Biosystems Engineering (2003) 85 (3), 365379

doi:10.1016/S1537-5110(03)00066-7
Available online at www.sciencedirect.com
SW}Soil and Water
Identication and Prioritisation of Critical Sub-watersheds for Soil Conservation
Management using the SWAT Model
M.P. Tripathi
1
; R.K. Panda
2
; N.S. Raghuwanshi
2
1
Soil and Water Engineering Department, Faculty of Agricultural Engineering, Indira Gandhi Agricultural University, Raipur 492 012,
Chhattisgarh, India; e-mail of corresponding author: drmpt64@hotmail.com
2
Agricultural and Food Engineering Department, Indian Institute of Technology, Kharagpur 721 302, West Bengal, India;
e-mail: rkp@agfe.iitkgp.ernet.in
(Received 12 September 2001; accepted in revised form 20 March 2003; published online 21 May 2003)
A few areas of the watershed are critical and responsible for high amount of soil and nutrient losses.
Implementation of best management practices is required in those critical erosion prone areas of the watershed
for controlling the soil and nutrient losses. Identication of these critical areas is essential for the effective and
efcient implementation of watershed management programmes. In this study, a calibrated Soil and Water
Assessment Tool (SWAT) model was veried for a small watershed (Nagwan) and used for identication and
prioritisation of critical sub-watersheds to develop an effective management plan. Daily rainfall, runoff and
sediment yield data of 7 years (19921998) were used in this study. Data related to nutrient losses for few
storm events of 1997 were also used. Besides these data, the topographical map, soil map, land resources data
and satellite imageries of the study watershed were used in this study.
A geographical information system was used for generating the watershed and sub-watershed boundaries,
drainage networks, slope, soil series and texture maps. Supervised classication method was used for land use/
cover classication from satellite imageries. The weighted average values of parameters such as runoff curve
number, surface slope, channel length, average slope length, channel width, channel depth, soil erodibility
factor and other soil layer data were taken for each sub-watershed to verify the model.
The calibrated SWAT model was veried for the monsoon season on daily basis for the year 1997 and
monthly basis for the years 19921998 for both surface runoff and sediment yield. It was also tested for the
available data on nutrient losses. Critical sub-watersheds were identied on the basis of average annual
sediment yield and nutrient losses during the period of 3 years 19961998. The erosion rates and their classes
were used as a criterion for identifying the critical sub-watersheds. Out of the 12 sub-watersheds, one sub-
watershed fell under moderate soil loss group and ve sub-watersheds fell under high soil loss group of soil
erosion classes whereas other sub-watersheds fell under slight erosion classes. The study revealed that the
SWAT model could successfully be used for identifying and prioritising critical sub-watersheds for
management purposes.
# 2003 Silsoe Research Institute. All rights reserved
Published by Elsevier Science Ltd
1. Introduction
The Resource considerations for implementation of
watershed management programmes or various other
reasons related to administration or even political
considerations may limit the implementation of manage-
ment programmes to a few sub-watersheds only. Even
otherwise, it is always better to start management
measures from the most critical sub-watershed, which
makes it mandatory to prioritise the sub-watershed
available. Watershed prioritisation is thus the ranking of
different critical sub-watersheds of a watershed accord-
ing to the order in which they have to be taken up for
treatment and soil conservation measures
The intensive study of individual watersheds is
necessary to enable management plans to be developed
and also to apply the results of one watershed, to
another with similar characteristics. Effective control of
ARTICLE IN PRESS
1537-5110/03/$30.00 365 # 2003 Silsoe Research Institute. All rights reserved
Published by Elsevier Science Ltd
soil and nutrient losses requires implementation of best
management practices in critical erosion prone areas of
the watershed. It can be enhanced by the use of
physically based distributed parameter models, remote
sensing technique and geographic information system
that can assist management agencies in both identifying
most vulnerable erosion prone areas and selecting
appropriate management practices.
Numerous studies have indicated that, for many
watersheds, a few critical areas are responsible for a
disproportionate amount of the pollution (Dickinson
et al., 1990; Dillaha, 1990; Maas et al., 1985; Storm et al.,
1988). Critical areas of non-point source pollution can
be dened both from the land resources and the water
quality perspectives (Maas et al., 1985). From the land
resource perspective, critical areas are those land areas
where the soil erosion rate exceeds the soil loss tolerance
value. Critical areas from the water quality perspectives
are areas where the greatest improvement can be
achieved with the least capital investment in best
management practices.
The average soil loss value of 164 t ha
1
yr
1
(Dhruva
Narayana, 1993) and permissible soil loss value of
112 t ha
1
yr
1
(Mannering, 1981) can be taken into
consideration for identifying the critical sub-watershed.
Priorities can be xed on the basis of ranks assigned to
each critical sub-watershed according to ranges of soil
erosion classes described by Singh et al. (1992) for the
Indian condition. They categorised the soil loss ranges
into different soil erosion classes. Soil erosion classes
such as slight (05 t ha
1
yr
1
), moderate (510 t ha
1
yr
1
), high (1020 t ha
1
yr
1
), very high (20
40 t ha
1
yr
1
), severe (4080 t ha
1
yr
1
) and very severe
(>80 t ha
1
yr
1
) were reported by Singh et al. (1992).
An average soil loss tolerance value of 90 t ha
1
yr
1
was used for Nomini Creek watershed located in
Westmoreland County, Virginia by Tim et al. (1992) in
their study for identifying the critical areas from the
land source prospective. They also considered a thresh-
old value for the loading rate P of 112 kg ha
1
yr
1
.
This threshold value of P loading was obtained from the
work of DelRegno and Atkinson (1988). Tim et al.
(1992) classied watershed areas into three classes i.e.
low, medium and high potential areas from both land
resource and water quality prospective. Also for nutrient
losses a threshold value of 10 mg
1
l for nitrate nitrogen
and 0.5 mg
1
l for dissolve phosphorous as described by
EPA (1976) can be considered as criterion for identify-
ing the critical sub-watersheds.
The Soil Conservation Department of Damodar
Valley Corporation (DVC) Hazaribagh, Bihar (India)
has demarcated 20 prioritised sub-watersheds out of 39
sub-watersheds for treating them with the appropriate
soil conservation measures (Misra, 1986). The prioriti-
sation of these sub-watersheds was based on an
empirical formula developed by DVC using a limited
stream ow record of only 3 years. They considered
three priority criteria, i.e. priority I (erosion index of 30
and above), priority II (erosion index lying between 15
and 30) and priority III (erosion index less than 15). The
actual formula used for calculating the erosion index I
E
is as follows:
I
E
5
A
u
3
A
gw
A
df

A
f
5
1
where: A
u
is the upland area in ha; A
gw
is the gullied
wasteland area in ha; A
df
is the denuded forestland area
in ha; and A
f
is the wood forestland area in ha.
Several techniques, ranging from manual overlay of
spatially -index mapped data to pollutant yield model-
ling, have been used to characterise and delineate critical
areas of non-point source pollution in complex land-
scapes. McHarg (1969) used a manual map overlay
system to display the common attributes of selected land
areas in order to make decisions on the type and degree
of land development that is commensurate with the
physical properties and limits of an area. Recently, there
has been a shift towards the use of computerised data
management systems to facilitate the delineation of
critical areas of non-point source pollution (Hession &
Shanholtz, 1988; Vieux, 1991). Some of the research
workers used the sediment yield index I
sy
method for
prioritisation of sub-watersheds (Karale et al., 1975,
1977). They used the following equation for computa-
tion of I
sy
.
I
sy
5
P
E
i
A
ei
D
r

A
w
100 2
where: I
sy
is the sediment yield index; E
i
is the weighing
value of erosion intensity mapping unit; A
ei
is the area of
the erosion intensity mapping unit in a watershed in ha;
D
r
is the delivery ratio; and A
w
is the total area of
watershed in ha.
Several physically based distributed parameter models
(ANSWERS, AGNPS, SHE, SWRRB and SWAT) have
been developed to predict runoff, erosion, sediment and
nutrient transport from agricultural watersheds under
various management regimes. Among these models, Soil
and Water Assessment Tool (SWAT) is the most recent
one used successfully for simulating runoff, sediment
yield and water quality of small watersheds. The SWAT
model is a distributed parameter, continuous model
developed by the USDA-ARS (Arnold et al., 1996,
1998). The SWAT model was tested mainly on monthly
and annual basis for predicting runoff and sediment
yield (Srinivasan et al., 1993; Srinivasan & Arnold, 1994;
Rosenthal et al., 1995; Bingner, 1996; Bingner et al.,
1997; Peterson & Hamlett, 1998). However, Tripathi
ARTICLE IN PRESS
M. P. TRIPATHI ET AL. 366
et al. (1999a, 1999b) tested the SWAT model on the
basis of daily runoff and sediment yield for an Indian
watershed, namely Nagwan in the Hazaribagh district of
the Bihar State in India. Limited research work on
identication of critical sub-watersheds and assessment
of the impact of management practices on runoff,
sediment yield and nutrient losses using SWAT has
been reported (Arnold et al. 1999; Santhi et al., 2001).
Also the testing of SWAT model on daily basis for the
nutrient losses has not appeared much in the literature.
Keeping the above facts in mind, the current study
was undertaken with the use of a veried model i.e.
SWAT to identify the critical sub-watersheds on the
basis of estimated sediment yield and nutrient losses of a
small watershed for the purpose of developing the
effective management plan.
2. Theoretical considerations
SWAT model is a distributed parameter model that
operates on a daily time step. The major goal of the
model development was to predict the impact of
management measures on water, sediment and agricul-
tural chemical yields in large ungauged basins (Arnold
et al., 1996). It is comparatively simple, user friendly,
physically based and distributed, which uses readily
available inputs. It is computationally efcient to
operate on large basins in a reasonable time. It is a
continuous time-scale model, capable of simulating
long-term effects of management change.
The SWAT model uses a command structure for
routing runoff and chemicals through a watershed
similar to the structure of problem-oriented lang-
uage for hydrologic modelling (known as the HYMO
model) (Williams & Hann, 1973). Specic commands
are there for routing ows through streams and
reservoirs; adding ows and inputting measured data
or point sources. Using a routing command language,
the model can simulate a basin subdivided into
hydrological response units (HRU), grid cells or sub-
watersheds.
The SWAT predicts surface runoff for daily rainfall by
using the Soil Conservation Service (SCS) curve number
(CN) method (USDA-SCS, 1972). Sediment yield is
computed for each subbasin with the Modied Uni-
versal Soil Loss Equation (MUSLE) (Williams &
Berndt, 1977). Subbasin nutrient yield and nutrient
cycling were taken from the EPIC model (Williams et al.,
1984) and modied as necessary for inclusion into the
SWAT model (Arnold et al., 1996). The SWAT model
allows for simultaneous computations on each subbasin
and routes the water, sediment, and nutrients from the
subbasin outlets to the basin outlet.
2.1. Nitrogen in subbasin
The amount of NO
3
-N contained in surface runoff is
estimated for each subbasin by considering the rst layer
(10 mm thickness) only. The total amount of water
leaving the layer is the sum of runoff, lateral subsurface
ow, and percolation.
Q
T
5Q O
1
Q
R1
3
where: Q
T
is the total water lost from the rst layer in
mm; Q is the runoff volume in mm; O
1
is the percolation
from the rst layer in mm; and Q
R1
is the lateral ow
from the rst layer in mm.
Amounts of NO
3
-N contained in runoff, lateral ow
and percolation are estimated as the products of the
volume of water lost and the average concentration of
NO
3
-N.
V
NO
3
5Q
T
C
NO
3
4
where: V
NO
3
is the amount of NO
3
-N lost from the rst
layer; and C
NO
3
is the concentration of NO
3
-N in the
rst layer.
Leaching and lateral subsurface ow in lower layers
are treated with the same approach used in the upper
layer except that surface runoff is not considered.
A loading function developed by McElroy et al.,
(1976) and modied by Williams and Hann (1978) for
application to individual runoff events is used to
estimate organic N transport by sediment. The loading
function estimates the daily organic N runoff loss based
on the concentration of organic N in the topsoil layer,
the sediment yield and enrichment ratio. The loading
function
Y
ON
50001YC
ON
E
R
5
where: Y
ON
is the organic N runoff loss at the subbasin
outlet in kg ha
1
; C
ON
is the concentration of organic N
in the top soil layer in g t
1
; Y is the sediment yield in
t ha
1
; and E
R
is the enrichment ratio.
The value of C
ON
is input to the model and is constant
throughout the simulation. Enrichment ratios are
logarithmically related to sediment concentration. The
logarithmic equation estimating enrichment ratio is
E
R
5x
1
c
x
2
a
6
where: c
a
is the sediment concentration in g m
3
; and x
1
and x
2
are parameters set by the upper and lower limits.
The enrichment ratio to approach 10, the sediment
concentration would be extremely high. Conversely,
a very low sediment concentration would cause
the enrichment ratio to approach inverse of the
sediment delivery ratio D
R
, which is the ratio of
subbasin sediment yield and gross sheet erosion. The
simultaneous solution of Eqn. (6) at the boundaries
assuming sediment concentrations range from 500 to
ARTICLE IN PRESS
IDENTIFICATION AND PRIORITISATION OF CRITICAL SUB-WATERSHEDS 367
250 000 g m
3
gives
x
2
5
log 1=D
R

2699
7
x
1
5
1
025
x
2
8
The technique used to estimate the denitrication rate
is described in the users manual of the SWAT model
(Arnold et al., 1996). The N mineralisation model is a
modication of the PAPRAN, which is known as a
simulation model of annual pasture production limited
by rainfall and nitrogen (Seligman & Keulen, 1981). The
daily amount of immobilisation is computed by
subtracting the amount of N contained in the crop
residue from the amount assimilated by the microorgan-
isms. Immobilisation may be limited by N availability. If
the amount of N available is less than the amount of
immobilisation predicted, the decay rate constant is
adjusted. To estimate the N contribution from rainfall,
SWAT uses an average rainfall N concentration of
8 ppm for all locations for all storms. The amount of N
in rainfall is estimated as the product of rainfall amount
and concentration.
Crop use of N is estimated using a supply and demand
approach. The daily (day i) crop N demand can be
computed using
C
NDi
5C
NB

i
B
i
C
NB

i1
B
i1
9
where: C
NDi
is the N demand of the crop in kg ha
1
; C
NB
is the optimal N concentration of the crop; and B is the
accumulated N in kg ha
1
.
The crop is allowed to take N from any soil layer that
has roots. Uptake starts at the upper layer and proceeds
downward until the daily demand is met or until all N
has been depleted. If the soil cannot supply the daily
N demand for legumes, the decit is attributed to N
xation.
2.2. Phosphorus in subbasin
The SWAT approach is based on the concept of
partitioning pesticides into the solution and sediment
phases (Knisel, 1980). As P is mostly associated with the
sediment phase, the soluble P runoff equation can be
expressed in the simple form
Y
SP
5
001C
LPP
Q
k
d
10
where: Y
SP
is the soluble P in kg ha
1
lost in runoff
volume Q in mm; C
LPP
is the concentration of soluble P
in soil layer in g t
1
; and k
d
is the P concentration in
the sediment divided by that of the water in m
3
t
1
.
The value of C
LPP
is input to the model and remains
constant. The value of k
d
used in SWAT is 175.
Sediment transport of P is simulated with a loading
function. The P loading function is
Y
P
5001YC
P
E
R
11
where: Y
P
is the sediment phase of P loss in runoff in
kg ha
1
and C
p
is the concentration of P in the topsoil
layer in g t
1
.
The P immobilisation model, also developed by Jones
et al. (1984) are similar in structure to the N
immobilisation model. The daily amount of immobilisa-
tion is computed by subtracting the amount of P
contained in the crop residue from the amount
assimilated by the microorganisms.
2.3. Channel sediment routing
The sediment routing model consists of two compo-
nents operating simultaneously (deposition and degra-
dation). Deposition in the stream channel is based on
the fall velocity of the sediment particles (Arnold et al.,
1990). With a temperature of 228C and a sediment
density of 12 t m
3
, Stokes Law for fall velocity
becomes
V
f
5411 d
2
12
where: V
f
is the fall velocity in mh
1
and d is the
sediment particle diameter. The depth y
f
that sediment
of particle size d will fall during time, T
T
is
y
f
5V
f
T
T
13
The sediment delivery ratio D
R
through the reach is
estimated with the following equations.
D
R
5
1 05y
f
d
q
; y
f
d
q
14
D
R
5
05 d
q

y
f
; y
f
> d
q
15
where: d
q
is the depth of ow.
Finally, deposition S
D
is calculated with the equation:
S
D
5S
IN
1 D
R
16
where S
IN
is the sediment entering the reach.
Stream power is used to predict degradation in the
routing reaches. Williams (1980) used Bagnolds (1977)
denition of stream power to develop a method for
determining degradation in channels. Bagnold dened
stream power P
S
with the equation
P
S
5gqS
w
17
where: g is the density of the water; q is the ow rate;
and S
W
is the water surface slope. By applying stream
power to bed load predictions (Bagnold, 1977) and
ARTICLE IN PRESS
M. P. TRIPATHI ET AL. 368
estimating model parameters (Williams, 1980), the
equation for sediment reentrained, D
r
, is
D
r
5a
sp
g
15
F
dur
wd
q
S
W
V
c

15
18
where: a
sp
is a parameter dependent on maximum
stream power for the reach; F
dur
ow duration in h;
and V
c
is the velocity in the channel.
The parameter a
sp
can be estimated with the equation
a
sp
5g
w
qS
c

05
mx
19
where: S
c
is the slope of the channel and the subscript
mx refers to the maximum ow expected in the reach for
extreme events. The value of q is assumed to equal some
maximum rainfall intensity (250 mmh
1
) and a
sp
be-
comes
a
sp
56944gA
D
S
c

05
20
where: A
D
is the drainage area into the reach in km
2
. All
of the stream power is used for reentrainment of loose
and deposited material until all of the material has been
removed. When this occurs, degradation of the bed
material D
B
begins and is calculated by
D
B
5KCD
r
21
where: K and C are MUSLE (Williams & Berndt, 1977)
factors for the stream channel.
Total degradation D
T
is the sum of the reentrainment
and bed degradation components. This amount is also
allowed to be redeposited before reaching the basin
outlet.
D
T
5D
r
D
B
1 D
R
22
Finally, the amount of sediment reaching the basin
outlet, S
OUT
, is
S
OUT
5S
IN
S
D
D
T
23
where: S
IN
is the sediment entering the reach.
2.4. Nitrate and phosphorus routing
Once NO
3
-N enters a stream it is considered a
conservative material for the duration of an individual
runoff event (Williams, 1980). Thus, NO
3
-N routing is
simply a matter of adding the yields from all subbasins
to determine the basin yield.
The loading function approach is also used in routing
organic N from the subbasin outlets to the basin outlet:
Y
ON

Bj
5001 Y
B

j
Y
CON

j
E
RR

j
24
where: (Y
ON
)
B
is the organic N runoff loss at the basin
outlet in kg ha
1
; Y
B
is the sediment yield reaching the
basin outlet from subbasin j in t ha
1
; Y
CON
is
the concentration of organic N in the sediment reaching
the subbasin j outlet in g t
1
; and E
RR
is the enrichment
ratio for the channel routing from subbasin j to the
channel outlet.
The delivery ratio for the channel routing is calculated
from
D
R
5
Y
SB

j
Y
B

j
25
where: Y
SB
is the sediment yield at the subbasin outlet in
t ha
1
; and Y
B
is that sediment yield from subbasin j
after it has been routed to the basin outlet in t ha
1
.
As with NO
3
-N routing, once soluble P enters a
stream it is considered a conservative material and
routing is accomplished by adding the yields from all
subbasins to determine the basin yield.
Again, the loading function approach is used in
routing P from the subbasin outlets to the basin outlet.
Y
P

Bj
5001 Y
B

j
C
PSB

j
E
RR

j
26
where: (Y
P
)
B
is the P yield at the basin outlet in kg ha
1
;
and C
PSB
is the P concentration in the sediment reaching
the subbasin j outlet in g t
1
.
3. Materials and methods
3.1. Study area and data collection
The selected Nagwan watershed (9246 km
2
) is located
in Upper Damoder Valley Corporation (DVC) in the
Hazaribagh district of Bihar, India. Location map of the
study area is shown in Fig. 1. The watershed receives an
average annual rainfall of 1256 mm, out of which the
monsoon season (JuneOctober) contributes more than
80% rainfall. Rainfall and runoff data for 7 years (1992
1998) from the gauging station at Nagwan sediment
observation post were collected from DVC, Hazaribagh.
IRS-1B (LISS II) satellite data with date of pass 19th
October 1996 were collected and used for land use/land
cover classication. Topographic maps (1:50 000) were
collected from Survey of India, Calcutta, and the soil
resources data were collected from DVC, Hazaribagh,
for use in the study.
The software available at Regional Remote Sensing
Service Center (RRSSC), Kharagpur, such as the
Environmental Analysis and Scientic Interface (EASI)
and Picture Analysis, Correlation and Enhancement
(PACE) commonly known as EASI-PACE (PCI Inc.,
1994) for terrain analysis and image processing were
used. Extracted data were processed with the help of
Excel package of Microsoft Ofce 97 and all the input
data le were generated in DOS using UTIL programme
which was built-up with the SWAT model used in the
study (Arnold et al., 1996, 1998).
ARTICLE IN PRESS
IDENTIFICATION AND PRIORITISATION OF CRITICAL SUB-WATERSHEDS 369
3.2. Extraction of watershed parameters for the model
The digitised contour vectors were gridded using the
EASI-PACE and converted to grided data. The grids
were interpolated using the k-nearest neighbours meth-
od for generating the digital elevation model (DEM).
The DEM of the watershed was prepared in 30 m by
30 m resolution. Many researchers have also used DEM
of 30 m by 30 m resolution and obtained satisfactory
results (Bingner, 1996; Sharma et al., 1996; Tiwari et al.,
1997; Wang & Hjelmfelt, 1998).
The watershed and sub-watershed boundaries, drai-
nage networks and slope map were generated using the
procedure described by Jenson and Domingue (1988).
The area delineated by the algorithm was 9023 km
2
against the manually judged area of 9246 km
2
. The
automatically delineated watershed was used for
further study. Since SWAT works on sub-watersheds
basis, the delineated watershed was subdivided into 12
sub-watersheds on the basis of topography using
procedure similar to that used for delineation of main
watershed (Fig. 2). In this study Nagwan watershed was
coded as WS and sub-watershed were coded as WS1 to
WS12.
Supervised land use classication method was used
for land use/land cover classication. The identied land
use classes were upland paddy (rice), low land paddy,
orchards, deep water, shallow water, closed forest, open
forest, fallow land, grasses/shrubs, upland crops and
settlements. Soil textures and soil series maps were also
generated. There are mainly three soil series (Harina,
Bhuswa and Atia) in the watershed. The predominant
soil of the watershed is silt loam. Sandy loam, clay loam,
ARTICLE IN PRESS
WS1
WS2
WS3
WS8
WS11
WS6
WS9
WS7
WS5
WS4
WS10
WS12
Outlet
N
Watershed boundary
Sub-watershed boundary
Drainage channel
Fig. 2. Geographic information system delineated sub-water-
sheds
Nagwan Watershed
in India
DELHI
PATNA
DVC
HAZARIBAGH
CALCUTTA
LEGEND:
O-Outlet/gauging station
Not to the scale
N
85
.
43E
24
.
12N
85
.
43E
23
.
99N
85
.
25E
23
.
99N
Fig. 1. Location of Nagwan watershed in India
M. P. TRIPATHI ET AL. 370
loam, loamy sand and silty clay loam are the other types
(as per the USDA nomenclature) of soil found in the
watershed. Layerwise soil properties under different soil
series of the watershed are given in Table 1.
The weighted curve number for each sub-watershed
was calculated using land use/land cover map, soil
texture maps and standard curve numbers for the Indian
conditions (Dhruva Narayana, 1993). Other input
parameters of the delineated sub-watersheds, such as
overland and channel slope, channel length and average
slope length were extracted using the various maps
including contour map, sub-watershed map, slope map
and drainage map (Table 2).
3.3. Model calibration and validation
The SWAT model has already been calibrated and
validated by Tripathi et al. (1999a, 1999b) for Nagwan
watershed on monthly and daily basis for runoff and
sediment yield for the year 1991 and 1992, respectively.
The calibrated values for hydraulic conductivity of
alluvium for surface runoff and channel runoff were found
to be 64 and 10, respectively, and values of Mannings
roughness coefcient n for overland ow and channel ow
were found to be 0065 and 0040, respectively.
Since data for nutrient losses were available for few
events during the monsoon season of the year 1997 only,
ARTICLE IN PRESS
Table 1
Soil properties under different soil series of Nagwan watershed
Soil
Series
Depth,
mm
Bulk,
density,
g/cm
3
Available
water
capacity
Organic
carbon,
%
Course
sand,
%
Fine
sand, %
Silt,
%
Clay,
%
Harina 150 155 014 047 401 388 117 79
370 145 014 041 383 397 118 124
690 131 012 021 126 288 191 396
1050 125 011 016 179 267 103 438
Bhushwa 150 136 020 055 47 204 544 188
370 131 018 051 116 315 362 194
690 133 014 025 48 281 253 405
1050 147 011 016 84 233 235 434
Atia 150 131 020 054 54 333 357 236
370 131 018 053 58 290 328 309
690 133 014 024 22 228 282 457
1050 140 014 021 19 235 274 461
Table 2
Sub-watershed wise input data for the SWAT model
Sub-watershed Area,
km
2
Slope,
%
Curve number Av. slope
length, m
Channel
length, km
Channel
slope
K
factor
P
factor
C
factor
1991 1996
WS1 1719 22 731 836 4643 960 0005 028 060 10
WS2 933 30 687 710 4938 528 0008 019 050 10
WS3 627 21 769 797 4816 180 0001 022 060 10
WS4 989 22 589 550 4564 540 0004 026 060 10
WS5 1467 21 668 689 3958 600 0005 021 060 10
WS6 354 28 771 801 4923 225 0001 019 050 10
WS7 946 31 715 690 5170 576 0005 024 050 10
WS8 424 23 677 689 5743 294 0006 019 060 10
WS9 310 29 731 747 4378 225 0008 023 050 10
WS10 723 33 671 665 4547 540 0009 023 050 10
WS11 480 29 777 792 4794 336 0009 017 050 10
WS12 051 91 731 668 2908 090 0006 025 060 10
WS
*
9023 23 710 720 4617 1386 0005 021 060 10
Note: K, soil erodibility factor;
P, soil conservation practice factor;
C, crop management factor.
*
Whole Nagwan watershed.
IDENTIFICATION AND PRIORITISATION OF CRITICAL SUB-WATERSHEDS 371
the model was again tested for the year 1997 for daily
runoff and sediment yield beside nutrient losses for the
few available events at the outlet of the Nagwan
watershed.
Since model was used to identify critical sub-water-
sheds according to the annual sediment yield. Therefore
it was also validated on monthly basis for simulating the
surface runoff and sediment yield for monsoon seasons
of multiple years (19921998). The model performance
was evaluated on the basis of test criterion recom-
mended by ASCE Task Committee (1993). Graphical
and statistical methods were also used for evaluating the
model performance. The numerical performance criteria
are briey described below.
Martinec and Rango (1989) recommended that the
criteria should be as simple as possible. The percent
deviation of runoff volumes D
V
is one goodness-of-t
criterion.
D
V
5
V V
0
V
100 27
where: V is the measured yearly or seasonal runoff
volume and V
0
is the model computed yearly or seasonal
runoff volume. D
V
can take any value. However, smaller
the number better the model results. D
V
would equal
zero for a perfect model.
The second basic goodness-of-t criterion recom-
mended by ASCE Task Committee (1993) is the Nash
Sutcliffe coefcient or coefcient of simulation efciency
C
OE
(Nash & Sutcliffe, 1970):
C
OE
51
P
n
i 51
Q
i
Q
0
i

2
P
n
i 51
Q
i
Q
2
28
where: Q
i
is the measured daily discharge; Q
0
i
is the
computed daily discharge; and Q is the average
measured discharge values. The values for C
OE
can be
varied from 0 to 1, with 1 indicating a perfect t. A value
for C
OE
equal to zero indicates that the model was
simulating no better than using the average of the
observed data. Martinec and Rango (1989) recom-
mended using Q for the year or season to avoid
unrealistically high values of C
OE
in low runoff years.
Successful application of a calibrated hydrologic
watershed model depends on how well the model is
veried. The parameters required for the model were
extracted from the analysis of DEM, soil map and
satellite imagery. Those parameters were then used in
the SWAT model. The observed daily rainfall and
temperature data for 1 year (1997) and multiple years
(19921998) were used for the validation of the model
on daily and monthly basis, respectively. The model was
also validated for the nutrient losses for the 12 events of
1997. Since land use data for the year 1997 was not
available, estimated curve number (CN) values of 1996
were given as input to the model. The soil series and soil
texture were different for different sub-watersheds.
Therefore, weighted average of all the soil resource data
were given as input to the model.
3.4. Identication and prioritisation of critical
sub-watersheds
The veried model was applied for identifying and
prioritising the critical sub-watersheds of the Nagwan
watershed. Since average rainfall data of 3 years (1996
1998) were signicantly similar to the average rainfall
data of 13 years (19861998), simulations were per-
formed using rainfall data of 3 years for identication
and prioritisation of the critical watersheds on the
basis of average annual sediment yield and nutrient
losses. The ranges of erosion rates and their classes
suggested by Singh et al. (1992) were inferred for
identication of critical sub-watersheds. The critical
sub-watersheds were then prioritised and proposed for
evaluating the management scenarios to reduce the
runoff rate, sediment yield and nutrient losses from the
watershed.
A particular sub-watershed may get top priority due
to various reasons but often, the intensity of land
degradation is taken as the basis. This approach of
prioritising watersheds based on actual sediment yield
rates may be possible only when the number of sub-
watersheds to be prioritised is less and necessary data
are available. Further, this method will be helpful when
the sediment yield potentials of different sub-watersheds
do not have considerable variation.
The critical sub-watersheds were identied on the
basis of average annual sediment and nutrient losses
from the sub-watersheds during the period of 1996
1998. In this context, mean annual sediment yields were
simulated for each sub-watershed of Nagwan watershed
using SWAT model. Priorities were xed on the basis of
ranks assigned to each critical sub-watershed according
to ranges of soil erosion classes described by Singh et al.
(1992) (Table 3). Soil loss values of identied critical
sub-watersheds were compared with the average soil loss
(1635 t ha
1
yr
1
) of India (Dhruva Narayana, 1993)
and prescribed permissible upper limit of soil loss
(112 t ha
1
yr
1
) (Mannering, 1981) for justication of
results. Also for nutrient losses a threshold value of
10 mg l
1
for nitrate nitrogen and 05 mg l
1
for dis-
solved phosphorous as described by Environmental
Protection Agency (EPA) were considered as criterion
for identifying the critical sub-watersheds (EPA, 1976).
Identied critical sub-watersheds were arranged in
descending order and then priorities were xed for their
management.
ARTICLE IN PRESS
M. P. TRIPATHI ET AL. 372
4. Results and discussion
4.1. Surface runoff
The graphical representation of validation results for
the daily runoff is shown in Fig. 3. The graphs show that
the magnitude and temporal variation of simulated
runoff matched closely with the observed runoff values
for the entire monsoon season of 1997. Timings of
occurrence of the peaks for both observed and simulated
runoff matched well. However, the model overpredicted
runoff for few rainfall events of high magnitude.
The descriptive statistics for both measured and
simulated daily runoff are given in Table 4. The mean
values of observed and simulated runoff were not
signicantly different at 95% level of condence, the t-
calculated (053) being less than the t-critical (198).
Similarity in means and standard deviation revealed
similarity in the frequency distributions of observed and
simulated runoff. However, the highest peak runoff
value predicted by the model was slightly higher than
that of the observed value. The per cent deviation D
v
value indicated that the model was underpredicting
runoff by 46%, whereas a high value (087) of the
NashSutcliffe simulation efciency C
OE
showed a close
agreement between the measured and simulated runoff.
Daily predicted runoff values for the monsoon season
of 1997 were plotted against the measured values and
their distribution along with the 1:1 line is shown in
Fig. 4. The results showed that the distribution of
observed and simulated runoff was uniform throughout
the season. Regression analysis between the observed
and simulated runoff values resulted in a high value
(091) of the coefcient of determination r
2
indicating a
close relationship between measured and simulated
runoff (Fig. 4, Table 4).
4.2. Sediment yield
The time series of observed and simulated daily
sediment yield of the Nagwan watershed for the
validation period (June 1October 31, 1997) were
compared graphically as shown in Fig. 5. The time to
peaks of simulated sediment yield matched consistently
well with the time to the measured peak sediment yield
throughout the season. However, the model slightly
under predicted few peak events of sediment yield.
Daily predicted sediment yields were plotted against
the measured values and their distribution along with
the 1:1 line as shown in Fig. 6. The simulated sediment
yields were distributed uniformly about the 1:1 line for
both lower and higher values of observed sediment yield.
Regression analysis was also performed between the
observed and simulated sediment yield values and the
best-t line is shown in Fig. 6. A value 089 of r
2
indicates a close relationship between the measured and
simulated sediment yields. The NashSutcliffe simula-
tion efciency of 089 indicated that there was good
agreement between the observed and simulated sediment
yields during the validation period of 1997. The overall
deviation between the simulated and observed sediment
yields was found to be 143%.
The descriptive statistics for both measured and
predicted daily sediment yields are given in Table 4. A
close agreement between means and standard deviation
of measured and simulated sediment yields indicates that
the frequency distributions for the occurrence of sedi-
ment yields were similar. Comparison of means using
Students t-test (t-calculated value of 006 and t-critical
value of 198) revealed that the mean values of observed
and simulated sediment yields were not signicantly
different at 95% condence level. However, the max-
imum sediment yield predicted by the model was slightly
lower than the observed maximum sediment yield.
4.3. Monthly distribution of runoff and sediment yield
The results of measured and simulated monthly runoff
for the monsoon seasons of 19921998 were compared
ARTICLE IN PRESS
Table 3
Area under different classes of soil erosion by water in India
Serial no.
Soil erosion
classes
Soil erosion
range, t ha
1
yr
1
Area, km
2
1 Slight 05 801 350
2 Moderate 510 1 405 640
3 High 1020 805 030
4 Very high 2040 160 050
5 Severe 4080 83 300
6 Very severe >80 31 895
0
25
50
75
100
1-Jun 1-Jul 1-Aug 1-Sep 1-Oct
Time, days
R
u
n
o
f
f
,

m
m
R
a
i
n
f
a
l
l
,

m
m
0
50
100
150
200
250
Fig. 3. Observed and simulated runoff hydrograph for model
validation (June to October, 1997); ( ), rainfall; ( ),
observed runoff; ( ), simulated runoff
IDENTIFICATION AND PRIORITISATION OF CRITICAL SUB-WATERSHEDS 373
and the results are given in Table 5. Results showed that
the model was underpredicting the monthly runoff
during monsoon season by 60%. Though there was
signicant difference between the observed (6063 mm)
and simulated (5696 mm) means of monthly runoff
because t-calculated (239) was higher than t-critical
(203) at 95% level of condence, a high value (097) of
coefcient of determination r
2
showed close relationship
between observed and simulated monthly surface runoff
for the validation period.
The results of measured and simulated monthly
sediment yield for the monsoon season of 19921998
were also compared (Table 5). Results showed that most
of the time the model predicted sediment yield was close
to the observed sediment yield during each month of
ARTICLE IN PRESS
Table 4
Statistical analysis of the observed and simulated daily surface runoff and sediment yield for monsoon season (JuneOctober) of the
year 1997
Statistics Runoff, mm Sediment yield, t ha
1
Observed Simulated Observed Simulated
Mean 280 268 0026 0022
Standard deviation 706 816 0072 0067
Maximum peak 5033 5700 0538 0460
Total 42907 40940 3885 3330
Count 153 153 153 153
t-calculated 053 006
t-critical (two tailed) 198 198
Coefcient of determination r
2
091 089
Deviation, % 46 143
Coefcient of simulation efciency C
OE
087 089
0
.
0
0
.
2
0
.
4
0
.
6
0
.
8
1
.
0
1-Jun 1-Jul 1-Aug 1-Sep 1-Oct
Time, days
S
e
d
i
m
e
n
t

y
i
e
l
d
,

t

h
a

1
0
50
100
150
200
250
R
a
i
n
f
a
l
l
,

m
m
Fig. 5. Observed and simulated sediment yield for model
validation (June to October, 1997); ( ), rainfall; ( ),
observed sediment yield; ( ), simulated sediment yield
0
10
20
30
40
50
60
70
0 10 20 30 40 50 60 70
Observed runoff, mm
S
i
m
u
l
a
t
e
d

r
u
n
o
f
f
,

m
m
Fig. 4. Comparison between observed and simulated daily runoff
for model validation (June to October, 1997); ( ),
regression line; ( ), 1:1 line
0
.
0
0
.
1
0
.
2
0
.
3
0
.
4
0
.
5
0
.
6
0
.
0 0
.
1 0
.
2 0
.
3 0
.
4 0
.
5 0
.
6
Observed sediment yield, t ha
1
S
i
m
u
l
a
t
e
d

s
e
d
i
m
e
n
t

y
i
e
l
d
,

t

h
a

1
Fig. 6. Comparison between observed and simulated daily
sediment yield for model validation (June to October, 1997);
( ), regression line; ( ), 1:1 line
M. P. TRIPATHI ET AL. 374
monsoon season (r
2
5089). In general, the simulated
monthly sediment yield compared well with measured
values during monsoon season of the validation years
because means were similar at 95% condence level
(Table 5). The results also showed that the model was
overpredicting monthly sediment yield by 40%.
Based on the above results, it could be inferred that
the model was accurately validated for predicting daily
as well as monthly surface runoff and sediment yield
from the Nagwan watershed.
4.4. Nutrient losses
Data for the nutrient losses were not available for the
year 1991. Hence the prediction capability of nutrient
losses of the model could not be calibrated (Tripathi,
1999). Therefore, the model was evaluated by perform-
ing validation only for the year 1997 with the available
nutrient loss data of 12 events in this study. The
nutrients considered for validation were organic nitro-
gen, phosphorus, nitrate nitrogen (NO
3
-N) and soluble
phosphorus.
The graphical comparison between observed and
simulated nutrient losses is shown in Figs 710 and
descriptive statistics is given in Table 6. It was seen that
the observed and simulated values of nutrients including
organic nitrogen, phosphorus, NO
3
-N and soluble P
were uniformly distributed about 1:1 line (Figs 710).
Observed and simulated means of organic nitrogen,
phosphorus, NO
3
-N and soluble P were not signicantly
different at 95% level of condence, since t-calculated
were found to be less than t-critical in all the cases
(Table 6). The D
v
values were found to be 158, 117, 37
and 125%, respectively, for organic N, P, NO
3
-N and
soluble P indicating that the model was predicting
nutrient losses satisfactorily. For organic N, P, NO
3
-N
and soluble P the r
2
values were 082, 086, 089 and 082,
respectively,which is a good agreement between ob-
served and simulated values of nutrient losses.
ARTICLE IN PRESS
Table 5
Results of statistical analysis for observed and simulated monthly surface runoff and sediment yield (19921998)
Statistics Surface runoff, mm Sediment yield, t ha
1
Observed Simulated Observed Simulated
Mean 6063 5696 077 080
t-calculated 239 052
t-critical (two tailed) 203 203
Coefcient of determination r
2
097 089
Deviation, % 60 40
Coefcient of simulation efciency C
OE
098 079
0
0
.
2
0
.
4
0
.
6
0
.
8
0
.
0 0
.
2 0
.
4 0
.
6 0.8
Observed organic N, kg ha
1
S
i
m
u
l
a
t
e
d

o
r
g
a
n
i
c

N
,

k
g

h
a

1
Fig. 7. Comparison of observed and simulated organic nitrogen
for model validation; ( ), regression line; ( ), 1:1 line
0
.
00
0
.
05
0
.
10
0
.
15
0
.
20
0
.
25
0
.
30
0
.
00 0
.
05 0
.
10 0
.
15 0
.
20 0
.
25 0
.
30
Observed nitrate, kg ha
1
S
i
m
u
l
a
t
e
d

n
i
t
r
a
t
e
,

k
g

h
a

1
Fig. 8. Comparison of observed and simulated nitrate nitrogen
for model validation; ( ), regression line; ( ), 1:1 line
IDENTIFICATION AND PRIORITISATION OF CRITICAL SUB-WATERSHEDS 375
On the basis of above validation results for the
nutrient losses, it was conrmed that model was
predicting reasonably well and can be used for manage-
ment and planning of the Nagwan watershed for further
agricultural development.
4.5. Identication and prioritisation of critical
sub-watersheds
Identication and prioritisation of critical sub-water-
sheds based on actual sediment yield rate may be
possible only when sediment data are available.
Mean annual sediment and nutrient losses were
simulated for each sub-watershed of Nagwan watershed
using SWAT model (Table 7) and priorities were
xed as described earlier. Results showed that out of
the 12 sub-watersheds, the WS5 fell under moderate soil
loss group of soil erosion classes (510 t ha
1
yr
1
). The
WS6, WS7, WS9, WS10 and WS12 fell under high soil
loss group of soil erosion classes (1020 t ha
1
yr
1
),
whereas other sub-watersheds fell under slight erosion
classes.
None of the sub-watersheds fell under very high,
severe or very severe erosion classes. This may be
because of the fact that the study watershed is having an
average slope of 23% only. The study watershed might
have got stabilised as contour and graded bunds and
terraces already exist in the watershed. However, the
sub-watershed WS12 resulted in maximum sediment
yield, which is also more than average soil loss
1635 t ha
1
yr
1
of India (Dhruva Narayana, 1993).
This may be due to the high average surface slope of
91% with undulating topography. Sub-watersheds
WS6, WS7, WS9 and WS10 exceeded the prescribed
permissible upper limit of soil loss of 112 t ha
1
yr
1
(Mannering, 1981).
The results indicated that the dissolved nutrient
losses including NO
3
-N and soluble P were within the
permissible limit of 10 and 05 mg l
1
, respectively (EPA,
1976; Tim et al., 1992). The trend showed that losses of
nutrient attached with the sediment were proportional
to losses of sediment from the watershed (Table 7).
Annual losses of nutrient attached with sediment were
found to be more in case of WS12 followed by WS9,
WS7, WS10 and WS6. The nitrate nitrogen and soluble
phosphorous in runoff were not high as compared to
other sub-watersheds.
On the basis of annual sediment yield and nutrient
losses, sub-watersheds WS6, WS7, WS9, WS10 and
WS12 were found to be critical. After arranging the
critical sub-watersheds in ascending order, considering
the annual sediment yield and nutrient losses from each
sub-watershed, priorities were xed. The sub-watershed
that comes rst is given the top priority for developing
the management plan to reduce the soil and nutrient
losses.
As a result the critical sub-watersheds WS12, WS9,
WS7, WS10 and WS6 were selected and recommended
to adopt the management measures in that order to
reduce the sediment and nutrient losses and to conserve
the rainwater within the watershed for sustainable crop
production. Other sub-watersheds were not considered
for management practices because those sub-watersheds
were not yielding sediment more than moderate or slight
erosion classes.
ARTICLE IN PRESS
0
.
0
0
.
1
0
.
2
0
.
3
0
.
4
0
.
5
0
.
0 0
.
1 0
.
2 0
.
3 0
.
4 0.5
Observed P, kg ha
1

S
i
m
u
l
a
t
e
d

P
,

k
g

h
a

1

Fig. 9. Comparison of observed and simulated organic P for
model validation: ( ), regression line; ( ), 1:1 line
0
0
.
01
0
.
02
0
.
03
0
.
04
0
.
05
0
.
00 0
.
01 0
.
02 0
.
03 0
.
04 0
.
05
Observed soluble P, kg ha
1
S
i
m
u
l
a
t
e
d

s
o
l
u
b
l
e

P
,

k
g

h
a

1
Fig. 10. Comparision of observed and simulated soluble P for
model validation; ( ), regression line; ( ), 1:1 line
M. P. TRIPATHI ET AL. 376
5. Conclusions
The study conrmed that the Soil and Water
Assessment Tool (SWAT) model could accurately
simulate runoff, sediment yield and nutrient losses
particularly from small agricultural watersheds. The
simulated data closely matched with their observed
counterparts in the study. The study also revealed that
all the sub-watersheds of a small agricultural watershed
do not contribute to the discharge, sediment yield and
nutrient losses measured at the outlet. The SWAT model
could identify the critical sub-watersheds, which are
major contributors of these parameters. The SWAT
model can successfully be used for prioritisation of the
critical sub-watersheds in order to develop multi-year
management plan to reduce the runoff, sediment and
nutrient losses from a small agricultural watershed.
Acknowledgements
Authors wish to acknowledge the CSIR, New Delhi,
for providing nancial assistance to conduct this study.
Er. Kamal Misra, Director (Soil Conservation) DVC,
Hazaribagh, Bihar, and Project Co-ordinator, IGBP
Watershed Management, New Delhi, are also ac-
knowledged by the authors for providing the data to
conduct the above study. The facilities and support
provided by the Department of Agricultural and Food
Engineering, IIT, Kharagpur and RRSSC, IIT Campus
are sincerely acknowledged.
References
Arnold J G; Srinivasan R; Muttiah R S; Williams J R (1998).
Large area hydrologic modeling and assessment part I:
model development. Journal of the American Water
Resources Association, 34(1), 7389
Arnold J G; Srinivasan R; Ramanarayanan T S; DiLuzio M
(1999). Water Resources of the Texas Gulf Basin. Water
Science and Technology, 39(3),121133
Arnold J G; Williams J R; Nicks A D; Sammons N B (1990).
SWRRB: A Basin Scale Simulation Model for Soil and
Water Resources Management, pp 125. Texas A & M
University Press, College Station, TX
Arnold J G; Williams J R; Srinivasan R; King K W (1996). Soil
and Water Assessment Tool, Users Manual USDA,
ARTICLE IN PRESS
Table 7
Model output for identication of the critical sub-watersheds of the Nagwan watershed (19961998)
Sub-watershed
Area,
km
2
Runoff,
mm
Sediment
yield, t ha
1
Organic
N, t ha
1
Organic
P, t ha
1
NO
3
-N, t ha
1
Soluble P, t ha
1
WS1 1723 46768 441 500 238 271 027
WS2 929 27478 366 415 199 142 017
WS3 632 42214 439 503 239 229 025
WS4 993 14359 168 196 094 075 009
WS5 1471 28449 700 750 358 115 018
WS6 352 45613 1287 1326 624 204 028
WS7 947 29296 1347 1385 656 119 018
WS8 424 33156 333 384 184 174 020
WS9 298 37398 1463 1486 699 159 023
WS10 722 26227 1280 1323 627 105 017
WS11 478 41665 467 530 253 225 025
WS12 054 25634 1882 1840 879 126 016
Table 6
Statistical analysis of the observed and simulated nutrient losses for the 12 events of the year 1997
Statistics Organic N Phosphorous NO
3
N Soluble P
Obs. Sim. Obs. Sim. Obs. Sim. Obs. Sim.
Mean, kg ha
1
0326 0275 0168 0148 0100 0096 0017 0015
Standard deviation, kg ha
1
0233 0255 0145 0135 0071 0085 0015 0014
Maximum, kg ha
1
0686 0750 0440 0400 0220 0240 0047 0040
Total, kg ha
1
3917 3300 2015 1780 1194 1150 0206 0180
Count 12 12 12 12 12 12 12 12
t-calculated 166 147 039 115
t-critical (two tailed) 220 220 220 220
Coefcient of determination r
2
082 086 089 082
Deviation, % 158 117 37 125
Note: Obs., observed; Sim., simulated.
IDENTIFICATION AND PRIORITISATION OF CRITICAL SUB-WATERSHEDS 377
Agriculture Research Service, Grassland, Soil and Water
Research Laboratory, 808 East Blackland Road Temple,
TX 76502
ASCE Task Committee (1993). Criteria for Evaluation of
Watershed Models. Journal of Irrigation and Drainage
Engineering, 119(3), 429442
Bagnold R A (1977). Bedload transport in natural rivers. Water
Resources Research, 13(2), 303312
Bingner R L (1996). Runoff simulation from Goodwin Creek
watershed using SWAT. Transactions of the ASAE, 39(1),
8590
Bingner R L; Garbrecht J; Arnold J G; Srinivasan R
(1997). Effect of watershed division on simulation of runoff
and ne sediment yield. Transactions of the ASAE, 40(5),
13291335
DelRegno K J; Atkinson S F (1988). Nonpoint pollution and
watershed management: a remote sensing and geographic
information system approach. Lake and Reservoir Manage-
ment, 4(2), 1725
Dhruva Narayana V V (1993). Soil and Water Conservation
Research in India, pp 146151. Indian Council of Agricul-
tural Research, Krishi Anusandhan Bhavan, Pusa, New
Delhi
Dickinson W T; Rudra R P; Wall G J (1990). Targeting
remedial measures to control nonpoint source pollution.
Water Resource Bulletin, AWRA, 26(3), 499507
Dillaha T A (1990). Role of best management practices in
restoring the health of the Chesapeake Bay: 1990-advances
in estuarine sciences. Report No. CBP/TRS4/90, US
Environmental Protection Agency, Chesapeake Bay Pro-
gram/Chesapeake Res. Consortium, Gloucester Point, VA
pp 5781
EPA (1976). Quality Criteria for Water. Environmental
Protection Agency, Washington, DC
Hession C W; Shanholtz V O (1988). A geographic information
system for targeting non-point source agricultural pollution.
Journal of Soil and Water Conservation, 43(3), 264266
Jenson S K; Domingue J O (1988). Extracting topographic
structure from digital elevation data for geographic infor-
mation system analysis. Photogrammetric Engineering and
Remote Sensing, 54(11), 15931600
Jones C A; Cole C V; Sharpley A N; Williams J R (1984). A
simplied soil and plant phosphorus model, I. Documenta-
tion. Soil Science Society of the American Journal, 48(4),
800805
Karale R L; Bali Y P; Narula K K (1977). Priority watershed
for soil conservation works in Matatila catchment. Journal
of Indian Society of Soil Science, 25, 207
Karale R L; Bali Y P; Singh C P (1975). Photo interpretation
for erosion assessment in the Beas catchment. Photo
Nirvachak, 4, 30
Knisel W G (1980). CREAMS: a eld scale model for
chemicals, runoff, and erosion from agricultural manage-
ment systems. Conservation Research Report No. 26,
USDA-SEA, Washington, DC, pp 643
Maas R P; Smolen M D; Dressing S A (1985). Selecting critical
areas for nonpoint source pollution control. Journal of Soil
and Water Conservation, 40(1), 6871
Mannering J V (1981). The Use of Soil Tolerances as Strategy
for Soil Conservation. Soil Conservation Problem and
Prospect, pp 337349. R. P. C. Morgan John Wiley &
Sons, Chichester, England
Martinec J; Rango A (1989). Merits of statistical criteria for
the performance of hydrological models. Water Resources
Bulletin, American Water Resource Association, 25(20),
421432
McElroy A D; Chiu S Y; Nebgen J W, et al. (1976). Loading
functions for assessment of water pollution from nonpoint
sources. Environment Protection Technology Service, EPA
600/276151
McHarg I L (1969). Design with Nature. Natural History
Press, New York, NY
Misra N (1986). Modelling runoff and sediment yield from
small watersheds. PhD Thesis, Agricultural Engineering
Department, Indian Institute of Technology, Kharagpur,
India, pp 5556
Nash J E; Sutcliffe J V (1970). River ow forecasting through
conceptual models, part 1}a discussion of principles.
Journal of Hydrology, 10(3), 282290
PCI Inc. (1994). EASI-PACE Image Analysis System Manual,
Version 53. PCI Incorporated, Richmond Hill, Ontario,
Canada; Vols. I & II, variously paged
Peterson J R; Hamlett J M (1998). Hydrological calibration of
the SWAT model in a watershed containing fragipan soils.
Journal of the American Water Resources Association,
34(3), 531544
Rosenthal W D; Srinivasan R; Arnold J G (1995). Alternative
river management using a linked GIS-hydrology model.
Transactions of the ASAE, 38(3), 783790
Santhi C; Arnold J G; Wliiiams J R; Dugas W A; Srinivasan R;
Hauck L M (2001). Validation of the SWAT model on a
large river basin with point and nonpoint sources. Journal of
American Water Resources Association, 37(5), 11691188
Seligman N G; Keulen H V (1981). PAPRAN-A simulation
model of annual pasture production limited by rainfall and
nitrogen. In: Simulation of Nitrogen Behaviour of Soil-Plant
Systems, (M. J. Frissel; J. A. VanVeen, eds) pp 192221.
Proceeding of the Workshop, Wageningen, The Nether-
lands, JanuaryFebruary 1980
Sharma K D; Menenti M; Huygen J; Vich A (1996). Modeling
spatial sediment delivery in an arid region using thematic
mapper data and GIS. Transactions of the ASAE, 39(2),
551557
Singh G; Ram Babu; Pratap Narain; Bhushan L S; Abrol I P
(1992). Soil erosion rate in India. Journal of Soil and Water
Conservation, 47(1), 9799
Srinivasan R; Arnold J G (1994). Integration of a basin scale
water quality model with GIS. Water Resources Bulletin,
AWRA, 30(3), 453462
Srinivasan R; Arnold J G; Rosenthal W; Muttiah R S (1993).
Hydrologic modelling of Texas Gulf Basin using GIS. In:
Proceedings of Second International GIS and Environmen-
tal Modelling, Breckinridge, CO pp 213217
Storm D E; Dillaha T A; Mostaghimi S; Shanholtz V O (1988).
Modeling phosphorous transport in surface runoff. Trans-
actions of the ASAE, 31, 117127
Tim U S; Moostaghimi S; Shanholtz V O (1992). Identication
of critical nonpoint pollution source areas using geographic
information systems and water quality modeling. Water
Resources Bulletin, AWRA, 28(5), 877887
Tiwari K N; Kanan N; Singh R; Ghosh S K (1997). Watershed
parameters extraction using GIS and remote sensing for
hydrologic modelling. ASIAN-PACIFIC Remote Sensing
and GIS Journal, 10(1), 4352
Tripathi M P (1999). Hydrological modelling for effective
management of a small watershed. PhD Thesis, Agricultural
and Food Engineering Department, Indian Institute of
Technology, Kharagpur, India, variously paged
ARTICLE IN PRESS
M. P. TRIPATHI ET AL. 378
Tripathi M P; Panda R K; Raghuwanshi N S (1999a). Runoff
estimation from a small watershed using SWAT model.
Hydrological Modelling, Proceeding of International Con-
ference on Water, Environment, Ecology, Socio-Economics
and Health Engineering, held at Seoul, Korea from October
18 to 21, 1999, pp 143152
Tripathi M P; Panda R K; Raghuwanshi N S (1999b).
Estimation of Sediment Yield from a small Water-
shed Using SWAT Model. In: Proceeding of Civil
and Environmental Engineering Conference New
Frontiers and Challenges, held at Asian Institute of
Technology, Bangkok, Thailand from November 8 to 12,
pp I-87I-96
USDA-SCS (1972). National Engineering Handbook, Hydrol-
ogy Section 4, Chapters 410
Vieux B E (1991). Geographic information system and
nonpoint source water quality and quantity modelling.
Hydrological Processes, 5(1), 101113
Wang M; Hjelmfelt A T (1998). DEM based overland ow
routing. Journal of Hydrologic Engineering, 3(1), 18
Williams J R (1980). SPNM, A model for predicting sediment,
phosphorous and nitrogen yields from agricultural basins.
Water Resources Bulletin, AWRA, 16(5), 843848
Williams, J R; Berndt H D (1977). Sediment yield prediction
based on watershed hydrology, Transactions of the ASAE,
20(6), 11001104
Williams J R; Hann R W (1973). HYMO: Problem-oriented
Language for Hydrologic Modelling}Users Manual.
USDA, ARS-S-9
Williams J R; Hann R W (1978). Optimal operation of large
agricultural watersheds with water quality constraints.
Texas Water Resource Institute, Texas A & M University,
Technical Report No. 96
Williams J R; Jones C A; Dyke P T (1984). A modeling
approach to determining the relationship between erosion and
soil productivity. Transactions of the ASAE, 27(1), 129144
ARTICLE IN PRESS
IDENTIFICATION AND PRIORITISATION OF CRITICAL SUB-WATERSHEDS 379