The Arctic has experienced warming at a rate two or more times faster than the global average since the mid-20th century, leading to an accelerating pace of Arctic sea ice melt, particularly in autumn. Numerous studies have explored the impacts of this sea ice loss on both regional and remote climate patterns. For example, autumn sea ice loss in the Barents-Kara Seas is linked to increased frequency of cold Eurasian winters and snowstorms, influencing Eurasian blocking, East Asian winter monsoon circulations, and land processes. Sea ice decline in the Kara-Laptev Seas even affects subsequent summer precipitation over East Asia. Conversely, reductions in the Chukchi-East Siberian Sea correlate with cooler temperatures in central North America, and changes around Greenland are tied to temperature variability in northern Europe and eastern North America. However, some studies have challenged these findings, reporting weaker or less consistent effects of Arctic sea ice loss on mid-latitude weather patterns. While the climatic effects of Arctic sea ice loss remain debated, the underlying physical drivers are a subject of intense research. Anthropogenic forcing is a primary contributor to the long-term trend of sea ice decline. However, internal climate variability also plays a crucial role. For instance, anomalous summertime anticyclonic circulation over Greenland and the Arctic Ocean is estimated to contribute significantly to September Arctic sea ice loss. Several atmospheric modes, including the Arctic Dipole, the Arctic Oscillation/North Atlantic Oscillation, and the Pacific North American pattern, have also been identified as influential factors in Arctic sea ice variability. This study focuses on the Asian-Pacific Oscillation (APO), a key mode of atmospheric internal variability in the extratropical Asian-Pacific sector, particularly prominent in summer. The APO is characterized by a seesaw pattern of upper-tropospheric temperatures between Asia and the North Pacific. The study investigates the potential link between the summer APO and autumn sea ice variability in the eastern Arctic using reanalysis data and simulations from the Community Earth System Model Large Ensemble (CESM-LE).

Existing literature demonstrates a strong link between Arctic sea ice loss and various climate phenomena. Studies highlight the connection between autumn sea ice decline and increased frequency of cold winters and snowstorms in Eurasia, impacting weather patterns through mechanisms such as changes in Eurasian blocking patterns, alterations in East Asian winter monsoon circulations, and influencing land processes. The relationship between sea ice reduction and subsequent summer precipitation over East Asia has also been examined, with the seasonal persistence of snow depth and soil moisture playing a role. Additionally, research has explored the links between sea ice reductions in the Chukchi-East Siberian Seas and cooling in North America, and the connection between sea ice changes around Greenland and temperature changes across northern Europe and eastern North America. However, other studies have raised concerns regarding the strength and robustness of these relationships, questioning the extent to which Arctic sea ice loss directly influences mid-latitude weather.

This study employs a combination of observational analyses and climate model simulations to investigate the relationship between the summer Asian-Pacific Oscillation (APO) and autumn sea ice variability in the eastern Arctic. **Reanalysis Data:** The analysis utilizes monthly sea ice concentration (SIC) data from the Met Office Hadley Centre Sea Ice and SST dataset (HadISST1), sea surface temperature (SST) data from the Extended Reconstructed SST version 5 (ERSST v5), and atmospheric reanalysis data from the National Centers for Environmental Prediction/National Center for Atmospheric Research (NCEP/NCAR). Data are used to analyze the period 1950-2019. **CESM-LE Simulations:** To validate the observational findings, the study uses the Community Earth System Model Large Ensemble (CESM-LE) simulations. Forty ensemble members are used, each subjected to the same historical forcing but with different initial conditions. A subset of these members, exhibiting a significant relationship between summer APO and autumn SIC consistent with observations, was selected as the BMME ensemble. A second subset showing a contradictory relationship constitutes the WMME ensemble. **Statistical Analyses:** Maximum covariance analysis (MCA) is used to identify the dominant modes of co-variability between summer upper-tropospheric eddy temperature (UTT) over the Asian-Pacific sector and autumn SIC in the eastern Arctic. Regression analysis examines the relationships between autumn SIC anomalies and the summer APO index. Wavelet coherence analysis investigates the co-variability across different timescales. Lead-lag correlations are applied to determine the temporal relationship between the APO and North Pacific/North Atlantic SSTs. Partial correlations are used to distinguish the unique influences of North Pacific and North Atlantic SSTs. Analysis of the horizontal wave activity flux helps identify the propagation of atmospheric waves. Monte Carlo tests with 1000 simulations are used to assess the statistical significance of correlations and regressions.

The study found a significant negative correlation (r = −0.59, p < 0.01) between the summer APO index and the autumn sea ice concentration index (SICI) in the eastern Arctic. A positive phase of the summer APO (warmer UTT over Asia, cooler UTT over the North Pacific) was associated with reduced autumn sea ice in the eastern Arctic. This relationship is supported by squared wavelet coherence analysis showing significant in-phase coherency across various timescales (5–8 year, 8–12 year, and 16–24 year bands). The physical mechanism linking the summer APO to autumn sea ice loss involves a sequence of atmospheric and oceanic processes. A positive summer APO leads to warmer SSTs in the mid-latitude North Atlantic. This warming persists into autumn, influencing atmospheric circulation patterns over the Barents-Kara-Laptev Seas (strong anticyclonic anomalies) and the East Siberian Sea (weak lower-tropospheric cyclonic anomalies). This circulation pattern enhances the transport of warm, moist air into the eastern Arctic, leading to increased lower-tropospheric humidity, downwelling longwave radiation (DLR), and surface air temperature (SAT). These increased humidity, DLR, and SAT are significant contributors to autumn sea ice melt. The correlations of summer APO with the autumn low-level humidity, DLR, and SAT averaged over the eastern Arctic are 0.52, 0.58, and 0.56 (p<0.01), respectively. The analysis of CESM-LE simulations supports the observational findings. The BMME ensemble, which captured the observed negative correlation between summer APO and autumn SIC, showed similar patterns to the observations: positive summer APO leading to warmer SSTs in the North Pacific and North Atlantic, with persistent warming in the North Atlantic into autumn. The WMME ensemble which simulated a positive correlation did not show persistent SST warming in the North Atlantic in autumn. This difference suggests the crucial role of North Atlantic SSTs in the lagged influence of the summer APO on autumn sea ice.

This study demonstrates a robust linkage between the summer APO and subsequent autumn sea ice loss in the eastern Arctic, mediated by North Atlantic SSTs. The positive phase of the summer APO induces warming in the mid-latitude North Atlantic, which persists into autumn. This warming triggers a chain of atmospheric changes leading to increased moisture transport and enhanced thermodynamic conditions conducive to sea ice melt in the eastern Arctic. The results highlight the importance of internal climate variability, specifically the APO, in modulating Arctic sea ice change, adding to our understanding of factors beyond anthropogenic forcing. The recent decadal shift towards a positive summer APO phase likely exacerbates the autumn sea ice loss driven by global warming. However, future projections of the APO show a weakening trend, which could potentially counteract some of the effects of anthropogenic warming on sea ice. This adds to the complexity of predicting future Arctic sea ice extent. The study focuses on the role of the North Atlantic SSTs as a key pathway in this process, acknowledging that other pathways may also play a role and warrant future investigation.

This research provides strong evidence that the summer Asian-Pacific Oscillation (APO) significantly influences autumn sea ice extent in the eastern Arctic through a pathway involving persistent North Atlantic SST warming. The findings highlight the importance of considering both anthropogenic forcing and internal climate variability when projecting future Arctic sea ice changes. While the study points to a mechanism for accelerated sea ice loss, future research is needed to fully explore all the complexities and uncertainties involved in predicting future Arctic sea ice changes. Improving the representation of the APO in global climate models is important to reduce uncertainty in these predictions.

The study acknowledges several limitations. The primary focus is on the pathway of influence through the North Atlantic SSTs; other processes might be involved. The interpretation of the CESM-LE simulations might be limited by the specific selection of BMME and WMME ensembles; a broader range of simulation results may reveal additional insights. The study focuses on the relationship between summer APO and autumn sea ice, neglecting potential influences from other seasons or the influence of sea ice loss on other aspects of Arctic climate.