Arctic winter sea ice could shrink by an extra million square kilometres, ancient clues suggest
Researchers compared 27 climate models with evidence from a warmer Arctic 6,000 years ago. The models overestimated how much winter ice survives.
By Newsheir Staff, Newsroom
· 1 min read
In this article
Climate models may be underestimating how much winter sea ice the Arctic will lose, according to a study published this month in Nature Geoscience by researchers at the University of Cambridge and CIMAT in Mexico.
Reading the past to test the future
The team built the largest collection yet of a chemical marker produced by algae that live in sea ice, taken from ocean sediment cores. The cores record the mid-Holocene, about 6,000 years ago, when natural shifts in Earth's orbit made the Arctic warmer than in recent centuries.
The researchers ran 27 climate models for that same period and compared the results with the sediment evidence. The models consistently showed more winter sea ice than was actually there.
What changes
Weighting the models by how well they matched the ancient evidence removed that bias and narrowed the uncertainty in future projections by up to 85%. The adjusted projections suggest Arctic winter sea ice will shrink by about 1 million square kilometres more than earlier estimates.
Most past research has focused on the summer minimum, when the ice is at its smallest. Winter is when the models disagree most. The revised weighting also suggests that changes in snowfall over southeast Greenland and rainfall around the Mediterranean may be underestimated, which could matter for weather in Europe.
The caveat
The method assumes that how well a model matches a past warm climate tells you how well it predicts the future. The authors use the result to narrow the range of projections, not as a single forecast.
Sources and references
AI disclosure This article was prepared with the help of AI tools and reviewed by our editors. How we use AI.
Spotted an error? Request a correction. Read our editorial standards.
Format: Research