Runoff Modeling in an Agro-Forested Watershed Using Remote Sensing and GIS

A hybrid technique was used for the runoff production and its routing in an agro-forested watershed located within the Kanha National Park in Central India with the use of remote sensing and geographic information system (GIS) data. In this technique, a modified Soil Conservation Service curve number (SCS-CN) method and a two-dimensional overland flow model were combined. Modified SCS-CN method estimated daily net rainfall fractions were used as an input to the overland flow model along with other remote-sensing-derived inputs such as the digital elevation model (DEM), rainfall, and roughness factor for routing of the produced runoff. The model works on a cell basis and routs produced runoff from one cell to next following the maximum downslope directions. The flow model uses the diffusive wave approximations of the St. Venant equations for routing surface water. The model was tested by calibrating the Strickler coefficient K, which is inversely proportional to resistance to flow, and comparing the observed and simulated daily change in the water levels for two gauging sites. The calibrated average values of K for different subcatchments were 15.7, 21.7, 23.4, and 28.4 for Kurkuti, Sijhora, between the gauging sites, and downstream catchments, respectively. The model was tested for some statistical parameters like the Nash Sutcliffe coefficientand RMSE using residuals between observed and simulated data, and found to be within the acceptable limits. The results show that the hybrid technique works well to extend the application of curve number to address the routing phase of runoff. DOI: 10.1061/(ASCE)HE.1943-5584.0000466. (C) 2012 American Society of Civil Engineers.