Assessing the impacts of land use and land cover change on stream flows of river Kyambura

Abstract

The impacts of land-use land cover changes on hydrological systems has received considerable attention from researchers and practitioners world over. At catchment level, land-cover changes and their effects on the hydrologic processes remain a major challenge. River Kyambura catchment located in the western part of Uganda has experienced noticeable land use and land cover changes but their impact on hydrological processes are not yet well understood. This study aimed at assessing the impacts of land use and land cover changes on the streamflows of River Kyambura using a modelling approach. A Soil Water Assessment Tool (SWAT) model of River Kyambura was set up, calibrated and validated using daily streamflow data obtained from the Directorate of Water Resources Management for the Kyambura gauging station located at lat. 00011'22"S, lon. 030006'11"E. Coefficient of Determination (R2) was used as the objective function as a measure of model performance during calibration and validation. The model calibration and validation processes gave R2 values of 0.78 and 0.688 respectively. ArcSWAT was also used to classify the land use-cover maps from Landsat satellite images acquired for the year 1989, 2009 and 2019 respectively using the Maximum Likelihood Algorithm for Supervised Classification. The Kappa coefficient objective was used to determine the classification accuracy for the different landuse images. Coefficients of 0.86, 0.96 and 0.96 were obtained for the 1989, 2009 and 2019 imagery respectively. The land use maps of 1989, 2009 and 2019 were analysed for land cover changes. The results showed that more significant land cover changes occurred after 2009 with an increase in vegetation cover of 24.53% and a decrease in barren land of 27.69%. Built up areas had a 3.50% increment whereas open water decreased by 0.33%. The effects of land-use changes on stream flow were assessed by using the calibrated SWAT model. The model results showed a reduction in water yield for both the first (1989–2009) and second (2009–2019) periods. These periods were marked with an increase in vegetation cover. The water yield was also found to be directly proportional to precipitation, surface runoff, lateral flow, groundwater flow, percolation and evapotranspiration in the period of 1989 to 2019. The results obtained in this study can provide useful information for water resource planning and management as well as soil and water conservation in the Kyambura catchment.