Wahlenbergbreen glacier in Svalbard, Norway, is seen in September 2013, as its flow began speeding up. At the peak of the surge, beginning in 2015, the ice advanced at 9 meters per day, bulldozing everything in its path. A glacier’s shape, the nature of its bed, and climate change can interact to trigger normally quiescent glaciers to surge.

HEÏDI SEVESTRE

Why slow glaciers can sometimes surge as fast as a speeding train—wiping out people in their path

ARU, TIBET, AND SVALBARD ARCHIPELAGO—On the high plateau of western Tibet, 17 July 2016 started out as a lovely day. "The air smelt particularly fresh after a heavy rainfall the night before," says Dradül, the chief of Aru village, who like many Tibetans goes by one name. Then, Dradül got a chilling phone call. In a torrent of words, a villager described how an avalanche of ice had just "swallowed the grassland," wiping out a rich pasture where villagers, including some of Dradül's relatives, were tending yaks and sheep.

Dradül rushed out to Aru Co Lake, 5100 meters above sea level, to help with the rescue. The grassland had disappeared, entombed under a 30-meter-high wall of ice. Eyewitnesses said the glacier had barreled in like a fast train, dumping enough ice to fill 40,000 Olympic-size swimming pools. The surge crushed nine people, including Dradül's sister and her four children, as well as hundreds of head of livestock.

Thousands of glaciers perch near human settlements, and in recent decades, dozens of surges have claimed lives. One of the worst calamities occurred in 2002, in the Caucasus Mountains of southern Russia, when Kolka Glacier rumbled into a valley, killing 140 people. Anecdotes, and even some preliminary tallies, suggest surges are becoming more frequent. Just 2 months after the Aru disaster, Chinese scientists were on hand when a surge from an adjacent glacier engulfed another swath of grassland. No one was injured that time. But the back-to-back surges were "simply astounding," says Yao Tandong, a glaciologist at the Chinese Academy of Sciences's Institute of Tibetan Plateau Research in Beijing. "It makes you wonder what's going on."

Most surges, broadly defined as a flow at least 10 and often hundreds of times faster than a glacier's usual pace of advance, are quieter affairs. Many are imperceptibly slow, but others attain staggering speeds. In 1953, for example, Kutiah Glacier in Pakistan advanced 12 kilometers over 3 months. Besides overwhelming settlements, glacier surges can threaten distant communities. They can block rivers, creating lakes that can later unleash floods, and by depleting glacier mass, they can threaten the flow of meltwater that downstream towns and farms may depend on.

Now, by studying glaciers from Tibet to the Arctic islands of the Svalbard archipelago in Norway, researchers are starting to understand why some glaciers swing between extremes of stagnation and crushing flow, and how surges may be predicted. Until recently, most glaciologists believed that a glacier's physical characteristics, such as its thickness and shape, and the properties of the terrain it sits on determine whether it can surge. Now, they believe an external factor also plays a major role: water from precipitation and melting. Pooling on the surface, it can infiltrate the glacier through crevasses and reach its base, warming, lubricating, and, ultimately, releasing the ice.

No one thinks that's the full explanation. "It's a combination of factors that determine whether and how often a glacier can surge," says Andreas Kääb, a remote-sensing expert at the University of Oslo (UiO). But a major role for meltwater suggests that "we are likely to see more glacier surges in a warming world," Yao says. "This poses a serious challenge to hazard management."

Studying surging glaciers could also offer insights into grander-scale ice flows with global consequences: the movements of the ice sheets in Antarctica and Greenland, which can change abruptly, altering the ice discharges that affect sea level. "The underlying physics are the same," says Thomas Schuler, a UiO glaciologist.

Glacier surges have both fascinated and perplexed scientists for decades. "If you think of glaciers as a bank account, then a surge is a massive spending spree," Kääb says. All glaciers have to shed mass that has accumulated in their upper reaches. "Some glaciers just flow faster, but others are unable to for whatever reasons," he says. "They are kind of stuck until the mass accumulated for decades or even centuries gets unleashed in a spectacular way."

Just over 1% of our planet's glaciers—some 2300 in all—are known to undergo these precipitous movements, though the number is likely to rise as glaciers come under closer surveillance by remote sensing. They are concentrated in geographic hot spots including Svalbard, Canada's Yukon territory, Alaska, western Tibet, and the Karakoram and Pamir mountain ranges of Central Asia. This geographic pattern only deepens the puzzle. For instance, some experts think glaciers in the Karakoram are prone to surging because of their steepness; as mass builds up from heavy snowfalls near the top of a glacier, for example, gravity alone may trigger a surge. But this cannot explain why Svalbard, where the terrain is relatively flat, abounds in surging glaciers.

Even glaciers right next to each other can have totally different personalities. Jack Kohler, a glaciologist at the Norwegian Polar Institute in Tromsø, points to a pair of adjacent, massive glaciers on Svalbard: Kongsvegen and Kronebreen. "They are like twin brothers, but one surges and the other one doesn't," Kohler says. "It's a total mystery."

In September 2015, Wahlenbergbreen glacier in Svalbard, Norway, surged at nearly 9 meters a day, swallowing up a beach.

HEÏDI SEVESTRE

To understand the deeper dynamics of surges, researchers have tried to witness them firsthand, but it hasn't been easy. In 1980, glaciologist Garry Clarke of the University of British Columbia in Vancouver, Canada, thought the odds of catching a surge were good at Trapridge Glacier in the Yukon, which had a dramatic surge 4 decades earlier. He noted that the glacier's upper reaches were getting steeper and crevasses were multiplying—often a sign of instability. "It looked really primed to unleash another surge," Clarke says. His team installed instruments to monitor everything from ice temperature to water pressure and conductivity beneath the ice. "We were really hoping to capture the start of an energetic surge," he says. "All we had to do was to wait. But that moment never came."

On Svalbard, however, Schuler and his colleagues have had better luck. In 2004, they began monitoring Europe's largest ice field by area: Austfonna ice cap, a monster that is 560 meters thick in spots and straddles 8500 square kilometers, roughly the area of Puerto Rico. They were not expecting a surge; their goal was to assess fluctuations in ice mass. But 3 years later, they saw crevasses forming. In the summer of 2007, they installed GPS receivers on metal stakes drilled into the glacier. Then, says Schuler, "Things got more and more interesting."

As the researchers reported in 2015 in The Cryosphere, Austfonna's movement accelerated each year in early July and slowed in late August. Faster speeds broadly correlated with the number of days of above-freezing air temperatures. But year after year, after the glacier slowed in August, its movement was faster than it had been before the speedup. "It got pushed to a higher level every summer," Schuler says. At the same time, its crevasses were deepening and extending. Suddenly, in the autumn of 2012, the glacier failed spectacularly. Over the following months, it gushed 4.2 cubic kilometers of ice—enough to fill 1.7 million Olympic-size swimming pools—into the Barents Sea. "It was the surge of the century," Schuler says.

Based on the correlation between warming and speedup, Schuler and his colleagues suspect that the trigger for the surge was meltwater that trickled down through crevasses and accumulated at the glacier's base, summer after summer. As the infiltrating water froze, the latent heat it released warmed the surrounding ice. "This alone can change glacier dynamics quite drastically" because warm ice flows a lot faster than its subzero counterpart, Schuler says. And as more water accumulated beneath Austfonna, the increasing pressure, like a hydraulic jack, lifted the glacier from its bed.

Ultimately the cold ice anchoring Austfonna's tongue to the ground disintegrated. "That was the critical part that held the ice back," says Jon Ove Hagen, a UiO glaciologist. Its loss unleashed the surge.

Glaciers gain mass in their upper reaches, where snowfall is heavier, and lose it at their snouts, where the ice breaks up and melts (right). Most glaciers flow steadily, but some get stuck and accumulate mass (center), then release it in a surge. A surging glacier can race down a valley or mountain, growing thinner and longer (left). Then, anywhere from days to years later, the glacier’s speed ebbs and it begins thickening again. Mud and debris flow Receiving area Sectional view Regular glacier Presurge Postsurge Ice front/snout Reservoir zone Crevasse Subglacial stream Terminal moraine Crevasse Ponding water Meltwater Sediment or rock Surging glaciers are riddled with crevasses , especially in their lower reaches. When the surge ends, meltwater that built up under the glacier before the surge may sweep mud and debris from its snout. Trickling down Meltwater plays a key role in triggering surges. Pooling on the glacier’s surface, it can seep down into crevasses. There it can refreeze, releasing heat that softens the ice; it can also pool at the base of the ice. Steady state In a “normal” glacier, meltwa- ter drains effi- ciently from its base, carrying away heat and leaving the ice anchored to its bed. Buildup to a surge If drainage is poor or melting accel- erates, meltwater can accumulate under a glacier, warming the ice and lifting it off the ground. Aftermath Once the surge releases the meltwater, the glacier subsides onto its bed, and the cycle begins again. Glacial surface down cracks in the ice. Meltwater travels Ice Meltwater drain channel Steady meltwater flow under glacier releases heat and pressure. Pooled meltwater Hydraulic lifting Ice warming Glacier settles Majority of meltwater expelled during surge. A glacier unleashed Meltwater can freeze on its way down.
C. BICKEL/SCIENCE

Download this graphic

The Austfonna study was a revelation. "If water is important for triggering a surge—as we are increasingly realizing—then climate change must have an impact," says Heidi Sevestre, a glaciologist at the University of St. Andrews in the United Kingdom, who was not involved in the study.

Analyses of the Aru surges also point to a climate link. In western Tibet, annual snowfall totals have risen steadily since the 1990s, especially at higher elevations, as strengthening westerly winds bring more precipitation. "Glaciers are accumulating mass," Yao says. "The bank accounts are getting fatter." Meanwhile, average air temperatures in the region rose 1.5°C over the past 5 decades, nearly twice the global average. The warming has boosted the amount of meltwater shed by the Aru glaciers by 50%, 3D computer modeling suggests. "This means that more water can go through the cracks and eat the ice away," says Adrien Gilbert, a UiO glaciologist who described his team's findings at the Third Pole Science Summit last July in Kunming, China.

Satellite imagery revealed new crevasses in the Aru glaciers in 2010, which grew deeper and more extensive summer after summer. The final trigger for the surges may have been unusually heavy rains and snows during the 40 or so days before the first surge. The precipitation "might be the straw that broke the camel's back," says Yao, who led a study of the Aru surges that appeared last February in the Journal of Glaciology.

The modeling by Gilbert's team suggests that, like Austfonna, the Aru glaciers surged after their frozen tongues became unmoored. "You need a huge amount of water to cause the failure," Kääb says. "But as soon as the water finds its way out, the surge stops. It's a relief for the glacier." Dradül and other villagers confirm that after both surges, the foothills were flooded as water gushed from the snouts of the glaciers.

Sevestre and her St. Andrews colleague Douglas Benn have incorporated the effects of meltwater and precipitation into a broader picture of why some glaciers surge, and where surges are likely to occur. "To stay out of trouble, glaciers have only one job to do: to keep in balance," Sevestre says. This means shedding any heat they gain from the air and the ground, from water that infiltrates through cracks, and from friction generated as the ice advances.

In a 2015 Journal of Glaciology paper, the duo—then at the University Centre in Svalbard—presented a modeling study showing that glaciers can most easily maintain a thermal balance at climatic extremes: in cold, dry climates, where they can shed heat to the frigid air, and in warm, humid environments, where they discharge heat in steady streams of meltwater. By contrast, glaciers in intermediate conditions can easily get out of kilter, accumulating internal heat until enough meltwater builds up at their base to trigger a surge.

That picture could help explain the geographic pattern of surging glaciers. And it suggests that a change in climate can be a tipping factor in glacier behavior, as Sevestre and Benn write in their paper: "Glaciers may change from ‘normal’ to surge-type and vice versa under cooling or warming climates." Or, as Adrian Luckman, a glaciologist at Swansea University in the United Kingdom, puts it, glacier behavior "may be a lot more fluid than we thought."

Move the slider to see Tibet's Aru range in June and September, 2016, before and after two massive slides.

NASA EARTH OBSERVATORY IMAGES BY JOSHUA STEVENS AND JESSE ALLEN, USING LANDSAT DATA FROM THE U.S. GEOLOGICAL SURVEY, MODIFIED COPERNICUS SENTINEL DATA (2016) PROCESSED BY THE EUROPEAN SPACE AGENCY

On a late spring day on Svalbard, Christopher Nuth guns his snowmobile through the fog, shifting his weight like an acrobat to keep the bounding machine on course. The UiO glaciologist is on his regular commute from Ny-Ålesund, an international science village, to Kongsvegen—one of two glaciers with strangely contrasting behaviors.

A bone-jarring hour later, Nuth reaches his field site. Visibility has improved, and Kongsvegen looks serene, its smooth surface sloping gently toward a fjord. "Kongsvegen's calm appearance is deceptive," Nuth says. "It has a history of erratic behavior." In 1948, the glacier disgorged a massive amount of ice and debris—there are no reliable estimates—into the fjord and sideways toward its neighbor, Kronebreen. Then it slowed to become one of the world's pokiest glaciers, creeping about 2 centimeters a day. Kronebreen, in contrast, exhibits no such vagaries, maintaining a steady pace of 3 meters a day.

Now, Kongsvegen may become unruly again. Its upper reaches, where snow accumulates, are getting steeper, and it is picking up speed. Kneeling in the snow, Nuth connects solar panels to a GPS receiver anchored into the ice. A network of sensors here and on Kronebreen will allow his team to measure the ice's motion to millimeter accuracy.

Preliminary findings show that in warm and wet conditions, the glacier's surface can rise as much as a third of a meter—a massive hydraulic effect from meltwater pooling under the ice, Nuth believes. "It's quite amazing," he says. "That's a lot of water to lift hundreds of meters of ice." Often, the ice flows faster when it rises—more evidence for the role of meltwater in surges. Next year, the team will install seismometers to listen in on the flow of water through and beneath the ice, and to crevasse formation.

The scientists on Svalbard hope that wiring up the neighboring glaciers will reveal why they move at wildly different rates. They also hope the findings will yield broader insights into ice flow mechanics, which could save lives and aid forecasts of sea level changes. Nuth believes Kongsvegen may be on the brink of letting loose, as it did 70 years ago. "We may even capture a glacier surge," he says. "You never know."

Jane Qiu is a science journalist in Beijing. Her trips to Aru, Tibet, and Ny-Ålesund on the Svalbard archipelago were supported by journalism fellowships from the International Water Management Institute and the European Geosciences Union, respectively.