SWAT Literature Database for Peer-Reviewed Journal Articles

Title:Evaluation of the SWAT model’s hyrodology component in the piedmont physiographic region of Maryland 
Authors:Chu, T.W. and A. Shirmohammadi 
Journal:Transactions of the ASAE 
Article ID: 
URL (non-DOI journals): 
Broad Application Category:hydrologic only 
Primary Application Category:hydrologic assessment 
Secondary Application Category:none 
Watershed Description:346 ha Warner Creek (north central Maryland) 
Calibration Summary:monthly (1994-95) r2/E values: flow = .66/.52 surf. runoff = .43/.35 subsurf. flow = .56/.27 ---------------------------- adjusted monthly (1996-99) r2/E values: flow = .69/.68 surf. runoff = .43/.35 subsurf. flow = .57 & .53 
Validation Summary:monthly (1994-95) r2/E values: flow = .69/.63 surf. runoff = .88/.77 subsurf. flow = .47/.42 --------------------------- adjusted monthly (1996-99) r2/E values: flow = .68/.67 surf. runoff = .88/.77 subsurf. flow = .42/-0.02 
General Comments:Preliminary simulations showed that SWAT underpredicted subsurface flow & total streamflow. A water budget analysis showed considerable groundwater flow from outside the watershed. After further adjustments to account for the extra subsurface flow, SWAT still had difficulty handling extreme wet conditions. Overall, the predictions were acceptable for long-term simulations but not for daily. Extensive graphical comparisons are provided. Four sets of monthly tabulated statistics are presented: initial results, adjusted results, and both sets repeated with 1997-1999 for the validation period instead of 1996-1999. 
Abstract:Continuous water quality monitoring is expensive and spatially impractical in mixed land use watersheds. Mathematical watershed-scale models are among the best tools available for analyzing water resources (quantity and quality) issues in spatially diverse watersheds. Although existing watershed-scale models provide some reasonable guidelines, their application without proper validation has resulted in some misconceptions about such models. This study used six years of hydrologic data to calibrate and validate the capability of the SWAT (Soil and Water Assessment Tool) model in predicting surface and subsurface flow for a 340 ha watershed in the Piedmont physiographic region of Maryland. Previous studies have indicated that most existing models only handle subsurface flow bounded by the surface topography, thus neglecting the possible subsurface flow contribution from the outside of watershed, which appears to be a great model deficiency considering the major pathway of pollutant loadings via subsurface flow. Preliminary simulations showed that SWAT underestimated subsurface flow and total streamflow, especially during wet periods. A water budget analysis, therefore, was performed to quantify various components of the hydrologic cycle within the watershed. The resulting imbalance in water budget analysis suggested a considerable groundwater contribution from outside the watershed, especially during wet years. Adjustments to measured base flow and streamflow were made to exclude the extra groundwater recharge from outside the watershed, thus comparing the model predictions with appropriate measured data. However, SWAT seemed to be unable to simulate the extremely wet hydrologic conditions, even after adjustments to measured data. Overall, the hydrology component of the SWAT model is able to perform an acceptable prediction of long-term simulations for management purposes, but fails to have reasonable predictions for short time intervals (i.e., daily).