Eyes on Earth Episode 49 – Fire Atlas

JANE LAWSON:

Hello everyone, and welcome to another episode of Eyes on Earth. Our podcast focuses on our ever-changing planet and on the people at EROS and across the globe who use remote sensing to monitor the health of Earth. My name is Jane Lawson, and Iíll be hosting todayís episode. 

We will be talking about how EROS has been involved in an effort to map fires for the National Park Service. A fire atlas for each of nine parks includes maps showing burn acreage and severity for every fire within each park since 1984. Yellowstone, for example, has had nearly 100 fires, ranging from a few acres to thousands. The project used data from Landsat for earlier years, and both Landsat and Sentinel-2 for more recent fires. The National Park Service Fire Atlas is part of a larger interagency project between EROS and the USDA Forest Service thatís called Monitoring Trends in Burn Severity, or MTBS. MTBS maps the burn severity and extent of large fires that are at least 1,000 acres in the western United States and 500 acres in the east. 

Our guest today is here to talk about the process of creating the National Park fire atlases and how they will help with park management. Krishna Bhattarai has spent 12 years involved in mapping fires for the MTBS project at EROS and now serves as a technical lead, working with analysts and overseeing project quality. He took the lead role in completing the National Park Service fire atlas project, making assignments to analysts and training them while ensuring quality of the delivered product. Krishna, welcome to Eyes on Earth.

BHATTARAI:

Thank you, Jane.

LAWSON:

First of all, letís talk about what a Fire Atlas is. What information gets included for each fire, and how does it look on the map?

BHATTARAI:

Fire atlas is a compilation of fire information for one area. So, thereís a fire pattern, fire extent and frequency. Fire atlas is really a great resource to understand a complete fire history of one place. When we create data for each fire, that includes fire perimeter and burn severity map. Fire perimeter includes fire name, fire ID, fire date, fire year, and burn acreage information. Burn severity map provides a color classified map, and the patterns are reflected in six different colors¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨ representing different severity classes. Red color represents high severity, and dark green color represents unburned.

LAWSON:

So, either the area with the low severity didnít get burned at all or just a little bit compared to the very high severity, which was probably just completely burned. Is that correct?

BHATTARAI:

Yep, yep.

LAWSON:

So, youíre taking Landsat imagery from before the time of the fire and then after? Or how does that work?

BHATTARAI:

We do the initial assessment. We check the Landsat image just before the fire and after the fire. If we do the extended assessment in a high biomass area or big forest area, then we check the next yearís growing season. For example, a fire in 2000, we pick 2001, and then a pre-image from 1999.

LAWSON:

Why does the National Park Service want to map the fires in this way? What do they hope to do with the information?

BHATTARAI:

Fires have been mapped extensively by multiple agencies, but they do have different objectives. Those are not very useful for NPS. They have a size criteria. For example, if we talk ¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨MTBS, they do not map below 500 acres in the eastern (U.S.) and below 1,000 acres in the western (U.S.). They are not mapping the small fires. But NPS fires are very small. NPS fires, they range from like a half-acre to over a thousand acres. They want to map this fire because nobody catches those (small) ones, and they need those data for their unit. So thatís why they are working with MTBS at EROS to create this fire atlas for their own specific boundary.

LAWSON:

Otherwise it would be hard for them to get this information. Especially on those small fires, like ten acres.

BHATTARAI:

Yes. Because, in this case, I know we map a lot of fires with one acre size. There are many fires that we map very low acreage. And then it goes to like the Yellowstone fire atlas. They had very large fires over there. They want to map these fires within their unit boundary, so they can develop a reliable data source for the historical and current fires.

LAWSON:

Right. They can look back. What can they do in the future if they know this information about the past fires?

BHATTARAI:

Local fire managers must have good knowledge on the fire history. Only at that point can they make a sustainable strategy for the future. I believe these data sets are a very reliable resource for them to understand past fire behavior that we had. That may help them to understand fire history, current activities and how to plan for future management.

LAWSON:

This sounds like a valuable tool, then. Do you want to tell me about the process for completing one of these atlases? How long does it take to map the information for one fire? And then how about for a park the size of Yellowstone? Because you did do Yellowstone.

BHATTARAI:

The process, the first time, once we get our information, then we scene selection and processing. And then, once we get data ready for mapping, then we do the perimeter delineation using ArcMap, then we do the burn severity interpretation. Analysts group that into four classes, such as high severity, moderate, low and then unburned severity. And then we start bundling and deliver to the request. This is a simple process to complete mapping. But the time for mapping one fire depends upon the size of the fire. A small fire, it may take 20 minutes. But if it is a bigger one, it may take more than a day. Mostly we work a very small fire, normally one hour for one fire that includes all scene selection, processing, everything. So, I just say one to two hours for one fire.

LAWSON:

So, for an average?

BHATTARAI:

I would like to mention here that each fire includes multiple different data sets. Landsat, pre-image subset, post-image subset, fire perimeter, and burn severity map. We also create a KMZ file that is Google Earth-supported. Once we open the KMZ file, it opens all the data, mostly burn severity map and pre- and post-imagery, on Google Earth. So any user can view the fire on the Google Earth high-res imagery.

LAWSON:

So had you worked with the National Park Service before on MTBS? Or was this a new sort of relationship?

BHATTARAI:

EROS has been working a long time with the Park Service. Every year we receive a request for mapping. multiple individual fires. And we had been working like that, but at this time last year, they had a proposal creating fire atlas for each unit. We had a relationship for a long time, but this is a different kind of project.

LAWSON: So, they wanted nine units, right?

BHATTARAI:

Yes. They had initially provided nine units to us. So now, yeah, they are completed.

LAWSON:

Besides Yellowstone, I understand some of the nine parks included Zion National Park, Saguaro National Park, Pipestone National Monument and Tallgrass Prairie Preserve. Did a particular park or fire stand out to you as more interesting or challenging to work with?

BHATTARAI:

This is kind of new. Because they wanted different data sets, and this was the new stuff that I was working this time, that was kind of interesting. They compiled data from different sources, and they had a hard time to compile, organize and manage (it) properly. They provided fire information, for like five fires, in one single entry. Fire, burn date, date rendered, in one single entry for the same year. We had a hard time (finding out) which one went with which fire. That was a kind of challenging part. But the other one, it was really interesting to work because it was a new task. The final product, we hope, is going to be a really great resource for the local park managers. So, we were very excited to work on this one.

LAWSON:

Wonderful. Do you have any closing thoughts that you would like to share?

BHATTARAI:

Local managers, they could use this data in different ways to create a long-term plan within the park. Since they are in digital form, they can take data from other sources like topographic, climate data, human data, vegetation data like land cover data, to analyze the history of fire, how fire impacted that area, and how they are going to create a new plan to resolve those impacts. And also a very good resource for how they are going to plan long-term strategy within the park.

LAWSON:

Well, thank you, Krishna for joining us for this episode of Eyes on Earth.

BHATTARAI:

Thanks for having me.

LAWSON:

And thanks to the listeners as well. Check out our EROS Facebook and Twitter pages to watch for newest episodes. You can also subscribe to us on Apple Podcasts. This podcast is a product of the U.S. Geological Survey/Department of the Interior.