Located on the northwest coast of Ellesmere Island just above the 82nd parallel, Milne Fiord is about as far from Toronto as Canada is wide . The roughly 55-kilometre-long and 5-kilometre-wide fiord is buttressed on either side by steep mountains and rock outcrops. At the mouth of the fiord is the Milne Ice Shelf — ice shelves are large slabs of floating ice that are attached to land but, unlike land-fast sea ice, are extensions of land-based glaciers. Moving away from the sea-level ice shelf and further into the fiord, the elevation builds up to 1000 metres. The sheer rock faces that line the mouth of the fiord slowly become enveloped as the glacier eventually joins a large inland ice cap.

A sense of scale is at play that doesn’t let the mind comprehend what it’s seeing. This thought is only compounded by the numerous tributary glaciers feeding into the fiord that rival many of their remaining icy cousins, and pale in comparison to Milne glacier itself.

The incomprehensible scale of Milne Fiord isn’t just limited to its physical characteristics; there’s a sense of geological time in motion in this place. The ice shelves of northern Ellesmere were likely formed between 4,000 and 5,500 years ago, making them the oldest sea ice in the northern hemisphere and roughly as old, if not older, than the Pyramids of Giza — a hard concept to grasp for those who, relying on the otherwise ephemeral nature of ice, relegate spilled ice cubes to the area underneath their fridge with a swift kick.

While the scale is incomparable, the reality is ice can and does melt under the right conditions, whether that be a cube or an ice shelf. At the start of the 20th century, there was a single 8,600 square kilometre ice shelf encompassing the entire northwestern coast of Ellesmere Island.

“Those ice shelves that are along the coast have been in decline over the last 100 years or so,” says Mueller. During that century, the shelves extent decreased by approximately 90 per cent. By the early 2000s, only six significant shelves were left , but even those were not safe.

“In the last 20 years, they’ve been rapidly reducing in size,” adds Mueller. Between 1999 and 2015, northern Ellesmere Island’s ice shelves further decreased by roughly 40 per cent , with significant breakup events happening in 2003, 2005, 2008, 2010 , 2011 and 2012 — the 2005 event eliminated the mighty Ayles ice shelf. Despite all of this, Milne remained. Milne was considered less vulnerable to breakup as it was roughly twice as thick as other ice shelves and more sheltered by land in Milne Fiord.

While the satellite images that first alerted White to the shelf’s collapse were helpful, the consequences of the calving event are far more nuanced than what can be glimpsed remotely. “Remote sensing is a super amazing tool,” says Mueller. “You can see what the satellite sees as it flies over, but you can’t take a profile of the ocean.” This is why Mueller has been going there for the better part of 15 years — and is returning to figure out what happened during the breakup.

In July 2022, when Mueller and his colleagues first stepped into the large common room of the Polar Continental Shelf Program base — the launching point for any High Arctic research work — they were greeted by a work of art not dissimilar to something you might find scrawled in the notebook of a highschooler who endured a little too much detention. A whiteboard in the corner of the room read, “Record Breaking 15 days at PCSP Resolute.” It was surrounded by elaborate doodles of stick figures and an airplane that read “jinxed,” the illustration of a knitting fox with a caption above it that said “What the fog!?” and scrabble-looking tiles arranged together to read “Get us out.”

The first three days at the base in Resolute Bay, Nunavut, were spent preparing and planning for the next 20 days of fieldwork. But very quickly, three days turned into 5, then six and longer. At some point, the days felt like a riff off Groundhog Day, set at a facility that sometimes felt like a strange summer camp and at times felt like a sleepy retirement home. It was the perfect situation to allow the big questions surrounding the state of the fiord to fester. One of these questions was: had a rare Arctic ecosystem nestled between the ice shelf and the glacier vanished?

Unlike the rough and rolling surface of the glacial ice, there are distinct patches of year-round flat ice in the middle of the fiord that look similar to what might be found on a local pond. Like a local pond, beneath that ice is a layer of freshwater. However, the similarities stop there because the freshwater under that ice in Milne fiord floats on top of denser seawater and is prevented from going anywhere by the ice shelf that acts like a natural dam. It’s called an epishelf lake. Effectively cut off from the elements, these lakes are part of unique ecosystems.

“We’ve looked in the past at the whole coastline and basically pieced together that there could have been up to 17 of these lakes,” says Mueller . As the ice shelves continued to fracture and collapse, so did the epishelf lakes. In 2002, as a PhD student, Mueller discovered the loss of the 4,500-year-old epishelf lake in Disraeli Fiord, which had been dammed in by the then recently fractured Ward-Hunt Ice Shelf. That work represented the first time an epishelf lake had been measured before and after draining.

Milne’s epishelf lake was the last known remaining in the Arctic. In 2011, a colleague of Mueller’s had placed a mooring — a long line of instruments spaced out at critical intervals that take measurements such as salinity, temperature and depth — in the Milne epishelf lake to record what was happening in real-time. It had stayed in the fiord ever since.

“As far as I know, it’s the longest-running mooring in a Canadian Arctic fiord, and that’s a fantastic record to have,” says Mueller but vitally, he adds, “it may well have caught this event.”
If it did, it would be the first and last time such an event could be captured in real-time in the Arctic. The other questions that festered while in Resolute Bay were: if it was still there, could it be recovered? And: was it still recording?

As the red and white De Havilland Twin Otter banked right and began its descent, spotlights of sun moved about the cabin, illuminating a mountain of scientific equipment, camping gear and a handful of eager researchers. Clutching his notebook with both hands, Mueller pressed his face against one such spotlight, looking out onto the ice below with the same intent and wonder as a young child on their first plane ride. For Mueller, it was both a fond glance at an old friend and an important aerial opportunity to observe how Milne Fiord had changed over the previous three years.

Even before Mueller disembarked from the plane, one of the stand-out questions had already been answered by the first half of the team that had arrived a few days before. The Milne Fiord epishelf lake was no more. The last of its kind in Canada, gone.

The focus was now on retrieving the mooring, which meant waiting an unknown number of days for a helicopter to show up — another side-effect of the weather delays in getting to the field. In the meantime, the team would focus on work that could be carried out by hiking to sites on the epishelf lake ice, glacier and glacial tongue. 

“When I first started at Milne Fiord, I was very interested in the ice shelf,” says Mueller. “But I soon realized that this is a fiord system that is not only the ice shelf but also the epishelf lake, the ocean below (which is warming), the atmosphere above (which is warming) and the entire watershed that supplies freshwater into the lake… we’ve been taking that on as a system.” 

In part, that’s what glaciologist Anna Crawford, who was also part of the Milne expedition, is hoping to understand better. Just as the ice shelf acted as a natural plug for the epishelf lake, which in turn served as a type of buffer for the glacial system, Crawford’s thought is that the sea ice acts as a kind of cast that provides a level of support for all of the internal stresses, strains and fractures in the shelf. Although this has implications for Milne and the other remaining Ellesmere ice shelves — which will experience more open water events in the future — the broader value of understanding the interaction between sea ice and ice shelves is much larger, and is something Crawford is working on by using models.

Earth system models simulate the physical, biological and chemical processes that are all part of Earth.  “These are the models that are used to project the climate out around the globe centuries into the future.” However, as Crawford also points out, there’s one big issue, “Right now, sea ice and ice shelves don’t talk to each other in those models.” Her challenge is finding a way to make that representation possible. 

When the unmistakable sound of the helicopter finally echoed through Purple Valley, the camp filled with energy. Within hours, the mooring had been found and retrieved. In even less time, its data was quickly downloaded to discover that it had been running throughout the calving event and subsequent draining of the epishelf lake. The camp was both elated and saddened by what they were able to see: the real-time draining of the epishelf lake — data that will both help them better understand the processes at play within the fiord system. It’s also data that could provide insight into how other epishelf lake systems in the Antarctic might respond to a changing climate. 

Over the next handful of days, the majority of work continued to be carried out on foot as heavy fog had grounded the helicopter until hours before it was set to depart, when the fog began to burn off. In an instant, the camp exploded into action trying to accomplish as much as possible at the farther reaches of the fiord.

With the helicopter gone for good and the work that didn’t require it effectively completed, the team made the call to depart the now stunning, purple-hued valley a few days early before another bout of fog or rain would prevent the Twin Otter from landing on the recently snow-free moraine airstrip. 

Just as the Milne Fiord team members were readjusting to the pace of daily life at home, White was gearing up to track the remnants of the ice shelf. These ice islands had drifted approximately 700 kilometres southwest of Milne Fiord, near Borden Island.

In 2021, a year after the breakup of the Milne Ice Shelf, a joint project between the Canadian Ice Service and the National Marine Network began planning to place tracking beacons on those remnants — just as had been done when other ice shelves collapsed. To do this, White, in conjunction with the Royal Canadian Air Force, flew over the islands and carefully threw parachute-laden beacons out of the back of the plane. “We all sit and watch as it hopefully lands on the island,” says White. “They have never missed.”

While White can broadly locate the islands from satellite data, the beacons, which have a two-year lifespan, provide extra resolution. “The nice thing about having the beacons is that you get that data on an hourly basis,” says White. “So we are able to create this big database that can be used for modelling and research [and] it sort of fills in the gaps between what I might see on a weekly basis.”

So far, the islands that have been tagged range in size from 28 square kilometres to one square kilometre, but White doesn’t know how many there might be. “There could be 50, there could be 100, there could be 500,” she says. “As we’re flying over, I’m seeing more and more, and the pieces that I’m discovering while out on the flights are the ones that are too small to see in the resolution of the imagery that we’re acquiring.”