David Holland (left) at Helheim Glacier, August 2018 (Lucas Jackson/Reuters)
15
A Key
Almost from the moment they put ICESat up in the sky, NASA’s scientists watched with concern as the satellite suffered technical problems. The machine had been equipped with three lasers that could send pulses of light down from space to measure changes in Greenland’s and Antarctica’s ice. Each laser was intended to last a year or two. The first laser failed after thirty-six days. The second one burned out after a hundred days. The space agency devised an emergency plan to prolong the life of the third, but when that laser failed, too, there was no point keeping ICESat aloft. In 2010, the spacecraft received instructions to leave orbit, and not long after, most of it burned up as it reentered the atmosphere. The parts that didn’t vaporize—a few bits of space debris—dropped into the Barents Sea north of Siberia.
GRACE was still circling the earth in good health; every month it sent down information about the demise of Greenland’s and Antarctica’s ice. What GRACE couldn’t always provide, though, was a fine-grained picture of where the losses were coming from. That had mostly been ICESat’s job. Thus the engineers and scientists now found it harder to understand which glaciers in Greenland were thinning the most, or precisely how (and how fast) the process seemed to be moving. A few glaciologists considered the situation urgent. “We are slowly going blind in space,” Robert Bindschadler, a NASA veteran, told a journalist at the time.1 Yet by that point, space agency administrators were discussing remedies. One solution was nicknamed “Gap Filler,” which soon became known as IceBridge. The moniker arose from the idea that NASA needed to “bridge” the “data gap” between the now-defunct ICESat and a replacement satellite that was probably many years away from launch.2 “Headquarters approached us and said: ‘How about you take on this gap-filler concept?’ ” John Sonntag, who eventually became the IceBridge mission scientist, recalls. “So we kind of invented it as we went along.”
The plan for the new program was straightforward: Fly an airplane over the world’s two great ice sheets for several hundred hours each year. Cover Greenland for several months in the spring; survey Antarctica for a few months in the fall. And on those flights, use an arsenal of tools to measure changes in the ice sheets below, along with the sea ice floating nearby. In many respects, IceBridge resembled the NASA flights over Greenland a decade before, which were the first to use lasers to take a measure of the ice sheet’s decline. Right away the mission planners considered it necessary to map out a number of routes that could represent a “sampling” to determine ice losses. These were routes that could be repeated with precision every year, so that any changes could be easily tracked.
Within a few years, the IceBridge air campaigns had become routine. Every April, IceBridge was stationed at the main airport on Greenland’s southwest coast. Originally known as Bluie West Eight during World War II, the airport had for decades been known by its Danish name, Sondrestrom. But a small village of about five hundred people had grown up around the site, and in the early 1990s the town assumed its Inuit name: Kangerlussuaq. In turn, the airport had been renamed Kangerlussuaq international airport. Not far from the runway’s terminus, an old military barracks—a rambling, two-story, no-frills concrete structure—had been converted into a dormitory for scientists coming in and out of Greenland. It was known as KISS, which stood for Kangerlussuaq International Science Support. On a typical day during the IceBridge campaign, John Sonntag would get up early at KISS, put on an olive green flight suit, and begin his morning with a visit to the town’s weather office, a suite of rooms near the small grocery store in town. A Danish meteorologist, gazing at a number of computer screens, would give Sonntag Greenland’s regional forecast and a packet of maps. Then Sonntag would weigh his options. “On any day you have a quiver of flight plans you can follow,” he would explain, “and you pick the highest priority mission you have on a given day if the weather supports it.”
The local weather in Greenland can vary from glorious to deadly. Sometimes, too, a forecast for low clouds or fog can scuttle a mission, since they can interfere with sensing equipment. To Sonntag, the highest priority missions were ones routed above locations where ice losses had been most extreme. These were the glaciers—Petermann in the northwest, Zachariae Isstrom in the northeast, and Jakobshavn in the west—that observations showed had been thinning or could be most at risk. So in the early morning, Sonntag would spend a few minutes mulling over possibilities until he picked his route. Then he would inform the pilots. Usually, with temperatures at Kangerlussuaq hovering around zero, the aircraft engines would need to be warmed up. Preflight checks had to start. The rest of the science team, shivering, would gather on the runway near the plane by eight A.M.
A typical IceBridge flight—eight hours in duration—is like flying at very low altitude from New York to Houston and then back again without landing. Except of course it isn’t. For one thing, you’re flying a big military plane—a P-3 or C-130—over glaciers that flow for miles from the central ice sheet to the sea on winding courses that run between dark, desolate mountain peaks. “These are places that only a few hundred people in the world have ever seen,” IceBridge pilot Jim Lawson observed one day, almost to himself, mid-flight. You’re wearing flight headphones most of the time so you can talk over the din in the cabin, and also so you can hear Sonntag, sitting in front on the flight deck, call out the sights below. One day you pass over Nansen fjord on the east coast, near where the Norwegian landed in 1888 to begin his crossing. Another day you swing down the west coast, by the calving glaciers of Eqip, where Paul-Émile Victor’s team landed in 1949. Amidst the wide white expanse of the ice cap, you might pass over Dye-3, the drilling site from the early 1980s, or above Summit Camp, the place where the GISP-2 core was drilled in the early 1990s. Somewhere nearby, you buzz over the blank stretches of ice where Eismitte and Station Centrale once existed.
It seems as though history becomes scenery, and scenery becomes data. In the airplane cabin, the science team consults its screens as information rolls in from their sensing instruments. An altimeter bolted to the aircraft floor sends laser pulses down to map the ice elevation below. An ice-penetrating radar peers through layers of ice to bedrock. A digital camera records high-resolution images of the icescape below. Even with a good weather forecast, it may get turbulent up in the air. Sonntag recalls how he once encountered turbulence over the peaks of east Greenland unlike anything he’d ever experienced. “The wind flows down these mountains, and it starts to eddy and swirl,” he says. “I remember looking back into the cabin after one bump—the coffeepot had shattered, laptops are on the floor, someone’s lunch is on the ceiling. A couple people are on the floor. Even parts of the floor panels had come off.” He cut the mission short that day, turned back, and returned to Greenland’s west coast for dinner and an evening jog. On another IceBridge flight over Antarctica’s Weddell Sea he didn’t have that option. At twenty-five thousand feet, in pitch dark and hundreds of miles from any airfield, the plane lost two of its four engines. “They just stopped,” Sonntag says, “and I looked out the window and saw that the props weren’t spinning.” The plane began losing altitude. “And then the pilots rang their alarm bell, which sounds like a school bell. And the pilot does not ring that bell unless things go seriously wrong.” Instructions from the flight deck came down: Get your exposure suits on. “So we get on these rubber suits, what we call the ‘Gumby suits,’ thinking we are going to ditch.” Before they ended up in the Weddell Sea, though, the pilots discovered that a lower altitude could thaw a frozen fuel line, which happened to be the source of the problem. They restarted the engines. In twenty years of polar flights, Sonntag says, “that was the closest I’ve ever gotten to…you know.”3
Usually, the challenges are more mundane. Buffered from the cold, windy misery of the ice sheet, measuring Greenland by IceBridge means living fifteen hundred feet above the ice at a speed of about three hundred miles per hour—all day, six days a week, ten weeks in a row. Monotony and fatigue can become the biggest burdens as the survey overflies thousands of miles of ice and rock. Inside the plane, the larger purpose of the long days in the air, to better understand the vanishing glaciers, can start to recede. Approaching Helheim Glacier in east Greenland one afternoon—a glacier named for the word in Norse mythology that describes the world of the dead—Sonntag announces through the headsets, “Okay, science crew, get ready.” It is the signal that the technicians in the cabin need to focus on their instrument readings. Down below, the glacier is a snow-covered river of ice, four miles wide and twenty-five miles long, which flows out from the central ice sheet like a chute, curving between snow-flecked mountains toward the black ocean, where it breaks apart.
One of the technicians in charge of data storage, Aaron Wells, sits up from a nap and rubs his eyes. The members of the science crew have heard Sonntag’s instructions, and someone has quipped into their headset: “I can’t tell what the Helheim looking at.” Wells responds with a sigh and logs on to his laptop. He says into the headset, “I don’t know what the Helheim doing with my life.”
For the past fifteen years, since the GRACE satellite measurements began, Greenland has been shedding about 286 billion tons of ice, on average, every year.4 A large amount of that ice is pushed into the ocean through “outlet” glaciers like Helheim. While the losses can be obvious to sensing instruments, they can often be imperceptible to the eye. In the last few years, however, Sonntag has noticed clear incontrovertible changes. Meltwater lakes pool on the ice earlier in the spring season, and at higher elevations, than ever before. The lakes cover the white surface like giant beads of aquamarine, attesting to the fact that the warming Arctic air is now creeping onto Greenland’s ice sooner—and deeper into the middle of the ice sheet. “But the biggest change is the Jakobshavn Glacier,” Sonntag remarks. “It looks different every year. And especially in comparison from the early 1990s to now, it doesn’t even look like the same place.”5
Flying over the glacier one day, Sonntag points to Jakobshavn below: a fracturing river of ice, flowing from a channel on the western edge of the ice sheet. The glacier ends at a wall, its calving front, which from above resembles the shape of a seagull. The cliff face measures about three hundred feet high and is about six miles across; it is the place where Jakobshavn’s icebergs break off and begin their journey down an ice-choked fjord and into a larger body of water known as Disko Bay, where tourists can take iceberg cruises to ogle at their stupendous size and variations in color—alabaster white, riven with deep blue hollows. In time, these icebergs are carried out to the North Atlantic, where they melt. Like Helheim, Jakobshavn is an outlet glacier, and in its workings you can understand a fundamental aspect of how Greenland’s ice sheet loses mass, as well as how the world’s oceans rise in height. Between 2000 and 2011, Jakobshavn alone was responsible for a millimeter in global sea level rise.6 In August 2015, it calved several icebergs over two days that cut nearly five square miles of ice from the glacier’s front.7
Mid-flight, in the IceBridge cockpit, Sonntag pulls open his laptop to display some color-coded maps. They compile the past few years of remote-sensing data. On his screen, Jakobshavn’s glacier appears as a massive drain, increasingly sucking the western quadrant of the ice sheet toward it. “This is ominous,” he explains. “Because there is nowhere for that ice to go but into the ocean. And when there’s nowhere for the ice to go but the ocean, it means sea level rise.”8 As Jakobshavn loses ice, Sonntag says, it retreats. One way to visualize the retreat is to imagine that as the end of this river of ice breaks off into icebergs, the river itself pulls back from the coast. Or you might think of someone’s outstretched arm, a fist on the end to resemble an ice cliff, being pulled back toward their body.
Jakobshavn is perhaps the world’s most studied glacier. Its scientific history reaches back several centuries, at least to 1719, when a whale fisherman coming through the area named Feykes Haan noted a fjord “that is always full of ice, with frightfully tall icebergs, but from where they come is unknown.”9 Hinrich Rink, the first scientist to exhaustively study Greenland’s ice sheet, arrived here in 1850. He made his way to the calving front—at that point in time, many miles farther into the bay—to take notes about the position and behavior of the glacier. A few decades later, Amund Helland was the first scientist to measure its speed. By 1900 or so, many European scientists understood Jakobshavn to be special—unusually fast, unusually productive, perhaps faster and more productive, in fact, than any glacier in the world. Its icebergs were credited for sinking the Titanic. But they were also picturesque, and a familiar presence to Knud Rasmussen, who spent his childhood in a house on the north bank of Jakobshavn’s fjord, watching icebergs from the glacier move slowly past his front window every day.
In 1913, Alfred Wegener came here for a seven-day canoe trip. It was just a few weeks after his expedition party had finished crossing the ice sheet and was rescued by a passing ship. “The entire fjord in front of the glacier,” Wegener wrote in his notebook, “all the way through the banks of the fjord’s opening, [is] densely covered with icebergs and calving remnants.”10 Even then, he saw that Jakobshavn’s calving front was moving steadily back toward the ice sheet. And when Wegener visited again in 1929, just a year before he made his fateful trip to the center of the ice sheet, he noticed that Jakobshavn had retreated even farther. In subsequent years—through the era of Paul-Émile Victor’s explorations, for instance, and Henri Bader—the glacier continued sliding backward. And then, as the remote-sensing era began in Greenland in the late 1990s, interest in Jakobshavn sharpened further. The glacier’s retreat, measured by satellite and aerial radar, continued—thousands of feet per year. But its speed and iceberg production multiplied. By the turn of the millennium, it was considered the fastest and most active big glacier in the world.
Around that time, it caught the attention of Bob Thomas, the NASA scientist who had spent his early years in Antarctica driving dogsleds and who had bet his career in the 1990s on the idea that a remote-sensing airplane mission could measure the Greenland ice sheet. Thomas was interested in why Jakobshavn’s speed had increased. In its retreat, the glacier had lost an ice shelf that had extended out into the fjord—that was at least part of the explanation. Many years before, Thomas had been one of the first glaciologists to theorize that an ice shelf, a thick floating ledge of ice that extends past the edge of a glacier, is enormously important to how a glacier behaves. As he saw it, the shelf acts like a buttress that holds up the wall of a cathedral, which is to say the shelf prevents the glacier from moving a lot of ice out into the ocean. And yet once that shelf—that buttress—breaks off, ice from the glacier pours out. Some glaciologists preferred to compare the importance of ice shelves to something else: The effect of a cork popping out from a wine bottle tipped on its side. In this metaphor, the cork is the ice shelf, and the glacier’s backlog of ice is the wine.
Thomas had started wondering what role the water temperatures in the fjord played in breaking up the ice shelf. Maybe warmer water had accelerated its disintegration? He started talking about it with a young researcher named David Holland at Columbia University’s Lamont-Doherty laboratory. Holland liked Thomas’s approach to research, especially his belief that theory couldn’t get you too far in science. It was necessary to observe the natural world exhaustively, by remote sensing and by field visits, year after year, so as to answer the difficult questions about its behavior. As Holland remembers it, in 2003 Thomas called to say: “This glacier at Jakobshavn has just fallen apart. It’s going crazy.”
“That’s fascinating,” Holland replied. But he was being ironic. Holland was an oceanographer and mathematician, and had no idea what Thomas was talking about. Most of the time Holland worked on computer models that tried to make sense of the ocean’s effects on the atmosphere, and vice versa. He had grown up in Newfoundland, Canada, and when he wasn’t thinking about math or oceans he was, by his own admission, thinking about hockey. He knew quite a bit about the polar explorers from the nineteenth century who had come to Canada searching for the Northwest Passage and met tragic ends; in some respects, his studies of their journeys had led him toward his own career in polar research. At the time, though, Holland knew very little about ice sheets or Greenland.
“Jakobshavn is a glacier in Greenland,” Thomas explained. “I’m going to get a dogsled team. I’m going there in the winter. You should come.” Thomas explained to Holland that he was intent on researching the fjord that the glacier emptied into and that he planned on bringing a CTD—a standard device that oceanographers drop into the water to measure its conductivity, temperature, and depth. “It was a very romantic idea Bob had,” Holland explains. “The last measurement in the Jakobshavn Fjord was done in the first international polar year, in 1882 or so. And that was done by this guy Hammer from Denmark.” Holland’s point was that even as Jakobshavn’s glacier front had been measured repeatedly, the waters around it had not.
It turned out that Thomas never went to Greenland that winter with a dogsled team. But Holland visited a few years later, in spring 2007, with his wife and fellow researcher, Denise. At that point, the prospect of collecting observations about the water, as well as the calving front, had caught his interest. He and Denise arranged for a pilot to fly them by helicopter from a small airport on the west coast to a location about thirty miles inland. The pilot dropped them off on a rocky scarp of land near the glacier’s calving front. Together, they began measuring the temperature of the fjord and watching the ice. They stayed for about a week. And when they were ready to go back, Holland called the helicopter pilot with a satellite phone, and the pilot said, “I’m not coming today. Good night.” Holland considered this mystifying at the time, but he later found it amusing—a good example of the surprises and frustrations of polar fieldwork. Eventually, when he made it back to the United States, he and Denise agreed to start a long record of data collection. They intended to see if an observational record would help better explain the glacier’s behavior.
By the time of Holland’s first visit it was known that the to and fro movement of glaciers could be affected by many things apart from ice shelves. Gravity and friction—for instance, whether a glacier is moving through a narrow mountain pass—play a part. So do bumps and hills in the bedrock deep underneath the ice. Some recent studies had showed that glaciers could be sped along by meltwater when it flows underneath the glacier, between bedrock and ice. Holland had even come to think that chaos could play a role, too, as ripples from obscure long ago events set in motion unpredictable circumstances. “The weather could have been different 150 years ago on one day,” Holland would say, “and this glacier could be very different because of it.”11
In sum, glaciers were about as complex as a system could get. But on that first trip to Jakobshavn, Holland didn’t think we understood glaciers very well at all. And he reasoned that we might only get so far with GRACE or the IceBridge missions. If we wanted to make accurate predictions about the future of the world’s ice sheets, we would also need to get out in the field and watch Jakobshavn as it collapsed and retreated, even if it took twenty or thirty years. To measure Jakobshavn from the water’s edge, then, would be to exercise persistence in pursuit of a higher ideal. It would test the notion that in understanding Greenland’s fastest and most famous glacier, you might unravel some of the mysterious laws governing the behavior of all the dying glaciers in the world.
Every year since, Holland has returned to Jakobshavn to build up an observational record of the ocean and the ice. He has approached the task from a variety of angles. On the banks of the icy fjord near the calving front, he and Denise set up devices at several locations to collect seismic information. The anticipation is that a rumbling seismic “signature” might lend more clarity to what happens before, during, and after the glacier breaks off into icebergs. One of these camps is known informally as Camp Fox, since Holland once saw a fox there; another is known as Camp Huge Rabbit, because he once spotted a large Arctic hare hopping about there. His main camp—the one near where the helicopter drops him off every June—is situated on a spectacular overlook just south of Jakobshavn’s calving front. He has set up two insulated, prefabricated “igloo” huts there that can each house two people. He has also installed a small radar system, along with solar panels and small wind turbines, so he can focus it on the calving front twenty-four hours a day, seven days a week.
The flight to the campsite is thirty minutes by helicopter. The Hollands bring tables, chairs, tents, and sleeping gear. They haul solar panels and batteries for the new radar, which is packed tight in a blue steamer trunk. A drone system for aerial surveying is brought along. There is food in coolers, food in crates, and large stores of coffee, tea, bread, and Nutella. There is a Coleman stove, as well as generators, a portable toilet, stocks of water in big blue jugs, gasoline for generators in big red jugs, and butane for the stove in big yellow jugs. The chance a polar bear will come through is slight—“Why would a bear come here?” the helicopter pilots asks. “There’s no food, no plants, no life”—but they bring bear alarm fences and rifles as a precaution.
At the camp, Holland walks around with a sense of purpose and a low center of gravity. As a high school student in Newfoundland, he played hockey seriously, even after he lost his two front teeth to a puck. “I’ve never been hit by a crowbar,” he says, “but a slap shot in your face? Being punched is a joke compared to that.” Any ferocity he might have demonstrated in the rink is hard to discern out here. In the field, Holland is jokey and genial. Usually he wears a black fleece zip-up sweatshirt and sports a baseball cap with a fabric flap to protect his neck from sunburn; even in the glaring sunlight of the Greenland spring, he rarely wears sunglasses. Instead he prefers spectacles—horn-rimmed, which give him the appearance of an IBM computer scientist from the early 1960s. Attending to small tasks, Holland walks constantly from one end of camp to another as he consults a small, bound booklet that he keeps in the front pocket of his cargo pants. He removes the booklet constantly to scribble out small notes—scientific insights, sometimes, but usually bits of information such as phone numbers, logistical reminders, or dates. There may be a tool he should have brought but didn’t, and needs to have for the following spring.
By the first evening—the end of the first week of June—the station is mostly set up. Camping by Jakobshavn’s calving front is akin to camping on a spot overlooking a river, except in this case the river is many miles wide and clogged with both large icebergs and shattered ice, the latter of which is so profuse that it forms a frozen stew, known as “mélange,” which covers the fjord waters here. The sun never sets. The visitors watch the glacier’s calving front with a constant sense of expectation. It rumbles ominously sometimes during the night, with a sound like distant thunder. In daytime it tends to emit an occasional and sharp rifle crack, the recoil of a split somewhere deep within itself or the evidence that a small chunk of ice has been shed. Though the front seems high at three hundred feet, the glacier actually descends below the mélange and fjord another three thousand feet or so. Jakobshavn is among the deepest fjords and glaciers in the world. Therefore a true calving event, one that cleaves the glacier from top to bottom, is bigger and more explosive—not a rumble, not a crack—and unleashes the energy of several atomic bombs. One morning a few years back, the Hollands awoke at three A.M. and had the good fortune to capture on video the breaking of a massive berg that measured twice the length of the Empire State Building. “It was very noisy, very spectacular,” David recalls. “We were impressed.”12
The glacier’s riverbanks, instead of sand or grass, comprise a hilly moonscape of dust and rocks, ranging in size from skipping stones to boulders as large as refrigerators. Decades before, the big stones must have been moved about effortlessly, and then left behind, by the retreating movement of the glacier. In fact, this area happens to be called “new land”—that is, land that was covered in ice as recently as twenty or thirty years ago but has now been unveiled thanks to the recession of the ice sheet and glacier. Gray, colorless, and devoid of vegetation, new land is grim. Tens of thousands of years ago, when glaciers retreated from northern Europe and North America, many parts of the world looked like this—a place that exists before life arrives.
As it is, new land makes simple tasks unpleasant. When a cold wind blows down from the ice sheet, microparticles of till, a fine rock flour ground down by the glacier, kick up into the air. A film of grit descends on the lenses of eyeglasses, on the keys of a computer keyboard, on the skin of a boiled egg at breakfast. It settles on the cusps of molars. Meanwhile, the dry lunar wastes of new land seem even stranger in contrast to what’s around: The sound of water, flowing and rushing, circles the campsite. It comes from rivulets that trickle down the edges of the ice sheet a few hundred yards behind the campsite, as well as from fast-moving rivers in front of the campsite, hidden from sight, which run a few hundred yards down the hill and pulse under the edges of Jakobshavn’s glacier itself. What seems like wind, in fact, is just water running, far off and nearby, hard to place but difficult to ignore, an enormous, unceasing white noise, which is merely the sound of ice melting, all day, all night, everywhere.
A few years back, Holland decided to informally call his efforts in Greenland HELISHE. “The whole idea was that this was going to be difficult, and we actually liked the word hellish,” he explains. HELISHE was not a true acronym, but Holland didn’t care, since it amused him anyway. H is for hydrography; I is ice; S is seismic; E is for experiment. “It was important to have the acronym first,” Holland says, “and the words fit pretty good.”
HELISHE is conducted every June at Jakobshavn and every August on the other side of Greenland, on the east coast at Helheim Glacier. The experiment has two main areas of inquiry. The first involves the study of what Holland calls “the mechanical part”—the calving process that produces icebergs. “It’s a fracture process not really understood,” he says, “but people are making progress, so we are trying to contribute to that.” The other aspect of the study, meanwhile, involves the arrival of those warm waters that not only melt the face of the glacier but can also erode the glacier’s perch on hard land—what’s known as its “grounding line”—many thousands of feet below the surface. “The waters have a reactionary force, and the glacier responds,” Holland says.
These two parts of the Jakobshavn research subdivide into a variety of different endeavors. In addition to Holland’s seismic sensors at Camp Fox and Camp Huge Rabbit, his radar at the main camp focuses on the calving front to measure the moment-by-moment movement of the glacier, which in June 2016 seemed to be moving about sixty feet per day. Every summer, moreover, Holland spends time dangling from helicopters in a dry suit so he can drop probes into the icy mélange of the fjord in front of the glacier; his intent is to measure the temperature and salinity of the water at various depths. More recently, he has also begun dropping probes into gaps in the glacier itself, behind the calving front and closer to the ice sheet, so as to measure “stretching properties” of the ice and thereby gain insight into what may be happening as the ice flows and strains before breaking off. All of this work is finally complemented by a week in a boat, which follows the week of camping by Jakobshavn. Holland cruises around Disko Bay, far from the fjord and the calving front, to deploy temperature probes and other robotic tools. The point is to gather more data on the salinity and temperature profile of the surrounding waters at various depths and see how they infiltrate the fjord and affect the glacier.
“The thing about the helicopter and the boat is that it’s human intensive,” Holland says. “You do get data, but it takes a lot of work.” It can also be dangerous. Even if a captain could pilot a small boat through the icy mélange to get to the calving front—a place where a big iceberg may break off—it could be deadly. Over the last few years, Holland has settled on a more efficient way to get information all year round. With the help of a Greenlandic marine biologist named Aqqalu Rosing-Asvid, he augments his field observations by outfitting temperature sensors on local seals and halibut that swim in the fjord. The seals tend to go down about twelve hundred feet or so. “So they’re constantly covering the top half of the fjord,” Holland says. “And by some miracle of nature, the halibut cover the bottom half of the ice fjord.” That means they swim in the icy depths down to three thousand feet or so.
Using a satellite network, the seals phone in their data several times a day—information that is routed to a rooftop antenna near Holland’s office at NYU, in Manhattan’s Greenwich Village. “The cost per measurement is really small,” Holland says. “And the data—well, I don’t know if any technology can compete with seals for that purpose.”
One afternoon, a pilot arrives in a small helicopter, and Holland takes short flights across the fjord, to Camp Fox and Camp Huge Rabbit, to upgrade the seismic instruments. Meanwhile, his students keep the radar focused on the calving front and set up the solar panels and batteries. In the near distance, Jakobshavn rumbles and cracks. As the sun dips toward the horizon late at night, the temperature drops into the thirties, but when the sun is high during the afternoons the approach of summer seems obvious. It’s getting warmer. The mosquitoes are coming out. The ice sheet can sense that, too. Day by day, the sound of rushing water grows louder.
The past decade has already yielded evidence to Holland that warming waters act as a kind of control knob to accelerate the speed of the glacier.13 But the “mechanical” aspects of iceberg formation—how and why an iceberg breaks off—remain largely mysterious. Several recent studies suggest that once a glacier front reaches a height of about three hundred feet, which is about the height of Jakobshavn’s front, the cliff can no longer sustain itself, and the ice inevitably breaks apart, or calves, from a kind of material failure. Sometimes this failure might be accelerated by meltwater on top of the glacier that eats its way into the glacier from above and hydrofractures the end of the iceberg.14 But Holland believes there are likely other reasons glaciers calve and that we don’t yet understand the conditions that might determine the process of breakage. Amongst polar researchers, the pursuit of these “calving laws,” as they’re sometimes called, comprises a kind of holy grail of glaciological research. To find them would be akin to being able to predict the conditions for, say, an earthquake in San Francisco or Tokyo. At the moment, that prediction happens to be impossible.
“As I see it,” Holland says one afternoon, “the nature of the problem is one of sea level, and sea level is affected on fast time scales when warm water comes near glaciers in Greenland or Antarctica.”15 There is a crucial aspect to what he is saying that can be easily missed: He is looking at changes on fast time scales. To a glaciologist, fast might mean decades or even centuries. In either case, however, it happens to be the reason why glaciers known as “outlet” or “marine terminating” carry such importance. They flow from the edges of ice sheets and end at the ocean. And in a hotter world, they have the potential to quickly move massive amounts of ice from ice sheets into the ocean.
But no one seems to know precisely how quickly. Holland’s belief is that the work at Jakobshavn—whether it takes another ten or twenty years—will provide information to create better computer models. Without these observations, the task of projecting Greenland’s future, or estimating how high seas will rise in the next one hundred to two hundred years, may not be feasible. One evening, sitting in the dining tent at the calving front, Holland says, “In oceanography, the models are basically complete, and in meteorology they’re essentially complete. But in glaciology they’re incomplete. So, you write down an equation, and someone could come over with a flashlight and say, ‘Where did you come up with that?’ ”
They’d be right to do so, he believes. Even though climate models have become better at predicting future temperatures and precipitation patterns, he says, ice sheet models have not reached nearly the same level. The reasons why glaciers move and break (buffeted invisibly by oceans, air, gravity, bedrock hills, and the properties of ice) have proven extraordinarily complex. And Holland doesn’t believe that we can build good models from theories and ideas—we need to look first to nature, so that it can teach us how it works. Pointing outside the tent to the calving front, Holland says, “This is the way it is, and your model has all these ways it could be different. If you’re going to build a computer model that has all these different choices you’ve made, which are a thousand million different choices, why would I believe the one you just came up with?” The best way to make a good model is to see what happens, he insists. Then we can build it into the code.
Perhaps thirty or fifty years from now, he says, the glaciology models will be robust.
He knows that is a long time—maybe too long—to wait. For one thing, the stakes for the glacier at Jakobshavn appear to be significant. It drains about 7 percent of the Greenland ice sheet, and its continuing retreat and collapse could someday boost sea levels by more than a foot. But in looking at Jakobshavn, Holland is also thinking about a glacier in Antarctica, located in the western part of the continent’s massive ice sheet, called Thwaites. It was named many years ago for Fredrik Thwaites, a geologist from the University of Wisconsin.
Thwaites is sometimes described as the Doomsday Glacier.16 Like Jakobshavn, it has recently come in contact with warming ocean water. And like Jakobshavn, it is precariously poised on a deep bed that gets even deeper as it moves inland, making a sudden retreat (and a massive dump of icebergs into the ocean) much easier. The collapse of Thwaites, however, would likely have a far greater impact on sea levels than Jakobshavn, since it is about the size of Great Britain. The result would be at least two feet of sea level rise, and a possible and subsequent collapse of the entire West Antarctic ice sheet, which holds enough ice for about twelve feet of sea level rise. In 2014, two scientific studies based largely on remote-sensing data, one led by Eric Rignot at the University of California, Irvine, another led by Ian Joughin at the University of Washington, declared that Thwaites appeared to be in or close to the early stages of collapse. There was no settled conclusion on how fast it would happen—many decades or even many centuries seemed possible. Nevertheless, the glacier was starting to move massive amounts of ice into the ocean. “The collapse of this sector of West Antarctica,” Rignot declared, “appears to be unstoppable.”17
Holland has been to Antarctica seven times. For a long while, he was interested in a retreating glacier next to Thwaites called Pine Island. He spent ten years planning an expedition there, only to later realize that Thwaites was probably less stable, although both glaciers may in fact be in perilous condition. “That’s an important glacier,” Holland says of Pine Island, “but I think Thwaites is really the focus. We picked the guy next door.” The problem is that you can’t just return to these places on a whim for a closer look. The difficulty of getting to Thwaites—possibly the most remote glacier on earth, and a place only a few dozen human beings have ever walked upon—makes anything beyond remote sensing exceedingly difficult. In part, this is why predictions about its speed of decline contain large margins of error. “You cannot make a laboratory model for Thwaites,” explains Holland. “And if you build a computer model, you have to make so many choices that the permutations are on a scale of a billion.”
Holland thinks that we may not be able to predict the effects of Thwaites in time. The difficulty in gathering facts about it leads, in turn, to deficiencies in our computer models for ice sheets. “It’s hard to imagine we can gain knowledge quick enough,” he says. “If Thwaites was to go bad, and it collapses, I think we’ll be observing it.” Perhaps only then will we understand it. Meanwhile, in Jakobshavn, Holland has found a real-world model for collapse. He can get here in a few days and visit several times a year. He can watch every aspect of the ice and measure every aspect of the surrounding waters.
“I think this is a faith article—that Jakobshavn is an analog for Thwaites,” he says one evening, walking out to the high rocky bluff where his radar is trained on the calving front. Still, he doesn’t seem to be someone who acts on faith. He starts taking out his notebook from the front pocket in his cargo pants to jot down a reminder. The dust is kicking up a bit again. The calving front is quiet for now. “I don’t have any other cards to play,” he says.