Potential evapotranspiration products, Streamflow simulations, Hydrological drought detection, Hydrological components, Hydro-climatic regions, Africa, Physical geography, GB3-5030, Geology, QE1-996.5

Study region: This study focuses on three hydroclimatic river basins in Africa: Upper Blue Nile Basin (UBNB), Oubangui River Basin (OURB), and Upper Zambezi River Basin (UZRB). Study focus: Accurately estimating potential evapotranspiration (PET) and selecting suitable data sources for hydrological modeling present significant challenges. This study investigates the performance of three gridded PET (GPET) products, GLEAM, GLDAS, and ERA5-Land in hydrological applications, specifically focusing on hydrological simulation and drought assessments. The hydrological performance of these GPET products was evaluated using the Génie Rural à 4 paramètres Journalier (GR4J) model. New hydrological insights for the region: The results indicate that GLEAM PET outperforms other products in simulating daily streamflow for UZRB and OURB, while ERA5-Land PET is more effective in the UBNB. Seasonally, GLEAM PET exhibits the highest accuracy and lowest error in streamflow simulation across all three regions. PET products influence UBNB streamflow less than OURB and UZRB. Simulated other hydrological components using GLEAM PET show the highest agreement with actual evapotranspiration (AET-GLEAM 3.7a) and terrestrial water storage anomalies (GRACE-TWSA) across all three regions. GLEAM effectively detects temporal drought indices (SSI-9, SSI-12). Streamflow simulations with GPET products in the UBNB and UZRB capture drought events more accurately than in the OURB. These findings emphasize the importance of selecting appropriate GPET products to enhance hydrological simulation and drought assessment.