In previous post I ended rather abruptly by saying that one of the problems with the vagueness of the definition of a heat wave, is that they are difficult to compare. What I meant by that is that those of the public, who don’t catch the nuances of those different definitions, think that is being talked about exactly the same thing, while they are in fact distinctly different.
For example, when the weather (wo)men in our region talk about a heat wave, they mean a period of five consecutive days with temperatures over 25 °C of which three days with temperatures over 30 °C at Uccle (Belgium) or De Bilt (the Netherlands). We have no real problem with that because in our country we are used to 20 °C, 25 °C temperatures or even higher. But we don’t cope well with 30+ °C temperatures, especially when they last more than a couple days. These are rare occurrences in our region and we are not used to them.
Bottom line: when our weather services declare a heat wave, it is hot and we need to take measures to cope with such (rare) heat. That is the significance of the definition: it captures the extremes of a region. Extremes that could have an impact those who are weak like small children, old people or those who are ill. In our current societal structure and in our climate, we just aren’t accustomed to these temperatures.
Now enter that new definition of a heat wave: the average temperature of three consecutive days. There is no regional threshold anymore, because the investigators wanted to compare between the different countries with different local definitions. Now we see an upwards going graph:
When I followed the link in the tweet of Jill Peeters I landed on a Climate Central page: Climate Change Ups Chances of Europe Heat. The study was a cooperation of an “international group of climate scientists”, also from the KNMI. Indeed, there was a page on the KNMI website also: More frequent heat waves due to climate change (Dutch). None of the two made the effort to provide a link to that study.
Climate Central said that the study was build on peer reviewed methods. Looks like a nice way of saying it is not peer reviewed itself, yet still be able to include the magical words.
Let’s compare the official De Bilt heat wave data with that new definition. With the new definition we see an ever increasing graph and it then is easily concluded that heat waves are steadily rising since the beginning of the measurements. The first clue comes however when noticing that the graph doesn’t even use the word “heat wave”, but the description is “Summer heat”. Yet the word “heat wave” is used extensively in the text and by the people/sites/blogs that mentioned the study.
This is how the graph looks with the official KNMI heat wave data:
Quite a different picture. The reason is of course that both are different things altogether. The first thing we notice is that there is no gradually increase. There are two distinct periods in which heat waves occurred: from 1911 until 1950 and from 1975 until now. Before that first cycle, there is a ten year period in with no heat waves coming above the threshold. Between the two there is a gap of about 25 years. This is also touched in a previous post a couple years ago: Climate begins at the 1950s, which describes why there is the perception that the number of heat waves is going up.
So if we look at the observed data, there is no real problem with increasing heat waves, at least not in numbers of heat waves over time. If anything can be concluded, it is that there were more heat waves in the period from 1911 until 1950. Less periods of higher temperatures are coming above the threshold in our current cycle. The number of heat waves in the current group apparently does not give much cause for alarm.
Then let’s look at the duration, the number of days of those heat waves:
The same picture here. The number days of heat waves in the first group was higher than this of the current group. If there is anything that we can conclude is that the number of heat wave days per year decreased, not increased.
Again our current cycle pales in comparison with the period of 1911 until 1950. Not much reasons for alarm here.
But isn’t global warming causing temperatures increasing more earlier in the year?
When we look at the data, again we see heat waves occurring earlier in the first cycle. More, in that cycle we see the earliest more later in the year. The current cycle of heat waves has a more narrow band. Nothing unusual here. The heat wave of last week was about in the middle of the historical data.
Finally, what about temperatures? Are we seeing heat waves with higher temperatures nowadays?
That is not what the data shows. Temperature in the first or the second group have about the same range. And again, the current heat wave had a maximum temperature in the middle of that range.
If we compare the averaged temperature data and call that a heat wave, it will be confusing for the public. The public could wrongly assume they are talking about really hot days for which measures are needed, while they are actually talking about higher occurrences of warmer weather, something most of us don’t necessarily find a problem.
Any way we slice it, from the official KNMI heat wave data, the heat waves from the past seem to be more extreme than the current ones. This new definition confuses two things: warmer temperatures (on the rise) and extreme temperatures which are potentiality dangerous (and not on the rise). Yet they are called by the same name and the public, when no background is given, could get confused into believing that extreme temperatures are on the rise. Which, according to the observations, would be an incorrect conclusion.
Trust, yet verify 