SWAT model calibration approaches in an integrated paddy-dominated catchment-command

The applicability of Soil and Water Assessment Tool (SWAT) for rainfall-runoff modelling in the river-reservoir integrated catchment-command areas is limited due to the limited description of paddy-field dynamics. To model the dominant paddy-field hydrology in most of the catchments of Asian countries with tropical monsoon type climatology, the popular SWAT couples a pothole module in place of the conventional Curve Number (CN) approach. Although few pothole approaches are existing in the literature, the efficacy of these modules have not been evaluated for catchment-scale applications. Moreover, the existing pothole modules are more conceptual in nature with many calibration parameters, that could yield erroneous outcomes in catchment-scale applications. To address these issues, the applicability of the pothole module was assesses and evaluated in a 12,014.70 km2 river catchment-reservoir command iof eastern India. The simulation results reveal that for streamflow generation, the revised pothole approach performed the best with the NSE of 0.62, followed by the CN approach (NSE=0.60) and the original pothole approach (NSE=0.59). The water balance analysis resulted small evapotranspiration (ET) values (0.01 mm/day) for the CN approach and existing approach-based model simulation. The daily streamflow prediction uncertainty was the highest in case of existing pothole approach with R-factor estimate of 0.73, followed by 0.58 and 0.56 in case of CN approach and revised pothole approach, respectively. Overall, the CN approach performs better in case of rain fed paddy field, whereas the revised pothole approach simulates the daily-scale streamflow better in the irrigated paddy conditions.