In the 1980s, a group of scientists in a small research building in Temple, Texas, began a journey to mimic how watersheds work. Today, a major product of their efforts-SWAT-is used throughout Texas, the nation and the world.
SWAT, the Soil and Water Assessment Tool, is a sophisticated computer model that predicts the impacts of weather, soils, land use and land management on water supplies as well as nonpoint and point source pollution in small to large watersheds. Information such as rainfall amount, soil type, and the amount of nutrients and pesticides applied to the land over the years are fed into the model. Geographical Information Systems (GIS) are also integrated into the computer program that has 400 to 500 mathematical equations with more than 50,000 lines of computer code.
The model predicts how much water, sediment, nitrogen, phosphorus, pesticides, bacteria and other pollutants are running off the land and getting into lakes and rivers, and the impact different water management decisions could have.
"We're trying to mimic nature," said Dr. Jeff Arnold, research leader and agricultural engineer for the Grassland Soil and Water Research Laboratory in Temple, part of the U.S. Department of Agriculture's Agricultural Research Service (ARS). "We give our best estimate of what's going to happen.
"And SWAT's estimation gives decision makers a tool to solve water quality problems, Arnold said.
Dr. Raghavan Srinivasan, director of Texas A&M University's Spatial Sciences Laboratory and professor in the Departments of Ecosystem Science and Management, and Biological and Agricultural Engineering, agreed.
"For policy makers, models such as SWAT can serve as virtual laboratories for testing the effectiveness of alternative environmental policies and pollution control programs," he said.
SWAT is a continuation of models developed over 30 years at the ARS laboratory in Temple. Srinivasan, who was at Blackland Research and Extension Center at the time, and Arnold worked together with other Experiment Station, USDA Natural Resources Conservation Service (NRCS) and university scientists to develop SWAT and its national spatial databases and GIS interfaces.
"Srinivasan took these large-scale databases-the soils data and land use and topography-and really pulled all those together in a form we could use in the models," Arnold said. He explained that GIS takes the maps and associated data and spatially pulls everything together, automatically develops all the SWAT inputs and then displays the information on maps. "Without that application, the model's use would be very limited."
One of the big drivers for developing the model was determining environmental impacts of different conservation practices, Arnold said. Congress, through the Resources Conservation Act, requires NRCS to report to Congress every 10 years on the status of the nation's soil and water resources and evaluate conservation practices.
So, in the early 1990s, the researchers sold NRCS on the idea of using EPIC (Erosion Productivity Impact Calculator), SWAT and GIS to model the 48 contiguous states, simulating the effects management activities have on water quantity and quality in watersheds, Srinivasan said. The result was the Hydrologic Modeling of the United States or HUMUS project, a $1 million five-year project. Before HUMUS, NRCS relied on field observations and statistical approaches to estimate the impacts of conservation programs.
"This was the very first time we were able to model the continent using EPIC and SWAT models," Srinivasan said. "There were no models at that time for such a large area."
The latest national assessment for NRCS, Conservation Effects Assessment Project, or CEAP, is currently being developed. In this assessment, the scientists are using SWAT to incorporate outputs from APEX (EPIC's successor), which simulates cultivated lands. SWAT routes pollutants such as sediment, nutrients and pesticides through streams, rivers and lakes to the sea, Arnold said.
"The USDA Farm Program has spent billions of dollars supporting conservation practices with the farmers, and they want to know the environmental impacts of these practices," Arnold said.
Because of the success of the HUMUS project, the SWAT research team started working with the U.S. Environmental Protection Agency (EPA) and state environmental agencies in the late 1990s to evaluate the environmental impacts of pollutants at the local watershed level in support of the federal Clean Water Act. EPA's Office of Science and Technology developed a tool kit-Better Assessment Science Integration point and Non-point Sources or BASINS-to help states analyze their impaired water bodies, estimate Total Maximum Daily Loads (TMDLs) and evaluate various best management practices suggested for each pollutant. SWAT, GIS and the databases are part of that tool kit.
Today, many states across the country use SWAT in their TMDL programs.