Submitted to: Annual International SWAT Conference
Publication Type: Proceedings
Publication Acceptance Date: 5/13/2009
Publication Date: 12/15/2009
Citation: Baffaut, C., Sadeghi, A.M. 2009. Bacteria Modeling with SWAT for Assessment and Remediation Studies – A Review. In: 2009 International SWAT Conference, August 5-7, 2009, Boulder, Colorado. Texas Water Resources Institute Technical Report No. 356, Texas A&M University System, College Station, Texas. p. 180-187. Interpretive Summary: Simulating bacteria fate and movement in watersheds with computer models has started only recently. We reviewed five applications in which the Soil and Water Assessment Tool, an existing watershed-scale computer simulation model, was used to simulate bacteria transport. The review highlights similarities between these applications: sampling protocols and amounts of bacteria spread on the land. Results point to the need for designing stream sampling procedures that allow bacterial analysis of storm water samples and for additional field studies to determine input parameters. These results are useful for scientists who are planning research in this field and for model users who look for information on the model’s capabilities and limitations.
Technical Abstract: A module to simulate bacteria fate and transport in watersheds was first tested in SWAT 2000 and fully integrated to the SWAT2005 code. Since then, few investigators have utilized SWAT to model bacteria or pathogens fate and transport in spite of bacteria impairment being a major impairment of streams and rivers in the United States. In this paper, bacteria modeling applications from Missouri, Kansas, and Georgia in the U.S. and from Brittany in France were reviewed, highlighting the modeling successes and the challenges. These applications include watersheds that range from 16 km2 in Georgia to 3,870 km2 in Missouri. In all cases, land use included agricultural (cropland and pastures) and forested land with a mix of point and nonpoint sources. Nonpoint sources included indirect (manure deposited on land) and direct contributions from cattle or wildlife to the streams. In some cases, urban and residential contributions were also taken into account. Strategies to represent the different sources were determined and compared. Changes to the model’s code that were necessary to handle contributions from urban areas were reviewed. Calibration methods, parameter sensitivity and goodness of fit were compared. Research needs in the following areas should be addressed: data collection to characterize runoff event bacteria concentrations and understanding of the runoff extraction and transport of bacteria. Equations that represent these processes and parameterization of these equations need to be addressed.