IMPACT OF LAND USE LAND COVER CHANGE ON HYDROLOGY OF UPPER AKAGERA CATCHMENT, SOUTH EASTERN RWANDA

Abstract

Water is valued resource to be beneficially managed and management of water resources most often deals with complex systems composed of many interconnected parts. One of the challenges faced in contemporary water resources management is land use and land cover changes hence, modeling the hydrology of watersheds is required for effective watershed management strategy. This study was conducted to assess land use/cover change of the catchment, land use maps of the study area were derived from satellite images processed by GIS and ERDAS software for the year 1987, 1997 and 2007. SWAT model was used to simulate stream flow, to determine the impact of LuLc change on this catchment stream flow. The LuLc change analysis has shown that there is no significant change for land use such as water body, bare land and forest land. For bush land, it shows a significant decreasing trend 39.58%, 31.92% and 24.25% for 1987, 1997 and 2007 years respectively. For cultivated land there is an increasing trend , 10.77%,15.60% and 20.43.The same is true for settlement showing 6.36%,9.10% and 11.85% for 1987,1997 and 2007 year. Sensitivity analysis has shown that Gwqmn, Esco, Cn2, Sol_Awc, Ch_K2, Revapmn are high sensitive parameters. The models were calibrated from 1980 to 1999 and validated from2000 to 2008. The model performance criteria for calibration showed that R 2 was found to be 0.742, Nash-Sutcliffe Coeff. 0.76. For the validation period R 2 was found to be 0.793, Nash-Sutcliffe Coeff. 0.87, which show good performance of the model. The mean annual surface flow for 2007 land cover was increased by 28% than that of 1987 land cover. Similarly the 1997 land cover mean annual surface flow was increased by 9% than that 1987 land cover. These land use changes did result in significant changes in stream flow. This emphasizes the need for assessment of the impact of land use/land cover on the catchment system and hydrological parameter. Therefore, it is highly compulsory to launch a good system of both hydrological and meteorological information systems.