Analysis of catchment hydrologic response under changing land use The case of upper Molo river catchment, Kenya

Abstract

Change in land use has a direct effect in catchment hydrologic response. It is caused by human intervention to enhance and diversify their livelihood needs, and at the same time get economic benefits from the land resources. These interventions result in changes in surface runoff, soil erosion and sediment yield among others. If the change in land use is not well managed then it will affect the quantity and quality of water resources as well as production potential of the land. Based on this ground this study was formulated to investigate the effects of land use changes on catchment response, in particular surface runoff and sediment yield. Such a study required continuous hydrologic data such as stream flow and sediment yield for a number of gauging stations within the study catchment. However, most catchments in Kenya do not have adequate data to accomplish such study. In this study upper Molo River catchment in eastern Mau was used because of its consistent stream flow data. In this catchment there has been significant reduction in stream flow during dry season and flooding in the rainy season. This study investigated a modelling approach for predicting the changes in catchment response as a result of land use change. Soil and Water Assessment Tool (SWAT) was identified as suitable model and used to simulate the catchment response under different land use types. The input data used were digital elevation model (DEM), land cover, soils and rainfall. The DEM was processed in Arc View GIS and land cover maps derived from satellite image using ERDAS 8.5 imagine software. The land cover analysis results show that forest cover reduced by 48% as a result of increase in agricultural and settlement areas between the years 1986 to 2001. Simulation analysis carried out for 1986 and 1995 land cover maps, show an increase in surface runoff of 13.3%. In the simulation the data set was divided into two; 1980 to 1989 for calibration and 1991 to 2000 for validation. Conceptual parameters derived during calibration were used in the model to simulate streamflow for the two data sets and gave a Nash Sutcliffe coefficients of 0.87 and 0.72 respectively. The sediment yield values were 1.5t/ha for the calibration and 2.7t/ha for validation periods respectively. These results show insignificant change in the catchment response but demonstrated the effects of land use changes on catchment response. It is therefore concluded that land cover change of less 48% have insignificant change on catchment hydrologic response.