How Antarctic Sea Ice Grew for Decades Before Suddenly Collapsing

The planet was warming. The Southern Ocean was warming. And yet, from the late 1970s right through to 2015, the sea ice ringing Antarctica kept spreading. More of it each decade, not less. Climate scientists watched this happen and, broadly speaking, couldn’t quite explain it, not convincingly anyway, because their models kept predicting the opposite. Then in 2016 something flipped. Sea ice extent dropped to a record low and has never recovered, sitting stubbornly well below where it ought to be even now.

That reversal was not, it turns out, a random climate hiccup. A new study from Stanford University has traced the whole arc, the decades of unexpected growth and then the sudden collapse, to a single underlying system: a kind of oceanic thermostat, wound tight by rainfall and then abruptly released by wind.

The Southern Ocean is peculiar in ways that matter enormously for the rest of the planet. It’s where the deep global overturning circulation draws breath, pulling heat and carbon dioxide down into the abyss and distributing them around the world. Sea ice governs much of how that exchange happens. What goes on down there, in other words, doesn’t stay down there. Earle Wilson, an assistant professor of Earth system science at Stanford’s Doerr School of Sustainability and lead author of the new study, published March 23 in the Proceedings of the National Academy of Sciences, has spent years trying to pin down what the Southern Ocean has been doing. The answer hinged on two forces pulling in opposite directions at once.

The first force was freshwater. As the climate warmed, precipitation over the Southern Ocean increased, rain and snow falling on water that was already frigid and exposed to the atmosphere. Fresh water is less dense than salt water, so this influx sat on the surface, forming a lid. Beneath that lid, warmer, saltier water, the kind that had been circulating up from depth, was effectively locked away. The upper ocean stratified. Two layers formed, nearly insulated from each other, and the deep warmth couldn’t reach the surface to melt ice. “For a while, precipitation was winning until upwelling took over,” Wilson says.

The second force was upwelling driven by wind. Stronger and stormier conditions around Antarctica, likely connected to broader climate trends, meant that the curl of wind across the water was increasing. That curl drives Ekman upwelling, a process in which surface waters are pushed sideways and deeper water rises to replace them. Between 2014 and 2016, upwelling rates in the Weddell Sea nearly tripled. The freshwater lid, which had been holding back decades of accumulated ocean heat, began to lose the battle.

The data that allowed Wilson’s team to see all this came from a somewhat overlooked corner of the global ocean monitoring network. Since the early 2000s, thousands of autonomous floats collectively known as the Argo array have been drifting through the world’s oceans, diving to depth, measuring temperature and salinity, then resurfacing to transmit their readings by satellite. Most of these floats operate in open water. Some, though, travel beneath the Antarctic sea ice, logging data through the winter before coming up for air in summer. Lexi Arlen, a PhD student in Wilson’s Polar Ocean Dynamics Group, helped compile and analyse roughly two decades of this under-ice data, a dataset rarely touched for sea ice research. “It was very exciting to be able to use a combination of data and idealized modeling to explain both the observed expansion and retreat phases of sea ice,” Arlen says.

What the floats revealed was, in retrospect, striking. The thermocline beneath the Weddell Sea, the boundary where warmer circumpolar deep water sits below the colder surface layer, had been steadily shoaling and warming since around 2008. By 2015, that subsurface warm water was sitting closer to the surface than at any point in the observational record and was warmer than it would be eight years later, when sea ice reached its record minimum in 2023. The warmth was there, piled up and waiting. The freshwater lid was the only thing keeping it from getting out. When that lid weakened, when surface salinity ticked upward as upwelling brought salt to the surface, the heat vented. Sea ice declined and has not meaningfully recovered since.

Wilson’s team tested this narrative with a simplified computer model of the Weddell Sea ice-ocean system, running it with observed variations in precipitation and upwelling. The simulations showed sea ice expanding until around 2010, then declining, which broadly matches the satellite record. The steepness of the 2016 collapse was probably amplified by atmospheric circulation patterns that the model doesn’t capture, but the underlying mechanism plausibly accounts for most of the multiyear trend.

Not everywhere, though. The Pacific sector of the Southern Ocean, wrapping around from the Antarctic Peninsula to the Ross Sea, behaved differently. There, the ocean interior actually cooled after sea ice declined rather than warming, the opposite of what the model would predict. Wilson describes this as an unanswered part of the puzzle; other processes, perhaps changes in how sea ice drifts, perhaps more turbulent storm-driven mixing, appear to be more influential in that region.

Wilson notes that the ocean’s long memory sets it apart from the atmosphere as a driver of sea ice change. Weather can swing ice extent around from week to week. But the ocean can build up heat over years, hold it, and then release it in a sustained pulse that keeps ice suppressed for a decade or more. That is roughly what appears to have happened. The question now is whether the current low-ice state is a temporary shifted equilibrium or the opening chapter of something more lasting.

The practical stakes are considerable. Antarctic sea ice controls how much solar energy the Southern Ocean absorbs. It also influences how quickly the West Antarctic Ice Sheet, which sits on bedrock below sea level in places, loses mass to the ocean. Sea level projections for the coming century are sensitive to both processes, and those projections are only as good as scientists’ understanding of what governs sea ice from one decade to the next. Wilson plans to keep monitoring the Argo float data and develop a fuller predictive theory, one that might finally let modelers anticipate the Southern Ocean’s next move rather than explain it after the fact.

Whether the ice recovers, and when, depends on how precipitation and upwelling evolve in a world that is, on balance, still warming. At the moment, upwelling seems to be winning.

DOI / Source: https://doi.org/10.1073/pnas.2530832123