More than half of hot extremes due to climate change

Extreme weather is part of the chaotic nature of weather and emerges through a complex interplay of many factors. Will the weather become more extreme as climate change progresses? Yes and no. It would be wrong to conclude that climate change has no effect on the frequency of such events based simply on the fact that weather extremes existed in the past. However, it is also clear that what is often referred to as ‘global weirding’, or the idea that all weather phenomena are becoming increasingly extreme, falls short. After all, it is not scientifically proven that all weather extremes, such as hail storms and tornadoes, are becoming more frequent.

Because extremes are rare by definition, a localised change in their frequency is statistically difficult to prove. But when all the measuring stations around the world are pooled, a clear picture emerges: there has been a global trend towards more frequent and intense hot extremes since the 1950s. In addition, significantly more stations have recorded increases than decreases in heavy precipitation.

Loaded dice

After Europe’s summer-long heatwave in 2003, scientists concluded that although hot periods of this type can occur without human influence, global warming has more than doubled the likelihood of these hot summers in Europe. In turn, these scientists attributed more than half of the probability of the hot summer in 2003 to warming caused by humans. Like someone cheating with loaded dice in order to roll more sixes, so did warming increase the odds of a heatwave that summer.

Instead of merely determining what fraction of risk of a single hot summer can be attributed to humans, we can also ask ourselves what proportion of all extreme heat and precipitation events occurring worldwide is due to global warming. We examine this question in a study recently published in the scientific journal Nature Climate Change [1]. We show that more than half of the hot extremes worldwide and nearly a fifth of precipitation extremes can be attributed to global warming. Not one of these events is solely the direct result of warming, but warming increases their frequency. And the less common and more extreme the hot extreme or heavy rainfall event, the more this can be attributed to a man-made contribution.

Difference between 1.5˚C and 2˚C

With each increment of warming, the frequency of hot extremes and heavy precipitation events worldwide rises sharply. If temperatures rise globally by 2˚C, we expect twice as many extreme heat events worldwide than we would with a 1.5˚C increase. These global warming targets, which are discussed in climate negotiations and which differ little at first glance, therefore have a great influence on the frequency of extremes.

We use climate models to quantify the heat and precipitation extremes and we know that these also have their weaknesses; for example, when simulating  persistent high pressure systems. Small-scale processes such as convection, that is for instance upward air transport during thunderstorms, are not resolved in these models. Rather, they are merely approximated. Nevertheless, observations of extremes available from the last several decades correspond well with the models.

On the other hand, comparable statements for other types of extreme events, such as hail storms or tornadoes, are much more difficult. Numerous studies on these events do not find a significant increase attributable to climate change. Generally it is statistically more difficult to exclude an external factor than to confirm it. Many small-scale extreme weather events, such as hail storms, also fall literally through the mesh size of the climate models and observational networks.

Basis for comprehensive risk assessment

A substantial proportion of all globally occurring hot extremes and heavy rainfall events can be attributed to warming primarily caused by humans. Since a heat or precipitation event does not have the same socio-economic impact everywhere in the world, it is necessary to combine our approach with regional information on exposure and vulnerability in order to carry out a comprehensive risk assessment. This type of risk assessment could serve as an important scientific basis for decisions on warming targets or even for global questions of liability.