Extreme cold snaps and global warming: A speculative explanation

  • Date: 09/09/21
  • Dr David Whitehouse, GWPF Science Editor

Another example of a climate science story in which significant uncertainties are underplayed, results overhyped and the speculative nature of initial research poorly portrayed.

The U.S. Southern Plains cold snap in February 2021 resulted in the collapse of the Texas energy grid with damages estimated at nearly USD 200 billion, substantially higher than that caused by recent hurricanes. While global warming is predicted to result in increases in some weather extremes, such as heat waves and heavy precipitation events, it is not generally considered to be responsible for severe winter weather at mid-latitudes.

Climate models are also in trouble because of reports that in the Northern Hemisphere intense snowfall events and unusual cold snaps have become more frequent, according to some research. The models do not explain this.

A new paper links the Texas cold snap to changes happening in the Arctic and suggests that warming there is a likely driver of the recent extreme winter weather. Writing in the journal Science, Cohen at al. suggest a link between stratospheric winds and severe mid-latitude cold spells. They say that despite the atmosphere generally getting warmer winds from the polar regions, they can have a very large cooling effect in certain circumstances, concluding that colder winters are possible with global warming.



It all depends upon the behaviour of the so-called stratospheric polar vortex (SPV), an atmospheric feature defined by strong westerly winds that circle the Arctic and, under normal conditions, keep its cold air contained.
It’s suggested that accelerated warming in the Arctic – so-called Arctic amplification (AA) – is driving these events by disrupting the SPV. Cohen et al. identifies a “stretched” SPV state that allows cold polar air to move farther south than usual. According to them the frequency of such a state has increased in recent decades, doubling in the 1980-2020 period.

They extend their research using regression analysis to find a link between these excursions and Arctic Sea Ice and Eurasian snow cover. At this stage this is, of course, only indicative of a correlation – not causation. To ‘establish’ the connection they use models: two simulations are performed. Firstly, they reduce the albedo over Eurasia and secondly add atmospheric heating over the Barents-Kara Seas. They found that the atmospheric response to doing this ‘resembled’ a stretched polar vortex state thus ‘providing evidence’ for causality.

Sensible commentators will be careful here as it is tempting to see this paper as definitive; after all, the authors themselves say this is the first paper of its kind to see this connection. For example, the BBC report gives the wrong impression although its description meanders from saying these factors ‘are linked’, but later on changes this to ‘indicates.’ I also don’t agree with Professor Jennifer Francis that the “study takes this controversial hypothesised linkage and moves it solidly towards accepted science.” Whatever the eventual outcome, what has not been reported is that the effect is small compared to natural variability.

This area of research remains much more speculative than those who reported the Science paper would lead you to believe. It’s another example of a climate science story in which significant uncertainties are underplayed, results overhyped and the speculative nature of initial research poorly portrayed.

The scientists hope that this paper will provide the basis for an early warning of cold snaps. Unusually, people will have the opportunity to cast judgement on the quality of such warnings. The lead author’s main job is providing such forecasts to subscribers.

Feedback: david.whitehouse@thegwpf.com

Recent Popular Articles


We use cookies to help give you the best experience on our website. By continuing without changing your cookie settings, we assume you agree to this. Please read our privacy policy to find out more.