What happened
On 30 October 2026, researchers at the IBS Center for Climate Physics at Pusan National University in South Korea published a study in Nature Climate Change on how the Greenland and Antarctic ice sheets respond to warming over very long periods. Yun Kyung-Sook and Axel Timmermann ran 100,000-year simulations at two, four, six, eight and twelve times pre-industrial carbon dioxide levels, coupling a full Earth system model to an ice-sheet model so that the climate and the ice could change each other.
Running a full climate model for 100,000 years is not possible, so the team alternated: the ice-sheet model ran for 1,000 years using a climate taken from a 10-year climate-model run, then passed its new ice shape back to the climate model, and so on, the paper explains.
The results differed sharply by region. Warming above about 2°C led to the long-term break-up of the West Antarctic ice sheet, and Greenland’s ice disappeared once warming exceeded about 6°C. But the East Antarctic ice sheet stayed comparatively stable even at warming above 10°C, the institute said.
Why it matters
Antarctica holds ice equivalent to about 58 metres of global sea-level rise, and most of that is in the east, the paper notes. The Antarctic and Greenland ice sheets are both losing mass today: observations for 2002 to 2022 show losses of about 255 billion tonnes a year from Antarctica and 127 billion tonnes from Greenland, together adding about 1.1 millimetres a year to sea level. Earlier studies using ice-only models had suggested that East Antarctica could melt completely at between about 6°C and 10°C of warming.
The new simulations found a brake on that loss. Warmer air carries more moisture, and a southward shift in the westerly winds pushes more of it towards Antarctica, where it falls as snow on the high interior. More cloud also cuts the sunlight reaching the ice. When the researchers left these changes out, the Antarctic ice sheet lost about 70 per cent of its volume under the strongest warming. With them included, the loss was about 20 per cent, according to the institute’s summary.
‘Simplified climate forcing may overlook regional processes that strongly influence whether an ice sheet retreats, stabilizes or survives,’ said Axel Timmermann, the centre’s director.
Our results demonstrate that the long-term future of an ice sheet cannot be understood from temperature alone.
Yun Kyung-Sook, lead author, IBS Center for Climate Physics
The wider picture
This is not a reason to relax. The same simulations show warming of 2°C or more driving major ice loss from West Antarctica, worth about 2.5 metres of sea-level rise over the long term, and about 1 metre from Greenland, the paper reports. The researchers stress that parts of East Antarctica are also retreating, and Asia Economy reported that the team said the risk of sea-level rise should not be underestimated.
Other recent work points the other way for West Antarctica in the near term. A model from the University of California, Irvine and NASA’s Jet Propulsion Laboratory projects that Thwaites Glacier alone could add up to 30 millimetres to sea level by 2100, twice previous estimates, even without further warming, Phys.org reported. In GGND’s view, the two studies fit together: the slow-moving giant in the east may be sturdier than feared, while the faster, ocean-exposed ice in the west is not.
Could it work elsewhere?
The method could. The paper argues that coupled climate and ice-sheet models, in which the atmosphere responds as ice sheets change, should become standard for long-term sea-level projections. The authors say similar atmospheric effects may matter for other ice sheets, although Greenland behaved very differently in their runs.
What remains unresolved
- Coarse resolution. The authors say the model cannot fully represent melting beneath ice shelves or the behaviour of individual outlet glaciers, where ocean-driven loss is fastest.
- Timescales. The simulations cover thousands of years at fixed carbon dioxide levels. They are not forecasts for this century and are not observations.
- Reported figures. The institute’s summary gives the coupled Antarctic loss as about 20 per cent; readers should treat it as an approximate model result.
- Other models. One model is one view. Other research groups will need to test whether the snowfall and cloud brake appears in their simulations too.
What happens next
The authors call for higher-resolution simulations to test how the atmospheric effects interact with ocean-driven melting at the coast. Meanwhile, fieldwork continues: the British Antarctic Survey’s 2026–27 season began in late October. GGND will follow how other modelling groups respond to the findings.
Sources
- Nature Climate Change Atmospheric stabilization of the East Antarctic ice sheet under warming, Yun and Timmermann (30 October 2026)
- Phys.org / Institute for Basic Science Snowfall and clouds may slow East Antarctic ice loss in a warmer world (30 October 2026)
- Asia Economy East Antarctic Ice Sheet Could Survive 10-Degree Global Warming (30 October 2026)
- Phys.org / University of California, Irvine Thwaites Glacier could contribute twice as much sea-level rise as previously estimated (13 October 2026)
Analysis and context are by The GGND Team. Where we draw our own conclusions, we say so. To report an error, contact us.
