JIM SMITH:
The most exciting thing to me about LANDFIRE—it’s a diverse product suite that can be applied in more ways than I can possibly describe. Virtually any application that deals with what’s on the land or is occurring on the land can utilize LANDFIRE program products in some way, now for 20 years. And to have a 20-year record of that is pretty amazing.

TOM ADAMSON:
Hello everyone and welcome to another episode of Eyes on Earth, a podcast produced at the USGS EROS Center. Our podcast focuses on our ever-changing planet and on the people here at EROS and across the globe who use remote sensing to monitor and study the health of Earth. My name is Tom Adamson.

The LANDFIRE program was established in 2004 to provide data to help with decisions about wildfire management. The data, partly derived from satellite imagery and generated at EROS, is used to map disturbance to the landscape and provide information about vegetation. LANDFIRE has now been around for 20 years. Today, we’re talking with a few of the people who have been with LANDFIRE that entire time. I’ll let them introduce themselves.

SMITH:
Jim Smith. I’m the LANDFIRE program lead for The Nature Conservancy.

BRIAN TOLK:
I’m Brian Tolk. I am a data analyst slash scientist with the LANDFIRE disturbance team. I am a subject matter expert on disturbance.

BRENDA LUNDBERG:
I’m Brenda Lundberg and I am the LANDFIRE data administrator, and we collect reference data for LANDFIRE, including vegetation and plot data and disturbance and treatment perimeters.

DEBORAH LISSFELT:
Deborah Lissfelt and my primary role on LANDFIRE is working with the reference database.

ADAMSON:
Jim, could you please give us a brief overview of the LANDFIRE project? What was the problem that it was originally designed to solve?

SMITH:
Let’s just say the genesis of the LANDFIRE program was the 2004 National Fire Plan under President George W. Bush. During those years, it was recognized that the wildland fire issue was really starting to grow. More acres burned more intensively, and we didn’t really have the data or strategy needed to tackle that particular problem. So a group that included individuals from the Federal land management agencies, the U.S. Geological Survey and The Nature Conservancy, got together and conceived of the LANDFIRE project at that time. At its heart, the LANDFIRE program is a data integrator and data creator with the goal of supporting wildland fire and natural resource management communities by providing a comprehensive suite of products for every acre in the continental U.S., Alaska, Hawaii, and primary island territories.

ADAMSON:
How is The Nature Conservancy involved in LANDFIRE? What’s its role?

SMITH:
Let me first mention, Tom, that the LANDFIRE program is really funded by the Department of the Interior Office of Wildland Fire and the U.S. Forest Service, Fire and Aviation Management Branch, each of which designates a business lead for the program. I’m here representing The Nature Conservancy’s LANDFIRE team, which is a part of our organization’s North America region science team.

Along with the USGS EROS Science Center, we’re major partners to the LANDFIRE program, and TNC’s roles primarily pertain to communications, user support, applications development, sometimes production assistance, and to being subject matter experts for ecological products and questions.

ADAMSON:
So, Jim, can you name a few of the big benefits of LANDFIRE? How does it help?

SMITH:
Well, in my mind it’s simply put that the major benefit of the LANDFIRE program is that its rich suite of products is there for everyone, everywhere, publicly available free of charge. When someone in some area has better local data for their ownership, LANDFIRE products can extend that data coverage outside their boundaries. Also where information is absent or incomplete, LANDFIRE products can provide foundational data that can be used directly after some local review and sometimes revision or perhaps combined with local data where there are thematic or geographic gaps in their datasets. It’s sometimes hard to believe, but there still is not sufficient data out there to really support wildland fire and natural resource management across the country and LANDFIRE is helping fulfill that need.

ADAMSON:
Let’s bring in Brian. Do you have any examples of benefits of how LANDFIRE helps?

TOLK:
If it wasn’t very clear, one of the things is just having this comprehensive suite of fire supporting geospatial data for all the U.S. and territories, and, you know, one of the problems that we were trying to overcome was what happens if you’re managing a particular area in the country, whether it be BLM land or Forest Service land, and a fire starts outside of that area but then burns towards your area. Are those areas outside? Does the data match? Is it the same spatial data? Are we mapping the same things?

And LANDFIRE brings that continuity together so that all that data is available in one place, it can be accessed easily, and throughout the years, you know, we’ve just gotten better at that and a lot more folks have been leveraging our data in that way and using it. Jim mentioned the products that we have. I went through earlier and counted up and we’ve got over 30 mainstream fire-relevant geospatial data products that we’re providing. So that’s how I think that another way that we provide benefits and help to the fire community and natural resources community.

ADAMSON:
I can’t help but think of the word clearinghouse, or maybe central location is a better way to put it, is that fair to say LANDFIRE is kind of a central location for all of this data and these products?

TOLK:
I think that’s very fair. Yeah, everything can be accessed through the landfire.gov website.

ADAMSON:
Brian, I’ll stick with you for this one. You mentioned that you’re kind of a disturbance product subject matter expert. So can you break that down? What is meant by disturbance and what does LANDFIRE provide in that realm?

TOLK:
Right. The term gets thrown around generically, but what we’re doing is trying to find changes on the landscape, either physical or biological. So if something has been modified on the landscape, it’s going to change how a fire might burn through that area, or it could be used for a lot of other projects or natural resources type things as well, but I should mention that when we talk about a disturbance, we’re talking about like natural events like wildfire, tornadoes, insect outbreaks. And then you have human activities like logging, urbanization, ag conversion, and those kind of things.

So what we’re trying to do is we look at this as a two-pronged approach. We have satellite imagery that we go through and curate so that we find where these changes are occurring, and we have a team of mappers that goes through, and we try to weed out what’s real change versus what seasonal change due to things like just a dry year versus a wet year. And ultimately, we come up with a final product that has identified, OK, there has been fuel reduction in these areas and then we use it through modeling and other state transition type activities to update the vegetation of fuels products which then are leveraged by the fire community. We also incorporate treatment activities that are reported by federal, state, and local management agencies, and what that is things like prescribed fire activity, chemical treatments, if there’s been thinning or logging activity, we can get that information in there. A lot of times, we’ll identify that with our satellite imagery too, but many of these might not be discernible from satellite, and therefore having that information in there is beneficial so that agencies can track what kind of activities have occurred over time and give them a little more insight into what is occurring in there. We have annual disturbance products. Then we also have a 10-year historical disturbance product which takes those annual products and puts them together into a single package to make it easier to discern. And then we also have a fuel disturbance product which is specific to fuel activity and how the fuel models have changed and can be modified.

ADAMSON:
Let’s bring in Brenda and Deb as well. Brenda, another product is vegetation layers. Can you describe what that is?

LUNDBERG:
Yeah. LANDFIRE produces several different vegetation layers. Some of the current vegetation layers that we produce include existing vegetation cover and existing vegetation height, which contain lifeform vegetation cover and height estimates for trees, shrubs, and herbs. We also produce two existing vegetation type layers. We have one existing vegetation type layer based on systems, which is a terrestrial ecological systems that shows the distribution of those across the landscape, and that was a classification scheme developed by NatureServe. We also produced existing vegetation layer using NVC groups, which shows the distribution of the U.S. National vegetation classification groups across the landscape.

LANDFIRE also does produce some potential vegetation layers, and those include the biophysical settings layer, which represents vegetation systems that may have been dominant on the landscape prior to Euro-American settlement. And we also have an environmental site potential layer, which is vegetation that could be supported by a given site based on the biophysical environment. So we have quite a few different vegetation layers that we produce.

ADAMSON:
I’ve heard both Brenda and Deb described as data hawks. Does that sound fair to call you that?

LUNDBERG:
Yes, it does.

ADAMSON:
OK. Deb, how about you take this one? How do you deal with all of the data for these products?

LISSFELT:
Well, there’s kind of two sides to this coin. I mean, there’s the gathering process and then there’s the ingesting or processing of the data. So on the sort of the processing and ingesting of the data side of things, part of it is 20 years of familiarity with a lot of these data products between Brenda and myself. We’ve been looking at this data for years and that’s helped us with our efficiencies. And then we’ve also in the LANDFIRE teams we’ve developed some GIS tools and some process developments that have really increased our efficiency that’s allowed us to now do an actual annual update, which was not possible in years previously. And then we also access a lot of data online. There’s been some—a lot of consolidation amongst online databases so that we can access things a little bit more easily, more efficiently. 

LUNDBERG:
We just collect a lot of data from a lot of different sources like federal, state, and local sources, and it comes in so many different formats. So one of the big, you know, lifts we do is we get it into a like format and into our LANDFIRE format. So the data can be compiled and analyzed.

ADAMSON:
OK, you’re making this usable for people.

LUNDBERG:
Yes, exactly.

ADAMSON:
Let me go back to Jim with this. With all of this data, I imagine there might be places where this is useful for things other than just fire management.

SMITH:
Well, that’s certainly true, Tom. And one of the major reasons that The Nature Conservancy is involved in the LANDFIRE program is because of the value of these foundational datasets for not only wildland fire management applications, but natural resource management and conservation applications as well. It can be used to evaluate the costs and benefits of different resource management scenarios restoration scenarios. It can be used in ways that we never imagined, because when we built these datasets, they were not available. And now we have these 30-plus datasets that we can use in so many ways, like assessing vegetation condition continent wide. It was just not possible before LANDFIRE began. So yes, there are probably more nonfire applications of LANDFIRE program products than fire.

ADAMSON:
Jim, this year also marks the first time LANDFIRE is offering annual updates. What are the benefits to users of offering this?

SMITH:
Truthfully, if you ask any data consumer, they would probably want to have the most current information possible. It’s always helpful to know what is there now and what is happening now, but really for our key user communities such as wildland fire, having timely information is really critical. Knowing which areas have already experienced the wildland fire, or when that fire occurred and how severe it was, is really, really important to understanding the risks of additional fires and what the postfire effects were and what are the rehabilitation needs. This kind of application would also apply to nonfire disasters like hurricanes and tornadoes that can have intensive and extensive impacts. The importance of knowing the impacts of sudden landscape changes is the primary reason LANDFIRE is now delivering products annually to really help reduce the threats to life and property.

ADAMSON:
Brian, do you have anything to add to that?

TOLK:
Just from a fire perspective, I would say just having—we were providing intermediate products that are provided earlier in the year than our final products, which has basically modified datasets so including the events that Brenda and Deb manage. But the idea there is just to let land managers have that information ahead of their fire season so that they can have that when it comes to their modeling or any other activities that they need to prepare for the upcoming year. Everybody would like it sooner rather than later.

ADAMSON:
Yeah, indeed. And this must mean that we’re processing and gathering more data, more information, even quicker. So Brenda, can you speak to that? How do we gather so much data so quickly?

LUNDBERG:
We do put out a yearly data call asking for data disturbance and treatment perimeters and vegetation and field plot data. And you know, we send it out to our contact list and many of those folks do submit their data on a yearly basis, so we rely a lot on those data contributions. A lot of agencies and other sources have moved their data into online databases, so that means we can just go ahead and collect the data when we need it, which is super convenient. And then you know, we’ve made improvements in how we process data. For example, our disturbance and treatment perimeter data always went through a topology process, which we did by hand, which was very long and time consuming. And we’ve made improvements and we now have a topology tool that does it all automatically. And so it’s been a huge time saver and it’s allowed us to process data lot faster and be able to do these annual updates.

LISSFELT:
Additionally, I would just mention that the online sources of data, for instance, we use the National Interagency Fire Center for a lot of our fire data. Previously we would access fire data from many different sources online, but there’s been sort of a consolidation of fire data online into single, you know, databases, so it’s keeping us from duplicating our efforts from accessing the same fire data from different sources across all online sources. So that’s really contributed to the efficiency of accessing data and not duplicating efforts.

ADAMSON:
How do we know the data is accurate?

TOLK:
From a disturbance standpoint, I should preface this by saying there’s approximately 9 billion 30-meter pixels that cover the lower 48 states alone. So you know it’s not unusual to have something go awry and, you know, maybe slip through the cracks. And we’ve gotten a lot better over time of making sure that those sorts of things don’t happen. I look at it from our standpoint, what we’re doing is we kind of have a four-prong approach to this or a four step. We have visual inspection. When we’re doing the mapping activities, once we’re done, we kind of zoom out and look at overall patterns and try to identify if something doesn’t look right, similar to what Brenda was saying and Deb, that we’ve been doing this for a long time, And we have some very experienced individuals doing the work and so we can kind of sometimes pick out like this just doesn’t look right. It looks like there’s way too much occurring in this area considering what we’ve seen in the past. So that will come into play. We also have automated quality assurance checks, too, that you know look for inconsistencies in the way the data is coming together, so we’ll find things that way. We’ve also done work where we’ve compared to other datasets like NLCD and LCMAP where we look to see if what we’ve mapped jives with what they have. In the past, and if things don’t match up right, then we’ll take a better look at it. And then finally we do listen to users and by providing user feedback, it’s not uncommon that folks will say this isn’t looking right in this area. They’re the experts in the ground, they know what’s happening in their particular locations that they manage. So we’ll step back and take a look at it and see if it’s something systematic on our end that that we will adjust or if it’s just part of the data that’s going into it and determine whether changes into those areas are warranted or not.

ADAMSON:
Brian, you mentioned user feedback. Let me let Jim speak to that as well. What else do you hear from users of the data and how else do you incorporate their feedback?

SMITH:
My teammates and I are so lucky to work with the user community of LANDFIRE. And I can honestly say that some users are absolutely ecstatic and say that they really couldn’t do what they do without LANDFIRE products. Now some users have indicated that products, you know, need some revision for local use. And you have to remember that LANDFIRE, the LANDFIRE program, and LANDFIRE products were developed for specific purposes related mostly to national, regional, and large landscape applications. So they do not necessarily match the needs of every user or application as they’re delivered out-of-the-box, but they can be edited. They can be reviewed. They can be modified. And I can say that LANDFIRE products have been cited in more than 1,000 published journal articles, according to Google Scholar, which indicates, to me at least, that the products are useful for a wide variety of purposes, you know, outside the original intent and original design of the program.

So I feel like we can honestly say that they’ve had value to these user communities, but we do hear sometimes that things need some work and that we listen to them and we’ve made many, many, many changes over the years, big changes, not minor changes, big things have changed in the program, because of what we hear from users and one of them is annual deliveries. People really wanted annual deliveries and it took a lot of work and a lot of effort there at the USGS EROS Science Center to really make that happen. So that’s a great example of listening to users and really delivering what they need.

ADAMSON:
If you don’t mind, let me jump back to something that Brian mentioned. He mentioned that there are 9 billion 30-meter pixels across the lower 48 states. That’s a reference to the Landsat data, Landsat 30-m pixels. So is Landsat kind of the primary input to the data that LANDFIRE uses?

TOLK:
We grew up on Landsat, obviously, that’s been our bread and butter along the way, but one of the major improvements that we’ve made here in the in the last three to four years has been the incorporation of what’s called a Harmonized Landsat Sentinel data. So Sentinel is a somewhat of a companion satellite. And it’s based out of Europe, the European Space Agency manages that. Folks realized early on that if they coordinated those two datasets, that the amount of coverage that the world would receive is tremendous. So that was a huge benefit. If you think about it. When we were back using just one satellite, we had a pass once every 16 days. And if that area that we were trying to map was cloudy, that meant we had to wait 16 days later to get the next image. As time went, we added Landsat 9 and 8 data, so that doubled the amount. So we were looking at every eight day coverage, and then once we added the Sentinel data and we were looking at, depending on where you’re at in the United States or territories, we would have coverage every up to every two or three days. So we are able to just really bump up the quality of our imagery, our image composites as we call them, producing a lot more cloud-free data and it made a world of difference when it comes to our mapping and making our disturbance and vegetation products to build from.

SMITH:
I think it’s important to add in a couple of really important points about the LANDFIRE program, particularly about key wildland fire systems that ingest LANDFIRE data. One called the Wildland Fire Decision Support System, and one called the Interagency Fuels Treatment and Decision Support System, that are basically dependent upon the LANDFIRE programs and they are key platforms. If there is a wildland fire incident or during planning for wildland fire treatments or silvicultural treatments, those systems are really, really important. Those depend upon LANDFIRE program products to actually exist and operate.

ADAMSON:
Now I’d like to ask all of you the same question—What’s the most exciting thing about LANDFIRE from your perspective? Let’s start with you, Brenda.

LUNDBERG:
It’s been exciting to see all the improvements and innovations that have happened throughout the years. You know, I think one of my favorite moments is actually when we finished LANDFIRE national. It was such a Herculean effort. We were developing these products for the first time for the whole United States. There are a lot of challenges and the LANDFIRE team was able to come together and figure out solutions. There was just a real sense of accomplishment when we completed LANDFIRE national and produced our original map layers. So that’s one of my favorite moments.

ADAMSON:
Deb, how about you?

LISSFELT:
As Brenda was saying it, you know, I think that the monumental effort that it took to, you know, get to annual updates is what I’m most excited about. You know, between the collaboration amongst the teams, you know it, it was a huge effort, took a long time to get there, is what is most exciting for me.

ADAMSON:
And Brian, how about you?

TOLK:
It’s always frustrating if you’re trying to find data and information and you’re searching on the web to look for something that can support your work, and once you go out there and you think you finally have found something and then download it and you find out that it’s not in the same projection or it’s not in the same pixel space, it’s different resolution data. Or it’s got some weaknesses that you might not have realized when you were first pulling it in. That kind of stuff can be very frustrating and so to have the data that we do, everything is geolocated, everything’s in the same projection, everything is easy to use and pull into applications. All that stuff is really exciting, and I think that it just goes to show what the value of the products that we produce and how they’re being used, and it makes you feel like your work is worthwhile. 

I’m excited about, you know, the increase in the amount of imagery that’s going to be going up there towards the end of the decade, we’re going to be launching more Landsat satellites and we’ll be able to incorporate those into our composite imagery. There’s been some discussions about even getting higher resolution data and which just gives us more precision. And that would be huge, I think, into all the activities we’ve talked about, whether it be fire or any kind of other research activities or natural resource mapping.

ADAMSON:
And Jim, what is most exciting to you about LANDFIRE?

SMITH:
Well, I got a couple of things, I think, Tom, one of them is, you know, despite this being sometimes called the information age, believe it or not, you know useful data are still scarce. You know it may not seem like it at times, but when we’re talking about data that spans a continent that really can support a myriad of applications, that dataset did not really exist before LANDFIRE, so the most exciting thing to me about LANDFIRE is that the products are all lands, regardless of ownership, that the data is foundational, and by that I mean it’s a diverse product suite that can be applied in more ways than I can possibly describe. Virtually any application that deals with what’s on the land or is occurring on the land can utilize LANDFIRE program products in some way, now for 20 years. And to have a 20-year record of that is pretty amazing. I believe LANDFIRE is actually a model for interagency interorganizational collaboration. There are tremendous complicated things that we have to deal with on a daily basis on a monthly basis from year to year, and the Department of the Interior, the U.S. Forest Service, the USGS, and The Nature Conservancy and others have really found a way to collaborate and be effective and efficient over that time, which is really amazing. So I consider that to be an exciting thing about the program. I’m excited about the continued evolution of the product suite. You know, as technology and user needs change, and we all know that both are going to change in ways that we cannot fully foresee right now. So LANDFIRE program leadership really recognizes the need to evolve and respond to users to remain useful. I’ve been in the data business for 40 years. I’ve been a LANDFIRE partner for 20 years, and I have literally seen faces light up when someone understands the diverse nature of this product suite and what they can accomplish with comprehensive datasets and models. It still sometimes gives me chill bumps when I’m working with someone and introduce them to the products and really see their ideas begin to percolate and see what they can do. So those are my favorite moments from the past is dealing with users and seeing them light up when they see what they can do.

TOLK:
One thing I was wondering, did you want to acknowledge the folks that have been on the project in the past? You know, I feel like our leadership has been really strong over the years.

SMITH:
Yeah, sure, Brian, it’s a good point, and we have been extremely fortunate to have leaders from the Department of the Interior and U.S. Forest Service that have really helped us through the years. And that would be Henry Bastian, who has been the business lead for the DOI for probably going on 20 years, 19 1/2 or 20 years. Dan Crittenden, who was the original U.S. Forest Service business lead, who retired a few years ago. Then Frank Fay and now Jim Menakis and without good leadership and good representation of programs like this to those to those agencies, we would not have survived and wouldn’t have thrived like we have over the years, Brian, so thanks for reminding me to acknowledge what they have contributed to our work and have made this possible.

ADAMSON:
I’d like to thank Jim Smith, Brian Tolk, Brenda Lundberg, and Deborah Lissfelt for joining us on this episode of Eyes on Earth, where we talked about the LANDFIRE program’s valuable data that it has been providing now for the last 20 years. And thank you, listeners. Check out our social media accounts to watch for all future episodes. You can also subscribe to us on Apple and YouTube podcasts.

VARIOUS VOICES:
This podcast, this podcast, this podcast, this podcast, this podcast is a product of the U.S. Geological Survey, Department of Interior.