ASSESSMENT OF WATER BALANCES USING SWAT MODEL AND AGRICULTURAL WATER DEMAND ANALYSIS OF GIDABO RIVER CATCHMENT, RIFT VALLEY BASIN, ETHIOPIA

Abstract

To use water resources sustainably, it is important to understand the quantity of water resource spatially and temporally. The work presented here attempts to model water balance of one of Gidabo sub-watershed of Ethiopian Rift Valley Basin using the Soil and Water Assessment Tool (SWAT. Statistical model performance measures, the coefficient of determination (R2 ) and Nash-Sutcliffe Efficiency (NSE) were used to evaluate the correlation between the observed and simulated monthly stream flow. The result shows an acceptable performance of SWAT model in simulating the watershed hydrology as evidenced by the NSE value of 0.79 and R2 value of 0.80 at calibration (using data from1990-2009) and NSE = 0.71 and R2 =0.82 at validation (using data from2010- 2014). In terms of water balance components, the long-term annual precipitation is found to be in the magnitude of 1435.58 mm; whereas the surface runoff; total water yield; and actual evapotranspiration, are estimated as 62.20; 1189.73, and 228.40 mm respectively. In order to link the water balance estimation with the local agricultural water demand, wheat and potato are selected for they are common crops for irrigation in the watershed. While taking the month of March, which is the month of maximum agricultural water demand (11.6 m 3 /s), the watershed is found to be water-surplus as indicated by the water availability of the critical month (12.22 m3 /s) to be higher than the demand. This research implies that the use of models in supporting decisions related to water resources development and management is paramount important to understand how the system functions