Anda di halaman 1dari 10

Impact Assessment and Project Appraisal

ISSN: 1461-5517 (Print) 1471-5465 (Online) Journal homepage: http://www.tandfonline.com/loi/tiap20

Use of geographic information systems in an


environmental impact assessment of an overhead
power line

L. L. Warner & R. D. Diab

To cite this article: L. L. Warner & R. D. Diab (2002) Use of geographic information systems in
an environmental impact assessment of an overhead power line, Impact Assessment and Project
Appraisal, 20:1, 39-47, DOI: 10.3152/147154602781766807

To link to this article: http://dx.doi.org/10.3152/147154602781766807

Published online: 20 Feb 2012.

Submit your article to this journal

Article views: 1073

View related articles

Citing articles: 5 View citing articles

Full Terms & Conditions of access and use can be found at


http://www.tandfonline.com/action/journalInformation?journalCode=tiap20

Download by: [5.254.65.186] Date: 17 July 2017, At: 03:12


Impact Assessment and Project Appraisal , volume 20, number 1, March 2002, pages 3947, Beech Tree Publishing, 10 Watford Close, Guildford, Surrey GU1 2EP, UK

GIS analysis

Use of geographic information systems in an


environmental impact assessment of an
overhead power line

L L Warner and R D Diab

T
Can geographic information systems (GIS) help HE ASSESSMENT OF IMPACTS on the
in power line route selection in the ecologically environment demands the analysis of large,
sensitive and socially complex area of Kranskop complex data sets of a multi-disciplinary na-
in KwaZulu-Natal, South Africa? Transforming ture. Geographic information systems (GIS) have the
data into digital form and preparing layers for ability to store, manipulate, analyse and display
the GIS was time-consuming. Weights were as- large sets of complex, geographically referenced
data and are therefore well suited to spatial applic a-
signed to layers using a pairwise comparison ma- tions of this nature and complexity (Maquire et al,
trix, and combined into a single composite cost 1991).
surface using the multi-criteria evaluation mod- Many advantages of employing GIS in environ-
ule of the IDRISI GIS package, whose auto- mental impact assessments (EIAs) have been noted
mated route selector was unsuccessful, so (for example, Eedy, 1995; Antunes et al, 1996; Joao
computer-assisted methods were used. Cost indi- and Fonseca, 1996). Some of the attributes listed are
ces, area analysis and impacts of routes were the ability to handle large data sets in a spatial
readily compared. Changing weights had no ma- framework and the facility for cartographic display,
jor impact. The views of interested and affected which is a powerful means of conveying results in a
parties influenced our results; how can this best way that is easily interpreted (Cassettari, 1993).
be achieved in future? Positive attributes of GIS Layers of data can be combined or eliminated. Areas
are its interactive nature and its overlaying and of environmental constraint are easily identified and
can be assigned different weights. Buffers can be
displaying selected information at relevant times. placed around intrusions, such as power lines or ac-
cess roads, or around sensitive areas, such as nesting
Keywords: environmental impact assessment; geographic habitats.
information systems; power lines The GIS is dynamic and continuous updating of
information over time and space allows for planning
L L Warner and R D Diab are at the School of Life and Envi- flexibility, auditing and cumulative impact assess-
ronmental Sciences, University of Natal, Durban 4041, South ment as new developments take place. Different
Africa; Tel: +27 31 260-2300; Fax: +27 31 260-1391; E-mail: scenarios may be generated with minimal effort and
diab@nu.ac.za. time to provide multiple choices and options for de-
The Human Sciences Research Council in Durban is ac-
knowledged for the use of their facilities and for assistance in
cision-makers. Evaluation of impacts is available
digitising data. Mr Frank Sokolic of the School of Life and through methods such as area analysis and three-
Environmental Sciences, University of Natal, Durban is thanked dimensional viewshed analysis.
for writing the customised software. Balanced against these positive attributes, there

Impact Assessment and Project Appraisal March 2002 1461-5517/02/010039-9 US$08.00 IAIA 2002 39
Use of GIS in EIA of power line

Box 1. Glossary of terms

Cadastral
A cadastral is a register of property and a cadastral survey is
one in which property is surveyed. The word cadastral is also
loosely used to describe large-scale mapping in which prop-
erty divisions are shown (Monkhouse, 1970).

Orthophoto
An orthophoto is a map that contains an aerial photograph
which has been geo-referenced and corrected for distortions
resulting from topographic relief, camera lens characteristics
and camera attitude.

Raster
A raster data structure is one in which spatial data are repre-
sented by a grid of uniform cells, each of which contains a
value or values pertaining to the area covered by that cell.

Servitude
A strip of land of varying width that buffers a power line.

Viewshed
A viewshed shows all areas that are visible from a particular
point or series of points.

are a number of problems associated with the use of Figure 1. Location map of the study area in KwaZulu-Natal
GIS. Reasons for its capabilities not being fully util-
ised were noted by Joao and Fonseca (1996) and
include the time taken to prepare the database, the and social characteristics of the study area; to outline
initial purchase costs of hardware and software and the approach adopted in the GIS application; to
the data collection and conversion (to digital format) assess the results of the GIS application and to
costs. Other limiting factors are the level of technical highlight any problems encountered; and to make
expertise required to operate the system, the general recommendations for the future use of GIS in envi-
lack of availability of data in a digital format and the ronmental assessments.
difficulty of integrating layers of information pos-
sessing different resolutions and accuracy.
According to Antunes et al (1996), GIS is widely Background to the case study
used for the presentation of baseline environmental
information and to a lesser extent for impact evalua- The case study selected is the routing of a 132 kV
tion. Joao and Fonseca (1996), in a survey on the use overhead power line in the Kranskop area of the
of GIS by EIA practitioners, found that it was not province of KwaZulu-Natal on the east coast of
restricted to the production of maps, but was also South Africa. The component of the project relevant
being used for analysis and modelling. However, to this paper is the routing between Hermannsburg
they also noted that GIS was not used in the public and Middledrift, via a substation at Kranskop (Fig-
participation phase. It is often employed to convey ure 1). The route was considered in two parts: Sec-
the results of an EIA to the public, but evidence was tion I between Hermannsburg and Kranskop; and
not found for the use of GIS to incorporate the pub- Section II, from Kranskop to Middledrift, which is
lics point of view during data preparation and in the situated on the Tugela River.
analysis or modelling phases. The need for the power line arose from the
The purpose of this paper is to explore the viabil-
ity of using GIS as a tool in EIA. It was used in a
post-EIA study, in which one of the objectives was
to compare the optimum power line route as identi-
fied by the GIS method with that recommended by The need for the power line arose from
the original EIA. The EIA case study chosen as a the national Department of Water
basis for this investigation is the routing of an over-
head power line in an area where a large variety of Affairs and Forestrys envisaged
biophysical and socially sensitive characteristics scheme to pump water from the
exist which present an enormous challenge in the Tugela River to the urban-industrial
identification of a suitable power line route. GIS is
used to aid the identification of potential power line node of Richards Bay to forestall a
routes and then the selection of an optimum route. critical shortage of water in the area
Specific objectives of this paper are: to provide a
brief overview of the project and the environmental

40 Impact Assessment and Project Appraisal March 2002


Use of GIS in EIA of power line

national Department of Water Affairs and Forestrys This forest type has a high conservation status and
envisaged water pumping scheme (Mhlathuze Water further fragmentation, which is a major cause of
Augmentation Scheme), which is designed to pump species decline, is undesirable.
water from the Tugela River to the urban-industrial In the Tugela valley, a number of plant communi-
node of Richards Bay (Figure 1) to forestall a critical ties exist, some of which are in a relatively natural
shortage of water in the industria l area. For the pur- state, whilst others are severely disturbed. Of major
poses of this paper, the background and justification interest in this area is a colony of Cape Griffon vul-
for selecting the Hermannsburg substation as the tures, which use the steep cliff faces of the escarp-
starting point for the provision of power are not ment, in the vicinity of The Kop and Ntunjambili, as
relevant. The reasons were fully discussed in the nesting and roosting sites (see Figure 2).
EIA report by Diab et al (1995). The main land use on the plateau is agriculture
The diversity of environmental and social factors (maize, sugar cane and some vegetables) and silv i-
that had to be considered in this particular EIA made culture, with some natural grasslands to the north
it ideal as a case study to test the application of GIS. and west of Kranskop. Most of the farms are either
Topography represents a major constraint, as a steep privately or company owned. Issues raised by the 26
escarpment (Figure 2) marks the interface between landowners on the plateau related to objections to
the high-lying, gently undulating plateau (>1100 m) additional power lines traversing their property on
and the broken, rugged and inaccessible terrain of the grounds of loss of productive farmland, interfer-
the Tugela valley. ence with aerial crop spraying and overhead irriga-
In terms of hydrology, the upland plateau is char- tion systems, and dissatisfaction with compensation.
acterised by a series of shallow water courses, which They were also concerned about negative impacts on
give rise to a number of wetlands. Many of these are wattled crane and vulture populations and on sites of
being rehabilitated by local farmers in the hope of historical value and tourism potential.
attracting waterfowl. Sightings of the endangered East of Kranskop towards the escarpment and be-
wattled crane have been recorded in the area. low, towards the Tugela River, the land is under the
Rivers draining off the escarpment towards the jurisdiction of the Bhengu Tribal Authority. It con-
Tugela River are in steep-sided valleys that descend sists chiefly of scattered rural homesteads, with the
steeply in their upper courses. On the plateau, rivers exception of the peri-urban settlement of Ntunja m-
are gently sloping and shallow, with seasonal flood- bili, which is densely settled and includes a mission
ing of the many associated wetlands. Soils on the station and schools. The prime concern of the local
plateau are hydromorphic, whereas, closer to the tribal inkosi (chief) and community leaders was that
escarpment and in the Tugela valley, they are shal- the project should benefit the people of the area. The
low and highly erodible, posing a problem for the exact routing of the power line was of secondary
construction of access roads. importance, although some concern was expressed
Much of the natural vegetation on the plateau has about negative impacts on indigenous forests along
been modified by agricultural activity, but in the some routes.
north-east, areas of natural grassland (sourveld ) do Below the escarpment, there is little agricultural
exist. Patches of forest, transitional between temper- activity except for limited subsistence farming, and
ate and coastal, exist close to the escarpment. Being natural vegetation (disturbed natural bushveld con-
transitional, the forest is particularly species rich and sisting of trees, shrubs and grasses) occurs through-
contains a number of threatened endemic species. out. Tourism development in the area is limited at
Epiphytic orchids, classed as rare, are in evidence. present, but its scenic beauty, particularly The Kop,

Figure 2. Study area showing positions of substations, roads, rivers and escarpment

Impact Assessment and Project Appraisal March 2002 41


Use of GIS in EIA of power line

affords great opportunities. A number of sites of his- smaller data elements to be mapped or represented,
torical (Anglo-Zulu war of 1879) and archaeological this grid size was deemed adequate. Furthermore, as
(early Iron Age) interest are located in the area. the actual built power line may deviate marginally
All these factors, together with the views of inter- from the recommended route, a smaller grid size
ested and affected parties (I&APs) mentioned above, seemed unnecessary.
were considered in the original EIA and in the route Digital cadastral data and digital elevation data at
selection process (Diab et al, 1995). The intention of 200m intervals were obtained from the Surveyor
this paper is to incorporate these factors into a GIS, General in Pietermaritzburg and Mowbray respec-
which can then be used as an objective tool in the tively. The standard module, SLOPE in IDRISI, was
route-selection process. then used to calculate the slope between each adja-
cent grid square. Areas with a slope greater than a
critical value could be readily identified and ex-
Data and methodology cluded from the consideration of possible power line
routes.
The GIS package IDRISI (version 1.0) was selected Data were assembled into layers; these were then
for the analysis, since it is a raster-based GIS, which combined into a single composite map or surface,
is suited to overlay mapping and the analysis of con- which is a measure of suitability, using the module
tinuous spatial data. The package can handle the multi-criteria evaluation (MCE) in IDRISI.
volume of data to be processed, point data can be
accommodated and viewshed analysis is available. Creating layers
Functions that facilitate the combination of weighted
layers, automatic route selection and area analysis Four steps were followed to create the layers. First
are also included in the standard package. was the assignment of weights to each factor (layer),
Land-use cover was digitised using a combination according to its relative importance. It is highly
of 1:10,000 aerial photographs and site visits. Aerial likely that factors have different degrees of impor-
photographs of varying scales were used to verify tance and this should be reflected in their relative
boundaries that enclose the different land uses, as weights in the composite map of suitability. Impor-
well as the location of individual rural homesteads. tance can be assigned qualitatively and this can be
Rivers and roads were captured from 1:10,000 or- subjective, as is illustrated in the route-selection
thophotos and served as a backdrop for the transfer method developed by Sankoh et al (1993), where
and accurate mapping of data. The geographical lo- protection worthiness is classified as very high (3),
cations of vulture breeding and roost sites, wetlands, intermediate (2) and low (1).
historical forts and archaeological sites (full details Alternatively, values can be assigned subjectively,
given in Diab et al, 1995) were assembled from a yet quantified in a way that is transparent, can be
variety of mapped sources and included on the geo- substantiated and is available for debate and possible
referenced land-use maps. amendment. Such a method is the analytical hierar-
A 25m grid size was used for the rasterised data. chy process of Saaty (1977), which uses a pairwise
This choice was influenced by the minimum width comparison matrix. This forms the basis of the
of 70m for important wetland and indigenous forest IDRISI module WEIGHT. It uses a nine-point con-
areas and would ensure that these small data ele- tinuous rating scale, whereby criteria are rated
ments would be represented. As there were no against each other by means of a pairwise comparison

Table 1. Pairwise comparison table and resulting weights for factors combined to produce a single suitability surface for power
line routing

Rural Access Cultural/ Cadastral Topography Aesthetics Land Environmental Weight


homesteads historical boundaries use

Rural homesteads 1 0.0192


Access 3 1 0.0274
Cultural/ historical 3 3 1 0.0426
Cadastral 5 3 3 1 0.0637
boundaries
Topography 7 5 3 3 1 0.1038
Aesthetics 7 7 5 5 3 1 0.1768
Land use 7 7 5 5 3 3 1 0.2611
Environmental 9 9 7 5 5 3 1 1 0.3055

Note: To interpret the table, rows should be compared with columns: for example, the environmental factor is rated as extremely more
important (9) than rural homesteads but of equivalent importance (1) when compared with land use

42 Impact Assessment and Project Appraisal March 2002


Use of GIS in EIA of power line

matrix (Eastman, 1995). The scale used a range from The relative weights sum to 1 as required by MCE.
9 (extremely more important) to 1 (equal In this case, the consistency ratio (also calc ulated by
importance). WEIGHT), which is a measure of the probability
The question posed in completing the table (Table that the values were randomly assigned, is 0.07. Ac-
1) by the authors is how much more important is cording to Saaty (1977), a consistency ratio less than
one factor than each of the others, with regard to 0.10 is good, whereas, a value greater than 0.10 is an
locating a power line? It is recognised that there is indication that the matrix entries should be
a certain amount of subjectivity involved in this re-evaluated.
process, however, the ranking of the factors that The second step in preparing the layers is the
were recorded in the original EIA study (Diab et al, quantification of non-numeric data such as land-use
1995) was used as a basis for the derivation of the types. The pairwise comparison method provides a
table. I&APs and specialists had been asked to rank means for determining relative suitability of catego-
factors and the task here was to combine these rank- ries within a layer and hence a score can be assigned
ings and assign weights, such that the factor that to each category. In this study, land-use categories in
gave rise to the most comment and concern was ac- increasing order of importance are: grasslands;
corded the highest weight. ploughed land; maize or vegetables; sugar cane;
The layers are, in ascending order of importance: bushveld; forestry; dams; other water bodies; in-
rural homesteads; access; cultural and historical digenous forest; peri-urban area of Ntunjambili; ur-
sites; cadastral boundaries; topography (represented ban area of Kranskop; landing strips; and formal
by slope); aesthetics (distance from viewpoint homesteads.
within the viewshed); land-use cover; and environ- Thirdly, the standardisation method of Eastman
mentally sensitive elements (vultures, wetlands and (1995) was applied, where scores were transformed
a proposed conservation area). The presence of vul- to standardised values such that 0 represents the
tures along the escarpment emerged as the most least suitable score and 100 the most suitable.
important consideration and thus environmentally Finally, constraints, defined as restrictions, were
sensitive elements were rated as extremely more applied. For example, a power line may not be con-
important (9) when compared with rural homes and structed within the restricted zone of an airfield.
access, but of equivalent importance (1) when com-
pared with land use. Combining factors
Tourist potential and the outstanding natural
beauty of the area surrounding The Kop were the The combination of factors into a single composite
main reasons for the high ranking of the aesthetic suitability surface is performed by the multi-criteria
factor. It was rated as very strongly more important evaluation module (MCE) of IDRISI. The procedure
(7) than rural homesteads and access and moderately is a weighted linear combination where each factor
more important (3) than cadastral boundaries. is multiplied by a weight and the results added to
Slope was rated as strongly more important (5) give an overall measure of the suitability for each
than access because the sheer cliffs of the escarp- grid cell. Factors were developed so that the higher
ment may preclude the construction of the power the social cost (as opposed to benefit) the lower the
line, while access to areas without existing roads suitability.
may be possible through the option of building the Therefore, subtracting the suitability surface from
necessary access roads. Cadastral boundaries were the maximum value of 100 provides a measure of
ranked as moderately more important (3) than the cost, which should be interpreted not in terms of
cultural factor and access, since routing along a monetary units but as a measure of unsuitability.
boundary minimises impact on land use and in some Once composite suitability and cost surfaces have
instances may facilitate access along existing tracks, been compiled, two general methods of route sele c-
which are used to maintain fences and boundaries. tion, referred to as the automated and computer-
While power lines which span cultural sites may assisted methods, may be employed to identify an
be aesthetically displeasing, careful positioning of optimum route.
tower sites away from the cultural site, and precau-
tions taken during construction to avoid damage to
the site, will minimise impact. Large numbers of
individual rural homesteads in the Tugela valley The multi-criteria evaluation module
posed a problem for routing of the power line, since of IDRISI created a single composite
dwellings should not be located within the servitude.
However, the EIA study (Diab et al, 1995) indicated suitability surface: each factor is
that relocation would be possible in some cases, if multiplied by a weight and the results
compensation and land were made available and added to give an overall measure of
provided that ancestral burial grounds were not
involved. the suitability for each grid cell
The weights in the last column of Table 1 are
determined by using the WEIGHT module of IDRISI.

Impact Assessment and Project Appraisal March 2002 43


Use of GIS in EIA of power line

This study initially set out to use the automated Table 2. Costs of potential routes for Section I
method using the COSTPUSH and PATHWAY
modules of IDRISI to derive an optimal power line Group Route Total cost Length Cost
route. However, because of inconsistencies and (km) index
anomalies that are fully described in Warner (1998)
and were reported to the software supplier, it was Southern IA 29565 14.651 2.02
decided not to pursue this method because of time IB 29414 15.128 1.94
constraints. Northern IC 48393 19.298 2.51
The alternative computer-assisted method consists ID 47870 19.247 2.49
of visually scanning a composite cost surface IE 47135 19.556 2.41
and demarcating possible alternative paths. By dis-
IF 46077 19.320 2.38
playing the cost surface on a computer screen,
together with the start and end points of the power IG 51635 22.189 2.33
line, the route can be digitised as a series of Impact study IH 47442 17.219 2.76
connected straight lines. The end points of the lines Note: Cost is defined as 100-suitability the higher the cost,
correspond to points at which a decision has to be the lower the suitability
made about a change in direction so that the criteria
of least cost and shortest distance are continuously
satisfied. 100. The total cost of a route is the sum of grid cells
In this manner, computer-assisted power line intersected by the digitised and buffered route. Rea-
routes are identified. As it was not possible to visu- sons for lower costs along the southern routes
ally determine which of two possible routes around included the fact that they crossed relatively flatter
an unsuitable area would have the lowest terrain and there is less forestry and more agricul-
costdistance value, a number of alternatives were tural land.
digitised and subjected to further evaluation for se- The optimum route based on this analysis was
lection of the optimal route. Route IB (Table 2). By comparison, the recom-
mended route of the original EIA was Route IH.
Reasons for this apparent anomaly lie in the strong
Results influence of the views of I&APs in the original EIA
study.
Seven alternative routes for Section I were identified I&APs, in particular the local farmers associa-
using the computer-assisted method. The route rec- tion, expressed strong collective opposition to addi-
ommended in the original EIA was based on field tional power lines on their farmland situated to
visits, and views of I&APs and the client. Routes are the south of Kranskop. Although views of I&APs
divided into two broad groups: those located to the influenced factor weightings they were not included
north of Kranskop (6) and those to the south (2). as a quantified layer in the GIS analysis.
Figure 3 displays only three selected routes, the oth- In hindsight, they could have been included as
ers being minor modifications of those presented. follows: farms that were unsuitable because of
The southern routes had shorter line lengths by ap- practices associated with intensive agriculture could
proximately 25% and overall lower cost indices have been delimited; farms with existing power lines
(Table 2). and where farmers were not in favour of additional
Cost can be interpreted as the inverse of suita- lines could have been identified and located; and the
bility, hence the higher the cost, the lower the road could have been buffered by an amount accept-
suitability. The cost surface is calculated by IDRISI able to local residents for the purpose of keeping the
by subtracting the composite suitability surface from power lines from view. These new attributes would
then have been added as a separate layer. Alterna-
tively, the environmental layer, which in this appli-
cation assigned values to intrinsic aspects of the
landscape, could have been redesigned to incorpo-
rate views of I&APs.
The discrepancy noted between the optimum
route identified by the GIS analysis and that rec-
ommended in the original study underlines the
importance of incorporating views of I&APs in a
GIS analysis and has emerged as a strong recom-
mendation of this study.
Focusing attention on the northern routes, it is
noted that Route IH, recommended by the original
Figure 3. Composite suitability surface for Section I of the
power line routing between substation SS1 EIA, has the highest cost index of all the alternative
(Hermannsburg) and SS2B (Kranskop) showing routes (Table 2), suggesting that there were others
three selected power line routes that were associated with fewer impacts. Route IH is

44 Impact Assessment and Project Appraisal March 2002


Use of GIS in EIA of power line

Table 3. Route evaluation for Se ction I

Factor Attribute or Criterion Route ID Route IF Route IG Route IH

Environmental Indigenous forest 0 0.625 1.687 0.063


Wetlands 4.063 1.625 0 5.563
Land use Grassland 70.250 90.813 102.125 67.438
Ploughed fields 2.875 1.813 1.375 3.250
Maize or vegetables 6.250 6.500 10.000 16.125
Sugar cane 26.750 27.688 29.437 25.750
Indigenous bushveld 8.375 14.250 25.250 20.438
Forestry 105.063 78.938 53.000 60.938
Slope Slopes >20 0 0 2.562 4.875
Slopes 1020 37.063 30.875 56.062 42.750
Slopes <10 188.000 197.750 192.938 158.188
Cadastral Farm boundaries 38.813 33.938 33.563 25.125
Access More than 800m from a road 58.750 56.813 98.313 75.688
Other information # rural homesteads in servitude 2 2 1 2
# formal homesteads 1 1 1 1
# farm dams 1 1 1 0
# roads crossed 5 5 3 5
# bends in the route 16 18 19 11

Note: Units are in hectares unless indicated otherwise

the shortest of all the northern routes and it is likely no impact on wetlands, a lower impact on forestry
that the consideration of line length and, by implic a- and cultivated lands other than sugar cane, and trav-
tion, economics had a major influence on the rec- erses terrain with generally lower slope angles. On
ommendation of the EIA study. the other hand, Route IH has a lower impact on in-
Identification of the optimum route on the basis of digenous forest, is more accessible and has fewer
the cost index provides no understanding of the im- bends.
pact on individual attributes, or of the trade-offs The availability of this type of information
required between attributes when selecting one route highlights the trade-offs that were made in an im-
over another. Furthermore, the cost index is influ- plicit manner in the selection of a route prior to the
enced by the weighting accorded to factors, which is availability of this GIS information. Certainly, the
acknowledged to be a subjective process. quantification of impacts in this manner makes
This latter point will be addressed by means of a trade-offs far more explicit and makes a valuable
sensitivity analysis discussed later in the paper. contribution to the decision-making process.
However, it is possible to determine the impact of a For Section II of the power line study between
route on an attribute using area analysis as shown in Kranskop and Middledrift, six alternative routes
Table 3. Using the IDRISI module AREA, the area were identified, of which one was situated to the
of intersection of the buffered route (determined by north of the view site, The Kop, and the remainder
the required servitude) on a partic ular attribute is were routed to the south. The seventh route investi-
computed. gated was that recommended by the original EIA.
Comparing the recommended route of the original Figure 4 displays four selected routes, the rest being
EIA (IH) with the best of the northern routes accord- minor modifications of IIG in the area below the
ing to cost index (IG), it is noted that Route IG has escarpment.
Route IIA situated to the north of the view site,

Table 4. Costs of potential routes for Section II

Group Route Total Length Cost


cost (km) index

Northern IIA 56013 19.426 2.88


Southern IIB 63310 20.279 3.12
IIC 65494 20.571 3.18
IID 65055 20.717 3.14
IIE 63348 20.184 3.14
IIF 55873 18.573 3.01
Figure 4. Composite suitability surface for Section II of the Impact study IIG 65664 20.942 3.14
power line routing between SS2B (Kranskop) and
SS3 (Middledrift) showing four selected power Note: Cost is defined as 100-suitability the higher the cost,
line routes the lower the suitability

Impact Assessment and Project Appraisal March 2002 45


Use of GIS in EIA of power line

Table 5: Route evaluation for Section II

Factor Attribute or Criterion Route IIA Route IIB Route IIF Route IIG

Land use Indigenous forest 0.125 7.500 2.625 11.000


Wetlands 0 0 3.563 0
Grassland 8.25 10.375 3.813 6.813
Sugar cane 16.938 0.188 12.375 6.375
Indigenous bushveld 189.938 208.938 183.000 205.063
Forestry 1.688 11.313 8.625 12.500
Aesthetics Visible and within 3 km of The Kop 43.500 0 0 0
Slope Slopes >20 3.375 10.375 11.938 11.813
Slopes 10-20 50.875 52.188 56.188 58.375
Slopes <10 172.063 179.438 146.438 175.250
Cadastral Farm boundaries 7.063 9.813 15.000 9.000
Access More than 800m from a road 152.813 167.250 128.813 143.938
Other information # rural homesteads 6 7 9 10
# roads crossed 6 3 3 3
# bends in the route 14 13 16 15

Note: Units are in hectares unless otherwise indicated.

The Kop, emerged as the route with the lowest cost varying the suitability of a particular attribute. For
index and Route IIG was the recommended route of example, one of the concerns evident in the impact
the original EIA (Table 4). Sections of Route IIA are study was the traversing of commercial forestry by
more than 3 km from an existing road making access the proposed power line. Widely divergent views on
difficult. The low weighting given to the access fac- the importance of commercial forestry were appar-
tor would account for the relatively low cost index ent, with landowners assigning it a very high ranking
of Route IIA despite its inaccessibility. compared with other I&APs.
Comparing the impacts of the Section II routes us- Hence, forestry emerged from the pairwise com-
ing area analysis, it is noted that Route IIG generally parison procedure and standardised routine, which
experie nces more impacts than the other routes and aimed to capture the collective ranking, with a suit-
hence it is surprising that it was the recommended ability of 83, with indigenous forest having a value
route of the EIA study (Table 5). If GIS analysis had of 35. Based on the premise that forestry might in-
been used in the orig inal EIA study, it is likely that a tuitively be expected to have a less suitable rating,
different alternative would have been recommended. the suitability of commercial forestry was reduced in
increments of 10 until it reached 35 (equivalent to
indigenous forest). Composite cost surfaces were
Sensitivity analysis calculated with the new values, and the total costs of
the selected routes re-evaluated.
Sensitivity tests were performed to gain some insight The changes caused no alteration in the ranking
into the effect of assigned weights on the MCE and of the routes until the last change (suitability = 35),
the ensuing cost analysis. This was achieved by when Route IH was ranked fifth, instead of eighth
(Table 6). Route IB was consistently the best overall
route for Section I. The order only changed when
Table 6. Section I routes ranked according to cost index for
varying forestry suitability values
Table 7. Section II routes ranked according to cost index for
varying forestry suitability values
Route Rank Forestry suitability
location order
Rank order Forestry suitability
83 73 63 35
83 73 63 35
South 1 IB IB IB IB
2 IA IA IA IA 1 IIA IIA IIA IIA
North 3 IG IG IG IG 2 IIF IIF IIF IIF
4 IF IF IF IF 3 IIB IIB IIB IIB
5 IE IE IE IH 4 IIG IIG IIG IIG
6 ID ID ID IE 5 IIE IID IID IID
7 IC IC IC IC 6 IID IIE IIE IIE
8 IH IH IH ID 7 IIC IIC IIC IIC

46 Impact Assessment and Project Appraisal March 2002


Use of GIS in EIA of power line

forestry was assigned the same suitability value as delineation of areas of unsuitably steep topography,
indigenous forest. can be quickly and easily performed by a GIS.
Similarly, for Section II, Route IIA was consis- Combining different technical, aesthetic, social
tently the best route (Table 7). The ranking of routes and environmental data requires subjective input and
IID and IIE reversed when suitability was decreased professional judgement. The GIS package cannot
from 83 to 73, however, the rankings remained make decisions unaided but is a useful aid to deci-
consistent for further decreases in the suitability of sion-makers. By using the weighting procedure de-
forestry. This is most likely a result of the small per- scribed in this paper, adjustments can be readily
centage of forestry occurring in Section II (4%). made and a sensitivity analysis performed. Alterna-
tive composite cost surfaces can be generated with
minimum effort, once the data set has been
Discussion and conclusion established.
On balance, it appears that there are still a number
GIS has been used to combine a number of factors of problems limiting the application of GIS in EIAs.
(layers) into a single composite surface, which has Chief of these is the length of time involved in dig i-
facilitated comparison of alternative power line tal data capture and the transfer of readily available
routes and an evaluation of impacts on, inter alia , digital data to the format required. However, in the
land-use types and environmentally sensitive areas. absence of pressing time constraints and for very
Experience has been gained in the relative weighting large and important projects, it would seem that the
of factors and, through the sensitivity analysis, the flexibility afforded by the GIS for generating alter-
influence that the weights have on the final product. native composite cost surfaces outweighs these
Two sections of a power line route were consid- disadvantages.
ered. In both, the optimum route identified by the
GIS analysis differed from the recommended route
of the original EIA study. Indeed, for Section I the References
original recommended route had the highest cost
index of all alternative routes. It was speculated that Antunes, P, R Santos, L Jordao, P Goncalves and N Videira
the recommendation of the original EIA was influ- (1996), A GIS-based decision support system for environ-
mental impact assessment, Proceedings of the 1996 Inter-
enced by the strong collective opposition to addi- national IAIA Conference Improving Environmental
tional power lines on their farms that was voiced by Assessment Effectiveness: Research Practice and Training,
the landowners. 1723 June, Estoril, Portugal, pages 451456.
Cassettari, S (1993), Introduction to Geo-information Manage-
The failure to incorporate I&AP views as a sepa- ment (Chapman and Hall, London).
rate factor in this GIS analysis was highlighted and Diab, R D, W N Ellery and J Combrink (1995), Environmental
suggestions were made as to how they could be impact study: proposed power line from Hermannsburg via
Kranskop and Middledrift to Sitilo, report commissioned by
incorporated in future. It was concluded that, if a Bosch and Associates, Durban.
GIS analysis had been available at the time of the Eastman, J R (1995), IDRISI for Windows Users Guide, Version
original EIA, a different alternative would have been 1.0 (Clark University, Worcester) pages 8-18-15.
Eedy, W (1995), The use of GIS in environmental assessment,
recommended. Impact Assessment, 13(2), pages 199206.
It is also possible to make some general observa- Joao, E, and A Fonseca (1996), The role of GIS in improving
tions about the feasibility of using GIS as a tool in environmental assessment effectiveness: theory vs. practice,
Impact Assessment, 14(4), pages 371387.
power line route selection. Transformation of data Maguire, D J, M F Goodchild and D W Rhind (editors) (1991),
into digital form and preparation for use in the GIS Geographical Information Systems: Principles and Applica-
are time-consuming tasks, confirming the experience tions (Longmans, London).
Monkhouse, F J (1970), A Dictionary of Geography (Edward Ar-
of other workers. Although IDRISI caters for the nold Publishing, 2nd edition).
transfer of data from other software packages, in Saaty, T L (1977), A scaling method for priorities in hierarchical
practice it was found that custom-designed software structures, Journal of Mathematical Psychology, 15, pages
234281.
was necessary at a number of stages. This is an im- Sankoh, O A, D Bonner, T Determann, J Gersten, A Gropp, I
portant consideration when project deadlines are Hoelters, N Krueger, U Lehmann, C Marty, K Strauss and H
short. However, time-consuming and complex man- Wegner (1993), Finding and assessing route alternatives,
Journal of Environmental Management, 38, pages 323334.
ual operations, such as viewshed analysis, which Warner, L L (1998), Power line route selection: a case study
was incorporated into the aesthetic layer, and the using GIS, unpublished MSc thesis, University of Natal.

Impact Assessment and Project Appraisal March 2002 47

Anda mungkin juga menyukai