Monday, 6 October 2008
George R. Brown Convention Center, Exhibit Hall E
Rajith Mukundan, University of Georgia, University of Georgia, 3111 Miller Plant Sciences, Athens, GA 30602 and David Radcliffe, Crop & Soil Sciences, University of Georgia, University of Georgia, 3111 Miller Plant Sciences, Athens, GA 30602
Soil
data is an integral part of watershed scale hydrologic models. Soil properties
such as bulk density and hydraulic conductivity affect processes such as
infiltration and surface transport of water and pollutants. The Soil and Water
Assessment Tool (SWAT) tool, a widely used physically based watershed scale
model for water and pollutants uses the State Soil and Geographic (STATSGO)
database as the default dataset for soil information. We compared the
performance of this 1:250,000 scale resolution database with the Soil Survey
Geographic (SSURGO) database having 1:24,000 scale resolution in a calibrated
SWAT model for the Broad River watershed in Georgia. The SSURGO soil database
was incorporated into the SWAT database after preprocessing the original data
using the preprocessing tool. Spatial Modeling was performed using digital
elevation model (DEM), land use layer, soil layer and daily rainfall data. The
model was simulated for the period from 2005 January to 2007 December using
daily time steps. Our preliminary results indicate that the model predicted
higher flow during storm events with SSURGO database. However, STATSGO
predicted more sediment during storm events. These results could be due to the
effect of differences in spatial representation of hydraulic conductivity
values in the two soil databases.