Eyes on Earth Episode 69 - Thirty Years of Land Change in the U.S.

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. Land change is a constant. Even land areas that see little in the way of major change can see disruptions from storms, heat waves, wildfires, or invasive species. But major changes aren't uncommon, either. Each year in the U.S., farm fields are converted to tracts of suburban homes, wetlands fill in to become more permanent bodies of water, and shrublands burn to be replaced with grasslands. A team of researchers here at EROS just published a comprehensive look at land change in the U.S. The Land Change Monitoring, Assessment, and Projection initiative, which we call LCMAP, uses the deep archive of the Landsat satellite record to produce science products that characterize land cover and change across the conterminous United States. As of this month, LCMAP's dataset runs from 1985 to 2020. Our guest today, EROS research geographer Roger Auch, was the lead author of the just-released land change study. Roger has investigated land cover and land use for more than 20 years, including as part of a team that traversed the Nation documenting change in the late 1990s and early 2000s for a project called Land Cover Trends. In the recent paper, Roger and his fellow researchers dug into LCMAP's initial results to look at year-by-year land cover class change from 1985 to 2016. Along the way, they learned some very interesting details about how the nation's landscapes have evolved. Roger, welcome back to Eyes on Earth. It's been a while. 

ROGER AUCH:

Good morning, John. 

HULT:

Great to have you back. Before we get into the results that the team found here, let's define our terms. You were looking at land covered class change. And there's a difference between land cover land change, land cover class change? What do we mean when we talk about land cover and class change? What level of change are we talking about with this particular paper?

AUCH:

This paper focuses mostly on what is traditionally called somatic land cover change, or as you said, a class label change.

HULT:

So when we talk about the thematic land cover class, we're talking about things like cropland, this 30-by-30 meter pixel in the science product ... this is cropland, or this is tree cover, or it's developed, or something along those lines, those major classes?

AUCH:

Generally, most land cover class legends are a mix of land cover and land use. So your example of cropland is both, right? The land cover his tilled ground. That's the cover, but it's also a use. And so with the accounting that goes on, people want to know changes in developed land or changes in agricultural land. It really is a mix of cover and use.

HULT:

Just to be crystal clear on this. If we're talking about changes to cropland, in the paper that you guys published, we're not talking about changing from corn to soybeans, we're talking about changing from cropland to another use, another land cover altogether, something like grassland, right?

AUCH:

Yes, that's correct. 

HULT:

Well, let's talk a little bit more about LCMAP. It has 10 Science products. Primary land cover and annual land cover change were the main products for this study. Tell us about LCMAP in general.

AUCH:

LCMAP right now has eight land cover labels: developed, cropland, grass/shrub, tree cover, water, wetland, snow/ice, and barren. What LCMAP also brings to bear, which is new, is being able to look at changes in the condition of those thematic land cover classes. Time of spectral change product represents a significant break in the expected spectral pattern across time. When Collection 1.0 first came out, one of my fellow researchers mapped that for the conterminous United States, and it was about 25% had that type of change. Now that change may end up being a land cover label change as well. But let's take an example the mountain pine beetle outbreak in the Rocky Mountains circa 2010. You have a pixel that has half of its trees infected by the pine beetle and die, alright? So you have half this pixel filled with healthy trees and half of it filled with dying and dead trees. That most likely makes a spectral time break, but it may not create a label change because it's still tree cover.

HULT:

So LCMAP does offer several ways to look at how the landscapes are changing in some interesting ways. Because you mentioned one of your colleagues finding 25% of the land change, but that's not what you found when you looked at class change. What were the top line results of your study? How much of the U.S. saw a land cover label change or a class change since 1985, and what were the most obvious trends?

AUCH:

So our paper dealt with 1985 through 2016 for the conterminous United States, which we call CONUS. For CONUS, it was about 11 and a half percent, plus or minus. The vast majority of CONUS did not have a land cover label change. But that 11 and a half percent is more complex than that, because about 40% of that footprint had more than one change. The leading types of land cover label change that happened affects tree cover. It's removal, and its regrowth. Tree cover to grass/shrubs, and grass/shrub to tree cover. They basically balanced out, harvesting going on clear cut, harvesting down in the southeast quite a bit. But also then. fire impacts, right? So burn down part of a national forest in the Rocky Mountains, that area is probably going to convert the grass/shrub, and it might take 25-30 years to come back out. It's percentage change as a composition of the CONUS only went down about six tenths of 1%.

HULT:

Let's dig into that a little bit. Where did that slight decrease come from? If it balances out on the timber harvest side, where did the slight decrease come from? Was it the fire? or was there some, some other factor going on there?

AUCH:

Two main places. One is, in the eastern half of the United States, much of that is tree covered. And so when you add on to urban areas, generally you're converting tree cover to developed. But Massachusetts ... Massachusetts has very little cropland left. And so most of the conversion in the Northeast, and a lot of it in the south, is from tree cover. There's a loss there. In the western United States, where we've seen more forest fires, it takes longer for those burned areas to recover. And so if you have more fires, then you have more forest land going to this kind of grass/shrub early regrowth type of situation.

HULT:

Okay, so it is the fires and the developed land together that account for that slight loss. On the developed land side, you notice some pretty interesting things that sort of aligned with economic trends in the United States?

AUCH:

We knew that gains in developed land took a hit with the Great Recession, because the National Land Cover Dataset has shown a decrease in developed land between their 2006 and 2011 product versus their 2001 and 2006 product. What LCMAP's annual data step showed was that the 1990s through 2006 were pretty robust for gains and developed land. And if you think about what's going on during that time, the 1990s was pretty good economic period. Same into the early 2000s.We saw a great drop in developed increase, starting with the Great Recession. We knew that hit was going to happen. What surprising was that so far, up until at least 2016, the rate of new developed land had not returned to its higher amounts seen in the 1990s through 2006. What's happening there with the millennial generation becoming property owners, there's some evidence that shows that they're more focused on living in the already built environment, and maybe not going out on the suburban frontier. We'll have to see. And that's the one reason why LCMAP was important. It keeps monitoring. 

HULT:

When the Great Recession happened, you saw the drop off and you haven't seen it pick up to the same pace. And the question is why? Millennials ... do they want to live in houses that already exist? Are there other factors going on? And any of these trends that we're seeing in our lives are reflected in the land cover class change, at least on the developed side.

AUCH:

Yeah, so one of the things that land cover mapping can do is show what's happened in that aspect, and then other researchers can try to understand the drivers behind that, because that gets to be very complex. To say 'this caused that' can be difficult. Some things are fairly straightforward; other things may have connections that you wouldn't think of unless more threads are brought in.

HULT:

There are some more 'why' questions for some of these land cover class changes. Talk a little bit about open water and wetlands. What did you find on that side of things? It sounds like a lot of the open water wetlands change happened right here in the Dakotas is that right?

AUCH:

Even though surface water is a small fraction of the total CONUS, and its change is a small fraction of the total land cover class change, it does have some very interesting regional geographies. We saw a gain in water in some parts of the country and a loss and other parts in the Dakotas. North Dakota, South Dakota, we have these Prairie Pothole wetland areas. They're very dependent on interannual variability and precipitation. In the 1990s, there was about a decade long, very wet period, some of the wettest years, especially in North Dakota, that's been recorded since they started keep records back around the turn of the 20th century. It filled a lot of these basins that had been open water in the middle, but were more wetland condition on the shallower parts of the basin. And then those shallower wetlands became fully open water. Devils Lake, North Dakota expanded its surface area tremendously. In South Dakota, we have the Day County chain from Lake Pickerel down to Bitter Lake that all kind of flow together. I remember in high school, 40 plus years ago, duck hunting on a slough right on the edge of Bitter Lake. And I could wade that, three or four feet deep. Well, that same slough is probably 10 feet deep of water. And so that's one of the areas we saw expand a surface water. What will happen in the future, especially if this region dries out under climate change, we'll see a great contraction of that. In the western part United States, these closed hydraulic basins, we saw a contraction of surface water, because especially in the last decade or so of the study time period, the U.S. west generally had drought conditions. You're not getting precip or runoff in any of these places, that the water goes down.

HULT:

You saw a whole lot of additional open water surface water in that Prairie Pothole region. But that's balanced out on the CONUS-wide scale by these drier areas in the West. A lot of this, you talk about 11 and a half percent plus or minus for change, feels different depending on where you're standing. There could be a lot more change in your neighborhood than in another. It could be the difference between standing in the slough and wading through it, and now having 10 feet of water.

AUCH:

One of the things that I've learned over time in geography is, scale matters. Most people live in the local scale. They can see urban growth all around them and think that's happening everywhere. But if you scale up, that doesn't look as much. On the reverse hand, little bits of change everywhere can add up, depending on what you're trying to measure and what you're trying to inquire. Go back to the tree cover question. If you have small losses everywhere, cumulatively, that may be a substantial amount of carbon sequestration loss.

HULT:

Let's close this out talking about cropland. That was a net loss. But there's something interesting going on there too, isn't there?

AUCH:

We really saw two trends in CONUS cropland. One was, starting in the beginning of 85-86 and running up through about 2007, was a general trend of cropland loss. And then in 2008, there was a reversal. One of the big drivers was a federal policy called the Conservation Reserve Program. The legislation was passed in 1985, so the first year would have been 1986 that it got implemented. At its height, it had enrolled about 36 million acres.

HULT:

This is the CRP, Conservation Reserve, CRP program. And this is meant to preserve natural areas preserve grasslands, a financial incentive to crop less ground?

AUCH:

There's multiple reasons why this program came to be, but it started off originally as anti-erosion, trying to take marginal land out of production. It was also during the farm crisis of the 1980s, and so this allows some farm operators to have a guaranteed basis of income. But CRP also changed over time, became more targeted, more protection of waterways and certain wildlife habitats. The CRP program had an ebb and flow of gain, general gain, from '86 through 2007. After that, CRP had a decrease in enrollment almost every year. And that's what our research picked up

HULT:

From 2008 on, you're looking at gains in cropland, but because you factor in back to '85, there's still a decrease?

AUCH:

Yes. And one reason, too, is you're losing every year cropland to new development, and once that goes to new development, but it doesn't go back. You have cropland losing to places like the developed. In the south, they can turn their cropland into trees, to be forestry, because of its long growing season. Land owners, land operators down there can switch to a forestry land use, and then maybe after they cut the trees, they can convert that back if the markets are in better shape.

HULT: 

Speaking of the markets and speaking of the South, that reminds me ... I want to circle back to the south here on the open water side. Tell us about the catfish ponds.

AUCH:

It's kind of a local/regional story, primarily in west central Mississippi, the Mississippi alluvial plain, a very agricultural area for a long time. They tried to expand raising catfish into an industrial scale, into ponds, and then selling them to grocery stores. So in the 1980s there was a big boom. Well, catfish probably don't have as well-received a reputation as say, salmon or tuna. In the 2000 area and the more recent time periods, there's been a contraction of that catfish aquaculture. So these ponds either got bulldozed down, built in again, and maybe were put back into annual row crops, or they're empty. On the grand scale of CONUS change it's not much, but it is an interesting story.

HULT:

Once again, zooming in to the local scale, and just looking at open water, you can tease out a story about what kind of fish America wants to eat, more or less. Catfish was a shot these folks took, and it didn't quite catch on. And you can see that in the way the land is used in this area of Mississippi. Do you have any closing thoughts on land change on LCMAP, or just the value of land cover research in general? You've been at this a long time. I mean, why does it interest you and why does it matter? And why should it matter?

AUCH:

I started off as a historian first and then became a geographer. Something like LCMAP combines space and time. The old German philosopher Immanuel Kant used to talk about history needs space for it to happen, and space has a temporal, or historical aspect to it. So this is a combining both of them together. What we do in the land impacts us in both direct and indirect ways. How does it impact interactions with the atmosphere? And so maybe changes our climate over time? How does that change our water supply? So when we talk about majority of the American population living in metropolitan areas, those people need water, they need energy and they need food to exist in those concentrations, right? New York City, if it didn't have resources coming in from more rural areas, would be a tough place to keep 8 million people going. These questions are important to address and find out.

HULT:

We've been talking to Roger Auch about land change in the United States. It's been a fascinating conversation. Roger, thank you so much for joining us.

AUCH:

Thank you, John, for the invitation. 

HULT:

And thank you to the listeners, as well. Be sure to join us next time to learn more about satellites, remote sensing, land change and more. And definitely stay tuned for the next show on LCMAP, where we're going to be looking at Blue Oak forests in California. You can find all our shows on our website, U-S-G-S-dot-gov, forward slash E-R-O-S. You can also subscribe to us on Google podcasts. 

MULTIPLE SPEAKERS:

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

HULT:

You just did that for memory, hunh? 

AUCH: 

Yeah, but I may have turned the snow and ice around. I always get that one turned around.

HULT:

Still pretty impressive. I think.