Scientists have developed a forecasting method that predicts Arctic sea ice extent up to four months ahead, with a focus on the annual minimum in September. This approach outperforms existing models by integrating long-term climate patterns, seasonal cycles, and short-term weather influences. The tool aims to aid communities and industries reliant on Arctic conditions.
Arctic sea ice, which reflects sunlight to cool the planet and affects global weather patterns, is vanishing rapidly due to climate change. Researchers from the United States and the United Kingdom have introduced a new prediction system detailed in the journal Chaos, published by AIP Publishing. The model targets September, when sea ice reaches its lowest point, using data from the National Snow and Ice Data Center dating back to 1978.
The system treats sea ice changes as an interconnected process influenced by varying timescales: long-term climate memory, annual cycles, and rapid weather shifts. Tests using real-time data from September 2024 and historical records showed it provides more accurate forecasts one to four months ahead compared to other methods. By incorporating regional details across the pan-Arctic, the model handles year-to-year variations effectively.
"Indigenous Arctic communities depend on the hunting of species like polar bears, seals, and walruses, for which sea ice provides essential habitat," said author Dimitri Kondrashov. "There are other economic activities, such as gas and oil drilling, fishing, and tourism, where advance knowledge of accurate ice conditions reduces risks and costs."
Kondrashov added, "The model includes several large Arctic regions composing [the] pan-Arctic. Despite large differences in sea ice conditions from year to year in different regions, the model can pick it up reasonably accurately."
While long-term climate projections remain reliable, short-term forecasts have improved through this integration. The team plans to enhance the model by adding factors like air temperature and sea level pressure to better capture summer variability. The research, led by Dmitri Kondrashov, Ivan Sudakow, Valerie Livina, and Qingping Yang, appears in Chaos (2026; 36(2)), with DOI: 10.1063/5.0295634.