whether its nature is endorheic or not
(Babault et al., 2005a). However, the apex of sedimentation,
the effective base level in an endorheic basin can be expected to
be always higher than that of a basin opened to the sea.
Davis cycle of erosion in which a landscape is smoothed
close to sea level can therefore be valid at high elevation
whether drainage is endorheic or not, at least but not necessarily provided that tectonic uplift has ended (Babault et al.,
2007). The same is true for the formation of stepped pediments in continental and internally drained areas like those
of the Iberian range in Spain (e.g., Casas-Sainz and CortesGracia, 2002). The corollary of this conclusion is that where
remnants of low-relief erosional surfaces are bounded by
overfilled sedimentary basins, smoothing might have occurred
well above sea level. In such geological settings, rock uplift
cannot be inferred from the elevation of remnants of lowrelief erosional surfaces alone (Babault et al., 2005b, 2007).
The paleoelevation of the apex of sedimentation as well as the
distance from it and the fluvial paleogradient must be taken
into account to infer rock uplift from smooth relict landscapes. Another requisite to infer rock uplift from relict landscapes is the slow long-term erosion rates of the remnant
surfaces (e.g., Clark et al., 2006). High erosion rates of the
remnant surfaces would lead to an underestimation of the
rock uplift.
It is important to note that the above considerations of
landscape evolution allow the inferring of rock uplift, but in
most cases not surface uplift. When a high-elevation, lowrelief landscape is incised by the modern fluvial system, surface uplift is always lower than rock uplift (Molnar and England, 1990). Another important point to underline is that the
erosionally driven unloading of the crust induces a flexural
isostatic rebound at the regional scale. Consequently, a
fraction of the measured rock uplift has to be attributed to
the isostatic response of the missing masses, whereas the
108
remaining fraction has to be linked to one or more deep

processes in the crust or in the mantle (e.g., Abbott et al.,
1997; Leonard, 2002; Roy et al., 2009).
In the Andes, B2000 m of rock uplift have been deduced
from the study of channel longitudinal profiles and in particular the reconstruction of past longitudinal profiles in the
eastern Cordillera of Bolivia based on low-relief paleosurface
remnants (Kennan et al., 1997). As defined above, uplift of
low-relief paleosurface remnants, now incised by the current
drainage network, is rock uplift and not surface uplift. In their
topographic reconstruction at B10 Ma, Kennan et al. (1997)
obtained a downstream profile ranging from 500 to mostly
1500 m in elevation. Such results argue for the low-relief
erosional surfaces of Bolivia to have developed above 500 to
1500 m depending on their distance from the coastline, B25
to 325 km, respectively. Barke and Lamb (2006) reconstructed
the paleoprofiles of the Eastern Cordillera using values of
concavity and steepness (see Section 5.6.2.1.4) and they obtained a more accurate estimate of rock uplift at 17057695 m.
In the eastern Tibetan Plateau, Clark et al. (2006) deduced
a minimum of 3000 to 4000 m surface uplift of the highest
parts of the southeastern Plateau from the elevation of remnants of a high elevation, low-relief surface. Clark et al. (2006)
based their reasoning on the low values of erosion
(0.010.02 mm yr1) of remnant surfaces to argue that rock
uplift can be deduced from relict landscape elevations,
assuming a paleoslope to the east ranging between 0.00571
(104) and 0.0571 (103).
5.6.2.1.2
Drainage network development and

reorganization on a plateau
Although drainage development and reorganization might be

complex in some cases it can be used to unravel the tectonic
evolution of a region. Uplift influences drainages via baselevel
changes, drainage reversals, and capture processes. The dictum
Drainage systems have a heritage rather than an origin (in
Summerfield, 1991, Chapter 16) summarizes well the constant
interaction between tectonics, erosion, and deposition. This
view is opposed to an old and unrealistic one that considers
that a drainage network originally grows on an initial slope
before channels adapt their course to structure. Although
drainage does not generally start to develop on a flat, gently
sloping topography, overfilling of landscapes by lava flows or
sediments can lead to a smoothing of the local relief and to
the resetting of a previous channel network. The incorporation
in an orogen of a drainage network newly formed in an
overfilled foreland basin could, however, explain the similarity
of network geometry in orogens (Castelltort and Simpson,
2006).
Where the local relief is low, a trunk of a gently sloping
river can be reversed by only a slight regional tilt of the
topography. This occurs only if the rate of incision downstream of that trunk cannot balance the rate of rock uplift.
Plumes generate dome-like plateaus and consequently they
can trigger a modification of the drainage network associated
with drainage reversals and captures that eventually lead to a
radial drainage (Cox, 1989). Later, such drainage can be more
or less captured by a structurally controlled drainage that develops in a rift valley superimposed on the domal plateau like
that of the East African plateau (Cox, 1989; Moore and
Blenkinsop, 2002). The result of the channel network reorganization is a barbed drainage pattern with junction angles
greater than 901, see Figures 12.7 and 16.9 in Ollier (1981)
and in Summerfield (1991), respectively.
Rock uplifts of plateaus induce a gain in potential energy
that enhances the stream power (e.g., Whipple and Tucker,
1999) of the surrounding channels. Consequently, internally
drained areas with higher local base level are expected to be
captured by external transverse drainages with lower base level
and located on rims of plateaus. If so, stream piracies can
provide evidence of plateau uplift. Captures can be identified
by a series of drainage patterns and deposits that indicate
poorlyintegrated drainage. These include the occurrence of
hanging valleys with wind gaps in the drainage divide, and
wind gap sediments (if preserved) in the headwaters of a beheaded river, elbows of capture, and a river paleoflow direction
different than present flow direction (see Douglass et al., 2009,
for a compilation of the various criteria). Drainage reorganization by captures have been used to indicate surface uplift in
the southeastern Tibetan Plateau margin (Clark et al., 2004) and
in the Anatolian plateau (Nicoll, 2010).
5.6.2.1.3
River longitudinal profiles: Steepness indices
Rock uplift increases the potential energy and triggers an increase in the stream power (e.g., Whipple and Tucker, 1999).
Consequently, the increase in incision modifies the streams
morphologies, first at the rims of a plateau and eventually
inside the plateau. The downstream slopes of rivers increase
and in most cases knickpoints develop (Whipple and Tucker,
1999). Local slopes are not the only variable in bed morphology that rock uplift can potentially modify; channel width,
(e.g., Whittaker et al., 2007) and sinuosity (among others) can
also be affected and in turn they can change the relationship
between local slopes and rock uplift (See Wobus et al., 2006,
for a review). However, a relation between rock uplift and
downstream slopes has been highlighted in many different
natural landscapes. In the following the authors summarize
the methodology developed to derive rock uplift from digital
elevation models, which is based on the empirical power-law
that relates the downstream slope in a river bed to the contributing drainage area (Flint, 1974; Howard and Kerby,
1983):
S ks Ay
where S is the local downstream slope, A is the upstream
drainage area, y is the concavity index and ks is the steepness
index. Studies in California (Snyder et al., 2000) and in the
Siwalik Hills in front of the Himalaya (Wobus et al., 2006)
demonstrated a positive correlation between the steepness
indices of a set of channels with comparable morphologies
(normalized to a common concavity index), and spatially
uniform rock uplift rates encompassing each drainage basin.
In these studies, steady-state conditions are assumed, that is,
rock uplift is everywhere balanced by erosion. Inference of
rock uplift rates using steepness indices derived from digital
elevation data requires first a calibration area with known
uplift rates and similar channel morphologies. If not, the
spatial variability of steepness indices provides at least a
high spatial resolution estimate of the rock uplift pattern
(Hodges et al., 2004; Wobus et al., 2003). Special care must be

taken in the interpretation of steepness indices as a proxy for rock
uplift rates. Rock uplift rates not only modify the local slopes
leading to a simple relation between steepness indices and uplift
rates. Steepness index depends also on other factors (e.g.,
Whipple, 2004; Whipple and Tucker, 1999) including rock
strength (erosivity), sediment flux (tools for erosion or cover
of the bedrock that inhibits erosion), channel width, drainage
basin hydrology, river incision process(es) (plucking, abrasion,
and weathering) and critical shear stress for incision. For
example, where the pattern of the channel width, W, with
respect to the cumulative drainage area, A, varies downstream
in a different fashion from that commonly assumed,
W 0.35A0.5 (Montgomery and Gran, 2001), steepness indices
may not reflect rock uplift rates. Fluvial erosion is commonly
understood to scale with bed shear stress and water flow velocity.
Both variables depend on the local slope and on the channel
cross-sectional geometry. Consequently, local channel width and
water depth modifications can strongly alter the relationship
between steepness indices and rock uplift rates (e.g., Whittaker
et al., 2007). That some of the parameters listed above might be
expected to vary with tectonic rates and consequently alter the
significance of the steepness index, it is not the only limitation
to use it as a proxy for rock uplift rates. It is important to note
that some of the parameters listed above vary with lithology
and climate. However, local significant variations of steepness
indices values are thought to reflect changes in rock uplift rates,
where the spatial variations of the factors listed above are
negligible (Wobus et al., 2003).
In the eastern Tibetan Plateau, this method has been used
to infer the distribution of active rock uplift (Kirby et al.,
2003). However, since the obtained steepness indices are assumed to relate to the rates of rock uplift, no total amount of
rock uplift can be derived from this method alone until the
duration of the considered tectonic event is known.
5.6.2.1.4
Longitudinal paleoprofile reconstruction of

rivers
The total amount of rock uplift can be deduced from the reconstruction of a rivers paleoprofile. Two methods can be
used based on the longitudinal profile equation and paleomarkers of the river longitudinal profile such as river terraces.
The first method relies on the study of a series of rivers
characterized by segments bounded by a knickpoint. First, it is
assumed that an increase in fluvial incision in response with a
sudden base level fall, or an increase in the rate of a base level
fall (conceptually equivalent to an increase in rock uplift),
propagates upward along a channel and is associated with a
knickpoint that separates an unmodified upstream part from a
downstream segment that adapts its slopes to the new
boundary conditions imposed by the new rock uplift rate (e.g.,
Whipple and Tucker, 1999). Second, it is assumed that the
upstream part corresponds to a river profile that is adjusting or
adjusted to a prior state of base level fall. Third, incision rates
in the upstream part above the knickpoint must be slow, a
condition matched under dominated semiarid to hyperarid
climate conditions as indicated by the preservation of transient landscape, such that profile lowering is negligible and the
inferred rock uplift is not underestimated. Under those
109
assumptions an empirical power-law gradientarea relationship can be used (Flint, 1974; Howard and Kerby, 1983) along
with the well-known relationship between downstream distance, x, and cumulative drainage area (Hack, 1957) to reconstruct a river paleoprofile from the equation of a
longitudinal river profile elevation, z(x), in two dimensions
(Whipple, 2001; Whipple and Tucker, 1999):
1 1hy
zx ks ky
x1hy
a 1 hy L
zL;
hya 1;
xc r x r L
where z is the elevation, xc is the distance from the division to

the channel head, L is the distance of outlet from source, ks is
the steepness index, y concavity index, ka is Hacks coefficient,
and h Hacks exponent. The river paleoprofile is obtained from
a fit of the longitudinal river profile elevation equation, z(x),
to the preserved paleoprofile in the upstream parts. The difference in elevation between the reconstructed profile set high
up and the modern downstream segment at the outlet elevation provides an estimate of the rock uplift even if it is not
spatially uniform (e.g., Hoke et al., 2007).
In the Andes, at least 1000 m of relative rock uplift of the
Western Cordillera and adjacent central Andean plateau with
respect to the Central Depression, located between the Pacific
coast and the Western Cordillera, has been deduced using this
method (Hoke et al., 2007). Based on perched paleoprofiles
near to the Pacific Coast, Hoke et al. (2007) also proposed that
the Central Depression could have undergone a rock uplift of
1000 m. If this interpretation can be confirmed by another
proxy, and if the Central Depression is not tectonically decoupled from the Altiplano, then B2000 m of rock uplift of the
Western Cordillera and adjacent central Andean plateau can be
deduced from the paleoprofiles reconstruction. Using a similar
approach, 20002700 m of rock uplift have been estimated for
the Eastern Cordillera at the eastern edge of the paleodrainage
systems that drained the Eastern Cordillera as indicated by the
paleosurface remnants and the associated fluvial deposits
(Barke and Lamb, 2006). Barke and Lamb (2006) also inferred
rock uplift from upstream propagation of the modern river
profile and its difference in elevation with respect to the
paleosurface remnants, and the heights above the foreland
basin of the knickpoints that separate the paleodrainage system
from the modern downstream profiles. From the combination
of all these measures they obtained a value of 17057695 m for
rock uplift for the eastern edge of the Eastern Cordillera.
In the eastern margin of the Tibetan Plateau, Schoenbohm
et al. (2004) calculated the Red River incised B1400 m in the
regional low-relief landscape from the paleoprofiles of a series
of perched tributaries above the Red River. Since the paleoprofile of the Red River is estimated at 1500 m and considering that
it has risen from 0100 m, sea level and present day elevation
respectively, they deduced a value of 14001500 m for surface
uplift in the Red River region (Yunnan Province of China).
The second method to derive an amount of rock uplift is
based on paleomarkers of the river profile such as river terraces. The strong assumption is that rock uplift is balanced by
river incision (e.g., Burbank et al., 1996). However, such assumption is difficult to demonstrate in most cases, and because river incision is not only a response to rock uplift, then
110
this method has to be validated by other data. For example,

climate can be responsible for incision in the piedmont
leading to a fall of the base level, which results in an upstream
propagation of incision in the mountainous area without any
change in rock uplift (e.g., Carretier and Lucazeau, 2005).
Upstream integration of drainage areas by river network
growth also leads to deep incisions in the downstream parts of
a longitudinal profile (e.g., Carretier et al., 2009; Vassallo
et al., 2007). In both cases paleoriver profiles will overestimate
rock uplift.
5.6.2.2
Paleoaltimetry from Sedimentology
To understand the continental tectonic processes responsible

for surface uplift, and to understand how the feedbacks
between such tectonic processes and climate and erosion
processes modify the surface uplift, the appropriate paleoaltimetric methods need to be used. As pointed out by Clark
(2007), a need exists to assess the paleo-mean elevation of
region, not only the local paleoelevation. This comes from the
fact that, although an increase in local elevations, local relief
or drainage basin relief can be associated to some rock uplift,
it does not necessarily indicate a surface uplift (an increase of
the mean elevation) (Molnar and England, 1990).
5.6.2.2.1
Paleoaltimetry from marine sediments
The easiest way to infer rock uplift is from the elevation of the
youngest shallow marine sediments preserved in the studied
region (including eustasy). The choice of mapping the
youngest strata instead of an older strata preserved everywhere
has the advantage to minimize the underestimation of rock
uplift when the upper limit of the marine unit is partially
eroded (e.g., Pederson et al., 2002). Any localized incision or
spatially homogenous erosion into these marine sediments
implies that a fraction of the rock uplift is to be attributed to
the erosionally driven isostatic rebound. As a consequence the
deduced rock uplift does not match the necessarily lower
surface uplift. An important constraint for any tectonic model
of plateau uplift can be deduced from the surface uplift rate.
However, the ratio between the elevation and the age of the
youngest shallow marine sediments only gives a minimum
estimate of the mean rock uplift rate since a time-lag can exist
between the depositional age and the subsequent rock uplift.
In the following the authors show two examples of surface
uplift deduced from marine strata.
In the Colorado Plateau, the Late Cretaceous landscape
reconstruction represents the last known time surface elevation. Based on the widespread distribution of the Late
Cretaceous shallow-marine and coastal sediments, Roy et al.
(2009) estimated the Colorado Plateau underwent B1.9 km
of rock uplift using data from the previous work of Pederson
et al. (2002) and taking into account that the Late Cretaceous
sea level was 200 to 250 m higher than today. Roy et al. (2009)
estimated B0.5 km of net erosion (erosion minus sedimentation) during the Cenozoic, taking into account the amount
of erosion of the Late Cretaceous marine sediments and the
amount of preserved younger continental deposits and volcanic rocks. To the rock uplift associated with erosionalunloading, Roy et al. (2009) added the flexural response of the
plateau to the tectonic unloading at the plateau margins where
extension took place during the Cenozoic (Basin and Range

and Rio Grande Rift). They obtained B0.50.6 km and
B1 km of rock uplift associated with crustal unloading (erosion and extension) at the center of the plateau and at the
edges, respectively. They deduced that 41.6 km of the remaining mean rock/surface uplift must be explained by tectonic processes that occurred during the Cenozoic. In
summary, widespread evidence of sea level or slightly below
sea level paleoelevations are very useful where preserved to
deduce rock and surface uplift.
The Andes were also at sea level during the Cretaceous such
that the surface uplift of the Altiplano must have been later
(e.g., Hoke and Garzione, 2008). In contrast, on the southern
Tibetan Plateau, Fielding (1996) deduced 5 km of surface
uplift from Cretaceous marine limestone deposited in the
Lhasa block (Hennig, 1915; Norin, 1946). As Fielding (1996)
noticed, it is difficult to generalize the paleoelevation of an
extensive region like the Tibetan Plateau from just a small part
of it, given that Tibet results from the accretion of several
crustal blocks and volcanic arcs associated with oceanic subduction under the southern margin of Asia before the collision. If the Andes are a contemporary analog of what Tibet
was like before the collision between India and Asia, then
mean elevations were probably well above the sea level, at
least in some part of it (Fielding, 1996). Consequently, the
5 km surface uplift since the Cretaceous (Fielding, 1996)
might be an overestimation when applied to the whole
of Tibet.
5.6.2.2.2
Paleoaltimetry from paleohorizontality of

lacustrine sediments
Some shallow lake deposits form in flat regions where they

can be used to deduce differential rock uplift. Changes in
elevation between deposits of the same age and of the same
facies that cannot be accounted for by any variation in depositional depth indicate relative rock uplift. South of
the Eastern Moroccan Atlas belts and plateau, the remnants of
the lower Pliocene shallow lacustrine sedimentation closer
to the Atlas lie at B1200 m asl and the more distant at
B600 m asl. The B600 m difference of elevation have been
inferred by Babault et al. (2008) as a postdepositional tilting
to the south indicating up to B600 m of relative rock uplift of
the southern margin of the Atlas since the early Pliocene. If the
lake deposits are assumed to have not undergone subsidence
and the elevation of the lower deposits is considered to have
remained the same, such estimate corresponds to minimum
rock uplift.
5.6.2.2.3
Paleoaltimetry from paleoslopes: Large-scale

patterns of deposition
Postdepositional changes in slope can also be determined

from fluvial sediments or erosional landforms.
Since slopes in braided rivers are generally less than 1.141
(Paola and Mohrig, 1996) then greater slopes can be used to
deduce a tilt of such depositional surfaces. Moreover, paleoslopes can be back calculated in order to quantify more accurately the amount of tilting. The method is based on the
assumption that the basal shear stress at bankfull stage has to
be higher than the critical shear stress required to transport the
coarsest clast (Paola and Mohrig, 1996). The paleoslope is
then derived from measuring the median clast size in channel

scours and the depth of such ancient channels of preserved
coarse-grained braided rivers. Errors from this method come
from the underestimation of the paleodepth resulting in an
overestimation of the paleoslope by 30 40%, and from the
underestimation of the surface grain size leading to an
underestimation of the paleoslope by 30 40%. Paola and
Mohrig (1996) compared estimated values of slopes from
their formula to the actual slopes of modern braided rivers
and they show that the error in the estimated slopes is about a
factor of 2.
McMillan et al. (2002) have shown that the paleoslopes to
the east of the Rocky Mountains piedmont, estimated using
the method of Paola and Mohrig (1996), ranged between
B0.0061 to B0.061 during the MiocenePliocene and that
they have been subsequently increased up to B0.61. The Great
Plains are connected to the Rocky Mountains that are themselves connected to the south to the Colorado Plateau and
defined the Rocky Mountain orogenic plateau (McMillan
et al., 2006). As a consequence, the regional rock uplift associated with the tilt to the east observed in the Great Plains over
250 km has then been used to infer 680 m (410815 m) of
regional rock uplift that affected the whole Rocky Mountain
orogenic plateau (McMillan et al., 2002).
The same concept has been roughly applied to the western
flank of the Western Cordillera in the Andes of Chile. Based on
the models of Stanistreet and McCarthy (1993), the depositional slope of the OligoceneMiocene braided fluvial fan
that mantle the Paleozoic basement of the Precordillera of
northern Chile, in the western flank of the Altiplano have been
assumed to be less than 11. Today these piedmont sediment
dips varies between 21 and 41 to the West and such westward
tilts have been used to infer a post-10 Ma rock uplift of the
Western Cordillera (Mortimer, 1973; Riquelme et al., 2003)
ranging between 500 and 1400 m (Faras et al., 2005).
5.6.2.2.4
Grain size distribution in piedmont

sedimentation
If it is assumed that an increase in rock uplift of the upstream

mountainous areas induce both a change in the sedimentary
flux that enters the basins from the rivers draining the rim of a
plateau and the slope of the streams, then a change in facies can
be expected from fined-grained sediments to clasts and boulders as well as higher sedimentation rates. However, erosion
and transport also depend on water discharge, which is subject
to changes throughout the history of plateau building in response to regional or global climate changes (e.g., Blackstone,
1975; Frostick and Reid, 1989). Consequently the relation between grain size and rock uplift is not straightforward (e.g.,
Molnar and England, 1990) and overall, the method does not
allow any quantification of either rock uplift or surface uplift.
A change in depositional facies in the northwestern Tibetan
Plateau and southern margin of the Tarim basin is interpreted
as a response to a main rock and surface uplift of the Kunlun
Mountains with respect to the Tarim basin (Zheng et al.,
2000). The shift from distal alluvial fan or flood-plain environments to proximal debris-flow accumulation is associated with an increase in sedimentation rate from an average
B0.15 mm yr1 to 1.4 mm yr1, as well as growth strata
(Zheng et al., 2000). Such association of increase in grain size,
111
as well as sedimentation rate can be interpreted as the evidence for the main rock/surface uplift of the northwestern
Tibetan Plateau (Zheng et al., 2000).
5.6.2.3
Paleoaltimetry Data Based on Paleobotany
Three methods that analyze plant physiognomy have been

proposed to infer paleoaltimetry. Two of them use floras (1) to
estimate temperature then divided by lapse rate (the rate of
temperature change with elevation) to give an elevation, and
(2) to estimate enthalpy that is combined with gravitational
acceleration to deduce an elevation. The third method uses
stomatal frequency in leaves to deduce changes in CO2 partial
pressure, which is a function of elevation.
Within the first method, paleotemperatures from fossil
floras, three fundamentally different approaches are used (see
Meyer, 2007, for a review). The first two are based on the
climatic distribution of nearest living relatives (or the floristic
method), and on the correlation of plant physiognomic
characters with climate. The third differs by the way the lapse
rates can be utilized to derive a paleoelevation. Most of the
current methods use the plant physiognomic method rather
than the nearest living relatives, which is believed to be
less accurate. Following the work of Bailey and Sinnott (1916)
that demonstrate that leaf margins correlate with climate
in modern floras, the plant morphological characters used in
the physiognomic method include leaf size and the presence
or absence of teeth. An extensive database called the ClimateLeaf Analysis Multivariate Program (Wolfe, 1993, 1995;
Wolfe and Spicer, 1999) relates many physiognomic characters
of leaves from modern forests to climate variables (e.g.,
untoothed leaf margins correlate to warm mean annual
temperature and large leaf size correlates to high levels of
moisture). The main results of the Climate-Leaf Analysis
Multivariate Program is that the mean annual temperature is
the most accurate climatic parameter estimated (1s error
r1 1C). Giving that temperature change varies with elevation,
continentality, and the effect of broad uplifted land surfaces,
three different lapse rates exist depending on the spatial scale
considered with values ranging between 3 1C km1 in the regional (between the coast and the continental interior of a
continent) lapse rate method (Wolfe, 1992a), 5.5 1C km1 in
the global mean terrestrial lapse rate (e.g., Axelrod, 1966), and
3.64 to 8.11 1C km1 in the local lapse rate method (e.g.,
Meyer, 1992). Almost the double of palaeoelevation estimate
can be derived from the regional instead of the global lapse
rate, highlighting the major impact lapse rate has on the
paleoelevation calculations.
Global climate changes through time imply that the mean
temperature at a given altitude does not remain constant. As a
consequence, paleotemperature estimates at a given locality
must be compared to paleotemperature estimates at sea level
for isochronous floras. The late Eocene Florissant flora of
central Colorado is a good candidate to estimate the error of
the different methodologies employed to deduce paleotemperature from fossil floras and paleoatlimetry from lapse
rates methods since it has been studied by many workers using
one or more of these methods (Meyer, 2007). Excluding
low elevation estimates (455 m) derived from the global
112
lapse rate method (Axelrod, 1998) the estimates of paleoelevation range between 19003200 m (Meyer, 1992) and
19002300 m (Gregory, 1994; Gregory and Chase, 1992;
Gregory and McIntosh, 1996) for the local lapse rate method
and values of 2900 m asl (Wolfe, 1992b) and 4133 m asl
(Wolfe, 1994) are obtained using the regional lapse rate
method. All these studies lead to a high to very high paleoelevation range of 300071100 m asl. In the second method,
paleoelevation from fossil leaves is obtained from the moist
static energy, h (J Kg1) (Forest, 2007; Forest et al., 1995,
1999). In the mean annual temperature method, present lapse
rates are transferred to past climate with no meteorologically
sound physical basis and such empirical lapse-rates with large
spatial variations (Meyer, 1992; Wolfe, 1992a) are used to infer
paleoelevation leading to estimates that can vary by a factor of
two. Unlike the mean annual temperature method, the aim of
the moist static energy method is to quantify a variable with a
theoretically and by observation, well-constrained behavior
with altitude and longitude. At mid-latitudes the mean flow of
the atmosphere is from west to east. Since moist static energy is
conserved along air masses it is also generally constant across a
region with longitude. Finally, moist static energy is constrained by thermodynamics such that
h c0p T Lv q gZ;
where c0p is the specific heat capacity at constant pressure of
moist air, T is the temperature (in K), Lv is the latent heat of
vaporization for water, q is the specific humidity, g is the
gravitational acceleration, Z is altitude, and
H c0p T Lv q;
where H is moist enthalpy.
Assuming that h is invariant within the same latitudinal
band, if we can estimate the enthalpy at sea level and at a
site of unknown elevation Z for a particular latitude, then
such elevation is given by (Forest, 2007; Forest et al., 1995,
1999)
Z
Hsealevel Hz
:
g
Both enthalpy estimates are derived from fossil leaf assemblage with the Climate-Leaf Analysis Multivariate Program. The total expected error in this method comes from two
sources of error (Forest, 2007; Forest et al., 1999): (1) from the
standard deviation from zonal invariance (the longitudinal
variability) of h (yielding a minimum estimate of the error in
altitude of 460 m); and (2) from the uncertainty in predicting
mean annual enthalpy from fossil leaf physiognomy (yielding
560 m of uncertainty in altitude). Forest et al. (2007, 1999)
estimated the paleoelevation difference between two isochronous fossil assemblage sites is 7910 m assuming that the
expected errors apply to past.
Recently, Peppe et al. (2010) argued that many sources of
large potential error are related to this method. Specifically,
they demonstrated a significant bias toward underestimation
of leaf area in the Climate-Leaf Analysis Multivariate Program
data set affecting all predicted climatic variables such as enthalpy, leading to an uncertainty in mean annual enthalpy
higher than that predicted by Forest et al. (1999). Errors in

paleoelevation estimates when the leaf size bias is included are
72000 m or more. Peppe et al. (2010) concluded that among
others, this bias affects previously published paleoelevations
inferred by this method and that such paleoelevations are
unlikely to be accurate either in magnitude or estimated error
(e.g., Forest et al., 1995; Gregory-Wodzicki, 1997; Spicer et al.,
2003; Wolfe et al., 1997, 1998).
The third method is based on the stomatal frequency
change over altitudinal gradients (Kouwenberg et al., 2007;
McElwain, 2004). Atmospheric pressure decreases with altitude and is shown to depend on the molecular weight of air,
elevation, acceleration due to gravity, the gas constant and
mean July temperature (Jones, 1992). CO2 partial pressure is a
fraction of the atmospheric pressure that does not vary with
altitude (Gale, 1972). The globally averaged atmospheric
pressure at sea level has not changed significantly over the
Cenozoic and the available estimates of molar volume of CO2
at sea level are sufficiently accurate for the Cenozoic. Consequently, the CO2 partial pressure at any elevation is a
function of the CO2 partial pressure at sea-level and the atmospheric pressure divided by a constant (Beerling and Royer,
2002; McElwain, 2004; McElwain et al., 2002). McElwain
(2004) shows an inverse relationships between stomatal
density of plant leaves and the CO2 partial pressure, converting the stomatal density into a potential tool to infer
paleoaltimetry, independent of ecological or local climatic
variability. This is true for many species but not all. Therefore
the sensitivity of the selected fossil leaves needs to be tested
as well as the response limit to high or low CO2 partial
pressure due to morphological and physiological constraints
(Kouwenberg et al., 2007). This method applies to extant
floras since stomatal density of modern leaves of same or close
to the fossil species need to be calibrated to CO2 partial
pressure or elevation (Kouwenberg et al., 2007). McElwain
(2004) obtained a 7300 m of average error. However,
Kouwenberg et al. (2007) highlighted other types of errors
such as the light intensity-related uncertainties due to mixture
of sun and shade leaves and varying for different species, and
from uncertainties in exact sea-level CO2 concentrations. For
the species studied by McElwain (2004) and Kouwenberg et al.
(2007) estimated errors related to the light intensity give error
ranges of 162 and 558 m whereas errors introduced in estimating sea-level CO2 from proxy data when sea-level CO2
estimates on the same species are not available, hypothetically
range between 7450 m and 71200 m and could increase to
71400 m (Kouwenberg et al., 2007). When the uncertainty in
sea-level CO2 estimates is high, such method is not better than
the others based on paleobotany, whereas it is the best one
when sea-level CO2 estimates from proxies are accurate
or modern low-elevation leaf of extant species available
(Kouwenberg et al., 2007).
In summary, the paleoelevations inferred in the Rocky
Mountain orogenic plateau from paleobotanic methods indicate the plateau was already high in the late Eocene but the
accuracy of such estimates are low, more than 71100 m, and
the many factors that influence the character of the flora
prevent a reliable estimation of the accuracy of the paleoelevation. The new paleoaltimetric method based on stomatal
frequency and stomatal index may be a good alternative in the
particular cases where the uncertainty in sea-level CO2 estimates is low.

In the case of the AltiplanoPuna plateau of the Andes,
paleoelevation estimates from fossil floras (nearest-livingrelative and foliar-physiognomy methods) now lying at 4000
to 4300 m asl argue for low elevations at 1000 m asl in the
early Miocene (25 to 19 Ma) and ranging from 0 m to 2800 m
asl in the mid Miocene (B20 to B10 Ma) but with
large uncertainties ranging from 71200 to 72000 m (see
Gregory-Wodzicki, 2000 for a review).
5.6.2.4
5.6.2.4.1
Paleoaltimetry Data Based on Stable Isotopes

Paleoaltimetry data based on the
stable isotopic records (d18O and d2H) of
carbonates derived from meteoric and surface
waters
Empirical and theoretical approaches have used the d18O and

the d2H to deduce paleoelevation estimates. The condensation
of water vapor driven by the ascension of regional winds along
orographic obstacles creates a fractionation of H and O in
atmospheric water vapor, rainfall, and snowfall. At the equilibrium water-vapor transformation, heavy isotopes (18O and
2
H (D)) are removed from water vapor because they form
stronger bonds than those formed by lighter isotopes (16O
and 1H) which are more likely to break and to remain in the
gas phase. Consequently, rainfall is enriched in heavy isotopes,
whereas continuing rising air is depleted in heavy isotopes and
the observation of decreasing in d18O and d2H values in
rainfall as elevation increases has been used as a proxy for
paleoelevations.
Rowley et al. (2001) and Rowley and Garzione (2007) have
presented a theory that tracks the change in oxygen isotopic
composition of precipitation based on the equilibrium fractionation during Rayleigh distillation coupled with the
thermodynamics of atmospheric ascent and water vapor
condensation. In their model, however, the complex processes
involved in the conversion from condensate to precipitation
and its influence on isotopic compositions is based on an
empirical relationship (based on data from the Alps). They
show that low-elevation temperature primarily controls the
vertical distribution of condensation whereas low-elevation
relative humidity only results in a small contribution. The
low-elevation temperature, especially, and to a lesser degree
the relative humidity, consequently control the vertical distribution of the difference in isotopic composition (D(d18O))
between a low, preferably near sea level, composition of meteoric water and a potentially elevated sample of meteoric
water. Their model fits well with the modern vertical profiles
of d18O in precipitation (Rowley and Garzione, 2007). Hence,
they theoretically demonstrate that D(d18O) can be used as a
measure of elevation with a greater accuracy of the modeled
elevations at higher elevations, and a general tendency of the
modeled elevations to be underestimated. Where the lowelevation temperature and the relative humidity of local climate is used, instead of global means, the model is also more
accurate (Rowley and Garzione, 2007).
Studies that track paleoelevation changes through geological times generally use carbonates derived from surface of
ground waters that integrate precipitation waters. The quantity
113
of precipitation varies along months and seasons with precipitations of different isotopic compositions. It explains why
the mean isotopic composition at any given station is precipitation amount-weighted. Streams also spatially integrate
precipitation above the sample site with isotopic compositions that vary with elevation. But the frequency of elevations within a catchment (the hypsometry) is not a linear
function and must be computed individually. Furthermore,
the amount of precipitation also varies as a function of elevation (Anders et al., 2006; Rowley and Garzione, 2007).
Consequently, the isotopic compositions of surface water
reflect the amount of precipitation falling as a function of
elevation on that hypsometry: the precipitation amountweighted hypsometric mean elevation of Rowley and Garzione
(2007). An important conclusion is that the isotopic
lapse rate of precipitation and that determined by surface
waters are expected to be different (Rowley and Garzione,
2007).
In the following, important advantages, restrictions and
uncertainties inherent of the methods, and highlighted by
Rowley and Garzione (2007), are briefly summarized.
First, since the resulting paleoelevation corresponds to the
precipitation amount-weighted hypsometric mean elevation,
it does not give any information on the distribution of
drainage basin elevations and particularly on the maximum
elevations. For that reason estimates of paleoelevations from
low-elevation samples of fluvial origin taken in foreland
basins cannot discriminate between plateau-like topographies
and rugged mountains, both implying different tectonic
implications. Intermontane basins at higher elevations are
then considered to be better targets to reduce the effects of the
hypsometric and elevation dependence (e.g., Garzione et al.,
2006; Ghosh et al., 2006b).
Second, past climates were globally warmer than today
leading to a shallower isotopic lapse rate, which means that
the use of modern isotopic lapse rate underestimates paleoelevations in the cases past-climates were really warmer.
Third, evaporation of precipitation in the atmosphere before reaching the ground or in stream waters results in 18O
enrichment relative to 2H when plotted against the Global
Meteoritic Water Line. Data from regions where evaporation
occurs do not fit the model of Rowley et al. (2001), or Rowley
and Garzione (2007). Although sedimentary carbonates reflect
as well as stream waters various sources with hypsometric and
elevation-dependent precipitation amount effects as potentially evaporation effects, paleosol carbonates have the advantage to reflect rainfall that directly infiltrates through the
soil making them potentially better proxies for local rainfall
composition estimates.
Fourth, although isotopic compositions of precipitation
may be highly variable at short time scales, the low values of
carbonate sedimentation rates imply that samples integrate
isotopic compositions at least over thousands of years and
maybe tens of thousands of years. This is an advantage to
deduce the long-term variation of D (d18O) and the associated
changes of elevations.
Fifth, the isotopic composition of carbonates is different,
however, from that of surface waters from which they precipitate, with an additional fractionation of oxygen isotopes
that depend on the temperature (B 1% each 5 1C).
114
Consequently, the assumption of a given paleotemperature

when the carbonate formed adds a significant uncertainty for
calculations of paleoelevations. Mean annual temperature can,
for example, be deduced from the Climate-Leaf Analysis
Multivariate Program (Wolfe, 1993, 1995; Wolfe and Spicer,
1999). However, it may neither correspond to the temperature
of formation of paleosol carbonates since they are associated
with evaporation events that only occur during the dry season,
or during the growing season when evapotranspiration rates
are highest (Cerling and Quade, 1993; Liu et al., 1996), nor
the temperature of formation of lacustrine carbonates since
they form during the summer when the evaporation rates are
highest and the carbonate solubility is lowest (e.g., Duston
et al., 1986; Effler and Johnson, 1987). The recently published
carbonate thermometer based on 13C-18O bonds is however a
good alternative to deduce the mean temperature of carbonate
formation and finally, knowing d18O of carbonates, to calculate the d18O of meteoric water from which they form. Rowley
and Garzione (2007) also pointed out that diagenesis can also
alter d18O of carbonates where temperature of recrystallization
exceeds 40 1C, leading to useless carbonates for paleoelevation
estimates. In summary, the uncertainties in this method come
from various sources like the low-elevation temperature used
in modeled d18O of meteoric waters, the dispersion of the data
used to deduce empirical isotopic lapse rates, and the temperature of carbonate precipitation (see Rowley and Garzione,
2007 for more details), and are estimated to range between
7400 to 7500 m (1s errors) at paleoelevations higher than
4000 m asl to 71000 m at paleoelevations lower than 500 m
asl (errors are deduced from bootstrap simulations in both the
Andes and the Himalaya/Tibet). However, other errors come
from global or local climate changes, as well as from the
evaporation of surface waters, but their impact on the accuracy
of paleoelevation estimates is not easy to quantify (Rowley
and Garzione, 2007).
5.6.2.4.2
Paleoaltimetry data based on the abundance of

O C O (D47): The carbonate clumped isotope
paleothermometer
18 13 16
Recently a new technique has been developed to estimate

paleoelevations using a new paleothermometer: the clumpedisotope D47 (Ghosh et al., 2006a). The ionic group
13 18 16 2
C O O2 is a multiply substituted isotopolog in carbonate
minerals. The two heavy isotopes make it to have lower zeropoint energies than its isotopically normal and singly substituted relatives (12C16O3, CO3, COO2, and COO2). A direct
consequence is that rare isotopes are clumped by a thermodynamic driving force creating bonds with each other instead
of being randomly dispersed throughout the mineral lattice.
The abundance of multiply substituted molecules increases
with decreasing temperature (Wang et al., 2004). The abundance of 13C18O16O (mass 47) in CO2 produced by digesting
a carbonate mineral in phosphoric acid has been shown to be
proportional to the abundance of 13C18O16O2 in carbonates
(Ghosh et al., 2006a). It is hence possible to measure the
d18O, d13C, and abundance of D47 isotopologs (mostly
13 18 16
C O O) in the product CO2 (e.g., Ghosh et al., 2006b).
The abundance of 13C18O16O (mass 47) is measured relative
to 16O12C 16O (mass 44) and called R47. R47 is normalized by
the R47
random and called:
47
D47 R47
measured =Rrandom 1 1000,
with R47
random calculated from the random distribution of heavy
isotopes in the sample based on the measured d18O, d13C
values (Eiler and Schauble, 2004). The temperature-dependant
enrichment in 13C18O16O has been calibrated using synthetic
calcites precipitated under a temperature range of 1 1C to
50 1C, representative of surface conditions (Ghosh et al.,
2006a) leading to a relation between the new paleothermometer D47 and the temperature (T in 1K) of carbonate precipitation in nature:
D47 0:0592 106 T 2 0:02:
This new paleothermometer is particularly useful since
it allows paleoelevations to be obtained in three ways
(Ghosh et al., 2006b; Quade et al., 2007; Rowley, 2007).
Firstly obtained paleotemperature can be compared to the
measured temperature lapse rate to deduce paeloelevations,
especially in arid regions where d18O and d2H techniques
cannot be applied due to evaporation of precipitation before
reaching the ground. The second way to use the D47 and the
associated paleotemperature is to deduce the isotopic composition of meteoric water from which carbonate precipitates,
since the d18O of carbonate is known and can be used to
compare such values to the altitude dependence of d18O
(Garzione et al., 2006). The third way to obtain paleoelevations is to sample a vertical transect to obtain the modern
mean annual and mean summer surface temperature as a
function of the modern annual weighted mean d18O. The
same two variables are obtained from the D47 and can be
compared to the expected values estimated from the modern
vertical transect (Ghosh et al., 2006b). The similarity in slope
in these two dimensions between sampled data and the
modern mean annual trend is shown to reflect primarily the
change in elevation (which is the strength of this method) and
not a change in the global climate or latitude or changes in the
season of formation of soil carbonates (although these second-order effects must be corrected) (Ghosh et al., 2006b).
The modern meteoric water d18O values and surface temperatures vary with season during the year and soil carbonates
may grow during different seasons leading to an offset from
the modern mean annual trend (meteoric waters d18O vs.
temperatures). At each elevation, the seasonal variation of
meteoric waters d18O versus temperatures is shown to be
a series of lines almost perpendicular to the modern mean
annual trend in the graph of meteoric waters d18O versus
temperatures (cf. Figure 1 of Ghosh et al., 2006b). Linking
seasonal extreme values with the mean annual trend for
corresponding altitudes allows the generation of lines of
constant altitude (isohypse) from which paleoelevations can
be derived.
The uncertainties linked to the D47 method are significantly
lower than most of any other method. Ghosh et al. (2006b)
presented values of paleoelevations for the Bolivian Altiplano
from the Late Miocene to Pliocene of 2007200 m between
11.4 and 10.3 Ma, 25007500 m between 7.6 and 7.3 Ma, and
35007400 m between 6.7 and 5.8 Ma.
This method is based on the assumption that atmospheric
circulation patterns have not changed significantly in the past
10 to 15 Ma at the latitude of the Andes (Ghosh et al., 2006b).
Any change in climate and atmospheric circulations would
however modify the moisture supply, the amount of rain that
falls before the clouds reach the studied orogen and consequently the d18O values (Ghosh et al., 2006b; Poulsen et al.,
2010). Poulsen et al. (2010) used a global circulation model
that tracked the water vapor and rainfall isotopic composition
in space and time. Poulsen et al. (2010) also showed that
surface uplift of the Atliplano results in spatially heterogeneous
variations in the amounts of precipitations over the Andes,
with positive or negative variations. They showed that the decrease in rainfall d18O is not systematic with surface uplift.
Here the authors present only two examples from the paper
of Poulsen et al. (2010) to illustrate their results. First, as surface
uplift increases, precipitation d18O also increases in the eastern
flanks of the central Andes where moisture supply coming from
the enriched d18O of the Amazon Basin is enhanced as a consequence of intensification and a shift to the west toward the
Andes of the South American low-level jetstream. Second,
modifications of the South American low-level jet also lead to
large increases in precipitation in the northern central Andes
and consequently to a substantial decrease in meteoric water
d18O due to amount effects (the d18O of airmasses is correlated
to rainfall rates). Poulsen et al.s (2010) model shows that isotopic lapse rates can change through time from low to high
isotopic lapse rates, leading to an underestimation of paleoelevations and consequently an overestimation of a recent surface
uplift if modern isotopic lapse rates are applied to past d18O
values to estimate paleoelevations. Poulsen et al. (2010) concluded that the depletion in meteoric water d18O preserved in
Miocene soil carbonates of the central Andes (Ghosh et al.,
2006b) partially explains an increase in mean elevations. The
remaining part reflects the initiation and the intensification of
convective rainfall induced by orographic lifting of water vapor,
which is in turn enhanced by climate change (atmospheric
circulation changes in wind directions and speeds) and obviously by surface uplift (Poulsen et al., 2010).
5.6.2.5
Paleoaltimetry Data Based on Paleoatmospheric

Pressure from Basalt Vesicularity
Atmospheric pressure is simply the mass of air above a given

level, and the mass of air above a given level decreases with
elevation as well as the density of molecules that constitute our
atmosphere. Consequently air pressure decreases with elevation and it does so at a standard atmospheric lapse rate of
B100 mb km1 in the lower atmosphere. During the Cenozoic atmospheric pressure at sea level probably did not change
significantly. Since the atmosphere composition did not evolve
significantly, the standard atmospheric lapse rate can be considered constant. Everyone knows from weather forecast maps
that surface atmospheric pressures vary spatially, with anticyclonic centers of high pressure and cyclonic centers of low
pressure. Such meteorologic features move horizontally at
daily to season timescales as a function of insolation, a spatial
115
distribution also influenced by the landmass distribution and

the oceanic circulations. Since landmass distribution varies at
the geological timescale one can also expect variations in the
spatial distribution of highs and lows at the geological time
scale. However, modern surface atmospheric pressure changes
range between B970 mb in deep lows (hurricane) to
B1040 mb in strong highs at sea level, leading to maximum
modern climate-driven variations on the order of B70 mb
between highs and lows, a value reduced to a most likely
30 mb by Sahagian et al. (2002b). As a consequence, a change
in topographic elevation of 3 km will result in a mean surface
air pressure decrease of 300 mb leading to a signal 10-fold
higher than that of a likely climate-driven maximum change in
air pressure. Since the standard atmospheric lapse rate is constant through time (during at least the Cenozoic) and known,
a measure of a past surface air pressure is a direct measure of
paleoelevation independent of long term climate changes.
The pressure near the surface of a lava flow is mostly the
atmospheric pressure whereas at the bottom of a lava flow the
pressure is the sum of the hydrostatic pressure that the column
of lava exerts, plus the atmospheric pressure. Hence, the difference in internal pressure in bubbles at the base and top of a
lava flow depends on atmospheric pressure and lava flow
thickness (Sahagian et al., 1989, 2002a; Sahagian and Maus,
1994; Sahagian et al., 2002b). Because lava flow thickness is
easy to measure in the field, the size of bubbles in vesicular
basalts records the paleoatmospheric pressure, the variation
of which can be interpreted in terms of elevation changes,
assuming a known air pressure lapse rate (Sahagian and
Proussevitch, 2007; Sahagian et al., 2002a, and references
herein). The measurement of vesicle size in 2.5 cm core samples is realized using numerical analysis techniques (Proussevitch and Sahagian, 2001; Proussevitch et al., 2007) of slice
images obtained by X-ray computed tomography (Ketcham
and Carlson, 2001).
The uncertainties associated with the method are 7400 m
(s) based on elevation of modern Hawaiian lavas used to
validate the method with more accurate results with multiple
sampling. The accuracy of the technique asserted by another
approach, including errors from the measurements of vesicle
size and lava flow thickness, and the use of average sea-level
barometric pressure is 7410 m agreeing with the standard
deviation approach (Sahagian et al., 2002a).
This method has been applied to unravel the uplift history of
the Colorado Plateau (Sahagian et al., 2002a). Studied lavas
were emplaced between 800 and 2000 m regardless of the age
suggesting that basalts were erupted in all parts of the Colorado
Plateau since 25 Ma (Sahagian et al., 2002a). This means the
plateau was not flat during the past 25 Ma with a relief of at least
1200 m and a mean elevation higher than 800 m. Sahagian
et al. (2002a) obtained a rock uplift rate of 40 m Ma1 between
25 and 5 Ma (800 m of rock uplift), and an acceleration to
220 m Ma1 since 5 Ma (1100 m of rock uplift).
5.6.2.6
Paleoaltimetry Data Based on Cosmogenic

Nuclides
Cosmogenic nuclide production rates increase with elevation

above sea level (e.g., Lal and Peters, 1967). As such, this might
116
be used to determine a paleoaltimetry (e.g., Blard et al., 1995,

2005; Dunai, 2010; Gosse and Phillips, 2001; Libarkin
et al., 2002; Riihimaki and Libarkin, 2007; Schafer et al., 1999;
Van der Wateren and Dunai, 2001). First, the exposure
history of a sample must be well constrained to derive a
paleoelevation or an average uplift rate from its cosmogenic
nuclide concentration. Two methods have been used and
summarized by Riihimaki and Libarkin (2007) and Dunai
(2010).
The simplest one consists of determining the paleoelevation of a dated volcanic flow. The lava flow sampled must
have been preserved from erosion and unshielded by sediments during the time of exposure, which must be known
accurately. Duration of exposure can be deduced where the
sampled lava flow has been buried under a more recent,
sufficiently thick and also dated lava flow that prevents
further cosmogenic nuclides from being produced (Blard
et al., 2005). Dividing the paleoelevation estimate by the age
of the overlying lava gives a mean rock uplift rate since the
end of exposure. True rock uplift rate may be higher if the
uplift event lasted less than the time passed since the emplacement of the overlying lava flow. Uncertainties come
from the analytical methods of both cosmogenic nuclide
concentration and dating method used to constrain the lava
ages, and from the uncertainties in the depth-history of a
sample and duration of exposure (Riihimaki and Libarkin,
2007). In order to be able to measure a paleoelevation, the
duration of exposure must exceed tens of thousands years
and in some cases up to several hundred thousand years
(Dunai, 2010). The depth history of a sample is the primary
source of uncertainties. Riihimaki and Libarkin (2007) illustrated clearly that the high density contrast between atmosphere and rocks makes a 0.5 m uncertainty in depth
equivalent to more than 1000 m of error in altitude.
The only way to deduce complex exposure history where
a sedimentary layer may have once capped the sample site is
to perform depth profiles and to analyze the exposure history
by multiple nuclides (Riihimaki et al., 2006). At best
paleoelevation will not be more accurate than B200 m at a
low paleoelevation to B500 m at a high paleoelevation
(Riihimaki and Libarkin, 2007). This is a minimum range of
error because this estimate does not include any uncertainty
in the depth of the sample due to erosion or burying (Riihimaki and Libarkin, 2007), but Blard et al. (2005) argued
that the cosmogenic 3He paleoaltimeter can potentially reach
resolutions better than B1000 m. However, this method has
been seldom used for dating probably due to the usually
poorly constrained depth history of potential targets. The
second method is a test for recent rapid uplift (Brook et al.,
1995; Brown et al., 1991). Brook et al. (1995) have shown
that high concentration of 10Be in poorly dated glacial deposits of the Dry Valleys in Antarctica suggest that they were
deposited at their current high elevation, which deny any
recent high (0.431 mm yr1) rock uplift rate as previously
proposed. However, the method did not allow the discarding
of an alternative explanation in which the glacial sediments
are older and that they have been raised, with a moderate
uplift rate of 0.2 mm yr1, for example, allowing enough
exposure time to produce the high concentration of cosmogenic nuclides.
5.6.2.7
Cooling-History and Erosion Rates as a Proxy for

Rock or Surface Uplift
Cooling-ages can be deduced from thermochronologic

data (e.g., Reiners, 2007, and references herein). A cooling-age
defines a specific temperature (or range of temperatures)
during cooling history of a sample. Cooling-history T(t) can
also be tracked by thermochronological data. Moreover, a
thermal model of the crust allows converting temperatures
into depth and cooling histories that can be transformed
into exhumation histories (erosional denudation or tectonic
exhumation in normal fault footwalls). Thermochronologic
techniques and interpretations relevant to uplifts and topographic changes can be found in Reviews in Mineralogy and
Geochemistry (Reiners and Ehlers, 2005).
First, thermochronology describes the cooling of rocks
(exhumation when temperatures are converted into depth):
the movement of rocks relative to Earths free cooling surface,
which may raise or decrease in elevation through time. Consequently, thermochronology alone neither provides estimates
of paleoelevation, nor of rock and surface uplift. Second, the
cooling-ages do not directly relate to changes in mean elevation or mean local relief and changes in erosion rates (e.g.,
Reiners, 2007, and references herein). For that reason, in most
cases, it is the spatial changes in cooling-ages or coolinghistories that are used to unravel the topographic evolution,
that is, the changes in local relief. Third, thermochronologic
data allow amounts of mean elevation change or rock uplift to
be estimated only when combined with geologic data. Fourth,
channel incision and erosion on coupled hillslopes can reflect
transient response to tectonically driven rock uplift (e.g.,
Whipple and Tucker, 1999). But it can also reflect transient
response to climatically driven changes in sediment supply and
fluvial discharge (Molnar and England, 1990). Provided prior
sufficient thickening of the crust, moderate to high mean elevations are supported by a crustal root. Rock uplift in thickened settings is then the sum of the tectonic rock uplift and the
erosionally driven rock uplift (e.g., Whipple, 2009).
5.6.2.7.1
Vertical profiles of thermochronological data

combined with other lines of evidence as a
proxy for rock and surface uplift
Age-elevation relationships in a vertical profile show a normal

correlation where erosion rates and rock uplift rates are uniform and the slope of this relation is the erosion rate. Provided
additional geologic information on paleoelevation is available
(e.g., stratigraphic) and assuming a geothermal gradient, vertical profiles of thermochronologic cooling-dates can be used
to deduce rock-uplift and surface-uplift rates even if direct
paleoelevation markers like marine sediments are missing
(e.g., Ducea et al., 2003). In the Santa Lucia Mountains of
California (US), the last major transition from submarine
to subaerial sedimentation occurred at approximately 4 Ma.
Ducea et al. (2003) deduced the topography was at sea level at
4 Ma. Assuming a geothermal gradient of 25 1C km1, a mean
surface temperature of 10 1C and the (U-Th)/He closure isotherm for apatite at 76 1C, they deduced the closure depth in
the Santa Lucia Mountains of California (US) at 2640 m
below the sea level at 4 Ma. The current elevation (800 m)
of the 4 Ma closure depth (deduced from the vertical
profile) plus the paleoelevation of the closure depth at 4 Ma

( 2640 m) led to B3440 m of rock uplift since then
(0.86 mm yr1). The next step consisted of using the acquired
estimates of erosion rates (0.66 mm yr1) from the vertical
profile (0.35 mm yr1 for the period 42.3 Ma and B0.9
mm yr1 since 2.3 Ma) to calculate the corresponding surface
uplift (rock uplift minus erosion rates) (Molnar and England,
1990). Ducea et al. (2003) deduced both the rock uplift
(0.86 mm yr1) and the surface uplift (0.2 mm yr1) from
denudation rates and stratigraphic data. Such an approach
however makes the assumption that erosion rates are uniform
in space, at least at the scale of mean elevation, that is,
1000 km2 (at the same spatial scale than that of the crust, see
Section 5.6.1.1).
5.6.2.7.2
Horizontal profiles of thermochronological data

combined with other lines of evidence as a
proxy for rock/surface uplift
Spatially extensive sampling over horizontal distances is more

useful to constrain the local relief evolution rather than rock/
surface uplift evolution (e.g., Reiners, 2007, and references
herein). A paleorelief (paleo local-relief) for a specific location
gives a minimum estimate of paleoelevation above sea level.
But both the distribution of the paleorelief within the landscape and the history of change in base level must be known
to convert a local relief change in a mean elevation change
(Clark, 2007), a geodynamically more grounded measure of
elevations. As long as there is no direct relationship between
mean elevation or local relief and erosion rates there is no
direct relation between mean elevation and local relief.
In settings like high plateaus (e.g., Tibet, Altiplano), low-relief
in the internally drained regions dominates whereas mean
elevations are the highest in the world. Clark (2007) stressed
that local relief is expected to increase during a fall
in base level without an increase in mean elevation. The low
local-relief of the high plateaus will undoubtedly increase
when it is integrated into the drainage network draining the
edges of the plateaus down to the sea without a need for
the mean elevation to increase if any geodynamic process that
could account for it occurs in the meantime.
The crustal thermal field is composed of deep isotherms (T
4100 1C) parallel to the long topographic wavelengths (or
mean elevations). Lower isotherms progressively fold toward
the shallow depths where isotherms closely fit to the shortwavelength topography (local relief). The perturbation depth of
the disturbance is shown numerically to be a strong function of
the short wavelength of the topography, but it is weakly
dependent on its amplitude (Braun, 2002a, b, 2003; Braun
et al., 2006; Reiners et al., 2003). However, additional geologic
information is here also needed since any thermal perturbation
can also result from thrust kinematics and from changes in
erosion rates and rock uplift rates (e.g., Braun, 2005; Brewer and
Burbank, 2006; Ehlers et al., 2005; Huntington et al., 2007).
Many studies used the influence of topography on cooling-ages
or cooling-histories to demonstrate local relief changes through
time in orogens (Braun, 2002a; Clark et al., 2005a, b; Densmore
et al., 2007; Ehlers et al., 2006; House et al., 2001; Mancktelow
and Grasemann, 1997; Reiners et al., 2006; Stock et al., 2006;
Stockli, 2006; Stuwe et al., 1994). Because warping of isotherms
is higher for the lowest temperature thermochronometers, local
117
relief changes will be best evidenced by the lowest thermochronometers (see Valla et al., 2010).
In the Sierra Nevada of California, House et al. (1998,
2001) have shown using a constant-elevation sample transect
parallel to the orogen that the apatite helium cooling-ages for
the closure temperature of 6875 1C are roughly anticorrelated with modern valleys and ridges transverse to the
range. They interpret this spatial correlation as the evidence for
a long-wavelength (l 70 km) relief of 15007500 m higher
in the Late Cretaceousearly Tertiary than today. They also
correlated the local relief to a mean elevation higher than
3000 m asl by analogy with the modern Andes. Clark et al.
(2005b) showed that such an analogy is misleading and they
combined the results on local relief obtained from thermochronologic data with geomorphic data allowing them to
argue for a range elevation of only 1500 m asl during the same
time. This highlights the constant need from other lines of
evidence to better interpret thermochronologic data in terms
of paleotopography.
Braun (2002a) showed that decreasing or increasing local
relief can modify the age-elevation relation in vertical profiles
leading to, respectively, steeper or shallower slopes of the
linear regressions than the actual erosion rate. Braun (2002a,
b, 2005) defined a method of unraveling the transient state of
landscape evolution, that is, whether local relief increases or
decreases through time. It is based on an analysis of the ageelevation relationship over a continuous three dimensional
spectrum of topographic wavelengths. The main assumptions
are constant geothermal gradient, steady landforms, and no
horizontal advection of rock (Braun, 2002b). When applied to
the Sierra Nevada of California, Brauns (2002a) approach
leads to the same conclusion of House et al.s (1998, 2001),
that is, a local relief decrease since the Late Cretaceous.
However, it does not define the mean elevation change, which
is critical in the understanding of high plateaus.
The lithosphere flexural strength controls the isostatic-flexural rebound to erosion and consequently potentially controls
the spatial distribution of thermochronologic data. In the
postorogenic evolution of an orogen, a greater reduction in
topographic relief (at shorter and longer wavelengths, 12 and
150 km, respectively, for the case of the Dabie Shan, China) has
a similar effect on the spatial distribution of the thermochronologic ages as does a thinner lithosphere (Braun and
Robert, 2005). In their study of the postorogenic decay of the
Dabie Shan and following the spectral method of Braun
(2002a, b, 2005), Braun and Robert (2005) confirmed that a
low elastic thickness and flexural isostatic rebound can explain
the recent ages in the center of the orogen without any need for
a recent (Cenozoic) tectonic event and associated surface uplift.
5.6.2.7.3
Incision rates obtained from

thermochronological data combined with other
lines of evidence as a proxy for uplift
Local relief history can also be unraveled by measuring depth

in incision evolution of a canyon where it is incised within a
smooth and mainly uneroded paleolandscape. Under particular conditions local relief (canyon depth) can be used as a
proxy for rock uplift and surface uplift. First, depth in incision
is limited by rock uplift. The incision response time to rock
uplift is controlled by a geomorphic response time that can
118
reach several million years (e.g., Hoke et al., 2007; Schildgen

et al., 2010; Schlunegger et al., 2006). Second, the thermal
response to changes in the shape of Earth free cooling surface
is likely to lag B1 Ma (e.g., Schildgen et al., 2007). As a
consequence the incision amounts and rates inferred from
cooling-histories give a minimum rock uplift amount and rate
(Schildgen et al., 2007). Only if climate is steady or becomes
drier during the period of incision, rock uplift and the
resulting incision can be interpreted as tectonics in origin, and
they are accompanied with surface uplift (e.g., Schildgen et al.,
2007).
Incision history can be deduced from dating methods such
as low-temperature thermochronology of samples collected in
vertical profiles or collected along valley at river level. K/Ar or
Ar/Ar dates of tuffs or lava flows preserved above the current
river level are also used to deduce mean incision rates
since their formation (e.g., Thouret et al., 2007). But
thermochronology is not a unique technique. Cosmogenic
nuclide dates of terraces also give values of mean incision
since the abandonment of a terrace. Cosmogenic nuclide
techniques, however, generally allow dating of surfaces for the
last tens of thousands to hundreds of thousands years, rarely
up to a few million or tens million years (e.g., Dunai et al.,
2005).
In the western slopes of the Peruvian Altiplano, Schildgen
et al. (2007) have deduced from samples collected along a
valley at river level and their He ages, a minimum of 2.4 km of
erosion between 9 and 2.3 Ma in the Cotahuasi-Ocona Canyon, a canyon of up to 3 km depth. They interpret it as the
evidence for rock uplift of tectonic origin since climate shifted
to drier conditions in the meantime (e.g., Ehlers and Poulsen,
2009; Gaupp et al., 1999; Hartley, 2003; Rech et al., 2006).
Rock uplift can then be converted to surface uplift since erosion is restricted to canyon incision into a well preserved
paleosurface older than 1416 Ma.
At the eastern margin of the Tibetan Plateau, Kirby et al.
(2002) and Clark et al. (2005a) showed in age-elevation relations a break in slope at B512 Ma and 913 Ma, respectively. They interpreted the recent (late Miocene or early
Pliocene) rapid incision of rivers within a widespread relict
landscape, with mean elevations now above 1000 m asl
and up to 3000 m asl, slowly eroding since the Jurassic
as evidence for eastward plateau growth, that is, surface
uplift.
On the Colorado Plateau, in the upper reaches of the
Grand Canyon, Flowers et al. (2008) determined coolinghistories similar during the Early- to mid-Tertiary in samples
now separated by 1500 m in both elevation and stratigraphic
position. They interpret it as evidence for a proto-Grand
Canyon that was B1 km deep, implying a minimum rock
uplift of 1 km before mid-Tertiary and that the watershed of
the Grand Canyon already lay at high elevation by that time.
5.6.3
Tectonic Mechanisms and Associated Surface

Uplift Rates for Plateau Uplift, Regional
Warping, and Subsidence
Uplift of Earths surface expresses the isostatic response to

the thickness variation of the lithosphere resting on the
asthenosphere (e.g., Fleitout and Froidevaux, 1982), as well as

the differential buoyancy of the convective mantle resulting in
dynamic topography (e.g., Hager et al., 1985). So, whatever
the tectonic model envisioned to explain uplift, including the
case of dynamic topography, it implies a change of density at
depth. Density changes in the lithosphere may be achieved in
two ways: (1) changing rock densities; and (2) changing
the respective thickness of the crust and the mantle, more
precisely thickening the crust and thinning the lithosphere
mantle. Changing rock densities requires changing temperature such as increasing temperature induces decreasing density, increasing buoyancy forces and eventually surface uplift.
Rocks of the mantle lithosphere are much denser than the
rocks of the crust and the mantle lithosphere is much thicker
than the crust. Therefore, temperature increase in the lithospheric mantle will be much more efficient to Earths surface
uplift than it does in the crust. The boundary between lithosphere and asthenosphere is of thermal nature, so thinning of
the mantle lithosphere can be achieved either by mechanical
processes or a temperature increase. Crustal thickening requires mechanical processes, even if magma underplating
from underlying mantle can be considered as thermal
thickening regarding the role of temperature in magma production. For a normal geothermal gradient, the shear strength
of the crust is much lower than that of the lithospheric upper
mantle, and crustal thickening is easier than lithosphere
mantle thinning, considering mechanical processes. Finally,
there are multiple ways to change density at depth in the
lithosphere, which results in numerous models for plateau
surface uplift. The line of argument for plateau uplift has its
symmetrical thinking for subsidence: decreasing temperature
will induce increasing rocks density, decreasing buoyancy and
plateau subsidence, and thickening of the mantle lithosphere
as well, and so on.
Discussing exhaustively the numerous tectonic models that
have been proposed to explain the uplift of large areas at Earth
surface is beyond the scope of this chapter. First, we must
remember that we are dealing with plateaus as we recall the
definition in the first section, that is, large areas with flat
topography. We must also remember that two types of hypotheses sustain the interpretation of such features at Earths
surface: the acquirement of the flat topography precedes uplift;
and smoothing of topography results from the interaction of
surface processes (erosion/sedimentation) with tectonics
during uplift.
The question arises as to what kind of data allow us to
discern between the effective mechanisms that are proposed to
explain plateau uplift. We can group together these increasingly numerous models into three types: (1) crustal thickening
as the driving mechanism; (2) uplift due to the thinning of the
mantle lithosphere; and (3) the role of the deep convective
mantle as the cause of uplift and referring to dynamic
uplift.
Crustal thickening and thinning of the mantle lithosphere
may be achieved either by homogeneous or heterogeneous
deformation. Of course, end-member models imply hybrid
models. Symmetrical end-members models, that is, crustal
thinning and mantle lithosphere thickening, either by
homogeneous or heterogeneous deformation, exist to explain
subsidence.
Crustal thickening can be induced either by: (1) crustal

shortening beneath the uplifting plateau (e.g., Argand, 1924;
Bird, 1978; Dewey and Burke, 1973; Froidevaux and Ricard,
1987); (2) by lateral flow of ductile crust from immediately
adjacent thickened area (e.g., Bird, 1991; Zhao and Morgan,
1987); (3) by crustal underplating of material coming from
the continental margin and tectonically eroded by the subducting plate (e.g., Baby et al., 1997; Schmitz, 1994); and (4)
by magmatic addition (Gotberg et al., 2010; Kono et al., 1988;
Thorpe et al., 1981) but to a much smaller contribution than
shortening and, mid and lower crustal flow (e.g., Baby et al.,
1997; Francis and Hawkesworth, 1994; Giese et al., 1999;
Sheffels, 1990). Crustal thickening by thrust faults, that is,
involving strain localization, corresponds to a heterogeneous
deformation mode. Crustal shortening by homogeneous deformation will require homogeneous ductile rheology, which
implies a high geothermal gradient. As heterogeneous deformation is likely to produce strong topographic gradients
and relief roughness at Earths surface, mass redistribution by
erosion/sedimentation process is needed to maintain a flat
topography. Lateral flow of ductile crust may be assimilated to
a homogeneous deformation mode. In this case, flat topography may exist long before uplift, unlike the previous case
where syn- or late uplift sedimentation will be inevitably
observed at the plateau surface.
Thinning of the mantle lithosphere may develop in two
ways. First as mentioned above, the thermal boundary between lithosphere and asthenosphere migrates upward as the
geothermal gradient increases. Mantle plumes may generate
such processes resulting in surface uplift (e.g., Eaton, 2008;
Teixell et al., 2005). Second, the lithosphere mantle, which is
colder and denser than the underlying asthenosphere mantle,
may be removed from the buoyant crust and sink into the
asthenosphere. Whether such lithospheric mantle removal
is considered to result from progressive peeling away during
continental collision, or more or less sudden foundering by
convective removal after lithosphere thickening, it is named
lithosphere delamination (Bird, 1978) or detachment (e.g.,
Houseman et al., 1981).
The term dynamic uplift is often used when Earths
mantle, including the lithospheric mantle, is suspected of
being involved in surface uplift processes, but strictly speaking
it does not imply a thinning of the mantle lithosphere.
Dynamic topography refers to vertical stress generated by the
deep convective mantle that results in subsidence or uplift of
Earths surface (e.g., Gurnis, 1993; Hager et al., 1985; LithgowBertelloni and Silver, 1998).
Most interpretations of uplifted plateaus appeal to hybrid
models, that is, a more or less complex combination of the
different processes discussed above. For the same object, different
contradictory processes, in that they result in a conflicting uplift
history, are invoked. For example, as crustal thickening is related
to plate convergence, the rate of surface uplift controlled by this
process must match the rate of plate convergence and it implies
a gradual growth of the plateau (e.g., Liu-Zeng et al., 2008;
Tapponnier et al., 2001). Sudden detachment of a lithospheric
root should generate faster uplift and plateau uplift as a whole
(e.g., Garzione et al., 2006; Molnar et al., 1993).
To illustrate the complexity of this problem, the authors
provide the following, but not exhaustive, interpretations
119
Time
(a)
(b)
(c)
Crust
Lithosphere
Asthenosphere
50 km
Figure 11 Himalaya-Tibet tectonic history involving crustal

thickening and delamination of the mantle lithosphere first proposed
by Bird, P., 1978. Initiation of intracontinental subduction in the
Himalaya. Journal of Geophysical Research 83, 49754987.
that have been proposed for the surface uplift of the Tibetan,
Altiplano, and Colorado plateaus:
1. For the Tibetan Plateau (Figure 11): crustal thickening by
localized deformation (Dewey and Burke, 1973; Tapponnier et al., 2001) or by lateral flow of ductile crust (Clark
and Royden, 2000), crustal thickening and lithosphere
delamination (Bird, 1978), homogeneous deformation
of the whole lithosphere and consecutive detachment of
the mantle lithospheric root (England and Houseman,
1988; Houseman et al., 1981; Molnar et al., 1993).
2. For the Altiplano (for an exhaustive review of mechanisms
see Barnes and Ehlers, 2009): crustal thickening by lateral
flow of ductile crust (Husson and Sempere, 2003), crustal
thickening and detachment of the mantle lithosphere
(Garzione et al., 2006), crustal thickening, magma underplating, and thinning of the mantle lithosphere (Allmendinger
et al., 1997).
3. For the Colorado Plateau: crustal thickening (McQuarrie
and Chase, 2000), delamination of the lithospheric mantle
(Bird, 1979), lateral flow of ductile crust (Bird, 1991),
thinning of the lithosphere mantle due to mantle plume
(Eaton, 1987) or Laramide subduction (Spencer, 1996), or
by warming the lithosphere (Roy et al., 2004), dynamic
uplift (Liu and Gurnis, 2010; Moucha et al., 2009).
5.6.4
Plateau Uplift and Global Climate Change
Uplift of Earths surface as a cause of climate change goes back

at least to Dana (1856) who argued, through geomorphologic
120
and tectonic analysis, that the Paleozoic Appalachians had

been rejuvenated during the Late Cenozoic. A few decades
later, Chamberlin (1899) postulated that the simultaneous
occurrence of orogens and ice ages in Earths history was not a
coincidence but a causal relation: enhanced chemical weathering of aluminosilicate minerals, consecutive to intense
erosion of rocks in mountain belts, subtracts CO2 from the
atmosphere resulting in decreasing atmospheric greenhouse
effect and global cooling (e.g., Raymo, 1991; Raymo and
Ruddiman, 1992). As the authors are concerned with high
continental plateaus, erosion in such areas is weak by definition, and this process will be limited. The primary physical
effects of plateau surface uplift are changes in the atmospheric
circulations that are induced by the presence of an elevated
surface at cooler upper levels in the atmosphere. This may
result in increasing erosion and chemical weathering and
eventually in climate change. The authors will not discuss here
this indirect consequence of plateau surface uplift that is
largely debated elsewhere (e.g., Ruddiman, 1997).
Plateau uplift has three main effects on climate (Ruddiman
and Prell, 1997): (1) lapse-rate cooling of high-elevation
surfaces; (2) perturbation of jet stream meanders; and (3)
creation and intensification of monsoon circulations.
The main present-day high plateaus arose during the Late
Cenozoic, a period of progressive cooling of Earths climate
that culminates in the Late Neogene. So, it is difficult, indeed
even impossible, to determine the influence of a distinctive
plateau on global climate change during this period. Most of
works recognize the primary influence of the Tibetan plateau
because of its extent and elevation, which are far greater than
the other plateaus, and because of its particular latitudinal
position as well (Raymo and Ruddiman, 1992).
Lapse-rate cooling, that is, the atmospheric cooling with
increasing elevation (6.5 1C km1), refers to the fact that it
causes the high plateaus to be much colder than the nearby
plains. This is especially the case in winter because temperatures are further lowered by snow cover, through albedotemperature feedback. Depending on the extent and the
height of the plateau, and its latitudinal position, this effect
may result in nucleating continental ice sheets and global
cooling (Birchfield and Wertman, 1983).
The second effect concerns the increase in amplitude of the
meanders of the subtropical jet stream, as most high plateaus
are lying in lower middle latitudes. Uplifting plateaus create
obstacles to the west-to-east flow that results in the diversion
into a broad poleward meander around the obstacles, and
eventually large meanders develop upstream and downstream
from the high plateaus (e.g., Kutzbach et al., 1989; Rind and
Chandler, 1991; Ruddiman and Kutzbach, 1991; Ruddiman
et al., 1989).
The third direct effect of plateau uplift is the intensification, and even the creation, of seasonal monsoon circulation (Molnar et al., 1993; Ruddiman and Prell, 1997). The
symmetrical Hadley cells describe the atmospheric circulation
in a general west-to-east flow between equatorial and midlatitudes. Warm moist air rises in the intertropical convergence
zone. Condensation and precipitation release latent heat that
causes this now drier air to rise in the troposphere, then to
flow poleward. At approximately 301 N and S latitudes, it
flows east and sinks causing a high pressure area at the surface.
Because of adiabatic heating, this air dries further as evidenced

by the pervasive occurrence of deserts at this latitude. The
uplift of a high plateau such as the Tibetan Plateau at this
latitude has two effects. First, the general west-to-east flow is
altered, so that the high pressure zone at latitude 301 is displaced northward and a low pressure zone develops aloft in
northern India in the summer monsoon. Consequently, the
summer winds originate in the Arabian Sea moving to India,
and the eastern Himalaya receive increasing moisture owing to
the barrier erected by the plateau (orographic effect). Second,
the presence of the hot Tibet in summer causes air to rise near
and above the plateau and to flow toward the equator. This
creates a cell with a reverse sense of rotation compared to the
Hadley cell, that is, air is drawn from equatorial areas and
ascent above the plateau, producing the Indian summer
monsoon, characterized by moisture-bearing winds at surface
blowing from the southwest (from the sea) to the northeast
(to the continent).
In the case of southeast Asia, this perturbation of the general
circulation of the atmosphere due to the Tibet uplift, which
originates from lithosphere dynamics, is considered to have
a strong impact on Earths climate, although timing of
uplift and of related climate change are still strongly debated
(e.g., Dupont-Nivet et al., 2007; Molnar et al., 2006; Rowley
and Currie, 2006). The question remains open and is
best summarized by Molnar et al. (1993) who finally asked the
question: Did convective removal of the lower lithosphere
beneath the Tibetan Plateau force a rapid uplift at approximately 8 Ma that created a heat source sufficient to strengthen
the monsoon, which in turn increased precipitation and weathering in the Himalaya, causing sufficient extraction of CO2
from the atmosphere to induce global cooling?
5.6.5
Conclusion
Most of interpretations of uplifted plateaus appeal to models

with a more or less complex combination of the different
tectonic processes discussed above. Such a situation underscores the urgent need to obtain more pertinent and clear
records of the surface uplift and subsidence history of presentday and past plateaus, respectively, and to come back to
the difficulty of determining paleoaltitudes and structures
at depth.
Acknowledgments
This review was conducted with support from the MINIS N (CGL2006-07226 y
TERIO DE CIENCIA E INNOVACIO
CGL2010-15416). An anonymous reviewer helped clarify
aspects of the manuscript. The authors also thank Lewis
Owen for comments and corrections that improved the
manuscript.
References
Abbott, L.D., Silver, E.A., Anderson, R.S., et al., 1997. Measurement of tectonic
surface uplift rate in a young collisional mountain belt. Nature 385, 501507.
Al-Lazki, A., Eric, I., Dogan, S., Muawia, S., Niyazi, B., Randa, M., T., 2004. Pn
tomographic imaging of mantle lid velocity and anisotropy at the junction of the
Arabian, Eurasian, and African plates. Geophysical Journal International 158,
10241040.
Allmendinger, R.W., Jordan, T.E., Kay, S.M., Isacks, B.L., 1997. The evolution of the
AltiplanoPuna Plateau of the Central Andes. Annual Review of Earth and
Planetary Sciences 25, 139174.
Amante, C., Eakins, B.W., 2009. ETOPO1 1 Arc-minute global relief model:
procedures, data sources and analysis. In: Memorandum, N.T. (Ed.), National
Geophysical Data Center, Marine Geology and Geophysics Division, Boulder,
CO, 19 pp.
Anders, A.M., Roe, G.H., Hallet, B., Montgomery, D.R., Finnegan, N.J., Putkonen, J.,
2006. Spatial patterns of precipitation and topography in the Himalaya.
Geological Society of America Special Papers 398, 3953.
Argand, E., 1924. La Tectonique de lAsie. Compte-rendu du 13e Congre`s
Geologique International, Bruxelles, pp. 171372.
Armijo, R., Tapponnier, P., Mercier, J.L., Han, T.-L., 1986. Quaternary extension in
southern Tiber; field observations and tectonic implications. Journal of
Atalay, I., Koc- man, A., 1979. Kuzeydogu Anadolunun jeotektonik ve morfotektonik
evriminin ana -c izgileri. Jeomorfoloji Dergisi 8, 4175.
Axelrod, D.I., 1966. A method for determining the altitudes of Tertiary floras. The
Palaeobotanist 14, 144171.
Axelrod, D.I., 1998. The Oligocene Haynes Creek flora of eastern Idaho. University
of California Publications in Geological Sciences 143, 99.
Ayalew, L., Yamagishi, H., 2004. Slope failures in the Blue Nile basin, as seen from
landscape evolution perspective. Cosmogenic isotopes in geomorphology 57,
95116.
Ayele, A., Stuart, G., Kendall, J.-M., 2004. Insights into rifting from shear wave
splitting and receiver functions: an example from Ethiopia. Geophysical Journal
Babault, J., Bonnet, S., Crave, A., Van Den Driessche, J., 2005a. Influence of
piedmont sedimentation on erosion dynamics of an uplifting landscape: an
experimental approach. Geology 33, 301304. http://dx.doi.org/10.1130/
G21095.1.
Babault, J., Bonnet, S., Van Den Driessche, J., Crave, A., 2007. High elevation of
low-relief surfaces in mountain belts: does it equate to postorogenic surface
uplift? Terra Nova 19, 272277.
Babault, J., Teixell, A., Arboleya, M.L., Charroud, M., 2008. A Late Cenozoic age
for longwavelength surface uplift of the Atlas Mountains of Morocco. Terra
Nova 20, 102107. http://dx.doi.org/10.1111/j.1365-3121.2008.00794.x.
Babault, J., Van Den Driessche, J., Bonnet, S., Castelltort, S., Crave, A., 2005b.
Origin of the highly elevated Pyrenean peneplain. Tectonics 24, TC2010. http://d
x.doi.org/10.1029/2004TC001697.
Baby, P., Rochat, P., Mascle, G., Herail, G., 1997. Neogene shortening contribution
to crustal thickening in the back arc of the Central Andes. Geology 25,
883886.
Bai, D., Unsworth, M.J., Meju, M.A., et al., 2010. Crustal deformation of the
eastern Tibetan plateau revealed by magnetotelluric imaging. Nature Geosci. 3,
358362.
Bailey, I.W., Sinnott, E.W., 1916. The climatic distribution of certain types of
Angiosperm leaves. American Journal of Botany 3, 2439.
Barke, R., Lamb, S., 2006. Late Cenozoic uplift of the Eastern Cordillera, Bolivian
Andes. Earth and Planetary Science Letters 249, 350367.
Barnes, J.B., Ehlers, T.A., 2009. End member models for Andean Plateau uplift.
EarthScience Reviews 97, 105132.
Bastow, I.D., Nyblade, A.A., Stuart, G.W., Rooney, T.O., Benoit, M.H., 2008. Upper
mantle seismic structure beneath the Ethiopian hot spot: rifting at the edge of
the African low-velocity anomaly. Geochemistry Geophysics Geosystems 9,
Q12022.
flows: 1. Numerical models with applications to the tectonics of the
HimalayanTibetan orogen. Journal of Geophysical Research 109, B06406.
Baumont, D., Paul, A., Zandt, G., Beck, S.L., Pedersen, H., 2002. Lithospheric
structure of the central Andes based on surface wave dispersion. Journal of
Geophysical ResearchSolid Earth, 107.
Beck, S.L., Zandt, G., 2002. The nature of orogenic crust in the central Andes.
Beck, S.L., Zandt, G., Myers, S.C., Wallace, T.C., Silver, P.G., Drake, L., 1996.
Crustal-thickness variations in the central Andes. Geology 24, 407410.
Becq-Giraudon, J.F., Montenat, C., Van Den Driessche, J., 1996. Hercynian highaltitude phenomena in the French Massif Central: tectonic implications.
121
Becq-Giraudon, J.F., Van Den Driessche, J., 1994. Depots periglaciaires dans le
Stephano-Autunien du Massif Central; temoin de leffondrement gravitaire dun
haut plateau hercynien. Comptes Rendus de lAcademie des Sciences, Serie II.
Sciences de la Terre et des Planetes 318, 675682.
Beerling, D.J., Royer, D.L., 2002. Fossil plants as indicators of the phanerozoic
global carbon cycle. Annual Review of Earth and Planetary Sciences 30,
527556.
Best, M.G., Barr, D.L., Christiansen, E.H., Gromme, S., Deino, A.L., Tingey, D.G.,
2009. The Great Basin Altiplano during the middle Cenozoic ignimbrite flareup:
insights from volcanic rocks. International Geology Review 51, 589633.
Birchfield, G.E., Wertman, J., 1983. Topography, albedo-temperature feedback, and
climate sensitivity. Science 219, 284285.
Bird, P., 1978. Initiation of intracontinental subduction in the Himalaya. Journal of
Bird, P., 1979. Continental delamination and the Colorado Plateau. Journal of
Bird, P., 1988. Formation of the Rocky Mountains, Western United States: a
continuum computer model. Science 239, 15011507.
Bird, P., 1991. Lateral extrusion of lower crust from under high topography, in the
isostatic limit. Journal of Geophysical Research 96, 10,27510,286.
Blackstone, D.L., 1975. Late cretaceous and cenozoic history of laramie basin
region, Southeast Wyoming. Geological Society of America Memoirs 144,
249279.
Blard, P.-H., Lave, J., Pik, R., Quidelleur, X., Bourle`s, D., Kieffer, G., 2005. Fossil
cosmogenic 3He record from K-Ar dated basaltic flows of Mount Etna volcano
(Sicily, 381N): evaluation of a new paleoaltimeter. Earth and Planetary Science
Letters 236, 613631.
Braun, J., 2002a. Estimating exhumation rate and relief evolution by spectral
analysis of age elevation datasets. Terra Nova 14, 210214.
Braun, J., 2002b. Quantifying the effect of recent relief changes on age-elevation
relationships. Earth and Planetary Science Letters 200, 331343.
Braun, J., 2003. Pecube: a new finite-element code to solve the 3D heat transport
equation including the effects of a time-varying, finite amplitude surface
topography. Computers & Geosciences 29, 787794.
Braun, J., 2005. Quantitative constraints on the rate of landform evolution derived
from low-temperature thermochronology. Reviews in Mineralogy and
Braun, J., Robert, X., 2005. Constraints on the rate of post-orogenic erosional
decay from low-temperature thermochronological data: application to the Dabie
Shan, China. Earth Surface Processes and Landforms 30, 12031225.
Braun, J., van der Beek, P., Batt, G., 2006. Quantitative Thermochronology:
Numerical Methods for the Interpretation of Thermochronological Data.
Cambridge University Press, Cambridge.
Brewer, I.D., Burbank, D.W., 2006. Thermal and kinematic modeling of bedrock and
detrital cooling ages in the central Himalaya. Journal of Geophysical Research
111, B09409.
Brook, E.J., Brown, E.T., Kurz, M.D., Ackert, R.P., Raisbeck, G.M., and Yiou, F.,
1995.Constraints on age, erosion, and uplift of Neogene glacial deposits in the
Transantarctic Mountains determined from in situ cosmogenic 10Be and 26Al.
Geology 23, 1063-1066.
Brown, E.T., Edmond, J.M., Raisbeck, G.M., Yiou, F., Kurz, M.D., Brook, E.J., 1991.
Examination of surface exposure ages of Antarctic moraines using in situ
produced 10Be and 26Al. Geochimica et Cosmochimica Acta 55, 22692283.
Buck, W.R., 1991. Modes of continental lithospheric extension. Journal of
Burbank, D.W., Leland, J., Fielding, E., Anderson, R.S., Brozovic, N., Reid, M.R.,
Duncan, C., 1996. Bedrock incision, rock uplift, and threshold hillslopes in the
northwestern Himalayas. Nature 379, 505510.
Cahill, T., Isacks, B.L., 1992. Seismicity and shape of the subducted Nazca Plate.
Carretier, S., Lucazeau, F., 2005. How does alluvial sedimentation at range fronts
modify the erosional dynamics of mountain catchments? Basin Research 17,
361381.
Carretier, S., Poisson, B., Vassallo, R., Pepin, E., Farias, M., 2009. Tectonic
interpretation of transient stage erosion rates at different spatial scales in an
uplifting block. Journal of Geophysical Research 114, F02003.
Casas-Sainz, A.M., Cortes-Gracia, A.L., 2002. Cenozoic landscape development
within the central Iberian Chain, Spain. Cosmogenic Isotopes in Geomorphology
44, 1946.
Castelltort, S., Simpson, G., 2006. River spacing and drainage network growth in
widening mountain ranges. Basin Research 18, 267276.
Cerling, T.E., Quade, J., 1993. Stable carbon and oxygen isotopes in soil
carbonates. In: Swart, P., Lohmann, K.C., McKenzie, J.A., Savin, S. (Eds.),
122
Continental Indicators of Climate. American Geophysical Union, Jackson Hole,

WY, pp. 217231.
Chamberlin, T.C., 1899. An attempt to frame a working hypothesis of the cause of
glacial periods on an atmospheric basis. Journal of Geology, 545584.
Chorowicz, J., Collet, B., Bonavia, F.F., Mohr, P., Parrot, J.F., Korme, T., 1998. The
Tana Basin, Ethiopia; intra-plateau uplift, rifting and subsidence. Tectonophysics
295, 351367.
Clark, M.K., 2007. The significance of paleotopography. Reviews in Mineralogy and
Tang, W., 2005a. Late Cenozoic uplift of southeastern Tibet. Geology 33,
525528.
Clark, M.K., Maheo, G., Saleeby, J., Farley, K.A., 2005b. The nonequilibrium
landscape of the southern Sierra Nevada, California. GSA Today 15, 410.
Clark, M.K., Royden, L.H., 2000. Topographic ooze: building the eastern margin of
Clark, M.K., Royden, L.H., Whipple, K.X., Burchfiel, B.C., Zhang, X., Tang, W.,
eastern Tibetan Plateau. Journal of Geophysical Research 111, F03002.
Clark, M.K., Schoenbohm, L.M., Royden, L.H., et al., 2004. Surface uplift, tectonics,
and erosion of eastern Tibet from large-scale drainage patterns. Tectonics 23,
TC1006. doi:10.1029/2002TC001402.
Coffin, M.F., 1992. Emplacement and subsidence of Indian Ocean plateaus and
submarine ridges. Geophysical Monograph 70, 115125.
Coney, P.J., Harms, T.A., 1984. Cordilleran metamorphic core complexes: Cenozoic
extensional relics of Mesozoic compression. Geology 12, 550554.
Cox, K.G., 1989. The role of mantle plumes in the development of continental
drainage patterns. Nature (London) 342, 873877.
Dana, J.D., 1856. On American geological history; address before the American
Association for the Advancement of Science, August, 1855. American Journal of
Science and Arts 22, 305334.
Das, T., Nolet, G., 1998. Crustal thickness map of the western United States by
partitioned waveform inversion. Journal of Geophysical Research 103,
3002130038.
Davis, W.M., 1899. The geographical cycle. Geographical Journal 14, 481504.
Debayle, E., Leveque, J.-J., Cara, M., 2001. Seismic evidence for a deeply rooted
low-velocity anomaly in the upper mantle beneath the northeastern Afro/Arabian
continent. Earth and Planetary Science Letters 193, 423436.
DeCelles, P.G., Robinson, D.M., Zandt, G., 2002. Implications of shortening in the
Himalayan fold-thrust belt for uplift of the Tibetan Plateau. Tectonics 21, 1062.
Densmore, M.S., Ehlers, T.A., Woodsworth, G.J., 2007. Effect of Alpine glaciation
on thermochronometer age-elevation profiles. Geophysical Research Letters 34,
L02502.
Detrick, R.S., Sclater, J.G., Thiede, J., 1977. The subsidence of aseismic ridges.
Dewey, J.F., Burke, K.C.A., 1973. Tibetan, Variscan, and Precambrian basement
reactivation: products of continental collision. The Journal of Geology 81,
683692.
Dickinson, W.R., Snyder, W.S., 1978. Plate tectonics of the Laramide Orogeny.
Memoir Geological Society of America, 355366.
Dorbath, C., Granet, M., 1996. Local earthquake tomography of the Altiplano and
the Eastern Cordillera of northern Bolivia. Tectonophysics 259, 117136.
Douglass, J., Meek, N., Dorn, R.I., Schmeeckle, M.W., 2009. A criteria-based
methodology for determining the mechanism of transverse drainage
development, with application to the southwestern US. Geological Society of
Ducea, M., House, M.A., Kidder, S., 2003. Late Cenozoic denudation and uplift
rates in the Santa Lucia Mountains, California. Geology 31, 139142.
Dugda, M.T., Nyblade, A.A., Julia, J., 2007. Thin lithosphere beneath the Ethiopian
plateau revealed by a joint inversion of rayleigh wave group velocities and
receiver functions. Journal of Geophysical Research 112, B08305.
Dugda, M.T., Nyblade, A.A., Julia, J., Langston, C.A., Ammon, C.J., Simiyu, S.,
2005. Crustal structure in Ethiopia and Kenya from receiver function analysis:
implications for rift development in eastern Africa. Journal of Geophysical
Research 110, B01303.
Dunai, T.J., 2010. Cosmogenic Nuclides Principles, Concepts and Applications in
the Earth Surface Sciences. Cambridge University Press, New York.
Dunai, T.J., Lopez, G.A.G., Juez-Larre, J., 2005. Oligocene-Miocene age of aridity in
the Atacama Desert revealed by exposure dating of erosion-sensitive landforms.
Geology 33, 321324.
Duncan, C., Masek, J., Fielding, E., 2003. How steep are the Himalaya?
Characteristics and implications of along-strike topographic variations. Geology
31, 7578.
Dupont-Nivet, G., Krijgsman, W., Langereis, C.G., Abels, H.A., Dai, S., Fang, X.,
2007. Tibetan plateau aridification linked to global cooling at the
EoceneOligocene transition. Nature 445, 635638.
Duston, N., Owen, R., Wilkinson, B., 1986. Water chemistry and sedimentological
observations in Littlefield Lake, Michigan: implications for lacustrine marl
deposition. Environmental Geology 8, 229236.
Dziewonski, A.M., 1984. Mapping the Lower Mantle: determination of lateral
heterogeneity in P velocity up to degree and order 6. Journal of Geophysical
Research 89, 59295952.
Eaton, G.P., 1987. Topography and Origin of the Southern Rocky Mountains and
Alvarado Ridge. Geological Society, London, Special Publications 28, 355369.
Eaton, G.P., 2008. Epeirogeny in the southern Rocky Mountains region: evidence
and origin. Geosphere 4, 764784.
Ebinger, C.J., Bechtel, T.D., Forsyth, D.W., Bowin, C.O., 1989. Effective elastic plate
thickness beneath the East African and Afar plateaus and dynamic compensation
of the uplifts. Journal of Geophysical Research 94, 28832901.
Effler, S.W., Johnson, D.L., 1987. Calcium carbonate precipitation and turbidity
measurements in Otisco Lake, New York. Journal of the American Water
Resources Association 23, 7379.
Ehlers, T.A., Chaudhri, T., Kumar, S., et al., 2005. Computational tools for lowtemperature thermochronometer interpretation. Reviews in Mineralogy and
Ehlers, T.A., Farley, K.A., Rusmore, M.E., Woodsworth, G.J., 2006. Apatite (U-Th)/
He signal of large-magnitude accelerated glacial erosion, southwest British
Columbia. Geology 34, 765768.
Ehlers, T.A., Poulsen, C.J., 2009. Influence of Andean uplift on climate and
paleoaltimetry estimates. Earth and Planetary Science Letters 281, 238248.
Eiler, J.M., Schauble, E., 2004. 8O13C16O in Earths atmosphere. Geochimica et
England, P., Houseman, G.A., 1989. Extension during continental convergence, with
application to the Tibetan Plateau. Journal of Geophysical Research 94,
17,56117,579.
England, P., Molnar, P., 1990. Surface uplift, uplift of rocks, and exhumation of
rocks. Geology 18, 11731177.
England, P.C., Houseman, G.A., 1988. The mechanics of the Tibetan Plateau. The
Evolution of Passive Continental Margins in the Light of Recent Deep Drilling
Results 326, 301320.
Epis, R.C., Chapin, C.E., 1975. Geomorphic and tectonic implications of the postLaramide Late Eocene erosion surface in the Southern Rocky Mountains.
Geological Society of America Memoirs 144, 4574.
Faras, M., Charrier, R., Comte, D., Martinod, J., Herail, G., 2005. Late Cenozoic
deformation and uplift of the western flank of the Altiplano; evidence from the
depositional, tectonic, and geomorphologic evolution and shallow seismic
activity (northern Chile at 19 degrees 300 S). Tectonics 24, TC4001. http://dx.doi.
org/10.1029/2004TC001667.
Fielding, E., Isacks, B., Barazangi, M., Duncan, C., 1994. How flat is Tibet?
Geology 22, 163167.
Fielding, E.J., 1996. Tibet uplift and erosion. Tectonphysics 260, 5584.
Fleitout, L., Froidevaux, C., 1982. Tectonics and topography for a lithosphere
containing density heterogeneities. Tectonics 1, 2156.
Flint, J.J., 1974. Stream gradient as a function of order, magnitude, and discharge.
Water Resource Research 10, 969973.
Flowers, R.M., Wernicke, B.P., Farley, K.A., 2008. Unroofing, incision, and uplift
history of the southwestern Colorado Plateau from apatite (U-Th)/He
thermochronometry. Geological Society of America Bulletin 120, 571587.
Forest, C.E., 2007. Paleoaltimetry: a review of thermodynamic methods. Reviews in
Mineralogy and Geochemistry 66, 173193.
Forest, C.E., Molnar, P., Emanuel, K.A., 1995. Palaeoaltimetry from energy
conservation principles. Nature 374, 347350.
Forest, C.E., Wolfe, J.A., Molnar, P., Emanuel, K.A., 1999. Paleoaltimetry
incorporating atmospheric physics and botanical estimates of paleoclimate.
Francis, P.W., Hawkesworth, C.J., 1994. Late Cenozoic rates of magmatic activity in
the Central Andes and their relationships to continental crust formation and
thickening. Journal of the Geological Society 151, 845854.
Froidevaux, C., Ricard, Y., 1987. Tectonic evolution of high plateaus.
Fromm, R., Zandt, G., Beck, S.L., 2004. Crustal thickness beneath the Andes and
Sierras Pampeanas at 301 S inferred from Pn apparent phase velocities.
Geophysical Research Letters 31, L06625.
Frostick, L.E., Reid, I.A.N., 1989. Climatic versus tectonic controls of fan
sequences: lessons from the Dead Sea, Israel. Journal of the Geological Society
146, 527538.
Fukao, Y., Yamamoto, A., Kono, M., 1989. Gravity anomaly across the Peruvian
Andes. Journal of Geophysical Research 94, 38673890.
Gale, J., 1972. Availability of carbon dioxide for photosynthesis at high altitudes:
theoretical considerations. Ecology 53, 494497.
Gani, N., Gani, M., Abdelsalam, M., 2007. Blue Nile incision on the Ethiopian
Plateau: pulsed plateau growth, Pliocene uplift, and hominin evolution. GSA
Today 17, 411.
Garzione, C.N., 2008. Surface uplift of Tibet and Cenozoic global cooling. Geology
36, 10031004.
Garzione, C.N., Hoke, G.D., Libarkin, J.C., et al., 2008. Rise of the Andes. Science
320, 13041307.
Garzione, C.N., Molnar, P., Libarkin, J.C., MacFadden, B.J., 2006. Rapid late
Miocene rise of the Bolivian Altiplano: evidence for removal of mantle
lithosphere. Earth and Planetary Science Letters 241, 543556.
Gaupp, R., Kott, A., Worner, G., 1999. Palaeoclimatic implications of Mio-Pliocene
sedimentation in the high-altitude intra-arc Lauca Basin of northern Chile.
Gephart, J.W., 1994. Topography and subduction geometry in the central Andes:
clues to the mechanics of a noncollisional orogen. Journal of Geophysical
Research 99, 1227912288.
Ghosh, P., Adkins, J., Affek, H., et al., 2006a. 13C-18O bonds in carbonate minerals:
a new kind of paleothermometer. Geochimica et Cosmochimica Acta 70,
14391456.
Ghosh, P., Garzione, C.N., Eiler, J.M., 2006b. Rapid uplift of the Altiplano revealed
through 13C-18O bonds in paleosol carbonates. Science 311, 511515.
Giese, P., Scheuber, E., Schilling, F., Schmitz, M., Wigger, P., 1999. Crustal
thickening processes in the Central Andes and the different natures of the
Moho-discontinuity. Journal of South American Earth Sciences 12, 201220.
Gok, R., Michael, E.P., Ekrem, Z., 2007. Lithospheric structure of the
continentcontinent collision zone: eastern Turkey. Geophysical Journal
Gok, R., Sandvol, E., Turkelli, N., Seber, D., Barazangi, M., 2003. Sn attenuation in
the Anatolian and Iranian plateau and surrounding regions. Geophysical
Gosse, J.C., Phillips, F.M., 2001. Terrestrial in situ cosmogenic nuclides: theory
Gotberg, N., McQuarrie, N., Caillaux, V.C., 2010. Comparison of crustal thickening
budget and shortening estimates in southern Peru (12141 S): implications for
mass balance and rotations in the Bolivian orocline. Geological Society of
Gotze, H.J., Lahmeyer, B., Schmidt, S., Strunk, S., 1994. The lithospheric structure
of the Central Andes (2025S) as inferred from quantitative interpretation of
regional gravity. In: Reutter, K., Scheuber, E., Wigger, P. (Eds.), Tectonics of the
Southern Central Andes: Structure and Evolution of an Active Continental
Margin. Springer-Verlag, Berlin, pp. 2348.
Gregory, K.M., 1994. Paleoclimate and paleoelevation of the 35 Ma Florissant flora,
Front Range, Colorado. Palaeoclimates 1, 2357.
Gregory, K.M., Chase, C.G., 1992. Tectonic significance of paleobotanically
estimated climate and altitude of the late Eocene erosion surface, Colorado.
Geology 20, 581585.
Gregory, K.M., Chase, C.G., 1994. Tectonic and climatic significance of a late
Eocene low-relief, high-level geomorphic surface, Colorado. Journal of
Gregory, K.M., McIntosh, W.C., 1996. Paleoclimate and paleoelevation of the
Oligocene Pitch-Pinnacle flora, Sawatch Range, Colorado. Geological Society of
Gregory-Wodzicki, K.M., 1997. The late Eocene House Range flora, Sevier Desert,
Utah; paleoclimate and paleoelevation. Palaios 12, 552567.
Gregory-Wodzicki, K.M., 2000. Uplift history of the Central and Northern Andes: a
review. Geological Society of America Bulletin 112, 10911105.
Griot, D.-A., Montagner, J.-P., Tapponnier, P., 1998. Phase velocity structure from
Rayleigh and Love waves in Tibet and its neighboring regions. Journal of
Gurnis, M., 1993. Phanerozoic marine inundation of continents driven by dynamic
topography above subducting slabs. Nature 364, 589593.
Gurnis, M., Mitrovica, J.X., Ritsema, J., van Heijst, H.-J., 2000. Constraining mantle
density structure using geological evidence of surface uplift rates; the case of
the African Superplume. Geochemistry, Geophysics, Geosystems 1(7), 1020. http
://dx.doi.org/10.1029/1999GC000035.
Gurnis, M., Mueller, R.D., Moresi, L., 1998. Cretaceous vertical motion of Australia
and the Australian-Antarctic discordance. Science 279, 14991504.
Hack, J.T., 1957. Studies of longitudinal stream profiles in Virginia and Maryland.
United States Geological Survey Professional Paper 294, 4594.
123
Hager, B.H., Clayton, R.W., Richards, M.A., Comer, R.P., Dziewonski, A.M., 1985.
Lower mantle heterogeneity, dynamic topography and the geoid. Nature 313,
541545.
Haines, S.S., Klemperer, S.L., Brown, L., et al., 2003. INDEPTH III seismic data:
from surface observations to deep crustal processes in Tibet. Tectonics 22,
1001.
Harrison, T.M., Copeland, P., Kidd, W.S.F., Yin, A., 1992. Raising Tibet. Science
255, 16631670.
Hartley, A., 2003. Andean uplift and climate change. Journal of the Geological
Society 160, 710.
Heit, B., Yuan, X.H., Bianchi, M., Sodoudi, F., Kind, R., 2008. Crustal thickness
estimation beneath the southern central Andes at 301 S and 361 S from S
wave receiver function analysis. Geophysical Journal International 174,
249254.
Hennig, A., 1915. Zur Petrographie und geologie von Sudwest Tibet. In: Hedin, S.
(Ed.), Southern Tibet. Norstedt, Stockholm, 220 pp.
Hirn, A., Lepine, J.-C., Jobert, G., et al., 1984a. Crustal structure and variability of
the Himalayan border of Tibet. Nature (London) 307, 2325.
Hirn, A., Nercessian, A., Sapin, M., et al., 1984b. Lhasa Block and bordering
sutures; a continuation of a 500-km Moho traverse through Tibet. Nature
(London) 307, 2527.
Hodges, K.V., Wobus, C., Ruhl, K., Schildgen, T., Whipple, K., 2004. Quaternary
deformation, river steepening, and heavy precipitation at the front of the
Higher Himalayan ranges. Earth and Planetary Science Letters 220,
379389.
Hofmann, C., Courtillot, V., Feraud, G., Rochette, P., Yirgu, G., Ketefo, E., Pik, R.,
1997. Timing of the Ethiopian flood basalt event and implications for plume
birth and global change. Nature 389, 838841.
Hoke, G.D., Garzione, C.N., 2008. Paleosurfaces, paleoelevation, and the
mechanisms for the late Miocene topographic development of the Altiplano
plateau. Earth and Planetary Science Letters 271, 192201.
Hoke, G.D., Garzione, C.N., Araneo, D.C., Latorre, C., Strecker, M.R., Williams, K.J.,
2009. The stable isotope altimeter: do Quaternary pedogenic carbonates predict
modern elevations? Geology 37, 10151018.
Hoke, G.D., Isacks, B.L., Jordan, T.E., Blanco, N., Tomlinson, A.J., Ramezani, J.,
2007. Geomorphic evidence for post-10 Ma uplift of the western flank of the
central Andes 81300 -221 S. Tectonics 26, TC5021.
Hoke, L., Hilton, D.R., Lamb, S.H., Hammerschmidt, K., Friedrichsen, H., 1994. 3He
evidence for a wide zone of active mantle melting beneath the Central Andes.
Holt, W.E., Wallace, T.C., 1990. Crustal thickness and upper mantle velocities in the
Tibetan plateau region from the inversion of regional Pnl waveforms: evidence
for a thick upper mantle lid beneath Southern Tibet. Journal of Geophysical
Research 95, 1249912525.
Horton, B.K., 1999. Erosional control on the geometry and kinematics of thrust belt
development in the Central Andes. Tectonics 18, 12921304.
House, M.A., Wernicke, B.P., Farley, K.A., 1998. Dating topography of the Sierra
Nevada, California, using apatite (U-Th)/He ages. Nature 396, 6669.
House, M.A., Wernicke, B.P., Farley, K.A., 2001. Paleo-Geomorphology of the Sierra
Nevada, California, from (U-TH) /He Ages in Apatite. American Journal of
Science 301, 77102.
Houseman, G.A., McKenzie, D.P., Molnar, P., 1981. Convective instability of a
thickened boundary layer and its relevance for the thermal evolution of
continental convergent belts. Journal of Geophysical Research 86, 61156132.
Howard, A.D., Kerby, G., 1983. Channel changes in badlands. GSA Bulletin 94,
739752.
Huerta, A.D., Nyblade, A.A., Reusch, A.M., 2009. Mantle transition zone structure
beneath Kenya and Tanzania: more evidence for a deep-seated thermal upwelling
in the mantle. Geophysical Journal International 177, 12491255.
Hunt, C.B., 1956. Cenozoic geology of the Colorado Plateau. U.S. Geological
Survey Professional Paper 279, 99.
Huntington, K.W., Ehlers, T.A., Hodges, K.V., Whipp, Jr. D.M., 2007. Topography,
exhumation pathway, age uncertainties, and the interpretation of
thermochronometer data. Tectonics 26, TC4012.
Husson, L., Sempere, T., 2003. Thickening the Altiplano crust by gravity-driven
crustal channel flow. Geophysical Research Letters 30, 1243.
Isacks, B.L., 1988. Uplift of the Central Andean plateau and bending of the Bolivian
Orocline. Journal of Geophysical Research 93, 32113231.
James, D.E., Fouch, M.J., 2002. Formation and evolution of Archaean cratons;
insights from Southern Africa. Coastal tectonics 199, 126.
Jarvis, A., Reuter, H.I., Nelson, A., Guevara, E., 2008. Hole-filled SRTM for the
globe Version 4. Available from the CGIAR-CSI SRTM 90 m: http://
srtm.csi.cgiar.org.
124
Johnson, M.R.W., 2002. Shortening budgets and the role of continental subduction
during the IndiaAsia collision. Earth-Science Reviews 59, 101123.
Jones, H., 1992. Plants and Microclimate: A Quantitative Approach to
Environmental Plant Physiology. Cambridge University Press, New York,
456 pp.
Kennan, L., 2000. Large-scale geomorphology of the Andes; interrelationships of
tectonics, magmatism and climate. In: Summerfield, M.A. (Ed.), Geomorphology
and Global Tectonics. Wiley, Chichester, pp. 167199.
Kennan, L., Lamb, S.H., Hoke, L., 1997. High altitude palaeosurfaces in the Bolivian
Andes: evidence for late Cenozoic surface uplift. In: Widdowson, M. (Ed.),
Palaeosurfaces: Recognition, Reconstruction, and Interpretation. Geological
Society of London Special Publications, Bath, UK, pp. 307324.
Keskin, M., 2003. Magma generation by slab steepening and breakoff beneath a
subduction-accretion complex: an alternative model for collision-related
volcanism in Eastern Anatolia. Turkey. Geophysical Research Letters 30, 8046.
Ketcham, R.A., Carlson, W.D., 2001. Acquisition, optimization, and interpretation of
X-ray computed tomographic imagery: applications to the geosciences.
Computers & Geosciences 27, 381400.
Kind, R., Yuan, X., Saul, J., et al., 2002. Seismic images of Crust and Upper
Mantle beneath Tibet: evidence for Eurasian Plate subduction. Science 298,
12191221.
King, L.C., 1953. Canons of landscape evolution. Geological Society of America
Kirby, E., Reiners, P.W., Krol, M.A., et al., 2002. Late Cenozoic evolution of the
eastern margin of the Tibetan Plateau: inferences from 40Ar/39Ar and (U-Th)/He
thermochronology. Tectonics 21, 1001.
Kirby, E., Whipple, K.X., Tang, W., Chen, Z., 2003. Distribution of active rock
uplift along the eastern margin of the Tibetan Plateau: inferences from
bedrock channel longitudinal profiles. Journal of Geophysical Research 108,
2217.
Kono, M., Fukao, Y., Yamamoto, A., 1988. Mountain building in the Central Andes.
Rock Magnetism and Paleogeophysics 15, 7380.
Kono, M., Fukao, Y., Yamamoto, A., 1989. Mountain Building in the Central Andes.
Kosarev, G., Kind, R., Sobolev, S.V., Yuan, X., Hanka, W., Oreshin, S., 1999.
Seismic evidence for a detached Indian lithospheric Mantle beneath Tibet.
Science 283, 13061309.
Koulakov, I., Sobolev, S.V., Asch, G., 2006. P- and S-velocity images of the
lithosphereasthenosphere system in the Central Andes from local-source
tomographic inversion. Geophysical Journal International 167, 106126.
Kouwenberg, L.L.R., Kurschner, W.M., McElwain, J.C., 2007. Stomatal frequency
change over altitudinal gradients: prospects for paleoaltimetry. Reviews in
Kutzbach, J.E., Guetter, P.J., Ruddiman, W.F., Prell, W.L., 1989. Sensitivity of climate
to Late Cenozoic uplift in southern Asia and the American West: numerical
experiments. Journal of Geophysical Research 94, 1839318407.
Lal, D., Peters, B., 1967. Cosmic-Ray Produced Radioactivity on the Earth,
Handbook of Physics 46. Springer-Verlag, Berlin, pp. 551662.
Lamb, S., Hoke, L., 1997. Origin of the high plateau in the Central Andes, Bolivia,
South America. Tectonics 16, 623649.
Lamb, S., Hoke, L., Kennan, L., Dewey, J., 1997. Cenozoic evolution of the Central
Andes in Bolivia and northern Chile. Coastal Tectonics 121, 237264.
Leonard, E.M., 2002. Geomorphic and tectonic forcing of late Cenozoic warping of
the Colorado piedmont. Geology 30, 595598.
Li, C., van der Hilst, R.D., Meltzer, A.S., Engdahl, E.R., 2008. Subduction of the
Indian lithosphere beneath the Tibetan Plateau and Burma. Earth and Planetary
Li, S., Mooney, W.D., 1998. Crustal structure of China from deep seismic sounding
profiles. Tectonophysics 288, 105113.
Libarkin, J.C., Quade, J., Chase, C.G., Poths, J., McIntosh, W., 2002. Measurement
of ancient cosmogenic 21Ne in quartz from the 28 Ma Fish Canyon Tuff,
Colorado. Chemical Geology 186, 199213.
Lithgow-Bertelloni, C., Silver, P.G., 1998. Dynamic topography, plate driving forces
and the African superswell. Nature (London) 395, 269272.
Liu, B., Phillips, F.M., Campbell, A.R., 1996. Stable carbon and oxygen isotopes of
pedogenic carbonates, Ajo Mountains, southern Arizona: implications for
paleoenvironmental change. Palaeogeography, Palaeoclimatology, Palaeoecology
124, 233246.
Liu, L., Gurnis, M., 2010. Dynamic subsidence and uplift of the Colorado Plateau.
Geology 38, 663666.
Liu-Zeng, J., Tapponnier, P., Gaudemer, Y., Ding, L., 2008. Quantifying landscape
differences across the Tibetan plateau: implications for topographic relief
evolution. Journal of Geophysical Research 113, F04018.
Mackenzie, G.D., Thybo, H., Maguire, P.K.H., 2005. Crustal velocity structure across
the Main Ethiopian Rift: results from two-dimensional wide-angle seismic
modeling. Geophysical Journal International 162, 9941006.
Maggi, A., Priestley, K., 2005. Surface waveform tomography of the Turkish-Iranian
plateau. Geophysical Journal International 160, 10681080.
Mancktelow, N.S., Grasemann, B., 1997. Time-dependent effects of heat advection
and topography on cooling histories during erosion. Tectonophysics 270,
167195.
Masek, J.G., Isacks, B.L., Gubbels, T.L., Fielding, E.J., 1994. Erosion and tectonics
at the margins of continental plateaus. Journal of Geophysical Research 99,
13,94113,956.
McElwain, J.C., 2004. Climate-independent paleoaltimetry using stomatal
density in fossil leaves as a proxy for CO2 partial pressure. Geology 32,
10171020.
McElwain, J.C., Mayle, F.E., Beerling, D.J., 2002. Stomatal evidence for a decline in
atmospheric CO2 concentration during the Younger Dryas stadial: a comparison
with Antarctic ice core records. Journal of Quaternary Science 17, 2129.
McGlashan, N., Brown, L., Kay, S., 2008. Crustal thickness in the central Andes
from teleseismically recorded depth phase precursors. Geophysical Journal
McMillan, M.E., Angevine, C.L., Heller, P.L., 2002. Postdepositional tilt of the
MiocenePliocene Ogallala Group on the western Great Plains: evidence of late
Cenozoic uplift of the Rocky Mountains. Geology 30, 6366.
McMillan, M.E., Heller, P.L., Wing, S.L., 2006. History and causes of post-Laramide
relief in the Rocky Mountain orogenic plateau. Geological Society of America
Bulletin 118, 393405.
McNamara, D.E., Owens, T.J., Walter, W.R., 1995. Observations of regional phase
propagation across the Tibetan Plateau. Journal of Geophysical Research 100,
22,21522,229.
McNamara, D.E., Walter, W.R., Owens, T.J., Ammon, C.J., 1997. Upper mantle
velocity structure beneath the Tibetan Plateau from Pn travel time tomography.
McQuarrie, N., Chase, C.G., 2000. Raising the Colorado Plateau. Geology 28,
9194.
Menard, G., Molnar, P., 1988. Collapse of a Hercynian Tibetan Plateau into a late
Palaeozoic European Basin and Range province. Nature 334, 235237.
Meyer, H.W., 1992. Lapse rates and other variables applied to estimating
paleoaltitudes from fossil floras. Palaeogeography, Palaeoclimatology,
Meyer, H.W., 2007. A Review of paleotemperature lapse rate methods for estimating
paleoelevation from fossil floras. Reviews in Mineralogy and Geochemistry 66,
155171.
Molnar, P., 2005. Mio-pliocene growth of the Tibetan Plateau and evolution of East
Asian climate. Palaeontologia Electronica 8(1), 2A:23 pp.
Molnar, P., Houseman, G.A., England, P.C., 2006. Palaeo-altimetry of Tibet. Nature
444, E4.
Molnar, P., Tapponnier, P., 1978. Active tectonics of Tibet. Journal of Geophysical
Research 83, 53615375.
Montgomery, D.R., Gran, K.B., 2001. Downstream variations in the width of bedrock
channels. Water Resource Research 37, 18411846.
Moore, A., Blenkinsop, T., 2002. The role of mantle plumes in the development of
continental-scale drainage patterns: the southern African example revisited.
Mortimer, C., 1973. The Cenozoic history of the southern Atacama Desert, Chile.
Journal of the Geological Society 129, 505526.
Moucha, R., Forte, A.M., Rowley, D.B., Mitrovica, J.X., Simmons, N.A., Grand, S.P.,
2009. Deep mantle forces and the uplift of the Colorado Plateau. Geophysical
Myers, S.C., Beck, S., Zandt, G., Wallace, T., 1998. Lithospheric-scale structure
across the Bolivian Andes from tomographic images of velocity and attenuation
for P and S waves. Journal of Geophysical Research 103, 2123321252.
Nabelek, J., Hetenyi, G., Vergne, J., et al., 2009. Underplating in the Himalaya-Tibet
Collision Zone Revealed by the Hi-CLIMB Experiment. Science 325,
13711374.
beneath southern Tibet; synthesis of Project INDEPTH results. Science 274,
16841688.
Neuendorf, K.K.E., Mehl, Jr. J.P., Jackson, J.A., 2005. Glossary of Geology.
Alexandria, Virginia, 779 pp.
Nguuri, T.K., Gore, J., James, D.E., et al., 2001. Crustal structure beneath southern
Africa and its implications for the formation and evolution of the Kaapvaal and
Zimbabwe cratons. Geophysical Research Letters 28, 25012504.
Ni, J., Barazangi, M., 1983. High-frequency seismic wave propagation beneath
the Indian Shield, Himalayan Arc, Tibetan Plateau and surrounding regions;
high uppermost mantle velocities and efficient Sn propagation beneath
Tibet. Geophysical Journal of the Royal Astronomical Society 72,
665689.
Nicoll, K., 2010. Landscape development within a young collision
zone; implications for post-Tethyan evolution of the upper Tigris River system
in southeastern Turkey. International Geology Review 52, 404422.
Norin, E., 1946. Geological explorations in western Tibet. In: Hedin, S. (Ed.),
Reports from the Scientific Expedition to the Northwestern Provinces of China
Under the Leadership of Dr. Sven Hedin. Tryckeri Aktiebolaget, Thule,
Stockholm, 214.
Nyblade, A.A., Owens, T.J., Gurrola, H., Ritsema, J., Langston, C.A., 2000. Seismic
evidence for a deep upper mantle thermal anomaly beneath east Africa. Geology
28, 599602.
Nyblade, A.A., Pollack, H.N., Jones, D.L., Podmore, F., Mushayandebvu, M., 1990.
Terrestrial heat flow in East and southern Africa. Journal of Geophysical
Research 95, 1737117384.
Nyblade, A.A., Robinson, S.W., 1994. The African superswell. Geophysical Research
Letters 21, 765768.
Ollier, C.D., 1981. Tectonics and Landforms. Longman, London and New York, 324
pp.
Ollier, C.D., 1985. Morphotectonics of passive continental margins: introduction.
Zeitschrift fur Geomorphologie, Supplementbande 54, 19.
Ollier, C.D., Marker, M.E., 1985. The great escarpment of southern Africa. Zeitschrift
fuer Geomorphologie Supplementband 54, 3756.
Olson, P., Nam, I.S., 1986. Formation of seafloor swells by mantle plumes. Journal
Owens, T.J., Zandt, G., 1997. Implications of crustal property variations for models
of Tibetan plateau evolution. Nature 387, 3743.
Paola, C., Mohrig, D., 1996. Palaeohydraulics revisited: palaeoslope estimation in
coarse-grained braided rivers. Basin Research 8, 243254.
Park, Y., Nyblade, A.A., 2006. P-wave tomography reveals a westward dipping
low velocity zone beneath the Kenya Rift. Geophysical Research Letters 33,
L07311.
Pearce, J.A., Bender, J.F., De Long, S.E., et al., 1990. Genesis of collision
volcanism in Eastern Anatolia, Turkey. Journal of Volcanology and Geothermal
Pederson, J.L., Mackley, R.D., Eddleman, J.L., 2002. Colorado Plateau uplift and
erosion evaluated using GIS. GSA Today 12, 410.
Peppe, D.J., Royer, D.L., Wilf, P., Kowalski, E.A., 2010. Quantification of large
uncertainties in fossil leaf paleoaltimetry. Tectonics 29, TC3015.
Pik, R., Marty, B., Carignan, J., Lave, J., 2003. Stability of the upper Nile drainage
network (Ethiopia) deduced from (U-Th)/He thermochronometry; implications for
uplift and erosion of the Afar Plume dome. Earth and Planetary Science Letters
215, 7388.
Poulsen, C.J., Ehlers, T.A., Insel, N., 2010. Onset of convective rainfall during
gradual late Miocene rise of the Central Andes. Science 328, 490493.
Powell, J.W., 1875. Exploration of the Colorado River of the West and its
Tributaries. Publisher unknown, Washington, DC.
Press, F., Siever, R., Grotzinger, J., Jordan, T.H., 2004. Understanding Earth. W. H.
Freeman and Company, New York, NY.
Proussevitch, A.A., Sahagian, D.L., 2001. Recognition and separation of discrete
objects within complex 3D voxelized structures. Computers & Geosciences 27,
441454.
Proussevitch, A.A., Sahagian, D.L., Carlson, W.D., 2007. Statistical analysis
of bubble and crystal size distributions: Application to Colorado Plateau
basalts. Journal of Volcanology and Geothermal Research 164,
112126.
Quade, J., Garzione, C., Eiler, J., 2007. Paleoelevation Reconstruction using
Pedogenic Carbonates. Reviews in Mineralogy and Geochemistry 66,
5387.
Raymo, M.E., 1991. Geochemical evidence supporting chamberlin,T.,C. theory of
glaciation. Geology 19, 344347.
Raymo, M.E., Ruddiman, W.F., 1992. Tectonic forcing of late Cenozoic climate.
Nature (London) 359, 117122.
Rech, J.A., Currie, B.S., Michalski, G., Cowan, A.M., 2006. Neogene climate change
and uplift in the Atacama Desert, Chile. Geology 34, 761764.
Reiners, P.W., 2007. Thermochronologic approaches to paleotopography. Reviews in
125
Reiners, P.W., Ehlers, T.A., 2005. Low-Temperature Thermochronology; Techniques,

Interpretations, and Applications. Mineralogical Society of America, Alexandria
Virginia, 622 pp.
Reiners, P.W., McPhillips, D., Brandon, M.T., Mulch, A., Chamberlain, C.P., 2006.
Thermochronologic approaches to paleotopography. Geochimica et
Cosmochimica Acta 70, A525.
Reiners, P.W., Zhou, Z., Ehlers, T.A., Xu, C., Brandon, M.T., Donelick, R.A.,
Nicolescu, S., 2003. Postorogenic evolution of the Dabie Shan, eastern China,
from (U-Th)/He and fission-track thermochronology. American Journal of
Science 303, 489518.
Rigby, J.K., 1977. Southern Colorado Plateau, K/H Geology Field Guide Studies.
Kendall/Hunt Pub. Co., Dubuque, Iowa, pp. 148.
Riihimaki, C.A., Anderson, R.S., Safran, E.B., Dethier, D.P., Finkel, R.C., Bierman,
P.R., 2006. Longevity and progressive abandonment of the Rocky Flats
surface, Front Range, Colorado. Cosmogenic Isotopes in Geomorphology 78,
265278.
Riihimaki, C.A., Libarkin, J.C., 2007. Terrestrial cosmogenic nuclides as
paleoaltimetric proxies. Reviews in Mineralogy and Geochemistry 66,
269278.
Rind, D., Chandler, M., 1991. Increased ocean heat transports and warmer climate.
Riquelme, R., Martinod, J., Herail, G., Darrozes, J., Charrier, R., 2003. A
geomorphological approach to determining the Neogene to Recent tectonic
deformation in the Coastal Cordillera of northern Chile (Atacama).
Ritsema, J., Heijst, H.J.V, Woodhouse, J.H., 1999. Complex shear wave
velocity structure imaged beneath Africa and Iceland. Science 286,
19251928.
Ritsema, J., Nyblade, A.A., Owens, T.J., Langston, C.A., VanDecar, J.C., 1998.
Upper mantle seismic velocity structure beneath Tanzania, east Africa:
implications for the stability of cratonic lithosphere. Journal of Geophysical
Research 103, 2120121213.
Ritsema, J., van Heijst, H., 2000. New seismic model of the upper mantle beneath
Africa. Geology 28, 6366.
Roberge, J., Wallace, P.J., White, R.V., Coffin, M.F., 2005. Anomalous uplift and
subsidence of the Ontong Java Plateau inferred from CO2 contents of submarine
basaltic glasses. Geology 33, 501504.
Rowley, D.B., 2007. Stable isotope-based paleoaltimetry: theory and validation.
Reviews in Mineralogy and Geochemistry 66, 2352.
Rowley, D.B., Currie, B.S., 2006. Palaeo-altimetry of the late Eocene to Miocene
Lunpola basin, central Tibet. Nature 439, 677681.
Rowley, D.B., Garzione, C.N., 2007. Stable isotope-based paleoaltimetry. Annual
Rowley, D.B., Pierrehumbert, R.T., Currie, B.S., 2001. A new approach to stable
isotope-based paleoaltimetry: implications for paleoaltimetry and
paleohypsometry of the High Himalaya since the Late Miocene. Earth and
Roy, M., Jordan, T.H., Pederson, J., 2009. Colorado Plateau magmatism and
uplift by warming of heterogeneous lithosphere. Nature (London) 459,
978982.
Roy, M., Kelley, S., Pazzaglia, F., Cather, S., House, M., 2004. Middle Tertiary
buoyancy modification and its relationship to rock exhumation, cooling, and
subsequent extension at the eastern margin of the Colorado Plateau. Geology
32, 925928.
Ruddiman, W.F., 1997. Tectonic Uplift and Climate Change. Plenum Press, New
York, NY.
Ruddiman, W.F., Kutzbach, J.E., 1991. Plateau uplift and climatic change. Scientific
American 264(6672), 7475.
Ruddiman, W.F., Prell, W.L., 1997. Introduction to the Uplift-Climate Connection.
Plenum Press, New York, NY.
Ruddiman, W.F., Prell, W.L., Raymo, M.E., 1989. Late Cenozoic Uplift in Southern
Asia and the American West: Rationale for General Circulation Modeling
Experiments. Journal of Geophysical Research 94, 1837918391.
Sahagian, D., Proussevitch, A., 2007. Paleoelevation measurement on the basis of
vesicular basalts. Reviews in Mineralogy and Geochemistry 66, 195213.
Sahagian, D., Proussevitch, A., Carlson, W., 2002a. Timing of Colorado Plateau
uplift: initial constraints from vesicular basalt-derived paleoelevations. Geology
30, 807810.
Sahagian, D.L., Anderson, A.T., Ward, B., 1989. Bubble coalescence in basalt flows:
comparison of a numerical model with natural examples. Bulletin of Volcanology
52, 4956.
Sahagian, D.L., Maus, J.E., 1994. Basalt vesicularity as a measure of atmospheric
pressure and palaeoelevation. Nature 372, 449451.
126
Sahagian, D.L., Proussevitch, A.A., Carlson, W.D., 2002b. Analysis of

vesicular basalts and lava emplacement processes for application
as a paleobarometer/paleoaltimeter. The Journal of Geology 110,
671685.
Schafer, J.M., Ivy-Ochs, S., Wieler, R., Leya, I., Baur, H., Denton, G.H., Schluchter,
C., 1999. Cosmogenic noble gas studies in the oldest landscape on earth:
surface exposure ages of the Dry Valleys, Antarctica. Earth and Planetary
Schildgen, T.F., Balco, G., Shuster, D.L., 2010. Canyon incision and knickpoint
propagation recorded by apatite 4He/3He thermochronometry. Earth and
Schildgen, T.F., Hodges, K.V., Whipple, K.X., Reiners, P.W., Pringle, M.S., 2007.
Uplift of the western margin of the Andean plateau revealed from canyon
incision history, southern Peru. Geology 35, 523526.
Schlunegger, F., Zeilinger, G., Kounov, A., Kober, F., Husser, B., 2006. Scale of
relief growth in the forearc of the Andes of Northern Chile (Arica latitude,
181 S). Terra Nova 18, 217223.
Schmitz, M., 1994. A balanced model of the southern Central Andes. Tectonics 13,
484492.
Schoenbohm, L.M., Whipple, K.X., Burchfiel, B.C., Chen, L., 2004. Geomorphic
constraints on surface uplift, exhumation, and plateau growth in the Red River
region, Yunnan Province, China. Geological Society of America Bulletin 116,
895909.
Schulte-Pelkum, V., Monsalve, G., Sheehan, A., Pandey, M.R., Sapkota, S., Bilham,
R., Wu, F., 2005. Imaging the Indian subcontinent beneath the Himalaya. Nature
(London) 435, 12221225.
S- engor, A.M.C., Ozeren, M.S., Keskin, M., Sakinc, M., Ozbakir, A.D., Kayan, I.,
2008. Eastern Turkish high plateau as a small turkic-type orogen; implications
for postcollisional crust-forming processes in turkic-type orogens. Earth-Science
Reviews 90, 148.
Servant, M., Sempere, T., Argollo, J., Bernat, M., Feraud, G., Lo Bello, P., 1989.
Morphogenese et soulevement de la Cordillere Orientale des Andes de Bolivie
au Cenozoique. Comptes Rendus de lAcademie des Sciences, Serie 2(309),
416422.
Sheffels, B.M., 1990. Lower bound on the amount of crustal shortening, in the
central Bolivian Andes. Geology 18, 812815.
Shi, D., Zhao, W., Brown, L., et al., 2004. Detection of southward intracontinental
subduction of Tibetan lithosphere along the BangongNujiang suture by P-to-S
converted waves. Geology 32, 209212.
Simmons, N.A., Forte, A.M., Grand, S.P., 2007. Thermochemical structure
and dynamics of the African superplume. Geophysical Research Letters 34,
L02301.
Sine, C.R., Wilson, D., Gao, W., Grand, S.P., Aster, R., Ni, J., Baldridge, W.S., 2008.
Mantle structure beneath the western edge of the Colorado Plateau. Geophysical
Snyder, N.P., Whipple, K.X., Tucker, G.E., Merritts, D.J., 2000. Landscape response
to tectonic forcing: digital elevation model analysis of stream profiles in the
Mendocino triple junction region, northern California. Geological Society of
Sobel, E.R., Hilley, G.E., Strecker, M.R., 2003. Formation of internally drained
contractional basins by aridity-limited bedrock incision. Journal of Geophysical
Research 108, 2344.
Sonder, L.J., England, P.C., 1989. Effects of a temperature-dependent rheology on
large-scale continental extension. Journal of Geophysical Research 94,
76037619.
Sonder, L.J., Jones, C.H., 1999. Western United States extension: how the
west was widened. Annual Review of Earth and Planetary Sciences 27,
417462.
Spencer, J.E., 1996. Uplift of the Colorado Plateau due to lithosphere attenuation
during Laramide low-angle subduction. Journal of Geophysical Research 101,
13,59513,609.
Spicer, R.A., Harris, N.B.W., Widdowson, M., et al., 2003. Constant elevation
of southern Tibet over the past 15 million years. Nature 421, 622
624.
Stanistreet, I.G., McCarthy, T.S., 1993. The Okavango fan and the classification of
subaerial fan systems. Sedimentary Geology 85, 115133.
Stark, C.P., Stewart, J., Ebinger, C.J., 2003. Wavelet transform mapping of effective
elastic thickness and plate loading: Validation using synthetic data and
application to the study of southern African tectonics. Journal of Geophysical
Research 108, 2558.
Stock, G.M., Ehlers, T.A., Farley, K.A., 2006. Where does sediment come from?
Quantifying catchment erosion with detrital apatite (U-Th)/He thermochronometry.
Geology 34, 725728.
Stockli, D.F., 2006. Thermochronometric constraints on paleoaltimetry and

paleotopography case studies from the Colorado Plateau, Tibet, and Labrador.
Geochimica et Cosmochimica Acta 70, A617.
Strecker, M.R., Alonso, R.N., Bookhagen, B., Carrapa, B., Hilley, G.E., Sobel, E.R.,
Trauth, M.H., 2007. Tectonics and climate of the southern Central Andes. Annual
Stuwe, K., White, L., Brown, R., 1994. The influence of eroding topography on
steady-state isotherms. Application to fission track analysis. Earth and Planetary
Summerfield, M.A., 1991. Global Geomorphology. Prentice Hall Longman/Wiley,
London/New York.
Sutherland, R., Spasojevic, S., Gurnis, M., 2010. Mantle upwelling after
Gondwana subduction death explains anomalous topography and
subsidence histories of eastern New Zealand and West Antarctica. Geology 38,
155158.
Swenson, J.L., Beck, S.L., Zandt, G., 2000. Crustal structure of the Altiplano from
broadband regional waveform modeling: Implications for the composition of
thick continental crust. Journal of Geophysical Research 105, 607621.
Tabatabai Mir, S., Bergman, E., Gheitanchi, M.R., 2008. 3-Dimensional upper
mantle velocity structure for Iranian Plateau revealed by Pn and Sn tomography.
Journal of the Earth and Space Physics 33, 1324.
Tapponnier, P., Zhiqin, X., Roger, F., Meyer, B., Arnaud, N., Wittlinger, G., Jingsui,
Y., 2001. Oblique stepwise rise and growth of the Tibet Plateau. Science 294,
16711677.
Teixell, A., Ayarza, P., Zeyen, H., Fernandez, M., Arboleya, M.-L., 2005. Effects of
mantle upwelling in a compressional setting: the Atlas Mountains of Morocco.
Terra Nova 17, 456461.
ten Brink, U., Stern, T., 1992. Rift flank uplifts and Hinterland Basins: comparison
of the transantarctic mountains with the great escarpment of southern Africa.
Thorpe, R.S., Francis, P.W., Harmon, R.S., Anonymous., 1981. Andean andesites
and crustal growth. The Evolution of Passive Continental Margins in the Light of
Recent Deep Drilling Results 301, 305320.
Thouret, J.C., Worner, G., Gunnell, Y., Singer, B., Zhang, X., Souriot, T., 2007.
Geochronologic and stratigraphic constraints on canyon incision and Miocene
uplift of the Central Andes in Peru. Earth and Planetary Science Letters 263,
151166.
Tilmann, F., Ni, J., Team, I.I.S., 2003. Seismic imaging of the downwelling Indian
lithosphere beneath Central Tibet. Science 300, 14241427.
Turkelli, N., Sandvol, E., Zor, E., et al., 2003. Seismogenic zones in Eastern Turkey.
Turner, S., Hawkesworth, C., Liu, J., Rogers, N., Kelley, S., van Calsteren, P., 1993.
Timing of Tibetan uplift constrained by analysis of volcanic rocks. Nature
(London) 364, 5054.
Valla, P.G., Herman, F., van der Beek, P.A., Braun, J., 2010. Inversion of
thermochronological age-elevation profiles to extract independent estimates of
denudation and relief history I: Theory and conceptual model. Earth and
Van der Wateren, F.M., Dunai, T.J., 2001. Late Neogene passive margin denudation
history cosmogenic isotope measurements from the central Namib desert.
Global and Planetary Change 30, 271307.
Vandervoort, D.S., Jordan, T.E., Zeitler, P.K., Alonso, R.N., 1995. Chronology of
internal drainage development and uplift, southern Puna plateau, Argentine
central Andes. Geology 23, 145148.
Vassallo, R., Ritz, J.F., Braucher, R., et al., 2007. Transpressional tectonics and
stream terraces of the Gobi-Altay, Mongolia. Tectonics 26, TC5013.
Wang, X., Ni, J.F., Aster, R., et al., 2008. Shear-wave splitting and mantle flow
beneath the Colorado Plateau and Its boundary with the Great Basin. Bulletin of
Wang, Z., Schauble, E.A., Eiler, J.M., 2004. Equilibrium thermodynamics of multiply
substituted isotopologues of molecular gases. Geochimica et Cosmochimica
Acta 68, 47794797.
Wdowinski, S., Bock, Y., 1994. The evolution of deformation and topography of
high elevated plateaus 2. Application to the central Andes. Journal of
Weeraratne, D.S., Forsyth, D.W., Fischer, K.M., Nyblade, A.A., 2003. Evidence for an
upper mantle plume beneath the Tanzanian craton from Rayleigh wave
tomography. Journal of Geophysical Research 108, 2427.
Wernicke, B.P., Christiansen, R.L., England, P.C., Sonder, L.J., 1987.
Tectonomagmatic evolution of Cenozoic extension in the North American
Cordillera. Coastal tectonics 28, 203221.
West, M., Ni, J., Baldridge, W.S., Wilson, D., Aster, R., Gao, W., Grand, S., 2004.
Crust and upper mantle shear wave structure of the southwest United States:
Implications for rifting and support for high elevation. Journal of Geophysical
Research 109, B03309.
Whipple, K.X., 2001. Fluvial landscape response time: how plausible is steady-state
denudation? American Journal of Science 301, 313325.
Whipple, K.X., 2004. Bedrock rivers and the geomorphology of active orogens.
belts. Nature Geosci 2, 97104.
Whipple, K.X., Tucker, G.E., 1999. Dynamics of the stream-power river incision
model: implications for height limits of mountain ranges, landscape response
timescales, and research needs. Journal of Geophysical Research 104,
17,66117,674.
Whitman, D., Isacks, B.L., Chatelain, J.-L., Chiu, J.-M., Perez, A., 1992. Attenuation
of high-frequency seismic waves beneath the Central Andean Plateau. Journal of
Whitman, D., Isacks, B.L., Kay, S.M., 1996. Lithospheric structure and along-strike
segmentation of the Central Andean Plateau; seismic Q, magmatism, flexure,
topography and tectonics. Tectonophysics 259, 2940.
Whittaker, A.C., Cowie, P.A., Attal, M., Tucker, G.E., Roberts, G.P., 2007. Bedrock
channel adjustment to tectonic forcing: implications for predicting river incision
Widdowson, M., 1997. The geomorphological and geological importance of
palaeosurfaces. Coastal Tectonics 120, 112.
Wigger, P., Baldzuhn, S., Giese, P., et al., 1994. Variations of the Crustal structure
of the southern Central Andes deduced from seismic refraction investigations.
In: Reuter, H.J., Scheuber, E., Wigger, P.J. (Eds.), Tectonics of the Southern
Central Andes: Structure and Evolution of an Active Continental Margin.
Wilson, D., Aster, R., Ni, J., et al., 2005. Imaging the seismic structure of the
crust and upper mantle beneath the Great Plains, Rio Grande Rift, and
Colorado Plateau using receiver functions. Journal of Geophysical Research 110,
B05306.
Wittlinger, G., Masson, F., Poupinet, G., et al., 1996. Seismic tomography of
northern Tibet and Kunlun; evidence for crustal blocks and mantle velocity
contrasts. Earth and Planetary Science Letters 139, 263279.
Wittlinger, G., Vergne, J., Tapponnier, P., et al., 2004. Teleseismic imaging of
subducting lithosphere and Moho offsets beneath western Tibet. Earth and
Wobus, C., Whipple, K.X., Kirby, E., et al. , 2006.Tectonics from topography;
procedures, promise, and pitfalls. Special Paper Geological Society of America
398, pp. 5574.
Wobus, C.W., Hodges, K.V., Whipple, K.X., 2003. Has focused denudation sustained
active thrusting at the Himalayan topographic front? Geology 31, 861864.
Wolfe, J.A., 1992a. An analysis of present-day terrestrial lapse rates in the western
conterminous United States and their significance to paleoaltitudinal estimates.
U. S. Geological Survey Bulletin B 1964, 35.
Wolfe, J.A., 1992b. Climatic, floristic, and vegetational changes near the Eocene/
Oligocene boundary. In: Prothero, D.R., Berggren, W.A. (Eds.), Eocene-Oligocene
127
Climatic and Biotic Evolution. Princeton University Press, Princeton, pp.

421436.
Wolfe, J.A., 1993. A method of obtaining climatic parameters from fossil leaf
assemblages. U. S. Geological Survey Bulletin 2040, 171.
Wolfe, J.A., 1994. Tertiary climatic changes at middle latitudes of western
North America. Palaeogeography, Palaeoclimatology, Palaeoecology 108,
195205.
Wolfe, J.A., 1995. Paleoclimatic estimates from tertiary leaf assemblages. Annual
Wolfe, J.A., Forest, C.E., Molnar, P., 1998. Paleobotanical evidence of Eocene and
Oligocene paleoaltitudes in midlatitude western North America. Geological
Society of America Bulletin 110, 664678. http://dx.doi.org/10.1130/00167606.
Wolfe, J.A., Schorn, H.E., Forest, C.E., Molnar, P., 1997. Paleobotanical evidence
for high altitudes in Nevada during the Miocene. Science 276, 16721675.
Wolfe, J.A., Spicer, R.A., 1999. Fossil leaf character states; multivariate analyses.
In: Jones, T.P., Rowe, N.P. (Eds.), Fossil Plants and Spores; Modern Techniques.
Geological Society, London, UK, pp. 233239.
Xu, L., Rondenay, S., van der Hilst, R.D., 2007. Structure of the crust beneath the
southeastern Tibetan Plateau from teleseismic receiver functions. Physics of the
Earth and Planetary Interiors 165, 176193.
Yao, H., Beghein, C., van der Hilst, R.D., 2008. Surface wave array tomography in
SE Tibet from ambient seismic noise and two-station analysis II. Crustal and
upper-mantle structure. Geophysical Journal International 173, 205219.
Yilmaz, Y., Guner, Y., Saroglu, F., 1998. Geology of the quaternary volcanic centers
of the east Anatolia. Journal of Volcanology and Geothermal Research 85,
173210.
Yuan, X., Sobolev, S.V., Kind, R., 2002. Moho topography in the central Andes
and its geodynamic implications. Earth and Planetary Science Letters 199,
389402.
Yuan, X., Sobolev, S.V., Kind, R., et al., 2000. Subduction and collision processes
in the Central Andes constrained by converted seismic phases. Nature 408,
958961.
Zandt, G., Myers, S.C., Wallace, T.C., 1995. Crust and mantle structure across the
Basin and Range-Colorado Plateau boundary at 371 N latitude and implications
for Cenozoic extensional mechanism. Journal of Geophysical Research 100,
1052910548.
Zhao, W.-L., Morgan, W.J., 1987. Injection of Indian crust into Tibetan lower crust:
a two-dimensional finite element model study. Tectonics 6, 489504.
Zheng, H., Powell, C.M., Zhou, Z., Dong, G., J., 2000. Pliocene uplift of the
northern Tibetan Plateau. Geology 28, 715718.
Zhu, L., Owens, T.J., Randall, G.E., 1995. Lateral variation in crustal structure of the
northern Tibetan Plateau inferred from teleseismic receiver functions. Bulletin of
Zor, E., 2008. Tomographic evidence of slab detachment beneath eastern Turkey and
the Caucasus. Geophysical Journal International 175, 12731282.
Zor, E., Sandvol, E., Gurbuz, C., Turkelli, N., Seber, D., Barazangi, M., 2003. The
crustal structure of the East Anatolian plateau (Turkey) from receiver functions.
Biographical Sketch
Julien Babault is a lecturer in Earth Surface Processes in the Department of Geology, Universitat Auto`noma de
Barcelona, Spain. He holds an MSc degree in Geology, and a PhD in Geology from the Department of Geosciences
Rennes, Universite de Rennes 1. His research interest include mountain belts dynamics of erosion at continental
scale and dynamics of river network using field examples (Pyrenees, High Atlas of Morocco, Eastern Cordillera of
Colombia in the Andes), quantification of erosion rates and rock/surface uplift rates, and experimental modeling.
128
Jean Van Den Driessche is a professor of Geology at the University of Rennes1. He has a PhD from the University
of Montpellier and completed a State thesis in tectonics at the University Denis Diderot of Paris and the Institute
of Earth Physics of Paris. His teaching and research concern tectonics, lithosphere mechanics and dynamics of
continental erosion.
5.7 Tectonic Geomorphology of Active Folding and Development of

Transverse Drainages
EA Keller and DE DeVecchio, University of California, Santa Barbara, CA, USA
5.7.1
5.7.2
5.7.3
5.7.4
5.7.5
5.7.6
5.7.7
5.7.8
5.7.9
References
Introduction
Lateral Propagation of Reverse Faults and Related Folds
Geomorphic Evidence of Lateral Fold Propagation
Geomorphic Methods to Analyze Laterally Propagating Folds
Santa Ynez Mountains
Complex Lateral Propagation
Development of Transverse Drainage
Directivity of Earthquake Energy and Lateral Fold Propagation: A Hypothesis of Tectonic Extrusion
Conclusions
Glossary
Antecedent stream A stream that developed before
tectonic deformation of the terrain began, and maintained
its course continuously downcutting across the growing
tectonic features.
Bankfull channel width Refers to the horizontal width of
a stream when water completely fills the channel at the
point of incipient flooding, which has an approximate
return period of 1.5 years.
Blind reverse fault A compressional fault, where fault
offset does not rupture the surface. Surface deformation of
the uppermost rock layers is often characterized by folding.
Directivity In regards to an earthquake, is an effect
whereby seismic ground shaking is greater in the direction
of rupture propagation than in other directions from the
earthquake hypocenter.
Drainage density The ratio of the total length of all
streams and rivers within a specific drainage basin to the
area of that basin.
En echelon Refers to closely spaced parallel or subparallel
structural features, such as faults and folds, that overlap or
are offset in a step-like manner oblique to the overall
structural grain.
Hinterland Refers to the principle region of active tectonic
deformation, which has experienced significant uplift and
erosion in contrast to the foreland, which lies parallel to the
hinterland and is characterized by deposition and youthful
deformation.
Lateral propagation Describes tectonic processes related
to active faulting or folding, whereby faults or folds grow in
length in response to increased displacement or
amplification, respectively.
130
130
131
132
134
135
139
142
145
145
Segment boundary Refers to discrete zones of lateral fault

or fold termination, which are common in en echelon
(see definition, above) groups of structural features.
Segment boundaries are typically regions of low elevation
and therefore are often characterized by water gaps
(see definition, below) in active fold belts.
Stream capture Also called stream piracy, refers to a
process whereby a stream is diverted from its channel into
another stream, perhaps reversing flow direction. This
process is often initiated by headward erosion of one stream
valley to intersect another.
Stream diversion Refers to the process where the existing
channel of a stream or river is disrupted by tectonic uplift
and the resultant stream path is around the tectonic
topography.
Superposed stream Refers to a stream channel and its
valley within resistant bedrock that developed its course
through different preexisting alluvial cover. The resulting
channel may cross geologic structure such as a buried ridge.
Suture zone Refers to a structurally complex region where
two fault or fold segments have grown laterally together to
form a single structural element.
Tear fault Refers to a steeply dipping fault that has
strike-slip displacement that is typically orthogonal to the
structural grain, and in compressional regions is parallel to
the principle shortening direction. Tear faults accommodate
differential shortening between fault segments and are often
responsible for controlling the position of fault segment
boundaries.
Transverse drainage Refers to a stream network in an
actively deforming region, where the trunk stream is
oriented perpendicular to the structural grain.
Keller, A.E., DeVecchio, D.E., 2013. Tectonic geomorphology of active

folding and development of transverse drainages. In: Shroder, J. (Editor in
Chief), Owen, L.A. (Ed.), Treatise on Geomorphology. Academic Press, San
Diego, CA, vol. 5, Tectonic Geomorphology, pp. 129147.
http://dx.doi.org/10.1016/B978-0-12-374739-6.00088-9
129
130
Tectonic Geomorphology of Active Folding and Development of Transverse Drainages
Unit stream power A one-dimensional measure of the

rate of energy dissipation of a stream against the channel
bottom and banks (per unit width). Units are W m2. Unit
stream power is given by the equation: o rgQS
b where r is
density, g is gravity, Q is discharge, S is slope, and b is the
channel width.
Water gap A notch or valley across uplifted topography
through which water flows. The gap may be the result of an
antecedent drainage, superposition, or may occupy the

position of a fault or fold segment boundary (see definition,
above).
Wind gap A notch or valley across uplifted topography
through which water once flowed, but is now dry due to
stream capture or stream diversion. Wind gaps are also
referred to as an air gap.
Abstract
Two fundamental geomorphic questions are: (1) Do growing folds propagate laterally? and (2) how does transverse
drainage develop across an active fold belt? Research during the past decade has answered the first question, and it is now
known that some folds in fold-and-thrust belts definitely propagate laterally. Individual folds may propagate in one or both
directions, and two folds may propagate toward each other, producing a suture zone and a longer segmented structure.
Folds may also propagate laterally past one another to form an en echelon arrangement. Rates of lateral propagation and
incision of streams that traverse growing folds remain poorly constrained. Adequate chronology to quantify fold growth
and erosion is commonly not available and therefore remains one of the greatest limitations to understanding the evolution of surface folds. As a result, structural and geomorphic analyses are utilized, which are useful, but not adequate for
quantifying fold growth processes. Development of transverse drainage across folds is a complex process related to the rate
of uplift, lateral, and vertical fold growth (including change in fold width), development of water and wind gaps, lateral
diversion of channels, and change in channel geometry and unit stream power. Understanding lateral propagation has
important implications for earthquake hazard. If fault length is assumed to increase with fault displacement, then the folds
they produce should also increase in length. Hypothetically, as folds propagate laterally due to displacement on the
underlying faults, coseismic displacement along the faults themselves may have a preferred direction of propagation.
5.7.1
Introduction
Quantification of the geomorphology of active folds can

provide fundamental information about the rates of tectonic
processes and the spatiotemporal structural evolution of a
region undergoing compressional deformation. Consequently,
tectonic geomorphology of fold belts has been an area of intensive research in the past decade. Underpinning the geomorphic development of active folds is understanding the
development of transverse drainages across folded regions.
This chapter summarizes evidence for lateral and vertical fold
growth, the response of the drainage system to fold belt development, and the implications of lateral propagation to the
development of orogen-scale topography and seismic hazard.
The existence of lateral propagation of folds above concealed or blind reverse faults is an important hypothesis and is
generally accepted. The majority of evidence suggests that
folds do grow laterally and that propagation has a profound
effect on development of drainage through a fold belt (Burbank et al., 1996, 1999; Jackson et al., 1996; Boudiaf et al.,
1998; Delcaillau et al., 1998, 2007; Keller et al., 1999; Keller
and Pinter, 2002; Amos and Burbank, 2007; Chen et al.,
2007). What is not so clear is how the transverse drainages
develop in response to folding. In some cases, simple diversion of drainage with abandonment of water gaps seems fairly
straightforward. However, considerable complexity may be
associated with antecedent and superimposed streams and
links with lateral propagation, channel incision, stream capture, and stream diversion.
One of the most compelling transverse drainage features,
indicative of lateral fold growth, is the presence of wind gaps,
which develop from water gaps that have been defeated by

tectonic uplift in response to lateral propagation. Several factors
and processes are important for a water gap to remain open as a
fold grows vertically and in width (Burbank et al., 1996): (1)
rate of uplift; (2) rock resistance; (3) unit stream power (stream
power per unit channel width); and (4) nature and extent of
preexisting drainage. A water gap is most likely to persist in the
landscape if the uplift rate is low, rock resistance is low, upstream channel slope is moderate to steep, bankfull channel
width through the gap is relatively low, and upstream drainage
area (a proxy for discharge) is large (Burbank et al., 1999).
A paramount question concerning lateral propagation of
folding is: What is the potential significance of propagation of
faults and folds to earthquake process and hazard? The authors
present a hypothesis that as faults increase in length, so do the
folds they produce; and the authors hypothesize that a direction of earthquake rupture is more likely to be in the direction of lateral fold propagation. In Southern California,
apparently faults and folds tend to propagate to the west, south
of the San Andreas fault, in the vicinity of the Big Bend of the
fault; and to the east north of the fault. Understanding relations
between fault propagation and earthquake propagation is extremely important, as the earthquake hazard may be greatly
increased or reduced relative to rupture propagation direction.
5.7.2
Lateral Propagation of Reverse Faults and

Related Folds
Folding in fold-and-thrust belts (areas or regions characterized

by subparallel faults and folds) is commonly closely linked
with buried reverse faulting. As total fault displacement from

repeated earthquakes increases, so does total fault length. The
basic argument constructed to support this statement comes
from two observations (Cowie and Scholz, 1992; Jackson
et al., 1996): (1) the ratio of fault slip to rupture length during
an earthquake generally has a value of B105104 (rupture
length is generally 10 000100 000 times fault slip as a result
of an earthquake); and (2) the ratio of total displacement on a
fault to fault length is B103101 (fault length is generally
101000 times as long as total displacement). For these two
relationships to be satisfied simultaneously, faults are required
to increase their lengths as they increase total displacement
(Cowie and Scholz, 1992; Dawers and Anders, 1995; Hetzel
et al., 2004). Therefore, the rational conclusion is that
fault-related folds that develop above these faults propagate
laterally as they amplify.
Proving from geomorphic evidence that buried reverse
faults propagate laterally is difficult. However, the folds
(commonly linear anticlines) that buried reverse faults produce can provide (with adequate chronology) an estimate of
both the direction and rate of lateral propagation (e.g., Keller
et al., 1998; Jackson et al., 2002; DeVecchio et al., 2012). Both
active faulting and folding at the surface generally occur
together. As a result, fold scarps (the scarps produced by active
folding) have been mapped as fault scarps. Studies of individual folds suggest that horizontal shortening produces reverse faults that cause hanging-wall deformation and
development of anticlines (e.g., Yeats, 1986).
fold limbs. These six criteria are consistent with lateral

propagation, but do not prove propagation.
An additional useful criterion for correlating wind gaps
and water gaps to the process of lateral fold propagation is the
Hetzel criterion (Hetzel et al., 2004). If water gaps do, in fact,
represent the position of the topographic nose of the fold and
wind gaps record the previous location of the fold nose before
propagation, the intersection of the topographic envelope of
the fold with a horizontal line from the base of wind gaps
should correspond to the location of the lower and younger
wind gaps and the active water gap (Figure 1).
Fan-shaped patterns of tributary drainage that forms on the
flanks and nose of a propagating anticline have been shown as
valuable indicators of lateral propagation (Ramsey et al.,
2008). Figure 2 shows an elliptical-shaped fold at Time 1
before lateral propagation. The tributary channels are normal
to the elevation contours and form a fan shape pattern. With
lateral propagation (Time 2), the channels lengthen and are
reduced in slope; and if deeply incised, their curved patterns
will be inherited from Time 1. These channels will not be
normal to the contours. Where such a pattern is present (inherited drainage), lateral propagation is the likely cause.
Ramsey et al. (2008) documented folds in the Zagros Mountains that retain such information of earlier fold growth.
Rates of lateral propagation of active folding (based on a very
small number of studies where chronology is established) are
generally several times the rate of uplift and fault slip (Keller
et al., 1998). An example of a fold that is rapidly propagating
laterally is Wheeler Ridge, located at the southern end of the San
Joaquin Valley (Figure 3). A simple geologic cross section of
Geomorphic Evidence of Lateral Fold

Propagation
Faults and folds can be relatively long structures (tens of

kilometers) that are generally segmented. Segment boundaries
may be topographically expressed (geomorphic segmentation)
or be geologic structures, such as cross faults (geologic segmentation). Some faults and folds in the Zagros Mountains of
Iran exceed 50 km in length (Ramsey et al., 2008). In the
Zagros, for example, it can be demonstrated: (1) that fault
segments have grown by lateral propagation of a single segment, or (2) that several shorter segments grew together
through lateral propagation to form one longer continuous
structure with a suture zone that is located in a topographic
saddle along the crest of the fold. Some combination of
the two processes may also occur (Ramsey et al., 2008).
Lateral propagation of folds may be demonstrated or
strongly indicated by careful determination of geomorphic
relations between surficial processes (uplift, folding, weathering, and erosion) and earth materials (Jackson et al.,
1996). Keller et al. (1999) proposed six geomorphic criteria
useful in evaluating rates and direction of lateral propagation
of active folds (in the direction of lateral propagation): (1)
decrease in drainage density and dissection; (2) decrease in
elevation of wind gaps; (3) decrease in relief of the topographic profile along the crest of a fold; (4) development
of characteristic drainage patterns (resulting from drainage
diversion); (5) deformation of progressively younger deposits
or landforms; and (6) decrease in rotation and inclination of
AG1
AG1
(b)
AG2
WG
AG3
Same river produced all gaps
(a)
Elevation
5.7.3
131
Top
ogr
aph
ic e
nve
AG2
lope
AG3
WG
Distance
AG = Air (wind) gap
WG = Water gap
Figure 1 Idealized diagram illustrating the Hetzel criterion, which is

useful in demonstrating that a series of wind gaps and a water gap
are produced by the same river. (a) Propagation of the fold to the
right forces the active stream channel (water gap) to migrate in the
direction of propagation resulting in a series of wind (air) gaps that
are young to the right (AG 13). (b) Horizontal lines drawn from the
base of the wind gaps should intersect the topographic envelope of
the fold at the position of younger wind gaps and the water gap
confirming the interpretation that wind gaps form from lateral
propagation of the fold.
132
Time 1
Map view
Perspective
L1
Convex surface
(nose of fold)
Time 2
Present drainage
Inherited drainage
Fold length
Topgraphic
contours (map view)
L2
Figure 2 Idealized diagram showing fan-shaped drainages developed on the flanks and nose of a growing fold (plunge panel). Initially, drainage
is perpendicular to contours (Time 1). With entrenchment of these streams, they may persist in the landscape. If propagation occurs (Time 2),
the entrenched streams will not cross the contours at right angles but will be at a lesser angle and have a lesser channel slope. If this is
observed in the landscape, then it is powerful evidence of lateral propagation.
Wheeler Ridge is shown in Figure 4. The abundance of oil wells

that pass through the buried Wheeler Ridge reverse fault help to
establish the fault depth and geometry (Medwedeff, 1992).
At Wheeler Ridge, ages of folded alluvial fan surfaces across
the fold were used to determine the rate of lateral propagation
at B3 cm yr1 (Figure 3). Rates of uplift were estimated to be
approximately 3 mm yr1, indicating that the rate of lateral
propagation is 10 times the rate of uplift and, therefore, is a
rapid tectonic process (Keller et al., 1998). Lateral propagation
at Wheeler Ridge appears to be linked to a series of propagation events associated with tear faults (faults at a high angle
to the strike of the ridge) (Medwedeff, 1992; Mueller and
Talling, 1997). Evaluation of subsurface data at Wheeler Ridge
suggests that the position of major wind and water gaps coincide with tear faults that have a topographic expression
(apparent fault scarp with vertical relief) facing the direction
of lateral propagation. The change in relief (down to the direction of propagation) helps to establish stream position at
the base of the scarps of the tear faults.
The criteria discussed above are useful in recognizing lateral
propagation of folds but are not sufficient to prove lateral
propagation. This results because the criteria, given specific
scenarios, may be consistent with both fold propagation and
fold axis rotation models with different rates of surface uplift
along the fold during fold growth (Figure 5). Criteria 4 and 5 are
strong evidence of lateral propagation; and if at least two wind or
water gaps from the same stream, then this is strong evidence of
lateral propagation. Assuming that the stream can only be in one
place at a time, a model of fold rotation (Figure 5(b)) is unlikely
to produce two or more gaps at progressively lower elevations in
the direction of fold growth. In other words, there could only be
one gap and one diversion of drainage. Once a stream is defeated

by uplift in the rotation model, it would be deflected to the nose
of the fold (which, in that model, is fixed), and another gap
would not form. Even if multiple wind gaps could be produced
by rotation, the Hetzel criterion would not be satisfied (see
Figure 1). In general, lateral propagation of active anticlines and
anticlinal ridges can be visualized as illustrated on an idealized
diagram (Figure 6). A buried reverse fault and fold propagate
laterally to the left in Figure 6. Drainage from the steeper forelimb has been diverted around the nose of the fold. During
lateral fold growth, a wind and a water gap were produced.
Locations of gaps correspond to locations of cross faults that
segment the fold. Drainage density is greater in the higher, older
part of the fold where limb rotation is greater. The profile
through the wind gap is convex, representing folding of the
abandoned valley as lateral propagation continued. The idealized diagram is consistent with what was learned at Wheeler
Ridge (Figure 3) and Mission Ridge (discussed below), and is a
visual representation of the geomorphic criteria that help identify lateral fold propagation (Keller et al., 1999).
5.7.4
Geomorphic Methods to Analyze Laterally

Propagating Folds
Geomorphic methods employed to analyze lateral propagation of folds are keyed to the six geomorphic indicators of
lateral propagation of Keller et al. (1999), and include:
1. Construction of a detailed drainage map of the folded
forelimb and backlimb, as well as the plunge panel (nose
ramp). This is illustrated in Figure 3 for Wheeler Ridge and
133
Wheeler ridge direction of propagation
(a)
Drainage network
1185930W
Dissection
Drainage
density
(b)
1 km
Decreasing
Decreasing
Approx age 200 ka

Drainage
pattern
(South limb)
350000N
Wind gap
Water gap
Water gap
125 ka
60 ka
14 ka
1 ka
Deflection to east
Figure 3 Evidence for eastward propagation of the Wheeler Ridge anticline. (a) View toward the north of Aerial Wheeler Ridge (courtesy of
John S. Shelton). Note the eastward decrease in relief along the axis of the fold. (b) Other geomorphic indicators of lateral propagation of
Wheeler Ridge to the east (right). Modified from Keller, E.A., Gurrola, L., Tierney, T.E., 1999. Geomorphic criteria to determine direction of lateral
propagation of reverse faulting and folding. Geology 27(6), 515518.
Figure 7 for the Mission Ridge anticline in Santa Barbara.

This analysis may be completed by using aerial photographs, detailed topographic maps, and digital elevation
models (DEMs).
2. Analyzing drainage patterns to calculate drainage density,
which is a measure of degree of dissection. Drainage
density (Dd) is defined as the ratio of total length of
channels divided by the drainage basin area. Drainage
density on the south flank of Wheeler Ridge (backlimb)
decreases systematically from about 1518 km1 for surfaces older than 100 ka, to B0.4 km1 for surfaces of
1760 ka, and is nearly zero for Holocene surfaces (see
Figure 3). For Mission Ridge (see Figure 7), drainage
density decreases from about 8.0 to 3.3 km1, and limb
rotation decreases significantly in the direction of lateral
propagation. For Wheeler Ridge and the western propagation of Mission Ridge, the Hetzel criterion holds for
position of gaps. For the eastern end of Mission Ridge
(Figure 8, the ridge is more complex, and a simple model

of eastern lateral propagation does not explain the position
of gaps. Here, the morphology is dominated by Sycamore
Creek. The water gap that Sycamore Creek flows through
probably has been stable during fold growth. East of
Sycamore Creek, the fold may have propagated east at least
as far as Montecito Creek, but this is not supported by the
geomorphology (Figure 8).
3. Construction of topographic profiles along the crest; profiles normal to the fold and DEMs may assist in identifying
geomorphic parameters of folds. Elevations of wind gaps
along the crest of an anticline may be measured directly,
and these elevations are generally lower in the direction of
propagation (see Figures 1, 6, and 8) (Jackson et al., 1996;
Burbank et al., 1999; Keller et al., 1999). The topographic
profile along the crest of the fold may reveal the direction
of fold plunge, as is shown for Wheeler Ridge (Figure 3)
and western Mission Creek (Figure 8).
134
Wheeler ridge
Oil well
S
N
Tulare formation
San Joaquin formation
Etchegoin and
Chanac formations
Etchegoin and Chanac formations

San
ta
Mon
te
rey
formatio
formation
Margarita
Santa Margarita formation
Monterey formation
Wheeler ridge thrust

0
1 km
Figure 4 Geologic cross-section across Wheeler Ridge in the midportion of the ridge shown in Figure 3. The numerous oil wells that penetrate
the subsurface provide geologic control to understand the fault propagation fold that produced Wheeler Ridge.
Propagation
Fault displacement at T1
4. Establishing the chronology of Late Pleistocene and Holocene alluvium being folded is arguably the most difficult and
neglected aspect of studying active folding. However, with no
dates there can be no rates. Lacking dates, the understanding
of geomorphic processes will remain inadequate to address
active tectonics. Methods of dating alluvial deposits folded
over young anticlines have improved dramatically in recent
years (Keller and Pinter, 2002). As new dating methods are
discovered and developed, our understanding of active tectonic processes will significantly improve.
(a)
5.7.5
Santa Ynez Mountains
Rotation
(b)
Figure 5 Two models (upper-plan view; lower-profile along fold crest)

of fold growth. (a) Lateral fold propagation, where fault length increases
with fault displacement. (b) Fold rotation, where the fault tip is fixed
and limb rotation occurs in response to uplift, producing a displacement
gradient along the fault. Reproduced from Keller, E.A., Gurrola, L.,
Tierney, T.E., 1999. Geomorphic criteria to determine direction of
lateral propagation of reverse faulting and folding. Geology 27(6),
515518
The Santa Ynez Mountains (an anticlinorium above the

piedmont of Santa Barbara, CA, and part of the Santa Barbara
fold belt) is an example of a large, complex, linear structure
that is segmented (Tierney, 2002). Figure 9 illustrates the
important geomorphic and structural features of the range.
Notice the three, gently curved, geomorphic segments, two of
which are divided by the geology into subsegments. The range
crest is B100 km long, and individual segments range from
about 20 to 50 km. The segments decrease in elevation to the
west, and the westernmost segment (segment 1 on Figure 9)
appears to be propagating laterally to the west.
The present topography of the Santa Ynez Range developed
during the past 500 000 years (perhaps as young as 200 000
years) in response to displacement on one or more northdipping, buried reverse faults (Namson and Davis, 1988,
1992). Individual fault-and-fold segments apparently grew
simultaneously, but the segments are generally young to the
west (Tierney, 2002). The boundary between segments 1 and 2
is a major oblique left-lateral strike-slip cross fault known as
135
FL, Fs
FL, Fs
BL
BL
B
A
Dd
Ga
Dd
MFs
D
D
u
Gw
A
MF
C
AG
1 km
WG
5 km
Direction of lateral propagation
Gw Water gap
AG Air (wind) gap
FL
Forelimb
Dd Diverted drainage
BL
Backlimb
Fs
Fold scrap
Buried reverse
MF Mountain front
fault
u
MFs Mountain with fold scarp
Section
Cross fault u-up
Figure 6 Idealized diagram of a fold that is propagating laterally. Major geomorphic features are shown by the three cross-sections: AA0 is near the
terminus or nose of the fold (plunge panel) to the left; BB0 is through a wind gap, which has been deformed by folding since it was abandoned; CC0
is located at an older part of the fold, where a greater amount of shortening has occurred. Notice that the degree of dissection and drainage density
decreases in the direction of lateral propagation, as does the elevation of the topographic profile along the crest of the fold (DD0 ). Positions of wind
and water gaps coincide with cross faults that have both lateral and vertical components of slip. The steepest limb of the fold is the forelimb and forms
a relatively straight-prominent fold scarp. Modified from Keller, E.A., Pinter, N., 2002. Active Tectonics. Prentice Hall, Upper Saddle River, NJ, 362 pp.
the South Branch of the Santa Ynez fault. The Santa Ynez fault
is a major east-to-west trending, left-lateral oblique (south side
up) fault that generally follows the Santa Ynez River Valley just
north of the range. The boundary between segments 2 and 3 is
a series of cross folds (San Marcus folds in Figure 9). Apparently, the lateral growth of segment 3 to the west impacted
segment 2, causing cross folds to form (with perhaps faults at
depth). The Ramero Saddle is the boundary between the two
parts of segment 3. This may represent a suture zone, as noticed in one of the Zagros Mountain folds (Ramsey et al.,
2008). Thus, it appears that the Santa Ynez Range was formed
by amalgamation of several segments that grew together to
form a continuous structure that propagated to the west
(Tierney, 2002).
5.7.6
Complex Lateral Propagation
Lateral propagation of faults and folds with drainage diversions can be complex. A good example of this is the El Asnam
anticline in Algeria (Boudiaf et al., 1998). A series of maps

illustrates some of the processes related to the anticline, which
is propagating to the southwest (Figure 10).
The El Asnam anticline is underlain by the El Asnam reverse fault, which has been a very active structure, producing
earthquakes such as the 1980 (Mw7.3) event that caused extensive deformation (mostly uplift with lesser subsidence) in
the region. Geologic and geomorphic analysis using DEMs
suggest that an interesting series of drainage diversions occurred as the anticline has propagated laterally.
Figure 10(a) shows the position of the El Asnam anticline
in the recent geologic past when two water gaps cut across the
anticline. The much larger Ech Cheliff River forms a prominent water gap that is present till today. This river has not been
diverted, as it retained sufficient stream power to erode its
banks and bed as the anticline was uplifted. The second more
southerly water gap is that of the Fodda River. The more recent
situation is shown in Figure 10(b). As the fold continued to
propagate to the southwest, the Fodda River was diverted,
leaving behind a wind gap and a younger water gap. It is
136
Mission ridge direction of propagation

Wind gaps
Mission creek
500 m
(a)
1194230W
ek
n cr e
342630N
sio
Mi s
500 m
Drainage density (km/km2)
3.3
Decreasing
6.3
8.0
Drainage pattern
Deflection to West
Limb rotation (in degrees): Sedimentary units (surface)
(3.5)
(b)
1418 (6)
25 (12.0)
1194230W
94
Creek
X
Mis
73
342630N
AG
PC
Paleo channel
Mission
ridge
AG
Santa
Barbara
Mission
91
AG Air (wind) gap
107
X 91
94
115
114
95
104
ion
PC
Mission Ridge fault

91 Contours in meters
250 m
(c)
Figure 7 Evidence of lateral propagation of Mission Ridge to the west. (a) 1 m resolution hillshade of Mission Ridge north of the city of Santa
Barbara, CA, showing the position of several wind (air) gaps that developed from westward deflection of Mission Creek. (b) Three geomorphic
criteria suggest lateral propagation to the west. (c) Topographic contours along Mission Ridge showing the position and decreasing elevation of
wind gaps toward the west. The sum total of data suggests that Mission Creek is propagating laterally to the west. Modified from Keller, E.A.,
Gurrola, L., Tierney, T.E., 1999. Geomorphic criteria to determine direction of lateral propagation of reverse faulting and folding. Geology 27(6),
515518.
highly likely that this diversion resulted from the Fodda River,
as the Hetzel criterion (see Figure 1) is satisfied. We might
expect that with further lateral propagation, additional wind
and water gaps would form to the southwest. However, this
was not the case (Figure 10(c)). In 1980, a magnitude 7.3
earthquake uplifted the anticline approximately 5 m and
caused approximately 1 m of subsidence in the adjacent
Cheliff Basin. As a result, a topographic gradient to the north
developed. The earthquake also ponded drainage along the
fault, and, presumably, a tributary flowing to the northwest

captured the Fodda River. Thus, the recent history of the El
Asnam anticline is one characterized by lateral propagation
with drainage diversion in the direction of propagation and
then reversal of that drainage in response to a large earthquake
that changed the slope of the topography initiating stream
capture. Thus, we see that drainage diversion and stream
capture are important processes in the topographic and geomorphic history of the area.
137
A
A
ree
kc
A
200
Oa
nte
Mo
ree
cit
oc
ptu
aly
Eu
c
Mo k
nte
cit
oa
ill w
sh
ss
er
nti
clin
ind
ga
ap
dg
win
wa
pa
ny
ca
Sy
c
am
Ba
rk
ore
Mission ridge
anticline
VE = 5X
100
0
on
ion
iss
p
ga
Ol
dm
A
200
Wi
nd
Mi
ss
Elevation (m)
ion
cre
ek
win
ter
dg
ga
ap
1 km 2
100
4
Distance (km)
Figure 8 Longitudinal profile of Mission Ridge anticline showing the tectonic geomorphology. Shown are topographic expressions of water and wind
gaps, as well as a decrease in elevation to the west that suggest lateral propagation. The eastern part of the fold exhibits a more complex history.
Modified from Keller, E., Gurrola, L., 2000. Earthquake hazard of the Santa Barbara fold belt. Final Report to the US Geological Survey, NEHRP Award
527726, 78 pp. http://www.scec.org/research/98research/98gurrolakeller.pdf
Ca
rnia
lifo
Segment 1
avg. elev ~410 m
Segment 3a
avg. elev. ~1090 m
Segment 2a
avg. elev. ~760 m
Segment 2b
avg. elev. ~1000 m
Segment 3b
avg. elev. ~1250 m
34270N
Location of B
120320W
(a)
km 10
S. Branch Santa Ynez

fault (NE-SW)
12000W
Refugio
pass
Drainage
consistent
with lateral
propagation
1
km
San Marcoss
folds
NW-SE
Ramero
saddle
Segment
Segment boundary
Santa barbara
Inherited drainage
(b)
Figure 9 Geomorphology of the Santa Ynez Range in the Western Transverse Ranges. (a) The Santa Ynez Range, showing three gently
curved geomorphic segments, two of which may be subdivided into subsegments. Topographic elevation of the crest of the Santa Ynez Range
decreases systematically along the segments to the west, and boundaries between segments are structural. It is hypothesized that the segments
are older in the eastern part of the range and youngest to the west, and the entire range developed as segments interacted with each other
through time. (b) Entrenched fan-shaped drainage network on the westernmost segment of the Santa Ynez Range is suggestive of lateral
propagation to the west.
WG a
WG
100 m
0m
361400N
WG
WG
er
Riv
f
elif
hC
Ec
Ech Cheliff
Fo
dd
a
AG
WG
2 km
ve
Ri
13245E
(a) (Older)
WG a
100 m
0m
AG
ive
fr
elif
hC
Ec
WG
Ech Cheliff
dd
Fo
138
riv
AG
AG
er
(b) (Intermediate)
2 km
WG a
100 m
a
a'
0m
R
WG
AG
ive
fR
lif
he
hC
Ec
Ech Cheliff
AG
a
a
dd
Fo
er
riv
2 km
(c) (Present)
Direction of slope
AG
Air (wind) gap
WG
Water gap
Reverse fault
Fold/fault with direction
of lateral propagation
River flow direction
139
Intermontane mountain front (relic)

Active mountain front
(a)
Intermontane mountain front (relic)

Intermontane mountain front (active)
d
de
ol
f
,
ed fan
lift ial
p
U lluv
a
Alluvial
fans
(b)
Figure 11 Idealized diagram showing the development of buried reverse faults and overlying anticlines that, through time, propagate from
the hinterland toward a basin that has not been deformed. (a) An intermontane mountain front is present. (b) Buried reverse faults have
propagated into the basin, folding the overlying alluvial fan gravels forming a series of anticlines parallel to the range front. Notice that upper
parts of fans have been faulted and folded and incorporated into the mountain block with the development of a new active mountain front
(Courtesy of Burbank)
5.7.7
Development of Transverse Drainage
Until recently, our knowledge of fold-and-thrust belt evolution centered on the belief that fold growth migrated away
from the hinterland into an adjacent basin in a systematic way.
It was thought that only one thrust fault was active at a time;
and with continued shortening, deformation was transferred
(migrated) to a new structure farther into the basin. We now
understand that multiple faults and folds may be active simultaneously and that deformation does not often migrate
systematically into a basin (Figure 11; Burbank et al., 1999).
With the further understanding that folds commonly
propagate laterally, new ideas concerning development of

drainage across and around growing folds have been discussed. The evidence supporting new ideas about transverse
drainage across fold belts is largely geomorphic.
The origin of transverse drainage (drainage across structures from linear ridges to mountain ranges) is a longstanding
geomorphic problem (Bloom, 1998). Transverse drainage may
be antecedent, that is, it is present prior to uplift of a growing
structure and has sufficient stream power (proportional to the
product of discharge and channel slope, where drainage basin
area or channel length is a surrogate for discharge). For example, the Ech Cheliff River (Figures 10(a)10(c)) has been
Figure 10 Development of drainage across the El Asam anticline in Algeria. (a) The anticline is propagating laterally to the southwest, and two
water gaps were present. The larger gap is along the Ech Cheliff River, and a smaller gap has formed where the Fodda River crosses near the
nose of the fold. (b) With further lateral propagation, the Fodda River has been deflected to the southwest, leaving behind an air (wind) gap. (c)
The present topography has the Fodda River flowing north into the Ech Cheliff River. This diversion was in response to the M3 earthquake that
occurred in 1980. Uplift of about 5 m in the southwest with about 1 m of subsidence farther north produced a topographic gradient to the north.
Stream capture was most likely responsible for the diversion of the Fodda River to a point farther north on the Ech Cheliff River.
140
in the same location across the active El Asnam anticline

during fold growth; and thus, the river is antecedent to the
uplift of the anticline.
Alternatively, some streams are transverse to structure because of the presence of cover material that has a drainage
pattern that is let down over buried structures. As the cover
material is removed by erosion, the resulting transverse
drainage is said to be superposed. Superposition is difficult to
demonstrate unless some of the cover material remains.
Antecedent stream segments are obvious for drainage
across young fold belts where some of the folds migrate with
time into a largely undeformed basin. With the migration of
thrust faults and folds into a depositional basin, growing
faults and folds transform areas that were formerly depositional to growing folds with their associated erosional
processes (Burbank et al., 1999).
Drainage development associated with fold belts can be
complex. This results because folds may grow laterally as
well as vertically as a response to displacement of buried faults
within anticlines. Also, new folds may be added to a fold
belt as the series of folds and faults propagate into undeformed areas (commonly in adjacent basins, see Figure 11).
Streams draining from a mountain range (hinterland) deposit
alluvial fans that, as younger folds develop basinward, are
in turn folded. Thus, a series of parallel folds form, and previous cover materials (alluvial fan deposits) are folded. This
process is illustrated in Figure 12 for San Emigdio Canyon
and associated alluvial fans on the north flank of the San
Emigdio Mountains of Southern California. The oldest
mountain front, now an intermountain (relict) front, was the
apex for alluvial fan deposition about 100 ka. Today, the fan
gravels are folded by the Wheeler Ridge and Los Lobos faults
to the south.
A drainage that is defeated by uplift of a structure and flows
parallel to a fold axis will likely be further diverted in the
direction of propagation (Burbank et al., 1996, 1999; Jackson
et al., 1996). As diversions develop, streams are captured; and
the drainage basin area increases until sufficient stream power
is available to temporarily maintain a channel at the nose of
the fold where uplift has not yet occurred. As lateral propagation of the fold continues, this area becomes a water gap
and eventually may become a wind gap if the channel is defeated by uplift and/or stream capture. If defeat occurs, the
channel may be diverted again in the direction of lateral
propagation and, in the course of a fold development, may
make several passes around the fold as the drainage develops.
For some folds, several wind gaps may be produced in this
manner.
Evaluation of drainage patterns for specific folds may
suggest that some folds are propagating in two directions, that
adjacent folds are propagating toward each other, or that
younger folds are propagating parallel toward or against older
folds in a fold belt (Jackson et al., 1996; Ramsey et al., 2008).
Figure 13 shows the process of how drainage may be diverted
as two segments of a fold propagate laterally to form a
seemingly single structure. The profile along the ridge crest
has a topographic saddle where the suture zone developed.
Ramsey et al. (2008) reported folds in the Zagros Mountains
(fold belt) where two segments have grown together by lateral
propagation. The process of lateral fold growth may also
squeeze two rivers together to flow through a single gap.

This process will increase total and unit stream power, which
favors persistence of the gap even with further fold growth
(Burbank et al., 1999; Ramsey et al., 2008). Fold growth may
also involve the development of a series of folds that are en
echelon because the buried fault producing the folding is
segmented and en echelon. With fold growth of such a system,
individual folds may propagate laterally past one another
(Burbank et al., 1999).
Because rates of uplift are commonly larger than rates of
incision and because many more small streams are present
than large rivers that traverse growing folds, water gaps
are often defeated to become wind gaps. As an antecedent
stream flows across a growing fold, sufficient unit stream
power is necessary for a water gap to persist. With uplift
of a fold, the upstream limb of the fold (generally the backlimb) will experience a reduction in slope, which lowers
stream power, favoring defeat of the water gap. Across the
crest of the fold, on the downstream fold limb (generally
the forelimb) channel gradient increases, which increases
the stream power. In response to the uplift, the channel in the
water gap incises and narrows, sometimes producing a wineglass shape (wider at the top with stem-like incision of
a narrow channel lower in the channel cross section). Narrowing increases the unit stream power, favoring incision
(Amos and Burbank, 2007). Duvall et al. (2004) showed that
with constant rock resistance that a sevenfold increase in uplift
rate resulted in a factor of twofold increase in channel slope
and a threefold decrease in channel width. The narrower,
steeper channels would have a unit stream power increase of
about 5 times.
With all the possibilities, it is not surprising that the
processes that transform a water gap into a wind gap can be
complex. Two of several hypotheses, proposed by Keller et al.
(1999), for defeat of a water gap are: (1) uplift of the fold, as a
result of a single or series of large earthquakes, may block the
channel in the water gap, forcing a diversion in the direction
of lower topography, which is likely to be in the direction the
fold propagation; and (2) a tributary channel to the stream
crossing the nose of the fold on the hinterland side of the fold
that erodes headward parallel to the axis of the fold (the opposite direction the fold is propagating) may form. Headward
growth or extension of that drainage may capture the drainage
feeding a water gap. In many cases, a wind gap may form by a
combination of both tectonic and fluvial processes (uplift
diversion and capture of drainage).
Consideration of these processes allows for a hypothetical
explanation for the development of drainage across an actively
developing fold belt (see Burbank et al., 1999, for additional
explanations). The process is illustrated in Figure 14. A stream
establishes a channel across the path of the propagating
fold before it arrives, and that stream reach is antecedent to
the uplift. As fold growth continues, a water gap may form,
as discussed above. With this hypothetical model, drainage
is established across a developing fold belt by a series of
captures, diversions, and antecedent positioning. That is, the
process involves development of relatively short antecedent
reaches across the path of propagating folds, followed by defeat and stream deflection along relatively longer reaches
parallel to the fold axis (Keller et al., 1999).
141
California
bos fa
ult (bu
sf
ea
dr
An
Los Lo
n
Sa
Alluvial fan
Los Lo
ried)
bos fo
lt
au
lds
San Emigdio Cr.
nt
n
tai
fro
ve
cti
mo
Surface
Age
Q2, 0.51.0 ka
Plunging anticline
Q3, 4.77.5 ka
Traces of pleito fault
Q4a unknown
Q4b, 82125 ka
San E
Undifferentiated
quaternary deposits
and tertiary sedimentary rocks.
dio
mig
C
r.
Intermontane
mountain front
N
0
kilometers
(a)
2.5
N
San Joaquin Valley
San Emigdio Mts.
Holocene gravels
Late pleistocene gravels
Bedrock
Q4 Paleosol
x4 vertical exaggeration
Elevation (km)
2.0
1.5
Intermontane (relic)
mountain front
1.0
?
0.5
Pleito fault zone
0
(b)
Present active
mountain front
fold scarp
?
Wheeler ridge
Los Lobos ?
fault
fold
10
13
Distance (km)
Figure 12 (a) A map showing the Holocene and late Pleistocene deposits (Q2Q4) in San Emigdio Canyon in the southern end of the San
Joaquin Valley, CA. Black box in inset shows the location of San Emigdio Canyon with respect to the San Andreas fault in Southern California.
(b) Buried traces of the Wheeler Ridge fault and Los Lobos folds and topographic profile constructed along the crest of the west side of San Emigdio
Canyon. Late Pleistocene gravels are faulted and folded over the Wheeler Ridge fault and Los Lobos folds. Modified from Keller, E.A., Seaver, D.B.,
Laduzinsky, D.L., Johnson, D.L., Ku, T.L., 2000. Tectonic geomorphology of active folding over buried reverse faults: San Emigdio Mountain front,
southern San Joaquin Valley, California, 2000. Geological Society of America Bulletin 112(1), 8697.
142
Time 1
Time 1
Topographic profile
Basin
Antecedent
streams
Elev
Distance
A
Hinterland
Time 2
TD
Time 2
TD
TD
Elev
Water gap
WG
WG
Distance
D
C
TD
TD
TD
Time 3
TD
TD
Time 3
WG
AG
AG
Elev
Saddle
D
Distance
TD
Propagating
anticline
Capture
Figure 13 Idealized diagram showing the development of drainage

as two fault segments with anticlines propagate laterally toward one
another at Time 1. At Time 2 the distance between the propagating
folds narrows, and a water gap forms. With continued lateral
propagation (Time 3), the two folds may become sutured, and the
water gap is transformed into a topographic saddle. The result is a
longer, continuous fold where streams are diverted to both ends
where lateral propagation continues unimpeded.
5.7.8
Directivity of Earthquake Energy and Lateral

Fold Propagation: A Hypothesis of Tectonic
Extrusion
An earthquake may be considered as a process of rupture that

starts from an initial point on the fault plane (called the
C = Capture point
TD = Transverse
drainage
D = Deflected
drainage
WG = Water gap
AG = Air (wind) gap
Figure 14 Idealized diagram showing drainage that has developed

across a fold belt where most of the folds are propagating laterally.
This is only one possible scenario that might occur with complex
changes developing as folds propagate laterally and from the hinterland
into a basin. Through a series of antecedent positioning of streams and
of lateral deflections with stream capture, the drainage works its way
through a fold belt.
focus). Fault rupture does not occur instantaneously, and it

does not proceed in a uniform manner along the fault plane.
For example, during the Northridge (ca. Mw 6.7) event, the
earthquake ruptured the fault plane for B8 s, and the average
slip across the plane was about 1 m. During the earthquake,
the rupture propagated across the fault plane in a northwesterly direction at a speed of B3 km s1. However, the
propagation was not uniform; and some parts of the fault

plane experienced little or no slip, whereas others experienced
more than 3 m.
Directivity of seismic waves parallel to the rupture direction along the fault results in wave amplification when the
amplitude of seismic wave increases in the direction of fault
rupture. The direction in which the rupture occurs can greatly
affect the intensity of seismic shaking (strong ground motion)
(Benioff, 1955; Boore and Joyner, 1978). This principle is
illustrated in Figure 15, which shows that in the direction of
propagation of fault rupture the amplitude of the resultant
wave may be as high as 10 times the amplitude of the waves
in the reverse direction. This suggests that damages from
seismic shaking may be much greater in the direction of fault
rupture than in the opposite direction. For example, the 1994
Northridge earthquake, in the Western Transverse Ranges of
California, was up (north) and to the west, resulting in
stronger ground motion from seismic shaking to the northwest. The stronger ground motion to the northwest is believed
to be the result of the movement (fault rupture during the
earthquake) of a block of Earth moving upward and to the
northwest (US Geological Survey, 1996).
Shortening across the Western Transverse Ranges is the
result of right-slip and contraction in the vicinity of the
Big Bend of the San Andreas fault system (SAF). Right-slip
along the Big Bend produces (north and south of the SAF)
parallel to subparallel faults with predominantly left-lateral
strike-slip faults and left-oblique-slip reverse faults (if they
have a component of strike-slip). Major folds and (buried)
faults tend to propagate laterally to the west, south of the SAF,
and propagate to the east, north of the SAF. The authors hypothesize that the lateral propagation of fault rupture
during earthquakes tend to be to the west, south of the SAF;
and to the east, north of the SAF (Figure 16). At the north and
south of the extrusion zone, the plate bounding deformation terminates, and this may coincide with the left-lateral,
143
Santa Cruz Island and White Wolf faults, respectively

(Keller et al., 2000).
The concept of tectonic extrusion has been used in a variety
of tectonic settings, including the Mediterranean region
(McKenzie, 1972), Asia (Tapponnier et al., 1982), and the Los
Angeles Region (Walls et al., 1998). The driving mechanisms
for each of these locations are somewhat different, depending
upon specifics of the regional tectonic framework. In the case
of the Big Bend of the SAF, the extrusion is evidently driven by
the contraction produced by the bend in conjunction with the
orientation of left-lateral strike-slip faults, such as the White
Wolf, Santa Ynez, and Santa Cruz Island faults. An analogy
that has sometimes been drawn is the squeezing of a watermelon seed between the thumb and index finger until it pops
laterally. In the case of the Big Bend, we might envision two
watermelons seeds being squeezed, with one extruding to the
east, north of the fault, and the second to the west, south of
the fault.
In order for a fault to propagate laterally, it must rupture
and displace new ground in the direction of propagation.
Thus, earthquake ruptures would hypothetically tend to
propagate in the same direction as the growth of the fault and
associated folds. The hypothesis that direction of rupture
during earthquakes would tend to be the same as the direction
of propagation of the fault that produced the earthquake is
highly speculative. This leads to the question: Can we predict,
using geologic or geomorphic criteria, lateral direction of
earthquake rupture propagation? The answer to this question
is possibly yes in specific instances for buried reverse faults
that have produced an anticline with surficial expression
for which a single direction of lateral propagation may be
determined. In other cases, a fold may propagate laterally in
two directions, and some earthquakes may rupture laterally
in a direction opposite to the direction of the propagation of
the fold. Nevertheless, we hypothesize that the dominant
direction of lateral propagation of rupture during moderate
0
2
4
6
Wavelets
Down
8
X
Up
Resultant in the direction of rupture

Time
0
2
4
6
Wavelets
Active fault segment
Amplitude
0
Y
Amplitude
Direction of fault rupture
Resultant in the reverse direction

After H. Benioff, 1955
Figure 15 Concept of directivity increasing the amplitude of seismic waves and the direction of rupture propagation. Modified from Benioff, H.,
1955. Mechanism and strain characteristics of the White Wolf fault as indicated by the aftershock sequence. In: Oakeshott, G.B. (Ed.),
Earthquakes in Kern County, California during 1952. Division of Mines Bulletin, California, vol. 171, pp. 199202.
144
12000W
rnia
lifo
San Joaquin
Valley
An
dr
ea
lt
Extrusion
Big Ben
3500N
fa
u
Contraction
Ca
Sa
n
11900W
Mojave
Desert
Extrusion
SB
Kilometers
Contraction
0 10 20 30 40
3400N
Figure 16 Model of tectonic extrusion away from the Big Bend in the San Andreas fault (SAF). South of the SAF folds are propagating to the
west, whereas, north of the fault folds are propagating eastward. B, Bakersfield; M, Malibu; SB, Santa Barbara; V, Ventura. See text for
explanation. Modified from Keller, E., Gurrola, L., 2000. Earthquake hazard of the Santa Barbara fold belt. Final Report to the US Geological
Survey, NEHRP Award 527726, 78 pp. http://www.scec.org/research/98research/98gurrolakeller.pdf
to large earthquakes will be in the direction of the fold

propagation.
Although a very limited dataset exists with which to test
this hypothesis of lateral propagation of faulting and folding,
as well as rupture during earthquake, Table 1 provides some
examples from southern California earthquakes. The table lists
earthquakes and locations relative to the SAF, direction of
lateral propagation of rupture, and the lateral component of
displacement. Also indicated is whether a particular event
supports the hypothesis of westward propagation of structure
and rupture during earthquake, as well as the predicted leftoblique component of displacement.
The majority of earthquakes for which data are available
regarding direction of propagation of rupture and sense of
lateral displacement, generally supports the hypothesis that
fault rupture during moderate to large earthquakes tends
to have a westward component of propagation south of
the SAF, and an eastward component, north of the SAF. The
evidence is particularly good for the 1952 Mw7.5 Kern County
earthquake, which is north of the SAF and had left-oblique
displacement during the event the rupture propagated to
the northeast, as indicated by the pattern of aftershocks. The
1971 San Fernando Mw6.6 earthquake had a left-lateral
component of slip, which apparently increased from east

to west (Palmer and Henyey, 1971), but the direction of
propagation of the rupture was up and to the southeast.
The 1978 Santa Barbara earthquake was characterized by
left-oblique displacement, and evidence from the aftershock
locations indicate that the direction of propagation of the
rupture was to the northwest (Corbett and Johnson, 1982;
Yeats and Olson, 1984). Finally, the 1994 Northridge Mw6.7
demonstrated a component of northwest direction of lateral
propagation of rupture based upon aftershock locations (US
Geological Survey, 1996). Thus, the earthquakes of 1952,
1978, and 1994 (for which data exists concerning the direction of propagation of rupture) support the hypothesis that
there is a tendency for rupture propagation to have an eastward component north of the SAF and a westward component
south of the fault.
This hypothesis has important ramifications for earthquake
hazards because earthquake damage is generally most severe
in the direction of a propagating rupture (Benioff, 1955; US
Geological Survey, 1996) (Figure 16). Anticipating potential
direction of rupture during earthquakes will allow for better
modeling of potential damages to human structures as a result
of seismic shaking.
145
Table 1 Selected examples of Southern California earthquakes

Earthquake
Approximate
magnitude
1925 Santa
6.3
Barbara
1952 Kern Co. 7.5
1971 San
6.6
Fernando
Location
Direction of lateral Aftershock
relative to
propagation of
locations
San Andreas rupture
fault
Lateral component Supports

of displacement
hypothesis?
Yes (Y), No (N)
Comments references
South
West (?)
Not known
Many uncertainties
North
South
Northeast
Probably to
southeast
All northeast
Mostly
southeast
Probably
left-oblique
Left-oblique
Left-oblique
Y (displacement)
Y (displacement)
N (propagation)
1978 Santa
Barbara
5.9
South
Northwest
Northwest
Left-oblique
1994
Northridge
6.7
South
Northwest
Mostly
northwest
Not known
5.7.9
Conclusions
Study of the tectonic geomorphology of active faults and folds

from various locations across the world allows us to come to
the the following conclusions:
1. Investigations of the active tectonics of fold growth suggests
that folds may propagate laterally in one direction or two;
twofolds may propagate toward each other, producing a
suture zone with a longer segmented structure; and folds
may also propagate past each other to form en echelon
patterns.
2. A major challenge in the understanding of fold growth is
to establish the chronology that defines rates of lateral
propagation and incision of streams across growing folds.
3. Geomorphic criteria to recognize lateral propagation of
folding are straightforward, and several criteria are useful.
However, geomorphic criteria, although necessary to
demonstrate lateral fold propagation, are not sufficient to
prove the propagation. The strongest geomorphic evidence
to support lateral propagation includes the presence of
inherited drainage in the landscape that fans out from the
flanks and noses of actively growing, laterally propagating
folds. The presence of a series of wind gaps and water gaps
in the direction of lateral propagation that result from the
same stream being diverted laterally is strong evidence for
lateral propagation.
4. The development of transverse drainage across active folds
and fold belts is a complex process that involves the
interaction of several variables, including rate of uplift,
resistance of the rocks, unit stream power, and the nature
and extent of preexisting drainage.
5. A simple model for development of transverse drainage is
that a stream may establish a channel across the path of a
propagating fold before the fold arrives through lateral
propagation. That stream reach can be considered antecedent to the uplift. As fold growth continues, a water
gap may form which will either persist or eventually be
Oakeshott (1955)
Left-lateral
component of slip
apparently
increases east to
west (Palmer and
Henyey, 1971)
Corbett and Johnson
(1982), Yeats and
Olson (1984)
US Geological Survey
(1996)
defeated. Thus, drainages are established across the developing fold by a series of captures, diversions, and
antecedent positioning.
6. Understanding lateral propagation hypothetically may
have implications for better understanding the earthquake
hazard. If earthquake ruptures have a preferred direction of
propagation, then variability of seismic shaking and increase in shaking in the direction of propagation may be
better predicted. The hypothesis of tectonic extrusion states
that in the Big Bend area of the SAF, faults and folds tend to
propagate west, south of the fault, and to the east, north of
the fault.
References
Amos, C.B., Burbank, D.W., 2007. Channel width response to differential uplift.
Journal of Geophysical Research 112, F02010.
Benioff, H., 1955. Mechanism and strain characteristics of the White Wolf fault as
indicated by the aftershock sequence. In: Oakeshott, G.B. (Ed.), Earthquakes in
Kern County, California during 1952. Division of Mines Bulletin, California, vol.
171, pp. 199202.
Bloom, A.L., 1998. Geomorphology, Third ed. Prentice Hall, Upper Saddle River,
NJ, 482 pp.
Boore, D.M., Joyner, W.B., 1978. The influence of rupture incoherence on seismic
directivity. Bulletin of the Seismological Society of America 68(2), 283300.
Boudiaf, A., Ritz, J.-F., Philip, H., 1998. Drainage diversions as evidence of
propagating active faults: example of the El Asnam and Thenia faults, Algeria.
Terra Nova 35(10), 236244.
Burbank, D.W., McLean, J.K., Bullen, M., Abdrakhmatov, K.Y., Miller, M.M., 1999.
Partitioning of intermontane basins by thrust-related folding, Tien Shan,
Kyrgyzstan. Basin Research 11(1), 7592.
Burbank, D.W., Meigs, A., Brozovic, N., 1996. Interactions of growing folds and
coeval depositional systems. Basin Research 8(3), 199223.
Chen, J., Heemance, R., Burbank, D., Scharer, M., Miao, J., Wang, C., 2007.
Quantification of growth and lateral propagation of the Kashi anticline, southwest
Chinese Tien Shan. Journal of Geophysical Research 112, B03S16. http://dx.doi.
org/10.1029/2006/JB004345,2007.
Corbett, E.J., Johnson, C.E., 1982. The Santa Barbara California, earthquake of 13
August 1978. Bulletin of the Seismological Society of America 72A(6), 22012226.
146
Cowie, P.A., Scholz, C.H., 1992. Growth of faults by accumulation of seismic slip.
Journal of Geophysical Research 97(B7), 11,08511,095.
Dawers, N.H., Anders, M.H., 1995. Displacementlength scaling and fault linkage.
Journal of Structural Geology 17(5), 607614.
Delcaillau, B., Carozza, J., Laville, E., 2007. Recent fold growth and drainage
development: the Janauri and Chandigarh anticlines in the Siwlik foothills,
northwest India. Geomorphology 76, 241256.
Delcaillau, B., Deffontaines, B., Floissac, L., et al., 1998. Morphotectonic evidence
from lateral propagation of an active frontal fold; Pakuashan anticline, foothills
of Taiwan. Geomorphology 24, 263290.
DeVecchio, D.E., Keller, E.A., Fuchs, M., Owen, L.A., 2012. Late Pleistocene
structural evolution of the Camarillo fold belt: Implications for seismic hazard
and fold growth in Southern California: Lithosphere, vol. 4. doi:10.1130/L136.1.
Duvall, A., Kirby, E., Burbank, D., 2004. Tectonic and lithologic controls on bedrock
channel profiles and processes in coastal California. Journal of Geophysical
Hetzel, R., Tao, M., Niedernann, S., Strecker, M., Ivy-Ochs, S., Kubik, P., Gao, B.,
2004. Implications of the fault scaling law for the growth of topography: mountain
ranges in the broken foreland of north-east Tibet. Terra Nova 16(3), 157162.
Jackson, J., Norris., R., Youngson, J., 1996. The structural evolution of active fault
and fold systems in central Otago, New Zealand: evidence revealed by drainage
patterns. Journal of Structural Geology 18, 217234.
Jackson, J., Ritz, J.-F., Siame, L., et al., 2002. Fault growth and landscape
development rates in Otago, New Zealand, using in situ cosmogenic 10Be. Earth
and Planetary Science Letters 195(34), 185193.
Keller, E., Gurrola, L., 2000. Earthquake hazard of the Santa Barbara fold belt. Final
Report to the US Geological Survey, NEHRP Award 527726, 78 pp.
Keller, E.A., Gurrola, L., Tierney, T.E., 1999. Geomorphic criteria to determine direction
of lateral propagation of reverse faulting and folding. Geology 27(6), 515518.
Keller, E.A., Pinter, N., 2002. Active Tectonics. Prentice Hall, Upper Saddle River,
NJ, 362 pp.
Keller, E.A., Seaver, D.B., Laduzinsky, D.L., Johnson, D.L., Ku, T.L., 2000. Tectonic
geomorphology of active folding over buried reverse faults: San Emigdio
Mountain front, southern San Joaquin Valley, California, 2000. Geological
Society of America Bulletin 112(1), 8697.
Keller, E.A., Zepeda, R.L., Rockwell, T.K., Ku, T.L., Dinklage, W.S., 1998. Active
tectonics at Wheeler Ridge, southern San Joaquin Valley, California. Geological
Society of America Bulletin 110(3), 298310.
McKenzie, D., 1972. Active tectonics of the Mediterranean region. Geophysical

Journal of the Royal Astronomical Society 30(2), 109185.
Medwedeff, D.A., 1992. Geometry and kinematics of an active, laterally propagating
wedge thrust, Wheeler Ridge, California. In: Mitra, S., Fisher, G.W. (Eds.),
Structural Geology of Fold-and-Thrust Belts. Johns Hopkins University Press,
Baltimore, MD, pp. 328.
Mueller, K., Talling, P., 1997. Geomorphic evidence for tear faults accommodating
lateral propagation of an active fault-bend fold, Wheeler Ridge, California.
Journal of Structural Geology 19(34), 397411.
Namson, J., Davis, T., 1988. Seismically active fold and thrust belt in the San
Joaquin Valley, central California. Bulletin of the Geological Society of America
77, 359385.
Namson, J., Davis, T., 1992. Late Cenozoic thrust ramps of Southern California.
Final Report for 1991 Contract, Southern California Earthquake Center.
Oakeshott, G.B. (Ed.), 1955. Earthquakes in Kern County, California during 1952,
vol. 171. California Division of Mines Bulletin, 277.
Palmer, D.F., Henyey, T.L., 1971. San Fernando earthquake of 9 February 1971,
pattern of faulting. Science 172, 712715.
Ramsey, L.A., Walker, R.T., Jackson, J., 2008. Fold evolution and drainage
development in the Zagros Mountains of Fars Province, SE Iran. Basin Research
20(1), 2348.
Tapponnier, P., Peltzer, G., Ledan, A.Y., 1982. Propagating extrusion tectonics in
eastern Asia. EOS 63(45), 1097.
Tierney, T.E., 2002. Tectonic geomorphology of the greater Santa Ynez Mountains,
Santa Barbara County, California. Unpublished MA Thesis, University of
California Santa Barbara.
US Geological Survey, 1996. USGS response to an urban earthquake Northridge
94. USGS Geological Survey Open File Report 96263, 78 pp.
Walls, C., Rockwell, T., Mueller, K., et al., 1998. Escape tectonics in the Los Angeles
metropolitan region and implications for seismic risk. Nature 394, 356360.
Yeats, R.S., 1986. Active faults related to folding. In: Active Tectonics, National
Academy Press, Washington, DC.
Yeats, R.S., Olson, D.J., 1984. Alternate fault model for the Santa Barbara,
California earthquake of 13 August 1978. Bulletin of the Seismological Society
of America 74(5), 15451553.
Biographical Sketch
Edward A Keller is a professor, researcher, writer, mentor, and teacher to undergraduate and graduate students.
Currently, Kellers students are working on earthquake hazards, forms and processes of small coastal lagoons, and
geologic controls on habitat of the endangered southern steelhead trout. He was born and raised in California
(Bachelors degrees in Geology and Mathematics from California State University at Fresno, Masters degree in
Geology from the University of California at Davis); it was while pursuing his PhD in Geology from Purdue
University in 1973 that Ed wrote the first edition of Environmental Geology, a text now in the Ninth edition that
became the foundation of an environmental geology curriculum in many universities and colleges. Other widely
used textbooks include Active Tectonics with Nicholas Pinter, Second edition, 2002; Introduction to Environmental Geology Fifth ed., 2012; and Environmental Science, Eighth edition with Daniel Botkin, 2011. Keller
joined the faculty of the University of California at Santa Barbara in 1976 and has been there since, serving
multiple times as the chair of both the Environmental Studies and Hydrologic Science programs. During that
time, he authored more than 130 articles, books, and reports, including seminal works on pool-riffle dynamics in
alluvial and bedrock channels, role of large woody debris on stream channel form and process, and tectonic
geomorphology of active fold growth. Kellers academic honors include the Don J Easterbrook Distinguished
Scientist Award, Geological Society of America (2004), the Quatercentenary Fellowship from Cambridge University, England (2000), two Outstanding Alumnus Awards from Purdue University (1994, 1996), a Distinguished
Alumnus Award from California State University at Fresno (1998), and the Outstanding Outreach Award from the
Southern California Earthquake Center (1999).
147
Duane DeVecchio is adjunct faculty at the City College of San Francisco and is an Assistant Researcher at the Earth
Research Institute at the University of California Santa Barbara, where he earned his PhD in geology. DeVecchio
has a broad field-based background in the Earth sciences and likes to relate his experience of living in trailers in
the middle of nowhere mapping rocks. For his Masters Degree and post-Masters research, he conducted structural
and stratigraphic analysis, as well as numerical dating of upper plate volcanic and volcaniclastic rocks in southeast
Idaho and the central Mojave Desert of California, which record the Miocene extensional histories of these
regions. His PhD research was aimed at resolving fault slip rates and quantifying the earthquake hazard presented
by several active fault-related folds growing beneath urbanized Southern California. DeVecchios current research
interests focus on the timing and rates of change of Earths surface due to depositional and erosional processes
that result from climate change and tectonics. Since commencing his graduate education, Duane has devoted a
significant amount of time to becoming an effective communicator of science to todays students. He is a passionate
teacher and feels strongly that students need to develop the ability to critically evaluate data presented in numerous
forms from various sources. He believes this is particularly important in todays world where the internet offers
accessibility to vast amounts of information, yet the validity of this information is often questionable or misleading.
When DeVecchio is not conducting research or teaching, he enjoys whitewater rafting, rock climbing, snowboarding,
and camping with his partner, Christy.
5.8 Volcanic Landforms and Hazards

WD Huff and LA Owen, University of Cincinnati, Cincinnati, OH, USA
5.8.1
5.8.2
5.8.2.1
5.8.2.2
5.8.2.3
5.8.3
5.8.3.1
5.8.3.2
5.8.3.3
5.8.3.3.1
5.8.3.3.2
5.8.3.3.3
5.8.3.3.4
5.8.3.3.5
5.8.3.3.6
5.8.3.4
5.8.3.4.1
5.8.3.4.2
5.8.3.4.3
5.8.3.4.4
5.8.3.4.5
5.8.3.4.6
5.8.3.4.7
5.8.3.4.8
5.8.3.4.9
5.8.4
5.8.4.1
5.8.4.2
5.8.4.3
5.8.4.4
5.8.4.5
5.8.4.6
5.8.4.7
5.8.4.8
5.8.4.9
5.8.4.10
5.8.5
5.8.5.1
5.8.5.1.1
5.8.5.1.2
5.8.5.2
5.8.5.2.1
5.8.5.2.2
5.8.6
5.8.6.1
5.8.6.1.1
5.8.6.1.2
5.8.6.1.3
5.8.6.1.4
5.8.6.1.5
Introduction
Tectonic Settings
Subduction-related Volcanism
Hot-Spots
Zones of Extension
Variety of Volcanic Landforms
Classification of Volcanic Landforms
Major Controls on Landform Types
Major Types of Volcanic Eruption
Pelean eruptions
Plinian eruptions
Vesuvian eruptions
Vulcanian eruptions
Strombolian eruptions
Other eruptions
Major Types of Volcanoes
Cinder (Scoria) cone volcanoes
Composite volcanoes
Shield volcanoes
Lava domes
Caldera versus crater
Submarine volcanoes
Mud volcanoes
Geysers, fumaroles, hot springs
Lava tubes
Evolving Volcanic Landforms
Parcutin Cinder Cone
Kavachi, Solomon Islands
Surtsey
Iceland
Kilauea
Mount St. Helens
Mount Rainer
Krakatau
Mount Etna
Canary Islands
Ancient Volcanic Settings
Supervolcanic Eruptions
Toba caldera
Yellowstone caldera
Flood Basalts
Columbia river basalt group
Deccan traps
Volcanic Hazards
Direct Hazards
Pyroclastic flow
Tephra
Direct blast
Lava flows
Dangerous gases
150
150
150
151
152
152
152
152
154
154
154
155
155
156
156
157
157
158
160
160
161
163
163
164
166
167
167
169
169
169
169
170
171
171
175
175
175
175
176
178
179
179
180
181
181
181
181
182
182
183
Huff, W.D., Owen, L.A., 2013. Volcanic landforms and hazards. In: Shroder,
J. (Editor in Chief), Owen, L.A. (Ed.), Treatise on Geomorphology.
Academic Press, San Diego, CA, vol. 5, Tectonic Geomorphology,
pp. 148192.
148
http://dx.doi.org/10.1016/B978-0-12-374739-6.00089-0
Volcanic Landforms and Hazards

5.8.6.2
Indirect Hazards
5.8.6.2.1
Tephra
5.8.6.2.2
Lahar
5.8.6.2.3
Debris avalanches, landslides, and tsunamis
5.8.6.2.4
Climate change
5.8.6.3
Volcanic Hazard Mitigation
5.8.7
Future Challenges in the Study of Volcanic Landforms and Hazards
Acknowledgments
References
Glossary
Aerosol A suspension of fine solid particles or liquid
droplets in a gas.
Albedo Reflectivity or reflecting power of a surface.
Alkalic Pertaining to igneous rock that contain more
alkali metals (mainly sodium and potassium) than is
considered average for that series of rocks.
Basalt Fine grained volcanic rock, generally dark gray to
black in color, comprising mostly plagioclase feldspar
(465%), plus ferromagnesian minerals, with minor
amounts of quartz (o20%).
Basaltic Pertaining to, composed of, containing, or
resembling basalt.
Carbonatite Carbonate rock of magmatic origin.
Crust Outer-most layer of Earth defined on density,
composition, and seismic velocity. Base defined by the
Mohorovicic discontinuity, a zone of sharp change in
seismic velocity.
Debris flow Moving mass of rock and sediment, which
advances as a plastic flow.
Electric distance measuring equipment (EDM) Survey
equipment used to determine distances heights and
directions, which measures and stores data using a
combination of electronic means such as lasers and radar.
Geographical information systems (GIS) Computerized
technique designed to capture, store, manipulate, analyze,
manage, and present all types of geographically referenced
data.
Geyser A type of hot spring that intermittently erupts jets
of hot water and steam.
Glacioisostatic rebound Crustal uplift related to
unloading of the crust after glacier/ice sheets melt.
on Earth.
Hot-spots Volcanic center, 100200 km across, that
persists for a few tens of millions of years.
Hydrothermal Pertaining to hot water, or the interaction
and products of hot water.
Hydrovolcanic Pertaining to the interaction of water and
volcanic processes.
Ice core Core sample that has been collected by drilling
through a glacier or ice sheet.
Ignimbrite Rock formed by the widespread
deposition and consolidation of ash flows
(nuees ardentes).
149
183
183
183
183
185
187
189
189
189
Jokulhlaups Icelandic name for a glacial lake outburst

flood (GLOFs). The name is also often applied to GLOFs in
regions outside of Iceland.
K-bentonites Volcanic ash that has altered to expandable
clay (bentonite) and then adsorbed a significant amount of
potassium (K) during diagenesis.
Lahars Debris flow/mudflow comprising volcaniclastic
material.
Lapilli Solid and/or viscous pyroclastic material that has a
particle size range between 2 and 64 mm.
Limnic Pertaining to a body of freshwater.
Lithosphere Rigid outermost shell of Earth, comprising
the crust and upper part of the mantle.
Mafic Pertaining to igneous rocks that are chiefly
composed of ferromagnesian minerals.
Magma Mobile rock material generated within Earth and
capable of intrusion and extrusion; and forms igneous rock
when crystallized.
Mantle Zone of Earth below its crust and above its core;
beginning at depth of between 10 and 30 km and extending
to B2900 km.
Mantle plume Thermal column of abnormally hot rock
that forms from the core-mantle boundary and rises
through Earths mantle into the crust.
Nuees ardent An extremely fast flowing descending
gaseous cloud of volcanic ash.
Phreatomagmatic Pertaining to volcanic processes
involving magmatic gases and steam.
Pillow lavas Lavas that are generally shaped and of
the size of pillows, which form in subaqueous
environments.
Pyroclastic Pertaining to broken and fragmented rock
material formed during a volcanic eruption involving aerial
ejection.
Rhyolitic Pertaining to a group of igneous rocks that are
fine grained and contain abundant quartz (420%) and
alkali feldspars.
Scoria Ejected magma and/or rock fragments, 464 mm
in size, which are irregular in form and generally contain
vesicles (a cavity of variable shape in lava formed by the
entrapment of a gas bubble during solidification).
Seamount A discrete elevated part of the sea floor that
may be flat topped or peaked, rising more than 1000 m
above the surrounding area.
Stratosphere Second layer of the atmosphere, at an
altitude of above B10 km and below B50 km.
150
Tephra A general term for all pyroclastic material (ash,

lapilli, and scoria).
Tephrochronology The study and use of tephra to date
geologic successions.
Tholeiitic Pertaining to igneous rocks that are
oversaturated in silica, generally used for basalts.
Troposphere Lower layer of the atmosphere below

B10 km above Earths surface.
Tuff A general term for all consolidated pyroclastic rocks.
Volcanic bombs Also known as scoria.
Xenolith A foreign inclusion (commonly parts of the
intruded surrounding country rock) within an igneous rock.
Abstract
Volcanic centers are complex, dynamic, and continuously evolving landforms. As a consequence, volcanic landforms are
highly diverse and controlled by the tectonic setting, magma type and volume, surface environment, and their age. Active
volcanoes occur in most tectonic settings. Most of Earths active volcanoes, however, are subduction-zone related and occur
around the margins of the Pacific. Volcanic hazards are both direct and indirect, involving pyroclastic materials, lava flows,
poisonous gases, landslides, debris flows, floods, and climate change. Mitigating volcanic hazards and disaster management
require understanding of the geologic context and processes, monitoring of volcanic centers, and disaster preparedness.
5.8.1
Introduction
Active volcanoes and their associated features are dynamically

evolving landforms, produced and influenced both by Earths
internal and surface processes. Volcanic landforms are essentially the result of the tectonic setting, magma composition
and volume, surface environment, and the time the landform
has existed and evolved. Since these variables are large, the
varieties of different types of volcanic landforms are considerable. The influence that volcanism exerts on a regional
landscape is the result of a number of different factors. These
include the nature and pattern of various fissures and vents,
the length of time that volcanism is active, the relative age of
volcanism with respect to the various stratigraphic units, the
composition and physical characteristics of extruded materials, the volume of erupted material, and the amount and
extent of subsequent erosion. In some cases, lava flows and
thick tephras that accumulate over large areas may partially or
completely bury the preexisting landscape, whereas in others a
limited distribution of lavas and tephras may produce a distinctive assemblage of lava-capped hills and mesas.
The vibrant nature of volcanic landforms means that they
commonly constitute a significant natural hazard. The Smithsonian National Museum of Natural History (2011a) estimates
that there are about 20 terrestrial volcanoes actively erupting
on any given day. Roughly, 50 to 70 terrestrial volcanoes have
erupted over the course of the last year, and 160 erupt during a
single decade. There are 550 volcanoes that have been active in
all recorded human history, and an additional 1300 terrestrial
volcanoes have been identified that have erupted in the last
10 000 years. This makes the total about 1500 volcanoes that
have erupted in the last 10 000 years. But most of the actively
erupting volcanoes are on the deep sea floor, and thus more
difficult to detect and measure. If one assumes that three-fourth
of all eruptions are from underwater volcanoes, then a total of
about 6000 volcanoes have erupted worldwide in the last
10 000 years. Over the fullness of geologic time, this constitutes
a major contribution to the morphology of Earths landscapes.
This chapter examines the tectonic geomorphology of
volcanic landforms discussing their tectonic setting, the range
of forms, the nature and duration of how volcanic landforms
develop and evolve, and the nature of volcanic hazards and

volcanic hazard mitigation.
5.8.2
Tectonic Settings
Volcanoes are not randomly distributed over Earths surface.

Most are concentrated on the edges of continents, along island
chains, or beneath the sea forming long mountain ranges
(Figure 1). More than half of the worlds active volcanoes
above sea level encircle the Pacific Ocean to form the circumPacific zone known as the Ring of Fire. Volcanic landforms
occur in nearly every tectonic setting, but are most commonly
associated with subduction zones and hot-spots. However,
volcanism can occur even in immensely thickened crustal
regions once they begin to experience extension such as the
Tibetan Plateau. Moreover, volcanism is commonly associated
with other zones of extension within continental interiors
such as the Basin and Range Province in the western US. The
American West is one of the most unusual volcanic provinces
on Earth. It is an extraordinary wide region of contemporaneous volcanism, stretching B1800 km from the Rocky
Mountains west to the Cascades. It encompasses a wide variety
of volcanic landforms and diverse rock compositions, with the
only consistency being small basalt fields that are scattered
throughout the region. Intriguingly, except for the Cascadian
segment, there is no currently active plate subduction to drive
the volcanism, although there are hot-spots, persistently
leaky faults, and rifting. Commonly, areas of continental
volcanism are related to, or are a consequence of extension
produced by hot-spots such as in the East African Rift, the Afar
and the northern margin of the Red Sea. This section will
briefly outline the main tectonic settings for the common
types of volcanic landforms.
5.8.2.1
Subduction-related Volcanism
Approximately 94% of all terrestrial volcanoes occur in subduction zones with 90% of volcanoes occurring within only
nine countries (Indonesia, Japan, United States, Russia, Chile,
Philippines, New Guinea, New Zealand, and Nicaragua) in the
151
North America
plate
Eurasia
plate
Juan
de fuca
Philippine
plate
plate
Hawaiian
hot spot
Cocos
plate
Pacific
plate
Australia
plate
Pacific
plate
Eurasia
plate
India
plate
North
America
plate
Caribbean
plate
Arabia
plate
Nazca
plate
Antarctic
plate
Antarctic
plate
Africa
plate
South
America
plate
Antarctic
plate
Scotia plate
Figure 1 Tectonic plates and active volcanoes (red triangles) of the world. Reproduced from Tilling, R.I., Heliker, C., Swanson, D.A., 2010.
Eruptions of Hawaiian Volcanoes - Past, Present and Future, Second edition. USGS General Information Product 117, 63 pp.
Ring of Fire. Therefore, much of our discussion in this chapter

will focus on volcanic landforms associated with this setting.
The Ring of Fire occurs because most of the margins of the
Pacific Ocean coincide with converging margins along which
subduction is occurring. The convergent boundary along the
coasts of South America, Central America, Mexico, the
northwestern US (Northern California, Oregon, and Washington), western Canada, and eastern Alaska, are boundaries
along which oceanic lithosphere is being subducted beneath
continental lithosphere. This has resulted in the formation of
continental volcanic arcs that form the Andes Mountains, the
Central American Volcanic Belt, the Mexican Volcanic Belt, the
Cascade Range, and the Alaskan volcanic arc. The Aleutian
Islands (west of Alaska), the KurilKamchatka Arc, Japan,
Philippine Islands, and Marianas Islands, New Zealand, and
the Indonesian Islands, along the northern and western margins of the Pacific Ocean are zones where oceanic lithosphere
is being subducted beneath oceanic lithosphere. These are all
island arcs. Basaltic magmas are generated by flux melting of
the mantle overlying the subduction zone. Through magmatic
differentiation, basaltic magmas change to andesitic and
rhyolitic magma. Because these magmas are generally gas rich
and all have relatively high viscosity, eruptions in these areas
tend to be violent. Resulting volcanic landforms tend to be
cinder cones, stratovolcanoes, volcanic domes, and calderas.
Repose periods between eruptions tend to be hundreds to
thousands of years, thus giving people living near these
volcanoes a false sense of security (Rampino and Ambrose,
2000).
5.8.2.2
Hot-Spots
Chapter 5.9 describes hot-spots in great detail, but we will

briefly outline the locations of hot-spots and describe the differences between oceanic (on-ridge and off-ridge) and
continental hot-spots. Hot-spots are places within the
mantle where rocks melt to generate magma. The presence of a

hot-spot is inferred by anomalous volcanism, such as the Hawaiian volcanoes within the Pacific Plate. In the 1960s,
geologists were seeking ways to prove or disprove the new idea
of moving plates. Exploration of magnetic anomalies at midocean ridges provided strong support for seafloor spreading.
This led to studies of other ocean features to see how they
might be related to plate tectonics. While visiting Hawaii,
J. Tuzo Wilson, one of the founders of the theory of plate
tectonics, noticed some interesting features about ocean islands. On a map of the Pacific basin, Wilson found three linear
chains of volcanoes and submarine volcanoes (seamounts). Of
the three, the Hawaii-Emperor seamount chain was the most
well known. Wilson reviewed the reports that had been published on these island chains and recorded the age of each
island. An interesting pattern emerged. For each chain, the islands become progressively younger to the southeast. Active
volcanoes mark the extreme southeastern end of each chain.
Wilson proposed that the Hawaiian islands formed successively over a common source of magma called a hot-spot
(Wilson, 1963), and this model continues to guide our
thinking today. In the early 1970s, W. Jason Morgan proposed
what is now known as the plume hypothesis. This is the view
that hot mantle plumes rise from great depths through the
large-scale circulation systems caused by plate tectonics to
produce relatively fixed hot-spots. The need for a rapid supply
of hot mantle has made the so-called thermal boundary layer
at the coremantle boundary a prime candidate for the source
of the plumes. The greater density of the core prevents convection across that boundary, so heat must be transported
upward by conduction. But to conduct heat as rapidly as
convection transports it around the core and mantle, the
thermal boundary layer establishes a much greater temperature
gradient and hence this creates a plume (Morgan, 1971).
More recently, some geophysicists have suggested other
thermal boundary layers in the mantle as sources of heat
(Elkins-Tanton, 2005; Hooper et al., 2007). Whatever the
152
source, as a hot mantle plume rises and decompresses, it

partially melts in the shallow upper mantle, creating magma.
Once that magma breaches the surface, it flows as lava and
then solidifies into basalt. The chemistry of hot-spot magmas,
however, differs from those of the mid-ocean ridges in ways
that suggest a source that includes some relatively primitive
and presumably deep parts of Earth. In particular, hot-spot
magmas commonly have elevated values in the ratio of
primordial helium to radiogenic helium (3He/4He), which
indicates their source in a part of Earth that has not lost its
dissolved gases to the atmosphere. An active volcano resides at
one end of a hot-spot track. But wherever a major hot-spot
track can be traced back to its origin on the plate, it is found to
begin with a flood-basalt event. The largest volcanic events
created by our planet, flood-basalt eruptions occur every 20
million years on average. To account for the eruptions, many
researchers invoke a model in which the columnar plume
develops both a bulbous head and a trailing narrow tail while
rising through viscous mantle (Tarduno et al., 2009). Besides
the flood basalts, one predicted consequence of plumes is a
1- to 2-km rise in Earths surface a few million years before the
hot-spot erupts. That prediction makes the search for uplift
preceding flood volcanism an important line of research
(Hooper et al., 2007).
The Hawaiian hot-spot has been active for at least 70 Ma,
producing a volcanic chain that extends B6000 km across the
northwest Pacific Ocean. Hot-spots also develop beneath
continents. For example, the Yellowstone hot-spot has been
active for at least 15 Ma, producing a chain of calderas and
volcanic features along the Snake River Plain that extends
B650 km westward from northwest Wyoming to the
IdahoOregon border (Christiansen, 2001). There are dozens
of volcano hot-spots around the world, with the Hawaiian
Island chain just being the most well known. Others include
the Azores hot-spot, the Canary hot-spot and the East Australia
hot-spot. A particularly noteworthy hot-spot track extends
from India to the island of Reunion. About 66 to 68 Ma,
present-day India was above the hot-spot and great volumes of
basaltic lava erupted to produce the Deccan Traps. As the plate
moved northeast over the hot-spot more volcanic centers
formed: the Maldives from 55 to 60 Ma, the Chagos Ridge
from 48 Ma, the Mascarene Plateau from 40 Ma, and the
Mauritius Islands from 18 to 28 Ma. The youngest volcanoes,
Piton des Neiges and Piton de la Fournaise, formed in the last
5 Ma. The summits of these volcanoes make the island of
Reunion. Piton des Neiges is extinct and Piton de la Fournaise
is one of the most active volcanoes on Earth (Sen, 2001).
5.8.2.3
Zones of Extension
Extension zone activities include rift valley volcanism, midocean ridge volcanism, and Basin and Range-type volcanism.
Active volcanism is currently taking place along all of the
worlds oceanic ridges, but most of this volcanism is submarine volcanism and does not generally pose a threat to
humans. One of the few places where an oceanic ridge reaches
above sea level is at Iceland, along the Mid-Atlantic
Ridge, which is also coincident with a hot-spot. Here, most
eruptions are basaltic in nature, however, many are explosive
strombolian types or explosive phreatic or phreatomagmatic

types (see below for explanation of eruption types). Volcanism
also occurs in continental areas that are undergoing episodes
of extensional deformation. A classic example is the East
African Rift Valley, where the African plate is being split (Corti,
2009). The extensional deformation occurs because the
underlying mantle is rising from below and stretching the
overlying continental crust. Upwelling mantle may melt to
produce magmas, which then rise to the surface, commonly
along normal faults produced by the extensional deformation.
Basaltic and rhyolitic volcanism is common in these areas. In
the same area, the crust has rifted apart along the Red Sea, and
the Gulf of Aden to form new oceanic ridges (Davison et al.,
1994). This may also be the fate of the East African Rift Valley
at some time in the future. Other areas where extensional
deformation is occurring within the crust include the Basin
and Range Province of the western US (eastern California,
Nevada, Utah, Idaho, western Wyoming, and Arizona) and the
Rio Grande Rift, New Mexico. These are also areas of recent
basaltic and rhyolitic volcanism.
5.8.3
Variety of Volcanic Landforms
Generally speaking, different tectonic settings produce different types of volcanic landforms. These are classified on the
basis of the type of eruption and the shape and size of the
resultant landform. Below is a simplified classification, adapted from Rittmann (1962) and Summerfield (1991) that includes essentially all of the worlds best-studied volcanoes as
well as the geologic processes that both form them and act on
them after they have formed. Thus, a given volcanic landform
will be characteristic of the types of material it is made of,
which in turn depends on the prior eruptive behavior of the
volcano. Although later processes can modify the original
landform, it is possible to find clues in the modified form to
enable the formative processes to be determined.
5.8.3.1
Classification of Volcanic Landforms
Figure 2 shows some of the wide variety of different volcanic

landforms. This is loosely classified on the type and volume of
magma, and eruption style. These controls on landform type
are considered in more detail below.
5.8.3.2
Major Controls on Landform Types
Magma type and evolution of magma chambers have a profound control on landform types. Magma chambers are subterranean reservoirs containing molten silicate fluid. Some are
staging areas for volcanic eruptions, whereas others have no
link to Earths surface but instead undergo cooling and solidification to form coarse-grained plutons. When basaltic
magmas are repeatedly emplaced into continental crust, a
long-lived silicic magma chamber may form through the
processes of magmatic differentiation, partial melting, assimilation, or magma mixing (Schmincke, 2004). Differentiation involves the changing of magma composition by the
removal of denser early-formed ferromagnesian minerals
153
3a
2
1
3b
10
12
11
13
(a)
Type of magma
Fluid, very hot, basic

in composition
Increasing viscosity, gas

content, and acidity (high
proportion of silica)
Viscous, relatively cool,

acidic
Extremely viscous,
abundant crystals
Type of
activity
Effusive
Quantity of magma produced

Small
Lava flows1
Large
Exogenous9 domes
Collapse
Basalt domes and
shield volcanoes
lcelandic3a
Hawaiian3b
Scoria cones with

flows4
Mixed
Loose tephra cones
with thick flows5
Composite or stratovolcanoes6
Endogenous domes
(plug domes, tholoids)7
Ruptured endogenous domes

of thick lava
flows8
Explosive
Maars of tephra9
Maars with ramparts10
Explosive
mostly gas
Gas maars11
Explosion craters12
Volcanic fields with

multiple domes
Collapse
and explosion
calderas3
Ignimbrite sheets
(b)
Figure 2 The main types of volcanoes. Reproduced from Rittmann, A. 1962. Volcanoes and their Activity. Wiley, New York, 305 pp. (a)
Simplified sketches showing the main forms and (b) their relationship to magma type and volume. The numbers in the table relate to the
numbers next to the sketches.
through crystal settling. Partial melting of surrounding country rock produces magmas less mafic than their source rocks,
because lower melting point minerals are more felsic in
composition. Assimilation occurs when a hot magma melts
and incorporates more felsic surrounding country rock. The
evolution of magma chambers is of major petrologic interest
and has been studied extensively for many years. Experimental
and theoretical studies that have investigated the cooling of
magma chambers have demonstrated that a magma chamber

solidifies through formation of a mushy boundary layer
composed of crystals and interstitial melt along the chamber
walls (Naslund and McBirney, 1996).
Prior to the development of these ideas in the late nineteenth century, magmas were generally regarded as originating
from two distinct sources, one silica-rich and the other silicapoor. Intermediate lavas were explained as being a mixture of
154
these two sources. At the same time, some geologists suggested

these different magma sources originated from concentrically
distributed zones within Earth and others argued for a secular
change in erupted magma compositions. Early ideas about
differentiation began to emerge in the mid- to late-1800s with
the recognition of mineralogical and chemical similarities
among suites of igneous rocks, and then refined in the early
1900s by N.L. Bowen and others at the Geophysical Laboratories in Washington, DC who conducted experimental studies into the order of crystallization of the common silicate
minerals from a magma (Bowen, 1928; Young, 1998).
5.8.3.3
Major Types of Volcanic Eruption
The nature of an eruption has a profound influence on the

development of the landscape, including the destruction of
the volcanic landforms, creation of new volcanic centers, deposition of pyroclastic debris, and possible triggering of mass
movements. Explosive eruption types are named after infamous volcanic eruptions and are known as Pelean, Plinian,
Vesuvian, Vulcanian, and Strombolian eruptions (Schmincke,
2004; Figure 3).
A quantitative measure called the volcanic explosivity index
(VEI) was devised in 1982 to provide a relative measure of the
explosiveness of volcanic eruptions (Newhall and Self, 1982;
Table 1). The volume of material erupted, the eruption cloud
height, and qualitative observations such as cataclysmic and
explosive are used to determine the explosivity value. The
scale is open-ended with the largest volcanoes in history given
magnitude 8. A value of 0 is given for nonexplosive eruptions,
defined as less than 10 000 m3 of tephra ejected; and 8 representing a mega-colossal explosive eruption that can eject
1000 billion m3 of tephra and have a cloud column height of
over 50 km. The scale is logarithmic with each interval on the
scale representing a ten-fold increase in observed eruption
criteria. A total of 47 eruptions of VEI 8 magnitude or above,
ranging in age from Ordovician to Pleistocene, have been
identified, of which 42 occurred in the past 36 Ma. In 186 AD,
New Zealands Taupo volcano erupted with the fifth largest
Fissure icelandic
type
Strombolian
type
caldera-forming eruption in recorded history, which means

that there have not been any eruptions within the last 1000
years or so with a VEI of 8 (Mason et al., 2004).
5.8.3.3.1
Pelean eruptions
Pelean, or Nuees Ardente (firey cloud) eruptions occur when a

large quantity of gas, dust, ash, and lava fragments is ejected
out of a volcanos central crater. Pelean eruptions start when a
water-rich magma is suddenly released to the surface. The
magmas water flashes to steam, turning the magma into foam
and then ripping the foam into small shards. The shards and
pumice and gas flow together as if they were a liquid. At
600 1C, they are sometimes more fluid than water (Moore,
1967). The ejected material falls back toward Earth and then
travels down the side of the volcano at tremendous speeds,
typically faster than 150 km h1, as a pyroclastic flow
(Schmincke, 2004, see also discussion below). The initial
phases of the eruption are characterized by pyroclastic flows,
which constitute one of the most dangerous kinds of eruptions. Tephra deposits formed during pelean eruptions have a
lower volume and range than those produced during Plinian
and Vulcanian eruptions (see below). Viscous magma may
then emerge to form a steep-sided dome or volcanic spine in
the volcanos vent. The dome may later collapse, resulting in
additional flows of ash and hot blocks. The resulting landforms are a combination of these dome-formation, nuees
ardentes, and lateral blast events. Pelean eruptions derive their
name from Mount Pelee, the volcano that caused tremendous
destruction on Martinique, Lesser Antilles in 1902 (Francis,
1993; Schmincke, 2004). The eruption and related pyroclastic
flows killed more than 30 000 people, and the town of
St. Pierre was totally obliterated, constituting the worst volcanic disaster of the twentieth century.
5.8.3.3.2
Plinian eruptions
Plinian eruptions produce huge clouds of volcanic ash rising

up from a giant cinder cone. These eruptions are named after
the Roman statesman Pliny the Younger, who witnessed the
eruption of Italys Mount Vesuvius in 79 AD, which destroyed
Hawaiian type
Volcanian
type
Plinian type
Vesuvian
type
Pelean type
Figure 3 Schematic figure illustrating the main types of volcanic eruptions. Reproduced from Holmes, A. 1965. Principles of Physical Geology,
Second edition. Ronald Press, New York.
0
1
2
3
4
5
6
7
8
Source: Based on Newhall, C.G., Self, S., 1982. The volcanic explosivity index (VEI): an estimate of explosive magnitude for historical volcanism. Journal of Geophysical Research 87, 12311238.
Many
Many
3477
868
421
166
51
5 ( 2 suspected)
0
Kilauea
Stromboli
Galeras (1993), Mount Sinabung (2010)
Nevado del Ruiz (1985), Mount Etna (20022003)
Mount Merapi (2010), Grimsvo+ tn (2011)
Mount Veruvius (79 CE), Mount St. Helens (1980)
Krakato (1883), Mount Pinatubo (1991)
Thera (Minoan Eruption), Tambora (1815)
Yellowstone (640 ka), Toba (74 000 BP)
Constant
Daily
Weekly
Few months
41 year
410 years
4100 years
41000 years
410 000 years
Effusive
Gentle
Explosive
Severe
Cataclysmic
Paroxysmal
Colossal
Super-colossal
Mega-colossal
o10 000 m
410 000 m3
41 000 000 m3
410 000 000 m3
40.1 km3
41 km3
410 km3
4100 km3
41000 km3
Hawailan
Hawailan/Strombolian
Strombolian/Vulcanian
Vulcanian/Pelean
Pelean/Plinian
Plinian
Plinian/Ultra-Plinian
Ultra-Plinian
Supervolcanic
o100 m
1001000 m
15 km
315 km
1025 km
2035 km
430 km
440 km
450 km
Occurrences in last 10 000 years

Examples
Frequency
Plume
Description
Classification
Ejecta volume
VEI
Table 1 Volcanic eruption types
155
the towns of Pompeii and Herculaneum, killing thousands of

people (Giacomelli et al., 2003). Plinian eruptions are associated with volatile-rich dacitic to rhyolitic lava, and they
typically erupt from stratovolcanoes. A Plinian eruption can
continue for hours or days, sustaining a giant eruptive column. The thrust/force of the expanding gases results in the
volcanic material being ejected at speeds of several hundred
meters per second. Explosive eruption plumes such as those
generated at Mt. St. Helens and Mount Pinatubo can extend
high into the stratosphere. The highest Mount St. Helens
plume on May 18, 1980, for example, reached B31 km above
sea level (asl) and the highest Mount Pinatubo plume reached
45 km asl (Newhall et al., 1998). Plinian eruptions can also
generate pyroclastic flows reaching speeds of 700 km h1
(Shea et al., 2011). Pyroclastic flows have been known to travel
hundreds of kilometers from an eruption site. Lava flows may
be emitted when the eruption ends, and as the eruption end,
the whole summit area may collapse forming a caldera. Ash
falls, ash flows, and pyroclastic flows (nuees ardentes) predominate in landform construction.
5.8.3.3.3
Vesuvian eruptions
Vesuvian eruptions, named after Mount Vesuvius in Italy, are

very similar to Plinian eruptions, but they are not as large, so
they are sometimes called sub-Plinian eruptions. As in a Plinian eruption, Vesuvian eruptions consist of hot gas and ash
that can be ejected well into the troposphere and stratosphere.
Vesuvian eruptions occur when a volcano has had a long
period of dormancy. Over time, the pressure of volcanic gases
build up behind the solid lava plug that formed over the
volcanos vent. The eruption occurs when the gas pressure is so
high that the plug shattered and ash is ejected into the air.
After an eruption like this, ash rains down on the surrounding
countryside, commonly forming a thick ash cover. This ash
can be dangerous to breathe, kills crops, and its weight can
result in building collapse. The resulting landform is one or
more blanket layers of ash covering the preexisting topography, with a maximum thickness nearest the source vent and
decreasing with increasing distance (Heiken and Wohletz,
1985). The recent eruptions of Mount St. Helens, Mount
Pinatubo, and Chiles Puyehue are examples of large Vesuvian
eruptions (see below). Puyehue volcano erupted violently in
early June, 2011 following a series of earthquakes during the
previous day. Authorities evacuated some 3500 area residents.
Eruptions over the course of the weekend resulted in heavy
ashfalls, including falling on Argentine towns a hundred
kilometers away (Denverpost.com, 2011).
5.8.3.3.4
Vulcanian eruptions
Vulcanian eruptions are named after the eruption of Vulcano,

a small island in the southernmost Aeolian (Lipari) Islands in
the Tyrrhenian Sea off the northeastern coast of Sicily. The
term was first coined by Giuseppe Mercalli, who witnessed a
series of eruptions on the island of Vulcano between 1888 and
1890 (Schmincke, 2004). The ancient Romans thought that
this island was the forge of Vulcan, the son of Jupiter and the
blacksmith to the gods. Vulcanian eruptions are characterized
by the noise they make, which start off as a series of cannonlike blasts from the volcano vent. These explosions can release
high-velocity ejectiles comprising volcanic bombs and rocks.
156
In some cases, rocks 23 m across can fall several hundred

meters away from the volcano vent. The eruption also releases
an eruptive column of gas and ash that rises 510 km above
the volcano. The volcanoes that have Vulcanian eruptions are
often associated with a fairly viscous magma. The lavas flow
slowly and make it difficult for volcanic gases to escape and
are commonly associated with growing lava domes that can
generate pyroclastic flows when the dome collapses. Although
Vulcanian eruptions produce bombs and blocks, their main
deposit is ash, sometimes producing as much as a cubic
kilometer. This ash is very fine grained that is derived from
highly fragmented rock with a high gas content.
5.8.3.3.5
Strombolian eruptions
Strombolian eruptions are named after the volcano Stromboli

in Sicily, the northernmost of the Aeolian (Lipari) Islands, in
the Tyrrhenian Sea off the coast of Sicily. Strombolian eruptions are one of the least dangerous types of eruptions, comprising huge blobs of lava and hot rocks that burst from the
volcanos vent. As the lava hits on the sides of the volcano, it
streams down the slopes as lava streams. Strombolian eruptions occur when gas inside the volcano coalesces into bubbles,
called slugs (Bardintzeff and McBirney, 2000). These bubbles
grow large enough to rise through the magma column. Once
the bubbles reach the top of the magma column, they burst
because of the lower air pressure and eject magma into the air,
popping every few minutes. Since a Strombolian eruption can
continue for many years. Stromboli, for example, was known
as the lighthouse of the Mediterranean by Roman and Greek
sailors, since it has apparently been erupting almost continuously for more than 2400 years (Barberi et al., 1993). Another
famous Strombolian volcano is Mount Erebus in Antarctica,
which was erupting in 1841 when it was discovered by James
Ross, and has erupted frequently since then, upwards of 200
times between 1986 and 1990 alone (Esser et al., 2004).
5.8.3.3.6
Other eruptions
Other types of eruptions include hydrovolcanic eruptions,

fissure eruptions, Hawaiian eruptions, caldera collapses and
gas (limnic) emissions. A tragic example of the latter was the
August 21, 1986 Lake Nyos gas disaster in Cameroon which
resulted in the death of at least 1700 people due to a catastrophic release of CO2 and H2S from Lake Nyos. The bulk of
the gas was CO2 that had been stored in the lakes hypolimnion and was released when the lake waters overturned.
There was no significant direct volcanic activity involved, but
the CO2 was released from magmatic sources (Kling et al.,
1987). A similar incident that occurred at nearby Lake
Monoun killed 37 people two years earlier (Kling et al., 1987).
Hydrovolcanic processes are those involving explosive or
nonexplosive interaction of water and magma. Explosive
hydrovolcanic eruptions produce low-standing volcanoes with
bowl-shaped craters. Those that form beneath glaciers, such as
some that occur in Iceland, produce flat table-shaped volcanic
cones. Explosive hydrovolcanic eruption plumes commonly
collapse to form base surges that spread outward from the
base of the eruption column (Schmincke et al., 1973). Nonexplosive lava flows under water commonly form pillow lavas
or fragmental pillow breccias (Figure 4). The interactions
produce fragments known as hydroclastites. Historically, the
Figure 4 Pillow basalts exposed along the Kenai Peninsula in

Alaska. Height of the outcrop is 10 m.
name hyaloclastite refers to rocks with abundant quickly

chilled glass particles produced by nonexplosive spalling and
granulation of rinds of pillow lavas (Fisher, 1984; Fisher and
Schmincke, 1984). Explosive magma-water eruptions include
both phreatic eruptions, caused by conversion of ground water
to steam resulting in explosive expulsion of preexisting rock
fragments but no juvenile ejecta, and phreatomagmatic
eruption, caused by mixing of ascending magma with ground
or surface water resulting in expulsion of juvenile as well as
preexisting lithic material. Littoral explosions occur where
subaerial lava flows or hot pyroclastic flows intersect with
water along shorelines. Littoral cones are formed at the edge of
the sea from lava entering the water.
Various kinds of explosive hydroclastic processes can also
generate what are called maar volcanoes. Maars can occur
anytime that magma comes in contact with ground water, and
typically range in size from 10 m to 8 km across and 10 to
200 m deep. Maars are surrounded by a low rim composed of
loose fragments of both volcanic rock plus rock fragments torn
from the ground when the explosive eruption happened. The
crater rim can be asymmetrical with more material on the
downwind side, and the slopes can also be asymmetrical with
a steep inner side and a gentler outer slope. The four largest
known maars on Earth are found on the Seward Peninsula in
northwest Alaska (Beget et al., 1996). The Espenberg Maars on
the northern Seward Peninsula of Alaska were formed by a
series of Pleistocene basaltic eruptions through thick permafrost. The maars were excavated as much as 300 m into older
lithologies and range from 4 to 8 km in diameter (Self et al.,
1980). Hydromagmatic eruptions that derive water from
ground ice are evidently extremely explosive. The high heat
capacity of ice in permafrost modulates the supply of water
interacting with magma during the eruption, producing consistently low coolant-to-fuel ratios in an environment with a
sustained, abundant water supply (Beget et al., 1996). The
Espenberg Maars demonstrate that, under certain conditions,
eruptions that involve the interaction of lava and permafrost
are powerful enough to produce craters as large as small
calderas.
There are two types of vent landforms that develop from
the explosive interaction of magma and ground or surface
water, tuff cones, and tuff rings. Tuff rings (Heiken, 1971) have
a low-profile apron of tephra surrounding a wide maar crater
that is generally lower than the surrounding topography. The
tephra is commonly unaltered and thinly bedded, and is
generally considered to be an ignimbrite, or the product of a
pyroclastic density current (Sigurdsson et al., 1987; Valentine,
1987). Tuff rings are typically built around a volcanic vent
located in a lake, coastal zone, marsh or an area of abundant
groundwater. Tuff cones are considered to be a taller variant of
a tuff ring, and are steep-sloped and cone-shaped, and have
wide craters, which are formed of highly altered, thickly bedded tephra.
Bedding sags form by the impact of ballistically ejected
bombs, blocks and lapilli upon beds capable of being plastically deformed. They are common in hydroclastic deposits of
many maar volcanoes, tuff rings, and tuff cones. Beds beneath
the fragments may be completely penetrated, dragged down
and thinned, folded, or show micro-faulting (Heiken, 1971).
Deformation is commonly asymmetrical showing the angle
and direction of impact, if three-dimensional exposures are
available.
In contrast, Surtseyan eruptions take place mainly in
shallow seas and lakes. Because the water is shallow, as opposed to submarine eruptions, the pressure exerted by water is
relatively low. Surtseyan eruptions occur when an undersea
volcano has finally grown large enough to break the waters
surface; because water expands when it turns to steam, water
that comes into contact with hot lava explodes and creates
plumes of ash, steam, and scoria. Lavas created by a Surtseyan
eruption tend to be basalt, since most oceanic volcanoes are
basaltic. The classic example of a Surtseyan eruption was the
volcanic island of Surtsey, which erupted off the south coast of
Iceland between 1963 and 1965 (Wohletz, 1983, 1986).
Hydromagmatic activity built up several square kilometers of
tephra over the first several months of the eruption. Eventually, seawater could no longer reach the vent, and the eruption
transitioned to Hawaiian and Strombolian styles. This is described in more detail below. As a more recent example, in
March 2009, the Smithsonian National Museum of Natural
History (2011b) reported that several vents of the volcanic
island of Hunga Haapai near Tonga began to erupt. The onshore and offshore explosions created plumes of ash and
steam that rose to more than 8 km into the atmosphere, and
threw plumes of tephra hundreds of meters from the vents.
5.8.3.4
Major Types of Volcanoes
The geomorphic characteristics of the main types of volcanoes

are discussed in the following brief descriptions. The four
major types of central vent volcanoes are described below.
5.8.3.4.1
Cinder (Scoria) cone volcanoes
These are the simplest and most common type of volcano and
occur when particles and blobs of lava are ejected from a
volcanic vent. Cinder cones are steep-sided cones of basaltic
fragments and are smaller and simpler than composite volcanoes (Figure 5). The lava is blown violently into the air, and
the pieces rain down around the vent. Over time, this builds
up a circular or oval-shaped cone, with a bowl-shaped crater at
157
(a)
(b)
Figure 5 Example of cinder cones: (a) Parcutin cinder cone in the

Trans-Mexican Volcanic Belt and (b) Cinder cone in the Coso volcanic
field at the southern end of Owens valley, CA.
the top. Cinder cone volcanoes rarely grow larger than about
250 m above their surroundings. Streaming gases carry liquid
lava blobs into the atmosphere and fall back to Earth around a
single vent to form the cone.
Cinder cones typically have two different facies exposed in
their crater walls, a so-called wall facies and a crater facies
(Schmincke, 2004). The wall facies generally lie above the
basal breccias and phreatomagmatic layers and consist primarily of bombs, whereas the crater facies includes thinbedded deposits of fallout lapilli, phreatic deposits and, even
paleosols. Cinder cones are chiefly formed by Strombolian
eruptions that grow rapidly and soon reach their maximum
size. Most cinder cones have a bowl-shaped crater at the
summit and occur alone or in small to large groups or fields.
The longer the eruption occurs the higher the cone grows. If
gas pressure drops, the final stage cinder cone construction
may be a lava flow that breaks through the base of the cone. If
a lot of groundwater is present, then its interaction with
magma may result in a maar volcano rather than a cinder
cone. The shape of a cinder cone can be modified during its
life. When the position of the vent alters, aligned twin cones
develop. Nested, buried or breached cones are formed when
the power of the eruption varies. The lower part of the crater
floor is typically made up of rapidly erupted pieces and blobs
158
of lava that cool and coalesce to form what are called agglutinates (Schmincke, 2004). The agglutinate facies can be subdivided into two types. Type 1 is composed mainly of scoria
fragments, including spatter, with minor amounts of finer lapilli, which are generally well sorted. The scoria fragments are
up to 1 m in diameter with an average size of 2030 cm.
Scoria generally have black cores and a thin, reddish oxidized
crust, although some cores are also reddish. These fragments
show an elongated and flattened deformation structure, indicating plastic deformation during impact. The thickness of
individual layers is typically 0.23 m and they tend to be
laterally continuous. Welding and reverse grading are observed
in cases of layers more than 1 m thick. Layers less than 1-mthick show no grading or welding. These observations suggest
that a Type 1 agglutinate is of strombolian eruption origin.
Type 2 is poorly sorted and is characterized by many scoria
fragments, including spatter, in the ash matrix (Holm, 1987).
A classic example of a cinder cone is Paricutin, which
erupted out of a Mexican cornfield in 1943. Eruptions continued for the following 9 years and built a cinder cone 424 m
tall and covered 25 km2 of fields in lava flows and rubble.
Figure 6 shows a close-up of a typical Aa lava flow. Nearby
towns were eventually buried in ash by the eruptions
(McClelland et al., 1989).
Central Americas youngest volcano, Cerro Negro, was
born in April 1850 and has been one of the most active volcanoes in Nicaragua (McClelland et al., 1989). Cerro Negro is
the most recent of a group of four very young cinder cones
in the central Maribios Range 5 km northwest of Las Pilas
volcano. Strombolian eruptions at intervals of a few years to

several decades have constructed a 500-m-high basaltic cone
and an associated lava field. Cerro Negro is one of the bestknown examples of a cinder cone that has erupted more than
once. Its frequent discrete eruptions, separated by long quiescent periods, are in contrast to many other cinder cones, like
Paricutin, that were formed during a single long-term
eruption.
5.8.3.4.2
Composite volcanoes
Composite volcanoes or stratovolcanoes make up some of the

worlds most memorable mountains such as Mount Rainier,
Mount Fuji, and Mount Etna. Figure 7 shows two classic examples of stratovolcanos. These volcanoes have a conduit
system inside them that channels magma from deep within
Earth to the surface. The volcano is built up by the accumulation of material erupted through the conduit and increases
in size as lava and pyroclastic particles are added to its slopes.
US Geological Survey (2011b) estimates that out of 1511 volcanoes known to have erupted on Earth in the past 10 ka, 699
are stratovolcanoes. Stratovolcanoes can have clusters of vents,
with lava breaking through walls, or issuing from fissures on
the sides of the mountain and can grow thousands of meters
(a)
(b)
Figure 6 View of the typical texture on an Aa lava flow.
Figure 7 View of Mount Hood in Oregon (a) and Mount Vesuvius in

Italy (b), exemplifying typical stratovolcanoes.
tall. As was demonstrated by the 1980 eruption of Mount St.

Helens, composite volcanoes can explode violently. Stratovolcano shapes are more variable than these classic examples,
primarily because of wide variations in eruptive style and
composition. Some may contain several eruptive centers, a
caldera or perhaps an amphitheater as the result of a lateral
blast. Stratovolcanoes typically form at convergent plate
margins, where one plate descends beneath an adjacent plate
along a subduction zone. Examples of subduction-related
stratovolcanoes occur in many places in the world, but they
are particularly abundant along the rim of the Pacific Ocean, a
region known as the Ring of Fire. In the Americas, the Ring of
Fire includes stratovolcanoes forming the Aleutian Islands in
Alaska, the crest of the Cascade Mountains in the Pacific
Northwest, and the high peaks of the Andes Mountains in
South America.
Pyroclastic flows, lava domes, and cinder cones in some
places characterize stratovolcanoes. Generally, these types of
volcanoes erupt explosively and violently, sometimes
159
completely blowing their tops. Pyroclastic flows do not really

look much like flowing eruptions, but look more like big
explosions, complete with billowing clouds. During a pyroclastic flow, enormous volumes of extremely hot gases, ash,
and rocks rush down the side of a volcano like an avalanche.
Lava domes are rounded mounds that were built when super
thick magma that was too thick to flow squeezed up through a
volcanic
vent and cooled as it reached the air. Figure 8 provides a
schematic view of the typical processes associated with such
volcanoes.
Stratovolcanoes have been some of the most powerful and
destructive volcanoes in human history. They can send cubic
kilometers of rock and ash into the atmosphere in an extremely violent series of Plinian-type eruptions. The build-up
of extreme pressure beneath the surface from gases trapped in
the magma trying to escape eventually breaks loose in a violent paroxysm, liberating enormous volumes of overlying
rock, steam, and ash into the atmosphere. The build-up of
Eruption cloud
Prevailing wind
Eruption column
Ash (Tephra) fall
Landslide
(Debris avalanche)
Acid rain
Bombs
Lava dome collapse Lava dome
Vent
Pyroclastic flow
Pyroclastic flow
Fumaroles
Lahar (Mud or debris flow)
Lava flow
Ground
water
Silica (SiO2)
content
100%
68
63
53
Magma
type
Crack
Rhyolite
Dacite
Andesite
Basalt
Magma
Figure 8 Simplified sketch showing a volcano typical of those found in the Western United States and Alaska. Reproduced from Myers, B.,
Brantley, S.R., Stauffer, P., Hendley, J.W., 2008. What are volcano hazards? U.S. Geological Survey Fact Sheet 002-97. Revised March 2008, 2
sheets [http://pubs.usgs.gov/fs/fs002-97/].
160
pressure beneath the crust of the volcano can be so extremely

intense that the entire volcanic mountain might blow up,
leaving just the root of the original mountain remaining.
Some examples are discussed below.
In 1815, the Indonesian volcano Tambora erupted in a
series of extremely violent explosions that are said to be the
largest in recorded history. Ash was thrown 50 km into the
atmosphere and global temperatures around the world fell an
average of 3 1C, and in the higher latitudes the effect of the ash
in the atmosphere cooled summer temperatures so much that
1815 became known as the year without a summer. Tambora
and many other volcanoes occur in an island chain where the
Pacific Plate is subducting beneath the Indian plate. Tambora
ejected so much material when it erupted that the top of the
volcano caved in, forming an enormous caldera (Francis,
1993).
In 1883, the volcanic island of Krakatoa in the Sunda Strait
between the Indonesian islands of Java and Sumatra underwent a series of extremely violent eruptions that sent ash over
80 km up into the atmosphere and whose explosions could be
heard 3500 km away (Winchester, 2003). Some 36 000 people
lost their lives on Java and Sumatra when enormous tsunamis
generated by the tremendous release of energy from the
eruptions swept over the islands. The eruptions ejected so
much material from the interior of the volcano that following
the eruptions only a crescent-shaped fragment of the original
island remained once the caldera had collapsed.
5.8.3.4.3
Shield volcanoes
Some of the largest volcanoes in the world are shield

volcanoes. Shield volcanoes are large volcanoes that are
built almost entirely of fluid lava flows, and have broad
sloping sides and are generally surrounded by gently sloping
hills in a circular or fan-shaped pattern that looks like a warriors shield. Shield volcanoes are by far the largest volcanic
centers. This is illustrated in Figure 9 by comparing a large
composite volcano (Mount Rainier) with Mauna Loa and
Kilauea on Hawaii. The low-silica lava that pours out of
shield volcanoes is very fluid, so it can travel for great distances
down the shallow slopes of the volcano. Shield volcanoes
build up slowly over time, as a consequence of many hundreds of eruptions and numerous lava flows that typically
spread widely over great distances, and then cool as thin,
gently dipping sheets. The lava advances in all directions from
a central summit vent, or group of vents, building a broad,
gently sloping cone of flat, domical shape. Lavas also
commonly erupt from vents along fractures or rift zones that

develop on the flanks of the cone.
Shield volcanoes are produced by the action of gas, mainly
steam or water vapor, driven by heat from Earths core.
The pressure from the heat of the gas pushes the magma
upwards until it explodes. Molten magma shoots upward
from deep below the ocean floor and breaks through the
drifting plates. Eruptions at shield volcanoes are only explosive if water migrates into the vent, otherwise they are
characterized by low-explosive fountaining that forms cinder
cones and spatter cones at the vent. In a shield volcano,
90% of the volcano is lava rather than pyroclastic material.
Due to high magma supply rates, the lava is hot and
changes very little after it is generated. A common product of
hot-spot volcanism, shield volcanoes also occur along midocean ridge segments or affiliated with subduction-related
volcanic arcs.
The Hawaiian shield volcanoes are the largest mountains
on Earth (Regelous et al., 2003). Mauna Kea Volcano rises
4.2 km above sea level but extends B6 km below sea level
to meet the deep ocean floor. Its total height is slightly more
than 10 km, which is considerably higher than the tallest
mountain on land, Mount Everest in the Himalaya (8800 m
above sea level). Mauna Loa stands not quite as high as Mauna
Kea but is much larger in volume. The profile of the Mauna
Loa shield appears smooth, whereas the shield profile of
Mauna Kea has a more uneven appearance, reflecting the
growth of numerous small cinder cones on its upper slopes
after shield formation.
The largest shield volcano in the Solar System is Olympus
Mons on Mars. This giant measures 27 km above the surface of
Mars and is 550 km in width. A single volcanic hot-spot was
able to channel lava for billions of years, building up the
volcano to such a great size (Francis, 1993). In addition,
Olympus Mons likely achieved its great size because Mars
lacks plate tectonics and surface processes that could have help
lead to its degradation. The nature and dynamics of shield
volcanoes associated with hot-spots are described in much
detail Chapter 5.9
5.8.3.4.4
Lava domes
Volcanic or lava domes are created by small masses of lava that

are too viscous to flow very far. Unlike shield volcanoes, with
low-viscosity lava, the magma from volcanic domes just pile
up over and around the vent forming up into a dome, which
may grow by inflating from the inside or by squeezing out
Mauna Loa
19 000 Feet deep
Kilauea
5860 Meters deep
Mt. Rainer
200 km
(125 miles)
Figure 9 Comparison of Hawaiian shield volcanoes with a Cascadian stratovolcano. Note the isostatic downwarping of the ocean floor due to
the immense weight of Mauna Loa and neighboring Mauna Kea. Reproduced from Tilling, R.I., Heliker, C., Swanson, D.A., 2010. Eruptions of
Hawaiian Volcanoes - Past, Present and Future, Second edition. USGS General Information Product 117, 63 pp.
Figure 10 Obsidian dome in Long Valley, California, which formed

during the late Holocene.
Figure 11 Karymsky Lake caldera in Kamchatka. Its width is 2 km.
lobes or spines of lava. The dome grows by expansion of the

lava within, and the mountain forms from material spilling off
the sides of the growing dome. Lava domes can explode violently, releasing a huge amount of hot rock and ash. Lava
domes may be rounded, pancake-shaped, or irregular piles of
rock, depending on the type of lava they form from. They can
sometimes collapse and create pyroclastic flows, extrude lava
flows, or experience small and large explosive eruptions that
may even destroy the dome. A dome-building eruption may go
on for months or years, but they are usually repetitive. Redoubt volcano in Alaska and Chaiten volcano in Chile are
currently active examples of this type of eruption, and Mount
St. Helens developed several lava domes over a period of years
following its 1980 eruption. A classic example of a lava dome
is Obsidian Dome near Mono Lake in southern California,
which erupted in 1350 AD and comprises, as it names suggests, obsidian as well as rhyolite and pumice (Millar et al.,
2006; Figure 10).
5.8.3.4.5
Caldera versus crater
A volcano caldera is a depression in the ground created by the

collapse of land after a volcanic eruption. An example is
Karymsky Lake caldera in Kamchatka shown in Figure 11. It is
161
generally larger than a volcanic crater, which is the opening at

the top of a central vent volcano. In some cases, the caldera is
created slowly, when the ground sinks down after a magma
chamber is emptied. In other situations, the magma explodes
violently, and the caldera is the deep pit in the ground that
remains after the volcano blows its top off. A caldera forms
when the magma chamber beneath a volcano is emptied in a
large eruption. If the eruption happens quickly enough, the
empty chamber beneath is not strong enough to support the
weight of the volcano and it collapses down. This can happen
in a single cataclysmic event, or over the course of several
eruptions. Volcano calderas can be hundreds and even thousands of square kilometers in area.
There are many famous examples of volcano calderas.
Crater Lake in Oregon was created when a volcano, Mount
Mazama, erupted at B5.7 ka (Zdanowicz et al., 1999). Over
time, water filled the nearly 597 m-deep caldera, making it the
deepest lake in the United States, and the second deepest lake
in North America (Great Slave Lake in the Northwest Territories of Canada is the deepest). Another example of a volcanic caldera is the Yellowstone Caldera, which last erupted
640 000 years ago. It released 1000 km3 of rock, covering
much of western and central North America in two meters of
debris. Caldera eruptions on the Yellowstone scale have a
worldwide frequency of perhaps once every hundred thousand years. Somewhat smaller eruptions, on the scale of Crater
Lake in Oregon, are more frequent, perhaps every 1000 years
or less (Schmincke, 2004). Explosive caldera eruptions are
generally not isolated events. Rather, they are climactic stages
of magmatic processes that extended over hundreds of thousands of years. There are other examples of nonexplosive
volcano calderas. For example, the volcano Kilauea on the Big
Island of Hawaii has a magma chamber that is emptied by
large lava flows, and not by explosive eruptions. This causes
the ground to collapse down after the lava eruptions, sinking
many meters into the ground. Volcano calderas have been
seen on many other places in the Solar System. Although there
are not active volcanoes on Mars today, regions of Mars were
active billions of years ago, and there are many regions with
large calderas. Jupiters moon Io is an example of a place with
recent active calderas (Kerr, 2010).
Figure 12 schematically summarizes the range of forms
caldera collapses may take. These include four end-member
mechanisms of caldera collapse: (a) piston/plate, (b) piecemeal, (c) trapdoor, and (d) downsag. Chapter 5.9 provides
more details about caldera and craters for volcanic centers
associated with hot-spots.
A volcano crater is a circular depression around a volcanic
vent. This is where the lava, ash, and rock erupt out of a
volcano. In most situations, the crater is located at the top of
the volcano, and when the volcano does erupt, the material
comes out of this crater. But craters can also appear on the
flanks of volcanoes, particularly large ones, when side vents
reach the surface and erupt. These are known as flank craters,
while the crater on top is called a summit crater. A phreatic
eruption creates a crater without a volcanic cinder cone at all.
This is a situation where magma rises through water-saturated
rocks and causes steam to build up under the surface. The
pressure from all the steam causes the rock to explode outward
and create a volcanic crater.
162
Topographic rim
Structural rim
Piecemeal
Landslide
breccia
Chaotic
Piston/plate collapse
Funnel
(a)
(b)
Extension
Trapdoor
(c)
(d)
Downsag
Figure 12 Four end-member mechanisms of caldera collapse. (a) piston/plate, (b) piecemeal, (c) trapdoor, and (d) downsag. Reproduced from
Cole, J.W., Milner, D.M., Spinks, K.D., 2005. Calderas and caldera structures: a review. Earth-Science Reviews 69, 126.
One of the most famous volcanic craters in the world is

Diamond Head, located on Oahu in the Hawaiian Islands.
The ancient Hawaiians originally named it Leahi, meaning
brow of the tuna because of the shape of the silhouette of the
crater viewed from Waikiki (Pukui et al., 1974). British sailors
gave the current name when they came to the crater in the
1800s. When they first saw the crater at a great distance,
the calcite crystals in the lava rock appeared to glimmer in the
sunlight. The sailors mistakenly thought there must be diamonds in the soil. Diamond Head is a crater that has been
extinct for 150 000 years. The crater is 1000 m in diameter
with a 230-m summit.
A less well-known but equally fascinating crater is Kilbourne Hole, which occurs in a desert basin between the
Potrillo Mountains and the Rio Grande River in southern
Dona Ana County, New Mexico. In 1975, the Bureau of Land
Management designated it a National Natural Landmark.
Kilbourne Hole is roughly elliptical in shape, and is known as

a maar, a pit or depression caused by a volcanic explosion with
little material emitted except volcanic gas. The stratified, crossbedded pyroclastic surge deposits around the crater at Kilbourne Hole are spectacular. When the steam-saturated eruption column that forms during an explosive event
gravitationally collapses, a ring-shaped surge travels radially
outward along the ground away from the vent. The surge deposits may have formed as a consequence of a series of steam
explosions during the emplacement of the basalt (Padovani
and Reid, 1989; Hawley and Lozinsky, 1993). The crater
formed between 24 and 100 ka, and measures 2.7 km long
and over 1.6 km across, and is hundreds of meters deep. Kilbourne Hole has many characterstic mafic minerals remaining
from the eruption. The crusts of its volcanic bombs are dull
black or brown, but when broken open they commonly reveal
a brilliant, sparkling yellow and green interior of olivine glass
granules, reflecting the remarkable abundance of both crustal

and mantle peridotite/olivine-bearing xenoliths that are in
basalt bombs ejected during the eruption (Seager, 1987).
Basalt that erupted at about 7081 ka from a set of vents called
the Afton cones located north-northeast of Kilbourne Hole
flowed southward. The explosion that formed Kilbourne Hole
erupted through the distal edges of the Afton basalt flows,
indicating that the crater is younger than 70 to 81 ka (Hoffer,
1976).
Still another example in the same region is Sunset Crater
located just east of Flagstaff, Arizona in the San Francisco Volcanic Field (Smiley, 1958). Volcanism began at B6 Ma in the
western part of the field and has since migrated eastward producing over 600 volcanic vents. Activity culminated with the
eruption of Sunset Crater Volcano just over 900 years ago. The
US National Park Service is now preserving this 300 m-tall
cinder cone. Eruptive activity began along a 10 km-long NESW
trending fissure that produced a series of coeval vents including
Sunset Crater, Rows of Cones, Gyp Crater, and Vent 512. This
fissure eruption ended shortly after it began as activity became
focused on the northern end of the fissure. Strombolian eruption built Sunset Crater Volcano and blanketed an area of at
least 2000 km2 with cinders. Two lava flows, the Kanaa and the
Bonito flows, effused from the base of the cone. The Kanaa
flow extruded from the eastern base and continued in a
northeasterly direction along Kanaa wash for B8 km (Holm
and Moore, 1987). The Bonito flow effused from the northwestern base of the volcano and, at least in a broad sense,
resulted from three pulses of activity (Holm, 1987) with little
time elapsing between these events. Just west of the vent, lava
ponded in a low area accumulating lava to a depth of perhaps
over 30 m in the center to only a few meters along the margins
(Holm, 1987). This flow carried rafted mounds of red oxidized
agglutinate several hundreds of meters to the west and northwest (Holm, 1987). Continued lava fountaining filled in the
holes created by rifting, and ultimately produced a nearly
symmetric cone (Hanson et al., 2008).
5.8.3.4.6
Submarine volcanoes
Many volcanoes start out on land, rising above the surrounding landscape. But many more volcanoes get their start at
the bottom of the ocean. Eruptions beneath the sea surface are
the most common type on Earth, but are significantly underrepresented in the historical record. These submarine volcanoes can eventually grow into islands that rise above the
surface of the ocean. More than 5000 active underwater volcanoes have been identified, which account for more than 75%
of the total lava that erupts every year. Most of these are located
along the mid-ocean ridges and are very deep underwater, and
difficult to study, but some are located in more shallow water.
An underwater volcano erupts differently than a surface volcano because there is an unlimited amount of water to cool
down the lava. A shell of rock hardens around the lava almost
immediately, creating a type of formation called pillow lava.
Deeper than about 2000 m, the pressure of the water is so high
that it cannot boil. Underwater volcanoes build up over time
and can eventually reach the surface of the ocean. This is what
happened to form the Hawaiian Islands, where Earths crust
has drifted above an active plume, creating each of the islands
in turn. A new Hawaiian island, Loihi seamount, is forming
163
under the ocean about 48 km off the southeast coast of Hawaii

(Malahoff, 1987). Loihi seamount, sometimes known as the
youngest volcano in the Hawaiian chain, is already taller than
Mount St. Helens and will breech the ocean surface in a few
hundred thousand years. Both Loihi and Kilauea volcanoes sit
on the flank of Mauna Loa volcano, an older, larger, and still
active volcano on the Big Island of Hawaii. Loihi sits submerged in the Pacific off the southeastern coast of the Big
Island of Hawaii. Prior to the 1970s, Loihi was not known to
be an active volcano. Instead, it was thought to be a fairly
common old seamount volcano of the type that surrounds the
Hawaiian Islands. These latter volcanoes are similar in age
(80100 Ma) to the sea floor upon which the Big Island of
Hawaii sits. This sea floor was itself created some 6000 km
away on the undersea volcanic mountain chain known as the
East Pacific Rise. It has slowly moved northwestward to the
present location of the Hawaiian hot-spot.
5.8.3.4.7
Mud volcanoes
Mud volcanoes and mud volcanism remain some of the most

enigmatic and poorly understood natural phenomena despite
being known and described since the time of Pliny in 77 AD
(Dimitrov, 2002; Kopf, 2002). Thousands of mud volcanoes,
also known as cold seeps, exist worldwide, defining and affecting the habitat and the daily lives of millions of people
living among them. Moreover, mud volcanism and mud volcano distribution is intimately related to the formation and
the distribution of much of the worlds petroleum assets, thus
serving as an indicator for valuable natural resources. And
finally, mud volcanoes offer an insight into otherwise hidden
deep structural and diagenetic processes such as the formation
of gas hydrates, mineral dissolution and transformation,
degradation of organic material, and high pressure and high
temperature reactions. Mud volcanism generally involves the
generation and emission of voluminous quantities of both
methane and carbon dioxide such that most mud volcanoes
serve as efficient, natural sources of greenhouse-gases, and
consequently may conceivably play an important role in global climate dynamics. Like hydrothermal vents, underwater
mud volcanoes can have their own unique ecosystems filled
with life forms that get most of their energy from chemical
sources, such as methane.
Essentially, mud volcanoes are geological features through
which argillaceous material is altered and transported from
Earths interior and expelled onto its surface. Hence, mud
volcanism is the process by which such features are initiated
and sustained. However, mud volcanism is not one specific
process and mud volcanoes are not uniform features. Their
settings, driving forces, activity, materials, and morphologies
may vary. Mud volcanoes typically occur where stratified rocks
with an abundance of water-saturated, rapidly deposited finegrained sediment, are under pressure from overlying strata.
Furthermore, pressure exerted on these formations by deep gas
deposits, or by gas generated closer to the surface, can be a
factor in the formation of mud volcanoes. Gas is commonly
the driving force, pushing mud from deep within Earth upward toward the surface. The force is so great that it can lift
and break rock formations, so that pieces of rock are sometimes thrown out with the mud when the volcano erupts.
164
The term mud volcano system was coined by Stewart

and Davies (2006) to describe the set of structures associated
with a mud volcano and feeder complex that connects
the volcano to its source stratigraphic unit. The system is driven by pressure and a source of fluid, which may or may not
coexist with the mud source beds. Above the fluid source is a
feeder conduit, the detailed structure of which is largely unknown. It probably consists of a complex system of fractures
and mud-filled dikes that feed a fluid-sediment mix to Earths
surface (e.g., Planke et al., 2003). The fluid-sediment mix then
erupts to form the volcano. The plumbing of mud volcano
systems is poorly constrained. For instance, the mud and fluid
could coexist at the time of initiation, analogous to magma, or
the fluid could be transported from a deeper source, remobilizing mud at shallower stratigraphic levels (Deville et al.,
2006). Some mud volcano systems are thought to consist of
multiple mud chambers at different stratigraphic levels
(Mazzini et al., 2007; Planke et al., 2003), whereas other
models propose that mud volcano systems consist of significant masses of mud in the form of bulbous-shaped diapirs
(Brown, 1990; Milkov, 2000). External forces such as seismic
activity along active fault zones may trigger eruptions (Mazzini
et al., 2009; Manga et al., 2009). Most mud volcanoes stand
only a few centimeters to several meters high, but the landbased ones on Pakistans plains can rise to 100 m (Dimitrov,
2002). Some of these peaks have been known to spew gas
plumes that spontaneously combust, sending tongues of fire
high into the sky.
Although most common along active, convergent plate
boundaries, mud volcanoes also occur along passive continental margins, within continental interiors and throughout
deep-sea settings. About 2000 mud volcanoes have been described to date, however, as exploration of the deep seas
continues, this number is expected to increase substantially.
Estimations suggest everything from 7000 to 1 000 000 in total
(Milkov, 2000). Of all known mud volcanoes, more than half
(about 650 onshore and at least 470 offshore) can be related
to the Alpine-Himalaya active belt (Dimitrov, 2002). Starting
with the Mediterranean Ridge, this mud volcanic belt continues all the way down to the IndonesiaAustralia accretion
and collision complexes via Romania and the Black Sea, the
Caucasus/Caspian Sea region, Iran, Pakistan, India, and China
(Dimitrov, 2003). The western flank of the Pacific Ocean from
the Sakhalin Island/Sea of Okhotsk area in the north via
Japan, Taiwan, the Marianas, Melanesia, Samoa and Australia
to New Zealand in the south holds some 150 onshore volcanoes (Kopf, 2002; Dimitrov, 2003). The eastern flank of the
Pacific Ocean is markedly less dense in mud volcanoes, although examples are known from and around the Aleutian
Trench, Alaska, British Columbia, California, Costa Rica,
Ecuador, and inland Peru (Kopf, 2002; Dimitrov, 2003). The
Atlantic Ocean contains several hundreds of both onshore and
offshore mud volcanoes with the vast majority being concentrated along the Caribbean thrust belts and within the
Barbados accretionary complex (Dimitrov, 2003). Smaller
clusters have been reported from the Amazon and the Niger
deltas, along the Gulf of Cadiz and offshore Portugal and
Morocco. A few mud volcanoes have also been described from
the Mississippi and the Nile deltas, Lake Michigan, Greenland,
the North Sea, and the Netherlands (Kopf, 2002).
The Lusi mud volcano of east Java is a particular interesting

example. Several vents started to form in May 2006 increasing
to 34 by November 2006, increasing to B170 active vents by
2010 (Roberts et al., 2011). The Lusi mud volcano had a peak
flow of 180 000 m3 of mud per day and it is expected to
continue discharging for the next 2530 years (Davies et al.,
2011). The resultant mudflow was devastating for the communities in the region, destroying farmland and properties
south of the city of Sidoajo. Controversy surrounds the origin
of Lusi mud volcano. It is argued that hydrofracturing due to
drilling that took place a few days earlier than the eruption
lead to its formation. However, others argue that the
Yogyakarta earthquake and its aftershocks that also occurred
just a few days before the eruption were the major cause
(Mazzini et al., 2007). Whether the Lusi mud volcano was
human-induced or not, it provides an impressive example of
the dynamic nature of mud volcanoes and their effects on our
landscape and people.
5.8.3.4.8
Geysers, fumaroles, hot springs
Ascending hot magmatic fluids mix with near-surface

groundwater to form active geothermal systems with dramatic
surface features such as hot springs, solfataras, fumaroles,
mudpots, geysers, and hot acid lakes. These hydrothermal
features are common at stratovolcanoes and young calderas
and can persist for hundreds of thousands of years. Active
hydrothermal systems often precipitate deposits of native
sulfur and other minerals.
Archeological evidence shows that the first human use of
geothermal resources in North America occurred more than
10 000 years ago with the settlement of Paleo-Indians at Hot
Springs, Arkansas, who believed that this was the home of the
Great Spirit who brought forth the healing warmth of Mother
Earth (Lund, 1993). Here, also the first known commercial use
of geothermal energy in the United States occurred in 1830,
when Asa Thompson charged one dollar each for the use of
three spring-fed baths in a wooden tub. Within the next 20
years, more resorts and spas of increasing size grew up around
such springs. Erupting geysers provide spectacular displays of
underground energy suddenly unleashed, but their mechanisms are not completely understood. Large amounts of hot
water are presumed to fill underground cavities. The water,
upon further heating, is violently ejected when a portion of it
suddenly flashes into steam. This cycle can be repeated with
remarkable regularity, as for example, at Old Faithful Geyser in
Yellowstone National Park, which erupts on an average of
about once every 65 min (Figure 13). With over 10 000 geothermal features including hot springs, mudpots, fumaroles,
and geysers, Yellowstone National Park provides vast habitats
for a wide variety of thermotolerant and thermophilic organisms. Yellowstones geologic setting serves as the driving force
behind its numerous geothermal features. The center of Yellowstone is a volcanic caldera that covers a large portion of the
central area of the park. Volcanic activity below the caldera
and the surrounding area releases tremendous amounts of
heat to drive the hot springs, geysers, and other hydrothermal
features. Groundwater from snowmelt and other precipitation
in the greater Yellowstone area is heated by molten rock below
the surface. Once heated, the water finds a path through cracks
Figure 13 Old Faithful Geyser in Yellowstone National Park,

Wyoming. Eruptions launch between 14 000 and 32 000 l of boiling
water into the air to heights of over 50 m about every 90 min.
and fissures of lava flows to the surface, where it emerges in

geysers, hot springs, mudpots, and fumaroles.
Hot springs occur in many thermal areas where the
surface of Earth intersects the water table. The temperature and
rate of discharge of hot springs depend on factors such as the
rate at which water circulates through the system of underground channels, the amount of heat supplied at depth, and
the extent of dilution of the heated water by cool ground water
near the surface.
Geysers are hot springs that erupt periodically. The eruptions are the result of superheated water below ground becoming trapped in channels leading to the surface. The hottest
temperatures are at the bottom of these channels but the deep
water cannot vaporize because of the weight of the overlying
water and rock mass. Instead, steam is sent upward in bubbles,
collecting in the channels tight spots until they essentially
become clogged, leading to a point where the confined bubbles actually lift the water above, causing the geyser to overflow. This causes the pressure to decrease until suddenly
violent boiling occurs throughout much the length of the
column, producing a tremendous volume of steam that forces
the water out of the vent in a superheated mass. As the
eruption continues, the heat and pressure gradually decrease,
and the eruption stops when the water reservoir is depleted or
165
the steam runs out. The two types of geysers are fountain
geysers, which shoot water out in various directions through a
pool, and cone geysers, which shoot water out in a fairly
narrow jet, generally from a cone-like formation (Rhinehart,
1980).
Hot springs are similar to geysers, but their underground
channels are large enough to allow rapid circulation of water.
Rising hot water releases heat energy by evaporation or hot
water runoff, while convection currents return the cooler water
to the underground system, thus maintaining equilibrium.
The water and rocks in hot springs may be multicolored.
Brilliant rings of red, blue, brown, green, orange, and yellow
are formed by the different species of algae and bacteria that
flourish in the hot spring environment (Bryan, 2008). Normal
hot springs do not erupt, but may seem to churn and boil as
gases from underground pass through them. When the water
that feeds a hot spring passes through rocks underground it
often dissolves minerals from the rocks and brings them to the
surface. At typical hot spring temperatures of nearly 100 1C
even silica-rich compounds will dissolve. At the surface, water
cools and the silica forms geyserite, a white or grayish cauliflower-like deposit that creates rims or terraces around the
spring. This deposit is also known as siliceous sinter. If the
water passes through limestone, then it carries calcium carbonate to the surface. There the mineral crystallizes to form
travertine. This dense, white rock also creates ledges and other
rock formations around the spring.
Fumaroles are holes or vents from which steam rushes into
the air. They are similar to hot springs but lack the liquid
water. Either there is not enough water or the underground
rock is too hot and boils off all of the water so a pool cannot
form. The small amount of water that does seep into the area
is converted to steam and expelled from the vent, generally
creating a hissing noise. Hydrogen sulfide (H2S), one of the
typical gases issuing from fumaroles, readily oxidizes to sulfuric acid and native sulfur. This accounts for the intense
chemical activity and brightly colored rocks in many thermal
areas. All fumaroles require both heat and a source of gas or
water. They are most commonly supplied with heat and gas by
magma or masses of freshly ejected volcanic rock, and with
water by precipitation that seeps into the ground.
Subterranean-heated water also produces hot springs and
geysers. Hot springs are more common than fumaroles
and geysers. Geysers are distinguished from both hot springs
and fumaroles by their specialized plumbing systems, whereas
the difference between a hot spring and a fumarole is simply
the degree of heating. If the heat source is not strong enough
to boil water, the result is a hot spring. Even if water is boiled,
the resulting steam may be condensed by passing through
liquid groundwater before reaching the surface, in which case
the result is still a hot spring. Only if steam reaches the surface,
is a fumarole produced. Some vents are hot springs in the wet
season and fumaroles in the dry, when there is less groundwater to condense steam rising from below (Bryan, 2008).
Fumaroles whose gases are particularly sulfurous are
termed solfataras (Kilburn and McGuire, 2001). Furthermore,
some gas-emitting vents have temperatures below the boiling
point of water and emit mostly carbon dioxide and other gases
with little water vapor. Such dry, cool vents are termed mofettes to distinguish them from fumaroles.
166
Figure 14 Svartsengi geothermal plant on the Reykjanes Peninsula

in Iceland (lower image). Superheated water is vented from the
ground to drive turbines to generate electricity, and then used for a
municipal hot water heating system. The Blue Lagoon geothermal spa
(upper image) is fed by waters from the geothermal plant.
A deposit of hot and shattered rock laid down by an explosive volcanic eruption may cover many square kilometers
of ground and be hundreds of meters deep. During the years it
takes to cool, such a deposit may produce a vast field of fumaroles. This occurred in the Valley of Ten Thousand Smokes
in Alaska, where a thick ash and rock layer was laid down by a
large eruption in 1912. Immediately after deposition, this layer
was dotted by tens of thousands of fumaroles, some venting
from openings many feet across. Over the next half-century, as
the underlying mass cooled, most of these fumaroles became
extinct.
Reservoirs of steam or hot water in geothermal fields have
been tapped for electrical power in many volcanic regions
(Figure 14). The most common current way of capturing the
energy from geothermal sources is to tap into naturally occurring hydrothermal convection systems where cooler water
seeps into Earths crust, is heated up, and then rises to the
surface. When heated water is forced to the surface, it is a
relatively simple matter to capture that steam and use it to
drive electric generators. Geothermal power plants drill their
own holes into the rock to more effectively capture the steam.
There are three designs for geothermal power plants, all of
which pull hot water and steam from the ground, use it, and
then return it as warm water to prolong the life of the heat
source. In the simplest design, the steam goes directly through
the turbine, then into a condenser where the steam is condensed into water. In a second approach, very hot water is
depressurized or flashed into steam that can then be used to
drive the turbine. In the third approach, called a binary system, the hot water is passed through a heat exchanger, where it
heats a second liquid, such as isobutene, in a closed loop. The
isobutane boils at a lower temperature than water, so it is
more easily converted into steam to run the turbine (Tester
et al., 2006).
In 1892, the worlds first geothermal district heating came
online in Boise, Idaho, eventually growing to serve 200 homes
and 40 downtown businesses. Although this system was
unique for 70 years, there are now 17 geothermal district
heating systems in the United States and dozens more around
the world (Lund, 1993). Modern nonelectrical use of geothermal energy continues. For example, since the 1960s, the
Maoris of New Zealand have used geothermal energy for
cooking. France and many other European Union nations are
using geothermally heated water to heat thousands of homes.
In the early 1900s, the first instance of geothermal electric
power emerged. In Italy, Prince Piero Ginori Conti invented
the first geothermal power plant in 1904 at the Larderello dry
steam field and it is still in operation today. The first geothermal electricity plants in the United States were operated in
1960 at The Geysers in Sonoma County, California (Kagel
et al., 2007). In Iceland, the entire capital city of Reykjavik is
powered by the geothermal energy from the North Atlantic
Oceanic Ridge below the island.
Throughout the 1970s in the United States, government
backing through legislation, grants, and the actions of agencies
like the Department of Energy (DOE) furthered research, development, and continued acceptance of geothermal electrical
power into the electricity producing mainstream. Work in the
1970s laid the foundation for actualization of more advanced
types of geothermal steam production in the 1980s from flash
steam to binary vapor cycles. Net electrical output of these
plants reached 50 MW by the end of the decade. Progress
continued in the 1990s as the DOE increased its efforts to
promote the use of geothermal energy for electricity generation as well as space and process heating. By the mid-1990s,
the DOE had identified over 9000 thermal wells and springs
and 271 communities throughout the United States connected
to geothermal energy (Kagel et al., 2007).
5.8.3.4.9
Lava tubes
Lava tubes are natural conduits through which lava travels

beneath the surface of a lava flow. Tubes form by the crusting
over of lava channels and pahoehoe flows (Figure 15). During
long-lived eruptions, lava flows tend to become channeled
into a few main streams. Overflows of lava from these streams
solidify quickly and plaster on to the channel walls, building
natural levees or ramparts that allow the level of the lava to be
raised. Lava streams that flow steadily in a confined channel
for many hours to days may develop a solid crust or roof and,
thus, change gradually into streams within lava tubes. Lava
tubes are very efficient transporters of lava from the vent to the
flow front, and calculations show that lava flowing in a tube
cools by only about 1 1C km1 (Walker, 1991). At the flow
front, the lava behaves much like a river delta, forming small
distributary tubes that continue to branch until they consist of
167
(a)
Figure 15 View of a typical texture produced by a pahoehoe (ropy)

lava flow.
the same type of single flow-unit tubes, but smaller, that have
been forming the flow the whole way downslope. The amount
of gas in lava at the flow front is somewhat lower than that at
the vent, because, although lava tubes are good thermal insulators, they are not so well sealed that gas can be trapped.
Gas escapes through cracks in the tube and also through skylights, places where the roof of a tube has fallen in to expose
the flowing lava. Lava tubes can be many meters wide, and
typically run several meters below the surface. A broad lavaflow field commonly consists of a main lava tube and a series
of smaller tubes, which supply lava to the front of one or more
separate flows. When the supply of lava stops at the end of an
eruption or lava is diverted elsewhere, lava in the tube system
drains downslope and leaves partially empty conduits beneath
the ground. Inside the tube there can be various formations,
like lava stalactites, known as lavacicles. Pillars may stretch
from the top to the bottom of the lava tube. Basaltic lava
typically produces significant quantities of gas and steam, and
while flowing through the tube this causes the air pressure
inside to increase. This keeps the tube open but also may press
a hole into the covering crust, which is still hot and rather soft.
Those venting holes are called hornitos (Spanish for little
ovens) or spatter cones. The gas pressure from below throws
small pieces of lava out of the venting holes, which form up to
15 m high cones on top of the vent. The lava in those cones is
formed of layer above layer of spatters (Walker, 1991).
The lava will continue to move and drain the tube downhill when the volcano spots erupting. This leaves a partially
empty tube that is called a lava tube cave. Lava that is not able
to flow out of the tube will remain in the form of solidified
pools, flows, or even congealed, waterfall-like features. Collapses are rather common, as the roof of lava tubes is rather
thin, a result of the way it formed. As a result most lava tubes
do not last more than 10 000 years. Older tubes are extremely
rare. One exception, however, is Lava River Cave in Coconino
National Forest in Arizona (US Forest Department, 2011). This
km-long lava tube cave was formed B700 ka ago by molten
rock that erupted from a volcanic vent in nearby Hart Prairie.
(b)
Figure 16 Late Pleistocene Corona lava tube on Lanzarote in the

Canary Islands. (a) Opening in the dry section of the lava tube, which
stretches for B8 km, and (b) interior view of the lava tube.
Another is Corona lava tube on the Canary Islands (Canaries)

that is dated to B21 ka (Carracedo et al., 2003; Figure 16). In
addition to Hawaii, lava tubes also occur in the western U.S. in
Washington, California, Oregon, Nevada, Idaho, New Mexico,
Utah, and Arizona, in the Canary Islands, the Galapagos
Islands, Italy, Japan, Korea, Kenya, Mexico, and many other
volcanic regions (Bunnell, 2008).
5.8.4
Evolving Volcanic Landforms
Two major sets of processes dominate the formation of volcanic landforms, internal and external. Figure 17 schematically illustrates the main processes. A series of concise case
studies help illustrate the dynamic and evolving nature of
volcanoes.
5.8.4.1
Parcutin Cinder Cone
The Parcutin cinder cone (Figure 5(a)) is located in TransMexican Volcanic Belt in the Mexican state of Michoacan and
is arguably one of the most famous examples of an evolving
volcanic landform; unique because it was witnessed from its
168
Internal triggers
Filling and emptying of
summit conduits
Replenishment and tapping
of magma reservoir
Magma intrusion
Loading effects
Lateral spreading
Crustal
downwarping
Volcanic seismicity
(a)
External triggers
Climate effects
Changing
sea levels
Tectonic
seismicity
(b)
Brittle
failure
Basement uplift
and subsidence
Active fault movement
Vertical block displacement

due to filling/empty
Graben
of high level magma reservoir formation
Toppling due to
oversteeping
Downslope
sliding due to
lava loading
Horizontal block
displacement
Elastic displacement
due to filling/empty
of summit conduits or high-level
reservoir
Elastic
displacement due to
dyke emplacement
Elastic
deformation
(c)
Figure 17 Principal internal (a) and external (b) triggers responsible for generating deformation and Earth movements at active volcanoes, and
(c) mechanisms responsible for short timescale deformation in active volcanic terrains. Reproduced from McGuire, B., Saunders, S., 1993.
Recent Earth movements at active volcanoes: a review. Quaternary Proceedings 3, 3346.
very inception. The cinder cone began as a fissure eruption in a

cornfield on February 20, 1943, witnessed by Dionisio Pulido
and his wife and son as they ploughed their field. Within a
week, the cone had grown to B10 m-high, growing to about
336 m-high within a year by explosive pyroclastic eruptions.
The nearby villages of Paricutin and San Juan Parangaricutiro

were buried by lava and ash, and residents had to be excavated. Lava flows continued for the next 8 years covering an
area of B25 km2. During the last 6 months of this period, the
eruptions became more explosive, until Paricutin finally
169
became quiet and extinct gaining a height of 424 m. Luhr and

Simkins (1993) definitive book on Parcutin reproduces a
series of well-illustrated diverse geological, sociological, and
biological reports on this unique event.
5.8.4.2
Kavachi, Solomon Islands
Kavachi is one of the most active submarine volcanoes in the

south-west Pacific Ocean and has been located in the Bougainville and Solomon Island volcanic arc since 1939. The
volcano has become emergent at least eight times before being
eroded back into the sea (Smithsonian National Museum of
Natural History, 2011c, d). The volcano rises from a water depth
of B1.1 km and has a basal diameter of B8 km, and is roughly
conical in form (Smithsonian National Museum of Natural
History, 2011c). When the volcano erupted in 2003, a 15 m-high
island formed above the surface, but it disappeared soon after.
Impressive video footage of submarine and subaerial eruptions
have been made for Kavachi; many available on YouTube (e.g.,
YouTube 2010a, b). These help illustrate the continued dynamic
interplay between volcanic (endogenetic) and surface (exogenetic) processes in the evolution of volcanic landforms.
5.8.4.3
Surtsey
Surtsey is a volcanic island and part of the Vestmannaeyjar

submarine volcanic system that also produced the famous
eruption of Heimaey (Eldfell) in 1973. Surtsey is B1.5 km in
diameter and has an area of 2.8 km2 and is 33 km south of the
main island of Iceland and 20 km southwest of Heimaey.
The island is named for Surtur, a giant of fire in Icelandic
mythology (Kokelaar and Durant 1983). Surtsey is a classic
example of the growth of a new volcanic island. Episodic
eruptions began on November 8, 1963 and ended on June 5,
1967. The volcano grew from the sea floor, at a depth of
130 m, to sea level by November 15. During the first few days,
eruptions were not explosive and probably consisted of gentle
effusion of pillow lava. As the volcano grew toward sea level
the water pressure decreased and activity became explosive.
The early phases of the eruption were phreatomagmatic,
caused by the interaction of magma and water. Explosions
were closely spaced or steady jets, producing dark clouds of
ash and steam shooting tens or a few hundred meters above
the vent. At times, a column of ash and steam was carried
10 km above the growing island. A tuff ring was constructed
by glassy tephra that was deposited by base surges and by
fallout. This new island was unstable because it was made of
unconsolidated tephra. On January 31, 1964, activity shifted
400 m to the northwest and phreatic eruptions continued at a
new vent.
As the eruption progressed, a new tuff ring developed that
protected the vent from seawater. On April 4, 1964, this caused
the activity to change from phreatomagmatic explosions to lava
fountaining and the gentle effusion of lava flows. Lava flows
extended the island to the south and protected the underlying
tephra from wave erosion. This phase of the eruption ended on
May 17, 1965 and was quiet for more than a year. On August
19, 1966, activity resumed at new vents at the older tuff ring on
the east side of the island. More lava flows moved to the south
Figure 18 The 20-m-high Svartifoss waterfall in southern Iceland.

This is one of the numerous waterfalls that plunge over the trap
topography.
partially overlapping the older flows. The eruption stopped on

June 5, 1967; it lasted a total of 3.5 years. About 1 km3 of ash
and lava had been produced with only 9% of it above sea level.
The average temperature of the lava was 1140 1C and is composed of alkali olivine basalt. Between 1967 and 1991, Surtsey
subsided B1.1 m (Moore et al., 1992). The subsidence is
probably the result of compaction of the volcanic material that
makes the volcano, compaction of the sea-floor sediments
under the volcano, and possibly downwarping of the lithosphere due to the weight of the volcano.
5.8.4.4
Iceland
Iceland sits astride the Mid-Atlantic Ridge above the Icelandic

hot-spot. Located just south of the Arctic Circle, Iceland has
been extensively glaciated during glacial times, and ice caps
and glaciers are still prevalent. As a consequence, the volcanic
landscapes reflect the interaction of seafloor spreading,
uplift due to mantle plume processes and glacioisostatic
rebound, and glacial processes. The landscape is dominated
by extensive lava flows, which build up upon themselves resulting in stepped relief (trap topography; Figure 18). Subglacial volcanism results in volcanoes that a table-shaped
(Figure 19). Furthermore, subglacial volcanism commonly
results in the formation of subglacial lakes, which may burst
through outlet glaciers resulting in gigantic glacial lake outburst floods (jokulhlaups). One of the best studied was 1996
Grmsvotn lake of Vatnajojkull glacier that flooded Skeioara
river (Russell et al., 2007). Figure 20 shows Eyjafjallajokull
and Katla, which are typical of the volcanic centers on Iceland.
Eyjafjallajokull erupted during the winter and spring of 2010,
spewing vast amounts of ash into the atmosphere causing
widespread disruption to air traffic throughout Europe, and is
arguably one of the most costly volcanic eruptions of all time
(Sigmundsson et al., 2010).
5.8.4.5
Kilauea
Klauea is one of five shield volcanoes that make up the island

of Hawaii and is one of the most active volcanoes on Earth
(Figures 21 and 22). Klauea currently overlies the Hawaiian
170
Figure 19 Google EarthTM image of a table volcano in Iceland (no vertical exaggeration). Snow- and ice-covered Katla is present in the distant
upper right hand corner of the image.
Eyjiafiallajkull
Eruption of
Fimmvruhls
2000
Katla
Lava
dome
Magma intrusion
2000
Magma
chamber
4000
Sills
6000
0
2000
4000
6000
m
Figure 20 Cross-section through Eyjafjallajokull and Katla volcanoes in Iceland. The figure shows the location of the 2010 Eyjafjallajokull
eruption.
hot-spot and lies on a curving line of volcanoes that includes

Mauna Kea and Kohala, but excludes Mauna Loa that is to the
o cinder-and-spatter cone of Kilauea has
NW of Klauea. Puu O
been continuously erupting since 1983, making it the longestlasting rift-zone eruption of the past 200 years (Figure 22). As
such Kilauea is arguably the best-studied nonviolent volcanic
center. Lava less than 1000 years old covers 90% of Klauea (US
Geological Survey, 2011c). Eruptions from Kilauea and Mauna
Loa volcanoes are remarkably predictable. The lava chamber
fills, raising the ground and empties, deflating the ground with
astonishing regularity (Figure 22). Figure 23 shows the growth
profiles of Klaueas newest volcanic cone, built during the first 3
o-Kupaianaha eruption, illustrating the
years of the Puu O
geomorphic expression.
The immense mass of Hawaiian shield volcanoes has
isostatically depressed the ocean crust so that a deep ocean
trough exists around the Hawaiian islands (Figure 9). As
Moore et al. (1989) highlight, the mass and steep terrain

are inducing to large-scale mass failures of sections of the
Hawaiian islands and is illustrated by ancient giant submarine
landslides that occur offshore (Figure 24).
5.8.4.6
Mount St. Helens
Mount St. Helens, situated in the Cascades in Washington State

on the Northern American plate above the subducting Juan de
Fuca plate is probably one of the best-studied stratovolcanoes.
The catastrophic eruption on May 18, 1980, was the deadliest
and most costly volcanic event in U.S. history, and resulted in
75 fatalities and the loss of 250 homes, 47 bridges, 24 km of
railway, and B300 km of highway. Figure 25 shows the sequence of events on May 18, 1980. The massive debris avalanche (B3 km3 of debris) on the north flank of Mount
171
Mauna Kea
20 km
0
0
Hilo
10 miles
Site of
Kapoho
village
Mauna Loa
Mauna Loa
summit caldera
(Moku aweoweo)
Kilauea
t Zone
ast Rif
rthe
o
N
Hawaiian Volcano
Observatory
Kilauea
summit
caldera
Puu -Kupanianaha
Halemaumau
est
Ri
ft
Zo
ne
t hw
es
tR
ift
Z
Hi on
e
l
lt System
Fau
a
in
Mauna Ulu
Site of
Kaimu black
sand beach
Sou
Southw
on e
i ft Z
R
t
Eas
Kilauea Iki
Hawaii
Kohala
Mauna Kea
Area of
map
Hualalai
Mauna Loa
Kilauea
Puu o Mahana
green-sand beach
Lihi
10 km
20miles
Figure 21 Southeastern part of the Island of Hawaii showing the principal features and localities of Mauna Loa, Kilauea and Loihi volcanoes.
Reproduced from Tilling, R.I., Heliker, C., Swanson, D.A., 2010. Eruptions of Hawaiian Volcanoes - Past, Present and Future, Second edition.
USGS General Information Product 117, 63 pp.
St. Helens was triggered by a magnitude 5.1 earthquake due

to a sudden surge of magma. The avalanche reduced the
elevation of the mountain summit by B400 m and formed a
1.6 km-wide crater (Figure 26). The magma within the volcano
was released as a giant pyroclastic flow with a VEI of 5, flattening vegetation, and buildings over an area of B600 km2.
The collapsing volcano mixed with ice, snow, and water to
produce lahars that advanced down the Toutle and Cowlitz
Rivers. Figure 27 shows a generalized map of the impacts and
deposits from the 1980 eruption. Mount St. Helens continues
to evolve as illustrated in Figure 28.
5.8.4.7
Mount Rainer
Mount Rainer is another massive Cascadian stratovolcano.

Its location within 90 km of Seattle makes it one of the
most potentially hazardous volcanic centers in North
America. The most recent recorded volcanic activity was
during the latter part of the nineteenth century (Harris,
2005). However, the recent discovery and recognition

of ancient giant volcanic mudflows on the outskirts of
Seattle highlights the potential of future failures of Mount
Rainer with their catastrophic consequences (Harris,
2007). Figure 29 shows areas that could be affected by
debris flows, lahars, lava flows, and pyroclastic flows from
Mount Rainer.
5.8.4.8
Krakatau
Krakatau (name used for the island group and main island)
is located in the Sunda Strait between Java and Sumatra
in Indonesia, above the subduction zone of the Eurasian
and Indo-Australian plates. The cataclysmic August 27, 1883
eruption of Krakatau (frequently misstated as Krakatoa)
is probably one of the most infamous volcanic eruptions
of modern time. It provides important insights into the
development and evolution of caldera eruptions. A very
comprehensive and fascinating account is provided by
172
Pre-eruption inflation
Benchmark
displacements
Volcano
surface
after swelling
Stable
point
Stable
point
Magma
Volcano
surface
before swelling
Earthquakes
(a)
Summit Tilt of Kilauea, 19562009
600
Inflation
Kilauea Iki
eruption
600
Eruption
Intrusion Halemaumau 400
eruption
400
200
(b)
800
Deflation
Tilt (microradians)
Puu -Kupanianaha eruption

(still continuing in 2010)
Mauna Ulu
eruption
800
0
1956 1960
1965
200
1970
1975
1980
1985
1990
1995
2000
2005
0
2010
Year
Short-period
(tectonic)
earthquakes
Long-period
(volcanic)
earthquakes
43
42
44
46
45
47
30
20
10
0
No
data
No data
Number of volcanic
summit earthquakes
Eruptive
episode
41
40
30
20
10
0
2500
2000
No data
Summit
tilt
1500
1000
500
0
January
February
March
April
May
1000
800
600
400
200
0
Number of tectonic
summit earthquakes
Tilt, in
microradians
Puu -Kupanianaha eruption, JanuaryJune 1986
June
(c)
Hawaiian eruptions in recorded history

Summit
Mauna Loa volcano
2010
2000
1980
1960
1940
1920
1900
1880
1860
1840
1820
No
information
1800
Up to
1 month
Up to
1 week
Two days
or less
1780
Up to
1 year
Up to
1 month
Up to
1 week
Two days
or less
(d)
Kilauea volcano
Halemaumau
Crater (2008)
Duration of eruptions
Up to
1 year
Halemaumau crater, summit
Violently
explosive
Essentially continuous
activity
Total
inactivity
Mauna
Ulu
Puu Kupanianaha
Figure 22 The dynamic nature of Klauea and Mauna Loa Volcanoes. Reproduced from Tilling, R.I., Heliker, C., Swanson, D.A., 2010. Eruptions
of Hawaiian Volcanoes - Past, Present and Future, Second edition. USGS General Information Product 117, 63 pp. (a) Schematic cross-section
through the volcano. Magma entering the shallow reservoir exerts pressure on the volcano, causing earthquakes and distorting its shape from
the dashed-line profile to the solid-line profile. (b) Graph of tilt due to gradual inflation, followed by abrupt deflation associated with major
eruptions and intrusions. (c) Six-month segment of the tilt record reveals similar inflation-deflation patterns for the high-lava-fountaining
episodes of the Puu Oo-Kupaianaha eruption, even though the tilt changes and time intervals involved are much smaller (compare scales of the
two diagrams). Also shown are variation patterns of the two types of earthquakes that commonly precede and accompany Klauea eruptions. (d)
Graph summarizing the eruptions of Mauna Loa and Klauea volcanoes during the past 200 years. The Puu Oo-Kupaianaha eruption has
continued into the twenty-first century.
May 1986
y 1985
Jul
ar y 1 9
br u
85
Fe
9
8
1
4
ust
Aug
200
800
600
400
ember 1983
Dec
100
200
June 1983
Feet above pre-1983 surface
1000
300
Meters above pre-1983 surface
173
Pre-1983 ground surface
Figure 23 Growth profiles of Klaueas newest volcanic cone, built during the first 3 years of the Puu Oo-Kupaianaha eruption. Reproduced
from Tilling, R.I., Heliker, C., Swanson, D.A., 2010. Eruptions of Hawaiian Volcanoes - Past, Present and Future, Second edition. USGS General
Information Product 117, 63 pp.
Rift
zone
Displaced
flank segment
Dense rift core
Kilauea
South flank
Ocean floor
Basal
thrust
Oceanic crust
km
25
(a)
Kilometers
WNW
2
0
2
4
6
8
ESE
Volcanic pile
Sub-Etnean
basement
Intrusive core
Basal thrust
km
(b)
Figure 24 Schematic figures showing sliding of volcanic piles. Reproduced from McGuire, B., Saunders, S., 1993. Recent Earth movements at
active volcanoes: a review. Quaternary Proceedings 3, 3346. (a) Idealized model for rift-flank interaction at Kilauea. Seaward displacement of the
southern flank is attributed to uphill sliding of this part of the edifice along a thrust fault at the volcano-sea floor interface. Movement is
facilitated partly by the existence of a dense rift core of intrusive material, whereas continued dike emplacement along the rifts is accommodated
by repeated yielding of the southern flank. (b) Seaward spreading of the edifice at Mount Etna occurs along a basal thrust in a horizon of
subvolcanic clays. As at Kilauea, the model proposes lateral spreading accommodates new intrusions beneath the volcano resulting in the
formation of a growing intrusive core.
Winchester (2003) in his book, Krakatoa: The day the

world exploded: August 27, 1883. Winchester (2003) highlights that the 1883 Krakatau eruption was the first volcanic
eruption that became instant global news because of the
new technological innovations such as the transatlantic
telephone cables.
Before the cataclysmic 1883 eruption, Krakatau comprised
three main islands (Lang, Verlaten, and Krakatau), which were
the remnants of a previous caldera-forming eruption. Krakatau
island was B9 km long and B5 km wide and had three
volcanic cones. The cataclysmic 1883 eruption was preceded

by several years of seismic activity and a series of smaller
eruptions in June 1883. A series of four eruptions comprised
the cataclysmic eruption that was so violent that it produced a
pressure wave that was recorded on barographs around the
work (Winchester, 2003). The pyroclastic flows, volcanic ash
and tsunami resulted in widespread devastation throughout
the region (Winchester, 2003). The VEI is estimated to be 6,
and the amount of ejected rock, ash, and pumice was probably about 21 km3. The tsunami was so large that tidal gauges
174
Meters
South
3000
North
Summit
dome
Pre-collapse profile
(Bulge)
Pre-1980 profile
2500
2000
Goat rocks
dome
Magma
intrusion
(cryptodome)
1500
(a)
3000
2500
Summit
block
2000
1500
(b)
Bulge block
Graben
block
3000
2500
2000
1500
(c)
3000
2500
0.5
1.0 km
Pre-collapse profile
(bulge)
2000
1500
(d)
Figure 25 Schematic cross sections through Mount St. Helens showing the succession of events during the 1980 climactic eruption.
Reproduced from Lipman, P.W., Mullineaux, D.R. (Eds.), 1981. The 1980 Eruptions of Mount St. Helens, Washington. U.S. Geological Survey
Professional Paper 1250, 844 pp. (a) The volcano in the early morning of May 18. The bulging of the north flank is clearly shown by the pre1980 and precollapse profiles. (b), (c) Within about 30 s after the collapse, the debris avalanche, lateral blast, and vertical eruption begin as the
cryptodome is exposed. The bulge block was the first to slide, followed by the graben block. (d) Within the next 30 s, the summit block had slid
and the lateral blast had stopped, and the vertical eruption was in full fury.
recorded it as far away as the English Channel (Winchester,

2003). After the eruption, the island of Krakatau had almost
entirely disappeared, the southern half of its southernmost
crater and a 250-m deep caldera.
Krakatau has continued to evolve and in early 1928 a
new island, Anak Krakatau, rose above the sea in between
Lang, Verlaten, and Krakatau. Eruptions were initially of
pumice and ash, but were succeeded by lava flows. Anak

Krakatau and two additional islands were quickly
eroded away. Eventually, in 1930, erupting lava flows outcompeted the marine erosion and Anak Krakatau became established that continues to grow at a rate of B7 m per year,
with almost continuous Strombolian eruptions with a VEI of
1 to 2.

5.8.4.9
Mount Etna
5.8.4.10
Mount Etna, located on the east coast of Sicily, rising to

3329 m asl is the tallest active volcano in Europe, and is one of
the most active stratovolcanoes in the world. Mount Etna has
built up over the last 0.5 Ma from successive pyroclastic and
lava flows. A giant landslide has lowered the eastern flank of
Mount Etna. McGregor and Saunders (1993) suggest that the
volcanos future is tenuous as seaward spreading of the edifice
at Mount Etna is occurring along a basal thrust in a horizon of
subvolcanic clays (Figure 24).
175
Canary Islands
The Canary Islands are located just off the northwest coast of
mainland Africa B10 km west of southernmost Morocco. The
islands formed about the Canary hot-spot and are dominantly
of shield type, rising to 3718 m above sea level on Tenerife. Of
particular note are giant subaqueous landslides that surmount
the islands, which record times of catastrophic collapse and
illustrate the dynamic nature and far reaching hazards from
shield volcanoes (Figure 30).
5.8.5
Ancient Volcanic Settings
Several giant and dramatic volcanic events have had profound

influence on the present day geomorphology in regions
throughout the world. Clearly, volcanoes are responsible for
the existence of island arcs, and ultimately the creation of
continental crust by island arc accretion. However, these will
not be the focus of this section, as these are really beyond the
realms of tectonic geomorphology. Instead this section will
consider supervolcanic eruptions and flood basalts.
5.8.5.1
Figure 26 Google image of the north side of Mount St. Helens
showing the collapse of 1980 and the numerous volcanic landslides
and debris flows (no vertical exaggeration).
Green River
Supervolcanic Eruptions
The terms supervolcano was first introduced in 2000 by the

producers of a British TV program to describe the Toba eruption
of 74 000 years ago. The term implies a volcanic center that has
Riv
er
Green
Coldwater
Elk Rock
Creek
th
For
k
Spirit
Lake
Smith
dy
ud
Pyroclastic flow deposit
ift c
reek
5 miles
Sw
er
Riv
Mudflow deposit
Debris avalanche deposit
ek
Cre
Sout
h Fo
rk To
utle Ri
ver
Outline of crater
Lateral blast zone
ree
k
Toutle
Riv
er
C le
arw
ater C
No
r
Figure 27 Generalized map showing the impacts and deposits of the 1980 climactic eruption in the vicinity of Mount St. Helens. Reproduced
from Tilling, R.I. Topinka, L., Swanson, D.A., 1984. Eruption of Mount St. Helens: past, present and future. USGS Special Interest Publication, US
Government Printing Office, Washington DC, 56 pp.
176
September, 2003
October 14, 2004
198086
Lava dome
February 1, 2005
198086
Lava dome
198086
Lava dome
Deformed
glacier
Initial spine
Welt
Whaleback
(a)
N
Crater rim
500 m
2500
Altitude (m)
220 m
213 m
2300
200405
Dome
Empire state
building
New dome summit

February 1, 2005
150 m
Old dome summit
415 m
198086 Dome
383 m
1900
Glacier surface 2000
Glacier
Glacier bed
Debris
1980 crater floor
(b)
Figure 28 (a) Topographic changes of Mount St. Helens measured from aerial photographs and laser imaging (LiDAR) showing rapid growth of
a new lava dome, the whaleback-shape extrusion, and the deformed glacier. (b) Topographic profiles along a NS axis through Mount St. Helens
crater showing the new lava dome relative to the south crater rim, 1980 crater floor, 19801986 lava dome, and 2000 glacier surface.
Reproduced from Major, J., Scott, W.E., Driedger, C., Dzurisin, D., 2005. Mount St. Helens erupts again. USGS Fact sheet 20053036.
had an eruption of magnitude 8 on the VEI, meaning that the

measured deposits for the eruption is 41000 km3. Previous
terms such as mega-eruption have also been used to describe
the same phenomenon, volcanic eruptions far bigger than any
occurring in recorded human history such as Tambora (1815) or
Krakatau (1883). There are two types of supervolcano eruptions.
The first is an enormous outpouring of lava across a wide area.
This releases a large volume of material and so qualifies as a
supereruption, even though it does not involve the huge ash
clouds some people associate with typical volcanoes. The second
type is explosive, sending tephra across a wide area and causing
tremendous local destruction. Both types release a large volume
of gases as well as solid material, which may influence the atmosphere and potentially global climate. Historically documented eruptions do not, however, represent the full range of
intensity and magnitude of all explosive eruptions in geologic
history. Deposits in the geologic record provide compelling
evidence for eruptions that have been orders of magnitude larger
than ones witnessed by mankind. Ash layers in marine sediments are generally the best-preserved geologic record of explosive volcanism, and many of the Phanerozoic volcanic ash
layers represent enormous explosive eruptions that dispersed

fine ash and aerosols through the atmosphere over tens of
thousands of km2. Evidence of very large magnitude events
occur throughout the Phanerozoic. For example, Ordovician
explosive volcanism is recorded in both North America and
Europe by both near-vent pyroclastic flows and ignimbrites and
distal sequences of altered fallout tephras known as K-bentonites (Huff et al., 2010). To help illustrate how large supervolcanic eruptions have been, Figure 31 compares the volume of
erupted material from historical eruptions with those of important supervolcanic eruptions.
5.8.5.1.1
Toba caldera
A widely cited example of a supervolcano is the Toba Caldera

on the Indonesian Island of Sumatra. Other examples less than
1.5 million years old include the Valles Caldera in New Mexico,
the Long Valley Caldera in California, and the Yellowstone
Caldera in Wyoming (Figure 32). The Toba supervolcano has
erupted explosively a number of times over the past 1.2 million
years. By far the largest and most destructive of these occurred
around 74 000 years ago, and is generally referred to as the
177
Bellevue
Pug
Bremerton
sh
mi
wa
Du
Seattle
Lake
Washington
r
ve
Ri
oun
et s
Renton
North
Bend
r
ve
Cedar Ri
Kent
Gig
Harbor
Federal
way
Auburn
n Rive
r
ree
N
Tacoma
Sumner
Puyallup
Enumclaw
R
hite iver
W
Lake
Tapps
American
Lake
Buckley
Wilkeson
Greenwater
Mud
Mountain
Lake
Carbonado
Orting
r
Ca
Graham
bo
nR
Pu
ya
llu
pR
ive
Yelm
De
Lake
Kapowsin
McKenna
iver
Mount
Rainer
Lake Ohop
sc
hute
sR
Eatonville
iver
Shookumchuck
Reservoir
Alder Lake
4392 m
Alder
Ni
squ
ally
River
Ashford
Elbe
Debris flow hazard zone
Lahar hazard zone
0
0
10 miles
ve
Ri
itz
l
w
o
C
Flooding and post-lahar

sedimentation zone
Lava flow and pyroclastic
flow hazard zone
Morton
Randle
Packwood
10 km
Riffe Lake
Figure 29 Map showing areas that could be affected by debris flows, lahars, lava flows, and pyroclastic flows from Mount Rainer. Reproduced
from Driedger, C.L., Scott, W.E., 2008. Mount Rainer - living safely with a volcano in your backyard. USGS Fact sheet 20083062.
Youngest Toba Tuff or YTT eruption. The YTT eruption generated 2800 times more pyroclastic material than the moderate
Plinian eruption of Mount St. Helens in 1980. The eruption
produced 25003000 km3 of magma (dense rock equivalent)
and probably injected at least 1015 g of fine ash into the

stratosphere (Rampino and Self, 1993; Zielinski et al., 1996).
Pyroclastic flows covered about 20 000 km2 with lava reaching
both the Malacca straits and the Indian Ocean (Rose and
178
29 N
La Palma
CT
?
Playa
de la Veta
Roques
de
Garcia
Santa
Cruz
Orotava
Anaga
?
Teno
IV OV GV
Cumbre
Nueva
La Gomera
Guimar
CC
El Goifo
28
El Hierro
Canary
debris
flow
Tenerife
EG
Las
Playas I
Las
Playas II
Islands
Debris Avalanche
EI Julan
Slump
Debris flow
Saharan
Debris Flow
18
17
16 W
Figure 30 Summary of mapped landslides on the flanks of El Hierro, La Palma, and Tenerife. Reproduced from Masson, D.G., Watts, A.B., Gee,
M.J.R., Urgeles, R., Mitchell, N.C., Le Bas, T.P., Canals, M., 2002. Slope failures on the flanks of the western Canary Islands. Earth Science
Reviews 57, 135.
Chesner, 1987). Layers of tephra, identified as Toba tuff occur

in India, more than 3000 km away from Toba (Chesner et al.,
1991). Additional post-YTT eruptions include a series of lava
domes, the growth of the solfatarically active Pusukbukit volcano on the south margin of the caldera, and the formation of
Tandukbenua volcano at the NW-most rim of the caldera
(Chesner and Rose, 1991). Lack of vegetation suggests that this
volcano may be only a few hundred years old.
The supereruption of Toba has been closely linked to a
peak in SO2 in the GISP2 ice core (Zielinski et al., 1996), and
it has been proposed that the impact on the worlds climate
caused by the eruption of Toba was considerable (Rampino
and Self, 1993). Ice core data show that the weather was cooler
for several centuries after the Toba eruption, possibly due to
the large amount of stratospheric sulfur aerosol ejected by the
volcano, and following that, increased albedo. Rampino and
Self (1992, 1993) estimated that worldwide temperatures were
35 1C cooler, and down to 15 1C cooler in high latitudes, for
several years.
5.8.5.1.2
Yellowstone caldera
Volcanism in the Yellowstone region in the United States has,

during about the past 2 Ma, built an immense volcanic plateau that straddles a high mountain divide known as the Yellowstone Plateau volcanic field (Figures 33 and 34). This
volcanic region has evolved through three cycles of voluminous outpourings of rhyolite lava and volcanic ash, each of
them climaxing with one of Earths greatest pyroclastic-flow
eruptions and the resulting collapse of a central area to form a

large caldera (Christiansen, 2001). Other eruptions have
poured out basalt lava flows around the margins of the volcanic field. Volcanism in the Yellowstone region has a 17
million-year history of activity that has occurred progressively
from southwestern Idaho to Yellowstone National Park. At
least six other large volcanic centers along this path generated
caldera-forming eruptions; the calderas are no longer visible
because they are buried beneath younger basaltic lava flows
and sediments that blanket the Snake River Plain (Christiansen, 2001). Yellowstone is far from extinct; it is currently
the site of the largest heat-flow anomaly in the US and
earthquakes are a daily occurrence in the region, although
most are under 3.0 magnitude.
Historically, two types of volcanic events have taken place
in Yellowstone. These include caldera-forming eruptions that
occurred at B2.1 and 1.3 Ma, and a series of massive lava flows
that occurred between 640 and 70 ka (Figure 31; Christiansen,
2001). The Huckleberry Ridge Tuff erupted from the Island
Park Caldera in the southwest corner of Yellowstone at
B2.1 Ma. The Huckleberry Ridge Tuff was deposited as a series
of separate ash flows that were welded together while they were
still very hot. Some layers of fallout ash are preserved between
the flows. The different individual ash flows now appear as the
cliffs, and the ash fall as the thin layers forming slopes between
the cliffs. Much of the Huckleberry Ridge Tuff has been eroded
away since it erupted, but its original volume is estimated to
have been about 3000 km3. The Mesa Falls Tuff erupted at
t.
t. S s
M len
ai
Mazama Tambora
He 980 atm 2 toa
1
K 91 ka 3 7600 years
Mesa falls tuff
1815
1 ra 8
K 18
(From Island Park
ago (6800
Caldera)
Radiocarbon Yr. B. P.)
(1.5 Million years)
Lava Creek
640 000 years ago
179
Huckleberry ridge
2.1 Million years ago
Figure 31 Comparison of volumes of ash produced by major volcanic eruptions and supervolcanic eruptions. Reproduced from Good, J.M.,
1998. Interpreting the landscapes of Grand Teton and Yellowstone National Parks: recent and ongoing geology. Grand Teton National History
Association in corporation with the National Parks Service, Moose, WY.
Figure 32 Section through the Bishop Tuff at the northern end of

Owens Valley, CA. This huge succession (4100 m thick) of tuff
formed during the Long Valley caldera collapse B0.77 Ma.
Reproduced from Crowley, J.L., Schoene, B., Bowring, S.A., 2007.
U -Pb dating of zircon in the Bishop Tuff at millennial scale. Geology
35, 11231126.
B1.3 Ma and was created by the Mesa Falls eruption that

formed the Henrys Fork Caldera located in Idaho just west of
Yellowstone Park. It is the second most recent caldera-forming
eruption from the Yellowstone hot-spot and ejected 280 km3
of material (Christiansen, 2001). The Lava Creek ash-flow tuff
erupted with the collapse of the Yellowstone caldera at
B640 ka. Scientists first distinguished the Lava Creek Tuff from
the Huckleberry Tuff in the 1960s (Christiansen, 2001). This
was the first clue that Yellowstones history included more than
one caldera-forming eruption. All ash flow tuffs at Yellowstone
look very much like the Huckleberry Ridge, and can be difficult
to sort out. The Lava Creek Tuff has a total volume of about
1000 km3 with thicknesses up to hundreds of meters, some
filling preexisting canyons. Subsequent lava flows have slowly
filled in most of the caldera. Even now, in some places, nearly
molten rock resides as little as 38 km below the surface. Heat
from the volcanic activity makes its presence known by heating
ground water and creating the thermal features we now see.
Yellowstone National Park is home to some 10 000 thermal
features, over 500 of which are geysers. In fact, Yellowstone
contains the majority of the worlds geysers. The four basic
types of thermal features present in the park are geysers, hot
springs, fumaroles, and mudpots. Many of these are concentrated in Yellowstones major geyser basins: Upper, Midway,
Lower, Norris, West Thumb, Shoshone, and Heart Lake.
5.8.5.2
Flood Basalts
Flood basalts are the result of a giant volcanic basaltic eruption or series of eruptions that cover vast stretch of land or
ocean floor, and generally result in the creation of plateaus
and/or mountain ranges. Flood basalt eruptions appear to
occur sporadically throughout geologic time. Notable flood
basalts are described below.
5.8.5.2.1
Columbia river basalt group
Plateau basalts associated with hot-spots are discussed in

detail in Chapter 5.9, so we will offer just a brief commentary
180
114 W
112
108
110
45 N
44
43
42
50
100
150 km
Fault belts I-IV
Yellowstone crescent of
high terrain
Terrain
Crest of crescent
Active faults
Volcanic fields with

ages
Figure 33 Topography and geology of the Yellowstone Caldera showing the track of the Yellowstone hotspot over the last 10 Ma. Reproduced
from Pierce, K.L., Despain, D.G., Morgan, L.A., Good, J.M., 2007. The Yellowstone Hotspot, Greater Yellowstone ecosystem, and human
geography. In: Morgan, L.A. (Ed.), Integrated Geoscience Studies in the Greater Yellowstone Area - Volcanic, Tectonic, and Hydrothermal
Processes in the Yellowstone Geoecosystem. US Geological Survey, Professional Paper 1717, pp. 539. The volcanic fields align along the
Eastern Snake River Plain.
here on the nature of these significant landscape features.

During the late Miocene and early Pliocene, one of the largest
basaltic lava floods ever to appear on Earths surface covered
about 160 000 km2 of the U.S. Pacific Northwest. Over a period
of perhaps 1015 million years lava flow after lava flow poured
out, eventually accumulating to a thickness of more than
2000 m. Over 300 high-volume individual, lava flows have
been identified, along with countless smaller flows. Numerous
linear vents, some over 150 km long, show where lava erupted
near the eastern edge of the Columbia River Basalts. Older vents
were probably buried by younger flows. As the molten rock

came to the surface, Earths crust gradually sank into the space
left by the rising lava. The subsidence of the crust produced a
large, slightly depressed lava plain now known as the Columbia
Plateau (Hooper et al., 2007).
5.8.5.2.2
Deccan traps
The flood basalt province known as the Deccan Traps is located

on the Deccan Plateau in west-central India and is one of the
largest volcanic provinces in the world. The Deccan Plateau
Figure 34 South rim of Yellowstone National Park.
consists of more than 2000 m-thick succession of flat-lying

basalt lava flows and covers an area of nearly 500 000 km2. The
lavas are predominantly tholeiitic with rare alkalic mafic lavas
and carbonatites. Estimates of the original area covered by the
lava flows are as high as 1.5 million km2 and the volume of
basalt is estimated to be 512 000 km3. The Deccan Traps formed
at the end of the Cretaceous period, about 65 Ma, from eruptions associated with a deep mantle plume (Sen, 2001).
5.8.6
Volcanic Hazards
Volcanic hazards can be both direct and indirect. Direct

hazards are the immediate consequence of eruption and seismic
activity, whereas indirect hazards stem from a combination of
geomorphic and geologic processes on potentially active volcanoes that may occur at times other than during eruptions.
5.8.6.1
Direct Hazards
Direct hazards from volcanic eruptions can have catastrophic

consequences. These are described in turn below.
5.8.6.1.1
Pyroclastic flow
As discussed previously, pyroclastic flows can form when an

eruption column collapses, or as the result of gravitational
collapse or explosion on a lava dome or lava flow (Walker,
1984; Valentine, 1987; Sigurdsson et al., 1987; Decker and
Decker, 2005). Ignimbrites and nuees ardentes are two types
of pyroclastic flows. An ignimbrite contains mostly vesiculated
material, whereas a nuee ardente contains denser material.
These flows are commonly accompanied by a cloud of ash
elutriated from the flow called a co-ignimbrite plume. When
the incandescent ash particles are observed at night, the flow
looks like a glowing cloud moving away from the volcano.
Pyroclastic flows can move very fast. Small pyroclastic flows
can move as fast as 1030 m s1, whereas larger flows can
move at rates of 200 m s1 (Bryant, 1991). Nuees ardentes
have been known to extend 50 km from their source and
ignimbrites, because of the lighter weight material that they
carry, can extend 200 km from their source (Bryant, 1991).
At Mount Pinatubo in the Philippines, pyroclastic flow deposits were 220 m thick in some valleys but averaged 30 to
50 m-thick in others. Pinatubo unleashed three major
181
Figure 35 Tephra overlain by pyroclastic flow deposits near the

Grmsvotn volcano in Iceland.
destructive agents in 1991 that caused destruction to Central

Luzons infrastructure and rendered its vast agricultural lands
into virtual wastelands. These were ash falls, pyroclastic flows,
and lahars. Pyroclastic flows and lahars are the greatest volcanic
hazards. Pyroclastic flows can flow over ridges as much as a
kilometer in height (Bryant, 1991). More people have died due
to these hazards than any other volcanic hazard. Pyroclastic
flows can incinerate, burn, and asphyxiate people. Pyroclastic
flows travel long distances so their threat is far reaching. They
also can transform into lahars that travel even farther distances
from the volcano and can produce even greater hazards.
5.8.6.1.2
Tephra
Material ejected during an explosive volcanic eruption is

known as tephra (Figure 35), the largest pieces of which are
called blocks and bombs. Because these fragments are so large
they fall out near their source. Blocks and bombs as large as
700028 000 kg have fallen as far away as 1 km from their
source (Bryant, 1991). Smaller blocks and bombs can travel
much farther. Some of these blocks and bombs can have
velocities of 75200 m s1 (Bryant, 1991). Smaller ejecta such
as lapilli (264 mm) and ash (o2 mm) that are transported
upward by convection due to the heat of the eruption will fall
out farther from the volcano. Most particles greater than a
millimeter in size will fall out within 30 min of the time they
are erupted. The size of particles that fall out is largest near the
volcano and gets progressively smaller further from the volcano. The thickness of material generally decreases the further
away from the volcano. The smallest particles that are o10 mm
in diameter can stay in the atmosphere for several years after a
volcanic eruption.
The distance that tephra travels away from a volcano depends on the height of the eruption column, temperature of
the air, wind direction, and wind speed. An erupting column
that reaches above 10 km into the stratosphere will be sheared
by strong winds in this region and cause the eruption cloud to
spread out over a larger area. The temperature of the air during
an eruption will increase due to the hot material ejected into
the atmosphere, and this adds to the buoyancy. The eruption
of Krakatau in 1883 deposited ash over an area of
800 000 km2 (Bryant, 1991). Fallout from the eruption of
Tambora in 1815, which is the largest eruption in recent
182
history, was deposited as far away as 1300 km (Francis, 1993).

More than 10 cm of ash fell in India 3000 km away from the
Toba Caldera when it erupted B75 ka.
Tephra produces a wide range of hazards (US Geological
Survey, 2011a). When the ejected material is in the atmosphere,
it is electrically charged and often produces lightning. Large
ejecta shot ballistically from the volcano are also a hazard to
those unfortunate enough to be near the volcano. Other hazards are produced when the ash is deposited on the ground.
Ash can disrupt electricity, television, radio, and telephone
communication lines, bury roads and other manmade structures, damage machinery, start fires, and clog drainage and
sewage systems. Ash can produce poor visibility and cause respiratory problems (International Volcanic Health Hazard
Network, 2011). Often people living in areas affected by volcanic ash will wear masks with filters or wet cloth over their
nose and mouth to protect themselves from breathing in volcanic ash and fumes (Horwell and Baxter, 2006; Horwell,
2007). Studies of the percent of respirable (o5 mm) silica in
Paleozoic supervolcano tephras indicate that there would be a
major public health risk in the event of such large eruptions
today (Inanli and Huff, 2009). Ash falls vary widely in intensity, in size of the ash particles, and in the degree to which light
from the sun is obscured or blocked completely. Because of the
unexpected darkness during daylight hours, loud thunder and
lightning, and the occasional strong smell of sulfur during an
ash fall, many people may feel stunned and fearful of the
conditions and have a difficult time breathing if a well-sealed
shelter is not available. If a heavy ash fall continues for 12 to
60 h or more, roofs may collapse under the weight of the ash,
resulting in more confusion, injuries, and even death.
Ash is also a great hazard to airplanes. Ash from the 1982
eruption of Galunggung Volcano in West Java, Indonesia
caused engines in two jet airplanes to fail. Both aircraft
dropped by B7600 m before they could get their engines
started again. Damage can include pitting of windshields,
which causes them to look foggy. The fuselage, wings, engines,
and light covers can also be damaged by ash and are very
expensive for the airlines to repair. Volcanic eruption clouds
cause rerouting, cancellation, and delays in flights, which is
also expensive for the airlines. Ash on runways is another
problem because anytime a plane lands or takes off the ash is
re-suspended. Currently, much research is being done in using
satellites to track volcanic clouds.
5.8.6.1.3
removed or destroyed. Topography in this area had no affect on

the movement of material in the directed blast cloud. In a zone
between 12 and 30 km from the volcano, trees were flattened
and resembled toothpicks aligned in a uniform direction on
surrounding hillsides (Foxworthy and Hill, 1982).
Mount St. Helens is not the only volcano that has erupted
with a direct blast and current research shows that direct blasts
are not uncommon. The eruption of Mount St. Helens alerted
scientists to the warning signs and hazards of such an eruption. Knowledge of direct blast eruptions will help in the future, so that warnings can be given to people in areas that
might be affected by such a blast. Monitoring of volcanoes
with seismographs and instruments that would indicate
ground deformation can help identify hazardous zones and
indicate areas of possible safety.
5.8.6.1.4
Lava flows
Surprisingly, lava flows are probably the least hazardous to

human life of all processes in volcanic eruptions. Although they
typically destroy everything in their path, and thus are extremely
destructive to property, most move slowly enough that people
can generally move out of the way (Figure 36). The speed at
which lava moves across the ground depends on several factors,
including: (1) type of lava erupted and its viscosity; (2) the
steepness of the ground over which it travels; (3) whether the
lava flows as a broad sheet, through a confined channel, or
down a lava tube; and (4) rate of lava production at the vent.
Fluid basalt flows can extend tens of kilometers from an erupting
vent. The leading edges of basalt flows can travel as fast as
10 km h1 on steep slopes but they typically advance less than
1 km h1 on gentle slopes. But when basalt lava flows are confined within a channel or lava tube on a steep slope, the main
body of the flow can reach velocities 430 km h1 (Pinkerton
and Wilson, 1994). Viscosity is in large part controlled by the
silica content of the lava. Thus, andesite flows move only a few
kilometers per hour and rarely extend more than 8 km from
their vents. Dacite and rhyolite flows often form steep-sided
mounds called lava domes over an erupting vent. Lava domes
can grow by the extrusion of many individual flows over a
period of months or years. Such flows will overlap one another
and typically move less than a few meters per hour and clearly
lava flows are extremely hot, varying between 550 and 1400 1C.
Direct blast
The base of an explosively erupted column is known as the gas

thrust zone where material is shot ballistically from the volcanic vent. Above this zone is the convective thrust zone. This
zone is where heat buoys pyroclastic material upward toward
the top of the troposphere. Once the eruption column reaches
the stratosphere, shearing occurs and the material is spread
out in an umbrella shape (Francis, 1993).
The May 18, 1980 eruption of Mount St. Helens did not
initially produce such an eruption column. Instead, the initial
eruption was a directed blast. This blast was a result of depressurization triggered by an earthquake-initiated landslide on
the north flank of the volcano. The area affected by the directed
blast extended greater than 19 miles from the volcano. Everything within eight miles of the directed blast area was either
Figure 36 Slow-moving basalt flow crossing a highway in Kalapana,

Hawaii.
Another hazard associated with lava flows is they can melt snow
and ice that can result in flooding. Melting of ice beneath a
glacier may produce very large floods called jokulhlaups or
glacier lake outburst floods (Bryant, 1991).
5.8.6.1.5
Dangerous gases
Apart from the water vapor, most of the gases that come out
of volcanoes are poisonous. But they are usually only
dangerous to humans when they are very concentrated,
for example, near volcanic vents. Noxious sulfur dioxide gas
and other pollutants emitted from volcanoes react with oxygen and atmospheric moisture to produce volcanic smog
(vog) and acid rain (US Geological Survey, 2000). Vog poses a
health hazard by aggravating preexisting respiratory ailments,
and acid rain damages crops and can leach lead into household water supplies. Hydrochloric acid vapors can damage
eyes and lung tissue. It can even kill you. When volcanoes
pump out lots of carbon monoxide gas, it can collect in valley
villages. This gas is heavier than air and can kill people because it pushes the oxygen upward so people and animals
living in the valley cannot breathe properly. Volcanic gases can
be very harmful for plants and animals. When volcanoes
create clouds of gas containing mixtures of water and toxic
chemicals, this can turn into acid rain, which is deadly to
plants. When hydrofluoric acid is blasted out of a volcano with
lots of ash, the acid binds to the ash and falls on the ground. If
animals feed on the plants growing there, they can be
poisoned.
5.8.6.2
Indirect Hazards
Indirect hazards take numerous forms including the destruction due to tephra and volcanic landslides and debris
flows, and tsunami. Each is discussed in turn below.
5.8.6.2.1
Tephra
Tephra can also destroy vegetation, which can result in famine.

Famines are the largest indirect hazard produced by volcanic
eruptions. In 1815, after the eruption of Tambora that ejected
151 km3 of ash into the atmosphere, 80 000 people died due
to famine (Bryant, 1991; Francis, 1993).
5.8.6.2.2
Lahar
A lahar is a debris flow composed of a significant component

of volcanic materials (425%) (Fisher and Schmincke, 1984).
Lahars are volcanic mudflows, and they do not have to come
directly from volcanic activity. They occur when huge amounts
of volcanic ash, mixed with water flows down the side of a
mountain. They can flow at speeds more than 100 km h1,
following the path of a river valley, but with the weight of
concrete. Lahars are liquid when they are flowing, and then
harden almost solid when they stop. One cause of a volcano
lahar is an eruption, when volcanic ash mixes with a volcano
glacier, creating a muddy mixture. It is also possible for a lahar
to form when a lake or dam breaks, mixing water with ash
already on the side of a volcano.
Three major categories of lahars by origin are: (1) those
formed by the direct and immediate result of eruptions
through crater lakes, snow or ice, and heavy rains falling
183
during or immediately after an eruption on abundant

unstable loose material. They may also form by dewatering of
debris avalanches and by pyroclastic surges flowing over and
melting snow and ice; (2) lahars may form indirectly from
eruptions that commonly occur shortly after eruptions by
triggering of lahars from earthquakes or rapid drainage of
lakes dammed by erupted products; and (3) many lahars are
unrelated to contemporaneous volcanic activity, occurring by
mobilization of loose tephra during heavy rain or meltwater
on steep slopes of volcanoes during rain or meltwater from
snow seeping into loose debris at any stage of volcano conebuilding or cone degradation.
Moreover, volcanic terrain in areas of extinct volcanic
activity and without a volcanic edifice can give rise to
debris flows composed mainly of epiclastic volcanic fragments. Common nonvolcanic processes by which lahars
and other debris flows form are by heavy rains falling
upon loose debris or by loose debris becoming saturated
with water from melting snow, glaciers or heavy rains.
Water-saturated material can move downhill like wet concrete
when its internal strength is exceeded and can cause terrible
damage.
The eruption of Nevado del Ruiz in Columbia in 1985
sent lahars down the mountainside, burying the city of
Armero under 5 m of mud and debris, and killing B23 000
people. In the recent eruption of Mount Pinatubo in 1991,
700 people were killed by the lahars after the intense rainfall that followed the eruption. In the first few years following
the 1991 eruption, lahars deposited more than 3 km3 of
debris on the lowlands surrounding the volcano, burying
hundreds of square miles of land (Newhall et al., 1998).
A lahar coming off Mount Rainier in Washington sent a
wall of mud 140 m deep, covering a total area of
330 km2. 300 000 people now live in the area covered by that
lahar.
5.8.6.2.3
Debris avalanches, landslides, and tsunamis
Volcanic structural collapse in the form of avalanches, rock

fall, or landslides can be almost any size ranging from a few
loose rocks falling from the crater rim of a volcano to large
avalanches such as the one at Socompa Volcano in Chile that
covers 500 km2 (Francis, 1993). Large-scale debris avalanches
are associated with many volcanic areas including Augustine
Volcano in Alaska, Bandai-san in Japan, and Mount St. Helens
in Washington. Large-scale landslide deposits occur off the
coast of several Hawaiian Islands including Molakai and the
Big Island of Hawaii (Moore and Normark, 1994). A debris
avalanche is formed when an unstable slope collapses and
debris is transported away from the slope. Large-scale avalanches normally occur on very steep volcanoes. There are two
general types of debris avalanches, cold and hot. A cold debris
avalanche generally results from a slope becoming unstable,
whereas a hot debris avalanche is the result of volcanic activity
such as volcanic earthquakes or the injection of magma that
causes slope instability. For several months before Mount St.
Helens erupted in 1980, magma was injected into the north
flank of the volcano creating a bulge that was extremely unstable. An earthquake triggered the movement of land in this
area and the result was a fast moving debris avalanche
(Figures 25 and 27). Some large avalanches have been known
184
Figure 37 Google EarthTM image of the Laki Fissure (3 vertical exaggeration). Note that the fissures trend diagonally across the image from
left to right highlighted by small hills. The dark lower area of the images is basalt produced during the 1783 to 1784 eruption.
Increased
planetary
Albedo
Heterogeneous
chemistry
N2O5
NHO3
CIONO2 CIO
HCl
Stratosphere
hv and OH
SO2
HCl
SO2
Ash
Warming
H2SO4
Nucleation and
particle growth
Rainout
H2O, HCl, Ash
Removal
processes
Troposphere
Cirrus modification
Infrared
Surface cools
Figure 38 Schematic figure showing how a volcanic eruption such as Mount Pinatubo can produce significant perturbations to the Earthatmosphere system by injecting material into the stratosphere. Reproduced from McCormick, M.P., Thomason, L.W., Trepte, C.E., 1995.
Atmospheric effects of the Mt. Pinatubo eruption. Nature 373, 399404.
to carry blocks as large as three kilometers in length, several

kilometers from their source. Such avalanches can travel close
to 300 km h1 (Francis, 1993).
Landslide is a general term for mass movement. It can imply
a gradual movement rather than the more sudden movement of
an avalanche. The perfect example of possible large-scale
landsliding currently taking place is on the south side of the

island of Hawaii. On the south flank of Kilauea Volcano, there
are large fault scarps that are as high as 500 m. Movement along
these fault scarps is causing the south side of the island to fall
into the sea (Moore and Normark, 1994). The hazards that
avalanches and landslides can produce are numerous. Both
185
330
TOMS total ozone (DU)
320
310
300
290
280
270
Jan
Mar
May
Jul
Sep
7990 Nimbus 7 ozone range
Nov
1992 Nimbus 7 ozone
1993 Meteor 3 ozone

Figure 39 Variations in global mean stratospheric ozone measured by the total ozone mapping spectrometer (TOMS) before the Mount
Pinatubo eruption in 1991 [19791990 showing the mean (black line) and variance (blue band)] and the mean for 1992 and 1993. Reproduced
from McCormick, M.P., Thomason, L.W., Trepte, C.E., 1995. Atmospheric effects of the Mt. Pinatubo eruption. Nature 373, 399404.
processes can travel large distances and can wipe out everything
in their paths. They can dam rivers and lakes and produce
flooding. Moreover, the mixture of debris from a landslide or
avalanche with water may produce lahars that can affect people
living in valley areas far away from the volcanos summit
(Francis, 1993). Another important hazard that can result from
avalanches and landslides are tsunamis. When these waves
reach coastal areas they can cause major destruction. Tsunamis
are not produced only by collapsing landmasses, but can also
be produced by volcano-related earthquakes and explosions,
and by lahars or pyroclastic flows that have entered the sea. In
1883, when Krakatau erupted, the main cause of death was due
to tsunamis. These great waves reached as high as 35 m and
killed more than 30 000 people (Pararas-Carayannis, 2003).
5.8.6.2.4
Climate change
Volcanic eruptions can also have a profound impact on climate. However, most eruptions are only large enough to effect
global climate for a few years, with mean annual temperatures
dropping by about 1 1C. The drop in temperature is mainly
due to a reduction of insolation by atmospheric aerosol
loading. Notable historical eruptions that have resulted in net
global cooling have been associated with the 17831784
Icelandic Laki Fissure eruption (Figure 37) and the 1815
Mount Tambora eruption (Franklin, 1784; Thordarson and
Self, 2003; Oppenheimer, 2003). More complex mechanisms
exist such as a reduction in stratospheric ozone due to increased SO2 loading and can lead to net cooling, as associated
with the 1990 Mount Pinatubo eruption (Figures 38 and 39).
Volcanic hazard management
Noncrisis
Risk reduction
Risk analysis
Land-use
planning
Mitigation
measures
Preparation
Volcano
surveillance
Contingency
planning
Public
education
Crisis
Crisis management
Volcano
surveillance
Warnings
and public
information
Emergency
response
measures
Recovery
Figure 40 Volcanic hazard management planning and

implementation.
Supervolcanic eruptions may have a more profound effect on

climate. The Toba supervolcanic eruption of B6977 ka, for
example, is argued to have plunged Earth into a decade long
volcanic winter that may have created a bottleneck in human
evolution (Rampino and Self, 1992; Gibbons, 1993).
186
Remote
sensing
Hydrology
Gas
Seismicity
Ground
deformation
Geophysical
measurements
Stream/
Other
hydro-met. Lake gage
37%
45%
AFM
8%
Downhole
pressure/temperature
10%
Geochemical
2%
Microbarometer
1%
Broadband
12%
Other
Geophysical
2%
Visual
2%
Hydrological/
Meterological
14%
Accelerometer
17%
Short-period
70%
Seismic
52%
Deformation
28%
Borehole
strainmeters
5%
Tiltmeters
10%
CGPS
79%
Figure 41 (a) Different types of methods for helping to monitor and predict volcanic eruptions. (b) Pie charts showing percentages of
instrument types used for volcano monitoring in the United States as of 31 December 2008. The three categories containing the most instruments
(seismic, deformation, and hydrological) are subdivided by instrument type. Reproduced from Guffanti, M., Diefenbach, A.K., Ewert, J.W., Ramsey,
D.W., Cervelli, P.F., Schilling, S.P., 2008. Volcano-monitoring instrumentation in the United States, 2008. USGS Open-File Report 20091165.
CGPS continuously recording global positioning system.

5.8.6.3
Volcanic Hazard Mitigation
A volcanic risk is defined in terms of the potential hazard

produced by an eruption, degree of vulnerability of people
and infrastructures and value of the territory exposed by future
eruptions. A volcanic hazard is measured in terms of the
probability of an eruption event occurring at a given place and
time, which in the extreme situations can be effusive, when a
volcano produces slow moving lava flows, and explosive,
when it produces high rising volcanic clouds of gas and ash
and ground-hugging pyroclastic flows that move at very high
speeds along the slopes of the volcano (Giggenbach, 1996).
Mitigation of volcanic hazards in densely populated areas can
be based on emergency plans or interdisciplinary systems integration guidelines. An emergency plan is designed to wait
for an eruption and manage evacuation, whereas an interdisciplinary approach aims at reorganizing the territory to
minimize the impact of future eruptions and evacuation. The
latter approach is advantageous for a densely populated area
because of the difficulty of forecasting future eruptions more
than several days or weeks in advance. Furthermore, it can be
extremely difficult to evacuate a large number of people who
may be ignorant about the risk or do not wish to evacuate.
Thus, it is necessary to possess reliable data that define the risk
before concluding that a population around a dangerous
volcano needs to be resettled or the area isolated from future
developments (Suryo and Clarke, 1985). Figure 8 summarizes
Aerosols
Concentration/mass
Size distribution
Visible
Post-disaster
Reconnaissance
Impact on society
Visible, IR, Radar
some of the hazards that may be expected from a typical

composite cone volcano.
The management and mitigation of volcanic hazards
should not only occur during crisis periods. It is also important that management of volcanic hazards is initiated and
undertaken in periods of noncrisis, prior to an eruption occurring. Preplanning will ensure that the mitigation measures
employed in response to a crisis are successful. Figure 40
illustrates the different aspects of volcanic hazard management under crisis and noncrisis or pre-event conditions
(Myers and Driedger, 2008).
It is essential to carry out extensive geological investigations
of potentially active volcanoes during periods when those
volcanoes are in repose. Using tephrochronology to study
a volcanos eruptive record enables scientists to reconstruct
how each volcano has erupted in the past (Alloway et al.,
2006). From this information, it is possible to ascertain
the types and magnitude of hazards posed by the volcano,
and to determine how frequently the volcano has erupted in
the past. This information is fundamental, and is always the
starting point when planning and preparing for a future
eruption.
Several new and enhanced technologies are now being
applied to volcano surveillance, including real-time processing
seismic systems, satellite telemetry, use of global positioning
systems, geoelectrical methods, and volcanic-gas monitoring.
SO2
Concentration/mass
UV, Microwave
Active lava
Temperature
Heat flux
Morophology/texture
Visible, IR, Radar
187
Airborne ash
Cloud height/position
Ascent rate/tracking
Concentration/mass
Visible, IR
Topography
Static
Ground deformation
Visible (stereo), Radar
Crater Lakes
Color
Temperature
Heat flux
Morophology/texture
Visible, IR
Mapping
Lithology
Land cover
New ash/lava Morophology/texture
Visible, IR, Radar
Figure 42 Summary of application of remote sensing to volcanology. Reproduced from Oppenheimer, C., 1997. Surveillance and mapping of
volcanoes and their emissions by satellite remote sensing. Geography 82, 317333.
These new technologies, combined with the increased number

and capability of volcano observatories worldwide, have clearly
advanced volcanology. The main monitoring methods are illustrated in Figure 41 highlighting the major types
of methods used in the United States. Ongoing surveillance
of active volcanoes enables scientists to note any changes
to a volcano, and if possible provide warning of an impending
eruption. A variety of geophysical and airborne technologies are
currently in use around the world. One is monitoring
of volcanic earthquakes. Measurement of ground deformation
is a second monitoring method used in volcanically active areas.
Scarpa and Gasparini (1996) highlighted that measurement of
ground deformation can be done in a number of ways:
4. Global positioning systems can be used to measure horizontal and vertical Earth shifts.
A summary of the applications of remote sensing to volcanology are shown in Figure 42.
Changes in gas chemistry, the rate of gas emission from
craters, and the chemistry of thermal spring waters can also be
used to detect changes in a volcano. Other evidence of unrest
can also be detected from changes in groundwater, lake levels,
rate of stream flow, and water temperature (Giggenbach,
1996).
The preparation of hazard maps also helps to determine
whether a volcano is potentially hazardous and to assess the
risk. For that purpose, detailed knowledge about the history
and characteristics of the specific volcano is indispensable,
which requires, among other things, topographic and geologic
mapping. Hazard maps show the pathways of eruption products to be expected (such as lava flows or pyroclastic flows) for
various eruption intensities. Figure 43 provides an example of a
hazard map for the island of Hawaii produced by the US
Geological Survey. Monitoring of volcanoes by satellites has to
1. Measuring distances with electric distance measuring

(EDM) equipment.
2. Ground tilting measurements are made by precise leveling
and using some of the volcanic lakes as large-scale, natural
spirit levels.
3. Horizontal control surveys using triangulation and trilateration techniques.
1
2
3
4
5
6
7
8
9
Kohala
9
Mauna kea
8
Increasing severity
of hazard
188
7
4
Hualalai
3
Mauna
Kilauea
Loa
1
2
6
5
1
3
30 km
2
6
20 miles
Lava flows since 1800
Figure 43 Map of the Island of Hawaii showing the volcanic hazards from lava flows. Severity of the hazard increases from zone 9 to zone 1
defined by the USGS. Reproduced from Tilling, R.I., Heliker, C., Swanson, D.A., 2010. Eruptions of Hawaiian Volcanoes - Past, Present and
Future, Second edition. USGS General Information Product 117, 63 pp. The shaded areas show land covered by flows erupted in the past two
centuries from three of Hawaiis five volcanoes (Hualalai, Mauna Loa, and Klauea).
increase in order to detect possible changes (e.g., temperature or

SO2 emission). Continuous monitoring is essential.
Recent studies using geographical information systems
(GIS) have yielded new and improved hazard maps for
pyroclastic flows and surges. The method inserts the predictive equations in a GIS program coupled with DEM topography, using the LAHARZ program of Schilling (1998). The
result is a GIS-method intended to map hazard zones for
volcano-induced mudflows (Widiwijayanti et al., 2009).
Widiwijayanti et al. (2009) used data from eruptions of a
variety of volcanoes including the Soufriere Hills Volcano in
Montserrat, Mount Merapi in Indonesia, Mount Unzen in
Japan, and Colima in Mexico to develop a predictive equation
for mapping the hazard. The equation considered the crosssectional area of the flows, the area covered and the amount of
material in the flow, and is embedded in the GIS. The program
creates maps that outline where the rock and ash will flow for
different prospective volumes of material expelled. This
method has already been used twice in eruption crises, once at
Merapi in 2006 and in Montserrat in 20062009 to quickly
and effectively create pyroclastic flow maps (Widiwijayanti
et al., 2009).
5.8.7
Future Challenges in the Study of Volcanic

Landforms and Hazards
Probably the most important areas for future attention is

disaster management, which is the process of addressing an
event that has the potential to seriously disrupt the social
fabric of the community. Recognition of volcanic landform
and their past effects is clearly essential for the determining
the nature of volcanic hazards. Disaster management is
similar to disaster mitigation, however, it implies a whole-ofgovernment approach to using community resources to
fight the effects of an event and assumes the community
will be self-sufficient for periods of time until the situation
can be stabilized (Johnston and Houghton, 1995). Through
disaster management, we cannot completely counteract
the damage but it is possible to minimize the risks through
early warning, provide developmental plans for recuperation from the disaster, generate communication and medical
resources, and aid in rehabilitation and postdisaster
reconstruction.
This approach is advantageous for a densely populated
area because of our inability to forecast future eruptions
more than several days or weeks in advance and evacuate
large number of people who may be ignorant about the risk
or do not wish to evacuate. Waiting for an unrest of a
hazardous volcano and then trying to manage the hazard may
only prove effective in nondensely populated areas where the
social, political, and economic factors are easily manageable.
Thus, as technology continues to develop and our ability to
understand and foresee sudden changes in volcanic behavior
improves, we also need to seek a more fundamental association between the scientific and public policy communities.
Only in this way can we hope to manage, advise and oversee
the safety of the constantly growing populations in volcanically active areas.
189
Acknowledgments
Thanks to Tim Phillips for drafting the figures for this article
and to Urlich Kamp and Jack Shroder for their very careful
reviews.
References
Alloway, B.V., Larsen, G., Lowe, D.J., Shane, P.A.R., Westgate, J.A., 2006.
Tephrochronology. In: Elias, S.A. (Ed.), Encyclopaedia of Quaternary Science.
Elsevier, London, pp. 28692898.
Barberi, F., Rosi, M., Sodi, A., 1993. Volcanic hazard assessment at Stromboli
based on review of historical data. Acta Vulcanologica 3, 173187.
Bardintzeff, J.-M., McBirney, A.R., 2000. Volcanology: Massachusetts. Jones &
Bartlett Publishers, 268 pp.
Beget, J.E., Hopkins, D.M., Charron, S.D., 1996. The Largest Known Maars on
Earth, Seward Peninsula, Northwest Alaska. Arctic 49, 6269.
Bowen, N.L., 1928. The Evolution of Igneous Rocks. Princeton University Press,
Princeton, NJ, 332 pp.
Brown, K.M., 1990. The nature and hydrogeological significance of mud diapirs
and diatremes for accretionary prisms. Journal of Geophysical Research 95,
89698982.
Bryan, T.S., 2008. The Geysers of Yellowstone. University Press of Colorado, 462 pp.
Bryant, E.A., 1991. Natural Hazards. Cambridge University Press, New York,
294 pp.
Bunnell, D., 2008. Caves of Fire: Inside Americas Lava Tubes. National
Speleological Society, 124 pp.
Carracedo, J.C., Singer, B., Jicha, B., et al., 2003. La erupcion y el tubo volcanico
del Volcan Corona (Lanzarote, Islas Canarias). Estudia Geologia 59, 277330.
Chesner, C.A., Rose, W.I., 1991. Stratigraphy of the Toba tuffs and evolution of the
Toba caldera complex, Sumatra, Indonesia. Bulletin of Volcanology 53,
343356.
Chesner, C.A., Rose, W.I., Deino, A., Drake, R., Westgate, J.A., 1991. Eruptive
history of Earths largest Quaternary caldera (Toba, Indonesia) clarified. Geology
19, 200203.
Christiansen, R.L. 2001. The Quaternary and Pliocene Yellowstone Plateau Volcanic
Field of Wyoming, Idaho, and Montana: U.S. Geological Survey Professional
Paper 729-G, 145 pp.
Corti, G., 2009. Continental rift evolution: from rift initiation to incipient break-up in
the Main Ethiopian Rift, East Africa. Earth-Science Reviews 96, 153.
Davies, R.J., Mathias, S.A., Swarbrick, R.E., Tingay, M.J., 2011. Probabilistic
longevity estimate for the LUSI mud volcano, East Java. Journal of the
Geological Society 168, 517523.
Davison, I., Al-Kadasi, M., Al-Khirbash, S., et al., 1994. Geological evolution of the
southern Red Sea rift-margin: Republic of Yemen. Geological Society of
America, Bulletin 106, 14741493.
Decker, R., Decker, B., 2005. Volcanoes. WH Freeman, New York, 320 pp.
Denverpost.com, 2011. Photos: Puyehue Volcano erupts in Southern Chile. http://
photos.denverpost.com/mediacenter/2011/06/photos-puyehue-volcano-erupts-insouthern-chile/. Accessed October 2011.
Deville, E., Guerlais, S.-H., Callec, Y., et al., 2006. Liquefied vs stratified sediment
mobilization processes: insight from the South of the Barbados accretionary
prism. Tectonophysics 428, 3347.
Dimitrov, L.I., 2002. Mud volcanoes the most important pathway for degassing
deeply buried sediments. Earth-Science Reviews 59, 4976.
Dimitrov, L.I., 2003. Mud volcanoes a significant source of atmospheric methane.
Geo-Marine Letters 23, 155161.
Elkins-Tanton, L.T., 2005. Continental magmatism caused by lithospheric
delamination. Geological Society of America Special Paper 388, pp. 449461.
Esser, R., Kyle, P., McIntosh, W., 2004. 40Ar/39Ar dating of the eruptive history of
Mount Erebus, Antarctica: volcano evolution. Bulletin of Volcanology 66,
671686.
Fisher, R.V., 1984. Submarine volcaniclastic rocks. In: Kokelaar, B.P., Howells, M.F.
(Eds.), Marginal basin geology: volcanic and associated sedimentary and
tectonic processes in modern and ancient marginal basins. Special Publication
of the Geological Society of London, 16, pp. 527.
Fisher, R.V., Schmincke, H.-U., 1984. Pyroclastic Rocks. Springer-Verlag, Berlin,
472 pp.
Foxworthy, B.L., Hill, M., 1982. Volcanic Eruptions of 1980 at Mount St. Helens:
The First 100 day. USGS Professional Paper 1249, 125 pp.
190
Francis, P., 1993. Volcanoes: A Planetary Perspective. Oxford University Press, New
York, 443 pp.
Franklin, B., 1784. Meteorological imaginations and conjectures. Memoirs of the
Literary and Philosophical Society of Manchester 2, 373377.
Giacomelli, L., Perrotta, A., Scandone, R., Scarpati, C., 2003. The eruption of
Vesuvius of 79 AD and its impact on human environment in Pompei. Episodes
26, 234237.
Gibbons, A., 1993. Pleistocene population explosions. Science 262, 2728.
Giggenbach, W.F., 1996. Chemical composition of volcanic gases. In: Scarpa, R.,
Tilling, R.I. (Eds.), Monitoring and Mitigation of Volcano Hazards. SpringerVerlag, Berlin, pp. 221256.
Hanson, S.L., Falster, A.U., Simmons, W.B., 2008. Mineralogy of fumarole deposits
at sunset Crater Volcano National Monument, Northern Arizona. Rocks and
Minerals 83, 534544.
Harris, S.L., 2005. Mount Rainier: Americas Most Dangerous Volcano. Fire
Mountains of the West, Third ed. Mountain Press Publishing Company,
Missoula, Montana.
Hawley, J.W., Lozinsky, R.P., 1993. Hydrogeologic framework of the Mesilla Basin in
New Mexico and western Texas: New Mexico Bureau of Geology and Mineral
Resources Open-file Report 323, 98 pp.
Heiken, G.H., 1971. Tuff rings: examples from Fort Rock-Christmas Lake
Valley, south-central Oregon. Journal of Geophysical Research 76, 5615
5626.
Heiken, G., Wohletz, K., 1985. Volcanic Ash. University of California Press,
Berkeley, 246 pp.
Hoffer, J.M., 1976. The Potrillo Basalt Field, South-central. New Mexico Geological
Society Special Publication, New Mexico, vol. 5, pp. 8992.
Holm, R.F., 1987. Significance of agglutinate mounds on lava flows associated with
monogenetic cones: an example from sunset crater, Northern Arizona. Geological
Holm, R.F., Moore, R.B., 1987. Holocene Scoria Cone and Lava Flows at Sunset
Crater, Northern Arizona. Geological Society of America Centennial Field Guide
Rocky Mountain Section, 393397.
Hooper, P.R., Camp, V.E., Reidel, S.P., Ross, M.E., 2007. The origin of the
Columbia river flood basalt province: plume versus nonplume. Geological
Society of America vol. 430, Special Paper, 635668.
Horwell, C.J., Baxter, P.J., 2006. The respiratory health hazards of volcanic ash: a
review for volcanic risk mitigation. Bulletin of Volcanology 69, 124.
Horwell, C.J., 2007. Grain-size analysis of volcanic ash for the rapid
assessment of respiratory health hazard. Journal of Environmental Monitoring 9,
11071115.
Huff, W.D., Bergstrom, S.M., Kolata, D.R., 2010. Ordovician explosive volcanism.
In: Finney, S.C., Berry, W. (Eds.), Earth system during the Ordovician Period.
Geological Society of America, Special Paper 466, pp. 1328.
Inanli, F.O., Huff, W.D., 2009. Quartz crystal size distribution of the Ordovician
Millbrig K-bentonite. Journal of Volcanology and Geothermal Research 184,
285291.
International Volcanic Health Hazard Network, 2011. The health hazards of volcanic
ash: a guide for the public. http://cav.volcano.info/files/health.pdf. Accessed
October 2011.
Johnston, D., Houghton, B.F., 1995. Living with volcanoes. Tephra 14, 2228.
Kagel, A., Bates, D., Gawell, K., 2007. A Guide to Geothermal Energy and the
Environment. Geothermal Energy Association Report, http://geo-energy.org/pdf/
reports/AGuidetoGeothermalEnergyandtheEnvironment10.6.10.pdf. Accessed
October 2011.
Kerr, R.A., 2010. Magnetics Point to Magma Ocean at Io. Science 327, 408409.
Kilburn, C., McGuire, B., 2001. Italian Volcanoes (Classic Geology in Europe 1).
Dunedin Academic Press Ltd., 174 pp.
Kling, G.W., Clark, M.A., Wagner, G.N., et al., 1987. The 1986 Lake Nyos Gas
Disaster in Cameroon, West Africa. Science 236, 169175.
Kokelaar, B.P., Durant, G.P., 1983. The submarine eruption and erosion of Surtla
(Surtsey), Iceland. Journal of Volcanology and Geothermal Research 19,
239246.
Kopf, A.J., 2002. Significance of mud volcanism. Reviews of Geophysics 40, 52.
Luhr, J., Simkin, T., 1993. Parcutin: The Volcano Born in a Mexican Cornfield.
Geoscience Press, Phoenix, AZ.
Lund, J.W., 1993. Spas and Balneology in the United States. Geo-Heat Center
Quarterly Bulletin 14(4), pp. 13.
Malahoff, A., 1987. Geology of the summit of Lo`ihi submarine volcano. In: Decker,
R.W., Wright, T.L., Stauffer, P.H. (Eds.), Volcanism in Hawaii. U.S. Geological
Survey Professional paper, vol. 1350, pp. 133144.
Manga, M., Brumm, M., Rudolph, M.L., 2009. Earthquake triggering of mud
volcanoes. Marine and Petroleum Geology 26, 17851798.
Mason, B.G., Pyle, D.M., Oppenheimer, C., 2004. The size and frequency of
the largest explosive eruptions on Earth. Bulletin of Volcanology 66,
735748.
Mazzini, A., Nermoen, A., Krotkiewski, M., Podladchikov, Y., Planke, S., Svensen,
H., 2009. Strike-slip faulting as a trigger mechanism for overpressure release
through piercement structures. Implications for the Lusi mud volcano, Indonesia.
Marine and Petroleum Geology 26, 17511765.
Mazzini, A., Svensen, H., Akhmanov, G.G., Aloisi, G., Planke, S., Malthe-Srenssen,
A., Istadi, B., 2007. Triggering and dynamic evolution of the LUSI mud volcano,
Indonesia. Earth Planetary Science Letters 261, 375388.
McClelland, L., Simkin, T., Summers, M., Nielson, E., Stein, T.C., 1989. Global
volcanism 19751985. Prentice Hall, Englewood Cliffs, NJ, and American
Geophysical Union, Washington, DC, 655 pp.
Millar, C.I., King, J.C., Westfall, R.D., Alden, H.A., Delany, D.L., 2006. Late
Holocene forest dynamics, volcanism, and climate change at Whitewing
Mountain and San Joaquin Ridge, Mono County, Sierra Nevada, CA, USA.
Milkov, A.V., 2000. Worldwide distribution of submarine mud volcanoes and
associated gas hydrates. Marine Geology 167, 2942.
Moore, J.G., 1967. Base surge in recent volcanic eruptions. Bulletin of Volcanology
30, 337363.
Moore, J.G., Clague, D.A., Holcomb, R.T., Lipman, P.W., Normark, W.R., Torresan,
M.E., 1989. Prodigious Submarine Landslides on the Hawaiian Ridge. Journal
Moore, J.G., Jakobsson, S., Holmjarn, J., 1992. Subsidence of Surtsey volcano,
19671991. Bulletin of Volcanology 55, 1724.
Moore, J.G., Normark, W.R., 1994. Giant Hawaiian landslides. Annual Review of
Earth and Planetary Science 22, 119144.
Morgan, W.J., 1971. Convection plumes in the lower mantle. Nature 230,
4243.
Myers, B., Brantley, S.R., Stauffer, P., Hendley, J.W., 2008. What are volcano
hazards? U.S. Geological Survey Fact Sheet 00297. Revised March 2008, 2
sheets [http://pubs.usgs.gov/fs/fs002-97/].
Myers, B., Driedger, C., 2008. Geologic hazards at volcanoes: U.S. Geological
Survey General Information Product 64, 1 sheet [http://pubs.usgs.gov/gip/64/].
Naslund, H.R., McBirney, A.R., 1996. Mechanism of formation of igneous layering:
in Layered Intrusions. In: Cawthorn, R.G. (Ed.), Developments in Petrology.
Elsevier, Amsterdam, vol. 15, pp. 143.
Newhall, C., Hendley, J.W., Stauffer, P.H., 1998. The Cataclysmic 1991 Eruption of
Mount Pinatubo, Philippines. U.S. Geological Survey Fact Sheet 11397, 2
sheets [http://pubs.usgs.gov/fs/1997/fs113-97/].
Newhall, C.G., Self, S., 1982. The volcanic explosivity index (VEI): an estimate of
explosive magnitude for historical volcanism. Journal of Geophysical Research
87, 12311238.
Oppenheimer, C., 2003. Climatic, environmental and human consequences of the
largest known historic eruption: Tambora volcano (Indonesia) 1815. Progress in
Padovani, E.R., Reid, M.R., 1989. Field guide to Kilbourne Hole maar: New Mexico
Bureau of Mines and Mineral Resources Memoir 46, pp. 174185.
Pararas-Carayannis, G., 2003. Near and far-field effects of Tsunamis generated by
the Paroxysmal Eruptions, Explosions, Caldera Collapses and slope failures of
the Krakatau Volcano in Indonesia, on August 2627, 1883. Journal of Tsunami
Hazards 21, 191222.
Pinkerton, H., Wilson, L., 1994. Factors controlling the lengths of channel-fed lava
flows. Bulletin of Volcanology 56, 108120.
Planke, S., Svensen, H., Hovland, M., Banks, D.A., Jamtveit, B., 2003. Mud and
fluid migration in active mud volcanoes in Azerbaijan. Geo-Marine Letters 23,
258268.
Pukui, M.K., Elbert, S.H., Mookini, E.K. (Eds.), 1974. Place Names of Hawaii,
revised and expanded edition. University of Hawaii Press, Honolulu, 312 pp.
Rampino, M.R., Ambrose, S.H., 2000. In: McCoy, F.W., Heiken, G. (Eds.), Volcanic
Hazards and Disasters in Human Antiquity. Geological Society of America,
Boulder, Special Paper 345, pp. 7182.
Rampino, M.R., Self, S., 1992. Volcanic winter and accelerated glaciation following
the Toba super-eruption. Nature 359, 5052.
Rampino, M.R., Self, S., 1993. Climate-volcanic feedback and the Toba eruption of
74 000 years ago. Quaternary Research 40, 269280.
Regelous, M., Hofmann, A.W., Abouchami, W., Galer, S.J.G., 2003. Geochemistry of
lavas from the Emperor Seamounts, and the geochemical evolution of Hawaiian
magmatism from 85 to 42 Ma. Joural of Petrology 44, 113140.
Rhinehart, J.S., 1980. Geysers and Geothermal Energy. Springer-Verlag, New York,
223 pp.
Rittmann, A., 1962. Volcanoes and Their Activity. Wiley, New York, 305 pp.
Roberts, K.S., Davies, R.J., Stewart, S.A., Tingay, M., 2011. Structural controls on
mud volcano vent distributions: examples from Azerbaijan and Lusi, east Java.
Rose, W.I., Chesner, C.A., 1987. Dispersal of ash in the great Toba eruption, 75 ka.
Geology 15, 913917.
Russell, A.J., Gregory, A.R., Large, A.R.G., Fleisher, P.J., Harris, T.D., 2007. Tunnel
channel formation during the November 1996 jokulhlaup, Skeijararjokull,
Iceland. Annals of Glaciology 45, 95103.
Scarpa, R., Gasparini, P., 1996. A review of volcano geophysics and volcano
monitoring methods. In: Scarpa, R., Tilling, R.I. (Eds.), Monitoring and
Mitigation of Volcano Hazards. Springer-Verlag, Berlin, pp. 322.
Schilling, S.P., 1998. LAHARZ: GIS programs for automated mapping of laharinundation zones. U.SG.S. Open-file Report 98638, 80 pp.
Schmincke, H.-U., Fisher, R.V., Waters, A.C., 1973. Antidune and chute and pool
structures in the base surge deposits of the Laacher See area, Germany.
Sedimentology 20, 553574.
Schmincke, H.-U., 2004. Volcanism. Springer, Berlin, 324 pp.
Seager, W.R., 1987. Caldera-like collapse at Kilbourne Hole maar. New Mexico
Geology 9, 6973.
Self, S., Kienle, J., Huot, J.P., 1980. Ukinrek Maars, Alaska, II. Deposits and
formation of the 1977 craters. Journal of Volcanology and Geothermal Research
7, 3965.
Sen, G., 2001. Generation of Deccan Trap magmas. Proceeding of the Indian
Academy of Sciences (Earth and Planetary Science) 110, 409431.
Shea, T., Gurioli, L., Houghton, B.F., Cioni, R., Cashman, K.V., 2011. Column
collapse and generation of pyroclastic density currents during the A.D. 79
eruption of Vesuvius: The role of pyroclast density. Geology 39, 695698.
Sigmundsson, F., Hreinsdottir, S., Hooper, A., et al., 2010. Intrusion triggering of
the 2010 Eyjafjallajokull explosive eruption. Nature 468, 426430.
Sigurdsson, H., Carey, S.N., Fisher, R.V., 1987. The 1982 eruptions of El Chichon
volcano, Mexico (3): Physical properties of pyroclastic surges. Bulletin of
Smiley, T.L., 1958. The geology and dating of sunset crater, Flagstaff, Arizona.
In: Anderson, R.Y., Harshbarger, J.W. (Eds.), Guidebook of The Black Mesa
Basin, Northeastern Arizona. New Mexico Geological Society, Socorro, NM,
pp. 186190.
Smithsonian National Museum of Natural History, 2011a. Global volcanism
program: How many active volcanoes are there in the world? http://
www.volcano.si.edu/education/questions/index.cfm?faq=03. Accessed October
2011.
Smithsonian National Museum of Natural History, 2011b. Global volcanism
program: Hunga Tonga-HUnga Haapai. http://www.volcano.si.edu/world/
volcano.cfm?vnum=0403-04 . Accessed October 2011.
Smithsonian National Museum of Natural History, 2011c. Global Volcanism
Program: Kavachi. http://www.volcano.si.edu/world/volcano.cfm?vnum=0505-06=.
Accessed October 2011.
Smithsonian National Museum of Natural History, 2011d. Global Volcanism
Program: Kavachi eruptive history. http://www.volcano.si.edu/world/
volcano.cfm?vnum=0505-06=&volpage=erupt. Accessed October 2011.
Stewart, S.A., Davies, R.J., 2006. Structure and emplacement of mud volcano
systems in the South Caspian Basin. AAPG Bulletin 90, 753770.
Summerfield, M.A., 1991. Global Geomorphology. Pearson, Prentice Hall, Harlow,
UK.
Suryo, I., Clarke, M.C.G., 1985. The occurrence and mitigation of volcanic hazards
in Indonesia as exemplified at the Mount Merapi, Mount Kelut and Mount
Galunggung volcanoes. Quarterly Journal of Engineering Geology and
Hydrogeology 18, 7998.
191
Tarduno, J., Bunge, H.-P., Sleep, N., Hansen, U., 2009. The Bent Hawaiian-Emperor
Hotspot Track: Inheriting the Mantle Wind. Science 324, 5053.
Tester, J.W., Anderson, B.J., Batchelor, A.S., et al., 2006. The Future of Geothermal
Energy, Impact, of Enhanced Geothermal Systems (Egs) on the United States in
the 21st Century: An Assessment. Idaho National Laboratory report. ISBN
0-615-13438-6, 346 pp.
Thordarson, T., Self, S., 2003. Atmospheric and environmental effects of the
17831784 Laki eruption: a review and reassessment. Journal of Geophysical
Research 108, 4011.
US Forest Department 2011. Coconino National Forest: Lava River Cave.
http://www.fs.fed.us/r3/coconino/recreation/peaks/lava-river-cave.shtml. Accessed
October 2011.
US Geological Survey, 2000. Volcanic air pollution a hazard in Hawaii. US
Geological Survey Fact Sheet, pp. 169197.
US Geological Survey, 2011a. Volcanic ash: effects and mitigation strategies.
http://volcanoes.usgs.gov/ash/health/). Accessed October 2011.
US Geological Survey, 2011b. Volcanic hazard program: Stratovolcano. http://
volcanoes.usgs.gov/images/pglossary/stratovolcano.php. Accessed
October 2011.
US Geological Survey, 2011c. Hawaiian Volcano Observatory: Klauea Perhaps the
Worlds Most Active Volcano. http://hvo.wr.usgs.gov/kilauea/. Accessed October
2011.
Valentine, G.A., 1987. Stratified flow in pyroclastic surges. Bulletin of Volcanology
49, 616630.
Walker, G.P.L., 1984. Characteristics of dune-bedded pyroclastic surge bedsets.
Journal of Volcanology and Geothermal Research 20, 281296.
Walker, G.P.L., 1991. Structure and origin by injection of lava under surface crust,
of tumuli, lava rises, lava-rise pits, and lava-inflation clefts in Hawaii.
Bulletin of Volcanology 53, 546558.
Widiwijayanti, C., Voight, B., Hidayat, D., Schilling, S.P., 2009. Objective rapid
delineation of areas at risk from block-and-ash pyroclastic flows and surges.
Bulletin of Volcanology 71, 687703.
Wilson, J.T., 1963. Evidence from islands on the spreading of ocean floors. Nature
197, 536538.
Winchester, S., 2003. Krakatoa: The Day the World Exploded: August 27, 1883.
HarperCollins Publishers, New York, 416 pp.
Wohletz, K.H., 1983. Mechanisms of hydrovolcanic pyroclastic formation: grain-size,
scanning electron microscopy, and experimental studies. Journal of Volcanology
and Geothermal Research 17, 3163.
Wohletz, K.H., 1986. Explosive magma-water interactions: thermodynamics,
explosion mechanisms, and field studies. Bulletin of Volcanology 48,
245264.
Young, D.A., 1998. N.L. Bowen and Crystallization-Differentiation: The Evolution of
a Theory: Mineralogical Society of America, Monograph 4, Washington, DC,
276 pp.
YouTube, 2010a. Kavachi: the Birth of an island. http://www.youtube.com/
watch?v=L4rcT6TWO-E. Accessed October 2011.
YouTube, 2010b. Etnatao: Kavachi a submarine volcano. http://www.youtube.com/
watchv=G4LaM8iUc28. Accessed October 2011.
Zdanowicz, C.M., Zielinski, G.A., Germani, M.S., 1999. Mount Mazama eruption:
Calendrical age verified and atmospheric impact assessed. Geology 27,
621624.
Zielinski, G.A., Mayewski, P.A., Meeker, L.D., Whitlow, S., Twickler, M.S., 1996.
Potential atmospheric impact of the Toba mega-eruption w71,000 years ago.
192
Biographical Sketch
Warren D. Huff is currently professor of geology at the University of Cincinnati, where he has been a faculty
member since 1963. Much of his research deals with the study of bentonites and K-bentonites, which are the
remains of explosively erupted volcanic ash layers. These layers are now altered largely to clay minerals, although
some original volcanic crystals remain. He studies both types of minerals to learn about the nature of the source
volcanoes, many of which are several hundreds of millions years old, as well as the natural processes by which the
volcanic ash layers have been buried in the earth and altered to their present form. In recent years, he has served as
a technical consultant to NASA on the subject of Martian soil clays.
Professor Lewis A. Owen received his PhD in Geomorphology from the University of Leicester, UK, in 1988.
Before joining the University of Cincinnati in 2004, he held positions at the Hong Kong Baptist University, Royal
Holloway University of London, and the University of California Riverside. Professor Owens research focuses
on the Quaternary geology and geomorphology of tectonically active mountain belts and their forelands, particularly in the HimalayanTibetan orogen and the Cordilleras of North and South America. He has also
undertaken research in other tectonically active regions, including the Red Sea margin in Yemen and the Atlas and
Anti-Atlas Mountains of Morocco. In 2011, Professor Owen was awarded the Busk Medal from the Royal Geographic Society for his field research in Quaternary history and geomorphology in tectonically active areas.
5.9 Hot Spots and Large Igneous Provinces

P Bache`lery, CNRS-IRD-Universite Blaise Pascal, Clermont-Ferrand, France, and Laboratoire GeoSciences Reunion Universite de
La Reunion, France
N Villeneuve, Laboratoire GeoSciences Reunion Universite de La Reunion, France
5.9.1
Introduction
5.9.1.1
Types of Hot Spot Magmatic Structures
5.9.1.2
What is a Hot Spot Volcano?
5.9.2
Hot Spot Volcanic Chains
5.9.2.1
Volcanic Chain Analysis
5.9.2.1.1
Long-lived age-progressive volcanism
5.9.2.1.2
Short-lived age-progressive volcanism
5.9.2.1.3
No age-progressive volcanism
5.9.2.2
Volcanic Chain Analysis. The Hawaii-Emperor Example
5.9.2.2.1
Young stage of the Hawaiian chain: The Southeastern Islands, from Hawaii to Kaula
5.9.2.2.2
Oldest stage of the Hawaiian chain: The Northwestern Islands, from Nihoa to the Kure atoll
5.9.2.2.3
Emperor Chain: From Colahan to Meiji seamounts
5.9.2.3
Hot Spot Chains and Plate Tectonics
5.9.3
Hot Spot Volcanoes
5.9.3.1
Edifices General Morphology
5.9.3.2
Construction versus Destruction
5.9.3.3
Model of Building
5.9.3.3.1
Submarine preshield stage
5.9.3.3.2
Shield-building stage
5.9.3.3.3
Postshield stage
5.9.3.3.4
Erosional and late volcanism
5.9.3.3.5
Atoll and seamounts
5.9.3.4
Volcanic and Tectonic Morphologies
5.9.3.4.1
Summit areas, calderas and pit craters
5.9.3.4.2
Cones, eruptive fissures and rift zones
5.9.3.4.3
Lava flow fields
5.9.3.4.4
Faults, slides, and slumps
5.9.3.5
Erosional Morphologies
5.9.3.5.1
Influence of volcano shape
5.9.3.5.2
Influence of volcanic activity
5.9.3.5.3
Influence of geology and lithology
5.9.3.6
Submarine Morphologies
5.9.3.6.1
Submarine volcanic activity
5.9.3.6.2
Formations built by volcaniclastic materials
5.9.3.6.3
Sedimentary volcaniclastic systems
5.9.4
Conclusion
Acknowledgments
References
Glossary
Aseismic ridge A long and linear, mostly-volcanic,
structure that spans across the ocean floor.
Acidic /Acid Acidic magmas or rocks are characterized
by high silica content (greater than 63% SiO2 by
weight), and the prevalence of alkali feldspars and/or
quartz.
195
196
197
198
198
199
199
199
199
200
200
200
201
204
204
207
208
208
209
209
209
209
209
209
212
215
216
217
217
218
218
221
222
223
223
227
228
228
Basic Basic magmas or rocks have low silica content

(between 45% and 52% SiO2 by weight) and are iron and
magnesium-rich, which yield mafic minerals (pyroxene,
amphibole, biotite, olivine).
Benmoreite/benmoreitic A volcanic rock type
intermediate in composition between alkali basalt and
trachyte. Benmoreite is a sodic variant of trachyandesite.
Bache`lery, P., Villeneuve, N., 2013. Hot spots and large igneous provinces. In:
pp. 193233.
http://dx.doi.org/10.1016/B978-0-12-374739-6.00084-1
193
194
Hot Spots and Large Igneous Provinces
Bomb (volcanic) A fragment of molten lava, larger than

64 mm in diameter, thrown out from a volcanic vent by the
rapid expansion of gases (commonly called explosion)
during an eruption. A bomb is a pyroclast.
Cinder cone A conical hill formed by the accumulation of
bombs, lapilli and cinders thrown from a volcanic vent by
the rapid expansion of gases.
Crater A more or less circular depression, generally atop a
volcanic cone, from which magma erupts or erupted.
Cumulate (magmatic) A rock formed by crystals
accumulation or segregation during the cooling and
crystallization of a magma chamber.
Differentiation Modification of the chemical
composition of a cooling magma by fractional
crystallization or another process.
Dike A blade-like igneous body, a few meters thick, that
is oriented perpendicularly of the pre-existing intruded
rock-layers. In most cases, a dike is more or less vertical.
Debris avalanche A large-scale landslide forming deposits
composed of fragmented rocks from the collapsing flank of
a volcano. Fan-like morphology and hummocky surface is
characteristic of debris avalanche deposits, together with
jigsaw cracks in the brecciated blocks.
Debris flow A water-saturated viscous flow of rock
fragments and mud.
Epiclastic Detrital material produced by the physical
disintegration (erosion or weathering) of preexisting rocks.
Felsic Refers to silicate magmas or rocks rich in light
chemical elements such as Si, Al, Na and K. The term
felsic combines the words feldspar and silica. It is
commonly used as a synonym for silicic or acidic.
Fractional crystallization A mechanism of crystalliquid
separation as magmas cool. While crystals separate to form
cumulates, the liquid remaining in the chamber must
progressively changes its composition, becoming enriched
in incompatible elements.
Gabbro A medium- or coarse-grained plutonic (intrusive)
rock primarily made of plagioclase feldspar and pyroxene.
Gabbro is formed by the slow crystallization of a
basaltic magma.
Guyot A flat-topped seamount.
Hawaiian (dynamism) An eruption style characterized by
continuous jets of lava forming fire fountains.
Hummocky (surface or topography) An irregular
morphology with hills or ridges, typical of the surface of
debris avalanche deposits.
Hydrothermal system A system, heated by a deep magma
body, characterized by circulation of hot water and
vapor.
Hyaloclastite A rock formed by agglomeration of small
glassy angular fragments from magma erupted under water,
ice or into water-saturated sediments.
Ignimbrite A rock formed from pyroclastic-flow deposits
(mixture of welded or unwelded volcanic ash and pumice
lapilli, with lithic fragments).
Intrusion A magma body injected under the Earths
surface that cools and crystallizes to form a solid rock.
Lava lake A lake of molten lava confined in a crater and
fed by an active magmatic conduit.
Lava tube A roofed-channel in which the lava can flow

long distances, without losing much heat.
Mafic Refers to a silicate magma, mineral or rock rich in
magnesium and iron.
Phreatic (eruption or explosion) A type of volcanic
eruption that occurs when high-level ground water is
rapidly heated and vaporized by a rising magma. The
resulting steam explosions eject fragments (from ash to
block) of country rocks, without juvenile magma, unlike
phreatomagmatic eruptions.
Phreatomagmatic (eruption) An eruption type that
results from interaction between magma and ground water
or surface water. Tephra usually consist of glassy juvenile
clasts with variable amounts of lithic fragments.
Pillow-lava Lava flows formed in the subaqueous
environment characterized by a chilled glassy margin, and
tubes and rounded lobes that resemble pillows.
Plinian (eruption) A very explosive eruption type named
after Pliny, a Roman natural philosopher who observed the
eruption of Vesuvius in 79 AD. Plinian eruptions are
potentially devastating eruptions, marked by sustained
column of gas, juvenile ash and pumice, with little lithic
fragments.
Pyroclast Particle formed during explosive activity of a
volcanic eruption.
Pyroclastic (deposit / material) Produced by the
explosive activity of volcanoes.
Pyroclastic flow Deposit of a pyroclastic density current.
It is a fast-moving hot suspension of particles (tephra,
pumice, block) and gases that flows down from a volcano,
driven by gravity and gas expansion.
Seafloor spreading A theory first proposed by the
geophysicist H. H. Hess to explain that oceanic crust
continuously formed at variable velocities along mid-ocean
ridges, and spreads out laterally away.
Seamount A submarine volcano.
Spatter-cone / -rampart Cone or rampart built around an
eruptive vent or fissure by fluidal basaltic lava fragments
deposited hot, and generally welded.
Stratovolcano A volcano, generally with steep slopes,
built by alternating accumulations of lava flows and
pyroclastic deposits, and by intrusions.
Strombolian (dynamism) An eruption style characterized
by discrete explosions at periodic intervals.
Surtseyan eruption Volcanic explosions in seawater
dominated by glassy tephra, as observed at Surtsey (Iceland)
in 1963, resulting in tuff cones.
Tephra A fragmented volcanic ejecta thrown by explosion
during an eruption.
Trachyte/trachytic A fine-grained acid and Si-saturated
volcanic rock that consists of essential alkali feldspar and
plagioclase, relatively minor quartz or feldspathoid and
with biotite, pyroxene or amphibole as accessory minerals.
Tsunami A big wave in the sea generated by a sudden
event like an earthquake or a landslide.
Ultrabasic Ultrabasic magmas or rocks that have very low
silica content (lower than 45% SiO2 by weight) and are
magnesium-rich, which yield mafic minerals such as olivine
and pyroxene.
Vent An opening through which volcanic material is

emitted.
Viscosity A value accounting for the resistance of a
material to flow in response to stress.
195
Volcaniclastic Refers to any material or deposit formed by

particles from the fragmentation of volcanic rocks and
sediment transport.
Abstract
This chapter examines the tectonic geomorphology of hot spot volcanism and specifically oceanic hot spot volcanism. The
best way to explain and understand volcanic edifices development above hot spots is by using nested scales: starting from a
large overview to progressively more and more detailed scales. The construction of hot spot volcanic chains and the
identification of morphologies from the deep sea floor to volcano summits are considered in detail. Mechanisms of
construction, evolution, and dismantling of volcanic edifices are examined with descriptions using well-known examples
and models. Limitations and imperfections within models for hot spot volcanism are mainly related to the specific context
and scale, and to magmatology. At the oceanic scale, the models are influenced by the geodynamic context (plate motion
and proximity of a subduction or an accretion zone). At the volcanic edifice scale, the models are constrained by the history
of the volcano development, for example, cones superimpositions, proximity of fractures, zone of weakness, stress field,
major mass wasting, and rift zones.
5.9.1
Introduction
Hot spots and large igneous provinces (LIPs) are regions with
intense volcanism that cannot be directly attributed to one of
the major types of magma-generating processes acting at tectonic plate boundaries (e.g., mid-ocean ridge magmatism or
subduction magmatism). With the birth of the seafloor
spreading and plate tectonics theories in the early 1960s,
most of the worlds active volcanoes were recognized to occur,
logically, along or near the plate boundaries, with the noticeable exception of some volcanic spots located within plate
interiors, far from the nearest boundary with any other tectonic plate. The Hawaiian archipelago was one among the best
examples of this new intraplate volcanism.
The hot spot hypothesis, first proposed by Tuzo Wilson in
1963 for the Hawaiian chain, appears as a clever scheme to
reconcile the intraplate volcanism with the plate motions,
and, as such, the hypothesis was rapidly welcomed. Marine
research benefited from significant technological developments in the 1940s and 1950s, and with the strong interest
generated by the studies on seafloor spreading, key data were
obtained, at the same time, on seamounts and island chains.
Wilsons theory of a hot spot located deep enough in the
mantle to appear fixed relative to tectonic plate motion, to
cause linear volcanic chains on the overriding plate to develop, was generalized by Morgan (1971, 1972) who introduced the concept of mantle plume. From the observation of
a dozen volcanic chains such as Hawaii, Morgan (1971, 1972)
proposed that the alignment of volcanic islands and seamonts
provides a trace of a fixed volcanic plume on a moving tectonic plate. The instabilities at the origin of these plumes develop on boundary layers as deep as the coremantle
boundary. Morgan (1971, 1972) was able to show that these
hot spots were deeply rooted in the mantle, moving very
little against each other over time. By scrolling over the hot
spot, the alignment of volcanoes get progressively older and
therefore colder and denser, and volcanoes progressively sink
below sea level, away from the hot spot and the area of active
volcanism.
The hot spot concept is based on ideas discussed during the

past few decades. A broad definition of hot spot magmatism
can be taken from Ito and Van Kekens synthesis (2007) who
stated that:
The original work by Wilson (1963, 1973), Morgan (1971, 1972) and
Crough (1978) established the concept of Hot spot as a broad swelling
of topography capped by volcanism, which, combined with plate motion,
generates volcanoes aligned in a chain and with ages that progress
monotonically. In some cases, these chains project back to massive volcanic plateaus, or large igneous provinces (LIPs), suggesting that hot spot
activity began with some of the largest magmatic outbursts evident in
geologic record. (Morgan, 1972; Richards et al., 1989; Duncan and
Richards, 1991)
The term hot spot is ambiguous, suggesting that heat

alone is responsible for magma production by partial melting
in the upper mantle, instead of the decompression melting
of hot low-density mantle material rising from Earths
interior.
Many debates and controversies have occurred about these
narrow deep-Earth upwellings and their links with midplate
volcanism termed hot spot after Morgan (1972). The debate is
still active about the nature and the actual depth of the source
region of mantle plumes. Do they originate from Earths
coremantle boundary or from shallower regions within the
upper mantle? What are the relative motions of mantle
plumes over geologic time? What are the relationships between hot spots and mantle plumes, and between hot spot/
plume and plate tectonics? Are all hot spots associated with a
deep mantle plume? Are Andersons (2005) alternative hypotheses, with a plate tectonic origin for many hot spots,
valid?
However, the purpose of this article is not to discuss the
geophysical characteristics of hot spots. Very good syntheses
can be found in McNutt and Caress (2007) and Ito and Van
Keken (2007). But, hot spots are considered here from a
geomorphic perspective, as areas of unusual and concentrated
volcanic activity, at present or in the past, forming large volcanic regions not related to present plate boundary processes,
196
and without reference to the underlying origins for that

activity.
5.9.1.1
Types of Hot Spot Magmatic Structures
Whatever the origin of this exceptional volcanism (plume or

non-plume), hot spots are windows into the mantle. They
produce lava from which geochemists may attempt to characterize the different reservoirs deep within Earth. Their
specific situation with regard to plate tectonics offers geophysicists a reference system in which the absolute displacement of the plates can be determined. Their role in mantle
convection is also highlighted by the intriguing relationship
they have with continental breakup (Storey, 1995).
LIPs and hot spot volcanoes not only are widespread on
Earths surface, but they are also common on other planets,
and on some moons, in our solar system, with spectacular
examples on Mars. Hot spot volcanism is an important process
in the evolution of terrestrial planets, contributing to their
cooling and chemical differentiation. On Earth, the magmatism away from tectonic plate boundaries is commonly considered as responsible for approximately 10% of the planetary
heat flux from the deep interior to the surface. This value,
which may have been higher in the past, is probably underestimated, taking into account the thousands of small volcanic
seamounts that still remain to be discovered.
Several types of hot spot volcanism occur on Earth.
The most common forms are volcanic chains comprising
seamounts and large volcanic islands within midplate settings,
but other types of hot spot structures include aseismic ridges
(Laughton et al., 1970), superswells (McNutt and Fischer,
1987), and LIPs (Coffin and Eldholm, 1994).
Knowledge and understanding of these different forms/
groups varies widely. LIPs are regions that have experienced
very large eruptions of flood basalts. LIPs in continental areas
are subject to frequent work because of their scientific interest
and their relative accessibility. Aseismic ridges and submarine
plateaus are poorly explored, and research on them usually
focus on their geochemical or geophysical characteristics.
High-resolution bathymetric data, which would allow for
detailed geomorphic study, are rare, and the underwater
world remains very poorly explored. In fact, more is known
about hot spot forms on the surface of Mars than on Earth.
The Deccan traps, Siberian traps, and Columbia River
Basalt provinces in continental environments, and the Ontong
Java plateau and KerguelenBroken Ridge plateau in oceanic
environments are good examples of these large magma effusions (Figure 1). Iceland, in the present time, provides the best
analog of this type of volcanic form. Historically, only two
flood basalt eruptions have occurred; both were in Iceland.
The Laki fissure eruption in 178384 is the most recent. This
eruption took place along a 27-km-long fissure that produced
long basaltic flows (estimated at 15 km3) and large amounts
of SO2 (122 millions of tons) during 8 months, which had a
strong impact on global climate. The AD 934940 Eldgja
eruption took place along a 30-km-long fissure opened in the
same region as the Laki eruption. This eruption was larger
than the Laki eruption, with an estimated volume of lava of
more than 19 km3. Large-volume basaltic eruptions (41 km3)
have a recurrence period of several hundred years forming

shield volcanoes (Icelandic type) at relatively low discharge
rates, or fissure eruptions characterized by high discharge rates
(Thordarson and Hoskuldson, 2008). Although large scale,
the two historic Icelandic eruptions are by far less in terms
of discharge rate than the discharge rates of the flood
basalt eruptions of the Columbia River province or the
Deccan traps.
The extremely voluminous amounts of magma associated
with LIPs are emitted in a short time. LIPs occur on both
continents and oceans. In a broad definition they include
continental flood basalts, volcanic passive margins, oceanic
plateaus, submarine ridges, seamount chains, and ocean-basin
flood basalts (Coffin and Eldholm, 2005). Recently, a more
precise definition of LIP was proposed by Bryan and Ernst
(2008) who stated:
Large Igneous Provinces are magmatic provinces with areal extents
40.1 Mkm2, igneous volumes 40.1 Mkm3 and maximum life
spans of B50 Myr that have intraplate tectonic settings or geochemical affinities, and are characterized by igneous pulse(s) of
short duration (B15 Myr), during which a large proportion
(475%) of the total igneous volume has been emplaced.
It follows that seamounts, seamounts groups, submarine

ridges, and anomalous seafloor crust are not examples of LIPs,
as they are the results of long-lived magmatism with low
discharge rates.
Mafic (Fe- and Mg-rich) magmas are dominant in LIPs, but
acidic magmas can also be present, especially for those that
occur in continental settings. Individual lava flows are difficult
to characterize given exposure problems. Much of the information about the size, the nature, the duration, or frequency
of LIPs eruptions is derived from the study of continental
flood basalt provinces, in particular the Columbia River
(176 Myr), Afar (3114 Myr), and Deccan (6760 Myr) flood
basalts. Basaltic magmas are emplaced, from long fissures
systems or small shields, as gently dipping lava flows, with
lengths that can reach hundreds of kilometers, and thicknesses
of several hundred meters. Lava flows constituting the
Colombia River series are typical of rapid and voluminous
effusions, burying the pre-existing topography. These flood
basalts are dominantly pahoehoe lava flows, but some aa
flows are present (see Section 5.9.3.1). Thicknesses of individual lava flows vary from 10 m to more than 100 m. The
thicker lava flows can fill river canyons. Thicker lava flows are
common in other LIPs such as the Deccan traps where volumes of individual flow can exceed 8000 km3 (Bryan et al.,
2010). In the Ethiopian flood basalt province, lava flows may
locally emanate from shield volcanoes armed with dikes
whose orientation is controlled by regional extension tectonics (Mohr and Zanettin, 1988).
Silica-rich eruptions produce lava flows or pyroclastic-rich
eruptions such as ignimbrites (welded tuffs). Their size is
equivalent to the size of their basaltic counterpart, but they
mainly occur during the initial stage of LIP magmatic activity.
The largest eruptive silica-rich eruptions are from the ParanaEtendeka and Afro-Arabian provinces, where large-volume
eruptions formed massive ignimbrite units (see Bryan et al.,
2010 for a review).
197
50
50
150
100
50
50
100
150
Figure 1 Global map of Large Igneous Provinces (LIPs) in magenta, including continental and oceanic flood basalt provinces, volcanic passive
margins, oceanic plateaus, submarine ridges, and seamount groups. Agulhas Ridge (Agul), Alpha-Mendeleyev Ridge (Alph), Angola Plain (Ango),
Austral seamounts (Aust), Azores (Azor), Balleny Islands (Ball), Bermuda Rise (Berm), Broken Ridge (Brok), Canary Islands (Cana), Cape Verde
Rise (Cape), Caribbean flood basalts (Cari), Caroline seamounts (Caro), Ceara Rise (Cear), Chagos-Laccadive Ridge (Chag), Chukchi Plateau
(Chuk), Clipperton seamounts (Clip), Cocos Ridge (Coco), Columbia River Basalts (Colr), Comores Archipelago (Como), Conrad Rise (Conr),
Crozet Plateau (Croz), Cuvier Plateau (Cuvi), Deccan traps (Decc), Del Cano Rise (Delc), Discovery seamounts (Disc), Eauripik Rise (Eaur),
Eastern central North Atlantic (Ecna), East Mariana Basin (Emar), Etendeka (Eten), Ethiopian flood basalts (Ethi), Foundation seamounts (Foun),
Galapagos/Carnegie Ridge (Gala), Great Meteor-Atlantis seamounts (Grat), Guadelupe seamount chain (Guad), Hawaiian-Emperor seamounts
(Hawa), Hess Rise (Hess), Hikurangi Plateau (Hiku), Iceland/Greenland-Scotland Ridge (Icel), Islas Orcadas Rise (Isla), Juan Fernandez
Archipelago (Juan), Karoo (Karo), Kerguelen Plateau (Kerg), Line Islands (Line), Lord Howe Rise seamounts (Lord), Louisville Ridge (Loui),
Madagascar flood basalts (Mafb), Madagascar Ridge (Mari), Madeira Rise (Made), Magellan Rise (Magr), Magellan seamounts (Mags), Manihiki
Plateau (Mani), Marquesas Islands (Marq), Marshall Gilbert seamounts (Mars), Mascarene Plateau (Masc), Mathematicians seamounts (Math),
Maud Rise (Maud), Meteor Rise (Mete), Midcontinent rift volcanics (Keweenawan) (Midc), Mid-Pacific Mountains (Midp), Musicians seamounts
(Musi), Naturaliste Plateau (Natu), Natkusiak flood basalts (Natk), Nauru Basin (Naur), Nazca Ridge (Nazc), New England seamounts (Newe),
Ninetyeast Ridge (Nine), North Atlantic volcanic province (Navp), Northeast Georgia Rise (Nege), Northwest Georgia Rise (Nwge), Northwest
Hawaiian Ridge (Noha), Northwind Ridge (Nowi), Ontong Java Plateau (Onto), Osborn Knoll (Osbo), Parana flood volcanism (Para), Phoenix
seamounts (Phoe), Pigafetta Basin (Piga), Pratt-Welker seamounts (Prwe), Rajmahal traps (Rajm), Rio Grande Rise (Riog), Roo Rise (Roor), Sala
y Gomez Ridge (Sala), Shatsky Rise (Shat), Shona Ridge (shon), Siberian traps (Sibe), Sierra Leone Rise (Sier), Shom Abyssal Plain (Shom),
Tahiti (Tahi), Tasmantid seamounts (Tasm), Tokelau seamounts (Toke), Tuamotu Archipelago (Tuam), Tuvalu seamounts (Tuva), Vitoria-Trinidade
Ridge (Vitr), Wallaby Plateau (Wall), Walvis Ridge (Walv), and Yemen Plateau basalts (Yeme). Reproduced from Coffin, M.F., Eldholm, O., 1994.
Large igneous provinces: crustal structure, dimensions, and external consequences. Reviews of Geophysics 32, 136.
Little data exist to enable describing the morphology of

oceanic plateaus. The submarine Ontong Java plateau, in the
western Pacific, is by far the largest flood basalt province to
have occurred during the past 200 Myr. It rises approximately
2000 m above the surrounding sea floor and covers an area of
approximately 2 million km2. Fitton et al. (2004) described
deep-sea basaltic pillow lavas and massive lava flows fed by
fissure eruptions, and glass-rich volcaniclastic deposits. The
glass-rich volcaniclastic deposits have been interpreted as
primary hydromagmatic deposits from Surtseyan eruptions,
indicating, for at least one part of the plateau, eruptions in a
shallow-enough ocean to produce subaerial eruption columns
(Thordarson, 2004).
The environmental effects of LIP emplacements are known
to have been global, releasing huge quantities of volcanic gases
(CO2, SO2, Cl, and F) into the atmosphere, resulting in
significant climate change that likely lead to mass extinction

events (Wignall, 2005).
5.9.1.2
What is a Hot Spot Volcano?
The variety of contexts, histories, and intraplate volcanic

structures reflects the variety of magmatic processes. The
Morganian view of deep mantle plumes remains viable, at
least for a limited number of hot spots. As stated previously,
during the past decade, many publications have presented
quite different scenarios in order to explain the origin and
nature of hot spots. The debate within the scientific community focuses on the characteristics and origin of the
magma, source, localization, depth, and anomalies. What is
however not discussed is the obvious link between oceanic
chains, massive volcanic plateaus, and/or LIPs, and their
198
common mantle magmatic source. This section will draw

heavily on two publications that are considered as key references: Ito and Van Keken (2007) and Courtillot et al. (2003).
Ito and Van Keken (2007) suggested the existence of 69 hot
spots. Courtillot et al. (2003) and McNutt and Caress (2007),
based on the catalogs of the most cited hot spots (Sleep, 1990,
2003; Steinberger, 2000), presented a total of 49 hot spots,
which have been active in the past million years. From this
selection, Courtillot et al. (2003) have applied five criteria for
the recognition of hot spots: (i) the existence of a linear volcanic chain with monotonous age progression; (ii) the existence
of plateau/trap/flood at the tracks source; (iii) large magma
buoyancy flux and reliability; (iv) the existence of high ratios of
3
He/4He; and (v) the existence of slow shear velocity anomaly
at 500 km depth below the hot spot trace. Only seven hot spots,
one in Africa (Afar Ethiopia), one in the Indian Ocean (La
Reunion), two in the Atlantic (Iceland and Tristan da Cunha),
and three in Pacific area (Easter, Hawaii, and Louisville) are
typical of a deep plume origin. It is also interesting to note that
20 hot spots meet at least four of the five criteria (Figure 2).
Numerous structures are associated with the hot spot
magmatism. The magnitude of the volcanism (from traps to
oceanic islands), their locations (not only from continental
areas to marine environments, but also from the tropics to
high latitudes), and their ages, produce a wide variety of
landforms and situations. It is difficult to describe them all

comprehensively.
We therefore choose to focus on oceanic basaltic shield
volcanoes, which are iconic structures of hot spot volcanism.
These are currently major sites of magmatism on Earth. Extensively studied, they show a wide variety of morphologies
and dynamics. In addition, they have the extreme advantage of
commonly being in geodynamic contexts undisturbed by
crustal tectonic or geochemical interactions due to the absence
of continental crust. So this chapter will mostly describe
morphologies related to large oceanic shield volcanoes and
volcanic chains. Huff and Owen provide additional discussion
in this volume on the geomorphology of volcanoes in hot spot
and other settings (see Chapter 5.8).
5.9.2
5.9.2.1
Hot Spot Volcanic Chains

Volcanic Chain Analysis
Three main criteria are generally considered as crucial in the

analysis of volcanic chains:
1. The total age of the volcano chain;
2. Morphometric aspects;
50
50
150
100
50
50
100
150
Figure 2 Global map of hot spots. Red triangle marks are the 69 hot spot estimated localizations by Ito and Van Keken (2007). Black circles
are the 49 Hot spots as defined by Courtillot et al. (2003). Red circles are the seven primary hot spots that meet Courtillot et al. (2003) criteria.
Projection Long/Lat WGS84. The following list includes names of hot spots and their abbreviations: Afar (AF), Amsterdam-St. Paul (AMS),
Ascension-Circe (ASC), Austral (AU), Azores (AZO), Baja (BAJ), Balleny (BAL), Bermuda (BER), Bouvet (BOU), Bowie-Kodiak (BOW), Cameroon
(CAM), Canaries (CAN), Cape-Verde (CAP), Caroline (CAR), Comores (COM), Conrad (CON), Crozet (CRO), Darfur (DAR), Discovery (DIS), East
Africa/ Lake Victoria (EAF), East-Australia (AUS), Easter (EAS), Eifel (EIF), Fernando-Do-Norona (FER), Galapagos (GAL), Great-Meteor (GM),
Gough (GOU), Guadalupe (GUA), Hawaiian-Emperor (HAW), Hoggar (AHA), Iceland (ICE), Jan-Mayen (JM), Japanese-Wake (JWK), JuanFernandez (JFE), Juandefuca Cobb (COB), Kerguelen (KER), Lord-Howe (LHO), Louisville (LOU), Line Islands (LIN), MacDonald (Cook) (CK),
Madeira (MAD), Magellan Seamounts (MAG), Marion (MAR), Marquesas (MQS), Marshal Islands (MI), Meteor shona (MET), Mountains (MPM),
Musician (MUS), New England (NEW), Pitcairn (PIT), Raton Foundation (Rat), La Reunion (REU), Sainte-Helena (SHE), Samoa (SAM), San-Felix
(SF), Shatsky (SHA), Shona (SHO), Sierra Leone (SL), Society (SOC), Socorro (SCR), Tarava (TAR), Tasmantid (TAS), Tibesti (TIB), TrindadeMartin-Vaz (TRI), Tristan (TRN), Tuamotu (TUA), Vema (VEM), and Yellowstone (YEL). Reproduced from Courtillot, V., Davaille, A., Besse, J.,
Stock, J. (2003). Three distinct types of hotspots in the Earths mantle. Earth and Planetary Science Letters 205, 295308.
3. The subdivision of the plate during installation of

the chain.
Ito and Van Keken (2007) defined three classes: (i) Longlived age-progressive volcanism; (ii) short-lived age-progressive
volcanism; and (iii) no age-progressive volcanism.
5.9.2.1.1
Long-lived age-progressive volcanism
Ito and Van Keken (2007) attributed this name to volcanic

chains older than 50 Myr. By this criterion, they listed 13 hot
spots, whereas Courtillot et al. (2003) defined 11 volcano
chains.
Fernando, in the Atlantic Ocean, is probably the oldest hot
spot on Earth. It is linked with the Central Atlantic Magmatic
Province, dating back to 201 Myr, while North America, Africa,
and South America were still joined. Crozet/Prince Edward, in
the SW Indian Ocean, likely originated in the Karoo province
in South Africa, some 183 Myr ago. The Parana flood basalts
of Uruguay and Brazil and the Etendeka igneous province
of Namibia was emplaced at 130 Myr, and was the first expression of the Tristan da Cunha hot spot, and the precursors
of the opening of the South Atlantic ocean (Peate, 1997).
The gigantic Ontong-Java plateau formed at 122 Myr, and is
the origin of the linear Louisville hot spot track (Antretter
et al., 2004) and its partially destroyed volcanic chain on
the South Pacific due to the setup of the Kermadec trench.
By linking the origin of the Kerguelen hot spot to the
Rajmahal plateau, north of the Indian Ocean, Courtillot et al.
(2003) gave it an age of 118 Myr. The Easter hot spot age in
the South Pacific appears to be approximately 200 Myr,
whereas a little further north, the Galapagos hot spot probably
initiated its activity in the Caribbean plateau at 90 Myr. In
the southwestern part of Indian Ocean, the Marion hot
spot seems to find its origins in the heart of volcanoes of
Madagascar at 88 Myr. The numerous studies of La Reunion
hot spot show its relationship to the Deccan flood basalts in
India at 65 Myr. The Iceland hot spot, on the north Atlantic
mid-ocean ridge, has origins in both Northern Europe at
56 Myr and Greenland at 61 Myr. Macnab et al. (1995), White
(1997), and Jones et al. (2002) observed an abnormally thick
oceanic crust between Iceland and Greenland on the one side,
and Iceland and the Faeroe on the other. Comparison of this
crust thickness and the seafloor-spreading rate suggests that
the Iceland hot spot was associated with an excess of magma
near the Mid-Atlantic Ridge since the time of continental
breakup (Wilson, 1973; White, 1988, 1997) and that a continuous supply of magma caused the abnormal thickness of
the crust.
The origin of the Hawaiian-Emperor hot spot is not possible to determine. Indeed, the volcanic chain terminates
at the subduction zone of the Aleutian trench. The Meiji
seamount at 81 Myr is the oldest relief feature along its track
and any plateau associated with the origins of the hot spot of
Hawaii no longer exists.
Significant differences are noted for long-lived, age-progressive volcanism between the criteria of Ito and Van Keken
(2007) and those of Courtillot et al. (2003). Indeed, the
Courtillot et al. (2003) list of hot spots older than 50 Myr does
not include the Comoros, St. Helena, Trinidad-Martin Vaz,
Madeira, and Canary hot spots. Courtillot et al. (2003) did not
199
associate any of these formations with flood basalt provinces.

Furthermore, they considered them to be volcanic chains, as in
the case of St. Helena.
In contrast, Ito and Van Keken (2007) did not consider
including the Crozet/Prince Edward hot spot or the Fernando
hot spot into this group, for which they did not determine the
date of origin or their progressive age. For Easter Islands, these
authors dated the origin of the hot spot to 6 Myr, whereas
Courtillot et al. (2003) suggested 118 Myr. These differences
can be explained only in terms of the location of the origin of
the hot spot. Thus, Vezzoli and Acocella (2009) suggested an
age of 4 Myr, whereas Bonatti et al. (1977) proposed an age of
140 Myr.
In the Pacific Ocean, Courtillot et al. (2003) did not take
into account the Marquesas hot spot. Courtillot et al. (2003)
did not establish the starting age for this hot spot, whereas
Duncan and McDougall (1974) suggested 110 Myr for Hess
Rise and 145 Myr for Shatsky Ridge.
5.9.2.1.2
Short-lived age-progressive volcanism
Ito and Van Keken (2007), citing previous authors, listed eight
volcanic provinces based on the criteria for hot spot identification with a short-lived age-progressive volcanism. Society
(4.2 Myr), Marquesas (5.5 Myr), Pitcairn (11 Myr), Carolina
(13.9 Myr), Foundation (21 Myr), Tarava (43.5 Myr), Pukapuka (27.5 Myr), and Tasmania (24.3 Myr) chains all fall into
this category.
Here also, however, differences exist. Courtillot et al.
(2003) raised a number of questions about the location of
the hot spots origin and they did not date their time of
formation.
5.9.2.1.3
No age-progressive volcanism
In the list of 49 hot spots proposed by Courtillot et al. (2003),

11 are characteristic of this group of no age-progressive
volcanism (Ascension, Azores, Balleny, Bermuda, Bouvet,
Canary Islands, Cape Verde, Comores, Discovery, Jan-Mayen,
and Socorro). This is based on: (i) early origins; (ii) proximity
of absolute pole of the motion plate or, alternatively, installation along the mid-ocean ridge; and (iii) lack of modern
dating.
5.9.2.2
Volcanic Chain Analysis. The Hawaii-Emperor

Example
The Hawaii-Emperor is undoubtedly the most representative

of hot spot chains and is located in the middle of the North
Pacific Ocean, stretching from 18130 N, 500 km south of the
Tropic of Cancer, to a longitude of 1631N in the subduction
zone of the Aleutian trench, some 1500 km from the Arctic
Circle. The chain of islands, atolls, seamounts, and guyots
linked to the Hawaiian-Emperor hot spot extend 6000 km to
the northwest, with nearly 125 volcanoes. The Hawaiian
archipelago consists of only 19 islands with a total area above
sea level of 16 636 km2. The Hawaiian-Emperor can be divided in three units: (i) the Southeastern islands; (ii) the
Northwestern Islands; and (iii) the Emperor chain.
200
5.9.2.2.1
Young stage of the Hawaiian chain: The

Southeastern Islands, from Hawaii to Kaula
5.9.2.2.2
The biggest emerged islands, also called High Islands, compose this 600-km-long archipelago (Figure 3). Hawaii, or Big
Island, is the largest of all islands composing the archipelago
having dimensions of 130 120 km. The two highest peaks,
Mauna Kea and Mauna Loa, rise to 4207 and 4170 m asl,
respectively. The ocean floor descends to 5600 m (East of
Big Island) so that Mauna Kea has a relative relief of 9800 m,
the highest mountain in the world. The exposed surface of the
island is approximately 36 500 km2. The Big Islands volume
exceeds 102 000 km3, whereas the emerged part is 1200 km3.
The second largest island is Maui with an area of approximately 1900 km2, rising to a maximum elevation of 3055 m
asl at Haleakala peak.
Generally speaking, the Southern Islands, including the
submarine counterparts, can be subdivided into three distinct
units:
1. Hawaii, which has two dominant directions of construction (N45501 and N1501601).
2. Maui, Kaoolawe, Lanai, Molokai, and Oahu setup in a
N1101 axis, but appear to be plurigenetic in detail with
Oahu separated from the other islands.
3. Kauai, Niihau, and Kaula are clearly aligned along an axis
of N601.
The Loihi seamount (B3000 m high with its top at
B1000 m below sea level), an active volcano located 35 km
south of Hawaii, can be added to the southeastern Islands.
Garcia et al. (2006) showed that a volcanic eruption occurred
in 1996. Loihi will probably be attached to the Big Island
when it extends above sea level, similar to the other volcanoes
that make up Hawaii (Mauna Loa, Kilauea, Mauna Kea,
Hualalai, and Kohala).
The nine islands of the Southeastern Islands of the Hawaiian chain cover about 142 103 km2 on the Pacific Ocean
seafloor and have a total volume of B335 103 km3.
Oldest stage of the Hawaiian chain: The

Northwestern Islands, from Nihoa to the Kure
atoll
This part of the chain consists of the 1730 km trace of the

Northwestern Island from Nihoa Island to the Kure Atoll
(Figure 4). Kure Atoll is generally considered as the northern
limit of the Hawaiian chain, mainly because it is the last island, but strictly speaking the chain extends to the Colahan
Seamount (Figure 5). The major geomorphic difference with
the Southwestern islands is that they are much older and more
eroded. Emerged parts of the volcanic edifices are significantly
smaller, and a large number of seamounts are visible.
The entire chain of Hawaii represents a volume total of
750 103 km3, with the Northwestern islands having a volume
of 410 103 km3.
5.9.2.2.3
Emperor Chain: From Colahan to Meiji

seamounts
The Emperor chain is a part of the extensive Hawaiian-Emperor chain exclusively composed of submarine volcanic
peaks, and stretching almost 2900 km across the Northwestern part of the Pacific Ocean (Figure 5). The HawaiianEmperor chain forms an L-shaped chain of volcanic islands
and seamounts; the Emperor part is located to the north of a
601 rotation bend in the chain. On a small scale, three preferential alignments can be recognized in the Emperor chain.
Between Colahan and Kammu seamounts, there are alignments previously observed in the terminal part of the
Hawaiian chain. For approximately 370 km, the main trend
of the volcanic peaks is N1101. Then, between the Kammu
and Detroit seamounts (after plate rotation marked by inherited faults), at almost 2150 km along the chain, the major
orientation becomes N1701. Finally, between the Detroit
Seamount and the KurilKamchatka Trench (where volcanic
edifices are subducted), for approximately 400 km, the trend
is N3101. Overall, the Emperor chain does not show any
Kauai
22 N
Kaula
Niihau
Molokai
Oahu
Maui
Lanai
Kahoolawe
Hawaii
20 N
Loihi seamount
162 W
160 W
158 W
156 W
154 W
18 N
152 W
Figure 3 Location map showing the Southeastern Hawaiian Islands shown on bathymetric DEM. This figure has been generated with
GeoMapApp (http://www.geomapapp.org) and the Global Multi-Resolution Topography (GMRT) Synthesis. Reproduced from Ryan, W.B.F.,
Carbotte, S.M., Coplan, J.O., et al., 2009. Global Multi-Resolution Topography synthesis. Geochemistry, Geophysics, Geosystems 10, Q03014.
doi:10.1029/2008GC002332. http://www.geomapapp.org. Projection Lat/Long WGS84.
Kure Atoll
30 N
Midway
Atoll Pearl
and Hermes
Atoll
Lisiansky
201
Laysan
Gardener
pinnacles
26 N
French
fregate
Maro
reef
Necker
Nihoa
Kauai
22 N
178 E
178 W
174 W
170 W
166 W
162 W
Figure 4 Location map showing the Northwestern Hawaiian Islands shown on bathymetric map. This figure has been generated with
GeoMapApp (http://www.geomapapp.org) and the Global Multi-Resolution Topography (GMRT) Synthesis. Reproduced from Ryan, W.B.F.,
doi:10.1029/2008GC002332. http://www.geomapapp.org. Projection Lat/Long WGS84.
real discontinuity along its length except in the JinguOjin

seamounts area.
5.9.2.3
Hot Spot Chains and Plate Tectonics
The morphological aspects of hot spot chains are characterized

by an alignment of volcanic edifices stretching over several
thousands of kilometers. Where the hot spot is currently active, the juxtaposition of volcanoes results in the growth of
oceanic islands. Further away from the active center, islands
are more eroded, and then with distance progressively become
atolls, seamounts, and guyots.
Another feature of hot spot chains is that their trend can
change along their length. In the case of Hawaii-Emperor
some edifices in the chain seem to be a part of a set, juxtaposed on the same basis, whereas some are clearly separated
from the others by sedimentary beds. These patterns can
be explained by changes in plate motion direction and speed,
together with the hot spot magma production rate, which may
vary with time.
Niu (2004) pointed out that the change in direction along
the Hawaiian-Emperor track occurred at approximately
43 Myr. It is notable that in the Pacific, the Louisville and
Macdonald chains also show a change in their trajectory at
approximately 43 Myr. A fourth case (Tuamoto) is noted by
Niu (2004) in this sector where the distinction between different hot spots is difficult to recognize.
Dalrymple and Clague (1976), Patriat and Achache
(1984), and Niu (2004) suggested that this sudden change in
Pacific plate motion by approximately 601 could have been
caused by the collision between India and Eurasia plates at
approximately 45 Myr. Lithgow-Bertelloni and Richards
(1998) argued that the IndiaAsia collision had no effect on
changing the direction of Pacific plate motion.
Plate boundaries have a major role along the track of the

hot spot chains. For the Hawaii and Louisville Island chains,
subduction zones are destroying the oldest volcanic edifices.
However, this is not the only effect of plate boundary presence. Easter Island is located in the South Pacific by 271 09 S
and 1091 22 W. This small island, with a triangle shape
(162 km2), is the latest sign of the Easter hot spot manifestation. The ages of the subaerial part of the Easter Island are
included in a range between 0.78 and 0.2 Myr (Vezzoli and
Acocella, 2009).
Limits of extensions of the chain linked to the Easter hot
spot are not well defined. Referring to its uncertain location,
Clouard and Bonneville (2001) stated that the Easter hot spot
is one of the most puzzling hot spots of the Pacific plate. As shown
in Figure 6, the Easter Island hot spot is complicated by: (i) its
uncertain location (Clouard and Bonneville, 2001); (ii) the
presence of several seamounts in the same zone; (iii) the
presence of different plumes supplying these seamounts
(Easter, Foundations, McDonald, Society, Marquesas, Pitcairn,
and Juan Fernandez) (Bonatti et al., 1977); (iv) Easter chain
setup on both sides of the Pacific Rise; and (v) the fracturing of
tectonic plates in microplates (Easter and Juan Fernandez
microplates) and the presence of several faults.
In the literature, authors have not always presented the
chain in its entirety. Hagen et al. (1990), Rappaport et al.
(1997), Steinberger (2002) and Watts et al. (2006), for example, only described it on the Nazca plate without considering the existence of seamounts linked to the Easter hot
spot on the Pacific plate.
In contrast, Morgan (1972), Bonatti et al. (1977), Epp
(1984), OConnor et al. (1995), Ito et al. (1995), and Clouard
and Bonneville (2001) explained, using chemical, geomorphic, and geological arguments, that the Easter Chain extends on the western side of the East Pacific Rise perhaps to the
Tuamotu Plateau. For Clouard and Bonneville (2005), the
202
54 N
KurilKamchatka Trench
Meiji
Detroit
Winnebago
Minnetonka
48 N
Jinmu
Saga
Showa
Yomei
Godaigo
Ojin
42 N
Nintoku
Jingu
Ojin
36 N
Koko
Toba
Gosanjo
Kimmei
Genji
Yuryaku
Daikakuji
Kammu
162 E
168 E
Abbott
Colahan
De Veuster
174 E
30 N
180 E
Figure 5 Location map showing seamounts and guyots of the

Emperor chain shown on bathymetric map. This figure has been
generated with GeoMapApp (http://www.geomapapp.org) and the
Global Multi-Resolution Topography (GMRT) Synthesis. Reproduced
from Ryan, W.B.F., Carbotte, S.M., Coplan, J.O., et al., 2009. Global
Multi-Resolution Topography synthesis. Geochemistry, Geophysics,
Geosystems 10, Q03014. doi:10.1029/2008GC002332. http://
www.geomapapp.org. Projection Lat/Long WGS84.
Easter volcanic chain extends toward the west, on the Pacific

Plate, on Crough seamount. According to Clouard and Bonneville (2005), the Easter chain stretches approximately
5000 km (1000 km to the west on the Pacific Plate and
4000 km to the east on the Nazca plate) and terminates in the
subduction of the Nazca plate beneath the South American
plate, where the oldest samples are dated at 39 Myr (Figure 6).
The effect of plate boundaries on the hot spot chain
morphology is different for Tristan Da Cunha. The Tristan da
Cunha volcanic chain is localized in the South of Atlantic. This
hot spot had built two distinct volcanic chains on two different plates and stretches for over 3000 km in an eastern
direction (Walvis Ridge) on the Africa plate, from Tristan
Island onto the Namibia mainland (Figure 7). Thus ages vary
from 0.01 to 129 Myr. On the South American plate, the hot
spot chain stretched westerly for 2200 km toward Brazil by the
Rio Grande Rise. The distribution of seamounts on either side
of the Mid-Atlantic Ridge is asymmetric. Indeed, the alignment
of seamounts is perfectly clear on the Walvis Ridge, whereas a
large gap exists to the west. The first seamount, with an age of
72 Myr, is observed to the west to over 1500 km from the MidAtlantic Ridge; the oldest, on the South American continent,
being approximately 129 Myr.
Good agreement exists for the oldest ages on both the
African and South American plates, and the Tristan Da Cunha
hot spot generation that is associated with ParanaEtendeka
continental flood basalt (Gibson et al., 2005) enclosed in the
Gondwana super continent. The initiation of the Mid-Atlantic
Ridge and the growth of the Atlantic Ocean are responsible for
symmetrical construction of the two volcanic chains on the
two separate plates (Oconnor and Duncan, 1990; Oconnor
and Roex, 1992; Gibson et al., 2005). Burke and Wilson
(1976) proposed that there is a significant slowdown in the
drift of Africa at approximately 30 Myr, although spreading
remained at constant velocity to generate a shift of the MidAtlantic Ridge westward contemporary to an acceleration of
the American plate motion. The result was a relative shift of
the Tristan hot spot toward the East, enclosed in the African
plate where it continued the construction of the volcanic
chain. Oconnor and Duncan (1990) gave a different explanation. They argued for movement to the west of the Mid-Atlantic Ridge between 70 and 50 Myr. Consequently, a 430 km
westward shift from its current position occurred. These ages
are consistent with the dates shown in Figure 7. This may also
explain the inconsistency of ages on the Walvis Ridge between
01 and 31 E.
Volcanic chains associated with La Reunion and Kerguelen
hot spots are different from the previous case. Here, these two
channels are divided into two parts due to the presence of the
Mid-Indian Ocean Ridges (Figure 8). La Reunion volcanic
chain begins between Pakistan (Mahoney et al., 2002) and the
Deccan Traps in India (Courtillot et al., 2003) and stretches to
approximately 101 S, building the ChagosLaccadive Ridge.
The youngest ages are 40 Myr (Duncan, 1990; Duncan et al.,
1989). Below the Central Indian Ocean Ridge, the trace of this
hot spot chain is still visible on the African Plate (more precisely on the Somalian plate) with the Mascarene Plateau
formation extended to La Reunion Island. The Mascarene
plateau and the ChagosLaccadive Ridge existed together as a
single ridge until the Eocene and later separated as a consequence of spreading along the Central Indian Ridge
(McKenzie and Sclater, 1971; Sclater et al., 1981).
For the Kerguelen hot spot chain, a first part of the Ninetyeast Ridge, stretches above the SE Indian Ocean Ridge on the
Indian Plate, whereas a second part, the Kerguelen Plateau,
stretches to the south on the Antarctica Plate. The initial stage
of the Kerguelen hot spot formed the Rajmahal Traps, whereas
the Southern Kerguelen Plateau formed at approximately 120
to approximately 110 Myr (Ito and Van Keken, 2007). From
100 Myr, seafloor spreading separated India, Antarctica, and
Australia (Sclater and Fisher, 1974). The central Kerguelen
Plateau and Broken Ridge began their forming as one entity at
95 Myr. The northward migration of the Indian plate allowed
203
10 S
16 S
22 S
28 S
34 S
40 S
118 W
108 W
98 W
88 W
78 W
Figure 6 Bathymetric map of the Easter chain. Mean ages of the seamounts are in Myr. This figure has been generated with GeoMapApp
(http://www.geomapapp.org) and the Global Multi-Resolution Topography (GMRT) Synthesis. Reproduced from Ryan, W.B.F., Carbotte, S.M.,
Coplan, J.O., et al., 2009. Global Multi-Resolution Topography synthesis. Geochemistry, Geophysics, Geosystems 10, Q03014. doi:10.1029/
2008GC002332. http://www.geomapapp.org. Projection Lat/Long WGS84.
10 S
South America
Africa
is
alv
72
122
128.2
Rio
65
38.7 38.5
Gr
an
127
de
115
69.8
40
85
30 S
129
72
129
123
128
rid
123
122
127
125
108
113
ge
127
20 S
46
38 37.2
Ris
0.01
64
61.5
50.3
31
Tristan da
Cunha
40 S
50 W
40 W
30 W
20 W
10 W
10 E
Figure 7 Bathymetric and topographic map of the Tristan da Cunha chain. Mean ages of the seamounts and lands are in Myr. This figure has
been generated with GeoMapApp (http://www.geomapapp.org) and the Global Multi-Resolution Topography (GMRT) Synthesis. Reproduced from
Ryan, W.B.F., Carbotte, S.M., Coplan, J.O., et al., 2009. Global Multi-Resolution Topography synthesis. Geochemistry, Geophysics, Geosystems
10, Q03014. doi:10.1029/2008GC002332. http://www.geomapapp.org. Projection Lat/Long WGS84.
the formation of the Ninetyeast Ridge along a transform

fault, between 38 and 85 Myr (Frey and Weis, 1995, 1996).
The Kerguelen and Ninetyeast ridgeBroken Ridge compartments separated at approximately 4045 Myr with the SE Indian Ridge propagation (Royer and Colin, 1992; Frey and
Weis, 1995, 1996).
As discussed above in this section, tectonic plate boundaries

play a main role in the shape of the volcanic hot spot chains:
leading to either continuity versus discontinuity, leading to
spreading perpendicular to the stretch axis, offsets and/or
bends in chain. However, the importance of plate motion, and
particularly the variation, over time must be considered.
204
20 N
20 S
40 S
60 S
30 E
50 E
70 E
90 E
110 E
Figure 8 Bathymetric map of the La Reunion (R) and Kerguelen (K) chains. Mean ages of the seamounts are in millions of years (Ma). This
figure has been generated with GeoMapApp (http://www.geomapapp.org) and the Global Multi-Resolution Topography (GMRT) Synthesis.
Reproduced from Ryan, W.B.F., Carbotte, S.M., Coplan, J.O., et al., 2009. Global Multi-Resolution Topography synthesis. Geochemistry,
Geophysics, Geosystems 10, Q03014. doi:10.1029/2008GC002332. http://www.geomapapp.org. Projection Lat/Long WGS84.
Figure 9 illustrates through several graphics the comparison of

distance versus ages for several different settings. This corresponds to a plate motion average analysis. Also plotted are
the current plate motions using DORIS Instrumentation
(Determination dorbite et de radiopositionnement integres
par satellite) (Cretaux et al., 1998) to compare instantaneous
and long period results.
This analysis is simplistic because it does not take into
consideration plate motion variations, changes in the rotation
poles, and hot spot shifting over geologic times. Even if the
current speed measured does not differ much from the average
speed, they are still different. In the case of Easter, La Reunion,
and Tristan Da Cunha, for example, the calculated speeds are
greater than these measured currently with DORIS. This is the
reverse for Iceland on the Eurasian Plate, and Louisville and
the chain of Ninetyeast. In contrast, the Hawaii-Emperor example values are fairly consistent.
5.9.3
5.9.3.1
Hot Spot Volcanoes

Edifices General Morphology
Hot spot volcanism is typically associated with basaltic shield

volcanoes or lava-shield volcanoes. They are complex structures principally built by fluid lava flows accumulation. Basaltic shield volcanoes are polygenetic structures with a broad
shield shape, several thousand meters high, and commonly
exceeding 1000 km3 in volume (Walker, 1993). The archetype
of basaltic shield volcanoes is Mauna Loa in Hawaii. Mauna
Loa is the largest volcano on Earth with an estimated volume
of 40 000 km3, rising to 910 km above the surrounding
ocean floor and more than 4 km above sea level. Its subaerial
slopes are mostly between 21 and 81, excluding scars from old
landslides. The submarine slopes are steeper than those above
sea level when they are directly the result of lava effusions, as
205
140
120
Age (Ma)
100
80
60
LA
WA
- 7.
8
A-
y = 0.0065x <=> 15.4 cm yr1

R2 = 0.95911
>
yr 1
120
.8
c
100
.1
TA
60
y = 0.0668x
40
cm
yr
TA
y
0 cm
> N
- 7.
y = 0.0116x - 21 639
<=> 8.8 cm yr1
R2 = 0.947
Hawaii-Emperor
<=> 11.8 cm yr1
NE
y = 0.0092x + 39 187
-3
R2 = 0.92823
<=> 7.0 cm yr1
OT
<=> 1.5 cm yr1
>
KO
R2 = 0.94132
Easter
0
140
KA
y = 0.0085x
y = 0.0097x <=> 10.3 cm yr1

R2 = 0.90164
20
Age (Ma)
y = 0.0166x - 15 699
R2 = 0.96339
<=> 7.6 cm yr1
1
EAS
40
80
NW
1 >
r
E
1 >
yr
m
.1 c
y
7 cm
cm
A - 7.6
-> NN
yr
YAR
y = 0.0532x <=> 1.9 cm yr1

R2 = 0.88381
20
Iceland
Ninetyeast
0
140
NN
LA
WA
cm
cm
.4
yr
re*
cm
yr
NN
lo
ga
y = 0.0059x + 33 565
40
R2 = 0.99561
y = 0.0298x
20
Louisville
NN
0
140
yr 1
>
120
y = 0.0363x
R2 = 0.91034
<=> 2.7 cm yr1
m
0c
>
NE
R2 = 0.97032
1000
2000
<=> 13.6 cm yr1
<=> 3.3 cm yr1
Runion
3000
4000
5000
6000
Distance from the most recent activity center (km)
-4
y = 0.0217x
R2 = 0.63792
<=> 4.6 cm yr1
-1
.9
80
L
HE
cm
100
CB
60
CA
Age (Ma)
5.0
->
Ba
UA
60
- 7.7
NW
>
RE
Age (Ma)
80
-2
y = 0.0169x
R2 = 0.95265
<=> 5.9 cm yr1
100
yr 1
->
120
40
20
Tristan Da Cunha
0
0
1000
2000
3000
4000
5000
6000
Distance from the most recent activity center (km)

Figure 9 Age vs. distance along the Line Islands. For Easter, Iceland, and Tristan Da Cunha red squares represent the eastern direction and
blue diamonds the western direction. For La Reunion red squares represent the part of the chain on the Indian Plate and Blue diamonds on the
Somalian Plate. For Hawaii, Blue Diamonds represent the South-Eastern Hawaiian Islands, red squares the North-Western Hawaiian Islands, and
green triangles the Emperor chain. Dotted lines represent current plate motions measure by DORIS tool (after Cretaux et al., 1998). Below or
above-dotted line read Station idplate motiondirection. CACB (Cachoeira - Brazil), EASA (Ester Island), HELA (Sainte Helene Island), KOKA
(Kokee Park - Kauai Island), META (Metsahovi - Finlande), OTTA (Ottawa - Canada), REUA (La Reunion Island), WALA (Wallis Island), and YARA
(Yarragadee - Australia) are the DORIS station used. For La Reunion, the motion of the Indian plate is obtained using the Bangalore GPS station.
Reproduced from Bilham, R., Larson, K., Freymueller, J., 1997. GPS measurements of present-day convergence across the Nepal Himalaya.
Nature 386, 6164, and Cretaux, J.-F., Soudarin, L., Cazenave, A., Bouille, F. 1998. Present-day tectonic plate motions and crustal deformations
from the DORIS space system. Journal of Geophysical Research 103, 30,16730,181.
206
such as for Loihi seamount (101 to 151 Malahoff, 1987) or

the Puna ridge in Hawaii. But the slopes are gentler when the
submarine slopes are the result of huge landslides or are
mainly built by debris avalanche deposits and turbidite fans,
like those of the submarine slopes of La Reunion Island (11 to
81 Saint-Ange et al., 2011).
Much less numerous than the andesitic stratovolcanoes,
shield volcanoes are located mostly in oceanic settings. The
term shield volcano was originally used to name the volcanoes
of Iceland that have a flat and rounded profile. Shield volcanoes are built almost entirely from the accumulation of fluid
alkaline or tholeiitic basaltic lava, and the injection of intrusive rocks of the same composition. More silica-rich lavas,
with compositions ranging from hawaiite to trachyte or
phonolite, represent, when present, no more than 2% of the
total volume of the edifice. Pyroclastic-rich units generally
appear with strong magma differentiation just before a volcano becomes inactive. The percentage of pyroclastic material
is estimated at less than 1% for the emerged part of the volcanic edifice, reflecting the preponderance of effusive dynamisms in the construction of these emerged parts.
Two main types of lava flows surfaces are generally distinguished for basaltic lavas, aa and pahoehoe, although
intermediary situations may exist (slab lava), and all types can
be present in the same lava field. Aa lava flows have extremely
irregular surfaces, generally covered by decimetric spiny fragmental debris (clinkers) of frozen crust (Figures 10(a) and
10(b)). Pahoehoe (Figures 10(c) and 10(d)) lava flows have
smooth and continuous surfaces. Both types commonly coexist on basaltic shield volcanoes, in some places within a
single lava flow, even if pahoehoe lava flows are more frequent
in some lava shields such as for Kilauea, whereas aa lava flows

are more common on others such as for the Galapagos volcanoes or Karthala volcano in Grande Comore. A third type of
lava flow, blocky lava, is characterized by an irregular and
fractured surface, and covered by decametric blocks up to
meters across. On the shield volcanoes, this type is only found
during the late stages of volcanism as lava is differentiated to
become more silica-rich in composition.
Three types of morphologies are classically described for
shield volcanoes. It is generally difficult to make a sharp distinction between the different types of shield volcanoes because of the diversity of lava shields. For example, shield
volcanoes such as La Reunion volcanoes, regarded as Hawaiian-type, have many differences from the Mauna Loa and
Kilauea volcanoes.
Icelandic volcanic shields are small, low-angle, nearly circular shields, with a morphology little affected by erosion.
Icelandic shields are built by thin basaltic pahoehoe lava flows
fed by short eruptive fissures. Most of them occur during a
single eruption.
Volcanic shields in Galapagos are characterized by their
upwardly convex shape, resembling a tortoise shell, with
slopes exceeding 201 at midslope, and a clear flattening near
the summit. Another characteristic component of these volcanoes is the development of concentric eruptive fissures,
together with radial fissures, near a summit caldera, which is
commonly very deep. Unlike the Hawaiian shield volcanoes,
no well-defined rift zone occurs within the Galapagos shields.
In addition, gravitational detachments are less likely to occur
during the growth of these volcanoes. The amount of pyroclastic material is low, except for the last stage of activity
2m
(a)
1m
(c)
10 m
(b)
6m
(d)
Figure 10 Lava flow pictures obtained on Piton de la Fournaise volcano. (a) and (b): Aa lava; (c) and (d) Pahoehoe lava. Photographs by
Nicolas Villeneuve.
during which many cinder cones may form on the flanks and
top of the volcano.
The Hawaiian shield volcanoes are the largest volcanic
edifices. Their size distinguishes them from other shield volcanoes. The underlying lithosphere is isostically depressed,
causing downward flexure and development of a circumferential moat in which sediments accumulate. Hawaiian
shield volcanoes are built by successions of thin fluid lava
flows and are characterized by low subaerial slopes (usually
less than 151). Structures including calderas, rift zones, and
flank landslides are described in more detail in Section 5.9.4.
Several structures contribute to shaping the general
morphology of shield volcanoes. During the major phase of
building, shield volcanoes have frequent eruptions, emitting
basaltic lavas mostly from fissure vents. These eruptive fissures
opened at the summit area and on the flanks of the volcano,
and radiate from the summit. Numerous eruptive fissures are
concentrated along preferential zones of fissuring, called rift
zones. Rift zones are identified by lines of vents, fissures, and
diverging flows, and can be several tens of kilometers long for
only few kilometers wide, like those on Mauna Loa in Hawaii
or Karthala volcano in Comoros. At depth, swarms of near
vertical dikes underlie the rift zones. Dikes form as magma
advances laterally from a shallow magma chamber located
below the center to the volcano flanks. Larger intrusive bodies
may also exist within the rift zones, but this feature
is only associated with young Hawaiian shield volcanoes,
in particular Kilauea. The rift zones are true backbones for
the volcano, draining most of the magma that reaches the
surface.
The top of the volcano, where rift zones converge, commonly has a caldera. This caldera is generally formed by the
coalescence of multiple zones of collapse. Collapse craters or
calderas have very steep walls. Several subcircular and small
(less than 1 km in diameter) zones of collapse, called pit
craters, may exist within the caldera itself or along the rift
zones. Calderas and pit craters form in connection with the
drainage of magma from underlying shallow magma chambers or intrusive bodies.
Many of the largest landslides known on Earth occur on
volcanoes, in particular shield volcanoes. If they are initiated
into the emerged part of the volcanic edifices, they extend
dramatically in the submarine domain. Flank landslides on
shield volcanoes are largely controlled by the orientation of
rift zones and the gravitational stress field caused by the
weight of the volcano and by the buttressing effect when one
shield volcano grows against the flank of another (Fiske and
Jackson, 1972).
A lateral displacement of the free flank of the volcano
accommodates dike intrusions. This mechanism appears from
the early stages of volcano formation and is active as long as
the rift zones remain the main axes for the advance of magma
to the surface. On many volcanoes in the archipelago of
Hawaii, landslides extend offshore up to several tens of kilometers and have involved hundreds of km3 of debris avalanche and debris flow material (Moore et al., 1989). The
general term mass wasting is commonly used to designate the
sudden collapse and sliding away of the flank of a volcano. It
may refer to block landslide when part of the volcano advances semicoherently downhill above an underlying weaker
207
level. This corresponds to a rotational slide (slump) that tends

to move slowly. Or, it may refer to avalanche when the detached mass moves rapidly and breaks into smaller debris as it
moves downslope. During a long period of their growth,
shield volcanoes are affected by such gravity destabilizations
as they compensate for the overloading and oversteepening
of the slopes of the volcanic edifice. In a way, the products of
destabilization significantly contribute to the construction
of volcanic edifices, especially with regard to the submarine
flanks.
5.9.3.2
Construction versus Destruction
The morphological characteristics of a volcanic edifice strongly

depend on the rheology of the magma. For oceanic shield
volcanoes, basic (mafic) magmas volumetrically dominate.
This is due to the supply from mantle-derived magmas, the
lack of acidic crust that could change the composition of
primary magmas, and the mode of evolution of the magma by
fractional crystallization, which favors the basic compositions
rather than intermediate or acidic (felsic) compositions.
Consequently, little mention is made of the magma properties, because their values tend to be rather uniform among
different basaltic (basic) volcanoes. Viscosity and yield
strength determine the geometry and structures of lava flows
or domes and intrusions. Gas content, viscosity, and rheology
determine the explosivity of eruption and the type of volcanic
deposits. Toward the end of a volcanos life, eruptions
become less frequent, and usually more explosive, which
corresponds to a change in composition of the lavas from
alkali basalts to hawaiites, mugearites, and minor amounts
of trachyte. This may change the overall shape of a volcano
when a long period of activity with differentiated acidic
magmas leads to the emplacement of lava domes or pyroclastic flows. This is partly the case for the volcanism of the
Piton des Neiges in La Reunion where the construction of a
late stratovolcano is distinguished from the basaltic shield
(Stieltjes, 1985).
The evolution of basaltic shield volcanoes extend over
millions to tens of millions years. Phases of growth and destruction often alternate. Terrestrial and submarine erosion, as
well as flank landsliding, are as important as the volcanic activity in the building and shaping of the volcano morphology.
The recurrence of flank landslides has been recognized as one
of the most important dismantling processes on shield volcanoes since the discovery of huge deposits related to flank
collapse on Hawaii (Moore, 1964; Moore et al., 1989). Such
landslide deposits have been identified for other volcanic islands such as in the Canary Island (Carracedo, 1999; Masson
et al., 2002), La Reunion Island (Lenat et al., 1989; Oehler
et al., 2007; Le Friant et al., 2011), or, in another context,
Stromboli (Kokelaar and Romagnoli, 1995; Romagnoli et al.,
2009a) and the Lesser Antilles (Deplus et al., 2001).
Erosion is another major shaping process of shield volcanoes, particularly in the tropics. Terrestrial sedimentary
processes are highly active, favored by the wet tropical climate
and amplified by seasonal cyclonic conditions. Hurricanes
produce huge rains and torrential floods, causing efficient and
intense land erosion, and very high-concentrated sediment
208

5.9.3.3
transport (Bret et al., 2003; Garcin et al., 2005). Turbidity

currents are an important sediment transport process in the
oceanic realm, shaping the seafloor around volcanic islands
(Saint-Ange et al., 2011; Sisavath et al., 2011). Volcanic islands
are well known for their huge volcaniclastic production linked
to giant landslides, effusive or pyroclastic eruptions, and high
erosion rate (Louvat and Allegre, 1997; Masson et al., 2002;
Moore et al., 1994; Oehler et al., 2007; Rad et al., 2007;
Schmincke and Sumita, 1998; Schneider et al., 1998; Trofimovs et al., 2006). Recent studies underlined the particularities of erosion of basaltic terrains. Incision rates are very high,
similar to those for active orogenic areas, with rates ranging
from B0.5 to 3.4 m/kyr for La Reunion island (Louvat and
Allegre, 1997).
As mentioned previously, Hawaii volcanoes are considered
as the archetype of shield volcanoes, but other shield volcanoes may deviate from the general Hawaiian model. This
is illustrated with simple comparison of the morphology
of the Big Island of Hawaii, consisting of six volcanic
edifices (including Loihi seamount), and La Reunion, made
up of three volcanic edifices. A clear difference is apparent
with the relative extension of their submarine components
(Figure 11). The submarine slopes of Hawaii are widely constructed by effusions of lava, mainly along the rift zones. The
products of destabilization are present, forming well-defined
structures. In La Reunion, however, products of destabilization
largely dominate. Few portions are directly built by the volcanic activity. This results in a broad spreading of volcaniclastic products, and the submarine base of the volcanic
edifice is much larger than the size of the island above
sea level.
Model of Building
Based on numerous investigations on the volcanoes of the

Hawaiian chain, Peterson and Moore (1987) proposed a
multistage evolution for Hawaiian-type volcanoes that can be
divided into five phases (see Peterson and Moore, 1987).
5.9.3.3.1
Submarine preshield stage
This early stage of shield volcano formation is, of course, the

least studied. It is associated with low magma production
comprising alkaline lavas. The first magmatic emissions in the
deep ocean, due to the significant contrast in density between
oceanic water-saturated sediments and basalt rising to the
surface, takes place under a thick layer of sediments and epiclastic debris before flowing on the sea floor and building a
new volcanic edifice (Williams and McBirney, 1979; Staudigel
and Schmincke, 1984). With the growth of the volcanic edifice, a magma reservoir develops to rapidly form a caldera.
When the shield volcano develops on the sea floor, isolated
from other pre-existing edifice, dikes feeding the flank fissure
eruptions have no preferred orientation imposed by a neighboring edifice, and they radiate in all directions from the
magma chamber. Rift zones appear very early in the building
of oceanic volcanoes (Binard et al., 1991), under the influence
of gravitational stress or according to older structures printed
in the underlying lithosphere. Two or three rifts may develop,
which influence the formation of the future volcano. Rapidly
the magma production rate increases and the magma changes
from alkalic to tholeiitic. The lavas erupted are basalts and
olivine-rich basalts.
15
10
2300
CS
2140
2250
CS
1
2200
CS2
CS1
2200
CS3
2
CS
7.5
1
CS
12.5
CS
2260
5
2150
2.5
100 km
17300
17400
Elevation in meters
100 km
5660
0
6000
4000
2000
0
2000
4000
6000
2320
50
100
150
250
200
CS1
CS2
Distance in kilometers
5740
5820
17200
CS3
6000
4000
2000
0
2000
4000
6000
50
100
150
200
CS2
250
CS3 CS1
Figure 11 Greyscale slope map representation of Big Island of Hawaii and La Reunion including bathymetric and aerial Digital Elevation Models.
Databases used in this figure are the Global Multi-Resolution Topography (GMRT). Synthesis of Ryan, W.B.F., Carbotte, S.M., Coplan, J.O., et al.,
2009. Global Multi-Resolution Topography synthesis. Geochemistry, Geophysics, Geosystems 10, Q03014. doi:10.1029/2008GC002332, for
Hawaii, and swath bathymetry from FOREVER and ERODER cruises for La Reunion (on land topography from IGN France). The projection is
Mercator in kilometers. Data from Geomap app.

5.9.3.3.2
Shield-building stage
With the increase in the magma production, the eruptions

become more frequent. The volcano reaches the climax of its
activity, building 9095% of the edifice. Tholeiitic to transitional basalts, gabbros, and ultrabasic bodies are emplaced.
Structures such as calderas, pit craters, rift zones, and flank
landslides develop. The first period takes place underwater
with water/magma interactions depending on water depth. At
great depth, submarine lava flows (pillow lava, pahoehoe, and
rare aa types) alternate with the products of their fragmentation. Then, when the top of the volcano approaches the sea
level, hyaloclastic breccia dominates.
The major phase of building of the oceanic shield volcano
continues with the emergence and the subaerial growth of the
edifice. The formation of the volcano is then the result of its
effusive manifestations as well as intrusive activity within the
volcanic edifice (dikes, sills, magma reservoirs). The magma
emission rate is high, for example, 0.1 km3 yr1 for Kilauea
(Dzurisin and Koyanagi, 1984), 0.02 km3 yr1 for Etna
(Wadge and Guest, 1981), 0.01 km3 yr1 for the Piton de la
Fournaise (Lenat and Bache`lery, 1987), and 0.001 km3 yr1 for
the Karthala (Bache`lery and Coudray, 1993).
5.9.3.3.3
Postshield stage
The decline in activity (decrease in the eruption frequency

and magma production rate) accompanies the return to more
alkaline lavas (Frey et al., 1990), first basaltic, then hawaiitic, mugearitic, and trachytic, constituting a well-defined
magmatic series mostly due to fractional crystallization. The
transition from tholeiitic to mildly alkaline or alkaline magmatism may be marked by a period of eruptive quiescence.
However, many recurrent eruptions and intrusions reflect a
progressive instability of the mantle supply. During the alkaline period, the transition between the basaltic and hawaiitic
compositions can be rapid such as for Mauna Kea and Kohala
in Hawaii, or as for Piton des Neiges in La Reunion, or more
gradual with many eruptions such as those for Haleakala in
Hawaii. Magmatic differentiation, associated with the decrease
in eruptive activity of the volcano, indicate a strong slowdown
in the mantle supply, or even a complete stop.
The viscosity and the gas content of the magma increase as
the magma differentiates and becomes more acidic. Eruptions
become more explosive with viscous lava and Strombolian
dynamic, explosive eruptions generating pyroclastic flows and
surges, ultimately leading to the emplacement of domes. The
volcanic activity is commonly restricted to the summit part of
the volcano, creating cinder cones and large quantities of
pyroclastic sediments covering the top of the shield volcano,
as is the case for Mauna Kea in Hawaii, La Grille in Grande
Comore, and Floreana in the Galapagos.
With the more differentiated volcanism, characterized by
silica-rich magmatic compositions, the eruptive dynamics are
varied and in some places become very explosive (characterized by ignimbrites). During this phase, the shield volcano
may take on the appearance of a stratovolcano.
5.9.3.3.4
Erosional and late volcanism
Erosion accompanies the building of subaerial shield volcanoes. Its importance is small because of the extreme frequency of eruptions during the growth phase, but erosion
209
becomes more important as the volcano becomes less volcanically active. Lava flows advance within the depressions
and valleys carved by erosion.
As volcanic activity ceases, erosion becomes the most efficient geodynamic process, sometimes for several million years
(erosional stage). The height of the volcano above sea level
greatly declines and the relief becomes rugged. Coral reefs may
develop around the island in tropical regions.
The volcanic activity can resume sporadically (Renewed
volcanism stage), for example, as in Koolau and Haleakala in
Hawaii, Anjouan and Mayotte in the Comoros, Mauritius.
Magmas emitted during this stage are strongly alkaline and
undersaturated in silica (basanites, tephrites, phonolites,
nephelinites) or hypersthene normative (alkali olivine basalts)
(MacDonald and Katsura, 1964). Silica-saturated lavas (trachytes) also characterize the latest eruptions of Mayotte
(Upton, 1982). Volcanic activity can continue for several
million years, but with very few eruptions, and even including
periods of millions of years without any eruption, as is the
case for Mauritius Island.
5.9.3.3.5
Atoll and seamounts
After volcanic activity ceases the combined effects of erosion

and subsidence reduce the height of the volcanic complex to
sea level. In tropical regions, coral growth may form a complete barrier around a volcanic complex, which may entirely
disappear under sedimentary and biogenic rocks (atoll stage).
If the rate of subsidence is greater than the rate of coral reef
growth then the island sinks beneath the sea level and will
become progressively covered by the marine sediments (late
seamount stage).
5.9.3.4
5.9.3.4.1
Volcanic and Tectonic Morphologies

Summit areas, calderas and pit craters
The summits of shield volcanoes are flat-topped areas commonly occupied by craters or calderas. The terms caldera,
caldera complex, and caldera-like structure covers several
landform types. The different morphologies reflect, to varying
degrees, the fact that the summits of volcanoes, directly above
the shallow magma reservoir, are areas of subsidence. The
term caldera or caldeirao (literally cauldron) was first introduced in the Canary Islands to designate any large cauldronshaped depression, but it has a very broad meaning. Collapse
caldera, explosion caldera, or erosion caldera can be distinguished based on the mode of formation.
For shield volcanoes, different types of structures are present depending on the stage of development of the volcano.
Collapse calderas or drainage calderas (Lockwood and
Hazlett, 2010) develop during the submarine preshield stage
and the shield-building stage, whereas explosion calderas may
accompany the emplacement of more differentiated magmas.
Erosion calderas can form over a long period, early, in conjunction with major tectonic phases such as flank landslides,
or later, when erosion becomes dominant over volcanic
activity.
Erosion processes will be discussed further in Section
5.9.3.5. Attention is given here to collapse calderas, which
represent the most common type of caldera on shield
210
volcanoes, and have the best-preserved morphologies. However, a distinction must be made between two different types
of basaltic calderas: small calderas (generally less than 5 km in
diameter), fully bounded with subvertical walls, and larger
calderas (up to 1520 km in width), with horseshoe
morphologies, commonly open to the sea.
Calderas are generally described as great open depressions,
more or less circular or elliptical, that commonly shape the
summit of active volcanoes, generally formed during large
lateral eruptions or intrusions, which affect the main magma
chamber (MacDonald, 1965; Williams and McBirney, 1979;
Lipman, 2000; Cole et al., 2005). They differ from pit craters,
another type of collapse crater. Pit craters and basaltic calderas
resemble each other both in form and origin. Pit craters have
vertical cliffs and smaller sizes, usually 10s to 100s of meters in
diameter (Walker, 1988; Carter et al., 2007). The structures of
a summit pit crater and their dynamics have been studied by
applying analog models (Roche et al., 2001). These results, as
well as volcanological observations, show that they are linked
to secondary and very shallow magma storages, unlike calderas that have a direct link with the shallow magma chamber
of the volcano. The 2007 collapse of the Dolomieu caldera
at Piton de la Fournaise occurred along a ring fault system and
is clearly attributed to the emptying and roof collapse of
a magma chamber located at a shallow depth (Michon et al.,
2007; Massin et al., 2011).
It is widely accepted that basaltic calderas are formed by an
upward propagation of a disruption in the roof of a shallow
magma chamber during or following an eruption. Explosions
play a secondary and minor role in the formation of these
calderas, and collapse is the major component involved in
their formation. This is supported by the lack of thick tephra
deposits on their rims. Geophysical observations on some
recent examples of caldera formation, at Fernandina, Galapagos, in 1968 (Simkin and Howard, 1970), Myakejima,
Japan, in 2000 (Geshi et al., 2002), and Piton de la Fournaise,
La Reunion Island, in 2007 (Michon et al., 2007; Peltier et al.,
2009; Michon et al., 2009; Staudacher et al., 2009) favor an
incremental collapse of the rock column into the magma
chamber (Michon et al., 2011). In many cases, collapse of a
new caldera or within a caldera accompanies voluminous
flank eruptions, lateral magma intrusion, or long-lived eruptive activity. This suggests that the collapse is related to large
amount of magma withdrawing from the shallow magma
chamber. The Fernandina caldera (Galapagos), a caldera larger
(5 6 km) than the typical Hawaiian calderas like those
of Kilauea, Mauna Loa, or Karthala in Grande Comore
(Figure 12), is associated with concentric fractures parallel to
the caldera rims, which are regarded as the result of the emplacement of cone sheet intrusions (Walker, 1984; Chadwick
et al., 2011).
The size of basaltic calderas varies from approximately
1 km wide, as, for example, Dolomieu caldera at Piton de la
Fournaise (Figure 12(a) and (b)), to more than 16 km as is
the case for Kauai volcano (Hawaii) or for Nuku Hiva Island
(Polynesia) where caldera-like structure have been identified.
Many basaltic shield volcano calderas appear to have been
developed by coalescences of several pit craters. Mokuaweoweo caldera on top of Mauna Loa (Hawaii), Karthala volcano
caldera (Grande Comore), and Olympus Mons caldera on
Mars clearly consist of several pit craters or small calderas

that have combined to form a large caldera. The largest caldera on top of a shield volcano is the nested caldera of
Olympus Mons on Mars. This comprises six coalescent collapses that define a depression of 80 65 km (Figure 13).
The multicoalescent collapse shape is relatively common in
shield volcanoes as for Karhala (Figure 12(c) and (d)).
The walls of a caldera are steep cliffs (Figure 12(b) and
(d)) with banks and talus at their base formed by blocks and
larger portions falling from the cliffs. The observed height of
the walls varies from approximately 0.8 km (Fernandina,
Galapagos) to zero when lava flows erupted later partly or
completely fill the caldera. Commonly, curved cracks and fissures open outside the caldera, more or less parallel with the
caldera walls. The morphology of the caldera floor mainly
depends on its age, and of the volcanic activity inside it. Soon
after the collapse, calderas have a typical funnel shape with
vertical walls in the upper part, perched terraces remaining
from the precollapse floor, and talus of blocks at the bottom
due to the collapse. Months to years after the collapse, intracaldera eruptions contribute to its progressive filling by accumulation of new lava flows. Eruptions within the caldera take
place from fissures cutting the caldera floor, and in some
places extending up the caldera wall on the upper slopes of the
volcano. The floor of the caldera becomes flat, just covered by
talus of blocks falling from the cliffs. Pit craters, open into the
main caldera, may contain a lava lake or a lava pond.
The history of a basaltic caldera consists of a series of
collapses and refills. Caldera collapse is a short, indeed instantaneous, event, whereas caldera infill is a slow process.
When totally filled, lava overflows the rim of the caldera and
the caldera edges become buried and hidden (Merle et al.,
2008).
With the decline of the volcanic activity, during the postshield stage, volcanoes should tend to fill their caldera,
building up a cover of lava flows and scoria cones. With the
decrease in magma supply from the mantle, it becomes increasingly difficult to maintain an active shallow magma
chamber, located high in the edifice. The possibility to form a
new collapse caldera by magma draining is thus decreased
(Mauna Kea Hawaii, La Grille Grande Comore).
As magma differentiates to become benmoreitic and trachytic in composition, calderas can be formed in connection
with explosive eruptions, which are typically subplinian
or plinian in character. But, without a recent example of formation of such a caldera, it is difficult to determine a detailed
picture of their formation. However, these calderas seem to
have formed by collapse, not by explosive volcanic activity.
Large amounts of acidic magma were expelled in a very short
time possibly leading to the collapse of the roof of a deep
magma chamber. Pyroclastic flows and ignimbrites deposits
are present on the flanks of some differentiated shield volcanoes such as Piton des Neiges on La Reunion Island (Rocher
and Westercamp, 1989) or Sao Miguel on Acores (Guest et al.,
1999). Their recognition is always difficult since these calderas
are old structures. Caldera rims may form steep scarps, but
they are commonly widely reworked by erosion. Sometimes,
the rim is indistinct, as younger products from caldera-filling
eruptions bury them. Caldera infilling may include thick lava
flows or pyroclastic units, more easily distinguishable in the
211
7 651 000
S
7 650 500
Do
Bo
Slope %
50
7 650 000
PP
500 m
500 m
40
7 649 500
365 500 366 000 366 500 367 000 367 500
365 500 366 000 366 500 367 000 367 500
(a)
30
(b)
20
8 700 500
10
8 700 000
2
0
8 699 500
8 699 000
1
8 698 500
8 698 000
8 697 500
1000 m
00
00
00
00
00
00
00
00
0 5 21 0 21 5 22 0 22 5 23 0 23 5
3
3
3
3
3
32
3
00
32
(c)
1000 m
00
00
00
00
00
00
00
00
0 5 21 0 21 5 22 0 22 5 23 0 23 5
3
3
3
3
3
3
32
00
32
(d)
Figure 12 (a) Very high-resolution aerial photography (10 cm resolution) of the summit of the Piton de la Fournaise volcano. (b) Grayscale
slope map representation of the summit of the Piton de la Fournaise. Bo Bory Crater; Do Dolomieu Crater where the last and the biggest
caldera collapse occurred in 2007; SSoufrie`re Crater; PP: Petit plateau with a flat topography resulting from a lava lake active during the early
nineteenth century (Lacroix, 1936). (c) Satellite image of the summit Karthala volcano (r Google Earth). (d) Grayscale slope map representation
of the summit of the Karthala volcano with its coalescent collapses: (1) Choungou Chahale Pit Crater; (2) Choungou Chagnoumeni Pit Crater.
Projections WGS84 UTM40S (Piton de la Fournaise volcano) and UTM38S (Karthala volcano) in meters.
landscape, and that could lead to an inversion of relief as a

result of erosion, such as for Kauai (Hawaii) or the Marquesas
Islands (MacDonald, 1972).
If the formation of a caldera is classically attributed to the
vertical collapse of the overlying roof during or consecutively
to the emptying of a shallow magma chamber, the large dimensions of many calderas on basaltic volcanoes and uncertainties about subsidence geometries lead to divergent
interpretation concerning the processes of subsidence. Large
(620 km across) horseshoe-shaped calderas, open to the sea,
exist on many oceanic island volcanoes. They can be formed
both during the differentiated period, dominated by intermediate and acidic magmas, and during the basaltic period of
activity. A good example of the former is the Las Canadas
caldera on Tenerife (Canary Islands Marti et al., 1994). Piton
des Neiges volcano (La Reunion) also exhibits such a caldera
during its period of differentiated volcanism (Chevallier and
Vatin-Perignon, 1982). These calderas are associated with large
volumes of pyroclastic deposits including ignimbrites. The

latter type refers to calderas formed during the basaltic period
of growth of shield volcanoes, without any contribution of
explosive acidic volcanism. Such large basaltic calderas,
with typical horseshoe shapes, are known for several edifices,
including those of Polynesia, Canary Islands, and Cape
Verde. The size of these large basaltic calderas is not consistent
with the classic interpretation of the collapse of the roof
of a shallow magma chamber, although these are sometimes
nested calderas, which may have been formed by the coalescence of several collapse units. The Enclos FouqueGrand
Brule complex at Piton de La Fournaise (La Reunion Island)
provides an excellent current example of this type of
sector collapse. A mechanism such as the one discussed for
Dolomieu caldera (Michon et al., 2007) is unlikely for the
Enclos FouqueGrand Brule caldera complex because of the
large size (B14 km 9 km) of the collapse structure, which
cannot be linked to magma withdrawal from a single shallow
212
2 400 000
500 km
2 200 000
2 000 000
1 800 000
1 600 000
1 400 000
1 200 000
1 000 000
800 000
00
00
0
16
00
00
0
18
00
00
0
20
00
00
0
22
00
00
0
24
00
00
0
26
00
00
0
28
00
00
0
30
00
00
0
32
00
00
0
34
Figure 13 Shaded relief representation of the Olympus Mons Digital Elevation Model (DEM, NASA).
magma body. In such cases, other mechanisms have to be

involved.
The formation of the Enclos FouqueGrand Brule structure
is one the greatest scientific controversies over the formation
of Piton de la Fournaise (vertical versus lateral collapse
Bache`lery and Michon, 2010). Four interpretations have been
successively proposed (Figure 14):
1. A nested caldera collapses for the Enclos Fouque depression, whereas the Grand Brule exhibits the scars of
lateral landslides with the Grandes Pentes as their headwall
(Bache`lery, 1981; Figure 14(a));
2. A giant landslide that was also partly responsible for
the debris avalanche deposits on the submarine flanks
(Duffield et al., 1982; Gillot et al., 1994; Labazuy, 1996;
Figure 14(b));
3. A lateral movement (Grand Brule) that triggers a vertical
collapse in the summit area (Enclos Fouque depression)
above hydrothermally altered rocks (Merle and Lenat,
2003; Figure 14(c));
4. The collapse of the Grand Brule due to the downward drag
related to a dense intrusive complex below the east flank
and the collapse of the Enclos Fouque as the consequence
of the deformation of the hydrothermal system of the
volcano (Michon and Saint-Ange, 2008; Figure 14(d)).
None of these hypotheses is entirely convincing. The formation of large-scale depressions on a basaltic volcano, which
cannot be simply explained by the size of the magmatic
shallow structures, needs more explanations. Large calderas
are most commonly associated with crustal extension as

continental rifting or volcanic arcs (Lipman, 2000). In the case
of Piton de la Fournaise, the formation of this caldera-like
structure also intervenes in an extensive context, created by the
location of Piton de la Fournaise on the eastern flank of Piton
des Neiges and the remains of an older volcano that is now
largely dismantled (Gailler and Lenat, 2010). The Piton de la
Fournaise, a young volcano (less than 450 ka; Merle et al.,
2010), underwent gravitational destabilizations of its eastern
flank, as for the well-documented southern flank of Kilauea
in Hawaii, causing slumps or debris avalanches. This is also
almost commonly reported for other large basaltic calderas
identified in Polynesia, the Canary Islands, Hawaii, and Cape
Verde (for example). The mechanisms invoked to explain
the origin of these large calderas are commonly imperfect.
In the absence of strong lithological or structural elements,
dismantling and lateral spreading of the volcanic edifice
probably plays an important role in the formation of these
large basaltic calderas, where lateral and vertical movements
are associated.
5.9.3.4.2
Cones, eruptive fissures and rift zones
During the prealkalic period, and apart from some very few
exceptions (phreatic and phreatomagmatic explosions that in
some cases occur at the summit craters or near the shoreline),
eruptions of basaltic shield volcanoes are passive or nonexplosive. This reflects the low content in volatiles in the
magma and the high fluidity of basic magma, which allows
the lava to flow freely both below the surface and during
34
Elevation (m)
38
Grand Br
Enclos
Grandes pe
ntes
42
West
4000
30
172
(a)
176
180
East
Enclos
2000
4000
8000
2 km
184
Grandes
Pentes Grand
Brl
North
213
12 000
16 000
Distance (m)
188
192
West
4000
Elevation (m)
42
38
34
East
Enclos
Grandes
Pentes
2000
Grand
Brl
0
4000
8000
12 000
16 000
Distance (m)
30
172
176
180
184
188
192
(b)
West
Elevation (m)
42
38
East
4000
Enclos
Grandes
Pentes
Grand
Brl
2000
Hydrothermal system
0
Gliding horizon
V V
V V V V
34
4000
0
30
172
176
180
184
188
8000
12 000
16 000
Distance (m) Intrusive complex of

Les Alizs volcano
192
(c)
West
Elevation (m)
42
38
East
4000
Enclos
Grand
Brl
2000
Hydrothermal system
34
V V
0
30
172
(d)
Grandes
Pentes
176
Lateral
landslide
180
184
Vertical
collapse
188
4000
16 000
8000
12 000
Distance (m) Intrusive complex of
Les Alizs volcano
192
Location of
the profile
V V V V
Post collapse lava flows
Figure 14 Schematic views and cross-sections of the four different models that are proposed to explain the formation of the large Enclos
FouqueGrand Brule structure. Reproduced from Bache`lery, P., Michon, L., 2010. The Enclos caldera: how did it form? Proceedings of the
IAVCEI Third Workshop on Collapse Calderas, La Reunion, pp. 105108, and Bache`lery, P., 1981. Le Piton de la Fournaise (Ile de la Reunion).
Etude volcanologique, structurale et petrologique, Ph.D. thesis, University Clermont Ferrand II, France.
7 670 000
7 660 000
as
dE
Enclos
Fouqu
So
uth
Ea
7 650 000
7 645 000
st
2 160 000
Loihi
500
Gd Brul
r
No
7 655 000
00 00
10 20
n
zo
ft
Ri
Ri
Chenal Vincent
ft z
on
one
Rift z
SW
2 180 000
East
1500
1000
00
Ri
ft
zo
ne
2 200 000
Kilauea
a
un
idg
20
2 220 000 Mauna

Loa
7 665 000
2 240 000
7 640 000
00
2 140 000
7 635 000
10 km
50 km
2 120 000
7 630 000
00
00
00
00
00
00
00
00
00
0 0 80 0 60 0 40 0 20 0 00 0 80 0 60 0 40 0
1
1
2
2
1
2
2
2
7
7
7
7
7
7
7
7
1
1
1
1
1
1
1
1
00
00
00
00
00
00
0 0 75 0 80 0 85 0 90 0
37
3
3
3
3
50
36
0
73
(a)
Mauna
Kea
50
2 260 000
Rift zones have been defined from the volcanoes of the Hawaiian archipelago, where they are the main constructional
features. Each Hawaiian shield volcano has two narrow, long,
and well-defined rift zones extending from the summit caldera
to the base of the volcanic edifice that extend as far as 150 km,
as in the case of Hana ridge of Haleakala volcano, and the
association of Puna Ridge and the East Rift Zone of Kilauea
(Figure 15 A). Dike-fed flank eruptions are concentrated at the
rift zone, forming linear swarms of fissures and cones, to
produce an elongated prominent relief.
This structural architecture is the result of both the strong
gravitational forces on the magma transit from the central
zone toward the flanks, and the location, high in the edifice,
of the shallower magma chamber from which most of the
magma advances. This transfer of magma is well illustrated by
the subsidence, commonly measured in the summit area
during rift eruptions when magma migrates outward from the
central reservoir.
Two facts can be deduced from such a subsidence: (i) a
direct link exists between the magma pathways and storages in
the center of the volcano and the rift zone, what is often called
the plumbing system of the volcano; and (ii) the shallowness
of the upper central magma chamber, which is, of course, part
of the plumbing system, whose deformations have a knockon
effect to the surface.
The equilibrium between the density of the ascending
magma and the density of the surrounding rocks control the
location of the shallow magma chamber. Ryan (1987) proposed the concept of neutral buoyancy to consider the critical
role of density in the construction of a magma chamber.
00
eruptions. Volcanic eruptions are mostly from fissures, fed

by subvertical dikes. Effusive eruptions of Hawaiian and
Strombolian types build cones and ramparts around the
vents. The percentage of pyroclastic material accompanying
eruptions is generally very small.
Eruptions mainly occur at the summit or near the summit
rather than on the lower flanks. They typically begin with
strong, and more or less continuous, fountaining from a series
of nearly uninterrupted en echelon fissures. For most of
the eruptions, the initial magma production rate is high, and
welded spatter cones and spatter ramparts rapidly build along
the eruptive fissure from molten lava falling from fountains.
Fountains are incandescent jets thrown into the air due to the
hydrostatic pressure of the liquid and the rapid expansion
of bubbles of gas previously dissolved in the magma. After
few days, as eruptions continue, the production rates begin to
decrease. Explosive activity change from a continuous degassing to more rhythmic, than sporadic, explosions. This
Strombolian activity starts when gas escapes less easily, and
the bursting of larger bubbles throw up fragments of molten
lava. Ejecta may reach the surface partly as fluid or essentially
solid. Cinder cones develop then, consisting almost wholly of
loose pyroclastic fragments deposed in superimposed layers
formed by the successive explosions.
Eruptive fissures most commonly radiate from the summit
caldera and extend down the flanks of the volcanic edifice into
the sea. However, some volcanoes differ from this model such
as the Galapagos (Ecuador) where a notable circumferential
pattern of eruptive fissures occurs around the summit caldera,
in addition to more typical radial fissures on the lower flanks.
500
214
(b)
Figure 15 Shaded relief and contour lines representations of Kilauea (a) and Piton de la Fournaise (b) Rift Zones. Note that due to the
difference in size, two different scales are used. Projection Mercator in meters. The Kilauea DEM has been generated from the Global MultiResolution Topography (GMRT) Synthesis. Reproduced from Ryan, W.B.F., Carbotte, S.M., Coplan, J.O., et al., 2009. Global Multi-Resolution
Topography synthesis. Geochemistry, Geophysics, Geosystems 10, Q03014. doi:10.1029/2008GC002332. Piton de la Fournaise DEM has been
generated with data from IGN France (on land topography) and swath bathymetry from FOREVER and ERODER cruises (offshore).
Walker (1989) applied this concept for different volcanic settings to demonstrate its universality for understanding the
ascent and emplacement of magma bodies within a volcano.
However, the existence of lithological barriers with a decrease
in density must also be taken into account. At shallow levels,
the magma accumulates where its density matches the density
of the surrounding rocks. In long-lived shield volcanoes, a
dense body generally exists below the summit area corresponding to older hypovolcanic complexes of intrusions and
cumulates.
On other shield volcanoes such as those of the Hawaiian
chain, rift zones are not always so well defined. Galapagos
volcanoes do not have well-defined subaerial rift zones, although radial fissures are concentrated in a diffuse zone of
fissuring distributed over an area that is several kilometers
wide (Chadwick and Howard, 1991). Conversely, three active
volcanic rift zones have been recognized in the submarine
slopes, aligned, for two of them, with the diffuse subaerial rift
zones (Geist et al., 2006). Their morphology differs from the
subaerial fissure zone, as they have steeper along-axis and
across-axis slopes.
Rift zones also have been recognized for volcanoes of the
southeastern extension of the Society (Teahitia hot spot) and
Austral (Macdonald hot spot) volcanic chains. Submarine
volcanic edifices display morphology with rift zones aligned
with main structural directions corresponding to the orientations of ancient ridges or faults of the underlying plate
(Binard et al., 1991). Binard et al. (1991) proposed that the
formation of the seamounts occurs in an organized fashion
along pre-existing preferential structural orientations that are
associated with zones of crustal weakness.
On La Reunion Island, several rift zones are also present
(Figure 15(b)) but, as in the Galapagos Islands, they are not
prominent narrow linear features, but are broad zones of fissuring, larger near the seashore than near the summit, and
rapidly breaking off on the submarine flanks (Bache`lery, 1981;
Oehler et al., 2007). These rift zones are structures much less
active than those of the volcanoes of Hawaii. Eruptions localized along the rift zones are rare in comparison to those located
at the top or near the top of the volcano. The rift zones are not
places of magma storage, and there are no pit craters. The
eruptive fissures opening up on the rift zones are not parallel to
each other, unlike Hawaii, but radially span up to 451. They do
not extend far offshore, implying that the magma does not
extend very far from the top of the volcanic edifice. The
northeast rift zone of the Piton de la Fournaise shows a peculiar
morphology in submarine part, with a top with a smooth
and regular morphology, and a base very rough and cut
(Figure 15(b)). This reflects a complex history, and a significant
difference in age between an older (43 Myr; Smietana et al.,
2010) and a younger rift zone (probably less than 0.5 Myr).
5.9.3.4.3
Lava flow fields
Lava flows are the most common features of shield volcanoes

because of the dominant basaltic composition of their magmas. The general morphology (see Section 5.9.3.1) of lava
flows depends on several physical parameters including
magma composition, gas and crystal content, temperature,
and the eruption dynamic (morphology of the vents, effusion
rate, degassing, duration), plus the pre-existing volcanic
215
topography. The combination of these factors determines the

rheology within the discrete body of lava emplaced as a dynamically continuous unit, and ultimately all characteristics
that shape the lava flow field as surface, length, thickness, and
velocity.
During the main shield-building stage, lava composition
varies little, and hence the lava viscosity at the vent also varies
little. Then, the eruptive features and the morphology of the
surface over which the lava flow emplaces are the main factors
controlling the shape of the lava flow or the lava flow field. A
critical factor is the capability to form a network of lava
channels and lava tubes.
Lava tubes are tunnel-like features where the active lava
body moves under a hardened roof. Taking into account the
poor ability of basalts to conduct heat, lava tubes are very
efficient structures to retain the heat of the lava body. Consequently, tube-fed lava may remain at elevated temperature at
a constant effusion rate for a longer time than in open flow,
and thus can travel greater distances. Lava tubes are known to
develop only in fluid flows of basaltic composition. They are
common in pahoehoe flows, but can also form in aa flows.
The worlds longest known lava tube is the Kazumura Cave at
Kilauea (Hawaii), which stretches for more than 59 km.
Lava channels are more open structures, without hardened
crust on top, but that frequently develop a thin crust in the
middle of the flow surface. Open-flow channels form very
frequently in aa flows, especially if the length of the eruption
is long and the output rate is high.
On shield volcanoes, lava tubes and channels are responsible for the emplacement of most of the basaltic lava flows,
and they play a fundamental role in transferring lava from the
source vent to the lava flow front. Eruption rate and volume
are considered as important parameters in controlling whether
a lava tube will form during an eruption and so produce a
long lava flow (Walker, 1973). Sheet-like lava flows, without a
well-defined lava channel, generally develop during short fissure eruptions, or during a brief event during an eruption
involving a rapid overflowing due to an increase in the effusion rate or the breaching of a spatter cone containing a
lava lake.
The length of an eruption is a factor frequently controlling
the width of the lava flow field. Longer lasting eruptions are
associated with complex lava flow fields where both aa and
pahoehoe surface types can coexist; the first corresponding to
the higher effusion rates.
The preflow topography is also an important factor controlling the flow morphology. On flat areas, lava flows spread
out and accumulate. In closed low-lying areas like the floor of
a pit crater or a caldera, the accumulation of lava can result in
the formation of a lava pond that is like a lava lake but not
directly related to the magma-feeding path.
Steep slopes considerably increase the lava velocity as well
as the formation of narrower lava flows, which lead to increased flow turbulence resulting in a more efficient cooling
and degassing. This can change pahoehoe lava flows in aa
lava flows when the effusive rate is high enough. When a lava
flow invades an eroded area, pre-existing valleys may channelize and concentrate the flow, resulting in thick accumulations, commonly with a prismatic base. Whatever the type of
lava flow, numerous ruptures and deformations occur within
216
the moving lava flow. They express internal rheological and

pressure changes induced by rapid cooling from the crust exposed to air and continuous magma supply. As a result, different morphologies may develop due to shearing within the
flow and strong deformation of the crust that produces pressure ridges and tumuli, in some places injected by molten lava.
Pahoehoe flows (Figure 10 (c) and (d)) that are characterized
by the continuous solidification of the crust at the surface, and
molten lava that continues to flow beneath, exhibit the
greatest number of surface features. Many terms are used to
describe the diversity of surface types, including ropy, billowy,
imbricate surface, hummocky surface with lobes and toes, and
shelly pahoehoe surface.
5.9.3.4.4
Faults, slides, and slumps
Gravitational collapse and slumping are known to be important processes on oceanic islands, and they play a major
role in the formation of the submarine flanks of islands. Most
shield volcanoes experience considerable modifications due to
gravitational collapse of their flanks, mainly by massive
slumps and huge landslides as, for example, in Hawaii
(Moore, 1964; Moore et al., 1989; Moore et al., 1994), the
Canary Islands (Krastel et al., 2001; Mitchell et al., 2002), La
Reunion (Lenat et al., 1989; Lenat et al., 1990; Cochonat et al.,
1990; Bachelery et al., 1996), Cap Verde (Day et al., 1999),
Marquises (Filmer et al., 1994; Clement et al., 2003), La
Societe (Clouard et al., 2001; Clement et al., 2003; Hildenbrand et al., 2004), the Galapagos (Chadwick and Howard,
1991; Naumann and Geist, 2000; Mitchell, 2003) and Tristan
da Cunha (Holcomb and Searle, 1991).
Moore et al. (1994) suggested that there are at least 68
major landslides that exceed 20 km in length between Midway
and Hawaii. Some landslides have volumes as much as
5000 km3. Around La Reunion Island, Oehler et al. (2007)
and Le Friant et al. (2011) have identified several debris avalanche deposits resulting from the sliding and spreading of the
aerial and submarine volcanic formations. Oehler (2005)
observed at least 47 landslides around La Reunion Island due
to the flank instability. In the Canary Islands, 18 landslides
have been identified around most of the islands within the
archipelago (Krastel et al., 2001).
Two types of landslides are common (Varnes (1978) nomenclature). First, there are rotational slides, or slumps, which
are slow phenomena, characterized by a significant thickness
of moved material (410 km) and large scarps. This type of
landslide is characterized by a deep detachment and the preservation of the internal cohesion of the slipped material.
Slumps do not produce clear amphitheater structure as their
heads are commonly filled by the addition of new volcanic
material. Rotational landslides are limited upstream by high
fault scarp. Second, there are debris avalanches, which are the
result of sudden and rapid detachment of a significant portion
of a volcano, mobilized by the action of gravity along a basal
shear plane (Ui, 1983; Siebert, 1984). Debris avalanches are
fast-moving shallow landslides, which are able to flow over
long distances, beyond the base of the volcanic edifice itself.
Their thickness is less than those of slumps, and their bases
are less deeply rooted. An amphitheater-like structure usually
marks the landslide head. Deposits have hummocky surfaces,
shaped by the presence of blocks (portions of the initial cone
less destabilized) that are up to several tens of kilometers wide.

The most distal part of the deposits are fine grained that can
sometimes develop into debris flows or turbidity currents.
Debris avalanche deposits and slump structures commonly
spread out far in the deep sea, but have their origin in the
subaerial part of the edifice or close to the coast. The headwalls of these lateral mass movements are strongly marked by
steep slopes (up to 301) and, in some cases, dipping-seaward
normal faults. These faults are considered as the result of instabilities, with the unsupported flanks sliding toward the
ocean, due to repeated forceful intrusions of magma into the
rift zones (Swanson et al., 1976). Movement likely occurs on a
low-angle fracture zone or slippery decollement surface. This,
with gravity-driven forces, leads to the slow spreading of the
volcanic edifice whose geometry is determined by both magmatic transfers and the specific stress field of the volcano that
depends on its morphology and the influence of neighboring
volcanoes. Major sector collapses have occurred repeatedly in
the growth of some volcanoes.
Except for a few extraterrestrial volcanoes such as Olympus
Mons on Mars (Figure 13; Crumpler et al., 1996), the most
impressive of these mass-wasting structures are those of the
Hawaiian volcanoes (Figure 16). Alika slide on Mauna Loa
(Big Island of Hawaii) and Nuuanu slide at Koolau volcano
(Oahu Island) have spread 90 and 160 km across the seafloor,
respectively. This is due to their large size and great height.
Debris avalanche deposits contain large sections of the volcanic shield than can be as much as 35 km long and 1.5 km
thick (Lipman et al., 1988; Moore et al., 1989; Kanamatsu
et al., 1998; Lockwood and Hazlett, 2010). Tsunamis likely
were produced by such rapid seaward displacement of materials (Moore and Moore, 1984).
Debris avalanche deposits are characterized by two depositional facies: (i) blocks that are fractured and deformed,
but more or less preserved, fragments from the source volcano;
and (ii) matrix that consists in a mixture of small volcanic
fragments derived from various parts of the source volcano.
Hummocky topography, natural levees, marginal and distal
cliffs, jigsaw cracks, and an amphitheater at the source are
characteristic features of a debris avalanche deposit (Ui et al.,
2000).
Slumps are less catastrophic events, with more progressive
movements. Displacements observed and measured on the
southern flank of Kilauea volcano (Hawaii) provide the best
examples. They determine the active Hilina fault system that
can be at the origin of consequent earthquakes as in 1975 with
a magnitude-7.2 earthquake (Lipman et al., 1985).
The relationship between slumps or slides and debris
avalanches is not clearly established. These two processes are
involved in the spreading of volcanic edifices. Slumps may be
precursors of catastrophic landslides. Slumps and landslides
may occur simultaneously but in different locations within a
volcano. Alternatively, they may develop at different times
during the growth of a volcanic edifice.
The largest slumps and debris avalanches are present on
emerged volcanic structures, with the highest elevations.
Submarine volcanoes whose height exceeds 2500 m are
also affected. Smaller volcanoes are generally more stable
(Mitchell, 2003). This difference reflects the fact that, for
shield volcanoes, the destabilization is first controlled by a
Nuuanu slide
217
2230 N
Tuscaloosa
seamount
22 N
u
ola
Ko
Wailau
slide
2130 N
Oahu Island
Molokai Island
21 N
Lanai Island
15830 W
158 W
15730 W
157 W
Figure 16 Bathymetric and aerial DEM of Oahu and Molokai islands and the debris fields for the giant Nuuanu and Wailau landslides from
Koolau and East Molokai volcanoes (thin dashed lines show slide boundaries). Projection Lat/Long WGS84. This figure has been generated with
GeoMapApp (http://www.geomapapp.org) and the Global Multi-Resolution Topography (GMRT) Synthesis. Reproduced from Bilham, R., Larson,
K., Freymueller, J., 1997. GPS measurements of present-day convergence across the Nepal Himalaya. Nature 386, 6164, and Cretaux, J.-F.,
Soudarin, L., Cazenave, A., Bouille, F., 1998. Present-day tectonic plate motions and crustal deformations from the DORIS space system. Journal
gravitational disequilibrium rather than by a direct intrusive

process. The height of the source area of the landslide, the
initiation of the avalanche near or above sea level, the confinement of the debris avalanche during transport and deposit,
and ultimately, differences in building rates of the islands, can
be factors that control the degree of fragmentation of the
debris avalanche (Mitchell et al., 2002).
5.9.3.5
Erosional Morphologies
Erosional morphologies are very important for helping to

classify volcanoes. Thouret (1999), in his classification of
volcanoes, defined a specific group corresponding to volcanic
landforms resulting from eruptive and/or erosional processes.
The alternation between rapid growth (due to eruptive
dynamism) and erosion processes (weathering, gravity and
tectonic) produces landscape scars that are easily identifiable.
Degraded or/and fragmented volcanic materials are made
available for such carriage by rivers and drainage streams. For
these processes, endogenous and exogenous characteristics
and parameters are predominant.
5.9.3.5.1
Influence of volcano shape
Slope steepness and direction largely control erosion. For a

simple volcanic body, the pattern of erosion will be radial for
an initial stage and will later evolve as function of the
erosional history. Ollier (1988) described four broad stages

(Figure 17) of volcanic edifice erosion:
(i) The intact stage, where the edifice still has its original
conic shape and start to develop a radial drainage
network;
(ii) The planeze stage, where the radial drainage network
erodes the edifice to form a central erosional depression
and roughly triangular plateaus (known as planezes) left
intact between the converging drainage lines;
(iii) The residual stage, where the planezes have been
completely eroded and only a well-developed network of
radial rivers remain; and
(iv) The skeletal stage, where the original shape of the edifice has completely disappeared, and only a few relics of
the old edifice remain.
The Ollier (1988) model is very conceptual and can only
be applied to a very simple volcanic edifice such as Tristan Da
Cunha Island that has a perfect radial drainage.
In most cases, volcanoes, and particularly for complex
shield volcanoes, such as La Reunion, Mauritius, Big Island,
Oahu, Molokai, Canary Islands, and Tahitii, can simultaneously have morphologies characteristic of different erosion
stages, in relation to its geology, sealevel changes, and paleoclimatic history. More complex volcanic landscapes quickly
generate more complicated fluvial erosion. In cases where
several volcanic edifices are juxtaposed, erosional morphologies
218
(a)
Intact stage
(b)
Planezes stage
(c)
Residual stage
Rivers
(d)
Planezes
Truncated rivers
on the planezes
Skeleton stage
Figure 17 Main stages of volcanic edifice erosion. Modified from Ollier, C., 1988. Volcanoes. Blackwell, Oxford, 228 pp. with permission from Wiley.
3
3
2
Figure 18 Influence of lava flow emplacements on preexisting drainage network. Reproduced from Ollier, C., 1988. Volcanoes. Blackwell,
Oxford, 228 pp, with permission from Wiley. Numbers 1 to 5 are explained in text.
are directly related to the volcano growth chronology. A new

volcanic edifice that grows in the vicinity of an older one
will modify the drainage pattern. For example, Ollier (1995)
showed several examples of drainage modifications in catchments affected by the formation and erosion of huge volcanoes.
5.9.3.5.2
products, pyroclastic flows, and debris flows are set up on the

slopes of volcanic edifices. The susceptibility of these materials
to erosion is greater, and erosion rates may reach very high
levels. A closely related consequence is the greater erosive
power of the drainage system that is richer in sediment particles (Saint-Ange, 2009).
Influence of volcanic activity
New lava flows advance into pre-existing thalwegs to partially or

completely fill the valleys. Differential erosion due to the different strength of materials can result in relief inversion (1 and 2
in Figure 18), and/or changes in drainage pattern and the development of new river valleys (3 in Figure 18). In addition,
lava flows can dam lakes and/or produce waterfalls (respectively,
4 and 5 in Figure 18) causing abrupt changes in the river profile.
As noted earlier, the eruptive dynamism of hot spot volcanism is mainly characterized by low-viscosity lava flows.
However, in some situations or circumstances, explosion
5.9.3.5.3
Influence of geology and lithology
Geologic context plays an important role in controlling surface

erosion. Apart from the main radial component, the orientation of the drainage network is primarily related to geologic
structures. Clearly, geological structures, such as faults, intrusions, rift zones, and calderas (described in Section 5.9.3.4)
play a major role on the river pattern that influence erosion
processes (Lamb et al., 2007; Karatson et al., 1999).
The mechanical and hydraulic behaviors of the different
lithologies and their variation across the edifice cause
differential and specific erosion. La Reunion and Mauritius illustrate different geomorphological and landscape aspects. The
massive Kohala edifice is also used here for its typical
morphologies.
The geology and volcanic history of Mauritius has been the
subject of numerous studies (Lacroix, 1936; De Chazal and De
Baissac, 1950; Simpson, 1950; McDougall and Chamalaun,
1969; Baxter, 1972; Sentenac, 1964; Upton, 1982; Perroud,
1982; Saddul, 1995).
Four distinct stages separated by quiescence periods have
been recognized:
219
An eroded relief marks the geomorphology of Mauritius.

Nevertheless, quantitative information about erosion, sedimentation, and transport of sediments are very limited. Nigel
and Rughooputh (2010) gave results of measurements of
sediment loadings of 1 to 3 t ha1 yr1. They also report results
of soil loss of B10 t ha1 yr1 for a sugarcane plantation.
Morphological aspects of Mauritius are similar to those of
the Skeletal stage of Olliers (1988) model. Higher lands, between 300 and 800 m asl with slopes 420% (Saddul, 1995),
represent approximately 300 km2 or 16% of the total territory.
The residual forms, ridges, peaks, and crests shaped by erosion
are common in the landscape. Three main units are recognizable that relate to the massive erosion of Old lavas and
Early volcanic series (Figure 19):
(i) Old lavas series, which date between 5 and 19 Myr.

(ii) Early volcanic lavas series. This phase of activity extends
over approximately 1.8 Myr, from 3.5 to 1.7 Myr.
(iii) and (iv) Late volcanic lavas series. After a period of approximately 1 Myr of quiescence, two separate episodes
(0.7 Myr to 0.5 Myr and 0.1 Myr to 0.025 Myr) took
place with a new long period of calm of 0.4 Myr. The
youngest lavas cover 75% of the Island (Nigel and
Rughooputh, 2010).
(i) The North West Mountain Range where the highest point
reaches 823 m asl (Figure 19, Cross-Section Mau1). This
set is composed of the Old lava series.
(ii) The SouthSouthwest mountain range. This is the largest
mountain of Mauritius and with the highest point
(828 m asl). This mountain is particularly massive,
Mau4
7 690 000
7 780 000
10 km
10 km
7 680 000
Mau1
Northing (m - UTM40S)
7 770 000
7 670 000
7 760 000
Mau2
7 660 000
7 750 000
7 650 000
Mau3
7 740 000
7 640 000
La Runion
320 000
340 000
330 000
360 000
350 000
7 730 000
370 000
Elevation (m)
Elevation (m)
3000
2500
1000
1500
1000
500
0
Mauritius
530 000
Easting (m - UTM40S)
540 000
550 000
560 000
570 000
Easting (m - UTM40S)
Vertical exaggeration : X3
La Runion
Run3
0
10
20
1400
1200
1000
800
600
400
200
0
30
Run1
Run2
40
50
10
20
Mau2
30
Distance (km)
60
Vertical exaggeration : X6
Mauritius
Mau1
580 000
40
Mau3
50
Mau4
60
Figure 19 Shaded relief, contour lines and cross sections representations of La Reunion and Mauritius Islands. Piton des Neiges (PDN) and
Piton de la Fournaise (PDF). Projection WGS84/UTM40S in meters (La Reunion DEM generated from IGN data. Mauritius DEM generated from
SRTM data).
220
completely cut by deep valleys and narrow canyons

(Figure 19, Cross-Section Mau3). This is composed of
the Old lavas and Early volcanic series.
(iii) The Eastern relief that reaches 750 m asl, comprising a
high 25-km-long EW ridge from the center of the island
to the east coast. Four crests stretch out radially,
delimiting the summits of four convex spurs separated
from each other by short drainage patterns.
The most notable is the presence, in the center of the island, of a large circular plateau surrounded by diverging slopes
(48%; Saddul, 1995) toward the coastal plains (Figure 19,
Cross-Section Mau2). In this area, at elevations between 500
and 600 m asl, are located the remains of the recent volcanic
activity (Late volcanic lava series). This is a very good example
of surface rejuvenation due to recent volcanic events. As
shown in Figure 19, the recent lavas have smooth surfaces,
and some cones are still well preserved in the central region
(Cross-Section Mau4). Perroud (1982) described the morphology of the eruptive vents. All have similar shapes and
are tens of meters high. Twenty of them are clearly visible in
the landscape today, and evenly align along a N201 axis across
the island.
From the center of the island, except on the three main
mountain chains, short rivers and canyons radiate outward
and characterize/control the hydrological network. Proag
(1995) identifies 25 watersheds characterized by many major
rivers that drain from the central plateau and stretch for the
most part toward the east coast. The longest perennial river is
the Grand River South East that flows for more than 25 km.
Finally, much of the coastline consists of coastal plains
backed by a large lagoon. Compared with La Reunion,
Mauritius reflects a more mature stage of volcanic island.
La Reunion is built of two central volcanic massifs: the
younger, active, Piton de la Fournaise volcano and the older,
now dormant, Piton des Neiges volcano. Piton des Neiges
emerged more than 2 Myr ago and currently stands as a shield
volcano at 3070 m asl, 47000 m above the seafloor. The
activity of the younger Piton de la Fournaise volcano began
approximately 0.45 Myr on the southeastern flank of Piton
des Neiges and is still active, with its summit reaching an
elevation of 2631 m asl. La Reunion is mainly composed of
basaltic rocks (intermediate between tholeiitic and alkaline
magmas). The two central massifs are similar in composition,
but with different ages (McDougall, 1971; Bache`lery, 1981;
Upton, 1982; Gillot and Nativel, 1982; Gillot and Nativel,
1989; Bret et al., 2003; Merle et al., 2010).
The main river systems on La Reunion also drain the edifices of Piton des Neiges and Piton de la Fournaise in a general
radial pattern. Using Olliers (1988) model, Piton de la
Fournaise is in the initial stage, whereas Piton des Neiges is in
the planeze stage (Figure 19).
Three remarkable large erosional features are visible on
Piton des Neiges (Figure 19), which are known as cirques
(Cirque de Salazie, Cirque de Cilaos, and Cirque de Mafate).
These major erosional depressions are thought to have their
origin in the period when the effusive rates of Piton des Neiges
began to decrease at approximately 0.22 Myr.
Nearly 0.012 Myr (Deniel et al., 1992), magma production
on Piton des Neiges stopped and only erosive mechanisms
were active, which widened and deepened the cirques to their

current state today, accounting for a volume loss of 4250 km3
(Fe`vre, 2005). Piton de la Fournaise has 19 drainage basins,
but only three rivers are perennial (Rivie`re de lEst, Rivie`re
Langevin, and Rivie`re des Remparts). All three rivers are fed by
springs resulting from the combination of highly permeable
rocks in the upper section of the active volcano, and high
rainfall (Join et al., 2005).
As shown on Figure 19, the geomorphologies of La
Reunion and Mauritius are quite different in many ways.
However, there are similarities, which include smoothsurfaced recent lava flows in the center of these two islands,
between Piton des Neiges and Piton de la Fournaise in
La Reunion Island and in all the center of Mauritius. Crosssections Run2, Mau1, Mau2, Mau3, and Mau4 show a generic
convexoconcave shape. However, Run1 and Run3 crosssections are very different from those in Mauritius.
Run1 cross-section shows western slopes, which are not
only smoothed by a relatively recent volcanic activity, but also
slightly scared by erosion. The center of the island is very
rugged since erosion is very intense, caused by the intensive
rainfalls, and favored by a specific lithology (weathered rocks
and breccias). In the east, very high rainfall promotes particularly intense erosion. The Run3 cross-section shows a
young volcano where erosion is still little.
Figure 20 illustrates the differential erosion in the heart of
a dormant volcano (this corresponds to the middle part
of Run1 cross-section in Figure 19). Figures 20(a) and (e)
highlight how resistant intrusions are to erosion (a. intrusive
body, e. Dike). Figure 20(b) illustrates the erosion on Piton
des Neiges, where steep slopes are prone to mass movement.
Figures 20(c), (d) and (f) show that debris flow products are
available to erosion.
Kohala volcano on Big island (Hawaii) illustrates the effects of an inherited structure on surface erosion. Big Island is
composed of five (Kohala, Mauna Kea, Hualalai, Mauna
Loa, Kilauea) or seven volcanoes (adding Mahukona and
Loihi seamounts) according to Moore and Clague (1992).
Figure 21 shows a much eroded area located between two sets
of large canyons on the north of Kohala volcano, the Pololu
and Honokane canyons that limited the much eroded area
westward, and the Waimanu and Waipao eastward (respectively, 1, 2, 3, and 4 in Figure 21). The average rainfall in this
area is B4 m y1. This is much more than on the other side of
the rift zone that marks the relief of Kohala volcano (Lamb
et al., 2007). This is part of the explanation for the existence of
this deep drainage pattern. Another controlling factor is the
age of the surface. Kohala is the oldest volcano of the Big
Island of Hawaii. The last terrestrial eruption of Kohala volcano occurred at 0.25 Myr (Dalrymple, 1971, McDougall and
Swanson, 1972; Moore and Clague, 1992; Wolfe and Morris,
1996), and the last eruption of Mauna Kea occurred at 3.3 kyr
(Porter et al., 1977). Mauna Loa, Kilauea, and Hualalai volcanoes are still active. The last eruption of Hualalai volcano
occurred in 180001 (Moore and Clague, 1992). In addition,
it is now recognized that an important landslide is responsible
for the general subsidence of the northern flank of Kohala
volcano. Major lineaments on the summit of the volcano are
the marks of the normal faults associated with this landslide.
The Pololu, Honokane, Waimanu, and Waipao canyons are
(a)
(b)
(c)
(d)
(e)
221
(f)
Figure 20 Pictures of Piton des Neiges cirques and summit erosion. (a) Pain de sucre intrusive body in Cilaos cirque; (b) Western part of the
Piton des Neiges volcano; (c) & (f) Erosion in badlands in Salazie and Mafate cirques; (d) Very deep canyon in Mafate; (e) Dike injected below
the Piton des Neiges summit.
the lateral limits of this landslide. This landslide therefore

creates weaknesses where erosion is enhanced. The scars of
erosion are also present on the seafloor. Cross-sections in
Figure 21 highlight that: (i) the landslide on northern flank
of Kohala causes asymmetry in slopes between the north and
south (CS1); (ii) submarine slopes of Kohala have a convexoconcave morphology (CS1), whereas those of Mauna Kea
are affected by two significant breaks in slope (CS2); (iii)
Pololu, Honokane, Waimanu, and Waipao canyons are not
comparable with the erosion scars elsewhere in Big Island; and
(iv) the depth of the canyons is much greater on land than on
the submarine flank.
5.9.3.6
Submarine Morphologies
Oceanic volcanic islands are like icebergs (Schmincke, 2006);

what is above sea level is only part of their full volume. For
example, only 2.5% and 3% of the total volume of Gran
Canaria and La Reunion are above sea level, respectively. Since
a large part of the geomorphology of volcanic islands is below

the sea level, studying their submarine forms is important for
revealing the history and processes that produce the islands.
Clearly, the shoreline is not an impassable geological
boundary and the transfer of volcanic materials from land to
the sea is a major process in the evolution of these islands.
During the 1990s, scientific cruises increased around many
volcanic islands both in the context of subduction around the
Antilles (Deplus et al., 2001), or in the context of hot spot
with cruises near the Canary Islands (Masson, 1996;
Schmincke and Sumita, 1998; Schneider et al., 1998; Watts
and Masson, 1995), the Hawaiian Islands (Garcia, 1993;
Garcia and Hull, 1994; Moore et al., 1994; Naka et al., 1993),
and La Reunion Island (Lenat et al., 1989; Cochonat et al.,
1990; Ollier et al., 1998). These provided many important
new insights into the evolution of volcanic islands related to
hot spots.
As such, as in other very active geodynamic contexts,
in volcanic settings, processes of building and processes
of dismantling by erosion are coeval. Sometimes erosion
222
N
CS1
20.4
Shoreline
CS2
20.35
1
20.3
CS3
20.25
20.2
CS4
20.15
.8
55
.75
55
.7
55
.65
55
.6
55
.55
55
.5
55
.45
55
.4
55
.35
55
.3
55
.25
55 W
Elevation in meters
2000
1000
0
1000
CS4
10
20
30
40
CS3
50
60
CS1
CS2
2000
Figure 21 Kohala volcano on Big Island Hawaii. 1, 2, 3, and 4 represent respectively Pololu, Honokane, Waimanu and Waipao canyons.
Projection Lat/Long WGS84. Database used in this figure is the Global Multi-Resolution Topography (GMRT). Synthesis from Ryan, W.B.F.,
doi:10.1029/2008GC002332.
dominates, whereas sometimes volcanic activity rejuvenates

the relief. Construction phases and stages of erosion or
dismantling may alternate or exists when some parts of
the volcanic edifice are isolated from the main area of activity
by tectonics (Merle et al., 2010). In terms of volume of
material, the deposits from flank landslides (slumps and
debris avalanches) are higher by several orders of magnitude
in comparison with the pure products of erosion. But,
throughout the construction of a volcano, alternation of
erosion, during which develops a hydrographic network,
and filling by the volcanic activity leading to a rejuvenation
of the landscape, tends to underestimate the production of
epiclastic material by only taking into account the existing
relief.
5.9.3.6.1
Submarine volcanic activity
At great ocean depths, the dynamics and morphology of the

lava flows are determined by the surface effect of quenching
experienced by the magma erupted into the water, the limiting
effect of the high water pressure on the expansion of gases and
the output rate of magma. Seamounts and the base of large
oceanic shield volcanoes are made of dense pillow lavas and
accumulations of fragments of pillow lavas. Pahoehoe lava
flows are also common. However aa lava flows are rare or
absent. Hyaloclastites exist at depths below 2000 m, which is
mainly due to contraction, with the sudden cooling of the
magma and the effect of fragmentation during transport by

high-density currents.
Studies on the seamounts and guyots (Lonsdale and
Spiess, 1979; Smoot, 1982; Fornari et al., 1988; Batiza et al.,
1984; Vogt and Smoot, 1984; Clague and Dalrymple,
1987; Binard et al., 1991) show that structures like calderas
and craters, as well as rift zones, exist on young shield
volcanoes and originate underwater. Investigations on the
submarine Loihi volcano, off the coast of the Big Island of
Hawaii (Malahoff et al., 1982; Malahoff, 1987) also revealed
the existence, at the top of the volcano, of a summit depression (possibly a summit caldera), of three pit craters, including Peles pit that formed in 1996 (Davis and Clague,
1998), and of large-scale landslides affecting the flanks of
the volcano on the sides of two main rift zones. So, the formation of such structures, including calderas, is not limited
to the final stage of construction of Hawaiian basaltic shields,
but can be repeated throughout their evolution. The summit
of Loihi is almost 1000 m below sea level. The pressure
exerted by the water column is enough to prevent strong
explosive hydromagmatic activity. Loihi seamount is mainly
built by submarine lava flows (pillow lava, pahoehoe, and
aa flows). As submarine lava flows generally travel less
far than the subaerial flows, because of the rapid cooling
imposed by the cold water, the slopes of Loihi are 101 to
151 steeper than the slopes of the shields above sea level.
The lower slopes of Loihi are mostly covered by breccia

where small landslides occur.
When the volcanic edifice reaches a shallow depth (a few
hundred meters below sea level), then pyroclastic and epiclastic materials become dominant, and explosive hydromagmatic activity becomes the main style of eruption. Magmatic
gases and steam formed by the contact of magma and water
result in violent explosions and fragmentation of the magma
emitted, creating hyaloclastites and explosion breccias. Glassy
and little-indurated deposits are associated with materials reworked under marine action. Slope failures are common,
which help transport clastic debris to the base of the shield
volcano.
5.9.3.6.2
Formations built by volcaniclastic materials
Slumps and debris avalanches that most commonly originate

in the terrestrial part of the island have already been discussed
in Section 5.9.3.4.4. These formations extend widely on the
submarine slopes of volcanic islands, extensively involved in
the construction of submarine slopes. Debris avalanches are
formed by rupture of a flank of the volcano following its deformation and propagate in the submarine realm, covering the
immerged flanks of the island and the surrounding abyssal
plain. Debris avalanches move quickly and can spread over
great distances (Deplus et al., 2001; Masson et al., 2002;
Moore et al. 1994; Oehler et al., 2007).
The submarine eastern flank of Piton de la Fournaise is
largely built by debris avalanche deposits dating back to
150 kyr, which diverge from the subaerial portion of Piton de la
Fournaise volcano (Bachelery et al., 1996; Bache`lery et al.,
2003; Labazuy, 1996; Lenat et al., 1989; Oehler, 2005; Oehler
et al., 2004; 2007; Lenat et al., 2009; Le Friant et al., 2011;
Figure 22). The more recent landslide (possibly less than 5 kyr)
is related to the Grand Brule depression and corresponds to the
excavation of the Chenal Vincent and the deposit of the RalePousse fan. The eastern Plateau (Plateau sous-marin est on
Figure 22) relates to an older deposit whose edges are largely
cut by secondary landslides with presumed source areas within
previously emplaced avalanche deposits. Broad valleys with
steep walls cut these debris avalanche deposits. These mass
movements have no continuity with the terrestrial hydrographic network, and develop from both sides, and on the
submarine Eastern Plateau. This reflects the very high dynamics
of erosion affecting the submarine flanks.
The Chenal Vincent begins at the coast and measures
15 km and trends SWNE. Its width is 5.6 km at the beginning
of the valley. Several other valleys extend the Chenal Vincent
partly cutting into the deposits of the Rale-Pousse debris
avalanche, which suggests a recent age. The bottom of the
Chenal Vincent is itself cut by a network of small channels of
erosion (Figure 23) with volcaniclastic deposits. On top, two
sets of lava flows from Piton de la Fournaise are clearly
identifiable on acoustic imagery. These lavas feed volcaniclastic submarine sedimentation when they break up. This is
an excellent example of the dynamics of submarine proximal
slopes of an active volcanic island, where volcanic and sedimentary processes interact to form the relief. High-resolution
sonar imagery of the sea floor has led to better characterization
of their morphology and identification of a number of structures and processes (Cochonat et al., 1990; Ollier et al., 1998;
223
Saint-Ange, 2009). Numerous geomorphic features are observed such as superficial slump, slide scars, scours, gullies,
and dunes (Figure 24).
5.9.3.6.3
Sedimentary volcaniclastic systems
Submarine slopes represent a significant pathway for sediment

transfer from coastal areas to the deep sea. Since the early
1980s, the spectacular advances in the oceanographic technologies (multibeam bathymetry, sidescan sonar imaging,
high-resolution seismic reflection data, and piston cores) have
enabled much more accurate description of the submarine
flanks of volcanoes. This research has highlighted the importance of sediment transfer down submarine volcanic
slopes. On the submarine slope, in the prolongation of the
main river mouths, sedimentary processes are mainly dominated by erosion and sediment transport due to gravity instabilities induced by high sediment supply. In many cases,
these gravity-induced events create turbidity currents incising
progressively deep submarine canyons across the upper slope
and forming deep-sea fans composed by turbidite deposits. In
complete sedimentary systems (source to sink), the only area
dominated by depositional processes is located in the distal
marine environments at the base of the submarine slope
where the sediment is finally deposited. Detailed studies of
submarine volcanic slopes have shown the importance of
sedimentary processes in the evolution of volcanic islands and
the adjacent basins. Volcaniclastic turbidite deposits have already been identified off volcanic islands in subduction context such as in Stromboli (Casalbore et al., 2010; Romagnoli
et al., 2009b) and the Lesser Antilles (Deplus et al., 2001; Le
Friant et al., 2009), and also for oceanic shield volcanoes such
as Hawaii (Garcia and Hull, 1994) and the Canary Islands
(Acosta et al., 2003) where volcaniclastic turbidites are visible
for more than 400 km from these islands.
Off Hawaii and the Canary Islands, two main processes
have been proposed to explain the genesis of turbidites: (1)
large flank landslides; and, to a lesser degree, (2) feeding by
rivers. Funck and Schmincke (1998) in the Canary Islands and
Moore et al. (1989) in Hawaii show that the main submarine
canyons cutting the submarine slopes of these islands are
situated at the outlet of subaerial canyons. Krastel et al. (2001)
and Mitchell et al. (2003) also concluded that rivers can feed
turbiditic currents off the coast of these islands. Alibes et al.
(1999) and Garcia and Hull (1994), however, argued that
large flank landslides would be the main process causing
turbidity currents off the volcanic islands such as the Canary
Islands and Hawaii.
Recent work on the submarine slopes of La Reunion is the
only study, until now, that has imaged a complete sedimentary
system from its emerged source to the abyssal plain. The
physiography and climate of La Reunion Island provide favorable conditions to generate turbidity currents in the continuity of river floods generated by heavy rainfalls. Five
volcaniclastic deep-sea fans, extending to more than 200 km
from La Reunion, were discovered in 2006 during the FOREVER and ERODER oceanic surveys. Their geomorphic features (canyons, channels, and sediment waves) are typical of
turbidite systems (Saint-Ange et al., 2011; Sisavath et al., 2011).
These sedimentary systems are connected on land to the main
erosional features of the island (cirques and major rivers),
Ri
La de de
Dra
gue
10
00
2000
21 S
Vall
Rl e du
-Pou
ss
224
50
0
es
Rl-pouss
e
ed
ll
Va
De
pre
Gd. Brl
en
inc
V
Ch
Plateau
sous-marin
Est
F.O
Cou
loir D
olto
Pe
RF
nt
SE
Z-
2120 S
ss
ion
no
500
rd
Z-N
RF
on
rg
o
sG
tro
Ci
s
let
Ga
0
5550 E
(a)
10
km
5615 E
Sediment area
Vall
e du
R
La ide d
e
Dr
ag
ue
Rl
21 S
Debris avalanche
deposits
-Pou
ss
lle
des
nes
rgo
Go
pr
D
ent
inc
.V
Ch
es
sio
nN
or
Va
Plateau
sous-marin Est
Piton de la Fournaise
debris avalanche deposits
Co
ul
2120 S
oi
rD
ol
to
n
tro
Ci
nte lets
Pe Ga
5550 E
Cliff
(b)
10
km
5615 E
Debris avalanche deposits
extension limit
Blocks
alignment
Figure 22 Simplified morphostructural sketch of submarine east flanks of Piton de la Fournaise: (a) Shaded relief bathymetric map; and (b)
morphostructural sketch. Av: debris avalanche; F.O: Western Formation. Isobaths: 250 m. Projection Lat/Long WGS84. Modified from Saint-Ange,
F., 2009. La sedimentation volcanicastique en contexte de point chaud (le de La Reunion, Ocean Indien). Ph.D. thesis, Universite de La Reunion,
France, 279 pp.
RFZ NE
50
0
RFZ NE
225
00
2112 S
1
500
fl
c
R
c
fl
Ch
500
2116 S
00
Plateau
sous-marin Est
Pe
n
te
Lave flows
ot
ro
RFZ-SE
RFZ-SE
(a)
ent
inc
lv
ena
Channels
ga
let
Plateau Est
recent DAD
2
km
5554 E
5548 E
RFZ NE
50
RFZ NE
00
2112 S
500
fl
R
c
fl
500
2116 S
00
RFZ-SE
RFZ-SE
2
km
(b)
5548 E
5554 E
Figure 23 Proximal submarine area of the Piton de la Fournaise (data ERODER 1 survey). (a) 25-m resolution contour map. (b) 25-m
resolution sonar imaging map; white acoustic facies: low reflective, black acoustic facies: highly reflective; B: Block; c: channel; fl: lava
formations, R: Ripples on edge of Chenal Vincent. Modified from Saint-Ange, F., 2009. La sedimentation volcanicastique en contexte de point
chaud (le de La Reunion, Ocean Indien). Ph.D. thesis, Universite de La Reunion, France, 279 pp. Projection Lat/Long WGS84.
including the Saint-Etienne River, which supplies sand

derived from the Piton des Neiges volcano toward the Cilaos
deep-sea fan. A detailed study of the area near the coast, at
depths of between 50 and 200 m, offers a new vision of the
transfer of sediment to the deep sea (Figure 25). This mapping
shows a detailed set of canyon heads cutting into the narrow

continental shelf that characterizes the coast of a young volcanic island. Canyon heads have different morphologies,
sometimes broad and straight, or very narrow and incised.
Some of the canyons extend back to the coast where they are
226
2115 S
2115 S
Scours
Blocks
Scours
Dunes
Gullies
Gullies
2116 S
2116 S
5550 E
Megablocks
250 500
m
555050 E
5550 E
250 500
m
555050 E
Figure 24 SAR mosaic (Fournaise 2 survey) on south of the Grand Brule. White arrows represent the flow direction. Modified from Saint-Ange,
F., 2009. La sedimentation volcanicastique en contexte de point chaud (le de La Reunion, Ocean Indien). Ph.D. thesis, Universitede La Reunion,
France, 279 pp. Projection Lat/Long WGS84.
0 125 250
500
750
1000
metres
Depth m
3
30
60
80
120
160
245
Figure 25 Very-high-resolution bathymetric DEM near La Reunion Island coast at the mouth of the St Etienne river (Cilaos Cirque Catchment
Data from BATHYBAB survey). Reproduced from Babonneau, N., Bache`lery, P., Delacourt, C., et al., 2011. Direct sediment transfer from land to
deep-sea at La Reunion Island: insights from new shallow-marine multibeam data, to be submitted to Geology.
in direct connection to the river systems (Babonneau et al.,

2011).
The Cilaos turbidite system is fed by several sources: direct
feeding by the river as shown by hyperpycnal sequences that
are recognized in sediment cores (Saint-Ange, 2009), local

slope instabilities in the coastal area triggered by waves, and
deeper local submarine instabilities (Sisavath et al., 2011). The
relationship between the activity of these turbidite systems
227
Canyon area
21 S
Proximal fan
Distal fan
R1 Volcanic ridge
Sedimentary high
Sediment waves
22 S
R1
SR
R2
R3
R4
0
53 E
54 E
25
50
100
km
55 E
Figure 26 Mosaic of aerial DEM and submarine imagery map of the turbiditic system of Cilaos. Reproduced from Sisavath, E., Babonneau, N.,
Saint-Ange, F., et al. (2011). Morphology and sedimentary architecture of a modern volcaniclastic turbidite system: the Cilaos fan, offshore La
Reunion Island. Marine Geology 288(14), 117. Projection Lat/Long WGS84.
and the largest hurricanes is not completely established, but

observations of flood impact suggest a direct sediment transfer
toward the canyon head during extreme floods.
The Cilaos deep-sea fan is the best developed and is located
southwest of La Reunion Island, off the Cilaos cirque
(Figure 26). This fan is a complex turbidite system extending
for approximately 300 km onto the abyssal plain from
the coast to the Mahanoro fracture zone, southwest of La
Reunion to a depth of 4500 m below sea level. The structure of
the Cilaos deep-sea fan is strongly defined by the preexistent
seafloor morphology. In the south of La Reunion Island,
volcanic and sedimentary features shape the ocean floor on
the abyssal plain (Deplus et al., 2007; Deplus et al., 2009).
Volcanic features consist of a series of elongated ridges (named
R1 to R4 on Figure 26) comprising volcanic crests and isolated
seamounts. The larger ridge (named SR on Figure 26) corresponds to a sedimentary accumulation on volcanic high.
Two main units form the Cilaos turbidite system: (1) the
canyon area and (2) the deep-sea fan. On the acoustic backscatter image (Figure 26), both structures are characterized by
different reflectivity. The canyons area is a very high reflectivity
facies (dark tones) near the island, whereas the deep-sea fan
forms a wide zone of low reflectivity (light tones) that extends
over the abyssal plain (Figure 26). The system, as a whole,
covers an area of 15 000 km2. The canyons fed the fan that is
divided into a lobate proximal and a channelized distal fan
(Sisavath et al., 2011). The deep canyons allow the transport of
sediment and have pronounced morphology. At the foot of
the volcanic edifice, very low relief (generally less than the
resolution of bathymetric data) marks the presence of sedimentary units of the turbidite systems (i.e., Sisavath et al., 2011
for comprehensive description). Although some channels are
visible on the bathymetry, the true extension of the turbidite
system is particularly obvious from the acoustic imagery
(Figure 26). Such turbidite fans highlight the importance
of the marine volcaniclastic sedimentation, which is too often
underestimated in the studies of the evolution of volcanic

islands.
5.9.4
Conclusion
Hot spots and LIPs are an important expression of magmatism

on the terrestrial planets, and their study is a major challenge
in interpreting the mechanisms involved in planetary interiors. The complexity of hot spot magmatism, the diversity
of their morphologies, and the difficulty of setting up instrumental networks on related volcanoes, especially in oceanic
settings, suggest that it will be many years before we more fully
understand the nature and dynamics of the geomorphology of
hot spots.
Long-lived age-progressive volcanism that builds up volcanic chains is commonly seen as the archetype of hot spot
volcanism. Because of the variability of ages, hot spot volcanic
island chains exhibit a large range of geomorphic features,
from high islands with active volcanism and young relief,
to severely eroded volcanic edifices, atolls, seamounts and
guyots. The Hawaii-Emperor chain provides one of the best
examples of hot spot volcanic island chains. Similarities exist
between volcanic edifices concerning their structure and evolution over time, but great differences can result due to
differing degrees of endogenous and exogenous activity. Processes of erosion such as spreading or mass wasting are important as they largely contribute to the shaping of hot spot
volcanoes. The evolution of basaltic shield volcanoes extends
over million to tens of million of years. Phases of growth and
destruction commonly alternate. Processes contributing to the
construction of the submarine parts of the volcanic edifices are
being increasingly studied, but remain poorly understood.
Flank landslides, erosion, sediment transfer are among the
most active processes on these islands, in addition to volcanic
activity.
228
For two decades, new technologies such as development of

high-resolution imagery and dynamic surveying of volcanoes
are adding much to our knowledge. Satellite images and
measurements are allowing deformation to be quantified and
repeat mapping after each surface modification to be undertaken, as well as modeling the topography of the ocean floor.
With new instrumentation on research vessels, it is now possible to have high-resolution images of the seafloor to help
quantify the nature of submarine slopes of shield volcanoes to
study their dynamic processes.
There would be future improvements for the understanding of volcanic edifice construction and destruction with the
development of free multisource databases. Challenges will
include the development of algorithms to manage a large
volume of data in real time and to develop models that integrate the complex data sets resulting from numerous measurements and observations.
Acknowledgments
We thank Francky Saint-Ange, Nathalie Babonneau, and
Emmanuelle Sisavath, who allowed us to use several of their
figures. Data on the submarine flanks and the surrounding
abyssal plain of La Reunion (Indian Ocean) were obtained
during the FOREVER (1), ERODER (2) and BATHYBAB
(3) cruises led by C. Deplus (1), B. Savoye (2), P. Bachelery
(2), and N. Babonneau (2) (3). Reviews by Jean-Franc- ois
Lenat and Warren Huff helped to greatly improve the manuscript. And finally, many thanks to Lewis Owen for the
encouragement, patience, and invaluable help in carrying out
this work.
References
Acosta, J., Uchupi, E., Munoz, A., Herranz, P., Palomo, C., Ballesteros, M. ZEE
Working group, 2003. Geologic evolution of the Canarian Islands of Lanzarote,
Fuerteventura, Gran Canaria and La Gomera and comparison of landslides at
these islands with those at Tenerife, La Palma and El Hierro. Marine
Geophysical Researches 24, 140.
Alibes, B., Rothwell, R.G., Canals, M., Weaver, P.P.E., Alonso, B., 1999.
Determination of sediment volumes, accumulation rates and turbidite
emplacement frequencies on the Madeira Abyssal Plain (NE Atlantic): a
correlation between seismic and borehole data. Marine Geology 160, 225250.
http://dx.doi.org/10.1016/ S0025-3227(99)00026-2.
Anderson, D.L., 2005. Scoring hotspots: the plume and plate paradigms. In:
Foulger, G, Natland, J.H., Presnall, D.C., Anderson, D.L. (Eds.), Plates, Plumes,
and Paradigms, Geological Society of America Special Paper 388, pp. 3154.
Antretter, M., Riisager, P., Hall, S., Zhao, X., Steinberger, B., 2004. Modelled
palaeolatitudes for the Louisville hot spot and the Ontong Java Plateau. Origin
and Evolution of the Ontong Java Plateau. Geological Society, London, Special
Publications, 229, 2130. 10.1144/GSL.SP.2004.229.01.03.
Babonneau, N., Delacourt, C., Bache`lery, P., et al., 2011, Apport de la bathymetrie
haute-resolution a` la transition Terre-Mer des grands syste`mes sedimentaires de
La Reunion. ASF Annual Conference, Association des Sedimentologues Franc- ais,
Dijon, France.
Bache`lery, P., 1981. Le Piton de la Fournaise (Ile de la Reunion). Etude
volcanologique, structurale et petrologique. Ph.D. thesis, Univ. Clermont Ferrand
II, France.
Bache`lery, P., Coudray, J., 1993. Carte volcano-tectonique (1/50000e) de la Grande
Comore et notice explicative, Ministe`re Franc- ais Departement de la Cooperation,
St. Denis de la Reunion.
Bachelery, P., Labazuy, P., Lenat, J.-F., 1996. Submarine and subaerial debris
avalanches deposits in Reunion island. Comptes Rendus de lAcademie des
Sciences Serie II, Fascicule A Sciences de la Terre et des Plane`tes 323,

475482.
Bache`lery, P., Michon, L., 2010. The Enclos caldera: how did it form? Proceedings
of the IAVCEI Third Workshop on Collapse Calderas, La Reunion, pp. 105108
Bache`lery, P., Robineau, B., Courteaud, M., Savin, C., 2003. Avalanches de debris
sur le flanc occidental du volcan bouclier Piton des Neiges (Reunion). Bulletin
Societe Geolologique France 174, 125140. http://dx.doi.org/10.2113/174.2.125.
Batiza, R., Fornari, D.J., Vanko, D.A., Lonsdale, P., 1984. Craters, Calderas, and
Hyaloclastites on Young Pacific Seamounts. Journal of Geophysical Research
89, 83718390. http://dx.doi.org/10.1029/JB089iB10p08371.
Baxter, A.N., 1972. Magmatic evolution of Mauritius Western Indian Ocean.
Unpublished Ph.D. thesis, Edinburgh, Scotland, In Saddul, P, 1995.
Binard, N., Hekinian, R., Cheminee, J.L., Searle, R.C., Stoffers, P., 1991.
Morphological and structural studies of the Society and Austral hotspot regions
in the South Pacific. Tectonophysics 186, 293312.
Bonatti, E., Harrison, C.G.A., Fisher, D.E., Honnorez, J., Schilling, J.G., Stipp, J.J.,
Zentilli, M., 1977. Easter volcanic chain (south east Pacific): a mantle hot line.
Bret, L., Fevre, Y., Join, J.-L., Robineau, B., Bachelery, P., 2003. Deposits related to
degradation processes on Piton des Neiges Volcano (Reunion Island): overview
and geological hazard. Journal of Volcanology and Geothermal Research 123,
2541.
Bryan, S.E., Ernst, R.E., 2008. Revised definition of Large Igneous Provinces (LIPs).
Earth-Science Reviews 86, 175202.
Bryan, S.E., Peate, I.U., Peate, D.W., et al., 2010. The largest volcanic eruptions on
Earth. Earth-Science Reviews 102, 207229.
Burke, L., Wilson, J.T., 1976. Hot spots on the earths surface. Scientific American
235, 4657.
Carracedo, J.C., 1999. Growth, structure, instability and collapse of Canarian
volcanoes and comparisons with Hawaiian volcanoes. Journal of Volcanology
and Geothermal Research Special 94, 119.
Carter, A., van Wyk de Vries, B., Kelfoun, K., Bache`lery, P., Briole, P., 2007. Pits,
rifts and slumps: the summit structure of Piton de la Fournaise. Bulletin of
Volcanology 69, 741756. http://dx.doi.org/10.1007/s00445-006-0103-4.
Casalbore, D., Romagnoli, C., Bosman, A., Chiocci, F., 2010. Potential tsunamigenic
landslides at Stromboli Volcano (Italy): insight from marine DEM analysis.
Geomorphology. doi:10.1016/j.geomorph.2010.10.026.
Chadwick, J.R.W.W., Jonsson, S., Geist, D.J., et al., 2011. The May 2005 eruption
of Fernandina volcano, Galapagos: the first circumferential dike intrusion
observed by GPS and InSAR. Bulletin of Volcanology 73, 679697.
Chadwick, Jr W.W., Howard, K.A., 1991. The pattern of circumferential and radial
eruptive fissures on the volcanoes of Fernandina and Isabela islands, Galapagos.
Bulletin of Volcanology 53, 259275. http://dx.doi.org/10.1007/BF00414523.
Chevallier, L., Vatin-Perignon, N., 1982. Volcano-structural evolution of Piton des
Neiges, Reunion Island. Bulletin of Volcanology 45, 285298.
Clague, D.A., Dalrymple, G.B., 1987. The Hawaiian-Emperor volcanic chain, part 1:
geologic evolution. U.S. Geological Professional Paper 1350, pp. 554
Clement, J.-P., Legendre, C., Caroff, M., Guillou, H., Cotten, J., Bollinger, C.,
Guille, G., 2003. Epiclastic deposits and horseshoe-shaped calderas in Tahiti
(Society Islands) and Ua Huka (Marquesas Archipelago), French Polynesia.
Journal of Volcanology and Geothermal Research 120, 87101.
Clouard, V., Bonneville, A., 2001. How many Pacific hotspots are fed by deepmantle plumes? Geological Society of America 29, 695698.
Clouard, V., Bonneville, A., 2005. Ages of seamounts, islands and plateaus on the
Pacific Plate. In: Foulger, G.R., Natland, J.H., Presnall, D., Anderson, D.L. (Eds.),
Geological Society of America Special Paper 388: Plates, Plumes & Paradigms.
Geological Society of America, Boulder, CO, pp. 7190.
Clouard, V., Bonneville, A., Gillot, P.Y., 2001. A giant landslide on the southern
flank of Tahiti Island, French Polynesia. Geophysical Research Letters 28,
22532256.
Cochonat, P., Lenat, J.F., Bache`lery, P., et al., 1990. Importance des depots
gravitaires dans la mise dun syste`me volcano-sedimentaire sous-marin (volcan
de la Fournaise, Ile de la Reunion). Compte Rendu de lAcademie des Sciences
de Paris 311(Serie II), 679686.
Coffin, M.F., Eldholm, O., 1994. Large igneous provinces: crustal structure,
dimensions, and external consequences. Reviews of Geophysics 32, 136.
Coffin, M.F., Eldholm, O., 2005. Large igneous provinces. In: Selley, R.C., Cocks,
R., Plimer, I.R. (Eds.), Encyclopedia of Geology. Elsevier, Oxford, pp. 315323.
Cole, J.W., Milner, D.M., Spinks, K.D., 2005. Calderas and caldera structures: a
review. Earth-Science Reviews 69, 126.
Courtillot, V., Davaille, A., Besse, J., Stock, J., 2003. Three distinct types of
hotspots in the Earths mantle. Earth and Planetary Science Letters 205,
295308.
Cretaux, J.-F., Soudarin, L., Cazenave, A., Bouille, F., 1998. Present-day tectonic
plate motions and crustal deformations from the DORIS space system. Journal
Crough, S.T., 1978. Thermal origin of mid-plate hot-spot swells. Geophysical
Journal of the Royal Astronomical Society 55, 451469.
Crumpler, L.S., Head, J.W., Aubele, J.C., 1996. Calderas on Mars: characteritics,
structure, and associated flank deformation. In: McGuire, W.J., Jones, A.P.,
Neuberg, J. (Eds.), Volcano Instability on the Earth and other Planets.
Geological Society of London Publication, 110, pp. 307348.
Dalrymple, G.B., 1971. Potassiumargon ages from Pololu volcanic series, Kohala
Volcano. Hawaii: Geological Society of America Bulletin 82, 19972000. http://d
x.doi.org/10.1130/0016-7606(1971)82[1997:PAFTPV]2.0.CO;2.
Dalrymple, G.B., Clague, D.A., 1976. Age of the Hawaiian-Emperor bend. Earth and
Davis, A.S., Clague, D.A., 1998. Changes in the hydrothermal system at
Loihi Seamount after the formation of Peles pit in 1996. Geology 26,
399402.
Day, S.J., Heleno da Silva, S.I.N., Fonseca, J.F.B.D., 1999. A past giant lateral
collapse and present-day flank instability of Fogo, Cape Verde Islands. Journal
of Volcanology and Geothermal Research 94, 191218.
De Chazal, M., De Baissac, J., 1950. Etude sur la geologie de lle Maurice.
Proceeding Royal Society Aris Sciences of Mauritius. Proceedings of the Royal
Society of Aris Sciences of Mauritius 1, 5372.
Deniel, C., Kieffer, G., Lecointre, J., 1992. New 230Th238U and 14C age
determinations from Piton des Neiges volcano, Reunion: a revised chronology
for the differentiated series. Journal of Volcanology and Geothermal Research
51, 253267. http://dx.doi.org/10.1016/0377-0273(92)90126-X.
Deplus, C., de Voogd, B., Dyment, J., Bissessur, D., Sisavath, E., Depuiset, F.,
Mercier, M., 2009. New insights on the oceanic lithosphere at
La Reunion hotspot volcano. EGU Geophysical Research Abstracts 11-5728,
Vienna.
Deplus, C., de Voogd, B., Dyment, J., Depuiset, F., Sisavath, E., 2007. Does the
Reunion hotspot volcano emplace on a fossil ridge or a fracture zone? EOS
Transactions of the American Geophysical Union 88(52), (Fall Meeting
Supplement, Abstract V31F-05).
Deplus, C., Le Friant, A., Boudon, G., et al., 2001. Submarine evidence for largescale debris avalanches in the Lesser Antilles Arc. Earth and Planetary Science
Letters 192, 145157.
Duffield, W.A., Stieltjes, L., Varet, J., 1982. Huge landslide blocks in the growth of
piton de la Fournaise, La reunion, and Kilauea volcano, Hawaii. Journal of
Duncan, R.A., 1990. The volcanic record of the Reunion hotspot. In: Duncan, R.A.,
et al., Proceedings of the Ocean Drilling Program, College Station, Texas,
Scientific Results, vol. 115, pp. 310, 10.2973/odp.proc.sr.115.206.
Duncan, R.A., Backman, J., McDonald, A., 1989. Reunion hotspot activity through
Tertiary time: initial results from Ocean Drilling Program Leg 115. Journal of
Volcanological and Geothermal Research 36, 193198.
Duncan, R.A., McDougall, I., 1974. Migration of volcanism with time in Marquesas
Islands, French Polynesia. Earth and Planetary Science Letters 21, 414420. htt
p://dx.doi.org/10.1016/0012-821X(74)90181-2.
Duncan, R.A., Richards, M.A., 1991. Hotspots, mantle plumes, flood basalts, and
true polar wander. Reviews of Geophysics 29, 3150.
Dzurisin, D., Koyanagi, R.Y., 1984. Magma supply and storage at Kilauea volcano,
Hawaii, 19561983. Journal of Volcanology and Geothermal Research 21,
177206.
Epp, D., 1984. Possible perturbations to hotspot traces and implications for the
origin and structure of the Line Islands. Journal of Geophysical Research 89,
1127311286.
Fe`vre, Y., 2005. Mecanismes et vitesses derosion a` lechelle geologique sur une le
volcanique jeune a` relief eleve La Reunion (Ocean Indien). Thesis Universite
de la Reunion, 234 pp.
Filmer, P.E., McNutt, M.K., Webb, H.F., Dixon, D.J., 1994. Volcanism and
archipelagic aprons: a comparison of the Marquesas and Hawaiian Islands.
Marine Geophysical Research 16, 385406.
Fiske, R.S., Jackson, E.D., 1972. Orientation and growth of Hawaiian volcanic rifts,
the effects of regional structure and gravitational stresses. Proceedings of Royal
Society of London, Series A 329, 299326.
Fitton, J.G., Mahoney, P., Wallace, P.J., Saunders, A.D. (Eds.), 2004. Origin and
evolution of the Ontong Java Plateau. Geological Society of London, London,
229, 1-8. doi:10.1144/GSL.SP.2004.229.01.01.
Fornari, D.J., Garcia, M.O., Tyce, R.C., Gallo, D.G., 1988. Morphology and structure
of Loihi Seamount based on Seabeam sonar mapping. Journal of Geophysical
Research 93, 1522715238.
229
Frey, F.A., Weis, D., 1995. Temporal evolution of the Kerguelen plume: geochemical
evidence from B38 to 82 Ma lavas forming the Ninetyeast Ridge. Contributions
to Mineral Petrology 121, 1228.
Frey, F.A., Weis, D., 1996. Reply to the Class et al. discussion of temporal
evolution of the Kerguelen plume: geochemical evidence form B38 to 82 Ma
lavas forming the Ninetyeast Ridge. Contributions to Mineral Petrology 124,
104110.
Frey, F.A., Wise, W.S., Garcia, M.O., West, H., Kwon, S.T., Kennedy, A., 1990.
Evolution of Mauna Kea Volcano, Hawaii: petrologic and geochemical
constraints on postshield volcanism. Journal of Geophysical Research 95,
12711300. http://dx.doi.org/10.1029/JB095iB02p01271.
Funck, T., Schmincke, H.-U., 1998. Growth and destruction of Gran Canaria
deduced from seismic refraction and bathymetric data. Journal of Geophysical
Research 103, 1539315407.
Gailler, L.-S., Lenat, J.-F., 2010. Three-dimensional structure of the submarine
flanks of La Reunion inferred from geophysical data. Journal of Geophysical
Research 115(B12105), http://dx.doi.org/10.1029/2009JB007193.
Garcia, M.O. ,1993. PliocenePleistocene volcanic sands from Site 842: products of
giant landslides. In: Wilkens, R., Firth, J., Bender, J. (Eds.), Proceeding of the
Ocean Drilling Program, College Station, TX, Science Results, vol. 136,
pp. 5363.
Garcia, M.O., Hull, D.M., 1994. Turbidites from giant Hawaiian landslides results
from Ocean Drilling Program site-842. Geology 22, 159162.
Garcia, M.O., Sherman, S.B., Moore, G.F., Goll, R., Popova-Goll, I., Natland, J.H.,
Acton, G., 2006. Frequent landslides from Koolau Volcano: results from
ODP Hole 1223A. Journal of Volcanology and Geothermal Research 151,
251268.
Garcin, M., Poisson, B., Pouget, R., 2005. High rates of geomorphological
processes in a tropical area: the Remparts River case study (Reunion Island,
Indian Ocean). Geomorphology 67(3-4), 335.
Geist, D., Fornari, D.J., Kurz, M.D., Harpp, K.S., Soule, S.A., Perfit, M.R., Koleszar,
A.M., 2006. Submarine Fernandina: magmatism at the leading edge of the
Galapagos hot spot. Geochemistry, Geophysics, Geosystems 7, Q12007. http://d
x.doi.org/10.1029/2006GC001290.
Geshi, N., Shimano, T., Chiba, T., Nakada, S., 2002. Caldera collapse during
the 2000 eruption of Miyakejima Volcano Japan. Bulletin of Volcanology 64,
5568.
Gibson, S.A., Thompson, R.N., Day, J.A., Humphris, S.E., Dickin, A.P., 2005.
Melt-generation processes associated with the Tristan mantle plume:
constraints on the origin of EM-1. Earth and Planetary Science Letters 237,
744767.
Gillot, P.Y., Lefevre, J.C., Nativel, P.E., 1994. Model for the structural evolution of
the volcanos of Reunion Island. Earth and Planetary Science Letters 122,
291302.
Gillot, P.Y., Nativel, P.E., 1982. KAr chronology of the ultimate activity of piton des
Neiges volcano, Reunion Island, Indian Ocean. Journal of Volcanology and
Gillot, P.Y., Nativel, P.E., 1989. Eruptive history of Piton de la Fournaise volcano,
Reunion Island, Indian Ocean. Journal of Volcanology and Geothermal Research
36, 5365. http://dx.doi.org/10.1016/0377-0273(89)90005-X.
Guest, J.E., Gaspar, J.L., Cole, P.D., et al., 1999. Volcanic geology of Furnas
Volcano, Sao Miguel, Azores. Journal of Volcanology and Geothermal Research
92, 129.
Hagen, R.A., Baker, N.A., Naar, D.F., Hey, R.N., 1990. A SeaMarc II survey of recent
submarine volcanism near Easter Island. Marine Geophysical Researches 12,
297315. http://dx.doi.org/10.1007/BF02428200.
Hildenbrand, A., Gillot, P.Y., Le Roy, I., 2004. Volcano-tectonic and geochemical
evolution of an oceanic intra-plate volcano: Tahiti-Nui (French Polynesia). Earth
Holcomb, R., Searle, R., 1991. Large landslides from oceanic volcanoes. Marine
Geotechnology 10, 1932. http://dx.doi.org/10.1080/10641199109379880.
Ito, G., McNutt, M., Gibson, R.L., 1995. Crustal structure of the Tuamotu Plateau,
151S: implications for its origin. Journal of Geophysical Research 100,
80978114.
Ito, G., Van Keken, P.E., 2007. Hotspots and melting anomalies. In: Volume 7,
(David Bercovici (Ed.), vol. 7, Schubert, G. (Ed.), Treatise on Geophysics,
Elsevier.
Join, J.L., Folio, J.L., Robineau, B., 2005. Aquifers and groundwater within
active shield volcanoes. Evolution of conceptual models in the Piton de la
Fournaise volcano. Journal of Volcanology and Geothermal Research 147,
187201.
Jones, S.M., White, N., Maclennan, J., 2002. V-shaped ridges around Iceland:
implications for spatial and temporal patterns of mantle convection.
230
Geochemistry, Geophysics, Geosystems 3, 1059. http://dx.doi.org/10.1029/

2002GC000361.
Kanamatsu, T., McMurtry, G.M., Herrero-Bervera, E., 1998. Stratigraphy and
sedimentology of the Nuuanu (Oahu) and Wailau (Molokai) giant submarine
landslides, Hawaii. EOS Transactions of the American Geophysical Union
79.(Abstract F1022).
Karatson, D., Thouret, J.-C., Moriya, L., Lomoschitz, A., 1999. Erosion calderas:
origins, processes, structural and climatic control. Bulletin of Volcanology 61,
174193.
Kokelaar, P., Romagnoli, C., 1995. Sector collapse, sedimentation and clast
population evolution at an active island-arc volcano: Stromboli, Italy. Bulletin of
Volcanology 57, 240262. http://dx.doi.org/10.1007/s004450050092.
Krastel, S., Schmincke, H.-U., Jacobs, C.L., Rihm, R., Le Bas, T.P., Alibes, B., 2001.
Submarine landslides around the Canary Islands. Journal of Geophysical
Research 106, 39773997.
Labazuy, P., 1996. Recurrent landslides events on the submarine flank of Piton de
la Fournaise volcano (Reunion Island). In: McGuire, W.J., Jones, A.P., Neuberg,
J. (Eds.), Volcano Instability on the Earth and Other Planets, Geological Society,
London, Special Publications, vol. 110, pp. 295306.
Lacroix, A., 1936. Le volcan actif de lle de la Reunion et ses Produits. Paris,
Gauthier et Villard ed. 297 pp.
Lamb, M.P., Howard, A.D., Dietrich, W.E., Perron, J.T., 2007. Formation of
amphitheater-headed valleys by waterfall erosion after large-scale slumping on
Hawaii. Geological Society of America Bulletin 119, 805822. http://dx.doi.org/
10.1130/B25986.1.
Laughton, A.S., Matthews, D.W., Fisher, R.L., 1970. The structure of the Indian
Ocean. In: Maxwell, A.E. (Ed.), The Sea. John Wiley and Sons, New York, vol.
4, pp. 543586.
Le Friant, A., Boudon, G., Arnulf, A., Roberson, R.E.A., 2009. Debris avalanche
deposits offshore St. Vincent (West indies): impact of flank-collapse events on
the morphological evolution of the island. Journal of Volcanology and
Le Friant, A., Lebas, E., Clement, V., Boudon, G., Deplus, C., de Voogd, B.,
Bachelery, P., 2011. A new model for the evolution of La Reunion volcanic
complex from complete marine geophysical surveys. Geophysical Research
Letters 38.http://dx.doi.org/10.1029/2011GL047489.
Lenat, J.-F., Bache`lery, P., 1987. Dynamics of magma transfers at Piton de la
Fournaise, a model based on 5 years of monitoring and on the historic
activity. Abstract, Hawaii Symposium on How Volcanoes Work, January 1987,
153 pp.
Lenat, J.-F., Boivin, P., Deniel, C., Gillot, P.Y., Bachelery, P. Fournaise 2 Team,
2009. Age and nature of deposits on the submarine flanks of Piton de la
Fournaise (Reunion Island). Journal of Volcanology and Geothermal Research
184, 199207.
Lenat, J.-F., Vincent, P., Bache`lery, P., 1989. The off-shore continuation of an active
basaltic volcano: Piton de La Fournaise (Reunion island, Indian Ocean);
structural and geomorphological interpretation from sea beam mapping. Journal
of Volcanological and Geothermal Research 36, 136.
Lenat, J.-F., Bache`lery, P., Bonneville, A., Galdeano, A., Labazuy, P., Rousset, D.,
Vincent, P., 1990. Structure and morphology of the submarine flank of an active
basaltic volcano: Piton de La Fournaise (Reunion Island, Indian Ocean).
Oceanologica Acta 10, 211223.
Lipman, P.W., 2000. Calderas. In: Sigurdsson, H., Houghton, B.F., Rymer, H., Stix,
J., McNutt, S. (Eds.), Encyclopedia of Volcanoes. Academic Press, San
Francisco, pp. 643662.
Lipman, P.W., Lockwood, J.P., Okamura, R.T., Swanson, D.A., Yamashita, K.M.,
1985. Ground deformation associated with the 1975 magnitude-7.2 earthquake
and resulting changes in activity of Kilauea volcano 19751977. US Geolgical
Survey Professional Paper 1276, pp. 145.
Lipman, P.W., Normark, W.R., Moore, J.G., Wilson, J.B., Gutmacher, C., 1988. The
giant submarine Alika debris slide, Mauna Loa, Hawaii. Journal of Geophysical
Research 93, 42794299.
Lithgow-Bertelloni, C., Richards, M.A., 1998. The dynamics of Cenozoic and
Mesozoic plate motions. Review of Geophysical Research 36, 2778.
Lockwood, J.P., Hazlett, R.W., 2010. Volcanoes; Global Perspectives. WileyBlackwell Publication, 539 pp.
Lonsdale, P., Spiess, F.N., 1979. Pair of young cratered volcanos on the East
Pacific rise. Journal of Geology 87, 157173.
Louvat, P., Allegre, C.J., 1997. Present denudation rates on the island of Reunion
determined by river geochemistry: basalt weathering and mass budget between
chemical and mechanical erosions. Geochimica et Cosmochimica Acta 61,
36453669.
MacDonald, G.A., 1965. Hawaiian calderas. Pacific Sciences 19, 320334.
MacDonald, G.A., 1972. Volcanoes. Prentice-Hall, Inc., Englewood Cliffs, NJ, USA,
510 pp.
MacDonald, G.A., Katsura, T., 1964. Chemical composition of Hawaiian lavas.
Journal of Petrology 5, 82133.
Macnab, R., Verhoef, J., Roest, W., Arkani-Hamed, J., 1995. New database
documents the magnetic character of the Arctic and North Atlantic. EOS
Transactions of the American Geophysical Union 76(45), 449.
Mahoney, J., Duncan, R., Khan, W., Gnos, E., 2002. Cretaceous volcanic rocks
of the South Tethyan suture zone, Pakistan: implications for the Reunion
hotspot and Deccan Traps. Earth and Planetary Science Letters 203,
295310.
Malahoff, A., 1987. Geology of the summit of Loihi submarine volcano. US
Geological Survey Professional Paper 1350, pp. 133144.
Malahoff, A., McMurtry, M.G., Wiltshire, J.C., Yeh, H.W., 1982. Geology and
chemistry of hydrothermal deposits from active submarine volcano Loihi, Hawaii.
Nature 298, 234239. http://dx.doi.org/10.1038/298234a0.
Marti, J., Mitjavila, J., Arana, V., 1994. Stratigraphy, structure and geochronology of
the Las Canadas caldera (Tenerife, Canary Islands). Geological Magazine 131,
715727.
Massin, F., Ferrazzini, V., Bachelery, P., Nercessian, A., Duputel, Z., Staudacher, T.,
2011. Structures and evolution of the plumbing system of Piton de la Fournaise
volcano inferred from clustering of 2007 eruptive cycle seismicity. Journal of
Volcanology and Geothermal Research 202, 96106. http://dx.doi.org/10.1016/
j.jvolgeores.2011.01.008.
Masson, D.G., 1996. Catastrophic collapse of the volcanic island of Hierro
15 ka ago and the history of landslides in the Canary Islands. Geology 24,
231234.
Masson, D.G., Watts, A.B., Gee, M.J.R., Urgeles, R., Mitchell, N.C., Le Bas, T.P.,
Canals, M., 2002. Slope failures on the flanks of the western Canary Islands.
Earth-Science Reviews 57, 135. http://dx.doi.org/10.1016/S00128252(01)00069-1.
McDougall, I., 1971. The chronology and evolution of the young volcanic island of
Reunion (Indian Ocean). Geochimica and Cosmochimica Acta 35, 261288. http
://dx.doi.org/10.1016/0016-7037(71)90037-8.
McDougall, I., Chamalaun, F.G., 1969. Isotopic dating and geomagnetic polarity
studies on volcanic rocks from Mauritius, Indian Ocean. Geological Society of
American Bulletin 80, 14191442.
McDougall, I., Swanson, D.A., 1972. Potassiumargon ages of lavas from Hawi and
Pololu volcanic series, Kohala Volcano, Hawaii. Geological Society of America
Bulletin 83, 37313737. http://dx.doi.org/10.1130/0016-7606.
McKenzie, D.P., Sclater, J.G., 1971. The evolution of the Indian ocean since the late
Cretaceous. Geophysical Journal of the Royal Astronomical Society 25,
437528.
McNutt, M.K., Caress, D.W., 2007. Crust and lithospheric structure hot spots and
hot-spot swells. Treatise on Geophysics, 445478.
McNutt, M.K., Fischer, K.M., 1987. The South Pacific superswell. In: Keating, B.H.,
Fryer, P., Batiza, R., Boehlert, G.W. (Eds.), Seamounts, Islands, and Atolls,
Geophysical Monograph Series. American Geophysical Union, Washington, DC,
vol. 43, pp. 2534.
Merle, O., Lenat, J.-F., 2003. Hybrid collapse mechanism at Piton de la Fournaise
Volcano, Reunion Island, Indian Ocean. Journal of Geophysical Research 108,
2166.
Merle, O., Mairine, P., Michon, L., Bache`lery, P., Smietana, M., 2010. Calderas,
landslides and paleo-canyons on Piton de la Fournaise volcano (La Reunion
Island, Indian Ocean). Journal of Volcanological and Geothermal Research 189,
131142.
Merle, O., Michon, L., Bache`lery, P., 2008. Caldera rim collapse: a hidden volcanic
hazard. Journal of Volcanological and Geothermal Research 177, 525530. http:
//dx.doi.org/10.1016/j.jvolgeores.2008.06.011.
Michon, L., Massin, F., Famin, V., Ferrazzini, V., Roult, G., 2011. Basaltic calderas:
collapse dynamics, edifice deformation, and variations of magma withdrawal.
Michon, L., Saint-Ange, F., 2008. Morphology of Piton de la Fournaise basaltic
shield volcano (La Reunion Island): characterization and implication in the
volcano evolution. Journal of Geophysical Research 113. http://dx.doi.org/
10.1029/2005JB004118.
Michon, L., Th, Staudacher, Ferrazzini, V., Bache`lery, P., Marti, J., 2007. April 2007
collapse of Piton de la Fournaise: a new example of caldera formation.
Geophysical Research Letters 34. http://dx.doi.org/10.1029/2007GL031248.
Michon, L., Villeneuve, N., Catry, T., Merle, O., 2009. How summit calderas
collapse on basaltic volcanoes: new insights from the April 2007 caldera
collapse of Piton de la Fournaise. Journal of Volcanological and Geothermal
Research 184(1-2), 138151.
Mitchell, N.C., 2003. Susceptibility of mid-ocean ridge volcanic islands and

seamounts to large-scale landsliding. Journal of Geophysical Research 108. http
://dx.doi.org/10.1029/2002JB001997.
Mitchell, N.C., Dade, W.B., Masson, D.G., 2003. Erosion of the submarine flanks of
the Canary Islands. Journal of Geophysical Research 108. http://dx.doi.org/
10.1029/2002JF000003.
Mitchell, N.C., Masson, D.G., Watts, A.B., Gee, M.J.R., Urgeles, R., 2002. The
morphology of the submarine flanks of volcanic ocean islands. A comparative
study of the Canary and Hawaiian hotspot islands. Journal of Volcanology and
Mohr, P., Zanettin, B., 1988. The Ethiopian flood basalt province. In: Macdougall,
J.D. (Ed.), Continental Flood Basalts. Kluwer Academic Publishing, pp. 63110.
Moore, J.G., 1964. Giant submarine landslides on the Hawaiian ridge. US
Geological Survey Professional Paper 501-D, pp. 9598.
Moore, J.G., Clague, D.A., 1992. Volcano growth and evolution of the island of
Hawaii. Geological Society of America Bulletin 104, 14711484. http://dx.doi.org
/10.1130/0016-7606.
Moore, J.G., Clague, D.A., Holcomb, R.T., Lipman, P.W., Normark, W.R., Torresan,
M.E., 1989. Prodigious submarine landslides on the Hawaiian Ridge. Journal of
Moore, J.G., Moore, G.W., 1984. Deposit from a giant wave on the island of Lanai,
Hawaii. Science 226, 13121315.
Moore, J.G., Normark, W.R., Holcomb, R.T., 1994. Giant Hawaiian landslides.
Morgan, W.J., 1971. Convection plumes in the lower mantle. Nature 230, 4243.
Morgan, W.J., 1972. Plate motions and deep mantle convection. In : Shagam, R.,
Hargraves, R.B., Morgan, W.J., Van Houten, F.B., Burk, C.A., Holland, H.D.,
Hollister, L.C. (Eds.), Studies in Earth and Space Sciences (Hess Volume),
Geological Society of America Memoir 132, pp. 722.
Naka, J., Tsugaru, R., Danhara, T., Tanaka, T., Fujioka, K., 1993. Sedimentary
processes of volcaniclastic sediments, Leg 136. In : Wilkens, R.H., Firth, J.V., et
al. (Eds.), Proceedings of the Ocean Drilling Program, College Station, TX,
Science Results, vol. 136, pp. 8595.
Naumann, T., Geist, D., 2000. Physical volcanology and structural development of
Cerro Azul Volcano, Isabela Island, Galapagos: implications for the development
of Galapagos-type shield volcanoes. Bulletin of Volcanology 61, 497514.
Nigel, R., Rughooputh, S., 2010. Mapping of monthly soil erosion risk of mainland
Mauritius and its aggregation with delineated basins. Geomorphology 114,
101114. http://dx.doi.org/10.1016/j.geomorph.2009.06.013.
Niu, Y., 2004. Origin of the 43 Ma Bend Along the Hawaiian-Emperor Seamount
Chain: problem and Solution. In: Hekinian, R., Stoffers, P., Cheminee, J.-L.
(Eds.), Oceanic Hotspots. Springer, New York.
OConnor, J.M., Duncan, R.A., 1990. Evolution of the Walvis Ridge-Rio Grande Rise
hotspot system: implications for African and South American plate motions over
plumes. Journal of Geophysical Research 95, 1747517502.
OConnor, J.M., Roex, A.P.L., 1992. South Atlantic hot spot-plume systems: 1.
Distribution of volcanism in time and space. Earth and Planetary Science Letters
113, 343364.
OConnor, J.M., Stoffers, P., McWilliams, M.O., 1995. Timespace mapping of
Easter chain volcanism. Earth and Planetary Science Letters 136, 197212.
Oehler, J.-F., 2005. Les destabilisations de flanc des volcans de lIle de la Reunion
(Ocean Indien): mise en evidence, implications et origines. Ph.D. thesis, Univ.
Clermont-Ferrand II, Clermont-Ferrand, France, 422 pp.
Oehler, J.-F., Labazuy, P., Lenat, J.-F., 2004. Recurrence of major flank landslides
during the last 2 Ma-history of Reunion Island. Bulletin of Volcanology 66,
585598.
Oehler, J.-F., Lenat, J.-F., Labazuy, P., 2007. Growth and collapse of the Reunion
Island volcanoes. Bulletin of Volcanology 70, 717742. http://dx.doi.org/
10.1007/s00445-007-0163-0.
Ollier, C., 1988. Volcanoes. Blackwell, Oxford, 228 pp.
Ollier, C., 1995. Tectonics and landscape evolution in southeast Australia.
Ollier, G., Cochonat, P., Lenat, J.-F., Labazuy, P., 1998. Deep-sea volcaniclastic
sedimentary systems: an example from La Fournaise volcano, Reunion Island,
Indian Ocean. Sedimentology 45, 293330.
Patriat, P., Achache, J., 1984. IndiaEurasia collision chronology has implications
for crustal shortening and driving mechanismof plate. Nature 311, 615621.
Peate, D.W., 1997. The ParanaEtendeka provinces. In: Mahoney, J.J., Coffin, M.F.
(Eds.), Large Igneous Provinces: Continental, Oceanic, and planetary Flood
Volcanism. American Geophysical Union, Washington, DC, pp. 145182.
Peltier, A., Staudacher, T., Bache`lery, P., Cayol, V., 2009. Formation of the April
2007 caldera collapse at Piton de La Fournaise volcano: insights from GPS
data. Journal of Volcanology and Georthermal Research. (Special issue: Piton de
231
La Fournaise).http://dx.doi.org/10.1016/j.jvolgeores.2008.09.009 (Special issue:

Piton de La Fournaise).
Perroud, B., 1982. Etude volcano-structurale des les Maurice (ocean Indien
occidental): origine du volcanisme. The`se, Univ. Sci. Et Med. Grenoble,
218 pp.
Peterson, D.W., Moore, J.G., (1987). Geologic history and evolution of geologic
concepts, Island of Hawaii. United States Geological Society Professional Paper
1350 pp., 149-189.
Porter, S.C., Stuiver, M., Yang, I.C., 1977. Chronology of Hawaiian glaciations.
Science 195, 6163. http://dx.doi.org/10.1126/science.195.4273.61.
Proag, V., 1995. The Geology and Water Resources of Mauritius. Mahatma Ghandi
Institute Press, Mauritius.
Rad, S.D., Alle`gre, C.J., Louvat, P., 2007. Hidden erosion on volcanic islands. Earth
Rappaport, Y., Naar, D.F., Barton, C.C., Liu, Z.J., Hey, N., 1997. Morphology and
distribution of seamounts surrounding Easter Island. Journal of Geophysical
Research 102(24), pp. 713724, 728.
Richards, M.A., Duncan, R.A., Courtillot, V.E., 1989. Flood basalts and hot-spot
tracks: plume heads and tails. Science 246, 103107.
Roche, O., van Wyk de Vries, B., Druitt, T.H., 2001. Sub-surface structures and
collapse mechanisms of summit pit craters. Journal of Volcanology and
Rocher, P., Westercamp, D., 1989. The salazie-cirque ignimbrite (Piton-des-Neiges
volcano, Reunion-island) chronostratigraphy, description and significance of
lithic fragments and eruptive mechanisms. Journal of Volcanology and
Geothermal Research 36, 177191. http://dx.doi.org/10.1016/03770273(89)90012-7.
Romagnoli, C., Casalbore, D., Chiocci, F.L., Bosman, A., 2009a. Offshore
evidence of large scale lateral collapses on the eastern flank of Stromboli, Italy,
due to structurally controlled, bilateral flank instability. Marine Geology 262,
113.
Romagnoli, C., Kokelaar, P., Casalbore, D., Chiocci, F.L., 2009b. Lateral collapses
and active sedimentary processes on the northwestern flank of Stromboli
volcano, Italy. Marine Geology 265, 101119.
Royer, J.-Y., Colin, M.F., 1992. Jurassic to Eocene plate tectonic reconstructions in
the Kerguelen Plateau region. In: Wise, S.W., et al. (Eds.), Proceedings of the
Ocean Drilling Program Science Research. ODP, College Station, TX, vol. 120,
pp. 917928.
Ryan, M.P., 1987. Elasticity and contractancy of Hawaiian olivine tholeiite and its
role in the stability and structural evolution of subcaldera magma reservoirs
and rifts systems. US Geological Survey Professional Paper 1350,
13951447.
Saddul, P., 1995. Mauritius: A Geomorphological Analysis. Mahatma Ghandi
Institute Press, Mauritius, 340 pp.
Saint-Ange, F., 2009. La sedimentation volcaniclastique en contexte de point chaud
(le de La Reunion, Ocean Indien). Ph.D. thesis, Universite de La Reunion,
France, 279 pp.
Saint-Ange, F., Savoye, B., Michon, L., et al., 2011. A volcaniclastic deep-sea fan
off La Reunion Island (Indian Ocean): gradualism versus catastrophism. Geology
39, 271274.
Schmincke, H.-U., 2006. Volcanism. Springer, 324 pp.
Schmincke, H.-U., Sumita, M., 1998. Volcanic evolution of Gran Canaria
reconstructed from apron sediments: synthesis of VICAP project drilling. In :
Weaver, P.P.E., Schmincke. H.-U., Firth. J.V., Duffield. W. (Eds.), Proceedings of
the Ocean Drilling Program, College Station, Texas, Scientific Results, vol. 157,
pp. 443469.
Schneider, J.-L., Brunner, C.A., Kuttner, S., 1998. Epiclastic sedimentation during
the upper Miocenelower Pliocene volcanic hiatus of Gran Canaria: evidence
from Sites 953 and 954. In : Weaver, P.P.E., Schmincke, H.-U., et al. (Eds.),
Proceedings of the Ocean Drilling Program, College Station, TX, Scientific
Results, vol. 157, pp. 293313.
Sclater, J.G., Fisher, R.L., 1974. Evolution of the East Central Indian Ocean, with
emphasis on the tectonic setting of the Ninetyeast Ridge. Geological Society of
Sclater, J.G., Fisher, R.L., Patriat, P., Tapscott, C., Parsons, B., 1981. Eocene to
recent development of the southwest Indian Ridge, a consequence of the
evolution of the Indian Ocean triple junction. Geophysical Journal of the Royal
Astronomical Society 64, 587604.
Sentenac, R., 1964. Recherches deau souterraine a` lle Maurice, MSIRI Occasional
Paper 17, pp. 1-21.
Siebert, L., 1984. Large volcanic debris avalanches: characteristics of source areas,
deposits, and associated eruptions. Journal of Volcanology and Geothermal
232
Simkin, T., Howard, K.A., 1970. Caldera collapse in Galapagos Islands, 1968.
Science 169, 429437.
Simpson, E.S.W., 1950. The geology and mineral resources of Mauritius. Colonial
Geology and Mineral Resources 1, 217235.
Sisavath, E., Babonneau, N., Saint-Ange, F., et al., 2011. Morphology and sedimentary
architecture of a modern volcaniclastic turbidite system: the Cilaos fan, offshore La
Reunion Island. Marine Geology. doi:10.1016/j.margeo.2011.06.011.
Sleep, N.H., 1990. Hotspots and mantle plumes: some phenomenology. Journal of
Sleep, N.H., 2003. Simple features of mantle-wide convection and the interpretation
of lower-mantle tomograms. Comptes rendus Geoscience 335, 922.
Smietana, M., Bache`lery, P., Hemond, C., 2010. Heterogeneity in the mantle source
of La Reunion Island. Geochemica and Cosmochemica Acta 74(12 Suppl. 1),
A972.
Smoot, N.C., 1982. Guyots of the Mid-emperor chain mapped with multibeam
sonar. Marine geology 47, 153163.
Staudacher, T., Ferrazzini, V., Peltier, A., et al., 2009. The April 2007 eruption and
the Dolomieu crater collapse, two major events at Piton de la Fournaise (La
Reunion Island, Indian Ocean). Journal of Volcanological and Geothermal
Research 184, 126137. http://dx.doi.org/10.1016/j.jvolgeores.2008.11.005.
Staudigel, H., Schmincke, H.-U., 1984. The Pliocene Seamount series of La Palma,
Canary Islands. Journal of Geophysical Research 89, 11951215.
Steinberger, B., 2000. Plumes in a convecting mantle: models and observations for
individual hotspots. Journal of Geophysical Research 105, 1112711152.
Steinberger, B., 2002. Motion of the Easter Hot Spot relative to Hawaii and
Louisville hot spots. Geochemistry, Geophysics, Geosystems 3, 8503. http://dx.d
oi.org/10.1029/2002GC000334.
Stieltjes, L., 1985. Coulees historiques du Piton de la Fournaise, le de la Reunion,
Ocean Indien. BRGM.
Storey, B.C., 1995. The role of mantle plumes in continental breakup: case histories
from Gondwana. Nature 377, 301308.
Swanson, D.A., Duffield, W.A., Fiske, R.S., 1976. Displacement of the south flank of
Kilauea Volcano: the result of forceful intrusion of magma into the rift zone. US
Geological Survey Professional Paper 963, 39 pp.
Thordarson, T., 2004. Accretionary-lapilli-bearing pyroclastic rocks at ODP Leg 192
site 1184A, a record of subaerail phreatomagmatic eruptions on the Ontong
Java Plateau. In: Fitton, J.G., Mahoney, P., Wallace, P.J., Saunders, A.D. (Eds.),
Origin and Evolution of the Ontong Java Plateau. Geological Society of London,
London, 229, pp. 83111.
Thordarson, T., Hoskuldson, A., 2008. Post glacial volcanism in Iceland. Jokul 58,
197228.
Thouret, J.C., 1999. Volcanic Geomorphology an overview. Earth-Science Reviews
47, 95131.
Trofimovs, J., Amy, L., Boudon, G., et al., 2006. Submarine pyroclastic deposits
formed at the Soufriere Hills volcano, Montserrat (19952003): what happens
when pyroclastic flows enter the ocean? Geology 34, 549.
Ui, T., 1983. Volcanic dry avalanche deposits identification and comparison with
non-volcanic debris stream deposits. Journal of Volcanology and Geothermal
Ui, T., Takarada, S., Yoshimoto, M., 2000. Debris avalanches. In: Sigurdsson, H.,
Houghton, B.F., Rymer, H., Stix, J., McNutt, S. (Eds.), Encyclopedia of
Volcanoes. Academic Press, San Francisco, pp. 617625.
Upton, B.G.J., 1982. Oceanic Islands. Ocean Basins and Margins 6, 585648.
Varnes, D., 1978. In : Schuster, R.L., Krizek. R.J. (Eds.), Slope Movement Types
and Processes. Landslides, Analysis and Control. Special Report 176, National
Academy of Sciences, Washington, DC, pp. 1133.
Vezzoli, L., Acocella, V., 2009. Easter Island, SE Pacific: an end-member type of
hotspot volcanism. Geological Society of America Bulletin 121, 869886.
Vogt, P.R., Smoot, N.C., 1984. The Geisha Guyots: multibeam bathymetry and
morphometric interpretation. Journal of Geophysical Research 89,
11,08511,107. http://dx.doi.org/10.1029/JB089iB13p11085.
Wadge, G., Guest, J.E., 1981. Steady-state magma discharge at Etna 197181.
Nature 294, 548550.
Walker, G.P.L., 1973. Lengths of lava flows. Philosophical Transactions of the Royal
Society A 274, 107118.
Walker, G.P.L., 1984. Downsag calderas, ring faults, caldera sizes, and incremental
caldera growth. Journal of Geophysical Research 89, 84078416. http://dx.doi.or
g/10.1029/JB089iB10p08407.
Walker, G.P.L., 1988. Three Hawaiian calderas: an origin through loading by shallow
intrusions? Journal of Geophysical Research 93, 1477314784.
Walker, G.P.L., 1989. Spongy pahoehoe in Hawaii a study of vesicle-distribution
patterns in basalt in basalt and their significance. Bulletin of Volcanology 51,
199209. http://dx.doi.org/10.1007/BF01067956.
Walker, G.P.L., 1993. Basaltic-volcano systems. In: Prichard, H.M., Alabaster, T.,
Harris, N.B.W., Neary, C.R. (Eds.), Magmatic Processes and Plate Tectonics.
Geological Society Special Publication, 76, pp. 338.
Watts, A.B., Masson, D.G., 1995. A giant landslide on the north flank of Tenerife,
Canary Islands. Journal of Geophysical Research 100, 2448724498.
Watts, A.B., Sandwell, D.T., Smith, W.H.F., Wessel, P., 2006. Global gravity,
bathymetry, and the distribution of submarine volcanism through space and
time. Journal of Geophysical Research, 111. http://dx.doi.org/10.1029/
2005JB004083.
White, R.S., 1988. A hot-spot model for early Tertiary volcanism in the N Atlantic.
Geological Society Special Publications 39, 313.
White, R.S., 1997. Riftplume interaction in the North Atlantic. Philosophical
Transactions of the Royal Society of London, Series A 355, 319339.
Wignall, P., 2005. The link between large igneous province eruptions and mass
extinctions. Elements 1, 293297.
Williams, H., McBirney, A.R., 1979. Volcanology. Freeman, Cooper and Co., San
Francisco Publ., Berkeley, CA, 397 pp.
Wilson, J.T., 1963. A possible origin of the Hawaiian Islands. Canadian Journal of
Physics 41, 863870.
Wilson, J.T., 1973. Mantle plumes and plate motions. Tectonophysics 19,
149164.
Wolfe, E.W., Morris, J., 1996. Geologic map of the Island of Hawaii. US Geological
Survey Geologic Investigations Series Map I-2534-A, scale 1:1,000,000.
Biographical Sketch
Dr. Patrick Bachelery is a professor of volcanology at the University of La Reunion, La Reunion Island, France. His
present research interests focus on the petrology and evolution of intraplate oceanic islands. He was director of the
Piton de la Fournaise Volcanological Observatory, and participated in several marine expeditions. He has published some 60 scientific papers.
233
Dr Nicolas Villeneuve is an assistant professor in the Department of Geography and the GeoSciences Laboratory at
the University of La Reunion, La Reunion Island, France. He received his PhD in 2000 in geophysics with an
emphasis in geographic information science, geodesy, lava flow simulations, geomorphology in the volcanic field
and risk assessment at the Institut de Physique du Globe de Paris. His present research in dynamic geomorphology includes remote sensing and model methodologies applied to quantification of volcanic phenomenon,
specifically: volcanotectonics ground deformations; lava flow volume budgets; SO2 degasing budgets; and
erosion and sediment transfers from the summit of volcano to the deep sea. He has published some 20 scientific
papers.
5.10 Tectonic Geomorphology of Normal Fault Scarps

D Nash, University of Cincinnati, Cincinnati, OH, USA
5.10.1
5.10.2
5.10.3
5.10.4
5.10.5
5.10.6
References
Introduction
Basin and Range Province
Slope Retreat Versus Recline
Modeling the Decay of Transport-Limited Scarps
Limitation of the Geometric Model for Normal Fault Scarp Decay
Summary
Symbols and abbreviations

/
c
h
t
Internal friction angle [1]

Cohesion [M L1 T2]
Initial scarp height [L]
Elapsed time [T]
a
r1 , r3
Glossary
Active Rankine stress state Results when soil mass fails
as with a decrease in the lateral confining pressure causing
lateral extension. Scarps may form in unconsolidated
materials as a result of normal faulting in this stress
state.
Cohesionless material Unconsolidated materials
containing relatively little cohesive materials (e.g., clay)
composed primarily of gravel and sand. Although there are
no completely cohesionless materials, some glacial outwash
and alluvial fan gravels may be treated as cohesionless.
Composite scarp Scarps formed by repeated offset along
the same or similar normal fault shear surface.
Debris slope Apron of debris formed at the base of a
retreating free face. Debris slopes are frequently at the angle
of repose of the debris if composed primarily of
cohesionless material (e.g., sand and gravel).
Detachment-limited surface Scarp surface on which
material is removed as rapidly as it is weathered or loosened
from the free face of the scarp. Generally detachmentlimited surfaces are unvegetated.
Fan-head entrenchment Canyons eroded into the apex
and upper portions of alluvial fans. The entrenchment may
result from tectonic displacement, climate change, or other
causes.
Fault scarp A linear topographic feature resulting directly
from displacement along a fault. A fault scarp, as opposed
to a fault-line scarp is not produced by differential erosion
of rock units juxtaposed by fault movement.
235
238
239
243
246
248
248
Linear scale factor; ratio between size of two

similar landforms [1]
Slope angle at a point on hillslope or scarp [1]
Maximum and minimum principal stresses
[M L1 T2]
Fault-line scarp Scarp produced by differential erosion

along a fault.
Free face Detachment-limited scarp produced in an active
Rankine stress state that retreats back at some threshold
angle.
Geometric model Model for scarp degradation employing
an explicit relationship between scarp height and
degradation rate.
Hillslope diffusion model Model for scarp degradation
assuming that the downslope debris flux at a point on a
transport-limited scarp is linearly proportional to the
gradient of the scarp at that point and that continuity
conditions prevail. The model is similar in form to those for
heat and ground-water diffusion.
Morphologic dating Method for dating a degraded scarp
by matching its current observed morphology with the
morphology predicted by a calibrated model for scarp
degradation. It requires that the initial morphology of the
scarp be known or accurately estimated.
Multiple scarp Superposition of two or more smaller
scarps to form a single larger scarp.
Ravel Rapid downslope movement of material loosened
from the surface of a free face comprised of gravel or sand. A
sand run is a form of raveling.
Scarp degradation or decay A general term applied to
changes in the morphology of a scarp as it ages, sometimes
referred to as slope evolution. Degradation is used in
preference to erosion because the latter term is usually
associated with wash processes.
Nash, D., 2013. Tectonic geomorphology of normal fault scarps. In:

pp. 234249.
234
http://dx.doi.org/10.1016/B978-0-12-374739-6.00090-7
Tectonic Geomorphology of Normal Fault Scarps
Scarp recline The slope angle of the scarp decreases as it

degrades. These scarps are generally transport-limited. The
crestal convexity and basal concavity of simple scarps
generally also become more rounded as the scarp reclines.
Scarp retreat The scarp moves back parallel to itself as it
degrades. Detachment-limited scarps generally retreat,
progressively burying themselves with a basal debris slope.
Simple scarp A scarp consisting of an offset straight base
and crest formed in an active Rankine stress state.
235
Splintered scarp A scarp face formed by en echelon

displacements along a normal fault.
Transport-limited surface Scarp surface on which
loosened debris is produced faster than the transportational
processes are able to remove it. As a result the surface
becomes mantled with loosened debris and frequently is
vegetated.
Abstract
Scarps may be produced by all types of fault displacement. Although most fault scarps have a complex initial morphology, a
scarp produced by normal faulting of unconsolidated or poorly consolidated frictional materials (i.e., sands and gravels)
may have a relatively simple initial morphology consisting of a detachment-limited free face, sloping around 601, which
retreats back, progressively burying its base with raveled debris. Once the free-face burial is complete, the scarp crest and
base become more rounded and the scarp reclines provided the scarp remains transport limited and erosional processes are
only a function of the gradient. In some situations, scarp age may be determined by morphologic dating.
5.10.1
Introduction
This chapter examines aspects of fault scarps at the local scale.

The local scale ranges from millimeter dilatational cracks, for
example, in playa surfaces in the Mojave Desert, California
after minor seismic activity, to meter-high scarp displacements
of lava flows on Mars (not local but perhaps local scale).
A fault scarp is difficult to define. Is the head scarp of a rotational slump produced by landsliding a normal fault scarp?
Is the thrust ridge at the base of the same slump a fault scarp
produced by reverse faulting? Gravity tectonics tends to blur
the distinction between tectonic and surface processes.
In his examination of the study of geomorphology for the
period 18901965, Gerrard (2008) noted the difficulties
posed by the use of particular terms when their definition was
incomplete or unclear when they were originally introduced.
Gerrard (2008) gave the terms fault scarp and fault-line
scarp as examples. The latter term was used by Davis (1913) to
distinguish a scarp resulting from fault displacement (fault
scarp) from one produced by differential erosion along a fault
line (fault-line scarp). Even this distinction causes difficulty: is
the range that has retreated many kilometers back from the
range-front fault that produced it a fault scarp or a fault-line
(a)
(b)
scarp? The term fault scarp is used here for a linear topographic feature resulting from displacement along a fault. So
defined, a fault scarp may result from dip- or strike-slip displacement or any combination of both (Figure 1). By this
definition, the observed relief of a fault scarp may be equal to
or less than the cumulative offset along the generating fault or
faults but may not exceed it. Once produced, the scarp may
recline or retreat and still be considered a fault scarp even if
not precipitous. It is not, however, applied to linear landforms
produced by differential erosion of rock units exposed by
faulting because the observed offset of such features may be
entirely unrelated to the original fault offset (Figure 2). Differential erosion may, in fact, produce an escarpment with
relief that is the reverse of the original offset.
Range-front faulting resulting from tensional stress, although occurring around the world (and Solar System), is
particularly well displayed in the Basin and Range structural
Province of the western US. In contrast to other local fault
features, normal faulting in the Basin and Range generally
results in northsouth trending horsts and grabens. Because
the Basin and Range Province extends more than 2000 km
from the desert of central Mexico to the cool temperate climates of western Montana, it provides an ideal setting for
(c)
Figure 1 As defined here, a simple fault scarp may result from (a) normal, (b) reverse, (c) strike-slip fault displacement or any combination of
dip- or strike-slip fault movement. Reproduced from Stewart, I.S., Hancock, P.L., 1990. What is a fault scarp? Episodes 13, 256263, with
permission from International Union of Geological Sciences.
236
Figure 2 A fault scarp (A) results directly from fault displacement whereas a fault-line scarp results from differential erosion (B). Modified from
Stewart, I.S., Hancock, P.L., 1990. What is a fault scarp? Episodes 13, 256263, with permission from International Union of Geological
Sciences.
160 W 140 W 120 W 100 W 80 W 60 W 40 W 20 W
20 E
40 E
60 E
80 E 100 E 120 E 140 E 160 E 180
90
80 N
80 N
70 N
70 N
60 N
60 N
50 N
50 N
40 N
40 N
30 N
30 N
20 N
20 N
10 N
10 N
0
10 S
10 S
20 S
20 S
30 S
40 S
50 S
Cenozoic
Mesozoic
Paleozoic
Precambrian
30 S
40 S
50 S
60 S
60 S
170 W 150 W
130 W 110 W
90 W
70 W
50 W
30 W
10 W 0
20 E
40 E
60 E
80 E
100 E
120 E
140 E
160 E
180
Figure 3 World rifts mapped and inventoried by Sengor and Natalin (2001).
observing the erosion of normal fault scarps formed in similar

alluvial fan materials in a wide range of climates. Initially
scarps form in bedrock but erosion causes them to retreat back
from the generating fault so that subsequent faulting forms
scarps in the alluvial fans developed at the base of the range.
The scarps produced in coarse-grained and relatively cohesionless, fan materials are relatively simple. Insights gained
from this examination are applicable to some of the other 568
rifts S- engor and Natalin (2001) inventoried around the globe
(Figure 3).
237
Figure 4 Extent of the Basin and Range Province mapped by Dickinson, W.R., 2002. The Basin and Range Province as a composite extensional
domain. International Geology Review 44, 138, with permission from Taylor and Francis.
238
Figure 5 North American extent crustal stress measured by a variety of methods (NF: normal faulting, TF: thrust faulting, SS: strike-slip
faulting, U: unknown). Reproduced from Heidbach, O., Tingay, M., Barth, A., Reinecker, J., KurfeX, D., Muller, B., 2010. Global crustal stress
pattern based on the World Stress Map database release 2008. Tectonophysics 482, 315.
The term erosion is most frequently associated with wash

processes. Because wash processes are specifically excluded
from this chapter (for reasons to be presented later), the terms
decay or degradation will be used instead to refer to the
morphologic changes to the fault scarp as it ages.
5.10.2
Basin and Range Province
Although there is some disagreement on the extent of the

Basin and Range Province, particularly the northern and
southern boundaries, it covers much of the western US and
extends from central Mexico to just south of the Canadian
border (Figure 4). The structure and geologic history of the
province are quite complex and are comprehensively detailed
by Dickinson (2002). McPhees (1981) book, Basin and
Range, is a very readable, nontechnical treatment of the
geology of the province.
Crustal extension of the Basin and Range Province began in
the mid-Tertiary and is ongoing. The observed extant stress
pattern (Figure 5) is generally eastwest tension producing a
series of northsouth trending horsts and grabens. Dutton
(1885, p. 116) memorably noted that Whoever has examined, even cursorily, the map of Western America must have
noticed the following arrangement of the mountain masses:
The great belt of cordilleras coming up through Mexico and

crossing into US territory is depicted as being composed of
many short, abrupt ranges or ridges, looking upon the
map like an army of caterpillars crawling northward
(Figure 6).
For much of their length, the range-front faults cut through
alluvial fans or bajadas formed at the base of the upthrown
fault blocks (Figure 7). This is particularly evident where
ranges have been more eroded suggesting substantial retreat of
range fronts back from the faults that form them. In his
Canons of Landscape Evolution, King (1953) argued that
escarpments generally retreat rather than recline with time,
particularly in dryer environments, and used the Drakensberg
escarpment in Natal as an example: Wall-like in form, (the
Drakensberg escarpment) has retreated more than (250 km)
westward to its present position. The wall is still (1200 m) or
more high with black, forbidding precipices in its upper part
along the borders of Basutoland where the crest rises above
(3000 m). It is quite unmistakable, and shows little sign of
flattening in its 120 million years of existence. The rate at
which it has migrated averages about a foot in 200 years.
Kings argument that the Drakensberg demonstrates scarp
retreat has been subsequently attacked but recently Moore and
Blenkinsop (2006) and Blenkinsop and Moore (2013) report
supporting evidence for it. In contrast to the apparent retreat
Quaternary
Qv
Quaternary volcanic rocks
Tpc
Pliocene continental
Tpv
Pliocene volcanic rocks
Tpf
Pliocene felsic volcanic rocks
Ti
Tertiary intrusive rocks
Tmv
Miocene volcanic rocks
Tmf
Miocene felsic volcanic rocks
Txc
Paleocene continental
Kg
239
Cretaceous granitic rocks
Kg1
Lower Cretaceous granitic rocks
Jmi
Jurassic mafic intrusives
IMz
Lower mesozoic
IMze Lower mesozoic eugeosynclinal

Trg
Triassic granitic group
TrPe
Triassic and permian eugeosynclinal
uPz
Upper paleozoic
uPze Upper paleozoic eugeosynclinal
100
200
300
400
IPze
Lower paleozoic eugeosynclinal
IPz
Lower paleozoic
500
Kilometers
Figure 6 Northsouth trending horst and grabens of the Basin and Range Province. Reproduced from Reed, J.C., Bush, C.A., 2007. About the
geologic map in the National Atlas of the United States of America. U.S. Geological Survey Circular 1300, 52 pp., with permission from USGS.
of the bedrock range front, the much more recent and

ephemeral fault scarps formed in the fronting alluvial fans
appear to recline rather than retreat. Causes for scarp retreat
versus recline are examined next.
5.10.3
Slope Retreat Versus Recline
For much of the history of the discipline, geomorphologists

have debated whether hillslopes, of which scarps produced by
240
50
100
150
200
Kilometers
Figure 7 Range-front, normal faults believed to be sources of significant activity during the Quaternary (red) are generally closer to the ranges
base for the less eroded and presumably younger ranges than they are for the more eroded and presumably older ranges. Reproduced from
Machette, M.N., Haller, K.M., Dart, R.L., Rhea, S.B., 2003. Quaternary fault and fold database of the United States, U.S. Geological Survey OpenFile Report 03-417 (on line database http://qfaults.cr.usgs.gov/faults/), with permission from USGS. Topographic data 1 arc second OEM from
Seamless, http://seamless.usgs.gov/
241
Upper original surface

Crest
Free face
Debris
slope
Wash
slope
Toe or
base
Fault zone
Figure 10 Retreat of the abandoned face of an obsidian sand and

gravel quarry north of West Yellowstone, Montana.
Lower original surface

Figure 8 Morphologic nomenclature used by Wallace (1977) to
describe normal fault scarp. Reproduced from Wallace, R.E., 1977.
Profiles and ages of young fault scarps, northcentral Nevada.
Geological Society of America Bulletin 88, 12671281, with
permission from The Society.
1
, s
3
3
s=

c+
tan
2
c
3
3
1
1
Figure 9 Failure of an unconsolidated material in an active Rankine
stress state produces a scarp face at a slope angle, a, of 451 f=2.
s1 and s3 are the primary and secondary principal stresses. Failure
occurs when s3, the lateral confining pressure, is reduced to some
value s30 where the resulting Mohrs circle touches the Mohrs failure
envelope, producing a scarp face sloping at a. s, c, and f are
strength, cohesion, and internal friction angle respectively. s and t
are the normal and shear stresses on a plane at some angle a to the
secondary principal stress direction.
normal faulting are most certainly an example, retreat or recline

with time. Examples of both types of behavior are easily found
(e.g., Nash, 1986). The reason for these two fundamentally
different patterns was first recognized by G.K. Gilbert, Geology
of the Henry Mountains first published in 1877. Gilbert drew a
distinction between vegetated and unvegetated hillslopes. On
unvegetated hillslopes, debris loosened from the rock face is
immediately removed. Gilberts Law of Structures prevails on
these slopes and the resulting profile stands at some angle related to the resistance of the underlying rock to weathering and
erosion. Slopes underlain by more resistant rock are steeper
whereas those underlain by less resistant rocks are gentler. The
reflection of rock strength in the slopes developed on them is
easily seen in the gentle slopes on shales and the steeper slopes
on sandstones in arid canyons carved into the Colorado Plateau
(e.g., Grand Canyon).
In contrast to unvegetated slopes, Gilbert (1877) observed
that vegetated slopes develop a colluvial cover held in place by
plant roots. On these slopes, material loosened from the rock
face is not immediately removed and the hillslope becomes
mantled with loosened debris. The limit to lowering of the
slope surface is directly related to the rate at which the
mantling material moves down slope rather than to the resistance of the underlying rock to weathering and erosion. On
vegetated slopes, every slope segments receives debris from the
upslope segment above it and discharges debris to the segment
below it. Gilbert (1877) observed that the rate at which
loosened debris moves downslope is a positive function
of slope. This being the case, a progressive increase in elevation
will occur where there is a downslope decrease in gradient
at concave-up portion of a slope profile. Conversely, where the
gradient increases downslope, the rate of debris transport
also increases resulting in a net removal of debris from
convex-up portions of a slope profile. Under these circumstances, the midsection of a simple scarp profile will recline
with time.
More recent authors have renamed Gilberts vegetated and
unvegetated hillslope types to transport-limited and detachment-limited. A transport-limited hillslope is mantled
with loosened debris so the rate at which the surface may be
lowered or raised at a point on its profile is the relative rate at
which debris is brought to the point from upslope to the rate
at which it is moved downslope from that point. A detachment-limited hillslope, on the other hand, is one on which
loosened debris is removed from the hillslope surface as rapidly as it becomes detached from it. The term weatheringlimited has also been applied to unvegetated hillslopes but
this term is misleading because scarps formed in completely
weathered or unconsolidated material (e.g., alluvium) may be
detachment-limited, for example, on a river cut bank.
Because weathering rates and growth of vegetation are
faster in humid climates, hillslopes there tend to be transportlimited whereas in arid and semiarid climates hillslopes tend
242
Figure 11 Scarp formed by normal faulting during the 1959 Hebgen Lake, Montana earthquake in alluvium along Cabin Creek. On left,
photograph taken shortly after earthquake (USGS photographic Library, ID. Hadley, J.B. hjb0028c) and on right photographed in 1978. The
detachment-limited free face, although nearly buried, is still visible.
Figure 12 Retreating, detachment-limited, free face retreats parallel

to itself, progressively being buried by the basal debris slope. Fisher
(1866) developed the first model of this system. The diagram shown
was produced by a digital model, Nash (1981).
to be found in arid climates and detachment-limited hillslopes occur in humid climates.

In the following discussion, the nomenclature used by
Wallace (1977) for the parts of normal fault scarps in unconsolidated material (Figure 8), are applied. Parallel retreat is
observed on scarps in unconsolidated materials formed in an
active Rankine stress state (Figure 9). Failure results from reduction or reversal of the lateral confining earth stresses such
as those produced by crustal tension (normal faulting),
undercutting by fluvial activity (cut banks), or wave action
(coastal bluffs). In two dimensions (profile), the primary
principal stress, s1, results from the vertically-acting weight of
the material and the horizontally-acting secondary principal
stress, s3, is the lateral confining stress. Failure or scarp formation occurs when s3 is sufficiently reduced that the resultant Mohrs circle touches the Mohrs failure envelope
(Terzaghi et al., 1996). The initial failure surface forms a detachment-limited free face at a slope angle
a 451
f
2
1
5 Vertical exaggeration
Figure 13 Degradation of a transport-limited scarp predicted by
hillslope diffusion model. Note progressive rounding of base and
crest and reduction of the scarps gradient.
where f is the internal friction angle (angle of repose). If the

material is entirely cohesionless (i.e., c 0), the scarp will immediately collapse to f. Virtually no naturally occurring, unconsolidated material will be completely cohesionless, however,
so it will take some finite period of time for the scarp to ravel to f.
For clast-supported alluvium, observed values of f generally fall within the range of 301 to 351 so initial scarp angles
are around 601. If the scarp continues to be undercut by wave
or fluvial activity, it can maintain this precipitously steep
failure angle. If, however, debris raveled from the scarp face
accumulates at the base, such as occurs on an unworked
face of a gravel quarry f (Figure 10) or normal fault scarp
(Figure 11), the scarp will retreat back at an angle of 451 f=2
progressively burying its base.
Several models have been developed for the progressive retreat and burial of a scarp, the earliest and perhaps the most
elegant by Fisher (1866). Fisher assumed the free face is vertical
and there is no change in the density of the material as it is
removed from the free face and deposited on the debris slope,
he also assumed a closed system with no loss of debris. Nash
(1981) developed a more flexible but less elegant computer
model of the same process that permits modeling of loss of

material from the system and a nonvertical free face (Figure 12).
The time necessary for the retreating free face to be completely buried varies. In more cohesive material, the free face
may persist for more than a century. Wallace (1977) found the
free face of a normal fault scarp formed by the 1915 Pleasant
Valley, Nevada earthquake was still quite evident nearly sixty
years later in 1974 and is likely to persist for considerably
more than a century. In contrast, the free face produced by
normal faulting of alluvium along Cabin Creek near Hebgen
Lake, Montana during the 1959 earthquake was completely
buried in about three decades (Figure 11).
250
60
Diffusion age (m2)
200
40
20
150
0
0
10
243
20
100
50
NZ North-facing
NZ South-facing
0
10
20
30
40
Height (m)
50
60
70
Figure 14 Clarke and Burbank (2010) determined the morphologic

age (their diffusion age) of a large number of New Zealand terrace
scarps presumed to be formed at nearly the same time. If the simple
hillslope diffusion model is valid, the morphologic ages should be
nearly the same. Instead there is a strong linear correlation between
morphologic age and scarp height (inset is an expanded view of the
cluster of scarps between 0 and 22 m). Reproduced from Clarke,
B.A., Burbank, D.W., 2010. Evaluating hillslope diffusion and terrace
riser degradation in New Zealand and Idaho. Journal of Geophysical
Research 115, F02013, doi:10.1029/2009JF001279, with permission
from AGU.
5.10.4
Modeling the Decay of Transport-Limited

Scarps
A great many studies have modeled transport-limited hillslopes including Culling (1960), Bucknam and Anderson
(1979), Nash (1980a, b), Colman and Watson (1983),
Andrews and Hanks (1985), Hanks (2000), and Clarke and
Burbank (2010). The most popular model assumes that the
downslope flux of debris (volume per length of contour per
time so units of [L2 T1]) at a point on a hillslope is linearly
proportional to the gradient at that point (an assumption
notably not made by Gilbert, 1877). Because the resulting
model is very similar in form to diffusion equations used for
hb
ha
Figure 15 Two similar cones used as examples of very simple landforms. (ha/hb), (ha/hb)2, and (ha/hb)3 are the ratio of all linear, area, and
volume ratios of the cones, respectively. If the downslope debris flux is solely a function of gradient (i.e., independent of scale), then both cones
will undergo identical patterns of decay but at different rates. The ratio of the times necessary for both to reach some slope angle a, (ta/tb), is
equal to (ha/hb)2. Adapted from Nash, D.B., 2005. A general method for morphologic dating of hillslopes. Geology 33, 693695, with permission
from Geological Society of America.
244
Degraded
scarp
Fresh
scarp
(a)
(b)
(c)
Figure 16 Stewart and Hancocks (1990) nomenclature for nonsimple scarps produced by normal faulting: multiple (a), splintered (b), and
composite (c). Reproduced from Stewart, I.S., Hancock, P.L., 1990. What is a fault scarp? Episodes 13, 256263, with permission from
International Union of Geological Sciences.
Figure 17 Normal fault scarp resulting from 1983 Borah Peak,

Idaho earthquake (2.8 m rod for scale). The offset along the scarp
decreases as it approaches its terminus. (USGS photographic Library,
ID. Machette, M. N. 0007.)
heat flow, electrical flow, ground-water flow, and chemical

diffusion, it is generally referred to as the hillslope diffusion
model. The model has been applied with varying success to
the observed degradation pattern of transport-limited fluvial
terrace scarps (e.g., Nash, 1984; Colman and Pierce, 1986;
Nash and Beaujon, 2006; Clarke and Burbank, 2010), wavecut bluffs (e.g., Nash, 1980b; Hanks et al., 1984; Hanks and
Wallace, 1985), and fault scarps (e.g., Nash, 1980a; Hanks
and Wallace, 1985; Hanks and Schwartz, 1987; Hanks and
Andrews, 1989). Some have sufficient confidence in the
models validity that they have calculated the period of time
elapsed since a hillslope became transport-limited by matching the observed morphology of the hillslope with the
morphology predicted by an appropriately calibrated diffusion model, a process Clarke and Burbank (2010) referred to
as morphologic dating.
A number of procedures have been developed for morphologic dating (see Clarke and Burbank, 2010 for a review of
the most commonly used).
The hillslope diffusion model predicts the rounding of
the basal concavity and crestal convexity as well as the reduction in scarp gradient that is observed on degraded scarps
(Figure 13). Despite similarities between the pattern of change
predicted by the model and the observed pattern of scarp
degradation, there are important differences. Notably, the
morphologic age determined for scarps of the presumed
same age calculates significantly higher ages for higher scarps.
Colman and Pierce (1986), Nash and Beaujon (2006), and
Clarke and Burbank (2010) found the diffusion model fails in
notable ways. All find a fairly linear relationship between
morphologic age and scarp height for scarps of the same age
(Figure 14).
Nash (2005) proposed a general model for the degradation
of transport-limited hillslopes that Clarke and Burbank (2010)
termed the geometric model. The model makes no assumption about the relationship between the downslope debris flux
and slope gradient. The only assumption is that debris flux is
completely independent of scale. In other words, the downslope flux of debris at a point with a slope angle of a on a 10meter high scarp would be the same as the corresponding
point with a slope angle of a on a 100-meter high scarp. The
debris flux could be any function of gradient as long as it was
independent of scale. If this assumption is correct then two
otherwise identical landforms of different scales will go
through the identical pattern of degradation but at different
rates that are a simple function of their different scales.
Given two simple, similar cones (Figure 15), the first of
which is some scale factor x smaller or larger than the second,
any measured parameter of length made of the first cone will
correspond to that length multiplied by x on the second.
Similarly, any measure of a parameter of area and of volume
made of the first cone will correspond to the same measure of
the second cone multiplied by x2 and x3, respectively. Whatever the scale-independent function of gradient assumed, the
debris flux will be the same at the corresponding point on
each cone but the volume of material that must be moved for
the second cone to decay to the similar scale morphology as
the first must be x3 more or less than the first. A similar segment of a contour line on the second cone will be x times
245
13.0 m
12.5 m
10.0 m
7.5 m
5.0 m
2.5 m
0.0 m
Figure 18 Very simple normal fault scarp with differing offsets along its length (initial morphology above, degraded morphology below). Lines
correspond to profiles with initial offsets of approximately 3, 5, 7, 9, 11, and 13 m, respectively. (Disclaimer: the morphology of the fault block
used for Figures 1822 is for illustrative purposes. It bears little semblance to any fault scarp likely to be found in nature.)
h: 13, t:1
h: 13, t: 200
h: 13, t: 400
h: 13, t: 600
h: 13, t: 800
h: 13, t: 1000
Figure 19 Changes in morphology for a 13-meter high scarp predicted by the simple hillslope diffusion model after 0, 1, 2, 4, 6, 8, and 10
centuries (drawn with 25-times vertical exaggeration).
longer or shorter than the same segment on the first cone so

only x2 times more or less material must pass through that
segment on the second cone to achieve the same morphology
as the first cone. This being the case, it will take the second
cone x2 times as much time to achieve a similar decayed
morphology as the first cone.
This geometric model has the potential to be particularly
appropriate for predicting the pattern of degradation of
transport-limited scarps produced by normal faulting. In

contrast to a fluvial terrace scarp or wave-cut bluff, normal
fault scarps, by their nature, range across a variety of heights
(offsets) along their length (Figures 16b and 17). The scissor
offset at either end of the fault strand will increase from zero
at either terminus to some maximum offset along its length.
Given this broad range of offsets, the changes in morphology
observed along a degraded scarp of a single age, will show the
246
11.8
13.1
15
17.8
22.8
30
h: 2.93, t: 0
h: 5, t: 200
h: 6.96, t: 400
h: 9, t: 600
h: 11, t: 800
h: 13, t: 1000
Figure 20 The same changes in scarp morphology predicted for a 13-meter high scarp over 10 centuries (Figure 19) are observed on a scarp
of a single age but varying in height (ergodic substitution) (drawn with 25-times vertical exaggeration).
11.8
11.8
11.8
11.7
11.7
11.7
h: 2.93, t: 51
h: 5, t: 148
h: 6.96, t: 287
h: 9, t: 479
h: 11, t: 716
h: 13, t: 1000
Figure 21 After 10 centuries, the gradient of a 13-m high scarp is predicted by the hillslope-diffusion model to decrease from 301 to 11.791.
The geometric model predicts that it will take 1000 (11/13)2 715.976 years for a 11-m high scarp to degrade to the same angle. Note: Because
a numerical model is used to generate these figures, only scarp heights at discreet nodes and profiles at integer years are available so that there
are small variations in slope angle (drawn with 25-times vertical exaggeration).
identical pattern of decay of a given offset over a broad range

of times as is demonstrated in the illustrations that follow
(sometimes referred to as an ergodic substitution of time and
space). In the following demonstration, all figures are generated with a very simple hypothetical normal fault scarp
with an initial slope angle of 301. A simple numerical, finitedifferences, hillslope diffusion model, with a diffusivity of
1 m2 yr1 and a node spacing of 1 m is used (this simple
model was implemented with an Excel spreadsheet available
from the author). Because this model uses discreet rather than
continuous time steps, the predicted morphology is only
available for integer years and scarp profiles are only available
at nodes so their initial heights, although close, are not always
exact integers. If it takes a scarp with an initial height of
ha a time period ta to decline to a particular slope angle and
it takes a similar scarp with an initial height of hb a time tb
to reach the same slope angle then the geometric model
states
2
ta
ha
tb
hb
2
Figures 1822 demonstrate use of eqn [1] for geometric

modeling.
5.10.5
Limitation of the Geometric Model for

Normal Fault Scarp Decay
The geometric model for normal fault scarp decay is quite

simple and does not require a specific function relating debris
flux to gradient other than that function must be independent
of scale. That assumption, however, should be made with
caution. It is unlikely that the debris flux will be independent
of scale if the scarp is subject to wash erosion. As Gilbert
(1877) observed, the water flowing across a point on a hillslope will be a direct function of the upslope drainage area so
it is also strongly scale dependent. Scarps with a more rounded
or extensive basal concavity than crestal convexity suggest
wash processes may be significant. The geometric model
should not be applied to scarps with a significant wash
slope or rilled or gullied debris slope (Figure 8). Scarps with
a continuous cover of vegetation are less likely to be
247
02
t:
6
t:
4
t:
2
30
88
.9
.5
.7
72
t:
1
t:
3
0.
6
t:
89
.1
30
h: 2.93
h: 5
25
h: 6.96
Midsection slope angle ()
h: 9
h: 11
20
h: 13
15
10
0
0
200
400
600
800
1000
Time
Figure 22 Decrease in maximum scarp angle predicted for different scarp heights along the modeled normal fault scarp in Figure 18. If it takes
602 years for a 13 m high scarp to decline from 301 to 151, the geometric model predicts will take 89.1 years for a 5 m scarp to decline by the
same amount.
Figure 23 Renewed movement of fault superimposes younger scarp

on a prehistoric one (USGS photographic Library, ID. Witkind, I.J. 5ct
wij00005).
significantly affected by wash processes than unvegetated or

sparsely vegetated ones. Scarps formed in clast-dominated fan
material will be less susceptible to wash than will those
formed in more cohesive debris flow deposits.
Application of the geometric model to normal fault scarps
poses problems beyond debris fluxes that are not scale
independent, particularly the assumption that the initial
morphology of the scarp is similar along its varying height.

A relatively short period of time is generally required for the
initial scarp formed at a slope angle of 451 f=2 to ravel back
to f. Although this generally occurs fairly rapidly (decades) for
scarps in more cohesive materials, it may take significantly
longer. Higher scarps may also require significantly more time
to ravel back to f.
Many, if not most, normal fault scarps formed by rangefront faulting of alluvial fans recur at the same spot during
episodic movement (Figure 23). If fan erosion and deposition
do not obliterate completely the earlier scarp (i.e., plane the
surface flat), then the resulting palimpsest is not amenable to
geometric modeling. It is very difficult to differentiate scarps
produced by a single event from those produced by multiple
events from the morphology of the resultant degraded scarp
alone. Trenching of the scarp is required to recognize whether
recurrent fault movement is involved. If the scarp has been
produced by recurrent offsets, it may be possible to analyze
the flight of terraces produced by fan head entrenchment resulting from faulting. It should be noted, however, that it is
generally uncertain whether fan head entrenchment has resulted from tectonism, or change in land cover or climate.
Application of the geometric model to scarps produced by
normal faulting of fan material is also complicated by the fact
that the scarps formed are not always a simple offset producing a single scarp. Examination of scarps produced by the
248
1959 Hebgen Lake earthquake finds that roughly one half

were not simple but rather consisted of two or more breaks
close together (Figure 1(a)) or were simple homoclinal flexures without any break. Once these scarps have been degraded, their morphology will be indistinguishable from
much older simple scarps. Again, trenching will be necessary
for differentiating initially simple scarps from those with
multiple or no scarps.
5.10.6
Summary
In contrast to landforms produced by compressional or shear

faulting of consolidated materials, scarps produced by normal,
range-front faulting of alluvial fans frequently have a relatively
simple initial morphology. If the fan material is clast supported and relatively cohesionless and if the scarp is not
subject to erosion by wash processes, the decay of the scarp
may be relatively simple. In the initial stage, the scarp will have
a steep, detachment-limited free face with a slope of about
601. This face will retreat back parallel to itself, progressively
burying its base with a debris slope at the angle of repose
debris. This process is generally fairly rapid (decades) but may
take longer for more cohesive materials. Once the free face has
been buried, the scarp becomes transport-limited.
Numerous models have been proposed for describing the
decay of transport-limited hillslopes, of which the hillslope
diffusion model is the most widely used. Despite its wide
acceptance, the apparent failure of the diffusion model when
applied to high scarps is problematic. Unlike the hillslope
diffusion model, the geometric model makes no assumption
about the functional relationship between gradient and debris
flux. The only assumption made by this model is that debris
flux is scale independent, and only a function (any function)
of gradient. If this assumption is valid, then the morphologic
differences observed along a fault scarp over a range of offsets
are identical (geometrically similar scale models) to the
changes of a single height scarp over a range of times.
The geometric model provides both the means for studying
the pattern of scarp decay but may, under tightly constrained
circumstances, be used for dating some fault scarps produced
by normal faulting of alluvial fans.
References
Andrews, D.J., Hanks, T.C., 1985. Scarp degraded by linear diffusion: Inverse
solution for age. Journal of Geophysical Research 90(B12), 1019310208.
Blenkinsop, T., Moore, A., 2013. Tectonic geomorphology of passive margins and
continental interiors. In: Shroder, J. (Editor in Chief), Owen, L.A. (Ed.), Treatise
on Geomorphology, vol. 5, pp. 7192.
Bucknam, R.C., Anderson, R.E., 1979. Estimation of faultscarp ages from a
scarpheight slopeangle relationship. Geology 7, 1114.
Clarke, B.A., Burbank, D.W., 2010. Evaluating hillslope diffusion and terrace riser
degradation in New Zealand and Idaho. Journal of Geophysical Research 115,
F02013. http://dx.doi.org/10.1029/2009JF001279.
Colman, S.M., Pierce, K.L., 1986. Glacial Sequence near McCall, Idaho
weathering rinds, soil development, morphology, and other relative-age criteria.
Colman, S.M., Watson, K., 1983. Age estimated from a diffusion equation model
for scarp degradation. Science 221, 263265.
Culling, W.E.H., 1960. Analytical theory of erosion. Journal of Geology 68,
336344.
Davis, W.M., 1913. Nomenclature of surface forms on faulted structures. Geological
Dickinson, W.R., 2002. The Basin and Range Province as a composite extensional
domain. International Geology Review 44, 138.
Dutton, C.E. 1885. Mount Taylor and the Zuni plateau. In: Sixth Annual report of
the United States Geological Survey to the Secretary of the Interior, 18841885.
pp. 113198.
Fisher, O., 1866. On the Disintegration of a Chalk Cliff. Geological Magazine 3,
354356.
Gerrard, J., 2008. Geology and landforms, Chapter 2. In: Chorley RJ, Brunsden D,
Cox NJ, Goudie AS. (Eds.), The History of the Study of Landforms: Quaternary
and Recent Processes and Forms (18901965) and the Mid-Century
Revolutions.
Gilbert, G.K., 1877. Report on the geology of the Henry Mountains. US. Govt.
Printing Office, Washington, D.C, 160 p.
Hanks, T.C., 2000. The age of scarp like landforms from diffusion-equation
analysis. In: Noller, J.S., Sowers, J.M., Lettis, W.R. (Eds.), Quaternary
geochronology: Methods and Implications, American Geophysical Union
Reference Shelf Series 4, pp. 313338.
Hanks, T.C., Andrews, D.J., 1989. Effect of far-field slope on morphologic dating of
scarplike landforms. Journal of Geophysical Research-Solid Earth and Planets
94, 565573.
Hanks, T.C., Bucknam, R.C., Lajoie, K.R., Wallace, R.E., 1984. Modification of wavecut and faulting-controlled landforms. Journal of Geophysical Research 89,
57715790.
Hanks, T.C., Schwartz, D.P., 1987. Morphological dating of the Pre-1983 fault scarp
on the lost river fault at Doublespring Pass Road, Custer County, Idaho. Bulletin
of the Seismological Society of America 77, 837846.
Hanks, T.C., Wallace, R.E., 1985. Morphological analysis of the Lake Lahontan
shoreline and beachfront fault scarps, Pershing County, Nevada. Bulletin of the
King, L.C., 1953. Canons of landscape evolution. Geological Society of America
McPhee, J., 1981. Basin and Range. Farrar, Straus & Giroux, New York, 216 p.
Moore, A., Blenkinsop, T.G., 2006. Scarp retreat versus pinned drainage divide in
the formation of the Drakensberg escarpment, Southern Africa. South African
Nash, D.B., 1980a. Morphologic dating of degraded normal fault scarps. Journal of
Geology. 88, 353360.
Nash, D.B., 1980b. Forms of bluffs degraded for different lengths of time in Emmet
County, Michigan, USA. Earth Surface Processes 5, 331345.
Nash, D.B., 1981. Fault: A Fortan program for modeling the degradation of active
normal fault scarps. Computers and Geoscience 7, 249266.
Nash, D.B., 1984. Morphologic dating of fluvial scarps and fault scarps near West
Yellowstone, Montana. Geological Society of America Bulletin 95, 14131424.
Nash, D.B., 1986. Morphologic dating and degradation of fault scarps. Active
Tectonics, Panel on Active Tectonics. National Academy of Science, p. 181194.
Nash, D.B., 2005. A general method for morphologic dating of hillslopes. Geology
33, 693695.
Nash, D.B., Beaujon, J.S., 2006. Modeling degradation of terrace scarps in Grand
Teton National Park, USA. Geomorphology 75, 400407.
S- engor, A.M.C., Natalin, B.A., 2001, Rifts of the world. In: Ernst, R.E., Buchan,
K.L., (Eds.), Mantle plumes: their identification through time,Geological Society
of America Special Paper 352, Boulder, Colorado, pp. 389482.
Stewart, I.S., Hancock, P.L., 1990. What is a fault scarp? Episodes 13, 256263.
Terzaghi, K., Peck, R.B., Mesri, G., 1996. Soil Mechanics in Engineering Practice,
Third ed. Wiley, New York, 549 p.
Wallace, R.E., 1977. Profiles and ages of young fault scarps, north central Nevada.
249
Biographical Sketch
David B. Nash received his doctorate in Geology from the University of Michigan and did
postdoctoral research in thermal remote sensing at the NASA Jet Propulsion Laboratory. He
was a member of the National Research Councils Panel on Active Tectonics. He joined the
faculty of the Department of Geology at the University of Cincinnati in 1977. His research
areas include geomorphic processes and hydrogeologic modeling.
5.11 Landslides Generated by Earthquakes: Immediate and

Long-Term Effects
DK Keefer, University of Maine, Orono, ME, USA
5.11.1
5.11.2
5.11.2.1
5.11.2.2
5.11.2.2.1
5.11.2.2.2
5.11.2.2.3
5.11.3
5.11.4
References
Introduction
Overview of Landslide Occurrence in Earthquakes
Numbers and Classification of Earthquake-Induced Landslides
Relation of Landslide Occurrence to Ground Shaking, Topography, Materials, and Hydrologic
Conditions
Distribution of landslides related to earthquake magnitude and shaking intensities
Topographic, hydrologic, and geologic indicators of landslide occurrence
Detailed landslide distribution from an individual earthquake: example from the 1989 Loma Prieta, California
earthquake (M 6.9)
Geomorphic and Postearthquake Effects of Earthquake-Induced Landslides
Conclusions
Glossary
Avalanche Complex type of landslide consisting of a
stream of soil particles or rock fragments, typically mixed
with water, moving at high velocity via a combination of
sliding and flow, and in some cases, other mechanisms such
as falling.
Basal shear surface The basal boundary of a landslide
along which sliding (i.e., differential displacement, or
concentrated shearing) takes place.
Block slide Type of landslide in which blocks of material
move primarily by translational sliding, or shear
displacement, along a well-defined planar basal shear
surface.
Colluvium/colluvial soil Type of regolith or soil material,
which is typically heterogeneous, formed primarily by
gravity-induced mass movement or sheet erosion on slopes
of at least moderate inclination.
Earthquake magnitude (M) A measure of the energy
released at the source of an earthquake, determined from
measurements on seismographs. Moment magnitude (M)
is based on the seismic moment of the earthquake that
is equal to the rigidity of the earth materials along the
causative fault multiplied by the average amount of slip
on the fault during the earthquake and the size of the
area that slipped. Richter surface-wave magnitude (Ms) is
based on measurements of Rayleigh surface waves with a
period of nearly 20 s. Moment magnitude is considered to
be a more accurate measure of energy release in large
earthquakes (M47.5).
Fall Type of landslide in which material moves primarily
by free fall, bouncing, and rolling.
251
251
252
255
256
258
259
259
263
263
Flow Type of landslide that moves primarily by fluid

like flow.
Geographic Information Systems (GIS) Hardwareand software-based systems for digitally storing, analyzing,
and presenting geographic data, most commonly in
map form.
Landslide The downslope and lateral movement of a
block or mass of earth materials, including solid particles
and entrained fluids, driven by gravity, in some cases
in conjunction with such other forces as earthquake
shaking.
Landslide Concentration Number of landslide sources
per unit of surface area.
Lateral Spread Type of landslide consisting of blocks of
relatively intact surficial material moving, primarily
laterally, on fluidized zones of liquefied, granular material
or remolded, sensitive clay.
Liquefaction Substantial loss of strength or stiffness in
soils due to transient increases in pore-fluid pressures
induced by earthquake shaking. Liquefaction typically
occurs in saturated cohesionless sediments (such as wellsorted sands) and can result in excessive settlements of
buildings, failure of earth dams, large lateral movements of
the ground, and landslides.
Pore-water pressure The pressure of the interstitial water
in the pores or voids between solid particles in an earth
material.
Regolith General term for the entire surface mantle of
unconsolidated, loose earth material than nearly
everywhere covers firm, intact bedrock.
Keefer, D.K., 2013. Landslides generated by earthquakes: immediate and

long-term effects. In: Shroder, J. (Editor in Chief), Owen, L.A. (Ed.), Treatise
on Geomorphology. Academic Press, San Diego, CA, vol. 5, Tectonic
250
http://dx.doi.org/10.1016/B978-0-12-374739-6.00091-9
Landslides Generated by Earthquakes: Immediate and Long-Term Effects
Residual soil Unconsolidated earth material formed

largely by weathering of rock in situ, without significant
transport.
Rill A small hillside channel, typically the first type of
channel to be formed by surface runoff.
Rock Rock refers to consolidated bedrock that was
relatively hard, firm, and intact before the initiation of
landslide movement.
Sensitive clay Clay that loses significant strength when
disturbed, or remolded.
Shaking intensity Strength of earthquake-induced
ground shaking at a particular locality as determined from
perceptions of people, and effects on human structures and
the natural environment. Intensity is commonly reported
on the Modified Mercalli Intensity scale, which has grades
251
ranging from I (only felt by a few people) to XII (essentially

total destruction).
Slump Type of landslide in which blocks of earth material
move primarily by sliding, or shear displacement, with a
significant component of headward rotation, along a welldefined, concave-upward basal shear surface.
Soil Any loose, unconsolidated, or poorly cemented
aggregate of mineral grains, which may or may not contain
organic material. Soil comprises the entire regolith and all
human-placed fills.
Talus The accumulated mass of generally coarse (bouldersized), and angular rock fragments deposited little by little
at the base of a slope chiefly by falling, rolling, and sliding
of material detached from higher on the slope.
Abstract
Earthquake-induced landslides are of many different types and occur in virtually all geologic environments from steep and
high glaciated rock slopes to gently sloping river flood plains and coastal areas underlain by soft sediment. The largest
earthquakes can trigger tens of thousands of landslides throughout regions of tens or even hundreds of thousands of square
kilometers, moving millions or billions of cubic meters of material down slopes. Such landslides can greatly alter the
landscape in a variety of ways and pose hazards that persist long after the earthquake shaking that produced them.
5.11.1
Introduction
Landslides are among the most widespread of earthquake effects. In great earthquakes, landslides can number in tens of
thousands and occur hundreds of kilometers from the causative fault rupture. They can cause significant changes in the
landscape and, by such processes as stripping vegetation off
slopes and damming rivers, they can produce effects that
persist long after the earthquake shaking has stopped. They
have also killed thousands of people in recent earthquakes,
and they are among the leading causes of earthquake-related
economic losses. Thus understanding and forecasting their
occurrence is important in both scientific and practical terms.
Earthquake-triggered landslides have a wide range of characteristics and originate in many different types of terrain from
steep and high, glaciated rock slopes to gently sloping river
floodplains underlain by soft, liquefiable sediment. The largest
and most mobile of these landslides have volumes of more
than 100 million m3 and may travel distances of up to more
than 10 km at velocities that can exceed 100 km hr-1. Dense
concentrations of generally smaller landslides, triggered by
large or great earthquakes, can strip surficial regolith and even
bedrock from vast slope areas, leaving the slopes vulnerable to
subsequent failure and erosion while transporting large
amounts of sediment into fluvial systems.
The purposes of this chapter are to review existing knowledge on earthquake-induced landslides, to describe their occurrence, and to discuss their immediate and long-term
geomorphic effects. Data for the chapter are derived from
several different types of sources, including: (1) General syntheses and analyses at worldwide or regional scales (Seed,
1968; Youd and Hoose, 1978; Keefer, 1984a, 1994, 1999,
2002; Rodrguez et al., 1999; Papadopoulos and Plessa, 2000;
Prestininzi and Romeo, 2000; Bommer and Rodrguez, 2002;
Hancox et al., 2002); (2) Reports of postearthquake field investigations, mostly carried out during the past 50 years (for
citations to original sources to many of these reports, see especially Keefer and Tannaci, 1981 and Rodrguez et al., 1999);
(3) Studies of long-term geomorphic effects of landslides
caused by a few recent earthquakes (Ramachandra Rao, 1953;
Pain and Bowler, 1973; Adams, 1980; Pearce and Watson,
1986; Moseley et al., 1992; Dadson et al., 2003, 2004, 2005;
Keefer and Moseley, 2004; Lin et al., 2003, 2006; Meurnier
et al., 2008); and (4) Detailed geomorphic analyses of a few
particularly large earthquake-induced landslides (See especially Evans et al., 2009a, b).
The following section of this chapter summarizes the
characteristics of earthquake-induced landslides and general
relations between landslide occurrence on the one hand and
earthquake magnitude and shaking intensity on the other.
This is followed by a discussion of the geomorphic effects of
these landslides.
5.11.2
Overview of Landslide Occurrence in

Earthquakes
The numbers, distribution, and characteristics of landslides

produced by an earthquake determine the landslide-related
geomorphic impact of the event. All three of these measures in
turn depend on both the seismic ground-motions generated
by the earthquake and by the nature of the affected terrain. In
general, the numbers of landslides and geographic extent of
landside occurrence correlate with earthquake magnitude and
shaking intensities, whereas the specific characteristics of the
landslides produced, and thus the local geomorphic impacts,
correlate to a great degree with the local nature of the terrain.
252
5.11.2.1
Numbers and Classification of

Earthquake-Induced Landslides
Whereas total numbers of landslides have been determined for

only a relatively few, recent earthquakes, these available data
show that large earthquakes may produce tens of thousands of
landslides and that total numbers correlate with magnitude
(Figure 1). The largest number of landslides estimated to have
been caused by one of these events is approximately 50 000
(Figure 1). However, no earthquakes with moment magnitude, M47.7 are included in these data, and it is almost certain that great earthquakes with magnitudes in the range of
89.5 can trigger even larger numbers of landslides. For example, based on estimates of total landslide volume (Mathur,
1953), the M 8.6 Assam, India earthquake of 1950 almost
certainly triggered well more than 100 000 landslides. Earthquake-induced landslides may be classified in a number of
ways; one widely used classification (Keefer, 1984a) classified
these landslides into fourteen individual types grouped into
three main categories, largely on the basis of type of material
and nature of movement (Table 1). These types of landslides
differ greatly in their overall abundance in earthquakes
(Table 2), and each type of landslide occurs in a particular
suite of geologic and geomorphic environments. By far the
most abundant landslides triggered by earthquakes (Table 2)
are highly disrupted and shallow falls and slides (Figure 2)
that may involve either bedrock or regolith (soil) material
(Category I landslides in Table 1). Occurring on relatively
steep slopes, they typically strip vegetation and layers of
regolith or rock a few decimeters to a few meters thick from
the upper parts of steep slopes, depositing the transported
material on nearby talus accumulations, fans, or valley floors.
Whereas volumes of most individual landslides of these types
are in the range from less than 1 m3 to 10 000 m3 (Table 1),
landslides of these types may occur from closely spaced
sources (Figure 2(b)), and their deposits may coalesce, collectively affecting large slope areas and moving large volumes
of material.
The large end-members of the category of highly disrupted
landslides are rock avalanches complex landslides, with
Number of landslides (N)
100 000
10 000
1000
log N = 1.2312 M 4.8276
R2 = 0.72
100
10
1
5.0
5.5
6.0
6.5
7.0
7.5
8.0
Earthquake magnitude (M)

Figure 1 Earthquake magnitude vs. number of landslides.
Reproduced from Keefer, D.K., 2002. Investigating landslides caused
by earthquakes a historical review. Surveys in Geophysics 23,
473510, with permission from Springer.
individual volumes of more than 500 000 m3 (Table 1,

Figure 2(a)). They incorporate a variety of movement mechanisms to transport streams of highly disaggregated earth
materials distances of up to several kilometers at velocities of
tens or hundreds of kilometers per hour. Such rock avalanches
are relatively rare in earthquakes (Table 2), but they may be
highly destructive to human life and property. They may also
cause particularly significant geomorphic changes, altering the
local topography and drainage network over areas of several
square kilometers. One particularly well-studied earthquaketriggered rock avalanche occurred during an M 7.9 earthquake
in 1970 from a source on Nevado Huascaran, the highest peak
in the Peruvian Andes (Cluff, 1971; Plafker et al., 1971; Plafker
and Ericksen, 1978; Evans et al., 2009a). Plunging through an
eventual vertical drop of more than 3000 m, this complex rock
avalanche began as a fall of ice and rock from a near-vertical
slope. Progressively incorporating substantial amounts of
snow, ice, water, and colluvium, this rock avalanche bulked to
a final volume of 58.2 million m3 (Evans et al., 2009a). The
rock avalanche became in its lower reaches a debris flow
moving at an average velocity in the range of 5085 m s1
(Evans et al., 2009a), transporting material 17 km to the Santa
River, burying the town of Yungay and three nearby villages,
killing 6000 people (based on a reanalysis of the census data
by Evans et al., 2009a), and substantially altering the local
landscape. More distant effects occurred as material from the
rock avalanche entered the river, producing a debris flood/
hyperconcentrated flow that travelled to the Pacific Ocean,
180 km from the rock-avalanche source (Evans et al., 2009a).
Category II landslides remain more coherent during
movement than landslides in Category I (Table 1, Figure 3).
Some Category II landslides move translationally on planar
basal shear surfaces (block slides), some move with a significant component of rotation on curved basal shears (slumps)
and some move by a combination of translational shear and
internal flow (slow earth flows) (Table 1, Figure 3). Except for
slow earth flows, landslides of these types have more localized
geomorphic effects than Category I landslides, as they commonly transport material downslope a few tens of meters or
less (Table 1). Typical morphologies consist of a steep main
scarp at the upslope margin, a zone of depletion immediately
downslope, where the surface of the displaced landslide body
is below the original ground surface, and a zone of accumulation where deposited material, commonly consisting of a
series of large blocks separated by internal fissures and scarps,
accumulates above the original ground surface (Figure 3).
Though less abundant in earthquakes than Category I
landslides (Table 2), these Category II landslides tend to be
larger, with typical volumes in the range between 100 m3 and
a few million m3 (Table 1), and so they can collectively displace a significant portion of the total volume of material
moved downslope during an earthquake. One of the largest
Category II landslides attributable to a historical earthquake
involved 112 million m3 of material. Triggered by a Chilean
earthquake in 1575, this landslide dammed the San Pedro
River upstream from the city of Valdivia, which was destroyed
by a flood when the landslide dam breached 4 months after
the earthquake (Davis and Karzulovc, 1961, 1963).
Of the landslides in Category III, all of which move with a
significant component of fluid-like flow (Table 1), the most
Table 1 Characteristics of earthquake-induced landslidesa

Name
Type of Movement
Internal Disruption
Water
Content
Typical
Depths
Minimum Typical Velocities

Slope ( 1)
Typical Volumes
Typical Displacements
D U PS S
Disrupted landslides (Category I)
Rock falls
X X X
X Shallow
40
Disrupted
Translational sliding
rock slides
High
X X X
X Shallow
35
Rock
Complex, involving
avalanches sliding, flow, and
occasionally free
fall
Soil falls
Bouncing, rolling,
free fall
Very high
X X X
X Deep
25
High or very high
X X X
X Shallow
40
High
X X X
X Shallow
15
Very high
X X X
X Shallow
25
Rock slumps Rotational sliding
Slight or moderate
? X X
X Deep
15
Rock block
slides
Slight or moderate
? X X
X Deep
15
Soil slumps
Rotational sliding
Slight or moderate
? X X
X Deep
Soil block
slides
Slight or moderate
? ? X
X Deep
Slow earth
flows
and internal flow
Slight
Soil
Complex, involving
avalanches sliding, flow, and
occasionally free
fall
Most less than 1 104 m3;

maximum reported
2 107 m3
Rapid to very rapid
maximum reported
2 109 m3
Very rapid to extremely 5 1052 108 m3 or more
rapid
Extremely rapid
May fall to base of steep source slope and

move as far as several tens or hundreds of
meters farther, on relatively gentle slopes
May slide to base of steep source slope and
several tens or hundreds of meters farther,
on relatively gentle slopes
Several kilometers
Most less than 1,000 m3;

Most come to rest at or near bases of steep
maximum volumes not well
source slopes
documented
Moderate to rapid
May slide to base of steep source slope and
maximum reported
several tens or hundreds of meters farther,
4.8 107 m3
on relatively gentle slopes
Very rapid to extremely Volumes not well documented; Several tens of meters to several kilometers
rapid
maximum reported
beyond steep source slopes
1.5 108 m3
Extremely rapid
Coherent landslides (Category II)
10
Most between 100 and a few

million m3; maximum at least
tens of millions of m3
Slow to rapid
Most between 100 and a few
million m3; maximum at least
tens of millions of m3
Slow to rapid
Most between 100 and
1 105 m3; occasionally
1 105 to several million m3
Slow to very rapid
1 105 m3; maximum
reported 1.12 108 m3
Very slow to moderate; Most between 100 and
occasionally, with
1 106 m3; maximum
very rapid surges
reported between 3 107 and
6 107 m3
Typically less than 10 m; occasionally 100 m or

more
Typically less than 100 m; maximum
displacements not well documented
Typically less than 10 m; occasionally 100 m or
more
(Continued )
253
X Generally
shallow;
occasionally deep
Slow to rapid
High or very high
Disrupted
soil slides
Bouncing, rolling,
free fall
254
Table 1 Continued
Name
Type of Movement
Internal Disruption
Typical
Depths
Minimum Typical Velocities

Slope ( 1)
Typical Volumes
Typical Displacements
D U PS S
Lateral spreads and flows (Category III)
Soil lateral
spreads
Translation on fluid
basal zone
Rapid soil
flows
Flow
Generally moderate;
occasionally slight
or high
Very high
Subaqueous Generally lateral

Generally high or very
landslides
spreading or flow;
high; occasionally
occasionally sliding moderate or slight
a
X Variable
0.3
? ? ?
X Shallow
2.3
X Variable
0.5
Typically less than 10 m; maximum reported

600 m
1 105 m3; largest reported
9.6 106 m3
Very rapid to extremely Volumes not well documented; A few m to several km
rapid
largest are at least several
million m3
Generally rapid to
Volumes of not well
Not well documented, but some move more
extremely rapid;
documented; largest are at
than 1 km
occasionally slow to
least tens of millions of m3
moderate
Very rapid
Notes: Rock signifies bedrock that is relatively firm and intact before landslide initiation, and soil signifies loose, unconsolidated or poorly cemented aggregates of particles that may or may not contain organic materials. Internal disruption: slight
signifies landslide consists of one or a few coherent blocks; moderate signifies several coherent blocks; high signifies numerous small blocks and individual soil grains and rock fragments; very high signifies nearly complete disaggregation into
individual soil grains or small rock fragments. Depth: shallow signifies generally o3 m deep; deep signifies generally 43 m deep. Water content: D dry; U moist but unsaturated; PS partly saturated; S saturated. X indicates landslides with
this water content have been definitely identified. ? indicates landslides with this water content have not been verified but are inferred to occur based on questionable reports or indirect evidence. Velocity: very slow 1 106 3 10 6 m min 1;
slow 3 10 6 3 10 5 m min 1; moderate 3 10 5 0.001 m min 1; rapid 0.001 0.3 m min 1; very rapid 0.3 180 m min 1; extremely rapid 4180 m min 1. (Terminology after Varnes, 1978)
Source: Reproduced from Keefer, D.K., 2002. Investigating landslides caused by earthquakes a historical review. Surveys in Geophysics 23, 473510, with permission from Springer.
Water
Content
255
Table 2 Relative abundance of earthquake-induced landslides

Landslide type, listed in order of decreasing total numbers
Very abundant:4100 000 in 40 historical earthquakes
Rock falls
Disrupted soil slides
Rock slides
Abundant: 10 000 to 100 000 in 40 historical earthquakes
Soil lateral spreads
Soil slumps
Soil block slides
Soil avalanches
Moderately common: 1000 to 10 000 in 40 historical earthquakes
Soil falls
Rapid soil flows
Rock slumps
(a)
Uncommon: 1001000 in 40 historical earthquakes

Subaqueous landslides
Slow earth flows
Rock block slides
Rock avalanches
Source: Reproduced with permission from Keefer, D.K., 1984a. Landslides caused by
earthquakes. Geological Society of America Bulletin 45, 406421.
common are soil lateral spreads, which consist of coherent

blocks of material that move, largely horizontally, on subsurface layers of liquefied, granular sediment or fluidized,
sensitive clay (Figure 4(a)). Movement of lateral spreads can
produce areas of fissured and displaced ground stretching for
hundreds of meters or more along channel banks or free faces
in alluvial or coastal environments.
More completely liquefied or fluidized mud flows and
debris flows collectively classified as rapid soil flows (Keefer,
1984a, Table 1) may also be generated by earthquakes
(Figure 4(b)), though these are rarer (Table 2). Notable
earthquake disasters in central Asia, including the 1920
Gansu, China event (M 7.9) and the 1949 event near Khait in
the Soviet Tajikistan (magnitude, M 7.4) triggered massive
flows in loess that swept long distances down valleys at high
velocities, killing probably tens of thousands of people in the
China event (Close and McCormick, 1922; Wang and Xu,
1984) and 6400 people in the Khait event (Evans et al.,
2009b). In the case of the Khait event, a single massive loess
flow, formed by the coalescing of hundreds of flows from
multiple sources, buried a large valley for a distance of 20 km
under an estimated 245 million m3 of loess to an average
depth of 10 m (Evans et al., 2009b).
Earthquakes also cause submarine landslides, in numbers
that remain unknown. These landslides can remove large
portions of the distal margins of deltas, generating tsunami
waves and reconfiguring coastlines. The 1964 Alaska earthquake (M 9.2) was particularly notable among historical
earthquakes for the generation of submarine landslides, dozens of which occurred in deltaic sediments composed of sand
and gravel derived from glacial outwash (Keefer, 1984a).
Volumes of most are unknown, but one that affected the port
of Valdez involved approximately 75 million m3 of saturated
silty sand and gravel from the outwash delta at the head of the
Valdez fiord (Coulter and Migliaccio, 1966).
(b)
Figure 2 Category I landslidesdisrupted falls and slides. (a) Rock

avalanches and rock slides in glacial cirque in the eastern Sierra
Nevada of California, triggered by the Mammoth Lakes earthquake
sequence of May 1980 (maximum M 6.1). Reproduced from Keefer,
D.K., 1984b. Rock avalanches caused by earthquakes: source
characteristics. Science 223, 12881290, with permission from AAAS.
(b) Multiple Category I landslides strip regolith and vegetation off
slopes in the Central Mountain Range, Taiwan as a result of the 21
September Chi-Chi earthquake (M 7.5).
5.11.2.2
Relation of Landslide Occurrence to Ground

Shaking, Topography, Materials, and Hydrologic
Conditions
Relations between landslide occurrence, on the one hand, and

topography, slope materials, and hydrologic conditions on the
other hand are complex, although the general principles governing slope failure during earthquakes were explicitly identified 60 years ago or more: In a classic paper on the mechanism
of landslides, Tergazhi (1950) classified earthquake shaking as
an external cause of slope failure. During an earthquake, stresses
induced by ground shaking interact with the existing stresses
within a slope (determined largely by slope height and inclination) and the effective shear resistance of the slope material
(determined largely by cohesive and frictional strength and by
pore-water pressures). The interactions between slope topography, material properties, and local ground shaking can be
quite complex (Geli et al., 1988; Ashford and Sitar, 1997; Del
Gaudio and Wasowski, 2007), with field evidence and analytical studies generally suggesting that ground shaking is more
256
Liquefied
zone
Slump
Block slide
Slow earth flow
(a)
(a)
Soil lateral spread
(b)
Figure 3 Category II landslidescoherent slides. (a) Types of

Category II landslides. Slumps involve primarily rotational movement,
block slides translational movement, and slow earth flows
combinations of translational sliding and distributed internal flow.
Source of slump and block slide diagrams: Modified with permission
from Varnes, D.J., 1978. Slope movement types and processes. In:
Schuster, R.L., Krizek, R.J. (Eds.), Landslides Analysis and Control.
Transportation Research Board National Academy of Sciences Special
Report 176, Washington, DC, pp. 1133, and Source of slow earth
flow diagram: Karen H. Keefer. (b) Oblique aerial view of rotational
slump triggered by the 17 January 1994 Northridge, California (M 6.7)
event. Note unpaved road along upper part of image above landslide,
for scale. Reproduced from Keefer, D.K., 1999. Earthquake-induced
landslides and their effects on alluvial fans. Journal of Sedimentary
Research 69, 84104, with permission from Geoscience.
severe on ridge crests and areas of steep slope than on adjacent

areas of gentle slope or level ground (Paolucci, 2002; Sepulveda
et al., 2005a, b). Many studies have also shown that ground
shaking can induce complex, transient changes in both porewater pressures and shear resistance in some materials that influence slope stability under seismic conditions (See, e.g., Seed,
1968). With sufficient surveying, subsurface sampling, testing
of slope materials, evaluation of pore-water pressures, and
input of predicted seismic ground motions, the potential for
slope failure may be evaluated on a quantitative basis. However,
given the large numbers of potential slope failures and vast
areas affected by landsliding during large earthquakes, the
geomorphic impacts of such landsliding currently may be better
described by relating landslide distributions to general seismic
parameters, such as earthquake magnitude and shaking intensities, and to observable geologic, topographic, and hydrologic
attributes of slopes.
5.11.2.2.1
Distribution of landslides related to

earthquake magnitude and shaking intensities
As with the numbers of landslides, the areas throughout

which landslides are triggered by earthquakes correlate with
(b)
Figure 4 Category III landslideslateral spreads and flows. (a) Lateral

spread, consisting of coherent blocks moving on subsurface zone of
liquefied granular material, or fluidized sensitive clay. Modified with
permission from Varnes, D.J., 1978. Slope movement types and
processes. In: Schuster, R.L., Krizek, R.J. (Eds.), Landslides Analysis
and Control. Transportation Research Board National Academy of
Sciences Special Report 176, Washington, DC, pp. 1133. (b) Oblique
aerial view of rapid soil flow (mud flow) in volcanic ash triggered by
14 September 1984 Nagano-ken-Seibu, Japan earthquake (M 6.2).
Mud flow is approximately 1 km long. Reproduced from Keefer, D.K.,
1999. Earthquake-induced landslides and their effects on alluvial fans.
Journal of Sedimentary Research 69, 84104, with permission from
Geoscience.
earthquake magnitudes (Figure 5). With rare possible exceptions (See Mathews and McTaggart, 1978; Voight, 1978;
Keefer, 1984a; Rodrguez et al., 1999; Papadopoulos and
Plessa, 2000, for such possible exceptions involving earthquakes as small as M 2.9), earthquakes with Mo4 do not
produce landslides. In the largest earthquake for which the
limits of landslide occurrence have been determined (1964
Alaska, M 9.2) landslides occurred throughout an area of
approximately 269 000 km2 (Plafker et al., 1969). Of course,
the density or concentration of landslides varies greatly within
the overall areas affected, and detailed analyses of landslide
257
1 000 000
Area affected by landslides (A)
100 000
Rodrguez et al., 1999
10 000
1000
100
Keefer, 1984
10
1
4
4.5
5.0
5.5
6.0
6.5
7.0
7.5
8.0
8.5
9.0
9.5
distributions in several recent earthquakes show that in many

cases, most landslides occur in zones of high landslide concentration near the earthquake-generating fault ruptures
(Harp and Jibson, 1996; Fukuoka et al., 1997; Bozzano et al.,
1998; Esposito et al., 2000; Keefer, 2000, 2002; Bommer and
Rodrguez, 2002; Hancox et al., 2002; Jibson et al., 2004).
Other measures of the overall distribution of landslides triggered by earthquakes are maximum distances from epicenters
or fault ruptures at which landslides have been triggered,
which are as great as several hundred kilometers for the largest
earthquakes (Figure 6), and minimum shaking intensities
associated with landslide occurrence (Figure 7).
Keefer (1984a) did not find any significant differences in
measures of landslide distribution attributable to regional
differences in seismic attenuation, and in a detailed regional
study of earthquake-induced landslides in Central America,
Bommer and Rodrguez (2002) concluded that areas affected
by landslides there were similar to those in other parts of
the world. In contrast, Hancox et al. (2002) found that areas
affected by landslides and maximum epicentral distances
for landslides in New Zealand earthquakes tended to be
smaller on average than indicated by the worldwide data,
whereas minimum earthquakes magnitudes and minimum
shaking intensities associated with landslide occurrence
there tended to be higher. Suggestions that some regions may
rather produce earthquake-triggered landslides over anomalously large areas and great distances come from studies
of a few recent events in the Colorado Plateau region
of western North America (Harp et al., 1993; Jibson and Harp,
1996), and Quebec in eastern North America (Rodrguez et al.,
1999).
Maximum distance of landslides from earthquake epicenter (km)
Figure 5 Earthquake magnitude vs. area affected by landslides. Diamonds are data from individual historical earthquakes and solid line is upper
bound through these data, showing maximum area likely to be affected by landslides, reproduced with permission from Keefer, D.K., 1984a.
Landslides caused by earthquakes. Geological Society of America Bulletin 45, 406421. Squares are data from individual historical earthquakes,
reproduced from Rodrguez, C.E., Bommer, J.J., Chandler, R.J., 1999. Earthquake-induced landslides: 19801997. Soil Dynamics and Earthquake
Engineering 18, 325346, with permission from GSA, and dashed line is upper bound through these data.
1000
500
200
100
50
20
10
5
2
1
0.5
0.2
0.1
4.0 4.5 5.0 5.5 6.0 6.5 7.0 7.5 8.0 8.5 9.0 9.5
Figure 6 Earthquake magnitude vs. maximum distance from the

epicenter to landslides in three main categories. Reproduced with
permission from Keefer, D.K., 1984a. Landslides caused by
earthquakes. Geological Society of America Bulletin 45, 406421.
Dashed line is upper bound distance of Category I landslides
(disrupted landslides); dot-dash line upper bound for Category II
landslides (coherent slides); and dotted line for Category III
landslides (lateral spreads and flows).
258
Number of earthquakes
12
10
8
6
4
2
0
IV
V
VI
VII
VIII
Minimum Modified Mercalli Intensity for
disrupted landslides
IV
V
VI
VII
VIII
Minimum Modified Mercalli Intensity for
coherent landslides
(a)
IX
9
8
7
6
5
4
3
2
1
0
(b)
IX
10
9
8
7
6
5
4
3
2
1
0
(c)
IV
V
VI
VII
VIII
IX
Minimum Modified Mercalli Intensity for lateral
spreads and flows
Figure 7 Minimum Modified Mercalli Intensities of ground shaking

at which landslides in the three main categories were reported in
historical earthquakes, Black bars are data from Keefer, 1984a and
gray bars are data from Rodrguez, 1999. (a) Category I landslides
(disrupted slides and falls). (b) Category II landslides (coherent
slides). (c) Category III landslides (lateral spreads and flows).
Reproduced from Keefer, D.K., 2002. Investigating landslides caused
by earthquakes a historical review. Surveys in Geophysics 23,
473510, with permission from Springer.
5.11.2.2.2
Topographic, hydrologic, and geologic

indicators of landslide occurrence
Relations between the occurrence of earthquake-induced

landslides and local topographic, hydrologic, and geologic
conditions are also exceedingly complex, with landslides occurring in environments ranging from high and steep, glaciated slopes composed of well-indurated bedrock to soft and
unconsolidated flood-plain sediments underlying areas of essentially level ground. Despite this complexity, some generalizations may be made. The highly disrupted landslides in
Category I that comprise the vast majority of all earthquakeinduced landslides (Table 1) originate on steep slopes. Minimum inclinations of source slopes for these landslides
range from approximately 151 for disrupted soil slides to approximately 401 for rock falls (Keefer, 1984a). Additionally,
the large and highly mobile rock avalanches are restricted to
slopes more than approximately 150 m high, and they almost
always originate on slopes undercut either by active fluvial
erosion or by geologically recent or active glacial erosion
(Keefer, 1984b).
Where Category I landslides involve bedrock, virtually any
type of rock may make up the landslide body. However, rock
masses with certain attributes are particularly susceptible. Rock
materials that are probably most susceptible to slope failure in
earthquakes are those that are so intensely weathered that they
are gradational to residual soils and consist of heterogeneous
mixtures of cobbles and boulders surrounded by fine matrix
(Keefer, 1993). These intensely weathered rocks have produced especially abundant landslides in rain forest areas of
high annual precipitation (Pain, 1972; Nielsen et al., 1977;
Garwood et al., 1979; Schuster, 1991; Schuster et al., 1996).
Next highest in susceptibility are weakly cemented rocks
(Keefer, 1993), notably including young volcanic ashes exposed on steep slopes such as are abundant throughout seismically active areas of Central America (Harp et al., 1981;
Rymer, 1987; Bommer and Rodrguez, 2002; Rose et al., 2004)
and young, poorly indurated sedimentary rocks such as are
widespread throughout the California Coast Ranges (Harp
and Keefer, 1990; Harp and Jibson, 1996; Keefer, 1998, 2000)
and other areas of active tectonic uplift. Category I landslides
may also involve most or all types of regolith materials, but
Keefer (1984a) found they were most abundant in sandy residual and colluvial soils that were unsaturated as well as in
saturated volcanic materials containing layers of clay that were
sensitive, losing significant strength when disturbed. The coherent, Category II landslides may also involve either bedrock
or regolith (Table 1). Those in regolith, which are the most
numerous, occur in both alluvial environments where they
typically form along channel banks and bluffs in granular
materials and in colluvial environments on slopes as gentle
as approximately 51 (Table 1), where they typically involve
finer grained materials (Keefer, 1984a). These landslides can
reactivate existing landslides, but documented instances of
such reactivations caused by earthquakes are relatively rare
(Keefer, 1984a). The less common Category II landslides involving bedrock occur on slopes steeper than approximately
151 (Table 1), most commonly in poorly cemented, closely
jointed, sheared or weathered rock, especially where prominent discontinuities dip out of slopes, forming weak planes
where basal shear surfaces can form (Keefer, 1984a). Of the
partly to completely fluidized landslides in Category III, the
most common are lateral spreads (Table 2). These mostly affect low-lying, alluvial and coastal, sedimentary deposits with
a sandy texture, and commonly involve very gently sloping
ground (down to slopes of less than 11 (Youd, 1975; Keefer,

1984a)). The more completely fluidized mud flows and debris
flows can also originate on gentle slopes (down to approximately 21; Keefer, 1984a). In addition to the loess of central
Asia, described above, susceptible materials include a variety
of alluvial, colluvial and coastal sediments, most commonly
with a sandy texture, and certain types of volcaniclastic sediments consisting of interbedded layers of scoriaceous gravel
and sensitive fine ash (Keefer, 1984a).
A number of methods have been developed to produce
maps showing the potential for earthquake-induced landslide
generation across a landscape, given that landscapes geology,
topography, and hydrology and seismicity. These are based on
physical slope-stability analyses, statistical correlations, empirical data, or combinations of these approaches (Wieczorek
et al., 1985; Keefer, 1993; Jibson et al., 1998, 2000; Keefer
and Wang, 1998; Wang et al., 1998; Miles and Ho, 1999;
Hofmeister et al., 2000; Luzi et al., 2000; Miles and Keefer,
2000, 2007, 2009a, b; Romeo, 2000; Jibson and Jibson, 2003;
Murphy and Mankelow, 2004; Jibson, 2007, 2011; Lee et al.,
2008; Jibson and Michael, 2009).
5.11.2.2.3
Detailed landslide distribution from an

individual earthquake: example from the 1989
Loma Prieta, California earthquake (M 6.9)
The development of Geographic Information Systems (GIS)

technology has enabled comprehensive and detailed analyses
of the distribution of landslides triggered by earthquakes. One
of the first earthquakes for which such a GIS-based analysis
was carried out was the 17 October 1989 Loma Prieta earthquake (M 6.9), which occurred in the San Francisco BayMonterey Bay region, in the Coast Ranges of central
California. Fault rupture, involving largely strike-slip movement with a smaller thrust component, took place on a segment of the near-vertical San Andreas Fault, in a mountainous
area south of San Francisco. In the epicentral area are a wide
variety of rocks of different ages, lithologies, and degrees of
induration.
In the analysis of landslides caused by this earthquake,
Keefer (2000) related the Landslide Concentration (defined as
the number of landslide sources per square kilometer of surface area) to distances from the epicenter (Figure 8(a)) and
fault rupture, slope inclinations (Figure 8(b)), and lithologies
(Figure 8(c)). Landslide Concentrations showed strong direct
statistical correlations with slope inclination (Figure 8(b))
and strong inverse correlations with distance from the epicenter (Figure 8(a)) and fault rupture. Relations between Landslide Concentration and lithology were more
complex, but statistically significant differences in Landslide
Concentration were evident between the best indurated and
the least indurated rocks in the affected area (Keefer, 2000;
Figure 8(c)). Analysis of the distribution of slope inclinations
within each mapped geologic unit showed that the anomalously low Landslide Concentrations in unconsolidated and
semiconsolidated materials (Figure 8(a)) were attributable to
the relatively gentle slopes associated with them (Keefer,
2000). Other recent earthquakes for which GIS-based analyses
of landslide distribution have been performed include the
1994 Northridge, California (Jibson et al., 2000; Parise and
Jibson, 2000); 1999 Chi-Chi, Taiwan (Laio and Lee, 2000;
259
Wang et al., 2002; Khazai and Sitar, 2003; Meurnier et al., 2008);
2004 Niigata Ken Chuetsu, Japan (Sato et al., 2005); and 2005
Kashmir (Sato et al., 2006; Kamp et al., 2008, 2010; Owen
et al., 2008) earthquakes as well as the 2008 Wenchuan, China
earthquake, for which many such studies are currently
underway (see, for example, Sato and Harp, 2009 and Xu
et al., 2009 for some initial results).
5.11.3
Geomorphic and Postearthquake Effects of

Earthquake-Induced Landslides
Earthquake-induced landslides of course produce immediate,

local geomorphic effects from their removal, transport, and
deposition of earth material and from such other effects as
fissuring of adjacent slope areas. The most typical effects vary
with the type of landslide that occurs (Table 1). Category I
landslides commonly strip vegetation, surficial soil, regolith,
and in many cases bedrock from sources high on ridge flanks,
producing a local parallel slope retreat while adding loose
sediment to talus accumulations, fans and pediments, valley
floors, and directly into channels. Occurring in large numbers,
especially in great earthquakes, such landslides can greatly
alter the slope, topography, and configuration of low-order
drainage networks, while choking higher-order channels with
sediment. More complex geomorphic effects associated with
the largest of these landslides, as described above for the
Nevado Huascaran rock avalanche, may extend for distances
of many tens of kilometers downstream from the landslide
sources. The more coherent Category II landslides, which
typically transport material only a few meters or tens of
meters, fissure and disrupt slope areas, producing hummocky
topography and such effects as sag ponds, which alter the local
hydrology. Of the Category III landslides, lateral spreads
commonly create complex networks of open fissures, grabens,
and channels in the soft sediment of river flood plains or in
coastal environments, which may in turn alter the configurations of bluffs, river channels, or low-lying coastal areas. Mud
flows and debris flows can move material long distances,
scouring near their sources and creating both channel and
overbank deposits farther downslope and downstream. As
described above, subaqueous landslides may carry away large
parts of the distal margins of deltas. One general measure of
the geomorphic effect of earthquake-induced landslides is the
total volume of material moved downslope by the landslides
triggered by an earthquake. Perhaps the earliest determination
of such a volume for an individual earthquake was that of
Mathur (1953), who estimated that the M 8.6 Assam, India
event in 1950 moved 47 billion m3 from slopes, still the
largest known amount from any historical earthquake. The
variability in production of landslide material by individual
earthquakes, however, is also illustrated by studies after two
other earthquakes in the Himalaya, as the 1991 and 1999
Garhwal earthquakes (M 6.8 and 6.6, respectively) were determined to have dislodged relatively small amounts of
landslide material from slopes (Owen et al., 1996; Barnard
et al., 2001).
Keefer (1994) presented data on total volume of landslide
material produced by 15 historical earthquakes (Figure 9) and
concluded that for earthquakes occurring in plate-boundary
260
14
4.0
13
11
3.5
LC = 14.675re1.3284
(R2 = 0.97)
Landslide concentration, LC (landslides km2)
12
10
9
8
7
6
5
4
3
2
3.0
2.5
2.0
1.5
LC = 1.5897 tan + 0.17045

(R2 = 0.86)
1.0
0.5
LC = 0.15159 exp(3.2602 tan)

(R2 = 0.90)
1
0
0
(a)
0
5
10 15
20
25
30
Distance from epicenter, re (km)
0.2
(b)
0.4
0.6
Weakly cemented sandstone and siItstone

Moderately cemented mudstones, siltstones, and shales
Moderately cemented sandstones
Heterogeneous sedimentary rock sequence
Franciscan Complex
Mindego Basalt
Granitic rocks
Metasedimentary rocks
Mafic volcanics
Serpentinite
Well-indurated igneous, metamorphic, and Franciscan rocks
0.8
Tangent of slope angle ()
Unconsolidated and semiconsolidated sediments
Holocene undivided
Aromas Sand
Santa Clara Formation
Purisima Formation
Santa Cruz Mudstone
Monterey Formation
Lambert Shale
San Lorenzo Formation
Rices Mudstone Member
Twobar Shale Member
Locatelli Formation
Santa Margarita Sandstone
Temblor Sandstone
Lompico Sandstone
Vaqueros Sandstone
Zayante Sandstone
Butano Sandstone
Great Valley Sequence
(c)
35
LC = 5.9033 tan
2.8225
(R2 = 0.71)
0.5
1.0
1.5
2.0
2.5
1.0
1.2
1011
1010
Volume (m3)
109
log10 V=1.45 M 2.50
R2 = 0.88
108
107
106
105
104
5.0
5.5
6.0
6.5 7.0 7.5

Magnitude
8.0
8.5
9.0
Figure 9 Earthquake magnitude vs. total volume of landslides

generated by earthquake for 15 historical earthquakes. Solid line is
best-fit linear-regression line through data. Reproduced from Keefer,
D.K., 1994. The importance of earthquake-induced landslides to
long-term slope erosion and slope-failure hazards in seismically
active regions. Geomorphology 10, 265284.
regions containing widespread slopes susceptible to landslide

generation, total volume of landslide material, V, could be
related to earthquake magnitude, M, by the linear regression
equation
log 10V 1.45M 2.50: R2 0.876 for 5.3rMr8.6 [1]
By combining a relation such as eqn [1] with regional magnitude-frequency data on earthquake recurrence, it is possible
to estimate the long-term contribution of earthquake-induced
landslides to sediment production, erosion, and denudation
in a region. The first such study was carried out by Adams
(1980) for the Southern Alps of New Zealand. Using data
Adams (1980) developed on volumes of landslides produced
by three historical earthquakes and combining those data with
estimates of earthquake recurrence rates and measurements of
sediment produced by other processes, he concluded that
landslides triggered by earthquakes generated between 41 and
68 percent of all sediment produced in that region. In a later
261
study, using a similar method, combining eqn [1] with

regional earthquake-magnitude-frequency relations, Keefer
(1994) modeled the potential long-term contribution of
earthquake-induced landslides to erosion rates in several
regions (Figure 10). In many of these regions including the
San Francisco Bay region, the Sierra Nevada-Great Basin
region, onshore California as a whole, the island of Hawaii,
New Guinea, Turkey, and Peru, Keefer (1994) found that the
erosion rate from earthquake-induced landslides is of the
same order as the erosion rate from fluvial sediment discharge
from those regions, indicating that landslides generated by
earthquakes are a major component of the total regional
erosion. Additional analysis of erosion rates due to earthquake-induced landslides by Malamud et al. (2004) indicated
that typical seismically induced erosion rates in active subduction zones are 2007000 mm ka1 and in regions adjacent
to plate-boundary strike-slip fault zones is 10700 mm ka1.
Continuing geomorphic impacts from earthquake-induced
landslides in the months or years following an earthquake are
evident from several studies in a wide variety of environments.
Probably the earliest such study was that of Simonett (1967),
who described post-event effects from a 1935 earthquake
(Richter surface-wave magnitude, Ms 7.9) in Papua New
Guinea in the following way: Granitic areas have...extensive
gullying of deeply weathered bedrock, once landsliding breaches the subsoils. In the years following the earthquake,
stream valleys have become choked with alluvial fill, derived
from continued gullying of weathered granite. After another
earthquake in Papua New Guinea (1970, Ms 7.1), in which
landslides denuded 25% of slope area throughout a region of
240 km2, Pain (1972) and Pain and Bowler (1973) reported
that within a month after the earthquake slopes denuded by
the landslides had been extensively rilled and virtually all
landslide deposits had been eroded from lower slopes by intense rains, whereas within 6 months more than half of all
landslide material generated by the earthquake had been
carried out to sea by fluvial erosion. Similarly, in New Zealand
following the 1986 Inangahua earthquake (Ms 7.1), Adams
(1980) found that about half of the sediment attributable to
the earthquake-generated landslides was transported to the sea
within 19 months. In contrast, Pearce and Watson (1986)
found that even 50 years after the 1929 Buller, New Zealand
earthquake (Ms 7.6), most of the earthquake-generated
landslide material was still in storage close to the landslide
sources. During the M 8.6 Assam, India earthquake of 1950,
much of the 47 billion m3 of earthquake-generated landslide
material (Mathur, 1953) was transported onto valley floors or
into channels. There the landslide deposits created many
temporary dams that later burst, causing repeated floods that
aggraded low-lying areas by as much as 8 m (Mathur, 1953);
one such landslide dam burst 4 days after the earthquake,
inundating approximately 800 km2 and killing 500 people
Figure 8 Landslide Concentration, LC number of landslide sources/km2 of surface area compared to distance from epicenter (in 1-km-wide
bands), slope inclination, and areas underlain by various geologic formations in epicentral region of 17 October 1989 Loma Prieta, California
earthquake (M 6.9). (a) Distance from earthquake epicenter vs. LC: solid line is exponential best-fit curve to data. (b) Slope inclination vs. LC:
solid line is exponential best-fit curve to data; dashed lines are linear-regression best fits to slope inclinations, for intervals respectively less than
and greater than 341. (c) LC for various geologic formations in epicentral region. Reproduced from Keefer, D.K., 2000. Statistical analysis of an
earthquake-induced landslide distributionFthe 1989 Loma Prieta, California event. Engineering Geology 58, 231249.
262
Mean erosion rate (mm ka1)
600
500
400
300
200
100
ut
he
So
ra
er
Si
Sa
Fr
ev
an
ci
ad
sc
a-
re
ba
at
re
Ba
gi
on
si
n,
,C
al
ifo
r
ni
a
C
al
ifo
rn
C
rn
ia
al
C
ifo
al
rn
ifo
ia
rn
(a
ia
ll
on
sh
Is
or
la
nd
e)
of
H
aw
N
ai
ew
i
W
es
Ze
te
a
rn
la
n
N
d
ew
G
ui
ne
a
Ira
Tu
n
rk
(s
ey
in
ce
70
Ira
0
n
C
(1
E)
90
C
en
1
tra
19
lJ
81
ap
)
C
en
an
tra
Ti
(s
be
lJ
in
ce
ap
t
an
71
5
Pe
(s
C
in
ru
E)
ce
an
15
d
86
ad
ja
C
ce
E)
nt
re
gi
on
s
Figure 10 Erosion rates from earthquake-induced landslides, calculated for several regions. Modified from Keefer, D.K., 1994. The importance
of earthquake-induced landslides to long-term slope erosion and slope-failure hazards in seismically active regions. Geomorphology 10, 265284.
(Poddar, 1953). Substantial changes occurred in the channel

of the regions main river, owing to aggradation from landslide
material and to extensive fissuring of the channel banks
(Poddar, 1953). Effects of the fluvial transport of sediment
derived from the landslides were observed a year after the
earthquake, when the annual monsoon produced unprecedented flooding in the region owing to extensive postearthquake channel aggradation (Gee, 1953). Following the
more recent 2005 Kashmir earthquake (M 7.3) in the Himalaya, in contrast, studies conducted by repeat photography in
the 22 months after the earthquake, showed an increase in
landslide area at only 9% of earthquake-induced landslide
sites, whereas little or no change had occurred at 80% of the
sites (Khattak et al., 2010).
In a particularly well documented example of sediment
transport related directly to an individual earthquake-triggered
landslide, millions of cubic meters of sediment, derived from
the Nevado Huascaran rock avalanche and other landslides
triggered by the M 7.9 earthquake in 1970, were transported to
the mouth of the Santa River (Plafker et al., 1971; Evans et al.,
2009a). There, the sediment built a temporary delta front. By
1975, the delta front had prograded seaward by as much as
1 km, and much of the sediment had been transported
northward by the prevailing longshore currents, forming a
prominent beach ridge (Moseley et al., 1992). Farther south
along the Peruvian coastal plain, pervasive geomorphic effects
from earthquake-induced landslides and related ground failures were observed following the 2001 (M 8.4) earthquake,
(Keefer and Moseley, 2004). There a shattered landscape was
created, where: (1) abundant landslides delivered hundreds
of cubic meters of loose sediment to each linear kilometer
of dry arroyo channel (2) shallow hillside rills feeding
those channels were widened by extensive caving of their
banks; (3) subsurface pipes connecting the rills with the

arroyos collapsed, lengthening and integrating the channel
system; and (4) many upslope areas not affected directly by
landslides experienced extensive cracking and shattering of the
surficial regolith, rendering such areas susceptible to subsequent erosion. Such erosion as well as significant transport
of sediment through these shattered drainage systems was
already taking place during the first rainstorm to occur after
the earthquake (Keefer and Moseley, 2004).
The most extensive studies of postearthquake erosion and
sediment movement have been carried out for the 1999 ChiChi, Taiwan earthquake (M 7.5), which triggered approximately 26 000 landslides throughout an area of 6525 km2
(Wang et al., 2002). Dadson et al. (2003, 2004, 2005) analyzed sediment concentrations in the rivers draining the epicentral region following the earthquake and found up to a
fourfold increase in sediment concentration during the winter
(typhoon-free) seasons, with the largest increase occurring in
an area adjacent to the fault rupture. These sediment concentrations gradually decreased to preearthquake levels over a
period of approximately 3 years, whereas sediment concentrations associated with the frequent typhoons that affect the
region remained elevated. Comparisons of pre- and postearthquake landslide occurrence associated with typhoons
also showed greatly elevated levels of landslide occurrence; in
a study area in the epicentral region, Lin et al. (2006) found
that landslide occurrence associated with postearthquake
Typhoon Toraji, as measured by the total surface area disturbed by landslides, was five times greater than landslide
generation associated with preearthquake Typhoon Herb, even
though Herb brought more precipitation.
As with the Assam earthquake, many other historical
earthquakes have triggered landslides that blocked rivers,
creating landslide-dammed lakes, altering the local geomorphology and creating potential dam-burst flood hazards
that persist long after earthquake shaking has ended. Examples
of recent earthquakes that produced such dams include
the 1999 Chi-Chi, Taiwan (M 7.5) (Chen et al., 2003; Chang
et al., 2006), 2005 Kashmir, Pakistan-India (M 7.3) (Harp and
Crone, 2006; Dunning et al., 2007; Owen et al., 2008), and
2008 Wechuan, China (M 7.9) earthquakes (Xu et al., 2009;
Wu et al., 2010). In Taiwan, the 120 106 m3 Tsaoling rock
avalanche dammed the Ching-Shui river for a length of more
than 4 km, creating a lake of approximately 45 106 m3 in
volume (Chen et al., 2003) and the 51 106 m3 Jiufengershan
landslide created three smaller lakes (Chang et al., 2006).
The largest landslide triggered by Kashmir earthquake, a
rock avalanche with a volume estimated at 68 106 m3 to
80 106 m3 (Harp and Crone, 2006; Dunning et al., 2007;
Owen et al., 2008) blocked the Tang and Karli rivers at a
locality where they converged, creating lakes that had grown
to lengths of 400 and 800 m long, respectively, within approximately 2 months after the earthquake (Harp and Crone,
2006). The Wenchuan, China earthquake created more than
250 landslide dams (Xu et al., 2009), the largest of which
resulted from a landslide with a volume of 742 106 m3 (Wu
et al., 2010). Whereas in all these events, the identified landslide dams were created by large landslides, several landslide
dams created by the 1989 Loma Prieta, California earthquake
(M 6.9) show that such dams may also be created by relatively
small landslides (Schuster et al., 1989). In addition to creating
lakes that may last for greatly differing lengths of time, geomorphic effects of landslide dams include deposition of
lacustrine and deltaic sediments upstream from the dams,
diversion of channels by the landslide deposits themselves,
shifting of channels downstream from the dams due to
transport of sediments from the landslide dams or their associated lakes, and increased potential for landslide generation along the shores of the lakes created by the landslide
dams (Schuster, 2006).
5.11.4
Conclusions
Landslides are pervasive and widespread consequences of

earthquake shaking, with the largest earthquakes producing
thousands or tens of thousands of landslides over areas of tens
of thousands of square kilometers, or more. These landslides
are of many different types, with differing characteristics that
determine their local geomorphic effects. The most numerous
of these landslides, shallow falls and slides in regolith and
bedrock, commonly strip vegetation and surficial material off
slopes, transporting material to talus accumulations, fans, or
valley floors at the bases of the steep slopes on which they
originate. In aggregate, such landslides can effect large slope
areas, altering the local and regional landscape in ways that
persist long after the earthquake shaking has ceased. Other
types of landslides can disrupt and alter vast stretches of river
flood plains, coastal lowlands, and offshore areas. The largest
earthquake-induced landslides, which have volumes of more
than 100 million m3 and transport material for distances on
the order of kilometers, can have profound local geomorphic
effects as well posing substantial risks for human populations.
263
Existing knowledge about earthquake-induced landslides

derived from post-earthquake field investigations, regional
and worldwide synthesis of these data, and statistical and
analytical studies has illuminated general landslide characteristics and abundance, relations to earthquake magnitude
and shaking intensities, the geologic environments most likely
to generate the landslides, some details of likely landslide
distributions across the landscape, and at least some of their
more prominent geomorphic effects. However, given the
complexity of both the occurrence of earthquakes that produce landslides and the tectonically active landscapes in which
these landslides occur, much work remains to be done in refining and extending our understanding of these landslides
and their geomorphic effects.
References
Adams, J., 1980. Contemporary uplift and erosion of the Southern Alps, New
Zealand. Geological Society of America Bulletin 91(1), 24 and 91(2), 1114.
Ashford, S.A., Sitar, N., 1997. Analysis of topographic amplification of inclined
shear waves in a steep coastal bluff. Bulletin of the Seismological Society of
America 87, 692700.
Barnard, P.L., Owen, L.A., Sharma, M.C., Finkel, R.C., 2001. Natural and humaninduced landsliding in the Garhwal Himalaya of Northern India. Geomorphology
40, 2135.
Bommer, J.J., Rodrguez, C.E., 2002. Earthquake-induced landslides in Central
America. Engineering Geology 63, 189220.
Bozzano, F., Gambino, P., Prestininzi, A., Scarascia-Mugnozza, G., Valentini, G.,
1998. Ground effects induced by the Umbria-Marche earthquakes of SeptemberOctober 1997, central Italy. Proceedings of the Eighth International Congress of
the International Association for Engineering Geology and the Environment,
Vancouver, Canada. A.A. Baklema, Rotterdam, pp. 825830.
Chang, K-J., Chan, Y- C., Taboada, A., Dominguez, S., 2006. Postseismic surface
processes in the Jiufengershan landslide area, 1999 Chi-Chi earthquake
epicentral zone, Taiwan. Engineering Geology 86, 102117.
Chen, T-C., Lin, M-L., Hung, J-J., 2003. Pseudostatic analysis of Tsao-Ling
rockslide caused by Chi-Chi earthquake. Engineering Geology 71, 3147.
Close, U., McCormick, E., 1922. Where the mountains walked: an account of the
recent earthquake in Kansu Province, China, which destroyed 100,000 lives.
The National Geographic Magazine 41, 445464.
Cluff, L.S., 1971. Peru earthquake of May 31, 1970; engineering geology
observations. Bulletin of the Seismological Society of America 61, 511533.
Coulter, H.W., Migliaccio, R.R., 1966. Effects of the Earthquake of March 27, 1964
at Valdez, Alaska. U.S. Geological Survey Professional Paper 542-C, 35 pp.
Dadson, S.J., Hovius, N., Chen, H., et al., 2003. Links between erosion, runoff
variability and seismicity in the Taiwan orogen. Nature 426, 648651.
Dadson, S.J., Hovius, N., Chen, H., et al., 2004. Earthquake-triggered increase in
sediment delivery from an active mountain belt. Geology 32, 733736.
Dadson, S., Hovius, N., Pegg, S., Dade, W.B., Horng, M.J, Chen, H., 2005.
Hyperpycnal river flows from an active mountain belt. Journal of Geophysical
Research 110, 13. F04016.http://dx.doi.org/10.1029/2004JF000244 F04016.
Davis, S.N., Karzulovc, J.K., 1961. Deslizamientos en el valle del ro San Pedro
Provincia de Valdivia Chile. Universidad de Chile, Facultad de Ciencias Fisicales
y Mathematicos, Anales, Santiago, vol. 20, 57 pp.
Davis, S.N., Karzulovc, J.K., 1963. Landslides at Lago Rinihue, Chile. Bulletin of
Del Gaudio, V., Wasowski, J., 2007. Directivity of slope dynamic response to
seismic shaking. Geophysical Research Letters 34, L12301. http://dx.doi.org/
10.1029/2007GL029842.
Dunning, S.A., Mitchell, W.A., Rosser, N.J., Petley, D.N., 2007. The Hattian Bala
rock avalanche and associated landslides triggered by the Kashmir Earthquake of
8 October 2005. Engineering Geology 93, 130144.
Esposito, E., Porfido, S., Simonelli, A.L., Mastrolorenzo, G., Iaccarino, G., 2000.
Landslides and other surface effects induced by the 1997 Umbria-Marche
seismic sequence. Engineering Geology 58, 353376.
Evans, S.G., Bishop, N.F., Fidel, L.S., Valderrama, P.M., Delany, K.B., Oliver-Smith,
A., 2009a. A reexamination of the mechanism and human impact of catastrophic
264
mass flows originating on Nevado Huascaran, Cordillera Blanca, Peru in 1962

and 1970. Engineering Geology 108, 96118.
Evans, S.G., Roberts, N.J., Ischuk, A., Delany, K.B., Morozova, G.S., Tutubalina, O.,
2009b. Landslides triggered by the 1949 Khait earthquake, Tajikistan and
associated loss of life. Engineering Geology 109, 195212.
Fukuoka, H., Sassa, K., Scarascia-Mugnozza, G., 1997. Distribution of landslides
triggered by the 1995 Hyogo-ken Nanbu earthquake and long runout mechanism
of the Takarazuka Golf Course landslide. Journal of the Physics of the Earth 45,
8390.
Garwood, N.C., Janos, D.P., Brokaw, N., 1979. Earthquake-caused landslides: a
major disturbance to tropical forests. Science 205, 997999.
Gee, E.P., 1953. The Assam earthquake of 1950. In: Ramachandra Rao, M.B. (Ed.),
A Compilation of Papers on the Assam Earthquake of August 15, 1950. National
Geophysical Institute, The Central Board of Geophysics Publication 1,
Hyperabad, India, pp. 101107.
Geli, L., Bard, P-Y., Jullien, B., 1988. The effect of topography on earthquake
ground motion: a review and new results. Bulletin of the Seismological Society
Hancox, G.T., Perrin, N.D., Dellow, G.D., 2002. Recent studies of historical
earthquake-induced landsliding, ground damage, and MM Intensity in New
Zealand. Bulletin of the New Zealand Society for Earthquake Engineering 35,
5994.
Harp, E.L., Crone, A.J., 2006. Landslides triggered by the October 8, 2005,
Pakistan Earthquake and associated landslide-dammed reservoirs. U.S.
Geological Survey Open-file Report 20061052, 13 pp.
Harp, E.L., Jibson, R.W., 1996. Landslides triggered by the 1994 Northridge,
California, earthquake. Bulletin of the Seismological Society of America 86,
S319S332.
Harp, E.L., Jibson, R.W., Keefer, D.K., 1993. Seismically induced landslides
triggered at extraordinary distances: evidence from the Springdale, Utah,
landslide. Geological Society of America Abstracts with Programs 25(6), A32.
Harp, E.L., Keefer, D.K., 1990. Landslides triggered by the earthquake. In: Rymer,
M.J., and Ellsworth, W.L. (Eds.), The Coalinga, California, Earthquake Sequence
of May 2, 1983. U.S. Geological Survey Professional Paper 1487, pp. 335348.
Harp, E.L., Wilson, R.C., Wieczorek, G.F., 1981. Landslides from the February 4,
1976, Guatemala Earthquake. U.S. Geological Survey Professional Paper 1204-A,
35 pp.
Hofmeister, J., Wang, Y., Keefer, D.K., 2000. Earthquake-induced slope instability:
methodology of relative hazard mapping, western portion of the Salem Hills,
Marion County, Oregon. Oregon Department of Geology and Mineral Industries
Special Paper 30, 73 pp.
Jibson, R.W., 2007. Regression models for estimating coseismic landslide
displacement. Engineering Geology 91, 209218.
Jibson, R.W., 2011. Methods for assessing the stability of slopes during
earthquakes-A retrospective. Engineering Geology 122, 4350.
Jibson, R.W., Harp, E.L., 1996. The Springdale, Utah, landslide: an extraordinary
event. Environmental and Engineering Geoscience 2, 137150.
Jibson, R.W., Harp, E.L., Michael, J.A., 1998, A method for producing
digital probabilistic seismic hazard landslide maps: an example from the Los
Angeles, California, area. U.S. Geological Survey Open-file Report 98113,
17 pp.
Jibson, R.W., Harp, E.L., Michael, J.A., 2000. A method for producing digital
probabilistic seismic landslide hazard maps. Engineering Geology 58, 271289.
Jibson, R.W., Harp, E.L., Schulz, W., Keefer, D.K., 2004. Landslides triggered by the
2002 Denali Fault, Alaska, earthquake and the inferred nature of the strong
shaking. Earthquake Spectra 20, 669691.
Jibson, R.W., Jibson, M.W., 2003. Java programs for using Newmarks method and
simplified decoupled analysis to model slope performance during earthquakes.
US Geological Survey Open-File report 2003-5, version 1.1.
Jibson, R.W., Michael, J.A., 2009. Maps Showing Seismic Landslide Hazards in
Anchorage, Alaska. U.S .Geological Survey Scientific Investigations Map, 3077.
101, 631642.
Kamp, U., Owen, L.A., Growley, B.J., Khattak, G.A., 2010. Back analysis of landslide
susceptibility zonation mapping for the 2005 Kashmir earthquake: an assessment
of the reliability of susceptibility zoning maps. Natural Hazards 54, 125.
Keefer, D.K., 1984a. Landslides caused by earthquakes. Geological Society of
Keefer, D.K., 1984b. Rock avalanches caused by earthquakes: source characteristics.
Science 223, 12881290.
Keefer, D.K., 1993. The susceptibility of rock slopes to earthquake-induced failure.
Bulletin of the Association of Engineering Geologists 30, 353361.
Keefer, D.K., 1994. The importance of earthquake-induced landslides to long-term

slope erosion and slope-failure hazards in seismically active regions.
Keefer, D.K. (Ed.), 1998. The Loma Prieta, California, earthquake of October 17,
1989 landslides. U. S. Geological Survey Professional Paper 1551-C, 185 pp.
Keefer, D.K., 1999. Earthquake-induced landslides and their effects on alluvial fans.
Journal of Sedimentary Research 69, 84104.
Keefer, D.K., 2000. Statistical analysis of an earthquake-induced landslide
distributionthe 1989 Loma Prieta, California event. Engineering Geology 58,
231249.
Keefer, D.K., 2002. Investigating landslides caused by earthquakesa historical
review. Surveys in Geophysics 23, 473510.
Keefer, D.K., Moseley, M.E., 2004. Southern Peru desert shattered by the great
2001 earthquake: implications for paleoseismic and paleo-El NinoSouthern
Oscillation records. Proceedings of the National Academy of Sciences of the
United States of America 101, pp. 1084511176.
Keefer D.K., Tannaci, N.E., 1981. Bibliography on landslides, soil liquefaction, and
related ground failures in selected historic earthquakes. U.S. Geological Survey
Open File Report 81-572, 38 pp.
Keefer, D.K., Wang, Y., 1998. A method for predicting slope stability for earthquake
hazard maps. Proceedings of the Sixth U.S. National Conference on Earthquake
Engineering, Earthquake Engineering Research Institute, Seattle, Washington, 15 pp.
115, 102108.
Khazai, B., Sitar, N., 2003. Evaluation of factors controlling earthquake-induced
landslides caused by the Chi-Chi earthquake and comparison with the
Northridge and Loma Prieta events. Engineering Geology 71, 7995.
Laio, H-W., Lee, C-T., 2000. Landslides triggered by the Chi-Chi earthquake. Asian
Conference on Remote Sensing, ACRS 2000, 4 pp.
Lee, C-T., Huang, C-C., Lee, J-F., Pan, K-L., Lin, M-L., Dong, J-J., 2008. Statistical
approach to earthquake-induced landslide susceptibility. Engineering Geology
100, 4358.
Lin, C-W., Liu, S-H., Lee, S-Y., Liu, C-C, 2006. Impacts of the Chi-Chi earthquake
on subsequent rainfall-induced landslides in central Taiwan. Engineering
Geology 86, 87101.
Lin, C-W., Shieh, C-L., Yuan, B-D., Shieh, Y-C., Liu, S-H., Lee, S-Y., 2003. Impact
of Chi-Chi earthquake on the occurrence of landslides and debris flows:
example from the Chenyulan River watershed, Nantou, Taiwan. Engineering
Geology 71, 4961.
Luzi, L., Pergalani, F., Terlien, M.T.J., 2000. Slope vulnerability to earthquakes at
subregional scale, using probabilistic techniques and geographic information
systems. Engineering Geology 58, 313336.
Malamud, B.D., Turcotte, D.L., Guzzetti, F., Reichenbach, P., 2004. Landslides,
earthquakes, and erosion. Earth and Planetary Science Letters 229, 4559.
Mathews, W.H., McTaggart, K.C., 1978. Hope rockslides, British Columbia, Canada.
In: Voight, B. (Ed.), Rockslides and Avalanches, 1 Natural Phenomena. Elsevier,
Mathur, L.P., 1953. Assam earthquake of 15th Aug., 1950. A short note on factual
observations. In: Ramachandra Rao, M.B. (Ed.), A Compilation of Papers on the
Assam Earthquake of August 15, 1950. National Geophysical Institute, The
Central Board of Geophysics Publication 1, Hyperabad, India, pp. 5660.
Meurnier, P., Hovius, N., Haines, J.A., 2008. Topographic site effects and the
location of earthquake induced landslides. Earth and Planetary Science Letters
275, 221232.
Miles, S.B., Ho, C.L., 1999. Rigorous landslide hazard zonation using Newmarks
method and stochastic ground motion simulation. Soil Dynamics and
Earthquake Engineering 18, 305323.
Miles, S.B., Keefer, D.K., 2000. Evaluation of seismic slope-performance models
using a regional case study. Environmental and Engineering Geoscience 6,
2539.
Miles, S.B., Keefer, D.K., 2007. Comprehensive Areal Model of Earthquake-Induced
Landslides: technical specification and user guide. U.S. Geological Survey
Open-File Report 20071072, 69 p.
Miles, S.B., Keefer, D.K., 2009a. Evaluation of CAMEL comprehensive areal
model of earthquake-induced landslides. Engineering Geology 104, 115.
Miles, S.B., Keefer, D.K., 2009b. Toward a comprehensive areal model of
earthquake-induced landslides. American Society of Civil Engineers Natural
Hazards Review 10, 1928.
Moseley, M.E., Wagner, D., Richardson, III J.B., 1992. Space shuttle imagery of
recent catastrophic change along the arid Andean coast. In: Johnson, L.L. (Ed.),
Paleoshorelines and Prehistory: An Investigation of Method. CRC Press, Boca
Raton, FL, pp. 215235.
Murphy, W., Mankelow, J.M., 2004. Obtaining probabilistic estimates of

displacement on a landslide during future earthquakes. Journal of Earthquake
Engineering 8, 133157.
Nielsen, N.N., Furomoto, A.S., Lum, W., Morrill, B.J., 1977. The Honomu, Hawaii,
Earthquake Report of Inspection. National Academy of Sciences, Washington,
D.C., 70 pp.
Owen, L.A., Kamp, U., Khattak, G.A., Harp, E.L., Keefer, D.K., Bauer, M.A., 2008.
Landslides triggered by the 8 October 2005 Kashmir earthquake.
Owen, L.A., Sharma, M.C, Bigwood, R., 1996. Landscape modification and
geomorphological consequences of the 20 October 1991 earthquake and the
July August 1992 monsoon in the Garhwal Himalaya. Zeitschrift fur
Pain, C.F., 1972. Characteristics and geomorphic effects of earthquake initiated
landslides in the Adelbert Range, Papua New Guinea. Engineering Geology 6,
261274.
Pain, C.F., Bowler, J.M., 1973. Denudation following the November 1970 earthquake
at Madang, Papua New Guinea. Zeitschrift fur Geomorphologie Supplementband
18, 92104.
Paolucci, R., 2002. Amplification of earthquake ground motion by steep topographic
irregularities. Earthquake Engineering and Structural Dynamics 31, 18311853.
Papadopoulos, G.A., Plessa, A., 2000. Magnitude-distance relations for earthquakeinduced landslides in Greece. Engineering Geology 58, 377386.
Parise, M., Jibson, R.W., 2000. A seismic landslide susceptibility rating of geologic
units based on analysis of characteristics of landslides triggered by the 17 January,
1994 Northridge, California earthquake. Engineering Geology 58, 251270.
Pearce, A.J., Watson, A.J., 1986. Effects of earthquake-induced landslides on
sediment budget and transport over a 50-year period. Geology 14, 5255.
Plafker, G., Ericksen, G.E., 1978. Nevados Huascaran avalanches, Peru. In: Voight,
B. (Ed.), Rockslides and Avalanches, 1 Natural Phenomena. Elsevier, Amsterdam,
pp. 277314.
Plafker, G., Ericksen, G.E., Fernandez, J.C., 1971. Geological aspects of the May 31,
1970, Peru earthquake. Bulletin of the Seismological Society of America 61,
543578.
Plafker, G., Kachadoorian, R., Eckel, E.B., Mayo, L.R., 1969. Effects of the
Earthquake of March 27, 1964 on Various Communities. U.S. Geological Survey
Professional Paper 542-G, 50 pp.
Poddar, M.C., 1953. A short note on the Assam earthquake of Aug. 15, 1950.
In: Ramachandra Rao, M.B. (Ed.), A Compilation of Papers on the Assam
Earthquake of 15th August, 1950. National Geophysical Institute, The Central
Board of Geophysics Publication 1, Hyperabad, India, pp. 3842.
Prestininzi, A., Romeo, R., 2000. Earthquake-induced ground failures in Italy.
Ramachandra Rao, M.B. (Ed.), 1953. A Compilation of Papers on the Assam
Earthquake of August 15, 1950. National Geophysical Institute, The Central
Board of Geophysics Publication 1, Hyperabad, India.
Rodrguez, C.E., Bommer, J.J., Chandler, R.J., 1999. Earthquake-induced landslides:
198097. Soil Dynamics and Earthquake Engineering 18, 325346.
Romeo, R., 2000. Seismically induced landslide displacements: a predictive model.
Rose, W.I., Bommer, J.J., Lopez, D.L., Carr, M.J., Major, J.J. (Eds.), 2004. Natural
hazards in El Salvador. Geological Society of America Special Paper 375, 502 pp.
Rymer, M.J., 1987. The San Salvador earthquake of October 10, 1986geologic
aspects. Earthquake Spectra 3, 435463.
Sato, H.P., Harp, E.L., 2009. Interpretation of earthquake-induced landslides
triggered by the 12 May 2008, M 7.9 Wenchuan earthquake in the Beichuan
area, Sichuan Province, China using satellite imagery and Google Earth.
Sato, H.P., Hasegawa, H., Fujiwara, S., Tobita, M., Koarai, M., Une, H., Iwahashi, J.,
2006. Interpretation of landslide distribution triggered by the 2005 Northern
Pakistan earthquake using SPOT 5 imagery. Landslides 4, 113122.
Sato, H.P., Sekiguchi, T., Kojiroi, R., Suzuki, Y., Iida, M., 2005. Overlaying
landslides distribution on the earthquake source, geological and topographical
265
data: the Mid Niigata prefecture earthquake in 2004, Japan. Landslides 2,

143152.
Schuster, R.L. (Ed.), 1991. Research Council Natural Disaster Studies 5. National
Academy Press, Washington, D.C., National.
Schuster, R.L., 2006. Impacts of landslide dams on mountain valley morphology.
In: Evans, S.G., Scarascia Mugnozza, G., Strom, A., Hermanns, R.L. (Eds.),
Landslides from Massive Rock Slope Failure. NATO Science Series: IV:
Earth and Environmental Sciences. Springer, Heidelberg, vol. 49, pp. 591616.
Schuster, R.L., Nieto, A.S., ORourke, T.D., Crespo, E., Plaza-Nieto, G., 1996. Mass
wasting triggered by the 5 March 1987 Ecuador earthquakes. Engineering
Geology 42, 123.
Schuster, R.L., Wieczorek, G.F., Hope, D.G. II. Landslide dams in Santa Cruz
County, California, resulting from the earthquake. In: Keefer, D.K. (Ed.), The
Loma Prieta, California, Earthquake of October 17, 1989 Landslides. U.S.
Geological Survey Professional Paper 1551-C, pp. 5170.
Seed, H.B., 1968. Landslides during earthquakes due to soil liquefaction.
Proceedings of the American Society of Civil Engineers, Journal of the Soil
Mechanics and Foundations Division 94, 10531122.
Sepulveda, S.A., Murphy, W., Jibson, R.W., Petley, D.N., 2005a. Seismically induced
rock slope failures resulting from topographic amplification of strong ground
motions: the case of Pacoima Canyon, California. Engineering Geology 80,
336344.
Sepulveda, S.A., Murphy, W., Petley, D.N., 2005b. Topographic controls on
coseismic rock slides during the 1999 Chi-Chi earthquake, Taiwan. Quarterly
Journal of Engineering Geology and Hydrogeology 38, 189196.
Simonett., D.S., 1967. Landslide distribution and earthquakes in the Bewani and
Torricelli Mountains, New Guinea a statistical analysis. In: Jennings, J.N.,
Mabbutt, J.A. (Eds.), Landform Studies from Australia and New Guinea.
Cambridge University Press, Cambridge, pp. 6484.
Tergazhi, K., 1950. Mechanism of landslides. In: Paige, S. (Ed.), Application of
Geology to Engineering Practice: Berkey Volume. Geological Society of America,
Varnes, D.J., 1978. Slope movement types and processes. In: Schuster, R.L.,
Krizek, R.J. (Eds.), Landslides Analysis and Control. Transportation Research
Board National Academy of Sciences Special Report 176, Washington, DC,
pp. 1133.
Voight, B., 1978. Lower Gros Ventre slide, Wyoming, U.S.A. In: Voight, B. (Ed.),
Rockslides and Avalanches, 1 Natural Phenomena. Elsevier, Amsterdam, pp.
113166.
Wang, G., Xu, B., 1984. Brief introduction of landslides in loess in China.
Proceedings of the Fourth International Symposium on Landslides, Toronto, vol.
1, pp. 197207.
Wang, W-N., Nakamura, H., Tsuchiya, S., Chen, C-C., 2002. Distributions of
landslides triggered by the Chi-chi earthquake in central Taiwan on
September 21, 1999. Landslides Journal of the Japan Landslide Society 38,
318326.
Wang, Y., Keefer, D.K., Wang, Z., 1998. Seismic hazard mapping in EugeneSpringfield, Oregon. Oregon Geology 60, 3141.
Wieczorek, G.F., Wilson, R.C., Harp, E.L., 1985. Map Showing Slope Stability
during Earthquakes in San Mateo County, California. U.S. Geological Survey
Miscellaneous Investigations Series Map I-1257-E.
Wu, S., Wang, T., Shi, L., et al., 2010. Study on catastrophic landslides triggered
by 2008 Great Wenchuan Earthquake, Sichuan, China. Journal of Engineering
Geology 18(2), 145159.
Xu, Q., Fan, X-M., Huang, R-Q., Van Westen, C., 2009. Landslide dams triggered
by the Wenchuan Earthquake, Sichuan Province, south west China. Bulletin of
Engineering Geology and the Environment 68, 373386.
Youd, T.L., 1975. Liquefaction, flow and associated ground failure. Proceedings
of the First U.S. National Conference on Earthquake Engineering. Earthquake
Engineering Research Institute, Ann Arbor, Michigan, pp. 146155.
Youd, T.L., Hoose, S.N., 1978. Historic ground failures in Northern California
associated with earthquakes. U.S. Geological Survey Professional Paper 993,
177 pp.
266
Biographical Sketch
Dr. David Keefer has a global reputation in both landslide and earthquake science. By bringing new insights and
methods to mapping and assessing hazards he has played a key role in helping public decision makers to improve
mitigation and avoidance while reducing costs. Much of Keefers career has been spent working at the United
States Geological Survey. He retired from the USGS in 2008 and is currently adjunct professor in the Climate
Change Institute, University of Maine. He is best known for his pioneering work on seismically induced landslides. In 1984, he produced the defining paper on this topic by analyzing in detail data on earthquake-triggered
landslides from a wide range of environments. His study generated key relationships between landslide occurrence and parameters such as the earthquake magnitude and the distance to the fault. This work has been
replicated on a number of occasions since, using data collected employing modern techniques. Keefers studies of
earthquake-induced landslides have developed into a wide range of themes but active involvement in areas
recently affected by earthquakes has remained a key component of his work. Substantial contributions have
included the definition of parameters for determining seismic slope-failure susceptibility from field, laboratory
and analytical studies, and the development of new methods for mapping earthquake-induced landslide hazards.
He was also a pioneer in the creation of landslide early warning systems, most notably in the San Francisco Bay
area, that developed the capability to issue public warnings to potentially affected communities. Keefer is the
author or co-author of 140 scientific publications focused on natural hazards, including several articles published
in such prestigious journals as Science and the Proceedings of the US National Academy of Sciences. He has been a stalwart in organizing international
symposia and edited volumes of research papers for the worldwide research communities that study earthquake-induced landslides, thereby defining much of
the current state-of-the-art in this field. Keefer is an outstanding and enthusiastic scientist on many levels. The quality of his work and his warm and open
nature have inspired countless others to work in this area. He recently received the Sergey Soloviev Medal of the European Geosciences Union for his work in
the area of natural hazards research and mitigation.
5.12 Paleoseismology
H Kondo, National Institute of Advanced Industrial Science and Technology, Tsukuba, Ibaraki, Japan
5.12.1
Introduction
5.12.1.1
Scope of Paleoseismology
5.12.1.2
Paleoseismologys Relation to Tectonic Geomorphology
5.12.2
Earthquake Recurrence Models
5.12.2.1
Characteristic Earthquake Model
5.12.2.2
Time-/Slip-Predictable Model
5.12.3
Recent Methodological Developments in Paleoseismology
5.12.3.1
Remote Sensing Technologies
5.12.3.2
Fault Trenching
5.12.3.3
Numerical Dating
5.12.4
On-Fault Paleoseismology
5.12.4.1
Fault Length and Dimension
5.12.4.2
Timing of Paleoearthquakes and Slip per Event
5.12.4.3
Long-term Slip Rate
5.12.5
Off-Fault Paleoseismology
5.12.5.1
Marine Terraces
5.12.5.2
Paleoliquefaction Features
5.12.5.3
Tsunami Deposits (Tsunamiites)
5.12.5.4
Earthquake-Triggered Landslides
5.12.5.5
Earthquake-Triggered Cracks
5.12.6
Contribution to Seismic Hazards
5.12.7
Challenges
Acknowledgments
References
Glossary
Bayesian analysis Statistical method that uses evidence to
update the state of uncertainty over competing probability
models.
Colluvial wedge A tapering slope deposit that
accumulates against a fault scrap shortly after an
earthquake.
Colluvium Loose sediment that build up against
slopes.
Coseismic Occuring during an earthquake.
Delta foresets Inclined beds of sediment that are
deposited on the front of a delta.
Digital elevation models (DEMs) Computer model or
three-dimensional representation of a terrains surface.
Fault scarps Topographic expression, usually a cliff of
break in slope, because of the displacement of the land
surface due to faulting.
Footwall Body of rock that occurs below a fault plane.
Flame structure Soft sediment deformation structure that
looks like an upward-pointing flame, and is usually a few
centimeters in size.
268
269
269
270
271
272
273
273
273
274
278
278
281
285
285
285
285
287
288
290
290
294
294
294

on Earth.
Hangingwall Body of rock that occurs above a fault plane.
Impermeable Pertaining to the inability for water to
permeate through a rock.
Interplate Between tectonic plates.
Interseismic Between earthquake events.
Intraplate Within a tectonic plate.
Involutions Soft sediment deformation structures that are
circular or sub-circular in form, usually a few centimeters
in size.
Lichenometry Method of aging a rock surface using lichen
growth.
Light detection and ranging (LiDAR or lidar) An optical
remote sensing technology, ground based or airborne, that
can measure distance, which is commonly used to produce
digital elevation models.
Kondo, H., Owen, L.A., 2013. Paleoseismology. In: Shroder, J. (Editor in

Chief), Owen, L.A. (Ed.), Treatise on Geomorphology. Academic Press, San
Diego, CA, vol. 5, Tectonic Geomorphology, pp. 267299.
http://dx.doi.org/10.1016/B978-0-12-374739-6.00092-0
267
268
Paleoseismology
Liquefaction Process resulting in unconsolidated

sediment being transformed into a substance that acts like a
liquid.
Megathrust earthquakes Amongst worlds largest
earthquakes, which occur in subduction zone settings.
Meizoseismal zone Area of maximum damage during an
earthquake.
Microfabric Microscope arrangement of grains or crystals
within a rock or sediment.
Mole tracks Series of small hills that can be traced across
the landscape and were produced during an earthquake.
Poissonian probability distribution The probability
of a given number of events occurring in a fixed interval of
time and/or space if these events occur with a known
average rate and are independent of the time since the
previous event.
Sag pond Small pond or lake produced by subsidence
related to extension along a strike-slip fault.
Sand blows A cone of sand, centimeters to many meters
in diameter, ejected to the surface, possibly during an
earthquake.
Sand dikes A vertical or subvertical body of sand intruded

into surrounding sediment possibly during an earthquake.
Seismogenic fault A fault that has, and can, produce
earthquakes.
Shutter-ridge An isolated ridge formed by displacement
of a spur by strike-slip faulting.
Sills Tabular intrusion of sediment that is parallel to the
intruded beds, which can be produced during an
earthquake. The term can also be used to describe the
intrusion of magma and formation of an igneous sill.
Strike A horizontal line on a bedding plane or fault.
Subduction zone An area on Earth where one tectonic
plate is moving under another tectonic plate.
Surface rupture Surface expression of an active fault
produced during an earthquake.
Tephrochronology Study and use of tephra to date
geologic successions.
U-series geochronology Dating of biogenic and
inorganic carbon that involves three U-series decay chains.
Abstract
Paleoseimology utilizes aspects of tectonic geomorphology, sedimentology, and stratigraphy to determine the location,
timing, and displacement of past earthquakes. Most studies concentrate on providing data on Late Quaternary earthquakes
to extend the historical seismic record. Earthquake recurrence models utilize paleoseismic data to help in earthquake
hazard assessment. Recent methodological developments, including remote sensing, geodesy, fault trenching, and numerical dating, have helped accelerate knowledge and study of past earthquakes. Paleoseismic studies include on-fault and
off-fault studies. The former involves mapping active fault traces and fault trenching. Off-fault studies mainly include
studying and examining uplift of marine terraces, paleoliquefaction features, earthquake-triggered landslides, and tsunami
deposits. Seismic hazard analysis is concerned with the long-term forecast of large earthquake and/or determining or
estimating the amount of ground motion likely during a future earthquake. Effective seismic hazard mapping includes
developing ground acceleration maps based on high-resolution, detailed geomorphic and Quaternary geologic mapping.
Notable future challenges for paleoseismology include reconstructing the timing and magnitude of past earthquakes
beyond a few tens of thousands of years to enhance understanding of the nature of earthquakes with long recurrence
intervals, improving models for earthquake cycles, and applying paleoseismic knowledge for effective seismic hazard
mitigation.
5.12.1
Introduction
Tectonic geomorphology has many applications, ranging from

academic topics including understanding the nature of landscape development and the record of surface processes across
Earths history to applied applications that include seismic
and volcanic hazard mitigation. Reconstruction of past seismic
events from geomorphic, sedimentologic, and/or stratigraphic
data has become a major area of growing interest within tectonic geomorphology and this area of study, known as
paleoseismology, has evolved during the past few decades.
This is particularly relevant because increased legislation in
many seismically active regions mandates earthquake hazard
assessment before undertaking major construction and/or
development projects. A good example is the Seismic Hazard
Mapping Act of 1990 that was passed in California in the
aftermath of the 16 October 1989 Loma Prieta earthquake
(Lew and Real, 2004). Major compilations of paleoseismology
include Paleoseismology (McCalpin, 1996, 2009a), which provides comprehensive reviews of a wide range of paleoseismic
topics, and Special Publication of the Geological Society,
London, volume 316 (Reicherter et al., 2009), which presents
a series of useful case studies.
This chapter reviews the contribution of tectonic geomorphology and allied disciplines including sedimentology,
stratigraphy, and geochronology to paleoseismic studies and
provides a framework for geomorphologists involved in tectonic studies. The chapter starts with describing the scope of
paleoseismology and follows by a discussion of earthquake
recurrence models, which are essential for the effective application of paleoseismic data for hazard assessment. It then
discusses the methods used to collect paleoseismic data, and
the geomorphic and associated evidence for past earthquakes
on faults and other features. This is followed by a brief
discussion of the contributions of paleosiesmology to other
disciplines, and the future challenges and opportunities.
Paleoseismology
269
Seismology
instrumental data
Paleoseismology: geological data
Historical data
10 000 years BP
5000
Hazard
0
Assessment
Figure 1 The temporal realm of paleoseismic study. After Morner, N.-A., 2011. Paleoseismology: the application of multiple parameters in four
case studies in Sweden. Quaternary International 242, 6575. Present seismic activity is recorded by seismic instruments, seismic activity in the
recent past is recorded by historical information, and paleoseismic activity is recorded primarily by geologic and geomorphic records.
5.12.1.1
Scope of Paleoseismology
Paleoseismology aims to reconstruct the timing, magnitude,

and location of past earthquakes based on geomorphic and
geologic evidence across a range of timescales and spatial dimensions (McCalpin, 1996; Yeats and Prentice, 1994). The
timescales of seismologic, historic, and geomorphic/geologic
studies overlap (Figure 1). Paleoseismological research, however, is mainly concerned with the study of the history of
earthquakes recorded in tectonic landforms, sediments, and
rocks, which result from Earth surface deformation due to coseismic and inter-seismic faulting, generally on timescales
ranging from 102 to 105 years. However, the range of paleoseismological timescales can extend from today to several
million years ago; for example, ranging from studies undertaken just after a recent earthquake to studies of microfabric in
ancient bedrock. In essence, paleoseismological studies provide the only data that yield information on the nature of
pre-instrumental and prehistoric earthquakes.
The Principle of Uniformitarianism (Hutton, 1795) is one
of the overriding foundations for paleoseismological studies,
and as such, modern observations are used to explain past
tectonic landforms and structures to reconstruct former
earthquake events. Another basic principle (or assumption) of
paleoseismology is that an earthquake is associated with a
particular fault displacement and that an earthquake fault has
been or will be active over several seismic cycles, that is,
earthquakes recur in both space and time. Knowledge of past
earthquake occurrence allows paleoseimologists to determine
the potential hazard of future earthquakes, including the
timing, size, and location of likely earthquakes, and to estimate the magnitude and distribution of potential ground
motion and, for those earthquakes affecting human-made
structures, the likely destruction. All these are essential for
hazard mitigation, particularly for reducing injury and death
to humans and their animals, and the destruction of land and
property.
A paleoseismologist determines the best type of data to
obtain and how it would contribute to seismic hazard analysis
before undertaking a paleoseismic study. These data could, for
example, include historical documentation, remote sensing
information, geodetic surveys, geomorphic mapping, and/or
fault trenching, and might be used to produce a map of the
extent of an active fault and/or likely ground acceleration.
While undertaking paleoseismic studies, it is important not to
lose sight of its ultimate role and contribution to earthquake
forecasting and prediction; and this chapter briefly reviews the

application of paleoseismology to practical seismic hazard
analysis to help illustrate this point.
5.12.1.2
Paleoseismologys Relation to Tectonic

Geomorphology
Paleoseismology is interdisciplinary and can include aspects of

tectonic geomorphology, seismology, geodesy, geochronology,
earthquake geology, Quaternary geology, structural geology,
archeology, history, engineering, architecture, economics,
sociology, and politics. Each field has its own methods and
approaches, and each considers different timescales. In some
cases, the different timescales and approaches result in different interpretations and explanations, despite the notion
there should be a unique answer to the problem being
addressed for the particular timescale being considered. As
such, there is clearly a need to examine the nature of
past earthquakes on a variety of timescales to help understand
different interpretations and explanations. Of particular
note is the relationship between paleoseismology and engineering, which includes geotechnical and civil engineering, as
well as architecture and structural design, and social sciences.
These fields are generally linked with applications to various
seismic hazards analysis, and the practical use of paleoseismology for public well-being, specifically safe building
design, construction, and remediation/retrofitting in seismically active regions. These aspects of the application of
paleoseismic studies are sometimes far from the original
scientific interests of the paleoseismologist or tectonic
geomorphologist, but they are an essential part of the
paleoseismologists job.
Tectonic geomorphology is particularly useful for paleoseismological studies. First, tectonic geomorphology provides
the data to help reconstruct cumulative displacement and
deformation on timescales that can extend back hundreds
of thousands or even millions of years. This is particularly
valuable for the study of large earthquakes that have long recurrence intervals. Most tectonic landforms are created during
co-seismic faulting at the surface and are the result of the
accumulation of fault movement. Cumulative faulting and
deformation due a succession of earthquakes may produce
topographic relief and a variety of landforms including fault
scarps, offset channels, shutter ridges, sag ponds, and mole
tracks. In addition, if the earthquake source fault is in a
270
Paleoseismology
nearshore-marine setting, such as along a subduction zone,

surface deformation of the hangingwall may be recorded
as uplifted landforms including marine terraces, beach
ridges, and emerged shell assemblages (see details in Section 5.12.5.1). Even when the magnitude of earthquake is
relatively small (oBM6.5), for example, produced by a blind
thrust, the cumulative deformation can be recorded as a deflection in the longitudinal river terrace profile, or a warp
across a pediment or an alluvial fan surface. In essence, most
tectonic landforms are directly produced by paleoearthquakes
and reflect repeated earthquake events. This allows the landforms to be used to obtain information on past earthquakes
if individual earthquake events can be differentiated, for
example, by using a combination of different age and/or
different types of landforms.
Second, it is possible to locate and map the trace of
an active fault using tectonic geomorphic data by, for
example, mapping fault scarps, flexure scarps, deflected
streams, shutter ridges, sag ponds, and/or break-in-slopes.
Locating the trace of an active fault is important because future
earthquakes are likely to occur along it. As described in more
detail below, the length and size of active faults make it possible to estimate the size of past earthquakes and provide a
means of predicting the likely size of future earthquakes.
Furthermore, one of the most common methods to reconstruct the timing of past earthquakes is fault trenching (see
details in Section 5.12.3.2). Tectonic geomorphology is essential for selecting sites for fault trenching, which requires
locating appropriate sites within a few meters of the active
fault. Thus, paleoseismology and tectonic geomorphology are
intimately related.
5.12.2
Since in reality it is generally difficult to acquire the necessary

field data along the entire length of a fault to predict its recurrence, Wallaces (1970) formula helps provide a more
practical way to assess the nature of faulting.
Wallaces (1970) proposed relationship (eqn [1]) relates
back to the elastic rebound model that was proposed by Reid
(1910) after the 1906 San Francisco earthquake. Reids (1910)
model assumed that elastic strain energy builds up in the region around a fault before an earthquake due to crustal
movements very far from the fault plane, with the energy increasing linearly over time (Figure 2). When the accumulated
strain reaches the critical value of fault strength, an earthquake
occurs to release the accumulated strain in the surrounding
region. Wallaces (1970) relationship (eqn [1]), predicated on
observations and field data, is consistent with the elastic rebound model as a first-order approximation (e.g., Yonekura,
1975; Yoshikawa et al., 1981; Thatcher, 1986).
Matsuda (1975) proposed the first earthquake recurrence
model for large earthquakes produced by active faults in
Japan. This model is known as the Koyuu-Jishin Setsu and
is an expanded version of Wallaces (1970) relationship.
Matsuda (1975) showed, using historically ruptured faults in
Japan, that there is a positive relationship between the length
Interseismic
strain accumulation
Coseismic
strain release
UX
UX
Horizontal
displacement
Earthquake Recurrence Models
du
1
All the terms in this relationship can be determined in the

field, based on geomorphic and geologic data (see details in
Section 5.12.3). Where two terms can be determined from the
field data, it is possible to estimate the remaining term(s).
2D-shear
strain
duX dy
dUX/dy
One of the fundamental goals of paleoseismology is to develop earthquake recurrence models for forecasting the occurrence of future destructive earthquakes for practical seismic
hazard analysis. Robert Wallace in the U.S. and Tokihiko
Matsuda in Japan pioneered paleoseismic research for understanding the nature of faulting for practical seismic hazard
analysis. These pioneers realized that field data are usually not
sufficient to evaluate the past seismic activity along the entire
length of an active fault and that there is a need to develop
recurrence models for large earthquakes to more fully assess
seismic hazards, even if any particular model might have large
uncertainties.
Wallaces major contributions to earthquake geology and
geomorphic studies stemmed from his studies along the San
Andreas fault. He proposed (Wallace, 1970) a simple relationship between earthquake recurrence interval (R), the
co-seismic displacement (D), the long-term slip rate (S), and
the creep rate (C; if the fault creeps) at a site such that
R D=S C
Fault plane
du
/
y
y
2D
Dislocation model:
cross section
y
D
Slip du
Slip rate du/dt

Figure 2 Model of the earthquake cycle for a strike-slip fault.
Reproduced from Burbank, D.W., Anderson, R.S., 2001. Tectonic
Geomorphology. Blackwell Science, Oxford, 274 pp., and modified
from Thatcher, W., 1986. Cyclic deformation related to great
earthquakes at plate boundaries. In: Reilly, W.I., Harford, B.E. (Eds.),
Recent Crustal Movements of the Pacific Region. Bulletin of the
Royal Society of New Zealand, vol. 24, pp. 245272, with permission
from Wiley and NZJR. The far-field strain (large arrows in top panel)
of one block with respect to the other remains constant through
time, as does the aseismic slip (du/dt in bottom panel) in the ductile
zone below the locking depth (D). Interseismic displacement is
greatest farthest from the fault, but most of the shear strain (dux/dy
in middle panel) occurs within two locking depths (2D) of the fault.
During coseismic displacement, the greatest displacement (du)
occurs along and near the fault and compensates for the slip deficit
developed during the interseismic interval. The co-seismic shear
strain is equal and opposite to the interseismic shear stain.
Paleoseismology
Observations
Cumulative slip
Variable-slip model
(a)
271
Variable displacement per event

at a point
Constant slip rate along length
Variable earthquake size
Distance along fault

Uniform-slip model
Cumulative slip
Constant displacement per event

at a point
Constant slip rate along length
Constant size large earthquakes:
more frequent moderate
earthquakes
(b)

Characteristic earthquake model
Cumulative slip
Constant displacement per event

at a point
Variable slip rate along length
Constant size large earthquakes:
infrequent moderate earthquakes
(c)
Figure 3 Models for slip accumulation (vertical axis) along a regional suite of faults. (a) Variable slip, (b) uniform slip, and (c) characteristic
earthquakes. Reproduced from Burbank, D.W., Anderson, R.S., 2001. Tectonic Geomorphology. Blackwell Science, Oxford, 274 pp., and modified
from Schwartz, D.P., Coppersmith, K.J., 1984. Fault behavior and characteristic earthquakes: examples from the Wasatch and San Andreas fault
zones. Journal of Geophysical Research 89, 56815698.
of an active fault and the magnitude of earthquakes that

are produced along it. This led to empirical laws to estimate
the size of earthquakes from the lengths of active faults. Following these laws, Matsuda (1975) stated that the largest
magnitude of earthquake for the specific fault or fault system
is repeated regularly and with the same amount of displacement. Combining Wallaces relationship, if the past slip rate
and displacement per event are known, then it is possible to
estimate the recurrence interval as well as the size and timing
of future large earthquakes.
Wesnousky et al. (1983) advanced the understanding of
earthquake recurrence by comparing the observed seismicity
for tens of years of instrumental time with the paleoseismic
record. Their study showed that the frequency of paleoearthquakes was much higher for smaller M4B6.5 earthquakes
than that predicted by the GutenbergRichter law (Gutenberg
and Richter, 1954), which expresses the relationship between
magnitude (M) and number of earthquakes as
log N a2bM
2
where N is the number of events having a magnitude ZM, and

a and b are constants.
Significantly, Wesnousky et al. (1983) showed that earthquake frequency distribution for a particular, single fault does
not necessarily satisfy the GutenbergRichter relationship, but
that the primary factor controlling the magnitudefrequency
distribution in regional seismicity studies is the relative
distribution of slip rates and lengths of preexisting faults.

Paleoseismological data are, therefore, extremely important in
forecasting destructive earthquakes of M4B7.0 rather than
smaller earthquakes that are generally recorded in seismogram
networks.
5.12.2.1
Characteristic Earthquake Model
Schwartz and Coppersmith (1984) proposed a characteristic

earthquake model in which a fault or a fault segment ruptures
repeatedly and exhibits approximately the same amount and
distribution of surface slip during each successive earthquake
(Figure 3). In this model, if the amount of slip is proportional
to earthquake magnitude (cf. Matsuda, 1975; Wells and
Coppersmith, 1994) then a fault will produce earthquakes of
the same size; earthquake magnitude is predictable and
knowledge of recurrence time is not required. The long-term
slip rate along a fault is assumed to vary due to the variability
of slip along strike. The model assumes persistence of a segment boundary that would stop rupture during earthquakes
during each seismic cycle. Therefore, if a fault produces characteristic earthquakes throughout progressive seismic cycles
then paleoseismic data for a single fault event would provide
valuable information on both past and future rupture patterns. Thus, the characteristic earthquake model and the assumptions within it are very important for predicting future
large earthquakes. This model predicts a higher frequency
of surface-rupturing earthquakes than those of instrumental
272
Paleoseismology
Stress
Periodic model
Time-predicable model
2
2
Cumulative coseismic
slip
(a)
Slip-predicable model
1
1
Predictable time
U
e
li
ts
t
ra
te
Unknown slip
sta
n
Co
nt
te
nt
Predictable time
sta
ra
n
Co
sta
ra
Unknown time
Co
Predictable slip
t
p
sli
p
sli
(b)
Time
Predictable slip
t
(c)
Figure 4 Models for earthquake recurrence. Reproduced from Burbank, D.W., Anderson, R.S., 2001. Tectonic Geomorphology. Blackwell
Science, Oxford, 274 pp., and modified from Shimazaki, K., Nakata, T., 1980. Time-predictable recurrence model for large earthquakes.
seismological studies, supporting the conclusions of Wesnousky et al. (1983).

Many different views exist on the repetition of slip distribution during successive large earthquakes (Figure 3). Sieh et al.
(1989) proposed, for example, a uniform slip model, in which a
fault or a fault section that produces relatively smaller slip
during an earthquake will repeat more frequently than a larger
slip section. The basic model is derived from surface slip data
associated with the 1857 M7.8 Fort Tejon earthquake, within
the double bend of the San Andreas fault. Sieh (1978b) conducted measurements of geomorphic offsets along the San
Andreas fault, which revealed that the slip distribution during
the 1857 event was not constant along the fault. Based on these
data, in addition to the chronology of paleoearthquakes, Sieh
(1978a) argued that a fault section recording lesser slip would
regain the slip deficit along the fault with more frequent
earthquakes during numerous seismic cycles. This requires a
constant slip rate along the fault and assumes that slip rate is
significantly different from that of the characteristic earthquake
model. Recently, the revised slip distribution on the San Andreas fault, improved by a light detection and ranging (LiDAR)
survey undertaken by Zielke et al. (2010), raises questions about
the repetition of paleoearthquakes proposed by Sieh (1978b).
Variable slip is another example of an earthquake model.
In this model, a fault or a fault segment randomly repeats
variable amounts of slip during each progressive earthquake
(see also the explanation in Schwartz and Coppersmith,
1984). If this model were representative of actual faulting, it
would not be possible to predict or forecast the occurrence of
earthquakes generated from such active faults. Shortage of
quantitative data makes it difficult at this time to fully evaluate
which model is the most valid.
5.12.2.2
Time-/Slip-Predictable Model
Shimazaki and Nakata (1980) proposed the time-/slippredictable model (which was originally called the timepredictable recurrence model for large earthquakes) in which a
fault or a fault segment produces large earthquakes that are
related to the size of the last earthquake (Figure 4). Both the
time-predictable and slip-predictable models require a
stable slip rate or stressing rate. With the time-predictable
model, when the size of the last earthquake is large, that is, the
slip is also proportionally large, then the time to the next
earthquake is large. In contrast, when the size of last earthquake is relatively small, the time to the next earthquake is
short. In this case, if the size of the slip and slip rate are
known, then it is possible to estimate the time to the next
earthquake; that is, it is time-predictable. However, it is not
possible to estimate the magnitude of the next earthquake.
In contrast, with the slip-predictable model, when the
elapsed time since the last earthquake is great, then the slip
will be larger and the size of the next earthquake magnitude
will be greater. Similarly, when the elapsed time is short, then
the amount of slip and the earthquake magnitude will be
smaller. In this case, if the elapsed time is known then it is
possible to estimate the amount of slip for the next earthquake; that is, it is slip-predictable. However, it is not possible
to forecast the time of the next earthquake.
The time-/slip-predictable model has greater advantages
over the other earthquake recurrence models for the application
of paleoseismic data. In particular, when the time-predictable
model is applied to determine the probability for future large
earthquakes, it yields greater accuracy than that of the characteristic earthquake model (e.g., Headquarters for Earthquake
Research Promotion, 2005). The time-predictable model was
developed before the characteristic earthquake model, however,
and the latter model has been widely applied throughout the
world. This is because the time-/slip-predictable model is based
on a paleoseismological dataset of uplifted Holocene marine
terraces. Therefore, the application of this model is likely limited to interplate megathrust earthquakes along subduction
zones. In contrast, the characteristic earthquake model is
more applicable to both intraplate and interplate faults. Many
seismologists and geodesists have tested the timepredictable model, including Nishenko (1985), Sieh et al.
(1989), Hori and Oike (1999), Murray and Segall (2002) and
Parsons (2006). Testing paleoseismological predictions of
Paleoseismology
273
time-/slip-predictable models at a site is very important. But

there are just a few sites on active faults where this has been
done because of the difficulties associated with acquiring long
records from field data (Liu-Zeng et al., 2006; Weldon et al.,
2004; Kondo et al., 2010).
5.12.3
Recent Methodological Developments in

Paleoseismology
During the past few decades there have been numerous

methodological advances that have helped accelerate the development of paleoseismology. These have included novel remote sensing technologies, new and improved dating methods,
and advanced fault trench logging methods. In addition, a
plethora of studies have been done on the landforms associated
with active faulting and folding that are produced by earthquakes (Figure 5). Chapters 5.6, 5.7, 5.8, and 5.10 provide
comprehensive discussions on the characteristics of tectonic
landforms associated with faults and folds.
5.12.3.1
Remote Sensing Technologies
The past 50 years have witnessed a mini-revolution in remote

sensing (National Research Council, 2008). In particular, the
development of global positioning systems (GPS) has allowed
rates of crustal displacement to be determined almost in real
time, and inexpensive handheld and differential GPS are allowing tectonic landforms to be surveyed with great ease (National
Research Council, 2010). An example of the use of differential
GPS to measure offset landforms is shown in Figure 6.
Over the past decade, airborne and land-based LiDAR have
also advanced our ability to survey tectonic landforms in
great detail (e.g., Frankel et al., 2007a,b; Kondo et al., 2008a;
Arrowsmith and Zielke, 2009; Blisniuk et al., 2010). LiDAR is
helping to accurately locate and trace active faults using tectonic
landforms and to quantify rates of vertical and horizontal
crustal displacement. Figures 7 and 8 provide examples of how
airborne LiDAR can be used to map active faults to help determine past fault displacements. In addition, high-resolution
digital elevation models (DEMs), produced from satellite and/
or geodetic data, are providing opportunities to search for active faults and surface ruptures in areas where it may be logistically too difficult to undertake extensive fieldwork (Avouac
et al., 2006; Fu et al., 2007; Walker, 2006). The development of
Google EarthTM during the past decade, with its easy and free
accessibility, has also helped to enhance tectonic geomorphic
mapping to identify and map active faults and associated
structures and landforms for paleoseismic studies.
5.12.3.2
Fault Trenching
In the 1970s, geologists began to examine paleoseismic evidence in more detail by excavating trenches across active faults.
Trenching has allowed paleoseimologists to examine faulted
and deformed sediment in great detail to help reconstruct
multiple earthquake events. Particularly noteworthy is the pioneering research of Sieh (1978a, b) on the San Andreas fault.
Following Siehs work, fault trenching methods have been
Figure 5 Images illustrating some of the tectonic geomorphic

evidence used to identify active faults for paleoseismic studies.
(upper panel) Google EarthTM image of a fault scarp of an active
normal fault in the Pamirs of western China. (middle panel) Oblique
aerial photograph of fault scarp of an active reverse fault that
traverses an alluvial fan in the Precordillera of Argentina. (lower
panel) Offset channel and small scarp along the Death Valley strikeslip fault in California. This offset was probably produced during a
single earthquake event.
274
Paleoseismology
2.5 1.0 m (RL)

2.0 0.2 m (V)
Levee 2
Levee 3
20.4 0.8
4.1 m (RL)
5.9 0.7 m (V)
420300
4069050 + Levee 2
Levee 3
Levee 4
0
Levee 4
78.8 2.8
33.8 0.8
40
5.3 2.4 m (RL)
4.4 0.4 m (V)
meters
20
40
meters
N
Levee 1
4068840 +
420300
4068840 + 4068840
420450
Levee 1
20.8 1.4
20.0 1.2
3.2 1.0 m (RL)

1.7 0.3 m (V)
11.9 0.6
Figure 6 Simplified differential GPS maps showing fault scarp trace and offset levee apices developed within late Quaternary surfaces and TCN
sample locations for sites along the Hunter Mountain fault in the Eastern California Shear Zone, California, USA. Reproduced from Lee, J.,
Stockli, D.F., Owen, L.A., Finkel, R.C., Kislitsyn, R., 2009. Exhumation of the Inyo Mountains, California: implications for the timing of extension
along the western boundary of the Basin Range Province and distribution of dextral fault slip rates across the Eastern California Shear Zone.
Tectonics 28, TC1001, 10.1029/2008TC002295. Magnitude of right lateral (RL) strike-slip and vertical (V) offset are shown. The differential GPS
survey points are indicated by gray circles; fault traces by thick black lines; levee apices by thin black lines; projection of levee apices by gray
triangles; arrows indicate relative motion across faults; black circles indicate TCN sample locations; and TCN ages are indicated by italic fonts.
widely applied across many regions of the world (e.g., Sieh,

1978a; Sieh et al., 1989; Schwartz and Coppersmith, 1984;
Wesnousky et al., 1999; Weldon et al., 2002; Lindvall et al.,
1989; Grant and Lettis, 2002; Pantosti et al., 1993; Okada et al.,
1992; Tsutsumi et al., 1991; Berryman et al., 1989; Meghraoui
et al., 1988, 2000; Kumar et al., 2001, 2006, 2010; Sarmiento
et al., 2011; Jayangondaperumal et al., 2011). McCalpin (1996,
2009b) described the methods of fault trenching in detail, including how to choose the fault trench site, excavate safely, and
log the trench. During the past several decades there has been a
growing recognition of the large variety of deformation structures that can be recognized in sediments within fault trenches
and natural excavations, including a variety of faults, fissures,
colluvial wedges, and liquefaction structures. Figure 9 illustrates some of these structures produced by past earthquake
events. Fault trenching is now a routine aspect of paleoseismic
work and site investigation for engineering and construction in
earthquake-prone countries (Figure 10). Figure 11 shows the
fresh vertical exposure of structure and stratigraphy afforded by
a typical fault trench. Commonly, multiple fault trenches are
excavated parallel or at oblique angles to each other to enable
paleoseismologists to make three-dimensional reconstructions
of deformed and displaced sediments (e.g., Kondo et al., 2003,
2010; Figures 12 and 13).
Over the past decade, development of the digital
camera has allowed fault trenches to be photographed and
photo-mosaics to be constructed almost immediately
(e.g., Rockwell and Ben-Zion, 2007; Kondo et al., 2008b;
Philibosian et al., 2009; Figure 14). New methods to collect

samples of fault trench sediments have also been developed.
This includes using the geoslicer, invented by Nakata and
Shimazaki (1997), which involves retrieving meter-size intact
slices of subsurface sediment by vertically driving a pair of
steel plates into the ground and removing the undisturbed
material. The geoslicer method allows paleoseismologists to
examine unconsolidated sediments as oriented samples and
has been applied to active faults (Kondo et al., 2003), liquefaction features in tidal marsh (Takada and Atwater, 2004),
and tsunami deposits in coastal marshes and lakes (Sawai
et al., 2004, 2008). Soil peels are another method to help
inspect paleoseismic structures. To produce a soil peel, resin is
sprayed onto the fault trench wall and is allowed to penetrate
some distance into the wall before drying and hardening.
Successive coats of resin are added and backed by cloth that
creates a layer strong enough to be removed for detailed
examination (McCalpin, 2009b).
5.12.3.3
Numerical Dating
Paleoseismology has benefited greatly in recent years from

advances in numerical dating (National Research Council,
2010). Noller et al. (2000) provided a useful overview of the
different types of Quaternary geochronological methods that
can be applied to paleoseismology. Since the Noller et al.
publication, much progress has been made in developing
Paleoseismology
Figure 7 Airborne laser swath mapping images of a displaced

alluvial fan along a stretch of the Death Valley fault at Red Wall
Canyon, California. Reproduced from Frankel, K.L., Brantley, K.S.,
Dolan, J.F., et al., 2007a. Cosmogenic 10Be and 36Cl geochronology
of offset alluvial fans along the northern Death Valley fault zone:
Implications for transient strain in the eastern California shear zone.
Journal of Geophysical Research Solid Earth 112, B06407,
10.1029/2006JB004350. (a) Present day and (b) alluvial fan restored
29779 m to its original, prefaulted configuration. The oldest
alluvial fan surface is shown in red. Sites where 10Be surface
samples were collected are shown by black squares and the locations
of 36Cl depth profile are shown by black triangles. Colored circles
indicate channels that correlate to each other when offset is
restored.
geochronological methods, and in particular optically stimulated luminescence (OSL) and terrestrial cosmogenic nuclide
(TCN) surface exposure dating are now widely used.
Application of numerical dating to paleoseismic studies
can be broadly divided into dating surface features (landforms) and subsurface dating. Subsurface dating is mainly
applied in fault trench work, and the most common method is
radiocarbon dating. Other methods include U-series, OSL,
and tephrachronology. Much progress has been made with
radiocarbon dating, particularly since the development and
widespread use of accelerator mass spectrometry (AMS)
radiocarbon dating, which allows very small (o1 g) amounts
of organic material to be dated compared with conventional
radiocarbon methods that generally require large (many
grams) amalgamations of organic material to obtain an age.
The use of AMS methods allows the age variation of organic
material within a single stratigraphic unit to be examined; for
example, to check whether the organic material within a unit
really reflects the true age of the sediment or whether it has
275
been derived from older deposits or had been exposed on

the landscape for a considerable time before being deposited
in the sediment. Improved calibration of radiocarbon ages has
also helped enhance paleoseismic studies by allowing radiocarbon ages to be compared directly with historical data and/
or other dating techniques. Programs such as OxCal (Ramsey,
2001) have made it very easy for paleoseismologists to calibrate their ages, estimate probability, and to apply statistical
methods such as Bayesian analysis (Biasi et al., 2002; Weldon
et al., 2004).
Use of luminescence dating (thermoluminescence (TL)
and OSL) in paleoseismology has a long history, but it has
only really been during the past decade that luminescence
dating has become widespread. In paleoseismic studies, TL
and OSL are used to date sediment, although they can also be
used to date pottery and burnt flint (Aitken, 1985, 1998). In
essence, luminescence dating determines the time elapsed
since a sediment sample was exposed to daylight (OSL dating) or heat (TL dating). The method relies on the interaction
of ionizing radiation with electrons in semi-conducting
minerals, resulting in metastable accumulation of charge.
Illumination or heating releases the charge as a measurable
emission of photons, that is, luminescence. Ages can be determined from o100 years to 4100 ka, with errors as low as
5%. Application of luminescence methods has accelerated
because of the development of the single aliquot regenerative
OSL method (Murray and Wintle, 2000). This method helps
assess whether the deposited sediment was adequately
bleached prior to its deposition, that is, the luminescence
signal was reset so that OSL dates are not overestimates of the
true age of deposition. Samples are collected by inserting
opaque tubes into freshly excavated exposures or by cutting a
block of sediment out of a section to ensure that the sampled
sediment is not exposed to light before processing in the
laboratory (Figure 15).
OSL has been proven successful in many areas throughout
the world (e.g. Crone et al., 1997; Lee et al., 2001; Kent et al.,
2005; Owen et al., 1999; Rockwell et al., 2000, 2010; Washburn et al., 2001; Wesnousky et al., 2005; Tsukamoto et al.,
2009; Ganev et al., 2010). Figure 16 shows an example of the
application of the OSL method to determine the timing of
past earthquakes on the Lone Pine fault along the Eastern
California Shear Zone in the Owens Valley. OSL methods have
been particularly successful in showing that earthquakes
cluster in time, as demonstrated by Rockwell et al. (2000) in
the Eastern California Shear Zone as illustrated in Figure 17.
Figure 17 also shows the work of Ganev et al. (2010) who
used OSL dating to show a close correspondence of earthquake event ages on the Calico fault in the Eastern California
Shear Zone with periods of increased seismic moment release
in the region determined by Rockwell et al. (2000).
The use of TCNs in paleoseismology has mainly been
confined to dating landforms and surface features.
Notable studies along active faults include those of Bierman
et al. (1995), Siame et al. (1997), Van der Woerd et al. (1998),
Zehfuss et al. (2001), Brown et al. (2002), Meriaux et al.
(2004), Chevalier et al. (2005), Matmon et al. (2005, 2006),
Frankel et al. (2007a,b), Le et al. (2007), and Blisniuk et al.
(2010). Marine terraces have been dated using TCNs by
Perg et al. (2001) and Saillard et al. (2009). TCN dating
276
Paleoseismology
(a)
(b)
(c)
(d)
594
Fault scarp
Alluvial fan surface
(m)
Artificial
modification?
Vetical offset:
1.8 m
y = 4.3E3*x+5.850E+2
584
0
594
f(x) = 4.3E3*x + 5.868E+2
500
(m)
b
1000
(m)
Alluvial fan surface
f(x) = 5.4E3*x + 5.862E+2

Vetical offset:
y = 5.4E3*x + 5.849E+2 1.3 m
584
0
500
(m)
1000
Figure 8 A fault scarp detected by LiDAR survey in a densely built-up area along the ItoigawaShizuoka Tectonic Line (ISTL) active fault
system, one of the hazardous fault systems in Japan. (a) Aerial photograph showing that man-made structures obscure the original topographic
surface for aerial photograph interpretation. (b) High-resolution DEMs produced by the LiDAR data show a hidden scarp (red arrows) in this
urban area. Note that the northern portion of the fault trace is in accordance with the moat of a historical castle, which was built in the
seventeenth century; therefore, the scarp should have been formed before this age. (c) Measurement of vertical displacement based on
topographic profile using the LiDAR data. (d) View of the fault scarp in its urban setting. Modified from Kondo, H., Toda, S., Okumura, K., Chiba,
T., 2008a. A fault scarp in an urban area identified by LIDAR survey: a case study on the Itoigawa-Shizuoka tectonic line, central Japan.
involves measuring the accumulation (concentration) of

TCNs that are produced by the interaction of cosmic ray
particles with minerals at Earths surface to determine ages on
stable geomorphic surfaces (Bierman et al., 1995; Gosse and
Phillips, 2001). The concentration of TCNs increases with time
in an exposed sample as a function of a depth-dependent
nuclide production rate, and for radiogenic TCNs, also a
function of their radioactive decay. Exposure ages for depositional or erosional surfaces range from a few decades to
several 100 ka for radiogenic TCNs (although in exceptional
environments TCN ages of several million years can be
obtained) and many million years for stable TCNs. The most
commonly used TCNs include 10Be, 26Al, 36Cl, and 21Ne.
Generally, large exposed boulders and cobbles or amalgamated surface clasts (i.e., cobbles or gravel) are sampled from
surfaces of interest, although bedrock fault scarp surfaces have
also been sampled. The sampled boulders/cobbles must be
demonstrably stable and lack weathering to produce reliable
results (Matmon et al., 2006; Meriaux et al., 2004; Figure 18).
Collecting sediment samples from pits has also been used
and is referred to as TCN dating using depth profiles. A depth
profile is an excellent way to avoid problems of inheritance of
Paleoseismology
277
Draping and overlap

of fault scarp
Unfaulted strara
Overlap unconformity
Fault plane
(a)
Older
offsets
Younger
offsets
Eroded
fault scarp
Older overlap
unconformity
Fault plane
(b)
Colluvial wedge or
landslide debris across
fault scarp
Fissure filled
with colluvium
Eroded
fault scarp
Fault plane
Fault plane
(c)
(d)
Folding due to shaking

and liquefaction
Draping strata
Sand blow
Injection
dike
Offset strata
(e)
(f)
Figure 9 Examples of stratigraphic and structural relationships that relate to past fault-related deformational events. Reproduced from Burbank,
D.W., Anderson, R.S., 2001. Tectonic Geomorphology. Blackwell Science, Oxford, 274 pp., and modified from Allen, C.R., 1996. Seismological
and paleoseismological techniques of research in active tectonics. In: Wallace, R.E. (Ed.), Active Tectonics: Studies in Geophysics. National
Academy Press, Washington, DC, pp. 148154.
Figure 10 Giant fault trench near the Mecca Hills in Southern

California. This was one of many deep trenches several hundred
meters long and tens of meters deep that were excavated for
paleoseismic investigations for a proposed housing development.
Figure 11 Fault trench through an active thrust fault in the

Precordillera of Argentina. The student paleoseismologist is
examining a trench log. Note the grid of strings that is pinned to the
trench wall that was used to help construct the log.
278
Paleoseismology
TCN from prior exposure since inheritance is commonly associated with sampling surface clasts (Anderson et al., 1996;
Hancock et al., 1999; Owen et al., 2011).
UTh disequilibrium dating has been particularly useful in
dating fossil coral to determine the age of marine terraces (e.g.,
Muhs and Szabo, 1982; Muhs et al., 1994; Keller and Gurrola,
2000) and it provides one of the least ambiguous geochronology tools for dating tectonic landforms, providing ages that
can range from approximately 5 to 350 ka. Recently, 230Th/U
dating of pedogenic carbonates has been successfully applied
to help examine slip rates along the southern end of the San
Andreas fault system (Behr et al., 2010; Fletcher et al., 2010).
5.12.4
Paleoseismology is subdivided into two main fields: on-fault

and off-fault paleoseismology. As the name suggests, on-fault
paleoseismology is concerned with examining landforms, sediments, and structures associated with a fault on land and/or in
shallow offshore settings, and this section describes the faulting
parameters needed for seismic hazard analysis, which generally
refers to earthquake 4MB7.0. Off-fault paleoseismology is
described in Section 5.12.5 and involves the analysis of feature
distant from an active fault. On-fault paleoseismological research is generally the only way to obtain data directly on fault
slip and specific past earthquakes. Surface ruptures associated
with large earthquakes provide the principle information on
surface faulting and seismic sources (e.g., Wesnousky, 2008 and
the references). On occasions, moderate-size shallow earthquakes, between MB6.5 and B7.0, have been produced by
blind thrusts (e.g., Stein and Yeats, 1989; Stein and King, 1984;
and, for example, the 1994 Northridge earthquake, M6.7) or
short (o20 km) faults (e.g., the 2004 mid-Niigata prefecture
earthquake, M6.6, the 2007 NotoHanto earthquake, M6.4, and
the 2007 Niigataken Chuestu-oki earthquake, M6.8; Okamura
et al., 2007; Awata et al., 2008; Shishikura et al., 2009). These
moderate-size quakes can produce notable surface deformation
with minor surface rupture.
5.12.4.1
Figure 12 Before-excavation view of a subtle fault scarplet and sag

pond at the Cevresu site along the North Anatolian fault system,
Turkey. To the left is mountain-side and to the right is basin-side;
therefore, the small subtle scarp is un-normally facing to the
upstream side near the distilled part of an alluvial fan. The site is
located on the surface rupture associated with the 1942 ErbaaNiksar
earthquake. Thick, fine sediment and buried channels can be
expected in a three-dimensional excavation. See also Figure 13.
On-Fault Paleoseismology
Fault Length and Dimension
Mapping active faults is essential to locate the potential

source(s) of large earthquakes and to estimate magnitudes.
Since the magnitude of an earthquake is linearly proportional
to the size (i.e., the dimension) of the fault plane (e.g.,
Matsuda, 1975; Wells and Coppersmith, 1994), the length and
dimension of faults are the primary parameters that are
required to estimate the potential earthquake magnitude.
Although an active, mapped surface fault is rather superficial
compared to the thickness of the entire seismogenic zone
Figure 13 After-excavation view of the Cevresu site along the North Anatolia fault system, Turkey. The excavated trenches are two faultcrossing and two fault-parallel trenches at this moment. The sediments exposed on the walls clearly indicate faults cutting whitish and brownish
sandy layers just beneath the small fault scarp, and the sag pond is filled with a gray silt layer at the down-thrown side.
Paleoseismology
279
Figure 14 An example of a trenching survey on a reverse fault, ruptured in the 2005 Kashmir earthquake in Pakistan. (a) The trench site is located
on the lateral ramp and almost in the middle of the surface rupture at Muzaffarabad. Cumulative vertical displacement and deformation on fluvial
terraces are identical. (b) Photo-mosaics of trench walls. The trench was excavated at the foot of fault scarp cutting a fluvial terrace where the
Kashmir earthquake produced B1.7 m vertical displacement. Two faulting events, including the 2005 event, were identified in the 6-m-deep trench.
Modified from Kondo, H., Nakata, T., Akhtar, S., et al., 2005. Long recurrence interval of faulting beyond the Kashmir earthquake around the
northwestern margin of Indo-Asia collision zone, Geology 36, 731734.
280
Paleoseismology
Group for Active Faults, 1980, 1991). Tectonic landforms

that are important to map include fault scarps, flexure scarps,
fault bulges, pressure ridges, back-tilted or warping terraces,
grabens, sag ponds, and mole tracks. In addition, mapping
offset Quaternary strata can also help precisely locate an active
fault. The definition of active fault may vary, but a good
definition is one that has moved during the Holocene (the
last B11.6 ka). The U.S. Geological Survey and New Mexico
Bureau of Mines and Mineral Resources (2006) refer to faulted
or folded units and landforms that affect Quaternary units
or surfaces simply as Quaternary faults, implying that they
have likely been repeatedly active throughout the Quaternary;
such faults are considered as neotectonic features. The initiation of active faulting is commonly discussed using geologic
data and consideration of the total cumulative displacement
along the fault (e.g., Hubert-Ferrari et al., 2002; Blisniuk et al.,
2010).
The length of a mapped fault is used to estimate the
magnitude (M) of earthquakes using empirical relationships, based on surface rupture length (SRL) datasets of
historical earthquakes. As such, Wells and Coppersmith
(1994) proposed that
(1520 km), the surface fault is the gateway into the direct
study of an earthquake fault.
Mapping active faults is generally conducted by mapping
tectonic landforms and Quaternary units (e.g., Research
Figure 15 Colluvial wedge in front of a normal fault in Death Valley,

Southern California. The gray tube just to the right of the geologic
hammer has been inserted into the colluvial wedge to collect
sediment for OSL dating. The tube is simply removed and the ends
sealed to ensure the sample is not exposed to light until it is
analyzed in the laboratory.
M 5:08 1:16log SRL
N 80E
3
S 80W
50 cm
50 cm
Fault strand f2
2.5 0.3 ka
3.3 0.3 ka
Fault
strand F1
3.0 14.0 ka
3.8 0.3 ka
4.4 0.4
ka
Fault strand F1
Trench floor
TWater table
Legend:
A: Soil horizon
B: Soil horizon
Clay
Fine sands with

pebbles
Medium sands
with pebbles
Course sands with
occasional pebbles
Trench roof
Basalt cinder
intermixed with coarse
to medium sand
Pebbles with sand
Fault zone with
developed shear planes
Figure 16 Fault trench log across the Owens Valley fault, California. Two earthquake events demonstrated by two separate fault traces can be
seen in the fault trench. The older event (seen toward the southwestern end of the trench) is older than B3.3 ka determined by OSL dating.
Reproduced from Lee, J., Stockli, D.F., Owen, L.A., Finkel, R.C., Kislitsyn, R., 2009. Exhumation of the Inyo Mountains, California: implications for
the timing of extension along the western boundary of the Basin Range Province and distribution of dextral fault slip rates across the Eastern
California Shear Zone. Tectonics 28, TC1001, 10.1029/2008TC002295.
Probability of moment release

(1025 dyne-cm yr1)
Paleoseismology
281
10
Eastern California shear zone
Calico fault earthquakes
8
6
4
2
0
0
???
???
6
Age (ka)
10
12
Figure 17 Compilation of paleoearthquake ages (shown by black bars) and of seismic moment release data (pale gray) in the Eastern California
Shear Zone through time. Ages are based primarily on OSL dating for the Calico fault; each number represents an earthquake event. Reproduced
from Ganev, P.N., Dolan, J.F., Blisniuk, K., Oskin, M., Owen, L.A., 2010. Paleoseismologic evidence for multiple Holocene earthquakes on the
Calico fault: implications for earthquake clustering in the Eastern California Shear Zone. Lithosphere 4, 287298.
seismogenic zone, which extends down to between B15 and

B20 km (Sibson, 1986; Scholz, 2002). The development of
seismic reflection surveys across faults now makes it possible
to estimate the down-dip width of a seismogenic fault, which
is important since the area of the fault is directly related to the
magnitude of an earthquake. The use of seismic reflection
surveys is mostly limited to reverse/thrust faults and normal
faults, where the fault plane is inclined and cuts sub-horizontal, thick strata. In some cases, the size and geometry of
fault plane at depth are uncertain and are greatly debated,
despite numerous methods of estimating, such as seismic reflection and gravity surveys, and seismological methods.
5.12.4.2
Figure 18 Sampling boulders for 10Be TCN dating along the San
Jacinto fault in southern California. This large inset boulder is ideal
for TCN dating.
Matsuda (1975) used surface-rupturing earthquakes generated by intra-plate active faults in Japan to propose the
following relationship:
log SRL 0:6M22:9
4
In a similar vein, Anderson et al. (1996) proposed using an

additional slip rate parameter (SR) to reduce the error in the
magnitude estimation proposed by Wells and Coppersmith
(1994) such that
M 5:1270:12 1:16 0:07log SRL
0:2070:04 log SR
5
Subsequent empirical relations have been proposed specifically for strong ground motion prediction, including the work
of Somerville et al. (1999) and Irikura and Miyake (2011).
Down-dip width is another important parameter in estimating the size of a seismogenic fault. Where the fault plane is
almost vertical, the width of the fault is nearly the same as the
Timing of Paleoearthquakes and Slip per Event
Reconstruction of the timing of past earthquakes is essential to

determine recurrence intervals or to determine if there has
been clustering of earthquakes in time. Clearly this is important for calculating the probability of future earthquakes
on individual faults. Recently developed dating methods are
described above in Section 5.12.3.3, and improving the accuracy and precision of the ages and timing of past earthquakes are very important to help reduce uncertainty in
hazard assessment (e.g., Scharer et al., 2007).
Fault trenching is the main technique used to determine
the timing of past earthquakes. Fault trenches are excavated
across a fault trace to examine the stratigraphy and structures
on the exposed walls (Figures 10, 11, and 14). In strike-slip
systems, evidence of paleoearthquakes is relatively simple,
as partially summarized in Figure 9. However, identifying
and logging structures can be difficult in some lithologies.
This can result in a degree of subjectivity in the interpretation
of stratigraphy and structures in the fault trench. To avoid
ambiguity and confusion, quantitative evaluation of uncertainties for individual paleoearthquakes is attempted, as
discussed in Scharer et al. (2007). In recent years, interest has
grown in reconstructing the slip and displacement per event
along the strike-slip faults. This may be undertaken by
reconstructing slip events on landforms such as river terraces
(e.g., Figure 19). Moreover, three-dimensional trenching
Paleoseismology
310470 cal yr B.P.

r3
t3 r
2
t2
ne
Ch
i
er
t3
t2
4
2
j
j
t4
rm
t2
t2
t1
tt ri
v
t5
t2 r1
Hu
Offset
Ch
an
an
m
m
t2
li
t4
25
0y
rB
.P.
r4
lj
t5
Wellington
fault
ne
282
Scale (meters)
0
0
10
20
30
40 m
Overbank silt and sand
Figure 19 Displaced fluvial terraces along the Wellington fault, New Zealand. Reproduced from Burbank and Anderson, 2002, and modified from
Van Dissen, R.J., Berryman, K.R., Pettinga, J.R., Hill, N.L., 1992. Paleoseismicity of the Wellington-Hutt Valley segment of the Wellington Fault, North
Island, New Zealand. New Zealand Journal of Geology and Geophysics 35, 165176. The lowest terrace, dated at o250 years, is not displaced by
this strike-slip fault, whereas the next highest terraces (t2) and the channels that cut across its surface show a displacement of B4 m attributable to
the last major earthquake. Note that any fault displacement of the riser (r1) between t2 and t1 was beveled off during the creation of t1. The amount
of offset of the risers increases systematically with each older terrace. Important controls on terrace width are exerted by the geometry of the river
channel prior to abandonment (note the triangular shape of t3). The vertical throw across the fault is smaller (B10%) than the horizontal
displacement.
Gully
10.50.5m
1252
125
52
Buried channel
19.90.5m
Higher fan surface

1251
DTE
DTC2
?
DTB
DTC
Depression
Tree line
4.05.0 m
1249
Lower fan surface

Gully
1251
DTA4
10.80.2m
DTA3
DTA2
DTA1
Fault
Lower fan surface

1250
DTD1
DTD2
DTD3
DTD4
Buried channel
15.30.1m
Artificial field
1249
124
49
0
10 m
Figure 20 An example of slip-per-event reconstruction in combination with human-made, geomorphic, and buried channel offset features on the
1944 BoluGerede earthquake rupture along the North Anatolian fault system, Turkey. The offset tree line is associated with the 1944 earthquake
and the B10-m-offset of gullies is cumulative for the last two events. In addition with these offset features at the surface, three-dimensional
trenching (rectangles in solid black lines) was performed. In the three-dimensional trench, two additional discrete buried channels were present
and a similar amount of the 1944 slip for the last four events was resolved. Modified from Kondo, H., Ozaksoy, V., Yldirim, C., 2010. Slip history
of the 1944 Bolu-Gerede earthquake rupture along the North Anatolian fault system: Implications for recurrence behavior of multisegment
earthquakes. Journal of Geophysical Research 115, B04316. doi:10.1029/2009JB006413.
is increasingly being used to reconstruct timing and amount

of slip. The trenching methods rely heavily on whether
offset features such as buried channel deposits or structural
relief are preserved at the site (Kondo et al., 2003; Weldon
et al., 2004; Liu-Zeng et al., 2006; Kondo et al., 2010; Figures
12, 13, and 20).
The timing of past earthquake events on reverse and/or
thrust faults is a little more difficult to reconstruct. Figure 21
illustrates one of the relations between reverse faulting/folding
and sedimentation, and the resulting paleoseismic structures.
Depending on the climatic conditions at the site and the

amount of vertical co-seismic fault displacement, it is rare that
co-seismic fault/fold scarp sediments overlap all inter-seismic
scarp sediments and, generally, the hangingwall becomes an
erosional area after the co-seismic uplift (Figure 22). Moreover, when the amount of vertical displacement is a few
meters or larger, the cumulative vertical slip for just a few
earthquakes may be close to or even exceed 10 m (Figure 14).
This means that it is technically difficult and dangerous to
excavate a fault trench to examine multiple earthquake
Paleoseismology
events. However, use of densely spaced boreholes and highresolution seismic reflection data can help with geologic
interpretation in these cases (Dolan et al., 2003; Ishiyama
et al., 2004).
In extensional settings, recognition of a colluvial wedge is
key to reconstructing the timing of faulting. McCalpin (2009c)
described a typical succession of events as colluvium is deposited from a fault scarp (Figure 23). Initially, the faulting
event creates a free face (which is entirely in the unit marked by
short dashes in Figure 23(a)). The first colluvial wedge is then
deposited, composed exclusively of material derived from the
free face (Figure 23(b)). The second faulting event creates a
basal tension fissure into which the earlier colluvial wedge drops
(Figure 23(c)). Then the second colluvial wedge fills the tension
fissures and progrades out onto the downthrown block
(Figure 23(d)). The lower part of the second colluvial wedge is
derived mainly from gravity and debris deposition from the
lower part of the free face and is composed of material derived
from the newly exposed unit. As the scarp continues to backwaste and decline, more colluvial material is derived from
the upper part of the scarp. The third faulting event creates a free
face composed of a lower, now exposed unit and a new antithetic fault farther outboard of the earlier antithetic fault, which
was not rejuvenated in this third event (Figure 23(e)). Most
of the free face now exposes uplifted units in the footwall.
Next, the third colluvial wedge buries the earlier two wedges
(Figure 23(f)). The succession of lithologies in the third wedge
roughly parallels the stratigraphic succession exposed in the
scarp face, that is, the basal portion is derived from the uppermost unit, and the upper (wash facies) portion is mainly derived
from the lower, now exposed unit. Although these successions
of events are simplistic, they eloquently encapsulate the basic
nature of the sedimentary tectonics at an active normal fault.
In a recent work, the timing of paleoearthquakes and the
amount of vertical slip per event on normal faults have been
successfully reconstructed using 36Cl TCN dating by progressively, sampling up the exposed fault plane (Schlagenhauf
et al., 2011).
Fold crest
Fo
Growth strata
reli
mb
Pre-growth strata
Detachment fold
(a)
Overlap
Offlap
(b) Uplift < accumulation
Backlimb
overlap, then offlap
Uplift < accumulation

then
(d) Uplift > accumulation
283
(c) Uplift > accumulation
Offlap, onlap, then overlap
Uplift > accumulation

then
(e) Uplift < accumulation
Figure 21 Conceptual diagrams of the geometric relations between

a growing fold and deposition on the flanks of the fold that can be
applied to co-seismic folding and faulting. (a) Nomenclature. (b)
Overlap when uplift rates are less than deposition rates. (c) Offlap
when uplift rates are greater than deposition rates. (d) Overlap then
offlap created by initially low uplift rates, which then become greater
than deposition rates. (e) Offlap, onlap, then overlap created by
initially high uplift rates which then decrease to less than the
deposition. Modified from Burbank, D.W., Verges, J., 1994.
Reconstruction of topography and related depositional systems
during active thrusting. Journal of Geophysical Research 99,
2028120297.
(a)
(b)
(c)
(d)
5m
(g)
(e)
(f)
Figure 22 Various types of reverse fault scarps observed along the Spitak fault ruptured in the 1988 Ms 6.9 Spitak, Armenia, earthquake. (a) Simple
reverse fault scarp. (b) Collapsed scarp after the earthquake. (c) Simple bulge. (d) Reverse fault scarp with strike slip component. (e) Reverse fault
scarp with shortening and back-thrust. (f) Reverse fault scarp with shortening and subhorizontal detachment near the surface. (g) En-echelon pressure
ridge due to secondary strike slip fault. (1) Bedrock; (2) unconsolidated Quaternary sediments; (3) turf. Reproduced from Philip, H., Rogozhin, E.,
Cisternas, A., Bousquet, J.C., Borisov, B., Karakhanian, A., 1992. The Armenian earthquake of 1988 December 7: faulting and folding, neotectonics and
paleoseismicity. Geophysical Journal International 110, 141158.
284
Paleoseismology
(a)
(b)
(c)
(d)
(e)
(f)
Figure 23 Schematic figures showing how the provenance of scarp-derived colluvium may vary through successive fault displacements.
Reproduced from McCalpin, J.P., 2009b. Field techniques in paleoseismology terrestrial environments. In: McCalpin, J.P. (Ed.),
Paleoseismology. Second ed. Academic Press, Burlington, MA, pp. 29118. See text for explanation.
Paleoseismology
Figure 24 View of a marine terrace on the Nicoya Penninsula in Costa

Rica. The terrace has been eroded into basalt and the paleoseismologist
is surveying its altitude and collecting samples for TCN dating.
5.12.4.3
Long-term Slip Rate
Tectonic geomorphic methods are advantageous for reconstructing long-term slip rates, but the age of a landform is
generally difficult to determine accurately and precisely. In
addition, the amount of displacement of a landform may have
large errors. McCalpin (2009a) described many of the errors
associated with determining long-term slip rates.
Using Wallaces (1970) formula (Section 5.12.2), slip
rate and slip-per-event yield the average recurrence interval. Slip
rate and recurrence interval provide an estimate of the slip-perevent at the site. Since slip rate is essential for understanding the
dynamics of a fault and earthquake hazard assessment, numerous researchers have determined slip rates in many tectonic
settings throughout the world (Ikeda and Yonekura, 1986;
Knuepfer, 1987; Niemi and Hall, 1992; Van der Woerd et al.,
1998; Ritz et al., 2003; Mason et al., 2006; Ferry et al., 2007; Le
et al., 2007). Many of these studies have benefited from the
recent development of high-resolution space and satellite
measurements to quantify cumulative displacement of landforms along faults (see Section 5.12.3.1) and from the newly
developed numerical dating methods (see Section 5.12.3.3).
5.12.5
5.12.5.1
Off-Fault Paleoseismology
Marine Terraces
Marine abrasion platforms, which when uplifted are referred

to as marine terraces, are erosional landforms that develop
due to landward retreat of the active sea cliff in response to
erosion by ocean wave energy (Anderson et al., 1999). A
285
marine terrace will develop whenever the rate of sea cliff retreat is significant compared with the rate of relative sea-level
change so that a planar surface can be beveled onto the
continent margin (Figure 24). The seminal work on marine
terraces was undertaken in the Huon Peninsula, New Guinea,
where 230Th/234U dating of tectonically uplifted corals was
used to reconstruct former sea-level changes in relation to the
marine oxygen isotope stratigraphy (Bloom et al., 1974;
Ota, 1975; Ota et al., 1993; Chappell et al., 1996). These
studies laid the foundation for determining uplift from sealevel histories derived from marine terraces.
Particularly noteworthy research was undertaken along the
Californian coast where it was recognized that marine terraces
do not simply correlate to sea-level highstands and therefore
reflect significant tectonic uplift (Upson, 1951; Muhs et al.,
1982, 1992; Sarna-Wojcicki et al., 1987; Rockwell et al., 1992;
Perg et al., 2001; Bookhagen et al., 2006; Muhs et al., 2012;
Gurrola et al., in press). One of the classic marine terrace
studies was conducted by Rockwell et al. (1992) near Point
Conception in Southern California and showed that five discrete marine terrace levels were related to uplift. Important
studies on marine terraces have been undertaken elsewhere,
notably in Chile (Saillard et al., 2009) and Costa Rica
(Gardner et al., 2001). These marine terraces reflect tectonic
uplift and are likely co-seismic, but individual seismic events
cannot be easily discerned. A particularly noteworthy study,
however, was undertaken by Bookhagen et al. (2006) who
related uplifted Holocene beach berms to paleoseimic events
on Santa Maria Island off the Chilean coast.
Estuarine and lagoonal settings, however, can provide highresolution data on paleoseismic events. Figure 25 shows three
of the principal mechanisms that can produce paleoseismic
evidence for earthquakes in coastal environments. Figure 25(a)
shows soils of vegetated upper intertidal tidal wetlands that are
co-seismically subsided into a lower intertidal zone and covered
by marine or estuarine sediments. Evidence of co-seismic suddenness of the event includes fragile aboveground stems and
lead bases of herbaceous plants entombed in growth position
in the marine sediments, tree ring cross-matches that show
synchroneity of death for the buried trees, and normal tree ring
growth suddenly terminated at tree death. Figure 25(b) shows
tsunami sediments deposited directly on a subsided surface and
Figure 25(c) shows liquefaction sediments vented directly on a
subsided surface.
Many tectonically active coasts occur in subduction zone
megathrust settings, which undergo interseismic and co-seismic
deformation during an earthquake cycle (Nelson et al., 1996;
Sawai et al., 2004). Figure 26 illustrates the characteristics of
this deformation cycle. Such a cycle would produce a succession
of transgressive and regressive sediments, which can be used to
help determine the duration of the earthquake cycle.
5.12.5.2
Paleoliquefaction Features
Liquefaction is the transformation of a granular material from

a solid state to a liquefied state as a consequence of increased
pore-water pressure (Obermeier, 2009). Liquefaction due to a
density inversion can occur in sediments as a consequence of
more dense sediment being deposited on top of less-dense
286
Paleoseismology
Subsidence
Buried soil of
subsided marsh
and forest
(a)
Tsunami
Sand layer on
buried soil
Liquification
Sand intruded
into and
extruded onto
buried soil
(b)
Before
earthquake
Minutes to hours
after earthquake
Minutes to years
after earthquake
Decades to centuries
after earthquake
(c)
Figure 25 Schematic figure showing three of the principal processes that produce the most diagnostic paleoseismic evidence of subduction
earthquakes in coastal environments. Reproduced from Atwater, B.F., Nelson, A.R., Clague, J.J., et al., 1995. Summary of coastal geologic
evidence for past great earthquakes at the Cascadia subduction zone. Earthquake Spectra 11, 118. See text for discussion.
sediment. This is often the case in sedimentary environments

where sediment is deposited rapidly. Liquefaction may also
occur because of slope instability, which is common where
sediments have been deposited on steep paleoslopes such as
delta foresets within lakes. Ground shaking during an earthquake can cause liquefaction of sediment and in such an instance liquefaction features are sometimes referred to as
seismites, which can be used in paleoseismic studies.
Figure 27 schematically illustrates some of the main liquefaction structures that can develop within thin, clean sand
strata bounded by impermeable strata (typically clayey silt).
Clearly, it can be difficult to unequivocally attribute liquefaction structures such as involutions, frame structures, sand
dikes, and sills to an earthquake event because of the different
mechanisms that can produce similar structures. Because of
this problem, Sims (1975) suggested criteria that should be
applied to deformational structures in lacustrine sediment
before assigning them a seismic origin. These include the
structures being confined to a particular stratigraphic horizon,

the structures should be regionally extensive, and they should
not be associated with any slopes. Similar criteria would also
apply to nonlacustrine sediments. Applying such criteria
allows a paleoseismic record to be produced from a succession
of sediments. A notable example using this approach was
undertaken by Hempton and Dewey (1983) using lacustrine
sediments in southeast Turkey, which were likely produced by
earthquakes along the East Anatolian fault.
Some liquefaction structures produced during large or bigger earthquakes are particularly impressive and include sand
dikes and sand-blows as illustrated in Figure 28. Such features
are encountered in many places within the meizoseismal zone
of the 18111812 New Madrid earthquakes (Obermeier, 2009).
Craters can also form during earthquakes such as that illustrated in Figure 29. Liquefied sediment may also form thin sills
and extend great distances horizontally, which are commonly
associated with large sand-blows (Figure 30).
Paleoseismology
Elastic strain accumulation (interseismic)
287
Earthquake rupture (coseismic)

Coast
Coast
Uplift
Subsidence
Upper contin
ental plate
Subducted oceanic plat

e
Shortening
Subducted oceanic plat S

e tra
in
ate
tinental pl
Upper con
Extension
rel
ea
se
Seismogenically locked
Transform zone
(a)
(b)
Land-level changes
Interseismic
RSL changes
Regional sl rise
RSL changes at coast
Postseismic
200 yr
1m
1m
200 yr
Preseismic
(c)
(d)
(e)
(f)
Figure 26 Schematic diagrams showing the pattern of (a) interseismic and (b) co-seismic deformation associated with a subduction zone
megathrust during an earthquake deformation cycle. Reproduced from McCalpin, J.P., Carver, G.A., 2009. Paleoseismology of compressional
tectonic environments. In: McCalpin, J.P. (Ed.), Paleoseismology. Second ed. Academic Press, Burlington, MA, pp. 315420 and Nelson, A.R.,
Shennan, I., Long, A.J., 1996. Identifying coseismic subsidence in tidal wetland stratigraphic sequences at the Cascadia subduction zone of western
North America. Journal of Geophysical Research 101(B3), 61156135. (c) Land-level changes at the coast during two earthquake deformation
cycles of different amplitude. (d) Relative sea-level (RSL) changes produced by the two cycles during a period of no change in regional (eustatic)
sea level. (e) A gradual rise in RSL during the cycles that does not include short-term or small-scale changes in local and regional sea level. (f)
RSL changes at a coast resulting from the sum of parts (d) and (e). In parts (a) and (b), the solid- and thin-dashed lines mark the megathrust; the
thick-dashed line is a transition zone between the locked and plastically deforming parts of the plate boundary. Shading shows the relative amount
of uplift and subsidence of the upper plate during the (a) interseismic and (b) co-seismic parts of the deformation cycle. The shaded band in part
(f) shows /0.5 m of uncertainty in estimates of former RSLs determined from geomorphic or stratigraphic data.
Clayey mud
Bulbous
intrusion
Dike
Fold
Clean fine sand

Each stratum
few cm thick
Clayey silt, firm
Dike
Detachment
Fault
Clean sand
Clayey silt, firm
Figure 27 Schematic figure showing liquefaction structures within thin clean sand strata bounded by impermeable strata. Reproduced from
Obermeier, S.F., 2009. Using liquefaction-induced and other soft-sediment features for paleoseismic analysis. In: McCalpin, J.P. (Ed.),
Paleoseismology. Second ed. Academic Press, Burlington, MA, pp. 497564.
In addition to providing the timing of past earthquake

events, the distribution of liquefaction structures can be
used as a natural seismic network to estimate the epicenter
of a past earthquake (Obermeier, 1996). Ambraseys (1988),
for example, suggested a broad relationship exists between
earthquake magnitude and the distance from an earthquake epicenter to the furthest liquefaction effect (venting
to the surface or ground fracturing) (Figure 31). This
allows an estimate of the magnitude of a former earthquake to be determined if liquefaction structures can be
located.
5.12.5.3
Tsunami Deposits (Tsunamiites)
The study of tsunami deposits (tsunamiites) emerged with the

work of Atwater (1987) along the coast of Washington and
Oregon states in the U.S., which were attributed to onshore
tsunami deposition linked to a large Cascadia offshore earthquake at c. AD 1700. In addition, Dawson et al. (1988) showed
evidence for tsunamiites in eastern Scotland that were attributed to a large, tsunami-triggered slide generated west of
Norway B8 ka. Distinguishing tsunamiites from tempestites
(storm deposits) is challenging, but several studies have
288
Paleoseismology
10 to 40 m
Stratum of silt
with organic debris
Fining-up sequence of
sand with minor silt
Laminations
within core
2 to 6 m cap of
silt and clay
0.3 to 0.6 m
Tabular sand-filled
fissure (dike), commonly
0.15 to 0.3 m wide;
many contain clasts of
sidewall material
Basal zone of sand

containing many
clasts of sidewall
material and wood
and coal fragments
Ground craking
caused by weathering
Bifurcating dikes
in weathered zone
Tabular intrusions
that pinch
together upward
Region in which
initial bedding
has been
destroyed
Clean sand extending

to a depth as much
as 30 m
Figure 28 Schematic vertical section showing dikes cutting through overbank silt and clay strata, and overlying sand-blow deposits.
Reproduced from Obermeier, S.F., 2009. Using liquefaction-induced and other soft-sediment features for paleoseismic analysis. In: McCalpin, J.P.
(Ed.), Paleoseismology. Second ed. Academic Press, Burlington, MA, pp. 497564. (a) Stratigraphic details of sediment vented to the surface. (b)
Dikes that pinch together as they ascend. (c) Characteristics of dikes in a fractured zone of weathering, in highly plastic clays.
provided data and criteria to help (Tuttle et al., 2004; Kortekaas

and Dawson, 2007). But even more challenging is distinguishing earthquake-triggered tsunamiites from those produced by other mechanisms such as landsliding and bolide
impacts. Einsele et al. (1996) described the various pathways
tsunami sediment is transported and deposited (Figure 32) and
Dawson (2000) provided a general overview, and Dawson and
Stewart (2007) provided a comprehensive review of tsumani
deposits throughout the geologic record. Important insights
into the deposition and characteristics of tsunamiites are being
obtained from observations made after the tsunamis of 26
December 2004 and the 11 March 2004.
Arguably, the most eloquent work on earthquake-triggered
tsunamiites has been on the coastal deposits in Washington
and Oregon states (Atwater, 1987, 1992, 1996; Atwater and
Moore, 1992; Atwater and Yamaguchi, 1991; Atwater et al.,
1991, 1995, 2005). These studies utilized coastal sediments, as
described above (see Section 5.12.5.1) and in Figure 25(b),
and highlighted the occurrence of great earthquakes and tsunami along the Cascadia subduction zone. The studies were
particularly instructive because they drew attention to the
hazard from great earthquakes and catastrophic tsunami along
the western U.S. and Canadian coastlines. These studies also
proved that the Japanese Orphan tsunami of 1700 originated
as a consequence of a Cascadia subduction-related earthquake
and illustrated the far-reaching effects of earthquakes (Satake
et al., 1996; Satake and Atwater, 2007).
5.12.5.4
Earthquake-Triggered Landslides
Chapter 5.11 provides a comprehensive review of the large

variety of different types of earthquake-triggered landslides
and stressed that even small earthquakes (M 4) can trigger
landslides. The largest earthquakes can trigger tens of thousands of landslides over regions that may cover many tens or
even hundreds of thousands of square kilometers (see Chapter
5.11). Chapter 5.11 reviews the broad relationships between
landslide distribution and earthquake magnitude, which has
the potential for a paleoseismic proxy. However, using landslides as paleoseimic proxy is challenging because landslides
can, of course, be produced by mechanisms other than
earthquakes.
One of the most convincing attempts to reconstruct past
earthquake events was undertaken by Bull (2009) who
examined rockfalls in the Sierra Nevada of California. Dating
boulder falls using lichenometry, Bull (2009) was able to reconstruct the timing of past earthquakes along the San Andreas fault over the past 200 years (Figure 33).
Giant landslides (41 million m3) are commonly attributed to earthquakes, but dating them and proving a seismic
origin is challenging. Dortch et al. (2009) provided an interesting study of four giant landslides in the Himalaya and
Transhimalaya of northern India (Figure 34). The landslides
were dated using 10Be TCN surface exposure methods and
Dortch et al. (2009) compiled data on all the giant landslides
Paleoseismology
289
E
Bh
Nonbedded
organic matter rich
sand
Bh
Bh
Layer 5: Structureless
sand
Undisturbed
soil
Layer 4: Bedded
sequence of
alternating clean
and organic matter
rich sand
Undisturbed
soil
Layer 3: Sand and

many small clasts
BC
Crater wall
Layer 2: Sand and scattered

clasts of clean C-horizon sand
and A and Bh horizon material
BC
Shattered zone
Graded
zone
C
Tube-like dikes containing
structureless clean sand
Crater
Layer 1 sand and large clasts of

Bh horizon material
C Bedded clean sand
30 cm
Figure 29 Schematic vertical section of a filled liquefaction crater that forms a bowl in three dimensions. Reproduced from Obermeier, S.F.,
2009. Using liquefaction-induced and other soft-sediment features for paleoseismic analysis. In: McCalpin, J.P. (Ed.), Paleoseismology. Second
ed. Academic Press, Burlington, MA, pp. 497564. The filled crater in this figure long predates the 1886 Charleston earthquake, on the basis of
thickness of the Bh horizon in the filled crater.
Floor mat
Loose silty sand

with trace clay
Silty clay
Soft sand
Soft clay
Sand
Soft, friable
clayey silt
Source sand
Figure 30 Schematic vertical section showing where sills form by liquefaction during an earthquake. Reproduced from Obermeier, S.F., 2009.
Using liquefaction-induced and other soft-sediment features for paleoseismic analysis. In: McCalpin, J.P. (Ed.), Paleoseismology. Second ed.
Academic Press, Burlington, MA, pp. 497564.
290
Paleoseismology
in the Himalayan and Transhimalaya that have been dated; a

mere 16 in total. The landslide ages clustered at approximately
78 ka, coincident with a period of enhanced monsoon precipitation, but the associated age errors were large. Dortch
et al. (2009) argued the clustering of ages showed that the
landslides occurred during times of strengthened monsoon
and that increased rainfall might have helped trigger the
landslides. Alternatively, the four landslides of the form and
size examined could have been triggered by earthquakes. As
5.12.5.5
Earthquake-Triggered Cracks
Other off-fault paleoseismic structures include meter-size

cracks produced during great earthquakes (Gonzalez et al.,
2008; Loveless et al., 2009; Figure 35). These are generally not
preserved because they are erased by erosion very shortly after
their formation (typically after a few thousand years). However, abundant surface cracks spanning hundreds of thousands
of years have been preserved in the Atacama Desert along
coastal stretches of northern Chile and southern Peru because
erosion rates there are practically zero due to the hyperarid
climate. Loveless et al. (2009) pointed out that cracks of the
kind in the Atacama have been observed to form during and/
or shortly after strong subduction earthquakes (Figure 35).
The preservation of cracks for long time periods throughout
the Atacama Desert demonstrates evidence for multiple episodes of crack reactivation, and shows changes in crack
orientation over spatial scales similar to the size of earthquake
segments (Loveless et al., 2009). Loveless et al. (2009) suggest
that the meter-scale cracks can be used to map characteristic
earthquake rupture segments that persist over many seismic
cycles to better constrain the persistence of asperities in other
arid, tectonically active regions.
5.12.6
10
Farthest liquification effects
(sand blows, fissures, and such) for
shallow-focus earthquakes
Farthest liquification effects for
deep-focus earthquakes
9
Earthquake moment magnitude (Mw)
such, this study illustrates the potential and challenges of

using landslides as paleoseismic indicators.
Farthest sand blows reported for

1811-12 new madrid earthquakes
5
10
20
50 100 200
Epicentral distance (radius) to
farthest liquification feature (km)
500
Figure 31 Relationship between earthquake moment magnitude

(MW) and distance from earthquake epicenter to the furthest
liquefaction effect (venting to the surface or ground fracturing) with
bound suggested by Ambraseys (1988) which is for estimating the
minimum value of magnitude. Curve separates data from earthquakes
worldwide that had focal depths o50 km from those having depths
450 km. Reproduced from Obermeier, S.F., 2009. Using liquefactioninduced and other soft- sediment features for paleoseismic analysis.
In: McCalpin, J.P. (Ed.), Paleoseismology. Second ed. Academic
Press, Burlington, MA, pp. 497564.
Tsunami wave
(from distal seismic event)
Contribution to Seismic Hazards
Seismic hazard analysis to help reduce losses due to destructive

earthquakes can be divided into two main aspects: first, longterm forecast of future large earthquakes and second, an estimate of the amount of ground motion that may occur during
an earthquake and the frequency or repeat time at which they
occur. The methodology and approaches that are adopted are
divided into probabilistic and deterministic analyses. In addition, the likelihood of an earthquake occurring is assumed to
be either time-dependent or time-independent. Details of the
utilization of paleoseismic data in seismic hazard analysis are
described in detail in McCalpin (1996, 2011). Underlying ideas
and assumptions related to the utilization of a paleoseismological dataset are briefly described below.
Bay, estuary, lagoon
Land surface
Tsumani deposits
Tsunami deposits
Pre-event sediment accumulation

(e.g. Foreshore sand, aeolian dunes, etc.)
Gravity mass flow

Figure 32 Schematic illustration of principal pathways of tsunami sediment transport and deposition. Reproduced from Einsele, G., Chough
S.K., Shiki, T. 1996. Depositional events and their records an introduction. Sedimentary Geology 104, 19.
Paleoseismology
291
Lichen size (mm)

10
7.28
95% confidence band for

probability density plot
0.8
Probability density %/yr
19.15
15
0.6
0.4
0.2
1925
1909
1890
1900
1906
1877
1857
1837
1875
1850
Calendric age (A.D.)
1890
1872
1857
1815
1825
Unknown
1800
1812
Dates of earthquakes known as of 1996

Figure 33 Modeled times of rockfalls dated using lichen (Rhizocarpon subgenus Rhizocarpon) from 10 Sierra Nevada sites, which requires
regional causes. Reproduced from Bull, W.B., 2009. Tectonic Geomorphology of Mountains: A New Approach to Paleoseismology. Blackwell,
Oxford, 316 pp. The 1872 and 1890 events are local earthquakes and 1812, 1857, and 1906 are San Andreas fault earthquakes. The 1837 peak
might represent a San Andreas earthquake of 1938.
Figure 34 Giant landslide at Patseo in the India Himalaya. This

landslide was dated to 7.970.8 ka by Dortch et al. (2009) using 10Be
TCN surface exposure methods and may have been triggered by a
large earthquake.
One of the main focuses in seismic hazard assessment has

been predicting the probability of earthquakes occurring
within the next few decades, and ultimately when in that time
they might occur. Probability maps are the most common
ways of presenting the likelihood of earthquake occurrence
(e.g., Headquarters for Earthquake Research Promotion, 2005;
Working Group on California Earthquake Probabilities, 2008).
The probabilities are calculated as either time-dependent and
time-independent, but reliable and accurate estimates rely on
the use of time-dependent models. However, if the specific
fault has a long recurrence interval (greater than several
thousand years) then the probability of an earthquake within

the next 50 years reaches the upper saturation value of a few
percent (e.g., Headquarters for Earthquake Research Promotion, 2005). This means that the probability of a large earthquake occurrence during a human lifetime is negligible.
However, such large, low-frequency earthquakes certainly have
the potential to cause severe damage and losses with
devastating effects. From a paleoseismologists viewpoint, the
recurrence of large earthquakes over geologic time is most
probably time-dependent, even if considering very high natural
variability. However, most seismologists supporting the
GutenbergRichter relation argue that earthquakes occur randomly following a Poissonian probability distribution.
These contrasting opinions on the frequency and occurrence
of large earthquakes are likely the result of the differing timespans and spatial contexts that are considered. In general,
paleoseismologists focus on a single fault or a fault system,
whereas seismologists consider a broader regional view across
many faults. Moreover, this means that paleoseismologists will
need to work toward formulating more sophisticated timedependent models to satisfy data acquired across different
timeframes.
Source characterization is the start and basis for predicting
the peak ground acceleration and velocity for a site and/or
region. Typical peak ground acceleration and velocity are presented as probabilistic ground motion maps (e.g., National
Seismic Hazard Maps for Japan Headquarters for Earthquake
Research Promotion, 2005; the U.S. National Seismic Hazard
Map Petersen et al., 2008). In this context, most of major
active (Quaternary) faults are included as source faults that
will produce ground motion. These seismic hazard maps are
292
Paleoseismology
(a)
(b)
(c)
Figure 35 (a) Giant cracks traversing part of the Atacama desert in Chile. These cracks are produced during great earthquakes and are
preserved for many hundreds of thousands of years because of extreme aridity. (b) Close up through an earthquake-triggered crack. The crack
fill stratigraphy indicates that the crack has been progressively opened by a succession of great earthquakes over hundreds of thousands of
years. (c) A crack opened during a recent earthquake.
prepared to satisfy the requirements of how ground shaking

might occur within the next few decades. The cell size used to
calculate the predicted ground motion is generally a few square
kilometers in area, constrained by the computing limitations.
Unfortunately, the resolution of a calculated cell size may not
be great enough for the user to know how the ground motion
will affect individual houses and specific buildings. Ground
motion is very sensitive to ground conditions, notably the
nature of the substrate (sediment and/or bedrock) and topography, which have a nonlinear response to seismic waves. As
such, high-resolution geomorphic and Quaternary geology
maps are very important in supplementing ground motion
maps for developing very detailed seismic hazard assessments.
Another type of ground motion prediction is derived from
considering an earthquake of a particular size for a specific
fault and seismic source. An example of the source characterization for a crustal earthquake is described in Irikura and
Miyake (2011). Active fault and paleoseismological data such

as fault dimension and slip heterogeneity are utilized for
constraining the strong ground motion prediction. Deterministic seismic hazard analysis is also undertaken for
specific buildings such as nuclear power plants. In such cases,
the maximum acceleration for a specific moment magnitude
on a specific fault is one of the main parameters used. But the
predicted maximum acceleration is sometimes unreliable because of the different timeframes considered for the particular
dataset that is being used.
With respect to off-fault paleoseismological datasets, Reicherter et al. (2009) provided a useful summary of how historical
and prehistorical records of earthquake ground effects can
be useful for seismic hazard assessment. Figure 36 summarizes
the main features and dimensions that are relevant to earthquake environmental effects assessment. Most of the landforms,
structures, and features in Figure 36 have been described above.
Paleoseismology
293
Figure 36 Environmental Seismic Intensity Scale 2007 chart summarizing the main features and dimensions of the more relevant environmental
effects of earthquakes. Reproduced from Reicherter, K., Michetti, A.M., Silva, P.G., 2009. Palaeoseismology: historical and prehistorical records of
earthquake ground effects for seismic hazard assessment. Special Paper of the Geological Society, London, vol. 316, p. 32.
294
Paleoseismology
Thus, it should be emphasized that data on active faults

and past earthquakes are fundamental elements for earthquake prediction and forecasting. Geomorphic and geologic
research is therefore the gateway for seismic hazard analysis. It
follows that acquiring reliable paleoseismic data with increasedaccuracy and precision is extremely important to help develop more reliable applications for the seismic hazard
mitigation.
5.12.7
Challenges
Paleoseismology is a relatively young discipline, yet it has

developed rapidly over the past few decades as plate tectonics
has evolved as the framework for geologic studies. The digital
revolution has also helped paleoseismology to advance significantly, including remote sensing, geodesy, digital photography, and microcomputing. Moreover, development of new
dating methods has enabled the age of previously undated
landforms and sediments to be determined to help define the
timing of past earthquakes and rates of horizontal and vertical
crustal displacement. However, there are many major challenges for paleoseismology to overcome in the coming decades. Notable ones include improving the accuracy and
precision of numerical dating to define the timing of past
earthquakes and the ages of landforms and sediments to determine slip rates. In particular, a need exists to develop dating
methodologies to help push the dating range back beyond the
normal radiocarbon range of about 30 ka. Unfortunately,
newly developing dating methods such as OSL and TCN have
large uncertainties associated with them for ages older than a
few tens of thousands of years (and even, in some cases, of
younger ages). This makes it very difficult to obtain accurate
and precise paleoseismic data beyond a few tens of thousands
of years. Therefore, pushing the record back further is clearly
going to be one of the great challenges for the future.
Improved fault trenching methodologies are needed for
areas that have previously been avoided because they have low
sedimentation rates and do not readily record every seismic
event. As new methods develop, however, we can begin to
explore environments that were once out of our reach, such as
the offshore realm or great depths underground. Densely
spaced and long paleoseismic records along a fault or within a
fault system are essential to reconstruct details of past rupture
history and paleoseismic sources.
With the advancements in field data acquisition, another
major challenge is in improving models for earthquake
cycles. As briefly described above, the recurrence models established during the 197080s are still being used and these
need to be continuously tested and refined. This will likely be
achieved as more abundant paleoseismic data are acquired.
In particular, the number of sites for slip-per-event reconstructions needs to be increased throughout the world to
improve knowledge on multisegment earthquakes, particularly for the very largest earthquakes. Combining the new
tools used in fault trenching and slip measurement including
GPS and ground/airborne LiDAR surveys will greatly help
increase data acquisition in the field throughout the world.
Furthermore, the development of seismological knowledge
and understanding the physics of earthquakes will lead to
further cohesive advances in paleoseismology. In particular,

improved computer modeling may also help to enhance and
refine existing models, especially for spatial expansion of field
data at a particular site.
Probably the great challenge is in using paleoseismic data
to help in seismic hazard mitigation; that is, transferring the
lessons learned from paleoseismology to policymakers, planners, and managers so that they can use them most effectively
in reducing the damaging effects of earthquakes. During the
preparation of this article, the 11 March 2011 Tohoku earthquake, M9.0, occurred off the Pacific coast of Japan. This gigantic megathrust earthquake produced widespread tsunami
and ground motions, greater than anything ever recorded on
seismograms in Japan. Ironically, extensive paleoseimological
research on paleotsunami deposits in the Tohoku region was
well underway before the 11 March 2011 earthquake (Satake
et al., 2007; Sawai et al., 2007; Shishikura et al., 2007). These
studies highlighted the existence of widespread tsunamiite on
land some 2.53.0 km away from the present coastline, and
that this correlated with the AD 869 Jogan tsunami that was
documented in historical records. These studies had already
been presented to the government authorities, but the paleoseismic evidence had not been fully accepted and widely reported. In fact, this paleoseismic evidence was the only
warning of the likely catastrophic event that was to come on 11
March 2011 and it clearly illustrates the importance of paleoseismic research. In this context, enhanced efforts should be
paid to improving paleoseismological research including
using it to educate, train, and disseminate knowledge to government agencies and officials as well as the public to help
protect against destructive earthquakes.
Acknowledgments
Thanks to Tim Phillips and Yuko Ono for drafting most of the
figures.
References
Aitken, M.J., 1985. Thermoluminescence Dating. Academic Press, London, 359 pp.
Aitken, M.J., 1998. An Introduction to Optical Dating. Oxford University Press,
Oxford, 267 pp.
Allen, C.R., 1996. Seismological and paleoseismological techniques of research in
active tectonics. In: Wallace, R.E. (Ed.), Active Tectonics: Studies in Geophysics.
National Academies Press, Washington, DC, pp. 148154.
Ambraseys, N.N., 1988. Engineering seismology: Earthquake engineering and
structural dynamics. Journal of International Association of Earthquake
Engineering 17, 1105.
Anderson, R.S., Densmore, A.L., Ellis, M.A., 1999. The generation and degradation
of marine terraces. Basin Research 11, 719.
Anderson, R.S., Repka, J.L., Dick, G.S., 1996. Dating depositional surfaces using
in situ produced cosmogenic radionuclides. Geology 24, 4751.
Arrowsmith, J.R., Zielke, O., 2009. Tectonic geomorphology of the San Andrea fault
zone from high resolution topography: an example from the Cholame segment.
Geomorphology 113(12), 7081. http://dx.doi.org/10.1016/
j.geomorph.2009.01.002.
Atwater, B.A., 1987. Evidence for great Holocene earthquakes along the outer coast
of Washington State. Science 236, 942944.
Atwater, B.F., 1992. Geologic evidence for earthquakes during the past 2,000 years
along the Copalis River, southern coastal Washington. Journal of Geophysical
Research 97, 19011919.
Paleoseismology
Atwater, B.F., 1996. Coastal evidence for great earthquakes in western Washington. In:
Rogers, A.M. et al. (Eds.), Assessing Earthquake Hazards and Reducing Risk in
the Pacific Northwest. U.S. Geological Survey Professional Paper 1560, 7790.
Atwater, B.F., Moore, A.L., 1992. A tsunami about 1000 years ago in Puget Sound,
Washington. Science 257, 16141617.
Atwater, B.F., Musumi-Rokkako, S., Satake, K., 2005. The Orphan Tsunami of 1700:
Japanese Clues to a Parent Earthquake in North America. University of
Washington Press, Seattle, WA, 144 pp.
Atwater, B.F., Nelson, A.R., Clague, J.J., et al., 1995. Summary of coastal geologic
evidence for past great earthquakes at the Cascadia subduction zone. Earthquake
Spectra 11, 118.
Atwater, B.F., Stuiver, M., Yamaguchi, D.K., 1991. Radiocarbon test of earthquake
magnitude at the Cascadia subduction zone. Nature 353, 156158.
Atwater, B.F., Yamaguchi, D.K., 1991. Sudden, probably coseismic submergence of
Holocene trees and grass in coastal Washington State. Geology 19, 706709.
Avouac, J.-P., Ayoub, F., Leprince, S., Konca, O., Helmberger, D.V., 2006. The
2005, Mw 7.6 Kashmir earthquake: Sub-pixel correlation of ASTER images and
seismic waveforms analysis. Earth Planetary Science Letters 249, 514528. http:
//dx.doi.org/10.1016/j.epsl.2006.06.025.
Awata, Y., Toda, S., Kaneda, H., Azuma, T., Horikawa, H., Shishikura, M., Echigo, T.,
2008. Coastal deformation associated with the 2007 Noto Hanto earthquake,
central Japan, determined by uplifted and subsided intertidal organisms. Earth
Planets and Space 60, 10591062.
Behr, W.M., Rood, D.H., Fletcher, K.E., et al., 2010. Uncertainties in slip-rate
estimates for the Mission Creek strand of the southern San Andreas fault at
Biskra Palms Oasis, southern California. Geological Society of America Bulletin
122, 13601377.
Berryman, K.R., Ota, Y., Hull, A.G., 1989. Holocene paleoseismicity in the fold and
thrust belt of the Hikurangi subduction zone, eastern North Island, New Zealand.
Biasi, G.P., Weldon, II R.J., Fumal, T.E., Seitz, G.G., 2002. Paleoseismic event
dating and the conditional probability of large earthquakes on the southern San
Andreas fault, California. Bulletin of the Seismological Society of America 92,
27612781.
Bierman, P.R., Gillespie, A.R., Caffee, M.W., 1995. Cosmogenic ages for earthquake
recurrence intervals and debris flow fan deposition, Owens Valley, California.
Science 270, 447450.
Blisniuk, K., Rockwell, T., Owen, L.A., Oskin, M., Lipponett, C., Caffee, M.C.,
Dortch, J., 2010. Late Quaternary slip rates gradient defined using highresolution topography and 10Be dating of offset landforms on the southern San
Jacinto Fault zone, California. Journal of Geophysical Research 115, B08401. htt
p://dx.doi.org/10.1029/2009JB006346.
Bloom, A.L., Broecker, W.S., Chappell, J., Matthews, R.K., Mesolella, K.J., 1974.
Quaternary sea level fluctuations on a tectonic coast; new 230Th/234U dates from
the Huon Peninsula, New Guinea. Quaternary Research 4, 185205.
Bookhagen, B., Echtler, H.P., Melnick, D., Strecker, M.R., Spencer, J.Q.G., 2006.
Using uplifted Holocene beach berms for paleoseismic analysis on the Santa
Maria Island, southcentral Chile. Geophysical Research Letters 33, 5.
Brown, E.T., Bendick, R., Bourles, D.L., Gaur, V., Molnar, P., Raisbeck, G.M., Yiou,
F., 2002. Slip rates of the Karakorum Fault, Ladakh, India, determined using
cosmic ray exposure dating of debris flows and moraines. Journal of
Geophysical Research 107, 2192. http://dx.doi.org/10.1029/2000JB000100.
Bull, W.B., 2009. Tectonic Geomorphology of Mountains: A New Approach to
Paleoseismology. Blackwell, Oxford, 316 pp.
Burbank, D.W., Anderson, R.S., 2001. Tectonic Geomorphology. Blackwell Science,
Oxford, 274 pp.
Burbank, D.W., Verges, J., 1994. Reconstruction of topography and related
depositional systems during active thrusting. Journal of Geophysical Research
99, 2028120297.
Chappell, J., Omura, A., Esat, T., McCulloch, M., Pandolfi, J., Ota, Y., Pillans, B.,
1996. Reconciliation of late Quaternary sea levels derived from coral terraces at
Huon Peninsula with deep sea oxygen isotope records. Earth and Planetary
Chevalier, J.-L., Ryerson, F.J., Tapponnier, P., Finkel, R.C., Van der Woerd, J.,
Haibing, L., Qing, L., 2005. Slip-rate measurements on the Karakorum fault may
Crone, A.J., Machette, M.N., Bradley, L.-A., Mahan, S.A., 1997. Late Quaternary
surface faulting on the Cheraw Fault, Southeastern Colorado. Geologic
Investigations Map, USGS Map I-2591.
Dawson, A.G., 2000. Tsunami deposits. Pure and Applied Geophysics 157,
875897.
Dawson, A.G., Long, D., Smith, D.E., 1988. The Storegga Slides: evidence for a
possible tsunami in eastern Scotland. Marine Geology 82, 271276.
295
Dawson, A.G., Stewart, I., 2007. Tsunami deposits in the geological record.
Dolan, J.F., Christofferson, S.A., Shaw, J.H., 2003. Recognition of paleoearthquakes
on the Puente Hills blind thrust fault, Los Angeles, California. Science 300,
115118.
Dortch, J., Owen, L.A., Haneberg, W.C., Caffee, M.W., Dietsch, C., Kamp, U., 2009.
Nature and timing of mega-landslides in northern India. Quaternary Science
Reviews 28, 10371056.
Einsele, G., Chough, S.K., Shiki, T., 1996. Depositional events and their records
an introduction. Sedimentary Geology 104, 19.
Ferry, M., Meghraoui, M., Karaki, N.A., Al-Taj, M., Amoush, H., Al-Dhaisat, S.,
Barjous, M., 2007. A 48-kyr-long slip rate history for the Jordan Valley
segment of the Dead Sea fault. Earth Planetary Science Letters 260,
394406.
Fletcher, K.E., Sharp, W.D., Kendrick, K.J., Behr, W.M., Hudnut, K.W., Hanks, T.C.,
2010. 230Th/U dating of a late Pleistocene alluvial fan along the southern San
Andreas fault. Geological Society of America Bulletin 122, 13471359.
Frankel, K.L., Brantley, K.S., Dolan, J.F., et al., 2007a. Cosmogenic 10Be and 36Cl
Implications for transient strain in the eastern California shear zone. Journal of
Geophysical Research Solid Earth 112, B06407. http://dx.doi.org/10.1029/
2006JB004350.
Frankel, K.L., Dolan, J.F., Finkel, R.C., Owen, L.A., Hoeft, J.S., 2007b. Spatial
variations in slip rate along the Death ValleyFish Lake Valley fault system
determined from LiDAR topographic data and cosmogenic 10Be geochronology.
Geophysical Research Letters 34, LI8303.
Fu, B., Lei, X., Hessami, K., Ninomiya, Y., Azuma, T., Kondo, H., 2007. A new fault
rupture scenario for the 2003 Mw 6.6 Bam earthquake, SE Iran: Insights from
the high-resolution QuickBird imagery and field observations. Journal of
Geodynamics 44, 160172.
Ganev, P.N., Dolan, J.F., Blisniuk, K., Oskin, M., Owen, L.A., 2010.
Paleoseismologic evidence for multiple Holocene earthquakes on the Calico
fault: implications for earthquake clustering in the Eastern California Shear Zone.
Lithosphere 4, 287298.
Gardner, T., Marshall, J., Merritts, D., et al., 2001. Holocene forearc block rotation
in response to seamount subduction, southeastern Pennsula de Nicoya, Costa
Rica. Geology 29, 151154.
Gonzalez, G., Gerbault, M., Martinod, J., Cembrano, J., Allmendinger, R., Carrizo,
D., Espina, J., 2008. Crack formation on top of propagating reverse faults of the
Chuculay Fault System northern Chile: insights from field data and numerical
modeling. Journal of Structural Geology 30, 791808. http://dx.doi.org/10.1016/
j.jsg.2008.02.008.
Gosse, J.C., Phillips, F.M., 2001. Terrestrial in situ cosmogenic nuclides; theory
Grant, L.B., Lettis, W.R., (Eds.) 2002. Paleoseismology of the San Andreas Fault
System. Bulletin of the Seismological Society of America, Special Issue 92(7),
25512554.
Gurrola, L.D., Keller, E., A., Chen, J., Owen, L.A., Spencer, J. Q., in press. Tectonic
geomorphology of marine terraces, Santa Barbara Fold Belt, California.
Geological Society of America Bulletin.
Gutenberg, B., Richter, C.F., 1954. Seismicity of the Earth and Associated
Phenomena, Second ed. Princeton University Press, Princeton.
Hancock, G., Anderson, R., Chadwick, O., Finkel, R., 1999. Dating fluvial terraces
with 10Be and 26Al profiles: application to the Wind River, Wyoming.
Headquarters for Earthquake Research Promotion, 2005. Report: National Seismic
Hazard Maps for Japan (2005). p.158.
Hempton, M.R., Dewey, J.F., 1983. Earthquake-induced deformation structures in
young lacustrine sediments, East Anatolian Fault, southeast Turkey.
Hori, T., Oike, K., 1999. A physical mechanism for temporal variation in seismicity
in Southwest Japan related to great interplate earthquakes along the Nantai
trough. Tectonophysics 308, 8398.
Hubert-Ferrari, A., Armijo, R., King, G., Meyer, B., Barka, A., 2002. Morphology,
displacement, and slip rates along the North Anatolian Fault, Turkey. Journal of
Geophysical Research 107(B10), 2235.
Hutton, J., 1795. Theory of the Earth with Proofs and Illustrations. Creech,
Edinburgh 2 volumes, 297 pp.
Ikeda, Y., Yonekura, N., 1986. Determination of late Quaternary rates of net slip on
two major fault zones in central Japan, Bulletin of the Department of Geography.
University of Tokyo 18, 4963.
Irikura, K., Miyake, H., 2011. Recipe for predicting strong ground motion from
crustal earthquake scenarios. Pure and Applied Geophysics 168, 85104.
296
Paleoseismology
Ishiyama, T., Mueller, K., Togo, M., Okada, A., Takemura, K., 2004. Geomorphology,
kinematic history, and earthquake behavior of the active Kuwana wedge thrust
anticline, central Japan. Journal of Geophysical Research 109, B12408. http://d
x.doi.org/10.1029/2003JB002547.
Jayangondaperumal, R., Wesnousky, S.G., Chaudhari, B.K. 2011. Paleoseismological
evidence for earlylate Holocene surface faulting along the northeastern
Himlayan front at the western flank of eastern syntaxis, India. Bulletin of
Seismological Society of America, in press.
Keller, E.A., DeVecchio, D., 2013. Tectonic geomorphology of active folding and
development of transverse drainages. In: Owen, L.A. (Ed.), Treatise on
Geomorphology. Academic Press, San Diego, vol. 5, 20 pp.
Keller, E.A., Gurrola, L.D., 2000. Earthquake hazard of the Santa Barbara fold belt,
California, USGS National Earthquake Hazard Reduction Program, final report,
award number#99HQGR0081: http://www.scec.org/research/98research/
98gurrolakeller.pdf
Kent, G.M., Babcock, J.M., Driscoll, N.W., et al., 2005. A 60 k.y. record of
extension across the western boundary of the Basin and Range province:
estimate slip rates from offset shoeline terraces and a catastrophic slide beneath
Lake Tahoe. Geology 33, 365368.
Knuepfer, P.L.K., 1987. Changes in Holocene slip rates in strike-slip environments.
U.S. Geological Survey Open File Rep 87-673, 249261.
Kondo, H., Nakata, T., Akhtar, S., et al., 2008b. Long recurrence interval of faulting
beyond the 2005 Kashmir earthquake around the northwestern margin of IndoAsia collision zone. Geology 36, 731734.
Kondo, H., Ozaksoy, V., Yldirim, C., 2010. Slip history of the 1944 Bolu-Gerede
earthquake rupture along the North Anatolian fault system: Implications for
recurrence behavior of multisegment earthquakes. Journal of Geophysical
Research 115, B04316. http://dx.doi.org/10.1029/2009JB006413.
Kondo, H., Toda, S., Imaizumi, T., et al., 2003. Recent non-characteristic behavior
along the Tanna fault based on three-dimensional trenching and Geoslicer
techniques. Zishin2 55.474424.
Kondo, H., Toda, S., Okumura, K., Chiba, T., 2008a. A fault scarp in an urban area
identified by LIDAR survey: a case study on the Itoigawa-Shizuoka tectonic line,
central Japan. Geomorphology 101, 731739.
Kortekaas, S., Dawson, A.G., 2007. Distinguishing tsunami and storm deposits:
an example from Martinhal, SW Portugal. Sedimentary Geology 200,
208221.
Kumar, S., Wesnousky, S.G., Jayangondaperumal, R., Nakata, T., Kumahara, Y.,
Singh, V., 2010. Paleoseismological evidence of surface faulting along the
northeastern Himalayan front, India; timing, size, and spatial extent of great
earthquakes. Journal of Geophysical Research 115, B12422.
Kumar, S., Wesnousky, S.G., Rockwell, T.K., Briggs, R.W., Thakur, V.C.,
Jayangondaperumal, R., 2006. Paleoseismic evidence of great surface-rupture
earthquakes along the Indian Himalaya. Journal of Geophysical Research 111,
B03304.
Kumar, S., Wesnousky, S.G., Rockwell, T.K., Ragona, D., Thakur, V.C., Seitz, G.G.,
2001. Earthquake recurrence and rupture dynamics of Himalayan Frontal Thrust,
India. Science 294, 23282331.
Le, K., Lee, J., Owen, L.A., Finkel, R.C., 2007. Late Quaternary slip rates along the
Sierra Nevada frontal fault zone, California: slip partitioning across the western
margin of the Eastern California Shear Zone/Basin and Range Province.
Lee, J., Spencer, J.Q., Owen, L.A., 2001. Holocene slip rates along the Owens
Valley fault, California: implications for the recent evolution of the Eastern
California Shear Zone. Geology 29, 819822.
the Inyo Mountains, California: implications for the timing of extension along
the western boundary of the Basin Range Province and distribution of dextral
fault slip rates across the Eastern California Shear Zone. Tectonics 28, TC1001.
http://dx.doi.org/10.1029/2008TC002295.
Lew, M., Real, C.R. 2004. Implementation of the seismic hazard mapping act in
California. In 13th World Conference on Earthquake Engineering, Vancouver,
BC., Canada, paper 3166.
Lindvall, S.C., Rockwell, T.K., Hudnut, K.W., 1989. Evidence for prehistoric
earthquakes on the Superstition Hills fault from offset geomorphic features.
Liu-Zeng, J., Klinger, Y., Sieh, K., Rubin, C., Seitz, G., 2006. Serial ruptures of the
San Andreas fault, Carrizo Plain, California, revealed by three-dimensional
excavations. Journal of Geophysical Research 111, B02306. http://dx.doi.org/
10.1029/2004JB003601.
Loveless, J.P., Allmendinger, R.W., Pritchard, M.E., Garroway, J.L., Gonzalez, G.G.,
2009. Surface cracks record long-term seismic segmentation of the Andean
margin. Geology 37, 2326. http://dx.doi.org/10.1130/G25170A.1.
Mason, D.P.M., Little, T.A., Van Dissen, R.J., 2006. Rates of active faulting during
late Quaternary fluvial terrace formation at Saxton River, Awatere fault, New
Zealand. Geological Society of America Bulletin 118, 14311446.
Matmon, A., Schwartz, D., Haeussler, P., Finkel, R., Lienhaemper, J.J., Stenner, H.D.,
Dawson, T.E., 2006. Denali fault slip rates and Holocenelate Pleistocene
kinematics of central Alaska. Geology 34, 645648.
Matmon, A., Schwartz, D.P., Finkel, R., Clemmens, S., Hanks, T., 2005. Dating
offset fans along the Mojave section of the San Andreas fault using cosmogenic
26
Al and 10Be. Geological Society of America Bulletin 117, 795807.
Matsuda, T., 1975. Magnitude and recurrence interval of earthquakes from a fault.
Zishin2 28, 269283.
McCalpin, J.P. (Ed.), 1996. Paleoseismology, First ed. Academic Press, San Diego,
588 pp.
McCalpin, J.P. (Ed.), 2009a. Paleoseismology, Second ed. Academic Press,
Burlington, MA, 613 pp.
McCalpin, J.P., 2009b. Field techniques in paleoseismology terrestrial
environments. In: McCalpin, J.P. (Ed.), Paleoseismology, Second ed. Academic
McCalpin, J.P., 2009c. Paleoseismology in extensional tectonic enviroments.
In: McCalpin, J.P. (Ed.), Paleoseismology, Second ed. Academic Press,
Burlington, MA, pp. 171270.
McCalpin, J.P., Carver, G.A., 2009. Paleoseismology of compressional tectonic
environments. In: McCalpin, J.P. (Ed.), Paleoseismology, Second ed. Academic
Meghraoui, M., Camelbeeck, T., Vanneste, K., Brondeel, M., Jongmans, D., 2000.
Active faulting and paleoseismology along the Bree fault, lower Rhine graben,
Belgium. Journal of Geophysical Research 105(B6), 1380913841.
Meghraoui, M., Philip, H., Albarede, F., Cisternas, A., 1988. Trench investigations
through the trace of the 1980 El Asnam thrust fault. Bulletin of the
Meriaux, A.-S., Ryerson, F.J., Tapponnier, P., et al., 2004. Rapid slip along the
central Altyn Tagh Fault: morphochronologic evidence from Cherchen He and
Sulamu Tagh. Journal of Geophysical Research 109, B06401.
Muhs, D.R., Kennedy, B.J., Rockwell, T.K., 1994. Uranium-series ages of marine
terrace corals from the Pacific coast of North America and implications for lastinterglacial sea level history. Quaternary Research 42, 7287.
Muhs, D.R., Rockwell, T.K., Kennedy, G.L., 1992. Late Quaternary uplift rates of
marine terraces on the Pacific coast of North America, southern Oregon to Baja
California. Quaternary International 1516, 121133.
Muhs, D.R., Simmons, K.R., Schumann, R.R., Groves, L.T., Mitrovica, J.X., Laurel,
D., 2012. Sea-level history during the last interglacial complex on San Nicolas
Island, California: implications for glacial isostatic adjustment processes,
paleozoogeography and tectonics. Quaternary Science Reviews 37, 125.
Muhs, D.R., Szabo, B.J., 1982. Uranium-series age of the Eel Point terrace, San
Clemente Island, California. Geology 10, 2326.
Murray, A.S., Wintle, A.G., 2000. Luminescence dating of quartz using an improved
single-aliquot regenerative-dose protocol. Radiation Measurements 32, 5773.
Murray, J., Segall, P., 2002. Testing time-predictable earthquake re-currence by
direct measurement of strain accumulation and release. Nature 419, 287291.
Nakata, T., Shimazaki, K., 1997. Geo-slicer, a newly-invented soil sampler for
high-resolution active fault studies. Journal of Geography (Chigaku Zasshi)
106, 59.
National Research Council, 2008. Earth Observations from Space: The First 50
Years of Scientific Achievements. The National Academies Press, Washington,
DC, 142 pp.
National Research Council, 2010. Landscapes on the Edge: New Horizons for
Research on Earths Surface. The National Academies Press, Washington, DC,
180 pp.
Nelson, A.R., Shennan, I., Long, A.J., 1996. Identifying coseismic subsidence in
tidal wetland stratigraphic sequences at the Cascadia subduction zone of western
North America. Journal of Geophysical Research 101(B3), 61156135.
Niemi, T.M., Hall, N.T., 1992. Late Holocene slip rate and recurrence of great
earthquakes on the San Andreas fault in northern California. Geology 20,
195198.
Nishenko, S.P., 1985. Seismic potential for large and great interplate earthquakes
along the Chilean and southern Peruvian margins of South America: a
quantitative reappraisal. Journal of Geophysical Research 90, 35893615.
Noller, J.S., Sowers, J.M., Lettis, W.R. (Eds.), 2000. Quaternary Geochronology:
Methods and Applications. AGU Reference Shelf 4, American Geophysical
Union, Washington, DC.
Obermeier, S.F., 1996. Use of liquefaction-induced features for paleoseismic
analysis an overview of how seismic liquefaction features can be
distinguished from other features and how their regional distribution and
Paleoseismology
properties of source sediment can be used to infer the location and strength of
Holocene paleo-earthquakes. Engineering Geology 44, 176.
Obermeier, S.F., 2009. Using liquefaction-induced and other soft-sediment features
for paleoseismic analysis. In: McCalpin, J.P. (Ed.), Paleoseismology, Second ed.
Academic Press, Burlington, MA, pp. 497564.
Okada, A., Watanabe, M., Ando, M., Tsukuda, T., Hirano, S., 1992. Estimation of
paleoseismicity in the Nobi active fault system, central Japan excavation study
of the Umehara fault, central strand in the Nobi active fault system. Joural of
Geography 101, 118 (in Japanese with English abstract).
Okamura, Y., Ishiyama, T., Yanagisawa, Y., 2007. Fault-related folds above the
source fault of the 2004 mid-Niigata Prefecture earthquake, in a fold-and-thrust
belt caused by basin inversion along the eastern margin of the Japan Sea.
Journal of Geophysical Research, B03S08. http://dx.doi.org/10.1029/
2006JB004320.
Ota, Y., 1975. Late Quaternary vertical movement in Japan estimated from deformed
shorelines. In: Suggate, R.P., Creswell, M.M. (Eds.), Quaternary Studies. Royal
Society of New Zealand, Wellington, pp. 231239.
Ota, Y., Chappell, J., Kelley, R., Yonekura, N., Matsumoto, E., Nishimura, T., 1993.
Holocene coral reef terraces and coseismic uplift of Huon Peninsula, Papua New
Guinea. Quaternary Research 40, 177188.
Owen, L.A., Cunningham, W.D., Richards, B., Rhodes, E.J., Windley, B.F.,
Dornjamjaa, D., Badamgarav, J., 1999. Timing of formation of forebergs in the
northeastern Gobi Altai, Mongolia: implications for estimating mountain uplift
rates and earthquake recurrence intervals. Journal of the Geological Society v.
156, 457464 Erratum: Journal of the Geological Society, 156, Contents xii.
Owen, L.A., Frankel, K.L., Knott, J.R., Reynhout, S., Finkel, R.C., Dolan, J.F., Lee,
J., 2011. Beryllium-10 terrestrial cosmogenic nuclide surface exposure dating of
Quaternary landforms in Death Valley. Geomorphology 125, 541557.
Pantosti, D., Schwartz, D.P., Valensise, G., 1993. Paleoseismology along the 1980
surface rupture of the Irpina fault: implications for earthquake recurrence in the
southern Apennines, Italy. Journal of Geophysical Research 98, 65616577.
Parsons, T., 2006. M47.0 earthquake recurrence on the San Andreas fault from a
stress renewal model. Journal of Geophysical Research 111, B12305. http://dx.d
oi.org/10.1029/2006JB004415.
Perg, L.A., Anderson, R.S., Finkel, R.C., 2001. Use of a new 10Be and 26Al
inventory method to date marine terraces, Santa Cruz, California, USA. Geology
29, 879882.
Petersen, M.D., Frankel, A.D., harmsen, S.C., et al., 2008. Documentation for the
2008 update of the United States National Seismic Hazard Maps. U.S.
Geological Survey Open File Report OF-08-118, p. 61.
Philibosian, B., Fumal, T.E., Weldon, R.J., et al., 2009. Photomosaics and logs of
trenches on the San Andreas Fault near Coachella, California. In U.S. Geological
Survey Open-File Report 2009-1039.
Philip, H., Rogozhin, E., Cisternas, A., Bousquet, J.C., Borisov, B., Karakhanian, A.,
1992. The Armenian earthquake of 1988 December 7: faulting and folding,
neotectonics and paleoseismicity. Geophysical Journal International 110,
141158.
Ramsey, B.C., 2001. Development of the radiocarbon calibration program OxCal.
Radiocarbon 43(2A), 355363.
Reicherter, K., Michetti, A.M., Silva, P.G., 2009. Palaeoseismology: historical and
prehistorical records of earthquake ground effects for seismic hazard
assessment. Special Publication of the Geological Society, London, vol. 316,
p. 324.
Reid, H.F., 1910. The mechanism of the earthquake. In: The California Earthquake of
April 18, 1906. Report of the State Earthquake Investigation Commission, vol. 2,
Carnegie Institution, Washington, DC, pp. 1192.
Research Group for Active Faults, 1980. Active Faults in Japan Sheet Maps and
Inventories. University of Tokyo Press, Tokyo, (in Japanese with English
abstract).
Research Group for Active Faults, 1991. Active Faults in Japan Sheet Maps and
Inventories, Revised ed. University of Tokyo Press, Tokyo, (in Japanese with
English abstract).
Ritz, J.F., Bourles, D., Brown, E.T., et al., 2003. Late Pleistocene to Holocene slip
rates for the Gurvan Bulag thrust fault (Gobi-Altay, Mongolia) estimated with
10
Be dates. Journal of Geophysical Research 108(B3), 2162.
Rockwell, T., Gath, E., Gonzalez, T., et al., 2010. Neotectonics and paleoseismology
of the Pedro Miguel fault zone in Panama: earthquake hazard to the Panama
Canal. Bulletin of the Seismological Society of America 100, 30973129.
Rockwell, T., Nolan, J., Johnson, D.L., Patterson, R.H., 1992. Ages and deformation
of marine terraces between Point Conception and Gaviota, Western Transverse
Ranges, California. SEPM Special Publication 48, 333241.
Rockwell, T.K., Ben-Zion, Y., 2007. High localization of primary slip zones in large
earthquakes from paleoseismic trenches: observations and implications for
297
earthquake physics. Journal of Geophysical Research 112, B10304. http://dx.doi.

org/10.1029/2006JB004764.
Rockwell, T.K., Lindvall, S., Herzberg, M., Murbach, D., Dawsaon, T., Berger, G.,
2000. Paleoseismology of the Johnson Valley, Kicjapoo, and Homestead Valley
Faults: clustering of earthquakes in the Eastern California Shear Zone. Bulletin
of the Seismological Society of America 90, 12001236.
Saillard, M., Hall, S.R., Audin, L., et al., 2009. Non-steady long-term uplift rates
and Pleistocene marine terrace development along the Andean margin of Chile
(311) inferred from 10Be dating. Earth and Planetary Science Letters 277, 5063.
Sarmiento, A.C., Wesnousky, S.G., Bormann, J.M., 2011. Paleoseismic trenches
across the Carson and Sierra Nevada rangefronts in Antelope Valley, California
and Reno, Nevada. Bulletin of the Seismological Society of America, 101. http://
dx.doi.org/10.1785/0120100176.
Sarna-Wojcicki, A.M., Lajoie, K.R., Yerkes, R.F., 1987. Recurrent Holocene
displacement on the Javon Canyon Fault; a comparison of fault-movement
history with calculated average recurrence intervals. U.S. Geological Survey
Professional Paper 1339.
Satake, K., Atwater, B.F., 2007. Long-term perspectives on giant earthquakes and
tsunamis at subduction zones. Annual Reviews of Earth and Planetary Sciences
35, 349274.
Satake, K., Sawai, Y., Shishikura, M., Okamura, Y., Namegaya, Y., Yamaki, S., 2007.
Tsunami source of the unusual AD 869 earthquake off Miyagi, Japan, inferred
from tsunami deposits and numerical simulation of inundation. American
Geophysical Union, Fall Meeting 2007, Abstract T31 G-03.
Satake, K., Shimazaki, K., Tsuji, Y., Ueda, K., 1996. Time and size of a giant
earthquake in Cascadia inferred from Japanese tsunami records of January
1700. Nature 379, 246249.
Sawai, Y., Fujii, Y., Fujiwara, O., et al., 2008. Marine incursions of the past 1500
years and evidence of tsunamis at Suijin-numa, a coastal lake facing the Japan
Trench. The Holocene 18, 517528.
Sawai, Y., Kamataki, T., Shishikura, M., et al. , 2008. Aperiodic recurrence of
geologically recorded tsunamis from the past 5,500 years in eastern Hokkaido,
Japan. Journal of Geophysical Research, 10.1029/2007JB005503.
Sawai, Y., Satake, K., Kamataki, T., et al., 2004. Transient uplift after a 17th-century
earthquake along the Kuril subduction zone. Science 306, 19181920.
Sawai, Y., Shishikura, M., Okamura, Y., et al., 2007. A study on paleotsunami using
handy geoslicer in Sendai Plain (Sendai, Natori, Iwanuma, Watari, and
Yamamoto), Miyagi, Japan. Annual Report on Active Fault and Paleoearthquake
Researches, Geological Survey of Japan/AIST 7, 3146.
Scharer, K.M., Weldon, III R.J., Fumal, T.E., Biasi, G.P., 2007. Paleoearthquakes on
the southern San Andreas fault, Wrightwood, California, 3000 to 1500 B.C.: a
new method for evaluating paleoseismic evidence and earthquake horizons.
Schlagenhauf, A., Manighetti, I., Benedetti, L., Gaudemer, Y., Finkel, R., Malavieille,
J., Pou, K., 2011. Earthquake supercycles in Central Italy, inferred from Cl
exposure dating. Earth Planetary Science Letters,. http://dx.doi.org/10.1016/
j.epsl.2011.05.022.
Scholz, C.H., 2002. The Mechanics of Earthquakes and Faulting, Second ed.
Cambridge University Press, New York, 496 pp.
Schwartz, D.P., Coppersmith, K.J., 1984. Fault behavior and characteristic
earthquakes: examples from the Wasatch and San Andreas fault zones. Journal
Shimazaki, K., Nakata, T., 1980. Time-predictable recurrence model for large
earthquakes. Geophysical Research Letters 7, 279282.
Shishikura, M., Echigo, T., Namegaya, Y., 2009. Evidence for coseismic and
aseismic uplift in the last 1000 years in the focal area of a shallow thrust
earthquake on the Noto Peninsula, west-central Japan. Geophysical Research
Letters 36, L02307.
Shishikura, M., Sawai, Y., Okamura, Y., et al., 2007. Age and distribution of
tsunami deposit in the Ishinomaki Plain, Northeastern Japan. Annual Report on
Active Fault and Paleoearthquake Researches, Geological Survey of Japan/AIST
7, 3146 (in Japanese with English abstract and figure captions).
Siame, L.L., Bourles, D.L., Sebrier, M., et al., 1997. Cosmogenic dating ranging
from 20 to 700 ka of a series of alluvial fan surfaces affected by the El Tigre
fault, Argentina. Geology 25, 975978.
Sibson, R.H., 1986. Earthquakes and rock deformation in crustal fault zones. Annual
Reviews of Earth and Planetary Science 14, 149175.
Sieh, K., Stuiver, M., Brillinger, D., 1989. A more precise chronology of earthquakes
produced by the San Andreas fault in southern California. Journal of
Sieh, K.E., 1978a. Pre-historic large earthquakes produced by slip on the San
Andreas Fault at Pallett Creek, California. Journal of Geophysical Research 83,
39073939.
298
Paleoseismology
Sieh, K.E., 1978b. Slip along the San Anreas fault associated with the great
1857 earthquake. Bulletin of the Seismological Society of America 68,
14211448.
Silva, P.G., Rodrguez Pascua, M.A., et al., 2008. Catalogacion de los efectos
geologicos y ambientales de los terremotos en Espana eb ak Escala
ESI-2007 y su aplicacion a los estudios paleosismologicos. Geotemas 6,
10631066.
Sims, J.D., 1975. Determining earthquake recurrence intervals from deformational
structures in young lacustrine sediments. Tectonophysics 29, 141152.
Somerville, P., Irikura, K., Graves, R., et al., 1999. Characterizing earthquake slip
models for the prediction of strong ground motion. Seismological Research
Letters 70, 5980.
Stein, R.S., King, G.C.P., 1984. Seismic potential revealed by surface folding: 1983
Coalinga, California, earthquake. Science 224, 867872.
Stein, R.S., Yeats, R.S., 1989. Hidden earthquakes. Science American 260,
4857.
Takada, K., Atwater, B.F., 2004. Evidence for liquefaction identified in peeled slices
of Holocene deposits along the lower Columbia River, Washington. Bulletin of
Thatcher, W., 1986. Cyclic deformation related to great earthquakes at plate
boundaries. In: Reilly, W.I., Harford, B.E. (Eds.), Recent Crustal Movements of
the Pacific Region. Bulletin of the Royal Society of New Zealand, vol. 24, pp.
245272.
Tsukamoto, S., Duller, G., Murray, A., Choi, J.-H., 2009. Application of
luminescence dating to Quaternary tectonic movements and environmental
change. Quaternary International 199, 1164.
Tsutsumi, H., Okada, A., Nakata, T., Ando, M., Tsukuda, T., 1991. Timing and
displacement of Holocene faulting on the Median Tectonic Line in central
Shikoku, southwest Japan. Journal of Structural Geology 13, 227233.
Tuttle, M.P., Ruffman, A., Anderson, T., Jeter, H., 2004. Distinguishing tsunami from
storm deposits in Eastern North America: The 1929 Grand Banks Tsunami
versus the 1991 Halloween Storm. Seismological Research Letters 75,
117131.
U.S. Geological Survey and New Mexico Bureau of Mines and Mineral Resources,
2006. Quaternary fault and fold database for the United States. Accessed Jan 9,
2006, from USGS web site: http://earthquake.usgs.gov/regional/qfaults/.
Upson, II J.E., 1951. Former marine shorelines of the Gaviota Quadrangle, Santa
Barbara County, California. Journal of Geology 59, 415446.
Van der Woerd, J.W., Ryerson, F., Tapponnier, P., et al., 1998. Holocene left-slip
rate determined by cosmogenic surface dating on the Xidatan segment of the
Kunlun fault (Qinghai China). Geology 26, 695698.
Van Dissen, R.J., Berryman, K.R., Pettinga, J.R., Hill, N.L., 1992. Paleoseismicity of
the Wellington-Hutt Valley segment of the Wellington Fault, North Island, New
Zealand. New Zealand Journal of Geology and Geophysics 35, 165176.
Walker, R.T., 2006. A remote sensing study of active folding and faulting in
southern Kerman province, S.E. Iran. Journal of Structural Geology 28,
654668.
Wallace, R.E., 1970. Earthquake recurrence intervals on the San Andreas fault.
Washburn, Z., Arrowsmith, J.R., Forman, S.L., Cowgill, E., Wang, X., Zhang, Y.,
Chen, Z., 2001. Late Holocene earthquake history of the central Altyn Tagh fault,
China. Geology 29, 10511054.
Weldon, II R.J., Fumal, T.E., Powers, T.J., Pezzopane, S.K., Scharer, K.M., Hamilton,
J.C., 2002. Structure and earthquake offsets on the San Andreas fault at the
Wrightwood, California, paleoseismic site. Bulletin of the Seismological Society
of America 92, 27042725.
Weldon, R.J., Scharer, K.M., Fumal, T.E., Biasi, G.P., 2004. Wrightwood and the
earthquake cycle: what a long recurrence record tells us about how faults work.
GSA Today 14, 410.
Wells, D.L., Coppersmith, K.J., 1994. New empirical relationships among
magnitude, rupture length, rupture area, and surface displacement. Bulletin of
Wesnousky, S.G., 2008. Displacement and geometrical characteristics of earthquake
surface ruptures: Issues and implications for seismic hazard analysis and the
earthquake rupture process. Bulletin of the Seismological Society of America 98,
16091632.
Wesnousky, S.G., Barron, A.D., Briggs, R.W., Caskey, S.J., Kumar, S., Owen, L.,
2005. Paleoseismic transect across the northern Great Basin. Journal of
Wesnousky, S.G., Kumar, S., Mohindra, R., Thakur, V.C., 1999. Holocene slip rate
of the Himalayan Frontal Thrust of India, Observations near Dehra Dun.
Tectonics 18, 967976.
Wesnousky, S.G., Scholz, C.H., Shimazaki, K., Matsuda, T., 1983. Earthquake
frequency distribution and the mechanics of faulting. Journal of Geophysical
Research 88(B11), 93319340.
Working Group on California Earthquake Probabilities, 2008. The uniform California
earthquake rupture forecast, version 2 (UCERF 2). U.S. Geological Survey Open
File Report 2007-1437: Also California Geological Survey Special Report 203.
104 pp. (http://pubs.usgs.gov/of/2007/1437/).
Yeats, R.S., Prentice, C.S. (Eds.), 1994. Special Section: Paleoseismology. Journal
of Geophysical Research, Solid Earth, 101(B3).
Yonekura, N., 1975. Quaternary tectonic movements in the outer arc on southwest
Japan with special reference to seismic crustal deformations. Bulletin of
Department of Geography, University of Tokyo 7, 1971.
Yoshikawa, T., Kaizuka, S., Ota, Y., 1981. The Landforms of Japan. University of
Tokyo Press, Tokyo.
Zehfuss, P.H., Bierman, P.R., Gillespie, A.R., Burke, R.M., Caffee, M.W., 2001. Slip
rates on the Fish Springs fault, Owens Valley, California, deduced from
cosmogenic 10Be and 26Al and soil development on fan surfaces. Geological
Zielke, O., Arrowsmith, J.R., Grant-Ludwig, L., Akc- iz, S.O., 2010. Slip in the 1857
and earlier large earthquakes along the Carrizo Plain, San Andreas Fault.
Science 327, 1119.
Biographical Sketch
Dr. Hisao Kondo is a research scientist at the Active Fault and Earthquake Research Center, Geological Survey of
Japan/National Institute of Industrial Science and Technology (AIST). Dr. Kondo earned his PhD at Department
of Geography, Hiroshima University, in 2004 and held positions as a postdoctoral fellow at the Active Fault
Research Center of the Geological Survey of Japan, and has been awarded a Japan Society for the Promotion of
Science Research Fellowship for Young Scientists. Dr. Kondos research focuses on the tectonic geomorphology,
paleoseismology, and Quaternary and earthquake geology in relation to active tectonics. In particular, Dr. Kondo
is interested in the repeatability of faulting behavior and the recurrence of large earthquakes along active fault
systems in Turkey and Japan. Dr. Kondo has also undertaken field research in Pakistan.
Paleoseismology
299
Prof. Owen received his PhD in geomorphology from the University of Leicester, UK, in 1988. Before joining the
University of Cincinnati in 2004, he held positions at the Hong Kong Baptist University, Royal Holloway
University of London, and the University of California Riverside. Prof. Owens research focuses on the Quaternary geology and geomorphology of tectonically active mountain belts and their forelands, particularly in the
HimalayanTibetan orogen and the Cordilleras of North and South America. He has also undertaken research in
other tectonically active regions, including the Red Sea margin in Yemen and the Atlas and Anti-Atlas Mountains
of Morocco. In 2011, Prof. Owen was awarded the Busk Medal from the Royal Geographic Society for his field
research in Quaternary history and geomorphology in tectonically active areas.
5.13 Glacially Influenced Tectonic Geomorphology: The Impact of the Glacial

Buzzsaw on Topography and Orogenic Systems
JA Spotila, Virginia Tech, Blacksburg, VA, USA
5.13.1
5.13.2
5.13.2.1
5.13.2.2
5.13.2.3
5.13.2.4
5.13.3
5.13.3.1
5.13.3.2
5.13.4
5.13.4.1
5.13.4.2
5.13.4.3
5.13.5
References
Introduction
Basics of Glacial Erosion
Erosional Processes
Glacial Erosion and Glacier Sliding
Rates of Glacial Erosion
Complexities and Exceptions
Glacial Erosion and Topography
Alpine Glacial Topography
Specific Models of Evolving Glacial Topography
Influence of Glaciers on Tectonics
The Glacial Buzzsaw Hypothesis
Glacial Erosion and ClimateTectonic Coupling
Orogen and Landscape Response to Glacial Climate Change
Discussions and Conclusions
300
300
301
302
302
304
304
304
306
308
308
311
312
313
314
Abstract
Glacial erosion is approximated as a simple linear function of basal sliding, but is actually dependent on a myriad of
spatially and temporally varying boundary conditions. Glacial erosion is more proficient than fluvial erosion, but both are
limited by the local rate of tectonic rock uplift. The unique characteristics of glacial topography can be linked to mechanics
of the erosion process and local conditions (e.g., selective linear erosion under polythermal ice conditions), as demonstrated in surface process models. Some kinematic aspects of glacial erosion are exclusive, including a focusing of erosion at
equilibrium-line altitude, aggressive headwall retreat, and relatively rapid topographic adjustment to climate change. As a
result of these characteristics, orogenic systems behave differently when dominated by glaciers. Under the right conditions,
rapid rock uplift and efficient glacial erosion limit topography, induce steady-state orogenic flux, and focus crustal strain,
thus imparting geodynamic significance to glacial climate. This hypothesis of a glacial buzzsaw has been verified both
topographically and exhumationally in a variety of settings worldwide, as well as in numerical models of orogenic evolution. Unraveling exceptions to this, including teflon peaks and glacial thresholds, is key to improving understanding of
the tectonic geomorphology of glaciated terrain.
5.13.1
Introduction
Glacial erosion is known to be very proficient and to have an

extreme influence on topography and tectonic systems (Hallet
et al., 1996; Brocklehurst and Whipple, 2002; Berger et al.,
2008a). As such, a comprehensive review of tectonic geomorphology would be incomplete without a dedicated discourse on the processes of glacial erosion and their effects.
This chapter reviews the basics of glacial erosion processes as a
foundation, given the important distinctions that these processes have relative to more familiar fluvial erosion. The
chapter implicitly considers glacial erosion to mean erosion by
Spotila, J.A., 2013. Glacially influenced tectonic geomorphology: the impact

of the glacial buzzsaw on topography and orogenic systems. In: Shroder, J.
300
temperate, alpine glaciers, and exclude erosion by polar glaciers and ice sheets unless otherwise stated. This review of
glacial erosion culminates with the physical basis for erosion
rules that are commonly used in surface process models and
highlights both the complexity and elegant simplicity of glacial erosion. Next, the chapter reviews how glacial erosion
influences topography, including examination of how glacial
erosion modifies pre-existing fluvial topography and the
timescales over which this occurs, as well as discusses whether
glaciers increase topographic relief. This leads to a review of
our understanding of specific elements of glacial topography,
such as longitudinal profiles, based on modeling. Finally, the
influence of glacial erosion on orogenic systems, based on
combined empirical and modeling-based approaches is explored. This includes the importance of glacial equilibriumline altitude (ELA), the glacial buzzsaw hypothesis, thresholds
in glacial erosion, and the partitioning of deformation and
reshaping of orogenic wedges.
http://dx.doi.org/10.1016/B978-0-12-374739-6.00093-2
Glacially Influenced Tectonic Geomorphology: The Impact of the Glacial Buzzsaw on Topography and Orogenic Systems
Erosional Processes
Bedrock erosion by glaciers is accomplished as a combination

of abrasion and quarrying. Abrasion, or scouring, occurs as
glaciers drag debris over the icebedrock interface. This is a
smoothing process, which removes rough edges and results in
streamlined bedrock forms, such as striations, grooves, and the
ice-flow-facing sides of roche moutonnees (cf. MacGregor
et al., 2009). Abrasion produces fine sediment by direct bed
erosion and the comminution of clasts that have become
entrained in the ice. The rate and efficiency of abrasion are
controlled by the basal sliding rate and the presence of
tools, or hard entrained debris (Hallet, 1979; Lliboutry, 1993).
Abrasion is efficient on slopes that face opposite the direction
of ice flow, but inefficient where ice-bed separation occurs
downstream of subglacial topographic features. Quarrying is
the glacial equivalent of fluvial plucking, and occurs as glaciers
fracture and remove large blocks from the bedrock floor.
This produces coarse sediment and increases bed roughness
(Hallet, 1996). The efficiency of quarrying depends on the
presence of rough edges or steps on the bed, basal water
pressure (Cohen et al., 2006), and the material properties of
the rock, particularly the spacing of discontinuities (Duhnforth
et al., 2010). Quarrying is more efficient downstream of
subglacial topographic features, and is in turn important for
generating new irregularities in the bed topography.
The relative importance of abrasion and quarrying depends
on local conditions, including the bedrock properties, subglacial
hydrology, topography, and temperature. Abrasion can have
nonlinear dependence on special conditions, such as the influx
of large debris to the glacier bed via the bergschrund or crevasses
or the production of sand grains via chemical and physical
weathering (Lliboutry, 1993). The efficacy of abrasion and
quarrying can vary over short distances beneath individual glaciers or over time at a given location, with timescales of variation
that can be decadal, seasonal, or even as short as days (Anderson
et al., 2006). Lliboutry (1993) suggested that quarrying should
dominate early during a glacial cycle, where there are numerous
rough edges and rock is fractured, and that abrasion should
gradually become more important as more intact bedrock is
exhumed. Although either process can dominate locally, they are
to some degree coupled, such as via the influence of quarrying
on abrasion by controlling the production of erosive tools
(Anderson et al., 2006). Despite local variation, however, there
are indications that quarrying is, on average, responsible for
more bed lowering than abrasion (Loso et al., 2004; Amundson
and Iverson, 2006). For example, Riihimaki et al. (2005) estimated that quarrying is responsible for at least 2/3 of erosion
beneath a small glacier in southern Alaska, based on the relative
yields of bedload and suspended load in glacial outwash.
A particularly important process that sets dynamic boundary conditions for glacial erosion is the subglacial hydrologic
system, which broadly includes variables of subglacial water
pressure, water storage, water discharge, flux of meltwater from
the surface, and subglacial channelization (cf. Fountain and
Walder, 1998; Clarke, 2005; Flowers, 2008; Kavanaugh, 2009).
Basal hydraulics are important for glacier sliding, which in turn
influences abrasion and quarrying. For example, supercooling
Erosion rate
discharge
5.13.2.1
Basics of Glacial Erosion
Sediment flux
5.13.2
301
Erosion
rate
Net basal
translation
(ice)
Water
discharge
Sediment
production
Sediment
evacuation
Sediment
storage
Spring
Sediment
discharge
Summer
Time
Fall
Figure 1 Conceptual relationships between glacier flow (net distance

of basal translation, or horizontal ice motion), erosion rate, water
discharge (at toe of glacier), and sediment budget (production,
discharge at toe of glacier, net evacuation and storage), based on
observations in the ablation region of the Bench Glacier in Alaska
(redrafted based on interpretations of Riihimaki et al., 2005). The net
ice motion is based on seasonal GPS observations of one mid-glacial
station (#2) in 2002, and was not included in the original conceptual
diagram. Notice the correlation between glacier sliding, erosion, and
water discharge, all of which peak in early summer. The water
discharge remains high throughout the melt season, although glacier
sliding and erosion both decrease. Sediment production is linked to
erosion, whereas sediment discharge lags behind sediment production
because it also depends on water discharge. Sediment evacuation
represents the net amount of sediment discharged. The rate of
sediment discharge eventually decreases at the peak of the melting
season, despite continued high meltwater discharge, because the
sediment production due to erosion has ceased and sediment storage
has been fully depleted.
and ice growth in overdeepenings can freeze a glacier to its bed,

thereby minimizing further erosion and grading the glacier
(Alley et al., 2003). Basal water pressure also controls the pore
pressure in bedrock cracks, which influences the ease of
quarrying (Cohen et al., 2006), as evident in the prevalence of
quarrying in cavities in the lee of subglacial topography where
fluids accumulate (Hallet, 1996). Subglacial hydrologic discharge and connectivity of outwash conduits are also critical for
removal of sediment produced by glacial erosion, which could
otherwise accumulate into a till layer that would effectively
shield bedrock from erosion (Hooke, 1991; Tulaczyk et al.,
1998; Alley et al., 2003; Riihimaki et al., 2005). Correlations
have been observed between subglacial water discharge and
both sediment yield and glacier sliding (Humphrey and Raymond, 1994; Bogen, 1996; Riihimaki et al., 2005) (Figure 1).
The importance of the subglacial hydrologic system, which is
largely fed by surface meltwater reaching the bed via crevasses
and moulins, is also apparent in seasonal variations in glacier
velocity, which tend to spike when subglacial fluids are most
abundant just before seasonal integration of subglacial
plumbing and major outwash events (Humphrey and Raymond, 1994). The lack of available meltwater in the accumulation area of a glacier may also contribute to the limitation in
glacial erosion above the ELA (Thomson et al., 2010). All these
factors make glacial erosion processes highly sensitive to subglacial hydrology, local climate, and other ambient conditions.
302
5.13.2.2
Glacial Erosion and Glacier Sliding
Sediment output (kg s1)
Given the complexity of glacial erosion processes, it may be

surprising that net glacial erosion can be effectively modeled as
a simple function of basal ice velocity. Erosion and sliding
velocity have been empirically correlated on individual glaciers
in Alaska (Figure 2), despite expected complexities including
sediment storage and a lag related to periodic sediment
flushing (Humphrey and Raymond, 1994; Riihimaki et al.,
2005). The erosionsliding correlation may be partly due to
the fact that both are influenced by the subglacial hydrologic
system; meltwater discharge affects both sediment yield and
the basal pore pressure that affects glacier flow (Riihimaki
et al., 2005). Nonetheless, observations imply that erosion and
sliding are linearly related, with B0.11 mm of erosion per
meter of glacier sliding. This general correlation is supported
by other field evidence. Brocklehurst and Whipple (2007)
showed that ice flux, by proxy of glacier area, influences how
valleys respond to changes in rock uplift rate in the Southern
Alps and Northwestern Himalaya. Amundson and Iverson
(2006) found that differential sliding velocity, by proxy of ice
balance velocity, explains the height of most hanging valleys in
mountainous western Canada. The hanging valleys appear to
result from the variation in erosion rate of glaciers of different
size and discharge. Thermochronometry also supports the link
between erosion and glacier sliding. Berger and Spotila (2008)
observed that exhumation is focused in a zone where the ELAs
of numerous glaciers intersect the windward flank of the St.
Elias Range, Alaska, consistent with observations in parts of the
Patagonian Andes (Thomson et al., 2010). Shuster et al. (2011)
also showed that thermochronometry and topography of
Fiordland, New Zealand, can be explained if glaciers fully reset
the landscape with erosion that scales specifically to glacier
sliding velocity (as opposed to ice discharge).
There are also theoretical reasons why erosion should scale
with glacier sliding. For example, ice velocity should govern the
flux of abrasive particles and the forces between particles on a
1000
100
10
10
100
1000
Average sliding speed ( 106 m s1)

Figure 2 Empirical correlation between sediment yield and average
basal sliding speed of the surging Variegated Glacier in Alaska. This
correlation is the widely cited basis for the commonly assumed
relationship between ice flux and erosion. Based on data of
Humphrey, N.F., Raymond, C.F., 1994. Hydrology, erosion, and
sediment production in a surging glacier: Variegated Glacier, Alaska,
198283. Journal of Glaciology 40, 539552.
glacier bed, thereby relating ice velocity to abrasion rate (Hallet

et al., 1979). Other factors that play a role may be of less
importance to counter the effect of ice flux on many temperate
glaciers. For example, ice thickness does not seem to be critical.
Abrasion rates are not dependent on normal stress exerted by a
glacier (Hallet et al., 1979). Likewise, given that ice flows in the
direction of steepest ice-surface gradient and that steeper slopes
translate to faster ice velocity, ice thickness tends to be relatively thin and of limited variation along temperate glaciers,
such that the basal sliding itself remains more important for
erosion (i.e., the thin-ice approximation) (Paterson, 1994;
Anderson et al., 2006). Note that the relationship of sliding
velocity and erosion also holds only for basal ice flux, not the
net motion of a glacier. For temperate glaciers, the fraction of
basal motion to surface or total motion can vary from nearly
zero (e.g., frozen-based glaciers) to 100% (Paterson, 1994).
The simple approximation that glacial erosion is linearly
related to basal velocity or unit ice flux has been validated in
numerous numerical and analytical models that have sought to
explain glacial topography with simulations of glacier flow and
dynamic ice mass balance (Harbor, 1992; Braun et al., 1999;
MacGregor et al., 2000; Tomkin and Braun, 2002; Tomkin and
Roe, 2007; Jamieson et al., 2008; Herman and Braun, 2008;
MacGregor et al., 2009). The constants in the erosion rules used
by these models vary, and some models incorporate additional
process-specific terms, such as the clast concentration at the
bed, clast comminution, and accumulation of till. Some numerical models also include terms for internal ice deformation,
dependent on variables including temperature and pore pressure, which filter out the glacier motion that does not contribute to erosion (Tomkin, 2007, 2009; Tomkin and Roe,
2007). Aside from these variations, each of these models uses a
linear relationship between erosion and sliding and is able to
accurately predict aspects of glacial topography, such as crossvalley or longitudinal profiles. Many of the basic characteristics
of glacial topography can be explained by attaching this simple
erosion rule to physics-based models of ice flow (Anderson
et al., 2006; Tomkin, 2009).
An important implication for the relationship of glacial
erosion and glacial sliding is that erosion should be greatest near
the ELA, where ice discharge is maximum (Paterson, 1994; Meigs
and Sauber, 2000). This has been shown in simple analytical
models of individual glacier profiles (e.g., Anderson et al., 2006)
and numerical models of entire glaciated orogens (Tomkin and
Roe, 2007) (Figure 3). In both cases, the concentration in ice
discharge and erosion at ELA are not sharp peaks, but rather
broad curves that attain a maximum at the ELA. Note that there
are exceptions to this result. Herman and Braun (2008) showed
that realistic topography and glacier networks can result in
concentrations of ice discharge that do not occur at the ELA, and
thereby predict concentrations of erosion elsewhere along the
glacier network. Nonetheless, this general phenomenon of erosion focusing at glacial ELA is the basis for the glacial buzzsaw
hypothesis, as discussed in Section 5.13.4.1.
5.13.2.3
Rates of Glacial Erosion
Long-standing questions about glacial erosion include how

fast it can proceed, how it compares to fluvial erosion, and
0.5
(a)
0.5
Dimensionless distance
5
Erosion rate
(mm yr1)
105
Glacier toe
0
ELA
3
Glacier toe
1
0
(b)
106
ELA
25
Distance (km)
50
Figure 3 Modeled concentration of ice flux and erosion near the ELA.
(a) Results of the analytical model of Anderson et al. (2006), showing
how dimensionless ice discharge peaks at the ELA. The accumulationarea ratio of 0.5 differs from typical values (B0.65), because this
simple model did not take into account down-glacier variations in
glacier width. Notice how the distribution of ice discharge is a broad
bell curve, rather than a sharp peak at ELA. (b) Results of the
numerical model of Tomkin and Roe (2007) for evolution of a glaciated
orogenic wedge (assuming an 81 taper). Erosion rate peaks at the ELA,
which in this case occurs closer to the glacier toe (AAR 0.6).
how it varies over time. Comparisons of topography have led

to a rough consensus that glacial erosion is more effective than
fluvial erosion at sculpting landscapes (see Section 5.13.3.1).
But how do rates of these processes compare?
A seminal paper promoting glaciers as more efficient eroders than rivers was Hallet et al. (1996). This study summarized data from around the world on glacial and fluvial erosion
rates, based on short-term (decades to centuries) sediment
yields. Glacial erosion rates were found to span four orders of
magnitude (0.01100 mm yr1, spanning polar to wet temperate glaciers) and scale with the extent of glacier coverage,
but the most rapid rates of erosion observed globally occurred
in areas of temperate modern glacial coverage (Figure 4). As a
result, this paper has emerged to be a favorite citation for the
quasiparadigm that glacial erosion is more proficient than
fluvial erosion (e.g., this paper had been cited nearly 300 times
in peer reviewed journals as of June, 2010, according to the
CSA Web of Science Citation Index). A similar range and
pattern of short-term erosion rates was also observed in a
more recent global review (Delmas et al., 2009). However,
there are limitations to using short-term sediment yields to
infer erosional proficiency. Sediment yield from glaciers not
only include primary bedrock erosion but also expulsion
of stored sediment, making short-term sediment yield a dangerous proxy that is potentially time-dependent (Harbor and
Warburton, 1993). Koppes and Hallet (2006) revised estimates of erosion rate to be lower for some of the same basins
used by Hallet et al. (1996), on the basis of expulsion of
stored sediment at a time of glacial retreat.
Another compilation that assessed erosion rates over different timescales suggests that more overlap exists between
fluvial and glacial settings. Koppes and Montgomery (2009)
Sediment yield, ton km2 yr 1
Dimensionless
discharge
303
104
b
103
c
d
102
101
100
101
102
103
104
105
106
Area, km2
Figure 4 Summary of the results of Hallet et al. (1996) that compare
short-term (B100 yr) unit sediment yield for glaciated and nonglaciated basins worldwide. Field (a) represents the area of high
sediment yield for temperate glaciated basins, including both tectonically
active and inactive mountain ranges. Other fields represent basins that
are not heavily glaciated: (b) partially glaciated Himalayan basins, (c)
fluvial basins with relief greater than 3 km from southern Asia, (d)
fluvial basins with 13 km relief. Notice how the glaciated basins have
the highest sediment yield of any basin compared. This is the oft-cited
comparison that glaciers are more proficient at erosion than rivers.
study found that high rates of short-term erosion (110 mm

yr1) occur in both fluvial and glacial basins, noting in particular the very high sediment yield of rivers in active tectonic
regions including Taiwan, the Himalaya, and the Pacific
Northwest. They observed that glacial erosion rates decrease as
the interval of measurement lengthens (see also Spotila et al.,
2004). Other studies have shown that glacial erosion rates
peak during periods of glacier retreat, melting, and accelerated
ice flow (Koppes et al., 2009; Delmas et al., 2009; Riihimaki
et al., 2005). Glacial erosion may be unsteady through time,
and glacial systems may even undergo periods of relatively
limited erosion, such as during advances or glacial maxima. In
contrast, fluvial erosion appears steadier through time, suggesting that long-term, mean erosion by glaciers may not exceed that by rivers. Indeed, the fastest long-term rates of
exhumation observed globally, B0.51 cm yr1, occur in both
glacial and fluvial settings (e.g., Burbank et al., 1996; Spotila
et al., 2001; Ehlers et al., 2006; Berger and Spotila, 2008).
However, the Koppes and Montgomery (2009) study also
has limitations, including a reliance on relatively limited
geographic areas for the highest fluvial erosion (e.g., Taiwan
and disequilibrium volcanic areas) and the unknown effect of
landuse on short-term sediment yield in fluvial areas that are
likely partially cultivated. The claim that fluvial rates are timeinvariant is also based on a comparison of spatially averaged
short-term rates (i.e., sediment yields) to point-measurements
304
of long-term rates (i.e., maximum exhumation rates based on

thermochonometry), thus weakening the comparison.
Thus, although there is a loose consensus that glaciers are
generally more effective at erosion than rivers, there are exceptions to this and it appears that, given the right conditions,
either system is capable of extreme erosion rates (up to 1 cm
yr1) that keep pace with rock uplift at rates determined by
tectonics. The variation in glacial erosion rates over time, and
the fact that no exclusive, controlled comparison of fluvial and
glacial rates can be made, also make it difficult to assess which
agent of erosion is more powerful. However, the fact that
glaciers exist means that there are places in the world where
fluvial erosion rates were not adequate to keep pace with rock
uplift on their own. Likewise, the persistence of glacial topography long after deglaciation and the rapid onset of glacial
topography in a glaciated fluvial landscape both suggest that
glaciers are on average more proficient (Anderson et al., 2006).
5.13.2.4
Complexities and Exceptions
As alluded to above, there are numerous complications to

glacial erosion, and the rule that glacial erosion is a linear
function of sliding velocity should be viewed with caution and
evaluated on a case-by-case basis. Unlike fluvial erosive power,
which increases with slope and downstream with increasing
discharge in a relatively predictable fashion, glacial erosion is
more subject to local conditions that can cause spatial and
temporal variation. As discussed, subglacial hydrology, including the water pressure, volume, temperature, and the integration and connectivity of subglacial plumbing, is one
complicating factor that can play a major role in both glacier
sliding, individual erosional processes, and the transport of
sediment produced by glacial erosion that can otherwise accumulate at the bed and protect bedrock from erosion (Humphrey and Raymond, 1994; Bogen, 1996; Alley et al., 2003;
Riihimaki et al., 2005). Glacial erosion is also particularly
dependent on bedrock properties, including the importance of
hard grains and clasts needed for abrasion and the influence of
fracture spacing on quarrying (Lliboutry, 1993; Augustinus,
1992; Glasser and Ghiglione, 2009; Duhnforth et al., 2010).
Patterns of ice flow, including the confluence of glacial valleys
that can result in irregular variations in ice discharge along
glacial profiles is another potential complication to glacial
erosion that has been illustrated in models that take into account two- or three-dimensional glacial networks (Herman and
Braun, 2008; Jamieson et al., 2008; MacGregor et al., 2009).
Another major factor is the ratio of total glacial motion that
is accommodated by basal sliding versus ice deformation (e.g.,
Paterson, 1994). Glacial erosion nearly stops when the basal ice
of a glacier becomes frozen to the bed, such that subglacial
meltwater becomes absent and bedrock no longer feels the
motion of the glacier. Extremely low rates of glacial erosion
have been well documented for polar glaciers and continental
ice sheets, which are largely frozen-based. For example, much of
Sweden exhibits preserved relict preglacial landforms that have
experienced minimal erosion over multiple glacial cycles based
on cosmogenic dating (Fabel et al., 2002; Hattestrand and
Stroeven, 2002; Stroeven et al., 2006). Similar observations of
minimal bedrock erosion have been made for the edge of the
Laurentide ice sheet in North America (Colgan et al., 2002).

Thomson et al. (2010) showed that frozen-based glaciers in the
Patagonian Andes have likely protected an active orogen from
erosion, leading to mean elevations that extend above ELA.
These observations do not mean that frozen-based glaciers
cannot erode at all, however. Studies of cold-based glaciers in
the Allan Hills, Antarctica, show that some glacial erosion is
possible, such as where the bed has significant topography and
prominent outcroppings may be plucked, or where cavities on
lee sides of features result in ice-bed separation and local
abrasion (Atkins et al., 2002; Davies et al., 2009).
The influence of ice-bed temperature on glacial erosion can
also lead to complex polythermal erosive conditions, in which
local basal melting may result in localized erosion. This can
lead to relief production, where wet-based glaciers promote
erosion of valleys, whereas frozen-based glaciers at higher,
colder altitudes inhibit summit erosion (i.e., selective linear
erosion; Sugden, 1978; Tomkin and Braun, 2002; Jamieson
et al., 2008). This has been observed and modeled in the
Torngat Mountains of northeastern Canada based on cosmogenic dating (Staiger et al., 2005). In this study, interfluves
covered with felsenmeer (frost-shattered bedrock and boulders)
experienced some (o2 m Ma1) glacial erosion proportional
to glacier sliding, but the coefficient in the erosion rule was
several orders of magnitude lower than for wet-based conditions (Staiger et al., 2005). Briner et al. (2006) similarly
showed the deep fjord valleys on Baffin Island had been produced by polythermal erosive conditions. Glaciers covering
interfjord plateaus were thin and frozen-based, leading to
minimal erosion based on cosmogenic dating. These conditions can lead to spontaneous creation of fjords at the edge of
continental ice sheets, where topographic steering of ice into
rapid outlet ice streams results in thicker, warmer-based ice, and
incision of deep canyons into a continental margin (Briner
et al., 2006; Kessler et al., 2008). The gentler gradient of valleys
relative to interfluves contributes to thicker, warmer-based ice at
lower elevations, an effect that can also occur along individual
glaciers, such as at overdeepenings (Jamieson et al., 2008).
The impact of polythermal ice conditions on glacial erosion illustrates the potentially nonlinear nature of glacial
erosion and the importance of variables that control the
thermal regime of the icesubstrate interface, including geothermal heat flux, surface temperature, ice thickness, ice velocity, topographic slope, bed morphology, strain heat release
from ice deformation, hydrology, preexisting permafrost conditions, and glacier networks (Hattestrand and Stroeven, 2002;
Alley et al., 2003; Staiger et al., 2005; Herman and Braun,
2008; Jamieson et al., 2008). The thermal state of the bed may
also change with time, possibly explaining observations of
cyclic periods of active erosion with the advance and retreat of
glaciers (Delmas et al., 2009; Koppes and Montgomery, 2009).
As a result, the relationship between glacier sliding and erosion should not be viewed as an absolute rule.
5.13.3
5.13.3.1
Glacial Erosion and Topography

Alpine Glacial Topography
The features of typical alpine glacial topography have been

widely described in previous studies (e.g., Sugden and John,
Figure 5 Photograph of a typical alpine basin that has been shaped

by glacial erosion in an tectonically active mountain range (Garnet
Canyon, Teton Range, Wyoming; photo by J. Spotila). The
deglaciated valley shows the characteristic U-shape. The longitudinal
profile is approximately linear in the lower reaches but steepens near
the headwall, where numerous irregular steps occur. Notice the steep
valley walls and oversteepened peaks situated in the interfluves. The
peak on the right is the Grand Teton, which is a classic example of a
teflon peak. From Foster, D., Brocklehurst, S.H., Gawthorpe, R.L.,
2010. Glacial-topographic interactions in the Teton Range, Wyoming.
Journal of Geophysical Research 115(F01007), 20.
1976; Brocklehurst and Whipple, 2002). The classical form of

a temperate, glaciated mountain range includes cirques,
U-shaped cross sections, hanging valleys, low slope longitudinal profiles that are punctuated by steep steps and overdeepenings, high headwalls that rise to aretes and horns and
are flanked by truncated spurs, and hillslopes that are oversteepened beyond the angle of repose (Figure 5). Anderson
et al. (2006) pointed out that glacial valley networks tend to
have bulbous ends, and are not dendritic like fluvial networks,
and that glacial topography tends to be knobby and bumpy
above the ELA, but becomes smoother further down glacial
valleys; a transition that might relate to basal ice conditions.
Glaciated topography is clearly distinguishable from the
typical V-shaped valleys and convex hillslopes of a fluvial
landscape (Kirkbride and Matthews, 1997). Recent studies
have quantified the magnitude of difference between glacial
and fluvial topographies, using parameters such as geophysical
relief, basin hypsometry, and the degree of curvature of
U-shaped valleys (Kirkbride and Matthews, 1997; Brocklehurst
and Whipple, 2002, 2004; Brook et al., 2006) (Figure 6). For
example, heavily glaciated topography tends to have extensive
area at low elevation (relative to the local maximum), and
hence low hypsometric integral, due to focusing of erosion
near ELA (Brocklehurst and Whipple, 2004). Basin hypsometry is thus a particularly useful metric for deducing the scale
of former glacier coverage and its impact on topography.
Studies like this illustrate that there is fundamental information about glacial erosion processes that can be learned
from just topography itself.
Insight into how glacial erosion affects a landscape, and
into whether glacial or fluvial erosion is more proficient in a
305
given mountain setting, can be gained by investigating how

glacial erosion modifies a formerly fluvial topography. This
type of study is well suited for spacetime substitution, by
comparing modern drainages dominated by either process in
a given location. Based on studies in the Olympic Mountains,
Washington (Montgomery, 2002), the Sierra Nevada, California (Brocklehurst and Whipple, 2002), the Bitterroot
Range, Montana (Naylor and Gabet, 2007), Idaho (Amerson
et al., 2007), the French Alps (van der Beek and Bourbon,
2008), and the Southern Alps (Brook et al., 2006), several
generalities can be drawn. Glacial erosion generally appears to
remove a greater volume of rock from topography than fluvial
erosion, as documented by increasing valley width, ridgevalley relief, valley cross-sectional area and the degree of
curvature of the valley floor, and basin geophysical relief with
increasing degree of glaciation (Figure 7). This suggests that
glaciers can be more proficient than fluvial erosion and that
glaciation can generate relief in mountain landscapes. Glaciers
also respond differently to increases in rock uplift. Although
fluvial profiles steepen to accommodate faster erosion as
rock uplift increases, glacial profiles steepen far less, suggesting
that other mechanisms must be at work to enable more
rapid erosion that keeps pace with increasing rock uplift
(Brocklehurst and Whipple, 2007).
The timescale at which glaciers will modify a pre-existing
fluvial topography is another interesting aspect of the topographic effect of glacial erosion. Space for time substitution
and comparison of fluvial and glacial topography coupled with
estimates of glacial occupancy time based on oxygen isotope
records have permitted estimates of the time needed for resetting fluvial topography in the Southern Alps. Results imply
that recognizable glacial topography can develop in as little as
200 ka after the onset of glaciation (Kirkbride and Matthews,
1997), and that a fully glacial topography, as indicated by the
curvature of cross-valley profiles, can develop in 400600 ka
(Brook et al., 2006, 2008). Numerical models of glacial
topography have produced comparable estimates for the
timescale of topographic resetting. Harbor (1992) showed
that V-shaped valleys can be converted to U-shaped by
glacial erosion in as little as 100 ka, but do not attain topographic steady-state until B500 ka. Other models similarly
suggest that only one or two glacial cycles are needed to sculpt
longitudinal profiles or topography that are clearly distinguishable as glacial (MacGregor et al., 2000; Jamieson et al.,
2008). In contrast, the time required for fluvial erosion to recover from glaciated topography can be longer (Hobley et al.,
2010). The fact that there are few transitional landscapes and
that areas that were once glaciated still exhibit glacial topography, whereas fluvial landscapes do not appear to be preserved once glaciated, implies that the timescale of topographic
resetting by glacial erosion is shorter than for fluvial erosion,
thus reinforcing the idea that glaciers are more proficient
agents of erosion (Brocklehurst and Whipple, 2002, 2004).
The observation that transition to a glacial topography
corresponds to an increase in relief suggests that relief increased during the late Cenozoic due to glaciation, consistent
with the hypothesis that climate change produced isostatic
peak uplift in mountain ranges worldwide (Molnar and
England, 1990). However, the scale of relief production due to
glacial erosion may not be adequate to generate significant
306
Relief
Fluvial
Glacial
800 m
400 m
2 km
0m
Relief
1.5
Modern ELA
1.0
LGM ELA
400 m
Pixels ( 1000)
800 m
0.5
2 km
0m
1000
2000
Elevation (m)
Figure 6 Typical hypsometry of glacial basins. The dark shaded images represent the distribution of sub-ridgeline relief for typical glacial and
fluvial basins of the Sierra Nevada. Notice how the glacial basin has a bulls-eye pattern of relief that is focused closer to the basin outlet, and the
basin itself is more equant in dimension. The diagram in the lower right shows the distribution of elevation for a glaciated basin in an active tectonic
region (Southern Alps; Brocklehurst and Whipple, 2007). Note how the majority of the land area lies between modern and paleo-ELA, whereas
nearly all the land area lies below the modern ELA, consistent with elements of the glacial buzzsaw hypothesis. This differs from the hyspometry of
fluvial basins, which commonly display tails, in which a very small fraction of the landscape continues to high elevations that far exceed the mean.
Adapted from Brocklehurst, S.H., Whipple, K.X., 2002. Glacial erosion and relief production in the Eastern Sierra Nevada, California. Geomorphology
42, 124; and Brocklehurst, S.H., Whipple, K.X., 2007. Response of glacial landscapes to spatial variations in rock uplift rate. Journal of Geophysical
Research 112(F02035), 18 p.
1500
Relief (m)
1000
500
0.1
100
10
Drainage area (km2)
1000
Figure 7 Valley relief vs. drainage area for glacial and non-glacial
basins of the Olympic Mountains. Adapted from Montgomery, D.R.,
2002. Valley formation by fluvial and glacial erosion. Geology 30,
10471050. Black triangles are glacial basins and open triangles are
fluvial basins. The consistently higher relief of glaciated basins
suggests that, at least in this case, glaciers are more proficient at
removing mass and sculpting topography than rivers.
(i.e., hundreds of meters) peak uplift. Several studies have

noted that the increase in relief due to glacial erosion is
size-dependent, and is only significant for the largest alpine
glaciers (Brocklehurst and Whipple, 2002; Montgomery, 2002;

Amerson et al., 2007; Brocklehurst et al., 2008). More common
smaller glaciers may thus not be as efficient at producing relief
and causing isostatic peak uplift. Whipple et al. (1999) also
suggested that relief production by glacial erosion is limited by
the scale of ice buttressing of rock slopes and the height of
hanging valleys, both of which should scale with the thickness
of mountain glaciers, which is typically only a few hundred
meters. Brocklehurst et al. (2008) confirmed this by finding a
correlation between ice thickness at the last glacial maximum
and the magnitude of relief production (for different measures
of relief) for ranges in the western US (Figure 8). The magnitude of relief production under a glacial climate also depends
on the preservation of interfluves, which is dependent on both
glacial and bedrock conditions as well as the proficiency of
weathering (see discussion of teflon peaks below). If erosion
and weathering of peaks and ridges are enhanced by climate
cooling, the magnitude of relief production should be limited
(Whipple et al., 1999; Brocklehurst and Whipple, 2002).
5.13.3.2
Specific Models of Evolving Glacial Topography
An effective validation of understanding of glacial erosion

processes is the successful modeling of evolving glacial
topography using simple to comprehensive, physically based
numerical and analytical models. Considerable recent progress in this area has significantly improved and validated
Drainage basin relief (m)
1600
1200
800
400
100
200
LGM ELA ice thickness (m)

Figure 8 Empirical correlation of net drainage basin relief and glacier
thickness at the last glacial maximum (LGM), based on data from the
western US. Adapted from Brocklehurst, S.H., Whipple, K.X., Foster,
D., 2008. Ice thickness and topographic relief in glaciated landscapes
of the western US. Geomorphology 97, 3551. This correlation implies
that any relief production associated with glaciation of former fluvial
valleys would also scale with ice thickness, consistent with the
prediction of Whipple et al. (1999) that glaciation should produce only
minimal relief production and isostatic uplift of peaks.
understanding of numerous aspects of glacial erosion and the

resulting characteristic topography.
Numerical models have been successful at explaining
features of glacial longitudinal profiles and why they fundamentally differ from their fluvial counterparts. Glacial longitudinal profiles exhibit lower concavity than fluvial profiles and
contain extensive flat reaches that are stepped, contain sediment and water-filled overdeepenings, and tend to display
junctions with tributaries that are mismatched in elevation (i.e.,
hanging valleys) (Oerlemans, 1984; Whipple et al., 1999;
Brocklehurst and Whipple, 2002; Anderson et al., 2006). The
comprehensive one-dimensional numerical models of MacGregor et al. (2000) simulated the evolution of long profiles,
starting with a pre-existing fluvial profile and modifying it with
glacial erosion. An ice mass-balance flow model coupled with
an erosion rule that scales with ice discharge produced realisticlooking glacial long profiles, which exhibited concavity in
upper reaches and were flatter in lower elevations. These
models also showed that overdeepenings are produced by excess ice flux at tributary junctions and that glacial erosion is
generally more effective than fluvial erosion at modifying long
profiles. Anderson et al. (2006) conducted similar modeling,
which showed that the peak in erosion near ELA, associated
with the peak in unit ice discharge, produces the flat profiles
that steepen only above the ELA. Likewise, steps and overdeepenings resulted from excess discharge, whereas hanging
valleys resulted from the lower discharge of tributaries. However, their models could not reproduce realistic accumulationarea ratios (AAR B0.65), unless they implicitly assumed a
typical along-profile variation in glacial valley width. This
suggests that features of glacial long-profiles are dependent on
not only unit ice discharge and basal ice velocity, but also the
processes that control the cross-profile shape of glacial valleys.
Glacial cross-profiles have been successfully modeled as
well. Harbor (1992) showed that the classical U-shape to glacial
valleys can be simulated by ice flow and an erosion rule
dependent on basal velocity. The key to producing this shape
307
are lateral variations in sliding velocity, which result from the

profile shape itself. Valleys initially respond to erosion by vertical incision, but this gives way to more rapid erosion at valley
margins due to the high erosion rate on steep slope angles of
valley walls. Harbor (1992) showed that U-shaped valleys tend
to adjust to the maximum ice discharge (i.e., modern glaciers
are undersized to the valleys) and develop rather quickly on
pre-existing fluvial profiles, as discussed above and corroborated by other studies (Brook et al., 2006). This, along with the
models of long profiles, show that there are feedbacks between
deepening and widening processes that combine to result in the
characteristic glacial valley form.
Another interesting aspect of glacial valley development is
the lengthening of profiles by active erosion at the head (i.e.,
headwall and cirque migration). Empirical studies have shown
that glacial headwall migration can outpace vertical glacial
erosion by several times, increases with the scale of glaciation,
and can outpace that of opposing fluvial valleys (Brocklehurst
and Whipple, 2002; Oskin and Burbank, 2005; Naylor and
Gabet, 2007; Shuster et al., 2011). These studies suggest that
headwall retreat is a fundamental aspect of glacial modification of valley topography (Figure 9). Some studies suggest
that this is the primary form of relief production in valleys due
to onset of glaciation (Whipple et al., 1999; Brocklehurst
and Whipple, 2002). Given that headwalls are situated above
the ELA, it is not immediately clear how glacial erosion can
accomplish rapid headwall retreat. MacGregor et al. (2009)
performed comprehensive numerical simulations of headwall
backwearing to explore how this occurs. Their model incorporated terms for blowing and avalanching snow, freezethaw weathering of the wall, and excessive influx of abrasion
tools at glacier heads. Even without taking into account rotational ice flow in cirques or complex hydrology that might
result in polythermal ice conditions in cirques (e.g., Hooke,
1991), their models successfully showed that headwall-specific
processes can produce backwearing (Figure 10). Cirques did
form at the heads of long profiles and caused rapid retreat of
headwalls, primarily due to the large influx of snow from
adjacent slopes, focusing of precipitation by the evolving
topography, effective quarrying of fractured bedrock, rapid ice
sliding that will occur on the steep cirque bed, abundant
physical weathering of exposed bedrock and the resulting
availability of erosion tools (MacGregor et al., 2009). These
results show that cirque formation and headwall retreat are
fundamental aspects of glacial erosion and can be simulated
with physically based models of ice flow, precipitation, and
glacial erosion. Headwall retreat will not be effective under all
conditions, however, and Brocklehurst and Whipple (2007)
showed that it can be outpaced by vertical glacial incision
where rock uplift rates are rapid.
Comprehensive surface process models, which take into
account multiple erosional processes, tectonic uplift, and climate, have also been proven effective at simulating the evolution of glacial topography. Braun et al. (1999) simulated
fluvial, hillslope, and glacial erosion under various conditions,
and showed that glacial erosion does excavate a larger volume
of rock than fluvial conditions and that the volume of ice that
can be supported by the landscape is dependent on both
elevation and the glacial topography itself. Their results also
showed that glacial erosion peaks near the end of glacial
308
maxima and that erosion in general is most rapid under

conditions of disequilibrium, such as when climate is changing. Their model also showed that glacial erosion can reach a
steady state, where the rates do not increase despite increase in
ice volume. Tomkin and Braun (2002) used a similar comprehensive modeling approach to test specifically for the
magnitude of relief generated by glacial conditions. Their results showed that relief production depends on climate; relief
increases when glaciers are cold-based at high elevations, but
Ridge
Glacier
decreases under other conditions. Other models have successfully shown that realistic two- and three-dimensional
glacial topography can be generated from simple glacial erosion rules and basic ice dynamics (Tomkin, 2009). When these
models begin with real topography, they show that glacial
erosion can actually be quite complex, spatially variable, and
dynamic; unlikely to develop a steady-state (Herman and
Braun, 2008). The power of these cutting-edge surface process
models is that they take into account so many processes and
conditions, as well as the coupling between them. For example, the MacGregor et al. (2009) model takes into account
clast production, comminution, sediment transport, flexure
and uplift, physical weathering, hydrology, climate variations,
orographic precipitation based on the evolving topography,
and, like all models, ice dynamics and glacial erosion.
Original
surface
Valley
5.13.4
5.13.4.1
Influence of Glaciers on Tectonics

The Glacial Buzzsaw Hypothesis
The observation that glacial erosion is highly proficient and

characteristically more effective than fluvial erosion has led to
the development of the idea that glaciers are capable of
eroding as fast as topography can be rebuilt via rock uplift. In
the simplest terms, the glacial buzzsaw hypothesis posits that
glacial erosion is able to limit topography under appropriate
conditions (where topography generically means elevated
land area, or mountain growth, and not necessarily a specific
topographic metric, such as mean elevation or relief). Obviously, this would have significant implications for landscape
evolution and orogenic systems if valid, and as a result
this idea has been the focus of intense study over the past
few years.
Headward erosion
Figure 9 Conceptual model of glacial landscape evolution in an

actively uplifting mountain range that is dominated by headwall
retreat (redrafted from Oskin and Burbank, 2005). The three-step
model shows sequential uplift of a range. The original surface is
denoted (dashed line), as inferred for the Kyrgyz Range based on a
preserved unconformity. Notice how the onset of glaciation (i.e.,
when the topography reaches the ELA) results in limit to the height
of the range. As more topography pushes above the ELA, glacial
cirques retreat by headwall erosion, pushing the divide far away from
the range crest (shown by the short dashed lines).
3000
Elevation (m)
Initial profile
2200
Final
glacier
surface
Final profile
1400
0
10
20
Distance (km)
30
40
Figure 10 Results of a numerical simulation of longitudinal valley evolution of a glaciated basin. Adapted from MacGregor, K.R., Anderson,
R.A., Waddington, E.D., 2009. Numerical modeling of glacial erosion and headwall processes in alpine valleys. Geomorphology 103, 189204.
Note how glacial erosion results in lowering of an initial profile (dotted line), resulting in a somewhat linear glacial profile at lower reaches and
an oversteepened profile at the glacier headwall. This oversteepening is what enables cirque retreat. Also note how the final glacier surface itself
is small; this is because glacial erosion has outpaced rock uplift and has thus been self-limiting. Model based on variable-temperature simulation
and assumed erosion to be linear with glacier sliding.
The buzzsaw hypothesis was developed by Brozovic et al.

(1997), on the basis of topographic trends in the Northwest
Himalaya and Karakoram. (Note that the label of buzzsaw
does not actually occur in their paper, but was linked to this
study via an abstract by the same authors.) Hypsometry for
these actively uplifting areas showed peaks in elevation distribution and minima in slope distributions at snowlines, or
the mean Quaternary ELA (Figure 11). The mean topography
paralleled a rise in ELA across the ranges, but did not correlate
to significant spatial gradients in tectonic rock uplift. Brozovic
et al. (1997) interpreted this to be the result of glacial and
periglacial processes placing an upper limit on altitude and
relief, and suggested that this resulted from the proficiency of
ELA
0.06
75%
30
0.04
50%
25%
Frequency
Slope (degrees)
50
0.02
10
4
6
Elevation (km)
Figure 11 Topography of the northwest Himalaya that formed the

empirical basis for the glacial buzzsaw hypothesis. Adapted from
Brozovic, N., Burbank, D.W., Meigs, A.J., 1997. Climatic limits on
landscape development in the northwestern Himalaya. Science 276,
571574. The frequency of elevation for their entire study area is
shown by the solid dark line. Distributions of slope, for the 25th,
50th, and 75th percentiles, are shown in the dashed lines. Notice
how elevation peaks near the glacial ELA, which also corresponds to
minima in all slope percentiles. Very little topography extends above
the ELA, all of which has very high slope.
309
glacial erosion and its tendency to peak at ELA. Their idea was
that glaciers concentrate erosion at ELA and keep pace with
tectonic uplift, preventing the majority of topography from
rising above mean snowlines. If too much topography rises
above ELA, more glaciers form, and thus erosion accelerates.
Similarly, a negative feedback loop exists, by which too much
glacial erosion undercuts a glacier, resulting in less-erosive
fluvial conditions that would in turn allow topography to rise
to the ELA. Evidence that glaciers undercut themselves in such
a self-defeating way includes the outcomes of numerical
simulations (MacGregor et al., 2000; Egholm et al., 2009) and
empirical observations of reduction in glacier size despite
climate change fostering glaciation in the Patagonian Andes
(Kaplan et al., 2009). The buzzsaw model also holds that
glacial topography itself reinforces the buzzsaw, in that isolated peaks (i.e., a few percent of the landscape), or topographic lightening rods, would rise high above ELA and
focus precipitation and avalanching snow back onto glacial
valleys, keeping the vast majority of the landscape near the
mean ELA. Glacial erosion may also inhibit fluvial erosion
downstream, by overloading rivers with sediment (Whipple
et al., 1999; Korup and Montgomery, 2008).
The topographic trends that inspired this concept had
been observed earlier, such as by Porter (1981). Since the
buzzsaw hypothesis was posed, other studies have recognized
similar topographic trends in other locations. These include
the Cascade Range, Washington (Mitchell and Montgomery,
2006), where mean swath topography clearly parallels the
rise in modern glaciers and paleo-ELA across the range
(Figure 12), but does not reflect a ten-fold variation in rock
uplift rate. Other locations where this has been observed include the Andes (Montgomery et al., 2001; Thomson, 2002),
the northern Basin and Range, where the glaciers are actually
quite small (Foster et al., 2008), northwest Tibet (Seong et al.,
2009), and the European Alps (Anders et al., 2010). Anders
et al. (2010) added the idea that cirques limit the elevation of
adjacent peaks, such that even peaks that rise above the ELA
will still be limited and show concordance to the ELA. Egholm
et al. (2009) capped this all with a global topographic
W
3
Modern glaciers
Elevation (m)
Cascade range
2
Cirque outlets
40
80
Distance (km)
120
160
Figure 12 Correlation of mean swath topography and glacial ELA in the Cascade Range, Washington, consistent with the predictions of the glacial
buzzsaw hypothesis. Adapted from Mitchell, S.G., Montgomery, D.R., 2006. Influence of a glacial buzzsaw on the height and morphology of the
Cascade Range in central Washington State, USA. Quaternary Research 65, 96107. The swath topography (their swath #2) runs from west to east
across the range. The elevation of modern glaciers (circles) and deglaciated cirque outlets (triangles) are shown, with gray bars representing the
regressions of both trends. Notice how the topography parallels both of these trends, although is situated closer to the elevation of modern glaciers.
This implies glaciers limit topography, and that the mean topography only extends a fixed height above the local LGM ELA.
310
analysis, which showed that theres almost no surface area on

the continents that rises above modern snowlines, regardless
of tectonic environment. They showed that hypsometric
maxima parallel ELA globally, although this is slightly different from as defined in the original buzzsaw hypothesis. Rather
than strictly limiting topography to ELA, the Egholm et al.
(2009) analysis identified clustering of topography at ELA. The
global analysis also revealed that peaks are generally less than
1.5 km above the modern snowlines, consistent with Anders
et al. (2010) finding in the Alps.
In addition to meeting the general topographic predictions, there is additional supporting evidence for the buzzsaw
hypothesis. Several studies have shown that sharp gradients in
tectonic rock uplift appear to have little effect on regional
topography in glacially dominated areas, where topographic
steady-states have been invoked, such as in the Patagonian
Andes (Thomson, 2002; Thomson et al., 2010) and in the
St. Elias orogen (Meigs and Sauber, 2000; Spotila et al., 2004;
Berger and Spotila, 2008). Note that similar observations have
also been made in nonglaciated areas, however (Spotila et al.,
2001). Other studies have also found that the influence of
ELAs position in a landscape is evident even in locations
which do not meet the strict requirement that ELA limits all
topography, such as in the Sierra Nevada, California (Brocklehurst and Whipple, 2004). Models have also confirmed that
glaciers may be able to keep pace with rock uplift under certain conditions (e.g., Egholm et al., 2009). For example, the
numerical surface process models of Tomkin (2007, 2009)
showed that there is climate-dependent elevation lowering,
similar to that predicted by the buzzsaw hypothesis.
There are exceptions and limitations to the buzzsaw hypothesis, however. First, it is important to point out that not
all active orogens reach glaciated elevations, indicating that
fluvial erosion can be adequate to erode topography before
the invocation of the glacial buzzsaw. Another potential
problem is circularity, given that the ELA itself is naturally
influenced by local topography and precipitation, so a correlation between ELA and topography could be the result of a
different causality (strem, 1972; Porter, 1977; Meigs and
Sauber, 2000). There are also some mountain ranges that do
not fit with the predictions of the buzzsaw hypothesis, such as
the Southern Alps, where elevation, relief, uplift rate all correlate with distance from the Alpine fault, but not the ELA
(Kirkbride and Mathews, 1997; Brocklehurst and Whipple,

2007) (Figure 13). Others have suggested that vigorous glacial
erosion can actually increase relief in active orogens, rather
than limit it (Brook et al., 2008). Other possible flaws with the
buzzsaw hypothesis include potential limits to glacial erosion,
as have been observed in comprehensive surface process
models (Braun et al., 1999) and proposed based on the
mechanics of specific processes (Lliboutry, 1993; Alley et al.,
2003; Staiger et al., 2005; Jamieson et al., 2008).
Another exception to the idea that glaciers limit topography, which was actually permitted in the original formulation of the hypothesis (Brozovic et al., 1997), is that isolated
peaks may rise high above ELA. For example, Meigs and
Sauber (2000) showed that despite a mean elevation of
1225 m above sea level (asl) in the St. Elias range, Alaska,
about half of the landscape rises above ELA, and a tiny fraction
of the landscape attained heights as great as 6050 m asl;
much higher than ELA. This has led to the development of
the idea of teflon peaks, which are conceptually similar to
the topographic lightening rods in the original Brozovic
et al. (1997) model. Teflon peaks, as proposed by Anderson
(2005) and developed further by Foster et al. (2008), rise
above the snowline because they are shielded from glacial
erosion by frozen basal ice conditions (akin to selective linear
erosion; Sugden, 1978; Thomson et al., 2010), because of
the inability of snow and ice to accumulate on the steep
bedrock slopes (hence like teflon; precipitation does not
stick but continues to fall to the valley below), and because
they replace valleys and space-out cirques and thus reduce
glacier drainage density (Brocklehurst and Whipple, 2007).
Teflon peaks will also accelerate erosion in valleys below, by
focusing precipitation and avalanching, providing tools for
erosion into the bergschrund, and insulating the valley glacier
from ablation by a blanket of rock fall debris (Brozovic et al.,
1997; Foster et al., 2008). The height of teflon peaks should
depend on bedrock material properties and the original
spacing of large valleys (Foster et al., 2010). A prevalence of
such peaks may indicate that the glacial buzzsaw is locally
defeated, such as where glaciers are small and rock uplift rates
are rapid (e.g., Teton Range; Foster et al., 2008) (Figure 5).
Teflon peaks are not ubiquitous features in glacial topography,
however. Anders et al. (2010) and Egholm et al. (2009)
showed that peak elevations are commonly limited by the
4
Elevation (km)
SW
NE
Southern Alps
Maximum elevation
Modern ELA
2
Mean elevation
1
0
LGM ELA
0
50
100
150
Distance (km)
200
250
Figure 13 An example of an actively uplifting mountain range that violates the glacial buzzsaw hypothesis: the Southern Alps, New Zealand. Adapted
from Brocklehurst, S.H., Whipple, K.X., 2007. Response of glacial landscapes to spatial variations in rock uplift rate. Journal of Geophysical Research
112(F02035), 18 p. Maximum and mean elevation from a swath along the trend of the central Southern Alps, parallel to the Alpine fault, do not
correlate with ELA (modern or LGM). The rise in elevation that bears no relationship to ELA is instead linked to a local increase in rock uplift rate.
altitude of cirques, akin to a glacial base level, which can

effectively prevent the unlimited growth of teflon peaks. Extremely high teflon peaks may require high rates of tectonic
rock uplift, where erosion of glacial headwalls may not be able
to keep pace, thereby resulting in the growth of extremely high
headwalls and overall steepening of basins (Brocklehurst and
Whipple, 2007).
Based on these exceptions, it is clear that there is no simple,
absolute glacial buzzsaw that will limit topographic growth
under all conditions. In some cases, glaciers appear to be
quite capable at erosion and do seem to limit topography.
Glaciers also naturally concentrate erosion at ELA. However, it
is clear that glaciers can be relatively ineffective at erosion
under some conditions, and in certain cases will not be capable of keeping pace with rock uplift. This leads to an interesting idea that the glacial buzzsaw may be scale dependent
and only activates after a certain threshold of glaciation has
been reached. Small glaciers may not have the discharge to
produce adequate sliding that can erode as rapidly as needed
(e.g., Kirkbride and Matthews, 1997; Amundson and Iverson,
2006). Topography that includes numerous large teflon peaks
and inherently small valleys may thus be poorly predisposed
for development of a glacial buzzsaw (Foster et al., 2008).
Foster et al. (2010) suggested that only basins 420 km2 in the
Teton Range were capable of hosting glaciers large enough that
could keep pace with rock uplift. Likewise, increasing rates of
tectonic rock uplift may require larger glaciers for the glacial
buzzsaw to be in effect, suggesting that glacial erosion and
the evolution of glacial topography also depend on rates
of tectonic uplift (Brocklehurst and Whipple, 2007; Foster
et al., 2008).
5.13.4.2
Glacial Erosion and ClimateTectonic Coupling
An effective glacial buzzsaw, by definition, should lead to

steady-state topography and steady-state flux in orogenic systems. This means that the rate of erosional efflux from a glaciated orogen could readily equate to the rate of tectonic
influx, thereby precluding tectonic accretion. Given the importance of a steady-state flux for the behavior of orogenic
wedges (Whipple and Meade, 2004), this implies that glacial
erosion, and the climatic conditions that enable it, have significant geodynamic implications for mountain building
throughout the geologic record. Atmospheric temperature, not
just precipitation, may accordingly be a controlling factor in
collisional tectonics (Tomkin and Roe, 2007).
The effect of erosion on orogenic systems has been explored through recent coupled process models. Numerous
analytical and numerical models have shown the importance
of erosion and the degree of coupling between surface processes and tectonics in orogenic wedges (Beaumont et al.,
1992; Koons, 1995; Willett, 1999). In an orogenic wedge, an
increase in erosional efficiency will cause narrowing and
lowering of the wedge relief and an increase in rock uplift rate
(Whipple and Meade, 2004; Roe et al., 2006). Orogen size and
dynamics can thus be controlled by climate. Spatially nonuniform erosion in these systems, such as should occur at
glacial ELA, can also significantly alter deformation pattern
and pressuretemperaturetime paths during exhumation
311
Leeward
Windward
Modern ELA
LGM ELA
Figure 14 Conceptual diagram of the effect of orographic

precipitation on the size of glaciers and ELA of a coastal mountain
belt. Adapted from Meigs, A., Sauber, J., 2000. Southern Alaska as
an example of the long-term consequences of mountain building
under the influence of glaciers. Quaternary Science Reviews 19,
15431562. The windward face of the range traps more precipitation,
resulting in larger glaciers. The decrease in precipitation away from
the coast results in a rise in both modern and LGM ELA and results
in smaller glaciers on the leeward side of the range. If glaciers
influence mean topography, orographically-induced asymmetry in
glacier distribution will have a major effect on the evolution of
mountain ranges.
(Whipple and Meade, 2004). Modeling of a glaciated orogenic

wedge suggests that orogen-scale exhumation will become
focused at and just below the mean ELA (Tomkin and Roe,
2007; Tomkin, 2007). This differs from fluvially dominated
orogens, in which rock uplift is focused near the core of the
range. Tomkin and Roe (2007) also showed that the dimensions of glaciated wedges are more sensitive to climatic fluctuations, such as precipitation rate, than fluvial wedges. Other
numerical models show that glacial erosion can be more important in eroding orogenic wedges than fluvial erosion.
Tomkin (2007) showed that total orogenic erosion is linearly
proportional to the degree of glacier coverage, and explored
this result for four test cases of real world glaciated orogens.
The results were broadly consistent with the glacial buzzsaw
effecting rapid erosion, implying that the onset of glaciation in
the late Cenozoic increased rates of rock uplift. Other models
have shown that the effect of glacial erosion on orogenic
evolution can actually be complex and dynamic (Herman and
Braun, 2008; Tomkin, 2009).
Case studies of real world glaciated orogens have also
shown the impact of glacial erosion on tectonic processes.
Work in southern Alaskas St. Elias orogen has shown that
the long-term exhumation is focused on the windward side of
the critical wedge, where precipitation is greater and glaciers
are larger (Meigs and Sauber, 2000; Spotila et al., 2004)
(Figure 14). Moreover, rapid (i.e., 5 mm yr1) exhumation
appears to be focused where the mean Quaternary ELA
intersects the windward flank, termed the ELA front, matching the predictions of numerical models (Berger and Spotila,
2008; Tomkin and Roe, 2007) (Figure 15). Exhumation rates
along this zone of focused denudation do not vary with the
size (surface area) of glaciers draining across it, however,
suggesting a local violation of the rule that glacier erosion
rate correlates to ice discharge. This may indicate that
small glaciers are able to keep pace with rock uplift by cutting
down to the base level established by larger glaciers,
such that tectonics ultimately sets rock uplift rates here, or
that erosion by large outlet glaciers is somehow limited, perhaps by overdeepenings (Berger and Spotila, 2008). Although
312
N
6
20
2
15
10
1
5
Prevailing winds
AHe age (Ma)
Elevation (km)
Modern ELA
LGM ELA
Max. 4
elev.
Mean
elev. 2
ELA front
Min
elev.
0
0
20
40
60
Distance (km)
80
100
Precipitation (m yr 1)
25
0
120
Figure 15 Relationship of the spatial distribution of exhumation and ELA in the St. Elias orogen, Alaska. Adapted from Berger, A.L., Spotila,
J.A., 2008. Denudation and deformation of a glaciated orogenic wedge: The St. Elias orogen Alaska. Geology 36, 523526. Exhumation is
represented by the apatite (UTh)/He ages (or AHe ages, in Ma), where younger ages mean more rapid exhumation. Note how exhumation is
most rapid (youngest ages; red box) where the modern and LGM ELA intersect the windward flank of the orogen (the ELA front; blue box). In
contrast, exhumation does not correlate with the broad distribution of precipitation, topography (maximum, minimum, and mean swath
topography shown), or distribution of convergent structures (not shown). This is consistent with ELA focusing long-term glacial erosion and for
the idea that glaciers can act as excavators in dictating patterns of denudation and strain in orogens.
thermochronometry of bedrock suggests that long-term exhumation rates do not peak near massive outlet glaciers that
cross the deformation front (e.g., the Bering glacier), detrital
dating by Enkelmann et al. (2009) suggests that more rapid
exhumation may occur beneath the largest glaciers, where
bedrock cannot be sampled (e.g., the Seward glacier). Work in
the St. Elias orogen has also suggested that glacial climate
forced a change in the dynamics of the entire wedge. Differences in long-term bedrock cooling rates from different thermochronometers show that a sharp acceleration of cooling
occurred not at the onset of glaciation but at B1 Ma, corresponding to a shift in the pattern of glaciation associated with
the change in Milkovitch forcing from 41 to 100 ka cycles
(Berger et al., 2008a; Ruddiman et al., 1986). Berger et al.
(2008a, b) speculated that this change forced the orogen to
narrow and reorganize, including onset of backthrusting (i.e.,
transition to a two-sided wedge) and termination of distal
structures via sediment burial. The St. Elias orogen thus provides an exceptional case study of the effects of long-term
glacial conditions on the evolution of a collisional mountain
belt. The degree to which glaciers act as orogen-scale excavators in controlling the pattern of erosional efflux from this
orogen implies a high degree of coupling between erosion
and tectonics in glaciated collision belts. Other locations
where similar inferences have been made include the Andes,
Southern Alps, Fiordland (New Zealand), and Olympic
Mountains (Thomson, 2002; Thomson et al., 2010; Tomkin,
2007; Shuster et al., 2011; Tomkin and Roe, 2007).
Like the interaction of glaciers and topography, the interaction of tectonics and climate under glacial conditions is
complex and dependent on the nature of coupling between
numerous processes and conditions. At the scale of individual
glacial valleys, glaciers (and therefore climate, by way of ELA)
control erosion and shape topography, but are in turn affected
by erosive processes and topography itself, for example,
shading, effect of rock debris cover on glacier albedo, input of
erosion tools into the bergschrund, topographic steering of
orographic precipitation, and blowing and avalanching snow
(Foster et al., 2008, 2010; Anders et al., 2008, 2010; MacGregor

et al., 2009). At the orogenic scale, glaciers are capable of
focusing denudation and influencing deformation and orogen
architecture, but in turn are heavily dependent on the rates of
rock uplift set tectonically (Berger and Spotila, 2008). Rock
uplift rate itself appears to be an important parameter in determining how glacial erosion will affect a landscape and how
landscapes will react to climate change (Berger and Spotila,
2008; Tomkin and Roe, 2007; Tomkin, 2009; Koppes and
Montgomery, 2009). Similarly, ice-mass balance and its related characteristics (e.g., glacier thickness) appear to be influenced by orogenic structure and evolution (Tomkin, 2003).
Thus, the climateerosiontectonics system, when glacially
dominated, is dynamic, and will evolve in a complex fashion
that is dependent on ambient conditions and resistant to the
imposition of absolute generalities.
5.13.4.3
Orogen and Landscape Response to Glacial

Climate Change
It has long been recognized that changes in glacial cover due

to climate change can modify landscapes and result in rock
uplift. The viscoelastic, isostatic response of a landscape to
changes in ice volume has been well documented in areas of
recent deglaciation, such as the postglacial surface uplift in
Fennoscandia or the Alps (Gudmundsson, 1999; Bungum
et al., 2010; Barletta et al., 2006). Glacial retreat in the past
century in the fjords of southern Alaska, in the vicinity of
Glacier Bay, has likewise resulted in spectacular (B3 cm yr1)
rates of surface uplift. Deglaciation also has the important
effect of destabilizing landscapes. Following ice retreat, hillslope erosion increases due to removal of ice buttresses,
whereas fluvial networks must adjust and modify the glacial
topography and convey large accumulations of liberated glacial sediment (e.g., Meigs and Sauber, 2000). Major changes
in glacial cover thus result in substantial reorganization and
adjustment of the entire erosion system.
The onset of glaciation in an orogen has also been recognized as having significant geodynamic and geomorphic
influence. Several studies have identified accelerations in exhumation rates related to the onset of glaciation, including in
the Coast Ranges in British Columbia (Farley et al., 2001;
Ehlers et al., 2006; Densmore et al., 2007) and the Svalbard
islands in the Arctic (Blythe and Kleinspehn, 1998). Changes
in glacial coverage, such as resulting from the change from 41
to 100 ka Milankovitch periodicity (Ruddiman et al., 1986) at
B1 Ma or other subordinate climate changes, have also been
linked to major (i.e., more than ten-fold) increases in denudation rates, such as in the Coast Ranges, southern Alaska,
the European Alps, Himalaya, and Fennoscandia (Shuster
et al., 2005; Haeuselmann et al., 2007; Berger et al., 2008a;
Rahaman et al., 2009; Dowdeswell et al., 2010). These increases in denudation should correspond to concomitant increases in sedimentation in adjacent basins. Such changes are
also consistent with the hypothesis that late Cenozoic atmospheric cooling enhanced erosion and so influenced
mountain building and tectonics on a global scale (Molnar
and England, 1990).
At the center of the original Molnar and England (1990)
chickenegg hypothesis for the relationship of climate and
tectonic change is the prediction that a transition to glacially
dominated erosion can increase topographic incision (i.e.,
increase topographic relief and decrease mean elevation)
and thereby result in isostatic (Airy) uplift of peaks. Although
the lithospheres response to a change in crustal volume
associated with incision-related unloading can theoretically
be significant (Montgomery, 1994), the magnitude of uplift in
real-world cases is limited by the narrow width of mountain
belts relative to typical flexural wavelengths and the magnitude of relief produced by onset of glacial conditions
(Small and Anderson, 1998; Pelletier, 2004; Cederbom et al.,
2004; Champagnac et al., 2009). In most cases, the magnitude
of relief production and peak uplift predicted for actual
mountain ranges are limited to a few hundred meters (Small
and Anderson, 1995; Whipple et al., 1999; Brocklehurst
and Whipple, 2002; Brocklehurst et al., 2008; van der Beek
and Bourbon, 2008). The contribution of peak height resulting from relief production and glacioisostatic compensation is increased, however, under polythermal ice conditions
typical of ice-streams at the margins of polar ice sheets
(Kaplan and Miller, 2003; Stern et al., 2005; Medvedev et al.,
2008).
An interesting issue related to the response of a landscape
or orogen to a change in glacial climate is the timescale
at which these systems react. How long did it take to adjust
and find a new equilibrium following the transition to
the glacial climate of the late Cenozoic? Are response
times short enough that landscapes will adjust to climate
change over individual ice ages (e.g., 100 ka)? Or, are response
times so sluggish that glaciated landscapes and orogens never
truly attain steady state? As discussed above, the timescale
at which glacial erosion can modify a pre-existing fluvial
topography is on the order of several hundred thousand years
(Harbor, 1992; Kirkbride and Matthews, 1997; MacGregor
et al., 2000; Brook et al., 2006, 2008; Jamieson et al., 2008).
This is relatively rapid (i.e., glacial topography can be manifest
after only one or two glacial cycles), but implies that
313
landscapes will not attain states of equilibrium at the periodicity of individual ice ages (Herman and Braun, 2008;
Tomkin, 2009). Tomkin and Roe (2007) showed that the
timescale of response of an orogenic wedge to glaciation is
longer. Their numerical simulations suggest that the dimensions and fluxes of orogenic wedges will adjust to onset of
glaciation, modeled as a step function change in erosivity,
to attain a new steady state only after a few million years.
This means that short term climate cycles will be averaged
out and not individually ascertainable in the dynamic
response of orogenic wedges. That orogens do not respond
instantaneously also suggests that interpretations of sudden
change in exhumation rate due to specific climate changes
observed in case studies should be taken with caution
(e.g., Berger et al., 2008a).
Although changes in glacial climate can have major impact
on landscape and orogenic systems, not every change need do
so. There is some indication that the effect of glacial erosion
on topography and orogenic architecture may be dependent
on thresholds, such that only climate change of a certain
magnitude will cause noticeable topographic or tectonic
modification. For example, studies of alpine glaciers have indicated that there is a threshold of glaciation (ice coverage, ice
flux, ice sheet integration, size of glacial valleys, etc.) that must
be met to significantly affect topography and effect the glacial
buzzsaw; climate change resulting in small glaciers may not be
sufficient (Brocklehurst and Whipple, 2004; Egholm et al.,
2009; Foster et al., 2008, 2010). Brocklehurst and Whipple
(2007) indicated that glaciers needed to be 30100 km2 in
order to be capable of incising rapidly and keeping pace with
tectonic uplift. This is consistent with the correlation between
glacial erosion and degree of glacier coverage (Hallet, 1996;
Montgomery, 2002). Examples of orogens that experienced
changes in denudation or architecture attributed to glacial
erosion that lagged behind the inception of glaciations also
suggest that there is a threshold of glacial erosion that must be
met before an erosion system is reorganized (e.g., Shuster
et al., 2005; Haeuselmann et al., 2007; Berger et al., 2008a;
Dowdeswell et al., 2010).
5.13.5
Discussions and Conclusions
Two contradictory themes for the influence of glacial erosion

on topography and mountain building have emerged from
this review. In many respects, glacial erosion appears complex,
so as to defy the application of simple rules or predictions
(Lliboutry, 1993). How glaciers erode depends on many
conditions, including the ice mass balance, basal thermal
conditions, subglacial hydrology, slope and morphology of
the bed, headwall processes, rock uplift rate, and bedrock
properties, many of which are coupled to erosion processes
and resulting topography (Herman and Braun, 2008;
MacGregor et al., 2009). In this sense, glacial erosion appears
to be a nonlinear, dynamic system. In contrast, many aspects
of glacial erosion and the resulting topography can be relatively simply explained, by assuming that glacial erosion
scales linearly with the basal ice velocity or unit ice flux
(Anderson et al., 2006). Using relatively simple erosion rules,
models have demonstrated how numerous aspects of glacial
314
topography evolve, such as cross-valley profiles (Harbor,

1992). At the same time, complex models that take into account numerous processes and conditions are required to
explain other phenomena, such as cirque retreat (MacGregor
et al., 2009). This implies that while a very rough approach
can explain some first-order aspects of glacial erosion, a
comprehensive understanding that is transportable to variable
environments requires a much more detailed analysis of
specific local conditions.
Glacial erosion is clearly an effective means of denuding
landscapes. Glacial erosion rates can be extremely rapid, and
there are many indications that glacial erosion is commonly
more effective than fluvial erosion. Glacial erosion generally
excavates a greater volume of rock from topography than is
observed in fluvial landscapes and onset of glaciation can
generate moderate relief (Montomery, 2002; Brocklehurst and
Whipple, 2002). This is consistent with the idea that late
Cenozoic cooling affected erosion, topography, and sedimentation in the many of the worlds orogens (Molnar and
England, 1990). However, both forms of erosion appear to
max-out at similar magnitudes, suggesting they are ultimately
trumped by geodynamically limited rock uplift rates (Koppes
and Montgomery, 2009). Glacial erosion also appears to be
considerably more variable in time than fluvial erosion. In
particular, glacial erosion seems particularly effective during
periods of melting or ice retreat (Riihimaki et al., 2005), and
in general erosion seems to accelerate during periods of climate or landscape transition. In some respects, however, the
question of whether glacial erosion is more proficient than
fluvial erosion is contrived and moot; both forms of erosion
will depend on ambient conditions and, by definition, the two
forms of erosion cannot codominate under the same conditions. That the glacial buzzsaw hypothesis has been validated in numerous (but not all) settingsz does imply that
glacial erosion has significant geodynamic importance, but
beyond this, comparisons between glacial and fluvial erosion
seeking a champion can be a distraction from more fundamental process-based questions that need to be addressed to
advance our understanding of glacial erosion itself.
Like fluvial erosion, glacial erosion strongly influences
tectonics, including dictating patterns of denudation and deformation in orogenic systems (Spotila et al., 2004; Berger
et al., 2008b; Berger and Spotila, 2008). There are important
differences in how these systems function, however, such as
the concentration of erosion midway along glaciers, the
greater penetration of orographic precipitation into mountains during glacial climates, and the extreme efficiency of the
glacial buzzsaw, where it occurs (Anderson et al., 2006; Anders
et al., 2008; Brozovic et al., 1997). As a result, orogenic systems may behave differently when dominated by glacial
conditions, and thus climate should be considered fundamental to evolution of orogens and other tectonic processes,
such as terrane accretion, throughout the geologic record
(Molnar and England, 1990). Exactly how glacial erosion will
have influenced a given tectonic setting depends on local
conditions, given the potential thresholds for onset of the
glacial buzzsaw (Foster et al., 2010). For example, some orogens may respond to the onset of glaciation, whereas others
may not respond until a certain degree of glaciation is met,
such as forced by a change in the periodicity of Milankovitch
cycles (Berger and Spotila, 2008a). Where it is present, however, the glacial excavator should be highly effective at denuding an orogen and lead to a high degree of coupling
between erosion and tectonics, including fostering the development of a steady-state flux (Tomkin and Roe, 2007).
As is typical in science, the most interesting lingering
questions related to glacial erosion and their effect on tectonics involve the exceptions to rules of behavior. A problem
of major importance for the evolution of alpine topography
and orogenic systems is how and why the glacier buzzsaw can
be violated, the role of thresholds, and what controls the
height and character of teflon peaks (Foster et al., 2010).
Equally revealing may be the exceptions to the general rule
that glacial erosion scales with basal ice flux, such as can result
from subglacial hydrology or the thermal regime of the icesubstrate interface (Riihimaki et al., 2005; Jamieson et al.,
2008). Understanding the temporal variation and timescale
dependence of glacial erosion is also a prime target of future
research. Another direction of investigation that will advance
our understanding will be a focus on what controls sediment
production in glacial landscapes. Studies of headwall retreat
and teflon peaks have shown the importance of periglacial
weathering, which in turn is important for the erosional
processes of the glacier itself (Oskin and Burbank, 2005;
Foster et al., 2010; MacGregor et al., 2009). Studies have
shown that periglacial erosion may contribute the majority of
sediment that is transmitted by a glacier (OFarrell et al.,
2009), and frost shattering may actually be a mechanism of
limiting relief in glacial landscapes (Hales and Roering, 2009).
Detailed studies focusing on these secondary aspects of the
glacial system are needed. Likewise, quantitative field studies
of the links between processes and conditions in glacial systems are needed to enhance physics-based erosion laws, such
as quantification of contributions of abrasion versus plucking,
measurements of input of clasts into the bergschrund, rates of
subglacial weathering, and the influence of bedrock properties
on topographic and orogenic evolution (Lliboutry, 1993;
Riihimaki et al., 2005). Finally, additional case studies of
the glaciated orogenic systems that exhibit a high degree
of coupling between climate, erosion, and tectonics, are
needed (e.g., the St. Elias orogen, the Southern Alps, and the
Patagonian Andes). These constraints will enable more complex, comprehensive, three-dimensional modeling of ice flow
and resulting erosion, in what amounts to a systems-based
approach to understanding glacial erosion and its effect on
evolving topography and tectonics.
References
Alley, R.B., Lawson, D.E., Larson, G.J., Evenson, E.B., Baker, G.S., 2003. Stabilizing
feedbacks in glacier-bed erosion. Nature 424, 758760.
Amerson, B.E., Montgomery, D.R., Meyer, G., 2007. Relative size of fluvial and
glaciated valleys in central Idaho. Geomorphology 93, 537547.
Amundson, J.M., Iverson, N.R., 2006. Testing a glacial erosion rule using hang
heights of hanging valleys, Jasper National Park, Alberta. Journal of Geophysical
Research 111(F01020), 8.
Anders, A.M., Mitchell, S.G., Tomkin, J.H., 2010. Cirques, peaks, and precipitation
patterns in the Swiss Alps: Connections among climate, glacial erosion, and
topography. Geology 38, 239242.
Anders, A.M., Roe, G.H., Montgomery, D.R., Hallet, B., 2008. Influence of
precipitation phase on the form of mountain ranges. Geology 36, 479482.
Anderson, R.S., 2005. Teflon peaks: The evolution of high local relief in glaciated
mountain ranges. Eos Transactions 86, Fall Meeting Supplement, Abstact
H33F-04.
Anderson, R.S., Molnar, P., Kessler, M.A., 2006. Features of glacial valley profiles
simply explained. Journal of Geophysical Research 111(F01004), 14.
Atkins, C.B., Barrett, P.J., Hicock, S.R., 2002. Cold glaciers erode and deposit:
Evidence from Allan Hills, Antarctica. Geology 30, 659662.
Augustinus, P.C., 1992. The influence of rock mass strength on glacial valley
cross-profile morphometry: A case study from the Southern Alps, New Zealand.
Barletta, V.R., Ferrari, C., Diolaiuti, G., Carnielli, T., Sabadini, R., Smiraglia, C.,
2006. Glacier shrinkage and modeled uplift of the Alps. Geophysical Research
Letters 33, L14307.
Beaumont, C., Fullsack, P., Hamilton, J., 1992. Erosional control of active
compressional orogens. In: McClay, K.R. (Ed.), Thrust Tectonics. Chapman and
Hall, New York, pp. 118.
Berger, A.L., Gulick, S.P., Spotila, J.A., et al., 2008b. Quaternary tectonic response
to intensified glacial erosion in an orogenic wedge. Nature Geoscience 1,
793799.
Berger, A.L., Spotila, J.A., 2008. Denudation and deformation of a glaciated
orogenic wedge: The St. Elias orogen Alaska. Geology 36, 523526.
Berger, A.L., Spotila, J.A., Chapman, J., Pavlis, T.L., Enkelmann, E., 2008a.
Architecture, kinematics, and exhumation of a convergent orogenic wedge: A
thermochronological investigation of tectonic-climatic interactions within the St.
Elias orogen, Alaska. Earth and Planetary Science Letters 270, 1324.
Blythe, A.E., Kleinspehn, K.L., 1998. Tectonically versus climatically driven Cenozoic
exhumation of the Eurasian plate margin, Svalbard: Fission track analyses.
Tectonics 17, 621639.
Bogen, J., 1996. Erosion rates and sediment yields of glaciers. Annals of
Braun, J., Zwartz, D., Tomkin, J.H., 1999. A new surface-processes
model combining glacial and fluvial erosion. Annals of Glaciology 28,
282290.
Briner, J.P., Miller, G.H., Davis, P.T., Finkel, R.C., 2006. Cosmogenic radionuclides
from fjord landscapes support differential erosion by overriding ice sheets.
Brocklehurst, S.H., Whipple, K.X., 2002. Glacial erosion and relief production in the
Eastern Sierra Nevada, California. Geomorphology 42, 124.
Brocklehurst, S.H., Whipple, K.X., 2004. Hypsometry of glaciated landscapes. Earth
Brocklehurst, S.H., Whipple, K.X., 2007. Response of glacial landscapes to spatial
variations in rock uplift rate. Journal of Geophysical Research 112(F02035),
18 p.
Brocklehurst, S.H., Whipple, K.X., Foster, D., 2008. Ice thickness and topographic relief in glaciated landscapes of the western US. Geomorphology 97,
3551.
Brook, M.S., Kirkbride, M.P., Brock, B.W., 2006. Quantified time scale for glacial
valley cross-profile evolution in alpine mountains. Geology 34, 637640.
Brook, M.S., Kirkbride, M.P., Brock, B.W., 2008. Temporal constraints on glacial
valley cross-profile evolution: Two Thumb Range, central Southern Alps, New
Zealand. Geomorphology 97, 2434.
Brozovic, N., Burbank, D.W., Meigs, A.J., 1997. Climatic limits on landscape
development in the northwestern Himalaya. Science 276, 571574.
Bungum, H., Olesen, O., Pascal, C., Gibbons, S., Lindholm, C., Vestol, O., 2010.
To what extent is the present seismicity of Norway driven by post-glacial
rebound? Journal of the Geological Society of London 167, 373384.
northwestern Himalaya. Nature 379, 505510.
Cederbom, C.E., Sinclair, H.D., Schlunegger, F., Rahn, M.K., 2004. Climate-induced
rebound and exhumation of the European Alps. Geology 32, 709712.
Champagnac, J.D., Schlunegger, F., Norton, K., Blanckenburg, F., Abbuhl, L.M.,
Schwab, M., 2009. Erosion-driven uplift of the modern Central Alps.
Clarke, G.K.C., 2005. Subglacial processes. Annual Review of Earth and Planetary
Cohen, D., Hooyer, T.S., Iverson, N.R., Thomason, J.F., Jackson, M., 2006. Role of
transient water pressure in quarrying; a subglacial experiment using acoustic
emissions. Journal of Geophysical Research 111(F03006), 13 pp.
Colgan, P.M., Bierman, P.R., Mickelson, D.M., Caffee, M., 2002. Variation in glacial
erosion near the southern margin of the Laurentide ice sheet, south-central
Wisconsin, USA; Implications for cosmogenic dating of glacial terrains.
315
Davies, M.T.L., Atkins, C.B., van der Meer, J.J.M., Barrett, P.J., Hicock, S.R., 2009.
Evidence for cold-based glacial activity in the Allan Hills, Antarctica. Quarternary
Delmas, M., Calvet, M., Gunnell, Y., 2009. Variability of Quaternary glacial erosion
rates A global perspective with special reference to the Eastern Pyrenees.
Densmore, M.S., Ehlers, T.A., Woodsworth, G.J., 2007. Effect of alpine glaciation on
thermochronometer age-elevation profiles. Geophysical Research Letters 34,
L02502.
Dowdeswell, J.A., Ottesen, D., Rise, L., 2010. Rates of sediment delivery from the
Fennoscandian Ice Sheet through an ice age. Geology 38, 36.
Duhnforth, M., Anderson, R.S., Ward, D, Stock, G.M., 2010. Bedrock fracture
control of glacial erosion processes and rates. Geology 38, 423426.
Egholm, D.L., Nielsen, S.B., Pedersen, V.K., Lesemann, J.-E., 2009. Glacial effects
limiting mountain height. Nature 460, 884888.
Ehlers, T.A., Farley, K.A., Rusmore, M.E., Woodsworth, G.J., 2006. Apatite (UTh)/
He signal of large-magnitude accelerated glacial erosion, British Columbia.
Geology 34, 765768.
Enkelmann, E., Zeitler, P.K., Pavlis, T.L., Garver, J.I., Ridgway, K.D., 2009. Intense
localized rock uplift and erosion in the St. Elias orogen of Alaska. Nature
Fabel, D., Stroven, A.P., Harbor, J., Kleman, J., Elmore, D., Fink, D., 2002.
Landscape preservation under Fennoscandian ice sheets determined from
in situ produced 10Be and 26Al. Earth and Planetary Science Letters 201,
397406.
Farley, K.A., Rusmore, M., Bogue, S., 2001. Post-10 Ma uplift and exhumation of
the northern Coast Mountains, British Columbia. Geology 29, 99102.
Flowers, G.E., 2008. Subglacial modulation of the hydrograph from glacierized
basins. Hydrological Processes 22, 39033918.
Foster, D., Brocklehurst, S.H., Gawthorpe, R.L., 2008. Small valley glaciers and the
effectiveness of the glacial buzzsaw in the northern Basin and Range, USA.
Foster, D., Brocklehurst, S.H., Gawthorpe, R.L., 2010. Glacial-topographic
interactions in the Teton Range, Wyoming. Journal of Geophysical Research
115(F01007), 20.
Fountain, A.G., Walder, J.S., 1998. Water flow through temperate glaciers. Reviews
of Geophysics 36, 299328.
Glasser, N.F., Ghiglione, M.C., 2009. Structural, tectonic, and glaciological controls
on the evolution of fjord landscapes. Geomorphology 105, 291302.
Gudmundsson, A., 1999. Postglacial crustal doming, stresses, and fracture
formation with application to Norway. Tectonophyscs 307, 407419.
Haeuselmann, P., Granger, D.E., Jeannin, P.-Y., Lauritzen, S.-E., 2007. Abrupt glacial
valley incision at 0.8 Ma dated from cave deposits in Switzerland. Geology 35,
143146.
Hales, T.C., Roering, J.J., 2009. A frost buzz-saw mechanism for erosion of the
eastern Southern Alps, New Zealand. Geomorphology 107, 241253.
Hallet, B., 1979. A theoretical model of glacial abrasion. Journal of Glaciology 23,
3950.
Hallet, B., 1996. Glacial quarrying: A simple theoretical model. Annals of Glaciology
22, 18.
Hallet, B., Hunter, L., Bogen, J., 1996. Rates of erosion and sediment evacuation by
glaciers: A review of field data and their implications. Global and Planetary
Change 12, 213255.
Harbor, J., Warburton, J., 1993. Relative rates of glacial and nonglacial erosion in
alpine environments. Arctic and Alpine Research 25, 17.
Harbor, J.M., 1992. Numerical modeling of the development of U-shaped valleys
by glacial erosion. Geological Society of America Bulletin 104, 13641375.
Hattestrand, C., Stroeven, A.P., 2002. A relict landscape in the centre of
Fennoscandian glaciation: Geomorphological evidence of minimal Quaternary
glacial erosion. Geomorphology 44, 127143.
Herman, F., Braun, J., 2008. Evolution of the glacial landscape of the Southern Alps
of New Zealand: Insights from a glacial erosion model. Journal of Geophysical
Research 113(F02009), 24 pp.
Hobley, D.E.J., Sinclair, H.D., Cowie, P.A., 2010. Processes, rates, and time scales
of fluvial response in an ancient postglacial landscape of the northwest Indian
Himalaya. Geological Society of America Bulletin 122, 15691584.
Hooke, R.L., 1991. Positive feedbacks associated with erosion of glacial cirques
and overdeepenings. Geological Society of America Bulletin 103, 11041108.
Humphrey, N.F., Raymond, C.F., 1994. Hydrology, erosion, and sediment production
in a surging glacier: Variegated Glacier, Alaska, 198283. Journal of Glaciology
40, 539552.
Jamieson, S.S.R., Hulton, N.R.J., Hagdorn, M., 2008. Modeling landscape evolution
under ice sheets. Geomorphology 97, 91108.
316
Kaplan, M.R., Hein, A.S., Hubbard, A., Lax, S.M., 2009. Can glacial erosion limit
the extent of glaciation? Geomorphology 103, 172179.
Kaplan, M.R., Miller, G.H., 2003. Early Holocene deleveling and deglaciation of the
Cumberland Sound region, Baffin Island, Arctic Canada. Geological Society of
Kavanaugh, J.L., 2009. Exploring glacier dynamics with subglacial water pressure
pulses; evidence for self-organized criticality? Journal of Geophysical Research
114, F01021.
Kessler, M.A., Anderson, R.S., Briner, J.P., 2008. Fjord insertion into continental
margins driven by topographic steering of ice. Nature Geoscience 1, 365369.
Kirkbride, M., Matthews, D., 1997. The role of fluvial and glacial erosion in
landscape evolution: The Ben Ohau Range, New Zealand. Earth Surface
Koons, P.O., 1995. Modeling the topographic evolution of collisional mountain
belts. Annual Reviews of Earth and Planetary Sciences 23, 375408.
Koppes, M., Hallet, B, 2006. Erosion rates during rapid deglaciation in Icy Bay
Alaska. Journal of Geophysical Research 111, F02023.
Koppes, M., Hallet, B., Anderson, J., 2009. Synchronous acceleration of ice loss
and glacial erosion, Glaciar Marinelli, Chilean Tierra del Fuego. Journal of
Koppes, M.N., Montgomery, D.R., 2009. The relative efficacy of fluvial and
glacial erosion over modern to orogenic timescales. Nature Geoscience 2,
644647.
Lliboutry, L.A., 1993. Monolithologic erosion of hard beds by temperate glaciers.
Loso, M.G., Anderson, R.S., Anderson, S.P., 2004. Post-Little Ice Age record of
coarse and fine clastic sedimentation in an Alaska pro-glacial lake. Geology 32,
10651068.
MacGregor, K.R., Anderson, R.A., Waddington, E.D., 2009. Numerical modeling of
glacial erosion and headwall processes in alpine valleys. Geomorphology 103,
189204.
MacGregor, K.R., Anderson, R.S., Anderson, S.P., Waddington, E.D., 2000.
Numerical simulations of glacial-valley longitudinal profile evolution. Geology
28, 10311034.
Medvedev, S., Hartz, E.H., Podladchikov, Y.Y., 2008. Vertical motions of the fjord
regions of central East Greenland: Impact of glacial erosion, deposition, and
isostasy. Geology 36, 539542.
Meigs, A., Sauber, J., 2000. Southern Alaska as an example of the long-term
consequences of mountain building under the influence of glaciers. Quaternary
Mitchell, S.G., Montgomery, D.R., 2006. Influence of a glacial buzzsaw on the
height and morphology of the Cascade Range in central Washington State, USA.
Molnar, P., England, P., 1990. Late Cenozoic uplift of mountain peaks and global
climate change: Chicken or egg? Nature 346, 2934.
Montgomery, D.R., 1994. Valley incision and the uplift of mountain peaks. Journal
of Geophysical Research 99, 13 91313 921.
Montgomery, D.R., 2002. Valley formation by fluvial and glacial erosion. Geology
30, 10471050.
Montgomery, D.R., Balco, G., Willett, S., 2001. Climate, tectonics, and the
morphology of the Andes. Geology 29, 579582.
Naylor, S., Gabet, E.J., 2007. Valley asymmetry and glacial versus nonglacial
erosion in the Bitterroot Range, Montana, USA. Geology 35, 375378.
Oerlemans, J., 1984. Numerical experiments on large-scale glacial erosion.
Zeitschrift fur Gletscherkunde und Glazialgeologie 20, 107126.
OFarrell, C.R., Heimsath, A.M., Lawson, D.E., et al., 2009. Quantifying periglacial
erosion: Insights on a glacial sediment budget, Matanuska Glacier, Alaska. Earth
Oskin, M., Burbank, D.W., 2005. Alpine landscape evolution dominated by cirque
retreat. Geology 33, 933936.
strem, G., 1972. Height of the glaciation level in northern British Columbia and
southeastern Alaska. Geografiska Annaler 54A, 7684.
Paterson, W., 1994. Physics of Glaciers. Third Ed. Elsevier, New York, 480 pp.
Pelletier, J.D., 2004. Estimate of three-dimensional flexural-isostatic response to
unloading: Rock uplift due to late Cenozoic glacial erosion in the western United
States. Geology 32, 161164.
Porter, S.C., 1977. Present and past glaciation threshold in the Cascade Range,
Washington, U.S.A., Topographic and climatic controls and paleoclimatic
implications. Journal of Glaciology 18, 101116.
Porter, S.C., 1981. Pleistocene glaciation in the southern Lake District of Chile.
Rahaman, W., Singh, S.K., Sinha, R., Tandon, S.K., 2009. Climate control on
erosion distribution over the Himalaya during the past B100 Ka. Geology 37,
559662.
Riihimaki, C.A., MacGregor, K.R., Anderson, R.S., Anderson, S.P., Loso, M.G.,
2005. Sediment evacuation and glacial erosion rates at a small alpine glacier.
Journal of Geophysical Research 110(F03003), 17 pp.
Roe, G.H., Stolar, D.S., Willett, S.D., 2006. Response of a steady-state critical
wedge orogen to changes in climate and tectonic forcing. In: Willett, S.D.,
Hovius, N., Brandon, M.T., Fisher, D.M. (Eds.), Tectonics Climate, and
Landscape Evolution. Geol. Soc. Amer. Special Paper 398. Geological Society of
America, Denver, CO, pp. 227239.
Ruddiman, W.F., Raymo, M.E., McIntyre, A., 1986. Matuyama 41,000-year cycles;
North Atlantic Ocean and northern hemisphere ice sheets. Earth and Planetary
Seong, B.Y., Owen, L.A., Yi, C., Finkel, R.C., Schoenbohm, L., 2009.
Geomorphology of anonymously high glaciated mountains at the northwestern
end of Tibet: Muztag Ata and Kongur, Shan. Geomorphology 103, 227250.
Shuster, D.L., Cuffey, K.M., Sanders, J.W., Balco, G., 2011. Thermochronometry
reveals headward propagation of erosion in an alpine landscape. Science 332,
8488.
Shuster, D.L., Ehlers, T.A., Rusmore, M.E., Farley, K.A., 2005. Rapid glacial erosion
at 1.8 Ma revealed by 4He/3He thermochronometry. Science 310, 16681670.
Small, E.E., Anderson, R.S., 1995. Geomorphically driven late Cenozoic rock uplift
in the Sierra Nevada, California. Science 270, 277280.
Small, E.E., Anderson, R.S., 1998. Pleistocene relief production in Laramide
mountain ranges, western United States. Geology 26, 123126.
Spotila, J.A., Buscher, J., Meigs, A., Reiners, P., 2004. Long-term glacial erosion of
active mountain belts: Example of the Chugach-St. Elias Range, Alaska. Geology
32, 501504.
Spotila, J.A., Farley, K.A., Yule, J.D., Reiners, P.W., 2001. Near-field transpressive
deformation along the San Andreas fault zone in southern California, based on
exhumation constrained by (UTh)/He dating. Journal of Geophysical Research
106, 30 90930 922.
Staiger, J.K.W., Gosse, J.C., Johnson, J.V., et al., 2005. Quaternary relief generation
by polythermal glacier ice. Earth Surf. Process. and Landforms 30, 11451159.
Stern, T.A., Baxter, A.K., Barrett, P.J., 2005. Isostatic rebound due to glacial erosion
within the Transantarctic Mountains. Geology 33, 221224.
Stroeven, A.P., Harbor, J., Fabel, D., et al., 2006. Slow, patchy landscape evolution
in northern Sweden despite repeated ice-sheet glaciation. In: Willett, S.D.,
Hovius, N., Brandon, M.T., Fisher, D.M. (Eds.), Tectonics Climate, and
Landscape Evolution. Geol. Soc. Amer. Special Paper 398. Geological Society of
America, Denver, CO, pp. 387396.
Sugden, D.E., 1978. Glacial erosion by the Laurentide ice sheet. Journal of
Sugden, D.E., John, B.S., 1976. Glaciers and landscape. Edward Arnold Ltd.,
London, 376 pp.
Thomson, S.N., 2002. Late Cenozoic geomorphic and tectonic evolution of the
Patagonian Andes between latitudes 42S and 46S: An appraisal based on
fission-track results from the transpressional intra-arc Liquine-Ofqui fault zone.
Thomson, S.N., Brandon, M.T., Tomkin, J.H., Reiners, P.W., Vasquez, C., Wilson,
N.J., 2010. Glaciation as a destructive and constructive control on mountain
building. Nature 467, 313317.
Tomkin, J.H., 2003. Erosional feedbacks and the oscillation of ice masses. Journal
of Geophysical Research 108(2488), 12 pp.
Tomkin, J.H., 2007. Coupling glacial erosion and tectonics at active orogens: A
numerical modeling study. Journal of Geophysical Research 112(F02015),
14 pp.
Tomkin, J.H., 2009. Numerically simulating alpine landscapes: The geomorphologic
consequences of incorporating glacial erosion in surface process models.
Tomkin, J.H., Braun, J., 2002. The influence of alpine glaciation on the relief
of tectonically active mountain belts. American Journal of Science 302,
169190.
Tomkin, J.H., Roe, G.H., 2007. Climate and tectonic controls on glaciated criticaltaper orogens. Earth and Planetary Science Letters 262, 385397.
Tulaczyk, S., Kamb, B., Scherer, R.P., Engelhardt, H.F., 1998. Sedimentary processes
at the base of a West Antarctic ice stream: Constraints from textural and
compositional properties of subglacial debris. Journal of Sedimentary Research
68, 487496.
van der Beek, P., Bourbon, P., 2008. A quantification of the glacial imprint
on relief development in the French western Alps. Geomorphology 97,
5272.
Whipple, K.X., Kirby, E., Brocklehurst, S.H., 1999. Geomorphic limits to climateinduced increases in topographic relief. Nature 401, 3943.
Whipple, K.X., Meade, B.J., 2004. Controls on the strength of coupling
among climate, erosion, and deformation in two-sided, frictional
317
orogenic wedges at steady state. Journal of Geophysical Research 109(F01011),

24 pp.
Willett, S.D., 1999. Orogeny and orography: The effects of erosion on the structure
of mountain belts. Journal of Geophysical Research 104, 28 95728 981.
Biographical Sketch
James A Spotila is a Professor in the Department of Geosciences at Virginia Tech, in Blacksburg, Virginia. His
interests are in active tectonics and geomorphology, particularly the behavior of continental deformation systems,
the interaction of tectonics and surficial processes, and long-term landscape evolution of mountain belts. His
specialties include thermochronometry ((UTh)/He dating) and traditional geomorphic techniques.
5.14 Tectonic Aneurysms and Mountain Building

PO Koons, University of Maine, Orono, ME, USA
PK Zeitler, Lehigh University, Bethlehem, PA, USA
B Hallet, University of Washington, Seattle, WA, USA
5.14.1
Introduction
5.14.2
Tectonic Aneurysm: Conceptual Model
5.14.2.1
Terminology
5.14.2.2
Geologic Expression of Tectonic Aneurysms
5.14.2.2.1
Physical setting
5.14.2.2.2
Geomorphic setting
5.14.2.2.3
Metamorphism
5.14.2.2.4
Observations in the fixed reference frame: geophysics
5.14.2.2.5
Timing, rate, and duration of metamorphism and exhumation
5.14.2.3
Discussion and Summary, Geologic Expression
5.14.3
Physics and Boundary Conditions of the Tectonic Aneurysm
5.14.3.1
Solid Earth: Mechanical Conditions for Tectonic Aneurysms
5.14.3.1.1
Rheological and thermal considerations
5.14.3.1.2
Mantle trajectory
5.14.3.1.3
Topographic stress
5.14.3.1.4
Discussion and summary of mechanical effects
5.14.3.2
The Surface Boundary
5.14.3.2.1
Erosional requirements
5.14.3.2.2
Energy sources and erosion
5.14.3.2.3
Fluvial systems
5.14.3.2.4
Glacial systems
5.14.3.2.5
Emergent behavior: Self-organized balance between uplift and erosion
5.14.3.2.6
Summary, surface boundary
5.14.4
Geodynamics of the Tectonic Aneurysm
5.14.4.1
Numerical Models
5.14.4.2
The Corner Model
5.14.4.3
The Generalized Macroscale Himalayan Model (GMHM)
5.14.4.3.1
Strain softening and drainage formation
5.14.4.4
Early Stages of Aneurysm Development: Southern Alaskan Example
5.14.4.5
Geodynamic Summary
5.14.5
Conclusions
5.14.5.1
Model Summary
Acknowledgments
References
Nomenclature
A
material constant inversely related to the effective

viscosity of a power law substance here used to
describe the rheology of the lower crust (eqn [1]
in main text)
radiogenic heat source (W m3) used in the
calculation of the conductive/advective thermal
equation (eqn [2] in main text)
CSEF
EI
Fc
319
321
321
321
321
322
325
327
327
330
332
332
332
334
334
334
334
335
335
335
337
338
340
340
340
340
340
342
342
344
344
344
345
345
Chugach St. Elias thrust fault

Erosion Rate Index derived from slope, area,
and precipitation data
integrated strength of crust that, for convergent
margins, is sensitive to the square of the
thickness of the upper frictional/brittle layer of
the crust (eqn [3] in main text)
Koons, P.O., Zeitler, P.K., Hallet, B., 2013. Tectonic aneurysms and
mountain building. In: Shroder, J. (Editor in Chief), Owen, L.A. (Ed.),
Treatise on Geomorphology. Academic Press, San Diego, CA, vol. 5,
Tectonic Geomorphology, pp. 318349.
318
http://dx.doi.org/10.1016/B978-0-12-374739-6.00094-4
Tectonic Aneurysms and Mountain Building
GMHM
hBD
L
m asl
NBGP
NPHM
Pe
Q
R
T
TRMM
General Macroscopic Himalaya Model. A fully

three-dimensional numerical model including
both Western and Eastern Syntaxes with
thermally-weakened crustal rheology driven by
observed far-field plate velocities
thickness of brittle layer (eqns [3, 4] in main
text)
characteristic length scale of conductive
boundary layer in Peclet number
meters above sea level
Namche BarwaGyala Peri massif of the Eastern
Syntaxis in southeastern Tibet, where the
Tsangpo flows in the gorge between the two
mountains
Nanga ParbatHaramosh massif of the Western
Syntaxis in northwestern Pakistan adjacent to
the Indus river
power-law exponent expressing the dependence
of stress on strain rate (1 linear; 45 B
perfectly plastic; eqn [1] in main text)
Peclet number; Vz L/k, expressing the relative
importance of advection and conduction in
determining the local thermal state
activation energy in the viscous rheological
description (eqn [1] in main text)
gas constant in viscous rheological description
(eqn [1] in main text)
temperature (K)
Tropical Rainfall Measuring Mission provides
spatial and temporal variance of precipitation
319
TSI
Topographic Stress Index that provides an

indication of the proximity to failure of the
upper crust due only to natural topography
and, consequently, describes one of the
nonlinear links among local terrain and
local strain
U
sliding velocity at glacier base used to calculate
energy dissipation along the rock ice interface
vertical component of velocity field
Vz
WHS, EHS Western and Eastern Syntaxes of the Himalaya
e_
shear strain rate in viscous rheological
description (eqn [1] in main text)
j
thermal diffusivity (m2 s1) (eqn [2] in main
text)
kinematic velocity field (m s1) 3D material
v
velocities
differential stress at base of the frictional/brittle
DrBD
upper crust, hBD, defined by the pressuredependent, MohrCoulomb failure envelope
(eqn [4] in main text)
q
density (kg m3)
g
gravitational attraction at Earth surface (m s2)
r
differential stress in viscous formulation of
eqn [1] in main text
components of stress tensor with normal and
rij
shear components identified in a Cartesian
coordinate system
maximum, minimum principal stresses
r1, r3
s
shear stress at glacierbed interface
Abstract
Some of the most spectacular topographies on the planet arise as a product of localized interactions between climate,
erosion, and tectonics. A notable example emerges when climate and tectonic systems mutually influence their thermal
structures. In the atmosphere, this influence can lead to orographic effects that focus precipitation, which is funneled by
rivers and glaciers into spatially restricted zones with a high potential for erosion. Focusing of this erosional power onto the
solid earth at deforming convergent margins then perturbs stress fields and the material properties of the lithosphere,
diverting the three-dimensional tectonic velocity field. The resulting rapid ascent of crustal material produces a hot and
correspondingly weak domain in the upper crust. This weak domain further concentrates strain, generating the zone of
stable, rapid uplift and exhumation that is referred to as a tectonic aneurysm. The tectonic aneurysm is characterized by
extreme relief where rapid exhumation has been sustained (for 3 Myr or more) and very young metamorphic assemblages
are exposed adjacent to exceptionally powerful fluvial (B200 kW m1 or more) or glacial (B40 kW m1) erosional agents.
This characteristic ensemble of metamorphic, surface, and tectonic features, which together constitute the tectonic
aneurysm, is present at varying degrees in the three examples that are reviewed: Nanga Parbat, northwest Pakistan; Namche
Barwa, southeast Tibet; and St Elias, southeast Alaska. Petrologic, geomorphic, and geodynamic investigations of these three
extreme regions, summarized here, illuminate the nonlinear interaction and cooperative behavior of tectonics and surface
processes, which are likely universal, but are most clearly manifested in these regions where the high process rates produce a
high signal-to-noise ratio.
5.14.1
Introduction
Earths regions of extreme topographic relief have exerted a

strong grip on human imagination for millennia and on scientific inquiry for centuries. Since at least Leonardo da Vinci,
observers have recognized that the landscapes of the very large
mountains contain fundamental information on Earth
processes. Generally, the mechanisms of erosion and tectonics have been categorized as being either constructive or
destructive, with landscape representing a temporal snapshot

of this resulting competition (Davis, 1899; Penck, 1924). Although this traditional duality of landscape formation provides a clearly adequate description for many if not most
landscapes, it has become increasingly clear that some distinctive features of landscape result from positive feedback
among tectonic and surface processes and therefore can be
considered more as cooperative phenomena that form as dynamic ensembles, sharing many of the characteristics of other
320
cooperative dynamic entities such as phase transitions or

cloud formation. Cooperative behavior is apparent at many
scales in landscape development, but nowhere is the product
of cooperation so spectacularly displayed as in those regions
where the highest mountains are associated with exceptionally
voracious erosive agents.
To probe the coupled influence of tectonics and erosion,
this chapter explores regions where the topographic and tectonic signals are among the largest recorded on Earth. These
regions exhibit the most extreme topographic relief and exhumation rates on Earth and are associated with some of
Earths most vigorous erosional regimes. In these areas of
coupled high mountains and deep gorges, such as Nanga
Parbat: Indus River, Namche Barwa: Tsangpo River, and Mt. St
Elias: Malaspina glacier, searches for a single causative factor
for the development of extreme relief, as either a product of
tectonics or erosion, have largely been unsuccessful. These
tectonicerosional processes are so deeply interdependent in
these regions that they must coevolve. A spectacular geologic
manifestation of this coevolution is a tectonic aneurysm, a
term we have chosen to provide a vivid sense of the localized,
progressive, self-sustained weakening and failure of initially
strong lithospheric boundary, akin to the failure of an overinflated balloon or artery (Zeitler et al., 2001a). This chapter
reviews the features of the tectonic aneurysm, the physics necessary to produce an aneurysm, and then investigates through
3D mechanical models the evolution of large-scale orogens
influenced by the model aneurysm conditions. En route, it
attempts to clarify outstanding issues of aneurysm behavior,
including their apparent rarity and their relationship with
other models of tectonicerosion interaction such as critical
wedge mechanics and channel flow (e.g., Beaumont et al.,
2001; see also Chapter 5.15).
Considerable prior work on plate tectonics and convergent
orogens provides a rich backdrop for the discussion of tectonic
aneurysms. Plate motions are now well understood in the
context of global geodynamics, but the specific processes responsible for deformation and metamorphism in convergent
settings have remained elusive. One important relatively recent step toward understanding continental tectonics was the
recognition that the surface boundary and Earthsurface
processes can strongly influence the structural and metamorphic history of deforming crust, at both regional and local
scales (e.g., Richardson and England, 1977; Koons, 1990;
Beaumont et al., 1992; Willett et al., 1993). Much of this insight was derived from models that modified the accretionaryprism/thrust belt critical wedge model ( Chapple, 1978; Davis
et al., 1983; Dahlen et al., 1984) itself an extension of Rankine
state theory (e.g., Terzaghi, 1943), to account for deformation
throughout the continental crust, including subaerial regions
where focused erosion can play an important role in reshaping
the orogenic wedge.
Early mechanical models of continental collision assumed
orogen-wide equilibrium conditions (i.e., the entire wedge
and bounding detachment are all at the verge of failure) and
were effective at defining the general relationship between
orogen-scale topography and stress (e.g., Chapple, 1978; Davis
et al., 1983; Koons, 1990; Willett et al., 1993; Whipple and
Meade, 2004). Such models predict that deformation will
preferentially partition away from regions of large topographic
loads into the low regions of an orogen, where the amount of

work against gravity necessary to lift Earths surface is less
(Koons, 1989, 1990). It follows from this prediction that since
strain would steer away from the highest mountains, their
geology should reflect this strain avoidance, through exposure
of relatively shallow metamorphic assemblages, low heat flow,
and relatively old radiometric ages. Geological and thermochronological observations in narrow orogens such as the
Southern Alps of New Zealand are in fact consistent with this
notion; for example, in the Southern Alps the youngest ages
and deepest metamorphic assemblages occur at the lower
elevations adjacent to the Alpine Fault, whereas higher elevations along the Main Divide are characterized by low-grade
metamorphic assemblages and older cooling ages (Adams,
1979; Tippett and Kamp, 1993). Other narrow orogens, conditioned by strongly asymmetric orographic precipitation,
show similar geological and thermochronological patterns,
bolstering confidence that the two-sided orogen provides a
useful mechanical description of terrestrial orogens (e.g.,
Beaumont et al., 1992; Willett et al., 1993).
However, at the same time that orogenic-wedge models
were being developed, petrological and thermochronological
investigations in regions of extreme topography, most notably
the Nanga ParbatHaramosh Massif (NPHM) in Pakistan,
were identifying highly localized concentrations of very
young, high-temperature metamorphic assemblages and decompression melts at high elevations (46000 m; e.g., Zeitler
et al., 1993). It would be difficult to obtain results that were
more inconsistent with predictions from the early versions of
equilibrium models for the two-sided orogen. Over the same
period, understanding of the surface processes that could
couple to tectonics improved in part due to increasing numbers and reliability of measurements of erosion rates from
fluvial and glacial catchments; these data showed that within
many orogens the distribution and rates of surface processes
vary greatly in space (Milliman and Meade, 1983; Bishop and
Shroder, 2000), and some of the highest known rates of erosion were from glaciated basins (Hallet et al., 1996).
At more local scales at specific localities, the mismatch
between observations and equilibrium model predictions
most likely stemmed from these models having an overly
simplified rheological structure, homogeneous boundary
conditions, an underlying assumption of omnipresent mechanical failure (Terzaghi, 1943), or 2D geometry, which renders
them inadequate for addressing a number of significant aspects of orogenic development. These include the generation
of locally extreme elevations, the growth of structures oriented
parallel to plate boundaries, the association of crustal melts
with domains of localized shear, and the anomalous pairing
of deep gorges with high peaks at plate corners where both the
strong structural and topographic gradients are clearly 3D.
Recent geodynamic models have addressed some of these
issues by incorporating nonlinear linkages between the solid
Earth, its surface, and climate in orogenic settings in two and
three dimensions. An important outcome of more fully developed models is that, mediated by the nonlinear nature of
rock rheology, topographic alteration of the crustal stress and
temperature fields can lead to significant feedback between
surface and crustal processes, opening up the interesting
possibility that much more general interactions might exist
between climate, surface processes, and orogen-scale dynamics. These feedback are notoriously difficult to demonstrate using observational data from systems that cannot be
experimentally manipulated, but they can be studied numerically. In fact, models of crustal deformation at several scales
show that in some settings the development of important
feedback is an inevitable outcome of the underlying physics
(e.g., Koons et al., 2002; Beaumont et al., 2001).
5.14.2
Tectonic Aneurysm: Conceptual Model
The conceptual model for the tectonic aneurysm is as follows

(from Zeitler et al., 2001a: p. 7): yin a deforming orogen,
local rheological variations will arise from deep and rapid
incision. The crust will weaken as the strong upper crust is
stripped from above by erosion and the local geotherm is
steepened from below by rapid uplift of hot rock. If this
weakening occurs where the crust is already close to failure, it
will focus particle paths such that local movement of material
will be concentrated into the weaker zone. Provided that efficient erosion continues, a positive feedback develops in
which flow of material into this weakened zone maintains
local elevation and relief, reinforcing the concentrated exhumation and bowing up isotherms, further weakening the
upper crust. This focusing of strain and rapid exhumation leads to metamorphic and structural overprinting of
the crust as high-temperature lower crustal rocks are isothermally decompressed, and also leads to development of
large mountains of limited spatial extent perched atop hot,
weak crust.
The presence of an efficient erosional agent, such as a large,
energetic river along a steep reach, and rapid crustal strain
are important requirements for active aneurysm behavior
(Figure 1). It is important to keep in mind that such behavior
depends on all elements being present: localized, sustained,
and rapid erosion, temperature-dependent rheology that responds to the exhumational steepening of the near-surface
thermal gradient, and the presence of or potential for significant tectonic strain. Further, once all elements are in play and
become coupled, it is in the very nature of feedback that
questions about cause and effect become irrelevant.
5.14.2.1
Terminology
In the discussion that follows, unless otherwise qualified, we

use velocity to indicate particle velocity within a Cartesian
reference frame (x normal to plate boundary; y parallel to
plate boundary; z vertical with up ). We refer to processes associated with Earth deformation related dominantly
to dynamics as mechanical and those associated with surface
denudation as erosional. Throughout this chapter, where
model results of specific physical regions are discussed, we
italicize model features to differentiate from natural ones.
5.14.2.2
Geologic Expression of Tectonic Aneurysms
Currently active aneurysms are not common and these features are not likely to be well preserved in the long-term
321
geological record because postorogenic exhumation removes

the upper portions of the orogenic welt that would be most
likely to record aneurysm development. This section synopsizes the geologic expression of three aneurysms for which
there are extensive observational data and with which we are
most familiar: the Nanga ParbatHaramosh Massif in the
westernmost Himalaya; the Namche BarwaGyala Peri
(NBGP) massif in the easternmost Himalaya; and a proposed
incipient aneurysm in the St. Elias range in southeastern
Alaska (note that the review focuses on the more fully developed Himalayan examples). Other sites of aneurysm development have also been proposed for the Pamir (Kirby et al.,
2007), the Alps (Garzanti and Malusa`, 2008), other locations
in the Pamirs and Himalaya (Pavlis et al., 1997; Thiede et al.,
2004; Montgomery and Stolar, 2006), and even Mars (Kargel,
2004). More detailed information about the well-studied
geology and geophysics of Nanga Parbat can be found in
Zeitler et al. (2001a) and references therein. Many groups have
studied this region extensively, and our brief discussion will
only address the most salient features relevant to the aneurysm
model; the reader is encouraged to survey the literature to gain
an appreciation for the depth of information and range of
opinions that are available. For Namche Barwa, further information can be found in Burg et al. (1998), Zeitler et al.
(2001b), Seward and Burg (2007), Finnegan et al. (2008), and
Booth et al. (2009). For details about the incipient feedback
that might be operating in the St. Elias range, see Enkelmann
et al. (2009) and for an alternate view see Spotila and Berger
(2010).
5.14.2.2.1
Physical setting
Plate scale. The tectonic aneurysms discussed herein are all

located within the indentor-corner portions of large orogens,
near the terminations of orogens where convergent deformation transitions to strike-slip tectonics at the plate edge.
Nanga Parbat is the westernmost peak in the Himalaya,
Namche Barwa the easternmost (Figure 2), and the incipient
southern Alaskan aneurysm is located at the eastern end of the
St. Elias range where the Yakutat terrane collision with North
America shifts from strike slip to convergent deformation
(Figure 2).
Tectonics and structure. Although the general settings of the
three aneurysms are similar (crustal syntaxis above a lithospheric indentor corner), their detailed tectonic contexts are
quite different. Namche Barwa and St. Elias occur quite close
to active-strike deformation associated with the plate edge, but
Nanga Parbat is located at some distance from the inferred
western termination of the Indian lithosphere. Although the
two Himalayan aneurysms are similar in that they expose Indian-lithosphere Precambrian basement that has been exhumed from beneath an overthrust Mesozoic arc terrane
(Figure 2), the overthrust rocks are quite different, with the
Lhasa block in SE Tibet comprising a collage of Gondwanan
basement and Gangdese-arc plutons, compared with the more
mafic plutons that comprise the bulk of the Kohistan terrane
in the Nanga Parbat region. In SE Alaska, Enkelmann et al.
(2009) proposed that aneurysm development is starting to
develop in the Yakutat terrane metasedimentary basement as
well as the Chugach metamorphic complex, inferred to be an
accretionary complex subsequently intruded by plutons.
322
Tr
E
ou vac
t o ua
f s tio
ys n
te
m
an
sv
er
se
riv
ad Rap
ve id
cti
on
ed s in
tr a
Pa
Fo
cu
s
R
er ap
os id
io
n
er
(M
me Anatexis,
tamorphism
C
0
35
T~
alion ition
ns
tr a
le
ict
Fr
cti
du
Figure 1 Schematic representation of the fully evolved tectonic aneurysm. Rapid advection along particle trajectories indicated by arrows leads
to topographic, thermal, and rheological anomalies. The rheology of the pressure-dependent upper crust resting on a thermally activated lower
crust is represented by Ds the transition at 350 1C. Reproduced from Brace, W.F., Kohlstedt, D., 1980. Limits on lithospheric stress imposed by
laboratory experiments. Journal of Geophysical Research 89, 62486252, with permission from AGU.
The Himalayan aneurysms occur as metamorphic culminations within antiformal basement massifs that are remarkably
similar in structural style and size (Figure 3). In plan view,
their outcrop area is small, B50 km or less in diameter.
Basement rocks within the massifs show a steep foliation and
are bounded by steep active faults on one or more sides and
an older, less-active, thrust-sense shear zone elsewhere. In the
case of Nanga Parbat the bounding structures define a compressional pop-up structure of crustal scale (Schneider et al.,
1999); at Namche Barwa there are fewer data and more disagreement over the nature of the boundary structures: for example, Liu et al. (2010) report the northwestern boundary of
Namche Barwa to be a normal fault, but our observations
suggest that the active bounding structures are steep nearvertical reverse faults. The inferred St. Elias aneurysm is buried
by ice, so fine details of its structural geology are not known,
but it appears that the locus of rapid erosion straddles the
Contact Fault (Enkelmann et al., 2009), a reactivated thrust

that is now part of the Fairweather transform fault system
(Bruhn et al., 2004). An important observation in all cases is
that extensional structures do not seem to be responsible for
the great exhumation associated with these features; faults
bounding the Himalayan massif are either reverse faults and
thrusts, or near-vertical faults, and crustal convergence is exceptionally rapid in the St. Elias region according to geodetic
surveys (Elliott et al., 2010).
5.14.2.2.2
Geomorphic setting
Both Himalayan tectonic aneurysms coincide with regions

where the largest rivers draining southern Tibet and the
northern flank of the Himalaya slice deep gorges through
the range as they descend steeply from the Tibetan plateau to
the plains to the south. The exceptional potential of these two
rivers to erode rapidly was clearly illustrated by Finlayson et al.
(a)
(b)
Nanga Parbat: Indus
323
(c)
Namche Barwa: Tsangpo
St Elias: Malaspina
Eurasian plate
Indian plate
Pacific plate
Figure 2 Three tectonic aneuryms: (a) North face of Nanga Parbat massif as seen from the Indus valley (view is from north). The Indus River
in the middle distance (B1100 m above sea level (m asl) and the summit (8126 m asl) define 7 km of relief over o25 km horizontally. (b) West
face of Namche Barwa rising from the Tsangpo in the left at o3000 m asl to the summit (7782 m asl) in a horizontal distance of o12 km. (c)
St. Elias (5429 m asl) B20 km from the Gulf of Alaska with the terminus of the piedmont glacier, Malaspina in the foreground (photo: S.
Campbell).
(2002: 219) using a spatially distributed rate-of-erosion index

(EI) based on models of bedrock river incision ythroughout
the Himalayas. The EI is derived from slopes and drainage
areas calculated from a ydigital elevation model ycombined
with precipitation data yto generate synthetic annual stream
discharges. These variables drive three generalized process
models to produce EI maps that, while differing in detail,
provide an internally consistent, spatially continuous index of
large-scale erosion rates. The tectonic aneurysm centered in
the Eastern Syntaxis is clearly associated with the most distinct
erosional hotspot, which is fueled by perhaps the most energetic river over a substantial region (few thousand square
kilometers) on the planet (Figure 4). This extreme fluvial
power is due primarily to the exceptionally large discharge,
which reflects the vast drainage area in southern Tibet, the
modest river width (100150 m), and the high elevation that
provides the water unusually large quantities of potential energy. The Tsangpo drains over 2 105 km2 and, hence, despite
the relative aridity of southern Tibet (basin-averaged effective
precipitation of only 0.3 m yr1), the discharge of the Tsangpo
approaches 2 103 m3 s1, based on hydrologic data from
1956 to 2001, before slicing through the Himalaya as it drops
over 2 km in a little over 100 km.
Another important dimension of the surface boundary is
that it contains clear and readily assessed clues about the
location, and spatial extent of aneurysms. Adjacent to and
within these regions, local channel characteristics and the
topography contain clear signatures of the active crustal processes associated with large gradients in crustal strain, rock
uplift, and erosion (and bedrock vulnerability to erosion).
These include extreme local relief well in excess of 4 km, high
peak elevations that extend well above the snowline, steep

slopes that sustain rapid mass wasting, and abrupt changes in
the character of river valleys. Exceptionally steep and narrow
river channels conveying large discharges result in massive
stream power dissipation that fuels the rapid incision of deep
gorges with over-steepened walls where uplift is localized
within aneurysms (Figure 5).
In contrast, upstream of an aneurysm, valleys tend to be
broad with wide alluvial channels that reflect the incipient or
intermittent damming of the rivers by the active uplift.
Moreover, the plan-view geometry of river channels may reflect aneurysm behavior. As high crustal strain rates with
considerable variability in magnitude and direction would be
expected within an active aneurysm, the horizontal components of strain would tend to distort the course of river
valleys, making them unusually tortuous. Indeed, Figure 6
shows distinct regions where the Yarlung-Tsangpo is exceptionally tortuous in the Eastern Himalayan syntaxis.
This topographic signature is particularly distinct for tectonic aneurysms that are fueled by fluvial erosion; it is more
subtle in heavily glaciated orogens because the bedrock valleys
and channels are veiled by massive glaciers and icefields. The
distinction between exhumation in orogens by glaciers and by
rivers is blurred, however, by: (1) the significant temporal
changes in ice cover for many orogens during the period of
exhumation (e.g., for currently active aneurysms, consider the
variability in Quaternary glaciations); and (2) the progressive
down valley increase in rates of erosion and sediment transport by water under glaciers, as the flux of subglacial water
increases steadily due to melting and fuels erosion by rivers
that continues beyond glacier margins. During glaciations,
324
70
75
80
85
90
95
35
35
30
30
25
70
Mai
nK
ara
Indian cover
metasediments
Indian basement
kor
um
75
th r
us
t/S
h
yok
s
Arc terrane
(Kohistan/Lhasa)
Suture or
terrane boundary
85
80
utu
r
ez
25
100
Jia
li
In
Su dus
tu
re
Fa
ult
Zo
ne
In
du
s-
Ts
an
gp
ur
po
re
NB
Tsangpo R
Ts
an
gp
I-T
t
su
su
tu
GP
R.
Nanga ParbatHaramosh massif
R.
95
0.01.5
1.52.5
2.55.0
510
1020
2040
> 40
on
e
Indus
Indus
90
Ar-Ar biotite
cooling ages (Ma)
HM
Fault
Ductile shear
zone
100
NP
s
du
In ture
Su
North
25 kilometers
Gyala PeriNamche Barwa massif
Figure 3 Comparison of geological observations of the Nanga Parbat and Namche Barwa aneurysms, drawn to the same scale. The dashed red
line encloses exposures of low-pressure granulites. At Nanga Parbat, the purple regions show the surface projection of anomalously low S-wave
velocities at depth; the green region shows the location of a similar anomaly observed for P-waves. Note that the regions of very young cooling
ages and inferred rapid exhumation have nearly the same area for each locality. Note also the coincidence between young cooling ages, young
UPb ages, low-P granulite metamorphism, the active bounding thrusts, and the geophysical anomalies. Geological observations are from
Schneider, D.A., Edwards, M.A., Kidd, W.S.F., Khan, M.A., Seeber, L., Zeitler, P.K., 1999. Tectonics of Nanga Parbat, western Himalaya:
synkinematic plutonism within the doubly vergent shear zones of a crustal-scale pop-up structure. Geology 27, 9991002; Zeitler, P.K., Koons,
P.O., Bishop, M.L., et al., 2001b. Crustal Reworking at Nanga Parbat, Pakistan: evidence for erosional focusing of crustal strain. Tectonics 20,
712728, and Booth, A.L., Chamberlain, C.P., Kidd, W.S.F., Zeitler, P.K., 2009. Constraints on the metamorphic evolution of the eastern
Himalayan syntaxis from geochronologic and petrologic studies of Namche Barwa. Geological Society of America Bulletin 121, 385407. The
geophysical data are from Meltzer, A.S., Sarker, G.L., Seeber, L., Armbruster, B., Beaudoin, B., 2001. Seismic characterization of an active
metamorphic massif, Nanga Parbat, Pakistan, Himalaya. Geology 29, 651654. For Nanga Parbat, cooling-age data are from compilations by
Schneider, D.A., Zeitler, P.K., Kidd, W.S.F., Edwards, M.A., 2001. Geochronologic constraints on the tectonic evolution and exhumation of Nanga
Parbat, western Himalaya syntaxis, revisited. Journal of Geology 109, 563583; Treloar, P.P.J., Rex, D.C., Guise, P.G., Wheeler, J., Hurford, A.J.,
Carter A., 2000. Geochronological constraints on the evolution of the Nanga Parbat syntaxis, Pakistan Himalaya. Geological Society of London
Special Publication, Tectonics of the Nanga Parbat Syntaxis and the Western Himalaya 170, 137162; George, M., Reddy, S., Harris, N., 1995.
Isotopic constraints on the cooling history of the Nanga ParbatHaramosh Massif and Kohistan arc, western Himalaya. Tectonics 14, 237252,
and UPb data are those compiled in Zeitler, P.K., Koons, P.O., Bishop, M.L., et al., 2001b. Crustal Reworking at Nanga Parbat, Pakistan:
evidence for erosional focusing of crustal strain. Tectonics 20, 712728. For Namche Barwa, cooling ages are those reported in Stewart, R.J.,
Hallet, B., Zeitler, P.K., Malloy, M.A., Allen, C.M., Trippett, D., 2008. Brahmaputra sediment flux dominated by highly localized rapid erosion from
the easternmost Himalaya. Geology 36, 711714, and the UPb data from Booth, A.L., Zeitler, P.K., Kidd, W.S.F., et al., 2004. UPb Zircon
constraints on the tectonic evolution of southeastern Tibet, Namche Barwa area. American Journal of Science 304, 889929; Booth, A.L.,
Chamberlain, C.P., Kidd, W.S.F., Zeitler, P.K., 2009. Constraints on the metamorphic evolution of the eastern Himalayan syntaxis from
geochronologic and petrologic studies of Namche Barwa. Geological Society of America Bulletin 121, 385407, and Burg, J.P., Nievergelt, P.,
Oberli, F., et al., 1998. The Namche Barwa syntaxis: evidence for exhumation related to compressional crustal folding. Journal of Asian Earth
Science 16, 239252.
70
75
80
85
90
95
325
100
35
35
30
30
25
25
70
75
80
85
90
95
0.5 mm yr1
100
25 mm yr1
Figure 4 Index of fluvial erosion rates illustrating schematically the spatial variation in potential erosion rates throughout the Himalaya. The
index is the river power, the product of river discharge and slope; it has been expressed very roughly in terms of erosion rates using a
Himalaya-wide comparison of river power and diverse erosion and exhumation data, assuming a highly idealized range where all other
parameters controlling river incision show insignificant spatial variation on the scale of the range. See Figure 14 for more precise maps of river
power. The river discharge was estimated here from generalized maps of precipitation (Finlayson et al., 2002) and a coarse digital elevation
model (GTOPO-30). Courtesy of D. Finlayson.
major valleys were overwhelmed with sediments produced by

the glaciers beyond their margins, large lakes formed behind
transient dams of ice and debris (and most likely led to
catastrophic outburst floods; e.g., Montgomery et al., 2004),
whereas high divides were probably protected from erosion by
a veneer of cold ice (e.g., Koons, 1989). The collective impact
of glacial fluctuations on exhumation is, therefore, difficult to
characterize precisely, and the role of glacial fluctuations and
periglacial processes in the rapid exhumation of tectonic aneurysms remains largely unexplored.
The postulated tectonic aneurysm in the St. Elias orogen is
active under a much heavier precipitation and generally cooler
conditions than the Himalaya. North Pacific storms pound the
coastline with high winds and wave action, and the heaviest
precipitation in North America falls along the coastal mountains; for example, winter snowfalls summing to420 m are
common along the coastal side of the range. Heavy precipitation combined with high latitude has largely buried the
orogen with glacial ice, making glacial erosion the primary
erosion process in the region (Figure 7; Meigs and Sauber,
2000; Jaeger and Nittrouer, 1999).
Erosion by the glaciers in this region is expected to be
exceptionally fast because they are among the most massive
and fastest temperate glaciers on the planet. For example, the
Bering glacier is the worlds largest temperate glacier with an
area ofB5200 km2 and length of 4190 km; alone, it contains
B15% of the glacier ice in Alaska (Molnia and Post, 2010a).

Measured surface velocities are high, peaking approximately
0.8 m day1 (300 m yr1) during quiescent periods (Bruhn
et al., 2010), and highly variable, reaching peak velocities of
22 m day1 (8000 m yr1) during surges (Molnia and Post,
2010b). Indeed, in the region, basin-averaged current erosion
rates reach and exceed 10 mm yr1 (Figure 8; e.g., Hallet et al.,
1996; Jaeger and Nittrouer, 1999; Koppes and Hallet, 2006)
and exhumation rates reach approximately 5 mm yr1 (e.g.,
Berger et al., 2008; Enkelmann et al., 2011). The products of
current glacial erosion can readily be determined by surveying
the sediments accumulating in the pro-glacial fjords and lakes,
provided subaerial and subglacial inputs of sediments (e.g.,
Koppes and Hallet, 2006; Cowan et al., 2010) are accounted
for where they are significant. The rates of erosion inferred
from proglacial sediment volumes tend to decrease with the
length of time over which they are averaged (Koppes and
Montgomery, 2009; Fernandez et al., 2011). The causes for this
trend, including episodic bedrock erosion and subglacial
sediment storage, are under current study.
5.14.2.2.3
Metamorphism
The Himalayan massifs both expose high-grade gneisses belonging to the Indian lithosphere (Zeitler et al., 1989;
Whittington et al., 1999a, b; Argles et al., 2003; Booth et al.,
2004), based on zircon geochronology as well as geochemical
326
(b)
(a)
(c)
(d)
(e)
Figure 5 Changing character of the YarlungTsangpo. (a) Broad alluvial reach upstream of the Tsangpo Gorge, with wide sand banks visible at
low water. (b) Big Bend reach of the Tsangpo gorge. Satellite scene (c) showing the broad, braided river (seen in (a)) narrowing and speeding up
into a narrow, turbulent channel (near upper right-hand corner of image) hardly visible in the adjacent satellite scene (d). (e) Raging torrent,
over-steepened slopes and deeply incision in the lower portion of the Tsangpo Gorge. Photos (b) and (e) J.-P. Burg.
analysis. Interestingly, due to the high metamorphic grade of

the Nanga Parbat rocks, they had long been assumed to
be correlative with the High Himalayan Crystallines, but
Whittington et al. (1999a) first showed that rocks of the Lesser
Himalayan Sequence must in fact have been the protolith.
This observation underscores the point that specific and localized processes like aneurysm development are important in
crustal reworking and shaping the observed geological record
of orogens.
At both Namche Barwa and Nanga Parbat, extensive
studies of metamorphic exposures within the massif document a high-grade metamorphic event that overprints the
older basement, marked by a sharp decompression path
(Figure 9) and abundant anatectic phases that at Nanga Parbat includes bodies as large as 1 km in size. Studies of the
melts show many of them originated from vapor-undersaturated decompression melting of host lithologies present
within the massif, although localized water-present melting
occurred as well in shear zones, at lower temperatures and
pressures (Butler et al., 1997; see Crowley et al., 2009 for a
thorough review and further references about the extensive
work at Nanga Parbat). Young Namche Barwa melts similarly
are mostly sourced from decompression melting of protolith
along with some fluid-present melting (Booth et al., 2004). At
Nanga Parbat in a focused zone in the center of the massif,
metamorphic assemblages reached cordierite-K-feldspar-sillimanite, low-pressure-granulite grade, overprinting the local
basement. At Namche Barwa, the lower pressure granulite

overprint occurs adjacent to high-pressure rocks in the center
of the massif that are also part of the recent metamorphic
episode (Liu and Zhong, 1997; Booth et al., 2009). Total exhumation associated with the Neogene metamorphic episode
amounted to at least 20 km at both Nanga Parbat and
Namche Barwa, although at Namche Barwa the highestpressure rocks in the core of the massif were exhumed from
depths of nearly 50 km.
Hot springs are abundant in and near the Himalayan
massifs and serve as evidence that rapid exhumation and rock
uplift have steepened the geotherm, especially considering
that some springs are found at anomalously high altitudes
(e.g., 44000 m at Nanga Parbat; Craw et al., 1994). At Nanga
Parbat, fluid inclusions in veins provide evidence for the
presence of a dry steam zone, consistent with rapid decompression of hot fluid (Craw et al., 1997). At Nanga Parbat,
there is active and extensive meteoric-water circulation
through the top 510 km of crust, partly driven by topographic head, which does not appear to penetrate below the
shallow brittleductile transition at 400450 1C; below this
transition the small amounts of fluids present are metamorphic in origin and occur as small unconnected parcels (see
summary by Chamberlain et al., 2002). A broadly similar
pattern seems to exist at Namche Barwa, although definitive
evidence for a dry steam zone has not been observed (Craw
et al., 2005).
4
Elevation (km)
30
3
25
20
15
10
5
0
500
700
900
1100
1300
Tortuosity (deg km1)
35
0
1500
Down river distance (1000-km reach)

Figure 6 Tortuosity defined as the change in azimuth of the river
course per downstream kilometer (5-km running averageblue), and
long profile of the YarlungTsangpo (red). Most irregular reach
coincides with steepest reach of the knickpoint presumably because
both high river power and fast inferred uplift increase with shear
strain rates.
(a)
(b)
Figure 7 (a) Much of the St. Elias range, Alaska, USA is buried in a
sea of glacial ice. In the forefront, the Malaspina piedmont glacier extends
to the Pacific. (b) Mt. St. Elias rises abruptly above the coastline with
heavily crevassed, ~10-km-long Tyndall glacier connecting the western
base of the mountain to nearby Taan fjord. Photograph by B. Hallet.
5.14.2.2.4
Observations in the fixed reference frame:

geophysics
Nanga Parbat received extensive geophysical study using a dense

56-station seismic array (Meltzer et al., 2001) as well as magnetotelluric study (Park and Mackie, 2000). Chamberlain et al.
(2002) integrated these observations with data on fluid flow
327
and hydrothermal activity. One motivation for this geophysical

work was to test whether the observed metamorphic and thermal anomalies are related to contact metamorphic effects associated with a large pluton (they are not), but because of the
high process rates and ongoing activity at Nanga Parbat, the
work provided a unique opportunity to combine observations
made in the moving rock reference frame with those made
using the fixed modern reference frame of geophysics.
Tomographic and receiver-function results. Seismic tomography at Nanga Parbat revealed that within the massif, P- and
S-wave propagation is anomalously slow (B510%) and there
is significant attenuation (Meltzer et al., 2001). These anomalies are confined to a largely vertical section directly beneath
the massif and are consistent with this region being
100200 1C hotter than the surrounding crust. Stations
throughout the massif saw frequent S-wave arrivals, ruling out
the presence of a large magma body, though some stations did
record some events whose arrivals were more harmonic than
impulsive and might hint at the presence of hydrothermal
systems. At Nanga Parbat, crustal thickness locally appears to
be relatively thin, at 40 km (Meltzer et al., 2001) and in
southeastern Tibet, Namche Barwa is located atop a regional
Moho upwarp, with Moho depths below the massif being as
shallow as approximately 58 km than the regional background of B70 km (Zurek, 2008).
Magnetotellurics. Surprisingly, the conductivity of the shallow
mid-crustal rocks responsible for the seismic anomaly is very
low (Park and Mackie, 2000), an atypical result as middle crust
in other active tectonic regions is relatively conductive. The
observed resistivities are consistent with dry granulites. Together
with the tomography, this result suggests that what is being
observed beneath Nanga Parbat is a zone of hot crust, previously dehydrated by metamorphic reactions, that does not
contain large magma bodies or zones of interconnected fluids.
The magnetotelluric data show that the shallowest crust is also
resistive but identifies narrow zones of higher conductivity,
presumed to represent fluids channelized along fault zones.
Seismicity. The seismic array deployed at Nanga Parbat
recorded numerous small earthquakes. Significantly, this
microseismicity is confined to the massif itself and to shallow
depths (Meltzer et al., 2001). The sharp lower cut-off in seismicity at depths of o10 km suggests that the local thermal
gradient is quite steep and that the brittleductile transition
zone is greatly shallowed. The base of this seismicity is bowed
upwards by B3 km across a zone 12 km in width, reaching
within 7 km of the local topographic surface. At its apex this
deflection coincides with the region of highest metamorphic
grade, youngest cooling ages (see below) and, thus, the inferred location of highest exhumation rates.
5.14.2.2.5
Timing, rate, and duration of metamorphism

and exhumation
For both Himalayan aneurysms, there is a fairly extensive body

of work on UPb and ThPb accessory-mineral dating of
metamorphic rocks and granitoid phases. At Nanga Parbat,
UPb dates on matrix monazites suggest metamorphic activity
was underway over the interval 114 Ma (Smith et al., 1992),
and migmatites give UPb zircon ages of 34 Ma (Zeitler et al.,
1993; Schneider et al., 1999). Granite phases give ages
spanning this time range, and reach ages well below 1 Ma
328
100.000
Gl > 10 yrs/5
Glacial/5
Erosion rate (mm yr1)
10.000
Himalaya
1.000
0.100
0.010
B.C., Slaymaker
Global, M&S, 92
0.001
1
10
102
103
104
105
106
107
Basin area (km )

Figure 8 Modern effective erosion rates derived from sediment yields from glaciated basins in coastal Alaska (red symbols) versus unglaciated
fluvial basins worldwide (other colors from global compilation of Milliman, J.D., Syvitski, J.P.M., 1992. Geomorphic/tectonic control of sediment
discharge to the ocean: the importance of small mountainous rivers. Journal of Geology 100, 525544); data were first presented by Hallet, B.,
Hunter, L., Bogen, J., 1996. Rates of erosion and sediment evacuation by glaciers: a review of field data and their implications. Global and
Planetary Change 12, 213235. The red symbols represent estimates of erosion rates over periods longer (large circles) and shorter (small
circles) than a decade. For Alaskan tide-water glaciers, these estimates are conservative because measured contemporary sediment yields were
divided by a factor of five to account liberally for systematic bias associated with current glacial retreat throughout the region (e.g., Koppes and
Montgomery, 2009; Koppes et al., 2009). Large glacial basins generally produce more sediment than other major basins including those
identified as having the highest sediment yields by Milliman and Syvitski (1992), and basins in British Columbia (Slaymaker, O., 1987. Sediment
and solute yields in British Columbia and Yukon: their geometric significance reexamined. In: Gardiner, V. (Ed.), International Geomorphology,
Part 1. Wiley, Chichester, pp. 925945) and Himalaya (Delcaillau, B., 1992. Les Siwalik de LHimalaya du Nepal Oriental: fonctionnement et
Evolution dun Piedmont. Memoires et Documents de Geographie. Editions du Centre National de la Recherche Scientifique, Paris, 205 pp.).
(Zeitler et al., 1993) with a remarkable age of only B0.7 Ma

being reported (Crowley et al., 2009). Although sampling of
Nanga Parbat granitoids has not been comprehensive, the
youngest reported ages occur in the central portions of the
massif, coincident with the metamorphic and seismic anomalies. At Namche Barwa, granitoids and anatectites are not as
abundant in the center of metamorphic massif as at Nanga
Parbat, and Neogene ages on dikes and migmatites have been
reported ranging from B10 Ma to B3 Ma (Burg et al., 1998;
Ding et al., 2001, Booth et al., 2009; Zhang et al., 2011). In this
same massif but at a more northerly location near the peak
Gyala Peri, granites and migmatites are much more common
and reconnaissance dating shows a span in ages from B10 to
B1 Ma (P.K. Zeitler, unpublished data).
Both Himalayan aneurysms are loci of very young cooling
ages. Notably, ArAr biotite ages as young as 1 Ma or less occur
in both settings, and these ages rise sharply across the boundaries of both massifs where these are defined by active brittle
fault zones (Figure 3; Schneider et al., 2001; Stewart et al.,
2008). Other lower temperature thermochronometers show
similar spatial patterns and correspondingly younger ages (e.g.,

Zeitler, 1985; Seward and Burg, 2007; Stewart et al., 2008).
Thermochronological data are hard to model quantitatively in
settings as complex as this because of the large topographic
relief and short-wavelength topography, very high process rates
that possibly involve transients, uncertainties in boundary
conditions such as the temperature of rocks entering the local
system at depth, unknown but probably important role of fluid
flow in redistributing heat in the uppermost crust, and uncertainties in both thermal parameters and diffusion kinetics.
However, overall, thermochronological data document the sort
of sharp acceleration in cooling rates resulting from near surface
conduction and convection that would be expected under
conditions of very rapid exhumation (e.g., Zeitler et al., 1989).
With biotite ages as low as 1 Ma, and zircon UTh/He and
apatite fission-track ages of below 0.5 Ma, cooling rates exceeding 400 1C Myr1 are likely to have pertained over the recent geological past of currently exposed bedrock.
Direct exhumation-rate estimates for the Himalayan aneurysms have been made in a number of ways across a range of
329
16
PT paths
14
All
12
Namche
Barwa
50
Nanga
Parbat
Solidi (NP bulk compositions)
~10 Ma
40
Leucogranite
Migmatite
10 Ma
~6 Ma (matrix monazite)
8 Ma
6
6 Ma
Ky
~1 Ma (Ar closure biotite,
fluid-inclusion trapping)
Sill 4 Ma
30
~1.7 Ma
(dehydration melting)
~3 Ma (matrix titanite)
Depth (km)
Pressure (kbar)
(in-situ monazite)
10
20
~1 Ma (vein crystallization)
13 Ma (granite crystallization)
And
10
~0.7 Ma (leucogranite
crystallization)
0
0
200
400
600
800
1000
1200
Temperature (C)
Figure 9 Summary in PTt space of petrological observations at Namche Barwa and Nanga Parbat. Rocks from both massifs show strong
decompression paths that pass through the low-pressure granulite field, consistent with model predictions for exhumation rates of several
millimeters per year or higher. Nanga Parbat petrological data are from Whittington, A.G., Foster, G.L., Harris, N.B.W., Vance, D., Ayres, M.W., 1999.
Lithostratigraphic correlations in the western Himalaya an isotopic approach. Geology 27, 585588; Whittington, A.G., Harris, N.B.W., Butler,
R.W.H., 1999. Contrasting anatectic styles at Nanga Parbat, northern Pakistan. Geological Society of America Special Paper 328, 129144; PT
observations and model PT path, in green) and from Crowley, J.L., Waters, D.J., Searle, M. P., Bowring, S.A., 2009. Pleistocene melting and rapid
exhumation of the Nanga Parbat massif, Pakistan: age and PT conditions of accessory mineral growth in migmatite and leucogranite. Earth and
Planetary Science Letters 288, 408420 (UPb dates, PT data and inferred PT path from two migmatite samples, magenta and blue). Namche
Barwa petrological and UPb data are from Booth, A.L., Chamberlain, C.P., Kidd, W.S.F., Zeitler, P.K., 2009. Constraints on the metamorphic
evolution of the eastern Himalayan syntaxis from geochronologic and petrologic studies of Namche Barwa. Geological Society of America Bulletin
121, 385407 (orange). The filled boxes at low temperatures and pressures are fluid-inclusion observations from Craw, D., Koons, P.O., Zeitler, P.K.,
Kidd, W.F.S., 2005. Fluid evolution and thermal structure in the rapidly exhuming gneiss complex of Namche BarwaGyala Peri, eastern Himalayan
syntaxis. Journal of Metamorphic Geology, 23, 829845 (Namche Barwa, orange) and Winslow, D.M., Zeitler, P.K., Chamberlain, C.P., 1994. Direct
evidence for a steep geotherm under conditions of rapid denudation, western Himalaya, Pakistan. Geology 22, 10751078 (Nanga Parbat, green).
Solidi were calculated by Crowley et al. (2009) for bulk compositions of a Nanga Parbat leucogranite and two different migmatite samples.
timescales. UPb and ThPb dating of metamorphic and anatectic phases, combined with PT estimates, suggest long-term
net exhumation rates of 46 mm yr1 or more maintained
over 510 Myr and 20 km of exhumation or more, occurring
across a fairly limited areal extent of some 25 25 km
(Figure 10; Zeitler et al., 1993; Booth et al., 2009). For Nanga
Parbat, Crowley et al. (2009) have refined this estimate and
argued that their data provide evidence for an acceleration
in exhumation rates to 1113 mm yr1 since 1.7 Ma, with
correspondingly lower rates of 12 mm yr1 active before then.
At intermediate timescales, PT estimates for fluid-inclusion
trapping together with thermochronologically derived cooling
histories provide another longer term estimate. For Nanga
Parbat, Winslow et al. (1994) combined such data to yield an
estimate of 36 mm yr1 for the past 1.0 Myr. Cosmogenicisotopes studies of bedrock terrace incision and basin-wide
erosion rate, integrated over thousands of years, yield estimates
of 12 mm yr1 for the interior of the Nanga Parbat massif
(Burbank et al., 1996) and nearly 4 mm yr1 for small basins
near to but not within the Namche Barwa massif (Finnegan

et al., 2008).
Sediment-production studies at Nanga Parbat yielded
values of approximately 5 mm yr1 or more for a shorter, essentially modern period (Shroder and Bishop, 2000). For
Namche Barwa, what amounted to thermochronological
provenance studies of detrital zircon fission-track ages gave an
estimate for recent mean erosion rates of 721 mm yr1
(Stewart et al., 2008), recently supported and refined by
Enkelmann et al. (2011) to 517 mm yr1. Finally, although
rigorous inversion of cooling-age data for exhumation rate is
fraught with unquantified uncertainties, using a simple 1D
steady-state assumption (e.g., Reiners and Brandon, 2006), the
youngest cooling ages in the Himalayan aneurysms suggest
exhumation rates of 510 mm yr1 over the past 1 Myr.
Determining duration of activity and the timing of initiation of the Himalayan aneurysms is not an easy task, for two
reasons. First, given the very large amounts of exhumation,
cooling-age data can only provide a bare minimum estimate
330
Figure 10 Tectonic sketch map of SE Alaska, summarizing results from fission-track dating of detrital zircons. Significant populations of young
zircons are present only in sand samples from the margins of the Malaspina glacier, suggesting a very localized source area. Warping of
structural trends shows location of the region above the indentor corner developed due to the collision of southern Alaska with the Yakutat
terrane. Reproduced from Enkelmann, E., Zeitler, P.K., Pavlis, T.L., Garver, J.I., Ridgway, K.D., 2009. Intense localized rock uplift and erosion in
the St Elias orogen of Alaska. Nature Geoscience 2, 360363, with permission from Nature.
as rocks that might have contained a signal of older pre-aneurysm cooling have been eroded away. Second, geodynamic
models for aneurysms (see below) raise the possibility that
when they are active, rock exposures within them will reflect
something of a petrologic and geochronological steady state,
with rocks moving through the system at similar rates and
experiencing similar styles and timing of overprinting deformation and metamorphism. Thus, the oldest high-temperature ages of metamorphism and granitoid emplacement
might still be minimum estimates. Keeping that caveat in
mind, at both Namche Barwa and Nanga Parbat, the distribution of ages (see above) suggests that anatexis began at
about 10 Ma, with accessory-mineral metamorphic ages and
emplacement ages for decompression melts documenting
decompression since 10 Ma, to an increasing degree (recall the
argument by Crowley et al. (2009) that exhumation at Nanga
Parbat has greatly accelerated since 2 Ma). At both sites, dating
of anatectic phases suggests decompression has been underway for at least the past 34 Myr, and the PTt results from
Namche Barwa (Booth et al., 2009) would argue that rapid
decompression was underway at 6 Ma or before.
In southern Alaska, the main evidence for the proposed
aneurysm is derived from fission-track analysis of detrital zircons, as direct sampling of low-elevation bedrock is severely
hampered by the thick ice cover (Figure 11). Enkelmann et al.
(2009) found very young ages for a substantial fraction of the
zircon in the bed load of rivers emanating from the Malaspina
glacier, with a youngest age of 0.4 Ma and a modeled peak at
23 Ma. Given known age relationships in the region and the
source area for Malaspina sediment, the likely source for these
zircons is a small region in the hinterland. Because the closure

temperature of such young, damage-free zircons is quite high,
possibly exceeding 300 1C, simple incision into preexisting
topography could not produce such young ages, arguing that
these zircons are sourced from a region of focused rock uplift
and rapid erosion. Another way to get such young ages with
lower erosion rates would be to have a very steep thermal
gradient, but the only plausible way to get such a steep gradient would in fact be to have an extended period of rapid
rock uplift and balancing erosion. The coincidence of this
focused rock uplift with a model prediction of high rock uplift
led Enkelmann et al. (2009) to propose this part of the St Elias
range to be an incipient tectonic aneurysm, though Spotila
and Berger (2010) argued that the young ages could simply
result from rapid erosion into an extruded transpressional
slice along one of the local strike-slip faults, with no feedback
with erosion being required.
5.14.2.3
Discussion and Summary, Geologic Expression
Active, mature tectonic aneurysms express themselves most

clearly through the coincidence of a number of geological
anomalies and features, all distributed over a relatively small
structural massif of only tens of kilometers in lateral extent
bounded at least in part by active steep reverse faults. These
coincident features include very young high-temperature cooling ages, thermal anomalies in the upper crust that manifest
themselves as hot springs and meteoric fluid flow, a young
low-pressure high-temperature granulite facies zone marked by
abundant granitoid melts, some of which originate in vapor-
Fero
exhumcaious
tion
7
Overthrust arc
450 C
b/d tra
ion
n sit
Orogencrossing
river
Weak
4
15
km
(a)
(b)
Figure 11 (a) View of 1 Ma leucogranite pluton exposed at Nanga

Parbat, western wall of Raikot valley. The granite is about 1 km in
lateral extent. (b) Young granite dikes intruding 1.8 Ga Nanga Parbat
basement, Rupal valley, south side of Nanga Parbat massif.
undersaturated decompression melting, and very high peaks

and extreme relief that is coincident with exceptionally energetic glacial or fluvial erosion. In the massif, exhumation rates
are very high, over 5 mm yr1 and were maintained over several
millions years, resulting in tens of kilometers of net exhumation
by erosion because evidence for extensive low-angle normal
faulting is lacking. Figure 12 illustrates some of these features in
the context of an idealized cross section.
We stress that the tight colocation of this set of diverse
features, which would ordinarily be considered quite unrelated, must be an emergent phenomena, an expression of
self-organization resulting from feedback that we infer based
on the observations reviewed. Although Namche Barwa and
Nanga Parbat share all these features, their detailed tectonic
contexts are not identical (e.g., orientation and location of
active structures with respect to the details of both the regional
tectonic features and the local drainage and topography and
regional tectonic features). Given variations in initial boundary conditions alone, we would not expect each aneurysm to
appear identical, and in our view it is the numerous observed
similarities that are remarkable.
The observation of young cooling ages was what first drew
attention to Nanga Parbat as a potentially unusual feature
(Zeitler et al., 1982), and it is worth making a comment about
such data. Finding young ages in active orogenic belts should
not be difficult, particularly for very low-temperature systems
Dehydration
melting
Devolatilization?
0
km
331
Orogenic
basement
30
km
v:h=1
Figure 12 Geologic manifestation of a fully developed tectonic

aneurysm, based on observations at Nanga Parbat and Namche
Barwa. In the presence of regional compressive strain (1), older,
previously metamorphosed basement rocks enter the local system
and encounter high strain associated with deformation within the
aneurysm (2), possibly leading to additional devolatilization. As these
rocks are rapidly uplifted along steep reverse-fault systems they
experience rapid decompression and undergo dehydration melting
(3), and low-pressure metamorphism (4), leading to formation of
migmatites and granite dikes and bodies. Rapid advection of rock
significantly uplifts isotherms and the brittleductile transition (5),
steepening the geotherm and locally weakening the crust. In
conjunction with rapid erosion, this results in the development of
high mountains atop weak crust (6); this high topography and large
relief in turn supports rapid erosion (7), by glacial carving and mass
wasting. Eroded material is carried out of the system by large
orogen-crossing rivers. The steep geothermal and large topographic
head drive extensive meteoric-water circulation systems (8) above
the brittleductile transition; a small fraction of this water may be
kneaded to greater depths along shear zones (9), and could
contribute to melting and vein development.
such as UTh/He dating of apatite, which closes to diffusion at

temperatures of below 70 1C. Isotherm compression beneath
valleys, the tendency of isotherms to become significantly
advected upward at exhumation rates much above 0.5 mm
yr1, the ability of fluids to redistribute heat, and the potential
for deep and rapid glacial and fluvial incision all make young
ages likely. Thus, it is the higher temperature dating systems,
such as ArAr analysis of biotite, or fission-track dating of
young zircons lacking radiation damage, that are most diagnostic of unusual or intense geologic activity. In the most
simplistic view, observing young ages for systems that close to
diffusion at well above 300 1C requires that a considerable
amount of rapid exhumation has taken place. Of course, such
young ages could also reflect a very steep thermal gradient, but
the principal way to develop such steep gradients over a significant region is for exhumation to be rapid for at least a few
million years, so in either case, young high-temperature ages
imply rapid and considerable exhumation. Equally important
consequence of such rapid exhumation is upward advection of
the brittleductile transition zone, and thus thermal thinning
of the crust of the type called for by the aneurysm model.
One other observation about the Himalayan aneurysms is
that relative to their immediate surroundings, the location of
these features must have been relatively stable. This is
332
particularly relevant to Namche Barwa, where given the massive knickpoint on the Tsangpo, exposures in the massif and
phenomena like the high heat flow and young cooling ages are
arguably related to vigorous headward cutting by the Tsangpo;
in the absence of localized uplift closely synchronized to
compensate for the fluvial cutting, the knickpoint would migrate upriver at rates on the order of 100 km per million years
(see Section 5.14.3.2.3). However, several lines of evidence
argue for stability and against any sort of migration. First, the
petrological data show that rocks at both massifs have recently
been exhumed from depths (420 km) that far exceed even the
great relief present in these features (57 km), so simple
headward cutting would not account for the deep exhumation. Second, both massifs were overthrust by distinctive
arc-terrane rocks that contrast sharply with the basement rocks
exposed in the massif. Any persistent migration of aneurysm
phenomena at either of the massifs would be revealed by
a trail of basement rocks, and this is not observed (e.g.,
Figure 3). If anything, one could interpret the patterns of
deformation and age relationships at both massifs to support
an inward-focusing of strain and high-grade metamorphic
activity (e.g., Schneider et al., 1999).
5.14.3
Physics and Boundary Conditions of the

Tectonic Aneurysm
qT
kr2 T !
v rT ax,y,z
qt
The essence of the tectonic aneurysm lies in the dual positive

feedback mechanisms that relate erosion rates to crustal
strength via:
1. the influence of topography and surface processes on the
distribution of strain during continental deformation
through both stress- and strain-related processes; and
2. the sensitive influence that vertical displacements of rocks,
relative to Earths surface, have on the temperature and
therefore the strength distribution of the upper crust.
This positive feedback between vigorous surface processes
and tectonics produces the distinctive behavior of tectonic
aneurysms. Conditions that must be met for such a feature,
therefore, arise in both the tectonic and erosional domains; in
the following we investigate the necessary and sufficient conditions for aneurysm growth.
5.14.3.1
5.14.3.1.1
Solid Earth: Mechanical Conditions for Tectonic

Aneurysms
Rheological and thermal considerations
For vertical displacements to significantly influence orogen

material strength, the crustal rocks involved must exhibit two
properties:
Requirement I: For appropriate conditions of pressure, temperature, and water activity, the crustal rheology must exhibit an inverse dependence of flow resistance on
temperature; for example,
e_ ADsn eQ=RT
exponent, 1 linear; 45Bperfectly plastic; R the gas

constant; T the temperature (K); and Q the activation
energy. The widely accepted form of quartz/feldspardominated crustal rheology (Brace and Kohlstedt, 1980)
meets the requirements of upper crustal pressure-dependence underlain by mechanically weaker material where the
weakness is generally a function of thermally activated
ductile slip systems (Figure 13). The vigorous debate on
continental crustal rheology, addressed by Handy et al.
(2007) is outside the scope of this chapter; we will assume
that the initial rheological structure of the continental crust
follows the general BraceKohlstedt form of Figure 13 and
eqn [1]). Although the details of rheological models might
vary, the assumption of such a temperature-dependent
strength inversion is reasonable, and this inversion is
crucial to aneurysm behavior. Although this inversion can
be produced by other material strength profiles such as salt
layers beneath silicate rocks (Montgomery and Stolar,
2006), we focus here on the thermally dependent inversion
associated with silicate materials.
Requirement II: The temperature field must be a function of
the kinematics; that is, the advective term in the following
equation ( rT) must dominate the thermal field and be
sensitive to the surface boundary conditions
where k is the thermal diffusivity (m2 s1), r the kinematic velocity field (m s1), and a the radiogenic source
(W m3).
At the simplest level, in order to thermally weaken the
upper crust, the advection of heat by material transfer from
the deeper, warmer part of the continental crust must be
greater than the rate of conductive cooling at the upper
surface as expressed in the 3D transient conductive
advective Fourier equation (eqn [2]; e.g., Albare`de, 1976;
Koons, 1987; Batt and Braun, 1999). For time frames in
which heat-production rates are negligible, this critical
relationship can be approximated for near steady-state
systems (e.g., t4B1 Ma) by the Peclet number, Pe VzL/k
(Vz is the vertical component of velocity field and L is the
characteristic length scale, here equated with the depth to
detachment) (Carslaw and Jaeger, 1959). Peclet values
4B 0.5 (e.g., Vz43 mm yr1; L415 km; kBcrustal silicates B3 106 m2 s1) indicate that there will be significant upward advective displacement of isotherms and of the
thermally dependent transition to dominantly ductile,
warmer, and weaker middle crust.
The strength of the upper continental crust originates in its
pressure-dependent frictional rheology, and the thickness of
this strong lid (hBD) is a critical parameter in determining the
overall integrated strength of the crust (Fc), which is proportional to the square of the thickness of this lid (Sonder and
England, 1986)
1
where A is the material constant inversely related to the

effective viscosity; s the differential stress; n the power-law
2
Fc D
hBD
Dsdz
0
hBD DsBD
2
3
Effect of transverse gap in topography

Crustal strength reduced by
Z (km)
High-strength
brittle layer
10
Topographic stress index (TSI)

along Himalayan front
~85%
= 2 hBD
Standard crustal
strength profile
hBD
70
20
80
85
90
95
100
FC ~ 85% of standard
35
35
30
30
= f(gh)
Ductile lower
layer
= f(e 1/T)
(a)
25
Effect of transverse gap and exhumation
Crustal strength reduced by ~ 50%
= 2 hBD
0
Z (km)
75
= Strength profile
beneath gap
h1
333
10
20
Standard crustal
strength profile
hBD
h
1
High-strength
brittle layer
= f(gh)
= Strength profile beneath

gap after thermal
thinning of 5 km
FC ~ 50% of
standard
Ductile lower
layer
= f (e 1/T)
(b)
70
75
80
85
85
86
90
95
25
100
30
29
28
27
26
82
83
84
87
88
89
(c)
Figure 13 Topographic and exhumation influence on strength and stress state of the upper crust. (a) Diagrammatic influence of a single incised
valley of 4 km depth on the stress state beneath the valley. (b) Continued exhumation into the eroding channel advectively thins the high-strength
upper crust producing aneurysm rheology (Koons et al., 2002). (c) Topographic Stress Index (TSI) along central Himalayan Front calculated
B3 km below sea level from GTOPO 30 DEM. After Koons and Kirby (2007). TSI in a region provides an indication of the proximity to failure as
a function only of the stresses produced by topography (i.e., 0no contribution to stress state from topography; 1Bat failure due to
topographic stresses). The Front-parallel pattern of high TSI is augmented by Front-normal concentrations that coincide with the large transHimalayan rivers and encourage the formation of river anticlines along those channels. Reproduced from Montgomery, D.R., Stolar, D.B., 2006.
Reconsidering Himalayan river anticlines. Geomorphology 82, 415.
The contribution of the frictional thickness to the overall integrated strength may be approximated as
rgh2BD rFc r2rgh2BD
4
depending on the yield strength of the ductile material (Koons

et al., 2002; Koons and Kirby, 2007).
In a convergent orogen, thermal thinning of the uppercrustal lid by advection rapidly leads to dramatic strength reductions and sharp localization of strain. Consequently, for
high-Peclet systems, the integrated strength of the crust is reduced on temporal and spatial scales that are defined by advection rates. A 50% reduction in integrated strength can occur
after only B1 Myr of vertical displacement of lower crust at
rates approaching 5 mm yr1 (Koons et al., 2002; Figure 13).
Thermal reduction of this frictional crustal layer therefore
constitutes the proximal cause of the tectonic aneurysm.
An important parameter controlling the maximum temperatures attained in an aneurysm is the temperature at which
mechanical detachment occurs, as this transforms a component of geologically rapid horizontal plate convergence into
rapid motion toward the surface. Because in active orogens the
characteristic time of tectonically driven advection is short
relative to that of radioactive decay, rapidly advecting material
generally does not have time to heat significantly during its
traverse through the crust and consequently carries its thermal

state with it into the cooler upper part of the crust. As observed
in the Himalayan aneurysms (see above), deeper levels of
detachment will produce a large-magnitude decompression
PT path and a significant shallow thermal anomaly that can
impact the petrology of rocks exhumed within the aneurysm
through metamorphism, vapor-absent decompression anatexis, and fluid flow; all of these are superimposed on the
mechanical and rheological phenomena that result from the
steepened upper-crustal thermal gradient, and potentially
feed back into strain through the influence of fluids and melts
on rheology. For regions in which shallower detachments
root at depths with relatively low temperatures (e.g., southern
Alaska, Koons, 2009), petrological effects will not be as
pronounced.
Although most of observations that led to the tectonic
aneurysm model originated from regions that have experienced significant thermal thinning and that approach thermal
steady state, thermal weakening can occur well before steady
state is approached. The greatest strength of the upper crust
resides at the base of the pressure-dependent frictional lid
(Sibson, 1982; Handy et al., 2007), which is the part of the
frictional crust most distant from the surface (B15 km), so
significant thermal thinning and resultant lithospheric weakening can occur before the base of the brittle crust begins to
334
cool and before a observable heat flow anomaly can be near

the surface. These early stages of strain concentration, examined numerically by Simpson (2004), are relevant to the incipient tectonic aneurysm in southern Alaska discussed above.
Under these far from steady-state conditions, the degree of
weakening is a function of total displacement and is dominated by the vertical component.
5.14.3.1.2
Mantle trajectory
In addition to focusing of vertical flow by erosional removal

described above, changes in mantle trajectory significantly
alter the vertical velocity of the upper crust. Consequently,
shallowing of the mantle alters the total amount of crust
displaced into the deforming zone and, if erosion is sufficiently vigorous, can enhance exhumation and alter the
thermalrheological structure of the upper crust. In the geodynamic models examined below Nanga Parbat, Namche
Barwa, and St Elias, we incorporate the mantle trajectory
identified from seismic and gravity investigations.
5.14.3.1.3
Topographic stress
In an orogen, an important nonlinear interaction can arise

between deformation and topography because large-scale relief will alter the orogens stress distribution and induce strain.
Generally, in a convergent margin where the dominant principal stress (s1) is primarily associated with the horizontal,
increases in stresses generated by vertical load (szzBrgh) reduce the differences of principal stresses (s1s3), moving the
local state further from failure and decreasing the tendency for
permanent strain. However, the coordinate stresses in the
valleys influenced by the topographic slopes (siz, sxx, syy)
generally tend to move the state of the crust closer to the
failure envelope, thereby increasing the potential for strain.
Although the relative importance of each stress component is
dependent on the location relative to the topography and the
relative orientation of slope and tectonic stresses (McTigue
and Mei, 1981; Liu and Zoback, 1992; Koons, 1994; Enlow
and Koons, 1998; Koons and Kirby, 2007), relief leads to stress
heterogeneity in the upper crust that can significantly influence strain in three dimensions. The magnitude of this effect
can be approximated through the Topographic Stress Index (TSI)
introduced to evaluate the theoretical proximity to failure of
the crust due to solely topographic stresses (Koons and Kirby,
2007). The TSI, varying between 0 (no topographic stresses)
and 1 (topographic stresses alone are sufficient to cause failure), provides a readily calculated index for natural topography and is analogous to the Factor of Safety measure long
employed for evaluating slope stability. The magnitude of the
effect of relief on local stress and the ability to influence strain
generally decay with depth and are predominantly a function
of the amplitude and wavelength of the relief (e.g., McTigue
and Mei, 1981; Miller and Dunne, 1996; Koons et al., 2002).
To perturb the tectonic strain pattern significantly, the stress
perturbation generated by topographic features (ridges or
valleys) must extend down through the highest-strength part
of the brittle crust, to near its base (Koons and Kirby, 2007).
This requires that the wavelengths of valleys (or ridges) exceed
the thickness of the pressure-dependent, strong uppercrustal lid.
The influences of topographic stresses on tectonic strain are

enhanced if the local tectonic stresses are sufficient to bring the
orogen close to failure. Consequently, topographic perturbations that are located in proximity to the high-strain edges of
orogens exert maximum effect on the strain state. Modern
examples of this effect are visible along the Himalayan Front
(Figure 13; Koons and Kirby, 2007). Along the arc parallel to
the front, TSI approaches unity due to the topographic stresses
generated by the dominantly south facing slopes of the
Himalaya. In addition, the major transverse rivers cutting
the Himalayan front have sufficiently long wavelength with
high-amplitude ridge-to-valley relief that TSI normal to
the plate boundary along these gorges also approaches one
(e.g., 46000 m EverestArun). Consequently, topographic
stresses along planes both parallel to and normal to the
IndianEurasian plate boundary are concentrated both at the
orogen front where tectonic strain rates are highest, as well as
at right angles to the plate boundary within the river anticlines occurring along the Himalayas major transverse gorges
(Lave and Avouac, 2001; Montgomery and Stolar, 2006).
5.14.3.1.4
Discussion and summary of mechanical

effects
Tectonic aneurysms arise from feedback associated with

coupling between both rheology and kinematics, and topography and stress. The rheologicalkinematic coupling is
dominated by the inverse dependence of material strength on
temperature above B350 1C together with the dependence of
the local temperature field on the velocity field. The critical
components of the deforming region that determine the
amount of advective weakening are the vertical velocity and
the temperature at which material is detached. For incipient
aneurysm development, the critical weakening contribution
derives from the total vertical displacement, or the product of
the advective term (eqn [2]) and time.
All topographic relief influences the upper-crustal stress
state, but along convergent margins the local topographic
magnitude and orientation of the extreme relief are capable of
bringing the local stress state close to failure throughout the
high-strength crustal lid. The proximity to failure along these
margins is further enhanced where large rivers normal to the
convergent margin breach the orogen, leading to concentrations of stress at these intersections as indicated in the high
values of TSI in Figure 13 where the Arun river crosses the
Himalayan front. These stress concentrations combine with
the rapid localized incision and localized ascent of hot and
weak crustal material to sustain an aneurysm. Early in the life
of an aneurysm, details of exhumation and topography will be
important in localizing strain, but once significant advective
weakening has occurred, thermal weakening and diversion of
advective flow will dominate, and the resulting aneurysm will
be very stable in time and space.
5.14.3.2
The Surface Boundary
In the discussion above on the influence of topography on

vertical velocity, we implicitly assumed that erosional processes are able to maintain elevations and the distribution of
topography within a continuously deforming orogen. Here we
examine the conditions required for this assumption to be

correct in a region of rapid uplift.
5.14.3.2.1
Erosional requirements
Two critical components of the aneurysm model are a means

of efficient erosion at the surface boundary, and the evacuation of this eroded mass away from the aneurysm. This erosion is crucial in developing topography and relief, and in
creating and then maintaining the steep thermal gradient that
leads to rheological weakening of the upper-crustal lid. Efficient local erosion processes are required that keep pace with
the rapid rock uplift within the aneurysms of several millimeters per year or more (this requirement comes from the
high Peclet numbers needed to advect isotherms upward,
thereby thermally weakening the crust). In addition, sufficient
energy is required in the erosive system such that it can flush
eroded debris away. In the examples presented earlier, at the
site of the proposed aneurysms there has been 1030 km or
more of crust exhumed over the past few million years, and if
not evacuated, this material would inhibit or prevent sustained local uplift. Moreover, the erosion and uplift must be
closely synchronized with one another presumably because of
the tight coupling between crustal and surface processes. If the
exhumation were not closely compensated by localized uplift,
erosion would cease long before it excavated a hole tens of
kilometers deep. However, if uplift was not offset by erosion,
uplifted domains would disrupt or block drainages. Thus,
aneurysm development requires a tight coupling between exhumation and uplift, and ample erosive power fueled by a
large, elevated drainage system plumbed to the external portions of the orogen.
5.14.3.2.2
Energy sources and erosion
At the simplest level, the vigor of rivers and glaciers in eroding

an orogen and evacuating crustal mass increase with the total
amount of energy available for doing work on the surface.
The source of this energy is the potential energy contained in
the water and ice that are delivered to the topography by the
hydrologic cycle. The amount of gravitational potential energy
available for erosion per unit time is simply a function of the
regions precipitation and elevation above a base level.
5.14.3.2.3
Fluvial systems
In fluvial basins, the potential energy contained in the water

and the associated erosion and sediment transport are all
sharply focused along a network of rivers. As a result, the rate
of bedrock incision by rivers depends sensitively on the
drainage basin size and geometry, as well as the precipitation
and the river gradient. In addition, the rate of incision depends on a multitude of site-specific factors that control the
efficiency of individual erosive processes. These include the
resistance of the bedrock to erosion, the amount of debris in
the system with its dual roles of providing tools for eroding
bedrock and shielding the bed, the channel width and shape,
and time variation in discharge and debris transport (e.g., Lave
and Avouac, 2001; Sklar and Dietrich, 2004, 2006; Finnegan
et al., 2008; Lague, 2010).
We now consider the essence of the spatial variation in erosion rates that we would expect over vast regions at the simplest
335
level, making two major assumptions. First, we assume that incision rates by rivers scale simply with the fluvial energy available
per unit time (fluvial power) and, hence, that the important sitespecific factors mentioned above do not, in general, confound
the geographic variation in erosion rates simply inferred from
energetic considerations. The site-specific factors directly control
absolute rates of erosion, and hence consistent with Finlayson
et al. (2002), the rates derived from energy consideration are best
viewed as erosion rates indices (EI; e.g., Figure 14); they would
require region-specific empirical calibration to yield actual erosion rates. Second, we assume that the pace of erosion of the
entire landscape is set by the incising rivers, which is to be expected for high-relief regions closely approaching a topographic
steady state where hillslopes have reached threshold angles and
pervasive mass wasting lowers the hillslopes and divides at the
same rate that the rivers cut through bedrock (e.g., Burbank
et al., 1996). The case of tectonicerosion coupling related to the
bedrock being more easily eroded along high strain zones is
considered briefly below.
Fluvial power serves as a very useful, convenient index of
fluvial erosion rates through the entire Himalaya. It can be
estimated for large regions using the river slope, which can be
obtained from digital elevation models (DEMs), and the discharge, which can be derived from DEM-based drainage areas
and precipitation estimates from remotely sensed data where
direct measurements are not available. For example, Anders
et al. (2006) estimated annual river discharge for much of the
Himalaya using Tropical Rainfall Measuring Mission (TRMM)
data and DEM; their estimates match closely all measured
Himalayan discharges. Figure 14 illustrates the spatial variation
of the fluvial erosion rate index for both the B1000-km-broad
regions that extend to each end of the Himalaya. This figure,
and similar ones presented by Finlayson et al. (2002), highlights the highly localized nature of the exceptionally rapid
erosion expected in the Himalayan syntaxes as a direct consequence of the discharge of the Tsangpo (Brahmaputra) and
Indus rivers being very large, much larger than that for other
Himalayan rivers that all drain markedly smaller drainage
basins. Rivers in these smaller basins tend to be steeper than the
syntaxial rivers, but not sufficiently so to compensate for their
much lower discharge. One potential exception is the Sutlej,
one of the major tributaries of the Indus, where the fluvial
power approaches the highest values along the Indus. It is
noteworthy that the reach of the Sutlej with high fluvial power
coincides closely with the region known to be characterized by
rapid exhumation defined by apatite fission-track ages of less
than 2 Ma (Thiede et al., 2005).
Fluvial power per unit length is probably the best index of
erosion rate as it is the total amount of energy needed to
transport the sediment that arrives in a reach of the river, as
well as to erode the bedrock in that reach. Power per unit area
is less likely to be a useful erosion index given the sediment
evacuation requirement. For example, it would be high for a
very steep and narrow river channel; notably values as high as
30 kW m2 have been reported for a B2-km reach of a relatively small, very steep 10-m wide river crossing a fault with
estimated throw rates of 1 mm yr1 in the central Italian Apennines (Whittaker et al., 2007). This erosion index would not
serve as a good measure of incision rate for a much bigger
system in a deep canyon with steep long hillslopes; a river with
336
Figure 14 River power per unit river distance (in kW m 1) along principal rivers in the Himalaya. (a) Orange and red colors signal location and
extent of the eastern Himalayan tectonic aneurysm, a highly localized domain (roughly 0.5% of B106 km2 region) centered on Namche Barwa
that coincides with known area of rapid exhumation (see Figure 19). (b) River power is also exceptionally high along main Indus (northern most
river) where it slices through the western tectonic aneurysm, which includes Nanga Parbat. Energetic river reaches also occur along the
southern-most tributary, which is the Sutlej. Interestingly, these reaches extend from about 77 to 791 E, which coincides with a region of rapid
exhumation defined by apatite fission-track ages of less than 2 Ma (Figure 1 in Thiede et al., 2005). River power computed from annual
discharge estimated from 8 years of TRMM precipitation data, and 4-km averaged river slopes from ASTER DEMs. Courtesy of H. Greenberg.
a modest discharge would have limited capacity to transport

and evacuate the sediment load, in which case the bed would
be shielded from erosion. The total fluvial power per unit
length is the better index because it reflects more precisely the
greater sediment transport capacity of larger rivers. These two

erosion indices are simply related, however, dividing the fluvial power per unit length by the river width yields the power
per unit area. Figure 14 suggests that fluvial power generally

exceeds B200 kW m1 in both Himalayan tectonic aneurysms. As these major rivers are typically on the order of
100 m wide where they are most energetic, the corresponding
minimum fluvial energy per unit area of the bed is about
2 kW m2 for an active tectonic aneurysm, which is in accord
with more detailed analysis of the eastern syntaxis (Finnegan
et al., 2008). Such power requires river discharge in excess of
103 m3 s1 (B30 km3 yr1), which can only be sustained by
an immense drainage basin receiving ample precipitation.
Assuming a net basin-averaged precipitation rate of 1 m yr1,
for example, the basin area must exceed 3 104 km2, and the
channel slope must be unusually steep for a river with this
large discharge; for both the Tsangpo and Indus rivers, the
channel slopes exceed 1% through their steeper reaches. Only
a few basins of this size can exist even in the largest orogen
currently active on the planet (the Himalaya are roughly
2000 km long and 100 km wide, amounting to an area of
2 105 km2).
5.14.3.2.4
two reasons. First, the rate of energy dissipation per unit area
of the bed, which is the product, the velocity-shear stress
product, Ut, of the sliding velocity and the basal shear stress,
can be related directly to glacial erosion rate. For example,
explicit models of glacial abrasion (Hallet, 1979, 1983) show
that the abrasion rate scales with the amount of energy that is
spent by the sliding ice, per unit time and per unit area, on
rock-to-rock friction at the glacier bed. Second, glacial power
varies primarily with basal ice speed because t is essentially a
constant; it varies over a narrow range close to approximately
105 Pa for contemporary temperate valley glaciers (e.g.,
Paterson, 1981). Although a quantitative description of the
spatial variation in glacial erosion rates remains elusive, erosion rates are expected to increase with the rate of sliding,
which generally increases with the cross-section-averaged ice
speed. This speed equals the flux of ice divided by the valley
width and average glacier thickness. Hence, rapid erosion is
generally expected where the ice flux is large because the
drainage basin is large and snowfall is ample, and where: (1)
the ice is funneled by the topography through a narrow constriction; and (2) the ice is relatively thin, which is most likely
to be where the glacier surface is steep.
Figure 15 illustrates the spatial variation in measured
surface ice speeds for SewardMalaspina and highlights a region of special interest, the Seward Throat, a narrow gap in the
range that culminates in Mt. St. Elias through which funnel
massive amounts of ice accumulating on the Seward icefield.
Figure 15 shows schematically the profile of glacial power
along a central flow line in this glacier system; the power estimated roughly from the geometry of the basin, mass conservation principles, and simple approximations of the spatial
variation in mass balance and basal shear stress. The latter was
characterized by a broad peak, approaching 1.4 105 Pa,
Glacial systems
Rapid erosion can also occur in an orogen where massive, fastmoving glaciers are fueled by large volumes of ice gathered
from vast accumulation areas, especially where they funnel
through narrow valleys. As in fluvial systems, the power per
unit length available to erode is most likely to be a useful
index of glacial erosion rate, and in addition, it invites comparisons between fluvial and glacial efficiencies in eroding
orogens. This use of glacial power as an index of erosion rates
is uncommon, but it is quite consistent with contemporary
basin-scale models of erosion by alpine glaciers, which assume
simply that erosion rates scale with basal ice speed (e.g.,
Harbor et al., 1988; Oerlemans, 1984; Tomkin and Roe, 2007;
Herman and Braun, 2008, Egholm et al., 2009). This is due to
6.72
Glacial power, SewardMalaspina glacier system
Bagle
Sew
y ice fi
eld
ard
Contact fault
ice
50
field
Power (kW m1)
UTM zone 7N northing (103 km)
60
6.70
A
CSEF
6.68
6.66
na
spi
F.
30
20
Malaspina glacier
0
10 km
Alaska
20
40
60
80
100
120
Distance from divide (km)
6.62
4.8
(a)
40
10
la
Ma
6.64
337
5.0
5.2
5.4
UTM zone 7N easting
5.6
(102
km)
5.8
(b)
Figure 15 (a) Surface ice velocities for SewardMalaspina glacier system; inset shows location in Alaska. Massive amounts of ice from the
Seward icefield funnel south through the Seward Throat, a narrow gap in the range that culminates in Mt. St. Elias, and then spreads out into the
large piedmont lobe defining Malaspina glacier. The velocities, derived from satellite data by E. Rignot, approach 2 km yr 1 in the Seward Throat.
The flowband used in a glacier dynamic analysis by R. Headley is highlighted, and the pink star marks the one site of known ice thickness. Major
faults labeled in italics (CSEF: Chugach St. Elias thrust fault), and glaciers are in bold. (b) Generalized glacier power per unit distance downglacier, guided by glacier dynamic analysis, shows strong peak in the Seward Throat area where rapid ice flow and high shear stresses, between
25 and 70 km, combine to produce exceptionally high basal power. Correspondingly, rapid erosion is expected in this energetic region, and fast
local uplift must offset the rapid erosion for this system to be sustained. Courtesy of R. Headley.
338
generally consistent with detailed numerical analysis of the

current ice dynamics through the Seward Throat constrained
by considerable data on contemporary surface velocity and ice
surface gradients, and by sparse data on mass balance and
ice thickness (R. Headley, personal communication, 2011).
Figure 15 shows a broad peak in glacial power, with values
exceeding 40 kW m1 through the Seward Throat, and dropping to much lower values outside of this region. This spatial
pattern of erosion rates could only be maintained in the long
term if localized rock uplift compensates for the erosion, and
indeed the required pattern of uplift corresponds well with the
highest ranges in the region. These results are also compatible
with the conclusion of Enkelmann et al. (2010) that the most
rapid and deepest exhumation in the region is localized near
the northern end of the Seward Throat, where a tectonic aneurysm may be emerging.
5.14.3.2.5
Emergent behavior: Self-organized balance

between uplift and erosion
At least in the Himalaya, an interesting association exists between the location of proposed tectonic aneurysms and large
knickpoints on large rivers. Aside from being regions where
river power and inferred incision rates reach peak values,
knickpoints merit special attention because they are among
the most dynamic elements of landscapes. In the absence of
differential bedrock uplift, knickpoints in bedrock channels
propagate upstream rapidly with little change in geometry
(e.g., Whipple and Tucker, 1999). This is evident from Figures
16(a)16(c) that show the evolution of a river profile similar
to those dissecting the aneurysms using a simple 1D, uniformdischarge model of bedrock incision. The modeled profile,
although highly idealized, generally resembles the profile of
the YarlungTsangpo river as it descends from SE Tibet and
cuts through the Himalaya (Figures 5, 6, and 14). The profile
evolution is simulated assuming a steady antiformal uplift and
that the rate of incision is simply proportional to fluvial
power; the rate of fluvial incision is scaled to reach a maximum value of 10 mm yr1 along the initial profiles. This
maximum incision rate appears quite reasonable in view of
previous estimates of the erosion rates by the TsangpoSiang
River (Brahmaputra) that average from 5 to 17 mm yr1 for a
large area (B4800 km2; Stewart et al., 2008; Enkelmann et al.,
2011).
In the absence of localized uplift, the simulated knickpoint
retreats upriver very quickly (Figure 16), at a rate of roughly
220 km Myr1. By adding modest rock uplift, peaking at
5 mm yr1 in the vicinity of the knickpoint, the profile
shape changes (Figure 16(b)) and the upper part of the profile
continues to migrate quickly upstream. As expected where
uplift offsets erosion closely, the knickpoint migration is arrested (Figure 16(c)). This simple model provides insights
into two of the most striking properties of the Himalayan
aneurysms: (1) their apparent stationarity with respect to
their immediate surroundings for at least three million years
and (2) the depth of exhumation estimated at 2050 km.
The stationarity is most evident in the eastern Himalayan
tectonic aneurysm where the bulls eyes of rapid erosion are
well defined over timescales ranging from a few years to a
few million years; there the locus of rapid erosion must not
have shifted spatially more than 10 or 20 km (Figure 16),
which is only a small percentage of the hundreds of kilometers of knickpoint migration that would be expected in
the absence of localized uplift across the river over the
millions of years of rapid exhumation in the region. Stationarity of the locus of rapid erosion is also evident from the
magnitude of erosion; even with erosion rates of order of
100 mm yr1, the aneurysm must be stationary for 23 Ma
to account for 2030 km of exhumation that occurred in this
region.
Figure 16(c) is consistent with the conceptual expectation
of strain and differential uplift being localized within a tectonic
aneurysm, and reflects the tight coupling between rock uplift
and the sort of rapid erosion that can arise from a large river
crossing a prominent knickpoint. This figure also illustrates two
less intuitive predictions for regions where antiformal uplift
closely matches erosion. First, in contrast with the upriver migration of knickpoints characteristic of fluvial systems with little
or no tectonic forcing, the upper reaches of a knickpoint can
migrate downriver as the profile steepens. Although empirical
support for this notion is frail, it seems to exist: a small set of
measurements of bedrock cooling ages directly upstream of the
current knickpoint on the Tsangpo river, where fluvial power is
currently low, are suggestive of rapid exhumation (Finnegan
et al., 2008). Second, upstream of the knickpoint, river gradients are commonly so low that fluvial incision may generally
have difficulty in keeping up with a downstream increase in
bedrock uplift rates. The resulting tectonic back-tilting of river
valleys would further decrease the gradient of the river and its
sediment transport capacity, thereby causing sediments to infill
valleys and create the wide alluvial reaches that are common
upstream of the Himalayan tectonic aneurysms (Figure 5). If
localized uplift in downstream reaches is fast enough, river
gradients could be reversed and large lakes could be developed.
The former is an intriguing possibility, especially for the eastern
Himalayan tectonic aneurysm, because of a long history of
studies citing substantial evidence for westward flow of the
YarlungTsangpo in southern Tibet (e.g., Burrard and Hayden,
1907; Cina et al., 2009).
We stress that the robust evidence for the stationarity of the
Brahmaputra knickpoint together with the lack of evidence for
extensive, long-lived tectonically dammed lakes (not to be
confused with a number of short-lived glacially dammed lakes
that have been documented (Montgomery et al., 2004))
strongly support the notion that fluvial erosion and uplift are
closely matched in this region. In fact, the evidence suggests
that, over the long term (e.g., 40.1 Myr) they must match
each other within a few percent; otherwise the hot spot of
erosion would migrate significantly or a large lake would exist
upstream of the tectonic aneurysm.
On shorter timescales, glaciers or landslides can dam rivers
and locally arrest erosion and knickpoint migration, as proposed by Korup and Montgomery (2008). As they anticipated,
directly downstream of such a dam the river would steepen
quickly. After just 0.1 Ma, however, a steep reach develops immediately downstream of the zone of no erosion, with an estimated drop of 50 m. Profile details in the upper of the
YarlungTsangpo knickpoint may well reflect this glacial stabilizing effect of damming by glaciers or major landslides. Over
the longer term, however, this localized arrest of erosion would
lead to profiles that are clearly unrealistic with abrupt elevation
339
7
Elevation and uplift (km)
No uplift
6
5
4
3
2
1
0
0
(a)
50
100
150
200
250
100
150
200
250
200
250
7
Uplift rate: 5 mm yr1
6
5
4
3
2
1
0
0
(b)
50
Maximum uplift rate: 11 mm yr1
6
5
4
3
2
1
0
0
(c)
50
100
150
Downstream distance (km)
Figure 16 River profile evolution as influenced by localized uplift. Blue and red curves show river profile and imposed antiformal uplift,
respectively: initial profile, dotted; after 0.5 Ma, dashed; after 1 Ma solid. Incision rate is proportional to river power and is scaled to reach a
maximum of 10 mm yr 1 in steepest reach of initial river profile. (a) In the absence of uplift, knickpoints migrate quickly upstream, over
200 km Ma 1. (b) With uplift peaking at 5 mm yr 1 in center of initial knickpoint, the knick point changes form and splits; the upper half
retreats quickly upstream, whereas the lower half becomes stationary. (c) Uplift now peaks at 12 mm yr 1 in the center of initial knickpoint. Note
that the upper section of the knickpoint now migrates downstream; an alluvial fill hundreds of meters thick would bury the valley upstream of the
antiform (extending to a downstream distance of B140 km) consistent with the broad alluvial reaches upstream of both of the Himalayan
tectonic aneurysms (e.g., Figure 5).
drops exceeding 1000 and 2000 m in just 0.5 and 1 Ma, respectively, directly downstream of the point of no erosion as
calculated using the same model and model parameters that
yielded Figures 16(a)16(c). Since none of the rivers in the
Himalaya (or elsewhere) exhibits such drops, despite rapid

erosion occurring for periods of at least 3 Ma (Burg et al., 1998;
Finnegan et al., 2008), we conclude that sustained uplift that
matches the erosion closely is required to produce the observed
340
river profiles and that glacial stabilization is ineffective on

timescales exceeding 100 ka.
5.14.3.2.6
Summary, surface boundary
In both fluvial and glacial systems, areas of extreme power

dissipation are localized where the fluxes of water or ice per
unit width are exceptionally large, which is most likely to
occur in the larger high-elevation drainage basins that gather
the precipitation. Available data suggest that all bedrock and
precipitation characteristics being equal, erosion rates approaching the apparent threshold for the onset of welldeveloped tectonic aneurysms, 5 mm yr1 over B103 km2, can
arise more readily and in smaller basins in regions eroded by
temperate glaciers than in regions eroded primarily by rivers.
The Alaska tectonic aneurysm illustrates this situation, as it
appears centered in a region draining a glacial area on the
order of B103 km2. In contrast, as mentioned above, the
tectonic aneurysms in the Himalaya that are dominantly
fueled by fluvial erosion are associated with much larger
drainage basins (B105 km2). Moreover, intriguingly, the
Alaska tectonic aneurysm is centered on a region where the ice
velocity is not the highest. This suggests that glacial erosion is
unusually efficient, which may well be because of the bedrock
being exceptionally vulnerable to erosion due to the pervasive
fracturing likely to occur in this region of localized crustal
deformation (Enkelmann et al., 2009).
5.14.4
Convergent velocity of Indian lithosphere relative to Eurasian

lithosphere at 38 mm yr1 from B65961 E) to western
model boundary.
Shallow Indian mantle trajectories in the regions of the two
syntaxes (WHS, EHS) (between B93.51 E and 961 E and
between B73.5 and 76.51 E; Zurek, 2008; Li et al., 2008).
Salt rheology applied in the vicinity of Salt Ranges (B65
and 73.51 E).
Advective thermal anomaly where exhumation 420 km
at Himalayan Front and at two syntaxes (Koons et al.,
2002).
Hot, weak lower crust and upper mantle beneath Tibet.
Permit strain-dependent upper crustal rheology (Buck and
Poliakov, 1998; Koons et al., 2010, 2011).
Geodynamics of the Tectonic Aneurysm
This section further explores the connections between the

mechanical and surface-process controls discussed in Section 5.14.2 with the larger scale evolution of plate convergence,
using numerical geodynamic models of the Himalaya and
southern Alaska to frame our discussion. Our intent is to
examine whether the erosional and mechanical controls discussed in Section 5.14.3 produce necessary and sufficient
conditions, when imposed at geodynamic scales, to capture the
primary characteristics of tectonic aneurysm behavior.
5.14.4.1
boundary behavior and embedded 3D volumes to examine

the detailed mechanical and thermal behavior near the plate
corners of interest.
To illustrate the influence of the thermalrheological
feedback that are central to the tectonic aneurysm mechanism, here we compare macroscale models with no feedback
(Corner model ) to one with the feedbacks that we have explored above. The Corner model does not utilize laterally
varying rheology and is driven at the base and to the south by
the collision of a model India continent and a Eurasian continent. We then present and focus on a generalized macroscale
Himalayan model (Figures 17(a)17(c)) driven by boundary
conditions and laterally varying rheological behavior compatible with recent geological and geophysical observations
discussed above. These include:
Numerical Models
We employ three-dimensional numerical models that solve for

the coupled thermal and mechanical fields and that are driven
by observed relative plate velocities for the IndianEurasian
and YakutatNorth American boundaries. The rheological
elasto-visco-plastic model follows the BraceKohlstedt structure discussed in Section 5.14.3.1.1, of pressure-dependent,
frictional MohrCoulomb upper crust resting on a thermally
activated plastic lower crust, which in turn rests on a model
mantle with strength approximated by an olivine flow law
(see Upton et al., 2009 for a more comprehensive description). As with our model plate geometries, our model
rheologies are relevant for the present geodynamic settings
and are not intended to capture the complex early history
of these orogens. Our modeling strategy employs a large
spatial domain (2000 km 1000 km for IndiaEurasia;
800 km 600 km for YakutatNorth America) to control
5.14.4.2
The Corner Model
In the absence of the aneurysm and salt rheological perturbations, displacement in the Corner model describes an approximately eastwest front rising from a near planar structure
in a manner compatible with numerous 2D models of orogenic growth (e.g., Willett, 1999). This planar bounding
structure, when intersecting with a spherical Earth, would
describe an arc similar to the one that defines the Perfect Arc
segment of the Himalaya between the syntaxes (Seeber and
Gornitz, 1983; Bendick and Bilham, 2001). The simple Corner
kinematic field is notable for the concave (toward India)
curvature in the horizontal velocity field as the convergent
margins transition into the strike-slip lateral boundary
(Figure 17). Spatially coincident with this curvature is a decrease in the total vertical displacement as one moves from the
convergent margin into the strike-slip zone (Figure 17(a)).
The change in the displacement fields is reflected in the shear
strain field (Figure 17(a)). This smoothly concave shape is
similar to that described previously by Koons et al. (2002) and
Braun and Yamato (2010) but fails to capture the distinctive
topographic horns or the re-entrant (to the north) shape of
the syntaxes.
5.14.4.3
The Generalized Macroscale Himalayan Model

(GMHM)
Application of a rheological model conditioned by the

erosionrheological feedback discussed above (GMHM)
GMHM:
Aneurysm rheology
Looking north
Corner model:
Laterally homogeneous rheology
341
Embedded model:
Surface material strength
Shear strain
rates
NBGP
km
Horizontal (N)
displacement
Model
conditions
Indian block
0 km
00
Eur
as
bloc ian
k
Indian
block
2000 km
NP
HM
Sal WHS
t
Ran
ge
100
10
2000 km
Vertical
displacement
NB
GP
EH
S
A
Cohesion
50 MPa
Indian block
100 kPa
A
Looking west
200
20
00
1000 km
km Shear strain
rates
0k
1000 km
NPHM
Cohesion
NBGP
Vertical
displacement
Horizontal (N)
displacement
N
(a)
(b)
(c)
Figure 17 North- and west-looking views of 3D mechanical investigation of the influence of erosion-induced rheological perturbations
characteristic of tectonic aneurysm behavior on the kinematics and topography of the Himalayan orogen. (a) Corner Model: convergence of India
lithosphere into Eurasia with laterally homogeneous rheology. Surfaces depict x-displacement, z-displacement, and shear strain rates at the model
surface as labeled. In these and following figures, maximum values are in blue, and stand out in relief. See text for discussion. (b) GMHM Model
of same geometry and influx velocities as in (a) Corner Model, but with rheological modifications associated with the nonlinear coupling
described in Section 5.14.2. Note the northward re-entrant at both syntaxes. Surfaces depict x-displacement, z-displacement, and shear strain
rates at the model surface as labeled. (c) Embedded Model of Eastern Syntaxis with strain softening upper crustal rheology. Strain concentration
leads to surface strength reduction pattern represented by the model cohesion field at the orogen surface that exhibits the distinctive northward
step observed in the EHS. Reproduced from Koons, P.O., Upton, P., Barker, A.D., 2012. The influence of mechanical properties on the link
between tectonic and topographic evolution. Geomorphology, 137, 168180.
strongly modifies the kinematic and topographic patterns.

Departure from the purely tectonically induced displacement,
parallel to the basal boundary condition, is associated with
the spatial variation in rheology, which results in northward
re-entrant steps in the Himalayan Front at both syntaxes
with deformation focused into the two tectonic aneurysms
(EHS, WHS). Both of these syntaxes disrupt the small
circle projection of the Perfect Arc. In the east, the transition
of the smoothly curved (toward India) shape of the
Himalayan Front to lateral strike-slip boundary seen in the
simple Corner model (Figure 17(b)) is disrupted by the re-
entrant step to the north into the area of the NBGP.

Coincident with this step is also the concentration of vertical
displacement with resulting extreme topography of the
syntaxial Horn, absent in the simple Corner model, as well as
the associated thermochronological anomaly (Figure 17(b))
characteristic of the fully developed tectonic aneurysm.
Where the Salt, the inclusion of Salt rheology in this
general model further disrupts the smooth Himalayan Front
with a step of the displacement field to the south into the
domain near the area of the NPHM where the Salt rheology
was employed.
342
Details of tectonic aneurysm relevant to many of the local

observations are clear in the smaller scale, high-resolution
models from the two syntaxes in which the distinctive thermal
and petrological character of Nanga Parbat and Namche
Barwa develop within the larger collision zone as a function of
localized advection, focused by efficient erosion conditions
(Figure 17(b)). As illustrated in the Nanga Parbat example,
advection of deep crustal material elevates the isotherm
leading to the formation of low-pressure, high-temperature
conditions within a sub-massif lozenge (Koons et al., 2002).
Material passing through this lozenge undergoes metamorphic
recrystallization and, in some instances, partial decompression
melting mentioned above (Section 5.14.2.2.3). This hightemperature advection signal is characteristic of a thoroughly
evolved aneurysm and is also observed in both natural and
model Namche Barwa (Figure 17(b)).
Increased erosion/aneurysm focusing in both syntaxes
leads to the distinctive domal thermochronological pattern
and extreme relief described above and that we have described
as a basic characteristic of aneurysm behavior.
5.14.4.3.1
Strain softening and drainage formation
In addition to being characteristic of the kinematic field of a

tectonic aneurysm, strain localization is also associated with
an irreversible change in the 3D distribution of those strength
parameters critical to fluvial, slope, and glacial erosion
(Koons et al., 2010). As demonstrated in experimental
and natural investigations, erosion rates are extremely sensitive to the cohesion (brittle strength) and grain size of
surface materials with fluvial incision rates inversely related to
the square of the cohesion (e.g., Sklar and Dietrich,
2004; Koons et al., 2011), and rates of glacial quarrying and
frost weathering being sensitively dependent on the fracture
mechanical properties of bedrock (Hallet, 1983; Hallet
et al., 1991; Hallet, 1996). Because cohesion and grain size
vary over several orders of magnitude during the cataclasis
associated with faulting, the evolution of surface and nearsurface strength field as a function of strain exerts a critical
control on height and orientations of slopes in deforming
orogens.
To identify the evolving surface distribution of bedrock
strength relevant for surface processes, we have also modeled
the strain softening of the upper 15 km of our General
Himalayan Model and can track the temporal evolution of the
3D strength field and compare its main features with the
natural topographic patterns of major river and valleys (Koons
et al., 2011). As seen in Figure 17(c), the strength field changes
distinctly at the eastern edge of the plate boundary, stepping to
the north before swinging sharply to the south along the lateral high strain zone. As discussed elsewhere (Koons et al.,
2011) the general path of the Tsangpo closely follows the
distinct weak region in the surface strength field as it approaches and transits through the gorges of the NBGP massif.
Further east, the striking parallelism and convergence of the
major valleys in the Three Rivers region likely reflect the
combination of the large strains themselves with the drainage
basins acting as passive strain markers (Hallet and Molnar,
2001), and the strain weakening of the rocks. Strong support
for the latter stems from the recent study by Henck et al.
(2011) reporting a two-order of magnitude increase in recent
(10Be) erosion rates with increasing proximity to the indentor

corner; this increase occurs across a B100-km-wide region
from the Yangtze to the Salween with relatively uniform current precipitation and river power.
5.14.4.4
Early Stages of Aneurysm Development: Southern

Alaskan Example
In the early stages of strain localization leading to an aneurysm, the relative roles of surface versus tectonic processes
are less clear than in the evolved syntaxes where the thermal/
rheological anomaly is well established. We have explored
these relative influences in geodynamic models of the lessevolved orogen of the St. Elias range in southern Alaska. Most
of the large-scale kinematic and topographic characteristics of
southern Alaska following collision by the Yakutat microplate
are captured at the macroscale by the 3D convergence of a
crustal block on the Pacific plate colliding with North America
(Bird, 1996; Koons et al., 2010) and are summarized in
Figure 18. Details of the influence of the preexisting topography of the Mt Logan massif and the present topography
including the continental shelf on strain localization have
been examined in an embedded 3D grid of the St. Elias region
(Hooks, 2009; Hooks et al., in press,).
At the macroscale, the main elements of the early stages of
convergence are the generation of convergent, lateral, and
subduction sub-boundaries (Figure 18). The convergent subboundary evolves into two teleconnected orogens, the inlet
and outlet orogens that form at locations that correspond with
the St. Elias and the Central Alaska Range, respectively.
Incoming material from the south separates into upward or
downward trajectories as it enters the Inlet Orogen (corresponding to the St. Elias range) with the downward advecting
material cooling the base of the Inlet wedge and the upward
velocities advectively heating the upper crust of the Inlet orogen
(Figure 18; Koons et al., 2010). Together these dual, nonlinear
advective thermal/mechanical anomalies localize strain and fix
the position of the associated topographic features of the major
subaerial mountain belts, which are linked on the lateral edge
by strike slip to oblique structures (Figure 18). In the highresolution embedded grid, the influence of the topographic
stresses is clearly significant in the position of strain localization with increased strain rates focused to the south by the
preexisting Mt Logan massif (Hooks et al., in press,). The influence of deep ice-filled valleys in the St. Elias region on strain
localization has yet to be explored but is likely to accentuate
topographic stresses. Of greater importance to the position of
strain in this nascent model orogen, however, are the mechanical controls resulting from the flow bifurcation described in
the macroscopic model. The position of vertical advection and
therefore topographic growth is dominated by the position of
the bifurcation that leads to a frictional sliver along the
base of the orogen wedge (Koons et al., 2010). The importance
of the position of this basal strength transition to topographic
evolution has been identified and investigated in twodimensional orogens by Beaumont et al. (1992), Willett et al.
(1993), and Willett (1999).
As we indicated in Section 5.14.2, the distinctive features of
the tectonic aneurysm dominantly result from the thermal
343
Embedded model
Macroscale model
65 N
60 N
Z
Model geometry
500 km
55 N
155 W
150 W
xes)
50 km
(z axis)
m (y a
600 k
145 W
140 W
135 W
400 km
800 k
m (x
axes
Model
geometry
300 km
te
y
x
ic
if
ac
pla
100 km 200 km 300 km 400 km 500 km

Model vertical velocity
Vertical velocity
(Vz) mm yr1
5.00
2.50
0.0
2.50
5.00
350 ky
Vertical velocity of macroscale model
(a)
(b)
Model vertical velocity superimposed on

thermochronology from figure 10
Figure 18 Three-dimensional summary models of the geodynamics of the St. Elias region in southern Alaska. (a) Macroscale Model (details
in Koons et al., 2010). Geometry of the plate boundary derived from Pavlis et al., 2004; Hooks, 2009; Koons et al., 2010) Convergence of
the Yakutat plate at Pacific plate far-field velocities (Elliott et al., 2010). Vertical velocity pattern defined by maxima at inflow and outflow
boundaries with concentration along the transition between inflow and lateral boundaries of model St Elias where the strain trinity accommodates the complex deformation. (b) Embedded Model (details in Hooks, 2009; Hooks et al., in press). Model vertical velocities (red;
mm yr 1) are superimposed on the thermochronological observations presented in Figure 10. Highest model velocities coincide with the
Seward Glacier.
advection/rheological feedback and can be identified as a

thermochronological anomaly. Although some recent thermochronological data hint at the development of the distinctive
thermal anomaly (Enkelmann et al., 2009; Hooks et al., in
press) as Spotila and Berger (2010) point out, evidence for the
high-temperature anomaly characteristic of the Himalayan
syntaxes is lacking. Given the sensitivity of the developing

advective anomaly to the total vertical displacement, and to
the temperature at the base of the advecting column of the
relatively cool Yakutat block, a high-temperature anomaly in
the St. Elias region is not predicted and would require an
alternative heat source, additional to the vertical advection.
344
Valley-bottom
sediment
Mean annual
river power (W m2)
> 360
180360
0180
Topographic
relief > 3300 m
Biotite cooling
ages < 2.5 Ma
N
30
es
>
10
M
a
Zircon cooling
ages < 1.5 Ma
ag
0M
>1
Zi
rc
on
s
ge
na
co
Zir
>
es
tite
10
Ma
ag
Bio
>
es
tite
29
94
10
Ma
ag
Bio
95 E
96
Figure 19 Comparison of surface-process metrics with long-term geological exhumation rates as expressed by cooling ages on biotite (ArAr)
and zircon (UTh/He) for the Namche BarwaGyala Peri massif. Reproduced from Finnegan, N.J., Hallet, B., Montgomery, D.R., Zeitler, P.K.,
Stone, J.O., Anders, A.M., Yuping, L., 2008. Coupling of rock uplift and river incision in the Namche BarwaGyala Peri massif, Tibet. Geological
5.14.4.5
Geodynamic Summary
The large-scale geodynamic modeling suggests that most of

the natural spatial variance in the kinematic, topographic,
geological, thermal, and chronological fields may be reproduced numerically by applying known, far-field plate velocities to a lithospheric rheological model formed under the
influence of the erosion/advective conditions described above.
Given the bounding conditions of mantle geometry and trajectory that we have chosen, the numerical solutions are sensitive to the rheological structure, which in turn is sensitive to
the surface erosion pattern. We must emphasize, however, that
we have limited our solutions to those with a restricted set of
bounding velocities, particularly with respect to the dip of the
upper boundary of the mantle. Geophysical investigations
justify the assumptions of shallowly dipping mantle (e.g.,
Meltzer et al., 2001; Zurek, 2008; Li et al., 2008; Bruhn et al.,
2004; Sol et al., 2007; Koons et al., 2010) that we have applied
near the syntaxes. It is important, however, to recognize that
not all mantle trajectories applied as bounding conditions,
particularly those associated with trench rollback, produce
aneurysm behavior.
5.14.5
5.14.5.1
Conclusions
Model Summary
Tectonic aneurysms manifest themselves as regions of limited

spatial extent in which large mountains coevolve with
deeply incised gorges producing exceptionally young, hightemperature metamorphic massifs perched above weak hot
crust (e.g., Namche Barwa, Figure 19). Their development and
continued existence rely on positive feedback between rock
uplift, erosion, nonlinear temperature-dependent rheology,
and crustal strain. Phenomena that are observed at or near

fully developed aneurysms include:
High heat-flow anomalies: very steep thermal gradients,

hot springs, young cooling ages, and high altitude exposure
of very young, high-grade anatectic metamorphic rocks.
Exceptionally rapid exhumation over areas exceeding
1000 km2, sustained for millions of years by exceptionally
energetic rivers and glaciers.
Decompression partial melting.
Association of high peaks and deep gorges, reflecting rapid
erosion and producing high relief and related stress
anomalies.
Relatively shallow Moho dip.
In the natural examples, the extreme relief, elevated heat

flow, and young radiometric ages characteristic of the tectonic
aneurysm are all associated with extreme concentrations of
erosional power, both fluvial and glacial. For the mature aneurysms associated with the Himalayan syntaxes, regional
stream power is arguably the greatest on Earth due to focusing
by one or more rivers of massive discharge derived from
very large catchment areas into relatively small regions; that
is the Indus and Astor rivers bounding Nanga Parbat;
YarlungTsangpo and Po Tsangpo in the Namche Barwa region. In both syntaxes, the foci of extreme stream power are
capable of maintaining approximately constant elevations
along and adjacent to the river channels in spite of the rapid
local uplift.
Similarly, glacial erosion in the St. Elias area is demonstrably capable of maintaining, and transiently deepening, the
glacial valleys that drain the southern slopes of the St. Elias
orogen. The nonlinearities present in the erosional/thermal/
rheological feedback that lead to the tectonic aneurysm perturb the near-horizontal rheological structure of the crust
sufficiently to produce effectively vertical weak zones that

permit the rheological structure to dominate over velocitybounding conditions in the distribution of crustal strain; the
weak zones act as effective strain guides. This influence of
perturbed rheology on crustal deformation is also postulated
to play a critical role in the generation of another form of
major erosion/deformation feedback, the channel-flow model
(Beaumont et al., 2001). The tectonic aneurysm and channel
flow models are examining different components of orogen
growth and are mutually complementary. We have restricted
our discussion to phenomena that are driven by the mix of
gravity and basal shear stress where the displacement of the
mantle in three dimensions is well coupled to surface kinematics, whereas during channel flow, crustal and mantle
kinematics are mechanically decoupled.
The feedback among erosional/thermal/rheological processes that naturally lead to aneurysm behavior result in some
nonintuitive behavior. In aneurysms, as rheological weakening
comes to dominate over tectonic driving forces, strain is partitioned into rather than away from high topographic massifs
leading to high mountains that sit astride weak, hot rocks. The
geomorphology of tectonic aneurysms is to a significant degree an emergent property that leads to extreme topography
and, hence, the observation that regions of exceptionally rapid
uplift and exhumation coincide closely with those being eroded by exceptionally energetic rivers is not an anomalous
coincidence. Although we have concentrated here on three
well-studied aneurysms that develop at high strain plate corner settings, the same behavior, more subtly displayed, is
manifested in numerous convergent settings. An obvious example is the Himalayan Front where the 8000 m peaks are
associated with the very deep gorges of major trans-Himalayan
rivers. The topographic anomalies at these intersections produce major 3D stress anomalies (Koons and Kirby, 2007),
high advective heat flow (Craw, 1990; Derry et al., 2009), and
local strain concentration in the form of river anticlines
(Oberlander, 1985; Fletcher and Hallet, 2004). The emergent
property evident along the major trans-Himalayan rivers is the
close balance between erosion and uplift. If erosion were not
closely balanced by uplift, the smooth curving front of the
Himalaya (e.g., Seeber and Gornitz, 1983; Bendick and Bilham, 2001) would be quickly disrupted as the major river
knickpoints would quickly migrate into Tibet; the front would
become festooned, as recognized by Lave and Avouac (2001).
Tectonic aneurysms demonstrate the rich dynamics that are
possible in the Earth system, and the interconnections and
feedback that underlie and link the solid earth, Earths surface,
and climate. Although small in extent, these features have the
potential to make disproportionate impact on such features
and phenomena as the sedimentary record of orogeny or the
orogenic influence on ocean chemistry. For instance, Stewart
et al. (2008) have argued that the Namche BarwaGyala Peri
aneurysm contributes 50% of sediment in the lower Tsangpo
although constituting only 2% of the drainage area. The extreme localization of rapid erosion, sediment evacuation, and
bedrock cooling in the aneurysms bear on: (1) common assumptions in geodynamic and geochemical studies of the
Himalaya about sources of sediment; and (2) plans for
hydroelectric development and flood management in southeastern Tibet and the heavily populated areas of eastern India.
345
Further, for active aneurysms, their high process rates, extreme

topography, and coincidence with large rivers make them
significant loci of geologic hazards, as well as ideal natural
laboratories for studying the underlying geodynamics. More
generally, the aneurysm model represents a contribution to
understanding the specifics of how collision leads to the
geochemical and structural reworking of the continental
lithosphere.
Acknowledgments
This represents contributions from scores of individuals over
many decades for which we are grateful. We extend our particular appreciation to our syntaxial colleagues: Dave Craw,
Page Chamberlain, Anne Meltzer, Bill Kidd, Phaedra Upton,
Dave Montgomery, and Terry Pavlis. We thank L. Wheeler,
A. Stinson, C. Mako, and S. Hanson for help with text and
figures; J.-P. Burg, H. Greenberg, B. Hooks, and S. Campbell
for providing unpublished figures and photographs; D. Findlayson, Y. Merrand, and R. Headley for sharing unpublished
model results; and E. Rignot and H. Conway for providing
access to data on the surface velocity for the Seward Throat
and ice thicknesses, respectively.
This work has been generously supported by the Continental Dynamics program of NSF and specifically funded in
part by NSF EAR 1009812, 0409884, and 0735402 (BH), NSF
EAR 0003462 (PZ), and NSF EAR-1009626 (PK).
References
Adams, C.J.D., 1979. Age and origin of the Southern Alps. In: Walcott, R.I.,
Cresswell, M.M. (Eds.), The Origin of the Southern Alps. Royal Society of New
Zealand, Bulletin, 18, pp. 7377.
Albare`de, F., 1976. Thermal models of post-tectonic decompression as exemplified
by the Haut Allier granulites (Massif Central, France). Bulletin Societe Geologie
France 18, 10231032.
Anders, A.M., Roe, G.H., Hallet, B., Montgomery, D.R., Finnegan, N.J., Putkonen, J.,
2006. Spatial patterns of precipitation and topography in the Himalaya.
In: Willett, S.D., Hovius, N., Brandon, M.T., Fisher, D.M. (Eds.), Tectonics,
Climate and Landscape Evolution. GSA, Special Paper 398, pp. 3953.
Argles, T.W., Foster, G.L., Whittington, A.G., Harris, N.B.W., George, M.T., 2003.
Isotope studies reveal a complete Himalayan section in the Nanga Parbat
syntaxis. Geology 31, 11091112.
Batt, G.E., Braun, J., 1999. The tectonic evolution of the Southern Alps, New
Zealand; insights from fully coupled dynamical modeling. Geophysics Journal
compressional orogens. In: McClay, K.R. (Ed.), Thrust Tectonics. Chapman and
Hall, New York, pp. 118.
Bendick, R., Bilham, R., 2001. How perfect is the Himalayan Arc? Geology 29,
791794.
Berger, A.J., Gulick, S.P.S., Spotila, J.A., et al., 2008. Quaternary tectonic response
to intensified glacial erosion in an orogenic wedge. Nature Geoscience vol. 1,
793802.
Bird, P., 1996. Computer simulations of Alaskan neotectonics. Tectonics 15,
225236.
Bishop, M.P., Shroder, Jr. J.F., 2000. Remote sensing and geomorphometric
assessment of topographic complexity and erosion dynamics in the Nanga
Parbat massif. In: Khan, M.A., Treloar, P.J., Searle, M.P., Jan, M.Q. (Eds.),
Tectonics of the Nanga Parbat Syntaxis and the Western Himalaya. Geological
Society, London, Special Publication, 170, pp. 181200.
346
Booth, A.L., Chamberlain, C.P., Kidd, W.S.F., Zeitler, P.K., 2009. Constraints on the
metamorphic evolution of the eastern Himalayan syntaxis from geochronologic
and petrologic studies of Namche Barwa. Geological Society of America Bulletin
121, 385407.
Booth, A.L., Zeitler, P.K., Kidd, W.S.F., et al., 2004. UPb Zircon constraints on the
tectonic evolution of southeastern Tibet, Namche Barwa area. American Journal
of Science 304, 889929.
Brace, W.F., Kohlstedt, D., 1980. Limits on lithospheric stress imposed by
laboratory experiments. Journal of Geophysical Research 89, 62486252.
Braun, J., Yamato, P., 2010. Structural evolution of a three-dimensional, finite-width
crustal wedge. Tectonophysics 484, 181192.
Bruhn, R.L., Forster, R.R., Ford, A.L.J., Pavlis, T.L., Vorkink, M., 2010. Structural
geology and glacier dynamics, Bering and Steller Glaciers, Alaska.
In: Shuchman, R.A., Josberger, E.G. (Eds.), Bering Glacier: Interdisciplinary
Studies of Earths Largest Temperate Surging Glacier. Geological Society of
America, Special Paper, 462, pp. 217234.
Bruhn, R.L., Pavlis, T.L., Plafker, G., Serpa, L., 2004. Deformation during terrane
accretion in the Saint Elias orogen, Alaska. Geological Society of America
Bulletin 116, 771787.
Buck, W.R., Poliakov, A.N.B., 1998. Abyssal hills formed by stretching oceanic
lithosphere. Nature 392, 272275.
northwestern Himalayas. Nature 379, 505510.
Burg, J.P., Nievergelt, P., Oberli, F., et al., 1998. The Namche Barwa syntaxis:
Evidence for exhumation related to compressional crustal folding. Journal of
Asian Earth Science 16, 239252.
Butler, R.W.H., Harris, N.B.W., Whittington, A.G., 1997. Interactions between
deformation, magmatism and hydrothermal activity during active crustal
thickening: a field example from Nanga Parbat, Pakistan Himalayas.
Mineralogical Magazine 61, 3751.
Carslaw, H.S., Jaeger, J.C., 1959. Conduction of Heat in Solids, Second ed. Oxford
University Press, USA, 510 pp.
Chamberlain, C.P., Koons, P.O., Meltzer, A.S., Park, S.K., Craw, D., Zeitler, P.K.,
Poage, M.A., 2002. Overview of hydrothermal activity associated with active
orogenesis and metamorphism; Nanga Parbat, Pakistan Himalaya. American
Chapple, W.M., 1978. Mechanics of thin-skinned fold and thrust belts. Geological
Cina, S.E., Yin, A., Grove, M., et al., 2009. Gangdese arc detritus within the eastern
Himalayan Neogene foreland basin: implications for the Neogene evolution of
the YaluBrahmaputra River system. Earth and Planetary Science Letters 285,
150162.
Cowan, E.A., Seramur, K.C., Powell, R.D., Willems, B.A., Gulick, S.P.S., Jaeger,
J.M., 2010. Fjords as temporary sediment traps: history of glacial erosion and
deposition in Muir Inlet, Glacier Bay National Park, Southeastern Alaska. GSA
Bulletin 122, 10671080.
Craw, D., 1990. Fluid evolution during uplift of the Annapurna Himal, central Nepal.
Lithos 24, 137150.
Craw, D., Chamberlain, C.P., Zeitler, P.K., Koons, P.O., 1997. Geochemistry of a dry
steam geothermal zone formed during rapid uplift of Nanga Parbat, northern
Pakistan. Chemical Geology 142, 1122.
Craw, D., Koons, P.O., Winslow, D., Chamberlain, C.P., Zeitler, P.K., 1994. Boiling
fluids in a region of rapid uplift, Nanga Parbat massif, Pakistan. Earth and
Craw, D., Koons, P.O., Zeitler, P.K., Kidd, W.F.S., 2005. Fluid evolution and
thermal structure in the rapidly exhuming gneiss complex of Namche
BarwaGyala Peri, eastern Himalayan syntaxis. Journal of Metamorphic Geology
23, 829845.
Crowley, J.L., Waters, D.J., Searle, M.P., Bowring, S.A., 2009. Pleistocene melting
and rapid exhumation of the Nanga Parbat massif, Pakistan: age and PT
conditions of accessory mineral growth in migmatite and leucogranite. Earth and
Dahlen, F.A., Suppe, J., Davis, D., 1984. Mechanics of fold-and-thrust belts and
accretionary wedges: cohesive Coulomb theory. Journal of Geophysical Research
89, 1008710101.
Davis, D., Suppe, J., Dahlen, F.A., 1983. Mechanics of fold-and-thrust belts and
accretionary wedges. Journal of Geophysical Research 88, 11531172.
Davis, W.M., 1899. The Geographic cycle. The Geographical Journal 14,
481504.
Delcaillau, B., 1992. Les Siwalik de LHimalaya du Nepal Oriental: fonctionnement et
Evolution dun Piedmont. Memoires et Documents de Geographie. Editions du
Centre National de la Recherche Scientifique, Paris, 205 pp.
Derry, L.A., Evans, M.J., Darling, R., France-Lanord, C., 2009. Hydrothermal heat
flow near the Main Central Thrust, central Nepal Himalaya. Earth and Planetary
Ding, L., Zhong, D., Yin, A., Kapp, P., Harrison, T.M., 2001. Cenozoic structural
and metamorphic evolution of the eastern Himalayan syntaxis (Namche Barwa).
Egholm, D.L., Nielsen, S.B., Pedersen, V.K., Lesemann, J.E., 2009. Glacial effects
limiting mountain height. Nature 460(7257), 884. http://dx.doi.org/10.1038/
nature08263.
Elliott, J., Larsen, C.F., Freymueller, J.T., Motyka, R.J., 2010. Tectonic block motion
and glacial isostatic adjustment in Southeast Alaska and adjacent Canada
constrained by GPS measurements. Journal of Geophysical Research. http://dx.d
oi.org/10.1029/2009JB007139.
Enkelmann, E., Ehlers, T., Zeitler, P.K., Hallet, B., 2011. Denudation of the Namche
Barwa Antiform, Eastern Himalaya, Earth and Planetary Science Letters. Earth
Enkelmann, E., Zeitler, P.K., Garver, J.I., Pavlis, T.L., Hooks, B.P., 2010. The thermo
chronological record of tectonic and surface processes interaction at the
YakutatNorth American collision zone in southeast Alaska. American Journal of
Science 310, 231260.
Enkelmann, E., Zeitler, P.K., Pavlis, T.L., Garver, J.I., Ridgway, K.D., 2009. Intense
localized rock uplift and erosion in the St Elias orogen of Alaska. Nature
Enlow, R.L., Koons, P.O., 1998. Critical wedges in three dimensions; analytical
expressions from Mohr-Coulomb constrained perturbation analysis. Journal of
Fernandez, R., Anderson, J., Wellner, J., Hallet, B., 2011. Timescale dependence of
glacial erosion rates, a case of study: marinelli Glacier, Cordillera Darwin,
southern Patagonia. Journal of Geophysical Research 116, F01020.
Finlayson, D.P., Montgomery, D.R., Hallet, B., 2002. Spatial coincidence of rapid
219222.
Yuping, L., 2008. Coupling of rock uplift and river incision in the Namche
BarwaGyala Peri massif, Tibet. Geological Society of America Bulletin 120,
142155.
Fletcher, R., Hallet, B., 2004. Initiation of gneiss domes by necking, density
instability, and erosion. In: Whitney, D.L., Teyssier, C., Siddoway, C.S. (Eds.),
Gneiss Domes and Orogeny. The Geological Society of America, Boulder, CO,
vol. 380, pp. 7995.
Garzanti, E., Malusa`, M.G., 2008. The Oligocene Alps: Domal unroofing and
drainage development during early orogenic growth. Earth and Planetary Science
Letters 268, 487500.
Hallet, B., 1979. A theoretical model of glacial abrasion. Journal of Glaciology
23(89), 2950.
Hallet, B., 1983. The breakdown of rock due to freezing: a theoretical model.
Proceedings of the Fourth International Conference on Permafrost. Fairbanks,
Alaska, USA, pp. 433438.
Hallet, B., 1996. Glacial quarrying: a simple theoretical model. Annals of Glaciology
22, 18.
glaciers: a review of field data and their implications. Global and Planetary
Change 12, 213235.
Hallet, B., Molnar, P., 2001. Distorted Drainage Basins as markers of crustal
strain East of the Himalaya. Journal of Geophysical Research 106,
1369713709.
Hallet, B., Walder, J.S., Stubbs, C.W., 1991. Weathering by segregation ice growth
in microcracks at sustained sub-zero temperatures: verification from an
experimental study using acoustic emissions. Permafrost and Periglacial
Handy, M.R., Hirth, G., Burgmann, R., 2007. Continental fault structure and
rheology from the frictional-to-viscous transition downward. In: Handy, M.R.,
Hirth, G., Hovius, N. (Eds.), Tectonic Faults Agents of Change on a Dynamic
Earth. MIT Press, Cambridge, MA, pp. 139182.
Harbor, J.M., Hallet, B., Raymond, C.F., 1988. A numerical model of landform
development by glacial erosion. Nature 333(6171), 347349.
Henck, A., Huntington, K.W., Stone, J.O., Montgomery, D.R., Hallet, B., 2011.
Spatial controls on erosion in the Three Rivers Region, southeastern
Tibet and southwestern China. Earth and Planetary Science Letters 303,
7183.
Herman, F., Braun, J., 2008. Evolution of the glacial landscape of the Southern Alps
of New Zealand: Insights from a glacial erosion model. Journal of Geophysical
Hooks, B.P., 2009. Geodynamics of terrane accretion within southern Alaska. PhD
Univ. Maine. pp. 204.
Hooks, B.P., Enkelmann, E., Koons, P.O., Upton, P. Pavlis, T.L., in press, 3D
Mechanical view of the relative importance of surface and tectonic processes in
the initiation of the St Elias tectonic aneurysm, Earth Planetary. Science Letters.
Jaeger, J.M., Nittrouer, C.A., 1999. Sediment deposition in an Alaskan fjord:
controls on the formation and preservation of sedimentary structures in Icy Bay.
Journal of Sedimentary Research 69, 10111026.
Kargel, J.S., 2004. Mars A Warmer, Wetter Planet. Springer-Praxis Publishing,
Chichester, UK, pp. 557.
Kirby, B., Schoenbohm, L., Chen, J., Lei, S., 2007. Investigating the tectonic
aneurysm model: measuring slip rates along the Kongur detachment fault in the
Chinese Pamir. American Geophysical Union 2007 Fall Meeting, T23D-1641.
Koons, P.O., 1987. Thermal and mechanical consequences of rapid uplift in
continental collision: an example from the Southern Alps. Earth and Planetary
Koons, P.O., 1989. The topographic evolution of collisional mountain belts: a
numerical look at the Southern Alps, New Zealand. American Journal of Science
289, 10411069.
Koons, P.O., 1990. The two-sided wedge in orogeny; erosion and collision from the
sand box to the Southern Alps, New Zealand. Geology 18, 679682.
Koons, P.O., 1994. Three-dimensional critical wedges; tectonics and topography in
oblique collisional orogens. Journal of Geophysical Research 99, 1230112315.
Koons, P.O., 2009. On the implications of low spatial correlation of tectonic and
climate variables in the western European Alps. Geology 37, 863864.
Koons, P.O., Hooks, B., Pavlis, T., Upton, P., Barker, A., 2010. Three-dimensional
mechanics of the Yakutat convergence in the southern Alaskan plate corner.
Tectonics 29, TC4008. http://dx.doi.org/10.1029/2009TC002463.
Koons, P.O., Kirby, E., 2007. Topography, Denudation, and Deformation: The Role of
Surface Processes in Fault Evolution. In: Handy, M.R., Hirth, G., Hovius, N.
(Eds.), Tectonic Faults Agents of Change on a Dynamic Earth. MIT Press,
Cambridge, MA, pp. 205230.
Koons, P.O., Upton, P., Barker, A.D., Barker, 2012. The influence of mechanical
properties on the link between tectonic and topographic evolution.
Geomorphology, 137, 168180.
Koppes, M., Hallet, B., 2006. Erosion rates during rapid deglaciation in Icy Bay,
Alaska. Journal of Geophysical Research 111, F02023.
Koppes, M., Montgomery, D.M., 2009. The relative efficacy of fluvial and
glacial erosion over modern to orogenic timescales. Nature Geoscience 2,
644647.
Koppes, M.N., Hallet, B., Anderson, J.A., 2009. Synchronous acceleration of ice
loss and glacial erosion, Glaciar Marinelli, Chilean Tierra del Fuego. Journal of
Lague, D., 2010. Reduction of long-term bedrock incision efficiency by short-term
alluvial cover intermittency. Journal of Geophysical Research 115, F02011.
Lave, J., Avouac, J.P., 2001. Fluvial incision and tectonic uplift across the
Himalayas of central Nepal. Journal of Geophysical Research 106,
2656126591.
Li, C., van der Hilst, R.D., Meltzer, A.S., Engdahl, E.R., 2008. Subduction of the
Indian lithosphere beneath the Tibetan Plateau and Burma. Earth and Planetary
Liu, Y., Siebel, W., Theye, T., Massone, H.J., 2010. Isotopic and structural
constraints on the late Miocene to Pliocene evolution of the Namche
Barwa area, eastern Himalayan syntaxis, SE Tibet. Gondwana Research 19,
894909.
Liu, Y., Zhong, D., 1997. Petrology of high-pressure granulites from the eastern
Himalayan syntaxis. Journal of Metamorphic Geology 15, 451466.
Liu, Y., Zoback, M.D., 1992. The effect of topography on the state of the crust:
application to the site of the Cajon Pass Scientific Drilling Project. Journal of
McTigue, D.F., Mei, C.C., 1981. Gravity-induced stresses near topography of small
slope. Journal of Geophysical Research 86, 92689278.
Meltzer, A.S., Sarker, G.L., Seeber, L., Armbruster, B., Beaudoin, B., 2001. Seismic
characterization of an active metamorphic massif, Nanga Parbat, Pakistan,
Himalaya. Geology 29, 651654.
347
Miller, D.J., Dunne, T., 1996. Topographic perturbations of regional stresses and
consequent bedrock fracturing. Journal of Geophysical Research 101,
2552325536.
Milliman, J.D., Meade, R.H., 1983. World-wide delivery of river sediment to the
oceans. Journal of Geology 91, 121.
Milliman, J.D., Syvitski, J.P.M., 1992. Geomorphic/tectonic control of sediment
discharge to the ocean: the importance of small mountainous rivers. Journal of
Geology 100, 525544.
Molnia, B.F., Post, A., 2010a. Introduction to the bering glacier system, Alaska/
Canada: early observations and scientific investigations, and key geographic
features. In: Shuchman, R.A., Josberger, E.G. (Eds.), Bering Glacier:
Interdisciplinary Studies of Earths Largest Temperate Surging Glacier. Geological
Society of America, Special Paper ,462, pp. 1342.
Molnia, B.F., Post, A., 2010b. Surges of the Bering Glacier. In: Shuchman, R.A.,
Josberger, E.G. (Eds.), Bering Glacier: Interdisciplinary Studies of Earths Largest
Temperate Surging Glacier. Geological Society of America, Special Paper, vol.
462, pp. 291316.
Montgomery, D.R., Hallet, B., Yuping, L., Finnegan, N., Anders, A., Gillespie, A.,
Greenberg, H.M., 2004. Evidence for Holocene Megafloods down the Tsangpo
River gorge, southeastern Tibet. Quaternary Research 62, 201207.
Montgomery, D.R., Stolar, D.B., 2006. Reconsidering Himalayan river anticlines.
Oberlander, T.M., 1985. Origin of drainage transverse to structures in orogens.
In: Morisawa, M., Hack, J.T. (Eds.), Tectonic Geomorphology. Allen and Unwin,
Winchester, MA, USA, pp. 155182.
Oerlemans, J., 1984. Numerical experiments of large-scale glacial erosion.
Zeitschrift fuer Gletscherkunde und Glazialgeologie. 20, 107126.
Park, S., Mackie, R., 2000. Resistive (dry?) lower crust in an active orogen, Nanga
Parbat, northern Pakistan. Tectonophysics 316, 359380.
Paterson, W.S.B., 1981. The Physics of Glaciers. Pergamon Press, New York,
380 pp.
Pavlis, G.L., Picornell, C., Serpa, L., Bruhn, R.L., Plafker, G., 2004. Tectonic
processes during oblique collision: insights from the St. Elias orogen, northern
North American Cordillera. Tectonics 23, TC3001. http://dx.doi.org/10.1029/
2003TC001557.
Pavlis, T.L., Hamburger, M.W., Pavlis, G.L., 1997. Erosional processes as a control
on the structural evolution of an actively deforming fold and thrust belt; an
example from the Pamir-Tien Shan region, central Asia. Tectonics 16, 810822.
http://dx.doi.org/10.1029/97TC01414.
Penck, W., 1924. Die Morphologische Analyse. Geographische Abhandlungen 2,
283.
Reiners, P.R., Brandon, M.T., 2006. Using thermochronology to understand
orogenic erosion. Annual Review of Earth and Planetary Science 34,
419466.
Richardson, S.W., England, P.C., 1977. The influence of erosion upon the mineral
facies of rocks from different metamorphic environments. Journal of the
Schneider, D.A., Edwards, M.A., Kidd, W.S.F., Khan, M.A., Seeber, L., Zeitler, P.K.,
1999. Tectonics of Nanga Parbat, western Himalaya: synkinematic plutonism
within the doubly vergent shear zones of a crustal-scale pop-up structure.
Geology 27, 9991002.
Schneider, D.A., Zeitler, P.K., Kidd, W.S.F., Edwards, M.A., 2001. Geochronologic
constraints on the tectonic evolution and exhumation of Nanga Parbat, western
Himalaya syntaxis, revisited. Journal of Geology 109, 563583.
Seeber, L., Gornitz, V., 1983. River profiles along the Himalayan arc as indicators of
active tectonics. Tectonophysics 92, 335367.
Seward, D., Burg, J.P., 2007. Growth of the Namche Barwa Syntaxis and associated
evolution of the Tsangpo Gorge: Constraints from structural and
thermochronological data. Tectonophysics 451, 282289.
Shroder, Jr. J.F., Bishop, M.P., 2000. Unroofing of the Nanga Parbat Himalaya.
In: Khan, M.A., Treloar, P.J., Searle, M.P., Jan, M.Q. (Eds.), Tectonics of the
Nanga Parbat Syntaxis and the Western Himalaya. Geological Society, Special
Publication, 170, pp. 163179.
Sibson, R.H., 1982. Fault zone models, heat flow, and the depth distribution of
earthquakes in the continental crust of the United States. Bulletin of the
Simpson, G., 2004. Role of river incision in enhancing deformation. Geology 32,
341344. http://dx.doi.org/10.1130/G20190.2.
Sklar, L.S., Dietrich, W.E., 2004. A mechanistic model for river incision into
bedrock by saltating bed load. Water Resources Research 40(6), W06301.
Sklar, L.S., Dietrich, W.E., 2006. The role of sediment in controlling steady state
bedrock channel slope: Implications of the saltation-abrasion incision model.
348
Slaymaker, O., 1987. Sediment and solute yields in British Columbia and Yukon:
their geometric significance reexamined. In: Gardiner, V. (Ed.), International
Geomorphology, Part 1. Wiley, Chichester, pp. 925945.
Smith, H.A., Chamberlain, C.P., Zeitler, P.K., 1992. Documentation of Neogene
regional metamorphism in the Himalayas of Pakistan using UPb in monazite.
Sol, S., Meltzer, A., Burgmann, R., et al., 2007. Geodynamics of the southeastern Tibetan plateau from seismic anisotropy and geodesy. Geology 35,
563566.
Sonder, L.J., England, P., 1986. Vertical averages of rheology of the continental
lithosphere: relation to thin sheet parameters. Earth and Planetary Science
Letters 77, 8190.
Spotila, J.A., Berger, A.L., 2010. Exhumation at orogenic indentor corners under
long-term glacial conditions: example of the St. Elias orogen, Southern Alaska.
Stewart, R.J., Hallet, B., Zeitler, P.K., Malloy, M.A., Allen, C.M., Trippett, D., 2008.
Brahmaputra sediment flux dominated by highly localized rapid erosion from the
easternmost Himalaya. Geology 36, 711714.
Terzaghi, K., 1943. Theoretical Soil Mechanics. John Wiley, New York, 510 pp.
Thiede, R.C., Arrowsmith, J.R., Bookhagen, B., McWilliams, M.O., Sobel, E.R.,
Strecker, M.R., 2005. From tectonically to erosionally controlled development of
the Himalayan orogen. Geology 33, 689692.
Thiede, R.C., Bookhagen, B., Arrowsmith, J.R., Sobel, E.R., Strecker, M.R., 2004.
Climatic control on rapid exhumation along the southern Himalayan front. Earth
Tippett, J.M., Kamp, P.J.J., 1993. Fission track analysis of the late Cenozoic vertical
kinematics of continental Pacific Crust, South Island, New Zealand. Journal of
Tomkin, J.H., Roe, G.H., 2007. Climate and tectonic controls on glaciated
criticaltaper orogens. Earth and Planetary Science Letters 262(34), 385397.
http://dx.doi.org/10.1016/j.epsl.2007.07.040.
Upton, P., Koons, P.O., Craw, D., Henderson, C.M., Enlow, R., 2009. Along strike
differences in the Southern Alps of New Zealand: consequences of inherited
variation in rheology. Tectonics 28, TC2007. http://dx.doi.org/10.1029/
2008TC002353.
Whipple, K.X., Meade, B.J., 2004. Controls on the strength of coupling among
climate, erosion, and deformation in two-sided, frictional orogenic wedges at
steady state. Journal of Geophysical Research 109, F01011.
Whipple, K.X., Tucker, G.E., 1999. Dynamics of the stream-power river incision
model: implications for height limits of mountain ranges, landscape response
timescales, and research needs. Journal of Geophysical Research 104,
1766117674.
Whittaker, A.C., Cowie, P.A., Attal, M., Tucker, G.E., Roberts, G.P., 2007. Bedrock
channel adjustment to tectonic forcing: implications for predicting river incision
Whittington, A.G., Foster, G.L., Harris, N.B.W., Vance, D., Ayres, M.W., 1999a.
Lithostratigraphic correlations in the western Himalaya an isotopic approach.
Geology 27, 585588.
Whittington, A.G., Harris, N.B.W., Butler, R.W.H., 1999b. Contrasting anatectic styles
at Nanga Parbat, northern Pakistan. Geological Society of America, Special
Paper, vol. 328, pp. 129144.
Willett, S., Beaumont, C., Fullsack, P., 1993. A mechanical model for the tectonics
of doubly vergent compressional orogens. Geology 21, 371374.
Willett, S.C., 1999. Orogeny and orography: the effects of erosion on the
structure of mountain belts. Journal of Geophysical Research 104,
2895728981.
Winslow, D.M., Zeitler, P.K., Chamberlain, C.P., 1994. Direct evidence for a steep
geotherm under conditions of rapid denudation, western Himalaya, Pakistan.
Geology 22, 10751078.
Zeitler, P.K., 1985. Cooling history of the NW Himalaya. Tectonics, 4, 127-151. 61
Zeitler, P.K., Chamberlain, C.P., Smith, H.A., 1993. Synchronous anatexis,
metamorphism, and rapid denudation at Nanga Parbat, Pakistan Himalaya.
Geology 21, 347350.
Zeitler, P.K., Johnson, N.M., Naeser, C.W., Tahirkheli, R.A.K., 1982. Fission-track
evidence for the Quaternary uplift of the Nanga Parbat region, Pakistan. Nature
298, 255257.
Zeitler, P.K., Koons, P.O., Bishop, M.L., et al., 2001b. Crustal Reworking at Nanga
Parbat, Pakistan: Evidence for erosional focusing of crustal strain. Tectonics 20,
712728.
Zeitler, P.K., Meltzer, A.S., Koons, P.O., et al., 2001a. Erosion, Himalayan
geodynamics, and the geology of metamorphism. GSA today: a publication of
the Geological Society of America 11, 48.
Zeitler, P.K., Sutter, J.F., Williams, I., Zartman, R.E., Tahirkheli, R.A.K., 1989.
Geochronology and temperature history of the Nanga ParbatHaramosh Massif,
Pakistan. In: Malinconico, L.L., Lillie (Eds.), Tectonics of the Western Himalayas.
Geological Society of America, Special Paper, vol. 232, pp. 122.
Zhang, Z., Dong, X., Santosh, M., Liu, F., Yiu, F., He, Z., Shen, K., 2011. Petrology
and geochronology of the Namche Barwa Complex in the eastern Himalayan
syntaxis, Tibet: Constraints on the origin and evolution of the north-eastern
margin of the Indian Craton. Gondwana Research, doi:10.1016/j.gr.2011.02.002.
Zurek, B., 2008. The evolution and modification of continental lithosphere dynamics
of Indentor Corners and imaging the lithosphere across the eastern syntaxis of
Tibet. Lehigh University, unpublished Ph.D. thesis.
Biographical Sketch
Peter Koons received a firm grounding in the response of the silicate Earth to continental convergence during MSc
research at the University of Otago, New Zealand, and a PhD at E T H, Zurich. He returned to Otago University to
an evolving paradigm, strongly influenced by New Zealand scientists, in which documented rates of both erosion
and Earth deformation were several orders of magnitude greater than those assumed in the generally accepted
framework. The dynamic implications of these high rates for nonlinear interactions among mountains and
metamorphism, climate and seismicity have kept Koons from going sane ever since. In projects from the Southern
Alps, the Himalayan syntaxes, and southern Alaska, he has been influenced by the creativity and obstinacy of
Zeitler and Hallet. Koons is currently professor of Geodynamics in the Department of Earth Sciences and Climate
Change Institute at the University of Maine where he continues to be entangled in the interactions of atmosphere
and silicate Earth at increasingly higher temporal and spatial frequencies.
349
Peter Zeitler is a Professor of Earth and Environmental Sciences and an Iacocca Professor at Lehigh University.
After considering an English major at Dartmouth College, Zeitler took an undergraduate degree in Earth Sciences
before earning his Ph.D. from Dartmouth in 1983. He then spent 5 years as a research fellow in isotope
geochemistry at the Research School of Earth Sciences of the Australian National University in Canberra. Zeitlers
research interests include development of techniques in geochronology, with a focus on noble-gas methods and
thermochronology and their application to tectonics. Zeitler is also interested in crustal geodynamics, the nature
and origin of mountains, and the geologic evolution of Asia, where he has worked for 30 years in the Himalaya,
Tibet, and Mongolia.
Bernard Hallet is a Professor at the University of Washington in the Department of Earth and Space Sciences and
the Quaternary Research Center. He received his PhD from UCLA in 1975, and promptly headed to the Himalaya
before joining the faculty at Stanford. In 1980, he accepted an invitation to join the Quaternary Research Center,
in part to follow the footsteps of Link Washburn and to capitalize on the laboratory facilities he established to
study periglacial phenomena. Hallet served as the Director of the Quaternary Research Center from 1998 to 2003.
He is fascinated by landscapes primarily in mountainous and polar regions, and the diverse glacial and periglacial
processes that shape. He worked on detailed analyses of the mechanics of glacial erosion, frost weathering, and
patterned ground formation until 1989 when a chance meeting with Koons in Otago, New Zealand, opened his
eyes to the rich linkages between climate, tectonics, and erosion in regions of extreme relief. His research trajectory
was permanently derailed with another chance meeting, this one with Zeitler and co-workers on one of those last
days of an AGU meeting in San Francisco where he was first introduced to a geophysical aneurysm. Hallets field
work has taken him to the St. Elias Mountains (Alaska, USA), the Himalaya (Tibet, India, and Nepal), Patagonia
and Tierra del Fuego (Chile), and the Antarctic Peninsula. A central theme addressed in his research is physical
self-organization in nature, which ranges in size and type from a few meters in frozen soil patterns to hundreds of
kilometers in the Himalaya, and other tectonically active mountain ranges, where the topography often tends
naturally to a state where rates of uplift and erosion are balanced.
5.15 The Influence of Middle and Lower Crustal Flow on the Landscape
Evolution of Orogenic Plateaus: Insights from the Himalaya and Tibet
KV Hodges and BA Adams, Arizona State University, Tempe, AZ, USA
5.15.1
5.15.2
5.15.3
5.15.4
5.15.4.1
5.15.4.1.1
5.15.4.1.2
5.15.4.2
5.15.4.2.1
5.15.4.3
5.15.5
Introduction
Development and Geophysical Characteristics of the Tibetan Plateau
Gravitational Potential Energy Gradients and the Dynamics of Middle Crustal Flow
Geomorphology and Tectonics of the Tibetan Plateau
Long-Wavelength Topographic Variations, South to North (Transect AA0 )
Are EW corrugations related to middle-lower crustal flow?
Has Tibetan middle crust extruded at the Himalayan front?
Long-Wavelength Topographic Variations, West to East (Transect BB0 )
Does the gradual west-to-east decrease in mean elevation indicate eastward flow of the middle crust?
Short-Wavelength Topographic Variations on the Tibetan Plateau
A Self-Consistent Model of the Cenozoic Topographic Evolution of the Tibetan Plateau, Assuming Lower
and Middle Crustal Flow
5.15.6
Feedbacks among Middle-Lower Crustal Flow, Landscape Evolution, and Climate
5.15.7
Conclusions
Acknowledgments
References
Abbreviations
DEM
g
GPE
GPEo
digital elevation model.

gravitational acceleration.
gravitational potential energy.
gravitational potential energy of an orogenic
system.
GPEs
gravitational potential energy of lithosphere
adjacent to an orogenic system GTOPO30, US
Geological Survey digital elevation model,
30 arcsec resolution HST (Himalayan Sole
thrust).
INDEPTH International Deep Profiling of Tibet and the
Himalaya.
ITS
Indus-Tsangpo suture zone.
Glossary
Anatectic Related to partial melting.
Asthenosphere The mechanically weak region of the
mantle of the Earth.
Cosmogenic radionuclide Elements produced by the
interaction of cosmic particles with target atoms (usually in
rock or sediment) at Earths surface. Useful to date surfaces
or buried sediment.
Decollement A shallowly dipping, regionally extensive
fault surface.
Denudation Unroofing via erosion or faulting.
L
MBTS
MCTS
MODIS
PT
SRTM
STFSz
DGPE
m
q
351
351
355
355
356
359
360
361
361
362
362
363
364
364
364
lithospheric thickness.
Main boundary thrust system.
Main central thrust system.
Moderate-resolution imaging
spectroradiometer.
Physiographic transition.
Digital elevation model (c. 90 m px1
resolution) based on data acquired during the
International Shuttle Radar Topography
Mission.
South Tibetan fault system depth.
Differential gravitational potential energy.
material flux.
density.
Detachment A low-angle extensional fault.

Deviatoric stress The nonhydrostatic component of a
stress field.
Duplex A system of overlapping faults between more
shallowly dipping boundary faults above and below.
Felsic Rich in silicate minerals.
Foreland The low-elevation region adjacent to a
mountain range.
Geochronologic Pertaining to the measurement of
geologic time.
Hodges, K.V., Adams, B.A., 2013. The influence of middle and lower crustal
flow on the landscape evolution of orogenic plateaus: insights from the
Himalaya and Tibet. In: Shroder, J. (Editor in Chief), Owen, L.A. (Ed.),
Treatise on Geomorphology. Academic Press, San Diego, CA, vol. 5,
Tectonic Geomorphology, pp. 350369.
350
http://dx.doi.org/10.1016/B978-0-12-374739-6.00095-6
Insights from the Himalaya and Tibet
Geodesy The study of the shape and size of

Earth. Geodetic techniques are commonly used to gage
the deformation field in active orogenic
systems.
Knickpoint The location of an abrupt change in the
gradient of a river channel.
Lithosphere The rigid outer shell of the solid Earth,
including the crust and upper mantle.
Mafic Poor in silicate minerals.
Magnetotelluric Relating to methods of measurement of
Earth structure and composition through the study of
natural electric and magnetic fields.
Monsoon Heavy seasonal rains.
Orogenesis Mountain building.
Orogenic Pertaining to the building of mountains.
Out-of-sequence thrust A thrust fault that forms at an
unexpected position within a thrust belt that is not
predicted by conventional kinematic models.
Petrologic Pertaining to the study of rocks.
Rayleigh-wave A type of surface wave associated with
earthquakes.
Rheology An expression of the mechanical strength of
geologic materials.
351
Ridge push the outward force exerted on the adjacent

plate by a mid-ocean ridge.
Rifting The pulling apart of the lithosphere.
Slab pull The gravitational force exerted on a lithospheric
plate by its leading edge as it plunges into the mantle along
a subduction margin.
Strike-slip fault A fault along which the two sides
predominately move horizontally relative to each other.
Subduction The underthrusting of one tectonic plate
beneath another.
Syntaxis A crustal-scale bend in the trend of a mountain
belt.
Tectonostratigraphic Pertaining to packages of rocks
separated by fault systems.
Thermochronologic Pertaining to the measurement of
the thermal history.
Thermomechanical model A computer model that
incorporates variations of the physical properties of
deforming materials as a function of temperature.
Tractional flow A force applied to tectonic plates by
convention in the mantle.
Transcurrent fault Synonymous with strike-slip fault.
Abstract
The Tibetan Plateau serves as a model for understanding the possible influence of middle and lower crustal flow on the
geomorphic and tectonic evolution of a continental plateau. Arguments are presented in favor of a weak middle (and, in
some areas, lower) crust beneath the plateau, and the development of tectonic models of the HimalayanTibetan orogenic
system that incorporate lateral flow of this weak crust under the influence of gravity are explored. Many long- and shortwavelength features of Tibetan and Himalayan landscapes are reasonably explained as a consequence of such flow.
5.15.1
Introduction
Local geology permitting, a favorite field trip stop for introductory students is a roadcut of deformed migmatitic rocks.
We teach them that such outcrops represent the deep continental crust, where pressures and temperatures are high enough
that rocks can flow like toothpaste on geologic timescales
during mountain building. In fact, large exposures of lower
crustal rocks in deeply eroded orogenic systems such as the
Scandinavian and East Greenland Caledonides testify to a
more complex rheology. Some tracts seem to have behaved
more or less rigidly during mountain building, whereas others
particularly those containing large proportions of anatectic
melt rocks (anatexites) are penetratively deformed, with
fabrics and mineral textures indicative of high strains. In regions of the lower crust dominated by such materials, it is easy
to image large-scale flow during orogenesis. Several lines of
geophysical evidence suggest that the crust beneath orogenic
plateaus is sufficiently hydrated to be relatively weak
and at high enough temperatures to be partially molten.
If the viscosity of this crust is low enough, and if zones of lowviscosity are sufficiently interconnected, it is theoretically
possible that channels may form that enable this weak crust to
flow laterally as a consequence of lateral pressure gradients
(Bird, 1991).
As reviewed by Hodges (2006) and Klemperer (2006),

middle or lower crustal flow has been invoked in many models
of the tectonic evolution of the Tibetan Plateau in recent years.
The concept has been exported to the Andean Plateau (Husson
and Sempere, 2003; Gerbault et al., 2005; Ouimet and Cook,
2010) as well as many Precambrian to Cenozoic orogenic systems (e.g., Brown and Gibson, 2006; Brown and Juhlin, 2006;
Cagnard et al., 2006; Hatcher and Merschat, 2006; Kuiper et al.,
2006; Xypolias and Kokkalas, 2006; Jamieson et al., 2007;
Rivers, 2008; Barreiro et al., 2010; Dorr and Zulauf, 2010;
Dumond et al., 2010; Gee et al., 2010; Li et al., 2010; Raimondo
et al., 2010). Although such models remain controversial, this
chapter explores how lateral flow of a fluid middle or lower
crust might be manifested in the surface geomorphology of
plateau interiors and margins. We focus on the Tibetan Plateau,
where most of these ideas were developed, as an analog for
other present-day plateaus like the Andean Plateau and plausible older plateaus in the geologic record.
5.15.2
Development and Geophysical

Characteristics of the Tibetan Plateau
Developed as a consequence of Cenozoic collision between

India and Eurasia, the Tibetan Plateau is one of the most
352
remarkable physiographic features on Earth. It has roughly

half the area of the continental USA and a mean elevation
above sea level of approximately 5000 m (Figure 1). Truecolor satellite imagery (Figure 2) illustrates the profound
precipitation gradients across plateau margins as dramatic
color differences in the landscape, from the browns of the
high plateau to the deep greens to the south and east.
This transition is most abrupt across the high mountain
ranges that mark the southern rim of the plateau and form one
of the most dramatic rain shadows on our planet as they block
powerful storms coming from the south. The color/precipitation gradient is far more gradual across the more gentle
topographic slope of the southeastern margin of the plateau,
indicating that summer storms of the Indian monsoon
sweep around the eastern end of the Himalayan ranges and
are funneled into this region through deep gorges formed as
major river systems drain the southeastern plateau (shown as
the Three Rivers region in Figure 2). The unusual arcuate
configuration of these gorges, and the fact that several of their
traces extend northwestward as topographic lineaments within
the plateau, attest to the fact that they follow major transcurrent fault systems (Molnar and Tapponnier, 1975;
Tapponnier and Molnar, 1976; Tapponnier et al., 1982).
Continentcontinent collision might be expected to result
in crustal thickening within the collision zone and, if
boundary conditions permit, the lateral expulsion of material
between the two colliding plates. Both processes feature
prominently in models that have been developed to describe

plateau evolution. Escape models invoke a special kind of
crustal flow, one in which rigid blocks of Tibetan crust slip
past one another along a network of transcurrent faults
and the ensemble of blocks experiences macroscopic granular
flow (Tapponnier et al., 1982; Peltzer and Tapponnier, 1988;
Replumaz and Tapponnier, 2003). Although the net amount
of lateral expulsion since collision is highly controversial
(Houseman and England, 1993; Tapponnier et al., 1982), and
geodetic data clearly demonstrate that the regions between
major transcurrent faults within Tibet do not behave completely rigidly (Chen et al., 2004; Zhang et al., 2004;
Gan et al., 2007), most recent incarnations of the escape
model provide a reasonable explanation of active surface deformation throughout much of Tibet (e.g., Meade, 2007;
Thatcher, 2007).
Such models do not, however, fully describe lithospheric
deformation in Tibet, particularly the development of an unusually thick crust beneath the plateau (6075 km; Tseng
et al., 2009). England and Houseman (1986, 1988), Houseman and England (1993), and England and Molnar (1997)
have demonstrated that continuum mechanical models treating the entire Tibetan lithosphere as a thin viscous sheet explain lithospheric thickening as well as limited expulsion of
upper crust. These continuum models presume, for simplicity,
a vertically averaged rheology of the lithosphere. However,
the crust beneath the Tibetan Plateau is unusually hot
Elevation (m)
41 N
8572
37 N
33 N
29 N
25 N
21 N
17 N
71 E
75 E
79 E
1
83 E
87 E
91 E
95 E
99 E
103 E
107 E
Figure 1 Digital elevation model (GTOPO30 1 km px ) of south Asia centered on the Tibetan Plateau. Major mountain ranges, basins, and
other geographic elements are shown for reference. The plateau is defined in this paper as the high elevation region bordered by the Himalayan
ranges to the south, the Karakoram to the west, the W Kunlun ShanAltyn TaghQilian Shan mountain systems to the north, and the Longmen
Shan and Three Rivers region to the east.
353
41 N
37 N
33 N
29 N
71 E
75 E
79 E
83 E
87 E
91 E
95 E
99 E
Figure 2 True-color composite image created using data from NASAs Terra satellite MODIS (Moderate-Resolution Imaging Spectroradiometer)
instrument. Major fault systems and physiographic features are indicated, as well as some of the more prominent rift systems in the plateau
interior.
Elevation (m)
8572
41 N
37 N
33 N
29 N
25 N
21 N
17 N
71 E
75 E
79 E
83 E
87 E
91 E
95 E
99 E
103 E
107 E
Figure 3 Major fault systems of the HimalayanTibetan orogenic system from the compilation of Taylor and Yin (2009). Major named faults
(black lettering) include: AT Altyn Tagh; J Jiali; KA Karakoram; KN Kunlun; KX Karakax; and MF Main Frontal. Yellow acronyms refer
to suture zones: BNS Banggong-Nujiang; ITS Indus Tsangpo; and JS Jinsha. White lines indicate the positions of topographic profiles
(actually 100 km-wide, only center of profiles shown here). Profiles AA0 and BB0 are shown in Figure 7. Profiles CC0 through JJ0 are shown
in Figure 8.
354
compared to most continental regions (Francheteau et al.,

1984). Moreover, that crust is a collage of compositionally
variable materials accreted to Eurasia in the course of preCenozoic orogenic events (e.g., Bally et al., 1980; Alle`gre et al.,
1984; Dewey et al., 1988; Burtman, 2010).
Together these characteristics suggest that the actual rheology of the Tibetan crust is likely to be considerably more
complex than vertically averaged models would imply.
Classical models of the strength of the continental lithosphere (Goetze and Evans, 1979; Brace and Kohlstedt,
1980; Kirby, 1983) emphasize the change in deformation
mechanisms that occur with depth as a consequence of
increasing temperature and changing lithospheric composition. They feature a brittle upper crust where rock strength
is limited by friction and increases with depth with little regard for compositional variation (Byerlee, 1978; Marone,
1998). At deeper levels and higher temperatures, quartz-rich
rocks exhibit ductile behavior and become progressively
weaker. For a given strain rate, a variety of factors can influence
the strength of rocks in this regime. Among the most important are composition (less felsic compositions are stronger:
Burgmann and Dresen, 2008), the degree of hydration (wetter
rocks are weaker: Kohlstedt, 2007), and the presence and
volumetric proportion of anatectic melt (partially molten
rocks are weaker: Renner et al., 2000; Rosenberg and Handy,
2005).
Figure 4 illustrates qualitative strength-depth profiles for
the crust and uppermost mantle lithosphere beneath southern
Tibet and central and northern Tibet. They are based on
available geological, geophysical, and geochemical data regarding compositions and thermal structure (cf., Klemperer,
2006). Seismic velocity data from southern Tibet (Wittlinger
et al., 2009) suggest the existence of a dry, mafic, and thus
relatively strong lower crust. We can reasonably infer that the

Tibetan middle crust is less mafic and more hydrated and thus
should be substantially weaker than the seismically active
upper crust or the drier, more mafic lowermost crust. In central
and northern Tibet, various seismic data suggest that the
entire middle and lower crust as well as at least the uppermost mantle are hot and weak (Ni and Barazangi, 1983;
McNamara et al., 1995; Wei et al., 2010).
A three-part, vertically stratified rheology of Tibet with a
laterally continuous weak zone in the middle or lower crust is
a venerable concept in tectonics, dating back at least as far
as the work of Bird and Toksoz (1977), who deduced the
existence of the weak layer from Rayleigh-wave attenuation
observations. After detailed geophysical surveys of southern
Tibet began as part of the US National Science Foundation
INDEPTH project in the 1990s, more direct evidence for a
weak middle crust became available (Nelson et al., 1996).
Since then, much seismic and magnetotelluric data have confirmed the presence of a fluid-rich or partially molten middle
crustal level beneath the southern half of the plateau, extending in some areas as far south as the crest of the Himalaya
(Li et al., 2003; Unsworth et al., 2005; Klemperer, 2006;
Ashish et al., 2009; Caldwell et al., 2009; Guo et al., 2009;
Huang et al., 2009a; Zhang and Klemperer, 2010). Knowing
the effects of fluids and melts on crustal rheology as we do,
these observations imply that the middle crust may be sufficiently fluid to sustain channelized lateral flow (Bird, 1991;
Nelson et al., 1996). The available geophysical data for
northern Tibet (see the review by Klemperer, 2006) seem to
indicate that much of the middle and lower crust there may
have sufficiently low viscosity to support wholesale lateral flow
rather than narrowly channelized flow in the middle crust
(e.g., Bendick and Flesch, 2007).
Strength
Southern
Tibet
Depth
Elastic
upper
crust
Elastic
upper
crust
Central and
Northern Tibet
Weak middle
crustal channel
Weak middle
crustal channel
Strong
lower
crust
Moho
Strong eclogitized Indian
crust and Indian mantle
lithosphere
Weak Eurasian mantle lithosphere
Figure 4 Generalized strength profiles for southern Tibet (left) and central and northern Tibet (right) based on a variety of sources, especially
Klemperer (2006) and Wei et al. (2010). Gray horizontal line indicates the position of the Moho. Note that an elastic upper crust and a strong
lower crust bound a relatively thin (perhaps 1025 km?), weak channel in the middle crust of southern Tibet. Farther north, the elastic upper
crust is inferred to extend deeper based on earthquake seismology in that region (Wei et al., 2010). An ensemble of geophysical data suggests
that the middle crust, lower crust, and upper mantle are relatively weak beneath central and northern Tibet, north of the Banggong-Nujiang suture
(Figures 3 and 6).
5.15.3
Gravitational Potential Energy Gradients and

the Dynamics of Middle Crustal Flow
One useful context for thinking about the tectonic and geomorphic implications of such flow derives from an analysis
of lateral gradients in gravitational potential energy within
and around the Tibetan Plateau (Hodges et al., 2001). The
gravitational potential energy per unit area of a column of
lithosphere (GPE) can be defined as the integrated vertical
stress due to the body force:
GPE
ZL
L zrzg dz
1
h
where L is lithosphere thickness, h is surface elevation, z is

depth, r(z) is density as a function of depth, and g is gravitational acceleration (e.g., Jones et al., 1996; Flesch et al.,
2007). Lithospheric thickening in orogenic systems results in a
change in gravitational potential energy of the system (GPE0)
relative to the surrounding, thinner lithosphere (Artyushkov,
1973). Thickening of the crust alone leads to an increase in
gravitational potential energy, although concomitant thickening of the denser mantle lithosphere counterbalances this
effect (Fleitout and Froidevaux, 1982). The convective removal
(Molnar et al., 1993) or delamination (Bird, 1991) of a
dense lithospheric root at some point in the systems evolution would result in a dramatic overall increase in the difference between gravitational potential energy of the system
and its surroundings. We can define this difference (DGPE) as:
DGPE GPE0 GPES
2
where GPES represents the gravitational potential energy of

adjacent lithosphere. Large values of DGPE result in large
differential stresses that are entirely due to lithospheric
thickness variations; these stresses contribute significantly to
the dynamics of continental deformation in orogenic systems
(e.g., Molnar and Lyon-Caen, 1988; England and Jackson,
1989; Jones et al., 1996; Humphreys and Coblentz, 2007). The
GPE of continental plateaus is characteristically high, and the
GPE of Tibet is the highest on Earth. As a consequence, Tibet
exerts a very large deviatoric stress on its surroundings, estimated at c. 56 TN m1 (6 1012 N m1) by Molnar and
Lyon-Caen (1988) and Copley et al. (2010), or at a more
modest c. 2.5 TN m1 by Ghosh et al. (2006). Either way,
these values are of comparable magnitude to or larger than
stresses exerted by processes such as ridge push, slab pull, or
tractional flow in the asthenosphere per unit length along
plate boundaries (Thatcher, 2009).
Within major orogenic plateaus, deviatoric stresses induced
by DGPE are tensional (Artyushkov, 1973; Fleitout and Froidevaux, 1982), providing an explanation for why the active
surface deformation of plateau interiors is frequently dominated by extensional and transcurrent faults rather than thrust
faults (Dalmayrac and Molnar, 1981; Dewey, 1988; Molnar
and Lyon-Caen, 1989). However, another phenomenon may
also help smooth out gravitational potential energy anomalies
beneath plateaus: the lateral flow of crust from areas with
high GPE to areas with low GPE. One can think of the
HimalayanTibetan orogenic system as a vast reservoir of
355
excess GPE, built by IndiaEurasia collision, that is, dissipating by the redistribution of mass outward from the plateau
through a combination of brittle faulting in the Tibetan upper
crust and lateral flow in the Tibetan middle crust (Westaway,
1995; Royden et al., 1997; Hodges et al., 2001; Shen et al.,
2001; Royden et al., 2008).
The flux of material outward from the plateau (n) can be
represented as:
v kDGPE,
3
where k is a proportionality constant. From this equation

analogous to Fouriers, Darcys, and Ficks laws for heat flow,
fluid flow, and chemical diffusion, respectively it follows that
the directionality of flow is influenced strongly by the gradient
in GPE in the absence of resistance. At present, the highest
GPE gradients in the Tibetan realm are along the southern
margin of the plateau, which may explain why GPE-related
deviatoric stresses within the plateau are NS tensional
(Ghosh et al., 2006). Hodges et al. (2001) noted that although
continued NS convergence across the Himalaya exerts a
resistance against wholesale southward spreading of the
plateau, a coupling of middle crustal extrusion with aggressive
erosion along the Himalayan range front could provide an
extremely effective outlet for energy dissipation.
Within the Tibetan Plateau, flow of material in a fluid
middle crustal channel could redistribute mass laterally as a
consequence of horizontal pressure gradients imposed by
topographic gradients above the channel (Bird, 1991).
Figure 5(a) illustrates the essential characteristics of such a
channel that tunnels laterally through the crust. The major
geomorphic expression of the laterally propagating tip of such a
channel should be a propagating topographic front. For more
detailed discussions of this process, readers are directed to Clark
and Royden (2000), Clark et al. (2005a), Grujic (2006), and
Medvedev and Beaumont (2006). Some workers have suggested
that laterally propagating channels may emerge or extrude
(Figure 5(b)) at plateau margins under specific tectonic and
climatic conditions (e.g., Hodges et al., 2001; Beaumont et al.,
2001; Hodges, 2006). In Section 5.15.4, the authors concentrate
on how the marginal and interior landscapes of the Tibetan
Plateau might relate to middle crustal flow, brittle deformation
in the upper crust, and climatic processes.
5.15.4
Geomorphology and Tectonics of the Tibetan

Plateau
Even before the Cenozoic collision of India and Eurasia that

built the HimalayanTibetan orogenic system, the tectonic
history of Tibet had been long and complex. The bedrock geology of the plateau (Figure 6) tells the story of successive accretion of tectonostratigraphic blocks to a growing
Eurasia over the TriassicEocene interval (e.g., Bally et al.,
1980; Alle`gre et al., 1984; Dewey et al., 1988; Burtman, 2010).
However, the geomorphology of the plateau is largely
reflective of Miocene-Recent tectonics subsequent to
IndiaEurasia collision. In order to best appreciate these relationships, it is useful to think of the physiography of Tibet in
terms of long-wavelength (4102 km-scale) and short-wavelength (r102 km-scale) topographic variations. In Sections
356
Surface uplift propagates

with channel front
MOHO
(a)
Frontal
thrust
Detachment
MOHO
(b)
Figure 5 Cartoon cross sections illustrating two modes of channel flow (after Hodges, 2006). (a) A weak channel (random hatch pattern) is
shown tunneling through stronger crust (gray) under the influence of gravity. Arrow shows direction of flow. Note the propagating wave of
surface uplift associated with the channel front. (b) Channel extrusion to the surface. This model can be used to illustrate current ideas for
Miocene (and possibly later) channel extrusion toward the Himalayan front as driven by the GPE contrast between the Tibetan Plateau to the
north and the Indian foreland to the south. The channel is bound by a normal fault system on top (the South Tibetan fault system) and a thrust
fault system on the bottom (Main Central thrust system in the Miocene Himalaya). Erosion along the range front removed mass (and therefore
excess GPE) from the system. Structures labeled i represent the accretion of down-going Indian Plate into the channel for recycling by channel
extrusion toward the range front.
41 N
500 km
Qilian
KunlunQaidam
Qaidam
Basin
37 N
SQS
KS
JS
Qiangtang
33 N
SongpanGanzi-Hoh Xil
BNS
Lhasa
ITS
29 N
Himalaya
Indian Foreland
79 N
83 N
87 N
91 N
95 N
99 N
Figure 6 Generalized terrane map of Tibet. Modified from Yin, A., Harrison, T.M. 2000. Geologic evolution of the HimalayanTibetan orogen.
Annual Reviews of Earth and Planetary Sciences 28, 211280. Major sutures, shown in thick black lines, separate named tectonostratigraphic
units (labeled in black). The sutures are identified with white labels; from north to south, they include the late Proterozoic (Devonian?) southern
Qilian suture (SQS), Late Triassic Kunlun suture (KS), Early Jurassic (?) Jinsha suture (JS), Late Jurassic Bangong-Nujiang suture (BNS), and
Eocene IndusTsangpo suture (ITS).
5.15.4.15.15.4.3, we make use of a series of topographic

profiles across the plateau (Figure 7) and more detailed
profiles across select plateau margins (Figure 8) to describe
salient features. The cross-plateau profiles in Figure 7 represent averages and ranges of 100 km-wide topographic swaths
(Figure 3) extracted from the Global 30 arcsec digital elevation dataset (GTOPO30), which has an effective resolution
of slightly better than 1 km px1. The more detailed plateaumargin profiles in Figure 8 are based on Shuttle Radar
Topography Mission (SRTM) data, with a resolution of slightly

better than 90 m px1.
5.15.4.1
Long-Wavelength Topographic Variations, South

to North (Transect AA0 )
High-resolution satellite images and DEMs of Tibet (Figures 1

and 2) reveal a land of low-relief surrounded by margins of
A
8000
A
Southern
Highlands
Himalaya
Elevation (m)
Northern
Highlands
6000
357
E. Kunlun Shan
PT1
ITS
Altyn Tagh
4000
QB
2000
PT2
PT3
200 km
B
Karakoram &
W. Kunlun Shan
Elevation (m)
8000
Longmen
Shan
6000
IR
4000
2000
400 km
Figure 7 Topographic profiles AA0 (south to north) and BB0 (west to east) across the Tibetan Plateau. Black line indicates mean elevation for
a 100 km-wide swath centered on the profile lines shown in Figure 3. Gray-shaded areas indicate two standard deviations of the mean elevations
within the swath. Note scale change between the two profiles. Important geographic and geologic features are indicated, some by acronyms: IR
Indus River gorge; ITS IndusTsangpo suture; and QB Qaidam Basin. On the Himalayan range front, major physiographic transitions PT1,
PT2, and PT3 (Hodges et al., 2001) are shown for reference.
extreme to relatively moderate relief (Fielding et al., 1994).

The Himalayan ranges, which correspond geographically to
the southern margin of the plateau, harbor our planets
highest mountains and most dramatic relief. The foothills
of the Himalaya (Lower Himalaya) rise from near sea level in
the Indo-Gangetic plains in the south to typical maximum
elevations of c. 2000 m across a physiographic transition (left
margin of Transect AA0 ) that marks the approximate surface
trace of the Himalayan Sole thrust (HST), the active decollement along which India is subducted northward beneath Tibet
(e.g., Seeber et al., 1981; Yeats and Lillie, 1991). However, an
even more profound physiographic transition, marked by
major knickpoints on rivers that flow southward off the plateau, commonly separates the foothills region from the
physiographic Higher Himalaya (Hodges et al., 2001; Wobus
et al., 2006). Although some researchers regard it as marking
the trace of a N-dipping, surface-breaking, out-of-sequence
thrust fault (Wobus et al., 2003, 2005; Hodges et al., 2004),
others regard it as the surface expression of the leading edge
of a blind duplex system developed above a structural ramp
in the HST (Cattin and Avouac, 2000; Bollinger et al., 2006;
Herman et al., 2010). All of these researchers nonetheless agree
that deformation at this physiographic transition, whether at

the surface or slightly deeper in the upper crust, has resulted in
a major discontinuity in the regional rock and surface uplift
patterns in Plio-Pleistocene times. To the north, the steep front
of the Higher Himalaya spatially coincides with: (1) rapid
denudation on the million-year and millennial timescales as
evidenced by thermochronologic and cosmogenic radionuclide data (e.g., Burbank et al., 2003; Wobus et al., 2003,
2005; Bojar et al., 2005; Blythe et al., 2007; Adams et al., 2009;
Dortch et al., 2011); as well as (2) rapid surface uplift as
measured by geodesy (Jackson and Bilham, 1994). Importantly, this front also bears the brunt of the South Asian
monsoon in the Himalaya (Barros et al., 2000; Barros and
Lang, 2003; Burbank et al., 2003; Putkonen, 2004; Bookhagen
and Burbank, 2006) and consequently supplies much of the
eroded material carried southward toward the plains by major
river systems draining off the southern plateau (e.g., Garzanti
et al., 2007; Gabet et al., 2008).
A third important physiographic transition marks the crest
of the steep Higher Himalayan front and the literal edge of the
plateau. The position of this transition, which we refer to as
PT1 (following Hodges et al., 2001), is commonly marked by a
358
Elevation (m)
8000
Elevation (m)
PT2
4000
2000
8000
PT3
50 km
D
6000
50 km
PT1
4000
PT2
2000
0
8000
Elevation (m)
PT1
6000
PT3
50 km
E
50 km
I
PT1
6000
4000
PT2
2000
0
8000
Elevation (m)
6000
50 km
J
PT1
(?)
4000
PT2 (?)
2000
0
PT3
50 km
50 km
PT3
50 km
Figure 8 Topographic profiles CC0 through and FF0 across the Himalayan front, GG0 across the W Kunlun ShanTarim Basin front, HH0
across the north flank of the Qilian Shan, II0 across the Longmen Shan-Sichuan Basin escarpment, and JJ0 across the diffuse SE Tibet margin.
Black line indicates mean elevation for a 100 km-wide swath centered on the profile lines shown in Figure 3. Gray-shaded areas indicate two
standard deviations of the mean elevations within the swath. Important physiographic transitions discussed in the text are indicated with
acronyms PT1. PT2, and PT3 (Hodges et al., 2001).
prominent knickpoint on rivers flowing off the plateau and

across the Himalayan range front. Hodges et al. (2001) noted
that the PT1 as well as the physiographic transition at the
foot of the Higher Himalaya (PT2), and at the foot of the
Lower Himalaya (PT3) correlate with fault traces in at least
some well-studied transects through the central part of the
range (e.g., Transect EE0 , Figure 8). The authors will return to
the possible significance of this correlation later on. It is
noteworthy that the topographic profile across the Himalayan
front changes significantly along strike. In many parts of
Bhutan (e.g., Transect CC0 , Figure 8), there is a welldeveloped topographic bench immediately south of PT2. In
the NW Himalaya (e.g., Transect FF0, Figure 8), PT2 is almost
indistinguishable. Such variations likely represent along-strike
variations in the topology of blind and surface-breaking fault
systems, as well as longitudinal precipitation gradients.
Despite the impressive height of the major Himalayan
peaks (including most of Earth0 s landscape above 8000 m),
the mean elevation of the Himalayan crest is actually only
slightly higher than the mean elevation of Tibetan plateau as a
whole; Transect AA0 shows a mean elevation of just under
5000 m, with a total relief of only approximately 1000 m between the Himalayan crest and the Qaidam Basin. Closer inspection of the profile and the digital elevation model in
Figure 1 reveals two broad, BEWtrending topographic
highs within the plateau between approximately 801 and
951 N, in addition to the Himalayan crest at the plateaus
southern margin and the Altyn TaghQilian mountain complex along the northern margin.
From the Himalayan crest, the mean elevation decreases
northward from c. 5000 m to approximately 4000 m over a
distance of roughly 200 km. A narrow topographic low occurs
in the EW-trending valley of the upper Yarlung Tsangpo River,
which eventually flows eastward, then southward, off the plateau to empty into the Bay of Bengal as the Brahmaputra
River. In southern Tibet, the river course generally follows
the trace of the Eocene IndusTsangpo suture (ITS), a feature
that marks the surface boundary between geology of the Indian realm to the south and the Eurasian realm to the north
(Bally et al., 1980; Shackelton, 1981; Tapponnier et al., 1981).
The high region immediately north of the Yarlung Tsangpo
valley corresponds to the Transhimalayan ranges, which are
359
Elevation (m)
41 N
8572
37 N
33 N
29 N
25 N
21 N
17 N
71 E
75 E
79 E
83 E
87 E
91 E
95 E
99 E
103 E
107 E
Figure 9 Region of internal drainage on the Tibetan Plateau and the distribution of remnants of a relict low-relief surface in the eastern plateau
and off-plateau regions to the east.
largely underlain by magmatic rocks of a CretaceousPaleocene continental arc that marked the southern margin
of Eurasia to IndiaEurasia collision (Dewey et al., 1988).
This Transhimalayan region or what we will refer to in this
chapter as the southern highlands of Tibet broadly corresponds to the Lhasa tectonostratigraphic terrane (Figure 6). It
has average elevations of 50005500 m and isolated peak
elevations up to more than 7000 m (not evident on Transect
AA0 ). Farther north, is another subtle, EW-trending topographic low that broadly corresponds to the surface trace of
the Late JurassicEarly Cretaceous Banggong-Nujiang suture
(Girardeau et al., 1984; Dewey et al., 1988).
North of the Banggong-Nujiang suture lies a northern
highlands region underlain by predominately Paleozoic
Mesozoic metamorphic and sedimentary rocks of the Qiangtang tectonostratigraphic block (e.g., Alle`gre et al., 1984; Dewey
et al., 1988; Yin and Harrison, 2000; Kapp et al., 2005). Collectively the southern and northern highlands and the intervening topographic lowland represent what is normally
regarded as the central Tibetan plateau. The vast area north
of the drainage divide of the southern highland is internally
drained (Figure 9) and remarkably flat, with local relief typically measured in the hundreds of meters or less (Liu-Zeng
et al., 2008; Dong et al., 2010).
This broad region transitions northward to elevations of
B3000 m across a narrow zone with high local relief corresponding to the Eastern Kunlun Shan. A second zone of high
relief, the Altyn Tagh, separates the Qaidam Basin from lowlands to the north. The Altyn Tagh, Western Kunlun Shan, and
Qilian Shan collectively mark the structurally complex
northern margin region of the Tibetan Plateau, where active

major left-lateral strike-slip structures like the Altyn Tagh and
Kunlun faults are kinematically linked with numerous NWstriking and both SW- and NE-vergent thrust faults (Burchfiel
et al., 1989; Duvall and Clark, 2010; Harkins et al., 2010). Like
the Himalayan realm along the southern plateau margin,
the northern margin of Tibet apparently became the locus
of contractional deformation at the time of IndiaEurasia
collision in Eocene time (Clark et al., 2010; Yin, 2010); as
noted in the first of those papers, the structural reactivation of
the area in MioceneRecent times implies a relatively stationary position of the northern plateau margin for the past
45 million years.
5.15.4.1.1
Are EW corrugations related to middle-lower

crustal flow?
Although subtle, the broad EW corrugation in the Tibetan

Plateau comprising the southern and northern highlands
and the depression between them has been interpreted in
different ways. Jin et al. (1994) suggested that these might
represent large-scale buckling of the Tibetan lithosphere.
Bendick et al. (2008) presented a model in which the corrugations represent long-wavelength flexures of an elastic upper
crust above a fluid lower crust that is being forced northward
by a rigid Indian indenter. They based this hypothesis on a
model of northern Tibet in which there is no single, midcrustal channel of low viscosity crust but instead a middle and
lower crust that is uniformly fluid (Bendick and Flesch, 2007).
In contrast, Pullen et al. (2011) interpreted the NS, longwavelength topographic variation as inherited from older
360
tectonic patterns. They noted spatial correlations among the

northern physiographic highlands, a Mesozoic structural
antiform in the Qiangtang tectonostratigraphic block, and
crustal-scale geophysical features. However, the evidence for
an extremely weak lower crust beneath northern Tibet is
compelling (Klemperer, 2006) and seems inconsistent
with the explicit suggestion by Pullen et al. (2011) that the
Mesozoic surface structural architecture persists downward
into the deep crust.
5.15.4.1.2
Has Tibetan middle crust extruded at the

Himalayan front?
High erosion rates along the southern flank of the plateau, coupled with the kinematic characteristics of major north-dipping
fault systems in the Himalaya, led Nelson et al. (1996) to propose
a tectonic model for the Miocene evolution of this part of the
orogenic system that involves the viscous flow of middle crust
southward from beneath southern Tibet and removal of that
material from the orogenic system by monsoon-driven erosion at the Himalayan front (Figure 5(b)). The viability of this
channel extrusion hypothesis hinges on the correctness of two
assumptions. The first is that there was a fluid middle crust in
southern Tibet during the Miocene, similar to the one inferred
to exist at present based on geophysical data. Proponents of
the southward extrusion model of Nelson et al. (1996) point
to the metamorphic core of the Himalaya commonly referred to as the Greater Himalayan sequence (Hodges, 2000)
as an excellent analog for what the modern Tibetan middle
crust might look like: high-grade (amphibolite to granulite
facies) metamorphic rocks that contain abundant Miocene
granites resulting from widespread anatectic melting. The
outcrop belt for the north-dipping Greater Himalayan
sequence conforms to the Himalayan range front between PT1
and PT2 (Figures 7 and 8). The second key assumption has
to do with the nature of the upper and lower contacts of the
Greater Himalayan sequence. If the sequence represents an
emergent channel, it should be bound at the top and bottom
by coeval, N-dipping fault zones with opposite vergence:
S-directed thrusting at the base and N-directed extension
(from a kinematic perspective) at the top. Evidence for
Miocene, large-displacement (order 102 km) thrusting at the
appropriate structural level has existed for decades (e.g.,
Le Fort, 1975), but the recognition of normal faulting near the
Himalayan crest has come substantially more recently and not
without some controversy.
Chinese and French geoscientists can be credited with the
first recognition of N-dipping, low-angle, normal-sense
detachments in the physiographic high Himalaya (Caby et al.,
1983; Burg et al., 1984), but the first suggestion that these
structures may be part of a major fault system all along the
crest of the Himalayan arc can be traced to the work of
Burchfiel and Royden (1985). Speculating that this system
was broadly coeval with major south-directed thrust systems
at deeper structural levels the Miocene Main Central
Thrust system or MCTS and Pliocene Main Boundary
Thrust system or MBTS (Hodges, 2000) Burchfiel and
Royden hypothesized that steep MiocenePliocene topographic gradients resulted in a stress field consistent with
synchronous slip on shallowly dipping thrust and normal
faults in a continentcontinent collisional setting, effectively
accommodating southward expulsion of material between the

contractional and extensional fault systems toward the orogenic foreland.
Subsequent research by an international community of
independent teams confirmed the existence of a South
Tibetan fault system (STFS) comprising shallowly and
steeply N-dipping normal faults and oblique transcurrent
faults near the range crest from as far east as Bhutan and as
far west as NW India (as reviewed by Burchfiel et al., 1992;
Hodges, 2000). Although the STFS appears to have initiated
slip in MiddleLate Miocene time in many areas (e.g., Hodges
et al., 1992, 1996; Edwards and Harrison, 1997; Searle et al.,
1997; Wu et al., 1998; Walker et al., 1999), an ever-increasing
body of evidence points to the existence of multiple detachments of multiple ages at the structural level of the STFS
(Burchfiel et al., 1992; Hodges et al., 1996; Kellett et al., 2009;
Searle, 2010), some of them as young as Pleistocene (Hurtado
et al., 2001). In as much as southward thrust faulting has been
continuous along the Himalayan front since at least Miocene
time (see Hodges, 2000 for a review), evidence for normal
faulting near the Himalayan crest over most if not all of the
same interval led Hodges et al. (2001) and Hodges (2006) to
speculate that the southward extrusion of south Tibetan
middle crust has been continual in the Himalaya for the past
20 million years. Beaumont et al. (2001) presented elegant
thermomechanical models for how the surfacing of a midcrustal channel could be a natural consequence of rapid,
monsoon-driven erosion along the Himalayan front, and
Hodges et al. (2001) noted how this coupling of tectonics and
erosion provides an extremely efficient mechanism for dissipation of the excess gravitational potential energy from the
orogenic system.
The channel extrusion hypothesis requires that both the
basal thrust fault system and the capping normal fault system
binding the Greater Himalayan channel have displacements
measured in the tens or hundreds of kilometers. Although
virtually all Himalayan researchers agree that this is true for
major thrust structures of the MCT and MBT systems, not
everyone agrees that the South Tibetan fault system is this
significant. Geologic relationships described by Herren
(1987), Burchfiel et al. (1992), De`zes et al. (1999), Long and
McQuarrie (2010), and Searle (2010) from Bhutan to NW
India seem to require at least 20 km of slip along South
Tibetan fault system detachments and, in some cases, much
more. The notion of very large-scale displacements is supported by a growing body of petrologic, geochronologic, and
thermochronologic data demonstrating dramatic differences
in the late Cenozoic thermal history of the Greater Himalayan
sequence and strata in the hanging wall of the STFS (e.g.,
Searle et al., 2003; Jessup et al., 2008; Cottle et al., 2011).
However, a few authors (e.g., Yin, 2006; Sachan et al., 2010)
regard the South Tibetan fault system as a relatively minor
feature of limited significance.
Additional arguments have been marshaled against the
channel extrusion hypothesis by a variety of authors (e.g.,
Harrison, 2006; Kohn, 2008; Yin et al., 2010). Some reflect a
concern that Greater Himalayan sequence rocks are of Indian,
rather than Eurasian, affinity. If the Greater Himalayan sequence is indeed the fossil remains of a Miocene channel
flowing from southern Tibet, why are the rocks not similar to
those exposed at the surface in the Transhimalayan ranges?

Channel extrusion advocates counter that the channel in
southern Tibet as opposed to central and northern Tibet,
north of the IndusTsangpo suture may be Indian plate
materials accreted across the HST during continental subduction at the time of IndiaEurasia collision (Hodges, 2006;
Harris, 2007). Another criticism (Kohn, 2008) is that the
pressuretemperaturetime history of the Greater Himalayan
rocks is different from that predicted by published thermomechanical models of channel extrusion (e.g., Jamieson et al.,
2004), but such models include so many explicit input parameters that the inconsistency of any specific model result
with geologic observations is as likely a reflection of parameter
choice and simplifying assumptions as it is an indication of an
invalid conceptual model. Still another expressed concern is
that the model requires coupled deformation and monsoondriven erosion, but erosion is focused on the Himalayan front,
whereas the Greater Himalayan sequence in some places occurs in klippen far south of the area of highest rainfall (Harrison, 2006). However, recent analysis of precipitation
patterns shows a broader NS distribution of heavy rainfall
than previously inferred (e.g., Bookhagen and Burbank, 2010).
Besides, there is no reason to assume that the latitude of heavy
precipitation relative to the range front has not migrated
substantially since the Miocene, and much reason to believe
that some of the klippen are related to complex duplexing or
out-of-sequence thrusting after initial emplacement of the
Greater Himalayan sequence in the MCT system hanging wall
(Robinson et al., 2003, 2006; Pearson and DeCelles, 2005).
Although southward channel extrusion is far from demonstrated unequivocally, none of the alternative models proposed to explain that the MioceneRecent structural
architecture of the Himalaya are as consistent with the preponderance of geologic evidence.
5.15.4.2
Long-Wavelength Topographic Variations, West

to East (Transect BB0 )
A second topographic swath shown in Figure 7 (Transect

BB0 ) begins in the Hindu Kush, traverses eastward through
the eastern Himalayan syntaxis and the northern highlands
region of Tibet, and ends just east of the Longmen Shan.
Unlike the southern and northern margins of Tibet, the
western margin is diffuse, as the high plateau dies out gradually into the high relief mountainous landscape of the
Western Kunlun Shan and Karakoram. The prominent low in
the mean elevation of this swath, approximately 200 km from
its western edge, corresponds to the dramatic valley of the
Indus River as it cuts through the Karakoram. The high dynamic range of the profile between 200 and 1000 km east of
its western edge represents the rugged topography of the
western plateau in the structurally complex region where the
active Karakoram, Altyn Tagh, and Karakax fault systems
converge (Burtman and Molnar, 1993; Robinson, 2009;
Burtman, 2010). High-frequency relief diminishes substantially over the northern highlands of the plateau but then
increases again as the mean elevation begins a gradual decrease from c. 5000 m to slightly less than 4000 m before
dropping precipitously at the Longmen Shan escarpment.
5.15.4.2.1
361
Does the gradual west-to-east decrease in

mean elevation indicate eastward flow of the
middle crust?
One of the most important geomorphic observations in the

eastern Tibetan Plateau is that the region displays an elevated,
low-relief landscape (Figure 9) that has been heavily dissected
by river systems (e.g., Kirby et al., 2002; Clark et al., 2004;
Schoenbohm et al., 2004; Clark et al., 2006). This relict
landscape, with a preserved relief of no more than a few
hundred meters, is generally marked by bedrock erosional
surfaces that are sometimes covered by a thin veneer of Tertiary or Quaternary sediments (Clark et al., 2006). On Transect
BB0 , the elevations of this landscape range from near 5000 m
in the west to approximately 4500 m in the east, but regions to
the north and south of the Longmen Shan show a greater
variation in elevation. This is shown by Transect JJ0
(Figure 8), in which the relict landscape elevation is comparable to the profile mean and drops from c. 4000 m to c.
2000 m over a distance of approximately 150 km. In fact, the
southeast plateau margin is everywhere marked by a very
gradual decline in mean elevation, although the local relief is
high because the plateau gradient is parallel to the flow directions of deeply incised river systems (e.g., the Mekong, Salween, and Yangtze), which follow Cenozoic strike-slip faults
as they drain off the plateau to the Bay of Bengal and the
South China Sea (Clark et al., 2004; Ouimet et al., 2010).
Clark and Royden (2000), Clark et al. (2004), Schoenbohm et al. (2006), and Royden et al. (2008) employed this
gradual slope in the low-relief surface to argue in favor of a
channel tunneling model (Figure 5(a)) in which the eastward
flow of a fluid Tibetan middle crust driven by pressure gradients imposed by overlying crust of varying thickness
helped accommodate lateral growth of the plateau. The
channel tunneling model is generally supported by the teleseismic receiver function analyses of Xu et al. (2007) for SE
Tibet that imply the existence of a mid-crustal low velocity
zone comparable to that in southern Tibet. In the original
model, which diverges markedly from classical escape models
for Tibet (e.g., Tapponnier et al., 2001; Replumaz and
Tapponnier, 2003), the major surface-breaking strike-slip fault
systems in this region are inferred to be restricted to the elastic
upper crust, although one could imagine a kinematic linkage
(cf., Schoenbohm et al., 2006); some of the larger fault systems may continue as steep shear zones to depth and segment
the mid-crustal fluid channel without disrupting its conductivity. Bai et al. (2010) presented magnetotelluric data indicative of high conductivity zones in the middle crust of eastern
Tibet, south of the Longmen Shan and adjacent Sichuan
Basin. They were interpreted as regions of crust weakened by
high fluid contents or partial melts. If these zones represent
the lower crustal flow channels envisioned by Clark and
Royden (2000), such features appear to be laterally discontinuous and possibly bound by the down-dip projections
of major transcurrent structures such as the Red River and
Xiaojiang faults. A consistent result was obtained by Huang
et al. (2009b) based on the seismic velocity structure of the
region. Clark and Royden (2000) also suggested that the steep
Longmen Shan escarpment, west of the Sichuan Basin
(Transect II0 , Figure 8), as well as the steep northern
boundaries of the plateau against the Tarim and Qaidam
362
basins (Transects GG0 and HH0 , Figure 8) are evidence for

thick, rigid crust beneath these basins serving as blockades
that divert the lateral flow of fluid Tibetan crust to the region
between the Qaidam and Sichuan basins and to the southeast
between the eastern Himalayan syntaxis and the Sichuan Basin
(Figure 1). This model was further elaborated by Clark et al.
(2005a) and Cook and Royden (2008). Stuwe et al. (2008)
and Dayem et al. (2009a; b) explored how strength discontinuities between Tibet and the regions to the north could
have helped concentrate strain on major thrust and transcurrent fault systems. Just as the channel expulsion model is not
accepted by all researchers working in Tibet, neither is the
channel tunneling model first proposed by Clark and Royden
(2000). For example, although Liu-Zeng et al. (2008) recognized the relict low-relief landscape within the eastern plateau,
they questioned whether or not it predated plateau uplift and
could be used as a passive marker to construct an argument of
eastward flow of mid-crustal channel. A number of authors
(Holt, 2000; Flesch et al., 2005; Sol et al., 2007) have argued
for strong coupling between the upper crustal and mantle
strain fields beneath Tibet, which can be interpreted to be
inconsistent with upper-crustal decoupling from a weak
middle crust, but may instead simply imply a similarity of farfield stresses that produce compatible deformation fields in
the upper crust and mantle (cf. Klemperer, 2006). More recently, Copley et al. (2011) presented numerical experiments
that assume a strong coupling between the down going Indian
plate and the Tibetan lower crust and can reproduce the observed deformation field in southern and southeastern Tibet
without appealing to channel flow. Of course, one viable
model does not preclude the validity of another, in this case
either channel tunneling or channel extrusion. Thermomechanical modeling by Cook and Royden (2008) demonstrated how a fluid lower crust can develop in front of an
on-coming rigid indenter and how lateral flow can be induced. However, using a different approach to thermomechanical modeling, Godard et al. (2009) showed how
escarpments such as that of the Longmen Shan could be
produced by wholesale gravitational flow of overthickened
crust, enabled by focused erosion at the plateau margin,
without the existence of a discrete channel of low-viscosity
middle crust. Rey et al. (2010) argued that the extent of eastward channel flow invoked by Clark and Royden (2000) and
others more than 1000 km seemed improbable, if not
impossible. At the same time, Rey and coworkers seemed to
accept the likelihood of southward extrusion. Suffice it to say
that the mode and extent of eastward middle crustal channeling beneath southern Tibet shall remain controversial for
some time.
5.15.4.3
Short-Wavelength Topographic Variations on the

Tibetan Plateau
Superimposed on the long-wavelength topographic variations

across the plateau are short-wavelength variations related to
roughly NS-striking normal faults and kinematically linked,
conjugate transcurrent faults (striking NWSE or NESW) that
collectively accommodate EW extension of the plateau interior (Armijo et al., 1986; Taylor and Yin, 2009). The result is
a number of small and discontinuous rift basins scattered
across the central and southern plateau that are flanked by

local uplifts with isolated peak elevations up to more than
7000 m (Figures 2 and 3). The highest concentration of these
rift basins is in regions of the plateau with mean elevations
more than 5000 m. Although most of the related structures are
presently active, their initiation ages are largely unknown. An
increasing body of evidence suggests, however, that the extensional regime that now characterizes most of the central
and southern plateau interior was established by at least
1413.5 Ma (e.g., Coleman and Hodges, 1995; Blisniuk et al.,
2001). Other than in some rift-flank uplifts, low-temperature
thermochronology provides little evidence of significant erosional or tectonic rock exhumation in the central Tibetan interior subsequent to the Eocene (e.g., Wang et al., 2007).
The depth to which the extensional regime of the upper
crust continues downward into the lower crust is limited; most
earthquakes associated with EW rifts in the plateau interior
have shallow (o18 km) hypocenters, and the middle and
lower crust are aseismic. Masek et al. (1994) demonstrated
how the morphology of rift-flank uplifts in Tibet implies a
low flexural rigidity for the elastic upper crust, with isostatic
compensation taking place deeper in the crust. Modeling by
Jimenez-Munt and Platt (2006) supported the argument that
only a relatively weak middle-lower crust could explain the
observed EW extension in the Tibetan upper crust given
overall convergence between India and Eurasia.
In the 1980s, researchers such as Armijo et al. (1986),
Mercier et al. (1987), and Peltzer and Tapponnier (1988) interpreted the EW extension of the Tibetan interior as a consequence of escape tectonics: Material in the plateau extended
and was displaced as rigid blocks eastward toward China.
Subsequent authors (e.g., McCaffrey and Nabelek, 1998; Yin
et al., 1999; Kapp and Guynn, 2004) appealed to longitudinal
variations in the collisional boundary conditions to relate
rifting in Tibet with the impingement of India. A recent study
of active extension in the plateau (Elliott et al., 2010) supported another long-standing hypothesis: that most of the
extension in the high plateau has been driven by GPE variations (e.g., Molnar et al., 1993). A reasonable extension of
this hypothesis is that eastward flow of a fluid middle or lower
crust beneath an elastic upper crust would result in tractional
forces compatible with EW upper crustal extension.
5.15.5
A Self-Consistent Model of the Cenozoic

Topographic Evolution of the Tibetan
Plateau, Assuming Lower and Middle Crustal
Flow
Although the geoscience community has not yet reached

consensus regarding whether or not a fluid middle or lower
crust exists beneath the Tibetan Plateau and whether or not
large-scale lateral flow of that fluid crust occurs we present
here a model for crustal deformation in Tibet that incorporates both tunneling and extrusion to illustrate its consistency
with geological and geophysical observations.
It seems likely that the region that is now southern Tibet
had already achieved significant elevation before c. 5045 Ma
collision of India and Eurasia (as reviewed by Molnar et al.,
2010). Shortly after collision, however, stresses were quickly
transmitted throughout the plateau region, such that contractional deformation along the southern (Himalayan) and
northern plateau margins began at roughly the same time (as
reviewed by Hodges, 2000; Dayem et al., 2009b).
By the Late Oligocene or Early Miocene, crustal thickening
in southern Tibet had proceeded to the point that thermal
conditions were right for the development of a fluid middle
crustal channel. By that time, something like the modern
monsoon rainfall pattern had been established and the
gravitational potential energy contrast between southern Tibet
and India was sufficient to drive development of the South
Tibetan fault system, thereby enabling gravitational potential
energy dissipation through southward channel expulsion and
monsoon-driven erosion.
It has been widely assumed that the onset of EW rifting
coincided with the plateau reaching uniformly high elevations
everywhere, that the onset of rifting was c. 811 Ma everywhere
on the plateau, and that only a catastrophic event such as the
abrupt convective removal of a dense lithospheric root
would have caused large-scale uplift over such a broad area
(e.g., Molnar et al., 1993). However, we know so little about
the initiation ages of most of the Tibetan rifts that it is impossible to evaluate whether or not they formed over a brief
time interval, though the available data suggest a range of
perhaps as much as 10 million years (c. 188 Ma). (Many of
the available constraints are reviewed by Lee et al. (2011),
though those authors reject age estimates older than c. 14 Ma
and younger than c. 10 Ma). There is no direct evidence to
support an argument that Tibet abruptly achieved its high
elevation. Plate reconstruction exercises performed by Molnar
and Stock (2009) provided evidence that IndiaEurasia convergence slowed significantly at c. 20 Ma and has stayed
roughly constant for the past c. 11 million years. They suggested that the sudden exertion of an outward force by Tibet
could provide a reasonable explanation of the change in
convergence, and implied that such an event could correspond
to the removal of a lithospheric root and be temporally associated with the large-scale uplift of Tibet. However, a different plate reconstruction exercise by Copley et al. (2010)
showed no evidence for Miocene slowing of IndiaAsia convergence that might be related to an abrupt change in the
evolution of Tibet. Regardless by the Middle Miocene, initiation of EW extension in the upper crust of Tibet probably
marked the beginning of lateral EW flow of material within a
mid-crustal channel of at least limited areal extent. It seems
likely that a fluid middle crust was widespread beneath Tibet
by Late Miocene time, and the fluid layer may have extended
to the base of the crust in northern Tibet, north of the
northern edge of the subducted Indian Plate (Owens and
Zandt, 1997; Nabelek et al., 2009). Once these fluid zones
were established, the redistribution of mass under the influence of gravitational potential energy gradients began in
earnest. This redistribution is manifested today in two longwavelength topographic characteristics of the plateau interior:
(1) enormous EW corrugations (Transect AA0 , Figure 7);
and (2) a general west to east lowering of the mean elevation
of the plateau related to the evacuation of the middle crustal
channel to the east and southeast off the plateau margin
(Transect BB0 , Figure 7). This eastward flow was kinematically linked to strike-slip and normal faulting in the upper
363
elastic crust. The flow was bound to the north by rigid crust
beneath the Tarim and Qaidam basins, where strain was localized on major transcurrent fault systems and transpressional mountain ranges marking the plateau margin.
Similarly, rigid crust beneath the Sichuan Basin served to divert crustal flow to the north and south, but an upstream
localization of strain at the western basin margin produced the
Longmen Shan (Kirby and Ouimet, 2011). Considerable attention has been focused on the strain field in the Longmen
Shan especially how late Cenozoic strain has been partitioned between the upper and middle crust in light of the
devastating effects of the 2008 Wenchuan (M 7.9) earthquake (Burchfiel et al., 2008; Kirby et al., 2008; Hubbard and
Shaw, 2009). If lateral middle-lower crustal flow remains an
important tectonic process in Eastern Tibet, the Neogene and
Quaternary upper crustal deformation in the Longmen Shan
can be explained as an upper crustal manifestation of localized
uplift and concomitant shortening related to the impeded
eastward flow of Tibetan crust (cf., Burchfiel et al., 2008).
Questions remain regarding when or if southward channel
extrusion ceased after eastward evacuation of the south Tibetan middle crust began. Most researchers presently favor
models in which channel extrusion is an early to middle
Miocene phenomenon. However, all of the tectonic and climatic ingredients necessary for southward extrusion still exist
today and the available geologic and geodetic data do not
preclude it (Hodges et al., 2001; Hodges, 2006). If southward
extrusion has been continual since the Middle Miocene, the
erosion rate at the Himalayan front has roughly balanced
the time-integrated rate of flux of material toward the front, so
that the position of the plateau margin has remained roughly
stationary over time.
5.15.6
Feedbacks among Middle-Lower Crustal

Flow, Landscape Evolution, and Climate
Orogenic plateaus develop as a consequence of crustal thickening and, in the eyes of many, the delamination or convective
removal of a dense lithospheric root. Both of these processes
raise temperatures in the middle and lower crust by concentrating radioactive heat-producing elements and by increasing
mantle heat input to the base of the plateau crust. A consequent, dramatic thermal weakening of the crust sets the
stage for middle or lower crustal flow, and gravitational potential energy gradients related to plateau uplift provide a
motive for flow (eqn [3]) and strongly influence the directionality of that flow. In effect, although lithospheric plate
interactions establish orogenic systems, the dynamics of orogenesis are strongly influenced by local variations in gravitational potential energy when an orogenic system becomes
fully formed. That influence is especially important when the
middle and lower crust are capable of lateral flow. In particular, depending on the magnitude of forces related to plate
convergence, DGPE may largely drive the ways in which orogenic plateaus grow.
The principal impact of middle or lower crustal flow on the
interior topography of plateaus is to limit local relief as the
existence of a weak channel at depth reduces the effective
elastic thickness of the upper crust (Bird, 1991; Masek et al.,
364
1994). However, closer analysis of the Tibetan Plateau interior

reveals subtle, long-wavelength topographic variations that are
likely related to the flow field. Although the notion that the
EW corrugations in Tibet are related to buckling above a fluid
substrate might be debatable compare Bendick et al. (2008)
and Pullen et al. (2011) the long-wavelength slope in the
mean elevation of the plateau from west to east can be interpreted with some confidence as a robust signature of eastward
evacuation of a crustal channel. Rugged, high-relief topography develops along plateau margins that are bound by rigid
crust that impede flow, as evidenced by the Longmen Shan,
Kunlun Shan, and Qilian Shan. It is now well-established that
the existence of the Tibetan Plateau exerts considerable control
on the weather patterns of South and East Asia and the
strength of the Indian and perhaps South Asian monsoons
(Hahn and Manabe, 1975; Kutzbach et al., 1993; An et al.,
2001; Molnar, 2005; Molnar et al., 2010). Thermomechanical
models of the Himalaya (Beaumont et al., 2001; Beaumont
et al., 2004) indicate how focused erosion along the southern
Tibetan Plateau margin could attract a mid-crustal fluid
channel toward the south, a process that may conspire with
shortening across the mountain front to produce the highest
mountains on Earth. More recently, Westaway (2009) has
suggested that the East Asian monsoon acts as a similar attractor for southeastward flow of the channel, where mass is
removed from the plateau in part by the monsoon-driven
incision of the major rivers draining the southeastern plateau.
5.15.7
Conclusions
Although questions remain regarding the rheology and dynamics of the middle and lower crust beneath the Tibetan
Plateau, the authors are impressed by the overall consistency
of the emerging geologic and geophysical database for the late
Cenozoic HimalayanTibetan orogenic system with tectonic
models that include crustal flow. The authors favor models in
which the upper crust is largely decoupled from the middle
and lower crust below (e.g., Royden et al., 2008). Although
lower- and middle-crustal dynamics may be influenced by
NS shortening related to collision, much of the late Cenozoic
strain at that level reflects a response to gravitational forces.
The authors are persuaded by the available evidence that fluid
middle crust of southern Tibet representing recycled, accreted Indian crust flowed southward and extruded at the
Himalayan front, attracted there by a steep gravitational potential energy gradient and focused erosion. This energy-dissipative mechanism appears to have been most active in Early
to Middle Miocene time, but its persistence into the more
recent past remains controversial. Eastward flow of material in
the channel toward regions of lower GPE began soon thereafter, and abundant mineral cooling ages from the northeastern plateau that are likely related to this process record the
arrival of the flow front between 13 and 8 Ma (e.g., Kirby et al.,
2002; Clark et al., 2005b; Ouimet et al., 2010; Zheng et al.,
2010; Lease et al., 2011).
Although deeper levels deformed by ductile flow, the upper
elastic crust responded by a complex combination of strikeslip and extensional faulting that collectively resulted in: (1)
eastwest extension in the western, central, and southern
plateau; (2) shortening in the northeastern plateau; and (3)

the translation of narrow-upper crustal blocks around the
eastern Himalayan syntaxis on arcuate strike-slip fault systems.
This pattern of faulting, established in Middle Miocene time,
resulted in upper crustal kinematics that are indistinguishable
from the modern strain field as measured by geodesy (Zhang
et al., 2004). Thus, the eastward displacement of Tibetan crust
is manifested by kinematically consistent upper crustal faulting and lower and middle crustal flow.
The relict, low-relief surfaces mapped by Clark et al. (2004)
place an important constraint on the relationship between
tectonics and landscape evolution across the plateau: In general, the rise of the Tibetan Plateau from its early Cenozoic
elevation to its current mean elevation of 5000 m most
likely sometime between 20 and 10 Ma was largely passive,
and topographic variations within the plateau are largely
manifestations of deformational processes related to viscous
flow of the middle crust. Along the plateau margins, landscapes have been shaped by three processes acting in concert
and amplifying one another: (1) convergent tectonics related
to IndiaEurasia collision; (2) lateral crustal flow driven by
local pressure gradients; and (3) a climate that is dominated
by spatially variable monsoon rainfall.
Acknowledgments
This work has been influenced by many discussions over the
years with the likes of Doug Burbank, Clark Burchfiel, Marin
Clark, Arjun Heimsath, Jose Hurtado, Eric Kirby, Simon
Klemperer, Peter Molnar, Wiki Royden, Lindsay Schoenbohm,
Mike Searle, and Kelin Whipple. KVH would like to thank his
current and former students and postodoctoral advisees at
both MIT and Arizona State for keeping him sharp. Our work
has been funded by National Science Foundation grants from
the Tectonics and Continental Dynamics programs, most
recently EAR-0711140, EAR-0838112, and EAR-1007929. We
thank Eric Kirby and William Ouimet for their insightful reviews of the manuscript.
References
Adams, B., Dietsch, C., Owen, L.A., Caffee, M.W., Spotila, J., Haneberg, W.C.,
2009. Exhumation and incision history of the Lahul Himalaya, northern India,
based on (U-Th)/He thermochronometry and terrestrial cosmogenic nuclide
methods. Geomorphology 107, 285299.
Alle`gre, C.J., Courtillot, V., Tapponnier, P., et al., 1984. Structure and evolution of
the Himalaya-Tibet orogenic belt. Nature 307, 1722.
An, Z., Kutzbach, J.E., Prell, W.L., Porter, S.C., 2001. Evolution of Asian monsoons
and phased uplift of the Himalaya-Tibetan plateau since Late Miocene times.
Nature 411, 6266.
Armijo, R., Tapponier, P., Mercier, J., Han, T., 1986. Quaternary extension in
southern Tibet: field observations and tectonic implications. Journal of
Artyushkov, E.V., 1973. Stresses in the lithosphere caused by crustal thickness
inhomogeneities. Journal of Geophysical Research 78, 76757708.
Ashish, Padhi, A., Rai, S.S., Gupta, S., 2009. Seismological evidence for shallow
crustal melt beneath the Garhwal High Himalaya, India: implications for the
Himalayan channel flow. Geophysical Journal International 177, 11111120.
Bai, D.H., Unsworth, M.J., Meju, M.A., et al., 2010. Crustal deformation of the
eastern Tibetan plateau revealed by magnetotelluric imaging. Nature Geoscience
3, 358362.
Bally, A. W., Allen, C. R., Geyer, R. E., et al. 1980. Notes on the Geology of
Tibet and Adjacent Areas Report of the American Plate Tectonics
Delegation to the Peoples Republic of China, U.S. Geol. Surv. Open File Rep.
80-501.
Barreiro, J.G., Catalan, J.R.M., Fernandez, R.D., Arenas, R., Garcia, F.D., 2010.
Upper crust reworking during gravitational collapse: the Bembibre-Pico Sacro
detachment system (NW Iberia). Journal of the Geological Society 167,
769784.
Barros, A.P., Joshi, M., Putkonen, J., Burbank, D.W., 2000. A study of the 1999
monsoon rainfall in a mountainous region in Central Nepal using TRMM
products and raingauge observations. Geophysical Research Letters 27,
36833686.
Barros, A.P., Lang, T.J., 2003. Monitoring the monsoon in the Himalayas:
observations in central Nepal, June 2001. Monthly Weather Review 131,
14081427.
flows: 1. Numerical models with implications to the tectonics of the
HimalayanTibetan orogen. Journal of Geophysical Research 109.
http://dx.doi.org/10.1029/2003JB002809.
Bendick, R., Flesch, L., 2007. Reconciling lithospheric deformation and lower
crustal flow beneath central Tibet. Geology 35, 895898.
Bendick, R., McKenzie, D., Etienne, J., 2008. Topography associated with crustal
flow in continental collisions, with application to Tibet. Geophysical Journal
Bird, P., 1991. Lateral extrusion of lower crust from under high topography, in the
isostatic limit. Journal of Geophysical Research 96, 10,27510,286.
Bird, P., Toksoz, M.N., 1977. Srong attenuation of Rayleigh waves in Tibet. Nature
266, 161163.
Blisniuk, P.M., Hacker, B.R., Glodny, J., et al., 2001. Normal faulting in central Tibet
since at least 13.5 Myr ago. Nature 412, 628632.
Blythe, A.E., Burbank, D.W., Carter, A., Schmidt, K., Putkkonen, J., 2007. PlioQuaternary exhumation history of the central Nepalese Himalaya: 1. Apatite
and zircon fission track and apatite (U-Th)/He analyses. Tectonics 26.
http://dx.doi.org/10.1029/2006TC001990.
Bojar, A.-V., Fritz, H., Nicolescu, S., Bregar, M., Gupta, R.P., 2005. Timing and
mechanisms of Central Himalayan exhumation: discriminating between tectonic
and erosion processes. Terra Nova 17, 427433.
Bollinger, L., Henry, P., Avouac, J.P., 2006. Mountain building in the Nepal
Himalaya: thermal and kinematic model. Earth and Planetary Science Letters
244, 5871.
Bookhagen, B., Burbank, D.W., 2006. Topography, relief, and TRMM-derived rainfall
variations along the Himalaya. Geophysical Research Letters 33. http://dx.doi.org
/10.1029/2006GL026037.
Bookhagen, B., Burbank, D.W., 2010. Toward a complete Himalayan hydrological
budget: Spatiotemporal distribution of snowmelt and rainfall and their impact on
river discharge. Journal of Geophysical Research 115, F03019.
Brace, W.F., Kohlstedt, D.L., 1980. Limits on lithospheric stress imposed
by laboratory experiments. Journal of Geophysical Research 85,
62486252.
Brown, D., Juhlin, C., 2006. A possible lower crustal flow channel in the Middle
Urals based on reflection seismic data. Terra Nova 18, 18.
Brown, R.L., Gibson, H.D., 2006. An argument for channel flow in the southern
Canadian Cordillera and comparison with Himalayan tectonics. In: Law, R.,
Searle, M., Godin, L. (Eds.), Channel Flow, Ductile Extrusion, and Exhumation
of Lower-Middle Crust in Continental Collision Zones. Geological Society
Special Publication 268, London.
Burbank, D.W., Blythe, A.E., Putkonen, J.K., et al., 2003. Decoupling of erosion and
precipitation in the Himalaya. Nature 426, 652655.
Burchfiel, B.C., Chen, Z., Hodges, K.V., Liu, Y., Royden, L.H., Deng, C., Xu, J.,
1992. The South Tibetan Detachment System, Himalayan Orogen: Extension
Contemporaneous With and Parallel to Shortening in a Collisional Mountain
Belt. Geological Society of America, Boulder, CO.
Burchfiel, B.C., Deng, Q., Molnar, P., Royden, L.H., Wang, Y., Zhang, P., Zhang, W.,
1989. Intracrustal detachments within zones of intracontinental deformation.
Geology 17, 748752.
Burchfiel, B.C., Royden, L.H., 1985. North-south extension within the convergent
Himalayan region. Geology 13, 679682.
Burchfiel, B.C., Royden, L.H., van der Hilst, R.D., et al., 2008. A geological and
geophysical context for the Wenchuan earthquake of 12 May 2008, Sichuan,
Peoples Republic of China. GSA Today 18, 411.
365
Burg, J.P., Brunel, M., Gapais, D., Chen, G.M., Liu, G.H., 1984. Deformation of
leucogranites of the crystalline Main Central Sheet in southern Tibet (China).
Journal of Structural Geology 6, 535542.
Burgmann, R., Dresen, G., 2008. Rheology of the lower crust and upper mantle:
evidence from rock mechanics, geodesy, and field observations. Annual Reviews
of Earth and Planetary Sciences 36, 531567.
Burtman, V.S., 2010. Tien Shan, Pamir, and Tibet: history and geodynamics of
Phanerozoic oceanic basins. Geotectonics 44, 388404.
Burtman, V.S., Molnar, P., 1993. Geological and geophysical evidence for deep
subduction of continental crust beneath the Pamir. Geological Society of
America Special Paper 281, 176.
Byerlee, J.D., 1978. Friction of rocks. Pure and Applied Geophysics 116, 615626.
Caby, R., Pecher, A., LeFort, P., 1983. Le grande chevauchement central himalayen:
nouvelles donnees sur le metamorphisme inverse a` la base de la Dalle du Tibet.
Revue de Geographie physiqueau et Geologie dynamique 24, 89100.
Cagnard, F., Durrieu, N., Gapais, D., Brun, J.P., Ehlers, C., 2006. Crustal thickening
and lateral flow during compression of hot lithospheres, with particular reference
to Precambrian times. Terra Nova 18, 7278.
Caldwell, W.B., Klemperer, S.L., Rai, S.S., Lawrence, J.F., 2009. Partial melt in the
upper-middle crust of the northwest Himalaya revealed by Rayleigh wave
dispersion. Tectonophysics 477, 5865.
Cattin, R., Avouac, J.P., 2000. Modeling mountain building and the seismic cycle in
the Himalaya of Nepal. Journal of Geophysical Research 105, 1338913407.
Chen, Q., Freymueller, J.T., Yang, Z., Xu, C., Jiang, W., Wang, Q., Liu, J., 2004.
Spatially variable extension in southern Tibet based on GPS measurements.
Journal of Geophysical Research 109. http://dx.doi.org/10.1029/2002JB002350.
Clark, M.K., Bush, J.W.M., Royden, L.H., 2005a. Dynamic topography produced by
lower crustal flow against rheological strength heterogeneities bordering the
Tibetan Plateau. Geophysical Journal International 162, 575590.
Clark, M.K., Farley, K.A., Zheng, D.W., Wang, Z.C., Duvall, A.R., 2010. Early
Cenozoic faulting of the northern Tibetan Plateau margin from apatite (U-Th)/He
ages. Earth and Planetary Science Letters 296, 7888.
Tang, W., 2005b. Late Cenozoic uplift of southeastern Tibet. Geology 33,
525528.
Clark, M.K., Royden, L.H., 2000. Topographic ooze: Building the eastern margin of
Clark, M.K., Royden, L.H., Whipple, K.X., et al., 2004. Surface uplift, tectonics, and
erosion of eastern Tibet from large-scale drainage patterns. Tectonics 23.
http://dx.doi.org/10.1029/2002TC001402.
Clark, M.K., Royden, L.H., Whipple, K.X., Burchfiel, B.C., Zhang, X., Tang, W.,
eastern Tibetan Plateau. Journal of Geophysical Research 111. http://dx.doi.org/
10.1029/2005JF000294.
Coleman, M., Hodges, K., 1995. Evidence for Tibetan Plateau uplift before 14 Myr
ago from a new minimum age for east-west extension. Nature 374, 4952.
Cook, K.L., Royden, L.H., 2008. The role of crustal strength variations in shaping
orogenic plateaus, with application to Tibet. Journal of Geophysical Research
Solid Earth 113. http://dx.doi.org/10.1029/2007JB005457.
Copley, A., Avouac, J.P., Royer, J.Y., 2010. India-Asia collision and the Cenozoic
slowdown of the Indian plate: implications for the forces driving plate motions.
Journal of Geophysical Research Solid Earth 115. http://dx.doi.org/10.1029/
2009jb006634.
Copley, A., Avouac, J.-P., Wernicke, B.P., 2011. Evidence for mechanical coupling
and strong Indian lower crust beneath southern Tibet. Nature 472, 7981.
Cottle, J.M., Waters, D.J., Riley, D., Beyssac, O., Jessup, M.J., 2011. Metamorphic
history of the South Tibetan Detachment System, Mt. Everest region, revealed by
RSCM thermometry and phase equilibria modeling. Journal of Metamorphic
Geology 29, 561582.
Dalmayrac, B., Molnar, P., 1981. Parallel thrust and normal faulting in Peru and
constraints on the state of stress. Earth and Planetary Science Letters 55,
473481.
Dayem, K.E., Houseman, G.A., Molnar, P., 2009a. Localization of shear along a
lithospheric strength discontinuity: application of a continuous deformation
model to the boundary between Tibet and the Tarim Basin. Tectonics 28.
http://dx.doi.org/10.1029/2008TC002264.
Dayem, K.E., Molnar, P., Clark, M.K., Houseman, G.A., 2009b. Far-field lithospheric
deformation in Tibet during continental collision. Tectonics 28. http://dx.doi.org/
10.1029/2008TC002344.
Dewey, J.F., 1988. Extensional collapse of orogens. Tectonics 7, 11231139.
Dewey, J.F., Shackleton, R.M., Chang, C., Sun, Y., 1988. The tectonic evolution of
the Tibetan Plateau. Philosophical Transactions of the Royal Society of London,
Series A 327, 379413.
366
De`zes, P., Vannay, J.C., Steck, A., Bussy, F., Cosca, M., 1999. Synorogenic
extension: quantitative constraints on the age and throw of the Zanskar
shear zone (NW Himalaya). Geological Society of America Bulletin 111,
364374.
Dong, G., Yi, C., Chen, L., 2010. An introduction to the physical geography of the
Qiangtang Plateau: a frontier for future geoscience research on the Tibetan
Plateau. Physical Geography 31, 475492.
Dorr, W., Zulauf, G., 2010. Elevator tectonics and orogenic collapse of a Tibetanstyle plateau in the European Variscides: the role of the Bohemian shear zone.
International Journal of Earth Sciences 99, 299325.
Dortch, J.M., Dietsch, C., Owen, L.A., Caffee, M.W., Ruppert, K., 2011. Episodic
fluvial incision of rivers and rock uplift in the Himalaya and Transhimalaya.
Dumond, G., Goncalves, P., Williams, M.L., Jercinovic, M.J., 2010. Subhorizontal
fabric in exhumed continental lower crust and implications for lower crustal
flow: athabasca granulite terrane, western Canadian Shield. Tectonics 29.
http://dx.doi.org/10.1029/2009TC002514.
Duvall, A.R., Clark, M.K., 2010. Dissipation of fast strike-slip faulting within and
beyond northeastern Tibet. Geology 38, 223226.
Edwards, M.A., Harrison, T.M., 1997. When did the roof collapse? Late Miocene
north-south extension in the high Himalaya revealed by Th-Pb monazite dating
of the Khula Kangri granite. Geology 25, 543546.
Elliott, J.R., Walters, R.J., England, P.C., Jackson, J.A., Li, Z., Parsons, B., 2010.
Extension on the Tibetan plateau: recent normal faulting measured by InSAR and
body wave seismology. Geophysical Journal International 183, 503535.
England, P., Houseman, G., 1986. Finite strain calculations of continental
deformation 2. comparison with the India-Asia collision zone. Journal of
Geophysical Research- Solid Earth and Planets 91, 36643676.
England, P., Molnar, P., 1997. Active deformation of Asia: from kinematics to
dynamics. Science 278, 647650.
England, P.C., Houseman, G.A., 1988. The mechanics of the Tibetan Plateau.
In: Shackleton, R.M., Dewey, J.F., Windley, B.F. (Eds.), Tectonic Evolution of the
Himalayas and Tibet. The Royal Society, London.
England, P.C., Jackson, J.A., 1989. Active deformation of the continents. Annual
Fielding, E.J., Isacks, B.L., Barazangi, M., Duncan, C., 1994. How flat is Tibet?
Geology 22, 163167.
Fleitout, L., Froidevaux, C., 1982. Tectonics and topography for a lithosphere
containing density heterogeneities. Tectonics 1, 2156.
Flesch, L., Holt, W., Silver, P., Stephenson, M., Wang, C.-Y., Chan, W., 2005.
Constraining the extent of crust-mantle coupling in central Asia using GPS,
geologic, and shear wave splitting data. Earth and Planetary Science Letters 238,
248268.
Flesch, L.M., Holt, W.E., Haines, A.J., Wen, L., Shen-Tu, B., 2007. The dynamics of
western North America: stress magnitudes and the relative role of gravitational
potential energy, plate interaction at the boundary, and basal tractions.
Francheteau, J., Jaupart, C., Xian, J.S., et al., 1984. High heat flow in southern
Tibet. Nature 307, 3236.
Gabet, E.J., Burbank, D.W., Pratt-Sitaula, B., Putkonen, J., Bookhagen, B., 2008.
Modem erosion rates in the High Himalayas of Nepal. Earth and Planetary
Gan, W.J., Zhang, P.Z., Shen, Z.K., et al., 2007. Present-day crustal motion within
the Tibetan Plateau inferred from GPS measurements. Journal of Geophysical
Research- Solid Earth and Planets 112. http://dx.doi.org/10.1029/
2005JB004120.
Garzanti, E., Vezzoli, G., Ando, S., Lave, J., Attal, M., France-Lanord, C., DeCelles,
P., 2007. Quantifying sand provenance and erosion (Marsyandi River, Nepal
Himalaya). Earth and Planetary Science Letters 258, 500515.
Gee, D.G., Juhlin, C., Pascal, C., Robinson, P., 2010. Collisional Orogeny in the
Scandinavian Caledonides (COSC). GFF 132, 2944.
Gerbault, M., Martinod, J., Herail, G., 2005. Possible orogeny-parallel lower
crustal flow and thickening in the Central Andes. Tectonophysics 399,
5972.
Ghosh, A., Holt, W.E., Flesch, L.M., Haines, A.J., 2006. Gravitational potential
energy of the Tibetan Plateau and the forces driving the Indian plate. Geology
34, 321324.
Girardeau, J., Marcoux, J., Alle`gre, C.J., et al., 1984. Tectonic environment and
geodynamic significance of the Neo- Cimmerian Donqiao Ophiolite, BangongNujiang suture zone, Tibet. Nature 307, 2731.
Godard, V., Cattin, R., Lave, J., 2009. Erosional control on the dynamics of lowconvergence rate continental plateau margins. Geophysical Journal International
179, 763777.
Goetze, C., Evans, B., 1979. Stress and temperature in the bending lithosphere as
constrained by experimental rock mechanics. Geophysical Journal of the Royal
Astronomical Society 59, 463478.
Grujic, D., 2006. Channel flow and continental collision tectonics: an overview.
In: Law, R., Searle, M., Godin, L. (Eds.), Channel Flow, Ductile Extrusion, and
Exhumation of Lower-Middle Crust in Continental Collision Zones. Geological
Society Special Publication 268, London.
Guo, Z., Gao, X., Yao, H.J., Li, J., Wang, W.M., 2009. Midcrustal low-velocity layer
beneath the central Himalaya and southern Tibet revealed by ambient noise array
tomography. Geochemistry Geophysics Geosystems 10. http://dx.doi.org/
10.1029/2009GC002458.
Hahn, D.G., Manabe, S., 1975. The role of mountains in the south Asian monsoon
circulation. Journal of Atmospheric Science 32, 15151541.
Harkins, N., Kirby, E., Shi, X., Wang, E., Burbank, D., Chun, F., 2010. Millennial
slip rates along the eastern Kunlun fault: Implications for the dynamics of
intracontinental deformation in Asia. Lithosphere 2, 247266.
Harris, N., 2007. Channel flow and the Himalayan-Tibetan orogen: a critical review.
Journal of the Geological Society of London 164, 511523.
Harrison, T.M., 2006. Did the Himalayan Crystallines extrude partially molten from
beneath the Tibetan Plateau? In: Law, R., Searle, M., Godin, L. (Eds.), Channel
Flow, Ductile Extrusion, and Exhumation of Lower-Middle Crust in Continental
Collision Zones. Geological Society Special Publication 268, London.
Hatcher, R.D., Merschat, A.J., 2006. The Appalachian Inner Piedmont: an exhumed,
strike-parallel, tectonically forced orogenic channel. In: Law, R., Searle, M.,
Godin, L. (Eds.), Channel Flow, Ductile Extrusion, and Exhumation of LowerMiddle Crust in Continental Collision Zones. Geological Society Special
Publication 268, London.
Herman, F., Copeland, P., Avouac, J.P., et al., 2010. Exhumaiton, crustal
deformation, and thermal structure of the Nepal Himalaya derived from the
inversion of thermochronological and thermobarometric data and modeling of
the topography. Journal of Geophysical Research 115. http://dx.doi.org/10.1029/
2008JB006126.
Herren, E., 1987. Zanskar shear zone: northeast-southwest extension within the
Higher Himalayas (Ladakh, India). Geology 15, 409413.
Hodges, K., Wobus, C., Ruhl, K., Schildgen, T., Whipple, K., 2004. Quaternary
deformation, river steepening, and heavy precipitation at the front of the Higher
Himalayan ranges. Earth and Planetary Science Letters 220, 379389.
Hodges, K.V., 2000. Tectonics of the Himalaya and southern Tibet from two
perspectives. Geological Society of America Bulletin 112, 324350.
Hodges, K.V., 2006. A synthesis of the Channel Flow-Extrusion hypothesis as
developed for the Himalayan-Tibetan orogenic system. In: Law, R., Searle, M.,
Godin, L. (Eds.), Channel Flow, Ductile Extrusion, and Exhumation of LowerMiddle Crust in Continental Collision Zones. Geological Society Special
Publication 268, London.
Hodges, K.V., Hurtado, J.M., Whipple, K.X., 2001. Southward extrusion of Tibetan
crust and its effect on Himalayan tectonics. Tectonics 20, 799809.
Hodges, K.V., Parrish, R., Housh, T., Lux, D., Burchfiel, B.C., Royden, L., Chen, Z.,
1992. Simultaneous Miocene extension and shortening in the Himalayan
orogen. Science 258, 14661470.
Hodges, K.V., Parrish, R.R., Searle, M.P., 1996. Tectonic evolution of the central
Annapurna Range, Nepalese Himalayas. Tectonics 15, 12641291.
Holt, W.E., 2000. Correlated crust and mantle strain fields in Tibet. Geology 28,
6770.
Houseman, G., England, P., 1993. Crustal thickening versus lateral expansion in the
India-Asia continental collision. Journal of Geophysical Research 98,
1223312250.
Huang, G.C., Wu, F.T., Roecker, S.W., Sheehan, A.F., 2009a. Lithospheric structure
of the central Himalaya from 3-D tomographic imaging. Tectonophysics 475,
524543.
Huang, R.Q., Wang, Z., Pei, S.P., Wang, Y.S., 2009b. Crustal ductile flow and its
contribution to tectonic stress in Southwest China. Tectonophysics 473,
476489.
Hubbard, J., Shaw, J.H., 2009. Uplift of the Longmen Shan and Tibetan plateau,
and the 2008 Wenchuan (M 7.9) earthquake. Nature 458, 194197.
Humphreys, E.D., Coblentz, D.D., 2007. North American dynamics and western U.S.
tectonics. Reviews of Geophysics 45. http://dx.doi.org/10.1029/2005RG000181.
Hurtado, J.M., Hodges, K.V., Whipple, K.X., 2001. Neotectonics of the Thakkhola
Graben and implications for Recent activity on the South Tibetan Fault System
in the central Nepalese Himalaya. Geological Society of America Bulletin 113,
222240.
Husson, L., Sempere, T., 2003. Thickening the Alltiplano crust by gravity-driven
crustal channel flow. Geophysical Research Letters 30. http://dx.doi.org/10.1029/
2002GL016877.
Jackson, M., Bilham, R., 1994. Constraints on Himalayan deformation inferred from
vertical velocity fields in Nepal and Tibet. Journal of Geophysical Research 99,
1389713912.
Jamieson, R.A., Beaumont, C., Medvedev, S., Nguyen, M.H., 2004. Crustal channel
flows: 2. Numerical models with implications for metamorphism in the
Himalayan-Tibetan orogen. Journal of Geophysical Research 109. http://dx.doi.or
g/10.1029/2003JB002811.
Jamieson, R.A., Beaumont, C., Nguyen, M.H., Culshaw, N.G., 2007. Synconvergent
ductile flow in variable-strength continental crust: Numerical models with
application to the western Grenville orogen. Tectonics 26. http://dx.doi.org/
10.1029/2006TC002036.
Jessup, M.J., Cottle, J.M., Searle, M.P., Law, R.D., Newell, D.L., Tracy, R.J., Waters,
D.J., 2008. P-T-t-D paths of Everest Series schist, Nepal. Journal of
Metamorphic Geology 26, 717739.
Jimenez-Munt, I., Platt, J.P., 2006. Influence of mantle dynamics on the topographic
evolution of the Tibetan Plateau: Results from numerical modeling. Tectonics 25.
http://dx.doi.org/10.1029/2006TC001963.
Jin, Y., McNutt, M.K., Zhu, Y., 1994. Evidence from gravity and topography data for
folding of Tibet. Nature 371, 669674.
Jones, C.H., Unruh, J.R., Sonder, L.J., 1996. The role of gravitational potential
energy in active deformation in the southwestern United States. Nature 381,
3741.
Kapp, P., Guynn, J.H., 2004. Indian punch rifts Tibet. Geology 32, 993996.
Kapp, P., Yin, A., Harrison, T.M., Ding, L., 2005. Cretaceous-Tertiary shortening,
basin development, and volcanism in central Tibet. Geological Society of
Kellett, D.A., Grujic, D., Erdmann, S., 2009. Miocene structural reorganization of the
South Tibetan detachment, eastern Himalaya: Implications for continental
collision. Lithosphere 1, 259281.
Kirby, E., Ouimet, W., 2011. Tectonic geomorphology along the eastern margin of
Tibet: insights into the pattern and processes of active deformation adjacent to
the Sichuan Basin. In: Gloaguen, R., Ratschbacher, L. (Eds.), Growth and
Collapse of the Tibetan Plateau. Geological Society, London, pp. 165188.
Special Publication 353.
Kirby, E., Reiners, P.W., Krol, M.A., et al., 2002. Late Cenozoic evolution of the
eastern margin of the Tibetan Plateau: Inferences from 40Ar/39Ar and (UTh)/He
thermochronology. Tectonics 21, 322.
Kirby, E., Whipple, K., Harkins, N., 2008. Topography reveals seismic hazard.
Nature Geoscience 1, 485487.
Kirby, S.H., 1983. Rheology of the lithosphere: Selected topics. Review of
Geophysical Space Physics 21, 14581487.
Klemperer, S.L., 2006. Crustal flow in Tibet: a review of geophysical evidence for
the physical state of the Tibetan lithosphere. In: Law, R., Searle, M., Godin, L.
(Eds.), Channel Flow, Ductile Extrusion, and Exhumation of Lower-Middle Crust
in Continental Collision Zones. Geological Society Special Publication 268,
London.
Kohlstedt, D.L., 2007. Properties of rocks and minerals constitutive equations,
rheological behavior, and viscosity of rocks. In: Gerald, S. (Ed.), Treatise on
Geophysics. Elsevier, Amsterdam.
Kohn, M.J., 2008. P-T-t data from central Nepal support critical taper and repudiate
large-scale channel flow of the Greater Himalayan Sequence. Geological Society
Kuiper, Y.D., Williams, P.F., Kruse, S., 2006. Possibility of channel flow in the
southern Canadian Cordillera: a new approach to explaining existing data.
Kutzbach, J.E., Prell, W.L., Ruddiman, W.F., 1993. Sensitivity of Eurasian climate to
surface uplift of the Tibetan Plateau. Journal of Geology 100, 177190.
Le Fort, P., 1975. Himalayas: the collided range. Present knowledge of the
continental arc. American Journal of Science 275-A, 144.
Lease, R.O., Burbank, D.W., Clark, M.K., Farley, K.A., Zheng, D.W., Zhang, H.P.,
2011. Middle Miocene reorganization of deformation along the northeastern
Lee, J., Hager, C., Wallis, S.R., Stockli, D.F., Whitehouse, M.J., Aoya, M., Wang, Y.,
2011. Middle to late Miocene extremely rapid exhumation and thermal
reequilibration in the Kung Co rift, southern Tibet. Tectonics 30. http://dx.doi.org
/10.1029/2010TC002745.
Li, S., Kusky, T.M., Zhao, G., Liu, X., Zhang, G., Kopp, H., Wang, L., 2010. Twostage Triassic exhumation of HP-UHP terranes in the western Dabie orogen of
China: Constraints from structural geology. Tectonophysics 490, 267293.
Li, S., Unsworth, M.J., Booker, J.R., Wei, W., Tan, H., Jones, A.G., 2003.
Partial melt or aqueous fluid in the mid-crust of southern Tibet? Constraints
367
from INDEPTH magnetotelluric data. Geophysical Journal International 153,

289304.
Liu-Zeng, J., Tapponnier, P., Gaudemer, Y., Ding, L., 2008. Quantifying landscape
differences across the Tibetan plateau: implications for topographic relief
evolution. Journal of Geophysical Research-Earth Surface 113. http://dx.doi.org/
10.1029/2007JF000897.
Long, S., McQuarrie, N., 2010. Placing limits on channel flow: insights from the
Bhutan Himalaya. Earth and Planetary Science Letters 290, 375390.
Marone, C.J., 1998. Laboratory-derived friction laws and their application to
seismic faulting. Annual Review of Earth and Planetary Sciences 26, 643696.
Masek, J.G., Isacks, B.L., Fielding, E.J., 1994. Rift flank uplift in Tibet: evidence for
a viscous lower crust. Tectonics 13, 659667.
McCaffrey, R., Nabelek, J., 1998. Role of oblique convergence in the active
deformation of the Himalayas and southern Tibetan plateau. Geology 26,
691694.
McNamara, D.E., Owens, T.J., Walters, W.R., 1995. Observations of regional phase
propagation across the Tibetan Plateau. Journal of Geophysical Research 100,
2221522229.
Meade, B.J., 2007. Present-day kinematics at the India-Asia collision zone. Geology
35, 8184.
Medvedev, S., Beaumont, C., 2006. Growth of continental plateaus by channel
injection: models designed to address constraints and thermomechanical
consistency. In: Law, R., Searle, M., Godin, L. (Eds.), Channel Flow, Ductile
Extrusion, and Exhumation of Lower-Middle Crust in Continental Collision
Zones. Geological Society Special Publication 268, London.
Mercier, J.-L., Armijo, R., Tapponier, P., Carey-Gailhardis, E., Han, T.L., 1987.
Change from late Tertiary compression to Quaternary extension in southern Tibet
during the India-Asia collision. Tectonics 6, 275304.
Molnar, P., 2005. Mio-Pliocene growth of the Tibetan Plateau and evolution of East
Asian climate. Palaeontologia Electronica 8, 123.
Molnar, P., Boos, W.R., Battisti, D.S., 2010. Orographic controls on climate and
paleoclimate of Asia: thermal and mechanical roles for the Tibetan Plateau.
Molnar, P., Lyon-Caen, H., 1988. Some simple physical aspects of the support,
structure, and evolution of mountain belts. In: Clark, S., Burchfiel, B.C., Suppe,
J. (Eds.), Processes in Continental Lithospheric Deformation. Geological Society
of America Special Paper 218, Boulder, CO.
Molnar, P., Lyon-Caen, H., 1989. Fault-plane solutions of earthquakes and active
tectonics of the Tibetan plateau and its margins. Geophysical Journal
20 Ma and its implications for Tibetan mantle dynamics. Tectonics 28. http://d
x.doi.org/10.1029/2008TC002271.
Molnar, P., Tapponnier, P., 1975. Cenozoic tectonics of Asia: effects of a continental
Nabelek, J., Hetenyi, G., Vergne, J., et al., 2009. Underplating in the Himalaya-Tibet
Collision Zone Revealed by the Hi-CLIMB Experiment. Science 325, 13711374.
beneath southern Tibet: Synthesis of Project INDEPTH Results. Science 274,
16841688.
Ni, J., Barazangi, M., 1983. Velocities and propagation characteristics of Pn, Pg,
Sn, and Lg seismic waves beneath the Indian Shield, Himalayan Arc, Tibetan
Plateau, and surrounding regions: high uppermost mantle velocities and efficient
Sn propagation beneath Tibet. Geophysical Journal of the Royal Astronomical
Society 72, 665689.
Ouimet, W., Whipple, K., Royden, L., Reiners, P., Hodges, K., Pringle, M., 2010.
Regional incision of the eastern margin of the Tibetan Plateau. Lithosphere 2,
5063.
Ouimet, W.B., Cook, K.L., 2010. Building the central Andes through axial lower
crustal flow. Tectonics 29. http://dx.doi.org/10.1029/2009TC002460.
Owens, T.J., Zandt, G., 1997. Implications of crustal property variations for models
of Tibetan Plateau evolution. Nature 387, 3743.
Pearson, O.N., DeCelles, P.G., 2005. Structural geology and regional tectonic
significance of the Ramgarh thrust, Himalayan fold-thrust belt of Nepal.
Tectonics 24. http://dx.doi.org/10.1029/2003TC001617.
Peltzer, G., Tapponnier, P., 1988. Formation and evolution of strike-slip faults, rifts
and basins during the India-Asia collision: an experimental aproach. Journal of
Pullen, A., Kapp, P., Gehrels, G.E., Ding, L., Zhang, Q., 2011. Metamorphic rocks in
central Tibet: lateral variations and implications for crustal structure. Geological
368
Putkonen, J.K., 2004. Continuous Snow and Rain Data at 500 to 4400 m Altitude
near Annapurna, Nepal, 19992001. Arctic, Antarctic, and Alpine Research 36,
244248.
Raimondo, T., Collins, A.S., Hand, M., Walker-Hallam, A., Smithies, R.H., Evins,
P.M., Howard, H.M., 2010. The anatomy of a deep intracontinental orogen.
Tectonics 29, TC4024.
Renner, J., Evans, B., Hirth, G., 2000. On the rheologically critical melt fraction.
Replumaz, A., Tapponnier, P., 2003. Reconstruction of the deformed collision zone
between India and Asia by backward motion of lithospheric blocks. Journal of
Geophysical Research-Earth Surface 108. http://dx.doi.org/10.1029/
2001jb000661.
Rey, P.F., Teyssier, C., Whitney, D.L., 2010. Limit of channel flow in orogenic
plateaux. Lithosphere 2, 328332.
Rivers, T., 2008. Assembly and preservation of lower, mid, and upper orogenic
crust in the Grenville Province-Implications for the evolution of large hot longduration orogens. Precambrian Research 167, 237259.
Robinson, A.C., 2009. Geologic offsets across the northern Karakorum fault:
implications for its role and terrane correlations in the western HimalayanTibetan orogen. Earth and Planetary Science Letters 279, 123130.
Robinson, D.M., DeCelles, P.G., Copeland, P., 2006. Tectonic evolution of the
Himalayan thrust belt in western Nepal: implications for channel flow models.
Robinson, D.M., DeCelles, P.G., Garzione, C.N., Pearson, O.N., Harrison, T.M.,
Catlos, E.J., 2003. Kinematic model for the Main Central Thrust in Nepal.
Geology 31, 359362.
Rosenberg, C.L., Handy, M.R., 2005. Experimental deformation of partially melted
granite revisited: implications for the continental crust. Journal of Metamorphic
Geology 23, 1928.
Royden, L.H., Burchfiel, B.C., King, R.W., Wang, E., Chen, Z., Shen, F., Liu, Y.,
1997. Surface deformation and lower crustal flow in eastern Tibet. Science 276,
788790.
Royden, L.H., Burchfiel, B.C., van der Hilst, R.D., 2008. The geological evolution of
the Tibetan plateau. Science 321, 10541058.
Sachan, H.K., Kohn, M.J., Saxena, A., Corrie, S.L., 2010. The Malari leucogranite,
Garhwal Himalaya, northern India: chemistry, age, and tectonic implications.
Schoenbohm, L.M., Burchfiel, B.C., Chen, L.Z., 2006. Propagation of surface uplift,
lower crustal flow, and Cenozoic tectonics of the southeast margin of the
Schoenbohm, L.M., Whipple, K.X., Burchfiel, B.C., 2004. Geomorphic constraints
on surface uplift, exhumation, and plateau growth in the Red River region,
Yunnan Province, China. Geological Society of America Bulletin 116,
895909.
Searle, M., Simpson, R.L., Law, R.D., Parrish, R.R., Waters, D.J., 2003. The
structural geometry, metamorphic and magmatic evolution of the Everest massif,
High Himalaya of NepalSouth Tibet. Journal of the Geological Society 160,
345366.
Searle, M.P., 2010. Low-angle normal faults in the compressional Himalayan
orogen; Evidence from the AnnapurnaDhaulagiri Himalaya, Nepal. Geosphere 6,
296315.
Searle, M.P., Parrish, R.R., Hodges, K.V., Hurford, A., Ayres, M.W., Whitehouse,
M.J., 1997. Shisha Pangma leucogranite, South Tibetan Himalaya; field
relations, geochemistry, age, origin, and emplacement. Journal of Geology 105,
295317.
Seeber, L., Armbruster, J.G., Quittmeyer, R.C., 1981. Seismicity and continental
subduction in the Himalayan arc. In: Gupta, H.K., Delaney, F.M. (Eds.), ZagrosHindu Kush Himalaya Geodynamic Evolution. American Geophysical Union,
Washington, D.C.
Shackelton, R.M., 1981. Structure of Southern Tibet: report on a traverse from
Lhasa to Kathmandu organised by Academia Sinica. Journal of Structural
Geology 3, 97105.
Shen, F., Royden, L.H., Burchfiel, B.C., 2001. Large-scale crustal deformation of the
Tibetan Plateau. Journal of Geophysical Research, B, Solid Earth and Planets
106, 67936816.
Sol, S., Meltzer, A., Burgmann, R., et al., 2007. Geodynamics of the southeastern
Tibetan Plateau from seismic anisotropy and geodesy. Geology 35, 563566.
Stuwe, K., Robl, J., Hergarten, S., Evans, L., 2008. Modeling the influence of
horizontal advection, deformation, and late uplift on the drainage development in
the India-Asia collision zone. Tectonics 27. http://dx.doi.org/10.1029/
2007TC002186.
Tapponnier, P., Mercier, J.L., Proust, F., et al., 1981. The Tibetan side of the IndiaEurasia collision. Nature 294, 405410.
Tapponnier, P., Molnar, P., 1976. Slip-line field theory and large-scale continental
tectonics. Nature 264, 319324.
Tapponnier, P., Peltzer, G., Le Dain, A.Y., Armijo, R., Cobbold, P., 1982. Propagating
extrusion tectonics in Asia: new insights from simple experiments with
plasticine. Geology 10, 611616.
Tapponnier, P., Zhiqin, X., Roger, F., Meyer, B., Arnaud, N., Witlinger, G., Jingsui,
Y., 2001. Oblique stepwise rise and growth of the Tibet Plateau. Science 294,
16711677.
Taylor, M., Yin, A., 2009. Active structures in the Himalayan-Tibetan orogen and
their relationships to earthquake distributio, contemporary strain field, and
Cenozoic volcanism. Geosphere 5, 199214.
Thatcher, W., 2007. Microplate model for the present-day deformation of Tibet.
Journal of Geophysical Research-Earth Surface 112.
Thatcher, W., 2009. How the continents deform: the evidence from tectonic geodesy.
Tseng, T.L., Chen, W.P., Nowack, R.L., 2009. Northward thinning of Tibetan crust
revealed by virtual seismic profiles. Geophysical Research Letters 36.
Unsworth, M.J., Jones, A.G., Wei, W., Marquis, G., Gokarn, S.G., Spratt, J.E., Team,
I.-M., 2005. Crustal rheology of the Himalaya and Southern Tibet inferred from
magnetotelluric data. Nature 438, 7881.
Walker, J., Martin, M.W., Bowring, S.A., Searle, M.P., Waters, D.J., Hodges, K.V.,
1999. Metamorphism, melting, and extension: age constraints from the High
Himalayan Slab of southeast Zanskar and northwest Lahaul. Journal of Geology
107, 473495.
Wang, Y., Zhang, X., Sun, L., Wan, J., 2007. Cooling history and tectonic
exhumation stages of the south-central Tibetan Plateau (China): constrained by
40
Ar/39Ar and apatite fission track thermochronology. Journal of Asian Earth
Wei, S.Q., Chen, Y.J., Sandvol, E., et al., 2010. Regional earthquakes in northern
Tibetan Plateau: implications for lithospheric strength in Tibet. Geophysical
Research Letters 37. http://dx.doi.org/10.1029/2010GL044800.
Westaway, R., 1995. Crustal volume balance during the India-Eurasia collision and
altitude of the Tibetan plateau: a working hypothesis. Journal of Geophysical
Research 100, 1517315192.
Westaway, R., 2009. Active crustal deformation beyond the SE margin of the Tibetan
Plateau: constraints from the evolution of fluvial systems. Global and Planetary
Change 68, 395417.
Wittlinger, G., Farra, V., Hetenyi, G., Vergne, J., Nabelek, J., 2009. Seismic
velocities in Southern Tibet lower crust: a receiver function approach for eclogite
detection. Geophysical Journal International 177, 10371049.
Wobus, C., Heimsath, A., Whipple, K., Hodges, K., 2005. Active out-of-sequence
thrust faulting in the central Nepalese Himalaya. Nature 434, 10081010.
Wobus, C.W., Hodges, K.V., Whipple, K.X., 2003. Has focused denudation sustained
active thrusting at the Himalayan topographic front? Geology 31, 861864.
Wobus, C.W., Whipple, K.X., Hodges, K.V., 2006. Neotectonics of the central
Nepalese Himalaya: constraints from geomorphology, detrital 40Ar/39Ar
thermochronology, and thermal modeling. Tectonics 25. http://dx.doi.org/
10.1029/2005TC001935.
Wu, C., Nelson, K.D., Wortman, G., et al., 1998. Yadong cross structure and South
Tibetan detachment in the east central Himalaya (891901E). Tectonics 17,
2845.
Xu, L.L., Rondenay, S., van der Hilst, R.D., 2007. Structure of the crust beneath the
southeastern Tibetan Plateau from teleseismic receiver functions. Physics of the
Earth and Planetary Interiors 165, 176193.
Xypolias, P., Kokkalas, S., 2006. Heterogeneous ductile deformation along a midcrustal extruding shear zone: an example from the External Hellenides (Greece).
Yeats, R.S., Lillie, R.J., 1991. Contemporary tectonics of the Himalayan Frontal
Fault System: folds, blind thrusts and the 1905 Kangra earthquake. Journal of
Structural Geology 13, 215225.
Yin, A., 2006. Cenozoic tectonic evolution of the Himalayan orogen as constrained
by along-strike variation of structural geometry, exhumation history, and foreland
sedimentation. Earth Science Reviews 76, 1131.
Yin, A., 2010. Cenozoic tectonic evolution of Asia: a preliminary synthesis.
Yin, A., Dubey, C.S., Kelty, T.K., Webb, A.A.G., Harrison, T.M., Chou, C.Y., Celerier,
J., 2010. Geologic correlation of the Himalayan orogen and Indian craton: Part
2. structural geology, geochronology, and tectonic evolution of the Eastern
Himalaya. Geological Society of America Bulletin 122, 360395.
Yin, A., Harrison, T.M., 2000. Geologic evolution of the Himalayan-Tibetan orogen.
Annual Reviews of Earth and Planetary Sciences 28, 211280.
Yin, A., Harrison, T.M., Murphy, M.A., et al., 1999. Tertiary deformational history of
southeastern and southwestern Tibet during the Indo-Asian collision. Geological
Zhang, P., Shen, Z., Wang, M., et al., 2004. Continuous deformation of the Tibetan
Plateau from global positioning system data. Geology 32, 809812.
Zhang, Z.J., Klemperer, S., 2010. Crustal structure of the Tethyan Himalaya,
southern Tibet: new constraints from old wide-angle seismic data. Geophysical
Journal International 181, 12471260.
369
Zheng, D.W., Clark, M.K., Zhang, P.Z., Zheng, W.J., Farley, K.A., 2010. Erosion,
fault initiation and topographic growth of the North Qilian Shan (northern
Tibetan Plateau). Geosphere 6, 937941.
Biographical Sketch
Soon after earning his PhD from the Massachusetts Institute of Technology in 1982, Kip Hodges participated in a
field trip in northwest India led by renowned Himalayan geologists KS Valdiya and Vikram Thakur and became
forever hooked on the Cenozoic tectonics of the Himalaya and Tibet. In addition to NW India, he has conducted
extensive field research in Bhutan, Nepal, and Tibet; 2011 marked the Silver Anniversary of his first field foray into
Tibet. After spending many years on the faculty at MIT, Hodges moved to Arizona State University in 2006 to
become Founding Director of its School of Earth and Space Exploration. His research interests include the
integration of a variety of field and laboratory techniques to the study of orogenic systems, the development and
application of new analytical techniques for noble gas geochemistry and geochronology, lunar and meteorite
geochronology, the design of new modalities for formal and informal science education, and the development of
new strategies for mixed human and robotic scientific exploration of the Solar System.
Byron Adams is a PhD candidate at Arizona State University in tectonic geomorphology. His primary interests lie
within the broad spectrum of surface processes and tectonics, as well as their interactions with climate. His specific
pursuits include developing and utilizing quantitative techniques (such as thermochronology and terrestrial
cosmogenic nuclide dating) to better understand geologic processes specific to geomorphology and tectonics.
Most of this research has been focused on understanding the development of the HimalayanTibetan orogenic
system.
5.16 Polygenetic Landscapes

U Kamp, University of Montana, Missoula, MT, USA
5.16.1
5.16.2
5.16.2.1
5.16.2.2
5.16.2.3
5.16.2.4
5.16.2.5
5.16.3
5.16.3.1
5.16.3.2
5.16.3.3
5.16.3.4
5.16.3.5
5.16.3.6
5.16.3.7
5.16.4
5.16.4.1
5.16.4.2
5.16.5
5.16.5.1
5.16.5.2
5.16.5.3
5.16.6
5.16.6.1
5.16.6.2
5.16.6.3
5.16.7
5.16.8
References
Introduction
Early Conceptual Models for Landscape Evolution
Davis Model: Cycle of Erosion
Penck Model: Uplift and Denudation Related
King Model: Pediplanation
Budel Model: Etchplanation
Critique of Early Landscape Evolution Models
System and Equilibrium Models
Geomorphic Systems
Tectonic Geomorphology
Dynamic Equilibrium and Dynamic Meta-Stable Equilibrium
Time-Dependent and Time-Independent Landforms
Landscape Sensitivity
High-MagnitudeLow-Frequency Processes versus Low-MagnitudeHigh-Frequency Processes
Critique of Equilibrium Concepts
Models for Feedback between Climate and Tectonics
Climatic Instability and Paraglaciation
Glacial Buzzsaw
Relief Production
Unloading and Uplift
Tectonic Aneurysm
Glacial Protection
Landscape Evolution and Scale
Continental Scale Landscapes
Regional-Scale Landscapes
Landform Scale
Mathematical and Computational Modeling
Conclusion
Abbreviations
ELA
GTFs
Mw
Equilibrium-line altitude
Glossary
Alluvial fans Fan-shaped landform that forms by
the deposition from a fast flowing stream that slows,
and spreads typically at the exit of a canyon into a
flatter plain.
Cascading system Geomorphic system in which
pathways of transport of energy or matter or both are
interconnected, together with such storages of energy
372
373
373
373
374
375
375
375
375
377
377
377
377
378
379
379
379
381
381
381
382
382
382
382
382
384
384
390
391
Geomorphic transport functions

Moment magnitude
and matter as may be required. Also process system or

flow system.
Characteristic form time Duration (time) over which a
particular landform will persist.
Complex system System composed of interconnected
parts that as a whole exhibit one or more properties
(behavior among the possible properties) not obvious from
the properties of the individual parts.
Kamp, U., Owen, L.A., 2013. Polygenetic landscapes. In: Shroder, J. (Editor
in Chief), Owen, L.A. (Ed.), Treatise on Geomorphology. Academic Press,
San Diego, CA, vol. 5, Tectonic Geomorphology, pp. 370393.
370
http://dx.doi.org/10.1016/B978-0-12-374739-6.00097-X
Computational modeling Mathematical model

constructed on a computer to study the behavior of simple
and/or complex systems.
Cycle of erosion Model for landscape development
proposed by William Morris Davis.
Dynamic equilibrium (steady-state equilibrium);
Dynamic meta-stable equilibrium Situation that is
fluctuating about some apparent average state, where the
average state itself is also changing through time.
Endogenetic process Internal processes that are driven by
geologic forces, essentially tectonic and isostatic
mechanisms.
Endrumpf landscape Highly denuded landscape with
little relief.
Epeirogenic uplift Vertical displacement of the crust that
occurs over large regions.
Equilibrium-line altitude (ELA) Line that separates the
ablation zone and the accumulation zone on a glacier. At
this altitude, the amount of new snow gained by
accumulation is equal to the amount of snow/ice lost
through ablation.
Etchplain Tropical planation surface where deep
weathering has etched a surface into bedrock and the
weathered product has been stripped away to reveal an
extensive bedrock surface.
Etchplanation Deep tropical weathering and exhumation
of bedrock to produce an etchplain.
Exogenetic process External (geomorphic) processes that
are essentially driven by climatic and gravitational forces
that include denudation and deposition.
Fan terraces Incised alluvial fans remnants.
Flow system Geomorphic system in which pathways of
transport of energy or matter or both are interconnected,
together with such storages of energy and matter as may be
required. Also process system or cascading system.
Form system Geomorphic system in which form variables
are deemed to interrelate in a meaningful way in terms of
system origin or system function. Also morphological
system.
Geomorphic system Sets of interconnected parts that
function together as complex wholes. The parts are
landforms, Earth-surface processes, and environmental
factors which influence or control processes and forms.
Geomorphic transport functions (GTFs) Mathematical
representation in space and time of the geomorphic
processes.
Geomorphometry Science of quantitative land surface
analysis.
Glacial buzzsaw Glacial erosional processes that limit
topographic elevations in extensively glaciated orogens.
Glacial protection Protection of the landscape from
erosion by covering of glacier ice.
Glacierization Covering of a land area by glacier ice.
Glacioisostatic rebound Lithospheric uplift due to
unloading resulting from melting of glacier ice.
Holocene Geologic epoch that began at 11 700 years ago
and continues to the present.
371
Inselberg(s) Isolated rock hill or ridge that rises

abruptly from a gently sloping or virtually level
surrounding plain.
Isostatic Refers to the state of gravitational equilibrium
between the lithosphere and asthenosphere such that
tectonic plates float at an elevation which depends on their
thickness and density.
Lag time Time taken to begin reacting to change.
Landscape sensitivity The ability of a landscape to resist
disturbance from natural and human agencies.
Littoral karren Small-scale dissolutional and bioerosional
forms in limestone ranging in size from submillimetre to
some tens of meters that develop in the coastal settings.
Mw moment magnitude Measurement of the energy
released during an earthquake and is proportional to the
slip on a faults times the area of the fault surface that has
slipped. The scale is logarithmic, such that for every increase
of 1 Mw the total energy increases by about 32 times.
Morphological system Also form system; sets of form
variables that are deemed to interrelate in a meaningful way
in terms of system origin or system function.
Paraglaciation Unstable conditions caused by a
significant relaxation time in processes and
geomorphic patterns following glacial climate.
Pediplain Extensive slightly inclined denudation plain.
Pediplanation Processes by which pediplanes are formed.
Piedmonttreppen Series of benches.
Polygenetic Landforms and landscapes that are the
product of numerous different processes.
Primary surface (Primarrumpf) Oldest peneplain on
Earth surface.
Process-form system Energy-flow system in which the
processes may alter form and, in turn, changed form alters
the processes. Also process-form system.
Process system Geomorphic system in which pathways of
transport of energy or matter or both are interconnected,
together with such storages of energy and matter as may be
required. Also flow system or cascading system.
Relaxation time Time taken to achieve a
characteristic form.
Simple system System that contains few variables.
Steady-state System where the properties do not change
with time.
System Scheme to help understand the relationships
between interacting or interdependent components that
comprise an integrated whole, which in geomorphology is
used to help study geomorphic processes and their
products.
Tectonic aneurysm Ability of glaciers and rivers to
accomplish localized rapid rock uplift by rapid erosion and
crustal weakening.
Yardang Streamlined hill carved from bedrock or
consolidated or semiconsolidated material by wind erosion
and deflation.
Younger Dryas Stade Brief cold climate period between
approximately 12.8 and 11.7 ka.
372
Abstract
Earths surface is continuously evolving and is the result of the complex interactions between Earths internal (endogenetic)
and external (exogenetic) processes. Endogenetic and exogenetic interactions result in landforms and landscapes that are
the product of numerous different processes, and thus are polygenetic in origin. Recent tectonicgeomorphic studies have
led to the realization that complex links, interactions, and feedback exist among tectonics, climate, and landscape
development. Key issues in the evolution of polygenetic landscapes included examining the role of tectonics and isostatic
uplift, the nature of exhumation rates, climate control and erosion distribution, glaciers and relief production, focused
erosion and spatial coincidence, geomorphometry and erosion modeling, and equilibrium concepts such as topographic
steady state. Notably, the new models based on these complex relationships include: the glacial buzzsaw model whereby
topography is limited by glaciation; the tectonic aneurysm model in which localized uplift at syntaxes is caused by
enhanced fluvial and glacial erosion that, in turn, weaken the lithosphere and enhance bedrock uplift; and glacial damming
leading to differential erosion and uplift. In addition to natural factors, human influences are now contributing to
polygenetic landscapes and pose greater challenges for predicting the nature of future geomorphic change.
5.16.1
Introduction
Earths surface is continuously evolving and is the result of the

complex interactions between Earths internal (endogenetic)
and external (exogenetic) processes. Endogenetic processes are
internal processes that are driven by geologic forces, essentially
tectonic and isostatic mechanisms; these lead mainly to relief
construction. Endogenetic processes control long-term landform and landscape evolution in two ways: (1) directly
through uplift, subsidence, and plate movements and (2) indirectly by influencing local, regional, and global climates
(Summerfield, 1991). Exogenetic processes are external (geomorphic) processes that are essentially driven by climatic and
gravitational forces that include denudation and deposition;
they lead mainly to relief reduction. The main Earth surface
processes that are responsible for most topographic features
operate within the following systems: weathering, karst, fluvial, aeolian, glacial, periglacial (freezethaw), coastal, mass
wasting, and extraterrestrial impact processes.
Endogenetic and exogenetic interactions clearly result in
landforms and landscapes that are the product of numerous
different processes, each of which operates on its own spatial
and temporal scales, and within oscillating climates. Throughout the past century it has become increasingly apparent that
landscapes and landforms are a function of numerous processes, that is, they are polygenetic in origin. They are a function
of a host of processes, and they need to be examined using a
multitude of methods ranging from remote sensing, field and
laboratory analysis, geochronology, geographic information
technology, and computational modeling.
Despite the advances in technology such as supercomputing we are still only beginning to glimpse the various
complicated interactions between processes in Earths atmosphere, lithosphere, hydrosphere, and biosphere. Moreover, we
are in a time of rapid environmental change induced by
human activities that is impacting essentially every environment on our planet. As such, human processes are increasingly
controlling the development of landforms and landscapes. By
their very nature, polygenetic landscapes represent geodynamic systems of highly scale-dependent interactions among
climate, surface, and tectonic processes (Koons et al., 2002;
Bishop et al., 2003, 2010). Clearly, complex feedback mechanisms link the different processes.
The operational scale dependency of numerous processes
makes it difficult to analyze these interactions and feedbacks
(Bishop et al., 2002, 2003). However, Tucker and Hancock

(2010) showed how landscape evolution models serve science
and society: (1) they represent the latest ideas about how
various surface processes interact to shape Earths surface
and transport crustal mass; (2) they allow for the visualization and quantification of consequences of various hypotheses about process dynamics; (3) they add in developing
insight into the geohistoric development of particular places;
and (4) they can be used in environmental monitoring and
management such as for the rehabilitation of degraded
landscapes.
The term polygenetic first came into use in the late twentieth and early twenty-first century when geomorphologists
began to link climatic and tectonic geomorphology to formulate new theories on how landscapes develop. Many of the
key issues in the evolution of polygenetic landscapes included
examining the role of tectonics and isostatic uplift (Gilchrist
et al., 1994), the nature of exhumation rates (Whittington,
1996), climate control and erosion distribution and timing
(Clift et al., 2008a, b; Rahaman et al., 2009), glaciers and relief
production (Harbor and Warburton, 1992; Hallet et al., 1996;
Whipple and Tucker, 1999; Whipple et al., 1999), erosion and
spatial coincidence (Zeitler et al., 2001a, b; Finlayson et al.,
2002; Finnegan et al., 2008), geomorphometry and erosion
modeling (Montgomery and Brandon, 2002; Bishop and
Shroder, 2004), and equilibrium concepts such as topographic
steady state (Tomkin and Braun, 2002).
This chapter reviews the various theories on how the study
of polygenetic landscapes developed, starting by reviewing
the early conceptual models of landscape evolution, which
although they implicitly presented the view that landscapes
are polygenetic, did not explicitly call the landscape polygenetic. The early models helped formulate many of the
theories that we are currently testing with modern methods.
The early research, of course, did not have the advantage
of knowledge of plate tectonics and modern Quaternary
paleoclimatology. This chapter continues by outlining the
newer models and theories, most of which are described
in considerable detail in earlier chapters of this treatise, and
links these to show how landscapes are polygenetic in
origin. Finally, some views on how to advance and use our
knowledge beneficially in a world significantly influenced by
human activities are presented. This chapter tries to provide a
holistic view of tectonic geomorphology and how it relates to
many other branches of geomorphology.
5.16.2
5.16.2.1
Early Conceptual Models for Landscape

Evolution
Davis Model: Cycle of Erosion
Probably the earliest conceptual model for landscape evolution was proposed by Davis (1889, 1899). His Cycle of
Erosion incorporated within his Geographical Cycle was a
theoretical, deductive construct that described a progressive
sequence of stages in river development under a humid temperate climate initiated by rapid uplift. Davis (1889, 1899)
saw landscapes as being a function of the trio of geographic
controls: regional geological structure (the initial condition),
geomorphic processes (sum of weathering and transport), and
stage of landform evolution (representing time in the sense of
landscape development relative to the completion of the entire cycle). The main framework was time (stage), thus, the
model was time-dependent. Assuming tectonic stability, the
river system gains in potential energy, incises into the landscape, and then erodes it while passing through youthful,
373
mature, to old age landforms/landscapes (Figures 1 and

2(a)). During this succession, surface gradients decline, and
the cycle progresses from a high-relief to a low-relief landscape
close to river base-level, the peneplain. The next uplift event
initiates a new cycle passing through all three stages of the
succession with time. The stage of evolution at any given time
can be expressed in terms of its mean elevation above a base
level. Whenever uplift occurs before the landscape cycled
through the entire succession, the landscape was said to be
rejuvenated, and youthful landforms coexist with older landforms and thereby create a polycyclic landscape. Davis (1889,
1899) concept of peneplanation dominated Anglo-American
geomorphology for several decades, but never gained wide
acceptance by geomorphologists in mainland Europe.
5.16.2.2
Penck Model: Uplift and Denudation Related
The Davisian Cycle of Erosion maintained its popularity

through the late nineteenth and early twentieth centuries until
(a)
(d)
(b)
(e)
(c)
(f)
(g)
Figure 1 The Davisian Cycle of Erosion under a humid climate. (a) Landscape with little local relief, a peneplain developed during the previous
cycle, is starting point of new cycle. (b) Uplift and river incision form an early youth landscape with largely uneroded uplands and irregular river
gradients. (c) In late youth uplands are reduced in area, rivers are graded, and narrow floodplains develop. (d) The mature landscape has
maximum local relief and uplands are eliminated altogether. Floodplains become wider and major rivers develop meandering channels. (e) By late
maturity local relief has been significantly reduced and floodplains are extensive. (f) In old age the entire landscape is graded and floodplains are
very broad. Mean land surface elevation is already close to base level. Monadnocks, erosional residuals in resistant lithologies, may survive. (g)
Uplift initiates new cycle. Reproduced with permission from Summerfield, M.A., 1991. Global Geomorphology: An Introduction to the Study of
Landforms. Wiley, New York, and Strahler, A.N., 1969. Physical Geography. Wiley, New York.
374
Youth
Maturity
Altitude
Uplift
Old age
Altitude of main
valley floors
Altitude of highest
divides
Peneplain
Base
level
Maximum
relief
Time
(a)
Waning
development
Altitude of
highest divides
Altitude of main
valley floors
Altitude
Waxing
development
Endrumpf
Base
level
Time
(b)
Altitude
Uplift
Youth
Old age?
Maturity
Altitude of main
valley floors
Altitude of
highest divides
Pediplain
Base
level
(c)
Pediplain
Time
Figure 2 Early models for landscape development. (a) Davis (1889, 1899) Cycle of Erosion. (b) Pencks (1924) uplift and denudation
related model, and (c) Kings (1957) pediplanation model. Base level is assumed to be fixed through time and temporal scale varies between
diagrams. Reproduced from Davis, W.M., 1889. The rivers and valleys of Pennsylvania. National Geographic Magazine 1, 183253; Davis, W.M.,
1899. The geographical cycle. The Geographical Journal 14, 481504; Penck, W., 1924. Die Morphologische Analyse: Ein Kapitel der Physikalischen
Geologie. Engelhorn, Stuttgart (In German); King, L.C., 1957. The uniformitarian nature of hillslopes. Transactions of the Edinburgh Geological
Society 17, 81102; Summerfield, M.A., 1991. Global Geomorphology: An Introduction to the Study of Landforms. Wiley, New York, with
permission from Wiley, and Thornes, J.B., Brunsden, B., 1977. Geomorphology and Time. Methuen, London, with permission from Methuen.
Penck (1924) proposed an alternative model for landscape

evolution. Pencks deductive model gained much support in
mainland Europe. Penck (1924) suggested that the Davisian
Cycle of Erosion was flawed because it assumed that uplift is
followed by erosion of a stable crust, but in fact active uplift
although episodic could continue for a considerable
time, and that various possible successions occur in the
landscape controlled by different rates of uplift and erosion
(Figure 2(b)). Pencks model was largely time-independent, it
was based on tectonic history, and it emphasized geologic
structure and rates of endogenetic and exogenetic processes.
Moreover, Penck (1924) proposed that uplift rates initially
increased slowly before reaching a maximum and then declined gradually. Whenever downcutting of rivers increases, for
example, a waxing development stage occurs and slopes
steepen progressively from the base upward to produce a
convex profile. In contrast, a waning development stage occurs
whenever downcutting decreases, and concave profiles develop. Pencks model described a succession of uplift of a
primary surface (Primarrumpf), the formation of a series of
benches (Piedmonttreppen), valley widening through slope
replacement, the formation of large residual hills (Inselbergs)

and pediments, and eventually the elimination of inselbergs
and formation of an Endrumpf landscape. Although Pencks
concept of uplift and denudation helped to integrate endogenetic processes into long-term landscape evolution, his
model failed to acknowledge the importance of many other
factors affecting landscape development such as changes in
stream discharge as a result of climate change.
5.16.2.3
King Model: Pediplanation
Debate over the relative merits of the Davisian Cycle of Erosion and Pencks model for landscape evolution continued
throughout the first half of the twentieth century, but was
reexamined by King (1957) who proposed a model similar
to Pencks but with a significant different explanation of slope
development. King (1957) proposed that instead of a slope
replacement, which takes place in balance with tectonic activity as described in Pencks concept, whole-slope parallel
retreat was a characteristic of steep semi-arid landscapes.
King (1957) divided the slope profile into four segments
(waxing slope, free face, debris slope, and waning slope),

of which one or more might be entirely absent. The resultant
succession of landscape evolution would result in an
expansion and coalescence of pediments, a process called
pediplanation that would produce pediplains, as the free
face retreats eliminating the higher surfaces (older pediments)
to create isolated erosional remnants (Figure 2(c)). King,
being trained in the Davisian school, emphasized the
importance of time more in his model than Penck (1924) did.
5.16.2.4
Critique of Early Landscape Evolution Models
In the following decades, the early descriptive concepts were

criticized, in the case of the Davisian model specifically for its
assumptions of rates and occurrence of uplift and climatic
stability. It soon became clear that stable landscapes are rather
rare and that landscapes that follow a simple unidirectional
sequence of forms probably do not exist. Instead, polycyclic
landscapes are likely to be the norm. However, although the
early models are essentially qualitative, they contain important concepts, which are also relevant in the context of quantitative models, particularly the ideas of slope decline, slope
retreat, and equilibrium.
5.16.3
5.16.3.1
Laterized etchplain
(a)
Dissected etchplain
Budel Model: Etchplanation
An expansion of landscape models came with Budels

(1957) morphogenetic concept of etchplanation. This gained
considerable influence in mainland Europe, especially in
Germany, and it led to the development of the subdiscipline
of climatic geomorphology. The concept asserts that, under a
particular climatic regime, certain geomorphic processes will
dominate to produce a characteristic topographic expression
(Figures 3 and 4). Budel (1957) argued that climatic controls
outweighed rock type as a landform factor because the resistance of a rock type to erosion is dependent on the climate
to which it is subjected. One of Budels key notions was that of
a double surface of leveling. Budel (1957) proposed that in
ancient, stable cratonic areas of the humid tropics, a subsurface landscape is essentially etched from the rock by deep
weathering and subsequent removal of the weathered regolith
an etchplain developed. This process often results in the
exposure of structurally defined compartments of resistant
rock. Since in the humid tropics the controlling factors vary
spatially, the weathering front is highly irregular in form and
depth. Although during periods of tectonic and climatic stability, rates of weathering and denudation are roughly in a
steady state, climatic or tectonic variations might lead to imbalance in the geomorphic system, and, for example, the
stripping of the entire weathering zone. In the now lowered
landscape, more water will increase weathering rates again.
5.16.2.5
375
(b)
Partially stripped etchplain
(c)
Dominantly stripped etchplain/etchsurface
(d)
Incised etchsurface
(e)
Bedrock
Indurated
horizons
(especially
ferricated)
Weathering
mantle
Figure 3 Budels (1957) etchplanation model for landscape

development. (a) Laterized etchplain with low local relief and limited
stream incision, underlain by a thick weathering mantle, including
indurated horizons, especially ferricretes. (b) Climate change and uplift
accelerated river incision and developed dissected etchplain with
well-defined valleys and only localized formation of tors. (c) Partially
stripped etchplain with widespread stream dissection and extensive
stripping of weathering mantle; formation of numerous tors. (d)
Dominantly stripped etchplain with only localized remnants of
weathering mantle in deep pockets. (e) Fluvial erosion as result of base
level change extensively modified basal bedrock and developed incised
etchsurface. Reproduced with permission from Budel, J., 1957. Die
Doppelten Einebnungsflachen in den feuchten Tropen. Zeitschrift fur
Geomorphologie N.F. 1, 201228 (In German); Summerfield, M.A.,
1991. Global Geomorphology: An Introduction to the Study of
Landforms. Wiley, New York, and Thomas M.F., 1974. Tropical
Geomorphology. Macmillan, London.
System and Equilibrium Models

Geomorphic Systems
In the 1950s and 1960s, the discipline of geomorphology

underwent a quantitative revolution. Technological enhancements were followed by new methodological approaches,
such as systems modeling. The concept of the system approach

began to dominate. Based on an approach first developed for
cybernetics, a system is a set of objects that are linked together
with functional relationships between the objects and
their attributes (Hall and Fagen, 1956). In geomorphology,
376
Equator
Gletscherzone
Exzessive
Talbildungszone
Ektropische
Talbildungszone
Subtropische Pedimentund Talbildungszone
Tropische
Flachenbildungszone
Figure 4 Budels climatemorphological zones of the world. Explanation of key (from top to bottom): Zone of Glaciers; Zone of Pronounced
Valley Formation; Extra-tropical Zone of Valley Formation; Subtropical Zone of Pediment and Valley Formation; Tropical Zone of Planation Surface
Formation. Reproduced from Budel, J., 1963. Klimatische Geomorphologie. Geographische Rundschau 15, 269285 (In German).
landscape elements at a particular scale represent the objects,

geomorphic processes are the relationships, and physical
properties such as slope gradient or drainage pattern describe
the attributes (Figure 5). In general, it is assumed that
the more objects and relationships are known and understood
the better the system model. As for most research, it also
means that system modeling is to some degree subjective
since the researchers select the included components and the
employed technology and methodology at their discretion.
Thus, system models are only useful for specific purposes.
In any case, system modeling requires the evaluation of
the model.
Huggett (2011) described the main types of geomorphic
systems according to complexity. A morphological system or
form system describes sets of form variables that are deemed
to interrelate in a meaningful way in terms of system origin
or system function (Figure 6(a)). In a process system or
cascading system or flow system, pathways of transport of
energy or matter or both are interconnected, together with
such storages of energy and matter as may be required
(Figure 6(b)). A process-form system or process-response
system represents an energy-flow system in which the processes may alter form and, in turn, the changed form alters the
processes (Figure 6(c)). A simple system could, for example,
describe a few bolders resting on a talus slope. In a complex
but disorganized system, the vast number of objects interact
in a weak and haphazard way, for example, the countless
Channel
Valley-side slope
Wind erosion and deposition
Debris
transport
nt
Debris
transport
rin
ro
gf
the
Interfluve
Waste mantle
Debris
production
Debris
production
a
We
Uplift or
subsidence
Figure 5 A hillslope as a geomorphic system, showing storages
(waste mantle), inputs (e.g., wind deposition and debris production),
outputs (e.g., wind erosion), throughputs (debris transport), and units
(channel, valley-side slope, interfluve). Reproduced from Huggett, R.J.,
2011. Fundamentals of Geomorphology. Routledge, London, with
permission from Taylor and Francis.
individual particles in a hillslope mantle. In a complex and

organized system, the objects interact strongly with one another; for example, a hillslope represented as a process-form
system. Quantitative and process geomorphology increasingly
focused on specific geomorphic processes, smaller scale
Form system
Flow or cascading
system
Rockfall
Cliff
(a)
1 23
Cliff
Covered
cliff face
at time 2
Talus
Talus
Lower
slope
Process-from or
process-response
system
377
Lower
slope
Time
3
2
1
(b)
(c)
Figure 6 A cliff and a talus slope can be interpreted as different types of geomorphic systems. (a) Form system, (b) flow or cascading system,
(c) process-form or process-response system. Reproduced from Huggett, R.J., 2011. Fundamentals of Geomorphology. Routledge, London, with
permission from Taylor and Francis.
(individual landforms), and shorter term (102 to 102 years)

geomorphic change rather than large-scale landscapes and
long-term (102 to 106 years) landscape evolution.
equilibrium to a new level. Thus, landscape change occurs

episodically rather than progressively gradual.
5.16.3.4
5.16.3.2
Tectonic Geomorphology
During the late 1960s and into the early 1970s, the revolutionizing theory of plate tectonics shed new light on how
landforms may develop by endogenetic processes, and this
ultimately led to the new discipline of tectonic geomorphology. Summerfield (1991) pointed out that one of the main
new ideas was that landscapes do not evolve in any systematic
manner, but oscillate around an equilibrium form. In the
1980s and 1990s, with the advent of new theoretical frameworks and new analytical, modeling, and dating techniques,
there has since been a return to the study of long-term landscape evolution. For example, whereas in the beginning numerical landscape evolution modeling was relatively simple,
today (super-)computers allow for more complex approaches.
5.16.3.3
Dynamic Equilibrium and Dynamic Meta-Stable

Equilibrium
Building on the early work of Gilbert (1877, 1909, 1914),

Hack (1960, 1975) suggested a time-independent system
model that described landscapes as being generally in a balance between denudation and either tectonic uplift or fluvial
downcutting. As a result, the landscape was essentially in a
state of dynamic equilibrium, essentially steady-state equilibrium, between force and resistance (Figure 7(a)). The
maintenance of this dynamic equilibrium assumes a rapid
adjustment to environmental changes. However, today we
know that this is not necessarily the case since some very old
relict landforms do survive (Schafer et al., 1999; Dunai et al.,
2005).
Schumm (1975) refined this idea and proposed a dynamic
meta-stable equilibrium model (Figure 7(b)). Phases of
steady-state equilibrium are characterized by constant smaller
adjustments, but the adjustment can reach an extrinsic
or geomorphic threshold, which shifts the steady-state
Time-Dependent and Time-Independent

Landforms
Much debate exists over whether particular landscapes are

continuously evolving or in a steady state. Landscapes can be
seen as evolving and time-dependent or are in a steady state
and are time-independent systems. The former is viewed in
terms of its stage of development; the latter is viewed in terms
of interactions between materials and processes. Timeindependent landscapes should have high relief and high
denudation rates and strong interactions among environmental components, which are given, for example, in active
orogenic belts. The highest parts of mountains are generally
in a true steady state, which involves crustal uplift and denudation that are in equilibrium, and elevation oscillates
around a constant value. As long as the environmental conditions remain unchanged, the time-independent landscape
will retain a constant form. Summerfield (1991) described
the Southern Alps and the mountains of Taiwan as examples
for extensive steady-state landscapes. A steady-state orogen
can become out of equilibrium by either changed crustal
uplift or changed denudation rates. In contrast, Summerfield
(1991) argued that intra-plate landscapes are very unlikely to
be in steady state, even though the very low rates of landscape lowering may give the impression that the landscape is
at elevation stability.
5.16.3.5
Landscape Sensitivity
As long as controlling conditions in a landscape remain constant over time and space, a strong interdependence between
process and form leads to the generation of characteristic
forms that make up specific landscape types (Summerfield,
1991). However, landscapes vary in both magnitude and frequency of external changes and in their ability to adjust to
such changes. Hence, controlling conditions rarely remain
constant sufficiently long for characteristic forms to develop.
Instead, transient forms develop. Brunsden (1980) introduced
378
Dynamic equilibrium
(a)
Dynamic metastable equilibrium
(b)
Figure 7 Equilibrium models for landscape development. (a) Hacks

(1960, 1975) dynamic equilibrium model is a time-independent
system that describes landscapes as being generally in a balance
between denudation and either tectonic uplift or fluvial downcutting.
The landscape change occurs progressively gradual. (b) In Schumms
(1975) dynamic meta-stable equilibrium model phases of steady-state
equilibrium are characterized by constant smaller adjustments, but
the adjustment can reach a threshold that shifts the steady-state
equilibrium to a new level. Thus, landscape change occurs
episodically rather than progressively gradual. Reproduced from
Hack, J.T., 1960. Interpretation of erosional topography in humid
temperate regions. American Journal of Science 258, 8097, with
permission from AAAS; Hack, J.T., 1975. Dynamic equilibrium and
landscape evolution. In: Melhorn, W.N., Flemal, R.C. (Eds.), Theories
of Landform Development. Allen and Unwin, London, pp. 87102,
with permission from Allen and Unwin; Schumm, S.A., 1975.
Episodic erosion: a modification of the geomorphic cycle. In:
Melhorn, W.N., Flemal, R.C. (Eds.), Theories of Landform
Development. Allen and Unwin, London, pp. 6985, with permission
from Allen and Unwin, and Huggett, R.J., 2011. Fundamentals of
Geomorphology. Routledge, London.
the concept that geomorphic time includes three categories:

lag time (time taken to begin reacting to change), relaxation
time (time taken to achieve a characteristic form), and characteristic form time (time over which characteristic forms
persist) (Figure 8). In reality, characteristic and transient forms
coexist in the landscape.
How sensitive is a landscape to change is greatly debated.
Brunsden and Thornes (1979: 476) explained that: the sensitivity of a landscape to change is expressed as the likelihood
that a given change in the controls of a system will produce a
sensible, recognizable and persistent response. The issue involves two aspects: the propensity for change and the capacity
of the system to absorb the change. Whether a landscape
undergoes a change depends on its sensitivity and the magnitude, frequency, and duration of changes in the controlling
variables. Summerfield (1991) summarized the two main
types of landscape sensitivity: (1) a highly sensitive, fastresponding, mobile geomorphic system that adjusts rapidly to
change and (2) an insensitive, slow-responding system that is
highly resistant to change. Although in an insensitive state
landscape transient forms will predominate, characteristic
forms are much more likely to be present in a sensitive
Geomorphic change
Relaxation time
Lag
Relaxation
Impulse
of change
Characteristic
form time
Lag
Impulse
of change
Time
Figure 8 Brunsdens (1980) concept describes that geomorphic

time includes three categories: lag time (time taken to begin reacting
to change), relaxation time (time taken to achieve a characteristic
form), and characteristic form time (time over which characteristic
forms persist). Reproduced from Brunsden, D., 1980. Applicable
models of long term landform evolution. Zeitschrift fur
Geomorphologie N.F. Suppl. 36, 1626, and Summerfield, M.A.,
1991. Global Geomorphology: An Introduction to the Study of
Landforms. Wiley, New York, with permission from Wiley.
landscape that spends a longer time in equilibrium with prevailing environmental conditions.
5.16.3.6
High-MagnitudeLow-Frequency Processes
versus Low-MagnitudeHigh-Frequency
Processes
Much has been debated over the relative roles of high-magnitudelow-frequency and low-magnitudehigh-frequency events
in landscape evolution (Brunsden and Jones, 1984), in particular with respect to endogenetic versus exogenetic processes.
However, it is difficult to precisely define what constitutes a
high-magnitudelow-frequency event because it requires the
scale and time frame to be more defined, which of course varies
dependent on what is being considered. For example, geomorphic processes in high mountains are commonly an order
of magnitude greater in size and effect than most other terrestrial settings (Figure 9). Tectonic processes such as major
(moment magnitude Mw 77.9) or great (MwZ8) earthquakes
are commonly regarded as high magnitudelow frequency because they are often devastating to the local people and produce
widespread landsliding. But, of course, earthquake magnitude
and intensity vary greatly (over many orders of magnitude) and
low-magnitudelow-intensity events generally go unnoticed.
During recent years the dramatic effects of earthquakes on
landscape developments have been highlighted by the 2005
(Mw 7.6) Kashmir earthquake, which triggered several thousand
landslides throughout a region of 47500 km2 (Kamp et al.,
2008; Owen et al., 2008). Kamp et al. (2008) found that the
coseismic landsliding was 510 times higher than the background landslide frequency. Furthermore, Khattak et al. (2010)
showed that the postseismic landslide activity returned to the
100 000
10 000
1 000 000
Key:
V rock avalanche-generated debris flows
1000
V rainfall-generated debris flows
Peak discharge (m3 s1)
Total debris flow volume (m3)
10 000 000
379
V rock avalanche-generated debris flow (Vmax)

Q rainfall-generated debris flows
Rock avalanche-generated debris flow (Qmax)
Rock avalanche-generated debris flow (Qmin)
100 000
10
100
1000
10 000
100
100 000
Return period (years)

Figure 9 Relationship between frequency and magnitude (frequencyvolume and frequencypeak discharge) of debris flows on Cheekye fan,
British Columbia. Reproduced from Jakob, M., Friele, P., 2010. Frequency and magnitude of debris flows on Cheekye River, British Columbia.
background landslide frequency and, thus, the landscape

equilibrium within only 2 years. In addition, the 2008 (Mw 7.9)
Wenchuan earthquake triggered numerous large landslides,
including one that buried more than 700 people (Liu and
Kusky, 2008; US Geological Survey, 2008), and many thousands of smaller landslides, yet their full impact has still to be
assessed.
5.16.3.7
explain. Bracken and Wainwright (2006) concluded that geomorphic equilibrium is a myth and suggested using it as
a concept of simplification or a metaphor that underpins
many geomorphic concepts and ideas.
5.16.4
Models for Feedback between Climate and

Tectonics
Critique of Equilibrium Concepts
As a critique of the many equilibrium terms that had been

proposed in geomorphic texts, Ahnert (1994) stated that only
two are needed: (1) dynamic equilibrium in the original sense
of Gilbert (1877) and (2) steady state. Ahnert (1994) defined
that dynamic equilibrium refers to the relationship between the
process components of a system, and that steady state refers to
the system as a whole. Bracken and Wainwright (2006) continued this debate and summarized the historical development
of the concept of geomorphic equilibrium that was introduced
by Gilbert over 100 years ago and in fact highlighted that
it has always been one of the central themes in geomorphology. However, Bracken and Wainwright (2006) pointed out
that there was a great deal of variability in the ways it was
portrayed and applied, thus inflicting confusion on the discipline. Terms such as dynamic equilibrium (Gilbert, 1877;
Chorley and Kennedy, 1971; Schumm, 1973; Ahnert, 1994),
quasi-equilibrium (Petts and Foster, 1985), dynamic metastable equilibrium (Tricart, 1965; Pitty, 1971; Schumm, 1975),
steady state equilibrium (Mackin, 1948; Rubey, 1952; Hack,
1960; Richards, 1982), disequilibrium and nonequilibrium
(Renwick, 1992; Thorn and Welford, 1994) had been complicated and confused by many because they have been applied
in ways greatly different from what they were originally used to
5.16.4.1
Climatic Instability and Paraglaciation
With the detailed ice core work over the past few decades it
has become increasingly apparent that climate has oscillated
very rapidly (on decadal to millennial timescales) and over high
magnitudes (often as much as a 7 1C change in mean annual
temperature). Changes from glacial to nonglacial times have
often taken place incredibly rapidly, for example, within a few
decades out of the Younger Dryas Stade into the Holocene. With
this realization and detailed geomorphic and sedimentological
studies it has been suggested that many of the major landscape
changes have taken place during periods of climatic transition.
As such, tectonic processes have taken a back seat in many
settings. Of particular note is the development of the paraglacial
concept that was initially presented by Ryder (1971a, b) and
Church and Ryder (1972) who were examining sediment
transfer in British Columbia and Baffin Island (Figure 10(a)).
They argued that nonglacial processes are directly conditioned
by glaciation and that landscape changes are greatest during
times of deglaciation. The term paraglacial has been applied
to describe the processes that operate and the landforms that
are produced during deglaciation when landscape readjusts to
new climatic, environmental, and geomorphic conditions. The
concept of paraglacial time was presented as the period when
Commencement
of deglaciation
10
Completion
of deglaciation
8
Sediment yield
(dimensionless scale)
380
6
End of
paraglacial
period
2
Geological norm
0
Time
Proglacial period
Paraglacial period
(a)
Sediment
sources
Primary
sediment
stores
Rockwalls
Drift-mantled
slopes
Valley-floor
glacigenic deposits
Coastal
glacigenic deposits
Rock-slope failure
Rockfall
Slope failure
Debris flow
Gullying
Snow avalanches
Debris flow
Fluvial reworking
Wave action
Nearshore currents
Rockslide deposits
Talus
(Rock glaciers)
Debris-flow deposits
Debris cones
Avalanche tongues
Alluvial fans
Alluvial fans
Valley fill
River incision
Terrace formation
Barrier beaches
Spits
Baymouth bars
Barrier islands
Back-barrier deposits
Fluvial reworking
Secondary
sediment
stores
Alluvial
Valley-fill deposits
Lacustrine deposits
(Bottom sediments,
deltas)
Coastal deltas
Fjord deposits
Barrier structures
Fluvial reworking
Reworking by waves and currents
Sediment
sinks
(b)
Alluvial
Valley-fill deposits
Lacustrine
deposits
Coastal and
nearshore deposits
Shelf and
offshore deposits
paraglacial processes are dominant, mainly fluvial erosion and

resedimentation, and mass movement. Ballantyne (2002) provided a comprehensive review of paraglaciation and highlighted
its importance for landscape development in numerous settings
introducing the concept of a paraglacial cascade (Figure 10(b)).
5.16.4.2
Glacial Buzzsaw
In the late 1970s Stephen Porter noted that the equilibriumline altitude (ELA) for alpine glaciation in the Andes parallels
the summit elevations of the range. This observation has led to
the development of the glacial buzzsaw model and has subsequently been applied to understanding the topography in
many mountains, including the Cascade Mountains in
Washington State and the Chugach Range in Alaska (Meigs
and Sauber, 2000; Montgomery et al., 2001; Mitchell and
Montgomery, 2006; Spotila et al., 2004; Egholm et al., 2009).
For the Himalaya, Brozovic et al. (1997) argued that topographic data from near Nanga Parbat show that glaciers are
rapidly eroding topography above the ELA exerting an altitudinal limit on mountain height (Figure 11). Chapter 5.13
discusses the glacial buzzsaw in more detail.
In contrast to glaciation limiting topography, many researchers have argued that the associated denudational unloading causes uplift that enhances peak elevations. However,
Whipple et al. (1999) argued that there are geomorphic limits
to climate-induced increases in topographic relief and that
neither fluvial nor glacial erosion is likely to induce significant
isostatic peak uplift. Whipple et al. (1999) provided a quantitative overview and constraints, including empirical evidence
to define the effects of a transition from fluvial to glacial
erosion. Whipple et al. (1999) showed that in almost all
nonglaciated landscapes an increase in erosivity of the fluvial
system is anticipated to result in a reduction in both trunk
stream and tributary valley relief. When coupled with the
constraint that hillslope relief rapidly attains a maximum in
active orogens, this observation implies that ridge
valleybottom relief will actually decrease under these conditions. Whipple et al. (1999) suggested that relief increase is
possible only if a given climate change induces a decrease in
erosivity along headwater channel segments in concert with a
simultaneous increase in erosivity downstream.
An onset of glaciation increases hillslope relief, valley
widening, and the formation of hanging valleys and overdeepenings, and relief reduction over short wavelengths. In
contrast, over long wavelengths there is a reduction in relief
along trunk and tributary valley profiles. If the upper reaches
have thin cold-based glaciers, ridges and peaks may be protected from erosion and glacier valley profiles may become
more concave, adding potentially to overall glacial relief.
381
However, for warm-based glaciers Whipple et al. (1999) argued that relief production associated with each of the various
glacial relief production mechanisms scales with ice thickness.
Whipple et al. (1999) also argued that relief production is
limited to several hundred meters and it is unlikely that climate change would induce significant amounts of isostatic
relief production.
Bishop et al. (2003, 2010) showed that the K2 region of the
central Karakoram has been dominated by glaciation and
glacierization, and that glacier erosion limits the relief at
intermediate altitudes, thereby decreasing the variability in
scale-dependent relief. Bishop et al. (2010) argued that relatively low slope angles at intermediate altitudes document the
erosion and redistribution of material by glaciers from 4000 to
6000 m asl. Moreover, these low slope angles represent actual
glacier surfaces or high-altitude erosion surfaces from former
glaciations; these are classic signatures of a glacial buzzsaw.
Bishop et al. (2010) suggested that it is clear that the K2,
Nanga Parbat, and Hunza regions have been significantly affected by glaciation, although the subsequent influence of
local river incision and mass movement can increase local
relief. However, relief might not be related linearly to erosion
and erosional efficiency. For glacier erosion, research has indicated that glaciation can limit relief up to some altitude
approaching the ELA and that erosion fluctuations at the ELA
can generate extreme relief (Brozovic et al., 1997; Bishop et al.,
2003, 2010).
This ability of glaciers and rivers to accomplish rapid erosion underlies a proposed explanation for localized rapid rock
uplift in the syntaxes of the Himalaya, a phenomenon that has
been dubbed the tectonic aneurysm (Zeitler et al., 2001a, b;
Koons et al., 2002). This is discussed briefly in Section
5.16.5.2 and in great detail in Chapter 5.14. However, Van der
Beek et al. (2009) suggested that the Deosai Plateau of the
Western Himalaya is very slowly eroding, which is in contradiction to common belief that most of the Tibetan Plateau has
been rapidly rising.
5.16.5
5.16.5.1
Relief Production
Unloading and Uplift
The balance between uplift and denudation, or the net effect

that various geomorphic processes have in modifying the
landscape, determines whether a point on the landscape becomes higher or lower with respect to sea level. The rates of
both uplift and denudation vary greatly with time and space
and are still difficult to measure.
Molnar and England (1990) argued that glacial erosion
generating local relief in the cores of mountain ranges might
Figure 10 The paraglacial concept. (a) Graphical representation of the pattern of paraglacial sediment reworking. Reproduced from Church,
M.A., Ryder, J.M., 1972. Paraglacial sedimentation: a consideration of fluvial processes conditioned by glaciation. Geological Society of America
Bulletin 83, 30593071. (b) Simplified figure of the paraglacial sediment cascade, showing the principal primary and secondary sediment stores
and main sediment transfer processes. The cascading system operates from the top to the bottom of the figure. The sediment is likely to enter
storage on numerous occasions during source-to-sink transport (e.g., through multiple episodes of floodplain deposition or longshore transport).
All terrestrial sediment stores are sufficiently robust to survive throughout the Holocene (i.e., over a timescale of 104 years), though most
experience a transition from net accumulation to net erosion within this timescale. Relatively minor transport processes such as wind action and
direct mass-movement of sediment into lakes are not included in the figure. Reproduced from Ballantyne, C.K., 2002. Paraglacial geomorphology.
382
actually initiate tectonic uplift because of erosional unloading

of the crust in the cores of ranges (Figure 12). Moreover,
Molnar and England (1990) suggested that the onset of the
Quaternary Ice Age and enhanced glaciation led to increased
uplift of Himalayan and Tibetan peaks as a consequence of
great glacial and associated erosion. As such there is potential
for a significant positive feedback loop in which glacial erosion leads to uplift, which, in turn, produces larger glaciers
and more erosion to enhance peak uplift. Molnar and Englands (1990) paper stimulated much debate and similar
connections have been suggested elsewhere (Small and
Anderson, 1998). However, whether localized erosion can
actually result in mountain peak uplift remains controversial
(Gilchrist and Summerfield, 1991).
In active orogenic zones uplift rates can vary greatly. An
extreme uplift within only seconds can happen during earthquakes. For the 2005 Kashmir earthquake, Fujiwara et al.
(2006) using InSAR data measured coseismic vertical displacements of up to 5 m in only seconds (Figure 13). However, generally rates of more periodic surface uplift are at a
different scale. Merritts et al. (1999) calculated surface uplift
rates along the Banda ArcAustralian plate collision zone,
eastern Indonesia, of 0.21.5 m ka1 between 130 ka and
today. Higher rates are known today for tectonically very active
high mountain ranges. For the Southern Alps of New Zealand,
Bull and Cooper (1986) calculated a rate of 58 m ka1 for
the period between 140 ka and today. Nanga Parbat in the
Pakistani Himalaya probably has the highest present-day uplift rate of 4610 m ka1 (Treloar et al., 1991). However, it
does not always follow that mountains with the highest uplift
rates have the highest local elevations. For example, in the
European Alps, the highest present-day uplift rates occur on
the Rhone and Rhine valleys (Summerfield, 1991).
In general, rates of epeirogenic uplift, that is, uplift that
affects large areas of Earths crust, seem to be much lower,
although there is uncertainty about short-term rapid uplift due
to sparse data. Summerfield (1991) described average rates in
the range of 10200 m Ma1, and put the rise of the Colorado
Plateau in the western USA at 500 m during the Late Cenozoic
at 100 m Ma1 and the rise of the Deccan Plateau of India at
600 m during the past 40 Ma at 15 m Ma1. However, higher
epeirogenic uplift rates are associated with glacioisostatic rebound after Pleistocene glaciation, which progressively decrease over time. Nikonov (1980) presented rates for the
former center of the Laurentide ice sheet of up to 100 m ka1
between 10 and 8 ka, and 30 m ka1 between 4 and 3 ka.
5.16.5.2
Tectonic Aneurysm
Throughout the 1990s the tectonic aneurysm model became

popular to help explain regions of anomalous topography
such as Nanga Parbat and Namche Barwa in the syntaxial
bend areas at the east and west ends of the Himalayan arc. The
model argues for dynamic interactions of focused erosion,
topographic stresses, peak uplift, rapid exhumation, thermal
weakening of the lithosphere, and deformation that leads to
localized feedback between erosion, deformation, and uplift
(Zeitler et al., 2001a, b; Koons et al., 2002) (Figure 14).
See Chapter 5.14 for more detailed discussions.
5.16.5.3
Glacial Protection
Some researchers have suggested that glaciers might actually

protect landscapes from enhanced erosion, especially if they
are cold-based (Sugden, 1978; Fabel et al., 2002). Glacial
cover has the potential to reduce the area susceptible to mass
movement, fluvial incision, and weathering, and if glacial
erosion is low then the total erosion rates would be less
during glacial times. A new twist on this theme, based on a
study of the geomorphology of Namche Barwa, was suggested
by Korup and Montgomery (2008) (Figure 15). Geological
evidence in the Namche Barwa region has shown that the
Tsangpo and its tributary streams, which traverse Namche
Barwa, have probably been stable, remaining at the same
positions over at least the past million years, effectively preserving the margin of the Tibetan Plateau (Finnegan et al.,
2008). This is counterintuitive since rivers in this rapidly uplifting region of eastern Tibet should progressively erode
toward their heads, resulting in the steady degradation of the
plateaus margins, and that over time knickpoints should
migrate upstream and become less apparent as erosion
progresses. Various theories have been invoked to help explain the preservation of the margin of the Himalayan
Tibetan orogen, including theories of differential rock uplift
matching erosion and other mechanisms, such as landsliding
that armors valley floors and helps protect them from river
erosion (Lave and Avouac, 2001; Ouimet et al., 2007). Korup
and Montgomery (2008) argued that during glacial times,
glaciers, their associated lakes, and sediment fills protected the
landscape from net erosion. Thomson et al. (2010) found that
in the southernmost Patagonian Andes regional glacial protection resulted in a constructive tectonic response to late
Cenozoic climate cooling, or in other words, a cooling climate
can act to enhance topographic relief by inhibiting erosion.
5.16.6
5.16.6.1
Landscape Evolution and Scale

Continental Scale Landscapes
With the development of plate tectonics and new remote

sensing technologies, landscapes can be examined on continental scales. Globally the world can clearly be divided
into tectonicgeographic realms, essentially including oceanic
basins, ocean margins, continental interiors, and mountain belts
(Figure 16). Tectonics is clearly the dominant factor in driving
continental-scale landscapes as it controls gross morphology;
however, climate also has a profound influence since it influences distribution of biota, ice and surface water, global wind,
and ocean systems.
5.16.6.2
Regional-Scale Landscapes
Regional-scale landforms comprise features such as alluvial

fans, marine and river terraces, rivers and valley systems that
can be traced over many to tens of kilometers. The complexity
of these landforms is well illustrated with marine terraces,
many of which form by autocyclic processes, tectonic uplift,
and at times of higher sea level (climatic factors). Figure 17
illustrates the characteristics of typical marine terraces and
8000
Mountains
Present snowline
Lowest last
glacial snowline
Andes
Mexican
volcanoes
Alaska
range
6000
Central
America
Sierra
Nevada
Antarctica
Cascade
range
4000
Scotia sea
2000
Patagonia
Brooks
range
Arctic ocean
Altitude (m asl)
Glaciers
383
0
0
Latitude
50
Slope (degrees)
90
S
0.06
DD
40
40
30
30
20
20
0.05
0.04
0.03
0.02
10
10
4100 m
3800 m
0.01
0.00
0
0
10
50
10
50
Slope (degrees)
60
50
DD
0.06
SN
40
40
30
30
20
20
0.05
0.04
0.03
0.02
10
10
4800 m
4100 m
0.01
0.00
0
0
(b)
30
10
Frequency
30
Frequency
60
90
N
(a)
10
Elevation (km)
Figure 11 Glacial buzzsaw model. (a) The present and minimum snowline elevations during the last glacial are plotted along a topographic transect
along the cordillera of North and South Americas illustrating the positive relationship between topography and snowline elevation. Reproduced with
permission from Broecker, W.S., Denton, G.H., 1990. What drives glacial cycles? Scientific American, January, 4956, and Skinner, B.J., Porter, S.C.,
Park, J., 2004. Dynamic Earth: An Introduction to Physical Geology. Wiley, Chichester. (b) Frequency distribution of altitude and slope angles for
selected physiographic areas within the western syntaxis of the Himalaya (DD dissected portion of Deosai Plateau, NP Nanga Parbat, G
Ghujerab mountains (northern Karakoram), SN area north of Skardu). The arrows on the horizontal axis indicate the mean elevation for each
region; values are inside the graphs. The dark vertical lines running from top to bottom in the graphs show the range of snowline altitudes for each
region. The three subhorizontal gray lines show the 25th (lower line), 50th, and 75th (upper line) percentile of slope distribution as a function of
altitude for each region. The light blue shaded areas on the slope distribution curves highlight the regions of moderate slopes that generally coincide
with the modal elevations. The graphs show the strong relationship between snowline elevation and topography, which suggests that glaciation
places limits on topography. Reproduced with permission from Brozovic, N., Burbank, D.W., Meigs, J.A., 1997. Climatic limits on landscape
development in the northwestern Himalaya. Science 276, 571574, and Owen, L.A., 2010. Landscape development of the HimalayanTibetan orogen:
a review. Special Publication of the Geological Society of London 338, 389407.
384
Decline in summit elevations
Increase in summit elevations
ES
Z
Z
S
ES = E
EV
Present
topography
Present
topography
U
U
Spatially uniform erosion
Original crustal block
Spatially nonuniform erosion
Figure 12 Model of relief generation via glacial erosion for the case of mountain erosion. The center panel represents initial condition. When
erosion (shaded area) is spatially uniform (ES EE), the sum of erosionally driven rock uplift (UE) and summit erosion (ES) results in lower
summit elevations (DZS) (left panel) where EE is the average erosion. When erosion is spatially variable (right panel), changes in summit
elevation are positive because rock uplift is greater than summit erosion (DZS ES UE (x,y)). Rock uplift is the same in each case because EE,
which drives rock uplift, is equal (shaded areas are the same size). The present geophysical relief (R) is the mean elevation difference between a
smooth surface connecting the highest points in the landscape (dashed line) and the current topography. It represents the average of valley
erosion (EV) minus summit erosion calculated at each point in the landscape, including summit flats. Reproduced from Owen, L.A., 2010.
Landscape development of the HimalayanTibetan orogen: a review. Special Publication of the Geological Society of London 338, 389407;
Small, E.E., Anderson, R.S., 1998. Pleistocene relief production in Laramide Mountain Ranges, western U.S. Geology 26, 123126, and modified
with permission from Molnar, P., England, P., 1990. Late Cenozoic uplift of mountain ranges and global climate change: chicken or egg? Nature
346, 2934.
their mode of formation. In regions such as Southern California, marine terraces can be traced over many tens to hundreds of kilometers (Muhs et al., 2002). These terraces are a
function of higher sea level during interglacial or interstadial
times, and continued coastal uplift.
River terraces provide other examples of polygenetic regional-scale landforms. Their formative processes are highly
complex as illustrated in Figure 18. In the Karakoram and
Hindu Kush of northern Pakistan, for example, Owen (1989)
and Kamp (2001) recognized numerous different types of terraces based on their sedimentology and origin which included
morainic, glaciofluvial, fluvial, lacustrine, mass-movement debris, and fan terraces. The fan terraces, for example, are alluvial
fans in the sense of those described by Bull (1968) and Patton
et al. (1970) and represent polygenetic landforms comprising
numerous sediments, such as mass-movement debris, glacial,
glaciofluvial, fluvial, lacustrine, and aeolian material. Most of
the terraces represent a multiple change of accumulation and
erosion periods, which dominantly reflect times of climatic instability, where rates of erosion, sediment transfer, and deposition dramatically change to produce terrace landforms.
depositional landforms is often somewhat arbitrary, and will

depend upon the temporal and spatial scale in which one is
interested. For example, a yardang is both a depositional and
erosional landform: It consists of layers of prior sediments,
which later have been eroded by aeolian activity. The same is
true for all types of valley terraces such as alluvial fans, river
terraces, kame terraces, and loess terraces. However, polygenetic landforms are not always the result of both depositional and erosional processes. Different process intensity
might develop different morphologies. As examples, in karst
landscapes the superposition and complex intermixing of
karren morphologies might produce more complex polygenetic karren forms; and in the coastal littoral zone the
combined effects from boring and grazing marine organisms
and salt weathering might produce littoral karren (Fornos
and Gines, 1996). Furthermore, the creation of polygenetic
landforms can, and often is, influenced not only by tectonics,
directly in terms of uplift and faulting, but also by joint development in bedrock as a consequence of local or regional
tectonic stress fields.
5.16.7
5.16.6.3
Landform Scale
Many landforms are the product of polygenetic processes and

have evolved at different times and under ever-changing climates; thus great care is needed in using landforms to infer
climate (Williams, 2009). Williams (2009: 803) explained
that, although often landforms evolved mainly by erosion or
deposition, the classical distinction between erosional and
Mathematical and Computational Modeling
In the past decades, the capitalization of increasingly powerful

computer systems in landscape evolution modeling has
greatly contributed to theory development and improvement
(Beaumont et al., 2004; Champagnac et al., 2007; Whipple,
2009). Without mathematical and computational modeling,
our ideas of how the different Earths spheres interact at the
surface interface would not be as sophisticated as they are
(cm)
385
500
400
300
200
100
0
100
200
300
(a)
(b)
Figure 13 Ground displacement resulting from the 2005 Pakistan earthquake. (a) Combined crustal deformation map superimposed on
topography and the location of known active faults. Positive values indicate the movement of deformation, in centimeters, toward the SAR
satellite in the LOS direction (upward or E-SE displacement). The red star shows the epicenter determined by the U.S. Geological Survey. The
contour lines are each 250 m. The black curves show the locations of active faults. (b) Oblique view of crustal deformation and active faults
(from the south). Reproduced with permission from Fujiwara, S., Tobita, M., Sato, H.P., et al., 2006. Satellite data gives snapshot of the 2005
Pakistan Earthquake. EOS, Transitions of the American Geophysical Union 87, 7377.
today. Some modeling processing steps require hours, days, or

even weeks to run, and easily terabytes of data are included.
New approaches to measure topography, particularly from
space, deliver even more data, and at increasingly large scale.
Tucker and Hancock (2010) divided the history of modeling landscape evolution into two phases, with the 1980s being
the turning point. Until the 1980s, landscape evolution
models described qualitatively the sequential evolution of a
landscape over geologic time. But by the end of the twentieth
century, landscape evolution models described mathematical

theories about the interactions between geomorphic processes
and the evolution of topography over time. Today, a model
has often come to refer to both the underlying theory and the
computer programs that calculate approximate solutions to
the equations (Tucker and Hancock, 2010: 28).
In recent decades, numerous mathematical and computational landscape evolution models have been introduced, for
example, APERO/CIDRE (Carretier and Lucazeau, 2005),
386
Evacution
ion
os
Er
w
Diversio
n of crusta l f l o
Rapid
a
Metam dvection.
orphis
Anate
m,
xis
Figure 14 Schematic representation illustrating the dynamics of a tectonic aneurysm, shown at a mature stage based on Nanga Parbat. The
model argues that the dynamic interactions of focused erosion, topographic stresses, peak uplift, rapid exhumation, thermal weakening of the
lithosphere, and deformation, which leads to localized feedback between erosion, deformation, and uplift. Reproduced from Zeitler, P.K., Meltzer,
A.S., Koons, P.O., et al., 2001b. Erosion, Himalayan geodynamics and the geomorphology of metamorphism. Geological Society of America
Today 11, 48.
Tibe
tan +
1. Glacier advance
Hima
layan
marg
in
Grea
t rive
r gor
ge
Hima
la
forela yan
nd
2. Sediment deposition
6. Glacier retreat
River
sediments
LAKE
Glacier and/or
moraine
3. Stream incision
Lake
sediments
Knick
point
4. Bedrock uplift
5. Reduction in
erosive power
Figure 15 Possible mechanism for the preservation of the edge of southeast Tibet. The figure schematically shows the following is the
sequence of events: (1) a glacier advance dams a river; (2) the deposition of sediments in the resulting upstream lake and stream both protects
the valley floor from erosion and reduces the streams erosive power; (3) streams draining through the glacier or moraine are confined, so
stream flow is strong and downstream erosion increases; (4) stream incision into the bedrock essentially weakens the crust and enhances
bedrock uplift to the surface; (5) as streams travel farther from the knick point, smaller gradients reduce their erosive power; (6) the glacier
retreats, but the events are repeated numerous times as glaciers advance and retreat in response to climate change. Note that the repetition of
this succession of events maintains equilibrium between erosion and bedrock uplift, essentially preserving the topography of the margin of
southeast Tibet. Reproduced from Korup, O., Montgomery, D.R., 2008. Tibetan plateau river incision inhibited by glacial stabilization of the
Tsangpo gorge. Nature 455, 786789; Owen, L.A., 2008. How Tibet might keep its edge. Nature 455, 748749, and Owen, L.A., 2009. Latest
Pleistocene and Holocene glacier fluctuations in the Himalaya and Tibet. Quaternary Science Reviews 28, 21502164.
N.
Co Am
rd eri
ille ca
ra
NORTH
AMERICAN
PLATE
387
Urals
EURASIAN PLATE
Alps
s
n
hia
lac
pa
Ap
CARIBBEAN
PLATE
ARABIAN PLATE
Tibet
PACIFIC
PLATE
Himalaya
SOUTH
AMERICAN
PLATE
COCOS
PLATE
PHILIPPINE
PLATE
INDO-AUSTRALIAN PLATE
Se
Dr
rra
ak
do
Ma
en
s
M bur
ts g
s
de
An
NAZCA
PLATE
Wn
Ghats
AFRICAN
PLATE
East
African
Rift
Valley
ANTARCTIC PLATE
SOUTH
SANDWICH PLATE
Rift valleys
Mid-oceanic ridges
Subduction zone
Mountain belts of Cenozoic age
Continental-continental collision zone
Partially eroded mountain belts of Mesozoic and Late Palaeozoic age
Transform plate boundary
Highly eroded mountain belts of Early Palaeozoic age
Oceanic spreading ridge and

associated transform faults
Lowlands and plateaus (continental platform)
Great escarpments and associated

mountains
Hotspots
Figure 16 Distribution and plate tectonic setting for the major mountain systems. Reproduced with permission from Summerfield, M.A., 1991.
Global Geomorphology: An Introduction to the Study of Landforms. Wiley, New York, and Owen, L.A., 2004. Cenozoic evolution of global
mountain systems. In: Owens, P.N., Slaymaker, O. (Eds.), Mountain Geomorphology. Edward Arnold, London, pp. 3358.
CASCADE (Braun and Sambridge, 1997), CAESAR (Coulthard

et al., 1997), CHILD (Tucker and Bras, 2000), DELIM
(Howard, 1994), DRAINAL (Beaumont et al., 1992), EROS
(Crave and Davy, 2001), GILBERT (Chase, 1992), SIBERIA
(Willgoose et al., 1991), and ZSCAPE (Densmore et al., 1998).
Describing each model is beyond the scope of this chapter, but
the reader is referred to reviews by Carson and Kirkby (1972),
Kirkby (1996), Coulthard (2001), Wilcock and Iverson
(2003), Martin and Church (2004), Codilean et al. (2006),
Bishop (2007) and Tucker and Hancock (2010). In their extensive review, Tucker and Hancock (2010) identified the
most popular numerical methods to solve the governing
equations in landscape evolution modeling including finite
difference (Willgoose et al., 1989; Howard, 1994; Tucker and

Slingerland, 1997), finite volume (Braun and Sambridge,
1997; Tucker et al., 2001a, b), finite element (Simpson and
Schlunegger, 2003), and cellular automaton methods (Chase,
1992; Favis-Mortlock, 1998; Crave and Davy, 2001).
Basic components of a computational landscape evolution
model are: (1) statement of continuity of mass, (2) geomorphic transport functions for hillslope processes, (3) representation of runoff, (4) geomorphic transport functions
for erosion and transport by water, (5) numerical methods
to obtain solutions to the governing equations, and (6) consideration of initial and boundary conditions on the system
(Tucker and Hancock, 2010). Although results from such
388
Terrace
surface
Pale
o-se
vium
Collu
M
depo arine
sits
Mo
der
Mod
nb
ern s
eac
Old
wave
platf -cut
orm
ea c
a clif
liff
Paleo-shoreline
angle
Mo
der
nw
pla avecut
tfor
m
Modern-shoreline
angle
(a)
Paleoshoreline
T2
Fault
Anticline
T1
T2
T2
SC
T1
Mo
T1
der
VE = 10
10
np
latf
orm
km
T2
SC
T1
(b)
Figure 17 Marine terraces. (a) General morphology of an erosional coastline: active features include the modern wave-cut platform, beach, and
seacliff. An older uplifted and now-inactive shoreline is preserved by the old wave-cut platform, its cover of marine sediment, and the eroded
remnant of the old sea cliff. (b) Faulting, warping, and erosion of an uplifted coastal terrace after formation. Reproduced from Keller, E.A., Pinter,
N., 2002. Active Tectonics: Earthquakes, Uplift and Landscape. Prentice Hall, Upper Saddle River.
modeling contributed much to our understanding of landscape evolution, the models have significant limitations.
For example, popular 2D vertical-column continuity frameworks are often simplistic so that they are unable to represent
landforms with vertical or overhanging faces such as
cliffs and gully headscarps, or unable to represent vertical
variations in weathering rates or shallow groundwater flow
(Tucker and Hancock, 2010). Three-dimensional modeling

approaches tried to overcome such limitations; for example,
Kirkby (1976) included variations in soil properties such
as porosity and mineral density, and Brimhall and Dietrich
(1987) modeled additional parameters like gravitational
compaction and contraction/expansions due to changes in
mineralogy.
Fill 1
Fill 1
389
Aggradational
terraces
Fill 2
Fill 3
Gravel cap
on strath
Degradational
terraces
(a)
Paired
terraces
Unpaired
terraces
(b)
Aggradational
terrace 1
Aggradational
Strath terrace
Aggradational
terrace 2
terrace 3
Degradational
terraces
Cut-and-fill history
Incision
Aggradation
Time
(c)
Figure 18 Formation of river terraces. (a) Aggradational and degradational fluvial terraces. (b) Paired and unpaired river terraces. (c) Complex
sequence of aggradational and degradational surfaces. Multiple cut-and-fill events are outlines in the right-hand box. Reproduced from Burbank,
D.W., Anderson, R.S., 2002. Tectonic Geomorphology. Blackwell Scientific, Oxford, with permission from Wiley.
The theoretical core of any landscape evolution model is its

geomorphic transport functions (GTFs), and they must include the description of processes and changes within the
critical zone (Tucker and Hancock, 2010), defined as the layer
bounded by the top of the forest canopy and the base of the
weathering horizon (National Research Council, 2001). Until
recently, the modeling of rates and patterns of disintegration
was almost neglected. The exponential decay function developed by Ahnert (1967) based on the theory from Gilbert
(1877) was proven to successfully predict the inverse relationship between regolith thickness and erosion rate for many
landscape types (McKean et al., 1993; Heimsath et al., 1997;
Small et al., 1999). Although much progress has been made in
recent years, many GTFs still require attention. Tucker and
Hancock (2010) identified what future landscape evolution
modeling must incorporate: (1) the mechanics and chemistry
of weathering processes; (2) more process-specific models for

the linear creep function including field and experimental
tests; (3) models that better handle deep-seated landsliding
and shallow, rapid landsliding; and (4) models that describe
the net effect of many landslides.
Another central issue in landscape evolution models is the
treatment of runoff dynamics (Tucker and Hancock, 2010).
Many modeling approaches use considerably simplified equations. For example, most landscape evolution models assume
quasi-steady flow. However, some approaches try to model a
more realistic nonsteady flow (Solyom and Tucker, 2004).
Models of drainage basin evolution now address feedback between climate, hydrology, and landscape evolution (Beaumont
et al., 1992; Huang and Niemann, 2006; Anders et al., 2008).
Although much has been achieved in the numerical
and computational modeling of landscape evolution, many
390
CLIMATE
Precipitation
EROSION
Temperature
Surface
Temperature
TOPOGRAPHY
Weathering
Altitude
Fluvial Incision
Slope Angle
Mass
Movements
Slope
Azimuth
Glacial
Processes
TEC TONICS
Relief
Crustal
Thickening
Uplift
Deformation
Figure 19 Interactions and selected feedback pathways for climate, erosion processes, and tectonics. Note that many parameters of the system
are not included and the magnitudes of the linkages change with time, and that positive and negative feedback exist for selected pathways.
Reproduced from Bishop, M.P., Bush, A.B.G., Copland, L., Kamp, U., Owen, L.A., Seong, Y.B., Shroder, Jr. J.F., 2010. Climate change and
mountain topographic evolution in the Central Karakoram, Pakistan. Annals of the Association of American Geographers 100, 772793.
questions remain unanswered or have not even been formulated. Tucker and Hancock (2010: 41), for example, listed some
knowledge gaps that require attention in future research: the
dynamics associated with varying grain-size distributions, the
role of lithology, the horizontal movement of geomorphic features due to processes such as scarp retreat and tectonic displacement, the role of biota, the dynamics of stream-channel
adjustment, erosion and transport by debris flows and other
mass movements, and the formation and evolution of the critical zoney as well as the general issue of understanding the
physical, chemical, and biological controls on rate coefficients.
Furthermore, Tucker and Hancock (2010) stated that the next
generation of models will need creative methods to handle
features and processes without introducing scale effects, and that
the community must develop new methods for describing and
quantifying landforms rather than just new numerical methods.
5.16.8
Conclusion
Recent tectonic geomorphic studies have led to the realization

that complex links, interactions, and feedbacks exist between
tectonics, climate, and landscape development. Figure 19

illustrates some of the factors and links between different
endogenetic and exogenetic systems that result in the polygenetic landscape on this planet. The new climate
versustectonic forcing hypotheses are difficult to relate directly, as numerous feedback mechanisms are involved, and
direct linkages among climate, erosion, and tectonics must be
established to demonstrate a tectonic response to climate
change. Notably, the new models based on these complex
relationships include: the glacial buzzsaw where glaciation
limits topography; the influence of climate on rock uplift by
denudational unloading of the crust; the tectonic aneurysm
model in which localized uplift at syntaxes is caused by enhanced fluvial and glacial erosion that, in turn, weaken the
lithosphere and enhances bedrock uplift; climate driven outof-sequence thrusting and focused erosion driving extrusion of
ductile crustal channels; and glacial damming leading to differential erosion and uplift. Probably the greatest challenges
are to examine the complex variability between regions and to
quantify the timing, rates, and magnitude of landscape evolution on various timescales ranging from 100 to 106 years. The
development and application of new technologies, including
satellite remote sensing and global positioning systems, plus

analytical methods such as numerical dating, will allow many
of the new theories and models to be tested.
Human-induced climate change is having profound effects
on geomorphic systems, and predicting the nature of future
change is challenging. As a reaction to such changes and
challenges, Paul Crutzen, the Nobel Prize-winning chemist,
suggested in 2002 that we had left the Holocene and had
entered a new Epoch the Anthropocene because of the
global environmental effects of increased human population
and economic development (Zalasiewicz et al., 2008). In a
similar vein, Haff (2003) introduced the term neogeomorphology to describe current and future human impacts on the
Earths surface. Examining the nature of natural and humaninduced factors in producing polygenetic landscapes is among
the greatest challenges faced in the study of the Earths surface
and internal processes in the coming years (Committee on
Challenges and Opportunities in Earth Surface Processes,
2010).
References
Ahnert, F., 1967. The role of the equilibrium concept in the interpretation of
landforms of fluvial erosion and deposition. In: Macar, P. (Ed.), LEvolution des
versants. Les Cong et Colloques de lUniversite de Lie`ge, Lie`ge, vol. 40, pp.
2241.
Ahnert, F., 1994. Equilibrium, scale and inheritance in geomorphology.
Anders, A.M., Roe, G.H., Montgomery, D.R., Hallet, B., 2008. Influence of
precipitation phase on the form of mountain ranges. Geology 36, 479482.
Ballantyne, C.K., 2002. Paraglacial geomorphology. Quaternary Science Reviews 21,
19352017.
compressional orogens. In: McClay, K.R. (Ed.), Thrust Tectonics. Chapman Hall,
New York, pp. 118.
flows: 1. Numerical models with applications to the tectonics of the HimalayanTibetan orogen. Journal of Geophysical Research 109, B06406.
Bishop, P., 2007. Long-term landscape evolution: linking tectonics and surface
processes. Earth Surface Processes and Landforms 32, 329365.
Shroder, Jr. J.F., 2010. Climate change and mountain topographic evolution in
the Central Karakoram, Pakistan. Annals of the Association of American
Bishop, M.P., Shroder, Jr. J.F. (Eds.), 2004. Geographic Information Science and
Mountain Geomorphology. Praxis-Springer, Chichester.
Bishop, M.P., Shroder, Jr. J.F., Bonk, R., Olsenholler, J., 2002. Geomorphic change
in high mountains: a western Himalayan perspective. Global and Planetary
Change 32, 311329.
Bracken, L.J., Wainwright, J., 2006. Geomorphological equilibrium: myth and
metaphor? Transactions of the Institute of British Geographers NS 31, 167178.
Braun, J., Sambridge, M., 1997. Modelling landscape evolution on geological time
scales: a new method based on irregular spatial discetization. Basin Research 9,
2752.
Brimhall, G.H., Dietrich, W.E., 1987. Constitutive mass balance relations between
chemical composition, volume, density, porosity, and strain in metasomatic
hydrochemical systems: results on weathering and pedogenesis. Geochimica et
Brozovic, N., Burbank, D.W., Meigs, J.A., 1997. Climatic limits on landscape
development in the northwestern Himalaya. Science 276, 571574.
Brunsden, D., 1980. Applicable models of long term landform evolution. Zeitschrift
fur Geomorphologie N.F. Suppl. 36, 1626.
391
Brunsden, D., Jones, D.K.C., 1984. The geomorphology of high magnitudelow

frequency events in the Karakoram Mountains. In: Miller, K. (Ed.), International
Karakoram Project. Cambridge University Press, Cambridge, pp. 343388.
Brunsden, D., Thornes, J.B., 1979. Landscape sensitivity and change. Transactions
of the Institute of British Geographers NS 4, 463484.
Budel, J., 1957. Die Doppelten Einebnungsflachen in den feuchten Tropen.
Zeitschrift fur Geomorphologie N.F. 1, 201228 (In German).
Bull, W.B., 1968. Alluvial fans. Journal of Geological Education 16, 101106.
Bull, W.B., Cooper, A.F., 1986. Uplifted marine terraces along the Alpine fault, New
Zealand. Science 234, 12251228.
Carretier, S., Lucazeau, F., 2005. How does alluvial sedimentation at range fronts
modify the erosional dynamics of mountain catchments? Basin Research 17,
361381.
University Press, Cambridge.
Champagnac, J.D., Molnar, P., Anderson, R.S., Sue, C., Delacou, B., 2007.
Quaternary erosion-induced isostatic rebound in the Western Alps. Geology 35,
195198.
Chase, C.G., 1992. Fluvial landsculpting and the fractal dimension of topography.
Chorley, R.J., Kennedy, B.A., 1971. Physical Geography: A Systems Approach.
Prentice Hall, London.
Church, M.A., Ryder, J.M., 1972. Paraglacial sedimentation: a consideration of
fluvial processes conditioned by glaciation. Geological Society of America
Bulletin 83, 30593071.
Clift, P.D., Giosan, L., Blusztajn, J., et al., 2008a. Holocene erosion of the Lesser
Himalaya triggered by intensified summer monsoon. Geology 36, 7982.
Clift, P.D., Hodges, K.V., Heslop, D., Hannigan, R., Long, H.V., Calves, G., 2008b.
Correlation of Himalayan exhumation rates and Asian monsoon intensity. Nature
Codilean, A.T., Bishop, P., Hoey, T., 2006. Surface process models and the links
between tectonics and topography. Progress in Physical Geography 30,
307333.
Committee on Challenges and Opportunities in Earth Surface Processes, National
Research Council, 2010. Landscapes on the Edge: New Horizons for Research
on Earths Surface. The National Academies Press, Washington D.C.
Coulthard, T.J., 2001. Landscape evolution models: a software review. Hydrological
Coulthard, T.J., Kirkby, M.J., Macklin, M.G., 1997. Modelling hydraulic sediment
transport and slope processes, at a catchment scale, using a cellular automaton
approach. In: Pascoe, R.T. (Ed.), Proceedings of the Second Annual Conference
GeoComputation. University of Otago, Dunedin, 97, pp. 309318.
Crave, A., Davy, P., 2001. A stochastic precipiton model for simulating erosion/
sedimentation dynamics. Computers and Geosciences 27, 815827.
Davis, W.M., 1889. The rivers and valleys of Pennsylvania. National Geographic
Magazine 1, 183253.
Davis, W.M., 1899. The geographical cycle. The Geographical Journal 14, 481504.
Densmore, A., Ellis, M., Anderson, R., 1998. Landsliding and the evolution of
normal-fault-bounded mountains. Journal of Geophysical Research 103,
203219.
Dunai, T.J., Lopez, G.A.G., Juez-Larre, J., 2005. OligoceneMiocene age of aridity
in the Atacama Desert revealed by exposure dating of erosion-sensitive
landforms. Geology 33, 321324.
Egholm, D.L., Nielsen, S.B., Pedersen, V.K., Lesemann, J.E., 2009. Glacial effects
limiting mountain height. Nature 460, 884888.
Fabel, D., Stroeven, A.P., Harbor, J., Kleman, J., Elmore, D., Fink, D., 2002.
Landscape preservation under Fennoscandian ice sheets determined from in situ
produced 10Be and 26Al. Earth and Planetary Science Letters 201, 397406.
Favis-Mortlock, D., 1998. A self-organizing dynamic systems approach to the
simulation of rill initiation and development of hillslopes. Computers and
Finlayson, D.R., Montgomery, D.R., Hallet, B., 2002. Spatial coincidence of rapid
219222.
Yuping, L., 2008. Coupling of rock uplift and river incision in the Namche
BarwaGyala Peri massif, Tibet. Geological Society of America Bulletin 120,
142155.
Fornos, J.J., Gines, A. (Eds.), 1996. Karren Landforms. Servei de Publicacions,
Universitat de les Illes Balears, Palma de Mallorca.
Fujiwara, S., Tobita, M., Sato, H.P., et al., 2006. Satellite data gives snapshot of the
2005 Pakistan Earthquake. EOS, Transitions of the American Geophysical Union
87, 7377.
392
Gilbert, G.K., 1877. Report on the Geology of the Henry Mountains United States
Geographical and Geological Survey of the Rocky Mountains Region. United
States Government Printing Office, Washington.
Gilbert, G.K., 1909. The convexity of hilltops. Journal of Geology 17, 344350.
Gilbert, G.K., 1914. The Transportation of Debris by Running Water. United States
Geological Survey Professional Paper 86.
Gilchrist, A.R., Summerfield, M.A., 1991. Denudation, isostasy and landscape
evolution. Earth Surface Processes and Landforms 16, 555562.
Gilchrist, A.R., Summerfield, M.A., Cockburn, H.A.P., 1994. Landscape dissection,
isostatic uplift, and the morphologic development of orogens. Geology 22,
963966.
Hack, J.T., 1960. Interpretation of erosional topography in humid temperate regions.
American Journal of Science 258, 8097.
Hack, J.T., 1975. Dynamic equilibrium and landscape evolution. In: Melhorn, W.N.,
Flemal, R.C. (Eds.), Theories of Landform Development. Allen and Unwin,
London, pp. 87102.
Haff, P.K., 2003. Neogeomorphology, prediction and anthropic landscape.
In: Wilcock, P., Iverson, R. (Eds.), Prediction in Geomorphology. American
Geophysical Union Monograph, Washington, pp. 1527.
Hall, A.D., Fagen, R.E., 1956. Definition of system. General Systems (Yearbook of
the Society for the Advancement of General Systems Theory) 1, 1828.
glaciers: a review of field data and their implications. Global and Planetary
Change 12, 213235.
Harbor, J., Warburton, J., 1992. Glaciation and denudation rates. Nature 356, 751.
Heimsath, A.M., Dietrich, W.E., Nishiizumi, K., Finkelm, R.C., 1997. The soil
production function and landscape equilibrium. Nature 388, 358361.
Howard, A.D., 1994. A detachment limited model for drainage basin evolution.
Huang, X., Niemann, J.D., 2006. An evaluation of the geomorphically effective event
for fluvial processes over long periods. Journal of Geophysical Research 111,
F03015. http://dx.doi.org/10.1029/2006JF000477.
Huggett, R.J., 2011. Fundamentals of Geomorphology. Routledge, London.
Kamp, U., 2001. Jungquartare terrassen und talentwicklung in chitral, ostlicher
hindukusch. Zeitschrift fur Geomorphologie 45, 453475 (In German).
101, 631642.
115, 102108.
King, L.C., 1957. The uniformitarian nature of hillslopes. Transactions of the
Edinburgh Geological Society 17, 81102.
Kirkby, M.J., 1976. Deterministic continuous slope models. Zeitschrift fur
Geomorphologie, Supplement Band 25, 119.
Kirkby, M.J., 1996. A role for theoretical models in geomorphology? In: Rhoads,
B.L., Thorne, C.E. (Eds.), The Scientific Nature of Geomorphology. Wiley,
Lave, J., Avouac, J.P., 2001. Fluvial incision and tectonic uplift across the Himalayas
of Central Nepal. Journal of Geophysical Research 106, 26,561-26,592.
Liu, J.G., Kusky, T.M., 2008. After the quake: a firsthand report on an international
field excursion to investigate the aftermath of the China earthquake. Earth
Magazine October, 4851.
Mackin, J.H., 1948. Concept of the graded river. Bulletin of the Geological Society
Martin, Y., Church, M., 2004. Numerical modelling of landscape evolution:
geomorphological perspectives. Progress in Physical Geography 28, 317339.
McKean, J.A., Dietrich, W.E., Finkel, R.C., Southon, J.R., Caffee, M.W., 1993.
Quantification of soil production and downslope creep rates from cosmogenic
10
Be accumulations on a hillslope profile. Geology 21, 343346.
Merritts, D., Eby, R., Harris, R., Edwards, R.L., Chang, H., 1999. Variable rates of
Late Quaternary surface uplift along the Banda Arc-Australian plate collision
zone, eastern Indonesia. Geological Society of London Special Publication 146,
213224.
Mitchell, S.G., Montgomery, D.R., 2006. Influence of a glacial buzz-saw on the

height and morphology of the Cascade Range in central Washington State, USA.
Montgomery, D.R., Balco, G., Willett, S.D., 2001. Climate, tectonics and the
morphology of the Andes. Geology 29, 579582.
Montgomery, D.R., Brandon, M.T., 2002. Topographic controls on erosion rates in
tectonically active mountain ranges. Earth and Planetary Science Letters 201,
481489.
Muhs, D.R., Simmons, K.R., Kennedy, G.L., Rockwell, T.K., 2002. The last
interglacial period on the Pacific Coast of North America: timing and
paleoclimate. Geological Society of America Bulletin 114, 569592.
National Research Council, 2001. Basic Research Opportunities in Earth Science.
Washington, D.C.
Nikonov, A.A., 1980. Determination of magnitudes and recurrence time for past
earthquakes from seismodislocations: the Pamirs and Tien Shan junction zone.
Doklady Akademii Nauk SSSR 250 2, 336341 (In Russian).
Ouimet, W.B., Whipple, K.X., Royden, L.H., Sun, Z., Chen, Z., 2007. The influence
of large landslides on river incision in a transient landscape: eastern margin of
the Tibetan Plateau (Sichuan, China). Geological Society of America Bulletin
119, 14621476.
Owen, L.A., 1989. Terraces, uplift and climate in the Karakoram Mountains,
Northern Pakistan: Karakoram intermontane basin evolution. Zeitschrift fur
Owen, L.A., Kamp, U., Khattak, G.A., Harp, E.L., Keefer, D.K., Bauer, M., 2008.
Landslides triggered by the October 8, 2005, Kashmir earthquake.
Patton, C.P., Alexander, C.S., Kramer, F.L., 1970. Physical Geography. Wadsworth,
Belmont.
Penck, W., 1924. Die Morphologische Analyse: Ein Kapitel der Physikalischen
Geologie. Engelhorn, Stuttgart (In German).
Petts, G.E., Foster, I.D.L., 1985. Rivers and Landscape. Edward Arnold, London.
Pitty, A.F., 1971. Introduction to Geomorphology. Methuen, London.
Rahaman, W., Singh, S.K., Sinha, R., Tandon, S.K., 2009. Climate control on
erosion distribution over the Himalaya during the past B100 ka. Geology 37,
559562.
Renwick, W.H., 1992. Equilibrium, disequilibrium and non equilibrium landforms in
the landscape. Geomorphology 5, 265276.
Richards, K., 1982. Rivers: Form and Process in Alluvial Channels. Methuen,
London.
Rubey, W.W., 1952. Geology and Mineral Resources of the Hardin and Brussels
Quadrangles (in Illinois). United States Geological Survey Professional Paper 218.
Ryder, J.M., 1971a. Some aspects of the morphology of paraglacial alluvial fans in
south-central British Columbia. Canadian Journal of Earth Sciences 8,
12521264.
Ryder, J.M., 1971b. The stratigraphy and morphology of paraglacial alluvial fans in
south-central British Columbia. Canadian Journal of Earth Sciences 8, 279298.
Schafer, J.M., Ivy-Ochs, S., Wieler, R., Leya, I., Baur, H., Denton, G.H., Schluchter,
C., 1999. Cosmogenic noble gas studies in the oldest landscape on Earth:
surface exposure ages of the Dry Valleys, Antarctica. Earth and Planetary
Schumm, S.A., 1973. Geomorphic thresholds and complex response of drainage
systems. In: Morisawa, M. (Ed.), Fluvial Geomorphology, Proceedings of the 4th
Annual Geomorphology Symposia Series. Allen and Unwin, Binghamton,
pp. 299311.
Schumm, S.A., 1975. Episodic erosion: a modification of the geomorphic cycle.
In: Melhorn, W.N., Flemal, R.C. (Eds.), Theories of Landform Development. Allen
and Unwin, London, pp. 6985.
Simpson, G.H.D., Schlunegger, F., 2003. Topographic evolution and morphology of
surfaces evolving in response to coupled fluvial and hillslope sediment
transport. Journal of Geophysical Research 108, B6. http://dx.doi.org/10.1029/
2002JB002162.
Small, E.E., Anderson, R.S., 1998. Pleistocene relief production in Laramide
Mountain Ranges, western U.S. Geology 26, 123126.
Small, E.E., Anderson, R.S., Hancock, G.S., 1999. Estimates of the rate of regolith
production using 10Be and 26Al from an alpine hillslope. Geomorphology 27,
131150.
Solyom, P.B., Tucker, G.E., 2004. Effect of limited storm duration on landscape
evolution, drainage basin geometry, and hydrograph shapes. Journal of
Spotila, J.A., Busher, J.T., Meigs, A.J., Reiners, P.W., 2004. Long-term glacier
erosion of active mountain belts: example of the ChugachSt. Elias Range,
Alaska. Geology 32, 501504.
Sugden, D.E., 1978. Glacial erosion by the Laurentide ice sheet. Journal of
Summerfield, M.A., 1991. Global Geomorphology: An Introduction to the Study of
Landforms. Wiley, New York.
Thomson, S.N., Brandon, M.T., Tomkin, J.H., Reiners, P.W., Vasquez, C., Wilson,
N.J., 2010. Glaciation as a destructive and constructive control on mountain
building. Nature 467, 313317.
Thorn, C.E., Welford, M.R., 1994. The equilibrium concept in geomorphology.
Annals of the Association of American Geographers 84, 666696.
Tomkin, J.H., Braun, J., 2002. The influence of alpine glaciations on the relief of
tectonically active mountain belts. American Journal of Science 302, 169190.
Treloar, P.J., Potts, G.J., Wheeler, J., Rex, D.C., 1991. Structural evolution and
asymmetric uplift of Nanga Parbat syntaxis, Pakistan Himalaya. Geologische
Rundschau 80, 411428.
Tricart, J., 1965. Les discontinuities dans le phenomenes derosion. International
Association of Scientific Hydrology 59, 233243 (In French).
Tucker, G.E., Bras, R.L., 2000. A stochastic approach to modeling the role of rainfall
variability in drainage basin evolution. Water Resources Research 36, 19531964.
Tucker, G.E., Hancock, G.R., 2010. Modelling landscape evolution. Earth Surface
Tucker, G.E., Lancaster, S.T., Gasparini, N.M., Bras, R.L., 2001a. The channel-hillslope
integrated landscape development model (CHILD). In: Harmon, R.S., Doe III, W.W.
(Eds.), Landscape Erosion and Evolution Modeling. Kluwer, New York.
Tucker, G.E., Lancaster, S.T., Gasparini, N.M., Bras, R.L., Rybarczyk, S.M., 2001b.
An objectoriented framework for distributed hydrologic and geomorphic
modeling using triangulated irregular networks. Computer and Geosciences 27,
959973.
Tucker, G.E., Slingerland, R., 1997. Drainage basin responses to climate change.
393
US Geological Survey, 2008. Earthquake Hazard Program: Magnitude 7.9 Eastern

Sichuan, China. http://earthquake.usgs.gov/eqcenter/eqinthenews/2008/
us2008ryan/#summary (accessed 17.1.2011).
Van der Beek, J., Van Melle, J., Guillot, S., Pecher, A., Reiners, P.W., Nicolescu, S.,
Latif, M., 2009. Eocene Tibetan plateau remnants preserved in the northwest
Himalaya. Nature Geoscience 2, 364368.
belts. Nature Geoscience 2, 97104.
Whipple, K.X., Kirby, E., Brocklehurst, S.H., 1999. Geomorphic limits to climateinduced increases in topographic relief. Nature 401, 3943.
Whipple, K.X., Tucker, G.E., 1999. Dynamics of the stream-power river incision model:
implications for height limits of mountain ranges, landscape response timescales,
and research needs. Journal of Geophysical Research 104, 1766117674.
Whittington, A.G., 1996. Exhumation overrated. Tectonophysics 206, 215226.
Wilcock, P.R., Iverson, R.M., 2003. Prediction in geomorphology. Geophysical
Monograph. American Geophysical Union, Washington DC.135,
Willgoose, G.R., Bras, R.L., Rodriguez-Iturbe, I., 1989. A physically based channel
network and catchment evolution model. Technical Report 322, Ralph M.
Parsons Laboratory, Massachusetts Institute of Technology, Cambridge.
Willgoose, G.R., Bras, R.L., Rodriguez-Iturbe, I., 1991. A physically based coupled
network growth and hillslope evolution model: 1 Theory. Water Resources
Research 27, 16711684.
Williams, M.A.J., 2009. Cenozoic climates in deserts. In: Parsons, A.J., Abrahams, A.D.
(Eds.), Geomorphology of Desert Environments. Springer, Dordrecht, pp. 799824.
Zalasiewicz, J., Williams, M., Smith, A., et al., 2008. Are we now living in the
Anthropocene? GSA Today 18, 48.
Zeitler, P.K., Koons, P.O., Bishop, M.P., et al., 2001a. Crustal reworking at Nanga
Parbat, Pakistan: metamorphic consequences of thermal-mechanical coupling
facilitated by erosion. Tectonics 20, 712728.
Zeitler, P.K., Meltzer, A.S., Koons, P.O., et al., 2001b. Erosion, Himalayan
geodynamics and the geomorphology of metamorphism. Geological Society of
America Today 11, 48.
Biographical Sketch
Associate Prof. Ulrich Kamp received his PhD in geography with an emphasis on geomorphology from the
Technische Universitat Berlin, Germany, in 1999. In the following years, he had positions at Freie Universitat
Berlin, University of Nebraska at Omaha, and DePaul University in Chicago before joining the University of
Montana in 2005. Prof. Kamps research is on alpine geomorphology and glaciology, particularly Quaternary and
recent glacial change. He has carried out fieldwork in the mountains of South and Central Asia, and in the Rocky
Mountains and the Andes.
Prof. Lewis A. Owen received his PhD in geomorphology from the University of Leicester, U.K., in 1988. Before
joining the University of Cincinnati in 2004, he held positions at the Hong Kong Baptist University, Royal
Holloway University of London, and the University of California Riverside. Prof. Owens research focuses on
the Quaternary geology and geomorphology of tectonically active mountain belts and their forelands, particularly
in the HimalayanTibetan orogen and the Cordilleras of North and South America. He has also undertaken
research in other tectonically active regions, including the Red Sea margin in Yemen and the Atlas and Anti-Atlas
Mountains of Morocco. In 2011, Prof. Owen was awarded the Busk Medal from the Royal Geographic Society for
his field research in Quaternary history and geomorphology in tectonically active areas.
7.1 Mountain and Hillslope Geomorphology: An Introduction

M Stoffel, University of Berne, Berne, Switzerland, and University of Geneva, Carouge-Geneva, Switzerland
RA Marston, Kansas State University, Manhattan, KS, USA
Abstract
The downslope movement of rock and earth materials under the primary influence of gravity is considered in a series of
contemporary volumes by many who specialize in various aspects of such phenomena. As one of the main processes in
action at the top of watersheds worldwide, the initiation and maintenance of mass movement is one of the most pervasive
process influences in landscape evolution everywhere. The 35 chapters in this volume constitute a review of the latest
understandings about the actions of mass movement on slopes in mountains, and as such may offer a fairly succinct
summary from which future research on related phenomena may spring.
Mountains and hillslopes are important constituents of the

terrestrial landscape, where the direct influence of gravity almost continuously transports unconsolidated sediments
within, across, or out of the mountain and hillslope environments via the mediums of air, water, and ice. The geomorphic processes occurring on mountains and hillslopes will
leave a large variety of erosional features and depositional
landforms. The present-day landforms are rarely the products
of contemporary processes alone but a mirror of processes
that have acted on mountain and hillslope environments
over time.
Mountains and hillslopes can be seen as a process-response
system, where external inputs come from solar radiation,
precipitation, atmospheric substances, and from the weathering of bedrock. Outputs to mountain and hillslope environments are driven by evapotranspiration, percolation of
water, and by sediment transport by glaciers and streams.
Balance between inputs and outputs exerts control over
slope form.
Mountain and hillslope systems are commonly composed
of noncohesive, coarse-grained sediments that constitute
landform features such as screes, talus cones, glacial outwash
deposits, or alluvial fans. In the literature, different processes
have been described by which mountain and hillslope materials can be transported through these systems. At the
microscale, rainsplash can be effective in moving material on
slopes, whereas rainwash and surface runoff have been demonstrated to cause considerable transport of surface sediments
on slopes and the formation of rills.
Another category of processes on mountains and hillslopes
is generically referred to as mass movement, or mass wasting,
although this latter term is frowned on by many as a somewhat imprecise and rather old-fashioned term. Mass-movement processes occur when instability develops in the
mountain and hillslope systems. Some failures involve spontaneous rapid events, generally driven by short-term triggering
mechanisms, but mass movement can also occur over longer
Stoffel, M., Marston, R.A., 2013. Mountain and hillslope geomorphology:

an introduction. In: Shroder, J. (editor in chief), Marston, R.A., Stoffel, M.
(Eds.), Treatise on Geomorphology. Academic Press, San Diego, CA, vol. 7,
Mountain and Hillslope Geomorphology, pp. 13.
periods of time. Mass movement occurs when the stresses

applied to the slope materials exceed their internal strength. A
large variety of factors can act as triggers for hillslope failure,
such as prolonged or heavy rainfall (i.e., saturation of soil
materials), earthquakes, or glacier debuttressing.
In noncohesive materials, mass movement commonly
occurs through the toppling, sliding, or flowing of some
particles, which can eventually develop into larger massmovement processes. In cohesive materials (e.g., slopes
formed from clays and silts), rotational slips (slumps),
earthflows, and the more liquid mudflows are the most typical
processes operating over very short time periods. Among
the slow gravitative movements in cohesive materials are soil
creep, a process driven by the cyclical effects of temperature
fluctuations, moisture variations, animal activities, and solifluction or gelifluction, the slow movement of soil caused
by freeze-thaw action. In bedrock, mass movement tends
to be dramatic and quick; it occurs where weaknesses exist
along bedding planes and joints and where gravitational
stresses and freeze-thaw processes decrease stability. The
release of lesser or greater rock fragments into the air is called
rockfall, and the larger scale translational downslope movement without great turbulence of rock fragments over a surface
is referred to as rockslide.
Recent decades have witnessed an exponential growth of
scientific research on mountain and hillslope environments.
This volume of the Treatise on Geomorphology addresses
the themes outlined in the previous paragraphs and aims
at providing a state-of-the-art overview of contemporary geomorphology in hilly and mountainous geomorphic systems.
The mountains and hillslopes of any landscape show the
direct influence of gravity that almost continuously transports
unconsolidated sediments slowly or quickly, and within,
across, or out of mountain and hillslope environments via the
agents of associated ice, water, and even air in some cases.
Mountains alone cover 24% of Earths land surface. This volume on Mountain and Hillslope Geomorphology provides
a state-of-the-art overview of contemporary geomorphology
in hilly and mountainous systems. The volume contains
35 chapters contributed by an international array of geomorphologists who specialize in mountain and hillslope
processes, materials, and landforms. The text is richly illustrated with examples from around the worldy and beyond.
http://dx.doi.org/10.1016/B978-0-12-374739-6.00143-3
Mountain and Hillslope Geomorphology: An Introduction
The volume begins by examining the materials that comprise hillslopes. In addition to bedrock, mountain and hillslope surfaces are generally covered by noncohesive, coarsegrained sediments that can be mobilized by gravity. In this
volume, the character of hillslope rocks, regolith, and soil are
summarized by Taylor (Chapter 7.2). The stress, deformation,
and rheology of earth materials are further discussed by Major
as these factors affect susceptibility to weathering and erosion.
Comprehensive coverage is devoted to the processes and rates
of soil and regolith production on hillslopes by physical
weathering in Matsukuras volume (Chapter 7.4) and chemical weathering in that by Mudd et al. (Chapter 7.5)
Regolith and soils on mountains and hillslopes are transported by the surface processes of overland flow as discussed
by Buda (Chapter 7.7), and rills and gullies by Gao (Chapter
7.11). The generation of floods and flood waves on hillslopes
is reviewed by Bracken (Chapter 7.8), and the analysis of
flash flood runoff response is discussed by Penna et al.
(Chapter 7.9) largely with European examples. Conceptualization in catchment modeling is covered by Loague and Ebel
(Chapter 7.10).
Mass movement is covered in considerable detail, including major styles of slope failure: creep by Heimsath
(Chapter 7.13), slides by Burns (Chapter 7.14), flows by
Davies et al., (Chapter 7.15) topples by Hungr (Chapter 7.16),
rockfalls by Luckman (Chapter 7.17), and long run-out
landslides by Korup et al. (Chapter 7.18). Detailed coverage in
this volume is further devoted to the causes of mass movement: overloading by Van Den Eeckhart et al. (Chapter 7.19),
water by Dikau (Chapter 7.20), changes in slope angle by
Claessens et al. (Chapter 7.21), glacier shrinkage by Geertsema
and Chiarle (Chapter 7.22), earthquakes by Jibson (Chapter
7.23), and lateral spreading by Pasuto and Soldati (Chapter
7.25). What governs the style, activity state, and distribution of
mass movement is also reviewed by Guthrie (Chapter 7.24),
along with methods of evaluating hazards and risks by Crozier

(Chapter 7.26), and methods to avoid and mitigate mass
movement by Corominas (Chapter 7.27). The modeling of
mass movement is covered by Dorren et al. (Chapter 7.28).
The controls on overall rates of denudation are evaluated by
Binnie and Summerfield (Chapter 7.6), as are overall rates of
sediment yield by Vanacker (Chapter 7.12), especially as they
relate to land use.
The shape of hillslopes changes over time as the resisting
framework of lithology, stratigraphy, structure, and vegetation
interact with the driving forces of land-surface erosion and
mass movement. How hillslopes change over time is reviewed
by Roering and Hales (Chapter 7.29). How hillslope processes
vary by climate is covered by Soldati and Borgalti (Chapter
7.30). The differences in how hillslopes change over time as a
function of climate is discussed for cold environments by
Goodfellow and Boelhouwers (Chapter 7.31), temperate environments by Schlunegger et al. (Chapter 7.32), seminarid
environments by Cammeraat (Chapter 7.33), arid environments by Owen (Chapter 7.34), and in tropical environments
by Romer (Chapter 7.35). Finally, attention is given to hillslope form-process adjustments in extraterrestrial environments by Lanza and Newsome (Chapter 7.36).
In sum, this volume of the Treatise on Geomorphology is
offered to the readers as a series of papers by some of the chief
active scientists who study the phenomena of the downslope
movements of the rock, debris, and earth materials under the
force of gravity. Much more can and will be written about the
various mass-movement phenomena in future years but this
group of papers represents a fair cross section of the latest
ideas and observations from the field at this time. The literature on mass movement is vast, but this volume offers a small
portal into some of the most recent thinking and observation
about a pervasive process on hillslopes and in mountain
geomorphic systems.
Biographical Sketch
Dr. Markus Stoffel is an assistant professor of geomorphology and currently works at the Institute for Environmental Sciences (IES), University of Geneva, and at the Institute of Geological Sciences at the University of Berne
(Switzerland), where he directs the Laboratory of Dendrogeomorphology (www.dendrolab.ch). His research
interests are in hydrogeomorphic and earth-surface processes, climate change impacts, and dendrogeomorphology. He has authored more than 90 peer-reviewed papers on geomorphic, hydrologic, cryospheric, and geologic
processes in mountain and hillslope environments, with a focus on time-series of frequency and magnitude and
process dynamics, as well as dendroecology and wood anatomy of trees and shrubs, and integrated water resources management. He was a guest editor of Earth Surface Processes and Landforms, Geomorphology, Journal of
Hydrology, Natural Hazards and Earth System Sciences, Science of the Total Environment, and has coedited books
dedicated to Tree Rings and Natural Hazards (2010) and Tracking Torrential Processes on Fans and Cones (2012) with
Springer.
Markus holds BSc and MSc degrees in Physical Geography, an MSc degree in Media and Communication
Sciences and a PhD in Dendrogeomorphology from the University of Fribourg (Switzerland), as well as a habilitation thesis degree (venia docendi) from the University of Berne (Switzerland). In 2010, he received a distinguished professorship (Professor honoris causa) in Physical Geography of the University Babes- -Bolyai
(Romania).
Mountain and Hillslope Geomorphology: An Introduction
Richard Marston is university distinguished professor of Geography at Kansas State University. During 201112, he
was appointed as a Jefferson Science Fellow by the National Academies to serve at the US Department of State as a
Senior Science Advisor in the Office of the Geographer and Global Issues. Dick specializes in geomorphology
(fluvial, hillslope, and glacial), mountain geography, and water resource geography. He served as the 102nd
President of the Association of American Geographers during 200506. He has been Co-Editor-in-Chief of the
journal, Geomorphology, since 1999. The American Institute of Hydrology first certified Dick as Professional
Hydrologist #488 in 1984. He is a Fellow of the American Association for the Advancement of Science, American
Geographical Society, Geological Society of America, Explorers Club, and Royal Geographical Society. The
Association of American Geographers recognized him in 2003 with the Ronald F. Abler Distinguished Service
Honors. Dick has been awarded more than 40 research grants and contracts, and has presented more than 380
papers or posters of research results at professional gatherings. He is coeditor of three books, and contributor to
more than 50 refereed publications and 50 technical or consulting reports.
7.2 Regolith and Soils of Mountains and Slopes

G Taylor, University of Canberra, Bruce, ACT, Australia
7.2.1
7.2.2
7.2.2.1
7.2.2.2
7.2.2.3
7.2.3
References
Introduction
Mountain Types
Types of Slopes
Processes Acting on Mountain Tops and Slopes
Nature of Regolith on Mountains and Slopes
Summary
Glossary
Alluvial Related to sediments deposited by a stream of
water.
Biota Includes all living organisms (animal, plant, and
microbes).
Caldera A large basin-shaped depression, more or less
circular in shape.
Chrysoprase A gem of finely crystalline silica (SiO2) that
contains trace amounts of nickel that makes it green
like jade.
Cirque A semicircular steep-walled depression situated
high on a mountain side at the head of a glacial valley.
Duricrust A hard surface crust made by the accumulation
of soluble materials (Si, Fe, Ca) precipitated at or very near
the earths surface.
Etchplain An extensive erosion surface formed by
differential erosion where harder rocks remain as highs and
the weathered or more weathered rocks are removed.
Inselberg An isolated peak or knob rising abruptly from a
surrounding plain (residuals on etchplains are an example).
Karst A landscape formed, most commonly on limestone
but not exclusively, by solution and characterized by domes
and towers, and sinkholes where water moved
underground.
4
5
8
8
12
18
18
Labile Rocks composed of minerals that are chemically

unstable during weathering (see below).
Moraine An accumulation of ice-carried debris from a
glacier. The sediments range from very large to very fine
glacial flour.
Mudflow A flow of rock debris mixed with water that
moves under gravity and deposits material that is composed
of mud and boulders mixed together.
Ophilite Mafic igneous rocks believed to represent
deformed and uplifted oceanic crust.
Pyroclastic Pertaining to fragmental rocks formed from
volcanic explosion and deposition from air.
Saprolite A thoroughly decomposed rock that retains the
original rock fabric or looks like the original rock but totally
rotted.
Stratovolcano A volcano or volcanic cone built of layers
of lava and pyroclastic material.
Subducted The process whereby one of the earths
tectonic plates is pushed below the adjacent plate.
Weathering The destruction of rocks by physical
disintegration or chemical/biological decomposition at or
near the earths surface.
Abstract
Mountains are by definition high points in the landscape. Because of this they tend to be predominantly regions of erosion,
which results in very little or a thin regolith of debris too coarse to move or weathered and rocks. Some mountains,
although high and steep, are areas where materials accumulate; volcanoes and sand (dune) mountains are the two places
where this is common. Mountain slopes, on the other hand, are regions of both erosion and deposition. Upper slopes tend
to be erosional but with coarse-grained debris streams moving across them. Regolith on upper slopes tends to be coarse.
Finer materials are flushed across the upper slopes. The lower slopes are regions of accumulation of coarse- to fine-grained
regolith materials moved downslope. Most regolith materials on mountains and slopes are moved by water and mass
movement. Soils on mountains and upper slopes tend to be minimal, although pockets of thicker and more well-developed
regolith/soils may occur in surface irregularities. The lower slopes have soils that vary from lithosols on coarse debris to a
variety of well-developed earths, podzolic, and darker organic-rich soils on the lower slopes.
Taylor, G., 2013. Regolith and soils of mountains and slopes. In: Shroder, J.
(Editor in Chief), Marston, R.A., Stoffel, M. (Eds.), Treatise on
Geomorphology. Academic Press, San Diego, CA, vol. 7, Mountain and
Hillslope Geomorphology, pp. 419.
http://dx.doi.org/10.1016/B978-0-12-374739-6.00145-7
Regolith and Soils of Mountains and Slopes
7.2.1
Introduction
7.2.2
Mountains and their accompanying slopes are complex, formed

by vastly different processes and of very different materials
(Ollier and Pain, 2000; this volume). In terms of regolith and
soils, then, they are complex and vary according to the type of
mountain/slope and the materials that make up the mountains.
Regolith is the incoherent mass of varying thickness covering the underlying rock (Merrill, 1897). In places, this covering is made up of material originating through rock
weathering in situ. In other instances, it is of fragmental and
more or less decomposed material drifted by wind, water, or
ice from other sources. It is this entire mantle of unconsolidated material, whatever its nature or origin, that we propose
to call the regolith (Eggleton, 2001). It comprises everything
between fresh rock and fresh air. Soil is the unconsolidated
mineral matter on or near the surface of the earth that has
been subjected to and influenced by genetic and environmental factors, such as climate (including temperature and
moisture effects), macro- and microorganisms, and topography, all acting to produce a product soil that differs from
the material from which it is derived in many physical,
chemical, biological, and morphological properties and
characteristics (Soil Science Society of America, 1978). Soils
form the uppermost part of the regolith unless they are buried
by subsequent regolith-forming events and are preserved as
palaeosols (e.g., bauxites buried in a basalt pile).
A mountain is any part of the earths landscape isolated
from the surrounding landscape that is sufficiently high
(4100 ft, or 300 m, according to Jackson, 1997) above it to be
deserving of a distinctive name. This is obviously a relative
definition; so where landscapes are relatively low relief, a
mountain will be much lower than a mountain in a high-relief
region of the earth. A slope is an inclined surface of any part of
the earths surface (Jackson, 1997).
This chapter outlines various mountain types, the processes
at work on mountains that influence the character or regolith
and hence soils, and the nature of common regolith materials
on mountains and their slopes.
(a)
Mountain Types
Nearly all mountains are erosional, but some are fault controlled. Their geological origin has little to do with the nature
of the mountain; rather the post-uplift history is important
the exceptions being fault-bounded and volcanic mountains.
Erosional mountains are the result of increased erosional
energy caused by uplift. The mechanisms of uplift are varied
from simple epirogenic movements to uplift resulting from
continental collision; the result is the same: erosion ensues.
The simplest types of mountains are structurally or duricrusted
plateaus. Where the bedrocks are essentially flat-lying,
mountains or plateaus are formed by the lateral erosion of the
bounding slopes and cliffs around the plateau (Figure 1).
Similar mountains or plateaus are also formed often
over highly deformed bedrocks where a relatively elevated,
low-relief landscape is deeply incised by streams and where
the plateau margins are retreating by lateral scarp retreat
(Figure 2).
Some eroded mountains represent deeply incised and
dissected old landscapes. By old here I am referring to places
like the Kimberley Plateau of northwestern Australia, where
Ollier et al. (1988) recorded the region as being a terrestrial
landscape since Proterozoic glaciers swept the area. Mountains
also exist that are either in part or almost entirely formed from
solution (e.g., Figure 3). Perhaps the worlds most famous
examples are the mountains along the Lee River near Guilin,
China (Figure 4).
Chocolate-box mountains (alpine-style mountains) are
mountains where glaciers have sculpted the pre-existing
landscapes into a series of deep valleys (often U-shaped) with
divides of sharp-peaked ridges and horns. Such mountains
occur in regions formerly or presently glaciated during the last
(and possibly earlier) glacial periods (Figure 5). Overall, they
retain little evidence of previous landscapes. However, Ollier
and Pain (2000) reported Mount Snowden in Wales as an
example of one mountain that had its summit preserved as
part of an old surface. Erosional mountains such as these
generally occur at elevations above the last glacial permanent
(b)
Figure 1 (a) A mountainous plateau with cliffed margins in flat-lying Triassic sandstones incised by deep valleys and topped in the center
distance by Miocene basalts, Gross Valley, Blue Mountains west of Sydney. (b) A plateau capped by siliceous duricrusts (silcrete) in
northwestern New South Wales. The bulk of the slopes are deeply weathered, flat-lying, Cretaceous marine sandstones and mudstones.
(a)
(b)
Figure 2 (a) Another classical bedrock structurally controlled isolated plateau or mesa, Table Mountain, Cape Town, South Africa. (b) A major
mesa or isolated plateau in central Australia is controlled by its flat-lying sediments. Mount Connor (photo: Barry McLashen).
(a)
(b)
(c)
Figure 3 (a) Plateau with some relief cut across steeply dipping Palaeozoic rocks, gorge of the Chandler River near Armidale, New South
Wales. (b) Domed erosional mountains formed in part by stream incision of Devonian quartzose sandstones and in part by solution of the
sandstones. Bungle Bungle Ranges, northwestern Australia. (c) Another example of the Bungle Bungle Ranges.
Figure 4 Tower karst or solution-eroded mountains of southern

China, near Guilin.
snow line and thus tend to form the higher mountain ranges
of the world.
The majority of these glacially eroded mountains occur on
bedrocks that are highly deformed and contain a large variety
of lithologies. They tend to occur at major crustal collision
zones such as the Alps, Rockies, Andes, Himalayas, and in
Japan. They also tend to be zones of significant tectonic and
volcanic activity and are mostly being actively uplifted today.
Inselbergs are isolated, relatively fresh bedrock mountains
protruding from a plain. Some controversy exists about their
origin, but most researchers (e.g., Thomas, 1994; Ollier and
Pain, 2000) believe etchplanation applies to the situation
where deeply weathered bedrocks are eroded from above the
weathering front, leaving behind parts of the bedrock not as
affected by the deep weathering.
Figure 6 is a much simplified cartoon from Ollier and Pain
(2000) that illustrates the formation of inselbergs. Inselbergs
(a)
(b)
(c)
Figure 5 (a) Glacial mountains in southeastern Iceland, where, although the glaciers have retreated, the ridge and horn divides are seen behind
a prominent cirque with prominent fluvial erosion features and the lower slopes and valley bottoms. (b) Typical glacial erosional mountains with
a cirque lake below the last remnants of a glacier and typical sharp ridge and peak topography on the divide. Aerial view in southwestern South
Island, New Zealand. (c) Again in the high Andes, east of the Altiplano, south of Cusco, Peru.
are commonly formed in granite and gneiss, but may also

occur in sandstones; and indeed some remaining volcano
pipes, where the rest of the volcanic edifice has been eroded,
could be said to be inselbergs as well.
Some would argue that volcanic necks are not inselbergs
(e.g., Kesel, 1973) because they are not formed by the lowering
of the land surface around them. If volcanic necks or plugs are
not formed by the erosion of the remainder of the volcanic
edifice (the landscape is lowered), how then are they formed?
Many volcanic necks meet most criteria of inselbergs and thus,
in my opinion, can well be discussed with them.
Volcanic mountains occur mainly along mid-ocean ridges
(e.g., Iceland), along continental crust margins where oceanic
crust is being subducted (e.g., Java), and on crustal materials
that migrate over hot-spots (e.g., Hawaii and the central volcanoes of eastern Australia) (Figures 7 and 8).
Volcanoes build two major landforms, either of which may
form mountains, cones, or calderas. Cones are roughly conical
features built of lava or lava and ash (pyroclastic material or
ejecta) in varying combinations. Generally, the more lava in a
cone the lower the angle of slopes it has, while those with
more ash have steeper slopes. The lava volcanoes are known as
shield volcanoes (Figure 9), whereas those with ash and lava
that form steeper-sided cones are stratovolcanoes (Figure 9).

The island of Hawaii is the largest mountain and shield volcano in the world; it covers some 5000 km2 and rises 8742 m
above the ocean bottom. Either type may occur on land or in
the ocean, where they may remain below the surface as seamounts. Many submarine volcanoes also rise above sea level
to form volcanic islands, many of which are substantial
mountains. Hawaii is a shield volcano and is the largest
mountain in the world, rising some 10 200 m above the ocean
floor and 4100 m asl. Many of the high mountains in the
world are stratovolcanoes (e.g., Fuji, Vesuvius, Erebus,
Aucanquilcha in Chile, Popocatepetl in Mexico, Pinatubo in
Luzon, and Kilimanjaro).
Calderas are another feature of volcanic mountains. These
occur where the central part of a volcano collapses into the
volcano or is blown out during a massive eruption. The best
known example is the eruption of Krakatau in the Sunda
Straight between Sumatra and Java in 1883. Here, the whole
island was virtually blown away, leaving just a few small islands marking the former volcanic mountain. Mt. Pinatubo,
during its eruption in the mountain, erupted some 25 m3 of
ash and debris from its summit, lowering the mountain by
some 800 m and leaving a huge caldera (Figure 10).
Ground surface
Weathering
front
Figure 6 Shows the formation of inselbergs by deep weathering

followed by saprolite stripping so that the preformed or unweathered
bodies of bedrock are proud of the surrounding regolith. Reproduced
from Ollier, C.D., Pain, C.F., 2000. The Origin of Mountains.
Routledge, London and New York, 345 pp, with permission from
Taylor and Francis.
Sand mountains occur almost throughout the world. They

are simply overgrown sand dunes that commonly reach
heights of over 200 m. The largest is Cerro Blanco dune in
Nazca, Sechura Desert, Peru, which is 1176 m high.
Echoing Sand Mountain near Dunhuang, China, in the
Gobi Desert is 800 km2 in area, stretching 40 km from east to
west and 20 km from south to north. The main peak is B500 m
high. The Big Daddy dune at Sossusvlei in the Namib Desert is
383 m high (Figure 11), and the coastal dunes on Moreton
Island in Queensland are 280 m high at their highest point.
7.2.2.1
Types of Slopes
The types of slopes that occur around mountains depend, to a

large degree, on the nature of the material being shed from the
mountain and the processes that distribute it. Slopes are of
several types: free slopes or cliffs of fresh or slightly weathered
rock; upper slopes that are usually concave and contain the
coarsest-grained materials shed from above; and lower slopes
that are either flat or convex and contain the finer materials
shed from above.
7.2.2.2
Processes Acting on Mountain Tops and Slopes
Mountains are regions of high relative relief; thus the dominant process is gravity: materials by and large move downhill,
two exceptions being that glaciers can move material from
valley heads and deposit them on valley walls and that wind
can blow materials to higher elevations. Additionally, rain falls
on mountains and brings with it minor solutes that, in arid
regions especially, can deposit on mountains and plateaus.
Gravity moves materials down mountain slopes after they
are weathered and released for erosion from the upper slopes
or free-faces (Figure 12 top left and bottom left). A precursor
to this downward movement is that the materials must be
physically and/or chemically weathered. On many mountains
with free-faces, much of this release is the result of physical
weathering releasing blocks of material that is free to fall
under gravity. This is not to say such processes need steep
slopes to be effective; gravity does move debris from upslope
over very low gradients (Figure 12 bottom right). Water also
moves mainly under gravitational forces and thus is available
to move particulate and soluble components of weathering
downslope, providing the particulates are fine-grained enough
to be transported.
Other gravity-driven processes on mountains are rockfalls,
landslides, mudflows, and the like. The exact nature of a slide
depends to a large degree on the nature of the materials
available for movement. Rockfalls tend to occur from freefaces where cliffs are undercut, allowing large blocks or slabs
to fall. Rockfalls also frequently trigger other landslides as they
traverse the mountain slope (Figure 12).
Weathering is an important process for regolith formation
on mountains and slopes, but it is not my intention to discuss
weathering here except inasmuch as it impacts on other processes. Weathering is essential to building a supply of erodible
materials on mountain slopes, and once all the weathered
material (or regolith) is eroded, the accumulation of material
on the slopes will cease. This is particularly common on steep,
rocky slopes such as those found on chocolate-box mountains
glaciated during the last glacial episode. Such mountain slopes
are said to be weathering limited. In some cases, mountains
are formed of particularly weathering-resistant rocks (quartzite
or quartz sandstone) or on some plateaux capped by duricrusts (silcrete or bauxite), which also slows the retreat of the
escarpments at the plateau edge. The degree to which mountain tops and slopes weather also depends on lithology and
the length of time the rocks have been weathering. Older
landscapes tend to contain larger stores of weathered rock and
therefore are less likely to form weathering-limited slopes.
Complementary to the age factor is lithology: labile rocks
(e.g., basalt, graywacke, shale) weather deeply compared with
less labile rocks (e.g., sandstone, rhyolite). So in labile rocks, a
store of regolith and therefore movable materials can be significantly built up more quickly than on less labile rock.
The total volume of water and rainfall intensity is important in shaping mountains and slopes and in undertaking
weathering. The more water that moves through the rocks and
regolith, the more rock and regolith will be weathered. Most
weathering occurs within the rocks/regolith and not at the
surface, so it is the groundwater and infiltrating water that is
most important in weathering. Rainfall intensity is important
as heavy, intense rain events are more capable of moving
surface materials downslope, particularly as the regolith of
the slopes becomes saturated. For example, in Figure 12 (top
left), the lighter colored strip over the coarse-grained scree
slope is a fine-grained deposit formed by throughflow
Figure 7 Inselbergs of various types. (Top left) Granite inselbergs from Rio de Janiero, Brazil, where the surrounding weathered gneiss leaves
mountains of granite as inselbergs. (Top right) Uluru (Ayres Rock) in central Australia, where the old land surface has been eroded, leaving
sandstone and conglomerate inselbergs. (Bottom left) A volcanic pipe or plug left as the remainder of the volcanic pile that has been eroded;
Bora Bora, Tahiti. (Bottom right) Classical inselbergs at Sossusvlei, Namibia.
Eurasian Plate
Eurasian plate
North American Plate
Cascade
-range
San Andreas fault Mid-atlantic
Ridge
Aleutian Trench
Ring of fire
Hawaiian Hot Spot

Java Trenceh-
Arabian Plate
Cocos PlateEast Pacific

Rise
Nazca
Plate
Indo-Australian Plate
South
American
Plate
African Plate
Pacific Plate
Antarctic Plate
Topinka, USGS/CVO, 1997, Modified from: Tilling, Heliker, and Wright, 1987, and Hamilton, 1976
Figure 8 Map of the earths major plates showing directions of movement (in places), spreading zones (black lines), volcanoes (red dots), and
the ring of fire around the Pacific Ocean where some of the worlds major volcanoes are concentrated (reproduced from USGS website).
10
Figure 9 Left: A portion of a shield volcano on the southern Icelandic coastal fringe fronted by the jokalup/fluvial plains of Skeidararsandur.
The shield is largely covered by a glacier. Right: The stratovolcano of Mt. Pelee behind the town of St. Pierre in northern Martinique, which was
destroyed in 1902 by a glowing avalanche of ash and debris. The present cone grew in a caldera about 9000 years ago; the margin of the caldera is
at the ridge directly in St. Pierre (http://www.volcano.si.edu/world/volcano.cfmvnum=160012=&volpage=photos&photo=106068).
Figure 10 (Left) Anna Krakatau in the caldera of the pre-1883 Krakatau, Rakata Island. In the background is a remnant of the pre-1883 rim of
Krakatau. The dark lava on Anna Krakatau was erupted during 1975 (http://www.volcano.si.edu/world/volcano.cfmvnum=060200=). (Right) The Mt.
Pinatubo caldera after its 1991 eruption removed about 800 m from its summit (http://en.wikipedia.org/wiki/Mount_Pinatubo).
Figure 11 The Namib sand dune field showing the high red dunes
that occupy much of the Namib Desert (photo: Helen Taylor).
bringing fine sediment to the surface and depositing it as slope

angles allow.
On lower slopes, water carrying sediment from upslope
generally deposits into alluvial fans, which form the convex
lower slopes. These fans are internally complex and composed

of a mixture of sediment sizes from large cobbles to muds.
They frequently contain buried soils (palaeosols) formed at
times from overbank deposits that are distant from the contemporary locus of sedimentation.
Ice, or at least glacial activity, can be highly erosive and
transport vast masses of sediment from mountains, and in the
process form classical glacial landscape features, such as sharp
mountain ridges (Figure 5) with large U-shaped valleys with
floors generally filled by glacial moraine (Figure 13). Mostly,
the mountains and slopes are left clean of debris or weathered rock after the glacier retreats. Most glacial landscapes
were created during the Last Glacial Maximum (B20 000
years ago) and, hence, most such mountains and slopes have
accumulated minimal regolith, except at slope bases and in
the valley bottoms, where it is till and later fluvial deposits.
Many researchers (e.g., Butt and Zeegers, 1992; Thomas,
1994) believe that overall climate is a major controlling factor
of regolith formation, but I (Taylor and Eggleton, 2001) contend that water is far more important than temperature and
perhaps seasonality. However, in mountain environments
temperature may be important, especially as it relates to
physical weathering of exposed rock faces. The production of
Figure 12 (Top left) A basalt free-face above a flat, steeply inclined debris slope of boulders north of Reykjavik in western Iceland. (Top center) Small rockfall during 1990 in the moderately weathered
volcanic rocks of Castle Peak, Tuen Mun, Hong Kong. (Top right) Landslide triggered by a small rockfall. (Center) On Castle Peak, Tuen Mun, western Hong Kong, the longest landslide recorded in the
territory. (Bottom left) A coarse-grained scree fan showing steeper slopes of the coarsest material with progressively lower slopes as the finer-grained material accumulates on the lower fan; west of
Queenston, New Zealand. (Bottom right) Metasediment particles being moved under the influence of gravity over a very low-angle slope in Kakadu, Northern Territory, Australia.
11
12
Figure 13 (Left) Clean U-shaped glacial valley with sharp ridge mountains in the background, Mildord Sound, New Zealand. (Right) Glacier on
the right side of the photo heavy with a basalt debris load, and morainic debris deposited on the left side of the photo, southern Iceland.
Figure 14 (Left) Trees and shrubs growing on near-vertical slopes of Milford Sound, New Zealand. (Right) Bromeliads clinging to vertical rock
faces at Aguas Calientes, Peru.
fragmental debris in cold and arid environments is very much

dependent on thermal effects as well, of course, as on the
lithology and rock structure.
Vegetation in particular and biota in general may be significant factors in forming regolith on mountains and slopes.
Many free-faces while they do not accumulate much regolith
are vegetated, particularly in wet and temperate regions. The
vegetation adds weight to the rocks as well as forces rocks
apart, making them more susceptible to movement downslope. Figure 14 shows such slopes in the very wet Milford
Sound region of New Zealand, where vegetation established in
cracks and crevices in the rock faces slowly force parts of
them free.
Finally, wind is a significant process in building mountains
of sand as well as adding to regolith forming relatively high in
the landscape. Sand mountains are simply dunes blown up in
deserts (e.g., Gobi, Namib) or along the coastal fringe (e.g.,
Fraser Island, Australia; Namibian coast). Also, dust (Figure 15)
mainly as medium- to coarse-grained silt can be blown
from more arid to more vegetated regions, including mountains
where it deposits and forms a fraction within the mountain

and slope regolith (Chartres and Walker, 1988; Brimhall et al.,
1991).
Dust plumes may also be generated from volcanic eruptions (Figure 15) and add to regolith along the locus of their
dispersion (e.g., 2010 eruption of Mount Eyjafjallokull,
Iceland).
7.2.2.3
Nature of Regolith on Mountains and Slopes
Mountain tops are very varied in their nature; some are relatively pointed (alpine glaciated) and have little or no flat
or gently sloping ground. Such mountain tops rarely contain
much, if any, covering of regolith (Figure 5). They may,
however, have considerable depth of physically or chemically
weathered regolith (e.g., mild chemical weathering along
opened joints to several tens of meters). Mountain tops
in tropical to temperate regions, even though relatively
peaked, may have considerable depths of regolith. An extreme
13
Figure 15 (Left) Dust storm crossing the western plains of New South Wales toward the Eastern Highlands, where much of the dust will deposit
before crossing the Tasman Sea and also depositing in New Zealand (photo: S.M. Hill). (Right) Dust plumes crossing the eastern Mediterranean
Sea. Much of this dust will deposit across the mountains of Syria, Turkey, and surrounding regions (http://earthobservatory.nasa.gov/
NaturalHazards/view.phpid=35994). (Bottom) A volcanic dust plume from April 2010 shut down air traffic in much of Europe; but more significantly,
this dust contributed to deposits in the mountains and slopes of the continent (http://cliffmass.blogspot.com/2010/04/
icelands-volcanic-eruption-what-will-be.html).
Figure 16 (Left) Mountain of regolith (mainly smectite, iron oxyhydrides, oxides, and quartz) developed in an ophilite bedrock, Marlborough,
Queensland. (Right) Granite saprolite (B3 m thick) with granite core stones preserved near the top of Brown Mountain in southern New
South Wales.
example occurs at Marlborough in northern Queensland

on the Tropic of Capricorn, where ophilitic rocks are so
weathered that the mountain is virtually all regolith, consisting of ophilitic saprolite containing smectite, hematite,
goethite, and microcrystalline quartz (Figure 16 left). This

mountain is built up by a surface coating and internal scaffolding of chalcedonic quartz from which chrysoprase is
mined. In more temperate climates, regolith can still be
14
Landscape
regime
Erosional
Thinner
(stripped)
Depositional
Valley fill
Open
drainage
Transported
less weathered
Depositional
wet
Active
sed.
mment
Regolith
nature
Thick
(residual)
Drainage
Good to
poor
Good
Good to
moderate
Moderate
to poor
Free
Regolith
process
Residual
features
in situ
weathering
solute
removal
In situ weathering
physical and solute
removal
Physical accumulation
weathering and
solute removal
Partial
weathering
limited
solute
flushing
Physical
and solute
removal
Figure 17 Rolling landscape processes typical of the tops of rolling highlands (plateau tops).
deep very close to mountain tops; for example, Brown

Mountain in southern New South Wales has thick granitic
saprolite and core stones developed near the top of the
mountain. Although this mountain sits adjacent to the major
Great Escarpment of eastern Australia (Ollier and Pain, 1994),
this relatively old regolith remains in place despite relatively
steep slopes and high rainfalls. This profile mainly owes its
retention on the mountain to the high infiltration capacity of
the regolith.
On plateaux and tableland mountains, the situation with
regolith is different, as the tops of these landscapes have flat to
undulating features and even additional erosional mountains
atop them (Figures 1, 2 and 3 left). In Figure 1 (left), the
mountain atop the deeply incised landscape is a basaltic volcanic pile. Generally, regolith on these plateaus and tablelands
is like anywhere else: they consist generally of hills and valleys
of various heights that have regolith characteristics, as shown
in Figure 17.
Similar processes and landscapes occur at various scales on
larger plateaux where bare saprock or even fresh rock highs are
found with sediments accumulating in the topographic lows
between them. A good example is the top of Table Mountain
at Cape Town, South Africa (Figure 18).
Mesas or smaller plateaus generally are flatter topped than
plateaus or tableland mountain ranges. Their tops are commonly covered in saprolite (Figure 19 top left and bottom left);
sometimes overland stony layers are ferruginised remnants of
the underlying saprolite and plasmic zone (Figure 19 top right)
that resist weathering and remain as a lag at the surface.
This marginal formation of duricrusts (Figure 20) means
that landforms, despite marginal erosion by collapse, can be
retained over very long time periods in a landscape.
The slopes below a free-face generally are covered by a
regolith of scree consisting of the duricrusts forming the
free-face and weathered (mostly highly weathered) bedrock

components, typically clay and silt-sized materials eroded
from the weathering profiles beneath the duricrust. Soils on
mesas like those shown in Figures 19 and 20 are thickest in
the centers of the mesas and vary depending on the bedrock.
In the case of siliceous duricrusts, soils generally are deep,
sandy, and poorly consolidated, often containing large
amounts of aeolian debris (mainly coarse, silt-sized materials). In the case of mafic bedrock, soils generally are red
earths and quite deep (up to several meters). Soils on the
slopes tend to range from skeletal to deep minimal clay- to
pebble-rich earths. Again, aeolian components make up a
significant portion of these soils, particularly in arid regions;
but wind-blown particles are also widely recorded from temperate regions (Brimhall et al., 1988; Chartres et al., 1988).
Karstic (solutional) mountains generally have little if any
regolith except in small pockets where the regolith consists of
remnants of the solution weathering. Slopes around such
mountains are similarly rare and, where they exist, their
regolith is a scree of collapsed blocks from the karstic
mountains although over time this debris dissolves, leaving
the mountains rising abruptly from the surrounding alluvial
plains (e.g., Figure 3 center and right, and Figure 4). In the
case of siliceous karst, the sandstones forming the towers have
a thin regolithic skin of SiO2 cement within the rock face
covering much more weathered rock. Soils are rare on karst
mountains, although very small pockets on solution cavities,
benches, and opened joint surfaces do contain organic soils at
lower elevations where vegetation can colonize and build up
accumulations (e.g., Figure 4). Some of these soils become
more extensively developed as vegetation becomes established
in small pockets and as rock masses extend over adjacent bare
rock surfaces enabling more vegetation to establish and the
area of soil cover to grow. Such mats of thin organic soils,
15
Figure 18 (Left) Quartz sandstone hillocks interspersed with sandy sediments accumulating in a microdepression that, in most cases, support a
shrubby vegetation atop Table Mountain, South Africa. (Right) The margin of the Mitchell Plateau of northwestern Australia, showing a 23 m
free-face of bauxite and a flat to undulating surface covered by bauxite pisoliths and aeolian quartz sand.
Figure 19 (Top left) A ferricrete-topped mesa underlain by a deep weathering profile north of Kalgoorlie, Western Australia. (Top right) A mesa
in the Simpson Desert, South Australia, covered by gibber (close-packed stony layer). (Bottom) Mesas capped by silcrete: left is from near Mt.
Isa, Queensland, and right from north of Broken hill, New South Wales. These cappings serve to protect the plateau (mesa) from erosion except
around their margins, where erosion proceeds by undercutting and collapse.
16
however, tend to be unstable and they collapse off the steep

slopes from time to time.
Alpine-style mountains commonly consist of greater or
lesser areas of less exposed, fresh to near fresh rock and
regolith formed from physical weathering in debris fans
(Figure 5) and/or collapsed soils and other larger debris in
regions where vegetation is prolific even on the steep slope
either below the snow line or in regions of high precipitation
(Figure 21). The soils tend to be skeletal and organic and
consist of coarse-grained debris from upslope and accumulated organic residues from the vegetation.
Most truly alpine mountain regions of the world are sufficiently high to have been glaciated during the last Quaternary glacial episode, and much of the terrain reflects this
activity, as discussed above (Figures 5 and 13). Features such
as U-shaped valleys, hanging valleys, aretes, cirques, etc. all
attest to former glacial erosion in these mountains, even in the
tropics (e.g., in New Guinea and West Papua).
Inselbergs commonly have a regolith, thicker or thinner,
and usually composed not of loose materials but of weathered
bedrock. Rarely saprolite, but commonly saprock, makes up
the surface layers of regolith although along joints and
Fresh rock/saprolite
Duricrust
Regolith no duricrust
Water movement
Figure 20 The marginal development of duricrusts (often silcrete,

ferricrete, or calcrete) on mesas.
fractures weathering may extend to some depth. A soil component of the regolith is mostly absent, but very small pockets
of sandy and stony soils may occur in small depressions.
Vegetation may colonize these small pockets and depressions
as well as become established in joint fractures across inselbergs, even though soils are rare. Cosmogenic isotope data
from Cockburn et al. (1999) showed that inselbergs in the
central Namib had lower mean rates of B5.0771.1 m M 1
years, a much higher rate than those reported from South
Australia. The worlds largest inselberg is Uluru in central
Australia (Figure 7), but it is vegetation-free and has an indurated regolith surface, allowing water to be shed rapidly
and, consequently, weathering to be minimized.
Volcanic mountains have a complex association of regolith
and soils. They are of different ages, and so have varying thicknesses of regolith, given other factors are equal. Within one volcanic edifice, the materials at the surface are likely to be of varying
ages as different volcanic events cover different areas of the
mountain. They occur in very dry climates (e.g., Atacama Desert
in Chile, and Antarctica), while many occur in the wet tropics
(e.g., Nicaragua and Java), so the volumes of water available for
weathering of the volcanic materials vary enormously.
Active volcanoes are usually a collection of new and
therefore fresh lavas, ash, and other volcanic debris covering
older and more weathered volcanic materials. In some volcanoes, only one product is erupted, for example, basaltic lava
and ash from Hawaiian shield volcanoes; but the age of the
basaltic debris varies from place to place so that the regolith is
a complex arrangement of deeply weathered to fresh debris.
Regolith depth under similar hydrological conditions will vary
with age, and soil depth and structure will be greater on older
basaltic materials. Where ashes occur, weathering will be faster
than on basalt so although the nature of the regolith will not
Figure 21 (Left) The dense forest vegetation on the lower slopes of the Southern Alps west of Christchurch, New Zealand, where the upper
peaks are rocky and covered by bedrock outcrops interspersed with debris slopes (also see Figure 5); but the lower slopes below the snow line
are densely forested and covered by a thicker regolith held in place by vegetation. In the foreground, landslips show where this relatively thin
regolith has moved downslope and was carried off in the high gradient braided streams. (Right) Well-vegetated hillslopes opposite Machu
Picchu, Peru. The slopes are covered with shrubs and small trees, but the vertical slopes are colonised by bromeliads that root on the rock
surface, anchoring themselves in fractures and small irregularities of the cliff.
Regolith
landform
unit
Cross-section and
description of
landform and regolith
17
Age
(based on
radiometric
dating)
Eceles
Stony rise
00.2 Ma
Rouse
Degraded
stony rise
0.21 Ma
Dunkeld
Plain with
gilgai
13 Ma
Clay
Deeply
weathered
and incised
plain
34 Ma
Hamilton
Deeply
weathered
45 Ma
and deeply
incised plain
Figure 22 (Left) Bauxite sandwiched between a fresh basalt flow of the Palaeogene Monaro shield volcano above and basalt saprolite below on
the Monaro, southern New South Wales. These bauxites are relatively common between flows, as are alluvial and lacustrine sediments, but few
are latterly continuous and reflect localized weathering and other regolith processes away from the locus of contemporary volcanic activity over
the volcano. (Right) Summary of regolith landform units from the basalt plains of the Newer Volcanic Province of southeastern Australia. Each
profile is B20 m vertically. Reproduced from Joyce, E.B., 1999. A new regolith landform map of the western Victorian volcanic plains, Victoria,
Australia. In: Taylor, G., Pain, C.F. (Eds.), New Approaches to an Old Continent. CRC LEME, Perth, WA, Australia.
vary greatly its thickness generally will be greater on ashes.

Although much older (Palaeogene), the shield volcanoes of
the Monaro in southern New South Wales (Figure 22 left),
show in the geological record (Taylor et al., 1990) the same
variability as we see on Hawaii today.
Stratovolcanoes are similar in that volcanic materials accumulate in different parts of the volcanic edifice at different
times and, consequently, regolith varies from fresh materials
to deeply weathered regolith over very short distances. Complex patterns of regolith and soil development mantle the
slopes. This can only be adequately demonstrated using
regolith landform maps (e.g., Pain et al., 1991) that show the
relationships between surface regolith and landscape elements. Few detailed regolith maps of volcanic terrains are
available, but Joyce (1999) presented one of the basaltic volcanoes of western Victoria, Australia (Figure 22 right). This
author is not aware of any regolith landform maps of active
volcanoes that present the complexity and nature of regolith
on volcanic mountains.
Soils on volcanoes vary equally complexly as regolith. Soil
types vary from lithosols on fresh lavas and other volcanic
debris to deep, usually red-earth soils in higher landscapes and
yellow on lower slopes. Alfisols are the ultimate expression of
the regolith/soil process, where most of the volcanic debris is
converted to aluminous and/or ferruginous soils and/or pans
(e.g., Figure 22). In humid regions, vegetation will colonize
relatively fresh volcanic debris rapidly and contribute organics
Figure 23 Partially eroded dunes with an unconformity evident as the

minimal hardpan in the center of the photograph. This surface is about
30 ka, littered with human artefacts, and represents a pause in dune
building at that time; from Innaminka in northern South Australia.
to the soils and water, leading to faster (thus deeper) soils and
regolith development. In arid regions, such is not the case, and
soils tend to be thin and skeletal (e.g., some regions in central
Iceland or in the Atacama).
Sand mountains have a relatively simple regolith; in fact,
they can be considered as being entirely composed of regolith
(aeolian sand and minor silt and clay). Their regolith varies
18
little; textures are sandy, but often the surface and near surface
is bound by algae (e.g., Lange et al., 1992; Eldridge and Leys,
2003). Dune slopes are similar to crests. One feature that does
vary in sand mountains is their color. Most commonly they
are red, but coastal dunes commonly are white to cream, becoming redder as one moves more inland. Some sand
mountains and dune systems contain an internal stratigraphy
that reflects their evolution, boundaries being marked by
slightly more induration caused by algae and accumulations
of fine-grained materials leached through the profile from
overlying sands (Figure 23). Soils on sand mountains are all
sandy but vary considerably depending on the amount of
vegetation cover and available water. Most commonly, they are
simple sandy skeletal soils with slight reddening or yellowing
in the B horizon and perhaps minimal organic accumulations
near the surface. In other cases, particularly in dunes near
coastal regions, podzols are common.
7.2.3
Summary
All mountains for the greater or lesser part of their life are,
almost by definition, erosional. The fact that erosion occurs at
least on their upper slopes and summits means that regolith in
these parts is thin and mostly composed of the materials
making up the mountain often coarse-grained and with
minimal soil. Plateaux and mesas are exceptions in that they
have low-relief summits on which new regolith can accumulate,
mainly from water movement of aeolian accession; but eventually this regolith will be eroded. Volcanic mountains accumulate new primary material from time to time, but rarely over
their whole surface at the same time, causing the regolith to be
highly varied in nature from place to place on the mountain.
Sand mountains have the potential to grow almost continuously, but changes in environmental conditions mean that they
grow and move in fits and starts, often eroding as much as is
deposited. Thus their regolith is thin and sandy.
The one part of mountains where erosion is not predominant is the lower slopes where debris moved from higher
up accumulates as scree and debris slopes or alluvial fans.
These areas have relatively complex regolith and soils because
much of the material is depositional and as the locus of
deposition changes over time already formed regolith is
eroded, new regolith is deposited, and a complex pattern of
regolith materials and soils therein evolves.
Butt, C.R.M., Zeegers, H., 1992. Climate, geomorphological environment and

geochemical dispersion models. In: Butt, C.R.M., Zeegers, H. (Eds.), Regolith
Exploration Geochemistry in Tropical and Subtropical Terrains. Elsevier,
Amsterdam, The Netherlands, pp. 324.
Chartres, C.J., Chivas, A.R., Walker, P.H., 1988. The effect of aeolian accession on
soil development on granitic rocks in south-eastern Australia. II: oxygen-isotope,
mineralogical and geochemical evidence for aeolian deposition. Australian
Chartres, C.J., Walker, P.H., 1988. The effect of aeolian accessions on soil
development on granitic rocks in south-eastern Australia. III: micromorphological
and geochemical evidence of weathering and soil development. Australian
Cockburn, H.A.P., Seidl, M.A., Summerfield, M.A., 1999. Quantifying denudation
rates on inselbergs in the central Namib Desert using in situ-produced
cosmogenic 10Be and 26Al. Geology 27, 399402.
Eggleton, R.A., 2001. The Regolith Glossary. CRC LEME, Perth, WA, Australia,
144 pp.
Eldridge, D.J., Leys, J.F., 2003. Exploring some relationships between biological
soil crusts, soil aggregation and wind erosion. Journal of Arid Environments 53,
457466.
Jackson, J.A. (Ed.), 1997. Glossary of Geology. American Geological Institute,
Alexandria, VA, 769 pp.
Joyce, E.B., 1999. A new regolith landform map of the western Victorian volcanic
plains, Victoria, Australia. In: Taylor, G., Pain, C.F. (Eds.), New Approaches to an
Old Continent. CRC LEME, Perth, WA, Australia.
Kesel, R.H., 1973. Inselberg landform elements: definition and synthesis. Reviue
Geomorphologie Dynamique 22, 97108.
Lange, O.L., Kidron, G.J., Budel, B., Meyer, A., Kilian, E., Abeliovich, A., 1992.
Taxonomic composition and photosynthetic characteristics of the biological soil
crusts covering sand dunes in the western Negev Desert. Functional Ecology 6,
519527.
Merrill, G.P., 1897. A Treatise on Rocks, Rock Weathering and Soils. Macmillan,
New York, 411 pp.
Ollier, C.D., Chan, R.A., Craig, M.A., Gibson, D.L., 1988. Aspects of landscape
history and regolith in the Kalgoorlie region, Western Australia. Bureau of
Mineral Resources Journal of Australian Geology and Geophysics 10,
309321.
Ollier, C.D., Pain, C.F., 1994. Landscape evolution and tectonics in southeastern
Australia. Australian Geological Survey Organisation Journal of Geology and
Geophysics 15, 335345.
Ollier, C.D., Pain, C.F., 2000. The Origin of Mountains. Routledge, London and New
York, 345 pp.
Pain, C., Chan, R., Craig, M., Hazell, M., Kamprad, J., Wilford, J., 1991. RTMAP
BMR Regolith Database Field Handbook. Minerals and Landuse Program, Series
1. Bureau of Mineral Resources, Geology and Geophysics, Record 1991/29,
City, ST/Country.
Soil Science Society of America, 1978. Glossary of Soil Science Terms. Soil
Science Society of America, Madison, WI, 36 pp.
and Sons, Chichester, UK, 375 pp.
Taylor, G., Truswell, E.M., McQueen, K.G., Brown, M.C., 1990. The early Tertiary
palaeogeography, landscape evolution and palaeoclimates of the Southern
Monaro, New South Wales, Australia. Palaeogeography, Palaeoclimatology,
Thomas, M.F., 1994. Geomorphology in the Tropics; A Study of Weathering and
Denudation in Low Latitudes. John Wiley and Sons, New York, 460 pp.
References
Brimhall, G.H., Lewis, C.J., Ague, J.J., Dietrich, W.E., Hampel, J., Teague, T.,
Rix, P., 1988. Metal enrichment in bauxites by deposition of chemically mature
aeolian dust. Nature 333, 818824.
Brimhall, G.H., Lewis, C.J., Ford, C., Bratt, J., Taylor, G., Warin, O., 1991. Quantitative
geochemical approach to pedogenesis: importance of parent material reduction,
volumetric expansion, and aeolian influx in lateritization. Geoderma 51, 5191.
Relevant websites
http://en.wikipedia.org/wiki/Dune#South_America
19
Biographical Sketch
Graham Taylor graduated with a B.Sc. (Hons) and M.Sc. from the University of New South Wales in 1971, worked
in the minerals industry, then graduated with a Ph.D. from the Australian National University in 1977. He has
taught at the Australian National University, the University of Canberra, and the University of Hong Kong. His
research began in oceanography, moving to fluvial sedimentology, and finally he developed a teaching and
research program in regolith geology. Graham has over 140 refereed publications including a major book on
regolith geology with Tony Eggleton. He retired as a Professor Emeritus from the University of Canberra in 2002.
7.3 Stress, Deformation, Conservation, and Rheology: A Survey of

Key Concepts in Continuum Mechanics
JJ Major, US Geological Survey, Vancouver, WA, USA
7.3.1
7.3.2
7.3.3
7.3.4
7.3.4.1
7.3.4.2
7.3.4.3
7.3.4.4
7.3.5
7.3.5.1
7.3.5.2
7.3.5.3
7.3.5.4
7.3.6
7.3.7
7.3.7.1
7.3.7.2
7.3.8
7.3.8.1
7.3.8.1.1
7.3.8.1.2
7.3.8.1.3
7.3.8.2
7.3.8.3
7.3.9
7.3.9.1
7.3.9.2
7.3.9.3
7.3.9.4
7.3.10
References
Introduction
Continuum
Force
Stress
Total Stress
Stresses Acting on Arbitrarily Oriented Surfaces
Pore Water, Hydraulic Head, and Pore-Water Pressure
Effective Stress
Deformation
Normal Strain
Shear Strain
Rotation
Strains in an Arbitrarily Oriented Coordinate System
Rate of Deformation
Conservation
Conservation of Mass
Conservation of Linear Momentum
Constitutive Relations
Linearly Elastic Material
Relationships between stress and normal strain
Relationships between shear stress and shear strain
Relationship between pressure and dilatation
Linearly Viscous Fluid
Plasticity the Coulomb Failure Rule
Example Application
Conservation of Momentum and Stress Equilibrium
Effective Stress and Effective Stress Equilibrium
Effective Stress and Elastic Strain
Displacement Formulation of Constitutive Relations and Groundwater Flow
Concluding Remarks
Glossary
Conservation laws The concept in physics that any system
evolves in such a way as to maintain equilibrium in the
balance of a measurable physical quantity. In continuum
mechanics, conservation equations typically state that the
rate of change of some quantity, such as mass, momentum,
or energy, within a definable volume is balanced by the
amount of the quantity produced (or consumed) within the
volume and the rate at which the quantity is transported
across the volume boundaries.
Constitutive equations Mathematical expressions of the
behavior of idealized materials that describe the relation
21
22
22
22
23
24
25
25
28
28
30
30
31
31
32
33
34
35
36
36
37
38
38
40
40
40
40
41
41
42
42
between physical quantities, such as stresses and strains or

stresses and strain rates.
Continuum mechanics The branch of mechanics that
deals with the analysis of an idealized matter that
completely fills the space it occupies and has properties
that, when averaged over appropriate spatial scales, vary
continuously across that space.
Newton A unit of measure of force, defined as kilogram
meters per squared second (kg m s2).
Pascal A unit of measure of stress, defined as newtons per
square meter (N m2).
Major, J.J., 2013. Stress, deformation, conservation, and rheology a survey

of key concepts in continuum mechanics. In: Shroder, J. (Editor in Chief),
Marston, R.A., Stoffel, M. (Eds.), Treatise on Geomorphology. Academic
Press, San Diego, CA, vol. 7, Mountain and Hillslope Geomorphology,
pp. 2043.
20
http://dx.doi.org/10.1016/B978-0-12-374739-6.00146-9
Stress, Deformation, Conservation, and Rheology: A Survey of Key Concepts in Continuum Mechanics
Rheology The science of the deformation and flow of

matter.
Scalar A physical quantity defined solely by its
magnitude.
Strain The distortion of a body, such as a change in length
or a change in angle between perpendicular elements,
caused by application of a stress.
21
Stress Surface force per unit area exerted on a plane

having a particular orientation within or on a body of
material.
Tensor A physical quantity defined by its magnitude,
direction of operation, and position in space.
Vector A physical quantity defined by its magnitude and
direction of operation.
Abstract
This chapter provides a brief survey of key concepts in continuum mechanics. It focuses on the fundamental physical
concepts that underlie derivations of the mathematical formulations of stress, strain, hydraulic head, pore-fluid pressure,
and conservation equations. It then shows how stresses are linked to strain and rates of distortion through some special
cases of idealized material behaviors. The goal is to equip the reader with a physical understanding of key mathematical
formulations that anchor continuum mechanics in order to better understand theoretical studies published in
geomorphology.
7.3.1
Introduction
Quantitative understanding of the physical processes that

drive landscape evolution from tectonic uplift to erosional
processes that gnaw the land is firmly rooted in the concepts
of stress, deformation, rates of deformation, conservation, and
rheology, which lie at the heart of continuum mechanics. The
idea that a basic understanding of continuum mechanics is
important for understanding physical processes in geoscience
is shared by many and has been advocated for at least several
decades (e.g., Johnson, 1970). Explicit applications of physics
to geomorphic problems date back to at least the late 1800s,
when GK Gilbert used basic concepts of force balance to explain intrusions in the Henry Mountains, Utah (Gilbert,
1877). However, routine treatment of these topics in dedicated
courses taught in geoscience or environmental science departments at the undergraduate, and even the graduate, level
remains elusive. A few textbooks have focused on fundamental
concepts of continuum mechanics with applications to geological problems (e.g., Johnson, 1970; Means, 1976; Middleton and Wilcock, 1994; Turcotte and Schubert, 2002;
Cuffey and Paterson, 2010), and recently a textbook has
emerged that is focused on the mechanics of landscapes
(Anderson and Anderson, 2010).
In this chapter, a summary of basic concepts of continuum
mechanics is presented, hopefully with enough transparency
that the reader can grasp the fundamental, underlying
physics. The purpose of writing this chapter is to distill the
concepts presented more broadly in textbooks into a concise
package. Consequently, our treatment of these concepts is necessarily brief. A comprehensive treatment of continuum
mechanics, but a challenging read, is provided by Malvern
(1969). Thorough and approachable discussions targeted at
students of Earth science are provided by Means (1976),
Middleton and Wilcock (1994), Turcotte and Schubert (2002),
and Anderson and Anderson (2010). Bird et al. (1960) provided a lucid discussion of fundamental concepts related to
fluid mechanics, and Bejan (1984) presented an excellent
treatment of the application of scaling analysis to problems in

fluid mechanics.
Discussion here is focused on key concepts in continuum
mechanics that a reader needs to grasp in order to understand
and evaluate theoretical papers in the geomorphic, and more
broadly the geologic, literature. After presenting concepts of
force, stress, deformation, strain, and rate of deformation, the
chapter discusses conservation laws of mass and momentum
and then discusses the properties of a few special cases of
idealized materials and how they respond to applied stresses.
Constitutive equations, expressions that relate stresses to strains
and rates of distortion, provide the linkage to the rheological
behavior of materials (or how they deform and flow) under
applied loads. This may be the first exposure to this material
for some readers, so to make the concepts discussed here
more concrete, the chapter concludes with an example application from the literature that relies on many of the concepts
presented.
In a survey of this type, certain topics are treated superficially or omitted altogether. Unlike more comprehensive
treatments in textbooks, discussions of material strength and
methods to measure strength, stress, and strain are omitted, as
are discussions of stress and strain ellipsoids, the concept of
Mohrs circle, and complex idealized materials. The interested
reader can find excellent discussions of these topics in Selby
(1982), Middleton and Wilcock (1994), and Turcotte and
Schubert (2002). Detailed Earth science applications of the
concepts discussed here occur, for example, in Johnson
(1970), Middleton and Wilcock (1994), Turcotte and Schubert
(2002) and Cuffey and Paterson (2010), and applications focused on geomorphology are presented in Anderson and
Anderson (2010).
The discussion presented here is mathematical by necessity.
However, the basic mathematical concepts discussed are
not extraordinarily difficult. Much of the mathematical development of stress and strain is geometrical in nature and
relatively easy to conceptualize in one or two dimensions.
A few slightly more sophisticated mathematical concepts are
22
used in some derivations, such as limits and gradients, but

they, too, are easily grasped. Although most derivations of
concepts are presented in one or two dimensions, we live in a
three-dimensional world and stresses, strains, and fluxes are
three-dimensional in nature. In some places, it is, therefore,
convenient to extend the mathematical presentation from one
or two dimensions to three dimensions through the use of
vector and tensor notation. The chapter generally avoids the
compact, but cryptic, notation associated with vector and
tensor operations used by many authors of research papers,
except in a few explicit instances. When a more compact notation is convenient to the presentation, it is linked explicitly
to an example of a fully developed component of an equation.
By following this strategy, we hope to demystify some of
the mathematical description of the concepts discussed, with
the goals of providing the reader a direct linkage between the
conceptual mathematics and the underlying physical principles, and an entrance to the theoretical literature published
with regard to geomorphic processes. Thorough treatments of
vector and tensor notation and mathematical concepts such as
divergence, gradient, curl, dot products, and dyadic products
can be found, for example, in Malvern (1969), Schey (1973),
Means (1976), and Middleton and Wilcock (1994). In this
chapter, scalar quantities (those defined solely by their magnitude) are represented by a single letter (e.g., m), vectors
(quantities having magnitude and direction) are represented
by an underscore (e.g., U), and tensors (quantities defined by
magnitude, direction, and position) by an overlying tilde (e.g.,
s~ ) or with indicial notation (e.g., sij).
7.3.2
Continuum
When applying concepts of mechanics to problems in Earth

sciences, we are interested in forces applied to, and deformation and flow of, idealized continuous bodies (a continuum).
But what constitutes a continuum; is such an idealization
appropriate or adequate? A continuum is an idealized region
of space filled with matter having properties that, when averaged over appropriate spatial scales, vary continuously across
that space. Under this simplifying concept, we disregard molecular structure of material. At the macroscopic scale, we ignore discontinuities (e.g., discontinuities between mineral
grains or fractures in rock bodies) and assume that material
can be adequately characterized by averaged properties.
Clearly, such idealized matter does not exist, as discontinuities
are present at virtually all scales. In some instances, discontinuities may place bounds on the region that can be
considered a continuum. However, if a scale appropriate for
the problem is selected, discontinuities at a smaller scale can
be tolerated, and average properties of the matter can be
assumed to vary smoothly across the scale of interest.
of mass and velocity, Newtons second law can be written as

F
1
where F represents the forces acting on the body, d/dt the total
derivative that represents rate of change with time, m the mass
of the body, and v its velocity. For constant mass, this expression becomes
F m
dv
dt
2
and because dv/dt is the definition of acceleration, we get the

familiar expression
F ma
3
From the above expression, we see that force has a unit of kg

m s2, which is called a newton (N).
In mechanics, two classes of forces can be defined: body
forces and surface forces. Body forces act equally on every
element of mass within a body and are proportional to its
mass or volume (Middleton and Wilcock, 1994; Turcotte and
Schubert, 2002). An example of a body force is the force of
gravity; the weight of an element is the product of its mass and
the acceleration of gravity, g. Surface, or contact, forces, on the
other hand, act on the bounding surface of a body. Unlike a
body force, the influence of a surface force is proportional to
the size of the area over which it acts; furthermore, it acts in a
specific direction and in a specific position (e.g., Middleton
and Wilcock, 1994; Turcotte and Schubert, 2002). An example
of surface forces acting on a body can be illustrated by envisioning a person pushing a box across a table. As the box is
pushed, there is not only a force acting perpendicular to the
face being pushed, but also a component of that force acting
tangentially along the surface of the box in contact with the
table. As long as the magnitude of the tangential force exceeds
the force of friction resisting sliding the box will slide. We can
also envision that a box having a small footprint might be
easier to push than one having a large footprint, because the
magnitude of the tangential force transmitted to the base of
the small box is focused across a smaller area and can more
easily overcome the frictional force resisting sliding. Because
the influence of a surface force is proportional to the size of
the area over which it acts, an inherent geometric effect influences the changes in a body caused by that force. Therefore,
having a common way of quantifying the effect of a surface
force regardless of the size of the area over which it acts is
useful. The best way to remove the geometric effect of a surface
force acting on a body is to normalize the force by the area
over which it acts, which leads to the concept of stress.
7.3.4
7.3.3
dmv
dt
Stress
Force
Newtons second law states that the time rate of change of

momentum of a body is proportional to the sum of the forces
acting upon that body (e.g., Johnson, 1970; Middleton and
Wilcock, 1994). Because momentum is defined as the product
Stress, by definition, is the surface force per unit area exerted

on a body of material and is given in units of newtons per
square meter (N m2), or pascals (Pa). Stress is a very useful
concept for understanding the impact a surface force has
on a body. Intuitively, the more broadly a surface force is
distributed, the lower the stress. For example, this is the

underlying physics behind snow shoes. A person trying to
walk across a soft, snow-covered surface commonly punches
holes through, and sinks below, the surface. With snow shoes,
however, that persons weight (the imposed surface force) is
distributed over a broader area of snow surface, and he or she
is able to stay atop the surface because of the lower stress
imparted. The examples thus far illustrate surface forces acting
over a finite area. The broader concept of stress with respect to
continuum mechanics, however, is related to the stress acting
at a point or the force acting on an area in the limit as the size
of the area diminishes to zero (e.g., Malvern, 1969; Middleton
and Wilcock, 1994).
7.3.4.1
Total Stress
Stress is conveniently resolved into two components: a normal

stress, which acts perpendicular to a surface, and a shear stress,
which acts tangential to a surface. Figure 1 shows the normal
and shear stresses acting on an elemental control volume
defined in accordance with the Cartesian x, y, z coordinate
system. Each stress is identified using a pair of subscripts: the
first subscript refers to the direction perpendicular to the
surface on which the stress is acting and the second subscript
refers to the direction of the stress. Thus, normal stresses have
identically paired subscripts, whereas shear stresses have unequal subscripts. Three stresses are defined on each surface
perpendicular to a coordinate direction: a normal stress acting
perpendicular to the surface and two orthogonal shear stresses
acting along the surface. Hence, nine different stresses act on
the three-dimensional volume, and assuming there is no acceleration, stresses equal in magnitude but acting in opposing
directions are imposed on the other three faces. The forces
associated with the normal stresses act to stretch or compress
the elemental volume, whereas the forces associated with the
y
s11
Bs
sij @ 21
s31
yx
yz
xy
xx
zy
shear stresses attempt to distort and rotate the elemental volume about each axis. For an element in equilibrium (not accelerating), the moments about each axis must balance.
Therefore, the shear stresses are symmetric; for example,
sxy syx. Whereas nine stresses are defined in three dimensions, only six of those stresses are independent.
To properly account for the direction in which the stresses
act, a sign convention for positive and negative stresses must
be defined. Here, a common convention that outward-directed
normal stresses are positive and inward-directed normal
stresses are negative is adopted. Thus, normal stresses are defined to be positive in tension and negative in compression.
Although common, this sign convention is not universal. For
example, in soil mechanics it is common to consider compressive stresses positive, because compression is the most
common state of soils dealt with by geotechnical engineers
(e.g., Lambe and Whitman, 1969). Despite the seeming convenience of aligning the positive sign convention with the
most common state of stress in Earth, there are both mathematical and physical reasons that trump this convenience.
Mathematically, the outward-normal direction on the Cartesian elemental volume has a positive sense of direction on a
positive face and a negative sense of direction on a negative
face (see Figure 1). The normal stresses acting in those positive
senses, thus, tend to pull the element in opposite directions
leading to a state of tensile stress. The adopted sign convention
also is consistent with physical changes that occur during
normal strain, in which elongation associated with tension (a
positive value because the ending state is longer than the
starting state) is defined as positive strain. In classical thermodynamics, expansion of a gas is defined as doing positive
work on its surrounding environment. Shear stresses are defined as positive if they act in a positive direction on a positive
face, and in a negative direction on a negative face (Figure 1).
Stresses acting at a point can be represented in mathematical
form as sij. This mathematical notation, in which i and j represent the coordinate axes as numbers (1, 2, 3) or letters (x, y,
z), defines a stress matrix, also known as a stress tensor:
0
yy
xz
zx
zz
z
Figure 1 Definition sketch of stresses acting on the surfaces of an
elemental volume defined in accordance with a Cartesian coordinate
system. Normal stresses, acting perpendicular to surfaces, are
defined positive in tension, and shear stresses are positive as shown.
See text for subscript nomenclature.
23
s12
s22
s32
1 0
sxx
s13
C
B
s23 A @ syx
szx
s33
sxy
syy
szy
1
sxz
syz C
A
szz
4
In this matrix, the terms along the diagonal represent the

normal stresses and the off-diagonal terms represent the shear
stresses. Because pairs of shear stresses must be equal for an
element in equilibrium, sij sji.
We can use the stress matrix to define the mean normal
stress acting on the element as
s

1
sxx syy szz
3
5
The mean normal stress can, furthermore, be equated with a

mechanical mean pressure acting on the element. Because
pressure is typically defined as positive in compression, we can
define the mechanical mean pressure as
p

1
sxx syy szz
3
6
24
This expression tells us that a positive mean normal stress

(tension) is equivalent to a negative mechanical mean pressure, whereas a negative mean normal stress (compression) is
equivalent to a positive mechanical mean pressure.
In some analyses, it is useful to separate the mean normal
stress (pressure) acting on a medium from the overall stress by
subtracting the mean normal stress from the total stress.
Separating stresses in this manner allows us to isolate explicitly those stresses that deviate from the mean stress. For an
incompressible material, the stresses that deviate from the
mean normal stress are those that cause deformation. Hence,
the deviatoric stress is defined as the difference between the
total stress and the mean normal stress (e.g., Engelder, 1994).
The deviatoric stress matrix can be written as
0
sxx p
sxy
syx
szx
syy p
szy
B
sD
ij @
0
sxx p
B
@ syx
szx
sxz
C
syz
A
szz p
1
sxy
sxz
syy p
syz C
A
szy
szz p
7
From this matrix, we see that only the normal stresses are
affected; shear stresses are unaffected by variations in mean
normal stress (Middleton and Wilcock, 1994; Turcotte and
Schubert, 2002). Clearly, another way to write the total stress
tensor is sij pdij sD
ij , where dij is the identity matrix,
known as the Kronecker delta, which is equal to 1 when i is
equal to j, and zero otherwise.
7.3.4.2
Stresses Acting on Arbitrarily Oriented Surfaces
Although defining stresses coincident with defined coordinate

directions is convenient, an ability to define stresses along
surfaces oriented arbitrarily with respect to an original coordinate system is necessary to make full use of the concept of
stress. Resolving stresses in a new coordinate system that is
rotated with respect to an original coordinate system involves
trigonometric manipulations of force balances, which involve
directional sines and cosines of the angle of rotation. Detailed
discussions of these manipulations are provided in Malvern
(1969), Means (1976), Middleton and Wilcock (1994), and
Turcotte and Schubert (2002). For stresses defined as positive
in tension and acting within the xy-plane associated with a
Cartesian coordinate system, the results are
sx0 x0 sxx cos2 y syy sin2 y sxy sin 2y
8
sy0 y0 sxx sin2 y syy cos2 y sxy sin 2y
9

1
sxx syy sin 2y sxy cos 2y
2
10
sx0 y0
In these expressions, the prime notation refers to stresses

acting on surfaces that are oriented arbitrarily, and the unprimed notation refers to stresses acting on surfaces oriented
with respect to the originally defined coordinate system. For
y 0, sx0 x0 sxx ; sy0 y0 syy ; and sx0 y0 sxy , which is required if
the coordinate systems are coincident. If we now suppose that

a condition exists such that there are surfaces on which the
shear stresses disappear, we can find the orientations of those
surfaces by setting sx0 y0 equal to zero in eqn [10] and solving
for y. That condition yields
tan 2y
2sxy
sxx syy
11
The direction y defined by eqn [11] is known as an axis of

principal stress. Because y satisfies eqn [11], y 901 (y p/2)
must also satisfy eqn [11] (because tan p 0). Hence, two
orthogonal surfaces exist (in two dimensions) for which the
shear stress is equal to zero. The normal stresses acting on
those two planes are known as the principal stresses; for one
plane, the normal stress achieves a maximum value known as
the maximum principal stress (commonly signified as s1), and
for the other it achieves a minimum value known as the
minimum principal stress (signified as s2 when considering
stresses in two dimensions, or s3 when considering stresses in
three dimensions). For the sign convention adopted, the
maximum principal stress is the most positive stress. Although this terminology is somewhat contradictory with most
Earth science convention (in which the greatest compressive
stress, here the most negative normal stress, is considered the
maximum principal stress), this is merely a matter of sign
convention and has no bearing on the underlying physical
concept.
Convenient expressions to have are those that relate normal and shear stresses for any arbitrary coordinate system in
terms of the principal stresses and the angle of the new coordinate system with respect to the principal stress axes. For
stresses in two dimensions having an xy coordinate system
oriented at an angle y to the principal stress axes, the stresses
sxx, syy, and sxy are related to the principal stresses and principal stress axes by (e.g., Middleton and Wilcock, 1994;
Turcotte and Schubert, 2002)
sxx
s1 s2 s1 s2
cos 2y
2
2
12
syy
s1 s2 s1 s2

cos 2y
2
2
13
1
sxy s1 s2 sin 2y
2
14
where y is the angle between the direction of s1 and the direction of sxx.
To find the surfaces on which shear stress sx0 y0 is a maximum value, differentiate eqn [10] with respect to y and set the
result equal to zero. Doing so gives
tan 2y
syy sxx
2sxy
15
Comparing eqn [15] with eqn [11] shows that the angle 2y
for the principal axis orientation and 2y for the maximum
shear stress orientation are negative reciprocals. This means
that the axes that maximize shear stress lie at 451 to the axes of
principal stresses (Turcotte and Schubert, 2002). Setting y

p/4 in eqn [14] yields
sxy max
s1 s2
2
16
Tensor notation expressions for three-dimensional stresses

acting on planes of arbitrary direction relative to some defined
coordinate direction are discussed in detail in Means (1976)
and Middleton and Wilcock (1994).
Thus far, discussion of stresses has been implicitly restricted to homogenous, nonporous media. In porous media,
stresses causing deformation can be influenced by the pressure
of the fluid that fills the pores. That influence necessitates
discussion of pore fluid, pore-fluid pressure, and the concept
of effective stress. Because the most common fluid in porous
material at Earths surface is water, the discussion below is
restricted to pore water and pore-water pressure.
7.3.4.3
Pore Water, Hydraulic Head, and Pore-Water

Pressure
Regolith (the mantle of fragmented material that overlies

bedrock) and highly fractured rock at Earths surface here
termed soil contain voids (pores) that are variously wetted
or filled with water (pore water). Forces acting on pore water
establish gradients of fluid potential, the work required to
move a unit quantity of fluid from a datum to a specified
position, and pore water flows in response to these gradients.
The concept of hydraulic head, a measure of the energy in a
fluid-filled porous medium, usefully describes pore-water potential. Total hydraulic head, or potential per unit weight of
fluid, can be defined in terms of two fundamental forms of
energy: potential energy, defined in terms of gravitational and
pressure potential energy, and kinetic energy, the energy associated with fluid motion (Hubbert, 1940; Ingebritsen and
Sanford, 1998). In a typical soil subject to Darcian (seepage)
flow, the flow velocity is usually very low and the kinetic energy is negligibly small compared to the total potential energy.
Thus, for an incompressible fluid (fluid having a constant
density; rw for water) the total hydraulic head (h) in a waterfilled soil is given by (Hubbert, 1940; Ingebritsen and Sanford,
1998)
hc
p
rw g
17
where c is the gravitational, or elevation, potential, and p/rwg

the pressure potential, in which p is the gauge pressure of the
pore water relative to atmospheric pressure and g gravitational
acceleration in the coordinate direction. Pore-water pressure,
therefore, constitutes one of the two dominant components of
the fluid potential in soils.
Pore-water pressure is isotropic, meaning that it is has the
same magnitude in all directions, but it varies with position
relative to the water table within a soil (the depth horizon
where pore-water pressure is atmospheric, which defines the
zero-pressure datum) and with the proportion of soil weight
carried by contacts among the soil grains (intergranular
contacts). Below the water table, pore-water pressure is
greater than atmospheric and positive; above the water table,
25
pore-water pressure is less than atmospheric and negative

owing to tensional capillary forces exerted on pore water (e.g.,
Remson and Randolph, 1962). If soil is saturated and water
statically fills pore space, then the pore-water pressure is
hydrostatic and varies with depth below the surface as a
function of the overlying weight of water. Pore-water pressure
can exceed or fall short of hydrostatic under hydrodynamic
conditions or if a soil compacts or dilates under load (e.g.,
Iverson et al., 2000; Major, 2000). Below the water table, soil
compaction will cause a transient increase in pore-water
pressure, the duration and magnitude of which are governed
mainly by the rate of compaction and the permeability of the
soil. An increase in pore-water pressure can lead to a loss of
soil strength. If compaction thoroughly disrupts intergranular
contacts, then the pore fluid may bear the entire weight of the
solid grains, and the soil will liquefy. In unsaturated soil, such
as occurs above the water table or (occasionally) when a saturated soil dilates (expands) under load, water does not fill
pore space completely, and pore-water pressure locally is less
than atmospheric. In that case, capillary and electrostatic forces cause water to adhere to solid particles. As soil water
content decreases, tensional force increases and negative porewater pressure (also known as suction) bonds solid particles,
increasing the soil strength. The magnitude of negative
pore-water pressure depends on soil texture and physical
properties, as well as on water content. Fine-grained soils have
a broader pore-size distribution and larger particle-surface area
than do coarse soils. As a result, fine-grained soils have a
greater range of negative-pressure potential because they can
hold more water than coarse soils and because the water
bonds more tightly to particle surfaces. Piezometers are used
to measure positive pore-water pressure; tensiometers commonly are used to measure negative pore-water pressure (e.g.,
Reeve, 1986).
7.3.4.4
Effective Stress
The behavior of porous media having fluid-filled pores depends not only on the total state of stress to which the material is subjected, but also on the pressure of the pore fluid.
The state of stress that causes solid-body deformation is the
stress that acts on the skeleton of solid material that makes up
the porous medium; however, that stress is modulated by the
pressure of the pore fluid. Therefore, when dealing with porous media the total state of stress is commonly partitioned
into components that describe the fluid pressure and the stress
acting on the solid skeleton. Such partitioning of stress leads
to the concept of effective stress, a concept partly recognized
by Charles Lyell as early as the late 1800s (Skempton, 1960),
but not explicitly articulated until Terzaghi (1923, 1943)
proposed a simple theoretical framework for soil consolidation. The concept of effective stress is elegantly simple and is
defined as the difference between total stress and pore-fluid
pressure. Effective stress should not be confused with deviatoric stress, which is the difference between the total stress and
mechanical mean pressure exerted on a homogenous, nonporous medium.
The mathematical formulation most useful for describing
effective stresses in soils and other compressible porous media
26
y
y
y=H
II
l
ta
To
pr
osta
Hydr
t=0
in
cip
al
re
st
tic flu
ss
al
ot
(T
essu
id pr
flu
id
pr
e)
r
su
es
re
ph
*
pt
y=0
pto
(a)
y = H Ut
Ut
U t
al
osta
t
To
Hydr
t = t
pr
al
cip
in
tic flu
re
essu
ss
re
st
id pr
III
ph
e
pt
y=0
II
yy
(b)
y
Original position of surface
U
static
ss
sure
pres
re
st
fluid
ph
t=
p =0
*
al
cip
rin
lp
ta
To
o
Hydr
y=HU
II
I
y=0
(c)
e
yy
is given by (e.g., Lambe and Whitman, 1969; Iverson and Reid,

1992; Middleton and Wilcock, 1994)
s0ij sij pdij
18
where s0ij are the effective stresses acting on the solid skeleton,
sij the total stresses acting on the porous medium, and p the
pore-fluid pressure. Note that the effective stress in this formulation appears to be an additive function of total stress and
pore-fluid pressure, but recall that normal stresses are defined
negative in compression whereas pore-fluid pressure is defined
positive in compression.
Partitioning the total stress in terms of effective stress and
fluid pressure illuminates crucial physical insights. Consider a
saturated porous medium in which water statically fills the
pores. If that saturated medium is now submerged beneath a
water surface, both the total stress exerted on the medium and
the pore-water pressure increase by an equal amount. As a
result, the effective stress remains unchanged. Hence, simply
increasing the fluid pressure does not cause a volume change
of the medium. Now consider a container of laterally confined
saturated porous material, lets say saturated sediment. If a
vertical load is added instantaneously to the sediment surface
across an impermeable barrier that prevents pore-water
drainage, the total stress within the sedimentary body increases. In response to that stress change, the sediment grains
attempt to pack closer together. However, because the pore
water cannot escape and because we shall assume that both
the water and the sediment grains are incompressible, particle
rearrangement cannot occur. As a result, the intergranular
stresses acting on the sediment grains cannot change, the
sedimentary body cannot compact, and the water pressure
increases by an amount equal to the change in total stress.
Again, we find that simply increasing pore-water pressure does
not cause volume change. Now consider the case in which the
vertical load is applied across a drainage panel atop the sediment body, which allows pore water to drain. Because the
water pressure within the sedimentary body has increased
above hydrostatic in this example, pore water flows toward the
drainage panel at the deposit surface (where the water pressure
is zero) in response to the change in gradient of the hydraulic
head caused by the change in pore-water pressure. As porewater seepage progresses, the pressure in excess of hydrostatic
is gradually diminished and transferred to the stress acting on
the sediment grains and the deposit compacts. Compaction,
or volume change, therefore, occurred in response to changes
27
in the intergranular, or effective, stress. Thus, in porous media,

the measureable effects from changes in stress, such as volume
change, distortion, or changes in shearing resistance, are due
exclusively to changes in effective stress (Terzaghi, as quoted in
Skempton, 1960).
We can solidify the thought experiment in more concrete
terms by examining fluid pressure and total stress within a
shallow, one-dimensional, water-saturated sediment deposit
in which the vertical coordinate direction, y, is defined positive
upward (Figure 2). If water statically fills the pores within the
sediment body, then the hydrostatic fluid pressure, ph, of a
column of water extending from the surface to a depth H y is
ph rw gH y
19
where rw is the density of water, g the gravitational acceleration in the coordinate direction, and H the coordinate value
identifying the body surface (e.g., Major, 2000). The total
stress acting on the sediment body, extending from the surface
to the same depth, is
syy rt gH y
20
where rt is the total mass density of the water-saturated sediment. (The negative sign follows the convention that total
stress is defined as negative in compression, whereas porefluid pressure is defined as positive in compression.) The total
mass density for the body can be written in terms of water
density, rw, grain density, rs, and porosity, n (assumed here to
be uniform throughout the depth of the shallow body), as
rt rw n rs 1 n
21
Substitution of eqn [21] into eqn [20], and some algebraic

manipulation, leads to
syy rw rs rw 1 ngH y
22
This expression shows that the total stress at depth in a column of uniformly porous, water-saturated sediment depends
on the weight of the overlying water plus the buoyant weight
of the column of overlying solids (e.g., Major, 2000).
Suppose now the saturated sediment is loaded rapidly but
with no change of stress at the deposit surface caused (e.g., by
rapid deposition of a uniform thickness of similar saturated
sediment). As a result, the pore-water pressure changes because the water that fills the pores is incompressible and it
Figure 2 Definition sketch of fluid-pressure and stress fields for one-dimensional, gravity-driven consolidation of a saturated sediment slurry
overlying an impermeable bed. ph is the hydrostatic fluid pressure; p the fluid pressure in excess of hydrostatic; pt the total fluid pressure; pt0 the
initial value of the total fluid pressure; syy the total normal stress; se the effective intergranular stress; and U the settlement of the deposit surface.
(a) Initial state of the saturated slurry. In this case, the total fluid pressure is initially equal to the unit weight of the slurry. A probe measuring fluid
pressure at any depth would measure values along line II. (b) Consolidation state at time Dt. Settling of buoyant grains generates excess fluid
pressure that dissipates through Darcian seepage (laminar pore-fluid flow indicated by squiggles). A probe measuring fluid pressure at any depth
would measure values along line III. As excess fluid pressure dissipates, the load is transferred to the sediment grains, which causes changes in
grain packing, and line III shifts to the left. The magnitude of excess fluid pressure is defined by the width of the region between the profiles of
hydrostatic pressure (I) and total fluid pressure (III). The total stress remains constant, and the effective intergranular stress represents the region
between the profiles of total fluid pressure (III) and total stress (II). As excess fluid pressure decreases, effective stress increases. (c) Static state.
At infinite time, the excess pore fluid pressure has dissipated. A probe measuring fluid pressure at any depth would measure hydrostatic values
along line I. The effective stress in this static state is defined by the region between the hydrostatic fluid pressure (I) and the total stress (II). Note
that the deposit surface settled by an amount U in response to changes in grain packing. Reproduced from Major, J.J., 2000. Gravity-driven
consolidation of granular slurries implications for debris-flow deposition and deposit characteristics. Journal of Sedimentary Research 70, 6483.
28
resists particle rearrangement. That resistance leads to a temporary increase in fluid pressure. The total pore-water pressure
can then be written as pt ph p, where ph represents the
hydrostatic portion of the pressure and p represents the water
pressure that is in excess of hydrostatic. Under rapid loading,
water does not drain instantaneously from the pores; instead,
it temporarily bears the weight of the new load. If the water
bears the entire total stress imposed on the system, the sediment is said to be liquefied. Setting the total water pressure
equal to the total stress in eqn [22] and recasting the
expression leads to
pt rw gH y rs rw 1 ngH y
Deformation
Body forces and stresses cause both solids and fluids to deform.
In general, solids will distort only a certain amount upon application of a stress or will not distort at all until the applied
stress exceeds a certain value. Fluid on the other hand here
restricted to liquid typically will distort immediately and
continuously upon application of a stress; some special classes
of liquids, however, will distort continuously only after an applied shear stress exceeds a certain value. These differences in the
effects that forces and stresses produce in solids and liquids lead
to the concepts of deformation, strain, and rate of deformation.
Solid-body deformation can be described in terms of a
displacement field, an array of displacement vectors that
connect the positions of the elements of a body before and
after deformation (Means, 1976; Middleton and Wilcock,
1994; Turcotte and Schubert, 2002). Under this description,
deformation includes rigid-body translation, rigid-body rotation, and distortion (strain). For example, if a displacement
vector translates or rotates but does not change length, then it
undergoes a rigid-body deformation but not a strain. Strain
requires a distortion a change in length of a displacement
vector or a change in angle between perpendicular elements of
the displacement vectors. Changes in length are related to
elongation or contraction of a displacement vector and are
referred to as normal strain. Angular changes between perpendicular elements of a displacement vector are referred to as
shear strain. Description of deformation in this manner differs
from its description in the engineering literature. Bird et al.
a
Ua
x
Ub
x
x
x + x
x + x
x
(Ux)a
23
The first term on the right-hand side of this expression is the

hydrostatic pressure and the second term is the nonequilibrium, or excess, water pressure. This expression shows
that the excess water pressure is equal to the buoyant unit
weight of the sediment (Figure 2(a)). Owing to the head
gradient that is established because of the nonequilibrium
water pressure, the pore water will flow down gradient, from
high head to low head. As it does, the excess water pressure
will diffuse and the effective stress acting on the solid skeleton
will increase (Figure 2(b)). At infinite time, all of the excess
water pressure will have dissipated, the sediment deposit will
have consolidated, and the effective stress will equal the difference between the total stress acting on the system and the
hydrostatic water pressure an equilibrium state in which no
further volume change can occur.
7.3.5
(Ux)b
Figure 3 Definition sketch of displacement and elongation of a

linear segment strained in the x-coordinate direction. For infinitesimal
strain, dx 0 Edx. Linear strain is defined positive in elongation.
(1960) and Malvern (1969), for example, treated deformation

as synonymous with strain or distortion and all rigid-body
motion as rigid-body displacement, not deformation.
7.3.5.1
Normal Strain
We begin our consideration of normal strain by examining the

x-direction displacement of a line having endpoints a and b
and length dx (Figure 3). The x-direction coordinates of a and
b are x and x dx, respectively. We now suppose the line is
displaced to a new position having endpoints a0 and b0 and
length dx0 . The respective x-direction coordinates of the endpoints of the displaced line are x0 and x0 dx0 (Figure 3).
Displacement of point a in the x-direction is given by
(Ux)a x0 x. Displacement of point b in the x-direction is
given by (Ux)b (x0 dx0 ) (x dx). If (Ux)a is equal to (Ux)b,
then the line has been translated rigidly in the x-direction but
not strained (i.e., dx0 dx). If the x-direction displacements of
points a and b are unequal, then the line has been strained in
the coordinate direction.
By using Taylors theorem (Zwillinger, 2003), we can represent the x-direction displacement of point b as
Ux b Ux a

q Ux
q 2 Ux dx2
dx
qx
q x2
2
higher order terms
24
This theorem states that the displacement of point b can be

represented as the displacement of point a plus the product of
the gradient of displacement in the x-direction and the original line length plus an expansion series of higher-order
terms involving products of derivatives of displacement and
powers of original line length. Substituting this expression and
the definition of (Ux)a into the expression defining (Ux)b
allows us to write an expression for the x-direction displacement of point b as
x0 dx0 x dx x0 x

q Ux
q 2 Ux dx2
dx
qx
q x2
2
higher order terms
25
29
which can be recast as

0

dx dx
q Ux q 2 Ux dx
higher order terms
dx
qx
q x2 2
26
By definition, the normal component of strain is related to a

change in line length as
e
Dl
l

dx0 dx q Ux q 2 Ux dx
higher order terms 28
exx
dx
qx
q x2 2
A simple characterization of the state of strain is obtained for
infinitesimal strains, those that are so small that only linear
relationships need be considered. For such small strains, we
can neglect all higher-order terms of the expansion series. In
practice, infinitesimal strain theory gives good results for
strains as large as a few tenths of a percent to perhaps B1%
(Malvern, 1969; Middleton and Wilcock, 1994). Consideration of infinitesimal strain is useful even in a body undergoing large strain, because an infinitesimal strain over an
infinitesimal increment of time can be related to an instantaneous stress and provide useful insight (Middleton and
Wilcock, 1994). Neglecting higher-order terms and taking the
limit as dx goes to zero, eqn [28] becomes
q Ux
qx
29
Infinitesimal normal strain, defined as positive in elongation,

is, thus, related to the gradient of displacement along a coordinate direction. In a similar manner,
q Uy
eyy
qy
q Uz
ezz
qz
30
With regard to strain identification, the axes of strain (and,

thus, the planes of strain) are indicated by the subscripts.
Repeated subscripts indicate normal strain along a coordinate
axis; unequal subscripts indicate shear strain within the plane
defined by the subscript axes.
Let us now consider a small elemental area having sides of
length dx and dy, respectively. Let each of those sides be
elongated by an amount dUx and dUy, but remain perpendicular (Figure 4). We now define the fractional change of
the area of the element as
DA
xA
A
Uy = (*Uy /*y)y

a
27
where Dl is the change of line length and l the initial, undeformed line length. Strain is, therefore, a dimensionless
quantity. We can relate the definition of normal strain to the
one-dimensional change of line length in the x-direction by
combining eqns [26] and [27] as
exx
x + Ux
31
y + Uy
y
x
x
UX = (*UX /*x)x
Figure 4 Definition sketch of dilation (areal expansion) of a twodimensional element. The dashed line shows the original dimensions
of the element. Dilatation (volume strain) is defined positive in dilation.
After Middleton, G.V., Wilcock, P.R., 1994. Mechanics in the Earth and
Environmental Sciences. Cambridge Press, Cambridge, 459 pp.
where A is the original area of the element. If we again use

Taylors theorem to represent the displacements of the endpoints of the respective vector components of the elemental
area and assume infinitesimal strain in order to neglect higher
order terms of the series expansion, we can define the new area
of the deformed element as
A1 dx dUx dy dUy

q Uy
q Ux
dx dy
dy
A1 dx
qx
qy

q
U
q Ux
q Ux q Uy
y
A1 A 1
qx
qy
qx qy
32
Substituting the last expression of eqn [32] into eqn [31] and
taking the limit as dx and dy go to zero yields
DA q Ux q Uy q Ux q Uy
A
qx
qy
qx qy
33
For infinitesimal strain, q Ux =q x and q Uy =q y are much less

than 1 and, thus, the product term is very small and can be
neglected. In two dimensions, the areal strain is thus given by
xA
q Ux q Uy
exx eyy
qx
qy
34
For large strains, product and higher-order terms may not be

small, and at least some of them must be retained and
evaluated. Extending this infinitesimal strain analysis to three
dimensions shows that the volume strain, or dilatation, of an
element is given by
x exx eyy ezz
35
Because normal strains are defined as positive in elongation,

dilation is positive. Compression, the opposite of dilation, is
negative.
30
7.3.5.2
Shear Strain
Shear strain is measured as a change in angle between lines

that were originally perpendicular. Consider an elemental area
that undergoes a distortion that produces angular changes, but
which leaves the sides of the area approximately the same
length (Figure 5). The angular change along the x-axis, a1, is
defined as positive in an anticlockwise direction because it
produces a displacement along the positive y-axis. Likewise,
the angular change along the y-axis, a2, is defined as positive in
a clockwise direction because it produces a displacement
along the positive x-axis. This produces a condition known as
pure shear, provided that a1 a2. From geometric considerations, we can define
tana1
tan a2
dUy
dx
36
dUx
dy
For very small angles, those that are small compared to one
radian, tan aEa, and therefore
a1 E
dUy
dx
a2 E
dUx
dy
which upon substitution and taking the limit as dx and dy go

to zero yields

1 q Uy q Ux
1 q Ux q Uy
eyx
39
exy
2 qx
2 qy
qy
qx
We see from this definition that exy eyx; hence, the components of shear strain are symmetric.
Similar derivations for exz and eyz lead to

1 q Uz q Ux
exz
2 qx
qz
40

1 q Uz q Uy
eyz
2 qy
qz
The total shear strain, or engineering component of the shear
strain, is the sum of the angular changes a1 and a2, and is twice
exy. Most geological analyses of mechanics of materials utilize
the definition of average shear strain.
We can now combine the definitions for normal strain and
average shear strain into a strain tensor as
0
37
We now define the average angular change from the original

right angle of the elemental area, or average shear strain, as
exy 12a1 a2
q Ux
B
B qx

B
B 1 q Ux q Uy
eij B
B2 qy
qx
B

B1 qU
q Uz
x
@
2 qz
qx

1
1 q Uy q Ux
1 q Uz q Ux
2 qx
2 qx
qy
qz C

C
C
q Uy
1 q Uz q Uy C
C
2 qy
qy
qz C
C

C
1 q Uy q Uz
q Uz
A
2 qz
qy
qz
41
38
7.3.5.3
Rotation
Angular deformation leads to the condition of pure shear if a1

is equal to a2. Angular deformation of an element can also
lead to solid-body rotation, which can occur without distortion or with distortion if a1 is unequal to a2. To examine
displacement relations in terms solely of rotation, consider an
elemental area that undergoes an anticlockwise rotation
(Figure 6), which we shall define as a positive rotation. In this
Ux
2
y
Ux
x
1
Uy
2
Figure 5 Definition sketch of shear strain of an elemental area

under a condition of pure shear (the element is elongated in one
direction and shortened at right angles to that direction). Although
the angle between originally perpendicular line segments has
changed, under infinitesimal strain the lengths of the strained and
unstrained line segments remain similar and the angular change is
very small compared to one radian. The dashed line shows the
original shape of the element.
y
x
Uy
1
x
Figure 6 Definition sketch of a rotated, but unstrained, elemental
area. The rotation angle is small compared to one radian.
case, displacement is positive in the y-direction and negative in

the x-direction, and we shall use a subscript naming convention (oyx) to indicate that the rotation is in the positive
y-direction and negative x-direction of the xy-plane. From
geometrical arguments, we can show that
sin a1
dUy
dUx
; sin a2
dx
dy
For very small angles, sin aEa, and therefore a1EdUy/dx and
a2E dUx/dy. We now define the average rotational deformation in a positive sense as
oyx 12a1 a2
42
Substituting the definition of a1 and a2 into eqn [42] and

taking the limit as dx and dy go to zero shows that positive
rotation in the xy-plane is given by

1 q Uy q Ux

43
oyx
2 qx
qy
Unlike the components of shear strain, the components of
rotation are not symmetric; instead, oyx oxy. Similar derivations lead to

1 q Uz q Uy

ozy
2 qy
qz

44
1 q Ux q Uz
oxz

2 qz
qx
Combining all the components of rotation into a rotation
tensor yields

1
1 q Ux q Uy
1 q Ux q Uz

B
2 qy
2 qz
qx
qx C
B

C
C
B
B 1 q Uy q Ux
1 q Uy q Uz C
C

0

oij B
B2 qx
2 qz
qy
qy C
C
B

C
B1 qU
q Ux
1 q Uz q Uy
z
A
@

0
2 qx
2 qy
qz
qz
0
describe infinitesimal strains occurring in solid bodies subject

to stresses (Turcotte and Schubert, 2002).
7.3.5.4
q Ux
B qx
B
q Ui B
B q Uy
B
B qx
q xj
B
@ q Uz
qx
q Ux
qy
q Uy
qy
q Uz
qy
q Ux
qz C
C
q Uy C
C
qz C
C
C
q Uz A
qz
46
From this analysis, we see that we cannot define distortion, or

strain, solely in terms of simple displacement gradients. To do
so is to mix both rotation and shear strain in the off-diagonal
terms (Middleton and Wilcock, 1994). Because solid-body
translations and rotations do not alter distances between
neighboring elements of a solid, expressions that link stresses
to strains concern only the strain tensor, which is sufficient to
Strains in an Arbitrarily Oriented Coordinate

System
As with stresses, having expressions for the changes in lengths

and rotation angles of line elements that are oriented in any
coordinate direction is useful. Defining strain in an arbitrary
coordinate direction also involves resolving trigonometric relationships. Similar to the discussion of stresses, the discussion
for strains focuses on two dimensions and is relative to strains
defined as positive in elongation. Strains within an xy-plane
referenced to a new coordinate system that makes an angle y
with the original xy-coordinate system can be written as
(Turcotte and Schubert, 2002; Middleton and Wilcock, 1994)
ex0 x0 exx cos2 y eyy sin2 y exy sin2y
47
ey0 y0 exx sin2 y eyy cos2 y exy sin 2y
48

ex0 y0 12 exx eyy sin2y exy cos 2y
49
Similar to the principal axes that were defined for stresses,

principal axes of strain can be defined along directions in
which shear strain is zero by setting ex0 y0 equal to zero in eqn
[49], which gives
tan 2y
2exy
exx eyy
50
Again, y and y p/2 define the directions of the principal

strains e1 and e2.
By setting exy 0, exx e1, and eyy e2, we can write expressions for the normal strains and shear strains in a coordinate system that makes an angle y with respect to the
principal strain (e1) axis as (Turcotte and Schubert, 2002;
Middleton and Wilcock, 1994)
45
If we now sum the strain tensor (eqn [41]) and the rotation
tensor (eqn [45]), we have a deformation tensor that defines
simple displacement gradients:
31
exx e1 cos2 y e2 sin2 y
51
eyy e1 sin2 y e2 cos2 y
52
exy 12e1 e2 sin 2y
53
In eqns [51][53], the prime notation has been dropped as it

merely indicates the components in a new coordinate system.
Tensor notation expressions for three-dimensional infinitesimal strains of lines rotated with respect to a defined
coordinate system are discussed in Means (1976).
7.3.6
Rate of Deformation
The discussion of solid-body deformation considered the kinematic behavior (analysis of motions without reference to the
forces involved) of a solid undergoing a very small geometric
change. Because the strains and rotations that defined deformation were very small we focused on displacements, displacement fields, and displacement gradients relative to
material coordinates in an undeformed reference state, a state
32
very similar to the final deformed state. When considering fluid

deformation, however, displacements of fluid elements are not
small, but instead are very large or indefinite. For example, if
one poured maple syrup on a plate and tilted the plate, the
syrup would flow as long as the plate remained tilted. Therefore, in fluids, we are concerned not with displacements, but
rather with rates of displacements, or velocities. Instead of
considering deformation, we consider instead rates of deformation and focus on velocities, velocity fields, and velocity
gradients. These fields are expressed in terms of spatial coordinates and time without regard to an initial state. Similar to
the discussion of strain and deformation, rate of deformation
in a broader sense than that typically used in the engineering
literature (e.g., Malvern, 1969) is used. Here, rate of deformation encompasses both rate of distortion (commonly called
rate of strain), which involves what we might think of as rate of
dilatation and rate of shear, and which Malvern (1969) calls
rate of deformation, and rate of rigid-fluid rotation, called
vorticity, which Malvern (1969) calls spin.
It is tempting to relate the rate-of-deformation tensor to the
time derivative of the deformation tensor (eqn [41]), but this
is not the general case (Malvern, 1969). As noted, deformation
is defined in terms of material coordinates relative to an initial
reference state, whereas the rate of deformation is defined
with respect to the present spatial coordinates of a particle.
Consider an observer recording fluid velocities at some point
in a system. That observer cannot see the initial position of, or
the trajectory followed by, a fluid particle, but can only record
the particle velocity as a function of spatial position and time.
The components of the rate-of-deformation tensor are, therefore, founded in terms of velocity of a particle instantaneously
at one point relative to a particle instantaneously at a neighboring position (Malvern, 1969), or spatial gradients of instantaneous velocity. The rate-of-deformation tensor and the
time derivative of the deformation tensor are equivalent only
when the displacements and displacement gradients are very
small. For fluid flow, however, displacements of fluid particles
are not small. Because rate of deformation involves spatial
gradients of velocity, it has a unit of s1.
Like the deformation tensor, the rate-of-deformation tensor can be partitioned into a symmetric tensor component
that represents the rate of distortion of a fluid element, and an
antisymmetric tensor component that represents a rigid-body
rate of rotation, or vorticity (Malvern, 1969; Tritton, 1988;
Middleton and Wilcock, 1994). Each tensor is comprised of
components that describe spatial gradients of velocity. The
rate-of-distortion tensor (commonly called the rate-of-strain
tensor in geosciences) is given by
0
q ux
B
B qx

B
B 1 q ux q uy
Lij B
B2 qy
qx
B

B1 qu
q
uz
x
@
2 qz
qx

1
1 q uy q ux
1 q uz q ux
2 qx
2 qx
qy
qz C

C
C
q uy
1 q uz q uy C
C
2 qy
qy
qz C
C

C
1 q uy q uz
q uz
A
2 qz
qy
qz
54
in which ui represents the velocity in the Cartesian coordinate

direction i. The similarity of the rate-of-distortion tensor and
the strain (or distortion) tensor is apparent. By analogy with

solid-body strain, a fluid dilatation rate (rate of volume
change) can be defined by the diagonal terms of the rate-ofdistortion tensor as

D Lxx Lyy Lzz
q ux q uy q uz
qx
qy
qz

55
We shall see subsequently that this sum of spatial gradients of

velocity along coordinate directions, also known as the divergence of the velocity field, is an important component of
the conservation of mass equation. The off-diagonal terms of
the rate-of-distortion tensor define rates of distortion related
to shear.
In an analogous manner, the rate-of-rotation, or vorticity,
tensor is given by

1
1 q ux q uy
1 q ux q uz

B
2 qy
2 qz
qx
qx C
B

C
C
B
B 1 q uy q ux
1 q uy q uz C
C

0

Oij B
B2 qx
2 qz
qy
qy C
C
B

C
B1 qu
q ux
1 q uz q uy
z
A
@

0
2 qx
2 qy
qz
qz
0
56
As with strain and rotation, the components of the rateof-distortion and vorticity tensors can be combined into a
rate-of-deformation tensor that defines rates of dilatation,
rates of shear, and rates of rotation. As with deformation, we
cannot define rates of distortion, the velocity gradients responsible for distorting the shape of a fluid element, solely in
terms of simple velocity gradients because to do so is to mix
both rates of shear and rates of rotation in the off-diagonal
terms.
7.3.7
Conservation
Having examined the concepts of stress, deformation, and

rates of deformation, the chapter turns to the concept of
conservation equations, which are sometimes called equations
of change because they describe changes in quantities such as
temperature, density, and velocity with respect to time and
space (Bird et al., 1960). Conservation is one of the key
principles of continuum mechanics used for solving geomorphic problems quantitatively (Anderson and Anderson,
2010). Conservation equations typically state that the time rate
of change of some quantity within a definable volume equals
the amount of the quantity transported into the volume
plus the amount of that quantity produced (or consumed)
within the volume minus the amount of the quantity transported out of the volume. Conservation of physical quantities
is natures equivalent of balancing a budget. For problems in
geomorphology, we are concerned particularly with conservation of mass and momentum, but some problems also rely
on conservation of energy and conservation of chemical species (a form of mass conservation) (e.g., Anderson and
Anderson, 2010).
7.3.7.1
Conservation of Mass
For a stationary elemental volume (one of fixed size having a

position fixed in space), the rate of mass accumulating within
the volume is equal to the mass flux into the volume minus
the mass flux out of the volume. We shall assume that no mass
is generated (or consumed) within the volume (i.e., we ignore
relativistic effects); hence, the only way to have mass accumulate in the volume is to have it flow across the volume
boundary. By definition, material density is the amount of
mass per unit volume; thus, we can represent mass as the
product of density and volume. Consider the pair of faces of
the elemental volume perpendicular to the x-axis, as shown in
Figure 7. The flow of mass entering the volume across the lefthand face (at x) is given by the product of material density, the
area of the face across which the mass flows, and the velocity
at which the mass flows (rux,dydz)x. Likewise, the flow of mass
leaving the volume across the right-hand face (at x dx) is
given by (rux,dydz)x dx. Similar expressions can be written for
the mass fluxes across faces perpendicular to the y and z axes.
Using the expressions for mass flux across all faces of the
elemental volume, we can write the principle of mass conservation in mathematical terms as

q
rdxdydz rux dydzx rux dydzxdx
qt
h
i
ruy dxdzy ruy dxdzydy
ruz dxdyz ruz ydxdyzdz

q
r dydz rux x rux xdx
qt
h
i
dxdz ruy y ruy ydy

dxdy ruz z ruz yzdz
y
uyy+y

qr
q rux q ruy q ruz
qt
qx
qy
qz
58
uzz
This equation, also known as the continuity equation, states

that the rate of change of density within an elemental volume
fixed in space is equal to the net rate of mass flux across its
boundaries divided by its volume (Bird et al., 1960). The
negative sign in front of the right-hand side of the equation
indicates that if the net gradient of mass flux decreases along
the coordinate directions within the volume, then the change
of density with time is positive (because mass accumulates). If,
however, the net gradient of mass flux increases along the
coordinate directions within the volume, the change of density
with time is negative (because mass is lost).
The derivation of eqn [59] assumed that the elemental
volume was fixed in space. Suppose, however, that we allow
the volume to float along with the medium that is transporting mass across the volume boundary. What happens to our
expression for rate of change of density? Insight to that
question can be gleaned if we carry out the differentiation of
the term on the right-hand side of eqn [59] and regroup terms.
Doing so yields
uxx+x
x
y
z
uzz+z
z
x
60
The term on the left-hand side of eqn [60] now describes rate
changes of density detected from the perspective of an observer floating along with the motion of the medium that is
transporting the mass. The first term describes the rate of
change observed from general variations with time, and the
following three terms describe an additional change related to
any spatial gradients of density. The rate of change observed
from spatial gradients of density is related to the local velocity
of the transporting medium and is known as the convective
rate of change. The term on the left-hand side of eqn [60]
represents a special kind of time derivative called the substantial or material derivative, or the time derivative following motion (Bird et al., 1960). In this instance, that special
derivative is expressed as
Dr q r
qr
qr
qr
ux
uy
uz
Dt
qt
qx
qy
qz
uxx
59
57
Because the volume is stationary, we can recast eqn [57] as

dxdydz
Dividing eqn [58] by dxdydz and taking the limit as these

dimensions go to zero yields:

qr
qr
qr
qr
q ux q uy q uz
ux
uy
uz
r
qt
qx
qy
qz
qx
qy
qz
33
uyy
Figure 7 Definition sketch of mass flux across the boundaries of an

elemental volume.
61
Using that special derivative notation, and with slight rearrangement, we can write eqn [60] as

Dr
q ux q uy q uz
r
0
Dt
qx
qy
qz
62
For an incompressible medium (commonly a valid approximation for many geomorphic problems), density is constant.
Thus, Dr/Dt 0, and, therefore, the sum of the spatial gradients of the velocity field in the second term (known as the
divergence of the velocity field) must also be zero. Equations
[59] and [62] are commonly written in a more compact
34
y
notation, using what is called the divergence operator, as
qr
rru
qt
and
Dr
rru
Dt
uyuy
y+y
uzuzz
63
For an incompressible material ru 0.
7.3.7.2
Conservation of Linear Momentum
A very powerful concept in continuum mechanics is the

principle of conservation of momentum. This concept states
that the time rate of change of momentum in an elemental
control volume equals the flow of momentum into the control
volume minus the flow of momentum leaving the volume
plus the sum of the forces acting on the volume. This is essentially a control volume formulation of Newtons second
law that accounts not only for the forces acting on the volume,
but also for the impact of momentum flowing across the
control volume boundary (Bejan, 1984). Let us again consider
a stationary elemental volume depicted in Figure 8(a), and
examine the x-direction component of the conservation of
momentum expression.
Momentum is the product of mass and velocity. From the
analysis of mass conservation, we saw that mass is represented
as the product of density and volume. We can, thus, express
momentum as the product of density, volume, and velocity.
Therefore, the flow of momentum entering the volume across
the left-hand face (at x) is given by (ruxuxdydz)x and the flow
of momentum leaving the volume across the right-hand face
(at x dx) is given by (ruxuxdydz)x dx. Note, however, that
x-direction components of momentum are also crossing other
faces of the volume. The x-direction component of momentum entering the volume across its bottom face (at y) is
given by (ruxuydxdz)y , and the x-direction momentum flowing
out the top face (y dy) is given by (ruxuydxdz)y dy . Similar
expressions can be written for the x-direction component of
momentum flowing across the rear and front faces of the
volume. Gathering these terms, we can write an expression for
the net convective flux of the x-direction component of momentum across the volume boundary:
rux ux dydzx rux ux dydzxdx rux uy dxdzy
rux uy dxdzydy rux uz dxdyz rux uz dxdyzdz
64
In addition to the convective flux of momentum across the

volume boundary, the rate of change of momentum is also
affected by the forces acting on the volume, of which there
are two types: body forces that affect all parts of the volume
equally and surface forces related to the stresses acting on
the volume. Stresses acting on the volume affect not only
its bulk momentum, but also they induce diffusive molecular
transfer of momentum across the volume surface (Bird et al.,
1960). From Figure 8(b), we see that the sum of the
x-direction components of these forces can be written as
x
y
z
uzuzz+z
x
z
uy uyy
(a)
y
yy
yx
xz yz
zx
xx
zx
yz
y
zz
x
xy
zy
zy
xy
zz
xx
x
xz
z
yx
yy
(b)
Figure 8 Definition sketches of momentum flux through, and

stresses acting on, an elemental volume. (a) Momentum fluxes
across the boundaries of the elemental volume. (b) Stresses acting
on the volume surface.
Noting that the time rate of change of momentum in the

x-direction can be written as q rux dxdydz=q t and using eqns
[64] and [65], we can write an expression for the x-direction
component of the conservation of momentum as
q
rux dxdydz rux ux dydzx rux ux dydzxdx
qt
rux uy dxdzy rux uy dxdzydy
rux uz dxdyz rux uz dxdyzdz
sxx dydzxdx sxx dydzx
sxx dydzx sxx dydzxdx syx dxdzy syx dxdzydy

szx dxdyz szx dxdyzdz rgx dxdydz
uxuxx+x
uxuxx
65
Here, the negative terms result from positively defined forces

acting in the negative x-direction.
syx dxdzydy syx dxdzy

szx dxdyzdz szx dxdyz
rgx dxdydz
66
If we assume the spatial dimensions of the elemental volume are constant, then we can divide eqn [66] by dxdydz, take
the limit as each of those dimensions go to zero, and write the
x-direction component of the conservation of momentum as
q rux
q
q
q
rux ux ruy ux ruz ux
qx
qy
qz
qt
q sxx q syx q szx
rgx
qx
qy
qz
67
Expressions for the y- and z-direction components of the

conservation of momentum are derived in a similar manner
and are
q ruy
q
q
q
rux uy ruy uy ruz uy
qx
qy
qz
qt
q sxy q syy q szy
rgy
qx
qy
qz
68
and
q ruz
q
q
q
rux uz ruy uz ruz uz
qx
qy
qz
qt
q sxz q syz q szz
rgz
qx
qy
qz
71
From the previously discussed definition of the material derivative, and using eqn [61], we can recast this equation as

Dux
Dr
q ux q uy q uz
r
ux
r
Dt
Dt
qx
qy
qz
q sxx q syx q szx
rgx
qx
qy
qz
Dux q sxx q syx q szx
rgx
Dt
qx
qy
qz
73
In three dimensions, and using vector and tensor notation, we

can write this form of the conservation of momentum equation as
Du
r
r~
s rg
74
Dt
This expression, which explicitly relates material accelerations
(left-hand term) to surface and body forces (right-hand
terms), is called Cauchys first law.
If we now recall that total stress can be written as the sum
of the mechanical mean pressure (mean normal stress) and
the deviatoric stress, we can write eqn [73] in an alternative
form as

Dux
qp
q sxx p q syx q szx
rgx 75
r
Dt
qx
qx
qy
qz
69
in which u represents the velocity vector field, s~ the stress

tensor (eqn [4]), and g the vector field for acceleration of
gravity.
An alternative form of eqn [70] is achieved if we carry out
the differentiation of the term on the left-hand side and the
first group of terms on the right-hand side and regroup the
terms. For simplification, consider the x-direction component
of the momentum conservation equation. Equation [67], after
carrying out the noted differentiation, can be rearranged as

q ux
q ux
q ux
q ux
ux
uy
uz
r
qt
qx
qy
qz

qr
qr
qr
qr
q ux q uy q uz
ux
ux
uy
uz
r
qt
qx
qy
qz
qx
qy
qz
q sxx q syx q szx
rgx
qx
qy
qz
is equal to zero. Thus, we can write the x-direction component

of the conservation of momentum equation as
We can extend this to three dimensions and recast eqn [74] as
We can write the three components of momentum conservation in a more compact manner using vector and tensor
notation as
q ru
rruu r~
s rg
70
qt
35
72
We now recognize the bracketed term as the mass conservation, or continuity, equation (eqn [62]), and that equation
Du
rp r~
s D rg
Dt
76
The reason for presenting the conservation of momentum

equation on the one hand in terms of total stresses and on the
other in terms of mechanical mean pressure and deviatoric
stresses is that in solid mechanics we are generally interested
in total stresses, whereas in fluid mechanics we are interested
in dealing explicitly with the pressure that acts on a fluid
element and the stresses that deviate from pressure and act to
change the shape of a fluid element. Hence, eqn [74] is the
form of conservation of momentum commonly used in solid
mechanics, whereas eqn [76] is the form used in fluid
mechanics.
7.3.8
Constitutive Relations
The discussion thus far has focused on generalized aspects of

forces and stresses acting on bodies, the influence of pore-fluid
pressure on stresses, and development of concepts of strain
and conservation. None of these concepts, however, tells us
precisely how a body will respond to an applied stress in
other words, precisely how stresses and strains (or rates of
distortion) are related. Linking stresses with strains or rates of
distortion requires a conceptualization of how a material acts.
Unlike the development of the concepts of stress, strain, and
conservation from fundamental physical principles of mass,
motion, and geometry, conceptual expressions of material
activities are idealizations rooted in empiricism. Although real
materials can exhibit very complex stressstrain relationships,
and complex idealizations have been proposed to model those
relationships, some relatively simple idealizations approximate activities of real materials fairly well. Expressions that
describe the relationships between stress and strain, or stress
and rates of distortion, are called constitutive equations.
36
Three basic idealizations are used to describe material

actions: elastic, plastic, and viscous. Real materials can be
represented by these idealizations, sometimes simply and
sometimes by invoking complex combinations of idealized
actions. In Earth sciences, in general, material actions are
commonly explored using combinations of idealized material
actions (e.g., Middleton and Wilcock, 1994). Yet, despite
clearly limiting assumptions, even the simplest of idealized
material actions commonly lead to substantial insights.
Special cases of idealized actions include linear elasticity,
ideal plasticity, and linear viscosity. A linearly elastic material
strains immediately upon application of a stress, and the
amount of strain is directly proportional to the stress. As long
as the stress remains, the strain persists; once the stress is removed, however, all of the strain is recovered. The classical
conceptualization of a linearly elastic body is a spring. In an
idealized plastic material, the body does not strain until the
stress equals a threshold value, known as the yield stress. After
an applied stress equals the yield stress, the material strains
indefinitely so long as the stress equals the threshold value.
Unlike a linearly elastic material, the amount of strain is not
directly proportional to the stress, nor is it recoverable upon
removal of the stress. The classical conceptualization of an
ideal plastic material (also known as a Saint Venant material)
is a block sitting on a surface. The block will not move until
the friction between it and the surface it is sitting upon is
overcome. Once the friction is overcome, the block continues
to move across the surface as long as the applied stress (in this
case a shear stress) is sufficient to overcome the friction.
A simple viscous material, typically a liquid, strains immediately and indefinitely upon application of a shear stress. The
shear stress is, therefore, not proportional to strain, but rather
to how fast the material is distorting. In an idealized linearly
viscous material, called a Newtonian fluid, the shear stress is
linearly proportional to the rate of distortion, or the velocity
gradient established within the fluid. Fluids for which that
relationship is nonlinear, or in which the fluid does not respond in a viscous manner until a yield stress is exceeded, are
called non-Newtonian fluids. The classical conceptualization
for a Newtonian viscous fluid is an oil-filled dashpot or
springless shock absorber. Another idealized behavior useful
to geomorphology, known as the Coulomb failure rule, does
not relate stresses to strains, but rather assumes a plastic-like
behavior and relates the shearing strength of a material to its
shearing resistance. This empiricism is commonly used to assess the stability of masses of rock and soil and to relate shear
stresses to normal stresses in some models of flowing granular
material. Next, the mathematical formulations of a few special
idealized materials are examined.
81 (or 9 9) components (e.g., Malvern, 1969; Means, 1976;

Middleton and Wilcock, 1994). The coefficients that relate
stresses to strains are called stiffnesses, whereas the inverse coefficients that relate strains to stresses are called compliances.
Fortunately, this complex array of coefficients can be simplified.
We saw that for both stresses and strains, only six of the tensor
components are independent, which reduces the array of stiffnesses and compliances to 36 independent components. This
array of coefficients can be further simplified, and written in
terms of two elastic moduli (e.g., Middleton and Wilcock,
1994), if we assume that the elastic material under consideration is isotropic, meaning that the elastic moduli are identical
in all directions. We shall see that four basic moduli are used to
describe the behavior of linearly elastic material, but that only
two of those moduli are independent.
An ideal linearly elastic material is described in terms of
Hookes law, which states that stress is linearly proportional to
strain (Timoshenko and Goodier, 1987). As a general concept,
the law can be represented for uniaxial normal stress as
s Ee
77
in which s represents a uniaxial normal stress (positive in

tension), e represents strain (positive in elongation), and the
constant of proportionality E is known as Youngs modulus.
Consider the strained body shown in Figure 9. If a tensional
normal stress is exerted in the y-coordinate direction, Hookes
law can be written as
syy Eeyy
78
Note, however, that Youngs modulus does not fully relate

the normal strains occurring in the body to the applied stress.
For most elastic materials, as the body elongates in the ydirection, it contracts and results in a negative strain in the
orthogonal directions. For example, the amount of contraction
yy
l y = yy l y
ly
lx
7.3.8.1
7.3.8.1.1
l x = xx l x
Linearly Elastic Material

Relationships between stress and normal strain
Linear elastic theory relates stresses to strains in an idealized

elastic body and is strictly applicable only to infinitesimal
elastic strains. Previously, we saw that nine components fully
define each tensor (or matrix) of stress and strain. For the
general case, the constitutive model that relates stress to strain
in an elastic body composes a tensor (or matrix) comprised of
yy
x
Figure 9 Definition sketch of normal strains in a two-dimensional
linearly elastic material subject to uniaxial stress. The dashed line
shows original shape of the element. Normal strains are defined
positive in elongation.
2
in the x-direction is proportional to the amount of extension in

the y-direction. The ratio of contraction to extension is known
as Poissons ratio and is expressed for uniaxial stress as
u
exx
eyy
37
79
u
exx syy
E
80
Elongation in the y-direction also causes contraction in the

z-direction. For an isotropic material, under uniaxial stress, the
amount of contraction in the z-direction is identical to
the amount of contraction in the x-direction. Hence, we can
write a similar expression relating strain in the z-direction to
normal stress in the y-direction
u
81
ezz syy
E
For an isotropic material subject to triaxial rather than uniaxial
stress, and following along similar reasoning, we can write the
normal strains along the coordinate directions in terms of
Youngs modulus and Poissons ratio as

1
sxx usyy szz
E

1
eyy syy usxx szz
E

1
ezz szz usxx syy
E
exx
82
From these simple derivations, we see that normal strains in

any coordinate direction are related to normal stresses applied
in all coordinate directions and not just to the normal stress
applied in the direction coincident with the normal strain.
Equation [82] can also be written in terms of principal stresses
and principal strains
1
e1 s1 us2 s3
E
1
e2 s2 us1 s3
E
1
e3 s3 us1 s2
E
Principal axis 2
Substituting eqn [79] into eqn [78] yields

x
xy

Principal axis 1
1
Figure 10 Definition sketch relating coordinate axes and shear

stress on a plane having its outward-normal pole rotated at an angle
y to the principal stress axes. The nature of the principal stresses
defines a condition of pure shear.
In the discussion of stress, we found that the maximum

shear stress occurs on a plane that makes a 451 angle to
the principal stress axes. Under the assumption that s1
s2, and for y 451, we find from eqns [12][14] that
sxx syy 0 and that sxy s1. From the finding that sxy s1
and the assumption that s1 s2, we find from eqns [84]
that the principal strains are related to the maximum shear
stress
1 u
sxy
E
1 u
sxy
e2
E
e1
85
Under the conditions specified, eqn [85] shows that

e1 e2. Using this result and with y 451, we find from
eqns [51][53] that
exx eyy 0
exy e1
83
86
From this result, eqn [85] yields
7.3.8.1.2
Relationships between shear stress and

shear strain
exy
Relationships among shear stress, shear strain, and elastic

moduli are derived by considering a special case of plane stress
that is known as pure shear (Turcotte and Schubert, 2002). For
this case, s3 0, and we shall assume s1 s2. Furthermore,
we shall consider the strain that occurs within a plane in a
coordinate system that is rotated with respect to the principal
stress axes (Figure 10). Under these constraints and using eqn
[83], we find that
1 u
s1
E
1 u
s2
e2
E
e3 0
e1
84
1 u
sxy
E
87
Owing to symmetry of shear stresses and shear strains, exy

eyx. By analogy, similar expressions for shear strains in the
orthogonal planes are written as
exz
eyz
1 u
sxz
E
1 u
syz
88
Unlike the relationships between normal stresses and

elastic normal strains, the elastic shear strains are related to
only a single component of shear stress and not to all components of shear stress. Furthermore, by recalling that exy is the
38
average shear strain in the xy plane and is equal to gxy (where

gxy is the total, or engineering, shear strain in the xy plane),
eqn [87] can be written as
E
g
sxy
21 u xy
89
sxy Ggxy
90
where G E/2(1 v) is the rigidity modulus, also known as

the shear modulus.
Relationship between pressure and dilatation
We can now define another elastic modulus, known as the

bulk modulus, which describes the relationship between
mean normal stress (or mechanical mean pressure) and volume strain (or dilatation). We can write a relationship between the mean normal stress (positive in tension) and
volume strain (positive in dilation) as

1
sxx syy szz kx
3
91
Substituting eqns [35] and [82] into eqn [91] gives

1
1
sxx syy szz sxx usyy szz
3k
E

1
syy usxx szz
E

1
szz usyy szz
E
92
which, after some algebraic manipulation, reduces to

k
E
31 2u
95
7.3.8.2
Linearly Viscous Fluid
For a Newtonian viscous fluid, stress is linearly related to the

rate of distortion. Consider, for example, the simplest case in
which an incompressible fluid is subject to a shear stress in
the x-direction and responds by distortion in the xy-plane
(Figure 11). In that case, all velocity gradients are zero except
for q ux =q y, and all components of shear stress are zero except
for syx. Normal stresses (i.e., pressure) acting on the fluid may
not be zero, but because the fluid is assumed incompressible
they do not affect fluid distortion. For the case considered, the
relationship between shear stress and rate of distortion can be
expressed as
q ux
96
tyx m
qy
Here, a slightly different notation for the stress (t instead of s)
is used, the reason for which will become apparent. The
constant of proportionality, m, that relates shear stress to velocity gradient is called the dynamic viscosity, which Newton
referred to as the lack of slipperiness between fluid layers
(Barnes et al., 1989). This coefficient has a unit that is the
product of stress and time, N m2 s, or Pa s. We can write the
relationship between shear stress and velocity gradient in a
more general form by noting that the rate of distortion by
shear in the xy-plane is written as

1 q ux q uy
97
Lyx
2 qy
qx
93
The coefficient k is the bulk modulus; its reciprocal is known

as the compressibility.
We have now defined four basic moduli for a linearly
elastic material, which relate stresses to infinitesimal elastic
strains: E, u, G, k. The derivations show that only two of these
moduli (E, u) are independent; the other two (G, k) are
interdependent and written in terms of Youngs modulus and
Poissons ratio.
French engineer Gabriel Lame (17951870) wrote the first
book devoted to the mathematical theory of elasticity and
derived a complete set of constitutive relations for an elastic
solid (e.g., Middleton and Wilcock, 1994). Commonly, his
general relationship is written in terms of stresses as a function
of strains as
sij 2Geij lekk dij
uE
1 u1 2u
Although the bulk modulus k can be derived in terms of l, we

see that l does not have a direct physical interpretation.
or
7.3.8.1.3
before and has the simple physical interpretation as the shear

modulus. The constant l is defined as
For the case under consideration q uy =q x is zero, and as a

result we can write the rate of distortion under shear as

1 q ux
98
Lyx
2 qy
y
yx
uxy
y
*ux /*y
94
where i and j represent the coordinate directions x, y, z; ekk

represents the sum of the normal strains (dilatation); and dij
the Kronecker delta, or identity matrix. The coefficients G and
l are known as Lames constants. The constant G is defined as
x
Fixed boundary
Figure 11 Definition sketch of the velocity gradient produced by
shear of a linearly viscous fluid in the xy-plane.
Substituting eqn [98] into eqn [96] yields

tyx 2mLyx
viscous fluid as
99
r

Dux
qp q
q ux
q
q ux q uy

2m
qy
Dt
qx qx
qx
qy
qx
For an isotropic, incompressible fluid eqn [99] can be generalized to three dimensions as
tij 2mLij
101
The reason for writing the viscous stress tensor as tij, to explicitly distinguish it from the notation used for the total stress
tensor, sij, should now be apparent. The reader may now
recognize the similarity between the total stress tensor defined
for a viscous fluid and the general total stress tensor previously
defined in terms of the mechanical mean pressure (mean
normal stress) and the deviatoric stress, sij pdij sD
ij :
The thermodynamic hydrostatic pressure p and the viscous
stress tensor tij are, in the general case, unequal to the mechanical mean pressure and the deviatoric stress (Malvern, 1969;
Middleton and Wilcock, 1994). For an incompressible fluid,
however, the mechanical mean pressure is identical to the
hydrostatic pressure, and hence the viscous stress tensor is
identical to the deviatoric stress. Even for a monatomic gas,
the mechanical mean pressure and hydrostatic pressure are
nearly the same and so are the viscous and deviatoric stresses
(Middleton and Wilcock, 1994). For an isotropic, incompressible, linearly viscous fluid, we can write eqn [101] as
sij pdij 2mLij
100
The tensor defined by eqn [100] is called the viscous stress

tensor; the similarity of the expression between stress and rate
of distortion in a linearly viscous material and between stress
and strain in a linearly elastic body (e.g., eqns [78] and [87]) is
evident. In a manner similar to that for a linearly elastic body,
a more general treatment of the relationship between stresses
and rates of distortion in a linearly viscous fluid can be derived, including coefficients analogous to Lames constants
(e.g., Middleton and Wilcock, 1994).
Let us consider further the stresses acting in a viscous fluid.
For a fluid at rest, we find from the definition of the rate-ofdistortion tensor that Lij 0. Hence, for a viscous fluid at rest
the viscous stress tensor is zero and, therefore, the stress
components txx tyy tzz 0. However, even in a static fluid,
a normal stress equal to the hydrostatic pressure (p) acts on all
fluid particles. Therefore, the total stress acting on a viscous
fluid is composed of two parts: a hydrostatic pressure and a
viscous stress. Thus, for a viscous fluid, the total stress tensor
can be written as
sij pdij tij
39
102
which represents the surface forces acting on a linearly viscous

fluid element.
Equation [102] can be used with Cauchys first law, the
conservation of momentum expression, to derive sets of
equations that describe pressure and velocity components of a
flowing isothermal fluid. Inserting eqn [102] into eqn [74] (or
eqn [100] into eqn [76]), we can write the x-component for
the conservation of momentum of an incompressible linearly

q
q ux q uz
m
rgx
qz
qz
qx
103
With some algebraic manipulation and under the assumption

that the viscosity of the fluid is constant, this expression can be
recast as
2

Dux
qp
q ux q 2 ux q 2 ux
m
r
2
2
2
Dt
qx
qx
qy
qz

q q ux q uy q uz
rgx
m
104
qx qx
qy
qz
For an incompressible fluid, conservation of mass dictates
that the bracketed third term on the right-hand side of the
equation, known as the divergence of the velocity field, is zero.
Thus, eqn [104] can be further simplified to
r
2

Dux
qp
q ux q 2 ux q 2 ux
m
rgx
Dt
qx
q x2
q y2
q z2
105
This expression is the x-direction component of an equation

known as the NavierStokes equation for an incompressible,
linearly viscous fluid of constant viscosity. The viscous stress
term on the right-hand side of eqn [105] can be written more
compactly using the Laplacian operator r2 as
r
Dux
qp
mr2 ux rgx
Dt
qx
106
This expression can be generalized to three dimensions as

r
Du
rp mr2 u rg
Dt
107
The NavierStokes equation, derived using the conservation of

momentum equation, is a control volume representation of
Newtons second law for an incompressible, linearly viscous
fluid. Therefore, eqn [107] states that the inertial acceleration
(left-hand term) equals the sum of forces (right-hand terms)
acting on the system. Because the left-hand term has dimensions of force per unit volume, it is convenient to think of this
dynamical equation as a balance of static forces (Tritton,
1988). Thus, the left-hand term can be viewed as an inertial
force, and the right-hand terms as pressure, viscous, and body
forces, respectively. If we now introduce the concept of the
Reynolds number (Re), which represents the ratio of inertial
forces to viscous forces in a fluid, we can highlight some
limiting applications of the NavierStokes equation. For lowRe flows, viscous forces greatly exceed inertial forces; thus, the
left-hand term is negligibly small and eqn [107] can be simplified as
rp mr2 u rg
108
This equation, known as Stokes equation, chiefly balances

pressure and viscous forces, and describes creeping flow. It has
40
been used in geoscience, for example, to obtain estimates of

the velocity of a magma as it ascends through the lithosphere
(Turcotte and Schubert, 2002). For high-Re flows viscous forces are assumed to be relatively unimportant, and eqn [107]
can be simplified to
r
Du
rp rg
Dt
109
which is known as Eulers equation. This equation, which

chiefly balances inertial and pressure forces, has applications,
for example, in river hydraulics.
7.3.8.3
One of the principal empiricisms in soil mechanics that is

used widely in geomorphology relates the mean shearing
stress acting on a potential failure surface in a soil mass to soil
cohesion, normal stress, and the angle of internal friction. This
empiricism, commonly referred to as Coulombs law or
Coulombs failure rule, is generally written as
110
where t is the mean shearing stress, C the apparent material

cohesion (nonfrictional component of the soil strength), s0
the effective normal stress (negative in compression) acting on
the potential failure surface, and f characterizes the friction
among soil particles and is called the angle of internal friction
of the soil. Apparent soil cohesion depends on electrostatic
forces that act between clay particles, on cementation of soil
particles owing to secondary mineralization, on surface tension in water films between particles, and on the strength of
roots that infiltrate soil surfaces (Selby, 1982; Iverson et al.,
1997; Schmidt et al., 2001). The dominant control on soil
(and rock) strength, however, is frictional resistance between
particles and the interlocking among particles, and the product s0 tanf determines the frictional component of shear
strength. In general, apparent cohesion of soils is small and
not an important contributor to soil strength except in very
clay-rich soils, in near-surface soil pervasively penetrated by
roots, or in soils where effective stresses are low.
The effect of pore-water pressure on the shearing strength
of soil is made explicit by substituting the expression for
effective stress eqn [18] into eqn [110], which gives
t C s ptan f
111
This deceptively simple expression is commonly used to assess

the factors that govern slope failure. However, this expression
is incomplete in that it does not account for the stress and
pore-pressure fields that determine the mean shear stress,
effective stress, and pore-fluid pressure acting on a potential
failure surface (Iverson et al., 1997). The magnitude and
spatial distribution of pore-fluid pressure (which is related to
the distribution of hydraulic head) and the spatial distribution
of solid-grain stress determines the Coulomb failure potential
of a soil (Iverson and Reid, 1992; Iverson et al., 1997).
Methods for measuring rock and soil strength, along with
typical values of cohesion and angles of internal friction,
are provided by Selby (1982) and Middleton and Wilcock
(1994).
Example Application
The concepts discussed in this chapter are widely used in

geomorphology, and more broadly in geosciences. The reader
is referred to textbooks by Johnson (1970), Means (1976),
Middleton and Wilcock (1994), Turcotte and Schubert
(2002), and Anderson and Anderson (2010) for several detailed examples. Here, one example relevant to hillslope geomorphology, namely an analysis of slope-failure potential,
which utilizes many of the concepts discussed, is presented.
7.3.9.1
Plasticity the Coulomb Failure Rule
t C s0 tan f
7.3.9
Conservation of Momentum and Stress Equilibrium
Iverson and Reid (1992) developed an elastic effective-stress

model for gravity-driven groundwater flow and slope-failure
potential. To develop the model, they considered an elastic,
porous medium that is isotropic and isothermal, assumed
that the pore water is isothermal and has uniform density
and viscosity, and restricted the model to plane strain and
two-dimensional groundwater flow. They began their mathematical formulation by developing equations for stress
equilibrium. In terms of coordinates in the x- and y-directions,
they wrote Cauchys first law (eqn [74]) for a porous medium
with steady fluid flow (stresses are independent of time and
there are no inertial accelerations) as
q sxx q syx
rt gx
qx
qy
q syy q syx
rt gy
qy
qx
112
where rt is the bulk density of the fluid-saturated porous

medium and other terms are as previously defined.
7.3.9.2
Effective Stress and Effective Stress Equilibrium
Because the model is for a porous medium that contains pore

water, Iverson and Reid (1992) needed to consider effective
stress. For this model, they adopted a general definition of
effective stress as
s0ij sij bpdij
113
in which p is the pore-water pressure, and i and j the indices

that represent x or y. We shall return to the coefficient b
shortly. Substituting eqn [113] into eqn [112], they wrote the
stress equilibrium equations in terms of effective stresses as
q s0xx q s0yx
qp
rt gx b
qx
qy
qx
q s0yy q s0yx
qp
rt gy b
qy
qx
qy
114
By choosing the y-coordinate to be positive upward, gravity

acts exclusively in the negative y-direction, and they defined
the hydraulic head in the medium as (see eqn [17])
h
p
y
rw g
115
in which rw is the pore-water density and g the magnitude of

gravitational acceleration. Solving for p and substituting into
eqn [114] yields effective-stress equilibrium equations with
body forces represented in terms of hydraulic head gradients
and buoyant unit weight
q s0yy q s0yx
qh
rt brw g brw g
qy
qx
qy
116
where (rt brw)g represents the hydrostatic buoyancy effects

of the pore fluid, and the head gradient terms represent the
seepage force associated with pore-fluid flow.
7.3.9.3
Effective Stress and Elastic Strain
Equations [116] contain three unknown stress components. To

determine those stress components, it is necessary to identify
an appropriate constitutive relationship that relates stresses to
strains in the solid porous medium. In this particular problem, Iverson and Reid (1992) adopted constitutive equations
for a porous elastic medium developed by Biot (1941), which
were recast in terms of conventional elastic moduli by Rice
and Cleary (1976). Note, however, that eqn [116] is cast in
terms of effective stress and not total stress. Because of this, the
constitutive equations need to be cast in terms of effective
stress as well. Iverson and Reid (1992) showed that Biots
constitutive equations for total stress reduced to Hookes-lawtype equations for effective stress only if the coefficient b in
eqn [113] is defined as b 1kb/ks, where kb and ks are the
elastic bulk moduli for the solid porous medium and the individual solid grains, respectively. Using that definition for b
leads to the following constitutive expressions relating elastic
(plane) strains to effective stresses:
i
1h
1 u2 s0xx u1 us0yy
E
i
1h
eyy 1 u2 s0yy u1 us0xx
E
1u 0
syx
eyx
E
assumption is reasonable for near-surface Earth materials and

allows eqn [117] to be interpreted in terms of Coulomb failure
potential.
7.3.9.4
q s0xx q s0yx
qh
brw g
qx
qy
qx
Displacement Formulation of Constitutive

Relations and Groundwater Flow
The stress equilibrium and constitutive equations for stresses

and strains are commonly solved subject to appropriate
boundary conditions for stress. However, the problem tackled
by Iverson and Reid (1992) dictates that boundary conditions
be specified in terms of displacements rather than stresses.
They, therefore, recast the governing equations in terms of
effective stresses and displacements. This is achieved by substituting the expressions relating strains to displacements
(eqns [29], [30], and [39]) into eqn [117] and manipulating
the result to obtain

E
q Ux
uE
q Ux q Uy
1 u qx
1 2u1 u q x
qy

q Uy
q Uy q Ux
E
uE
s0yy
1 u qy
1 2u1 u q y
qx

E
q Ux q Uy
s0yx
21 u q y
qx
s0xx
118
Substituting eqn [118] into eqn [116] with b 1, followed by

further algebraic manipulation, results in two equations for
the two components of displacement in terms of the elastic
moduli E and u
2

E
q Ux q 2 Ux
2
2
21 u q x
qy
2

uE
E
q Ux q 2 Uy
1 2u1 u 21 u
q x2
q xq y
exx
rw g
117
These expressions, which are identical to the standard constitutive expressions for a linear elastic material except that
they are cast in terms of effective stresses, show that effective
stress alone is responsible for the elastic deformation of the
porous medium.
Slope failure potential is typically assessed using the Coulomb failure rule. Numerous observations of material failure
show that the definition of effective stress most useful for
describing Coulomb failure is given by eqn [18] (Skempton,
1960; Lambe and Whitman, 1969; Iverson and Reid, 1992).
The discrepancy between that definition of effective stress
and the one used for defining the constitutive relations for a
poroelastic material (eqn [113]) can be eliminated if we assume that the constituent particles forming the porous medium are much less compressible than the porous medium as
a whole (i.e., kb{ks). In that case bE1, and eqn [113] approximates eqn [18] (Iverson and Reid, 1992). Such an
41
qh
qx
2

q Uy q 2 Uy
E
21 u q x2
q y2
2

q Uy q 2 Ux
uE
E
1 2u1 u 21 u
q y2
q xq y
rt rw g rw g
qh
qy
119
Recognizing the coefficients in this set of equations as

Lames constants, eqns [119] can be written as

G
2

q 2 Ux q 2 Ux
q Ux q 2 Uy
qh
rw g
l G
2
2
2
qx
qy
qx
q xq y
qx

q 2 Uy q 2 Uy
q x2
q y2
2

q Uy q 2 Ux
l G
q y2
q xq y
rt rw g rw g
qh
qy
120
42
The formulation thus far provides two equations in three

unknowns (Ux, Uy , h). Thus, another expression is required to
solve the problem. Iverson and Reid (1992) closed the problem by using the conventional equation for steady, twodimensional groundwater flow in a hydraulically isotropic
porous medium, as given by

q
qh
q
qh
K
K
0
qx
qy
qx
qy
121
where K is the hydraulic conductivity, a function of the permeability of the medium and the viscosity of the pore fluid
(e.g., Freeze and Cherry, 1979).
Equations [120] and [121] provide a mathematically
complete problem that can be solved subject to appropriate
boundary conditions in terms of displacements and hydraulic
head gradients. Iverson and Reid (1992) and Reid and Iverson
(1992) imposed the following boundary conditions: a periodic surface topography; no forces acting on the topographic
surface; a bottom boundary at a large, but finite depth; no flow
boundaries along the lateral and basal margins of the domain;
and hydraulic head at any point along the topography surface
equal to the elevation of the point. Numerical solution of the
mathematical problem subject to these boundary conditions
showed that gravity-driven groundwater flow produces a spatially variable body force field that influences the effectivestress distribution in hillslopes; that gravitational effects alone
produce topographically influenced elastic stresses that can
generate near-surface failures even in the absence of groundwater flow; and that near-surface failure potential is increased
as a result of groundwater flow, especially near the slope toe.
In addition to these broad conclusions, Iverson and Reid
(1992) discussed insightful subtleties of the mathematical
formulation of the problem posed.
7.3.10
Concluding Remarks
In this chapter, a brief survey of key concepts of continuum

mechanics is provided. The topics discussed covered total
stress, pore-fluid pressure, effective stress, strain, conservation,
and a few idealized constitutive relationships. The fundamental concepts were derived from considerations of mass,
motion, and geometry in an effort to show the physical
foundations that lie behind the sometimes complex mathematical formulations. We hope that exposing the underlying
physical principles of these concepts and laying out the derivations in some detail provides the reader an entrance into
the literature that discusses these topics in greater detail, and
that it equips the reader with a suitable foundation to better
understand theoretical papers published in geomorphology.
References
Chemistry of Landscapes. Cambridge University Press, Cambridge, 637 pp.
Barnes, H.A., Hutton, J.F., Walters, K., 1989. An Introduction to Rheology. Elsevier,
Amsterdam, 199 pp.
Bejan, A., 1984. Convection Heat Transfer. Wiley, New York, NY, 477 pp.
Biot, M.A., 1941. General theory of three-dimensional consolidation. Journal of

Applied Physics 12, 155164.
Bird, R.B., Stewart, W.E., Lightfoot, E.N., 1960. Transport Phenomena. Wiley, New
York, NY, 780 pp.
Cuffey, K.M., Paterson, W.S.B., 2010. The Physics of Glaciers, Fourth ed. Elsevier,
Amsterdam, 704 p.
Engelder, T., 1994. Deviatoric stressitis: a virus infecting the Earth science
community. Eos, Transactions of the American Geophysical Union 75(18),
209211.
Freeze, R.A., Cherry, J.A., 1979. Groundwater. Prentice-Hall, Englewood Cliffs, NJ,
604 pp.
Gilbert, G.K., 1877. Report on the Geology of the Henry Mountains. U.S.
Geographical and Geological Survey of the Rocky Mountains Region,
Washington, DC, 160 pp.
Hubbert, M.K., 1940. The theory of ground-water motion. Journal of Geology 48,
785944.
Ingebritsen, S.E., Sanford, W.E., 1998. Groundwater in Geologic Processes.
Iverson, R.M., Reid, M.E., 1992. Gravity-driven groundwater flow and slope failure
potential. 1. Elastic effective-stress model. Water Resources Research 28(3),
925938.
Iverson, R.M., Reid, M.E., Iverson, N.R., LaHusen, R.G., Logan, M., Mann, J.E.,
Brien, D.L., 2000. Acute sensitivity of landslide rates to initial soil porosity.
Science 290, 513516.
Iverson, R.M., Reid, M.E., LaHusen, R.G., 1997. Debris-flow mobilization from
landslides. Annual Reviews of Earth and Planetary Sciences 25, 85138.
Johnson, A.M., 1970. Physical Processes in Geology. Freeman, Cooper and
Company, San Francisco, CA, 577 pp.
Lambe, T.W., Whitman, R.V., 1969. Soil Mechanics. Wiley, New York, NY,
553 pp.
Major, J.J., 2000. Gravity-driven consolidation of granular slurries implications
for debris-flow deposition and deposit characteristics. Journal of Sedimentary
Research 70, 6483.
Malvern, L.E., 1969. Introduction to the Mechanics of a Continuous Medium.
Prentice-Hall, Englewood Cliffs, NJ, 713 pp.
Means, W.D., 1976. Stress and Strain Basic Concepts of Continuum Mechanics
for Geologists. Springer, New York, NY, 339 pp.
Middleton, G.V., Wilcock, P.R., 1994. Mechanics in the Earth and Environmental
Sciences. Cambridge Press, Cambridge, 459 pp.
Reeve, R.C., 1986. Water potentialpiezometry. In: Klute, A. (Ed.), Methods of Soil
Analysis Part 1FPhysical and Mineralogical Methods. American Society of
Agronomy, Madison, WI, pp. 545561.
Reid, M.E., Iverson, R.M., 1992. Gravity-driven groundwater flow and slope failure
potential. 2. Effects of slope morphology, material properties, and hydraulic
heterogeneity. Water Resources Research 28(3), 939950.
Remson, I., Randolph, J.R., 1962. Review of Some Elements of Soil-Moisture
Theory. US Geological Survey Professional Paper 411-D, pp. D1D38.
Rice, J.R., Cleary, M.P., 1976. Some basic stress diffusion solutions for fluidsaturated elastic porous media with compressible constituents. Reviews of
Geophysics and Space Physics 14, 227241.
Schey, H.M., 1973. Div, Grad, Curl, and All That. W.W. Norton, New York, NY,
163 pp.
Schmidt, K.M., Roering, J.J., Stock, J.D., Dietrich, W.E., Montgomery, D.R., Schaub,
T., 2001. Root cohesion variability and shallow landslide susceptibility in the
Oregon Coast Range. Canadian Geotechnical Journal 38, 9951024.
Selby, M.J., 1982. Hillslope Materials and Processes. Oxford University Press,
Oxford, 264 pp.
Skempton, A.W., 1960. Terzaghis discovery of effective stress. In: Terzaghi, K.,
Bjerrum, L. (Eds.), From Theory to Practice in Soil Mechanics. Wiley, New York,
NY, pp. 4253.
Terzaghi, K., 1923. Die Berechnung der Durchlassigkeitsziffer des Tones aus dem
Verlaug der Hydrodynamischen Spannungsercheinungen. Akademie der
Wissenchaften in Wein, Sitzungsberichte, Mathematisch-Naturwissenschaftliche
Klasse Part IIa 132(3/4), 125128.
Terzaghi, K., 1943. Theoretical Soil Mechanics. Wiley, New York, NY, 510 pp.
Timoshenko, S.P., Goodier, J.N., 1987. Theory of Elasticity, Third ed. McGraw-Hill,
New York, NY, 567 pp.
Tritton, D.J., 1988. Physical Fluid Dynamics, Second ed. Oxford University Press,
Oxford, 519 pp.
Turcotte, D.L., Schubert, G., 2002. Geodynamics, Second ed. Cambridge University
Zwillinger, D. (Ed.), 2003. Standard Mathematical Tables and Formulae, Thirty-first
ed. CRC Press, Boca Raton, FL, 910 pp.
43
Biographical Sketch
Jon Major is a research hydrologist with the US Geological Survey Cascades Volcano Observatory in Vancouver,
Washington. He received his PhD in 1996 from the Department of Geological Sciences at the University of
Washington. His research focuses on hydrogeomorphic responses to landscape disturbance, particularly in volcanic river systems. He has worked on groundwater flow in landslides, mechanics of deposition by debris flows,
posteruption sediment transport and streamflow hydrology, hydrogeomorphic response to dam removal, and
analyses of debris flow and flood hazards at volcanoes in Washington, Oregon, Alaska, El Salvador, Chile, and the
Philippines. He is a fellow of the Geological Society of America (GSA), and has received the GSA EB Burwell
Award (Engineering Geology Division research publication award), the GSA Kirk Bryan Award (Quaternary
Geology and Geomorphology Division research publication award), and a US Department of Interior Award for
Excellence of Service.
7.4 Influence of Physical Weathering on Hillslope Forms

Y Matsukura, University of Tsukuba, Tsukuba, Japan
7.4.1
7.4.2
7.4.2.1
7.4.2.2
7.4.2.3
7.4.3
7.4.3.1
7.4.3.2
7.4.4
7.4.4.1
7.4.4.2
7.4.5
7.4.5.1
7.4.5.2
7.4.5.3
7.4.5.4
7.4.5.5
7.4.6
References
Introduction: Modes of Physical Weathering

Physical Weathering and Its Effect on Geomorphic Processes
Weathering Products from Physical Weathering
Effect of Weathering on Mass Movement (Slope Processes)
Strength Reduction from Weathering
Sheeting Joints from Unloading (Pressure Release)
Microsheeting in Granite
Exfoliation Rates of Sheeting Joints for Granite Dome
Effect of Slaking on Structural Landforms and Mass Movement
Effect of Slaking on Rock-Controlled Landforms
Effect of Slaking on Landslides on Mudstone Hillslopes
Effect of Crystal Growth Weathering (Salt Fretting and Frost Shattering) on Landforms and Mass
Movement
Effect of Frost Weathering on Periglacial Landforms
Effect of Frost Weathering on Rockfall and Talus Slope Development
Effect of Salt Weathering on Tafoni and Pan Formation
Effect of Crystal Growth Weathering on Notch Formation
Effect of Frost Action and Notch Formation on Cliff Collapse
Conclusion
44
45
45
45
45
46
46
47
47
47
48
49
49
50
50
50
51
53
54
Abstract
Physical weathering plays an important role in reducing the strength and aiding in the disintegration of hillslope materials.
The processes of physical weathering affect many landforms, that is, (1) unloading makes sheeting joints on granite domes;
(2) slaking makes cuesta and hoodoos; (3) salt weathering makes notches, tafoni, and pans; and (4) frost action affects
periglacial landforms such as talus. Slaking promotes landslide and frost action, and enlarging tension cracks behind a
vertical cliff induces cliff collapse. To shed light on the detailed relations between weathering and the formation and
evolution of landforms, it is necessary for geomorphologists to accumulate data on the rate and type of changes in rock
properties from weathering (in particular, strength reduction).
7.4.1
Introduction: Modes of Physical Weathering
Weathering does not necessarily result in change in landform,

unless landform materials are removed. It nevertheless generally leads to a reduction in the resistance of rock. Weathering
can, therefore, be taken as facilitating other processes, in
particular, exogenetic processes.
Weathering (i.e., in situ alteration of rocks and minerals from
exposure to atmospheric agents) is important in reducing the
strength of rocks. It generally takes place by chemical or physical
(mechanical) agencies or both. Although these processes are
interrelated, they will be described separately for convenience.
This section only deals with physical weathering.
Physical weathering is a process of disintegration from
(1) unloading (i.e., pressure release), (2) slaking (i.e., swelling
Matsukura, Y., 2013. Influence of physical weathering on hillslope forms. In:

Shroder, J. (Editor in Chief), Marston, R.A., Stoffel, M. (Eds.), Treatise on
Geomorphology. Academic Press, San Diego, CA, vol. 7, Mountain and Hillslope
44
from wetting; shrinkage from drying), (3) thermal stress, (4) frost
action, and (5) pressure effects associated with salt-crystal growth.
Unloading was first proposed by Gilbert (1904) to be the
mechanism responsible for producing the so-called exfoliation
domes (e.g., Ollier, 1984; Yatsu, 1988). Residual stresses
are reduced, however, by subtle long-term adjustments from
creep or dislocation, and recrystallization may occur. Whether
residual stress genuinely exists in rocks is not clear but is
thought to reside in minerals that crystallized under pressure
at depth and were then brought to the surface by uplift
and erosion so that residual stresses were locked in. The
mechanisms of sheeting and exfoliation deserve deeper explanation based on principles of rock mechanics.
Clay minerals (such as smectite (montmorillonite), swelling chlorite, and interstratified illite/montmorillonite) tend
to swell upon wetting and to shrink as they dry. Materials
containing these clay minerals, such as mudstone and shale,
undergo considerable expansion on wetting (e.g., Matsukura
and Yatsu, 1982; Yatsu, 1988); and this expansion sometimes
initiates microcrack formation, the widening of existing
cracks, or disintegration of the rock mass (as stated in
http://dx.doi.org/10.1016/B978-0-12-374739-6.00147-0
Influence of Physical Weathering on Hillslope Forms
Section 7.4.2.3). In rock drying, evaporation of the water absorbed causes the expanded clays to shrink, and often,
shrinkage cracks form. Alternate swelling and shrinking associated with cyclic episodes of wetting and drying, and
the fatigue effect, culminates in the physical disintegration of
rocks. This process is known as wet/dry weathering, or slaking.
When a rock mass is heated, the temperature of its outer
few millimeters is much higher than farther in because of the
low thermal conductivity of rocks. As a result, the outer part
tends to expand more, generating thermal stresses and leading
to rock disintegration (e.g., Ollier, 1984; Yatsu, 1988). Alternate heating and cooling bring about a fatigue effect, which
accelerates this thermal weathering.
Frost action, which is characterized by the 9% expansion
upon freezing of water that occupies pores and interstices
within a rock body, is also responsible for the physical disintegration of rocks (Brunsden, 1976; Washburn, 1979; Yatsu,
1988). This is called frost weathering and is dominant in cold
environments where alternating freezethaw action commonly occurs. The fatigue effect obviously facilitates this type
of weathering.
In coastal and arid regions, crystal growth from saline solutions is common because of evaporation. Salt crystallization
in interstices within rock produces stress, leading to widening
of the interstices and resulting in granular disintegration. This
is known as salt weathering (e.g., Wellman and Wilson, 1965;
Goudie and Viles, 1997). When salt crystals that have formed
within pores are heated, or become saturated with water, they
expand and exert pressure against the constraining pore wall.
This also produces thermal stress (e.g., Cooke and Smalley,
1968) or hydration stress (e.g., Winkler and Wilhelm, 1970),
both of which are involved in salt weathering (Yatsu, 1988).
7.4.2
7.4.2.1
Physical Weathering and Its Effect on

Geomorphic Processes
Weathering Products from Physical Weathering
Frost shattering and salt weathering reduce rock to weathering

products comprising several sizes of fragment. Laboratory
experiments reproducing frost shattering with 500 or 1000

cycles cause limestone and schist to disintegrate into relatively
larger fragments and reduce mudstone and chalk to fine materials (Lautridou and Ozouf, 1982). Salt weathering in arid
areas in Tunisia and in Death Valley, CA, produces silt from
Yorkstone and fan gravel (Goudie and Day, 1981; Goudie and
Watson, 1984).
7.4.2.2
Effect of Weathering on Mass Movement (Slope

Processes)
The relation between weathering and mass movement is

summarized in Figure 1. The lithology of slope-forming materials reflects their mechanical strength, which is the
main factor determining the resisting force of slope materials,
FR, against driving forces because of the weight of slope
materials, FD; if FR4FD, then the slope should be stable,
whereas if FRoFD, then the slope is likely to fall. The value of
FR is less where discontinuities exist in a rock mass; these
include bedding planes, cracks, cleavages, joints, and faults.
Some of these are inherent in lithology and others are
tectonic in origin. In some areas, joints or faults are the
dominant factor directing slope processes (mass movement).
If slope materials are subjected to weathering, their mechanical strength decreases and discontinuities such as weathering
cracks or joints develop causing a reduction in FR. However,
no systematic quantitative studies have been made concerning
the weathering-induced deterioration of slope materials, except for a few rocks as mentioned in the following section.
7.4.2.3
Strength Reduction from Weathering
Mudstone and shale are highly susceptible to slaking and

undergo considerable expansion on wetting. This sometimes
initiates microcrack formation, the widening of existing
cracks, or disintegration of the rock mass. In Figure 2, the
maximum expansion strain emax is plotted against the strength
reduction index, Is, which is defined as Is (1Scw/Scd) 100
(%), where Scd and Scw, respectively, denote the unconfined
compressive strength of dry and wet cylindrical specimens.
Slope material
Lithology
Discontinuities
Mechanical strength
Resisting force of cliff material, FR
Driving force, FD
No
Yes
Weathering
and fatigue
No
FD > F R
Yes
Mass movement
Figure 1 Relation between lithological properties and mass movement.
45
46

10
40
35
Water content (%)
Mudstone
Sandstone
Tuff
Expansion strain, max (%)
Landslip zone IV
30
Zone III
25
Zone II b
Based on depth profiles
20
Zone II a
Zone I
15
10
50
100
150
200
250
300
Cu, kN m2
Figure 3 Relation between weathering, water content, and
undrained shear strength of Lias clay. Reproduced with permission
from Chandler, R.J., 1972. Lias clay: weathering processes and their
effect on shear strength. Geotechnique 22, 403431.
0.1
0.01
50
60
70
80
90
100
Strength reduction index, Is (%)

Figure 2 Relation between maximum expansion strain from water
absorption (emax) and strength reduction index (Is). Reproduced from
Sunamura, T., 1992. Geomorphology of Rocky Coasts. Wiley,
Chichester, 302 pp; original data from Takahashi, K., 1976.
Differential erosion on wave-cut bench. Bulletin of the Faculty of
Science and Engineering, Chuo University 19, 253316.
Here, Is 100% corresponds to fragmentation of the specimens when submerged into water, that is, Scw 0. Rock
samples exhibiting larger (expansion) strain will likely
undergo greater strength reduction.
Weathering of overconsolidated clays and mudrocks as an
influence on the stability and morphology of slopes is summarized by Taylor and Cripps (1987), Gerrard (1988), and
Selby (1993). Extensive work on the effect of weathering on
the shear strength of overconsolidated clays has been undertaken by Chandler (1969, 1972) on the Keuper Marl and
Upper Lias clay of England. Some of the results are shown
in Figure 3. In the Upper Lias clay, the average undrained
shear strength of the clay fell from 200 to 63 kN m2 as
weathering proceeded from zone I to zone III material. In
mudrocks from the Carboniferous period, Spears and Taylor
(1972) observed a reduction of 93% in effective cohesion and
a drop in f0 to 261 as a result of weathering. Similar weathering effects have been noted on Oligocene mudrocks in
Japan (Matsukura and Mizuno, 1986). The extent of weathering is progressive, in the following order: slopes 1, 2, and
3. Changes to clay minerals caused by weathering reduce the
values of the residual angle of shearing resistance, f0r , which is
19.41 in slope 1 (steep, less weathered) and 12.19.21 in

slopes 2 and 3 (gentle, much more weathered).
The weathering rates, including reduction in strength from
swelling pressure generated in hydration processes, were
examined in four dated lava domes of porous rhyolite by
Oguchi et al. (1999) and Oguchi and Matsukura (1999). They
concluded that rock strength reduced rapidly in the initial
stage of weathering (the first 20 000 years) and that mechanical strength was reduced by 7090% in the first 40 000
years, although chemical and physical properties changed by
only 1030%.
Berthelin (1983) and Pentecost (1991) suggested that algae
can loosen the cement between grains and that growth of
algae, together with the expansion and contraction of cells
during wetting and drying, loosens the grains of friable
sandstone. Biological activities such as the growth of tree roots
also affect the mechanical strength of slope materials (Selby,
1993). Many of the beneficial effects of trees in stabilizing
slopes may persist for a few years after deforestation because
tree roots decay gradually; about half of the root strength is
lost in 25 years. The decay of roots usually leaves many root
channels so that the soil becomes more permeable and the
water table can respond to storms more rapidly. Any actions
that impede soil drainage, such as diversion of runoff from
forestry roads into gullies or the deposition of spoil into
depressions, tend to cause instability.
7.4.3
7.4.3.1
Sheeting Joints from Unloading (Pressure

Release)
Microsheeting in Granite
Chigira (2001) reported microsheets from unloading in

Hiroshima granite in Japan. Microsheets generally run parallel
to major slope surfaces and dip gently downslope, forming
cataclinal overdip slopes. The cataclinal overdip slope accelerates creep movement of microsheeted granite, which in turn
loosens and disintegrates granite via the widening or
neoformation of cracks. This disintegration is also furthered

by stress release, temperature changes, and changes in the
water content near the ground surface. The surface regions
of the microsheeted granite are thereby loosened with a welldefined basal front that ultimately slides in response to heavy
rain.
7.4.3.2
Exfoliation Rates of Sheeting Joints for

Granite Dome
In the last decade, a technique has been developed for

monitoring near-surface residence times of material, based on
terrestrial in situ cosmogenic radionuclides. This makes it
possible to infer the erosion rates of bare bedrock surfaces.
Wakasa et al. (2006) measured the concentrations of cosmogenic 10Be and 26Al produced in situ at bare bedrock surfaces of
successive sheets developing on a granite dome in Korea

(Figure 4) and calculated the exfoliation rates of sheeting
joints. The exfoliation rate was derived using a simple model
in which the sheeting joints undergo intermittent denudation,
that is, peeling off along the bedrock face. The average exfoliation (erosion) rate from episodic peeling-off processes
was 5.6 cm ka1. They asserted that exfoliation of a sheet occurs episodically and that the frequency of such events depends on the size of the sheet. Freezethaw action and impact
of rockfalls and earthquakes are believed to trigger peeling
from a parent rock mass (e.g., Whalley, 1984; Bull et al., 1994;
Matsuoka and Sakai, 1999).
7.4.4
7.4.4.1
Figure 4 Granite dome of Insubong in Bukhansan, NE of Seoul. The

sheeting has a modification that has developed at the lower slope of
the dome. Reproduced with permission from Wakasa, S., Matsuzaki,
H., Tanaka, Y., Matsukura, Y., 2006. Estimation of episodic exfoliation
rates of rock sheets on a granite dome in Korea based on
cosmogenic nuclide analysis. Earth Surface Processes and Landforms
31, 12461256.
(a)
47
Effect of Slaking on Structural Landforms

and Mass Movement
Effect of Slaking on Rock-Controlled Landforms
Slaking has been identified as a cause of rock-controlled

landforms (differential erosion landforms) such as cuestas
and hoodoos. The significant papers by Suzuki et al. (1970,
1972) and Takahashi (1975) attributed the differential erosion
of wave-cut benches in the intertidal zone along several coasts
in Japan (Figure 5) to the difference in swelling potentials of
shale (mudstone) and tuff or sandstone. The rocky coast of
Arasaki, SW of Tokyo, is underlain by a steeply dipping
rhythmic alteration of mudstone and tuff. Erosion proceeds
differently between the two rocks with altitude. On the
benches, mudstone forms the furrows and tuff forms ridges,
producing a washboard-like relief with a peak-to-trough
height of about 1.6 m. Field and laboratory tests on fresh
samples of both rocks revealed that the mudstone is mechanically more erosion resistant than tuff (Table 1). The
mudstone is stronger than the tuff and suffers less abrasion
loss. However, the mudstone is easily jointed to form fragments of about 1 cm by wetdry slaking, whereas the tuff is
not. The mean joint spacing on the mudstone surface is almost a constant 1 cm above an altitude of 1.5 m; below this
level, it increases because fragments of mudstone formed by
wetdry slaking are frequently and rapidly washed away by
(b)
Figure 5 (a) Panoramic view taken at the low tide: Mudstone forms shallow furrows, and sandstone forms the ridges on the wave-cut benches
producing a washboard-like relief with an amplitude of about 40 cm at Aoshima Island, Nichinan coast in Japan. (b) Closeup view: The fragments
of mudstone produced by wetdry slaking are rapidly washed away by waves in the intertidal zone.
48
Table 1 Physical and mechanical properties of mudstone and tuff from Arasaki coast, Japan
Rock properties
Mudstone
Tuff
Physical
Apparent density
Porosity (%)
Coefficient of thermal linear expansion (106/1C)
Longitudinal wave velocity (km s1)
wet
2.64
39.3
6.6
2.0
2.63
34.0
4.4
1.3
Mechanical
Compressive strength (Mpa)
dry
wet
dry
wet
dry
dry
wet
23.5
1.4
4.3
0.6
7.4
7.2
13.0
17.0
0.8
1.5
0.5
3.0
15.0
33.0
Tensile strength (Mpa)

Shear strength (Mpa)
Abrasion loss by Loss Angeles test (%/100 rounds)
Source: Adapted from Suzuki, T., Takahashi, K., Sunamura, T., Terada, M., 1970. Rock mechanics on the formation of washboard-like relief on wave-cut benches at Arasaki, Miura
Peninsula, Japan. Geographical Review of Japan 43, 211222.
waves. These observations indicate that differential erosion

results not only from differences in lithology and mechanical
strength but also from dissimilarity in the response of rocks
to the distinctive geomorphological processes at various altitudes. Furthermore, weathering of the rock is significant as a
process that facilitates erosion and mass movement. Similar
conclusions arise from studies of microrelief on wave-cut
benches in other areas (e.g., Takahashi, 1975, 1976). Yatsus
(1966, 1971) concept of rock control is clearly substantiated
on the Arasaki coast, based on quantitative data for rock
properties and their response to erosive agents acting on the
rocks.
Tanaka et al. (1996) examined the formation of mushroomshaped rocks known as pedestal rocks or hoodoos in the
Drumheller badlands, Alberta, Canada, by measuring rock
properties in the field and laboratory. They found that the formation of hoodoos is controlled mainly by the susceptibility of
the rock to wetdry slaking, as follows. Hoodoos are composed
of three types of Cretaceous sedimentary rocks, two kinds of
sandstone and siltstone. Each hoodoo was divided morphologically into two parts, the caprock and the pillar (Figure 6).
The morphological characteristics of hoodoos involve differences in rock type: the caprock is made of sandstone with a high
resistance to slaking, and the pillar is made of both sandstone
and siltstone with a lower resistance to slaking (Figure 7);
slaking tests show that the rocks of the pillar disintegrate into
small flakes after only one or two cycles of wetting and drying.
The difference in susceptibility to slaking of these rock types
derives from the combination of pore-size distribution and
clay mineral composition. Rocks of the pillar have high clay
fraction (2030%), a large value of the specific surface area
(1121 m2 g1), and interstratified illite/smectite with large expansion from 12.5 to 17.7 A. The landform development of
hoodoos is speculated to be as shown in Figure 8.
7.4.4.2
Effect of Slaking on Landslides on Mudstone

Hillslopes
Several studies have considered the effect of weathering on the

strength of argillaceous soils in relation to landslide problems
(e.g., Chandler, 1969, 1972; Quigley et al., 1971; Hutchinson
Cf
Pf
Ps
Figure 6 Typical example of hoodoos, which are made of three

types of sedimentary rocks, such as Cf-, Pf-, and Ps-layers: Cf is cap
rock composed of fine sandstone, Pf is pillar rock composed of fine
sandstone, and Ps is pillar rock composed of alternating white and
brown siltstone thin beds. Reproduced from Tanaka, Y., Hachinohe,
S., Matsukura, Y., 1996. The influence of slaking susceptibility of
rocks on the formation of hoodoos in Drumheller Badlands, Alberta,
Canada. Transactions of the Japanese Geomorphological Union 17,
107121.
and Gostelow, 1976; Carson and Tam, 1977; Matsukura and

Mizuno, 1986; Enriquez-Reyes et al., 1993; Bachmann et al.,
2004; Wakatsuki et al., 2009). For instance, Wakatsuki et al.
(2009) examined many slope failures caused by Typhoon 10 of
2003 in northern Japan. Two types of failures occurred on
mudstone slopes. The first type is smaller slope failures

(soil slips) with a slip depth of 35120 cm on slopes underlain
by Miocene tuffaceous mudstone (Tm). The second type involves larger slope failures with a slip depth of 320370 cm on
slopes underlain by Cretaceous mudstone and siltstone (MSi).
A laboratory experiment involving wetdry weathering shows
that Tm rock disintegrates into finer materials than MSi rock.
X-ray diffraction indicates that the alteration of minerals from
rock to soil is not clear, so that chemical weathering is weak on
both slopes. These weathering characteristics imply the following mechanism for slope failures on the Tm slopes. The
100
Percentage remaining
90
80
70
Cf - 1
Cf - 2
Pf - 1
Pf - 2
Ps - 1
(Ps-b rich)
Ps - 2
(Ps-w rich)
60
50
40
30
20
10
finer slope materials are produced on the surface of the slopes,

reflecting the grain size of the slaking materials. These finer
materials result in a lower permeability of the soil layer that
restricts the downward movement of soil water and the deeper
front of wetdry weathering so that the soil layer does not
become thicker. In this slope, rising of the groundwater surface
toward the slope surface during heavy rainfall causes soil slips.
By contrast, in the MSi slope, the greater permeability of the soil
layer resulting from coarser materials promotes the downward
movement of soil water and wetdry weathering so that the soil
layer becomes thicker. Piping arising from the deep water
table triggers the larger slope failures.
Chang et al. (1996) examined the formation of southfacing bald mudstone slopes in southwestern Taiwan. When
subjected to water, slaking of the mudstone can easily take
place, and severe shrinkage can occur in dry seasons in Taiwan.
Heavily eroded natural mudstone slopes frequently become
bald (not covered by any plants or bushes) and are very susceptible to damage such as mudflow and slope failure. These
bald slopes are mostly south facing. Solar radiation, rather
than the direction of typhoons, is the main reason for the
formation of bald slopes that face south.
7.4.5
0
0
5
10
15
Number of wetting-drying cycles
20
Figure 7 Disintegration curves as a result of slaking tests for the

samples. Cf, Pf, and Ps, respectively, denote caprock fine sandstone,
pillar fine sandstone, and pillar siltstone. Reproduced from Tanaka,
Y., Hachinohe, S., Matsukura, Y., 1996. The influence of slaking
susceptibility of rocks on the formation of hoodoos in Drumheller
Badlands, Alberta, Canada. Transactions of the Japanese
Geomorphological Union 17, 107121.
7.4.5.1
Effect of Crystal Growth Weathering (Salt

Fretting and Frost Shattering) on Landforms
and Mass Movement
Effect of Frost Weathering on Periglacial
Landforms
Many monographs and books have considered the influence

of frost weathering on periglacial landforms (e.g., Washburn,
1979; Chorley et al., 1984; French, 2007). Periglacial landforms are treated in detail in Volume 8 of this treatise, so only
Joint
Joint
49
Gully erosion
Cf
Gully erosion
Pf
Ps
(b)
(a)
Slaking
Slaking
Cf
Fallen block
of Cf
Pf
Ps
(c)
(d)
Figure 8 Formation of hoodoos in Drumheller badlands. (a) and (b) Stage 1: Cf and Pf layers are sculptured into a rectangular pattern of
gullies. (c) Stage 2: Slopes composed of Pf and Ps are eroded laterally by slaking more severely than those of Cf. Flakes made by slaking peel
away through the action of rain or wind. (d) Stage 3: Progression of lateral erosion of pillars makes Hoodoos. Reproduced from Tanaka, Y.,
Hachinohe, S., Matsukura, Y., 1996. The influence of slaking susceptibility of rocks on the formation of hoodoos in Drumheller Badlands, Alberta,
Canada. Transactions of the Japanese Geomorphological Union 17, 107121.
50
a short summary of the influence of frost weathering on

periglacial landforms is given in this section.
Frost weathering gives rise to accumulations of angular
fragments, with forms that may be rod- and plate-like if rocks
have a notable preferential aligned fabric. Shattering is confined mainly to rock outcrops, and a regolith greatly reduces
its effectiveness. In cold deserts or alpine areas with no regoliths, the entire ground surface may shatter to produce a
surface of broken rock fragments known as a felsenmeer (from
German for rock sea).
Two principal ways exist in which ice in periglacial areas can
modify the landscape. The first is through landforms associated
with permafrost, or ground that is continuously below 0 1C in
winter and summer. The second is through landforms associated
with freezing and thawing on an annual or shorter-term basis.
A number of landforms are associated with the growth of segregated ice in the ground. Mounds are distinctive and are found
in many sizes and forms. These mounds are the result of the
growth of closed system ice lenses from water bodies, or opensystem hydrolaccoliths from ground water, that lift the ground
above them as they grow (e.g., Mackay, 1982). Some are only a
few meters across; but the better-known type, the pingo, may be
as much as 300 m in diameter and 60 m high.
Much more common in terms of areal coverage are ice-wedge
polygons that are responsible for a mosaic surface pattern in
extensive low-lying areas of the Arctic. The polygons are typically
1540 m across and have three to seven sides, which is consistent with a random pattern. Their sides are uplifted and separated
by a narrow ditch. Ice-wedge polygons are part of a family of
contraction-crack polygons. In arid areas, the cracks may be filled
not with water and ice but with sand or other debris.
A further distinctive active-layer landform is a sorted, patterned ground. Again, the mechanisms are far from understood, but apparently coarse stones move to the surface and
outward from freezing centers. If these centers are scattered
widely, stone circles form. If they impinge on one another,
then polygons of coarse stones form around the centers;
generally, they are 110 m in diameter, and the larger sizes
involve larger constituent stones. If sorting takes place on a
slope, the polygons are elongated downslope until at gradients above 4111 they form stone stripes.
Other characteristic landforms are associated with the flow
of regolith during the summer thaw. In permafrost areas, this
process is known as gelifluction (an icebound version of
solifluction). The word describes the slow flow downslope of
saturated regolith, as well as downslope displacement from
repeated freezing and thawing of particles on a slope. In
general, gelifluction takes place at a rate of 15 cm yr1 and
can occur on slopes as gentle as 11.
7.4.5.2
Effect of Frost Weathering on Rockfall and Talus

Slope Development
Rocks exposed to coldhumid climates are subject to frost

weathering, which promotes debris production, rockwall retreat, and talus development in mountain areas (e.g., Rapp,
1960a, 1960b). The mechanics of rockfall and its rate are
discussed in detail by Carson and Kirkby (1972) and Whalley
(1984). Extensive quantitative measurement of rockfall rates
has been made by Rapp (1960a, 1960b) in Spitsbergen and
northern Scandinavia. Between 1952 and 1960, by putting

wire nets over talus slopes, Rapp concluded that rockfall in
Karkevagge was responsible for the loss of 50 m3 yr1 of debris
from the rockwalls of the area studied. Recent progress in
monitoring technology and automatic data logging enables
the automated monitoring of rock temperature and the timing
and magnitude of joint widening (frost wedging) in the field
(e.g., Matsuoka and Sakai, 1999; Matsuoka, 2001). Statham
and Francis (1986) reviewed the development of scree
slopes and considered three aspects: (1) debris input from
the headwall, (2) scree slope accumulation and form, and
(3) subsequent long-term behavior through weathering.
7.4.5.3
Effect of Salt Weathering on Tafoni and Pan

Formation
Salt weathering is responsible for cavity formation, such as

tafoni and pans or tanks, in dry and/or coastal environments
(e.g., Goudie and Viles, 1997). Several studies using scanning
microscopy and X-ray diffraction analyses (e.g., Gill, 1981;
McGreevy, 1985; Matsukura and Matsuoka, 1991; Matsukura
and Tanaka, 2000) have detected salts in the hollowed rock
surface.
Although coastal tafoni with a variety of sizes have been
reported from many locations and on a variety of rocks in the
world (e.g., as compiled by Sunamura (1992)), we do not
have systematic information about the rate of tafoni development or the effect of rock properties on tafoni development.
Matsukura and Matsuoka (1996) investigated the rates of
tafoni growth in coastal environments at 24 study sites in
Japan, at which many tafoni of varying sizes occur on the
face of marine cliffs on uplifted shore platforms of various
rock types (Figure 9). The rate of tafoni growth was defined
as the mean value of the 10 largest depths of 30 tafoni at
each site divided by the period of their growth; that is,
from the time of emergence of the shore platform to the
present. The salt crystallization hypothesis indicates that rock
properties such as pore-size distribution and rock strength
influence the tafoni growth rate. Comparison between the rate
of tafoni growth and the measured rock properties using intact
rock samples collected from the adjacent shore platforms
reveals that tafoni growth is more rapid where rocks have a
large proportion of micropores and/or lower tensile strength
(Figure 10).
7.4.5.4
Effect of Crystal Growth Weathering on Notch

Formation
Matsukura and a co-worker studied the southwestern slope of

Asama volcano in Japan and concluded that notch formation
is different at the foot of south-facing and north-facing cliffs
(Matsukura and Kanai, 1988; Matsukura, 1990, 1991). In this
area, a number of dissected valleys cut into pumice flow deposits. These valleys have a rectangular profile, with a flat
bottom 10400 m wide and vertical cliff walls varying from 5
to 15 m high. Many valleys run from east to west, and all have
north-facing and south-facing vertical cliffs (Figure 11(a)).
Both cliff profiles have a cave-like notch at the cliff base about
1.5 m above the valley floor (Figure 11(b)). The zone of notch
Rate of tafonil growth, Dd /T (mm yr1)
51
1.0
7
2,3,4
0.5
18
8,9
0.1
0.05
20.21
10,11,12
13
15
24
5
Dd /T = 0.130 WSI0.648
19
(R2 = 0.848)
16
1
6
22
23
0.01
14
0.005
0.01
0.05 0.1
0.5 1
WSI, P/S t
10
Figure 10 Relationships between the weathering susceptibility index

and the rate of tafoni growth at Japanese coasts. Reproduced with
permission from Matsukura, Y., Matsuoka, N., 1996. The effect of
rock properties on rates of tafoni growth in coastal environments.
Zeitschrift fur Geomorphologie NF Supplement Bd 106, 5772 (http:
www.borntraeger-cramer.de).
Figure 9 Features of tafoni at several sites: (a) tafoni having a

depth of 40 mm develop on the surface of the sandstone shore
platforms; (b) tafoni having depth of 137 mm develop on the vertical
cliff made of tuff; and (c) tafoni having depth of 124 mm develop on
the cliff behind the conglomerate shore platform. Reproduced with
permission from Matsukura, Y., Matsuoka, N., 1996. The effect of
rock properties on rates of tafoni growth in coastal environments.
Zeitschrift fur Geomorphologie NF Supplement Bd 106, 5772 (http://
www.borntraeger-cramer.de).
formation coincides with the zone of high water content

derived from capillary action.
In early spring, many rinds with white powder efflorescence are observed on the surface of the notch in southfacing cliffs (Figure 12). These rinds are broken up by cracks
producing a polygonal pattern and peel away through the

action of wind and earthquakes. X-ray diffraction analysis
shows that this white powder comprises salts including halite
(NaCl), gypsum (CaSO4 2H2O), hexahydrite (MgSO4 6H2O),
and mirabilite (Na2SO4 10H2O). Groundwater quality and
meteorological data were analyzed, and the ground temperature was measured. The groundwater contains high concentrations of anions such as Cl and SO2
4 and cations including
Ca2 , Mg2 , and Na . On fine days in winter and spring, the
sunshine impinges on the south-facing cliff, and the temperature of the cliff surface becomes considerably higher than
the air temperature. From March to May, the weather is dry
with least mean minimum humidity. These conditions induce
salt weathering and consequent notch growth at the base of
the south-facing cliffs.
For the north-facing cliff, no salt fretting occurs. Frost
weathering is frequently observed on the notch surface in
winter. The mean air temperature is below freezing in winter
in this area. Fine days dominant in winter, but no direct
sunshine on the north-facing cliff, which therefore receives
little energy as radiation. These conditions gradually cool
the inner zone of the notch and generate a frozen layer that
grows parallel to the notch surface. The maximum frost
penetration depth in winter is estimated to be 2030 cm. Exfoliation occurs in spring as a result of thawing at the
boundary between frozen and unfrozen layers. Repetition of
freezethaw cycles over several years gradually reduces the
strength of the cliff materials, and the consequent fatigue
failure induces exfoliation.
7.4.5.5
Effect of Frost Action and Notch Formation on

Cliff Collapse
Rockfalls having volume up to 1 000 000 m3 occur rarely at

coastal cliffs in Britain, chiefly in Chalk, and also in the Upper
and Lower Greensand. The highest Chalk cliffs in Britain are
52
Joint-controlled
tension crack
Former cliff profile
Reach R
H
Shear surface
Debris
Notch
0.5-m deep c.35
+ 2.3 MHWS
Beach
Mean sea
level
(a)
0
Wave-cut platform
5m
2.1 MLWS
Figure 13 Cross section of the chalk fall in early 1966 at Joss Bay,
Kent, southern England. Reproduced from Hutchinson, J.N., 1972.
Field and laboratory studies of a fall in Upper Chalk cliffs at Joss
Bay, Isle of Thanet. In: Parry, R.H.G. (Ed.), Stress Strain Behaviour of
Soils. Proceedings of the Roscoe Memorial Symposium. G.T. Foulis,
Henley-on- Thames, pp. 692706, and Hutchinson, J.N., 1984.
Landslides in Britain and their countermeasures. Journal of Japan
Landslide Society 21, 125.
Figure 11 (a) Mt. Asama and rectangular profile of valley with flat
bottom. Insolation is directly on the south-facing cliff (the left-hand
side) and does not impinge on the north-facing cliff (right-hand side).
(b) Vertical cliff and notch at the cliff base. Reproduced from
Matsukura, Y., 1991. Notch formation processes and cliff instability
in pumice flow deposits on the Asama Mountain slopes, Japan.
Science Report of the Institute of Geoscience, University of Tsukuba,
vol. 12, pp. 3763.
Days of air frost (195665)
0
10
Number per month
(b)
Average number
per month
10
Dated chalk falls

(18101970)
Figure 12 Salt efflorescence and polygonal rind occurring on the

notch surface of the south-facing cliff. Reproduced with permission
from Matsukura, Y., Kanai, H., 1988. Salt fretting in the valley cliff of
the Asama volcano region, Japan. Earth Surface Processes and
Landforms 13, 8590.
0
Water surplus
50
June
May
April
March
February
January
December
October
November
September
Water deficiency
August
70
0
July
Total rainfall-PE (mm)

(PE = potential evapo-transpiration)
0
70
Effective rainfall
50 (191650)
Figure 14 Incidence of chalk falls, frost, and effective rainfall on the

Kent coast. Reproduced from Hutchinson, J.N., 1972. Field and
laboratory studies of a fall in Upper Chalk cliffs at Joss Bay, Isle of
Thanet. In: Parry, R.H.G. (Ed.), Stress Strain Behaviour of Soils.
Proceedings of the Roscoe Memorial Symposium. G.T. Foulis,
Henley-on-Thames, pp. 692706.
53
(m)
100 km
140
42
Open fracture
Toyohama
80
Unit III
Sapporo
40
142
144
60
Unit II
A
40
B
C
Unit I
20
Notch
Toyohama
Tunnel
0
(a)
(b)
Figure 15 (a) Cross section of cliff before the 1996 collapse at the western entrance to the Toyohama tunnel. (b) Collapse of the cliff at the
western entrance to the Toyohama tunnel. Blocks A, B, and C fell on 10 February 1996. Reproduced with permission from Matsukura, Y., 2001.
Rockfall at Toyohama tunnel, Japan, in 1996: effect of notch growth on instability of a coastal cliff. Bulletin of Engineering Geology and the
170-m high at Beachy Head in Sussex; failure modes can be

studied more easily and safely in the lower cliffs. In falls at Joss
Bay, Kent, in 15-m-high cliffs subject to strong toe erosion by
the sea, failures were preceded by the gradual opening of a
tension crack down preexisting joints to about the mid-height
of the cliff. This led to a shear failure at a high angle
(c. 451 f/2) through the root of the mass (Figure 13). The
process was obviously assisted by the formation of a wave-cut
notch about 0.5 m into the foot of the cliff (Hutchinson,
1972, 1984). In Figure 14, the incidence of these falls is
compared with the average monthly effective rainfall (defined
as the total precipitation minus potential evapotranspiration,
the latter estimated by HL Penmans method) and the average
number of days of air frost per month. The correlation between these climate factors and the incidence of falls is close,
and the falls are concentrated into the winter months of
October to April. Rainfall has the greater influence during the
earlier part of this period, and frost later on.
In Hokkaido, Japan, a high coastal cliff collapsed suddenly
at 8:08 a.m. on 10 February 1996 onto the western entrance of
the Toyohama tunnel (Figure 15). The collapsed rock which
was about 6 m high, 54 m wide, and 13 m deep destroyed
about 45 m of the tunnel. A bus and car passing the point of
collapse were crushed, and 20 people (including both drivers)
were killed. The cause of this collapse has been investigated
from the viewpoint of engineering geology. Many studies
proposed that hydraulic pressure and/or frost shattering induced tension cracking behind the cliff, triggering the collapse.
Some studies (e.g., Matsukura, 2001) have indicated that a
notch, formed at the base of the cliff, could increase the instability of the cliff.
A number of failures can be observed in the cliff on
the southern slopes of Asama volcano (considered in Section 7.4.5.4). These failures occur because of notch growth.
Field observations indicate that the process of notch growth
and consequent slope instability play a crucial role in geomorphological processes in this area. Cliff failures from notch
growth take place sporadically at both walls in the valleys.
Fresh failure-blocks and small scree slopes piled with these
failure materials commonly occur at the cliff base. The breadth
of the fallen blocks was measured, and also their physical
and strength properties. Two stability analyses were made:
for beam failure and for slab failures (Matsukura, 1988).
Comparison of the results with field data provides a useful
framework for predicting the critical depths of a notch.
7.4.6
Conclusion
Weathering is the basic process in landform evolution and

changes to hillslope form. Recent experimental work and
microscope investigations using new analytical methods have
advanced the study of weathering processes. Advance has,
however, been slower in exposing the detailed relation between
weathering and landform evolution and changes in hillslope
form, although symposia have recently taken place on the
theme of Weathering and landscape evolution and Weathering and slope movements (e.g., Turkington et al., 2005;
54
Ollier et al., 2007). To shed light on the detailed relations

between weathering and the formation and evolution of
landforms, geomorphologists must accumulate data on the
rate and type of changes in rock properties from weathering
(in particular, strength reduction).
References
Bachmann, D., Bouissou, S., Chemenda, A., 2004. Influence of weathering
and pre-existing large scale fractures on gravitational slope failure: insights
from 3-D physical modeling. Natural Hazards and Earth System Sciences 4,
711717.
Berthelin, J., 1983. Microbial weathering processes. In: Krumbein, W.E. (Ed.),
Microbial Geochemistry. Blackwell, Oxford and London, pp. 223262.
Brunsden, D., 1976. Weathering. In: Embleton, C., Thornes, J. (Eds.), Process in
Geomorphology. Edward Arnold, London, pp. 73129.
Bull, W.B., King, J., Kong, F., Moutoux, T., Phillips, W.M., 1994. Lichen dating of
coseismic landslide hazards in alpine mountains. Geomorphology 10, 253264.
University Press, London, 475 pp.
Carson, M.A., Tam, S.W., 1977. The land conservation conundrum of eastern
Barbados. Annals of the Associations of American Geographers 67, 185203.
Chandler, R.J., 1969. The effect of weathering on the shear strength properties

Treatise On Geomorphology - John F. Shroder (AP, 2013)

Diunggah oleh

Informasi Dokumen

Judul Asli

Hak Cipta

Format Tersedia

Bagikan dokumen Ini

Bagikan atau Tanam Dokumen

Opsi Berbagi

Apakah menurut Anda dokumen ini bermanfaat?

Apakah konten ini tidak pantas?

Hak Cipta:

Format Tersedia

Treatise On Geomorphology - John F. Shroder (AP, 2013)

Diunggah oleh

Hak Cipta:

Format Tersedia

2.

1 Quantitative Modeling of Geomorphology

A model can be a many great things: a concept, a hypothesis,

Treatise on Geomorphology, Volume 2

indeed paradigms, are of this kind, such as Daviss (1899)

Quantitative Modeling of Geomorphology

taxonomy, however, was instead developed within a general

versus dynamic: models that represent equilibrium

relatively rare: for example, stochastic, form-based dynamic

Quantitative Modeling of Geomorphology

like computer programming contests, where the goal is to

Structure of this Volume

The theme of this volume is on quantitative modeling of

Quantitative Modeling of Geomorphology

automated computational methods that mimic Darwinian

modeling topographic development of whole river basins and

Quantitative Modeling of Geomorphology

greatly improved by the valuable contributions of a large

Hooke, R.L., 2003. Predictive modelling in geomorphology: an oxymoron?

3.1 Remote Sensing and GIScience in Geomorphology: Introduction

Bishop, M.P., 2013. Remote sensing and GIScience in geomorphology:

Treatise on Geomorphology, Volume 3

Remote Sensing and GIScience in Geomorphology: Introduction and Overview

Geomorphometry The discipline that is concerned with

Object-oriented analysis Object-oriented analysis refers

Remote Sensing and GIScience in Geomorphology: Introduction and Overview

Earth science investigations are increasingly utilizing geospatial

From a more methodological perspective, advances in

Geospatial Technology and Fieldwork

Remote Sensing and GIScience in Geomorphology: Introduction and Overview

lithological systems. Telecommunications and computers

Figure 1 Terrestrial laser scanning at the Chalk Cliffs Natural Debris

Remote Sensing and GIScience in Geomorphology: Introduction and Overview

Deployment of terrain rovers with sensors is also being

Remote Sensing and Geomorphology

Photography and Videography

Ground photography has historically played a significant role

Remote Sensing and GIScience in Geomorphology: Introduction and Overview

Figure 3 Ground photograph of the north-facing Braldu Valley in

our ability to comprehensively document the landscape as

Optical imaging sensors record the magnitude of reflected

Remote Sensing and GIScience in Geomorphology: Introduction and Overview

The spatial resolution of a sensor is usually referred to as

DNl intgles ol;

where lc is the wavelength corresponding to the peak

where L represents the average radiance value and L represents

where es is the electronic signal, g is the electronic gain, and o

Reflectance properties and applications

The reflectance properties of surface features are based on their

Remote Sensing and GIScience in Geomorphology: Introduction and Overview

Spectral response weight (w)

Survey Spectral Library and the ASTER spectral library can be

Remote Sensing and GIScience in Geomorphology: Introduction and Overview

AVIRIS FWHM (nM)

AVIRIS wavelength center (m)

Imaging spectroscopy is also used in vegetation science and

Remote Sensing and GIScience in Geomorphology: Introduction and Overview

Each phase of water exhibits absorption features between

Microwave Remote Sensing

Unlike optical-based sensors that depend on reflected and

This treatment of the topic should provide readers with an

The Atmosphere and Climate Forcing

The linkages between climate and surface processes have long