B8, 2012 
1 Image Date 
| Sciences. 
08). Vegetation 
ion of Western 
thematical and 
Australian soil 
RO Publishing. 
ment: An Earth 
arson Prentice 
Evaluating the 
1tcrop wetlands 
t region: Stage 
vironment and 
)5). "Detecting 
thern Germany 
al for Nature 
J. & Harris, D. 
ver the State of 
roceedings 9th 
try Conference. 
dione, J. A. & 
om high-spatial 
t Valley Fever 
onment, 106(1): 
sing and Image 
ac. 
Data Online." 
| technique for 
Landsat data." 
197-2203. 
yrphic approach 
egetatio, (118): 
it surface water 
e composites." 
)46. 
Monitoring the 
sets on multi- 
nment, 115(9): 
International Archives of the Photogrammetry, Remote Sensing and Spatial Information Sciences, Volume XXXIX-B8, 2012 
XXII ISPRS Congress, 25 August — 01 September 2012, Melbourne, Australia 
Runoff simulation using distributed hydrological modeling approach, remote sensing and GIS 
techniques: À case study from an Indian agricultural watershed. 
V.M.Chowdary"", V.R.Desai?, Mukesh Gupta?, A Jeyaram!, Y.V.N.K.Murthy? 
'Regional remote Sensing Centre (East), NRSC, ISRO Kolkata, India -700156 
Department of Civil Engineering, IIT, Kharagpur, India- 721302 
*DDRC, National remote Sensing Centre, Hyderabad, India 
*Corresponding author (chowdary_isro@yahoo.com; muthayya.chowdary@gmail.com) 
Working Group, Theme: VIII/4:Water 
Keywords: Distributed hydrological model, GIS, Remote sensing, Curve number, Muskingum-Cunge technique 
ABSTRACT: 
Distributed hydrological modeling has the capability of simulating distributed watershed basin 
processes, by dividing a heterogeneous and complex land surface divided into computational 
elements such as Hydrologic Response Units (HRU), grid cell or sub watersheds. The present study 
was taken up to simulate spatial hydrological processes from a case study area of Kansavati 
watershed in Purulia district of West Bengal, India having diverse geographical features using 
distributed hydrological modelling approach. In the present study, overland flow in terms of direct 
runoff from storm rainfall was computed using USDA Soil Conservation Services (SCS) curve 
number technique and subsequently it served as input to channel routing model. For channel flow 
routing, Muskingum-Cunge flood routing technique was used, specifically to route surface runoff 
from the different sub watershed outlet points to the outlet point of the watershed. Model 
parameters were derived for each grid cell either from remote sensing data or conventional maps 
under GIS environment. For distributed approach, validation show reasonable fit between the 
simulated and measured data and CMR value in all the cases is negative and ranges from -0.1 to - 
0.3. Further, this study investigates the effect of cell size on runoff simulation for different grid cell 
sizes of 23, 46, 92, 184, 368, 736, 1472 m resolution. The difference between simulated and 
observed runoff values increases with the increase of grid size beyond 184 m more prominently. 
Further, this model can be used to evaluate futuristic water availability scenarios for an agricultural 
watershed in eastern India. 
1. INTRODUCTION 
Hydrological behaviour of any watershed can be predicted 
through modeling. Particularly, the conversion of excess rainfall 
into surface runoff is traditionally analyzed by means of lumped 
models and these models assume that excess rainfall and 
physiographical conditions over a watershed are uniform. Thus 
errors can be introduced while simulating rainfall-runoff 
process using lumped models. Hence to overcome this 
difficulty, distributed modelling approach was proposed (Diksi 
et al., 1984). This approach has the capability of simulating the 
heterogeneity of both rainfall spatial distribution and catchment 
characteristics, may offer a better approach for runoff 
hydrograph simulation. Recent developments in the remote 
sensing technology and geographical information systems make 
It possible to capture and manage a vast amount of data of 
spatially distributed hydrological parameters and variables. 
Linking GIS and the hydrological modeling is very essential to 
achieve the desired objectives. As distributed models are more 
widely use in practice, the need of scientific principal relating to 
spatial variability, temporal and spatial resolution, information 
content and calibration become more apparent. 
One of the most widely used techniques for estimating direct 
runoff depths from storm rainfall is the United States 
Department of Agriculture (USDA) Curve Number (CN) 
method (SCS 1972). Greene and Cruise (1995) and Ponce and 
Hawkins (1996) identified the CN method as one of the most 
popular tools for calculating runoff depths. The description of 
the flow process in the numerous distributed rainfall-runoff 
models may be classified into two basic kinds (Beven, 1985). 
One is the kinematic wave approach for simulating the overland 
and channel flow (Abbott et al., 1986; Morris, 1980). The other 
is the conceptual storage approach (Diskin et al., 1984; Beven et 
al, 1984). Yu (1990) has a detailed literature review of the 
distributed rainfall runoff models. The conventional