It was 10 p.m. one night in June of 2008 and we were cut off from our on-glacier camp. When we'd left that morning, it was sunny and the glacier surface was crisp and crunchy, perfect for our plastic boots and crampons. My graduate student and I had decked ourselves out in hats and sunscreen to protect us from the sun's reflection off the ice, then headed up-glacier to check on a weather station and a stream gauge to make sure we were ready to capture all of the data we needed once spring melt began.

It was a long day, and by the time we headed back down-glacier, toward our camp, my bad knee had started acting up and I couldn't walk without pain. To make the situation worse, we realized that things had changed while we were away from camp. The warm day had melted the snow on the lower glacier, leaving a shifting landscape of ice slurries in hidden surface channels surrounded by small hummocks of solid ice. We got within 1,500 feet of our camp and couldn't go any farther, as there were two slushy streams between it and us that were just too wide and deep to cross. My graduate student tried to cross anyway, but the slush was like quicksand. I had to yell at him to lie down and pull himself out by the arms instead of trying to kick out with his legs. Now he was curled up in the tinfoil blanket from the emergency kit, icy wet to the waist and moaning about not wanting to die. Thank goodness for 24-hour daylight.

I couldn't reach him, and neither of us could reach our camp. It was time to get out the satellite phone and call base camp in Resolute Bay, 250 miles southwest.

Not all Arctic researchers have had the luxury of satellite phones or shortwave radios. Take the English explorers at the turn of the century: If they got in trouble on their travels, they had only themselves to rely on. Searchers were only sent out if the expedition didn't return after several years. This is what happened to Captain Sir John Franklin, who set off to traverse the last unexplored section of the Northwest Passage in 1845. The story ends badly — Franklin and his crew died of zinc deficiency and scurvy, some resorted to cannibalism, and they left artifacts strewn across the Arctic tundra as they tried to walk for help. Franklin's wife pushed for a search for the missing expedition as early as 1848, but it was too late for Franklin and his crew. More than 50 expeditions searched for 12 years before concluding that every one of the 129 men was gone. The searchers found only graves, relics, and Inuit stories about the expedition. It wasn't until 2014 that Parks Canada and the Royal Canadian Geographical Society finally found one of Franklin's abandoned boats, the H.M.S. Erebus, on the seabed where the Inuit had always said it was. The wreck of the H.M.S. Terror, Franklin's second boat, was found in 2016 by the Arctic Research Foundation. Both sites are currently being explored in collaboration with the local Inuit.

The Arctic has long tempted researchers to take risks to answer questions of science. Consider Norwegian Fridtjof Nansen, who suspected that, rather than being static, sea ice was a dynamic system that moved ice from east to west across the Arctic Ocean, from Siberia to Greenland, and potentially over the North Pole itself.

To test his theory, in 1893 he sailed a custom-built, reinforced boat — the Fram — into the pack ice on Russia's Laptev Sea, allowing it to freeze there and drift westward with the sea ice. Nansen took a huge risk with this experiment: The sea ice could have crushed his ship's hull, or it could have drifted in an alternate direction, which would have left the crew without the food required to return home.

On the same trip, Nansen decided to make a run for the North Pole. In March of 1895, he raced north from the Fram with his companion, Fredrik Hjalmar Johansen, traveling by dogsled and carrying kayaks for open water travel. However, they were still 300 miles from the North Pole when they realized they had to turn back because they didn't have enough food. Their plans called not for catching up with the Fram, but for reaching the archipelago of Svalbard, where they hoped to find a willing sea captain to take them home. They made it as far as Russia's Franz Josef Land, in June of 1896, where they stopped to fix their kayaks after a walrus attack and then unexpectedly (and fortuitously) connected with explorer Frederick Jackson, who was leading an expedition there. They returned to Norway with Jackson, while the Fram itself was sighted west of Svalbard, exactly where Nansen had expected it to be as it drifted on the ocean currents. Unlike the Franklin expedition, everyone returned home alive — after being on the ocean and in each other's sole company for three years.

More than a century later, our 2008 Arctic field program was part of a worldwide scientific endeavor called the Fourth International Polar Year (IPY), which focused on studies of the polar regions. Previous Polar Years had occurred from 1882 to 1883, 1932 to 1933, and 1957 to 1958 (this one was also called the International Geophysical Year, or IGY); all had resulted in new and interesting findings about Arctic processes, like sea ice and glaciers. For IPY No. 4, we were part of a global team, each subset of which was studying a specific glacier. Ours was the Belcher Glacier, which originated on the Devon Island Ice Cap in Arctic Canada.

In summer, our team on the lower part of the glacier and another team on the upper reaches of the glacier measured snow and surface water to determine how water on the surface was generated; we also observed how that surface water entered the glacier. The team at the terminus examined how the water affected the glacier's forward motion. The main goal was to use our field data to develop and run a model of the glacier itself to see how it would respond to climate change, and to compare our results with those of the other international research teams. This was important for examining how fast polar ice sheets are declining and what that means for the future.

I had a bad feeling about the project from the beginning. Glacier travel can be done in pairs, but if someone falls into a crevasse — a deep crack that forms in the ice when the glacier flows over subglacial obstacles — you have to work pretty hard when you're the only person there to save them. I wasn't confident in my student's ability to save me from a crevasse, and I hadn't taken a refresher course in crevasse rescue myself to brush up on my own skills. A recurring knee injury that had been my bane during my three Arctic field seasons had also been acting up again just before this trip began. I was leery of camping on the ice in summer, as I was concerned about the amount of water generated on the glacier surface during spring and summer melt.

Turns out I was right about the problems with water, given our stranded situation. The upper glacier team experienced the same thing later in the season, but the slush formed while they were still in camp, so they just stayed put until it disappeared and crossable surface channels appeared. For my field partner and me, it was already too late to avoid the sudden ice streams — once he had been submerged, the only option was to evacuate as quickly as possible.

Read the rest of this story at Narratively.

Narratively is a digital publication and creative studio focused on ordinary people with extraordinary stories.