Skip to main content
ARS Home » Research » Publications at this Location » Publication #169183

Title: EVALUATION OF SWAT MODEL IN PIEDMONT PHYSIOGRAPHIC REGION

Author
item SHIRMOHAMMADI, A - UNIV. OF MARYLAND
item CHU, T - UNIV. OF MARYLAND
item Sadeghi, Ali
item MONTAS, H - UNIV. OF MARYLAND

Submitted to: Annual International SWAT Conference
Publication Type: Abstract Only
Publication Acceptance Date: 7/11/2005
Publication Date: 7/11/2005
Citation: Shirmohammadi, A., Chu, T.W., Sadeghi, A.M., Montas, H. 2005. Evaluation of SWAT Model in Piedmont Physiographic Region of Maryland [abstract]. Third International SWAT Conference, EAWAG Publication. p.11.

Interpretive Summary:

Technical Abstract: Hydrologic and water quality models, such as SWAT (Soil Water Assessment Tool), are being used to assess hydrologic and water quality response of mixed land-use watersheds. They are also being used to evaluate alternative management scenarios as they affect the health of stream systems. The recent US Clean Water Act on TMDLs (Total Maximum Daily Loads) issued by the USEPA to every US state has motivated scientists and action agencies to use models such as SWAT to develop TMDL plans. However, accuracy of the simulation results obtained by models such as SWAT depends heavily on the availability of reliable input parameter values and proper observed streamflow and water quality data. Observed data are necessary to calibrate and validate these models, and proper input parameter values are necessary for their reliable performance in a given land resource region. This study used about ten years of observed hydrologic and water quality data to evaluate the capability of the SWAT Model in simulating hydrologic and water quality response of a 340-ha agricultural watershed in the Piedmont region of Maryland. The watershed is mainly an agricultural watershed with three major dairy operators. Soils in the watershed are dominated by Penn silt loam series with moderate infiltration rate. The study also evaluated the impact of uncertainties in input parameters on simulation output. Results indicated that the SWAT Model is a strong annual predictor, but its use for shorter time intervals such as daily and monthly time steps needs to be revisited. Results also indicated that uncertainty in input parameter values does indeed affect the model's output, thus risks associated with such uncertainties need to be considered. For example, in using the SWAT Model for developing TMDL plans, one should consider such inherent uncertainties in computing "Margin of Safety."