A dynamic link between spring Arctic sea ice and the Tibetan Plateau snow increment indicator

Arctic sea-ice anomalies significantly impact local and remote climate extremes through thermodynamic and radiative processes. Recent decades have witnessed substantial sea-ice loss due to factors like oceanic warming, feedback mechanisms, and atmospheric energy transport. Arctic sea-ice variability is influenced by tropical and mid-to-high latitude ocean-atmospheric processes and global warming. While studies have focused on summer, autumn, and winter variability, spring Arctic sea-ice variability and its precursors remain less explored. Previous research suggests that TP snow cover anomalies could be a precursor to extratropical circulation systems due to their cross-seasonal climate impacts. Late autumn TP snow cover anomalies may influence winter Pacific North American teleconnection, and spring TP snow cover might affect the Aleutian Low. However, a direct physical link between TP snow conditions and Arctic sea-ice variability wasn't established. This study aims to address this gap by focusing on the TP SCE increment within October-November-December (OND), a refined indicator that captures interannual variability more effectively than the traditional OND averaged TP SCE. The improved indicator, referred to as the snow increment indicator, demonstrates a potential physical linkage with the following early spring Arctic sea-ice concentration (SIC) through sea-ice dynamic processes.

Existing literature extensively documents the impact of Arctic sea ice loss on various climatic phenomena, highlighting its influence on atmospheric processes. Studies have explored the links between Arctic sea ice variability and tropical and mid-to-high latitude climate systems, including the role of atmospheric circulation patterns and global warming. Research has also examined the influence of Tibetan Plateau (TP) snow cover anomalies on extratropical circulation and their cross-seasonal impacts. However, a direct link between TP snow cover and spring Arctic sea ice variability remained unclear, prompting the current investigation to bridge this knowledge gap.

The study analyzed data from 1979 to 2021, using datasets including: National Snow and Ice Data Center (NSIDC) and National Oceanic and Atmospheric Administration (NOAA) snow cover extent (SCE) data, snow depth from the Big Earth Data Platform for Three Poles, snowfall data from ERA5, sea-ice motion vectors from NSIDC, ocean reanalysis data (ORAS5), HadISST sea ice concentration and sea surface temperature, and NCEP-DOE Reanalysis-2 atmospheric data. The authors calculated the TP SCE increment within OND as the difference between December and October SCE. A late autumn SCE increment index (SCEII) was created using an area-weighted mean of the TP SCE increment within OND and a high-pass filter. The late autumn interannual component of the SCE index (SCEI) was similarly calculated using OND TP SCE. Correlation analysis, composite analysis, and information flow analysis were used to examine relationships between SCEII and Arctic SIC. A Community Earth System Model (CESM1.1.1) was used to simulate the impact of TP albedo forcing on atmospheric circulation. Wave activity flux, Rossby wave source (RWS), energy conversion, and Eliassen-Palm (EP) flux calculations were performed to analyze atmospheric wave propagation and dynamics. Statistical methods including high-pass filtering, Fourier harmonics decomposition, and Student's t-test were used for data analysis. The study removed ENSO signals to isolate the TP snow effect.

The study found that the SCEII (snow cover extent increment index) exhibited larger interannual variability, stronger diabatic cooling, faster snow-albedo feedback, and stronger TP local atmospheric circulation linkage than the SCEI (snow cover extent index). The positive TP SCE anomalies associated with SCEII persisted into the following winter and early spring. While simultaneous OND SIC anomalies were nonsignificant, lagged correlations revealed significant negative SIC anomalies in the following DJF and FMA, particularly in the northern Barents Sea, Kara Sea, and Laptev Sea. Information flow analysis confirmed causality from SCEII to Arctic SIC in FMA. The relationship between SCEII and Arctic SIC was independent of ENSO. The formation of a hemispheric-scale atmospheric wave train along the subtropical westerly jet (SWJ), connecting TP and Arctic, was observed, with eastward propagation of wave activity flux from the TP. CESM simulations confirmed that TP albedo forcing triggered similar wave train patterns. The SCEII-associated cyclonic anomalies in the Arctic intensified cyclonic wind stress, leading to outward sea-ice drift and reduced sea-ice thickness and concentration. The study also analyzed Rossby wave source (RWS) anomalies, energy conversion, and Eliassen-Palm flux to explain the wave train formation and maintenance. Analysis indicated that the TP snow-atmosphere interaction was maintained through baroclinic energy feedback, while both barotropic energy conversion and transient eddy activity played key roles in the wave train's maintenance. The northward propagation of negative geopotential height anomalies from the North Atlantic to the Arctic was linked to tropospheric and stratospheric coupling processes. The trend analysis showed that the snow increment indicator had a limited role in the sharply reduced Arctic SIC in recent decades, attributing the sharp decline largely to greenhouse gas-induced global warming. The study also investigated the impact of tropical Indian and North Atlantic Ocean SSTs, finding that the SCEII-SIC relationship remained robust even after removing these SST signals.

The findings highlight the importance of considering the rate of snow accumulation on the TP as a key factor influencing spring Arctic sea-ice variability. The new snow increment indicator (SCEII) proved superior to the traditional indicator in capturing the climatic impact of TP snow cover. The identified dynamic pathway – a TP-Arctic wave train and associated cyclonic anomaly leading to sea ice drift – provides valuable insights into the teleconnections between the TP and the Arctic. This research emphasizes the significant role of TP snow cover, not merely its extent, but its accumulation rate, as a predictor of spring Arctic sea-ice conditions. The robust statistical relationships and causal analysis strengthen the link between TP snow accumulation and Arctic sea ice, furthering our understanding of climate variability on a large scale. The impact of TP snow accumulation appears relatively independent of ENSO and tropical SSTs, suggesting a direct and distinct influence.

This study introduced a novel snow increment indicator (SCEII) for TP snow variability, demonstrating its superior ability to capture the impact of TP snow on Arctic sea ice compared to traditional indicators. The key finding is the robust link between faster TP snow accumulation and subsequent reduced spring Arctic sea ice through a TP-Arctic wave train and sea-ice dynamic processes. This research highlights the crucial role of TP snow accumulation dynamics in influencing Arctic climate. Future research could focus on refining the model simulations to improve the representation of snow-atmosphere interactions and further explore the role of other factors in modulating the relationship. Investigation into the long-term trends and potential changes in this linkage under continued climate change is also needed.

The study relied on observational datasets and model simulations, each having inherent limitations. The model simulations may not perfectly capture the complexity of atmospheric processes and the interactions between the TP and the Arctic. The causal analysis using information flow might not capture all complex feedback loops. The focus on the interannual time scale might limit the generalizability of findings to other time scales. Further research is needed to address these limitations and confirm the findings using independent datasets and methods. The limited resolution of some datasets might affect the accuracy of results particularly in high spatial variability regions.