SWAT Literature Database for Peer-Reviewed Journal Articles

Title:Evaluation of the SWAT model’s sediment and nutrient components in the Piedmont Physiographic Region of Maryland 
Authors:Chu, T.W., A. Shirmohammadi, H. Montas and A. Sadeghi 
Journal:Transactions of the ASAE 
Volume (Issue):47(5) 
Article ID: 
URL (non-DOI journals):https://naldc.nal.usda.gov/download/9719/PDF 
Broad Application Category:hydrologic & pollutant 
Primary Application Category:pollutant cycling/loss and transport 
Secondary Application Category:calibration, sensitivity, and/or uncertainty analysis 
Watershed Description:3.46 km^2 Warner Creek, a tributary of Little Pipe Creek (and the larger Monocacy River and overall Chesapeake Bay drainage area) that is located in Frederick County in north central Maryland, U.S. 
Calibration Summary:Monthly (April 1994 - 1995) r2/E values: sediment = .10/.05 nitrate = .27/.16 ---------------------------Monthly (April 1994 - 1996) r2/E values: soluble P = .39/-.08 
Validation Summary:Monthly (1996-99) r2/E values: sediment = .19/.11 nitrate = .38/.36 NH4-N = .38/-.05 TKN = .40/.15 total P = .38/.08 ---------------------------Monthly (1997-99) r2/E values: soluble P = ..65/.64 --------------------------- Annual (1994-99; both calib. & val. periods) r2/E values: sediment = .91/.90 nitrate = .96/.90 NH4-N = .80/.19 TKN = .66/-.56 Total P = .83/.19 soluble P = .87/.79 
General Comments:Some flow calibration & validation results repeated from Chu & Shirmohammadi (2004). Both graphical and statistical results shown for predicted versus measured sediment, nitrate, ammonium, TKN, and total P. Accounting of "outside subsurface flows" was made. Monthly sediment and nutrient predictions were considered poor, but annual predictions were judged to be good. 
Abstract:Mathematical watershed-scale models are among the best tools available for analyzing water resources (quantity and quality) issues in spatially diverse watersheds since continuous water quality monitoring is expensive and spatially impractical in mixed land use watersheds. However, models without appropriate validation may lead to misconceptions and erroneous predictions. This study used six years of hydrologic and water quality data to calibrate and validate the capability of SWAT (Soil and Water Assessment Tool) model in assessing nonpoint source pollution for a 346 ha watershed in the Piedmont physiographic region. The evaluation of the hydrology component of SWAT completed in a previous study pointed out that SWAT has no mechanism to account for subsurface flow contributions from outside the watershed. For this evaluation, all nutrient loadings leaving the watershed were adjusted to subtract the chemical transport via subsurface flow contributions from outside the watershed. Evaluation results indicated a strong agreement between yearly measured and simulated data for sediment, nitrate, and soluble phosphorus loadings. However, simulations of monthly sediment and nutrient loadings were poor. Overall, it was concluded that SWAT is a reasonable watershed-scale model for long-term simulation of different management scenarios. However, its use on storm-by-storm or even on monthly basis may not be appropriate for watersheds with similar physiography and size. Additionally, ignoring the subsurface contribution of water and chemicals from outside the watershed into the watershed aquifer could cause significant errors in model prediction. 
Keywords:Nutrients, Sediment, SWAT model, Water quality, Watershed scale