current issue web exclusives blog multimedia archive subscribe advertisers
Feature Articles
Editor's Note
Audubon View
Field Notes
Audubon Family
Audubon In Action
Green Guru
One Picture
Bookmark and Share

Global Warming/Dispatch
Smoke Signals
There’s more carbon locked up in the high north’s permafrost than the combined total released into the atmosphere by humans. The big melt going on right now in the Arctic could trigger a train of catastrophic events across the world.


Standing on the shore in Deadhorse, Alaska, looking out on the Arctic Ocean, a thin white line on the horizon divides the dark gray-blue water and the light blue sky. Sea ice. It’s late June, and the frozen mass is making its annual retreat here in the land of the midnight sun, a place that may in the next decade experience ice-free summers. A few days earlier oil company workers had spotted a polar bear and her two cubs nearby on the rocky shore—an increasingly common sight as sea ice thins and shrinks and the animals resort to scavenging for food on land. Although these are the images that a warming Arctic typically conjures, dwindling sea ice and starving polar bears are only part of the story. Rising temperatures are already melting the Arctic’s permafrost, and an all-out failure of the frozen foundation would not only irrevocably change this fragile landscape but also speed up warming across the earth.

As we drive south from Deadhorse, the flat coastal plains, a summer haven for more than 150 bird species, climb into the rolling foothills of Alaska’s North Slope. Here, about 120 miles south of the Arctic Ocean and 160 miles north of the Arctic Circle, scientists at Toolik Field Station have been studying the land, waterways, and wildlife for decades. I’m here on a Marine Biological Laboratory’s Logan Science Journalism Fellowship to learn from scientists who have had a front-row seat to the changes already under way in one of the fastest-warming places on the planet.

Take the lightning—a rare phenomenon in 1975, when the station was founded. Now, on any given summer afternoon, piles of thick cumulonimbus clouds build, obscuring the Brooks Range’s snow-covered peaks and releasing bright flashes that light up the darkened sky. Scientists have seen lightning strikes increase 20-fold with warmer temperatures, sparking previously unheard-of wildfires in the tundra (see “The Hottest Spot,” November 2003). The fires set in motion changes that thaw permafrost, a process well on its way as the region heats up. Permafrost holds water at the surface and creates aquatic habitat in what would otherwise be an arid environment. Unabated melting will drastically reshape the North Slope landscape and have profound effects on fish and wildlife. When ice in the permanently frozen ground liquefies—whether started by large disturbances like fire or small ones like a trickle of water—the ground collapses, forming enormous gullies that can be hundreds of feet wide and long and tens of feet deep. The number of these scars, called thermokarsts, is growing here and throughout the Arctic, freeing up carbon long locked away as the organic matter thaws and decomposes.

The Arctic’s soil and permafrost hold nearly twice as much carbon as the earth’s atmosphere, dwarfing the amount of carbon emitted to date by burning fossil fuels. Since the industrial revolution our dependence on coal and oil has ratcheted up the atmosphere’s carbon content, from 560 to 760 gigatons. Permafrost holds an estimated 1,400 gigatons of carbon. In addition to carbon dioxide, the frozen source is releasing methane, a greenhouse gas 25 times more potent, though it stays in the atmosphere for only a decade rather than for millennia. The gas is bubbling up from land and also from a large, previously overlooked source: permafrost submerged beneath the Arctic Ocean. It’s too early to say whether this is a newly observed steady leak, or if it signals the beginning of a flood of methane. But it underscores how important it is to understand whether thermokarsts, fire, or other mechanisms will unleash this stockpile from the frozen north. When it comes to contributors to climate change, we usually talk about coal-fired power plants or deforestation in the Amazon. But sitting at the top of the world is a force for change potentially more powerful than any other ever seen.


It might seem strange to describe a landscape that endures such harsh weather as “delicate,” but the tundra hangs in a precarious balance. North of Toolik, for instance, two dark, parallel tracks extend for miles across the tundra. A single vehicle that drove over the bumpy terrain in the 1940s made these marks. More than a half-century later, they’re still visible. “The Arctic is a place where everything is about the minutest change in energy balance,” says Breck Bowden, a University of Vermont aquatic ecologist. “All it takes is for a vehicle to crush the vegetation, change the albedo—the amount of light that’s being reflected away from the surface—make it a little bit darker because the soil is coming through. More energy is absorbed, and that area just continues to melt. And as it continues to melt, a new community of vegetation comes to grow in there, and we now have a permanent scar across the landscape.” 

Today researchers are seeing wounds across the tundra. Bowden and Michael Gooseff, an environmental engineer at Penn State University, first happened upon a thermokarst near Toolik by accident in 2003. Scouting the Kuparuk, a clear blue river that snakes its way through tundra east of the station, something unusual caught their attention. “We noticed that one of the tributaries was really, really turbid,” says Gooseff. “There was lots of muddy water.”

Vegetation grows on the “active layer,” the top section of the tundra that thaws during the summer and may be only a foot and a half deep; the underlying permafrost remains frozen year-round and can be hundreds of feet deep. “After a big rainstorm, you shouldn’t see a whole bunch of sediment running down a stream because there just isn’t much to give up,” says Gooseff. The scientists followed the muddy water upstream about 20 miles until they found what has since been named the Toolik River thermokarst. An underground ice wedge had melted, creating a tunnel as the water ran off. Then, perhaps a few days before Gooseff and Bowden came across it, the tunnel collapsed, forming a deep gully about 150 feet long with a waterfall at its head. “We have a great picture of Breck standing in the thermokarst on a rafted piece of tundra. Breck is, what, six-two, and he’s dwarfed by this waterfall.”

When thermokarsts occur near lakes and rivers, they inject sediment and nutrients into waterways, potentially altering the aquatic ecology. Even if they crop up on hills far from streams, the movement of tons of soil, and the resulting release of carbon, nitrogen, and phosphorous, creates gulches, thus changing water flow and plant species.

Thermokarst features near Toolik Lake, Alaska. Locations with names and black triangles are the subject of intensive study. Black dots are other thermokarst feature locations.
Courtesy of Andrew Balser

Thermokarsts are a natural part of the tundra landscape, but the discovery in 2003 prompted scientists to wonder if there really were more of them taking shape, or if they were just noticing them more. So they compared aerial photos of the area from 1985 and 2006. “There are actually more of them, and they’re forming at a greater rate,” says Bowden, who is interested in how sediment and runoff change water chemistry. “We think there are significantly more thermokarsts now than in the past.” Bowden is overseeing a first-of-its-kind, five-year, $5 million project to study thermokarsts on the North Slope, funded by the National Science Foundation’s Office of Polar Programs.

To get to the three core thermokarst sites near Toolik, there are two options: fly or walk. On a day when Gooseff’s colleagues have called dibs on the helicopter, we don mosquito nets, rain jackets, and rubber boots and hike three miles to a lake called NE-14. After nearly an hour and a half, we crest a final hill and see the lake and the two thermokarsts on its shore, one old and one new. Both are horseshoe-shaped with the ends pointing toward the lake, middle collapsed. The 30-year-old one to the west is “healed.” It’s no longer expanding, and its bottom and walls are covered in thick willow bushes (the formations stop spreading as steep sides gradually give way to gentle slopes). The thermokarst to the east is “active”—an apt description, since as we walk its perimeter, a suitcase-size chunk of tundra gives way, landing with a loud plop in the mud below.

We slide down the eroding slope, and Gooseff, who focuses on modeling how these structures form and how long they grow, steps expertly from one overturned tussock to another to avoid sinking into the thigh-deep mud. Permafrost looks like brown concrete and is frozen solid. When it melts, the result is a goopy mess. Moving away from the walls toward the drier center, Gooseff points to fresh moose, grizzly, and wolf tracks. Suddenly a wolf appears, perhaps startled out of its den by our presence. As I fumble for my camera, Gooseff starts howling—not, as I first thought, to lure the wolf back but rather in an attempt to draw the other scientists’ attention to it. But the four, squished into a two-person tent to avoid the swarming mosquitos, are oblivious as they pore over data.

Down in the thermokarst, Gooseff installs sensors to measure soil moisture. By scattering these and ground-temperature sensors at various depths, he’ll develop a profile of how subsurface temperature changes and how water moves through the soil. He’s also tracking other variables—wind speed and direction, barometric pressure, humidity, and air temperature—with meteorological stations at each site. Eventually he’ll plug these measurements into models to see what conditions cause the landscape to fail. “On the one hand,” says Bowden, thermokarsts are “an interesting indicator that things are changing in the Arctic. On the other hand, it raises some questions about, well, if we accelerate this process, what does it mean for the landscape processes in the Arctic?”


The cranberries growing on the tundra can be easily overlooked. Gazing out, the vast, rolling expanse appears to be a homogeneous sea of cotton grass. But crouch down and a rich vegetative mix emerges. The cotton grass is obvious because it grows on the tussocks that cover the tundra like raised paving stones. About eight inches tall and wide, the wobbly growths make any trek across the tundra a challenging feat. “Step on the top and break an ankle, or step in between and break an ankle,” Bowden says wryly. Growing around the tussocks, among lichens and mosses, is an array of familiar plants in stunted sizes, perhaps a few inches tall: birch, willow, bog rosemary, rhododendron, blueberry, cranberry. At a glance, the baby-pea-sized cranberries hardly seem worth the effort a 700-pound grizzly bear must have to exert to get its fill. But just one taste of the tart yet sweet burst of flavor makes plopping down to hunt for more seem like a splendid way to spend a sun-drenched afternoon.

On the hill above Toolik Field Station, Gaius Shaver has been running experiments for more than 30 years to determine how fertilizer changes the tussock tundra’s floral medley. His test plots, north of Toolik Lake, are accessible only by a mazelike wooden walkway that, at points, is little more than a web of two-by-sixes. As a rule, nobody ventures off the walkway, lest they crush someone else’s experiment. On an overcast morning Shaver, clad in mud boots, a Carhartt jacket, and blue jeans, confidently makes his way along the slippery walkway and up to his plots. These are very different from the nearby untouched tundra. Because nitrogen and phosphorous are scarce in the frozen tundra in forms useful to plants, yet crucial to their growth, Shaver wanted to see what would happen if he added the nutrients to small plots. It’s been more than two decades since he started the study, which has transformed the plots from calf-high grass-dominated tussocks to knee-high shrub-covered patches. In one section covered with clear plastic—a makeshift greenhouse—birches are thriving and waist high, but few, if any, other species abound. The clear lesson: When more food is available, shrub growth takes off, shading out plants like cranberries and cotton grass.

As the North Slope has warmed during the past half century, it has become shrubbier. While this rise in woody vegetation has been a boon for songbirds and moose—mammals once scarce as far north as Toolik, now regularly spotted there—it may help spur warming. “Shrubs tend to feed back into the local and regional climate,” says Michelle Mack, a University of Florida plant ecologist. During the summer, birch, willow, and alder reflect solar radiation, driving up atmospheric temperature. In winter, tall, branched woody plants trap snow that would otherwise blow across the tundra. This actually keeps the ground warmer and may allow soil microbes to remain active for longer, cycling nutrients (providing food for shrub growth) and releasing greenhouse gases.

Disturbances like thermokarsts could also expand shrub cover. As with the older thermokarst at lake NE-14, decades after these features form “you see shrubs, not tussocks,” says Mack. Once shrubs move in, she adds, it could be difficult for lost permafrost to become reestablished.

More woody material also means more carbon is being stored above ground, instead of in the soils. While plants may trap carbon for hundreds of years before it’s cycled back into the atmosphere, permafrost can store it for tens of thousands of years.

Since trees act as carbon sinks, soaking up CO2, a shrubbier tundra might seem like a plus—but it actually exhales more carbon than tussock tundra. Although the plants in Shaver’s fertilized plots have more biomass and take up more carbon dioxide than those in the control plots, that gain is offset by the loss of carbon and nitrogen from deep soils; the nutrient-rich plots saw a net loss of nearly 2,000 grams of carbon per square meter over 20 years, Mack and colleagues reported in Nature in 2004. “The tundra is moving toward a shrubbier community, which means it will hold less carbon overall,” says Mack. The release of that carbon into the atmosphere, in turn, will feed back into more warming.

All of that wood creates more potential kindling. “We know that 12,000 years ago, when the tundra was more shrubby, there were more fires,” says Mack. Now, for the first time since the woolly mammoth and sabertooth tiger went extinct, big wildfires are again raging on the North Slope.


“As an undergrad I took an Arctic seminar and was taught that there are no cumulus clouds in the Arctic, no thunderstorms,” John Hobbie, a founder of Toolik Field Station, recalls of his education in the 1950s. Today in the warmer modern Arctic, thunderstorms are regular summer events. “One day this year we had 272 lightning strikes on the North Slope,” says Syndonia Bret-Harte, an ecologist at the University of Alaska-Fairbanks and Toolik’s associate science director. Most of the flashes that hit the wet tundra don’t catch fire.

But the summer of 2007 was an exceptionally dry year, and in July a strike hit the grass near Toolik, smoldered for a few weeks, then exploded. Bret-Harte recalls watching the fire from camp. “When the wind was not blowing, you could see this big wall of smoke,” she says, “and that was awesome and beautiful but disturbing at the same time. When the wind shifted and brought the smoke into camp, then it was like being in this thick, acrid fog. It was gross.” The Anaktuvuk River fire, the largest ever recorded on the North Slope, burned until October, ultimately consuming nearly 350 square miles and releasing 2.2 million metric tons of carbon into the atmosphere—the amount the entire country of Barbados emits annually.

Walking around the severely burned area two years later, the previously scorched earth dusts my boots and pant legs. Already some cotton grass has come back, and the green stems and white tufts growing out of singed tussocks contrast sharply with the blackened terrain. Bret-Harte estimates that plants now cover about half of the ground. But, she adds, “I haven’t seen any of the normal mosses and lichens coming back.”

While Bret-Harte studies how fire changes the makeup and abundance of plant species, Adrian Rocha is exploring the carbon balance. Rocha, a physiological ecologist at the Marine Biological Laboratory, has installed equipment in severely and moderately burned areas, and in an undisturbed area, that records carbon dioxide flux, among other data. He found that the most severely burned tundra emitted roughly twice as much carbon as undisturbed tundra typically absorbs. Mack, who’s also working on the fire project overseen by Shaver, calculated that the fire consumed as much as the top 20 centimeters of organic soil, and that the carbon in that material was, on average, 35 years old. Initially, the researchers thought the carbon released by the fire might have been hundreds of years old, so the fact that it had been pulled out of the atmosphere only about a quarter-century ago was surprising, says Mack. “That made me think, maybe these tundra systems are more resilient than we think.”

Still, warmer, drier summers and more shrubs will likely mean more fires. And just two years after the Anaktuvuk River fire, the researchers are already witnessing major changes in the burn sites. “We’re seeing all these thermokarsts out there,” says Bret-Harte. “Disturbing the surface that much, as you do with a burn, and making it black and absorbing radiation, is bound to make a lot of ice start to melt.”


As the Arctic heats up and permafrost thaws, nutrients are seeping into waterways. Combined with warmer, drier, longer summers, these changes could spell trouble for aquatic dwellers. Parking a 14-passenger van laden with waders and equipment to tag fish, biologist Linda Deegan hops out and exclaims, “Look, it’s a glaucous gull!” pointing to an enormous white bird with perhaps a five-foot wingspan tracing the S curves of the Kuparuk River. “Maybe the grayling are here.” For 20 years Deegan has been charting their spring and fall migration. Closely related to salmon, Arctic grayling winter in lakes or rivers that don’t freeze solid, then head to shallower streams in the spring to spawn.

Like Shaver did with his vegetation plots, Deegan’s team, starting in 1985, manipulated the aquatic ecosystem, dripping phosphate into a section of the Kuparuk. The fertilizer fostered more aquatic plants, which supported more insects and thus provided more for the grayling to feed on. So grayling could benefit from bursts of nutrients that drain into waterways after fires. But Deegan doesn’t expect that to happen in the long run.

Along with fertilizer, fires and thermokarsts dump sediment into waterways, potentially smothering aquatic life. Even if insect populations grow, their hatches and grayling migration—which may be cued by light, not temperature—might be thrown out of sync, leaving the fish little to eat. And as the region heats up, water bodies will warm; if water temperatures rise too much, the cold-loving grayling “shut down,” says Deegan. “They become lethargic and just settle on the bottom and won’t move, even if there’s food available.” If droughts become more prevalent, as predicted, stream sections could dry up, stranding the fish. That happened on the Kuparuk in 2003, and the grayling population Deegan has been studying for two decades still hasn’t recovered. Before the drought, in a typical two-week window during fall migration, her team would count between 3,000 and 5,000 individuals. In 2009 they counted only 315. “I think my poor grayling are in deep trouble,” says Deegan. 

The animals that depend on grayling could face their own crisis. Lake trout devour grayling in mountain lakes. Grizzly bears, hungry after the long winter, dine on grayling in the spring. At that time of year the fish are also a staple of Native Americans’ diets. Plummeting grayling numbers could hurt fish-eating migratory birds, too, including red-throated loons and red-breasted mergansers. The grayling’s plight in the Kuparuk serves as a good barometer for other Arctic populations, says Deegan. “Grayling are the dominant freshwater fish around the pole. I’m sure there are places with populations that are doing fine. But we don’t think this is restricted to just the Kuparuk. We have evidence from another stream in the same region where this habitat fragmentation and separation from overwintering areas has been happening because of drought.” 


The vastness of the tundra is nearly overwhelming from the air. Rising from the helicopter landing pad at Toolik, I look out on the rolling hills and braided waterways. After a few minutes, any signs of human presence disappear. I’m struck by how our influences from far away are altering this remote place, and that if the warming/melting cycle continues, we’ll feel the influence of the Arctic. Droughts and floods will increase, water supplies will diminish as glaciers and snow cover decline, and coasts will erode as sea level rises. Nobody knows precisely how fast permafrost will melt or how quickly it will let loose the carbon it holds. It could happen over hundreds of years, or thousands.

There’s no easy fix for slowing the release of carbon from permafrost. We can’t put a lid over the tundra or tinker our way out of the problem. It’s a cycle we can’t control. Where we can make a difference, says Bret-Harte, is at home. “In theory we can make a treaty to say let’s not burn fossil fuels anymore. We can’t make a treaty to stop thermokarsts and fires.”

Back to Top