Since the 1990s, Arctic winter temperatures have shown an amplified warming trend referred to as Arctic amplification (AA). AA is strongest over the Arctic Ocean during the fall and winter seasons, while during the summer it is weaker and shifted over land. Over the same period, interior North America, Northern Europe and Northern Asia, land temperatures in winter actually cooled from 1990–2013 followed by more variable winters. The recent mid-latitude winter cooling period has coincided with an increase in severe winter weather events as shown in our previous Communications, Earth Environment article, Cohen et al. 2023.
The accelerated warming of the Arctic coupled with cooling or lack of warming in the mid-latitudes has resulted in a vigorous debate of whether AA and mid-latitude winter temperature cooling or muted trends are coincidental or physically related. The pattern of a warm Arctic and cold continents (WACC) is the strongest observational evidence that some unaccounted-for mechanism has been offsetting greenhouse gas-forced warming over the Northern Hemisphere (NH) mid-latitudes though the support from global climate models is mixed or even contrary to the idea that AA can force mid-latitude cooling.
In our initial observational study published in Cohen et al. 2018, we wanted to explore how strong is the relationship between a warm or cold Arctic and severe winter weather, not just extreme cold but also heavy snowfall. A metric was developed by scientists at NOAA in Mayes-Boustead et al. which aggregates severe winter weather, the Accumulated Winter Season Severity Index or AWSSI. The index is based on in-situ meteorological observations of heavy snowfall and below freezing temperatures across the US. It is reported as an accumulated daily value throughout the winter season (December, January, February) and informs comparisons of weather severity over time. The AWSSI integrates both intensity and duration of extremes in daily temperature, snowfall, and snow cover into one index to assess weather severity across seasons and locations.
Our initial analysis using AWSSI data demonstrated a strong and quasi-linear relationship between Arctic temperatures and severe winter weather in the Northeastern US where relatively cold Arctic temperatures are related to a decrease in severe winter weather and relatively warm Arctic temperatures are related to an increase in severe winter weather. Our initial analysis also found that in the early winter when the polar vortex was generally strengthening, severe winter weather in the northeastern US was decreasing. However, in mid- to late-winter when the polar vortex was found to be generally weakening, severe winter weather in the northeastern US was increasing. Though the AWSSI was very informative about revealing the relationship between Arctic temperatures and severe winter weather, it was limited to the US.
In this new study we extended the AWSSI from the US only, to global, by computing the same metrics using ERA5 reanalysis, which is an approximation of observational gridded data for the entire globe. We refer to this new global metric of severe winter weather derived from reanalysis as rAWSSI.
We found that the same strong and quasi-linear relationship between Arctic geopotential heights (a measure of atmospheric pressure) and severe winter weather in our initial study existed not only for the Northeastern US but also for Northeastern Europe, especially Scandinavia, and Northern Asia with the new rAWSSI gridded data and for a longer period of analysis. However, one important difference between the new study using ERA5 and the old study using NCEP/NCAR reanalysis, is that the relationship between Arctic temperatures and mid-latitude severe winter weather was more complicated. Though severe winter weather was more likely when the Arctic was warmer than normal, sometimes that was also true when the Arctic was colder than normal in the lower troposphere. However, in the upper troposphere and lower stratosphere the same strong and quasi-linear relationship between Arctic temperatures and severe winter weather was found as in the initial study. For maritime locations such as the Western US, Western Europe and East Asia we found a weaker relationship between Arctic temperatures or geopotential heights and severe winter weather.
Another important and notable finding that was consistent between the two studies was that when the Arctic warming was shallow and the polar vortex was in general strengthening, there was an overall decreasing trend in severe winter weather not only in the Northeastern US but also Northeastern Europe and Northern Asia. However, in the mid- to late-winter when Arctic warming extends into the upper troposphere and lower stratosphere and the polar vortex weakens, there are at least periods when severe winter weather is increasing in the period of AA. Though as we also show since 1950, severe winter weather is decreasing across all of the Northern Hemisphere, more strongly over the Arctic Ocean, with the one exception of a small region in Central Asia.
Though our study is statistical in nature and therefore we could not establish causality, one of our last findings of the study was that polar vortex variability is increasing along an axis from Central Asia to central North America. This is strongly suggestive that the polar vortex has become less circular and more elongated over the observational period and that the stretched or elongated polar vortex state has become more frequent. The stretched polar vortex state favors more severe winter weather in both Central Asia and central North America. Therefore, our analysis is further support that more frequent disruptions or weakenings of the polar vortex have either offset an overall decrease in severe winter weather or even at times contributed to an increase in severe winter weather periodically and/or regionally during the period of AA.
Whether AA is contributing to more severe winter weather periodically and/or regionally across the mid-latitudes of the Northern Hemisphere continues to be vigorously debated. The foundation of the debate is that observational studies have revealed an increasing trend in colder temperatures and/or heavier snowfall. In contrast many global climate models show either a more damped response in winter weather to Arctic warming than observed or even only supports milder and/or less snowy winters for the mid-latitudes due to AA. Our analysis suggests that one possible source for the contradictory findings between observations and models is the vertical extent of the warming. When the Arctic warming is shallow there is no influence on mid-latitude winter weather and only when the Arctic warming is deep then does severe winter weather in the mid-latitudes increase.
References:
Cohen, J., Agel, L., Barlow, M. & Entekhabi, D. No detectable trend in mid-latitude cold extremes during the recent period of Arctic amplification. Comm. Earth & Env. 4, 341 (2023). https://doi.org/10.1038/s43247-023-01008-9
Mayes-Boustead, B. E., Hilberg, S. D., Shulski, M. D. & Hubbard, K. G. The Accumulated Winter Season Severity Index (AWSSI). J. Appl. Met. Clim. 54, 1693–1712 (2015).
Cohen, J., Pfeiffer, K. & Francis, J. Warm Arctic episodes linked with increased frequency of extreme winter weather in the United States. Nat. Commun. 9 (2018). https://doi.org/10.1038/s41467-018-02992-9
Please sign in or register for FREE
If you are a registered user on Research Communities by Springer Nature, please sign in