Projecting water resources changes in potential large-scale agricultural investment areas of the Kafue River Basin in Zambia

Climate change impacts regional water availability through the spatial and temporal redistribution of available water resources. This study focuses on understanding possible response of water resources to climate change in regions where potentials for large-scale agricultural investments are planned in the upper and middle Kafue River Basin in Zambia. We used historical and projected precipitation and temperature to assess changes in water yield, using the Soil and Water Assessment Tool (SWAT) hydrological model. Some of the Coupled Model Intercomparison Project Phase 5 (CMIP5) climate model outputs for the Representative Concentration Pathway (RCP) 4.5 and 8.5 scenarios project a temperature warming range from 1.8 - 5.7 °C over the region from 2020 to 2095. Precipitation projection patterns vary monthly but tend toward drier dry seasons with a slight increase in precipitation during the rainy season as compared to the historical time series. The best five calibrated parameter sets generated for the historical record (1965 - 2005) were applied for two future periods, 2020 - 2060 and 2055 - 2095, to project water yield change. Simulations projected that the 90th percentile water yield would be exceeded across most of the study area by up to 800% under the medium-low (RCP4.5) CO2 emission scenario, whereas the high (RCP8.5) CO2 emission scenario resulted in a more spatially varied pattern mixed with increasing (up to 500%) and decreasing (up to -54%) trends. The 10th percentile water yield indicated spatially varied pattern across the basin, increasing by as much as 500% though decreasing in some areas by 66%, with the greatest decreases during the dry season under RCP8.5. Overall, available water resources in the study area are projected to trend toward increased floods (i.e. water yields far exceeding 90th percentile) as well as increasing drought (i.e. water yield far below 10th percentile) vulnerability. Because surface water is a primary source for agriculture in this region, planning must focus on simulating the potential range in spatial and temporal variability of water resources for different agricultural production schemes, their infrastructure requirements, and attendant influence on water resources in the basin.