RRWC

Soil and Water Assessment Tool for the River Raisin Watershed

The River Raisin Watershed Council has contracted with Stantec Consulting of Ann Arbor, MI to develop a hydrologic model of the River Raisin basin. Stantec has donated over 300 man-hours to assisting the RRWC in successfully obtaining grant funding for this project.

The MDEQ requires watershed management plans funded through Section 319 grants to quantify sources of pollutants and determine recommendations for improvements. Prioritization of watershed water quality problems and solutions includes solicitation of stakeholders for existing data and problem identification, gathering large amounts of data and, the use of a hydrologic GIS-based model.

In a watershed of this size, a GIS-based model will help us to efficiently estimate existing non-point source loads, and project the impacts of watershed land use changes and recommended improvements. We will use the GIS-based Soil and Water Assessment Tool (SWAT) to model existing and projected conditions in the watershed. SWAT was developed in the 1990’s by Dr. Jeff Arnold for the USDA to predict the long-term effects of land management practices in large watersheds containing various soils and land use types. Because it is a long-term model, SWAT is not designed to predict individual flood events or estimate the effects of large, accidental spills. Rather, SWAT is designed to predict the yields of water, sediment, and agricultural chemicals in a river based upon physical conditions over an extended period of time. Examples of physical input data for SWAT include weather data, soil properties, vegetation, topography, and land management practices applied within the watershed.

SWAT uses topographic data to divide the watershed into subbasins, which are spatially referenced to each other. Within each subbasin, SWAT then identifies hydrologic response units, or HRU’s, that have unique land cover, soil types, and management conditions. In addition to HRU’s, each subbasin is assigned information about its climate, groundwater, ponds, wetlands, and its reach, the main channel, stream, or river draining out of the subbasin. For the River Raisin SWAT model, elevation data was acquired from the USGS National Elevation Dataset. Rivers, lakes, and streams were provided by the USGS National Hydrography Dataset, and supplemented by the EPA Basins website. Minor errors in both datasets were manually corrected prior to integration into the model. SWAT identified 35 different subbasins within the River Raisin watershed. The following datasets were then incorporated into the model:

· Data for dams and reservoirs was collected from the Great Lakes Fishery Commission, the Michigan Department of Natural Resources, and from local officials in the watershed.

· Land Use data for the entire watershed was found at the NOAA Coastal Change Analysis Program website

· Soils data was provided by the US EPA State Soil Geographic Database.

· Daily precipitation, temperature, wind speed, dew point, and cloud cover for 1995 through 2005 were provided by the NOAA Climatic Data Center.

· The MSU Michigan Climatological Resources Program was the source of daily solar radiation data.

· The EPA Permit Compliance System database offered a wide variety of chemical and physical data pertaining to point source dischargers in the watershed. This data was supplemented through direct contact with dischargers, including many water treatment plant officials.

· Agricultural management practices where also considered when assessing the water quality in a watershed. Three generalized crop rotation schedules were used based upon data collected from the USDA’s National Agricultural Statistics Service and through conversations with officials at the Lenawee Soil Conservation District.

The accumulation of all collected data was integrated into the SWAT model, and the calibration phase was initiated. The SWAT model interpreted the input data, and the predicted results were compared to actual long-term data collected by USGS monitoring stations and water chemistry data from the Heidelberg College National Center for Water Quality Research. Locations included sites near Monroe, Manchester, and Adrian. The calibration process for the River Raisin SWAT model was modified and repeated dozens of times until the results of the model closely resembled the actual conditions that occurred at the known locations and times, therefore validating the model.

Now that a validated SWAT model is available for the Raisin River, a wide variety of Best Management Practices (BMP’s) may be simulated, and their effectiveness predicted. Examples of BMP’s that have been simulated by other SWAT models include pond or wetland construction and the installation of riparian areas, such as field boundaries and filter strips adjacent to watershed streams and rivers. Grade stabilization structures, strip intercropping, parallel terraces, tile drainage alteration, and contour planting are other scenarios that can be simulated. SWAT is particularly useful at predicting the effectiveness of agricultural management practices such as modifying tillage practices and fertilization schemes, and adjusting crop rotation schedules.

SWAT is one of the most technologically advanced watershed models available. To date, only one other watershed in Michigan has developed a SWAT model. The stakeholders and members of the River Raisin Watershed Council will soon be developing a list of BMP’s to be tested in the model. Many watershed councils conduct BMP’s based upon funding availability. Since all watersheds and subbasins are unique, the effectiveness of BMP’s can vary greatly based upon physical conditions at implementation locations. SWAT provides a tool for assessing the effectiveness of BMP’s at specific locations within the watershed. By using this tool, time and money can be saved by eliminating the application of ineffective management practices.