Testing and Evaluation of the Agricultural Non-point Source Pollution Model (AGNPS) on Mirera Sub-catchment of Lake Naivasha Drainage Basin, Kenya

Abstract

Sedimentation is a major threat to surface water bodies such as lakes and rivers where the eroded soils from the hill slopes are finally deposited. The extent of erosion depends on the magnitude of the surface runoff, which transports the eroded sediments. Increased sediment concentration in surface water bodies can be detrimental to aquatic systems by; reducing light penetration, clogging aquatic vegetation and spawning gravels, and reducing water quality. Suspended sediments act as a media of transport for nutrients thereby accelerating the rate of nutrient accumulation in surface water bodies. Due to lack of data, it is necessary that other techniques such as modeling be used to analyze the dynamics of surface runoff, sediment and nutrient simultaneously. The Agricultural Non-Point Source (AGNPS) pollution model was found to be suitable to analyze these three variables together hence it was applied in Mirera sub-catchment located in Lake Naivasha drainage basin, Kenya. Mirera sub-catchment of 1.82 kmz drains into a dam created through road construction works in 1998 and had accumulated sediments over the same period. Through use of GIS and remote sensing, it was possible to extract the biophysical parameters from the catchment for input into the AGNPS model. Grid cells of 50 m by 50 m were used and simulation done on cell basis using GIS ILWIS which gave the spatial distribution of surface runoff volume, sediment yield and phosphorus yield in form of thematic maps. The results showed that both runoff volumes and sediment yield were reasonably simulated with efficiencies of 0.88 and 0.92 respectively while the phosphorus simulations did not match well with the observed values at the catchment outlet. The results indicate the high potential of AGNPS model for analyzing the dynamics and constituents of flow required in catchment management. The model however requires fimher verification and testing especially with regard to simulation of nutrients.