Diffusion Wave Based Modeling of Watershed for Soil Erosion and Sediment Yield Using FEM, GIS, and Remotely Sensed Data

This paper describes the simulation of soil erosion and sediment yield from agricultural watershed using remotely sensed data and data base management tools like Geographical Information System (GIS). Finite Element Method (FEM) technique has been used for solving the governing partial differential equations. The GIS has been used to generate the finite element grid map, slope, DEM, Landuse (LU)/Land cover (LC) and Soil map. The diffusion wave based soil erosion and sediment yield model for overland flow and channel flow has been developed and applied to the Khadakohol and Harsul watersheds located in Nashik district of Maharastra state, India. Soil map is prepared based on ground truth information and the major soil type is sandy silt loam. Scanned topographical maps of 1:25,000 scale are registered in ERDAS IMAGINE version 9.1 image processing software. Watershed boundary, drainage and contours are digitized in ArcMap. DEM has been prepared based on contours and slope map has been generated from the DEM by using the slope option of Spatial Analyst in ArcMap. Remotely sensed data of IRS 1D LISS III with resolution of 23.5m has been used to extract LU/LC map of Khadakohol and Harsul watersheds. Input files of slope, Manning's roughness, Crop management factor and Erodibility factor have been prepared. Part of the available data has been used for model calibration and the other part for model validation. The developed model has predicted satisfactory results for both calibration as well as validation events, when compared with the observed data.