Skip to content
For WHOI personnel: COVID-19 Guidelines and FAQs

A sea of ancient ice

Scientists dust off historical accounts to tackle a long-standing Arctic mystery

By Evan Lubofsky | March 4, 2020

WHOI scientist Alan Condron and his colleagues rely on historical drawings like this one published in 1876 to gain a better sense for how thick Arctic sea ice was in the early 19th century. 

“Any man who has seen it [palaeocrystic ice] knows it is ancient, the same as you know an old man or an old woman is old.”

Major A. W. Greely, Royal Geographical Society, 1888

When Harper’s Weekly magazine reported the spotting of a seven-mile-long chunk of thick sea ice off St. Johns, Newfoundland, Canada in 1884, the story referred to the prairie-sized floe as a “monster ice island” and forewarned ship captains travelling in the area: “Woe to the mail steamer that shall crash against its sides or upon its hidden base.”

This was at the tail-end of the Little Ice Age, when vast areas of the Arctic Ocean were covered by seemingly-impenetrable slabs of ice, and icebergs would stray as far south as Bermuda.

Sea ice was irrefutably thicker during the 19th and early 20th centuries than it is today—warming in the Arctic has caused much of its ancient ice to vanish—but according to WHOI climate scientist Alan Condron, the actual thickness of the legacy ice has been a long-standing mystery in climate science circles.

“While we have been able to determine the amount of sea ice extent since 1979 with satellite data, we've only had continuous satellite observations of ice thickness since the early part of this century,” says Condron.  “Before that, we only have a few sporadic observations from U.S. Navy submarines taken during the Cold War in the late 1950’s. Prior to that, there is nothing.”

To bridge the gap, Condron used computer model simulations to estimate ice thickness during the Last Ice Age—21,000 years ago when ice sheets blanketed much of North America and Europe.  The models reported an average sea ice thickness of 30 meters, but along the coast of Northern Canada, the simulated thickness grew to 50 meters—roughly the height of the Leaning Tower of Pisa.

These estimates gave Condron a baseline sense for how chunky Arctic ice may have been, but there was a problem. The values generated by Condron’s models far exceeded those reported by other models simulating ice conditions during the same period.

“Other climate models were reporting average Arctic sea ice thicknesses of only seven to eight meters, and the thickest ice we could find in all of these models was just 16 meters thick," says Condron. "It was rather baffling that these models were growing ice that was only slightly thicker than the ice we commonly see today, particularly since we know conditions in the Arctic were much colder during the Last Ice Age than they are now.”

The discrepancy between Condron’s model and other sea ice models became an issue for Condron as he prepared his manuscript on an abrupt climate change study for publication. “At an early stage, one reviewer felt that the ice thicknesses we were reporting suggested there was something seriously wrong with our model,” he says.

He explains that in the sea ice modeling community, modelers often impose a limit on how thick they let ice "grow" in their simulations in order to ‘correct’ for errors in the model. For example, ice thickness can increase significantly in models if they don’t simulate ocean circulation accurately, or have trouble simulating clouds and the amount of solar radiation reaching the surface. “In these cases, you would cap the thickness in your model to the maximum thickness of modern-day sea ice,” says Condron.

With his research paper in a holding pattern, he began thinking about other ways to ground-truth his model results, which led to an epiphany.

“Nineteenth century Arctic explorers often described sea ice conditions in their dairies with descriptions and sketches,” he says. “So my thought was to inspect some of those historical accounts to see if they seemed consistent with our estimates.”

WHOI climate scientist Alan Condron examines an iceberg drifting south in the Labrador Sea. (Photo by Andrew Daly, © Woods Hole Oceanographic Institution)

As Condron, along with his co-authors Anthony Joyce and Raymond Bradley, began leafing through the dairies, they quickly noticed descriptive passages supporting their side of the story. One account, penned by Vice-Admiral Sir George Nares, the leader of the 1875 British Arctic Expedition, described “floes… of gigantic thickness with a most uneven surface and covered with deep snow.”

A corresponding drawing—that was published in 1876 in the British weekly newspaper “The Graphic”— shows a glimpse of the polar sea as traversed by two members of Nares’ party during their spring sledding expedition to reach the North Pole. Captain Nares was so struck by the unusual thickness of the sea ice his team encountered in the western Arctic, he coined the term “palaeocrystic” ice to describe it.

Bradley, a climate scientist and professor at the University of Massachusetts, notes that other 19th and early 20th century explorers had documented exceptionally thick, extensive islands of ice embedded in the Arctic pack.

“The ice had rounded, hummocky surfaces that rose as much as ten meters above sea level and stood out from the smaller floes of fractured sea-ice which the explorers generally had to deal with,” says Bradley.

Condron notes that based on some of the drawings alone, it became clear that sea ice was not only much thicker than it is today, but also very similar to what he was seeing in his model. He re-submitted his manuscript with an explanation of how it has become common practice in model simulations to artificially ‘cap’ sea ice thickness to avoid values that seem unrealistic.

A few weeks later, the paper was accepted.

The diary records helped Condron get his paper through, but he feels they served a more important purpose by putting modern-day sea ice loss in a broader historical perspective.

“While Arctic sea ice conditions have changed significantly in recent decades, the changes are even more dramatic when viewed in the context of the conditions that Nares and others encountered when they went looking for the North Pole,” says Condron. “Basically, we’ve gone from a situation where ice was 50 meters thick along parts of the Canadian Coast, with large pieces of this ice adrift in the Beaufort Sea, to the present-day situation, where we rarely see ice that exceeds five meters. And it’s all happened in a relatively short period of time.”

Funding for this research was provided through the National Science Foundation’s Arctic System Science Program.