As the name would suggest, permafrost in the far north was long considered to be, well, more or less permanent, emphasizing the stability of the Earth system.

But more recently, researchers have come to realize that this frozen ground is steadily disappearing, with the full consequences of this situation still being worked out for the global carbon budget. Some have likened permafrost to a bank account that has been growing fatter during our planet’s cold periods; now, however, that account is being withdrawn faster than it is being added to. To make matters worse, the pace of withdrawal is accelerating.

To deal with the problem, the first thing that researchers must do is get an idea of the current account.

Luckily, it turns out that scientists at the University of Alaska-Fairbanks have been monitoring the state of the region’s permafrost since the mid-1970s, when the first camps and airstrips were being built for what became the 800-mile long trans-Alaska oil pipepline. Researchers realized that these sites would be great, relatively easily accessed sites to drill deep boreholes and drop in devices to monitor the temperature of the earth underground. Installed in the early ‘80s, these sites continue to be monitored to this day by geophysicists such as the university’s Vladimir Romanovsky—along with new ones that he has installed over the last 30-plus years in Alaska and parts of his native Russia.

In this interview with me, Romanovsky describes what he has found, explains some of the limitations of climate modeling, and tells of what he has learned from a lifetime spent monitoring the frozen north. He delves into what he thinks would cause the system to accelerate from gradual thaw to abrupt thaw, and what the consequences would be.

Romanovsky concentrates especially on the tipping point of the Boreal permafrost zone—meaning a forest that grows in regions of the northern hemisphere with cold temperatures, and is made of cold-tolerant species. (The name comes from the ancient Greek god “Boreas,” who was the “bringer of winter” and the god of the Cold North Wind. The Wildlife Conservation Society ways says that this region contains some of the most extensive, intact ecosystems on Earth.)

Officially retired, Romanovsky spoke with me after he had just flown from Alaska to Washington, DC, to present his latest findings at a geophysicists’ conference.

(Editor’s note: This interview has been condensed and edited for brevity and clarity.)

Dan Drollette Jr.: You’re an emeritus professor with the Permafrost Laboratory.[1] Can you tell me what the lab is, and what you folks do?

Vladimir Romanovsky: Well, it was created here at the University of Alaska-Fairbanks, with the aid of professor Tom Osterkamp. He was one of the first physicists or geophysicists at the university’s Geophysical Institute—and he was my adviser also.

We study permafrost from the geophysical point of view. We measure temperature in the ground, in as many locations as we can—and temperature is the key. If you know ground temperature, you can tell if you’re dealing with permafrost or not, and then record if it’s constantly below zero or changes. You can see the dynamic of how the permafrost reacts to climate change. And when you see these dynamics, you can project changes into the future—so, this record not just explains changes that are happening already, but can help to project them into the future. That’s what we do.

Drollette: Before we go any further, we should probably define some terms. What’s a good explanation of permafrost for us non-experts?

Romanovsky: Permafrost is not just ice but a loose conglomeration of ice, soil, organic matter, and rocks—a mix of ice and what we call “earth materials.” The simple, loose definition of permafrost is that it’s earth material which has been continuously frozen for two or more years.

And by continuously frozen, I mean it doesn’t flow during the summer time.

The geophysicist’s definition is a bit more specific: Permafrost is any earth material at or below zero degrees Celsius for two years. By this definition, it’s kind of closely related to the concept of “frozen”— but sometimes there is no water involved, for example, so then it can’t really be called frozen because there is no ice.

Or perhaps the sample contains some chemical composition in the groundwater so that it is not pure water—for example, maybe there’s some salinity involved. Under those conditions, then the permafrost could happen below zero degrees Celsius.

And by geophysical definition, permafrost could even involve no ice at all—it could be like bedrock, if the bedrock is below zero degrees Celsius.

But we still call something like that “permafrost”—even if there is no ice—because the properties are pretty much the same.

Drollette: And what is boreal forest?

Vladimir Romanovsky: I would define boreal forest as a forest that grows in regions of the northern hemisphere that have cold temperatures; it is composed of a limited number of cold-tolerant species such as spruces, larches, pines, and firs. It has no warm-loving, big, wide-leaf trees like oaks, maples, or lindens—even when summers in the boreal forest sometimes get pretty warm. The key thing is that the winters are extremely cold, and that limits the species there to trees that can survive that kind of deep cold.

I don’t have the exact numbers at my fingertips, but the boreal forest is widespread: Most of Siberia is boreal forest, except for the northern part, where it becomes tundra[2]. Most of Alaska is covered in this boreal forest, except for the tundra in the north and a little bit of a different type of forest in the south of Alaska. And a big chunk of Canada’s forest is boreal.

Consequently, the boreal forest is very big—one of the largest intact ecosystems on Earth.

Drollette: Is boreal permafrost fundamentally different from other kinds of permafrost?

Romanovsky: No, it’s just permafrost in the boreal zone.

Although there are some differences in the soils—the vegetation and so on. But it’s really just more a matter of location—a geographical distribution.

Drollette: The reason I bring this up is that from the background reading I’ve done, it sounds like people are very concerned about the heating-up of the boreal permafrost right now.

Romanovsky: Well, ultimately that’s a concern about any kind of permafrost.

But boreal permafrost may likely be more susceptible to change, largely because the boreal zone is further south, where the ground temperature is generally a bit warmer already—closer to zero degrees Celsius. And zero degrees Celsius is important, because that is the natural threshold for permafrost to switch from permafrost to non-permafrost. At that point, the ice which is so-often present in the permafrost starts to melt, and the permafrost thaws.

That’s why many people believe that the boreal zone is more vulnerable to climate change—why the thawing of the permafrost in the boreal zone will start sooner than permafrost in a tundra zone.

It’s generally true, although the details get more complicated; there’s a lot of variability. For example, there are some tundra zones present in much lower latitudes as well, such as in Alaska’s Yukon-Kuskokwim Delta[3]. Though it’s fairly well south, that region is kind of a transitional area, with some Arctic tundra and some boreal forest.

But the big point is that the permafrost there is very warm, something like only one degree Celsius from starting to melt, so the vulnerability is more.

Bottom line: The best indication of vulnerability is the temperature—how close it is to the zero degrees threshold.

Drollette: Is it true that the Arctic[4] is warming three or four times faster than the rest of the planet?

Romanovsky: Yes, that’s true, and I will be talking about it tomorrow at the American Geophysical Union’s Fall 2024 meeting. I will be presenting my data from northern Alaska—a continuous record of ground temperature and air temperature for the last 40 years.

And the data shows that the trends in the northern part of this region—in the tundra—have been for a rate of warming of one-degree per decade for the last 40 years.

In other words, for every 10 years that go by, there’s a one-degree Celsius increase in air temperature. So after 40 years, there’s been a four-degree Celsius increase in the Arctic, or roughly about four times the global average temperature increase.

Of course there’s some variability: Sometimes it’s a little bit warmer, sometimes a little bit cooler. But 40 years is a good span of time to look for trends.

Drollette: Is there a theoretical maximum to how old permafrost can get? It seems like there’s always accounts about someone finding the newly melted remains of a mammoth[5] that had been frozen 30,000 years ago.

Romanovsky: Permafrost can be very, very old, particularly in Siberia. There are some estimates that the permafrost in parts of the Siberian north could be up to 2 million years old. In Antarctica, it’s possible the permafrost could be even older, because we know that in Antarctica, the ice sheets started to form about 40 million years ago. So there could be some places there where the permafrost developed around that time, and it has been sitting there ever since.

But most of the boreal permafrost is much younger; the typical permafrost age may be less than 100,000 years.

And in some locations, it could be just a few thousand years or less. For instance, we know that some of the permafrost is fairly recent and kind of shallow, the result of the Little Ice Age—which only lasted for something like 300 or 400 years, depending on who you talk to. As a result, it’s not a very thick layer, and this particular permafrost layer is now very actively thawing—sort of like last-hired, first-fired.

So, the youngest permafrost is thawing right now, and with further climate warming, then the older and older permafrost will get involved.

Drollette: A Nature Geoscience article[6] from about five years ago said that some researchers are expecting the abrupt thaw of some permafrost—and that probably something like 20 percent of the permafrost zone could be involved, melting much more rapidly than expected.

Romanovsky: Well, we can argue about how large an area will be affected, and what “abrupt” means.

What we’re seeing is pretty much that air temperature and ground surface temperature are increasing steadily. We see the winter temperature ate increasing faster than the summer temperatures—but summer temperature is increasing too. And because of that, the layer above the permafrost—which we call the “active layer”—is thawing every summer but then completely refreezing during the winter.

This layer is not considered part of the permafrost because it’s not frozen for two or more years; it’s only frozen in the cold part of the year. Because it’s not frozen in the warmer part of the year, we call it the “active layer” or the “active thermal layer”—there are different names, but it’s to distinguish it from permafrost, and the depth of this layer is about maybe from a half-meter to a meter, typically.

And with warmer summers, this layer is getting thicker and thicker—meaning that the upper part of the ground is thawing. Fifty years ago, this layer was 50 centimeters thick. Now, the summer thaw is 70 centimeters, meaning it’s extended 20 centimeters into what is formerly permafrost and transformed it into the active layer. At some point, the active layer will become so thick that it cannot be re-frozen during the winter, especially as the winter is getting warmer really fast—and in this case, a new layer forms between this active layer and permafrost, which does not freeze at all during the whole year, known as the “talik.” It’s a Russian word, actually meaning “not frozen.”

We estimate that the talik will be growing bigger and bigger in the future; by the second half of this century, this talik will be growing in many places at something like 10 centimeters per decade. So it’s kind of slow, and not really abrupt, but it’s happening.

However, there are some other practices which can thaw permafrost much faster, making for what I consider to be truly abrupt permafrost thaw—including surface processes which expose more and more permafrost to the impact of a warming atmosphere. It’s usually in the form of erosion: Every year, earth material flows; it’s not sitting in the same place, but removed by the impact of things such as water movement. Under these kinds of processes, in which erosion exposes new permafrost to warmer conditions, the rate of thawing of permafrost can be very fast—up to a meter or even a meter-and-a-half per year, or about 10 times faster than this other process.

These surface processes—in which the land is sliding down and exposing newer and newer layers of permafrost to the summer’s warm conditions—are kind of common, but they’re not everywhere. So that means that this abrupt thaw could happen only in some limited areas, and that’s where estimates are needed—estimates of how much of permafrost there is, how much ice is in that particular area’s permafrost, and how large is the potential area of this process. And that’s where there are still some arguments about abrupt thaw.

So when you see news articles about permafrost melting, and dramatic pictures of cliffs falling down, they are indeed good examples of some of these abrupt processes—but it doesn’t mean those things are happening everywhere, all across Alaska and Siberia. It’s just in a few places, usually on the banks of large rivers or the coast.

Although, with warmer climate, it may start to happen not just on the banks of big rivers, but also on small creeks.

And much less steep cliffs are starting to show signs of this happening, it’s starting to happen even on more gentle slopes farther removed from rivers. A good example of this is Denali National Park, where these landslides are start to happening much more often.

Drollette: Do trees help to prevent this from happening, by anchoring the ground with their roots and by providing shade?

Romanovsky: Well, trees are actually not really great protectors of permafrost. Because direct warming from the sun is not really a part of this.

Actually, layers of moss and organic peat are much better protectors of permafrost—and they are exactly typical features of the boreal forest.

They are what protects the permafrost in the boreal forest—much more than the trees themselves.

Drollette: Okay, so, it’s the moss that people should be worried about.

Romanovsky: Yes, moss and peat keep the permafrost more stable; there are very well understood physical practices behind it. As a geophysicist, I know what practices are involved. It’s maybe a little bit more complicated for lay readers, but moss and peat can be thought of as good insulation against the heat of the summer.

Essentially, to over-simplify a bit: What happens is that the thermal conductivity of ice is four times that of the thermal conductivity of water. So in the winter, the frozen moss and organic matter lets heat escape from the ground into the air more easily—and when it turns watery in the summer, that same wet moss and organic matter prevent the heat from the atmosphere from easily penetrating into the permafrost.

And what we are getting from our measurements bears this out: If you look at the mean annual temperature at the ground surface in the boreal forest versus what it’s like at one half-meter or one-meter depth below, it can be two or three degrees Celsius colder down below.

The cooling effect of this organic layer is very significant. And in a boreal forest, it’s very typical to have this protective layer. That’s why the boreal forest has increased stability for its permafrost—it’s not because of the trees, but because of what’s happening with the moss and peat associated with the trees, right below the ground surface.

Drollette: That’s fascinating—how what looks like subtle differences can have major effects. And that sort of relates to an impression I’ve been getting: It sounds like climate models in the past didn’t really take into account abrupt thaw of any kind; they were really looking at much more gradual change.

Romanovsky: And if you look a little bit farther back, they weren’t even considering the thawing of permafrost as any important process for the climate change. Twenty or 25 years ago, there really was no permafrost in the models of climate change. Researchers didn’t look at permafrost; they didn’t consider how it could potentially release carbon into the atmosphere when it thaws—regardless of whether it thawed gradually or abruptly. They just didn’t take melting permafrost into account. Then, around 2007, came the first estimates of how much carbon is sequestered in permafrost globally, and the number was surprisingly big—about twice what is in the atmosphere right now.

And now some people say there’s three times more carbon in permafrost than in the atmosphere—which is a big number. If only part of this carbon ended up in the atmosphere as greenhouse gasses like methane or carbon dioxide, then it will have a huge impact in terms of more global warming.

And in those first, 2007-era models, it was just simple—a look at long-term thawing of permafrost because of warmer climate. Now we are talking about not only should the process of thawing be included in the models, but also how abrupt thaw of permafrost should be included.

Because there really is no good representation of abrupt thaw in the models now.

And that’s understandable, because it’s much more difficult to include that in the global climate models: For one thing, the spatial resolution of these models is at best something like 100 kilometers. Meanwhile, this abrupt process is happening more on the scale of about 10-, 15-, or 20- meters.

And on this 10- to 15- to 20- meter scale, ice is melting, water is moving, and material is flowing. Consequently, things like slides occur, banks collapse, and sinkholes form with these water bodies on top. Sometimes these sinkholes collapse—what’s called a thermal karst.

But you can’t really see these physical processes when your grid cell is 100 by 100 kilometers. This problem of scale is called “parameterization,” it’s probably the biggest problem for climate models to overcome.  Abrupt thaw is much, much smaller than the resolution of their models, so it’s hard to factor in.

But people are working on it, and there is some progress.

Drollette: Any last comments before we sign off?

Romanovsky: The interplay between carbon emissions and permafrost is very important on a global scale, and that’s mostly what we’ve been talking about here.

But for people who live on permafrost, probably the more important thing is how the melting of the permafrost will impact their infrastructure now—their houses, road, bridges, pipelines and other structures. For them, the most urgent issue is to figure out how to deal with all this. I mean, when the ground under your house has collapsed, what do you do? Do you keep living there or just leave?

The number of people living in the permafrost region may be small—about 5 million people in the northern hemisphere. Compared to the 8 billion human beings who live on the planet, it is not that big a number. But for the people who live there, that’s their home, their life, their environment. And most of the people who live there want to continue to live there. They don’t want to be forced to leave.

And that’s a big challenge, because much of the infrastructure was built on permafrost.

Admittedly, with some expensive new engineering approaches, it may be possible to delay some of the worst effects of the thawing, and save some of the roads, highways, airports, gas pipelines, and electric lines.

But with this kind of warming, those kind of engineering solutions may be not sustainable. You can delay by a good bit—maybe by a matter of decades, and maybe even by as much as 50 years. But sooner or later, if the climate continues to warm, then you will not be able to keep living there, even with all these engineering solutions.

And for some villages in Alaska, this is already a big deal already, happening right now—but they just don’t have the money.

So, that’s another big question. It’s not just a matter of geophysics. It’s actually a social question, an economics question.

They are in big trouble right now, and they don’t know what to do.

Endnotes

[1] For more, see https://www.uaf.edu/news/thirty-years-on-semi-solid-ground.php

[2] Tundra—sometimes referred to as the “cold desert”—is the flat, treeless region of the Arctic, with permanently frozen subsoil as opposed to frozen for a two-year minimum. Tundra is dominated by mosses, lichens, dwarf shrubs, and herbs. In comparison, the boreal forest contains trees.

[3] The tundra is usually found at higher latitudes, sometimes as far as 75 degrees north. Because the Yukon-Kuskokwim river delta is found at 61 degrees, it’s at the lower extreme.

[4] According to Wikipedia, there are several definitions of the term “Arctic” and what area is contained within it: “The area can be defined as north of the Arctic Circle (about 66° 34’N), (which is) the approximate southern limit of the midnight sun and the polar night. Another definition of the Arctic, which is popular with ecologists, is the region in the Northern Hemisphere where the average temperature for the warmest month (July) is below 10 °C (50 °F).” The northernmost tree line roughly follows the southern boundary of this region.

[5] See “Well-Preserved, 30,000-Year-Old Baby Woolly Mammoth Emerges From Yukon Permafrost” in the July 2022 issue of Smithsonian magazine at https://www.smithsonianmag.com/history/well-preserved-30000-year-old-baby-woolly-mammoth-emerges-from-yukon-permafrost-180980388/

[6] See “Carbon release through abrupt permafrost thaw” in the February 3, 2020 issue of Nature Geoscience, by Merritt Turetsky et al, at https://www.nature.com/articles/s41561-019-0526-0