Submitted to: USDA-CSREES National Water Quality Conference
Publication Type: Abstract Only
Publication Acceptance Date: 11/15/2008
Publication Date: 2/9/2009
Citation: Baffaut, C., Sadler, E.J. 2009. Simulation of Lateral Flow with SWAT [abstract]. USDA-CSREES National Water Quality Conference. February 8-12, 2009, St Louis, Missouri. 2009. Available: http://www.usawaterquality.org/conferences/2009/ceap.html.
Technical Abstract: Calibration of the SWAT model for the Goodwater Creek Experimental Watershed (GCEW) showed that percolation through the restrictive claypan layer, lateral flow above that layer, and redistribution of excess moisture up to the ground surface were not correctly simulated. In addition, surface runoff and lateral flow were not sensitive to the low claypan hydraulic conductivity, which is the main reason these soils are poorly drained and experience a perched water table in early spring. A revised algorithm was proposed for reducing percolation from one layer when the layer below is close to saturation, for estimating lateral flow under saturated conditions, and for redistributing excess moisture above a restricting soil layer anywhere in the soil profile. The revised model was tested on the 72-km2 GCEW in northeast Missouri. The Nash-Sutcliffe efficiency and the Pearson correlation coefficient calibration statistics were slightly improved at the 3 gauging stations of the watershed for the calibration and validation periods. More importantly for the simulation of pollutant transport, the separation of surface runoff and groundwater flow was significantly improved: surface runoff and groundwater average differences were less than 15% and 75%, respectively, compared to 25% and 275% with the original algorithm. The improved simulation of surface runoff and lateral flow should allow a more accurate simulation of pollutant transport when soils have a restrictive layer that results in lateral flow or poor drainage and soil saturation. For watersheds in which lateral flow is a significant component of the water balance, the revised model should significantly improve confidence in model’s results.