The importance of including variability in climate change projections used for adaptation
July 2015 - A new study has adapted the methods used to produce the UK national climate projections (UKCP09) in order to include year-to-year variability. The updated approach is demonstrated for the England/Wales region and shows how the likelihood of extreme seasons that impact society evolves over the twenty-first century.
UKCP09 was an assessment of the likelihood of possible future UK long-term averages under three socio-economic scenarios. These projections were based on large ensembles of climate simulations that explored various sources of modelling uncertainty, and were constrained by historical mean climate and large-scale temperature trends. The headline message based on these projections is for "hotter, drier summers and milder, wetter winters" over the UK.
Due to the 30-year averaging used in UKCP09, this statement refers to typical summer and winter temperature and precipitation only. Occasionally though, the UK experiences extreme seasonal weather, such as the very cold winter of 2009-10 or recent wet summers. Such seasons that apparently contradict the headline message can cause confusion, even leading some people to claim that they disprove climate change. To remove any confusion, a new analysis by the Met Office Hadley Centre, published recently in Nature Climate Change, has added simulated year-to-year variability to projections for England/Wales under a medium emissions scenario. The projections now provide the range of seasonal temperature and precipitation that England/Wales might experience under climate change. They can be expressed in terms of the extreme hot, cold, wet or dry seasons which people can more easily associate with the heatwaves, cold spells, floods, and droughts that impact society.
The results of this analysis for the England/Wales region are consistent with UKCP09, but provide information with an added level of detail, to make the projections more relevant to people. We find for instance that cold winters do not disappear under climate change, and there is still a 20% chance for 2020 of a winter colder than the 1961-90 average. There is a gradual reduction though in the chance of cold winters under climate change, and by 2100 the chances of such a winter drop to 4% (as illustrated by a few plausible time series in the figure).
For summer rainfall, some realizations have strong drying signals (red), whereas others have a lot of very dry summers but can still produce a few very wet summers (blue). Overall, for the next 20 years there is still a 35-40% chance of getting a wetter than average summer. There is an increasing chance for drier summers under climate change, but the chance for a "very wet" summer (20% above the 1961-90 average) reduces only slowly from 18% to 10% over the 21st century.
Fair comparison of projections with seasonal averages
In 2009-10 England/Wales experienced its third coldest winter since 1950. According to the new analysis, the chance of experiencing such a cold winter temperature in 2009-10 was about 6%. This is within the projected range for that year, albeit at the cold end, so is not inconsistent with climate change and certainly cannot be used to disprove it. More extreme seasons have occurred over the historical period. Two cases include the 1962-63 winter, and the recent very wet winter of 2013-14. These examples, which show the potential for rare atmospheric circulation patterns to cause very extreme seasonal weather, are outside the projections. Individual seasons like these are not, on their own, enough to invalidate the projections. In contrast, if several of these events happened over a period of a few years, this would be a cause for concern. In this way, the new analysis offers a way to make fair comparisons with seasonal averages as they arise, and to monitor the credibility of the climate simulations that underpin the projections.
Revised headline messages
The importance of year-to-year variability is not new to climate science. What this research shows is that if climate scientists want to make their projections of future climate more useful to people, it is important to factor in year-to-year variability on top of the climate change signal. In this way, the projections can be presented in terms of seasonal extremes that people can relate to and in principle could be applied to other regions.
For stakeholders of UKCP09, this research suggests new headline messages are possible. For example, for winter temperatures this could be "Over England and Wales, we expect an increasing chance of warmer winters, with fewer colder ones". For summer precipitation it might be "Over England and Wales, we expect an increasing chance of dry summers, but only a modest reduction in the chance of very wet summers".
These new messages paint a picture of the England/Wales climate under climate change but factor in the consequences of year-to-year variability. While there is a trend towards warmer winters and drier summers, there will still be a lot of variations in weather from year to year. Cold winters and wet summers just become less likely, and we will still have to be prepared for them.
The full version of the research paper is available on the Nature Climate Change website.