On the edge of a shrinking ice cap in northern Canada, dead moss is emerging from the ice. Each sample carries a precise date of death. Put together, those dates tell a story that a team led by Columbia University's Nicolás Young published on October 22 in Science: roughly 2,000 years ago, the last remnant of North America's great ice sheet started growing again, and by about 1960 modern warming had stopped it.

A plant with a death date

The Barnes Ice Cap is a Delaware-sized dome of ice on central Baffin Island in Nunavut, and the last vestige of the Laurentide Ice Sheet, which covered much of Canada and the northern United States until about 20,000 years ago, according to Columbia's Lamont-Doherty Earth Observatory. Unlike many glaciers, it is frozen to the bedrock and does not slide, so its margin records climate rather than flow, Smithsonian Magazine noted.

When the ice grows, it kills what lies in its path. Radiocarbon dating gives each plant a "kill date," the moment permanent ice last advanced across that spot, as Young explained in the University of Colorado Boulder's account. As the ice retreats today, it exposes mosses, leaves and twigs that have been entombed for centuries, and the team dated them one by one.

The timeline

The results point to a three-step history. The ice cap's edges had pulled back to roughly their present position by about 3,000 to 3,600 years ago, depending on the account. They then expanded, starting around 2,000 years ago, when the climate became cold enough for ice to accumulate, according to the Lamont release. Plant material documents the expansion up to roughly a thousand years ago, while 186 previously published radiocarbon dates from neighboring ice caps indicate that regrowth began about a thousand years earlier, per New Atlas's summary of the Columbia work.

The growth continued until the late 1950s. Early studies of the Barnes Ice Cap in the 1950s and 1960s described it as exceptionally stable, Miller of the University of Colorado noted. Then it flipped. By the 1960s, summer losses exceeded winter gains, and retreat from the old margin has reversed about 2,000 years of expansion, the Columbia release says.

Was it really the next ice age?

That label is where a careful reader should slow down. The finding itself, the ice cap's regrowth and reversal, rests on dated plants. The claim that it marked the beginning of Earth's next glaciation is the authors' interpretation. The idea of a "failed ice age" has circulated among paleoclimate researchers for years, according to the University of Colorado Boulder, and Young says earlier work could not quite prove it, while the new data supply physical evidence.

The scale tells you how much weight to place on it. Young said it would have taken thousands of years before the cooling became noticeable, and about 30,000 years for the Laurentide to regain its continent-sprawling mass. Yahoo's summary of the paper said the authors challenge definitions that require much larger ice sheets before declaring the onset of a glaciation. So "an ice age was starting" means a trend that, left alone, might have taken tens of thousands of years to build, not a looming event that was cancelled.

Why the ice was growing

The underlying driver is orbital. Glacial cycles are governed by Milankovitch cycles, periodic changes in Earth's orbit, tilt and axis direction that alter how much sunlight reaches certain latitudes, Gizmodo explained, citing NASA. For more than a thousand years before the 1960s, those cycles were pushing the planet toward a new glaciation, according to the University of Colorado Boulder.

We are, technically, still in the Late Cenozoic Ice Age, which began about 34 million years ago, and the last glacial period ended around 11,700 years ago. The study concerns the next cold phase within that long era, not the start of a new geological one.

What comes next

Climate models suggest the Barnes Ice Cap is likely to disappear completely within the next 300 years, according to Gizmodo's summary of the literature. That matters for a reason beyond the ice itself. The Laurentide has anchored the growth of continent-spanning ice for eons, and the loss of its last remnant could change the dynamics of future glaciations.

In a perspective piece accompanying the study, University of Wisconsin-Madison glaciologist Shaun Marcott wrote that future emissions could forestall the next ice age for hundreds of thousands of years. That is a projection, not an observation, and it depends on the emissions path. UNEP's latest assessment puts current policies on course for about 2.8°C of warming by 2100, which would keep the system far from the conditions in which ice caps grow.

What this study does not show

It is one ice cap on one island, supported by a set of radiocarbon dates and a published regional record. It does not show that every glacier was growing 2,000 years ago, and it does not predict when the next glaciation would otherwise have come. The attribution of the reversal to greenhouse gases is the authors' reading of a 1960s turning point, set against an earlier period when natural cooling was the driver.

Still, the timing is the notable part. Ice that had been advancing for two millennia turned back within a human lifetime. Geological processes that normally play out over tens of thousands of years were reversed in decades.

The practical lesson is about the clock, not the ice age. The climate system's slowest feedbacks are measured in millennia, and they are now responding on the timescale of a mortgage. What the Barnes moss records is not a cancelled ice age so much as how quickly a natural trend can be overridden.

---