The motivation for this analysis was an historic Arctic outbreak into the United States (US) accompanied by a record-breaking blizzard in the Great Lakes region that occurred just prior to and during the Christmas holiday period in December 2022.  This extreme weather event resulted in dozens of fatalities and was one of the costliest weather events in the US during the 2022 calendar year.   This event was extensively covered by the media including this article in the NY Times (https://www.nytimes.com/2022/12/22/climate/polar-vortex-winter-cold-weather.html).  The article first presented ideas how climate change could be contributing to an increase in extreme winter weather but then presented the other side: that any contribution of climate change, but specifically Arctic change, may have contributed to an increase in cold extremes that was limited to the 1990’s and 2000’s and has disappeared over the past decade, citing a recent paper in Nature Climate Change (https://www.nature.com/articles/s41558-020-00954-y.epdf).

There has been considerable debate on the relationship between Arctic change and cold extremes in parts of the mid-latitude continents.  The Arctic has experienced accelerated warming in recent decades, observed to be between two to four times greater in the Arctic relative to the rest of the globe, which is known as Arctic amplification (AA).  Coincident with part or all of the period of AA, regions of the mid-latitude continents have experienced a cooling trend, with the most pronounced and consistent winter cooling trends in the interior of both Eurasia and North America. Whether the juxtaposition of the two opposite anomalies are coincidental or whether the two are dynamically linked, whereby the rapidly warming Arctic is contributing, at least in part, to more severe winter weather in parts of Eurasia and/or North America, has been debated for at least a decade and remains unresolved.

Given that the Arctic is the source region for cold extremes across the mid-latitudes, a common argument applied to mid-latitude cold events is that as the Arctic warms more rapidly than other regions of the globe, cold extremes will moderate or warm and occur less often. A recent report from the National Academies of Science, when assessing the change in frequency of all weather extremes, was most confident in the decrease of cold extremes.

For more than a decade, my co-authors and I have been making the argument that AA could be contributing to more severe winter weather.  Most recently, following another historic US cold air outbreak and record snowstorms in February 2021, we looked at the potential role of Arctic snow cover and sea ice changes  (https://doi.org/10.1126/science.abi9167).  This extreme weather event resulted in hundreds of fatalities and may have been the costliest winter weather event in US history based on insured losses.  Due to my involvement in operational seasonal forecasting, I follow winter weather on a daily basis across the Northern Hemisphere in real time.  I am aware of not only numerous extreme winter weather events in the US but also East Asia since the early 2000’s through winter 2022/23. The most recent severe winter was in 2007/08 in China, but China did record a new all-time lowest national temperature in January 2023.  In the US there are also many recent examples of extreme winter events especially in the past decade including the winters of 2013/14, 2014/15, 2017/18, 2018/19, 2020/21 and 2022/23. Based on these recent events, we wondered if the argument in the New York Times article that trends in cold extremes “have not continued over the past decade” was supported by the observed data.

Therefore, we did an in-depth analysis of cold extremes in many of the more heavily populated regions of the Northern Hemisphere mid-latitudes, including the Eastern US, Northeast Asia, and Northern and Southern Europe.  Trends in cold extremes were compared to pan-Arctic temperatures poleward of 70ºN three days prior to the observed cold extreme in the populated regions.  We computed trends for the period since 1960, which is at least three decades before AA commenced, and since 1990 and 2000 (both considered the start dates of AA).  Trends for all three periods were calculated through winter 2022/23 with cold extremes defined as the 5% coldest days in the observational record over the longest period beginning in 1960.  We computed both the trends in frequency of cold extremes and the magnitude of cold extremes.

The most robust finding of our trend analysis was that the Arctic is warming since 1960 and the warming is statistically significant (see Table 1).  However overall trends in cold extremes across the mid-latitudes were found to be mixed and mostly not statistically significant.  The most robust finding related to cold extremes was a statistically significant decrease in the frequency of cold extremes since 1960 with the exception of Northeast Asia. But for the Eastern US, Northeast Asia and Northern Europe we did not find a statistically significant trend in the frequency or magnitude of cold extremes during either period of AA. Therefore, an influence of overall global warming since 1960 was found to be consistent with a decrease in frequency in cold extremes though not on the magnitude. However no consistent influence on the frequency or magnitude of cold extremes during the period of AA was found.

Table 1. Temperature trends show divergence between Arctic warming and mid-latitude cold extremes. Temporal trend in CEUS, SSNC, NEUR and SEUR DJF 5% coldest days and corresponding (3 days previous) Arctic temperatures (℃/30 events) from ERA-5. Bolded values are statistically significant at 0.05 confidence level.

In fact, the only statistically significant trend in the magnitude of cold extremes associated with AA was a strengthening of cold extremes in the Eastern US since 2000 (see Table 1).  This is consistent with media coverage, the experience of many and the arguments that AA can also at least partially contribute to more severe winter weather and contrary to the argument that any trends in cold extremes have disappeared since 2010.

It is likely that our analysis will only extend the debate about the influence of AA on winter weather across the mid-latitudes rather than fostering a consensus.  But our analysis importantly illustrates that the influence of Arctic change on weather at lower latitudes is complex and poorly understood and requires further and extensive research into the physical causes of the observed relationships.