Eyes on Earth Episode 72 – Northward Shift of the Boreal Forest

LOGAN BERNER:

We really live in like the golden age of remote sensing right now, where there's all these different platforms. We have this amazing Landsat record that goes back decades, allowing us to look at these changes that have occurred anywhere on the planet. It's really quite striking. 

JOHN HULT:

Hello everyone, and welcome to another episode of Eyes On Earth. We're a podcast that focuses on our ever-changing planet, 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 for this episode, John Hult. The boreal forest, or taiga, is the largest land biome on the planet. It stretches across nearly 5.7 million square miles in the northern latitudes. That's nearly a quarter of all forested lands in the world. The spruce, pine, and larch of the boreal forest are sometimes referred, to along with the tropical rainforests, as the "lungs of the world." This sprawling biome also happens to be one of the most rapidly shifting in the face of climate change. Many studies have suggested that the boreal forest tree line is moving northward as temperatures warm, edging itself into the colder tundra. In the taiga's warmer zones, there is concern that disturbances and altered weather patterns could see trees replaced by grasses or shrubs. Today's guest took to remote sensing to study those changes. Professor Logan Berner works at Northern Arizona University's Global Earth Observation and Dynamics Ecosystems lab, also known as GEODE. It's clever, right? In a recent paper, Logan sifted through four decades of Landsat satellite data at 100,000 sites across the planet to quantify changes to the boreal forest. He learned that vegetation green-up in colder areas at the taiga's northern edges was about three times more common than browning down on the warmer southern boundaries. We're excited to have Logan here today to talk us through what that means, and what the implications might be. Logan, welcome to Eyes On Earth. 

BERNER:

Thank you, John. It's a pleasure to be here. Thanks for having me. 

HULT:

Yeah. We're excited to talk about this, because we've been seeing suggestions that this was happening. But your paper at actually looks at the data. Let's get started by talking about the boreal forest biome, kind of in general. What is it? Where is it? What do we find there? And what do we know about how it's responded to climate change so far? 

BERNER:

The boreal forest is one of Earth's major terrestrial biomes. In essence, it's a swath, an emerald wreath, of cold-tolerant, primarily coniferous trees, that stretches across Northern Alaska through Canada, and then in Eurasia from Fennoscandia: Finland, Sweden, through Siberia. Siberia holds much of the world's boreal forest, primarily large forests in, in northeastern Siberia. The boreal forest is this massive biome that is the coldest Northern forests on our planet. The boreal forest is unique in, in that is underlain primarily by permafrost. Most trees can't handle living in these cold, long winters and short cool summers. It really filters out a lot of the species that are present, and leaves this community of very cold adapted trees. 

HULT:

Those trees, as I understand it, the roots kind of grow outward to deal with the permafrost. Is that right?

BERNER:

Yeah. Yeah. So the permafrost in some areas can be very shallow, you know, perhaps 10 inches deep, and the trees really aren't able to push their roots very deep into those soils. And so instead they'll tend to spread them out a little bit farther.

HULT:

Very unique sorts of trees survive in this environment. 

BERNER:

Indeed. Yeah. 

HULT:

How have they responded to climate change so far?

BERNER:

As temperatures rise, we see that these permafrost soils, cold soils that have been frozen for a long time, in many areas they're beginning to thaw and that can release nutrients and increase the amount of rooting depth that the trees have access to. But it can also undermine the stability of the landscape and lead trees to fall over and the occurrence of these drunken forests, so to speak. We also see that fires are becoming a lot more frequent and severe, in both north America and Eurasia. And we also have this increase in insect outbreaks. But we also see that there are more direct effects that climate change is having on tree growth in particular. We know, for example, from dendroecological studies, tree ring studies and from others, satellite remote sensing, that forest growth seems to be becoming less limited by temperatures and more limited by water availability. That's especially true along these southern margins of the boreal forest, where conditions can be pretty warm, and often becomes fairly dry during the summers.

HULT:

A lot of change going on in the boreal forests, just in general in the Arctic, there's a lot of change. And, and this sort of tracks with what we understand about what's happening up there in the northern latitudes, seeing a lot of this change. Why is it so important to understand this, for people down here, living in warmer climates? 

BERNER:

There are a lot of different implications that that can have, both locally and globally. There are ways that the boreal forest influences our global climate. For example, the boreal forest stores a tremendous amount of carbon, not only in the trees, but also in these very organic carbon-rich soils. And as climate change unfolds, we see that those changes in carbon cycling could feed back on how the climate is changing. We see more fires occurring that are releasing this old carbon up into the atmosphere. That could cause the atmosphere to become warmer. But at the same time, we see shifts in tree growth and the rates at which trees are pulling carbon out of the atmosphere that could potentially slow down the rate of carbon accumulation in the atmosphere and warming. The boreal forest is also fairly dark in comparison to Arctic Tundra, for example. And so that means that a lot of incoming solar radiation will be absorbed by the planet, which makes the local surface heat up, and can further contribute to warming. It's also important to recognize that there are a lot of people that do live in these northern lands, and as the climate changes, their livelihoods are being impacted. We see that, for example, as the climate has gotten warmer, there has been an expansion of trees and shrubs in parts of these northern margins of the boreal forest. And this has influenced habitat suitability for animals like moose and caribou that are very important for people living in these northern lands. We also have potential impact on natural resources. For example, a significant portion of timber comes out of the boreal forest. And so as the boreal forest continues to change, there might be impacts on timber production. And then lastly, the boreal forest is a major component of Earth's biosphere. It comprises some of the last remaining lands on our planet that are minimally impacted by human activities.

HULT:

So we've got the folks that live up there, the human populations, obviously we don't want to forget about those folks, and their resources and livelihoods are tied to the boreal forest. We have the critters that live up there, the caribou and the moose and whatnot, that need those areas to survive. And most importantly, perhaps, at least from where I'm sitting down here, what happens up north doesn't stay up north. It has impacts on climate change, and ultimately all of us, right? 

BERNER:

Exactly. 

HULT:

Well, tell us about your study. Why did you decide to look into the question of a boreal biome shift specifically? What did you know going in, and what knowledge gaps were you looking to fill here? 

BERNER:

We were interested in investigating the extent to which there are early indications of an emerging boreal forest biome shift. So we think that this should include an increase in vegetation growth and recruitment of trees and shrubs along the cold northern margins of the boreal forest. In essence, where the forest turns into tundra. And we think that those changes should have led to an overall greening of the landscape that we can pick up using long-term satellite observations. On the other hand, along the warm, southern margins of the boreal forest, we expect that higher temperatures could contribute to increasing rates of tree stress and mortality, and that this could lead to kind of a browning of the landscape. And so we wanted to look into whether or not we could really detect these early indications of a boreal biome shift. We knew going into this, that other studies had shown, including some work that we have done in the past, that there have been these patterns of greening and browning, greening being a long-term increase in vegetation, greenness, and browning being the opposite. The long-term decrease in vegetation, greenness as observed by satellites. But a lot of this early work used much coarser spatial resolution satellite observations, such as those provided by the Advanced Very High Resolution Radiometers, or AVHRR. Now, despite the name, they're not high spatial resolution. Each one of those grid cells is about 8-10 kilometers on each side. And so there can be, potentially, a lot of different vegetation dynamics occurring within those grid cells that might compound our ability to detect these indications of an emerging boreal biome shift. There are a lot of wildfires. In the southern parts of the boreal forest, there is a lot of timber harvest and other impacts of people on these landscapes. And those disturbance, and every dynamics can lead to these greening and browning trends observed by satellites that make it difficult to detect early indications of a biome shift. A greening or browning trend could just be due to the impacts of the fire and subsequent vegetation recovery. For our study, we were really interested in trying to minimize the confounding effect that those disturbances cause, and instead focus on recently undisturbed boreal forest. 

HULT:

You're looking at areas that aren't disturbed. You don't want to be thrown off, you don't want your results thrown off by vegetation, responding after a fire. And you're using Landsat data, so you're not looking at 8-10 kilometer pixels, you're looking 30-meter baseball infield-sized pixels. That's how you got the setup. And what did you find? What did your study tell us about how the forest biome has changed over the last two and four decades? 

BERNER:

There's been a widespread increase in vegetation, greenness greening along the cold climatic margins of the boreal forest. So this long-term greening we interpret as an increase in vegetation growth and tree cover, potentially associated with recruitment of trees and shrubs into the Arctic tundra. So this could be an in-filling, for example, of trees in areas that already had some trees. What we found is that this tended to occur in the cold parts of the forest, it where tree cover was very low, but where the, there was a fair bit of soil nitrogen, which we found quite interesting because nitrogen can be very limiting to tree and shrub growth in these northern latitudes, where, as we mentioned before, the soils are very cold and thin. It was compelling that we found greening along these cold northern margins, and especially where there was a lot of nitrogen available that allowed plants to grow more, to put on new leaves and to photosynthesize more, all of which requires having access to nitrogen. These areas, where you hadn't seen trees before, the tree line is moving north. And they were sort of gobbling up that nitrogen that they wouldn't have been able to act access without these changes. There's a good chance that a lot of the greening that we're seeing along these northern margins is related to kind of an in-filling of trees, like the trees becoming thicker. I think the process of trees migrating up farther north into the Arctic tundra is a very slow process. We will likely see that play out over decades to centuries. 

HULT:

So if you were just looking at a Landsat scene over one of these areas where it's sort of thickened up, it would look perhaps like a forest, but you're looking for more trees. It's fascinating to think that you were able to parse those changes, where there actually were trees there, and now there are more, and you were able to figure that out using the Landsat observations.

BERNER:

There's tremendous amount of data processing that goes into all of this activity that lets us insure that we are using these 20-40 years of Landsat observations in a consistent manner through time, which is really important. If you want to look at long-term trends in vegetation, greenness for other properties of Earth's land surface, we focused on these metrics of vegetation, greenness that we know are influenced by a variety of underlying ecological phenomenon. They reflect vegetation productivity and the light absorption by vegetation, which is used to drive photosynthesis and productivity, but they're also inherently influenced by things like tree recruitment. You know, the growth of new trees and new canopy that we can see from the satellites, and from trees dying. These indicators of vegetation greenness provide kind of a holistic indicator of vegetation condition, but it can be challenging to understand what exactly is happening on the ground, what's driving these trends in vegetation greenness, because they can be related to a number of different ecological dynamics. 

HULT:

So what are the next steps? What other questions you're looking at? What other avenues of research does this study open up? I mean, you, you found that on the southern edge, there was some browning, right? So where do we look next? And what sorts of questions are we asking now? 

BERNER:

Yeah, so one of the other interesting things we found was that there's been this long-term browning along the warm southern margins of the boreal forest that is likely associated with not only decreasing tree growth and vegetation and productivity, but also increasing rates of tree mortality. We think that over the longer term, this kind of browning could be the early indication of forest loss, where trees might be replaced by more drought-tolerant shrubs, and grasses that could lead to this northward contraction of the southern margins of the boreal forest. We have a lot of other work currently going on that is somewhat related to this project. We're working with colleagues up in Alaska to explore the causes and extent of tree line migration across the Brooks Range in Northern Alaska, with colleagues who have spent the last couple of summers trekking over 2,000 miles across the Brooks Range, collecting very detailed information on vegetation composition and abundance that we can then link with these long-term satellite observations to better understand what it is that's contributing to this greening, and where greening has occurred within the Brooks Range. We're also looking at the extent to which recent rates of tree mortality can be predicted across Northern Canada and Alaska using Landsat satellite observations and climate data sets. So here we're building complex machine learning algorithms that can be used to try to forecast where trees might die over the next 10 to 20 to 30 years. And we have a number of other activities going on that stem from this research. One of the exciting ones is that as part of this project and some other work focused on greening and browning across the Arctic, we have put together a new open-source software package for Landsat data processing. This hopefully will help a much broader community of ecologists and other non-remote sensing specialists to more readily access process and analyze Landsat time series for sample locations, anywhere on Earth's surface. 

HULT:

Let's talk a little bit about Landsat. How did Landsat the archive contribute to your research? Would it have been possible to do this work without the depth of the archive and the quality that people have come to expect?

BERNER:

So I think this is one of the things that made our study really unique. It allowed us to zoom into this finer scale and focus on these undisturbed areas, which is not really possible using other satellite platforms like MODIS or AVHRR. Some of those provide fairly long-term records, you know, with MODIS going back to about 2000, and AVHRR going back to the early 1980s. But the spatial resolution makes it really challenging to isolate the influence of disturbance versus more direct physiological demographic impacts of climate change. We also capitalized on the dense record of Landsat observations that extends back four decades for the study. We focused on long-term changes in vegetation greenness at about a hundred thousand sampling sites in order to develop the annual time series of vegetation greenness for these sample sites. We had to mine over 42 million multi-spectral measurements, cloud-free, spatial measurements. And so it ended up being half a billion measurements overall to get down to this annual time series for 100,000 sites. From one standpoint, the portion of the boreal forest that we looked at was minuscule. Like, we only pulled out sample data for, I think it ended up being about, if you smashed all the pixels together, it'd be about a 90 square kilometer footprint. You know, for something that is about a 15 million square kilometer domain. But it's a large statistical sample. And so it allows us to make inferences about these macro-ecological changes that are occurring across the domain. 

HULT:

Great. Great. Logan, do you have any closing thoughts you'd like to leave us with, about the boreal forest, or the Landsat archive, or anything else that's on your mind?

BERNER:

As time goes on, we will likely see more and more impacts of climate change on the boreal forest and Arctic tundra. We'll likely see more fires, we'll see more heat waves and droughts and other atmospheric impacts that could really have unanticipated consequences for the boreal forest. I think it's really important that we improve our network of ground monitoring sites, as well as further develop satellite monitoring capabilities that allow us to understand how and why the forest is changing. And then all of these observations from the field and from satellites can be used to help develop and validate ecological models that we can use to look at how future changes in climate might impact the boreal forest going forward. And that'll give us a better understanding, not only of what has happened over recent decades, but where else we might be going into the future. 

HULT:

We've been talking to Logan Berner of Northern Arizona University about changes to the boreal forest biome. This has been a fascinating conversation. Logan, thank you so much for joining us. 

BERNER:

Thanks so much for having me.

HULT:

And thank you to the listeners, as well for joining us. Be sure to join us next time as well, to learn more about satellites, remote sensing, land change, and so much more. You can find all our shows on our website, usgs.gov/eros. That's U-S-G-S dot G-O-V, forward slash E-R-O-S. Or you can subscribe to us on Google podcasts. This podcast is a product of the U.S. Geological Survey, Department of Interior.