The largest impacts of increases in temperature will not be experienced due to changes in the mean, but by changes in the extremes. This is illustrated using UK heatwaves to show that the number of days which will require preventative action will increase by roughly a factor of 6 for a 2ºC increase in summer temperatures.
Summer heatwaves can be extremely uncomfortable and even fatal, especially to the elderly. For example, an estimated 40,000 extra people died across Europe because of the extreme heat in the summer of 2003. Crop yields can also be affected by extreme heat.
The 2003 event prompted action across Europe to improve warnings and reduce vulnerability, which may have helped reduce the number of extra deaths during a heatwave in 2006.
The temperature thresholds for alerts vary across the country. In the North East of England a daily maximum temperature of 28ºC and minimum temperature of 15ºC defines an extreme heatwave day. For the South East, the thresholds are 31ºC and 16ºC respectively, highlighting how previous experiences may have already helped people and infrastructure to adapt.
Both thresholds are important, but the number of extreme warm nights has increased far faster than the number of extreme hot days over the past 60 years for England & Wales (Figure 1).
Using longer averages (30-years) and the combination of thresholds (i.e. both a hot day & warm night), an extreme heatwave event occurred roughly once every two years in some regions during the 1955-1984 period, and far less frequently for most of the UK (Figure 2).
More recently (1985-2014), the number of events has increased slightly, but is still less than once per year on average almost everywhere. It should noted that heatwave days tend to occur in batches (e.g. 1976, 2003 & 2006), rather than being more regularly spaced throughout the periods.
If temperatures increased by 2ºC everywhere (both day & night) – which might sound quite appealing – then the number of extreme hot events increases markedly (Figure 2) with many areas experiencing an average of 2-4 days per year on which preventative action would be required – an increase by about a factor of 6.
Note that this is an illustrative example with a number of caveats. Daily minimum temperatures are projected to continue to increase more rapidly than maximum temperatures. A 2ºC change in temperature would also likely be accompanied by a change in atmospheric circulation patterns, which might increase or decrease the number of extreme heatwave situations. The projected decrease in rainfall in spring & summer might increase the chances of heatwaves as the soils will be drier.
An increase of 2ºC is projected to occur around mid-century for the RCP8.5 emissions scenario (which assumes no climate mitigation), and around the 2080s for the RCP4.5 emissions scenario which does include some climate mitigation. Whether such a change will end up causing more fatalities depends on how well we can adapt to the heat.
Notes: This analysis uses the E-OBS dataset. London experiences less extreme events because the warning thresholds there are set as 32ºC (max) and 18ºC (min), which occur less frequently. Whether this is appropriate, or perhaps an artefact of the gridded dataset used, rather than station data, might be worth considering further. Also note that Scotland and Northern Ireland are not part of the Met Office warning system and do not have defined temperature thresholds.
Thanks to Richard Betts for useful comments.