GUA PARA LA SELECCIN

DEL COEFICIENTE DE
RUGOSIDAD DE MANNING EN
CAUCES NATURALES
Gua PRCTICA QUE TE AYUDAR A ESTABLECER LOS COEFICIENTES
DE RUGOSIDAD EN TUS MODELO HIDRULICOS

Por Jordi Oliveras

GUA PARA LA SELECCIN DEL

COEFICIENTE DE RUGOSIDAD DE
MANNING EN CAUCES NATURALES
Por Jordi Oliveras
Twitter: @hidrojing
www.hidrojing.com

Este obra est bajo una licencia de Creative Commons

Reconocimiento-NoComercial-CompartirIgual 3.0 Unported.

Esto significa que puedes copiar, distribuir y

pblicamente la obra e incluso transformar la obra.

comunicar

Eso s, con unas condiciones:

Reconocimiento: Debes reconocer los crditos de la obra de

la manera especificada por el autor, es decir, atribuirme
la autora original.

No comercial:
comerciales.

Compartir bajo la misma licencia: Si alteras o transformas

esta obra, o generas una obra derivada, slo puedes
distribuirla bajo una licencia idntica a sta.

No

puedes

utilizar

esta

obra

para

fines

QU ENCONTRARS EN ESTE LIBRO?

Te las ves y deseas cada vez que tienes que seleccionar el coeficiente de
rugosidad de Manning en cauces naturales?
Cuando nos enfrentamos a la definicin de los condicionantes hidrulicos de un
modelo hidrulico, ya sea uni o bidimensional, una de las variables que ms
influyen en el comportamiento del modelo es precisamente el coeficiente de
rugosidad.
Y es que efectivamente, los que nos dedicamos a esto de la modelizacin
hidrulica sabemos que una ligera variacin en el valor de ese parmetro
puede tener consecuencias inadvertidas en nuestros modelos.
No en vano, cuando nos ponemos a calibrar un modelo hidrulico para
contrastar la bondad de su comportamiento comparndolo con datos de
avenidas reales, el parmetro que modificamos hasta conseguirlo es
precisamente el coeficiente de rugosidad de Manning.
Por esa razn, en este e-book he querido recopilar y ofrecerte una pequea
gua con mtodos de seleccin de la variable estrella de la hidrulica de
canales: seleccin clsica por tablas, seleccin en funcin de los usos de
suelo, seleccin segn mtodo Cowan.
Adems, como BONUS, se incluye al final dos documentos:
-

Roughness Characteristics of Natural Channels (USGS, edicin 1987)

Guide for Selecting Manning's Roughness Coefficients for Natural

Channels and Flood Plains (USGS, edicin 1989)

Quiz algunas de las cosas que explique ya las sabrs, y quiz otras no@ o
quizs ya las conozcas todas@ sea como sea, espero que la lectura de esta
pequea gua te resulte interesante, entretenida@ y si encima te es de utilidad,
mi satisfaccin ser completa@

SELECCIN CLSICA POR TABLAS Y DOCUMENTOS DE

REFERENCIA
Afortunadamente, desde hace muchos aos, mucha gente ha estado
estudiando y analizando diferentes situaciones de flujo obteniendo de este
modo una cierta relacin entre unas caractersticas fsicas de cauces y
mrgenes y unos valores de coeficiente de Manning representativos.
Estas relaciones se suelen encontrar fcilmente en Internet en forma de
tablas@ quiz una de la ms reconocidas es la que integra el conocido libro de
Hidrulica de Canales Abiertos de Ven Te Chow. En esa tabla se recogen
valores recomendados de coeficiente de rugosidad de Manning para diferentes
casos como conductos cerrados y canales revestidos o dragados.
Pero tal y cmo apunta el ttulo de este e-book, y de cara tambin a la
modelizacin, me voy a centrar en los valores recomendados para cauces
naturales, que adems son los que el propio HEC-RAS recomienda tambin en
sus tablas de ayuda, y cuyos valores se recogen en la siguiente tabla:

Por si no lo conoces, HEC-RAS es un prestigioso software de modelizacin

hidrulica del Cuerpo de Ingenieros del Ejrcito de Estados Unidos. Ese
software se empez a desarrollar para estudiar el comportamiento hidrulico de
los cauces de Estados Unidos, y precisamente en ese pas existe un
organismo, la U.S. Geological Survey, dedicado a recopilar informacin
medioambiental del mismo.
En toda esta informacin del USGS hay un apartado especfico sobre el agua,
y en l se puede encontrar documentos e informes relacionados con la
seleccin del coeficiente de rugosidad de Manning en los que se recogen
valores de n de Manning verificados y comprobados para diferentes
caractersticas de cauces. De todos estos te recomiendo los siguientes:
-

Una relacin de 23 caractersticas verificadas de rugosidad de cauces

naturales estudiados en tramos de cauces reales. En cada propuesta de
valor, un enlace lleva a una pgina con informacin de detalle
(descricpin, esquemas, propiedades e imagenes) de ese tramo. Haz
clic aqu para acceder a la web.

Un libro en pdf (editado en 1967) que vendra a ser el origen de la

anterior relacin, ya que en l se recoge la misma informacin tcnica y
grfica, pero esta vez de hasta un total de 50 cauces del territorio
norteamericano. Este libro lo tienes como apndice, al final de este ebook.

SELECCIN EN FUNCIN DE LOS USOS DE SUELO

Hoy en da tenemos una mayor accesibilidad a informacin cartogrfica y
territorial, disponemos de unas mayores y ms potentes tecnologas y
herramientas que mejoran y optimizan la generacin de modelos hidrulicos.
De aqu que hayan surgido propuestas de definicin de parmetros clsicos
adaptados a las nuevas fuentes de informacin@ y los coeficientes de
rugosidad de Manning no han sido una excepcin.
Se ha establecido en este sentido, una relacin entre los usos del suelo y los
coeficientes de rugosidad de Manning que los representan, segn dos tipos de
clasificacin: SIOSE (Sistema de informacin sobre Ocupacin del Suelo de
Espaa, de la Direccin General del Instituto Geogrfico Nacional) y CORINE
(CoORdination of INformation of the Environment, de la Agencia Europea del
MEdioambiente).
El primero est realizado a una escala de referencia 1:25.000 y el segundo a
una escala 1:100.000. Esto significa que la variabilidad de tipos de suelo para
una misma zona ser mayor en el SIOSE que en el CORINE.
Esta relacin uso de suelo vs. coeficiente de rugosidad de Manning se recoge
en unas tablas includas en la Gua Metodolgica para el desarrollo del Sistema
Nacional de Cartografa de Zonas Inundables.
A continuacin te adjunto esas tablas@ vers que en la parte de la tabla
correspondiente al uso de suelo segn SIOSE hay mis anotaciones en lpiz
que intentan completar esa informacin:
-

Aadiendo algunas tipologas de coberturas del propio SIOSE que no

aparecen en la tabla

Indicando el cdigo SIOSE correspondiente a cada descripcin o

nombre del uso

Comentar que esta informacin est referida al territorio espaol, con lo que en
pases de la Unin Europea o del resto del mundo, la definicin de rugosidades
de Manning deber basarse en otras fuentes de informacin, o simplemente en
la definicin tradicional segn las caractersticas del terreno.

SELECCIN POR EL MTODO COWAN

Disponer de toda esta informacin previa de valores ya establecidos,
verificados, contrastados y tabulados es realmente de mucha ayuda@ pero
claro, esos valores responden a unas caractersticas muy concretas que
seguramente presentarn similitudes con las que tiene el cauce que debes
modelizar@ pero al final cada caso es nico.
Pero@ cmo seleccionar el coeficiente de rugosidad de Manning en cauces
naturales@ si tu cauce no est en las tablas? Por ello es importante entender
los factores que afectan el valor n para adquirir un conocimiento bsico del
problema y disminuir la incertidumbre@
Y si nos centramos en los cauces naturales los factores que ms influencia
tienen en la determinacin del valor de la n de Manning son: el tipo y tamao
de los materiales que lo componen y las caractersticas morfolgicas del
mismo. Teniendo en cuenta estos condicionantes, disponemos de un mtodo
de obtencin del valor del coeficiente de rugosidad de Manning. El mtodo de
Cowan o multiparamtrico.
Cowan, en 1956, desarroll una expresin que permite determinar el valor del
coeficiente de Manning a travs de la interaccin de diferentes parmetros que
permiten describir o valorar caractersticas concretas de un curso fluvial. La
expresin es la siguiente:

En esta expresin, que sirve tanto para la caracterizacin de resistencia al paso

del flujo en cauce principal como en llanuras de inundacin, el valor del
coeficiente de rugosidad de Manning n depende de:
nb = un valor base de n para un cauce recto, uniforme y liso en funcin
del material del fondo del lecho / para una llanura de inundacin con
suelo sin vegetacin
n1 = factor de correccin para implementar el efecto de las
irregularidades superficiales tanto en canal principal como en llanura de
inundacin
n2 = un valor que aade las variaciones de forma y tamao de la seccin
del cauce. Para llanuras de inundacin este valor se considera 0.
n3 = un valor que implementa el efecto de obstrucciones en canal
principal y en llanura de inundacin
n4 = un valor que incorpora el efecto de presencia de vegetacin tanto en
canal principal como en llanura de inundacin

m = un factor corrector que implementa la sinuosidad del cauce. Para

llanuras de inundacin este valor se considera 1.

As, con este mtodo, partiendo de un

canal terico recto, uniforme y liso de
un material dado al cual le corresponde
un valor de coeficiente de rugosidad de
Manning determinado, a ste se le van
aadiendo valores que representan la
presencia
de
elementos
y
caractersticas que condicionan el paso
del flujo.
La Gua Metodolgica para el desarrollo
del Sistema Nacional de Cartografa de
Zonas Inundables, en su Anejo VI
presenta una tabla que relaciona cada
uno de los anteriores parmetros con
unos valores o intervalos de valores en
funcin de las caractersticas que
presentan cada uno de los aspectos
goemorfolgicos
expuestos
anteriormente.
Esta tabla de correlacin de la Gua
Metodolgica, que se basa en el
documento
Guide
for
Selecting
Manning's Roughness Coefficients for
Natural Channels and Flood Plains
(USGS) se cie nicamente a la
caracterizacin de la rugosidad en el
lecho del cauce.

Pero profundizando ms en el tema, en esa Guide for Selecting Manning's

Roughness Coefficients for Natural Channels and Flood Plains (USGS) se
efecta una ms extensa descripcin de cada una de esas variables y adems
se recomiendan intervalos de valores no slo para el lecho del cauce, sino
tambin para las llanuras de inundacin.

Valores base nb
Se distinguen dos tipos de canal: canal estable o canal de arena. El primero es
aquel cuyo lecho es firme (de tierra, grava, guijarros, cantos rodados, roca u
hormign) y que permanece sin cambios significativos en el mbito del rgimen
de flujo previsto.
En la siguiente tabla se recogen los valores base nb para ambos tipos de canal.

Los valores de Benson y Dalrymple implican condiciones consideradas como

medias o habituales, mientras que los valores de Chow tienen relacin con el
tramo ms suave que puede darse para un material en concreto.
Se emplea el mismo criterio para la determinacin del valor base nb en el caso
de las llanuras de inundacin.

Factores correctores del coeficiente de rugosidad

Los valores que se establezcan para el valor base nb son para cauces rectos
con secciones muy uniformes, y para llanuras de inundacin sin vegetacin.
Cualquier aspecto que implique irregularidades, cambios de alineacin,
obstrucciones, vegetacin o incremento de la sinuosidad no har ms que
aumentar su rugosidad.

Factores correctores en canal principal

Irregularidad (n1)

En los cauces principales en los que la relacin entre anchura y profundidad es

pequea, debe tenerse en cuenta la rugosidad en las bancadas laterales
(ribera) causada por la erosin, ondulacin o presencia de elementos
emergentes y races.
Chow (1959) y Benson y Dalrymple (1967) mostraron que situaciones de
riberas con ondulaciones pueden incrementar los valores de n hasta 002.
En situaciones de irregularidad severa, puede ser necesario que los
incrementos de valor de rugosidad sean an mayores.
Variaciones en seccin transversal (n2)

El valor de la rugosidad no se ve afectado de manera significativa si las

variaciones de forma y tamao de seccin transversal se producen de manera
gradual y uniforme (incluso si estas variaciones son de consideracin).
Las mayores variaciones de rugosidad estn asociadas a alternancias en el
tamao de las secciones y giros pronunciados, contracciones de seccin y
sinuosidad del canal de aguas bajas.
La incidencia del efecto del cambio de tamao en la seccin depende
principalmente de la cantidad de alternancias entre secciones grandes y
pequea, y en segundo lugar por la magnitud de esos cambios.
La presencia de una perturbacin puede implicar una variacin del valor de n
hasta decenas de metros aguas abajo, aunque en este tramo no haya
condicionantes que supondran una modificacin del valor en l.
Obstrucciones (n3)

Obstrucciones tales como troncos, tocones, piedras, escombros, y pilares de

puente perturban el rgimen de e implementan un incremento de la rugosidad.
La magnitud de ese incremento depende de la forma de la obstruccin; el
tamao de la obstruccin en relacin con la de la seccin transversal; y el
nmero, disposicin, y la distancia entre obstrucciones.

El efecto de las obstrucciones en el coeficiente de rugosidad depende de la

velocidad del flujo.
Cuando la velocidad de flujo es alta, una obstruccin incide en un mbito de
influencia mucho ms grande que la propia obstruccin, ya que su presencia
afecta al rgimen de flujo en una anchura considerable a ambos lados de la
misma.
El mbito de influencia en canales con pendientes suaves o poco empinadas
puede llegar a ser de tres a cinco veces el ancho de la propia obstruccin. Esto
provoca que, aunque las obstrucciones sean pequeas, si se encuentran lo
suficientemente cerca pueda darse el caso que sus mbitos de influencia se
superpongan.
Vegetacin (n4)

El grado de incidencia de la vegetacin sobre el valor de n depende de la

profundidad de flujo, el porcentaje de la permetro mojado cubierto por la
vegetacin, la densidad de la vegetacin por debajo la lnea lmina de agua en
avenida, el grado en que la vegetacin es aplastada por ese calado, y la
alineacin de la vegetacin con respecto al flujo. Vegetacin orientada en filas
paralelas a la direccin del flujo puede tener menos efecto que las filas de
vegetacin que se encuentran perpendiculares al flujo.
Los valores de ajuste mostrados en la tabla 2 se aplican a los canales que son
estrechos. En canales anchos, cuya relacin profundidad/anchura es pequea,
y no presentan vegetacin en la lecho, el efecto de la vegetacin de ribera es
pequeo, y el ajuste mximo es de aproximadamente 0005.
Si el canal es relativamente estrecho y presenta orillas escarpadas cubiertas
por densa vegetacin que se cierne sobre el canal, el mximo ajuste es de
aproximadamente 003. Los valores de ajuste ms grandes mostrados en la
tabla 2 se aplican slo en lugares donde la vegetacin cubre la mayor parte de
la canal.
Meandros (m)

El grado de sinuosidad m, depende de la relacin entre la longitud total de

canal sinuoso entre la longitud (en lnea recta) del valle por el que discurre.
Se considera la cantidad de meandros es baja cuando la sinuosidad est entre
1 y 12, media cuando est entre 12 y 15, y alta cuando supera el 15.
Segn Chow (1959), los meandros pueden incrementar los valores de hasta en
un 30 por ciento. El ajuste por meandros debe considerarse slo cuando el flujo
queda confinado en el canal ya que en caso de desborde, seguramente la
proporcin de flujo que circule por el canal sea pequea con respecto de la que
circula por la llanura de inundacin.

Tabla 2. Valores correctores en canal principal

Factores correctores en llanura de inundacin

Los valores de rugosidad en la llanura de inundacin deben determinarse de
manera independiente a los del canal principal ya que las caractersticas fsicas
y morfolgicas pueden diferir ostensiblemente entre ambas zonas.
En el caso de las llanuras, el factor de correccin por variacin de la seccin
transversal es 0, mientras que el factor por meandros es 1.
Irregularidad (n1)

La irregularidad de la superficie de una llanura de inundacin provoca un

incremento de la rugosidad de la llanura de inundacin. La presencia de
elementos fsicos como elevaciones y depresiones de la superficie,
encharcamientos y montculos aumentan la rugosidad de la llanura.
Situaciones de aguas poco profundas, en pastizales o matorrales con
superficies de suelo irregulares o en cultivos con alineacin perpendicular al
flujo pueden llegar a incrementar la rugosidad en 002.
Obstrucciones (n3)

Es complejo medir la contribucin que tienen sobre la rugosidad algunas

obstrucciones (como depsitos de escombros, troncos, races expuestas, o
rocas aisladas) pero pueden estimarse a partir de lo expuesto en Tabla 3, que
enumera los valores de incremento de rugosidad para diferentes porcentajes
de presencia de obstruccin
.Vegetacin (n4)
Los efectos de la vegetacin sobre la rugosidad se basan en la observacin,
anlisis y experiencia, y sus valores y criterios se recogen en la Tabla 3.
Aunque la determinacin del rea ocupada por troncos de rboles y otra
vegetacin de gran dimetro es relativamente fcil, la medicin de la superficie
ocupada por las vides bajas, zarzas, hierba o cultivos es ms difcil (Tabla 3).
En el caso de campos abiertos y cultivos en llanuras de inundacin, existen
varias referencias para determinar los factores de rugosidad. Ree y Cuervo
(1977) llevaron a cabo experimentos para determinar los factores de rugosidad
de canales de tierra con pendiente suave sembrados con trigo y otros pastos.
Estos factores de rugosidad pueden ser utilizados para estimar la rugosidad de
las llanuras de inundacin cultivadas.

Tabla 3. Valores correctores en canal llanuras de inundacin

Click here to return to USGS publications

Roughness

Characteristics
of Natural Channels

By HARRY II . BARNES, JR .
U.S. GEOLOGICAL

SURVEY WATER-SUPPLY PAPER 1849

Color photographs and descriptive

data for- 50 stream channels for
which roughness coe

cients have

been determined

UNITED STATES GOVERNMENT PRINTING OFFICE, WASHINGTON : 1967

DEPARTMENT OF THE INTERIOR

DONALD PAUL HODEL, Secretary

U .S . GEOLOGICAL SURVEY
Dallas L . Peck, Director

First printing 1967

Second printing 1977

Third printing 1987

For sale by the Books and Open-File Reports Section, U .S . Geological Survey,
Federal Center, Box 25425, Denver, CO 80225

Contents
Symbols____________________________________________________________________________________
Abstract____________________________________________________________________________________
Introduction____________________________________________________________________________
Acknowledgments____________________________________________________________________
Scope of report________________________________________________________________________
Field investigation__________________________________________________________________
Computation of reach properties
and roughness coefficients__________________________________________________
Formulas__________________________________________________________________________
Computation procedure________________________________________________
Application of roughness coefficients____________________________________
Presentation of information____________________________________________________
Columbia River at Vernita, Wash . (n = 0 .024)-----------Indian Fork below Atwood Dam,
near New Cumberland, Ohio. (n = 0.026)---------------Champlin Creek near Colorado City, Tex .
(n=0 .027)____________________________________________________________________

Clark Fork at St . Regis, Mont. (n = 0.028)-----------------Clark Fork above Missoula, Mont. (n=0 .030) ____________
Columbia River at The Dalles, Oreg. (n = 0 .030) -----Esopus Creek at Coldbrook, N.Y. (n=0.030) ______________
Salt Creek at Roca, Nebr. (n = 0 .030) -------------------------Blackfoot River near Ovando, Mont. (n = 0 .031)-------Coeur d'Alene River near
Prichard, Idaho (n = 0 .032) -------------------------------------Rio Chama near Chamita, N. Mex .
(n = 0.032 ; 0 .036) -------------------------------------------------------Salt River below Stewart Mountain Dam, Ariz.
(n=0 .032)____________________________________________________________________

Beaver Kill at Cooks Falls, N.Y. (n=0 .033) ____________

Clearwater River at Kamiah, Idaho (n=0 .033) ______

Page

vi

10
14
18

22
26

30

34

38
42

46
50
54

58

62

Presentation of information-Continued
Etowah River near Dawsonville, Ga .

Page

(n=0 .041 ; 0.039 ; 0.035)--------------------------------------------

66

(n = 0.036) ____________________________________________________________________

70

West Fork Bitterroot River near Conner, Mont.

Yakima River at Umtanum, Wash . (n = 0 .036) __________

Middle Fork Vermilion River near Danville, 111 .
(n = 0.037) ____________________________________________________________________
Wenatchee River at Plain, Wash . (n=0 .037)______________
Moyie River at Eastport, Idaho (n=0 .038) ---------------Spokane River at Spokane, Wash. (n = 0.038) -----------Tobesofkee Creek near Macon, Ga.

74

78
82
86
90

(n=0 .043 ; 0 .041 ; 0 .039) --------------------------------------------

94

(n=0 .041)____________________________________________________________________

102

Bull Creek near Ira, Tex . (n=0 .041)____________________________

Middle Fork Flathead River near Essex, Mont.
Middle Oconee River near Athens, Ga .

98

(n=0 .042 ; 0 .041 ; 0 .044) ------------------------------- ------------ 106

110

Beaver Creek near Newcastle, Wyo. (n = 0.043) __________

Catherine Creek near Union, Oreg. (n = 0.043) ---------Chiwawa River near Plain, Wash . (n = 0.043)-----------Esopus Creek at Coldbrook, N.Y . (n = 0 .043)-------------Grande Ronde River at La Grande, Oreg .

114
118
122

(n=0 .043)____________________________________________________________________ 126

Murder Creek near Monticello, Ga . (n = 0 .045) -------- 130

Provo River near Hailstone, Utah (n = 0 .045 ; 0.073)__ 134
Rolling Fork near Boston, Ky . (n = 0.046 ; 0 .097) ------ 138
South Beaverdam Creek near Dewy Rose, Ga .
(n=0 .052 ; 0.047) --------------------------------------------------------

142

Deep River at Ramseur, N.C. (n=0 .049) ____________________ 146

Clear Creek near Golden, Colo. (n=0 .050) ________________ 150
Chattahoochee River near Leaf, Ga .
(n=0 .051 ; 0 .074) -------------------------------------------------------- 154

South Fork Clearwater River near Grangeville,

Idaho (n=0 .051) ------------ ------------------------------------------- 158
Cache Creek near Lower Lake, Calif.
(n=0 .053 ; 0 .079) -------------------------------------------------------- 162

iv

Presentation of information- Continued

East Branch Ausable River at Au Sable Forks,

Page

Middle Branch Wemtficld River at Goss Heights,

Mama. (n==O.O56) ---------------------------------------------------------- 170
/m==0.0571_-- 174
Mission Creek near Cashmere,
Bpnn1c4
N.
C
.
[o==O.O5q).--------. 178
HavvRiver near
North Fork Cedar River near Lester, Wash .
(n = 0.059) -------------------------------------------------------------------- 182

Hominy Creek at Caod}er, N .C. /n = O .060\---------------Rook Creek Canal near Dnrbn. Moot /m=0,06O\-Merced River at Happy Isles Bridge, near
Yosemite, Calif. (n==O.O65) ---------------------------------------Pond Creek near Louisville, Ky. /m==0 .O70\---------------Boundary Creek near PortbD}, Idaho (n=0.073)-----Rock Creek near Drby' Moot . (m= 0.075) --------------Selected references -----------------------------------------------------------------Index------------------------------------------.----------------------'

186
190
194

198

202
206

210
211

Symbols
A
C
d

h
ha
hf
Oh,

K
L
n
Q
R
S
V

Area of channel cross section

Flow-resistance factor
Diameter of bed material
Water-surface elevation
Velocity head
Energy loss due to boundary friction
Upstream velocity head minus the downstream velocity
head
A coefficient
Cross section conveyance
Length of reach
Coefficient of roughness
Discharge
Hydraulic radius
Energy gradient
Average velocity

ROUGHNESS CHARACTERISTICS

OF NATURAL CHANNELS

By Harry H. Barnes, Jr .

Abstract
Color photographs and descriptive data are presented for 50 stream
channels for which roughness coefficients have been determined .
All hydraulic computations involving flow in open channels require an
evaluation of the roughness characteristics of the channel . In the absence
of a satisfactory quantitative procedure this evaluation remains chiefly an
art. The ability to evaluate roughness coefficients must be developed
through experience. One means of gaining this experience is by examin
ing and becoming acquainted with the appearance of some typical channels
whose roughness coefficients are known .
The photographs and data contained in this report represent a wide
range of channel conditions. Familiarity with the appearance, geometry,
and roughness characteristics of these channels will improve the engineer's
ability to select roughness coefficients for other channels .

INTRODUCTION
The principal objective of this report is to present descriptive
data and photographs for 50 different stream channels for
which roughness coefficients have been determined . This
information, which has been accumulated by the U.S . Geo
logical Survey during the past 15 years, was previously availa
ble only in a photographic slide library in three-dimensional
color. Numerous requests for copies of the slides from organi
zations and private individuals led to the justification for the
present report .

All hydraulic computations involving flow in open channels

require an evaluation of the roughness characteristics of the
channel . At the present state of knowledge, the selection of
roughness coefficients for natural channels remains chiefly an
art . There are no resistance diagrams or quantitative relation
ships available similar to those used for steady flow in uniform
pipes or for the frictional resistance of ships . Consequently the
ability to evaluate roughness coefficients for natural channels
representing a wide range of conditions must be developed
through experience.
The experience necessary for the proper selection of rough
ness coefficients can be obtained in several ways, namely (1) to
understand the factors that affect the value of the roughness
coefficient, and thus acquire a basic knowledge of the problem,
(2) to consult a table of typical roughness coefficients for chan
nels of various types (Woodward and Posey, 1941), and (3) to
examine and become acquainted with the appearance of some
typical channels whose roughness coefficients are known.
Photographs of channels of known resistance are thus useful
in estimating the roughness characteristics of similar channels .
The photographs and data presented in this report cover a wide
range in conditions . Familiarity with the geometry, appear
ance, and roughness characteristics of these channels will
improve the engineer's ability to select roughness coefficients
for other channels .

To the untrained beginner, the selection of a roughness

coefficient can be no more than a guess ; and different indivi
duals obtain different results . Furthermore, it is sometimes
difficult to convince either the layman or the scientist that
consistently reliable roughness coefficients can be selected by
trained engineers on the basis of sound judgment and experi
ence. Fortunately, even though the, selection of the coef
ficients is classified as an art, the accuracy of many selections
can be evaluated in exact engineering or statistical terms .
The ability to evaluate the roughness characteristics of
channels is important in the hydraulic work of the U.S . Geo
logical Survey . This ability, for example, is involved in the

methods (Benson and Dalrymple, 1966 ; Dalrymple and Ben

son, 1966 ; Bodhaine, 1966 ; Matthai, 1966 ; Hulsing, 1966)
which are used in defining the peak discharge of most major
floods . For this reason the Survey maintains a program
which both trains young engineers in the evaluation of channel
roughness and tests the accuracy of roughness coefficients by
veteran engineers . The results of these tests as reported by
Bailey and Ray (1966) indicate that trained engineers can
select roughness coefficients with an accuracy of plus or minus
15 percent under most conditions . These facts prove that
present methods are sound but that there is much room for
improvement . A quantitative procedure for determining the
roughness characteristics of channels has been the goal of
research in the Survey and in other organizations for many
years, but as yet little practical success has been achieved.

ACKNOWLEDGMENTS

Data contained in this report represent contributions by

many engineers of the U .S. Geological Survey . Much credit
is due Hollister Johnson and Tate Dalrymple who conceived
and promoted the Survey's program to verify roughness
coefficients in natural channels.
The author gratefully acknowledges the advice and assistance of R . W. Carter, W. R. Stokes, E. D. Cobb, and R. E .
Smith.

SCOPE OF REPORT

Information on the geometry and roughness characteristics

of 50 different stream channels is presented in the report . All
of the stream channels are considered to be stable . Sandchannel streams were not included in the report because
roughness coefficients for streams of this type have been
defined in terms of size of bed material and other variables
(Simons and Richardson, 1962) .
The 50 sites include a wide range of hydraulic conditions
from the boulder-strewn mountain stream of the western
conterminous United States to the heavily vegetated flatsloped stream of the southern conterminous United States .
3

The techniques used in field investigations at each site are

first discussed in the report . The procedures used in comput
ing the value of the roughness coefficient are then described .
The remainder of the report consists of the presentation of a
set of data and photographs for each of the 50 sites.

FIELD INVESTIGATION

Sites were selected for study after a major flood had occurred
in a given region . Each site met the following criteria:
1 . The peak discharge of the flood was measured by the
current-meter method, or determined from a well-defined
stage-discharge relation ;
2 . Good high-water marks were available to define the watersurface profile at the time of the peak ;
3 . A fairly uniform reach of channel was available near the gage;
4. The flood discharge was within the channel banks-that
is, extensive flow in flood plains did not exist .
A transit-stadia survey of each reach was completed shortly
after the flood . The necessary information was obtained in
this survey to plot accurately to a common datum (1) the
water-surface profile as represented by high-water marks, (2)
a plan view of the reach, (3) cross sections at intervals along
the reach . Surveying techniques used in this investigation are
described in detail by Benson and Dalrymple (1966) .
Photographs of the. reach were taken during the time of
the survey . The photographs shown in this report thus
represent conditions in the reach immediately after the flood.
A size description of the bed material at most of the sites
was determined by sampling methods (Wolman, 1954) . These
samples were in general taken several years after the flood
for which the roughness coefficient was determined. The
samples may or may not be representative of the bed material
at the time of the peak . Frequency distributions of bedmaterial size were determined by sieve analysis where the
medium size of the material was less than 50 mm and, where
the material was too large to sieve, by measuring the inter
mediate axis of particles selected at random from the bed
surface .
4

COMPUTATION OF REACH PROPERTIES

AND ROUGHNESS COEFFICIENTS
Formulas
Most open-channel flow formulas can be expressed in the
following general terms,
Q=C A Rz S'

(1)

where Q is the discharge, in cubic feet per second ; C is a

factor of flow resistance ; A is the cross-sectional area of the
channel, in square feet ; R is the hydraulic radius, in feet ;
and S is the energy gradient . The Manning equation, one
of the well-known variations of equation 1, was used as the
basis for computing the reach properties and roughness co
efficients given in this report. The Manning equation is
Q'

_
_

1 .486
AR2/3
n

5112

where n is a roughness coefficient and other variables in the

equation are as defined above .
The Manning equation was developed for conditions of
uniform flow in which the water-surface profile and energy
gradient are parallel to the streambed, and the area, hydraulic
radius, and depth remain constant throughout the reach .
For lack of a better solution, it is assumed that the equation
is also valid for nonuniform reaches, invariably found in
natural channels, if the energy gradient is modified to reflect
only the losses due to boundary friction. The energy equation
for a reach of nonuniform channel between sections 1 and
2 in figure 1 is (see p. J)
(h+h,), = (h+hv)2+(hf)1 .2+k(Ohv), .2
where
h =elevation of the water surface at the respective
sections above a common datum;
hr, = velocity head at the respective section = V 2/ 2g ;
hf = energy loss due to boundary friction in the reach ;
5

Ah v =upstream velocity head minus the downstream

velocity head ;
k(Ah v) = energy loss due to acceleration of velocity or
deceleration of velocity in a contracting or
expanding reach ; and
k = a coefficient taken to be zero for contracting
reaches and 0 .5 for expanding reaches.
In computing the values of n listed in this report the value of ,
the velocity head coefficient, was always considered to be 1 .00.
The friction slope S to be used in the Manning equation
is thus defined as
S=hy
L

Ah+Ah,-k(Ah,)
L

(4)

where Ah is the difference in water-surface elevation at the

two sections and L is the length of the reach.
In using the Manning equation the quantity (1 .486/n)AR2 i 3 ,
called conveyance and designated K, is computed for each
cross section. The mean conveyance in the reach between
any two sections is computed as the geometric mean of the
conveyance of the two sections . The discharge equation in
terms of conveyance is
Q= 11 K2 S
where S is the friction slope as previously defined .
In this investigation the average value of the Manning n was
computed for each reach from the known discharge, the water
surface profile, and the hydraulic properties of the reach as
defined by 2-17 cross sections. The following equation,
which is based on the same concepts and definitions as equa
tions 2-5, was used in these computations . The equation is
applicable to a multisection reach of M cross sections which are
designated 1, 2, 3, . . . M-1, M.
n=

1 .486 (h+ ha)1-(h+hv)M-[(k Ohv) 1.2+(k Ohv)2.3+ . . .-1

Q ~L .2 + L2 .3 . . . +L(m-1) .M
Z1 Z2 Z2 Z3
z(M-1)ZM

where Z =
6

AR 21'

(k Ahv)(M-1) .M] (6)

and other quantities are as previously defined.

Computation Procedure

A planimetric map of each reach was developed by plotting

from the notes of the field survey. The location of all high
water marks and cross sections was shown on the map . The
distances between cross sections were determined from the map.
The profile of the water surface through the reach was
developed separately along each bank by plotting the elevation
and stationing of high-water marks . The water-surface ele
vation at each cross section was determined as the average of
the water-surface elevation on each bank as taken from the
water-surface profiles.
Cross sections were plotted from the field notes, and data on
stationing, distance, ground elevations, depths, and top widths
were tabulated. The area and wetted perimeter for each
panel between given ground elevations were computed. The
area, wetted perimeter, and hydraulic radius for each cross
section were determined, and values of AR"' were computed.
The average value of n for each reach was then determined
by substituting the proper quantities in equation 6.
The computation procedure is virtually the same as for com
putation of discharge by the slope-area method . This pro
cedure is described in detail by Dalrymple and Benson (1966) .

APPLICATION OF ROUGHNESS
COEFFICIENTS

The values of n presented in the report are intended for use

in the Manning equation
V _ 1 .486 R2/3 5112
(English units)
n

or

R2/3 S112

(metric units)

The value of n may be converted to values of the Chezy C by

the relation
= 1 .486 R116
C
(English units)
n

and the value of C may then be used in the Chezy equation

V=CRS

All these equations are limited to turbulent flow in fully

rough channels . Flow in natural channels normally meets
this criterion .

PRESENTATION OF INFORMATION
A four-page set consisting of channel data, plan sketches
(not to scale) and cross sections, and photographs, is presented
for each reach in the following section of the report . Each
set of data is identified by the permanent gaging-station
number and name used by the U.S. Geological Survey in
publication of streamflow-records .
The data tabulation shows the location of the gage and a
reference cross section, the drainage area of the stream, the
date of the flood, the peak gage height at the gage and at sec
tion 1 during the flood, the peak discharge measured by current-meter method, the computed roughness coefficient for the
reach, a general description of the channel, the median bedmaterial size, d o, and the reference size for which 84 percent
of the bed material is finer, d84. The area, top width, mean
depth (area/top width), hydraulic radius, and mean velocity
corresponding to the water-surface elevation at the time of the
peak are listed for each cross section . The distance or length
between cross sections and the fall in water surface between
cross sections are also shown . Information for two or more
peak discharges is available at some sites. These data are
listed according to the magnitude of discharge . Data cor
responding to the largest discharge appear first.
At several sites a small percentage of the flow occurred in
the shallow flood plain adjacent to the main channel . For
each of these sites the data and computations reflect the flow
of the main channel .
Data for the site, Rolling Fork at Boston, Ky., are unique in
that roughness coefficients were computed for both the over
flow plain and the main channel .
Two color photographs taken immediately after the flood
are shown for each reach . The position of the camera and the

number of the picture are shown on the plan sketch by a

pointer, which shows the direction in which the camera was
pointed . The water level at the time of the peak is indicated
in some of the photographs by a horizontal rod or tape.
The initial station for cross sections is at the left bank. Plots
are arranged so that the left bank appears on the reader's left.
The water levels shown on the cross sections correspond to the
water-surface profile at the time of the peak as defined by
high-water marks.
Sites are arranged according to the value of the computed
roughness coefficient, in ascending order.

"*-Section 1

Section 2,,rj.
v

-1.

t
PLAN VIEW

hvl

1 ~ - --- -

rgy grad ient

Water surface

Datum
L
PROFILE VIEW
Figure 1. - Definition sketch of a slope-area reach.

I hf+k(Ah)
[hv2

n = 0 .024

12-4645 . Columbia River at Vernita, Wash.

Gage location .-In sec. 11, T. 13 N., R. 24 E., at the Richmond

ferry site, 0 .5 mile north of Vernita station . Gage presently
operated 50 miles upstream for station called Columbia
River at Trinidad, Wash. Section 1 is 5,000 ft upstream
from cableway at Vernita gage .
Drainage area.-89,700 sq mi, approximately .
Date offlood.-May 22, 1949 .
Gage height.-48 .33 ft at Trinidad gage ; 29 .5 ft (different
datum) at section 1 .
Peak discharge .-406,000 cfs.
Computed roughness coefficient. -Manning n=0.024.
Description of channel .-Bed consists of slime-covered cobbles
and gravel . The straight and steep left bank is composed
of cemented cobbles and gravel. The gently sloping right
bank consists of cobbles set in gra;-e1 and is free of vegetation .
Reach properties
Section
1 . . . . . . . . . . .

2. . . . . . . . . . .
3. . . . . . . . . . .

Notes.-

Area
(sq ft)

47,100
49,000
49,600

Top
width
(ft)

1,800
1,650
1,760

Mean
depth
(ft)

26 .2
29 .7
28 .2

Hydraulic
radius
(ft)

Mean
velocity
(ft per sec)

26 .16
29,56
28 .10

8.65
8.28
8.17

Length (ft)
between
sections

Fall (ft)
between
sections

. . . . . . . . . . . . . . . .
2,500
0 .48
2,500
.49

n = 0.024

CROSS SECTIONS

Plan sketch and cross sections, Columbia River at

Vernita, Wash.

n = 0.024

No. 67 upstream from top of bank at section 3,

Columbia River at Vernita, Wash.
12

n = 0.024

No . 66 upstream along right bank from section 3,

Columbia River at Vernita, Wash .
13


n = 0.026

3-1215 . Indian Fork below Atwood Dam, near

NewCumberland, Ohio
Gage location.-Lat 40 31'30", long 8117'20", in SEY4 sec. 28,

T. 15 N., R. 7 W., on left bank 500 ft downstream from

Atwood Dam, 0 .5 mile upstream from mouth, and 1 .5 miles
southeast of New Cumberland, Tuscarawas County. Sec
tion 1 is about 300 ft downstream from gage .
Drainage area.-70 .3 sq mi.
Date of food.-May 11, 1948 .
Gage height .-10.27 ft at gage ; 9.99 ft at section 1 .
Peak discharge .-768 cfs.
Computed roughness coefficient. -Manning n=0.026 .
Description of channel .-Bed and banks are composed of clay.
Banks are clear except for short grass and exposed tree roots
in some places.
Reach Properties

Section

1. . . . . . . . . . .

2. . . . . . . . . . .
3. .... ......

Notes.-

Area
(sq ft)

280
273
279

Top
width
(ft)

52
51
52

Mean
depth
(ft)

5.4
5.4
5.4

Hydraulic
radius
(ft)

4.87
4.82
4 .97

Mean
velocity
(ft per sec)

Length (ft)
between
sections

Fall (ft)
between
sections

.
2.74 . . . . . . . . . . . . . , . .
2.82
257
0.08
2.76
202
.05

n = 0.026

CROSS SECTIONS

12
10
8
6
4
2
10
8
6
4
2
10
8
6
4
2

10

20

30

40

50

60

WIDTH, IN FEET

Plan sketch and cross sections, Indian Fork below

Atwood Dam, near New Cumberland, Ohio.

15

n = 0.026

No. 327 upstream from right bank below section 3,

Indian Fork below Atwood Dam, near New Cumberland, Ohio.
16

n = 0.026

No. 329 upstream from right bank at section 2,

Indian Fork below Atwood Dam, near New Cumberland, Ohio.
17


n = 0.027

8-1235 . Champlin Creek near Colorado City, Tex.

Gage location .-Lat 3219', long 10049', on right bank 600 ft

downstream from South Fork, 5 miles southeast of Colorado
City, Mitchell County, and 5 .5 miles upstream from mouth .
Section 2 is 350 ft downstream from gage .
Drainage area.-158 sq mi .
Date off ood.-May 17, 1949 .
Gage height.-5 .05 ft at gage ; 4 .24 ft at section 2 .
Peak discharge.-2,390 cfs.
Computed roughness coefficient. -Manning n=0.027 .
Description of channel.-Bed consists of gravel deposits over
smooth to rough rock. Banks are covered with grass and
have a few outcrops .
Reach properties
Section
2. . . . . . . . . . .

3. . . . . . . . . . .
4. . . . . . . . . . .

Notes.-

Area
(sq ft)

412
344
307

Top
width
(ft)

85
79
70

Mean
depth
(ft)

4.8
4.4
4.4

Hydraulic
radius
(ft)

Mean
velocity
(ft per s_ "7)

4.71
4.20
4.24

5 .80
6.96
7.78

Length (ft)
between
sections

Fall (ft)
between
sections

. . . . . . . . . . . . . . . .
0.43
176
148
.71

n = 0.027

CROSS SECTIONS

a0
W

a
H
W
W

Z
Z

_Z

J
W

Plan sketch and cross sections, Champlin Creek near

Colorado City, Tex .
19

n = 0.027

No. 511 downstream along right bank from above section 2,

Champlin Creek near Colorado City, Tex.
20

n = 0.027

No. 512 upstream along left bank from below section 4,

Champlin Creek near Colorado City, Tex.

21


n = 0.028

12-3 545 . Clark Fork at St. Regis, Mont.

Gage location .-Lat 47 18'05", long 115 05'15", in center of

SW Y4 sec. 19, T. 18 N., R. 27 W., on left bank at St. Regis,
0.5 mile downstream from St. Regis River. Section 1 is
660 ft upstream from gage .
Drainage area.-10,709 sq mi .
Date offlood.-May 24, 1948 .
Gage height.-19 .96 ft at gage ; 20.42 ft at section 1 .
Peak discharge .-68,900 cfs.
Computed roughness coefficient.-Manning n = 0.028 .
Description of channel .-Bed consists of well-rounded boulders ;
d5o =135 mm, d84 = 205 mm. Banks are composed of gravel
and boulders, and have tree and brush cover.
Reach properties
Section
1 . . . . . . . . . . .

2. . . . . . . . . . .
3. . . . . . . . . . .

Notes.-

Area
(sq ft)

6,860
6,976
7,194

Top
width
(ft)

404
429
454

Mean
depth
(ft)

16 .98
16 .26
15 .85

Hydraulic
radius
(ft)

Mean
velocity
(ft per sec)

16 .70
16 .04
15 .64

10 .04
9.88
9.58

Length (ft)
between
sections

Fall (ft)
between
sections

. . .. . . . . . . . . . . . .
755
0.555
438
.32

n = 0.028

CROSS SECTIONS

a
0
W

20

H
W
W
W

10

Z
Z
F

W
J
W

Plan sketch and cross sections, Clark Fork at St. Regis, Mont.

23

n = 0.028

No. 22 downstream along right bank from section 2,

Clark Fork at St. Regis, Mont.
24

n = 0.028

No. 23 upstream along left bank from section 2,

Clark Fork at St . Regis, Mont.
25

n = 0 .030

12-3405 . Clark Fork above Missoula, Mont.

Gage location .-Lat 4652'40", long 113 55'40", in NWY4NWj

sec . 19, T . 13 N, R. 18 W., on right bank 3 miles downstream from Blackfoot River and 3 miles east of Missoula .
Section 1 is 405 ft upstream from gage .
Drainage area.-5,999 sq mi .
Date offlood.-May 23, 1948 .
Gage height.-13 .07 ft at gage ; 14 .54 ft at section 1 .
Peak discharge.-31,500 cfs .
Computed roughness coefficient. -Manning n = 0.030.
Description of channel.-Bed is composed of sand, gravel, and
boulders ; d5o =175 mm, d84 = 325 mm . Thick undergrowth
is along right bank and along the left bank in the lower
part of the reach.
Reach Properties
Section
1 . . . . . . . . . . .

2. ... . ... ...

3. ... . .. . .. .
4. ... . .. . ...

Notes. -

Area
(sq ft)

3,866
3,461
3,634
3,798

Top
width
(ft)

285
267
294
312

Mean
depth
(ft)

13 .56
12 .96
12 .36
12 .17

Hydraulic
radius
(ft)

Mean
velocity
(ft per sec)

13 .24
12 .64
12 .10
11 .95

8 .15
9 .10
8 .67
8 .29

Length (ft)
between
sections

Fall (ft)
between
sections

. . . . . . .. . . .. . .. .
305
0 .63
243
.25
297
.18

n = 0.030
PLAN SKETCH
2

/1

I
2

Gage

CROSS SECTIONS

a
0
W

a
0
W
W
z
z

0
a

W
J
W

WIDTH, IN FEET

Plan sketch and cross sections, Clark Fork above

Missoula, Mont .

27

n = 0.030

No. 18 downstream along left bank from above section 3,

Clark Fork above Missoula, Mont.
28

n = 0.030

No. 19 downstream through reach from bridge 400 ft above

section 1, Clark Fork above Missoula, Mont.
29

Guide for Selecting Manning's

Roughness Coefficients for
Natural Channels and Flood Plains

United States
Geological
Survey
Water-Supply
Paper 2339

Prepared in
cooperation
with the
United States
Department of
Transportation,
Federal Highway
Administration

Guide for Selecting Manning's

Roughness Coefficients for
Natural Channels and Flood Plains
By GEORGE J. ARCEMENT, JR., and VERNE R. SCHNEIDER

Prepared in cooperation with the

U.S. Department of Transportation,
Federal Highway Administration

A guide presenting step-by-step procedures for selecting

Manning's roughness coefficient, n, for natural channels
and flood plains. Photographs of flood-plain segments
can be used for comparison with similar flood plains
to aid in assigning n values

U.S. GEOLOGICAL SURVEY WATER-SUPPLY PAPER 2339

DEPARTMENT OF THE INTERIOR

MANUEL LUJAN, Jr., Secretary
U.S. GEOLOGICAL SURVEY
Dallas L. Peck, Director

Any use of trade, product, or firm names in this publication is for

descriptive purposes only and does not imply endorsement by the
U.S. Government

UNITED STATES GOVERNMENT PRINTING OFFICE: 1989

For sale by the

Books and Open-File Reports Section,
U.S. Geological Survey,
Federal Center, Box 25425,
Denver, CO 80225
Library of Congress Cataloging in Publication Data
Arcement, George J.
Guide for selecting Manning's roughness coefficients for natural channels
and flood plains.
(Water-supply paper / United States Geological Survey ; 2339)
"Prepared in cooperation with U.S. Department of Transportation, Federal
Highway Administration."
Bibliography: p.
Supt. of Docs. no. : I 19.13:2339
1. Channels (Hydraulic engineering). 2. Flood plains. 3. Frictional resistance (Hydrodynamics). I. Schneider, V. R. II. United States. Federal
Highway Administration. III. Title. IV. Title: Manning's roughness
coefficients for natural channels and flood plains. V. Series: U.S. Geological Survey water-supply paper ; 2339.

TC175.A67

1989

627'.4

88-600129

CONTENTS
Abstract 1
Introduction 1
Methods 2
Channel n Values 2
Base n Values (nb) for Channels 2
Adjustment Factors for Channel n Values 6
Irregularity (HJ) 6
Variation in Channel Cross Section (2) 6
Obstructions (n3) 6
Vegetation (4) 6
Meandering (m) 8
Rood-Plain n Values 8
Modified Channel Method 8
Adjustment Factors for Flood-Plain n Values 8
Surface Irregularities (HJ) 8
Obstructions (3) 8
Vegetation (n4) 8
Vegetation-Density Method 9
Techniques for Determining Vegetation Density 9
Indirect Technique 10
Direct Technique 10
Photographs of Flood Plains 13
Procedures for Assigning n Values 28
Steps for Assigning n Values 28
Reach Subdivision (Steps 1 and 2) 28
Channel Roughness (Steps 3-13) 28
Rood-Plain Roughness (Steps 14-23) 29
Examples of Procedures for Determining n Values 32
Summary 32
References Cited 37
FIGURES

1.
2.
3.
4.
5.

Diagram of a schematic and cross sections of a hypothetical reach of a

channel and flood plain showing subdivisions used in assigning n values 3
Graph showing relation of stream power and median grain size to flow
regime 5
Diagram showing forms of bed roughness in sand-bed channels 5
Graph showing effective-drag coefficient for verified n values versus the
hydraulic radius of wide, wooded flood plains 10
Chart of example measurement of vegetation density showing tree diameter
and location in representative sample area 12
Contents

III

6-20.

21.
22.

Photographs showing flood plains having verified n values:

6. Cypress Creek near Downsville, La., Vegrf=0.0067 13
7. Bayou de Loutre near Farmerville, La., Vegd=0.0061 14
8. Bayou de Loutre near Farmerville, La., Vegd=0.0015 15
9. Bayou de Loutre near Farmerville, La., Vegd=0.0012 16
10. Coldwater River near Red Banks, Miss., Vegd =0.0011 17
11. Coldwater River near Red Banks, Miss., Vegd=0.0090 18
12. Yockanookany River near Thomastown, Miss., Vegd=0. 0082 19
13. Yockanookany River near Thomastown, Miss., 500 ft east of
area shown in figure 12, Vegd =O.QQS2 20
14. Flagon Bayou near Libuse, La., Vegd =O.QQSl 21
15. Pea Creek near Louisville, Ala., Vegd=O.OOS5 22
16. Pea Creek near Louisville, Ala., V<?grf=0.0102 23
17. Tenmile Creek near Elizabeth, La., V<?^=0.0067 24
18. Sixmile Creek near Sugartown, La., Vegd=O.OOS4 25
19. Thompson Creek near Clara, Miss., Vegd=0.0l\5 26
20. Thompson Creek near Clara, Miss., 500 ft east of area shown
in figure 19, Vegd =O.OU5 27
Flow chart of procedures for assigning n values 30
Sample form for computing n values 33

TABLES

1.
2.
3.
4.

Base values of Manning's n 4

Adjustment values for factors that affect the roughness of a channel 7
Adjustment values for factors that affect the roughness of flood plains 9
Outline and example of procedures for determining n values for a hypothetical
channel and adjoining flood plain 35

METRIC CONVERSION FACTORS

For readers who wish to convert measurements from the inch-pound system of units to the metric system
of units, the conversion factors are listed below:
Multiply inch-pound unit

cubic foot per second (ft3/s)

foot (ft)
foot per second (ft/s)
foot per square second (ft/s2)
inch (in.)
square foot (ft2 )
pounds per square foot (lb/ft2)

IV

Contents

By

0.02832
.3048
.3048
.3048
25.40
.0929
4.882

To obtain metric unit

cubic meter per second (m3/s)

meter (m)
meter per second (m/s)
meter per square second (m/s2)
millimeter (mm)
square meter (m2)
kilograms per square meter (km/m2)

GLOSSARY

A
SA,.
C*
dS4
g
h
K
L
/
m
n

Cross-sectional area of flow (ft2).

The total frontal area of vegetation blocking the flow (ft2).
Effective drag coefficient for vegetation.
Particle diameter that equals or exceeds that of 84 percent of the particles (ft).
Gravitational constant (ft/s2).
Height of water on flood plain (ft).
Conveyance of a channel section (ft3/s).
Length of channel reach being considered (ft).
Length of representative sample area (ft).
Correction factor for meandering of channel or flood plain.
Manning's roughness coefficient, including boundary and vegetation effects
(ft176).

%
"Zrijdj
^
nl
n2
%
n4
n4 '

Base value of n for the surface material of the channel or flood plain (ft1/6).
Summation of number of trees in a sample area multiplied by tree diameter (ft).
Value of n, excluding the effect of vegetation (ft176).
Value of n for the effect of surface irregularity (ft1/6).
Value of n for variations in shape and size of channel or flood plain (ft176).
Value of n for obstructions (ft 176).
Value of n for vegetation (ft1/6).
Value of n used in determining n0 , representing vegetation not accounted for in
vegetation density (ft 176).
Hydraulic radius (ft).
Slope of energy-grade line (ft/ft).
Slope of water-surface profile (ft/ft).
Stream power ((ft-lb/s)/ft2).
Mean velocity of flow (ft/s).
Vegetation density (ft" 1 ).
Vegetation resistivity (ft~ l ).
Width of representative sample area (ft).

R
Se
Sw
SP
V
Ve8d
Vegr
w

Contents

Guide for Selecting Manning's Roughness

Coefficients for Natural Channels and Flood Plains
By George J. Arcement, Jr., and Verne R. Schneider
Abstract
Although much research has been done on Manning's roughness coefficient, n, for stream channels, very
little has been done concerning the roughness values for
densely vegetated flood plains. The n value is determined
from the values of the factors that affect the roughness of
channels and flood plains. In densely vegetated flood
plains, the major roughness is caused by trees, vines, and
brush. The /? value for this type of flood plain can be
determined by measuring the vegetation density of the
flood plain.
Photographs of flood-plain segments where n values
have been verified can be used as a comparison standard
to aid in assigning n values to similar flood plains.

INTRODUCTION

Roughness coefficients represent the resistance to

flood flows in channels and flood plains. The results of
Manning's formula, an indirect computation of streamflow,
have applications in flood-plain management, in floodinsurance studies, and in the design of bridges and highways across flood plains.
Manning's formula is
^1486^2/3,1/2

(1)

where
V=mean velocity of flow, in feet per second,
R= hydraulic radius, in feet,
Se = slope of energy grade line, in feet per foot, and
n=Manning's roughness coefficient.
When many calculations are necessary in using Manning's formula, using a conveyance term is sometimes
convenient. Conveyance is defined as
K=-1.486-AR2/3

(2)

where
K= conveyance of the channel, in cubic feet per
second,

A = cross-sectional area of channel, in square feet,

R = hydraulic radius, in feet, and
n= Manning's roughness coefficient.
The term K, known as the conveyance of the channel
section, is a measure of the carrying capacity of the channel
section.
Suggested values for Manning's n, tabulated according to factors that affect roughness, are found in Chow
(1959), Henderson (1966), and Streeter (1971). Roughness
characteristics of natural channels are given by Barnes
(1967). Barnes presents photographs and cross sections of
typical rivers and creeks and their respective n values.
It would be impractical in this guide to record all that
is known about the selection of the Manning's roughness
coefficient, but many textbooks and technique manuals
contain discussions of the factors involved in the selection.
Three publications that augment this guide are Barnes
(1967), Chow (1959), and Ree (1954). Although much
research has been done to determine roughness coefficients
for open-channel flow (Carter and others, 1963), less has
been done for densely vegetated flood plains, coefficients
for which are typically very different from those for
channels.
The step-by-step procedures described in this guide
outline methods for determining Manning's n values for
natural channels and flood plains. The n values are used to
compute the flow information needed by engineers in the
design of highways that cross these environments.
Aldridge and Garrett (1973) attempted to systematize
the selection of roughness coefficients for Arizona streams.
In this guide, we attempt to broaden the scope of that work;
in particular, to describe procedures for the selection of
roughness coefficients for densely vegetated flood plains.
There is a tendency to regard the selection of roughness coefficients as either an arbitrary or an intuitive
process. Specific procedures can be used to determine the
values for roughness coefficients in channels and flood
plains. The n values for channels are determined by
evaluating the effects of certain roughness factors in the
channels. Two methods also are presented to determine the
roughness coefficients of flood plains. One method, similar
Introduction

to that for channel roughness, involves the evaluation of the

effects of certain roughness factors in the flood plain. The
other method involves the evaluation of the vegetation
density of the flood plain to determine the n value. This
second method is particularly suited to handle roughness for
densely wooded flood plains. Photographs of flood plains
that have known n values are presented for comparison to
flood plains that have unknown n values.

METHODS

Values of the roughness coefficient, n, may be

assigned for conditions that exist at the time of a specific
flow event, for average conditions over a range in stage, or
for anticipated conditions at the time of a future event. The
procedures described in this report are limited to the
selection of roughness coefficients for application to onedimensional, open-channel flow. The values are intended
mostly for use in the energy equation as applied to onedimensional, open-channel flow, such as in a slope-area or
step-backwater procedure for determining flow.
The roughness coefficients apply to a longitudinal
reach of channel and (or) flood plain. A hypothetical reach
of a channel and flood plain is shown in figure 1. The cross
section of the reach may be of regular geometric shape
(such as triangular, trapezoidal, or semicircular) or of an
irregular shape typical of many natural channels. The flow
may be confined to one or more channels, and, especially
during floods, the flow may occur both in the channel and
in the flood plain. Such cross sections may be termed
compound channels, consisting of channel and flood-plain
subsections. Cross sections are typically divided into subsections at points where major roughness or geometric
changes occur, such as at the juncture of dense woods and
pasture or flood plain and main channel. However, subsections should reflect representative conditions in the reach
rather than only at the cross section. Roughness coefficients
are determined for each subsection, and the procedures
described herein apply to the selection of roughness coefficients for each subsection.
There are several means of compositing the results to
obtain an equivalent n value for a stream cross section.
These procedures, summarized by Chow (1959, p. 136),
use each of the following three assumptions: (1) the mean
velocity in each subsection of the cross section is the same;
(2) the total force resisting the flow is equal to the sum of
the forces resisting the flows in the subdivided areas; and
(3) the total discharge of the flow is equal to the sum of the
discharges of the subdivided areas. Also, the slope of the
energy grade line is assumed to be the same for each of the
subsections. In some cases, computing the equivalent n
value is not necessary. Instead, the subsection conveyances,
which are additive, are computed by employing assumption
3 to obtain the total conveyance for the cross section.
2

Roughness values for flood plains can be quite

different from values for channels; therefore, roughness
values for flood plains should be determined independently
from channel values. As in the computation of channel
roughness, a base roughness (nb) is assigned to the flood
plain, and adjustments for various roughness factors are
made to determine the total n value for the flood plain.
Seasonal variability of roughness coefficients should
be considered. Floods often occur during the winter when
there is less vegetation. Thus, the field surveys, including
photographs, may not be completed until spring when
vegetation growth is more dense. A variable roughness
coefficient may be needed to account for these seasonal
changes.
In developing the ability to assign n values, reliance
must be placed on n values that have been verified. A
verified n value is one that has been computed from known
cross-sectional geometry and discharge values.

CHANNEL n VALUES

The most important factors that affect the selection of

channel n values are (1) the type and size of the materials
that compose the bed and banks of the channel and (2) the
shape of the channel. Cowan (1956) developed a procedure
for estimating the effects of these factors to determine the
value of n for a channel. The value of n may be computed
by
n=(nb +n 1 +n2 +n3 +n4)m

(3)

where
nb =a. base value of n for a straight, uniform, smooth
channel in natural materials,
,
/ij =a correction factor for the effect of surface irregularities,
2 =a value for variations in shape and size of the
channel cross section,
3 =a value for obstructions,
n4 =a value for vegetation and flow conditions, and
m =a correction factor for meandering of the channel.
Base n Values (nh) for Channels

In the selection of a base n value for channel

subsections, the channel must be classified as a stable
channel or as a sand channel.
A stable channel is defined as a channel in which the
bed is composed of firm soil, gravel, cobbles, boulders, or
bedrock and the channel remains relatively unchanged
throughout most of the range in flow. Table 1 (modified
from Aldridge and Garrett, 1973) lists base nb values for
stable channels and sand channels. The base values of

Guide for Selecting Manning's Roughness Coefficients for Natural Channels and Flood Plains

ft
108-

Water surface

64-

20

40

60ft

CROSS SECTION 1

Segments
2
3

80ft

CROSS SECTION 2

Subsections
2
I
I

1,000'

I
I
I 30' I

1,000'

I
I

oooooo
oooooo
CROSS

ooooo
ooooo
o Cotton
o fields
ooooo
ooooo

SECTION
3
(Not to scale)

CROSS SECTION 3

Figure 1. A schematic and cross sections of a hypothetical reach of a channel and flood plain showing subdivisions
used in assigning n values.

Benson and Dalrymple (1967) apply to conditions that are

close to average, whereas Chow's (1959) base values are
for the smoothest reach attainable for a given bed material.
Barnes (1967) cataloged verified n values for stable
channels having roughness coefficients ranging from 0.024
to 0.075. In addition to a description of the cross section,

bed material, and flow conditions during the measurement,

color photographs of the channels were provided.
A sand channel is defined as a channel in which the
bed has an unlimited supply of sand. By definition, sand
ranges in grain size from 0.062 to 2 mm. Resistance to flow
varies greatly in sand channels because the bed material
Channel n Values

Table 1. Base values of Manning's n

[Modified from Aldridge and Garrett, 1973, table 1; , no data]

Median size of
bed material
(in millimeters)

Bed
material

Base n value
Straight
Smooth
uniform
channel2
channel1

Sand channels
0.2
Sand3 .......... ........
0.012
.3
.017
.4
.020
.5
.022
.6
.023
.8
.025
1.0
.026
Stable channels and flood plains
_
Concrete .......
0.012-0.018
Rock cut .......
Coarse sand .... ........
Fine gravel .....
........
Coarse gravel . . .
Cobble......... ........
Boulder ........ ........

1-2

0.025-0.032
0.026-0.035

2-64

0.028-0.035

64-256
>256

0.030-0.050
0.040-0.070

0.011
.025
.020
.024
.026

1 Benson and Dalrymple (1967).

2 For indicated material; Chow (1959).
3 Only for upper regime flow where grain roughness is predominant.

moves easily and takes on different configurations or bed

forms. Bed form is a function of velocity of flow, grain
size, bed shear, and temperature. The flows that produce
the bed forms are classified as lower regime flow and upper
regime flow, according to the relation between depth and
discharge (fig. 2). The lower regime flow occurs during low
discharges, and the upper regime flow occurs during high
discharges. An unstable discontinuity, called a transitional
zone, appears between the two regimes in the depth to
discharge relation (fig. 3). In lower regime flow, the bed
may have a plane surface and no movement of sediment, or
the bed may be deformed and have small uniform waves or
large irregular saw-toothed waves formed by sediment
moving downstream. The smaller waves are known as
ripples, and the larger waves are known as dunes. In upper
regime flow, the bed may have a plane surface and sediment
movement or long, smooth sand waves that are in phase
with the surface waves. These waves are known as standing
waves and antidunes. Bed forms on dry beds are remnants
of the bed forms that existed during receding flows and may
not represent flood stages.
The flow regime is governed by the size of the bed
materials and the stream power, which is a measure of
energy transfer. Stream power (SP) is computed by the
formula:
SP=62 RS^V
4

where
62= specific weight of water, in pounds per cubic foot,
/?=hydraulic radius, in feet,
Sw = water-surface slope, in feet per foot, and
V=mean velocity, in feet per second.
The values in table 1 for sand channels are for upper
regime flows and are based on extensive laboratory and
field data obtained by the U.S. Geological Survey. When
using these values, a check must be made to ensure that the
stream power is large enough to produce upper regime flow
(fig. 2). Although the base n values given in table 1 for
stable channels are from verification studies, the values
have a wide range because the effects of bed roughness are
extremely difficult to separate from the effects of other
roughness factors. The choice of n values selected from
table 1 will be influenced by personal judgment and
experience. The n values for lower and transitional-regime
flows are much larger generally than the values given in
table 1 for upper regime flow. Simons, Li, and Associates
(1982) give a range of n values commonly found for
different bed forms.
The n value for a sand channel is assigned for upper
regime flow by using table 1, which shows the relation
between median grain size and the n value. The flow regime
is checked by computing the velocity and stream power that
correspond to the assigned n value. The computed stream
power is compared with the value that is necessary to cause
upper regime flow (see fig. 2, from Simons and Richardson, 1966, fig. 28). If the computed stream power is not
large enough to produce upper regime flow (an indication of
lower regime or transitional-zone flow), a reliable value of
n cannot be assigned. The evaluation of n is complicated by
bed-form drag. Different equations are needed to describe
the bed forms. The total n value for lower and transitionalregime flows can vary greatly and depends on the bed forms
present at a particular time. Figure 3 illustrates how the total
resistance in a channel varies for different bed forms.
Limerinos (1970) related n to hydraulic radius and
particle size on the basis of samples from 11 stream
channels having bed material ranging from small gravel to
medium-sized boulders. Particles have three dimensions
length, width, and thickness and are oriented so that
length and width are parallel to the plane of the streambed.
Limerinos related n to minimum diameter (thickness) and to
intermediate diameter (width). His equation using intermediate diameter appears to be the most useful because this
dimension is the most easy to measure in the field and to
estimate from photographs.
The equation for n using intermediate diameter is
n=

(0.0926) R 1/6
1.16+2. 0 log (JL\

(4)

Guide for Selecting Manning's Roughness Coefficients for Natural Channels and Flood Plains

\dj

(5)

n>

3
3

n
a*

Q.

3
Q.
CT

c.
era

3
(/>
o^
cr
o.
o

n
O

3
U>

QTQ

m
DO

T)
O

C/)
H
3J
m

28).

66

and
mons

regilow
me

power
g
stmedireaman
and

Ri19
chardson, FiRel2.
gureation

L IMETERS

I
ZE

MEDIAN
GRAI
S

CD

CD
<Q
~3

CD

CD
_<Q

C
-a

STREAM POWER (62RS W V), IN FOOT-POUNDS

PER SECOND PER SQUARE FOOT

where
R= hydraulic radius, in feet, and
dS4 =the particle diameter, in feet, that equals or exceeds
the diameter of 84 percent of the particles
(determined from a sample of about 100 randomly distributed particles).
Limerinos selected reaches having a minimum amount of
roughness, other than that caused by bed material, and
corresponding to the average base values given by Benson
and Dalrymple (1967) shown in table 1.
Burkham and Dawdy (1976) showed that equation 5
applies to upper regime flow in sand channels. If a
measured dS4 is available or can be estimated, equation 5
may be used to obtain a base n for sand channels in lieu of
using table 1.

Adjustment Factors for Channel n Values

The nb values selected from table 1 or computed from
the Limerinos equation are for straight channels of nearly
uniform cross-sectional shape. Channel irregularities, alignment, obstructions, vegetation, and meandering increase
the roughness of a channel. The value for n must be
adjusted accordingly by adding increments of roughness to
the base value, nb , for each condition that increases the
roughness. The adjustments apply to stable and sand channels. Table 2, modified from Aldridge and Garrett (1973),
gives ranges of adjustments for the factors that affect
channel roughness for the prevailing channel conditions.
The average base values of Benson and Dalrymple (1967)
from table 1 and the values computed from equation 5 apply
to near-average conditions and, therefore, require smaller
adjustments than do the smooth-channel base values of
Chow (1959). Likewise, the adjustments (from table 2)
made to base values of Benson and Dalrymple (1967)
should be reduced slightly.
Depth of flow must be considered when selecting n
values for channels. If the depth of flow is shallow in
relation to the size of the roughness elements, the n value
can be large. The n value decreases with increasing depth,
except where the channel banks are much rougher than the
bed or where dense brush overhangs the low-water channel.
Irregularity (ir,)

Where the ratio of width to depth is small, roughness

caused by eroded and scalloped banks, projecting points,
and exposed tree roots along the banks must be accounted
for by fairly large adjustments. Chow (1959) and Benson
and Dalrymple (1967) showed that severely eroded and
scalloped banks can increase n values by as much as 0.02.
Larger adjustments may be required for very large, irregular
banks that have projecting points.
6

Variation in Channel Cross Section (n2)

The value of n is not affected significantly by

relatively large changes in the shape and size of cross
sections if the changes are gradual and uniform. Greater
roughness is associated with alternating large and small
cross sections and sharp bends, constrictions, and sideto-side shifting of the low-water channel. The degree of the
effect of changes in the size of the channel depends
primarily on the number of alternations of large and small
sections and secondarily on the magnitude of the changes.
The effects of abrupt changes may extend downstream for
several hundred feet. The n value for a reach below a
disturbance may require adjustment, even though none of
the roughness-producing factors are apparent in the study
reach. A maximum increase in n of 0.003 will result from
the usual amount of channel curvature found in designed
channels and in the reaches of natural channels used to
compute discharge (Benson and Dalrymple, 1967).
Obstructions (n3)

Obstructions such as logs, stumps, boulders,

debris, pilings, and bridge piers disturb the flow pattern in
the channel and increase roughness. The amount of increase
depends on the shape of the obstruction; the size of the
obstruction in relation to that of the cross section; and the
number, arrangement, and spacing of obstructions. The
effect of obstructions on the roughness coefficient is a
function of the flow velocity. When the flow velocity is
high, an obstruction exerts a sphere of influence that is
much larger than the obstruction because the obstruction
affects the flow pattern for considerable distances on each
side. The sphere of influence for velocities that generally
occur in channels that have gentle to moderately steep
slopes is about three to five times the width of the
obstruction. Several obstructions can create overlapping
spheres of influence and may cause considerable disturbance, even though the obstructions may occupy only a
small part of a channel cross section. Chow (1959) assigned
adjustment values to four levels of obstruction: negligible,
minor, appreciable, and severe (table 2).
Vegetation (n4)

The extent to which vegetation affects n depends on

the depth of flow, the percentage of the wetted perimeter
covered by the vegetation, the density of vegetation below
the high-water line, the degree to which the vegetation is
flattened by high water, and the alignment of vegetation
relative to the flow. Rows of vegetation that parallel the
flow may have less effect than rows of vegetation that are
perpendicular to the flow. The adjustment values given in

Guide for Selecting Manning's Roughness Coefficients for Natural Channels and Flood Plains

Table 2. Adjustment values for factors that affect the roughness of a channel
[Modified from Aldridge and Garrett, 1973, table 2]
Channel conditions

n value
adjustment1

Smooth
Minor

0.000
0.001-0.005

Moderate

0.006-0.010

Severe

0.011-0.020

Gradual
Alternating
occasionally

0.000
0.001-0.005

Alternating
frequently
Negligible

0.010-0.015
0.000-0.004

Minor

0.005-0.015

Appreciable

0.020-0.030

Severe

0.040-0.050

Small

0.002-0.010

Medium

0.010-0.025

Large

0.025-0.050

Very large

0.050-0.100

Minor
Appreciable
Severe

1.00
1.15
1.30

Degree of
irregularity

Variation
in channel
cross section
(2)

Effect of
obstruction
(3)

Amount of
vegetation

Degree of
meandering2
(m)

Example

Compares to the smoothest channel attainable in a given bed material.

Compares to carefully dredged channels in good condition but having slightly
eroded or scoured side slopes.
Compares to dredged channels having moderate to considerable bed roughness
and moderately sloughed or eroded side slopes.
Badly sloughed or scalloped banks of natural streams; badly eroded or sloughed
sides of canals or drainage channels; unshaped, jagged, and irregular surfaces
of channels in rock.
Size and shape of channel cross sections change gradually.
Large and small cross sections alternate occasionally, or the main flow
occasionally shifts from side to side owing to changes in cross-sectional
shape.
Large and small cross sections alternate frequently, or the main flow frequently
shifts from side to side owing to changes in cross-sectional shape.
A few scattered obstructions, which include debris deposits, stumps, exposed
roots, logs, piers, or isolated boulders, that occupy less than 5 percent of the
cross-sectional area.
Obstructions occupy less than 15 percent of the cross-sectional area, and the
spacing between obstructions is such that the sphere of influence around one
obstruction does not extend to the sphere of influence around another
obstruction. Smaller adjustments are used for curved smooth-surfaced objects
than are used for sharp-edged angular objects.
Obstructions occupy from 15 to 50 percent of the cross-sectional area, or the
space between obstructions is small enough to cause the effects of several
obstructions to be additive, thereby blocking an equivalent part of a cross
section.
Obstructions occupy more than 50 percent of the cross-sectional area, or the
space between obstructions is small enough to cause turbulence across most
of the cross section.
Dense growths of flexible turf grass, such as Bermuda, or weeds growing where
the average depth of flow is at least two times the height of the vegetation;
supple tree seedlings such as willow, cottonwood, arrow weed, or saltcedar
growing where the average depth of flow is at least three times the height of
the vegetation.
Turf grass growing where the average depth of flow is from one to two times the
height of the vegetation; moderately dense stemmy grass, weeds, or tree
seedlings growing where the average depth of flow is from two to three times
the height of the vegetation; brushy, moderately dense vegetation, similar to
1- to 2-year-old willow trees in the dormant season, growing along the banks,
and no significant vegetation is evident along the channel bottoms where the
hydraulic radius exceeds 2 ft.
Turf grass growing where the average depth of flow is about equal to the height
of the vegetation; 8- to 10-year-old willow or cottonwood trees intergrown
with some weeds and brush (none of the vegetation in foliage) where the
hydraulic radius exceeds 2 ft; bushy willows about 1 year old intergrown with
some weeds along side slopes (all vegetation in full foliage), and no
significant vegetation exists along channel bottoms where the hydraulic
radius is greater than 2 ft.
Turf grass growing where the average depth of flow is less than half the height
of the vegetation; bushy willow trees about 1 year old intergrown with weeds
along side slopes (all vegetation in full foliage), or dense cattails growing
along channel bottom; trees intergrown with weeds and brush (all vegetation
in full foliage).
Ratio of the channel length to valley length is 1.0 to 1.2.
Ratio of the channel length to valley length is 1.2 to 1.5.
Ratio of the channel length to valley length is greater than 1.5.

1 Adjustments for degree of irregularity, variations in cross section, effect of obstructions, and vegetation are added to the base n value (table 1)
before multiplying by the adjustment for meander.
2 Adjustment values apply to flow confined in the channel and do not apply where downvalley flow crosses meanders.
Channel n Values

table 2 apply to constricted channels that are narrow in

width. In wide channels having small depth-to-width ratios
and no vegetation on the bed, the effect of bank vegetation
is small, and the maximum adjustment is about 0.005. If the
channel is relatively narrow and has steep banks covered by
dense vegetation that hangs over the channel, the maximum
adjustment is about 0.03. The larger adjustment values
given in table 2 apply only in places where vegetation
covers most of the channel.
Meandering (m)

The degree of meandering, m, depends on the ratio of

the total length of the meandering channel in the reach being
considered to the straight length of the channel reach. The
meandering is considered minor for ratios of 1.0 to 1.2,
appreciable for ratios of 1.2 to 1.5, and severe for ratios of
1.5 and greater. According to Chow (1959), meanders can
increase the n values by as much as 30 percent where flow
is confined within a stream channel. The meander adjustment should be considered only when the flow is confined
to the channel. There may be very little flow in a meandering channel when there is flood-plain flow.

FLOOD-PLAIN n VALUES
Roughness values for channels and flood plains
should be determined separately. The composition, physical
shape, and vegetation of a flood plain can be quite different
from those of a channel.

Modified Channel Method

By altering Cowan's (1956) procedure that was
developed for estimating n values for channels, the following equation can be used to estimate n values for a flood
plain:

n=(nb +n l +n2 +n3 +n4)m

(6)

where
nb =a base value of n for the flood plain's natural bare
soil surface,
n l =a correction factor for the effect of surface irregularities on the flood plain,
n2 = a value for variations in shape and size of the
flood-plain cross section, assumed to equal 0.0,
n3 =a value for obstructions on the flood plain,
n4 =a value for vegetation on the flood plain, and
m = a correction factor for sinuosity of the flood plain,
equal to 1.0.
By using equation 6, the roughness value for the flood plain
is determined by selecting a base value of nb for the natural
bare soil surface of the flood plain and adding adjustment
factors due to surface irregularity, obstructions, and vege8

tation. The selection of an nb value is the same as outlined

for channels in Channel n Values. See table 3 for n value
adjustments for flood plains. The adjustment for crosssectional shape and size is assumed to be 0.0. The cross
section of a flood plain is subdivided where abrupt changes
occur in the shape of the flood plain. The adjustment for
meandering is assumed to be 1.0 because there may be very
little flow in a meandering channel when there is flood-plain
flow. In certain cases where the roughness of the flood plain
is caused by trees and brush, the roughness value for the
flood plain can be determined by measuring the vegetation
density of the flood plain rather than by directly estimating
from table 3 (see Vegetation-Density Method).
Adjustment Factors for Flood-Plain n Values
Surface Irregularities (n.,)

Irregularity of the surface of a flood plain causes an

increase in the roughness of the flood plain. Such physical
factors as rises and depressions of the land surface and
sloughs and hummocks increase the roughness of the flood
plain. A hummock is a low mound or ridge of earth above
the level of an adjacent depression. A slough is a stagnant
swamp, marsh, bog, or pond.
Shallow water depths, accompanied by an irregular
ground surface in pastureland or brushland and by deep
furrows perpendicular to the flow in cultivated fields, can
increase the n values by as much as 0.02.
Obstructions (n3)

The roughness contribution of some obstructions on a

flood plain, such as debris deposits, stumps, exposed roots,
logs, or isolated boulders, cannot be measured directly but
must be considered. Table 3 lists values of roughness for
different percentages of obstruction occurrence.
Vegetation (n4)

Visual observation, judgment, and experience are

used in selecting adjustment factors for the effects of
vegetation from table 3. An adjustment factor for tree trunks
and other measurable obstacles is described in the
Vegetation-Density Method. Although measuring the area
occupied by tree trunks and large diameter vegetation is
relatively easy, measuring the area occupied by low vines,
briars, grass, or crops is more difficult (table 3).
In the case of open fields and cropland on flood
plains, several references are available to help determine the
roughness factors. Ree and Crow (1977) conducted experiments to determine roughness factors for gently sloping
earthen channels planted with wheat, sorghum, lespedeza,
or grasses. The roughness factors were intended for application in the design of diversion terraces. However, the data
can be applied to the design of any terrace, or they can be
used to estimate the roughness of cultivated flood plains.

Guide for Selecting Manning's Roughness Coefficients for Natural Channels and Flood Plains

Table 3. Adjustment values for factors that affect roughness of flood plains
[Modified from Aldridge and Garrett, 1973, table 2]

Smooth

n value
adjustment
0.000

Minor

0.001-0.005

Moderate
Severe

0.006-0.010
0.011-0.020

Flood-plain conditions

Degree of
irregularity (n t)

Variation of
flood-plain
cross section
("2)
Effect of
obstructions
(n3)

0.0
Negligible

0.000-0.004

Minor
Appreciable
Small

0.005-0.019
0.020-0.030
0.001-0.010

Medium

0.011-0.025

0.025-0.050

Amount of
vegetation (4)

Very large

0.050-0.100

Extreme

0.100-0.200

Degree of
meander (ni)

1.0

Example
Compares to the smoothest, flattest flood plain attainable in a given bed
material.
Is a flood plain slightly irregular in shape. A few rises and dips or sloughs
may be visible on the flood plain.
Has more rises and dips. Sloughs and hummocks may occur.
Flood plain very irregular in shape. Many rises and dips or sloughs are visible.
Irregular ground surfaces in pastureland and furrows perpendicular to the
flow are also included.
Not applicable.
Few scattered obstructions, which include debris deposits, stumps, exposed
roots, logs, or isolated boulders, occupy less than 5 percent of the crosssectional area.
Obstructions occupy less than 15 percent of the cross-sectional area.
Obstructions occupy from 15 to 50 percent of the cross-sectional area.
Dense growth of flexible turf grass, such as Bermuda, or weeds growing where
the average depth of flow is at least two times the height of the vegetation,
or supple tree seedlings such as willow, cottonwood, arrowweed, or saltcedar
growing where the average depth of flow is at least three times the height of
the vegetation.
Turf grass growing where the average depth of flow is from one to two times the
height of the vegetation, or moderately dense stemmy grass, weeds, or tree
seedlings growing where the average depth of flow is from two to three times
the height of the vegetation; brushy, moderately dense vegetation, similar to
1- to 2-year-old willow trees in the dormant season.
Turf grass growing where the average depth of flow is about equal to the height
of the vegetation, or 8- to 10-year-old willow or cottonwood trees intergrown
with some weeds and brush (none of the vegetation in foliage) where the
hydraulic radius exceeds 2 ft, or mature row crops such as small vegetables,
or mature field crops where depth of flow is at least twice the height of the
vegetation.
Turf grass growing where the average depth of flow is less than half the height
of the vegetation, or moderate to dense brush, or heavy stand of timber with
few down trees and little undergrowth where depth of flow is below branches,
or mature field crops where depth of flow is less than the height of the
vegetation.
Dense bushy willow, mesquite, and saltcedar (all vegetation in full foliage), or
heavy stand of timber, few down trees, depth of flow reaching branches.
Not applicable.

Chow (1959) presents a table showing minimum,

normal, and maximum values of n for flood plains covered
by pasture and crops. These values are helpful for comparing the roughness values of flood plains having similar
vegetation.

Vegetation-Density Method
For a wooded flood plain, the vegetation-density
method can be used as an alternative to the previous method
for determining n values for flood plains. In a wooded flood
plain, where the tree diameters can be measured, the
vegetation density of the flood plain can be determined.

Determining the vegetation density is an effective

way of relating plant height and density characteristics, as a
function of depth of flow, to the flow resistance of vegetation. Application of the flow-resistance model presented
below requires an estimate of the vegetation density as a
function of depth of flow. The procedure requires a direct or
indirect determination of vegetation density at a given
depth. If the change in n value through a range in depth is
required, then an estimation of vegetation density through
that range is necessary.
Techniques for Determining Vegetation Density

Petryk and Bosmajian (1975) developed a method of

analysis of the vegetation density to determine the roughFlood Plain n Values

ness coefficient for a densely vegetated flood plain. By

summing the forces in the longitudinal direction of a reach
and substituting in the Manning's formula, they developed
the following equation:

n=nQ -

14 -

12

(7)

where
n0 = Manning's boundary-roughness coefficient, excluding the effect of the vegetation (a base n),
C* =the effective-drag coefficient for the vegetation in
the direction of flow,
2A,=the total frontal area of vegetation blocking the
flow in the reach, in square feet,
g=the gravitational constant, in feet per square second,
A=the cross-sectional area of flow, in square feet,
L=the length of channel reach being considered, in
feet, and
fl=the hydraulic radius, in feet.
Equation 7 gives the n value in terms of the boundary
roughness, 0 , the hydraulic radius, R, the effective-drag
coefficient, C*, and the vegetation characteristics, SA/AL.
The vegetation density, Vegd, in the cross section is
represented by
(8)

The boundary roughness, n0 , can be determined from

the follpwing equation:
(9)
The definition of the roughness factors nb and nl
through 3 are the same as those in equation 6 and are
determined by using table 3. The 4 ' factor, which could
not be measured directly in the Vegd term, is for vegetation,
such as brush and grass, on the surface of the flood plain.
The n/ factor is defined in the small to medium range in
table 3 because the tree canopy will prohibit a dense
undergrowth in a densely wooded area.
The hydraulic radius, R, is equal to the crosssectional area of flow divided by the wetted perimeter;
therefore, in a wide flood plain the hydraulic radius is equal
to the depth of flow. An effective-drag coefficient for
densely wooded flood plains can be selected from figure 4,
a graph of effective-drag coefficient for verified n values
versus hydraulic radius of densely wooded flood plains.
Indirect Technique

A vegetation resistivity value, Vegr, can be determined through indirect methods (Petryk and Bosmajian,
10

16

10

HYDRAULIC RADIUS ( R ), IN FEET

Figure 4. Effective-drag coefficient for verified n values

versus the hydraulic radius of wide, wooded flood plains.

1975). When flood data that include a measured discharge

and depth of flow are available, hydraulic analysis can be
made, and the roughness coefficients can be determined for
a flood plain. By rearranging equation 7 and by using the
hydraulic radius and n value computed from the discharge
measurement and an assumed WQ, the vegetation resistivity
for the reported flood can be determined from:
Vegr

AL

d.49)2/^3 '

(10)

The value of Vegr determined at this known depth of

flow can be used to estimate Vegr for other depths by
estimating the change in the density of growth. An estimate
of the change in density can be done from pictorial or
physical descriptions of the vegetation. By evaluating the
change in Vegr, an evaluation of the n value as a function of
flow depth can be determined.
Direct Technique

Tree trunks are major contributors to the roughness

coefficient in a densely wooded flood plain. Where trees are
the major factor, the vegetation density can be easily
determined by measuring the number of trees and trunk
sizes in a representative-sample area. The n value as a
function of height can be computed by using equation 7.
A representative-sample area must be chosen on the
cross section to represent the roughness of the cross section
accurately. The flood plain can be divided into subsections

Guide for Selecting Manning's Roughness Coefficients for Natural Channels and Flood Plains

on the basis of geometric and (or) roughness differences in

the cross section. The vegetation density is determined for
each subsection.
The sampling area must be representative of the
roughness coefficient of the cross section. By closely
examining the cross section in the field, a representativesample area can be chosen. Another way to more accurately
determine the roughness coefficient is to select several
representative areas and compare the results. Cross sections
should be divided into subsections when changes in roughness properties occur.
All of the trees, including vines, in the sampling area
must be counted, and the diameters must be measured to the
nearest 0.1 ft. Each tree diameter is measured to give an
average diameter for the expected flow depth of the sample
area.
Determining the area occupied by trees within the
sampling area is not difficult. A sampling area 100 ft along
the cross section by 50 ft in the flow direction is adequate to
determine the vegetation density of an area when the sample
area is representative of the flood plain. A 100-ft tape is
stretched out perpendicular to the flow direction in the
sample area. Every tree within 25 ft along either side of the
100-ft tape is counted. The position of the tree is plotted on
a grid system by measuring the distance to each tree from
the center line along the 100-ft tape, and the diameter of the
tree is recorded on the grid system (see fig. 5).
The area, SA,-, occupied by trees in the sampling area
can be computed from the number of trees, their diameter,
and the depth of flow in the flood plain. Once the vegetation
area, XA,-, is determined, the vegetation density can be
computed by using equation 8, and the n value for the
subsection can be determined by using equation 7 and
appropriate values for n0 , R, and C*.
Equation 8 can be simplified to
. h
Vegd= =
AL
hwl

(11)

where
2/V/,-=the summation of number of trees multiplied by
tree diameter, in feet,
h= height of water on flood plain, in feet,
w= width of sample area, in feet, and
/=length of sample area, in feet.
To compute n for a flood plain by using the direct
method for vegetation density, first choose a representative
sample area along the cross section. The Vegd of the sample
area is determined by measuring the number and diameter
of trees in the 100-ft by 50-ft area. This is done easily by
plotting the location and diameter of the trees, as in the
sample area on the grid shown in figure 5. The numbers

next to the dots in figure 5 are the diameters of the trees in

tenths of a foot; those numbers underlined are the diameters
of the trees in feet.
The following table presents data from Poley Creek.
The total number of trees listed by diameter are summarized.
Site: Poley Creek, cross section 2, March 14,1979
Total number
of trees

Tree diameter
in feet

< ,

128
65
10
9
8
7
5
6
2
3
1
1
1

0.1
.2
.3
.4
.5
.6
.7
.8
.9
1.0
1.1
1.3
1.4

12.8
13.0
3.0
3.6
4.0
4.2
3.5
4.8
1.8
3.0
1.1
1.3
1.4

Vegd=^='-^Z= (2 ' 9)(57 - 5) =0.0115

AL hwl (2.9)(50)(100)
where
,:= summation of number of trees multiplied by tree
diameter, in feet;
h = height of water on flood plain, in feet;
w= width of sample area, in feet; and
/=length of sample area, in feet.
A value for flow depth is determined for the flood plain and
is assumed to equal the hydraulic radius, R, for the flood
plain. An effective-drag coefficient, C*, is selected from
figure 4. The boundary roughness, n0 , is determined for the
flood plain by using equation 9, and the n for the flood plain
is computed by using equation 7.
n0 =0.025, C*=11.0, R=2.9 ft

n=n0

n=0.025

l + (0.01 15)(11.

149\ 2 /1

n=0.136
Flood Plain n Values

11

o
Si.

Tl

2.

o
i"

s*
sL

v>
s>

o
i'

i
p
o

70

ere

3>

8-

25

20

15

10

.1

.1

.3

' ,

i *1

1.

.*

<

<

.1

.1

.1
1

10

.*

*1

1.

1.3

*4

.1

- ( 1

1< ' 1 .1

t ,1

1
1

.2
2

2J

2.

1.
2.

.<

^ 1

1
.3

40

1.0
z^-

.1

'

2*2 -

,7

" *.2

Tree diameter in
tenths of a foot

EXPLANATION

1.0

1
f

2
1 '2

80

2*

.1

,7

.5

2..4
4
2, ,

1,

2(

'V

.6

1 1

1
<2

*1

< .1

^7
i
' *1 ' 2

^2 4
6'<1

Tree diameter in
feet, underlined

70

: v

.1

,1
'7

r2

IP
^ 4

4 >6

< 2

4 '2.1

1.4

LO

*1

*1

.1

'

1
H ^H
s

&-

60

.22.J

2 1

of se ction

50

Line
"X

4 >6

2 <

1*

.2

i i 1 3
i>
1
< ,3
cI
1
2

FLOOD-PLAIN WIDTH, IN FEET

1,

*1

1 ,1

,2
8* 1 .1

,5

i ' 1 .1

1.' 2

2 < 1

2 < >1

iV
1
'
1 .1

30

2.
2

.2

*1

1*

i/1

,2

1 ,1

Location of tree

20

2.

1 .. 1
1*1

*3

.1

1< ' 1
1' ' 9

1< ' 2

90

1
*2

100

^7
i 1
1*
1

1, " 1 .1

*1
1
2
2 1

- 2

-4 (1J

11

DESCRIPTION: Flood plain consists of hardwood trees up to

40 ft tall, including many smaller diameter
trees and some vines and ground cover. The
surface is fairly smooth and has a firm soil base.

Figure 5. Example measurement of vegetation density showing tree diameter and location in representative-sample area.

UJ

0
Z

1-

LL

UJ

S3

10

15

20

*2
22
*.

25 1**1

DATE: March 14, 1979

SITE: Poley Creek, cross section 2

PHOTOGRAPHS OF FLOOD PLAINS

The following series of photographs (figs. 6-20)
represents densely vegetated flood plains for which roughness coefficients have been verified. The coefficients for
these sites were determined as a part of a study on
computation of backwater and discharge at width constrictions of heavily vegetated flood plains (Schneider and
others, 1977). By using these photographs for comparison
with other field situations, n values can then be used to
verify n values computed by other methods.
Information appearing with the photographs includes
n value determined for the area, date of flood, date
photograph was taken, and depth of flow on the flood plain.
A description of the flood plain includes values of vegetation density, effective drag coefficient, and base roughness.
Several reports present photographs of channels for
which roughness coefficients are known that would be
helpful in determining roughness values of other areas.

Barnes (1967) presented photographs of natural, stable

channels having known n values ranging from 0.023 to
0.075; a few flood plains were included in the report.
Ree and Crow (1977) conducted experiments to
determine friction factors for earthen channels planted with
certain crops and grasses. The values that were determined
may be used to help estimate the roughness of flood plains
planted with the type of vegetation used in their experiments. Photographs and brief descriptions of the vegetation
are given, and a tabulation of the hydraulic elements is
included.
Aldridge and Garrett (1973) presented photographs of
selected Arizona channels and flood plains having known
roughness coefficients. Included with the photographs are
descriptions of channel geometry and the roughness factors
involved in assigning an n value for the site.
Chow (1959) presented photographs of a number of
typical channels, accompanied by brief descriptions of the
channel conditions and the corresponding n values.

Computed roughness coefficient: Manning's n=0.10

Date of flood: February 21, 1974
Date of photograph: February 13, 1979
Depth of flow on flood plain: 2.6 ft
Description of flood plain: The vegetation of the flood plain is primarily trees,
including oak, gum, and pine. The base is firm soil and has slight surface
irregularities. Obstructions are negligible (a few downed trees and limbs).
Ground cover and vines are negligible. Vegd=0.0067, and C,,=12.0. The
selected values are nt =0.025, n^O.005, n3 =0.005, and n0 =0.035.
Figure 6. Cypress Creek near Downsville, La. (Arcement, Colson, and Ming, 1979a, HA-603, cross section 3).

Photographs of Flood Plains

13

Computed roughness coefficient: Manning's n=0.ll

Date of flood: March 18, 1973
Date of photograph: February 14, 1979
Depth of flow on flood plain: 3.6 ft
Description of flood plain: The vegetation of the flood plain is primarily large,
tall trees, including oak, gum, ironwood, and pine. The base is firm soil and
is smooth. Obstructions are few, and ground cover and undergrowth are
sparse. Vegd=0.0067, and C_=8.8. The selected values are nfc=0.020,
n, =0.002, n3 =0.003, and 0 =0.025.

Figure 7. Bayou de Loutre near Farmerville, La. (Schneider and others, 1977, cross section 2).

14

Guide for Selecting Manning's Roughness Coefficients for Natural Channels and Flood Plains

Computed roughness coefficient: Manning's n=0.11

Date of flood: March 18, 1973
Date of photograph: February 14, 1979
Depth of flow on flood plain: 3.7 ft
Description of flood plain: The vegetation of the flood plain is primarily large,
tall trees, including oak, gum, and ironwood. The base is firm soil and has
slight surface irregularities and obstructions caused by downed trees and
limbs. Ground cover and undergrowth are negligible. Vegd=0.0075, and
C.=7.7. The selected values are nfc =0.020, ^=0.002, n3 =0.003, and
710=0.025.

Figure 8. Bayou de Loutre near Farmerville, La. (Schneider and others, 1977, cross section 3).

Photographs of Flood Plains

15

Computed roughness coefficient: Manning's n=0.11

Date of flood: March 18, 1973
Date of photograph: February 14, 1979
Depth of flow on flood plain: 3.7 ft
Description of flood plain: The vegetation of the flood plain is primarily trees,
including oak, gum, ironwood, and pine. The base is firm soil and has slight
surface irregularities and obstructions caused by downed trees and limbs.
Ground cover and undergrowth are negligible. Vegd=0.0072, and C_=8.0.
The selected values are nfc=0.020, n^O.002, n3 =0.003, and n0 =0.025.

Figure 9. Bayou de Loutre near Farmerville, La. (Schneider and others, 1977, cross section 3).

16

Guide for Selecting Manning's Roughness Coefficients for Natural Channels and Flood Plains

Computed roughness coefficient: Manning's n=0.11

Date of flood: February 22, 1971
Date of photograph: April 5, 1979
Depth of flow on flood plain: 3.0 ft
Description of flood plain: The vegetation of the flood plain is primarily trees,
including oak, gum, and ironwood. The base is silty soil and has slight surface
irregularities. Obstructions are few, and some flood debris is present. Ground
cover is short weeds and grass, and undergrowth is minimal. Vegd=0.0077,
and C=10.2. The selected values are nfc =0.020, n, =0.002, n4'=0.005, and
7io=0.027.

Figure 10. Coldwater River near Red Banks, Miss. (Colson, Arcement, and Ming, 1979, HA-593, cross section 2).

Photographs of Flood Plains

17

wn^

Computed roughness coefficient: Manning's n=0.11

Date of flood: February 22, 1971
Date of photograph: April 5, 1979
Depth of flow on flood plain: 3.0 ft
Description of flood plain: The vegetation of the flood plain is primarily trees,
including oak, gum, and ironwood. The base is silty soil and has slight surface
irregularities. Few obstructions and some flood debris are present. Ground
cover is short weeds and grass, and undergrowth is minimal. Vegd=0.0090,
and C,=8.6. The selected values are nfc =0.020, n, =0.003, n4'=0.005, and
no=0.028.

Figure 11. Coldwater River near Red Banks, Miss. (Colson, Arcement, and Ming, 1979, HA-593, cross section 2).

18

Guide for Selecting Manning's Roughness Coefficients for Natural Channels and Flood Plains

Computed roughness coefficient: Manning's n=0.12

Date of flood: April 12, 1969
Date of photograph: March 28, 1979
Depth of flow on flood plain: 4.0 ft
Description of flood plain: The vegetation of the flood plain is primarily trees,
including oak, gum, ironwood, and many small diameter trees (0.1 to 0.2 ft).
The base is firm soil and has slight surface irregularities. Obstructions are
negligible. Ground cover and undergrowth are negligible. Vegd=O.OOS2, and
Cf =7.6. The selected values are nfc =0.025 and rto=0.025.

Figure 12. Yockanookany River near Thomastown, Miss. (Colson, Ming, and Arcement, 1979a, HA-599, cross section 5).

Photographs of Flood Plains

19

Computed roughness coefficient: Manning's =0.12

Date of flood: April 12, 1969
Date of photograph: March 28, 1979
Depth of flow on flood plain: 4.0 ft
Description of flood plain: The vegetation of the flood plain is primarily trees,
including oak, gum, ironwood, and many small diameter trees (0.1 to 0.2 ft).
The base is firm soil and has slight surface irregularities. Obstructions are
negligible (a few downed trees and limbs). Ground cover and undergrowth are
negligible. Vegd=O.QOS2, and 0^=7.6. The selected values are 6=0.025
and n0 =0.025.

Figure 13. Yockanookany River near Thomastown, Miss., 500 ft east of area shown in figure 12 (Colson, Ming, and
Arcement, 1979a, HA-599, cross section 5).

20

Guide for Selecting Manning's Roughness Coefficients for Natural Channels and Flood Plains

Computed roughness coefficient: Manning's n=0.13

Date of flood: December?, 1971
Date of photograph: April 10, 1979
Depth of flow on flood plain: 3.2 ft
Description of flood plain: The vegetation of the flood plain is a mixture of
large and small trees, including oak, gum, and ironwood. The base is firm soil
and has minor surface irregularities and some rises. Obstructions are negligible (some exposed roots and small trees). Ground cover and undergrowth are
negligible. Ve&,=0.0087, and C=11.5. The selected values are nfc=0.025,
n, =0.003, n3 =0.002, and w0 =0.030.

Figure 14. Flagon Bayou near Libuse, La. (Arcement, Colson, and Ming, 1979b, HA-604, cross section 4).

Photographs of Flood Plains

21

Computed roughness coefficient: Manning's n=0.14

Date of flood: December 21, 1972
Date of photograph: March 13, 1979
Depth of flow on flood plain: 2.9 ft
Description of flood plain: The vegetation of the flood plain is a mixture of
large and small trees, including oak, gum, and ironwood. The base is firm soil
and has minor surface irregularities caused by rises and depressions. Obstructions are minor (downed trees and limbs and a buildup of debris). Ground
cover is negligible, and the small amount of undergrowth is made up of small
trees and vines. Vegd=0.0085, and C=15.6. The selected values are
nfc =0.025, n, =0.005, n3 =0.015, n4'=0.005, and 0 =0.050.

Figure 15. Pea Creek near Louisville, Ala. (Ming, Colson, and Arcement, 1979, HA-608, cross section 5).

22

Guide for Selecting Manning's Roughness Coefficients for Natural Channels and Flood Plains

Computed roughness coefficient: Manning's n=0.14

Date of flood: December 21, 1972
Date of photograph: March 13, 1979
Depth of flow on flood plain: 2.8 ft
Description of flood plain: The vegetation of the flood plain is large and small
trees, including oak, gum, and ironwood. The base is firm soil and has minor
surface irregularities caused by rises and depressions. Obstructions are minor
(downed trees and limbs and a buildup of debris). Ground cover is negligible,
and the small amount of undergrowth is made up of small trees and vines.
Vegd=0.0l02, and C_=15.6. The selected values are nfc =0.025, n^O.005,
n3 =0.015, n4 '=0.005, and n0 =0.050.

Figure 16. Pea Creek near Louisville, Ala. (Ming, Colson, and Arcement, 1979, HA-608, cross section 4).

Photographs of Flood Plains

23

Computed roughness coefficient: Manning's n=0.15

Date of flood: December 7, 1971
Date of photograph: April 12, 1979
Depth of flow on flood plain: 4.1 ft
Description of flood plain: The vegetation of the food plain is large and small
trees, including oak, gum, and ironwood. The base is firm soil and has minor
surface irregularities caused by rises and depressions. Obstructions are
negligible (some exposed roots). Ground cover is negligible, and undergrowth
is minimal. Vegd=0.0067, and C^=14.4. The selected values are nfc =0.025,
n, =0.003, 3 =0.002, and n0 =0.030.

Figure 17. Tenmile Creek near Elizabeth, La. (Arcement, Colson, and Ming, 1979c, HA-606, cross section 3).

24

Guide for Selecting Manning's Roughness Coefficients for Natural Channels and Flood Plains

Computed roughness coefficient: Manning's n=0.18

Date of flood: March 23, 1973
Date of photograph: April 11, 1979
Depth of flow on flood plain: 5.0 ft
Description of flood plain: The vegetation of the flood plain is large trees,
including oak, gum, ironwood, and pine. The base is firm soil and has
moderate surface irregularities caused by rises and depressions. Obstructions
are negligible (a few vines). Ground cover and undergrowth are negligible.
Vegd=0.0084, and C^IS.3. The selected values are nfc =0.025, n,=0.008,
n3 =0.002, and n0 =0.035.

Figure 18. Sixmile Creek near Sugartown, La. (Schneider and others, 1977, cross section 7).

Photographs of Flood Plains

25

Computed roughness coefficient: Manning's n=0.20

Date of flood: March 3, 1971
Date of photograph: March 29, 1979
Depth of flow on flood plain: 2.9 ft
Description of flood plain: The vegetation of the flood plain is a mixture of
large and small trees, including oak, gum, and ironwood. The base is firm soil
and has minor surface irregularities. Obstructions are minor. Ground cover is
medium, and the large amount of undergrowth includes vines and palmettos.
Vegd=0.0ll5, and C,,=22.7. The selected values are nfc =0.025, ni=0.005,
n3 =0.010, n4 '=0.015, and n0 =0.055.
Figure 19. Thompson Creek near Clara, Miss. (Colson, Ming, and Arcement, 1979b, HA-597, cross section 9).

26

Guide for Selecting Manning's Roughness Coefficients for Natural Channels and Flood Plains

Computed roughness coefficient: Manning's n=0.20

Date of flood: March 3, 1971
Date of photograph: March 29, 1979
Depth of flow on flood plain: 2.9 ft
Description of flood plain: The vegetation of the flood plain is large and small
trees, including oak, gum, and ironwood. The base is firm soil and has minor
surface irregularities. Obstructions are minor (some downed trees and limbs).
Ground cover is medium, and the large amount of undergrowth includes vines
and palmettos. Vegd=0.0ll5, and C,=22.7. The selected values are
n,=0.025, B2=0.005, 3=0.010, n4'=0.015, and ^=0.055.

Figure 20. Thompson Creek near Clara, Miss., 500 ft east of area shown in figure 19 (Colson, Ming, and Arcement, 1979b,
HA-597, cross section 9).

Photographs of Flood Plains

27

PROCEDURES FOR ASSIGNING n VALUES

When determining n values for a cross section, pans

of the procedure apply only to roughness of channels, and
other pans apply to roughness of flood plains.
The procedure involves a series of decisions that are
based on the interaction of roughness factors. A flow chart
(fig. 21) illustrates the steps in the procedure (see Steps for
Assigning n values). A form (fig. 22) is provided to help in
the computation of the n values. After using the procedure
a few times, the user may wish to combine steps or to
change the order of the steps. Experienced personnel may
perform the entire operation mentally, but the inexperienced
user may find the form in figure 22 useful. Steps 3 through
13 apply to channel roughness, and steps 14 through 23
apply to flood-plain roughness. The procedure is adapted
from the report by Aldridge and Garrett (1973) but is
extended to include assigning n values for flood plains.
Steps for Assigning n Values
Reach Subdivision (Steps 1 and 2)

1. Determine the extent of stream reach to which the

roughness factor will apply. Although n may be applied to
an individual cross section that is typical of a reach, the
roughness in the reach that encompasses the section must be
taken into account. When two or more cross sections are
being considered, the reach that applies to any one section
is considered to extend halfway to the next section. For
example, in figure 1, the n value for cross section 1
represents the roughness in reach A, and the n value for
cross section 2 represents the roughness in reach B. If the
roughness is not uniform throughout the reach being considered, n should be assigned for average conditions.
2. If the roughness is not uniform across the width of
the cross section, determine where subdivision of the cross
section should occur. Determine whether subdivision
between channel and flood plain is necessary and whether
subdivision of the channel or flood plain is also necessary.
If the roughness is not uniform across the width of the
channel, determine whether a base n should be assigned to
the entire channel cross section or whether a composite n
should be derived by weighting values for individual
segments of the channel having different amounts of roughness (see steps 4-10). When the base value is assigned to
the entire channel, the channel constitutes the one segment
being considered, and steps 5, 8, 9, and 10 do not apply.
Channel Roughness (Steps 3-13)

3. Determine the channel type stable channel, sand

channel, or a combination and whether the conditions are
28

representative of those that may exist during the design

event being considered. Look especially for evidence of bed
movement and excessive amounts of bank scour. If the
conditions do not appear to be the same as those that will
exist during the flow event, attempt to visualize the conditions that will occur. To estimate the possible range in n
values, compare the channel with other channels for which
n values have been verified or assigned by experienced
personnel (see photographs in Barnes, 1967).
4. Determine the factors that cause roughness and
how each is to be taken into account. Some factors may be
predominant in a particular segment of the channel, or they
may affect the entire cross section equally. The manner in
which each factor is handled depends on how it combines
with other factors. A gently sloping bank may constitute a
separate segment of the cross section, whereas a vertical
bank may add roughness either to the adjacent segment or to
the entire channel. Obstructions, such as debris, may be
concentrated in one segment of the channel. Isolated boulders should be considered as obstructions, but if boulders
are scattered over the entire reach, consider them in
determining the median particle size of the bed material.
Vegetation growing in a distinct segment of the channel
may be assigned an n value of its own, whereas roughness
caused by vegetation growing only along steep banks or
scattered on the channel bottom will be accounted for by
means of an adjustment factor that can be applied to either
a segment of the channel or to the entire cross section. If a
composite n is being derived from segments, the user
should continue with step 5; otherwise step 5 should be
omitted.
5. Divide the channel width into segments according
to roughness. If distinct, parallel banks of material of
different particle sizes or of different roughness are present,
defining the contact between the types of material is fairly
easy (see fig. 1, cross section 2). The dividing line between
any two segments should parallel the flow lines in the
stream and should be located so as to represent the average
contact between types of material. The dividing line must
extend through the entire reach, as defined in step 1,
although one of the types of bed material may not be present
throughout the reach. If a segment contains more than one
type of roughness, use an average size of bed material.
Where sand is mixed with gravel, cobbles, and boulders
throughout a channel, dividing the main channel is impractical.
6. Determine the type of material that occupies and
bounds each segment of channel and compute the median
particle size in each segment by using either method A or B
(below). If the Limerinos equation (eq. 5) is used, the size
corresponding to the 84th percentile should be used in the
computation.
A. If the particles can be separated by screening
according to size, small samples of the bed material are

Guide for Selecting Manning's Roughness Coefficients for Natural Channels and Flood Plains

collected at 8 to 12 sites in the segment of the reach. The

samples are combined, and the composite sample is
passed through screens that divide it into at least five size
ranges. Either the volume or weight of material in each
range is measured and converted to a percentage of the
total.
B. If the material is too large to be screened, a grid
system having 50 to 100 intersecting points or nodes per
segment is laid out. The width, or intermediate diameter,
of each particle that falls directly under a node is
measured and recorded. The sizes are grouped into at
least five ranges. The number of particles in each range is
recorded and converted to a percentage of the total
sample.
In the above sampling methods, the size that corresponds to the 50th percentile (table 1) or the 84th percentile
(the Limerinos method) is obtained from a distribution
curve derived by plotting particle size versus the percentage
of sample smaller than the indicated size. Experienced
personnel can make a fairly accurate estimate of the median
particle size by inspection of the channel if the range in
particle size is small.
7. Determine the base n for each segment of channel
by using table 1 or equation 5 or the comparison given in
step 3. Chow's (1959) base values (table 1) are for the
smoothest condition possible for a given material. The
values (table 1) of Benson and Dalrymple (1967) are for a
straight, uniform channel of the indicated material and are
closer to actual field values than are those of Chow. If a
composite n is being derived from segments, proceed with
step 8. If n is being assigned for the channel as a whole,
proceed to step 11.
8. Add the adjustment factors from table 2 that apply
only to individual segments of the channel.
9. Select the basis for weighting n for the channel
segments. Wetted perimeter should be used for trapezoidal
and V-shaped channels having banks of one material and
beds of another material. Wetted perimeter should be used
also where the depth across the channel is fairly uniform.
Area should be used where the depth varies considerably or
where dense brush occupies a large and distinct segment of
the channel.
10. Estimate the wetted perimeter or area for each
segment and assign a weighting factor to each segment that
is proportional to the total wetted perimeter or area. Weight
n by multiplying the n for each segment by the assigned
weighting factor.
11. Select the adjustment factors from table 2 for
conditions that influence n for the entire channel. Do not
include adjustment factors for any items used in steps 7 and
8. Consider upstream conditions that may cause a disturbance in the reach being studied. If Chow's (1959) base
values are used, the adjustment factors in table 2 may be
used directly. If base values are computed from the Limerinos equation (eq. 5) or are taken from Benson and Dalrym-

ple (1967), the adjustment factors should be from one-half

to three-fourths as large as those given in table 2. If n is
assigned on the basis of a comparison with other streams,
the adjustment factors will depend on the relative amounts
of roughness in the two streams. Add the adjustment factors
to the weighted n values from step 10 to derive the overall
n for the channel reach being considered. When a multiplying factor for meander is used, first add the other adjustments to the base n. Round off the n value as desired. The
value obtained is the composite or overall n for the channel
reach selected in step 1. When more than one reach is used,
repeat steps 1-13 for each reach.
12. Compare the study reach with photographs of
other channels found in Barnes (1967) and Chow (1959) to
determine if the final values of n obtained in step 11 appear
reasonable.
13. Check the flow regime for all sand channels. Use
the n from step 11 in the Manning's equation (eq. 1) to
compute the velocity, which is then used to compute stream
power. The flow regime is determined from figure 2. The
assigned value of n is not reliable unless the stream power
is sufficient to cause upper regime flow.
Flood-Plain Roughness (Steps 14-23)

14. As in step 1, the n value selected must be

representative of the average conditions of the reach being
considered. Determine if the flood-plain conditions are
representative of those that may exist during the design
event being considered. Compare the flood plain with other
flood plains for which n values have been determined (or
have been assigned by experienced personnel) to estimate
the possible range in n values. Compare with photographs
in this guide and in other references.
15. The n value for the flood plain can be determined
by using the measurement of vegetation density or resistivity. There may be cases where the roughness is determined
by a qualitative evaluation of the roughness by using
equation 6 and the adjustment factors in table 3. A decision
must be made as to which method will be used.
16. If there are abrupt changes in roughness on the
flood plain, subdivide the flood-plain cross sections. A
representative sampling area is selected for each subarea of
the flood plain.
17. Determine the factors that cause roughness and
how each is to be taken into account. Such factors as surface
irregularities and obstructions can be accounted for in the
boundary roughness, whereas vegetation can be accounted
for in the boundary roughness or by using the quantitative
method.
18. A base value, nb , for the flood plain's bare soil
surface must be chosen. A value for nb is chosen from table
1.
Procedures for Assigning n Values

29

1. Determine extent of reach to which roughness factor will apply.

2. Determine if and where subdivision between channel and flood plain is necessary.
Determine how base n will be assigned.

FLOOD-PLAIN
ROUGHNESS

CHANNEL ROUGHNESS

3. Determine channel type, and estimate conditions at time of flow event.

Compare the channel with photographs and descriptions of other channels.
For the entire channel

4. Determine the factors that cause roughness

and how each will be accounted for.

6. Determine type and size of bed material.

7. Assign a base n from tables, formulas, and
comparison with other channels and
verification photographs.

Assign a composite n for the entire channel, derived

from individual segments of the channel

4. Determine the factors that cause roughness

and how each will be accounted for.
5. Mentally divide channel into segments so
that the roughness factor within a segment
is fairly uniform.
6. Determine type and size of boundary
material in each segment.
7. Assign a base n for each segment from
tables, formulas, or comparison with other
channels and verification photographs.
8. Apply adjustment factors for individual
segments, if applicable.
9. Select the method for weighting n.

By wetted perimeter
10a. Estimate wetted perimeter for each
segment of channel,
b. Weight the n values by assigning
weighting factors that are proportional
to the wetted perimeter.

10a. Estimate area for each segment

of channel.
b. Weight the n values by assigning
weighting factors that are proportional
to the area.

11 Adjust for factors not considered in steps 7 and 8, including channel

alignment, change in channel shape, vegetation, obstructions, and
meander. Round off as desired for use in the Manning's equation.
12 Compare value determined with that for other channels and verification
photographs to test for reasonableness.
13. For sand channels: Check flow regime by computing velocity and stream
power for the above n ; determine regime from figure 2. The n from
table 1 is valid only for the upper regime flow.
Figure 21. Flow chart of procedures for assigning n values (modified from Aldridge and Garrett, 1973, fig. 3).

30

Guide for Selecting Manning's Roughness Coefficients for Natural Channels and Flood Plains

FLOOD-PLAIN ROUGHNESS

14. Determine type of flood plain, and estimate conditions at time of flow event; compare the
flood plain with photographs and descriptions of other flood plains.
15. Determine method to be used in assigning n to flood plain, whether
vegetation-density method will be used with boundary-roughness factors,
or boundary-roughness factors only.

16. Determine if roughness is uniform throughout flood plain, or whether flood

plain needs to be subdivided. (Following steps apply to each subdivision.)

17. Determine the factors that cause roughness and how each is to be accounted for.

18. Assign a base n/jfrom tables and comparison with

other flood plains and verification photographs.

19. Determine the adjustment factors from tables.

Boundary-roughness method

Boundary-roughness method with

vegetation-density method
20. Determine HQ value from tables and formulas.
21. Determine vegetation density of representative
sample area of flood plain.

22. Determine n for flood plain by using formulas.

23. Compare value determined with that for other flood plains and
verification photographs to test for reasonableness.
Figure 21. Flow chart of procedures for assigning n values (modified from Aldridge and Garrett, 1973,
fig. 3) Continued.

Procedures for Assigning n Values

31

19. Select the adjustment factors from table 3 for

conditions that influence roughness of the flood-plain subsection.
20. Determine the n0 value by equation 9, by using
the adjustment factors selected in step 19. The n4 ' value is
the adjustment factor for vegetation not accounted for by the
vegetation-density method.
21. The vegetation density of the sampling area is
determined by using equation 11 and measuring the crosssectional area occupied by the trees and undergrowth in the
sampling area. An estimate of the depth of flow on the flood
plain is necessary to determine the vegetation density and
the n value. By measuring two or three sampling areas in a
subsection, a more representative value for vegetation
density can be determined.
22. The n value for the flood-plain subsection is
determined by using equation 6 or 7, depending on which
method has been chosen. If the quantitative method is being
used, the n value for each subarea of the flood plain is
computed by using equation 7 and vegetation-density and
boundary-roughness values for each subarea.
23. Compare the study reach with photographs of
other flood plains in this report and in other references to
determine if the final values of n obtained in step 22 appear
to be reasonable.

Examples of Procedures for Determining n

Values
A sketch of a hypothetical channel and flood plain is
shown in figure 1, and procedures for determining n values
are outlined in table 4. The channel and flood plain together
are divided into three separate reaches (A, B, C), and each
reach has a cross section (1, 2, 3). The shape of each cross
section is shown in figure 1.
In cross section 1, the flow is confined to the channel.
The channel is composed of firm soil, and no subdivision of
the channel is necessary. Steps 1 through 13, in Steps for
Assigning n Values, are used in the computation of n for
cross section 1. These steps apply only to channel conditions.
Flow in cross section 2 is also confined to the
channel, which is composed of three distinct parallel bands
of (1) bedrock, (2) sand, and (3) gravel and cobbles. The n
value for each segment is determined and a composite n for
the channel is computed by weighting each segment n value
by the wetted perimeter. Again, steps 1 through 13 are used
in the computation of n for cross section 2.
The flow in cross section 3 is channel and flood-plain
flow. The cross section is divided into three subsections.
Subsection 1 is flood-plain flow through woods, subsection

32

2 is channel flow, and subsection 3 is flood-plain flow

through a cotton field.
In subsection 1, the flood plain is made up of dense
woods having little undergrowth. The procedure using the
vegetation density of the woods is used to determine the n
value for the flood plain. The vegetation density is determined from a representative-sample area of the wooded
flood plain. A boundary roughness, n0 , is determined from
equation 9 and the n value is determined by using equation
7. Steps 14 through 23 in Steps for Assigning n Values are
used in the computation of n for subsection 1.
Subsection 2 of cross section 3 represents channel
flow. The channel is composed of firm soil, and no
subdivision of the channel is necessary. Steps 1 through 13
are used in the computation of n for subsection 2.
Subsection 3 represents the flow of a flood plain
planted in cotton. There is no need to subdivide the
subsection. The depth of flow is equal to the height of the
vegetation. Steps 14 through 23 are used in the computation
of the n value for subsection 3 by using equation 6.

SUMMARY

This guide presents procedures for assigning reliable

n values for channels and flood plains. The roughness
coefficient, n, applies to a reach of a channel and (or) flood
plain and should be representative of that entire reach. A
channel and flood plain may need to be divided into
subsections and n values assigned to each subsection if one
cross section is not representative of the entire reach.
Channel roughness is determined by following a
series of decisions based on the interaction of roughness
factors. A base value is assigned to the channel, and
adjustments are made for certain roughness factors.
A similar procedure is used to assign n values to flood
plains. A base value related to certain roughness factors is
determined for the flood plain; then an option, based on the
measurement of vegetation density of the flood plain, is
used to determine the total roughness of flood-plain subsections. The vegetation density of the flood plain is determined from physical measurements of the vegetation in a
representative sample area of a flood-plain subsection.
Photographs of flood plains for which n values have
been established are presented to aid in the determination of
roughness coefficients. The photographs can be used for
comparison with field situations to help verify selected n
values.
Examples and step-by-step procedures for determining roughness coefficients for channels and flood plains are
presented in this guide. These procedures can be used in the
field to help assign reasonable n values for many types of
channels and flood plains.

Guide for Selecting Manning's Roughness Coefficients for Natural Channels and Flood Plains

Stream and location: ___________________________________________

Reach: ____________________________________________________
Event for which n is assigned: ________________________________________
1. Is roughness uniform throughout the reach being considered? If not, n should be assigned for
the average condition of the reach.
2. Is roughness uniformly distributed along the cross section? Is a division between channel and
flood plain necessary? (Channel roughness uses steps 3-13, flood-plain roughness uses steps 14-23.)
Is roughness uniformly distributed across the channel? If not, on what basis should n for the individual
segments be weighted?
3. Describe the channel. Are present conditions representative of those during the flood? If not,
describe the probable conditions during the flood.
4. How will the roughness-producing effects of the following on the channel be accounted for?
Bank roughness: _____________________________________________
Bedrock outcrops: _____________________________________________
Isolated boulders: _____________________________________________
Vegetation: ________________________________________________
Obstructions:
Meander:
5-10. Computation of weighted n for the channel
Segment Approximate
Wetted
number dimensions,
Area,
perimeter,
(ft)
and
(ft2)
(ft)
material
Width Depth

Median
grain size,
(mm)

Base n
Adjustfor
segment ments

Adjusted n
Adjusted Weight times weight
n
factor
factor

Sum
Weighted n =
Figure 22. Sample form for computing n values (modified from Aldridge and Garrett, 1973, fig. 4).

Summary

33

11-13. Computation of n for the channel

Adjustment factors for the channel
Factor

Adjustment

Describe conditions briefly

Irregularity, n1
Alignment, n2
Obstructions, n3
Vegetation, n4
Meander, m
Weighted n plus adjustments
Computed n =

14. Describe the flood olain.

Are present conditions representative of those during the flood?

If not, describe probable conditions during the flood.
15. Is the roughness coefficient to be determined by roughness factors only or is it to include
vegetation-density method?
16. Is roughness uniformly distributed across the flood plain?
If not, how should the flood plain be subdivided?
17-23. Computation of n for flood plain
Adjustment factors not using vegetation-density method
Subsection

Irregularity,
"i

Base n ,
nb

Obstructions,

Vegetation,

"3

A74

Computed n

Adjustment factors using vegetation-density method

Sub- Base n , Irregu- Obstruclarity,
tions,
section
nb
"i
"3

Vegetation,
n4

Boundary
roughness,
n0=
'V71 +n3+A74

Vegetation
density,
Vegd

Effective
drag,

c*

Hydraulic
radius,
R

Computed
n

Figure 22. Sample form for computing n values (modified from Aldridge and Garrett, 1973, fig. 4) Continued.

34

Guide for Selecting Manning's Roughness Coefficients for Natural Channels and Flood Plains

Table 4. Outline and example of procedures for determining n values for a hypothetical channel and adjoining flood plain
[Modified from Aldridge and Garrett, 1973, table 6]
Step

Item to be determined or operation

to be performed

Factors on which decisions are based and the results

CROSS SECTION 1

1
2

Extent of reach ....................... The reach extends one section width upstream of cross section 1 to midway between
cross sections 1 and 2. Designated as reach A (fig. 1).
Subdivision of cross section 1.......... Only channel flow, no overbank flood-plain flow. Assign a base nb to entire channel.
Channel roughness (steps 3-13)

4
5
6
7
8
9
10
11

12
13

(a) Type of channel................... A stable channel made up of firm soil.

(b) Conditions during flow event....... Assume channel conditions are representative of those that existed during the peak
flow.
(c) Comparable streams ............... None.
Roughness factors..................... Add adjustments for grass and trees in channel and for channel alignment.
Divide into segments.................. Not necessary.
Type of channel ...................... Firm soil.
Base nb .............................. Table 1 gives an nb value for firm soil of 0.020-0.032. Use 0.025.
Adjustment factors for segments ....... None.
Basis for weighting n ................. Not applicable.
Weighting factors and weighted n ...... Not applicable.
Add adjustments for entire channel..... Vegetation (4) weeds and supple seedlings along bottom of channel (table 2).
4 =0.005. Meander is minor, m=1.00.
n (nb +n l +n2 +n3 + n4)m.
n=(0.025+0+0+0+0.005)1.00.
n=0.030.
Compare with other streams ........... None.
Check flow regime.................... Not applicable.
CROSS SECTION 2

1
2

Extent of reach ....................... From midway between cross sections 1 and 2 to midway between cross sections 2 and
3. Designated as reach B (fig. 1).
Subdivision of cross section 2.......... Flow remains in channel, no overbank flood-plain flow. The channel is composed of
distinct bands, each having a different roughness. Derive n by weighting segments.
Channel roughness (steps 3-13)

5
6

8
9

(a) Type of channel................... Combinations of sand and stable channel. Consider that channel reacts as a stable
channel.
(b) Conditions during flow event....... Some movement of sand may have occurred during the peak flow, but assume that
channel conditions are representative of those that existed during the peak.
(c) Comparable streams............... None.
Roughness factors..................... (1) Bedrock may be accounted for by adding an adjustment factor to the n value for
the bed or as a separate segment. Use latter.
(2) Divide into segments according to the type of material.
(3) Boulder at head of reach add as an adjustment factor to composite n.
Divide into segments.................. The channel has three basic types of roughness caused by parallel bands of bedrock,
sand, and gravel and cobbles. Each band is a segment.
Type of material and grain size ........ (1) Bedrock slightly irregular, containing fairly sharp projections having a maximum height of about 3 in.
(2) Sand determined by sieve analysis, median particle size is 0.8 mm.
(3) Gravel and cobbles as determined by examination, the material is from 2 to
10 in. in diameter. As determined from 100-point grid system, the median particle
size is 6 in.
Base nb .............................. (1) Bedrock table 1 shows that nb for jagged and irregular rock cut is from 0.035 to
0.050. Assume that the projections have an average cut; nb for this segment is
0.040.
(2) Sand table 1 gives an nb value of 0.025.
(3) Gravel and cobbles table 1 shows that the base nb for cobbles ranges from 0.030
to 0.050. The median diameter is small for the size range. Use a base nb value of
0.030.
Adjustment factors for segments ....... None.
Basis for weighting n ................. Use wetted perimeter for basis of weighting n for the channel segments.

Summary

35

Table 4. Outline and example of procedures for determining n values for a hypothetical channel and adjoining flood
plain Continued
[Modified from Aldridge and Garrett, 1973, table 6]
_
Step
r

Item to be determined or operation

,
,
,
to be performed

...
. . . , . .
,
,
, .
.
Factors on which decisions are based and the results
CROSS SECTION 2-Continued

10
11

12
13

Weighting factors and weighted n ...... About 15 ft of the wetted perimeter is bounded by bedrock, about 25 ft by sand, and
about 25 ft by gravel and cobbles. The unadjusted n value is (0.2x0.040
+0.4X0.025+0.4X0.030)71.0=0.030.
Add adjustments for entire channel..... (1) Boulders at head of the reach are slight obstructions, add 0.002 (table 2).
(2) The bend near the lower end of reach A (fig. 1) causes slight irregularity; add
0.002 (table 2).
n=(nb +n l +n2 +n3 +n4)m.
n=(0.030+0.002+0+0.002+0)1.0.
n=0.034.
Compare with other streams ........... None.
Check flow regime.................... Sufficient sand was not present to warrant a check.
CROSS SECTION 3

1
2

Extent of reach ....................... From midway between cross sections 2 and 3 to one section width down stream of
cross section 3. Designated as reach C (fig. 1).
Subdivision of cross section 3.......... There is overbank flood-plain flow on both sides of the channel. Subsection 1 is
flood-plain flow through trees, subsection 2 is channel flow, and subsection 3 is
flood-plain flow through a cotton field. Assign a base nb to each subsection.
Channel roughness (steps 3-13) subsection 2

4
5
6
7
8
9
10
11

12
13

(a) Type of channel................... A stable channel made up of firm soil.

(b) Conditions during flow event....... Assume channel conditions are representative of those that existed during the peak
flow.
(c) Comparable streams ............... See photographs of similar channels in Barnes (1967, p. 16-17). Channel made up of
same type of material. Barnes used n of 0.026 for the channel.
Roughness factors..................... Trees along the bank should be considered as obstruction (3) for the channel.
Divide into segments.................. Not necessary.
Type of material and grain size ........ Firm soil (clay).
Base nb .............................. Table 1 gives a base nb value for firm soil of 0.020 to 0.030. Use 0.025.
Adjustment factors for segments ....... None.
Base for weighting n .................. Not applicable.
Weighting factors and weighted n ...... Not applicable.
Add adjustments for entire channel..... Obstructions (3) negligible scattered trees and tree roots along edge of channel
banks (table 2). 3 =0.003. Meander is minor, m=1.00.
n=(nb +n l +n2 +n3 +n4)m.
n=(0.025+0+0+0.003+0)1.00.
n=0.028.
Compare with other streams ........... Similar to channel in photographs in Barnes (1967, p. 16-17). The n value reported
was 0.026.
Check flow regime.................... Not applicable.
Flood-plain roughness (steps 14-23) subsection 1 (made up of trees)

14

15
16
17
18
19

36

(a) Type of flood plain................ A slightly irregular flood plain covered with hardwood trees. No undergrowth.
(b) Conditions during flow event....... Assume present conditions are representative of those that existed during the peak
flow.
(c) Comparable flood plains ........... Rood plain is similar to one shown in figure 14 of this report.
Method to be used in assigning n ...... Use the vegetation-density method. Need to determine a value for boundary roughness.
Subdivision of flood plain ............. The flood plain is uniform throughout.
Roughness factors..................... Trees are the major roughness factor; surface irregularity and some obstructions are on
flood plain.
Base nb .............................. Table 1 gives a base nb value for firm soil of 0.020-0.030. Use 0.020.
Adjustment factors .................... Irregularity is minor; a few rises and dips across the flood plain. n l =0.005 (table 3).
Obstructions are negligible, consisting of scattered debris, exposed roots, and
downed trees. 3 =0.004 (table 3).

Guide for Selecting Manning's Roughness Coefficients for Natural Channels and Flood Plains

Table 4. Outline and example of procedures for determining n values for a hypothetical channel and adjoining flood
plain Contin ued
[Modified from Aldridge and Garrett, 1973, table 6]
Step

Item to be determined or operation

to be performed

Factors on which decisions are based and the results

CROSS SECTION 3, subsection 1 Continued

20
21
22

Vegetation density of representative

sample area.
n for flood-plain subsection 1 .....

nQ =(nb +n l +n2 +n3 +n4 ')m.

n0 =(0.020+0.005+0+0.004+0)1.0.
n0 =0.029.
Vegd=0.0ll5 is an average value from three sampling areas.
R=2.9 ft.
Vegd=Q.Qll5.

n=0.029
23

n=0.137.
Compare with other flood plains ....... Photographs of similar flood plains found in this report (fig. 14).
Flood-plain roughness (steps 14-23) subsection 3 (cotton field)

14
15
16
17
18
19
20
21
22
23

(a) Type of flood plain................

(b) Conditions during flow event.......
(c) Comparable flood plains ...........
Method to be used in assigning n ......
Subdivision of flood plain .............
Roughness factors.....................
Base nb ..............................
Adjustment factors....................

Flood plain is a cotton field in full growth.

Conditions are similar to flood event.
None.
Assign n by evaluation of boundary roughness only.
No division of flood plain is necessary.
Roughness factors to be considered are surface irregularity and vegetation.
Table 1 gives a base nb value for firm earth of 0.020-0.030. Use 0.025.
Irregularity is moderate with furrows parallel to flow on flood plain, n^O.010 (table
3). Vegetation is cotton crop; depth of flow is about equal to height of vegetation,
n4 =0.040 (table 3).
0 ................................... Not applicable.
Vegetation density of representative
Not applicable,
sample area.
n for flood plain ..................... n=(nb +n l +n2 +n3 +n4 +)tn.
n=(0.025+0.01+0+0+0.040)1.00.
n=0.075.
Compare with other flood plains ....... Ree and Crow (1977, p. 39^40) assigned cotton fields an n value of about 0.08.

REFERENCES CITED

Aldridge, B.N., and Garrett, J.M., 1973, Roughness coefficients

for stream channels in Arizona: U.S. Geological Survey
Open-File Report, 87 p.
Arcement, G.J., Colson, B.E., and Ming, C.O., 1979a, Backwater at bridges and densely wooded flood plains, Cypress
Creek near Downsville, Louisiana: U.S. Geological Survey
Hydrologic Investigations Atlas, HA-603, scales 1:62,500
and 1:2,000, three sheets.
1979b, Backwater at bridges and densely wooded flood
plains, Flagon Bayou near Libuse, Louisiana: U.S. Geological Survey Hydrologic Investigations Atlas, HA-604, scale
1:4,000, five sheets.
1979c, Backwater at bridges and densely wooded flood
plains, Tenmile Creek near Elizabeth, Louisiana: U.S. Geo-

logical Survey Hydrologic Investigations Atlas, HA-606,

scales 1:24,000 and 1:4,000, three sheets.
Barnes, H.H., Jr., 1967, Roughness characteristics of natural
channels: U.S. Geological Survey Water-Supply Paper 1849,
213 p.
Benson, M.A., and Dalrymple, Tate, 1967, General field and
office procedures for indirect discharge measurements: U.S.
Geological Survey Techniques of Water-Resources Investigations, book 3, chap. Al, 30 p.
Burkham, D.E., and Dawdy, D.R., 1976, Resistance equation for
alluvial-channel flow: Proceedings, American Society of
Civil Engineers, Journal of the Hydraulics Division, v. 102,
no. HY10, p. 1479-1489.
Carter, R.W., Einstein, H.A., Hinds, Julian, Powell, R.W., and
Silberman, E., 1963, Friction factors in open channels,
progress report of the task force on friction factors in open
channels of the Committee on Hydromechanics of the
Hydraulics Division: Proceedings, American Society of
References Cited

37

Civil Engineers, Journal of the Hydraulics Division, v. 89,

no. HY2, pt. 1, p. 97-143.
Chow, V.T., 1959, Open-channel hydraulics: New York,
McGraw-Hill Book Co., 680 p.
Colson, B.E., Arcement, G.J., and Ming, C.O., 1979, Backwater
at bridges and densely wooded flood plains, Coldwater River
near Red Banks, Mississippi: U.S. Geological Survey
Hydrologic Investigations Atlas, HA-593, scales 1:24,000
and 1:8,000, three sheets.
Colson, B.E., Ming, C.O., and Arcement, G.J., 1979a, Backwater at bridges and densely wooded flood plains, Yockanookany River near Thomastown, Mississippi: U.S. Geological
Survey Hydrologic Investigations Atlas, HA-599, scales
1:62,500 and 1:8,000, nine sheets.
1979b, Backwater at bridges and densely wooded flood
plains, Thompson Creek near Clara, Mississippi: U.S. Geological Survey Hydrologic Investigations Atlas, HA-597,
scales 1:24,000 and 1:8,000, three sheets.
Cowan, W.L., 1956, Estimating hydraulic roughness coefficients:
Agricultural Engineering, v. 37, no. 7, p. 473-475.
Henderson, P.M., 1966, Open-channel flow: New York, MacMillan Publishing Co., Inc., 522 p.
Limerinos, J.T., 1970, Determination of the Manning coefficient
from measured bed roughness in natural channels: U.S.
Geological Survey Water-Supply Paper 1898-B, 47 p.
Ming, C.O., Colson, B.E., and Arcement, G.J., 1979, Backwater
at bridges and densely wooded flood plains, Pea Creek near

38

Louisville, Alabama: U.S. Geological Survey Hydrologic

Investigations Atlas, HA-608, scales 1:24,000 and 1:2,000,
three sheets.
Petryk, Sylvester, and Bosmajian, George, III, 1975, Analysis of
flow through vegetation: Proceedings, American Society of
Civil Engineers, Journal of the Hydraulics Division, v. 101,
no. HY7, p. 871-884.
Ree, W.O., 1954, Handbook of channel design for soil and water
conservation: Soil Conservation Service, U.S. Department of
Agriculture, SCS-TP-61, 40 p.
Ree, W.O., and Crow, F.R., 1977, Friction factors for vegetated
waterways of small slope: Agricultural Research Service,
U.S. Department of Agriculture, ARS-S-151, 56 p.
Schneider, V.R., Board, J.W., Colson, B.E., Lee, F.N., and
Druffel, Leroy, 1977, Computation of backwater and discharge at width constrictions of heavily vegetated flood
plains: U.S. Geological Survey Water-Resources Investigations 76-129, 64 p.
Simons, D.B., and Richardson, E.V., 1966, Resistance to flow in
alluvial channels: U.S. Geological Survey Professional Paper
422-J, 61 p.
Simons, D.B., Li, R.M., and Associates, 1982, Resistance to
flow in alluvial channels, chap. 6, in Engineering analysis of
fluvial systems: Fort Collins, Colorado, Simons, Li, and
Associates, p. 6.11-6.17.
Streeter, V.L., 1971, Fluid mechanics: New York, McGraw-Hill
Book Co., 5th ed. 705 p.

Guide for Selecting Manning's Roughness Coefficients for Natural Channels and Flood Plains