Modeling the impacts of land use and land cover changes on streamflow and sediment yield with SWAT model in Migori River watershed, Lake Victoria basin region, Kenya

The Soil and Water Assessment Tool (SWAT) model was used to simulate the impacts of LULC changes on streamflow and sediment yields of Migori River watershed. The Nash - Sutcliffe Efficiency (NSE), co-efficient of determination (R2), and percent bias (PBIAS) of the monthly streamflow and sediment outflow simulation for the calibration and validation showed good model performance. From 1980 to 2020, the results show a decrease of 40.63% in shrublands, 84.86% in grasslands, and 52.90% in forests. Additionally, there was an 82.03% reduction in water areas and a 38.44% decline in wetlands. In contrast, cultivated land, bare land, and built-up areas all experienced growth during this same time frame, expanding by 34.25%, 132.28%, and 461.20%, respectively. These LULC changes have contributed to the decline of annual actual evapotranspiration by 2.83%, annual potential evapotranspiration by 5.38%, and annual groundwater flow by 19.06%; whereas annual surface runoff, water yield, and sediment yields have increased over the same period by 32.57%, 4.82%, and 84.58% respectively. The negative impacts of LULC changes between 1980 and 2020 on the hydro-sedimentological processes are the main cause of flooding problems within the watershed. The study recommends addressing the sedimentation and flooding with comprehensive soil and water conservation interventions.