269-Influence of Digital Elevation Model Derived from Radar Remote Sensing on Prediction of Runoff, Soil Erosion and Deposition at Catchment Scale.

Abstract

Distributed soil erosion models integrate spatial variability of catchment characteristics into input parameters and processes to improve its predictive outputs for each storm event. Spatial variability obtained at a grid-cell size from various resolutions and sources of satellite remotely sensed data, influences generation and routing of surface runoff and sediment in a catchment. In the study, a GIS based dynamic LISEM (Limburg Soil Erosion Model) was parameterised at 20, 40, 60, 80 and 100 m grid-cell sizes from an InSAR DEM, a reference Cartometric DEM, a 20 m land use and land cover class map and 1: 250,000 scale National Soil Database along with field observations leading to creation of ten LISEM geospatial databases of 30 parameters each which were tested for a 6-hour storm at three soil moisture levels to investigate the influence of grid-cell size and source of DEMs on surface runoff and sediment movement as well as to identify sediment source areas in a catchment. The results showed that at increasing the grid-cell size from 20 to 100 m of a DEM, the slope gradient flattened by 28% and the drainage length shortened by 79%, which had competing effect on the runoff and sediment flow routing. The prediction of surface runoff and soil loss was improved significantly for the InSAR DEM as compared to the Cartometric DEM suggesting its suitability for distributed modelling. Grid-cell size of 20 m pin-points sediment source areas for effective implementation of conservation measures.