Science Advances: Teleconnection from Arctic warming suppresses long-term warming in central Eurasia

Over the past decades, the Arctic has experienced warming at a rate three to four times the global mean, a phenomenon referred to as Arctic amplification (AA) (1–6). In winter, the Arctic is warming especially fast in the Barents-Kara Sea (BKS) region (7). Accompanying the BKS warming, surface air temperature (SAT) in central Eurasia experienced sustained cooling for approximately 20 years, from the early 1990s to the early 2010s, exhibiting a pronounced warm Arctic–cold Eurasian (WACE) SAT trend pattern (7–9). Previous studies attributed cooling over central Eurasia to BKS warming associated with the sea-ice loss (7–16). It is argued that faster warming in the BKS than the lower latitudes reduces the meridional temperature gradient and potentially weakens zonal winds, inducing an anomalous anticyclonic response in the BKS and Ural regions. This shift potentially increases the possibility of cold winters over Eurasia (7–16).

Despite these findings, the causal linkage between BKS warming and Eurasian cooling remains debated (17–40). Numerous studies attribute interdecadal cooling trends in central Eurasia to internally generated anticyclonic circulation in the BKS and Ural regions, noting that climate models often underestimate the atmospheric circulation and mid-latitude cooling responses to the observed sea-ice reduction in the BKS (17–35). This inconsistency may be due to not only model deficiencies (28, 29) but also a misunderstanding of statistical inference (36, 37). As of now, the causal relationship between BKS warming and Eurasian cooling has yet to be conclusively resolved. The core issue lies in determining whether the observed WACE pattern is primarily driven by atmospheric internal variability, directly influenced by BKS warming, or a combination of both factors. Notably, both hypotheses point to critical anticyclonic circulation anomalies located in the BKS and Ural regions. This geographic overlap indicates that straightforward statistical analysis of observational data alone may not be sufficient to discern whether the Arctic–mid-latitude linkage results from internal dynamics or external BKS warming influences. Hence, effective approaches to separate the mid-latitude impacts of BKS warming from internal climate variability are essential to resolve the debate about the causality of the Arctic–mid-latitude linkage.

Using bespoke model experiments is one approach (41), but it is reliant on models being able to capture the relevant processes. Here, we introduce a complementary approach that can be applied to the observations called the extended dynamical adjustment approach (EDAA) to decompose the BKS SAT changes into winter atmospheric circulation–induced and residual thermodynamically induced components associated with anthropogenic forcings (Materials and Methods). By diagnosing the direction of surface turbulent heat flux (THF) and the changes in the preceding sea-ice concentration (SIC), we further validate the effectiveness of EDAA and support the inferred causality of the EDAA decomposition. We interpret the thermodynamically forced component as being a causal response to BKS warming (i.e., triggered by anomalous upward surface heat fluxes over the BKS during its positive phase), accompanied by the preceding SIC changes, and the winter atmospheric circulation–induced component as being a consequence of atmospheric internal variability that is a driver of, and not a response to, BKS SAT change (i.e., causing the anomalous downward surface heat fluxes over the BKS during its positive phase), without accompanied preceding SIC changes.