Modelling Runoff, Soil Erosion And Sediment Yield In Sosiani Catchment In Kenya Using Arcswat

Abstract

Sosiani catchment hosts Eldoret Town, a major agricultural, education and medical center in the Rift valley region of Kenya. Recent research has established that the quality of water in Sosiani River is rapidly deteriorating because of catchment erosion and wastes from industries. However, even though there have been efforts to control the pollution, non-point sources of pollution have not been characterized. The aim of this study was to simulate effect of rainfall magnitude on soil erosion, runoff and sediment yield in Sosiani catchment using ArcSWAT. The soil dataset used was obtained from Digital World Soil Map, weather datasets were obtained from Climate Forecast System Reanalysis (CFSR), DEM dataset was obtained from Open Topography, landuse/land cover dataset was obtained from USGS and river flow dataset was obtained from Water Resources Authority Kenya. The period of study was 1984-2012. Flow Calibration period was 1986-1990 while the validation period was 2008-2012. Sediment and soil erosion were not calibrated for lack of observed data. Flow calibration and validation results returned an R2 value of 0.90 and 0.60 and NS value of -19.92 and -14.90. The study observed that increase in rainfall magnitude increased runoff, soil erosion and sediment yield except under conditions of increasing vegetative land cover. The study concludes that the model did not simulate the catchment correctly because it oversimulated both low and high flows. It recommends that datasets for the catchment (observed rainfall and river flows) need to be thoroughly checked for integrity before attempting to use them for hydrologic modelling. It further notes that agricultural land is the main contributor of sediment into river Sosiani and therefore remedial measures such as agroforestry and cover cropping would be appropriate immediate actions to control soil erosion.