ABSTRACT This study evaluated the SWAT+ model in a Norwegian catchment with mixed forest-agriculture land use, tile drainage, and multiple lakes, and examined the added value of incorporating soft data as process-based constraints during calibration. The primary aim was to test whether such constraints improve hydrological consistency in addition to statistical fit. A stepwise methodology was applied, including parameter initialization, model verification, water balance soft calibration, and constraint-based hard calibration. We showed how each stage incrementally improved model performance. Three hydrological constraints were defined to represent water balance components (runoff coefficient), streamflow signatures (baseflow index), and expert knowledge of catchment behavior (tile flow ratio). Constraint-based calibration achieved slightly lower efficiency scores (NSE = 0.61, KGE = 0.72) than unconstrained calibration (NSE = 0.65, KGE = 0.77), reflecting the trade-off between optimizing performance metrics and ensuring realistic hydrological processes. The baseflow index was the most influential constraint, eliminating about 77% of non-behavioral simulations when assessed individually. The results also highlight the importance of lake initialization and the need for multiple performance metrics when tuning lake release parameters. Overall, integrating process-based knowledge strengthened internal consistency and increased confidence that SWAT+ performs well for the right reasons.
Shafiei et al. (Thu,) studied this question.
Synapse has enriched 5 closely related papers on similar clinical questions. Consider them for comparative context: