Integration of GIS, Remote Sensing and USLE to Estimate Soil Erosion Potential in Nzoia River Basin (Kenya).

The study describes the application of the universal soil loss equation model, to quantify soil erosion in Nzoia River Basin using the geographic information service and remote sensing. The Nzoia River Basin is located entirely on the Kenyan side of Lake Victoria Basin. The approach adopted involved calculation of six universal soil loss equation factors as distributed remote sensing and geographic information system data layers. The data included spatial raster layers of soil, land cover, rainfall and digital elevation models ranging from 30 m to 1000 m spatial resolutions to adequately represent the surface characteristics. The soil erosion distribution map was generated as a product of six raster layers. Results from the erosion estimation showed on average, potential erosion highest in the cropland and woodland mosaic (180 Mg ha-1 yr-1) and lower in deciduous broadleaf forest (23 Mg ha-1 yr-1) and shrubland (22.5 Mg ha-1 yr-1). About 84.2 % (10831.8 Km2) of the Nzoia River basin is under slight (0 – 5 Mg ha-1 yr-1) erosion potential while about 2.1 % (272.9 Km2) and 2.8 % (358.2 Km2) of Nzoia river basin is under severe (> 80 Mg ha-1 yr-1) and very high (20 – 80 Mg ha-1 yr-1) erosion potential respectively. Erosion sources such as unpaved and poorly graded roads plus point sources should be investigated to quantify the contribution of agricultural pollution to 2.5 x106 Mg yr-1 total suspended solids discharged by Nzoia River into Lake Victoria.