For most of the satellite record, Antarctic sea ice behaved differently from its northern counterpart. While Arctic sea ice declined steadily from the late 1970s onward — eventually losing roughly half its summer extent — Antarctic sea ice remained stable, and between 2007 and 2014 actually expanded slightly. The asymmetry was a puzzle. It was sometimes cited, by those sceptical of the broader picture of polar change, as evidence that the climate system was less coupled and less responsive than the standard accounts suggested.
The asymmetry no longer exists. Since 2015 Antarctic sea ice has declined sharply. In 2023 the winter sea ice extent reached the lowest level in the satellite record — so far below the long-term average that statistical analysis put the probability of the event occurring by chance at roughly one in 3.5 million. The summers of 2022, 2023, and 2024 each produced near-record-low minima. The 2025–26 austral summer minimum was the sixteenth lowest in 47 years of record-keeping; the broader picture is one of a system that no longer returns to its prior baseline.
A paper published in Science Advances in May 2026, led by Aditya Narayanan of the University of Southampton, identified what the authors call a “triple whammy” of interacting drivers behind the collapse. Warm subsurface waters that had previously been kept below the surface mixed layer are now reaching the surface and the underside of the ice. Atmospheric circulation patterns over the Southern Ocean have shifted in ways that reduce ice formation. And the ice that does form is thinner, breaks up sooner, and exposes more dark ocean to incoming solar radiation — which absorbs more heat, which prevents ice formation in the following season. The feedback runs in one direction.
A second paper, published in Nature Communications in early May by a research team based in Tromsø, Norway, identified a related but distinct mechanism affecting the underside of the major East Antarctic ice shelves. The shape of the seafloor under those shelves channels warm seawater along specific paths, accelerating melt in ways that previous models did not capture. The Fimbulisen Ice Shelf, the subject of the Norwegian work, is one of the cold-water ice shelves that had been considered most stable. The new findings suggest that “stable” was a description that depended on conditions that no longer obtain.
The Antarctic system supports an ecosystem whose centre is krill. Krill feed on the algae that grow on the underside of sea ice. Whales, seals, penguins, and seabirds depend on krill. When the ice goes, the algae go, the krill go, and the trophic cascade runs upward through every species that depends on it. The Commission for the Conservation of Antarctic Marine Living Resources, which regulates Southern Ocean fishing, has been negotiating reduced krill quotas for several seasons; the harvest pressure on krill from industrial fleets — primarily Chinese and Norwegian — has expanded in the same period.
What the May papers do not say is that any of this is reversible. The systems described have crossed thresholds beyond which the prior equilibria do not appear to be available. This is the same finding the Edinburgh team produced for the Arctic in the paper published this week. Two oceans, two regime shifts, two months. The pole that did not appear to be changing has caught up with the pole that was.