Impact of Changes in Temperature Variability on Changes in Extreme Warm Spells and Heatwaves

It is widely presumed that the increase in mean regional temperatures associated with global warming increases the risks of extreme warm spells and decreases the risks of extreme cold spells. Our investigation based on different types of global observational reanalysis datasets and large ensembles of atmospheric model simulations of the last half-century with prescribed observed SST and radiative forcing suggests a need to modify this presumption. While extreme cold spells became less frequent almost everywhere, extreme warm spells did not uniformly become more frequent; indeed, their likelihood remained the same or even decreased from the first half to the second half of the period in some regions. Such regions covered nearly 40 percent of the globe in both winter and summer. The basic reason for this was a decrease of temperature variability in those regions that negated the effect of the mean warming shift on extreme warm temperature probabilities. By the same token, the increase of temperature variability in the other 60 percent of globe (including many populated land areas) increased the likelihood of extreme warm spells and heatwaves well beyond what might have been expected from the mean temperature shift alone.

The consistency of the changes in extreme warm spell and heatwave probabilities among the different observational datasets and model simulations examined suggests that they are robust regional aspects of global warming associated with atmospheric circulation changes. They highlight the need for climate models to represent not just the mean regional warming signals but also the changes in regional sub-seasonal temperature variability. An important factor in capturing such changes in temperature variability is the ability of climate models to adequately capture the spatial variation of the mean SST warming, especially in the tropics.