A paper published this week by researchers at the University of Edinburgh, in the journal Communications Earth & Environment, concludes that the Arctic Ocean has undergone a regime shift driven by sea-ice loss — a shift the authors describe as one the system may not recover from.
The finding concerns the Arctic's nitrogen cycle. Nitrogen is the limiting nutrient for most ocean life — the substance that determines how much phytoplankton can grow, which determines how much krill and small fish can be supported, which determines everything that depends on them. The Arctic Ocean's nitrogen cycle has historically been regulated by the seasonal cover of sea ice. When the ice is present, the cycle behaves one way. When the ice retreats, it behaves another.
For most of the satellite record, the cycle held its historical pattern. Beginning around 2007, that pattern began to break. Since 2020 the regime has shifted decisively into a new state — one in which the cycle does not return to its prior behaviour even when seasonal ice partially recovers. The biological consequences are still being measured, but the trajectory is not contested. A different Arctic Ocean now exists. Whether it can support what the old Arctic Ocean supported is the question that remains.
The Edinburgh team frames the finding carefully. They are not predicting collapse. They are reporting that the system has crossed a threshold beyond which the previous equilibrium does not appear to be available. In ecological terms, this is what is meant by a tipping point — not catastrophe, but irreversibility.
The finding is the second major polar-ice paper this month. A study published in Science Advances in early May, led by researchers at the University of Southampton and UNSW Sydney, identified a comparable regime change in the Antarctic, where sea-ice levels have collapsed from their long stability and are now driving ocean-heat circulation patterns researchers describe as a fundamental reorganisation of the Southern Ocean system. Two papers, two oceans, two regime shifts, in the same month.
The Arctic and Antarctic do not behave the same way. Their ice forms differently, their ecosystems are differently structured, their climate roles are not symmetric. What they share is the property of having been considered stable for most of the period during which they have been measured, and of being measurably less stable now.
The practical consequences play out over decades. Arctic fisheries — already in transition as the cold-water species that anchor them move north or decline — are likely to look further altered by mid-century. Greenland's ice sheet, whose mass loss is governed in part by the temperature of surrounding waters, continues to thin. The polar bear and the seal and the species we know about are the visible part of an iceberg whose mass is largely planktonic and chemical.
The deeper finding in the Edinburgh paper, and the reason it matters beyond the Arctic, is what it says about the assumption of recoverability. The implicit framing of much climate policy has been that systems can be returned to prior states if the forcings that destabilised them are reduced. The Edinburgh team's data suggests this assumption is becoming less defensible for at least some major systems. What is lost may, in some cases, simply be lost.