RAINFALL-RUNOFF MODELING AND ITS PRIORITIZATION AT SUB-WATERSHED LEVEL USING SWAT MODEL: A CASE OF FINCA’AA, OROMIA, WESTERN ETHIOPIA

Abstract

Rainfall-runoff modeling is essential for soil and water conservation and for watershed management. The Soil and Water Assessment Tool, SWAT, model was used to predict average annual stream flow and identify vulnerable areas in the Finca’aa watershed at the sub-basin level. Meteorological, land use land cover, LULC, soil, 30 m resolution digital elevation model (DEM), and hydrological data were used as inputs for the model. There were 21 sub-basins and 205 HRUs in the simulated watershed. The sequential uncertainty fitting (SUFI-2) technique in the SWAT- calibration uncertainty program package was used for the sensitivity analysis, the calibration, and the validation of the model. The base flow alpha factor (ALPHA BF) and SCS runoff curve number (CN2) were the most sensitive stream flow parameters. The coefficient of determination, R2, and Nash-Sutcliffe, ENS, values for the calibration and the validation were 0.81 and 0.79, and 0.76 and 0.74, respectively. These values indicate the good fit between the observed and simulated stream flow data. The surface runoff during the rainy, the intermediate, and the dry seasons were 328.25, 77.525, and 39.31 mm, respectively. The average and maximum annual simulated surface runoff values were 449.81 and 567.4 mm, respectively. Agriculture was the main factor that caused high surface runoff in the watershed. Sub-basins 4 and 5 were runoff prone areas. The findings of the study call for the implementation of appropriate water management practices to conserve the soil and water resources of the watershed.