Impacts of Reservoir Outflow Estimation Methods in SWAT Model Calibration

Abstract

<abstract> <bold><sc>Abstract.</sc></bold> The purpose of this article is to identify how the performance of the Soil and Water Assessment Tool (SWAT) in predicting downstream flow is affected by using different reservoir outflow estimation methods in model calibration. This study was conducted within the Grenada Reservoir watershed (4,500 km²) in Mississippi. The SWAT model was calibrated and validated using long-term measured streamflow and reservoir outflow data. The performance of the SWAT model was evaluated using four statistics: the coefficient of determination (R²), the Nash-Sutcliffe efficiency (NSE), the percent bias (PBIAS), and the ratio of the root mean square error to the observation standard deviation (RSR). Based on the calibrated parameters, the impacts of the reservoir outflow estimation methods were evaluated at the watershed outlet, which is located below the Grenada Reservoir. According to the statistical analysis, the measured daily flow, the measured monthly flow, and the specific target release methods showed satisfactory results for downstream flow for both daily and monthly time steps, whereas the non-specific target release method showed unsatisfactory results. These results indicate that proper selection of the reservoir outflow approach in SWAT is important to accurately simulate the streamflow regulated by reservoirs. The results from our study can provide useful information for applying the SWAT model to watersheds that include flood-control reservoirs.