Extreme climate event attribution

In the early 2000s, a new field of climate science research emerged that began to explore the human fingerprint on extreme weather, such as floods, heatwaves, droughts and storms. Known as “extreme event attribution”, the field has gained momentum, not only in the science world, but also in the media and public imagination because of the power it has to link the seemingly abstract concept of climate change with our own tangible experiences of the weather.

The Carbon Brief

Climate attribution seems to have crept up on me and taken me completely off guard this year. I can’t help but see them as a group of scientist vigilantes who have set themselves up as the Interpol of the AGW debate, cataloguing any particular study or climate research either for, or against the great cause of AGW. They’ve certainly very vociferous and seem to have captured the ear of newspapers such as the Guardian, something that I would have always like to do, but only in my dreams.

I’ve already had an email exchange with Geert Jan van Oldenborgh, one of the leading lights at World Weather Attribution in March concerning the warm end to February in the CET series. He seemed approachable enough, but rejected my idea that December 2015 had been far more extreme a month in central England than February 2019. Fair enough, I am no climatological statistician, but I can knock out a good linear trend or moving average like the next man, and I think I can tell just by looking at two temperature graphs which is the more extreme.

Anyway putting that to one side, what caught my eye this morning and prompted this post was an article in the Guardian with the following explicit headline:

I downloaded the report “Human contribution to the record-breaking June 2019 heat wave in France” which prompted this article, and I can see why it’s still in peer review because apart from the ‘science’ or should that be the ‘statistics’ the English translation doesn’t read particularly well. The first difficulty of course when investigating the frequency of heatwaves is an exact definition. For heatwaves no one country or meteorological service can agree on one :

  • Should they last a minimum of five days or just three.
  • Should they be based on the maximum temperature or the daily mean?
  • And what temperature threshold are you going to set to consider any candidate day?

They set up their own new definition of a heatwave for the purposes of their study :

…we define the event over a period of three days — as the highest 3-day averaged daily mean temperature in June (TG3x). The daily temperature is taken as the average over France and at the city of Toulouse in southern France. 

World Weather Attribution

Setting their own criteria simplifies things, but I still can’t understand how it helps in the conclusion that the Guardian story leads with:

The record-breaking heatwave that struck France and other European nations in June was made at least five – and possibly 100 – times more likely by climate change

Damian Carrington – Environment Editor of the Guardian.

For the life of me I still can’t begin to understand paragraphs like this one in the report :

To evaluate these models we test whether the statistics of extreme heat in these models corresponds to the observed statistics. The test consists of fitting the models to the same GEV distribution as in the observations and comparing the scale (σ) and shape (ξ) parameters of the fits to the model data with the parameters of the fits described in the section observational analysis. We do not consider the position parameter as biases in this parameter can easily corrected without affecting the overall results.

“Human contribution to the record-breaking June 2019 heat wave in France”

At this point you may think that I’m not of the opinion that heatwaves are on the increase, but you would be wrong, because I do. What I don’t like is the overly complicated methodology used in reports such as this to prove, using some mysterious statistical process, that heatwaves are from five to one hundred times more likely to occur than in the past. I wrote an article the other day about heatwaves in central England, for which I analysed their occurrence since 1878 using the strict UKMO heatwave definition, and found that the number of “heatwave days” has increased from 2.5 days in 1878 to 5.6 days in 2019. If you read the article I think that even a layman would more easily understand the term “heatwave days” that I use to quantify the increase, rather than being told they are five to a hundred times more likely to occur, which I simply don’t believe.

You may also like...