Water Resources Modeling in Data-scarce Watersheds: Contribution of the SWAT Model and the SUFI2 Algorithm to the Study of the Thiokoye River Basin

Sustainable water resource management in West African watersheds is increasingly constrained by limited hydrometeorological data, high rainfall variability, and complex hydrological processes. These factors hamper accurate simulation and understanding of watershed behavior, making it difficult to develop effective and adaptive management strategies. This study addresses these challenges by applying the SWAT model, coupled with the SUFI2 algorithm, to the Thiokoye river basin, a sub-catchment of the Gambia River, with the goal of improving hydrological understanding in data-scarce environments. The model was calibrated over the period 1979–1992 and validated from 1998 to 2002 to capture climatic variability and test model robustness. Performance was assessed using multiple statistical indicators: Nash–Sutcliffe efficiency (NS), coefficient of determination (R²), Kling–Gupta efficiency (KGE), percent bias (PBIAS), and uncertainty metrics (p-factor and r-factor). Calibration yielded excellent results (NS = 0.98, R² = 0.98, p-factor = 0.90, r-factor = 0.27), while validation confirmed the model’s reliability (NS = 0.82, R² = 0.83, KGE = 0.81, PBIAS = − 13.2%, p-factor = 0.78, r-factor = 0.19). Sensitivity analysis identified CN2, ALPHA_BF, GW_DELAY, and CH_K2 as the most influential parameters affecting surface runoff, baseflow, and groundwater response. These findings demonstrate the effectiveness of the SWAT-SUFI2 framework for simulating hydrological processes in poorly gauged basins and highlight its potential as a valuable tool to support integrated water resource planning and climate change adaptation in vulnerable hydrological systems.