KCARE Research Today

Wearing down the problem: Building a model to solve ephemeral gully erosion

Erosion can be a major problem for Kansas farmers: with every downpour or major weather event, precious topsoil slides away from arable fields. This eroded soil can leave large, unplantable tracts in what should be rich agricultural land. One such type of erosion yields larger rills called ephemeral gullies: they often form in the same place year after year and although they can be removed, it all adds to the farmer’s management time and cost.

Not much is known about the effectiveness of current practices for the reduction of ephemeral gully erosion. Some producers use terraces or winter cover crops to mitigate the problem. Others try creating obstacles or artificial swales to stop gully formation. It’s a time-consuming process that eats away at the farmer’s ability to focus on planting, growing, and harvesting.

That’s where Dr. Aleksey Sheshukov and his team come in. They want to assess ephemeral gully erosion by monitoring soil loss for ephemeral gullies on several no-till fields in Kansas. This will help them to evaluate the factors contributing to this type of soil loss by creating a predictive model. Once this model is complete, the team will use it to create a set of best management practices for farmers to slow ephemeral gully erosion or, better yet, keep it from happening at all.

Student uses backpack camera to photograph field
The research team engineered their own tools, like the backpack shown, for taking accurate photographs of the ephemeral gullies. Photo courtesy Aleksey Sheshukov.

The team identified a no-till field at Pillsbury Crossing, only a few miles southeast of Manhattan. They observed, from aerial imagery and digital datasets, that the field contained three ephemeral gullies. The largest eroded section is big enough to fit two and a half football fields. The other two combined measure about the area of half a city block. This is a significant amount of land lost to the farmer, and a great deal of sediment to lose into nearby waterways.

Once identified, Dr. Sheshukov and his team used photogrammetry to solve the question of how much soil is lost during the erosion process. Photogrammetry uses the analysis of multiple photographs taken of the same area and them creating a three-dimensional model from those photos. It’s the same technique used to make many of the maps people use today. They surveyed the entire area of the gullies, placing plastic pins at regular intervals into the ground forreference elevation. Armed with a digital full-frame camera, researchers had to take six overlapping photographs for every square meter of ground surface to get enough photos to create a proper 3D image of the gully. That’s thousands of photos. To add another level of difficulty, the shots couldn’t be taken from eye level. The camera had to be 3 meters (nearly 10 feet) off the ground, so the team designed and self-manufactured a backpack frame on which to mount the camera. While wearing the backpack, the operator could wirelessly capture the images on a tablet using a Wi-Fi connection. The team went through this process nine times from July 2016 to July 2017.

They uploaded every photo onto several processing software, and the resulting images helped them to calculate how the elevations were changing: basically, how much soil was being washed away or brought from topsoil uphill into the gully. All three gullies showed soil loss during the crop growing period in 2016. Then they were filled back with sediment over the winter period, and flashed again with rains in June 2017, getting deeper and wider as each season wore on. Such oscillatory sediment behavior is characteristic for any ephemeral gully on Kansas fields.

Now that the gullies are mapped and measured, and continuous survey data is collected, the team starts building computer models to examine the dynamics of runoff events. They have most of the data to create a tRIBS model (TIN-based Real-Time Integrated Basin Simulator). The researchers began a process of computer model development by preparing a digital elevation model (DEM), land use/land cover maps, soil and other data for the three gullies. Together with the tRIBS model and in collaboration with the scientists at University of Michigan, a specific ephemeral gully model is in development and will be calibrated this year.

Hopefully, these models will play a key role in defining and developing the best practices for combating this type of erosion.

Leave a Reply

Your email address will not be published. Required fields are marked *