Black Hills Perfect Laboratory for Multi-Sensor Fire Science Research

Release Date:

After the Legion Lake Fire at Custer State Park in South Dakota’s Black Hills erupted in December 2017— burning across private, state, and federal landscapes all at once—scientists at the Earth Resources Observation and Science (EROS) Center discovered an important opportunity among the ashes.

Color photo of the Black Hills of South Dakota

Researchers from the USGS EROS Center are conducting research on burn severity mapping in teh Black Hills of South Dakota using Landsat and terrestrial Light Detection and Ranging (lidar) data.

The Legion Lake blaze torched 10,000 acres of private land, 8,000 acres in Wind Cave National Park, and 54,000 acres in Custer State Park. For EROS Geographer and Senior Scientist Birgit Peterson and her colleagues, the chance to map burn severity using 3D technology across those multiple land management boundaries was an unexpected outcome of the fire’s aftermath.

It turns out the Black Hills is the perfect laboratory for assessing how and why fires vary across different land management scenarios, Peterson and EROS software engineer Gail Schmidt say. It’s a smaller, well contained forested region that isn’t difficult to access from their home base at EROS in eastern South Dakota. And with its preponderance of wildland-urban interfacing and its variety of ecosystems, the area represents many larger regions of the country prone to forest fires, thus offering the potential to apply the results of their work elsewhere.

And then there is this: Schmidt is also the chief of the Rockerville Volunteer Fire Department in the Black Hills. When wildfires erupt out there, she’s often in the middle of the firefighting response. That’s provided her with personal connections to many land owners and managers in that part of the state, which certainly proved beneficial after the Legion Lake Fire.

“Gail has been really integral to what we’re doing because of her connections out there and her work with the fire department,” Peterson said. “She helped us a lot with the planning for this project. She showed us around when we got out there. And then having been on the (Legion Lake) fire ... making all those connections for us was invaluable.”

Historically, getting buy-in from the people managing different land interests for fire science work is not always easy. “Based on what I’ve seen in the past,” Schmidt said, “the private land owners in particular are less inclined to allow us to have access to that land.”

The reality is, lands abutting one another can have vastly different interests and management objectives, Peterson said. There may be recreational needs tied to one landscape, ranching needs on the adjacent property, and forestry needs just beyond that. But since all can be impacted by the same fire, research that helps to identify what’s on the ground at each site, how it may burn in a wildfire, and how to mitigate potential fire damage by addressing fuel loads before the first sparks fly, can provide important and useful knowledge.

“I think if we can show land owners, and the private land owners as well, what is going on pre- and post-fire on these lands, they can start to better understand why the state and the Forest Service do some of the land management practices they do, whether that’s thinning or prescribed burning,” Schmidt said. “A lot of people don’t truly understand that fire can be good. Out here in the Black Hills, fire needs to be part of the ecosystem about once every 10 years, and that’s hard to get the public to understand.”

To help foster that knowledge, and with access now to the private, state, and federal lands in the Black Hills, EROS researchers are taking several approaches in their studies. One is assessing the structural change that occurs on a landscape after a fire.

USGS EROS researcher in the Black Hills of South Dakota

Fire Specialist Josh Picotte with ASRC Federal Data Solutions, a contractor to USGS EROS, does burn severity field work in the Black Hills. 

Peterson and her group received some initial seed money from Nate Benson of the National Park Service to study the effect of the Legion Lake Fire in nearby Wind Cave National Park. They also got some USGS Surveys, Investigations and Research (SIR) funding to do the burn severity research. Then with Schmidt’s connections, they were able to expand their work to include Custer State Park and private lands.

“We were able to get on a large piece of land that was privately owned, a large piece of ranch land that was very representative of the areas that had burned,” Schmidt said. “I guess I feel like it was a good thing to have; I think it’s key to the study to say we have state, private, and national park land.”

Peterson said she and colleagues Jeff Irwin and Josh Picotte have been making what are called Composite Burn Index measurements, which were developed to compare burn severity on the ground to that revealed in Landsat data. Picotte and Peterson have put in several circular plots 15 meters in radius, which then she and Irwin plan to resurvey later using terrestrial lidar.

The way it works is, that terrestrial lidar sits atop a tripod and sends out a laser beam that bounces off leaves, branches, trunks and other vegetation structures, and then returns to the instrument, creating a three-dimensional cloud of points in the shape of the structure. The lidar can capture a sweep from the ground up to the top of the canopy as well, Peterson said.

Because they had no pre-burn data on the vegetation structures before the Legion Lake blaze, Peterson said they are doing some space-for-time substitutions. That means “we go to areas that are very similar to areas that were burned, but were left untouched and were not burned, and do comparisons there,” she said. As part of their initial plots, they are also sampling along a gradient of low to moderate to higher burn severity.

Capturing all that information will eventually help them understand better what the fuel load was before the fire versus how much was left afterward, Peterson and Schmidt say.

“Obviously with firefighters and emergency responders ... if we know what the conditions were like going into a fire like this, and then we can have the information on the severity of the burn post-fire, then that just allows us to be able to predict a little bit more what’s going to happen next time,” Schmidt said.

Besides the Legion Lake fire, Schmidt said there are also potential opportunities to conduct their studies around prescribed burns taking place in the Black Hills, where they can obtain pre- and post-fire data from the same sites.

While Peterson calls the point clouds generated from lidar supplemental to the large-scale ability of Landsat to do vegetation and burn severity mapping, the use of terrestrial lidar is nonetheless becoming an important option for mapping plant life, woody biomass and more without the need for a lot of interpretation, she said. It’s not just lidar at the local level, either. The recent launchings of two spaceborne lidar systems— ICESat-2 (Ice, Cloud and Land Elevation Satellite-2), and GEDI (Global Ecosystem Dynamics Investigation)—are driving national interest in lidar, too, Peterson said.

If the EROS team can integrate their work with other datasets and start painting a picture of why fire is different across varying landscapes or how land management practices impact those fires, “then that probably becomes a broader picture than many folks are looking at,” she said.

USGS EROS researcher in the Black Hills of South Dakota

USGS Geographer Jeff Irwin from EROS works with surveying equipment while doing burn severity studies in the Black Hills.

It's not likely that private landowners will ever get too worked up about the science behind the processing of point clouds, the two researchers say. But their work could potentially influence the way those folks manage fuel loads on their lands in the future, Peterson said.

“If we can go to them and say, ‘Here’s what your permission to let us go here allowed us to do, and how we might be able to get that feedback so those types of fires don’t happen on your lands again in the future,’ I think that’s an important component,” Peterson said.

It’s going to take a few years to finish their field work, she said. After that, they need to process and analyze the data to determine exactly what they’ve got. That work will begin with a lot of questions, such as, how are they going to look at the point clouds, and how are they going to pull out the information they want?

The easy answer is basically come up with a series of metrics, like splitting the point clouds up horizontally into layers and looking at how much vegetation structure inhabits each layer, Peterson said. But she sees potential for much more, and likes the idea of using what she calls deep-learning algorithms.

“Often, you can look at a point cloud, and your eye can pick out exactly what’s going on,” Peterson said. “You can look at a before-and-after and go, ‘Oh, here, this was here before. This wasn’t here before,’ and start quantifying.

“That’s really complicated for us to write an algorithm to do, but it’s something, I think, that the deep-learning process would be ideal for. That would be new for us, but I think it’s something that would have potential.”

If she is right, then the research Peterson, Schmidt and the others are pursuing today could indeed someday have fire load and fire prevention implications not only in the Black Hills, but beyond. Taken in concert with vegetation base maps put out by LANDFIRE, with post-fire burn mapping done through Monitoring Trends in Burn Severity (MTBS), and with other datasets, the prospects for lidar-based structure assessment offer another potentially important piece of the puzzle, Peterson said.

Ultimately, it’s blending datasets in a way that is not being done very much right now, she said, while trying to paint a picture of burn severity and fuel consumption in a way that identifies differences from one landscape to another adjacent one. “That might be a broader picture than (other researchers) are looking at,” Peterson said.

Related Content

Filter Total Items: 2
Date published: September 10, 2018
Status: Active

LANDFIRE-Landscape Fire and Resource Management Planning Tools

LANDFIRE (LF), Landscape Fire and Resource Management Planning Tools, is a shared program between the wildland fire management programs of the U.S. Department of Agriculture Forest Service and U.S. Department of the Interior, providing agency leaders and managers with a common "all-lands" data set of vegetation and wildland fire/fules information for strategic fire and resource managemnt...

Contacts: Christopher Torbert
Date published: March 2, 2016

Monitoring Trends in Burn Severity (MTBS)

MTBS is a multi-year project designed to map the burn severity and perimeters of fire across all lands of the United States for the period spanning 1984 through 2010. The data generated by MTBS will be used to identify national trends in burn severity and evaluate the effectiveness of the National Fire Plan and Healthy Forest Restoration Act.