The long haul
Antarctica is a puzzle that science is racing to solve. The continent shifts from stable to unstable, frozen to melting, without much warning—and we don’t know why, or how. This switch hasn’t taken place in the century we’ve been observing it. But Antarctica has its own records that go back millennia, buried in the sea floor beneath hundreds of metres of ice. To retrieve them, a New Zealand-led expedition journeyed to the heart of the Ross Ice Shelf—a featureless, inhospitable expanse the size of France.
Two sets of merino underwear, two fleece jackets, a windbreaker, a puffer jacket, five pairs of gloves, a hat, a balaclava, fleece pants, fleece-lined boots and a set of extreme-cold-weather gear—a bulky jacket, large enough to enfold two people, and chunky, clown-like boots—are issued to me at Antarctica New Zealand’s Christchurch headquarters.
I’m told to return the following day at 6.00am, wearing the extreme-cold-weather gear—ECWs, they call it—for the flight south.
The next day, I’m woken by the sound of my phone, buzzing in the darkness. It’s 4.45am.
“Ice flight delayed 24 hours due to weather,” a text message says.
The same message arrives five mornings in a row, and I learn my first lesson about Antarctica: dates are only estimates.
On the sixth morning, no message arrives. I put on the ECWs and head for the airport, where I board a United States Air Force Globemaster C-17 along with 80 other passengers. This is one of the first deliveries of scientists to the ice for the summer, as it’s still early in the season. Dressed in our ECWs, we’re ready for temperatures as low as -40°C.
In five hours, I’m transported from the spring warmth of Canterbury to a frozen world, and I emerge into a place that’s brighter and colder than I had imagined. I get my first glimpse of Mt Erebus, towering over Ross Island, and the expanse of white desert that is the Ross Ice Shelf. Beyond, the Transantarctic Mountains stretch out to the horizon.
I’m joining a mission that’s been two years in the planning—a deep-field traverse, a new way of transporting large-scale science projects further into the polar region than before. Normally, science teams are dropped off via plane, but this time, a convoy of heavy-tracked vehicles will drive over the ice. These can drag many more tonnes of equipment than a plane can carry, and can travel in almost any weather.
We’re taking a hot-water drill to the middle of the ice shelf. Then, the team will bore through 350 metres of ice—a distance greater than the height of Auckland’s Sky Tower—to where it meets the ocean below, then continue another 350 metres down to the sea floor. There, they’ll collect data about the ice, ocean currents and biodiversity, and take samples of sediment. Much of this is new; many of these measurements are being taken for only the first or second time. What lies beneath the ice shelf isn’t incorporated into our climate models or included in the historical record. If all goes well, the deep-field traverse will fill in some of the very large gaps in what we know.
One of the scientists conducting research this year is glaciologist Wolfgang Rack, a polar veteran who has spent 14 seasons on the ice, much of it studying the behaviour of ice shelves.
Rack’s previous work took place on the Antarctic Peninsula, a narrow, jagged spit stretching toward the southern tip of South America. Thirty years ago, it had three significant ice shelves—Larsen A, B and C. But the peninsula is warming faster than any other part of the continent. Rack points to an area of ocean on the map spread out on the table.
“That used to be Larsen A, right there,” he says. “It was the first to go.”
The collapse of the Larsens was thought to be insignificant on a global scale—ice shelves don’t affect sea level because they’re suspended on the water—but within a few years, scientists noticed that the glaciers that fed the ice shelf had sped up. And not by a little bit—they were flowing at least four times faster. The ice shelf had been acting like a plug, and now one of the blocks to glacial flow had been removed.
“What really shocked us was the rate of change of the glaciers,” says Rack. “We really were not expecting that.”
Today, of the three ice shelves, only Larsen C remains. But the missing Larsens are small on the scale of Antarctica, as are the glaciers and ice sheets they held back. The Ross Ice Shelf is 50 times the size of Larsen B, and it’s the plug for the West Antarctic Ice Sheet, which is up to 3000 metres thick. If it melts, global sea levels will increase by about five metres.
Rack and I are sitting in front of a large window that looks out over the Ross Ice Shelf and, gazing out, I wonder what this view might hold by the end of the century.
“I think you’ll see an ocean here,” says Rack. “The ice shelf will be gone.”
The Larsens were a lesson in how ice shelves react to atmospheric changes. But ice-shelf stability also depends on the temperature of the seawater underneath. So how warm is the ocean below the Ross Ice Shelf? It’s one of the measurements the deep-field traverse is planning to obtain.
The day before our departure, I find Rob Teasdale, the traverse’s team leader and mechanic, head down over a fuel pump, tools strewn across the workshop floor.
“Might be leaving the day after tomorrow now,” he jokes. He’s been planning this trip for more than two years, part of a team of Antarctica New Zealand staff and contractors. Back home, he works at the Mt Hutt ski field, maintaining snow groomers and tracked vehicles.
One of the traverse’s biggest logistical challenges, he says, has been planning the right amount of fuel.
“It’s really hard to know how much we’ll use—there are just so many variables. Too much and we’ll be weighed down and slow, too little and we won’t make it. It’s a real balancing act.”
In the end, Teasdale decided on 33,000 litres of AN8 aviation fuel, which is often used in diesel machinery—enough fuel to circumnavigate the globe at least 12 times in a small car. We’ll drag it behind our convoy in soft bladders attached to sleds.
After reaching the hot-water drilling site, Teasdale and a small team will continue to Siple Coast, at the edge of the Ross Ice Shelf, along a route no one has previously tried. There’s no water on Siple Coast—rather, it marks the dividing line between the ice shelf and the West Antarctic Ice Sheet.
The next day, Teasdale walks around the vehicles assembled for the traverse, checking the straps securing the sprawling load. Our convoy consists of three PistenBully tracked vehicles, similar to snow groomers, and a Hāgglund, which is smaller and has a double cab. We’re towing about 60 tonnes of fuel and equipment, enough to set up and maintain a small, self-sufficient village for two months. It looks like a classic Kiwi camping trip gone very, very wrong.
Warning: Heavily crevassed area. Road on right of flags. Proper equipment and training required.
There are a couple of grim reapers carved into the sign to emphasise the point. We’ve reached the shear zone, a region created by two sections of ice shelf shifting and grating against each other. It’s the end of our first day on the ice, so we pitch camp at the edge.
There’s a marked route through, where a path has been stabilised using explosives. Even so, over the next few days, we put on helmets and harnesses and no one leaves a vehicle without being attached by rope. As I gaze out the window, the shear zone looks smooth—crevasses leave no visual clues. Only ground-penetrating radar will help us find our way.
Mapping experts Dan Price and Lawrence Kees guide our convoy in an old yellow PistenBully. The radar, mounted on a boom in front of the vehicle, fires pulses of energy into the snow pack, and the return signal is displayed on a screen in the cab as a series of lines. Straight horizontal lines mean the terrain is safe, while close-together curves, forming an egg-like shape, signal a crevasse ahead. In safe areas, Price and Kees drive at 10 kilometres an hour, a speed at which they would have only seconds to stop if a crevasse was detected. In less-certain terrain they slow to walking pace.
When a crevasse is discovered, one person is belayed on a rope to probe the ground, determining the direction in which the crevasse is travelling—its ‘strike’. Price and Kees then draw a profile of the crevasse and decide whether to cross. If it’s narrower than one-third of the length of the vehicle’s track, then it’s considered safe to drive over. Anything wider, and we have to go around it. But we’re lucky: the route is good and we drive straight over them all.
“Coming through the shear zone, we could see crevasses on the screen all the way through,” says Price. “Driving through without following a route would be like walking through a minefield—you’d be throwing the dice every time.”
This shear zone is easy to cross compared to the next, on the other side of the ice shelf, which the team will traverse on their way to Siple Coast. To prepare for it, Price spent six months planning a route through the maze of crevasses, using satellite and radar imagery captured between 2011 and 2017. This involved a certain amount of guesswork, because the Ross Ice Shelf isn’t static—it moves up to 700 metres a year, bulldozing its way north toward the Southern Ocean under the pressure of hundreds of inlet glaciers and vast neighbouring ice sheets.
Price used ice-velocity data to redraw the maps and determine where the crevasses were likely to lie during the traverse.
“The images we have are just a guide, really,” he says. “It’s an imperfect system but it’s the best we can do.”
Days before the traverse departed, the United States Antarctic Program came to the rescue, redeploying a satellite over Siple Coast to capture up-to-date high-resolution images, helping Price to paint a more accurate picture of the crevasse field.
Life on the traverse is simple. The convoy crawls along at 10 kilometres an hour—the speed most people comfortably jog—for up to 18 hours a day. We stop only to eat, sleep, and refuel the vehicles. Of the 10 of us, each person has a critical and particular role. PistenBully driver Bruce Davies is a traverse veteran, having completed many similar journeys in Greenland. He’s also the only one of us who’s ever driven into a crevasse.
Jeff Dunne drives the Hāgglund, and he’s also in charge of setting up the deep-field camp, when we reach the hot-water drilling site, for the scientists who’ll be arriving by air. Teasdale and Kurt Roberts are diesel mechanics specialising in the heavy equipment we’re carrying. Richie Hunter had only just returned from a season guiding on Mt Everest when he was offered the job as the field trainer in charge of safety on the traverse. His five seasons in Antarctica are continuing a family legacy—his great-grandfather Bill MacDonald visited Ross Island in 1913 when the Terra Nova returned to meet Robert Falcon Scott’s expedition.
Ice-drilling expert Alex Pyne is in charge of the hot-water drill, and the expedition has its own chef, Lana Hastie—but her job won’t begin until the kitchen tent is constructed.
Meals on the traverse are mostly boil-in-the-bag army rations, eaten outside in the cold, or our team of 10 crams into a cabin designed for just four. Doing the dishes involves swishing warm water around your bowl and sculling it before it freezes. We nickname it ‘dish soup’.
The toilet is a bucket in a tiny uninsulated cubicle towed behind a PistenBully, and our sleeping quarters aren’t any warmer. Five of us share a space three metres wide by two metres high. To stay warm, I snuggle into a fleece liner, wriggle into two bulky sleeping bags, and finally, pull over an outer protective bag. Even then, four layers in, I’m only just warm enough.
On the third day, the temperature plummets to -34°C. To work outside, we have to ensure no skin is showing.
Looking out the window of the Hāgglund at the snow drifts being pushed around by the wind, I think about the early explorers crossing the Ross Ice Shelf. Our route closely follows that of Scott’s expedition in 1911 and 1912. The men who attempted to reach the South Pole had little protection from the elements, and their success depended not just on planning, but on the weather and their animals’ health.
Buried somewhere beneath the ice lie the remains of the Scott and Shackleton expeditions—food dumps never reached, men, dogs and horses who perished along the way, and the tent and bodies of Henry Bowers, Edward Wilson and Scott, now on their final journey, carried by shifting ice toward the open sea.
Out here, time has little meaning. We cross an endlessly flat, featureless expanse under 24-hour sunlight. In my mind, the traverse from Scott Base to the science campsite merges into one long day.
In reality, we’re four days in when Teasdale calls over the radio: “We’re here!”
‘Here’ looks a lot like the other 360 kilometres of the Ross Ice Shelf that we’ve crossed. Someone jokes about abandoning the mission and continuing on to the Pole: “We’d be the first motorised Kiwi team since Hillary.”
Instead, we spend the next two weeks building a canvas village, when the weather allows. This involves setting up more than 30 tents, ranging from the humble four-sided Scott Polar to the Polar Haven science tent, which is large enough to hold a four-bedroom house, with room to spare. The kitchen tent has running water, hot and cold, as well as 24-hour electricity, and the two largest science tents are also fully powered.
“There’s a huge amount of effort put in behind the scenes to allow the science to happen,” says Dunne, the camp manager. “Nothing in Antarctica is easy—everything here takes twice as long as it would at home.”
The average summer temperature for the Ross Ice Shelf is -7.5°C, but it’s not -7.5°C, at least not while we’re here. The thermometer rarely rises above -15°C.
Antarctica isn’t the fairy-tale land I’d imagined it to be. The ice sparkles in the sunlight, but the wind and temperature are brutal. Stepping out of a fire-warmed room at -30°C, it’s hard to breathe, and the cold burns any exposed skin. Surrounding the camp is a washed-out, desert-like landscape. No mountains, no hills. Not even a bump.
The only sign of life outside our camp is the United States South Pole resupply mission, which comes within a kilometre of us as it trundles past. Every summer, a team from the United States Antarctic Program makes the 2600-kilometre return journey from McMurdo Station to the Pole, where about 200 people staff the Amundsen/Scott South Pole Station. Other than the resupply crew, we’re probably the only form of life on the Ross Ice Shelf.
On top of it, at least. Far beneath our feet is a dark, not-quite-freezing world teeming with life: the Ross Ice Shelf Cavity, the largest-known enclosed space on Earth.
Now that the camp is established, more than 20 glaciologists, geophysicists, microbiologists and oceanographers fly in to meet us.
To open a portal into the world below, Pyne and his drilling team from the Antarctic Research Centre at Victoria University have designed and built a modular hot-water drill. It works like a shower head: it has a hose with a fine nozzle on the end of a weight, which sprays hot water downward to melt a hole. Just above the weight, a reamer sprays water outwards and, if needed, upwards, enlarging the narrow bore to the diameter of a soccer ball. The nozzle can deliver 180 litres of water per minute—by contrast, the average home shower uses eight litres per minute—at a temperature just below boiling point. This water is melted from nearby snow, which is shovelled into portable flubber tanks and heated by a series of boilers powered by three large generators. Once the hole is drilled, there’s a risk of it freezing over again, so drillers and scientists must work in tandem. Periodically the hole is reamed, to remove the ice that forms on its walls.
“For me it’s like going to Mars,” says lead oceanographer Craig Stevens from NIWA. “We know almost nothing about the ocean beneath the ice shelf—a hidden ocean the volume of the North Sea.”
Stevens and his team are studying the turbulence of the seawater beneath the ice shelf.
“When you look at clouds contouring across a mountain range in wave-like forms, those same types of processes happen under the ocean. The amount of heat transferred in those processes is critical to how much ice will melt below the Ross Ice Shelf.”
Of particular interest to Stevens is the area beneath the shelf where ice meets water, a mixing zone five to 30 metres thick. The water does not freeze, even though its temperature is between -1 and -2°C, because of its saltiness and the pressure of all the water and ice above. Channels in the ice direct the flow of the water beneath it, and the temperature of this zone determines the amount of ice that melts beneath the shelf.
It takes about two years for water to circulate underneath the Ross Ice Shelf. When it flows in from the Ross Sea, at the level of the seabed, it’s warm and salty, and it continues to the grounding line, where the shelf meets the ice sheet. There, the seawater rises, turns back and begins to melt the shelf from beneath as it flows along its underside back towards the open ocean.
“We believe the ice shelf is currently stable, but it wouldn’t take much change in the ocean currents to push warmer water beneath it—then we could start to see dramatic changes,” says Stevens.
We know the beginning and the end of the Ross Ice Shelf’s story, but the middle is murky. During the last ice age, which ended about 15,000 years ago, the ice shelf was an ice sheet, solid all the way to the sea floor and extending further north than it does today. As the planet warmed up, the ice sheet retreated, and then lifted off the sea floor, becoming a shelf.
“The end of the last ice age was basically a natural global-warming experiment,” says geologist Christian Ohneiser. “The world warmed and things changed—and it wasn’t forced by humans. If we can figure out how quickly the ice retreated during that warming experiment, we can maybe say something about how it will behave today. How quickly can things change without us mucking with the system? How quickly can the ice retreat without us pushing it?
“If we can figure out how Antarctica changed in the past, we know what it’s capable of and how it will react today.”
A team of geologists, led by Gavin Dunbar from Victoria University, is hoping to extract a sediment core—a metre-long tube of mud— from the sea floor. They only need one. Its layers of mud and rock and fossilised algae may unlock the mystery of the ice shelf’s past.
Sediment cores from the edge of the Ross Ice Shelf show that it lifted off the ocean floor about 8000 years ago. Dunbar wants to know how long it took for the middle of the ice shelf to start floating, too. Decades, centuries, millennia? The team has no idea; apart from one sediment core taken far to the south in the 1970s, all others have been extracted from the edges of the continent, or offshore.
“No one ever in the history of humanity has collected one from the middle of the ice shelf,” says Ohneiser. “If this is dated only a short time after the front of the ice shelf, we know the retreat is very fast. But if the lift-off is 3000 years ago, then maybe it’s not very sensitive.”
Figuring out how quickly Antarctica melts is crucial, because it’s probably going to melt again, says Ohneiser, and we probably won’t have a lot of warning.
“Antarctica does things in pulses. For centuries, it will do nothing, and sit there and be a nice cold iceberg. Then it will switch to a retreat mode when it dumps lots of ice in the ocean, and the ice streams are flowing. The really interesting thing is the switch. It’s not a ramp, it’s a binary—it’s an on-off switch.”
At the hot-water drilling site, weather complicates things. It’s foggy every day except Sunday, the one day of the week that the Twin Otter aircraft at Scott Base stays on the ground. This means a United States team with a small remote-controlled submarine isn’t able to fly in, and microbiologists on the ice aren’t able to fly out with freshwater they’ve drawn from the cavity.
But work goes on. Scientific tools are re-purposed to fill in for what is missing. The microbiologists drive back to Scott Base instead, and after two days on the road, arrive just in time to culture their samples. Stevens has brought some back-up gear, which captures the imagery the submarine was intended to collect. Weather stations are installed on a transect, long-term monitoring equipment is moored within the cavity, and science lead Christina Hulbe’s team drives around mapping the ice-shelf interior with radar.
Within the cavity, the ocean is a little warmer and fresher than it was in the 1970s, which doesn’t surprise anyone. But it’s full of eddies and ripples, which does.
“This is an ocean that doesn’t see the wind, and it doesn’t see the cooling from winter and summer that happened a year or two earlier,” says Stevens. “It’s entered this protected quiescent cavity, or so we thought. But actually, the tides still get in there and are lifting this 300-metre lid up and down. It’s actually bobbing up and down every 24 hours, and that sends out underwater waves.”
From this data, Stevens and his team will be able to test how well models can predict the flow of the ocean beneath the cavity.
“There are two classes of model here,” he says. “We’re trying to understand what we’ve observed with one, and then we take that to the bigger model, and try to predict the future. These future-looking models don’t have the ice-shelf cavities as part of them, but we’re pretty certain that these ice-shelf cavities have a number of effects, which will prove to be an important part of the climate system.”
The underside of the ice shelf also has a surprise for the scientists—a thin layer of ice crystals, which have formed in the freshwater closest to the ice.
“A big challenge for us now is to understand why they’re there,” says Hulbe. “We know the process that’s responsible for this. We don’t know why it’s happening at this spot.”
The deepest glacier ice, just above the new crystals, is shot through with sediment—not from the sea floor, but collected from the Antarctic continent 400 years ago. The underside of the shelf hasn’t melted enough to allow the sediment to fall to the sea floor.
“The ice has been holding itself in a kind of a reasonable equilibrium since leaving the coast maybe hundreds of years earlier—that’s a little bit unusual,” says Stevens. “It’s quite an exciting thing to find in terms of understanding where the system is going to go in the next 50 to 100 years.”
After three weeks at the hot-water drill site, Rob Teasdale and the traverse team depart for Siple Coast, leaving the drillers and scientists to their work. Teasdale has allowed 40 days for the round trip to the coast, but they reach it in eight days—two weeks ahead of schedule. Instead of turning back, they continue to a United States field camp on the West Antarctic Ice Sheet to collect supplies for next year.
The success of this route means that next summer, a hot-water drill site will be established at Siple Coast, at the grounding line where the West Antarctic Ice Sheet begins to float, becoming the Ross Ice Shelf. This area of the cavity is like a beach, says Stevens—below, water meets solid ground, causing wave-like action.
Meanwhile, Gavin Dunbar and the geology team extract 11 sediment cores—10 more than they were hoping for. Each of these successfully collects up to 65 centimetres of mud before striking compacted sediment, deposited during the last ice age, which is too solid for the equipment to pierce.
As this issue went to print, the all-important sediment cores were still in quarantine in Christchurch, awaiting their trip to various laboratories for analysis.
Then, the composition of the sediment layers will tell Dunbar and his team whether an ice sheet, an ice shelf, or the open ocean was above the sediment at the time it was formed. A compacted mix of mud, sand and gravel is created by the weight of an ice sheet; mud with the odd stone in it indicates an ice shelf (large stones tend to fall out near the shelf’s grounding line). If the sediment contains fossils of microscopic plants, then ice wasn’t covering the surface of the ocean, as they require sunlight to grow.
Dunbar says it’ll be a year or so before they’ve accurately radiocarbon-dated the cores for the date of ice sheet lift-off—the missing piece of the story. There will likely be few carbon-based materials in the sediment, and much of that will be recycled carbon that can’t be used to date the samples, so he’s expecting the process to take a while.
“It’s a real challenge because no one’s ever dated sediments from underneath this part of the Ross Ice Shelf,” he says. “That middle measurement is priceless.”
Back in Wellington, Craig Stevens continues to receive data via satellite from the Ross Ice Shelf Cavity, thanks to the measuring devices moored within. He gets a daily update of readings taken every half hour—temperature, salinity, current.
“Those three things are the fundamental properties—if we can get those, that gives us the sense of how the system is working.”
Depending on how the batteries last, he may receive data for several years.
But for now, the formidable task of interpretation begins. The scientists who worked together on the ice will have their first meeting on the mainland in March to share their findings, and the data gathered from the Ross Ice Shelf will likely occupy them for years.
“It’s a more interesting place than we thought—for reasons we didn’t think,” says Hulbe. “That’s the reason why you have to go to places that are hard to get to and do things that are hard to do. Because you can’t observe any of this from space. You can’t observe any of this by sending a robot to drive along the surface. You have to go there in person and do the hard work.”