Role of Atlantic multidecadal variability in modulating Arctic sea ice loss and wetting
Journal article, 2026
Arctic precipitation has increased in recent decades (hereafter, Arctic wetting), but the drivers remain uncertain. Using observations, reanalyses, and single-model initial-condition large ensembles (SMILEs), we show that enhanced evaporation due to sea ice loss has been the primary driver of Arctic wetting during 1979-2024, especially in the Atlantic sector. However, the externally forced component in most SMILEs explains only similar to 69% of sea ice loss and 75% of wetting in the observations and reanalyses. Further analysis reveals that the observed transition of one of the Northern Hemisphere's interdecadal internal variability-Atlantic multidecadal variability (AMV)-from a negative to a positive phase substantially enhanced Arctic sea ice loss, thereby accelerating wetting by about 31%. Under SSP3-7.0, if the AMV switches phase in the near future from the current +1 to a -1 standard deviation anomaly, then the rates of Arctic sea ice loss and wetting would slow by nearly 29 and 33%, respectively, relative to the externally forced response alone. These results underscore the pivotal role of AMV in modulating Arctic sea ice loss and wetting and highlight the need to account for AMV phase changes in near-term Arctic climate projections.
Author
Ziyi Cai
Fudan University
Qinglong You
Fudan University
James A. Screen
University of Exeter
Hans Chen
Chalmers, Space, Earth and Environment, Geoscience and Remote Sensing
Ruonan Zhang
Fudan University
Zhiyan Zuo
Fudan University
Deliang Chen
Tsinghua University
Judah Cohen
Atmospheric and Environmental Research, Inc.
MIT, Dept Civil & Environm Engn
Shichang Kang
Chinese Academy of Sciences
Weiming Ma
Pacific Northwest Natl Lab, Atmospher Climate & Earth Sci Div
Sergey K. Gulev
Russian Academy of Sciences
G. W. K. Moore
University of Toronto
Univ Toronto Mississauga
Renhe Zhang
Fudan University
Science advances
2375-2548 (eISSN)
Vol. 12 13 eady7595Subject Categories (SSIF 2025)
Meteorology and Atmospheric Sciences
DOI
10.1126/sciadv.ady7595
PubMed
41880515