Analysis of the changes in landcover, streamflow, and sediment yield in Kasai catchment of the Congo River Basin

The Kasai catchment is experiencing poor landuse planning and sediment deposition, thus impacting streamflow. The landuse and landcover change (LULCC) from 2000 to 2020 were estimated using GIS and RS methods.  The results revealed that bare areas, built-up areas, cropland, flooded vegetation, grassland, open water, and shrub cover areas increased by 0.06%, 0.5%, 0.8%, 16.9%, 1.9%, and 78.6%, respectively, whereas tree cover areas decreased by 0.5%. The observed flows and sediments were used to calibrate the SWAT plus. The performance indices for SWAT plus in flow simulation indicated that R², NSE, and PBIAS were 0.94, 0.87, and 0.55, respectively. Validation results showed R², NSE, and PBIAS values of 0.83, 0.67, and 2.6, respectively. Performance indices for SWAT plus in sediment simulation revealed that R², NSE, and PBIAS were correspondingly 0.81, 0.66, and 7.1. Validation results were 0.8 (R²), 0.66 (NSE), and -1.3 (PBIAS). QSWAT plus post-processing was utilised to estimate the monthly environmental flow. The results ranged from 83.04 m³/s to 10079.97 m³/s. The impact LULCC on the streamflow and sediment yield was assessed by running the model with different landcover maps while other variables were maintained constant. The results showed that there was a decrease in total flow by 2.6 % and an increase in suspended sediment yield by 87.3%, an increase in flow by 0.66 % and a decrease in suspended sediment by 36%, a decrease in total flow by 0.09 % and an increase in suspended sediment by 12.09 % which correspondingly to landuse and landcover of 2000 to 2005, 2005-2009, and 2000 to 2020.

Keywords: Environmental flow, GIS, model performance, Remote Sensing, QSWAT plus