Li et al. (2025) Atmospheric circulation regimes modulating Eurasian winter decadal cooling

Identification

• Journal: npj Climate and Atmospheric Science
• Year: 2025
• Date: 2025-10-22
• Authors: Juncong Li, Xiaodan Chen, Zhiping Wen, Shengping He, Yuanyuan Guo, Sihua Huang, Yu Zhu
• DOI: 10.1038/s41612-025-01228-0

Research Groups

• Department of Atmospheric and Oceanic Sciences and Institute of Atmospheric Sciences, Fudan University, Shanghai, China
• Geophysical Institute, University of Bergen and Bjerknes Centre for Climate Research, Bergen, Norway
• Key Laboratory of Polar Atmosphere-Ocean-Ice System for Weather and Climate, Ministry of Education, Fudan University, Shanghai, China
• Shanghai Key Laboratory of Ocean-Land-Atmosphere Boundary Dynamics and Climate Change, Fudan University, Shanghai, China
• Institute of Eco-Chongming, Shanghai, China
• Nansen Environmental and Remote Sensing Center, Bergen, Norway
• Shanghai Climate Center, Shanghai Meteorological Service, Shanghai, China
• Guangzhou Institute of Tropical and Marine Meteorology, China Meteorological Administration, GBA Academy of Meteorological Research, Guangzhou, China

Short Summary

This study identifies three comparable decadal Eurasian winter cooling episodes since 1901, demonstrating that the Scandinavian pattern (SCAND) and North Atlantic Oscillation (NAO) are intrinsic atmospheric regimes modulating these events, with a projected 30% likelihood of another analogous cooling episode by 2050 due to internal variability.

Objective

• To identify the specific internal atmospheric variability patterns responsible for decadal Eurasian winter cooling episodes, including the early 21st-century event, and to project the likelihood of their future occurrence.

Study Configuration

• Spatial Scale: Eurasia (40°–60°N, 45°–110°E), Barents-Kara Seas (70°–85°N, 30°–90°E), Northern Hemisphere (north of 20°N).
• Temporal Scale: Decadal (15-year rolling trends), historical (1901–2023 for reanalysis, 1850–2023 for LENS2 historical), future projections (2025–2099 for LENS2 and CMIP6), with a focus on winter (January–February mean).

Methodology and Data

• Models used: Community Earth System Model (CESM) 2 Large Ensemble (LENS2), Coupled Model Intercomparison Project Phase 6 (CMIP6) models (16 models, 29 members).
• Data sources:
  • Reanalysis: ECMWF Reanalysis V5 (ERA5), NOAA-CIRES-DOE 20th Century Reanalysis V3 (20CRV3).
  • Observational datasets: HadISST (Sea Surface Temperature), multi-source Surface Air Temperature (CRU, HadCRUT, NOAAGlobalTemp, GISTEMP, Berkeley Earth, UDel).
  • Variables: Surface air temperature (SAT), geopotential height (HGT), horizontal wind (U and V), sea ice concentration (SIC), precipitation.
• Methods: Decomposition of SAT/SST into internal variability (IV) and externally forced (EX) components, 15-year rolling trend analysis, one-dimensional blocking algorithm, horizontal wave activity flux (WAF) calculation, Rotated Empirical Orthogonal Function (REOF) analysis for SCAND and NAO patterns, Bayesian multilevel logistic regression model.

Main Results

• Three comparable decadal Eurasian winter cooling episodes (P1: 1917–1931, P2: 1958–1972, P3: 1998–2012) are identified in reanalysis, occurring with an approximate 40-year cycle and primarily governed by atmospheric internal variability.
• These cooling episodes are not always accompanied by pronounced Arctic warming, challenging the "warming Arctic-cooling Eurasia" (WACE) concept.
• A substantial positive Scandinavian pattern (SCAND) trend strongly favors Eurasian cooling (anti-correlation r = –0.82, p < 0.01 in reanalysis; r = –0.52 ± 0.12, p < 0.01 in LENS2).
• The North Atlantic Oscillation (NAO) trend modifies the cooling magnitude: NAO+ attenuates cooling, NAO0 results in moderate cooling, and NAO− amplifies cooling. For SCAND+ cases in LENS2, NAO+ leads to -0.48 °C decade⁻¹ cooling, NAO0 to -1.00 °C decade⁻¹, and NAO− to -1.54 °C decade⁻¹.
• A Bayesian multilevel logistic regression model projects an approximately 30% likelihood of at least one analogous strong Eurasian cooling episode (below -1.5 °C decade⁻¹) occurring before 2050 due to internal variability, after applying observational constraints.
• Anthropogenic warming (EX component) is projected to mask about 40–60% of internal variability-driven cooling episodes in the near to mid-term future (2025–2074), with only half of the projected cooling potentially overriding anthropogenic warming.
• No coherent Sea Surface Temperature (SST) patterns were found across the three cooling periods, but a robust linkage between SCAND+ trends and enhanced tropical eastern Atlantic convection was identified.

Contributions

• First-time identification of the periodicity (approximately 40 years) and specific atmospheric internal variability (SCAND and NAO) driving decadal Eurasian winter cooling episodes, including the early 21st-century event.
• Redefines the role of NAO as a modulator of cooling magnitude rather than an independent driver, in conjunction with the primary influence of the SCAND pattern.
• Develops a Bayesian multilevel logistic regression model to project the likelihood of future strong Eurasian cooling episodes, incorporating both internal variability and the masking effect of externally forced warming.
• Provides important insights for energy security strategies in midlatitude Eurasian regions by offering an early warning system for extreme cooling risks.

Funding

• National Key Research and Development Program of China (Grant No. 2022YFF0801701)
• National Natural Science Foundation of China (Grant No. 42405066)
• BASIC project (Grant No. 325440)
• Nansen Center´s basic institutional funding (Grant No. 342624) funded by the Research Council of Norway
• Sigma2, the National Infrastructure for Computational Science in Norway (projects NN10055K, NN8121K, NS10055K, and NS8121K)

Citation

@article{Li2025Atmospheric,
  author = {Li, Juncong and Chen, Xiaodan and Wen, Zhiping and He, Shengping and Guo, Yuanyuan and Huang, Sihua and Zhu, Yu},
  title = {Atmospheric circulation regimes modulating Eurasian winter decadal cooling},
  journal = {npj Climate and Atmospheric Science},
  year = {2025},
  doi = {10.1038/s41612-025-01228-0},
  url = {https://doi.org/10.1038/s41612-025-01228-0}
}