ASSESSMENT OF LAND USE / LAND COVER CHANGE IMPACT ON STREAM FLOW USING SWAT MODEL, THE CASE OF KULFO RIVER CATCHMENT, RIFT VALLEY BASIN, ETHIOPIA

Abstract

In order to conserve land and water resources and ensure its long-term sustainability, land use land cover change dynamics and its impact on the water resources of watersheds have emerged as key issues in hydrology. The main objective of this study was to investigate impact of land use land cover change on stream flow of Kulfo River. ERDAS IMAGINE-2015 using the maximum likelihood algorithm of supervised classification was used to evaluate changes in land use and land cover of study area. Land sat-5 images of year 2000 and 2010 as well as land sat-8 images of year 2022 obtained from USGS with spatial resolution of 30m were used for land use land cover classification. Accuracy of model to classify land features accurately and precisely was evaluated by making comparison matrix between classified classes with those collected by GPS and from Google earth. Kappa coefficient from comparison matrix has been used to evaluate performance of model and it has been determined over 80% for all classification periods verifying the accuracy of classification. From comparison observed from each periods of land use land cover maps, there are some significant changes on land use and land cover features of study area. Over the past study years, the watershed has undergone significant changes, including the expansion of agricultural land and towns at the expense of the reduction of grassland, forest, and shrub area. Those LULC maps of different periods were varied as SWAT inputs while maintaining the remaining inputs the same in order to evaluate its impact on stream flow using SWAT model. The sensitivity analysis, calibration and validation was conducted using Sequential Uncertainty Fitting (SUFI-2) within SWAT-CUP. The calibration results showed very good match of simulation with observation with Nash-Sutcliff efficiency (NSE) of 0.81, coefficient of determination (R2 ) of 0.83, root mean square error (RMSE) of 0.12 and percent of bias (PBIAS) of -0.1.Perspective objective functions during validation have been determined as 0.84,0.86,0.01 and 0.03 respectively. Well calibrated stream flow for different periods has been compared to evaluate impact of LULC changes on stream flow. It has been pointed out that some increases in agricultural land and urbanization promoted runoff formation by reducing water infiltration into deep soil, result in an increase in surface runoff and a decrease in ground water flow. It is also revealed by the study that increased surface due to land use land cover change is more significant during wet seasons than that of dry seasons. The study's methodology and findings will be pertinent to environmental policy makers and other relevant bodies which will involve in the development, occupancy, and management of resources related to water and land