USGS Accurately Projects Impacts of Transforming Agricultural Landscapes

Release Date:

When a forest gives way to a potato field, when landscapes once teeming with wheat and barley disappear into seas of switchgrass, there are always implications. 

Environmental implications. Economic ones. Land use change can affect local climate. It can alter biodiversity, even the quality of groundwater. Any number of new realities emerge when land—especially that used for agriculture—undergoes si

Which prompts the question: Can we project with any certainty how such transformations will ripple through Mother Nature or down Main Street in the future? The answer, says scientist Terry Sohl at the Earth Resources Observation and Science (EROS) Center, is "yes."

Sohl and fellow USGS scientists have modified the Forecasting Scenarios of Land-Use Change (FORE-SCE) model developed at EROS to create a crystal ball that projects the impacts of agricultural change across large regions of the U.S. Great Plains. Unlike traditional modeling that examines land use change pixel by pixel, Sohl and his group project the impact of change on entire fields—parcels called Common Land Units that the U.S. Department of Agriculture has identified based on land ownership or management boundaries.

Using that approach, Sohl and his colleagues can mimic the planting decisions farmers make by individual fields and parcels of land, then scale it up to regional and national levels. “That’s what makes this work unique,” Sohl said. “This gives us a huge advantage because we’re actually mimicking real landscape patterns, and how ownership boundaries affect those landscape patterns.”

Video Transcript

Projecting Future Land Use Possibilities

Unlike land cover mapping, where scientists focus on detailing what is actually on the landscape using remote sensing, Sohl’s group considers multiple possibilities for land use in the future, then models how each scenario may influence everything from water quality to climate impacts.

The Landsat archive and Agricultural Census Data provide historical land use and land cover inputs for their model. In assessing how those landscapes change over time, they will turn to the Land Cover Trends project that was developed at EROS and dates back to 1973, along with the National Land Cover Database and the USDA’s Cropland Data Layer.

Historical temperature and precipitation data provide a window into how climate may affect land use in the future, and with it such things as biodiversity. The work of others offers valuable inputs as well, such as economic scenarios of agricultural land use produced by the USDA and U.S. Department of Energy.

All that can figure into the modeling mix when Sohl’s group looks at, say, western Kansas and the panhandle of Texas, where a steadily declining aquifer has set the stage for changing land use. Farmers there may be considering transitioning from corn and soybeans to more dryland farming, such as wheat, Sohl said. Perhaps they want to get out of farming altogether and move toward grazing livestock.

“It’s a bit of an unknown, so we can run a scenario in both cases,” Sohl said. “The beauty of being able to do multiple scenarios is, it helps you adapt to the future, and potentially choose a future that’s best from an economic perspective, and from an environmental perspective.

“By running these multiple scenarios and showing potential outcomes, we can provide that information not only to the scientists, but also to the land managers and the politicians. When they’re making policy, they could say, ‘OK, from an economic perspective, from an environmental perspective, the writing’s on the wall. We’re losing irrigated land. This is our best path forward.’ ”

Converting Pine Trees into Potatoes

The model is already being embraced in places like central Minnesota and in the Upper Missouri River Basin.

For the past eight years, the North Dakota-based R.D. Offutt Co. has been converting pine forests into potato fields in the Upper Mississippi watershed, growing spuds for McDonald’s and other customers. That area of permeable soil, lakes, rivers, and woods north of the Twin Cities is significant because it sits over a large aquifer that could be polluted by farm chemicals and depleted by crop irrigation. Numerous ongoing studies are trying to ascertain how the forest-to-farmland transformation is affecting the area’s watershed, a basin that drains into the Upper Mississippi River and supplies drinking water for 1.7 million people in the Twin Cities.

George Kraynick with Water Quality and Laboratory Services for the city of Minneapolis said the modeling by Sohl’s group is an important part of their ongoing discussion. More phosphorous and nitrogen showing up in the watershed means the potential for more cyanotoxins in their source water. Treating those toxins is expensive, Kraynick said, so understanding the potential of what the future could hold is imperative.

What would it take to produce a billion tons of biomass for fuel production?

USGS EROS scientists developed a model that portrays what it would take for the U.S. to produce a billion tons of dry biomass for biofuel production. In this scenario, magenta patches in the 2030 image represent fields converted to perennial grass for use by a cellulosic-based ethanol industry. The black lines represent field boundaries from the USDA's Common Land Unit data. 

“To us, 2040 to 2050 is where the true value of Terry’s work is,” Kraynick said. “Toxins are hard to get out. For us to install any kind of equipment change to treat that costs millions of dollars. Being able to use (Sohl’s) forecast, saying ‘You don’t have the problem now, but 15 years from now the nutrient load increase is going to negatively impact you,’ well, we can get the technology in place even though it’s not here … because that’s a change we can’t make fast.”

Other communities tied to agriculture could benefit from those kinds of projections as well, say Jim Wickham and Sean Woznicki with the Environmental Protection Agency (EPA) in North Carolina. The two men have been working with Kraynick and others on the water quality issue in the Twin Cities. Their sense is that modeling land cover scenarios not only potentially helps to protect source water in the future, but could ultimately result in cost savings for water treatment plants as well.

“If we can get a view into these possible futures, maybe we can do a better job of planning land use and different conservation methods that would have benefits downstream,” Woznicki said.

Wickham added: “I think the agricultural Midwest in general often has problems because of the extent of ag development and then nutrient runoff. So I think (Sohl’s) model would be valuable just about anywhere.”

The Possibilities for Bioenergy Crops

Researchers at Montana State University (MSU), the University of Wyoming (UW), and the University of South Dakota (USD) are tapping into the model’s value in a different way. The three universities have been awarded $6 million from the National Science Foundation to study a scenario in the Upper Missouri River Basin that would introduce a land use system called Bioenergy with Carbon Capture and Storage, or BECCS.

If adopted, the BECCS system would transfer land use from traditional crops or native vegetation to bioenergy crops—corn for ethanol, oilseeds, switchgrass. Biofuel crops are seen as fast-growing and renewable sources of energy. BECCS would also capture carbon dioxide (CO2) produced when the crops are turned into energy and store it deep underground.

“We’re trying to understand how biofuels and biofertilizers can be used now and in the future to turn the Upper Missouri River Basin into a more carbon neutral region,” said Paul Stoy, an associate professor in environmental science at MSU. “Things are changing now. We’re trying to help producers realize cropping systems that are more beneficial to soil conservation and economic returns. Terry’s work can help us understand the changes we are and will be going through.”

The economics of bioenergy crops certainly is a big part of this study, Stoy said. There is a biodiversity angle, too, said David Swanson, a biology professor and director of the Missouri River Institute at USD. Swanson is looking at how switching from wheat to corn for ethanol could affect native bird populations, or amphibian populations. How about wheat to switchgrass? What happens to the water quality aspect of wetlands, creeks, and streams with the transformation to bioenergy crops?

“All the impacts we’re looking at as part of this grant are really based upon (Sohl’s) model,” Swanson said. “It would be a lot more challenging without his model. It’s a critical piece.”

In some ways, it is a unique piece, too, Stoy said. “His scenario-based model is much better, with a higher resolution, than anything else out there. I think others would tend to aspire to try to match what his group is doing.”

They can try—and a lot of different groups do. But again, too many efforts rely on very coarse spatial and thematic resolution. Too many efforts model changes at a pixel level rather than where the real action is—in entire fields being transformed by the decision making of land owners and managers.

The Landsat-based work of Sohl’s group simply offers them something more, Wickham, Woznicki, Stoy, Swanson, and the others say.

“You can’t obviously accurately assess the future,” Wickham said. “But Terry’s models are widely published, widely regarded, and widely respected. It’s a very high value, Terry’s work. That’s why we’re excited about it.”

Related Content

Filter Total Items: 1
Date published: May 16, 2018
Status: Active

Land Cover Projections

Scientists at EROS look at local and global drivers of land-use change to project how different scenarios will impact and change landscapes. Using the FOREcasting SCEnarios of Land-use Change (FORE-SCE) modeling framework provides spatially explicit projections of future land-use and land-cover change. 

Contacts: Terry Sohl