Assessing the Impact of Land-Use Change on Hydrological and Sediment Processes Using SWAT

Accurately simulating reservoir inflow under rapidly changing land-use conditions is critical for effective watershed and reservoir management, particularly in data-scarce tropical regions. This study applied the Soil and Water Assessment Tool (SWAT) to simulate river discharge and estimate inflow to the Sengguruh reservoir in the Upper Brantas sub-basin, East Java, using climate, topographic, land-use, and soil datasets. Model calibration demonstrated strong performance (the coefficient of determination R2 = 0.76, Nash–Sutcliffe efficiency NSE = 0.74, and percent bias PBIAS = –3.46 %), while validation yielded satisfactory results (R2 = 0.53, NSE = 0.52, and PBIAS = 6.94 %). Compared to an artificial neural network (ANN) model, SWAT demonstrated greater consistency and interpretability due to its process-based, spatially distributed structure. Simulation results indicate that ongoing land-use changes, particularly deforestation and urban expansion, have increased runoff and sediment transport, thereby influencing inflow dynamics and model reliability. The study also quantified sediment behavior, revealing strong nonlinear relationships between discharge and total suspended solids (TSS), and between discharge and sediment yield, with R2 = 0.99. Overall, this research addresses a key knowledge gap by assessing SWAT’s applicability for predicting inflows and sediment under evolving watershed conditions. It highlights the need to integrate dynamic land-use data and real-time monitoring to enhance model accuracy and decision-support capabilities.