Surface runoff prediction on Kenyan catchments using the Green-Ampt model and related soil parameters

Abstract

The need for more agricultural areas stemming from population pressure has been contributing to the increased forest and grassland-to-agricultural areas conversion in recent years in the upper Mara River Basin, Kenya. The deforestation of the Mau forest and settlement on hydrological sensitive areas of the watershed along with poor soil and water conservation practices have caused the high level of soil erosion and water quality deterioration. The various efforts to curve down this problem necessitate the knowledge of the rate of erosion, surface runoff as a function of the physical and hydrometeorological characteristics of the watershed. In support of this effort, a study was conducted to evaluate the effects of three land-use/land cover practices on runoff and sediment yield using runoff plots and two models: Water Erosion Prediction Project (WEPP) and Erosion 3D. In order to accomplish these objectives, runoff plots were established at three sites: Amala upstream; Amala downstream and Nyangores watersheds. In all of the three watersheds, three land uses (cropland, grassland and bare land) with similar slope steepness and soil types were considered. Using the data collected from the runoff plots, the WEPP and Erosion 3D models were evaluated. The plot scale research results indicated that erosion was high in all of the sites on cultivated lands. The highest total sediment yield over the study period was observed on cultivated land (162.38 g/m2) in the Nyangores watershed; and the lowest sediment yield (29.95 g/m2) was observed on grassland at Amala downstream site. The model evaluation indicated that both of the models perform well in estimation of runoff, however, the WEPP model performs better than Erosion 3D in estimation of soil erosion