Location: Southeast Watershed ResearchTitle: Multi-site Evaluation of Hydrology Component of SWAT in the Coastal Plain of Southwest Georgia) Author
Submitted to: ASABE Annual International Meeting
Publication Type: Abstract only
Publication Acceptance Date: 10/26/2007
Publication Date: 6/29/2008
Citation: Cho, J., Bosch, D.D., Vellidis, G., Lowrance, R.R., Strickland, T.C. 2008. Multi-site Evaluation of Hydrology Component of SWAT in the Coastal Plain of Southwest Georgia [abstract]. ASABE Annual International Meeting. Interpretive Summary:
Technical Abstract: Many concerns have been raised about the potential impacts of land use changes and development in ungaged watersheds. In this study, the feasibility of using the Soil Water Assessment Tool (SWAT) for predicting the hydrology of ungaged watersheds within the Coastal Plain of southwest Georgia was evaluated. SWAT is a physically based watershed scale model originally developed for predicting hydrologic and water quality responses in ungaged watershed. SWAT was calibrated on subwatershed K (LR-K) within the Little River Watershed (LRW) by varying five parameters. The optimized parameter set was then applied to a watershed of similar physical conditions, subwatershed J (LR-J), a watershed with different land use and soil conditions (LR-O), and larger watersheds within the same drainage system (LR-I, LR-F, and LR-B) without further calibration . Total relative error (RE) and Nash-Sutcliffe model efficiency (NSE) were used to evaluate model performance. Simulation results with RE = ±10% and NSE = 0.70 were considered to be good. On the calibration watershed (LR-K), RE was 1.21% and daily, monthly, and yearly NSE values were 0.80, 0.94, and 0.93, respectively. Validation results on LR-J were considered to be good with -5.55% of RE and 0.74, 0.89, and 0.88 of daily, monthly, and yearly NSEs, respectively. The poorest simulation occurred on LR-O with corresponding measures of 22.69%, 0.65, 0.83, and 0.50. SWAT performed well in simulating temporal trends of discharge within all three larger watersheds with daily, monthly, and yearly NSE values greater than 0.7 in all cases. Relative error in total runoff increased as size of containing watersheds increased with -8.71%, 8.35%, and 15.61% for LR-I, LR-F, and LR-B, respectively. The variations in total runoff can be attributed to increases in surface storages and evapotranspiration from shallow groundwater within the wetland areas, which are greater near the watershed outlet. The modeling results showed that SWAT is capable of simulating temporal trends of flows on ungaged watersheds with similar characteristics to LRW in the Coastal Plain of southwest Georgia.