Integrating Satellite based Remote Sensing Observations with SCS Curve Number Method for Simplified Hydrologic Modeling in Ungauged Basins

Accurate estimation of spatial and temporal distribution of the water resources parameters is a crucial for hydrologic models. During the last decades, engineers and planners have shown keen interest in applying Geographic Information Systems (GIS) and satellite based Remote Sensing (RS) technologies to extract land surface parameters, which acted as a threshold in early days, to obtain reasonable results in hydrologic modeling especially in ungauged basins. With the advancement of computer technology, GIS and RS have become efficient ways to integrate spatial and non-spatial databases for hydrological modeling. In this study, we present a hydrological model for the Bata River basin, a tributary of the Yamuna River, India developed at a 20 m grid resolution. Infiltration losses, basin hydrograph, and river routing are the main model components of the proposed hydrologic model. Satellite based RS and GIS techniques are used to estimate the runoff potential at each 20 m grid resolution. IRS LISS 111 multi-temporal satellite data in the cultivation seasons in Rabi and Kharif, IRS pan data, and field data are integrated with SCS Curve Number method to obtain the infiltration losses at each 20 m grid scale; SCS dimensionless hydrograph method is used to generate the basin hydrographs; the Muskingum hydrological routing method is used for river routing. The proposed model is applied for the Bata River basin for a two year period. Both lumped and distributed modeling approaches are considered. For the distributed modeling approach, the complete watershed is divided into 10 subareas, where hydrographs for the total basin is obtained by routing the subarea hydrographs along the river reaches. We demonstrate that the total release of water as runoff greatly depends on the temporal arrangement of rainfall events, their magnitudes, as well as the cultivation season. Further, the distributed hydrologic model attenuates the peak flood by 50% as compared to the lumped hydrologic models, whereas lumped model runoff simulation showed approximately ±25% deviation in runoff as compared to distributed model approach. Finally, the developed model is shown to be capable of simulating the runoff for a given rainfall event and deriving hydrographs for the required duration especially for ungauged basins in data poor-regions in the world.