Near the head of Allison Glacier near Kullorsuaq, western Greenland.
Near the head of Allison Glacier near Kullorsuaq, western Greenland. Margie Turrin/Lamont-Doherty Earth Observatory

It took five years to drill down to the bedrock beneath the sheet of ice covering Greenland. Down and down and down the exquisitely engineered drill went, through almost two miles of ice, until it came close to the bedrock, where the ice grew rougher with pebbles and rocks.

This was back in 1993, during the most ambitious ice core drilling project that had ever been tried. There were two parallel projects, with sites just a few miles apart, one run by an American team and the other by an European team, collecting cores that could be compared to one another. Only, when they closed in on the bedrock and the rocky ice started to destroy the endlessly designed drill crowns, the Europeans pulled back.

“And the Americans decided to slam that thing into the bedrock,” says Joerg Schaefer, a paleoclimatologist at Columbia University’s Lamont-Doherty Earth Observatory. That’s how they ended up with a five foot sample of Greenland bedrock, which Columbia calls “perhaps earth’s rarest geologic sample: the only bit of bedrock yet retrieved from the ice sheet’s base.”

The bedrock samples.
The bedrock samples. Schaefer et al Nature 2016

All those years ago, the American scientists already had an idea of why they might want the bedrock sample. When cosmic rays—radiation from outer space—hit the Earth, they produce neutron showers that fall to the ground and make very specific radionuclides in the first few feet of the ground. Ice, though, stops that from happening. If those particles made by cosmic rays could be measured in the Greenland bedrock, they would reveal how long it had been since that bit of land had been clear of ice—when the massive Greenland ice sheet had last melted away.

When the bedrock was first retrieved, though, the techniques for measuring those particles were not refined enough to reveal the ice sheet’s history. A pilot analysis in the ‘90s showed that the particles scientists were looking for, isotopes of beryllium and aluminium, were present in the bedrock. At some point, this piece of land had been exposed to the sky.  “But they couldn’t tell for how long and when it was ice free,” says Schaefer.

When he first came to the United States from Europe, Schaefer started working—“out of geochemical geek-ness”—with Robert Finkel, a pioneer of beryllium analysis and one of the scientists who did the pilot analysis, on refining methods for measuring these isotopes. Finally, after many years of work, they were convinced they could measure the very faint signals they might find in the world’s only bedrock sample from under the Greenland Ice Sheet.

The Greenland Ice Sheet.
The Greenland Ice Sheet. Hannah Grobe

Before they could try, though, they had to convince the sample assignment committee who oversees the ice core to let them have part of it. That took a year. They had to prove that their techniques were good enough that they’d be able to measure the small amounts of isotopes they’d be looking for. “I appreciate it in hindsight,” says Schaefer. “It was a pain in the neck, but you cannot just give these samples out and then they are gone.” Because, in this case, using the samples meant destroying them.

The beryllium and aluminum isotopes that Schaefer and his colleagues were looking for are contained in quartz. To measure the isotopes, first the scientists crush the rock and take the quartz out. After they decontaminate the quartz, to make sure any signal they measure is the signal they’re looking for, they digest it in acid, in order to isolate, using these highly refined chemical methods, the isotopes they are interested in. To actually measure the isotopes, they use mass spectrometers married to particle accelerators, machines so expensive, rare, and specialized that the team used one machine, at Lawrence Livermore National Laboratory, to measure the beryllium isotopes and another, at Purdue University, to measure the aluminum.

Near the head of Allison Glacier near Kullorsuaq, western Greenland.
Near the head of Allison Glacier near Kullorsuaq, western Greenland. Margie Turrin/Lamont-Doherty Earth Observatory

What they found, when all this was done, more than 20 years after the bedrock was first collected, was that for large parts of the Pleistocene, which stretched from about 11,700 years ago back and back and back to 2.6 million years ago, much of Greenland had no ice sheet covering it. (Their results are published in Nature.) Compared to some models for the ebb and flow of the ice sheet, their data shows that the ice sheet is much less stable than some people thought—and less stable than we all should hope for.

If the Greenland ice sheet melts away, it means the oceans of the world will rise more than 7 meters. That’s more than 22 feet of sea level rise, when even half that would change the shape of the places we live, putting large sections of coastal cities underwater. If the ice sheet is less stable than scientists thought, it’s more possible it will disappear as the climate changes. “You have a hard time sleeping when you think about what it is,” says Schaefer. “The Greenland ice sheet has been much less stable than we wish. It’s not what we like to see. But it was what it was.”