Historical Trends and Variability in Heat Waves in the United Kingdom

Abstract

Increases in numbers and lengths of heat waves have previously been identified in global temperature records, including locations within Europe. However, studies of changes in UK heat wave characteristics are limited. Historic daily maximum temperatures from 29 weather stations with records exceeding 85 years in length across the country were examined. Heat waves were defined as periods with unusually high temperatures for each station, even if the temperatures would not be considered warm in an absolute sense. Positive trends in numbers and lengths of heat waves were identified at some stations. However, for some stations in the south east of England, lengths of very long heat waves (over 10 days) had declined since the 1970s, whereas the lengths of shorter heat waves had increased slightly. Considerable multidecadal variability in heat wave numbers and lengths was apparent at all stations. Logistic regression, using a subset of eight stations with records beginning in the nineteenth century, suggested an association between the Atlantic Multidecadal Oscillation and the variability in heat wave numbers and lengths, with the summertime North Atlantic Oscillation playing a smaller role. The results were robust against different temperature thresholds.

A new heatwave definition for the UK

“…Stott et al ., (2004) demonstrate that it is very likely that human influence on the climate has doubled the likelihood of a European heatwave exceeding that of 2003. The UK historical climate observations have shown strong variability in the occurrence of heatwaves over time, which are partly associated with large‐scale modes of climate variability such as the Atlantic Multidecadal Oscillation (AMO, Sanderson et al ., ). Climate change trends will therefore be superimposed on this natural variability in our climate.…”

Summer temperature extremes in Europe: how does the definition affect the results?

“…The annual 90th percentile, that is calculated using TX values from an entire calendar year, was not included in the analysis, as it was close to the median of TX for summer, and consequently reflected average, instead of extremely warm, temperature conditions (Zhang et al 2011). Higher annual percentiles were successfully used to identify hot days by Tomczyk and Bednorz (2016) and Sanderson et al (2017), among others.…”

“…Researchers use various baseline periods to determine thresholds, above or below which air temperature is considered extreme. Baseline periods that are most frequently used are as follows: 1961-1990(e.g., Alexander et al 2006Rusticucci et al 2016;Hoy et al 2017;Wypych et al 2017), 1971-2000(e.g., Spinoni et al 2015Sanderson et al 2017), and 1981-2010(e.g., Russo et al 2015. Other, less standard periods are also used, and they often comprise the entire considered time period (e.g., Stefanon et al 2012;Lavaysse et al 2018;Tomczyk et al 2018).…”

“…However, when one takes a closer look, it comes to light that the percentiles were calculated for different timeframes within the year. And so, for example, Tomczyk and Bednorz (2016) and Sanderson et al (2017) calculated a percentile for an entire calendar year, Lhotka and Kyselý (2015a) determined a percentile for the summer season (JJA), and Revich and Shaposhnikov (2008) did it separately for each summer month. Yet, another approach is to determine a percentile for each calendar day separately using an x-day-window-centered method.…”

Circulation analogues and uncertainty in the time-evolution of extreme event probabilities: evidence from the 1947 Central European heatwave

“…Meanwhile, for every 20-year period since 1976-1995, there are substantive increases in the likelihood of witnessing such circulation analogues, with probability ratios reaching between three and six for in the most recent decades. While these results appear to qualitatively resemble corresponding changes in the Atlantic Meridional Oscillation over the same period (Sanderson et al 2017), decomposing the relative effects of low-frequency modes of variability versus changes to European aerosol or greenhouse gas emissions (King et al 2016;Undorf et al 2018) is beyond the scope of this analysis.…”