Meltwater on a city‑sized iceberg in the Southern Ocean is rapidly forming a giant pool on its surface – possibly a sign that it is close to breaking apart.

��è����Ƶs are captivated by the frozen colossus, known as A23a, because meltwater is collecting and being held on its surface in an unusual way.

Satellite images reveal a raised rim of ice running around the entire cliff edge of the tabular Antarctic iceberg, giving it the appearance of an oversized children’s play pool — except this one spans about 800 square kilometres, an area larger than Chicago.

In places, the ponded water appears a deep, vivid blue, suggesting depths of several metres. Across the whole of A23a, the water volume probably runs into billions of litres – enough to fill thousands of Olympic‑sized swimming pools.

at the University of Chicago says the rim effect is typical of the world’s largest icebergs.

“My theory is that the edges are bent, nose‑down, creating an arch‑like dam on the top surface that keeps the meltwater inside,” he says. “The bending is probably a combination of cliff-face undercutting by waves and melting, and the natural tendency for ice cliffs to bend over even if they would be perfectly vertical otherwise.”

Free newsletter

Sign up to The Earth Edition

Sign up to newsletter

The streaks of surface water visible in the satellite imagery are a relic of the way the ice once flowed when the iceberg was still attached to Antarctica’s coastline, he says.

A23a is an old iceberg. It calved from the Filchner–Ronne ice shelf in 1986 and was then more than five times its current size. For a while, it held the title of the world’s largest iceberg.

In recent years, however, it has drifted north into warmer waters and air, and is now undergoing relentless fragmentation. The sheer volume of meltwater pooling on its surface may finally break it apart. “If that water drains into cracks and refreezes, it will prise the berg open,” says at the British Antarctic Survey.

It could, he says, turn to mush almost overnight.

City-sized icebergs once drifted past the coast of Britain when ice sheets covering much of northern Europe were in rapid retreat about 18,000 to 20,000 years ago.

at the British Antarctic Survey and his colleagues have found the preserved scour marks these giants made as their undersides ploughed through seafloor sediments. The long, comb-like features are buried under mud in the North Sea but are still visible in the seismic survey data collected to search for oil and gas.

“We can estimate from the extent of the scours and what is known about ancient sea levels that these bergs were probably five to a few tens of kilometres wide and perhaps a couple of hundred metres thick – icebergs on the scale of a mid-sized British city,” says Kirkham.

In Antarctica, tabular or table-top icebergs are a spectacular sight. Some, like the recent behemoths known as A23a and A68a, would rival even small US states in terms of area. They calve from ice shelves – the wide, floating protrusions of glaciers that flow off the land into the ocean.

The recognition that tabular icebergs once existed in the North Sea is therefore a clear indication that the seaward margins of a British and Irish ice sheet also had ice shelves. And it means there could be some lessons for future Antarctic decline, says Kirkham.

In the North Sea, the straight tramlines of the big icebergs are over-written by squiggly troughs made by the narrow keels of much smaller ice blocks. In other words, there’s a “regime change” in which large icebergs are replaced by countless small icebergs as the ice shelves shatter in response to rising temperatures, says Kirkham.

Free newsletter

Sign up to The Earth Edition

Sign up to newsletter

Radiocarbon dating of the sediments shows this shift occurring over a period from 20,000 to 18,000 years ago.

The observation casts doubt on the idea that the calving of mega-bergs like A23a and A68a might herald the widespread collapse of Antarctica’s ice shelves.

at the University of Florida has tracked tabular iceberg size in satellite data from the mid-1970s onwards and found that the trend is essentially flat.

“James’s research underscores mine, which is that large calving events are not necessarily a sign of instability or cause for alarm,” says MacKie. “Rather, ice shelves disintegrate via death by a thousand cuts. We should be concerned when we stop seeing the large calving events.”