Evaluating the Performance of the SWAT Model in Hydrological Simulation Using Global and Local Land Use and Soil Data (Case Study: Chel-Chai Watershed)

Water, this vital element, plays an irreplaceable role in our lives. From drinking water supply to energy production, agriculture, industry, and numerous other sectors depend on it. However, human activities have profoundly impacted water resource availability. Under these circumstances, a deeper understanding of the hydrological cycle and river behavior becomes more crucial than ever. One key tool for hydrological cycle simulation is the SWAT (Soil and Water Assessment Tool) model. This study investigated the effects of using regional versus global soil and land use data on SWAT model performance in the Chel-Chai watershed. Monthly river discharge data from Lazoureh and Jangaldeh stations (2006-2020 for calibration; 1997-2005 for validation) were utilized. The results showed that at Lazoureh station, during the calibration phase, regional and global data showed similar performance; the NS and R indices for both data types were 0.60 and 0.77, respectively, and the MAE and RMSE errors were both 0.78 and 1.11. Consequently, no difference was observed between regional and global data in this phase. During the validation phase at Lazoureh station, regional data performed better than global data, reducing MAE and RMSE by 2.56% and 1.92%, respectively. At Jangaldeh station, during the calibration phase, regional data also outperformed global data. The NS and R indices for regional data were 0.73 and 0.90, respectively, while for global data they were 0.58 and 0.87. Regional data also showed better performance during the validation phase. The results demonstrate that regional data can provide more accurate river discharge estimates, particularly during validation phases. This study highlights the importance of spatial data resolution in hydrological modeling accuracy.