Eyes on Earth Episode 48 – Satellites and the Forest Census

Download
Right-click and save to download

Detailed Description

It’s possible to map vegetation type with Landsat, but getting the maps right requires more than satellites alone. The teams behind LANDFIRE use an extensive network of ground control points to check their work, thereby bolstering the reliability and utility of their multi-layer GIS product suite. On this episode of Eyes on Earth, we meet scientist Karen Schleeweis, who tells us about LANDFIRE’s largest single source of ground truth data, a U.S. Forest Service program called Forest Inventory and Analysis (FIA).

Details

Episode Number: 48

Date Taken:

Length: 00:16:07

Location Taken: Sioux Falls, SD, US

Credits

Guest: Karen Schleeweis, FIA Techniques Research Band Lead/LANDFIRE Liaison, U.S. Forest Service

Host: John Hult (Contractor for USGS EROS Center)
 

Transcript

KAREN SCHLEEWEIS:
You know, there is a phrase that I hear often where people will say things like, "Well anyone can make a map but not everyone can make a good map." And how you know it's a good map is you are training and validating with good field data. So, you need the truth. You need a source of truth to assess your map products. Otherwise maps are more art than science.
JOHN HULT:
Hello everyone and welcome to another episode of Eyes on Earth. We're a podcast that 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. I'm your host John Hult. Today, we're talking about a place where satellite data and on the ground data meet. Specifically, the reference data set used by the multi-agency partnership known as LANDFIRE. LANDFIRE data teams here at EROS build incredibly detailed GIS mapping products that help fire scientists and land managers track fire danger and post fire recovery across the United States. LANDFIRE data also contributes to the study of wildlife, ecosystems, carbon, and climate. The U.S. Forest Service is one of LANDFIRE's founding partners. More than a quarter of the more than one million on the ground reference points that are combined with Landsat satellite data to produce LANDFIRE come from partnerships with the US Forest Service. Today, we are talking to Forest Service Ecologist Karen Schleeweis. She's the Techniques Research Band Lead and LANDFIRE Liaison for a program called Forest Inventory and Analysis or FIA for short. It's essentially an annual census for the nations forests, and it plays a key role in making sure that LANDFIRE is as accurate, detailed, and reliable as possible. Karen, thank you for joining us.
SCHLEEWEIS:
Thank you for the invitation.
HULT:
First off, Karen, tell us about FIA. How long has this been around, and where did it come from?
SCHLEEWEIS:
You know, as much time as this podcast has, I could probably take five times that amount to talk about that question. But if I had to summarize, I'm going to say, FIA in our logo, says we're the nations forest census. But really, we're a survey. And to take that further, we are a collection of surveys, actually. So, we have the Forest Monitoring Survey, which is the kind of ground plots that LANDFIRE interacts with. But we also have a National Woodland Owner Survey, timber products, tropical island survey. Our Forest Monitoring Survey is a probability based, statistical sample of field plots.
HULT:
What does that mean in practice?
SCHLEEWEIS:
The analogy would be the Nation's people census. They go out and the idea is they go out, and they measure every person. That's what makes it a census. So then the numbers are actually from counts. They know the total population. In our case, we are going out and measuring a sample of the forest. And then we make estimates about the forest area and where it is at. People use it to answer questions like, "How much forest exists and where? Who owns it? How is it changing in terms of growth, mortality and removals?" FIA is kind of the program that collects, analyze and reports on all those data. But then how long has it been around? The short answer is way longer than me. It's been in continuous operation since 1930. It's really had a similar mission that whole time. As legislators tried to adapt to the Nation's need for reporting on forests and rangelands, they adapt somewhat of what we are responsible for. An example, the 1999 Farm Bill: in direct response to user needs, we enhanced the FIA from a periodic to an annual survey. 
HULT:
So it wasn't always every single year.
SCHLEEWEIS:
No, not at all.
HULT:
Ok. Well, thank you so much for explaining that. That explains why it's different from the U.S. Census, obviously. And it also gets into how are you able to do this? Because one of the questions that I had was, do you hire college kids to go wander the forest and tag every single tree in the country? That's really not what's going on.
SCHLEEWEIS:
Not in terms of the college kids part, unless they have been heavily trained and vetted. But we do send boots on the ground. That is what we are known for. But they are highly trained. It's the quality of the measurements and that we're using the same definitions that make it such a useful data set that people trust.
HULT:
But again, you are not looking at every single tree. You are doing samples. So give us an idea of the scope of FIA. How much data are we actually talking about? And can you tell us a little bit more about how it's collected? Is this the kind of thing where it's just constantly in motion, the plans for the next survey?
SCHLEEWEIS:
Definitely constantly in motion. We measure forest land on public and private forests. All the way out in Guam and the Pacific Islands to Puerto Rico and the U.S. Virgin Islands and all of the 50 states in between. And that's a good example of where we leverage partnerships. FIA is really strong because of the partnerships it has with state foresters and other various organizations. So that's kind of the spatial extent. The total number of plots in the survey? Roughly we have one permanent field plot every six thousand acres. So if you imagine draping the US in a hexagon grid, almost like a soccer ball, with hexagons with an area of roughly twenty-four hundred acres, we randomly locate a spot within that hexagon. That's kind of the spatial grain. Then we visit those plots roughly every five to ten years, depending on the location. Then sometimes our spatial intensity is much higher. There's been states who pay for 2- to 3 times the intensity. Then that will increase the number of plots and drop that distance from plot to plot. And then how's it collected? Those plots ... there is kind of an intense design of the plot itself and how we go out on it. But we do actually measure the diameter breast height-the DBH-of the trees, the species types, we collect invasive data. There's quite a lot of variables that we collect on the ground. And so we certainly, like you said, don't measure every tree, but we actually measure every tree on our plot.
HULT:
So, the plot itself ... how large is the plot by the way?
SCHLEEWEIS:
Each field plot consists of a roughly 24-foot radius circle subplot. That's a little hard to describe: four sub-plots within a macro plot, then we also have micro plots. But I think to get the rough idea, you could say the plots encompass or describe the area of an acre.
HULT:
So, one acre is a plot. There's one per every 6,000 acres. And again, you're looking at every tree within that plot, and there are a heck of a lot of them in the United States. There's a lot of data.
SCHLEEWEIS:
It is. Yep. We're actually one of the larger national forest inventories in the world. That's also related to the size of the country, it doesn't mean we necessarily have the densest amount of plots, but it is a lot of forestry data.
HULT:
How are these inventories used? Who uses them? Why is FIA important, more or less?
SCHLEEWEIS:
There is actually a cool paper that was published in 2018; the main author was Tinkham. He did analysis over the last three decades, finding roughly 300 and some research papers that used FIA data. Which really got to the nugget of who uses it particularly for research and why. And forest health and carbon cycle applications were the two dominant, but forest products, climate and remote sensing applications that were in there as well. Fuels and fire hazards kind of fall under that forest health, and that's one of the links to LANDFIRE. But our data supports a lot of operational needs. For example, state foresters, industry consultants, different environmental organizations even forest service officials. Journalists have used our data for different reasons. And then we get a lot of calls from private citizens. Like "what's this tree in my back yard? And is it native? Can I do this with it? Can I find out about this? Or someone planted this species and how tall can I expect it to get?"
HULT:
It really does run the gamut, from people who are trying to manage the forest, to understanding what's changing, invasive species, fires, etc., into general management, all the way to outside agencies. And if I'm a company and want to build something here, I want to know what's on the land. I can look at the FIA data and get a sense of what's going on in the forested land?
SCHLEEWEIS:
Very much. And then, you know ... it's funny, because when you ask why it's important, you can lead with all the uses but you can also say it's important because we focus on high standards. So people trust our data. We're consistent, and have been through time. And I think more than ever it feels like there is a greater need for information about the supply and the condition of the nation's natural resources, including the forest land and rangeland.
HULT:
You mentioned that your background is in remote sensing. We've been talking about boots on the ground, measuring actual trees. Actual human beings, walking up to a tree, knocking on it or whatever. How does this tie into LANDFIRE? How does this tie into remote sensing? Tell us how LANDFIRE came to use FIA data, and LANDFIRE obviously is a satellite-derived data set. How did those things come together? What's the history there?
SCHLEEWEIS:
They knew from the beginning that they needed comprehensive, field-based reference data to make plausible maps with remote sensing. As soon as you start to ask, "what field data do we have in the U.S.?" Particularly for forests, FIA inventory will come up quickly.
HULT:
Tell us how LANDFIRE might use FIA to improve its mapping products. I guess this is just a basic, "What are field plots good for?" sort of a question.
SCHLEEWEIS:
There are a lot of FIA scientists on that original prototype helping develop the methods for vegetation modeling or for canopy cover modeling. There's over 20 geo-spatial layers that LANDFIRE produces. And not all of them are forest specific. FIA, for the areas that are forested, it's really that data that helps them improve. That can be a number of things. The FIA, canopy and height measures are the gold standard for forest plot data. And they use that with a lot of LIDAR to improve their canopy cover and canopy height maps across the US.
HULT:
Lidar? What's that? Is that a pet that you can buy? Like a ferret or something? What is lidar?
SCHLEEWEIS:
Yeah, but you have to be real careful, because they can bite you. (laughing) Wow, so lidar. Light Detection and Ranging. Lidar's essentially a type of image resource where certain bandwidths are bounced off of objects, and then they calculate the time that (it takes to) come back. So it's like shooting a laser and counting how long it takes to return. Then you know how far away it is, or possibly how tall it is. 
HULT:
So that's the value of LANDFIRE. Knowing how tall the trees are, how tall the canopy is?
SCHLEEWEIS:
For one of their spatial products, yes. They have a tree height product, but they also have a lot of data that helps now support height for shrub, and even grass. And that comes from other partnerships they have. Knowing the height of vegetation helps when you're understanding or modeling how fire spreads through the landscape, how high the canopy crown is, and how intense the fire is before it has to get up into a crown and be a crown fire. So we do provide height measurements. But then there are other things we measure even more directly. We measure duff and litter so that fine fuels as well as coarse fuels. So, dead woody debris that's on the ground. That really plays into fire modeling. We measure those directly and LANDFIRE uses those to produce their fuel maps. Plots on the ground, you know exactly what's happening right there, but it doesn't tell you much say, 130 feet away. In order to build that relationship, you link the ground data with something like Landsat that has a 30-meter grid. Where all of the space is covered. They're getting a fine idea from the plot data and the LIDAR. Then they're making relationships or models with the Landsat data, and kind of predicting the relationship across the surface. That's where they get their maps.
HULT:
LANDFIRE released its ReMap for the conterminous United States last year, and that updated its base map and offered a host of other improvements for users. Tell us how FIA contributed to that process.
SCHLEEWEIS:
There're a couple different ways. First, I have to say that I feel like clapping, because the 2016 ReMap effort was enormous. And I think this suite of products really represents a new generation of LANDFIRE products. That said, there's some basic ways that FIA contributed. More plots always helps. We now have an all condition inventory. So that's where we take these really detailed measurements on land that doesn't quite meet our forest definition. So that it might have trees on it, but we wouldn't necessarily have taken such intense measurements there before. Additionally, since the last mapping effort, FIA has pushed into interior Alaska. LANDFIRE maps the 50 states and the insular areas but before, FIA was not able to get into these more remote locations. LANDFIRE was able to leverage all of these new plots to help with the vegetation and forest mapping. And there are other ways that FIA helps contribute through the various remapping efforts, including this last one. We'll go through and do updates to say, the algorithms or the kind of logic that they use in their models to map say, forest height or forest canopy cover. LANDFIRE does an amazing job of connecting with its user base to understand where its products are and aren't working. That helps direct some of the improvements and innovations. And then we will use FIA data in those locations if the questions are related to improvements over forest data to help answer some of the questions about why does the data look like this or how we can improve it.
HULT:
So is there a way for the public to access FIA data? Where can people learn more about this?
SCHLEEWEIS:
The quick answer is there's a national FIA website which is easy to find www.fia.fs.fed.us. There is an About Us page on there that really has a lot of the background information. But the national public database is there, as well. You can access our state reports. You can access things like our strategic plan or all kinds of things. I happened to mention that one research paper. But all of our research publications, you can also access on TreeSearch. That's an electronic data base that has free copies of all the FIA and all the forest service publications. So those might be two resources: the national FIA website and TreeSearch if you're interested in some of the research papers.
HULT:
TreeSearch for research, not trying to find the tree that you carved your name into in 1972 but for research. Good to know. And also, I just wanted to point out, on the website, there are some really complicated stuff in there. You can get real deep into the data and have to know what the heck you are doing. Something like the State Fact Sheets ... there are some really sort of simple rundowns that you can get an idea of what you have in your state for example as well, right?
SCHLEEWEIS:
There are. And you know, a really fun, I think really exciting new resource is that we are leveraging things like Tableau and fact sheets. So now we have more interactive state fact sheets. Or even the questions like my mom might ask me, "what do you mean by forest?" We'll have an interactive fact sheet on that. We're really starting to recognize the different types of audiences that are interested in the data that we have and not just kind of report to a classic say timber industry or state forester audience.
HULT:
Reaching out to the people. Meeting them where they're at, right?
SCHLEEWEIS:
That's the idea.
HULT:
We've been talking with Karen Schleeweis of the US Forest Service about Forest Inventory and Analysis. Karen, thank you so much for joining us.
SCHLEEWEIS:
It's a pleasure, John.
HULT:
We hope you come back for the next episode of Eyes on Earth. You can find our episode archive by visiting usgs.gov/eros. You can also find us on Apple podcasts or Google podcasts. Thanks for joining us. This podcast is a product of the US Geological Survey, Department of Interior.