GIS-BASED SURFACE IRRIGATION POTENTIAL ASSESSMENT: A CASE STUDY IN BILATE RIVER WATERSHED, SOUTH-WESTERN, ETHIOPIA

Abstract

The major problem associated with rainfall-dependent agriculture in Ethiopia is the high degree of rainfall variability and unreliability. As a consequence, food insecurity often turns into famine. Irrigation development is one of the key strategies to increase agricultural production and alleviate poverty. This study, therefore, aimed to assess the surface irrigation potential and land resources potential of the Bilate River Watershed for irrigation expansion. Watershed delineation, identification of potentially irrigable land, and estimation of irrigation water requirement and surface water resources availability of the study area were the steps followed to evaluate this irrigation potential. Irrigation potential was mapped by using GIS of the watershed; Arc SWAT model was used to estimate the water resources availability in the watershed; Analytical hierarchy process (AHP) comparison was used to conduct land suitability assessment and a CROPWAT 8.0 model was used to determine the crop water requirement for major crops of the study area. To identify potentially irrigable land, irrigation suitability factors such as soil physical properties, slope, land use/ land cover, and distances from the water supply (sources) are taken into account. The final results of the overall weighted overlay analysis of irrigation land suitability indicate that 317,841ha (64%) are highly suitable (S1), 153,459ha (30.91%) are moderately suitable (S2), whereas 25,039ha (5.044%) are not suitable (N) for surface irrigation development. The SWAT model was calibrated and validated. The observed monthly streamflow values have a coefficient of determination (R2 ) and Nash-Sutcliffe Coefficient (NSE) of 0.77 and 0.66 respectively for the calibration period and 0.81 and 0.64 for the validation period. The irrigation water demand required by five major selected crops which are grown in the study area throughout the growing season was found to be 663.04 m 3 /s. The annual dependably simulated streamflow was 225.14 m 3 /s. The estimated dependably available flow can potentially irrigate only an area of 53,645.77 ha for highly suitable areas. The result showed that the water demand of the crops was greater than the available dependably flow of the watershed. Therefore to increase the irrigation potential of the watershed sprinkler or drip irrigation methods can be used