Robert Vislocky, Ph.D.
An article published in Nature Communications (“Rapid Flips Between Warm and Cold Extremes in a Warming World”) has been making its rounds recently in several climate forums on social media even though it was published a year ago (see link below). In the article the authors claim that flips between warm and cold extremes are becoming more frequent and intense in a warming world.
According to the authors, rapid extreme temperature flips “allow a very short time for human and ecosystem preparedness to respond and adapt” and “can increase the risk of severe and even irreversible impacts on human health, infrastructure, air quality, and plant phenology.”
Of course the news media swarmed all over the article like a fly to honey and pimped up the scare. Check out some of the dire headlines below:
“Sweating to Shivering: Study Finds Rapid Swings in Temperature Have Increased”https://www.nytimes.com/2025/04/22/climate/temperature-swings-climate-change.html
“Study Finds Sudden Weather Whiplash Events on the Rise”
https://www.usu.edu/today/story/study-finds-sudden-weather-whiplash-events-on-the-rise
“From boiling hot to freezing cold: Climate crisis sparks sudden temperature changes”
Now it’s time for the reality check. Figure 1 below from the article shows the trends in the frequency (a, b), intensity (c, d) and transition duration (e, f) of extreme warm-to-cold and cold-to-warm flips for the period 1961 to 2023. The global trend is shown in the smaller embedded charts while the regional trends are shown as shaded areas on the maps. Extreme flips are defined as going from at least one standard deviation above normal to at least one standard deviation below normal (or vice-versa).
Figure 1. Trends in the frequency (a, b), intensity (c, d) and transition duration (e, f) of extreme warm-to-cold and cold-to-warm flips. Wu, S., Luo, M., Lau, G.NC. et al. Rapid flips between warm and cold extremes in a warming world. Nat Commun 16, 3543 (2025).
Now brace yourselves, according to their charts the number of extreme warm-to-cold flips increased from about 1.64 flips per year in 1961 to a whopping 1.82 flips per year sixty-two years later in 2023 (Fig 1a). That’s an increase of 0.29 extra extreme flip days per century! The number of extreme cold-to-warm flips increased from about 1.43 flips per year in 1961 to a colossal 1.51 flips per year in 2023 (Fig 1b). That’s an increase of 0.12 extreme flip days per century. With that kind of change I don’t know how life could possibly adjust (tongue-in-cheek humor).
While the authors may be able to show statistical significance of these frequency trends, in reality they are meteorologically and physically insignificant to the point where no human, animal, or plant would even notice the change (even if such change was real).
In addition to the flip frequency, the authors claim that the transition duration is decreasing too. This is the time it takes in days to go from a warm extreme to a cold extreme (or vice-versa). In the graphs it appears the average flip between extremes takes a little over four days or roughly 100 hours. After a whole century of warming the average flip between extremes would be about 2.5 hours less based on their trend lines (Figs 1e, 1f). Again, this change is entirely meaningless as is the intensity increase of 0.15 standard deviations per century (Figs 1c, 1d).
On top of the meaningless trends in observed extreme temperature flips, there are several questionable aspects of the study that raise red flags for concern. For starters, why did the authors begin their analyses in 1961? The temperature datasets that they used in the study all go back much further than 1961 (e.g., 1940 for the ECMWF ERA5 reanalysis dataset, 1940 for the NCEP/NCAR reanalysis dataset, and 1850 for Berkeley Earth). So why not start the trend in 1940, for example, in line with when the reanalysis data sets begin? Maybe it’s because the trends disappeared or reversed when longer datasets are utilized? Interestingly, although the 1961-2023 data shows a tiny increase in flip frequency, the data since 1980 shows absolutely no trend according to their charts in (Figs 1a and 1b). So it’s quite possible the 1961 start date was cherry picked to show a positive trend in flip frequency.
But wait, there’s more! The authors apply CMIP6 climate model simulations using four different emission scenarios out to 2100 (see Figure 2a to 2f below). Under SSP5-8.5 and SSP3-7.0 scenarios the authors show roughly a 7% increase in the frequency and intensity of flips between extreme warm and extreme cold temperatures (and vice-versa) by the end of this century. Accompanying those results is the usual appeal to emotions that these flips will disproportionately affect poorer nations. Nonetheless, both of those high emission scenarios have been recently declared unrealistic by the IPCC. Under more realistic emission scenarios that were tested by the authors, no meaningful trends in extreme flips are shown out to 2100. What’s even more interesting is that the historical climate model simulation from 1961 to 2023 that they use as a baseline actually shows a slightly declining frequency of extreme temperature flips during this period (Figs 2a, 2b, gray curve), which is contrary to their observational study in Figure 1 that showed a tiny increase.
The bottom line is this is a perfect example of another study from the climate community that presents very marginal results with significant red flags but is sensationalized by the news & social media who refuse to bat a critical eye at the study along with an apparently forgiving peer review.
