Modelling the impacts of climatic variables on the hydrology of the Upper Crocodile River Basin, Johannesburg, South Africa

Abstract

The impacts of climatic variables on the hydrology of the Upper Crocodile River Basin (UCRB) were modelled using the Precipitation-Runoff Modelling System (PRMS) for the period between 1996 and 2017. The values for the coefficient of determination (R2) and the Nash–Sutcliffe model efficiency coefficient (EN) gave a satisfactory correlation between the measured and simulated streamflow values, while a higher root mean square error (RMSE) was obtained for the validation period compared to the calibration period. PRMS was confirmed suitable for baseflow simulations because its mean annual recharge estimate of 5.3% was close to that of 7.5%, which was obtained using a separate filter parameter technique, both calculated from baseflow values. The climate scenario simulations revealed that the streamflow components are sensitive to temperature and rainfall. Higher decline in baseflow is observed when temperature increases than when rainfall decreases. Also, a smaller decline is observed in baseflow as compared to total streamflow, indicating the resistance of baseflow to drought. The scenario where temperatures were increased by 1.5 °C and rainfall decreased by 20% resulted in a decrease of 39% in total streamflow and 28% in baseflow. When the natural streamflow decreases by 39%, the wastewater discharge (5 m3/s) exceeds natural streamflow 80% of the simulation period. As urbanisation increases in the UCRB, wastewater will also increase. It is concluded that, while higher amounts of streamflow currently dilute wastewater, with changing climate variables and increasing urbanisation, less natural streamflow will be available to dilute wastewater, leading to highly polluted surface water. Therefore, better treatment and constraints on wastewater discharge are recommended.