Will climate change lead to life threatening heat waves in the UK?
Duncan Golicher
19/06/2022
library(tidyverse)
library(dygraphs)
library(xts)
library(lubridate)
library(rnoaa)
library(mgcv)
get_ghcnd <- function(id="UK000056225"){
require(reshape2)
require(rnoaa)
climate_data<-ghcnd(id)
meas<-paste("VALUE",1:31,sep="")
climate_data<-melt(climate_data,id=c("year","month","element"),m=meas)
climate_days<-as.numeric(gsub("VALUE","",climate_data$variable))
climate_year<-as.numeric(as.character(climate_data$year))
climate_month<-as.numeric(as.character(climate_data$month))
climate_data$date<-as.Date(sprintf("%04d-%02d-%02d",climate_year,climate_month,climate_days))
d<-data.frame(date=climate_data$date,element=climate_data$element,value=climate_data$value)
d<-na.omit(d)
d
}
max_min<-function(stat="UKE00105903")
{
require(mgcv)
d<-get_ghcnd(stat)
d$year<-year(d$date)
d$month<-month(d$date)
d$week<-week(d$date)
d$yday<-yday(d$date)
d %>% filter(element=="TMAX") %>% mutate(value=value/10) ->d1
d1 %>% filter(year<2023) ->d1
mod<-gam(data=d1,value~s(yday+s(year)))
d1$max_trend<-predict(mod)
d %>% filter(element=="TMIN") %>% mutate(value=value/10) ->d2
d2 %>% filter(year<2023) ->d2
mod<-gam(data=d2,value~s(yday+s(year)))
d2$min_trend<-predict(mod)
dd1<-data.frame(date=d1$date,max_trend=d1$max_trend,max=d1$value)
dd2<-data.frame(date=d2$date,min_trend=d2$min_trend,min=d2$value)
dd1<-merge(dd1,dd2)
dd1
}
# dd1<-max_min("UKE00105903")
# jerez<-max_min("SPE00119945")
# malaysia <-max_min("MYM00048615")
# orly<-max_min("FRM00007149")
# delhi<-max_min("IN022021900")
# save(dd1,jerez,malaysia,orly,delhi,file="maxmin.rda")
load("maxmin.rda")Introduction
The current heat wave in the Uk has led to the Met office issuing an extreme heat warning and emergency responses being adopted in many places.The reporting of the event has been very dramatic and typically loaded with emotionally charged adjectives.
https://www.bbc.co.uk/news/uk-62217282
Heat waves can, and do, lead to excess mortality. This is a particular concern in countries in which they are comparatively infrequent events. A major heat wave in France in August 2003 may have led to around 16,000 excess deaths over a period of a few days. Subsequent studies suggested that these deaths were concentrated in the immobile elderly population, particularly affecting bed ridden individuals living in apartments at the top of buildings (Vandentorren et al. 2006) with no live in carers. Healthy people can suffer from heat stroke and dehydration in hot weather. However they usually recover without fatal consequences. The spike in deaths during the French heatwave occurred almost exclusively indoors. Heat related illness in younger less vulnerable age groups are not necessarily associated with specific heat “waves,” but tend to occur throughout the summer months (Abasilim and Friedman 2021). Deaths in Czech republic during heat waves have also be found to be concentrated in the elderly (Urban et al. 2017). The most fatal characteristics of heat waves go beyond the numbers reported as “record breaking temperatures,” as these temperatures are registered outside homes. Heat waves cause excess mortality through accumulated heat indoors which impact vulnerable people who are unable to take appropriate measures without help.
Data source
The data for this analysis is downloaded from NOAA’s Global historical Climatology Network.
https://www.ncei.noaa.gov/products/land-based-station/global-historical-climatology-network-daily
This provides daily records of a range of observations. The most relevant for this analysis are the daily maximums, but daily minimums are also of interest. The longest continuous series in the UK is provided for Oxford, station code UK000056225.
Climate and weather
The distinction between climate and weather is rather tricky to define consistently. According to NOAA “Weather is what conditions of the atmosphere are over a short period of time, and climate is how the atmosphere”behaves" over relatively long periods of time. When we talk about climate change, we talk about changes in long-term averages of daily weather."
https://www.nasa.gov/mission_pages/noaa-n/climate/climate_weather.html
This seems straightforward. However the weather on any given day forms a continuous time series. If temperatures are recorded continuously with high temporal resolution they show high levels of serial autocorrelation. In other words, any single measurement can be predicted by the measurements just before it. Temperature also shows spatial autocorrelation. Measurements taken close together in space also tend to be similar. So when these measurements are aggregated in order to look at the climate decisions regarding the way they are lumped together have to be made. Taking monthly means from daily records is rather arbitrary, as calendar month is not a natural division of time. Taking annual means has a more solid basis, as this coincides with a natural cycle.
If the NOAA definition is accepted, then heat waves constitute a feature of the “weather,” whilst changes in yearly means are measures of “climate.” If we are to assume that heat waves (i.e. weather events) will become more frequent as a result of climate change then its worth checking the extent to which this sort of weather is related to overall climate statistics. A simple analysis for Oxford is to take the maximum daily high over the 365 days of each year and relate this to the mean of the daily maximums over the year in which the peak occurred.
Maximum temperatures recorded at Oxford and mean daily maximum temperatures from 1900 to 2019
There is a lot of random variability shown by the scatter around the trends. However there is an underlying, more or less linear, relationship between the yearly mean and the absolute maximum for the year. This may not hold for all meteorological stations. The effects of urban heat islands may also have to be taken into account. Heat islands may increase the peaks but also decrease the diurnal temperature ranges ((Kalnay and Cai 2003). The relationship between the temperature on the hottest day of they year and the mean for that year does seem to be strong enough statistically to assume that if the mean temperature increases by two degrees (the target that has been set to be avoided through “net zero” emissions control) then the peak temperature at Oxford may also increase. This will lead to expected maximums moving from from around 36 C degrees to around 38 C. The scatter does makes this unpredictable and the recorded peaks may well be higher. The BBC have reported a new record high for the UK today (19 July 2022) of 40C at Heathrow airport. The airport itself is a notorious heat island. This extreme would be rather less likely in Oxford.
https://www.bbc.co.uk/news/live/uk-62184978
If mean temperatures do rise further then the effect of increases in mean temperature on the probability of yet higher maximum temperatures being recorded has to be taken seriously. The question is, will this rise inevitably lead to increased heat related mortality?
Daily temperature variability
The reporting of the current heat wave concentrates attention on the peak measurement made on the day. This focus on the peak ignores the duration of the observation. Each day has a predictable range of incoming solar energy. The sun rises in the morning and sets at night. Temperatures typically peak shortly after the point when the sun is highest in the sky, then they fall at night, typically reaching a low just before dawn. The diurnal range is influenced by cloudiness and atmospheric humidity (invisible water vapour) that can trap heat and lower cooling during the night (Easterling et al. 1997).
Let’s look at the diurnal range for Oxford during the last heat wave in 2019 using a dygraph. You can hover on the graph to get the numbers. Double clicking the graph will expand the time range. Trends have been fitted to each time series.
The trend lines show the mean values for the daily highs and lows. Notice that the gap between the trends widens during the summer, as would be expected when incoming solar energy to drive the temperature increase is at it greatest. In Oxford the summer gap is around 10 degrees.
This implies that the temperature rises and falls during an average summer day at a rate of around one degree per hour, although the fall tends to be slower than the rise. The absolute peak temperature does not last very long. In fact, if the direct effect of uncomfortable temperatures is considered to be the main problem during a heat wave, the total time considered to be most dangerous only increases by a few hours each year even if the mean yearly temperature increases by two degrees. However this is rather disingenuous. The dangerous effects of heat for those most vulnerable to death is the accumulated heat that builds up within poorly ventilated houses and particularly in apartments. This will be more deadly if the daily minimum temperatures increased in line with the maximums. To test whether this tend to happen we can plot out the diurnal range against the peak observed.
There is a lot of scatter again, but once more an underlying relationship. As the peak becomes higher the mean gap between peak and trough also tends to widen. Days with peak temperatures over 30 degrees tend to have lows around 15 degrees below this value. This makes sense. Heat waves occur during dry weather when atmospheric humidity is low.
So although temperatures of above 35 in the middle of the day sound scary, they will tend to be accompanied by comparatively cool nights, although as the temperature has been pushed up even these nights will still be above average. We can expect a peak of 35C to be followed by a trough of around 20C. Taking appropriate measures can prevent vulnerable individuals suffering from dehydration and dangerously high accumulated body temperatures.
What is the pattern in consistently hot countries?
Hot and dry
Many people in the UK traditionally took summer holidays in the south of Spain in order to “enjoy the heat.” The Spanish population tend to view heat as a predictable annoyance to be avoided. One of the hot spots in Spain is Jerez de la Frontera in Andalusia. Here is the dygraph for Jerez in 2019
The gap between the peak daily temperature and nightly low has widened in comparison to Oxford to around 17C. This is expected as the humidity is low. The traditional Spanish way of dealing with the peak temperatures is to build houses with high thermal mass, paint them white to reflect the sun, shutter the windows during the day and open them to allow cooler air to enter in the night. This works very well during the summer, but can lead to cold interior temperatures during the winters.
The intense pre monsoon heat in New Delhi is more uncomfortable and lasts longer than the heat in Southern Spain. Fortunately the 15 degree gap betwen the daily high and daily low also holds and the extremes become lower once the rains begin. Moister air leads to lower daily highs, although the daily lows may slightly increase, lowering the diurnal range. The peak temperature of the moister air is rarely pushed above 40C.
Tropical moist heat
The tropics present a very different challenge. There are no heat “waves” in the humid tropics. There is simply constant heat. The diurnal range rarely exceeds 10C and even the nights are uncomfortable. Here is the dygraph for a typical station in Malaysia.
It is difficult to see how residents can avoid the effects of uncomfortable levels of accumulated heat without using air conditioning. Changes in the overall yearly mean have little impact on the peak temperatures as humidity acts as a buffer against any change. The resident population of the tropics has clearly had to adapt to live under conditions of near constant heat. Fortunately the absolute highs do not exceed human body temperature (36.5 to 37 C) so do not directly threaten life providing that constant access to water is available and strenuous exercion is avoided.
Conclusion
Heat waves in the UK could become more frequent under climate change scenarios without posing an increased threat to human life. Peak thermometer records are likely to be accompanied with comparatively cooler conditions during the night if the air is dry. Rather than concentrating public attention on “record breaking highs,” public health bodies should keep an eye on humidity levels and accumulated heat. Adopting Spanish style reponses during infrequent days of severe dry heat would help to keep vulnerable people safe and comfortable. Tropical heat is a condition of countries found at low lattitudes and will not occur in the UK.
You can investigate the patterns for most ghcnd stations using this shiny app.