The effects of land use change on hydrological responses in the Choke Mountain Range (Ethiopia) - A new approach addressing land use dynamics in the model SWAT

Land use change (LUC) is a very important issue considering global dynamics and their responses to environmental and socio-economic drivers. Especially in fast changing developing countries, it is a scientific challenge to predict land use changes and their effects on water availability, flood risk and erosion rates. To address these issues, catchment models must be able to deal with land use dynamics. Unfortunately, many models handle land use in a static state way. The objective of this research is to investigate the effect of dynamic land use implementation into SWAT ("Soil and Water Assessment Tool") by developing the supportive tool "Land use Update and Soil Assessment" (LUPSA) and to improve the overall SWAT abilities to handle LUC. A catchment in the Choke Mountain Range (Ethiopia) was selected as test case where significant land use change occurred during the last decades. These dynamics were addressed by 5 land use maps based on interpretation of aerial photographs and satellite imageries, while several years of climate and discharge data were available. The available data for the test application were carefully analyzed and found to be limited in quantity and quality. LUPSA was applied to feed the SWAT model for the period of 1973 to 2003 with yearly land use updates. The annual LUC varied between -6% and +360% for different classes. The impact of land dynamics on the hydrological response was observed and shown at the daily discharge, the total annual runoff and the peakflow.. Also a higher proportion of low flow rates was found and caused more water stress. Considering the high uncertainties, SWAT was not able to produce reliable results due to the bad data quality. Nevertheless, the implementation of land cover dynamics in SWAT led to a significant change in the model outputs and demonstrated improved capabilities to handle their impacts on water resources. Further model testing is strongly recommended.