Evaluating a finer resolution global hydrological model’s simulation of discharge in four West-African river basins

Abstract

Performance evaluation of hydrological models enables their consolidation, thereby allowing for the evaluation of water resource conservation approaches. This research aims to evaluate the performance of a finer resolution version (5 arcmin) of the PCRaster Global Water Balance (PCR-GLOBWB) for discharge estimation, in four basins in data-scarce West Africa; the Niger, Komadugu-Yobe, Jama'are, and Ogun. At the Ogun, discharge simulation was validated in a proxy basin, Oueme, which is hydrologically comparable. The model performance was evaluated using Nash–Sutcliffe efficiency (NSE), coefficient of determination (r2), Kling–Gupta efficiency (KGE), RMSE—observations standard deviation ratio (RSR), percent bias (PBIAS) and visual plots. PCR-GLOBWB was found to be suitable in all four basins but yielded better performance at three of the basins; the Niger, Jama'are, and Komadugu-Yobe (NSE, KGE, and r2 above 0.7) compared to the Ogun basin where a proxy validation approach was followed. Results at the Ogun underlined the importance of measured data in hydrological studies. Still, model performance was satisfactory in the Ogun. PCR-GLOBWB performances across the four basins, in the area, validate its reliability as a tool applicable for water resources management strategies and further investigation of impacts of climate variations on river dynamics.