The capability of the SWAT model in simulating paired benchmark catchments (Case study: Dehgin Watershed)

Quantitatively assessing the effects of watershed management on hydrological response and sediment yield remains challenging due to the lack of natural control catchments. This study employed a paired catchment framework: a natural control catchment and a treatment catchment with several check dams. Flow and sediment transport were modeled using the semi-distributed SWAT model. Manual calibration was performed based on monthly discharge and sediment data. In the control catchment, SWAT showed high accuracy for streamflow (calibration: R²=۰.۸۶, NSE=۰.۸۰; validation: R²=۰.۹۲, NSE=۰.۸۸). The treatment catchment exhibited weaker performance (calibration: R²=۰.۷۸, NSE=۰.۶۵ – good; validation: R²=۰.۸۰, NSE=۰.۵۸ – satisfactory) due to flow regime disturbance by check dams. For sediment, the control catchment showed moderate accuracy (calibration: R²=۰.۷۳, NSE=۰.۶۹; validation: R²=۰.۶۵, NSE=۰.۵۹). Interestingly, the treatment catchment performed better in sediment simulation (calibration: R²=۰.۸۳, NSE=۰.۷۱; validation: R²=۰.۸۵, NSE=۰.۶۵), attributed to sediment trapping behind dams. The reduced model efficiency in streamflow simulation for the treatment catchment (NSE decrease from ۰.۸۰ to ۰.۵۸) reveals a key limitation of SWAT in representing hydrological effects of small check dams. Conversely, sediment simulation was more successful due to statistical smoothing. These findings highlight the need to revise parameterization of small check dams in semi-distributed models or adopt hybrid approaches.

کلیدواژه ها:

Paired catchments ، Manual calibration ، SWAT Model