a red sunset reflecting on the water with stark white ice chunks floating

The Larsen C Ice Shelf Collapse Is Just the Beginning—Antarctica Is Melting

The massive iceberg that broke off the Larsen C Ice Shelf may be a harbinger of a continent-wide collapse that would swamp coastal cities around the world.

A startling sunset reddens the Lemaire Channel, off the west coast of the Antarctic Peninsula. The continent’s coastal ice is crumbling as the sea and air around it warm.
This story appears in the July 2017 issue of National Geographic magazine. It was first published on June 14, 2017 and updated on July 12 with news of the Larsen C break.

Seen from above, the Pine Island Ice Shelf is a slow-motion train wreck. Its buckled surface is scarred by thousands of large crevasses. Its edges are shredded by rifts a quarter mile across. In 2015 and 2016 a 225-square-mile chunk of it broke off the end and drifted away on the Amundsen Sea. The water there has warmed by more than a degree Fahrenheit over the past few decades, and the rate at which ice is melting and calving has quadrupled.

On the Antarctic Peninsula, the warming has been far greater—nearly five degrees on average. That’s why a Delaware-size iceberg just broke off the Larsen C Ice Shelf and why smaller ice shelves on the peninsula have long since disintegrated entirely into the waters of the Weddell Sea. But around the Amundsen Sea, a thousand miles to the southwest on the Pacific coast of Antarctica, the glaciers are far larger and the stakes far higher. They affect the entire planet.

The Pine Island Ice Shelf is the floating terminus of the Pine Island Glacier, one of several large glaciers that empty into the Amundsen Sea. Together they drain a much larger dome of ice called the West Antarctic Ice Sheet, which is up to two and a half miles thick and covers an area twice the size of Texas. The ice sheet is draped over a series of islands, but most of it rests on the floor of a basin that dips more than 5,000 feet below sea level. That makes it especially vulnerable to the warming ocean. If all that vulnerable ice were to become unmoored, break into pieces, and float away, as researchers increasingly believe it might, it would raise sea level by roughly 10 feet, drowning coasts around the world.

The ice sheet is held back only by its fringing ice shelves—and those floating dams, braced against isolated mountains and ridges of rock around the edges of the basin, are starting to fail. They themselves don’t add much to sea level, because they’re already floating in the water. But as they weaken, the glaciers behind them flow faster to the sea, and their edges retreat. That’s happening now all around the Amundsen Sea. The Pine Island Ice Shelf, about 1,300 feet thick over most of its area, is a dramatic case: It thinned by an average of 150 feet from 1994 to 2012. But even more worrisome is the neighboring Thwaites Glacier, which could destabilize most of the West Antarctic Ice Sheet if it collapsed.

“These are the fastest retreating glaciers on the face of the Earth,” says Eric Rignot, a glaciologist at the NASA Jet Propulsion Laboratory in Pasadena, California. Rignot has studied the region for more than two decades, using radar from aircraft and satellites, and he believes the collapse of the West Antarctic Ice Sheet is only a matter of time. The question is whether it will take 500 years or fewer than a hundred—and whether humanity will have time to prepare.

“We have to get these numbers right,” Rignot says. “But we have to be careful not to waste too much time doing that.”

Getting the predictions right requires measurements that can be made only by going to the ice. In December 2012 a red-and-white Twin Otter plane skimmed low over the Pine Island Ice Shelf. The pilot dragged the plane’s skis through the snow, then lifted off and circled back to make sure he hadn’t uncovered any crevasses. After the plane landed, a single person disembarked. Tethered to the plane by a rope and harness, he probed the snow with an eight-foot rod.

<p>A diver watches an emperor penguin as it swims nearby. The brown patches above are microalgae, which cling to sea ice and photosynthesize in the spring.</p>

A diver watches an emperor penguin as it swims nearby. The brown patches above are microalgae, which cling to sea ice and photosynthesize in the spring.

Photograph by Laurent Ballesta

Finally the scout was satisfied: There were no buried crevasses that might swallow a landing party. More scientists got out of the plane. The team, led by glaciologist Martin Truffer of the University of Alaska, proceeded to set up camp. Their plan was to spend two months on the ice shelf; they would be the first humans to spend even a single night. The ice had long been considered too dangerous to visit. But Truffer’s team wanted to bore holes all the way through the ice shelf, so they could measure the heat eating at it from the seawater below.

As the researchers lay in their tents at night, in the middle of a 4,000-mile arc of coastline that lacked a single permanent outpost, they heard loud pops and bangs coming from the ice. Each morning they saw new cracks, an inch wide and seemingly bottomless, cutting across its surface. During their five weeks of studying it, the ice under their boots thinned by another seven feet.

It took scientists a long time to realize just how quickly West Antarctica’s ice could melt. In part that’s because the most vulnerable glaciers are so well guarded. In front of the Pine Island Ice Shelf—the floating end of the glacier—the sea surface itself freezes each winter. In summer this fractured sea ice joins icebergs calved from the ice shelves to form a shifting palisade that historically kept ships at least a hundred miles from the ice shelf.

In March 1994 the U.S. icebreaker Nathaniel B. Palmer became perhaps only the second vessel ever to reach it. For a few days powerful winds parted the ice floes, creating a narrow, ephemeral passage for the Palmer to thread. With no accurate maps to guide them, the crew on the ship’s bridge eyed the sonar monitor nervously. It showed a chaotic seafloor of canyons and sharp ridges, including one that rose within 20 feet of the ship’s keel.

The Palmer would spend just 12 hours at the front of the ice shelf before encroaching sea ice forced it to retreat north. But that gave the crew enough time to lower scientific instruments through the water column. They made a disturbing discovery. Near the surface, a current was streaming out from under the ice shelf that was slightly less salty than the sea around it, because it was freshened by melted ice. (The ice is fresh because it originated as snow falling on West Antarctica.) And at depths of 2,000 to 3,000 feet, along a seafloor canyon that ran straight under the ice, warmer seawater was streaming in.

Stan Jacobs, an oceanographer from the Lamont-Doherty Earth Observatory in New York, quickly understood what was going on. The warm water was coming from the South Pacific, more than 200 miles north. It was so heavy with salt that it was following the floor of a submarine canyon, which sloped down toward the glacier. The glacier itself had carved that canyon, thousands of years ago during the Ice Age, when it and the other glaciers in West Antarctica advanced hundreds of miles out from their present-day positions.

Now that same canyon was channeling warm ocean water under the Pine Island Ice Shelf. Somewhere tens of miles inland, the warm water was finding the “grounding line”: the place where the glacier lifts off the seafloor and becomes a floating ice shelf. Hitting that wall of ice, the warm water was eroding it, producing a steady stream of melt-laden seawater. Because it was cooler and fresher, it was less dense, and so it was rising above the warmer, incoming water and flowing back out to sea just under the shelf.

By measuring the amount of this freshwater, the researchers could estimate how much ice was being lost. The melt rates “were just crazy,” says Adrian Jenkins, a glaciologist from the British Antarctic Survey in Cambridge. According to his calculations, the ice shelf was losing 13 cubic miles of ice per year from its underside; back near the grounding line, the ice was probably thinning up to 300 feet per year.

“It was just beyond our concept that a glacier would melt that fast,” Jenkins says.

Over the next 13 years he and Jacobs tried three times to return to Pine Island. Sea ice blocked them each time. When they finally got back there on the Palmer in January 2009, they found that the melt rate had increased by about 50 percent. This time they came equipped with a new tool: a yellow robotic submarine called Autosub3. Shaped like a torpedo and as long as a delivery truck, it could navigate autonomously under the ice shelf, out of contact with the ship, for up to 30 hours at a time.

On its first three dives, Autosub3 discovered that the ice shelf had thinned enough to lift off a submarine ridge that, running across its width, had once supported and stabilized the ice shelf. That had opened a gap that was allowing warm water to flow in and melt the underside of the ice even faster. On its fourth dive the yellow robot nearly died. When the crew winched it out of the water, they found its nose cone smashed and some of its delicate internal equipment damaged.

Technicians reconstructed what had happened from the sub’s navigation data. Thirty miles back, under the ice shelf, Autosub3 had strayed into a chasm on the underside of the ice. Searching for a way forward, it had smashed and scraped against the walls of the chasm—ultimately rising 500 feet up into the labyrinthine bowels of the ice shelf. Finally it had dropped back out and escaped into open water.

The sub’s sonar data, meanwhile, revealed the breathtaking landscape it had navigated. The bottom of the ice shelf was corrugated with not just one but many channels, which cut as far as 600 feet up into it. The walls of these inverted ice canyons were sculpted into terraces, ledges, and sharp corners, and along the ceiling of each ran a gaping crack that penetrated even farther into the ice.

“What the hell is going on?” Jenkins recalls thinking when he first saw the sonar maps.

What he and Jacobs came to realize was that the upside-down canyons had been carved, like rock canyons on land, by flowing water. Apparently the meltwater rising off the grounding line was still warm enough to melt more ice. And as it flowed for tens of miles along the underside of the ice shelf, back out to the open sea, it was melting a lot of it.

Large swaths of West Antarctica are hemorrhaging ice these days. The warming has been the most dramatic on the Antarctic Peninsula, a spine of ice-cloaked mountains that reaches 700 miles up toward the tip of South America. Catching the powerful winds and ocean currents that swirl endlessly around Antarctica, the peninsula gets slammed with warm air and water from farther north. Average annual temperatures on its west side have risen nearly 5 degrees Fahrenheit since 1950—several times faster than the rest of the planet—and the winters have warmed an astonishing 9 degrees. Sea ice now forms only four months a year instead of seven.

Since 1988, four ice shelves on the east side of the peninsula have disintegrated into armadas of icebergs. (The Larsen C Ice Shelf may one day do the same, judging from that Delaware-size ice chunk that's about to break off it.) Warmer air helped trigger these collapses by forming meltwater ponds on the surface of ice shelves; the ponds drained into crevasses, wedging them deeper into the ice. As the shelves have vanished, the glaciers they once stabilized have stampeded into the ocean, accelerating to two, five, even nine times their original speed. They’re relatively small glaciers and won’t raise sea level much—but their acceleration has reinforced concerns that the same thing might happen to the much larger glaciers along the Amundsen Sea.

The Amundsen Sea is farther south than the peninsula, and the air there is not as warm. The biggest threat to its glaciers is the mechanism Jacobs and Jenkins helped uncover: deep submarine canyons that channel warm water from the north under the ice shelves, and deep inverted canyons that focus the warmth on the underside of the ice.

A satellite survey last year of many Antarctic ice shelves—led by glaciologists Ted Scambos of the National Snow and Ice Data Center in Boulder, Colorado, and Helen Fricker of the Scripps Institution of Oceanography in San Diego—revealed that such melt canyons are common. They tend to fan out and steer warm water toward the edges of the shelves. The ice there is crucial: It rubs against the stationary banks and slows the flow of the shelf and the glacier behind it. But that edge ice is also thinner than the rest. This “is something that bears watching,” Scambos said in early 2016.

Ian Howat, of the Byrd Polar and Climate Research Center in Columbus, Ohio, is another glaciologist who’s watching Pine Island closely. Last November he reported two ominous new rifts spreading across the ice shelf that threaten to prune it to its shortest length in recorded history. As Howat looked back through monthly satellite photos, he realized that the rifts had been triggered by a singular event that had happened, unnoticed, three years before. The strip of torn-up ice anchoring the ice shelf to its northern bank had suddenly fallen apart, suggesting it had been undermined by melting from below. It blew out “just in a matter of days,” Howat says, “like a zipper, unzipping the side of the glacier.”

It’s unclear when the entire ice shelf might disintegrate. The “warm” water flowing underneath it from offshore is only 4 to 6 degrees Fahrenheit above freezing. But roughly 3,000 cubic miles of it arrives every year, which means the ice shelf is receiving an amount of heat that exceeds the output of a hundred nuclear power plants, operating 24/7.

When Truffer and his team camped on the shelf in December 2012, they could sense how it had already weakened. As the meltwater cuts deep into the ice from below, the unsupported ice sags, causing the entire shelf to bend and warp. Crevasses erupt along the lines of stress, on both the top and the bottom of the ice. The pops and bangs the researchers heard and the daily opening of new cracks bore witness to the ice’s gradual failure as it thinned and broke down beneath them.

As the Pine Island Ice Shelf has weakened and the glacier behind it has accelerated, the ice has stretched and thinned for 150 miles inland from the coast. The destabilizing effects spread farther into West Antarctica every year. “A little nudge can get you to several decades of retreating behavior that’s hard to reverse,” Truffer says.

In fact, research by Rignot and others over the past few years indicates that the collapse of several major glaciers flowing into the Amundsen Sea is now unstoppable. Between 2002 and 2009 alone, the ice shelf in front of the Smith Glacier thinned by 1,500 feet in some places, the one in front of the Pope Glacier by up to 800 feet. The grounding lines of the Amundsen glaciers have retreated so far—tens of miles in some cases—that they now rest on seafloor that slopes down toward the center of the ice sheet. Each increment of retreat exposes a greater ice surface to warm ocean water. It’s a runaway process—and scientists are urgently trying to figure out how fast it will run.

The ice shelves, Fricker says, “are the canary in the coal mine.” Because they’re already floating, they don’t raise sea level themselves when they melt—but they signal that a rise is imminent, as the glaciers behind them accelerate. Fricker and her team have found that from 1994 to 2012, the amount of ice disappearing from all Antarctic ice shelves, not just the ones in the Amundsen Sea, increased 12-fold, from six cubic miles to 74 cubic miles per year. “I think it’s time for us scientists to stop being so cautious” about communicating the risks, she says.

The retreat and hemorrhage of these glaciers “will accelerate over time,” agrees Rignot. “Maybe you don’t care much about that for the next 30 to 40 years, but from 2050 to 2100 things could get really bad, and at that point listening to scientists is irrelevant.” Yet after things get really bad, they could still get worse.

Most of the heat trapped by our fossil fuel emissions since the industrial revolution began in the 19th century has gone into the ocean. Most of the heat now hitting the Antarctic ice shelves, however, comes from another effect of climate change: Intensified circumpolar winds and currents have driven warmer water from offshore onto the continental shelf and under the floating ice. Much more ocean warming is yet to come, even if we begin to cut emissions. A lot more heat is on the way to Antarctica.

Scientists are especially concerned about the Thwaites Glacier, which by itself could raise global sea level four feet; last fall the British and American science foundations announced a coordinated $20 million to $25 million field campaign that will deploy ships, planes, satellites, and underwater robots to assess the glacier’s status starting in 2018. For now, the best estimates suggest that Antarctica will sweat off enough ice to raise global sea levels by 1.5 to 3.5 feet by 2100, depending on how quickly humans continue to pump out greenhouse gases. Throw in Greenland and other rapidly melting glaciers around the world, and sea level could plausibly rise three to seven feet by 2100.

But that’s not the worst case: Sea level won’t stop rising in 2100. Earth’s past offers worrisome clues to what the more distant future might bring. Geologists studying ancient shorelines have concluded that 125,000 years ago, when the Earth was only slightly warmer than today, sea levels were 20 to 30 feet higher. Some three million years ago, the last time atmospheric carbon dioxide was as high as it is today, and the temperature was about what it’s expected to be in 2050, sea levels were up to 70 feet higher than today. Yet a collapse of the Greenland and West Antarctic Ice Sheets would raise sea level only about 35 feet.

To consider the worst case, then, scientists must turn their eyes toward East Antarctica, home to more than three-fourths of all the ice on Earth.

This past January a twin-propeller DC-3 made a series of flights from Australia’s Casey Station along the East Antarctic coast. Built in 1944, the plane was packed with modern scientific equipment. As it flew over the Totten Glacier, a radar recorded the thickness of the ice. Another instrument recorded tiny changes in Earth’s gravitational field—clues to the topography of the seafloor under the glacier’s floating ice shelf. Now and then a crew member opened the plane’s rear door, knelt in the windy opening, and tossed out a torpedo-shaped object. As the device splashed into the water, it split in two: One part floated, sending radio signals back to the plane, while the other part reeled down 2,600 feet of wire, measuring the water temperature all the way down.

Until recently the East Antarctic Ice Sheet was considered secure; unlike West Antarctica, it sits on high ground. But mapping with ice-penetrating radar has revealed a low-lying region cut by glacially carved channels that drop as far as 8,500 feet below sea level—perfect for guiding warm ocean water deep into the heart of the ice sheet. The Totten Glacier is the largest coastal outlet in this region. If it collapsed, global sea level could rise 13 feet—“roughly as much as all of West Antarctica,” Rignot points out. “One glacier alone.”

In January 2015, the Australian icebreaker Aurora Australis became the first ship to reach the front of Totten. Like the Palmer at Pine Island in 1994, it found deep, warm water flowing under the ice shelf, at a rate of 4.5 cubic miles a day. The glacier is already losing a couple of cubic miles of ice per year—small potatoes, in Antarctic terms. But Donald Blankenship, a University of Texas glaciologist who oversees the aerial survey, fears it could blow up.

In 2016 his team reported evidence from the bedrock that Totten repeatedly has retreated 100 to 200 miles inland from its current position—meaning it might help explain why sea level was so much higher three million years ago. Blankenship’s surveys have also identified two seafloor grooves deep enough to let warm water under Totten’s ice shelf. Last January the team was refining those seafloor maps.

Totten will lose its ice more slowly than West Antarctica. The worst case coming out of Antarctica still seems to be centuries away. But it would mean abandoning many of the world’s largest cities, including New York, Los Angeles, Copenhagen, Shanghai, and dozens of others—and it’s looking less crazy all the time. “The fuse is lit,” says Blankenship. “We’re just running around mapping where all the bombs are.”

Writer Douglas Fox has traveled to Antarctica five times and has spent months on the ice there. This is his first feature for National Geographic magazine.

Read This Next

A photographic eulogy for Greenland's icebergs
An icy world is in meltdown, amid penguin population shifts
Chile’s glaciers are dying. You can actually hear it.

Go Further

Subscriber Exclusive Content

Why are people so dang obsessed with Mars?

How viruses shape our world

The era of greyhound racing in the U.S. is coming to an end

See how people have imagined life on Mars through history

See how NASA’s new Mars rover will explore the red planet

Why are people so dang obsessed with Mars?

How viruses shape our world

The era of greyhound racing in the U.S. is coming to an end

See how people have imagined life on Mars through history

See how NASA’s new Mars rover will explore the red planet

Why are people so dang obsessed with Mars?

How viruses shape our world

The era of greyhound racing in the U.S. is coming to an end

See how people have imagined life on Mars through history

See how NASA’s new Mars rover will explore the red planet