When someone would ask “Are heatwaves increasing?”, what would you answer? Probably something like “That’s common knowledge, we all know that!”. Sure, but why? “We hear it everywhere”.
I heard it many times before and had no doubt it was true. At first glance it seems straight forward and logical. When temperatures go up (global warming, you know), the frequency of extreme temperatures go up. It has been told that the frequency of heatwaves is unprecedented and even ramping up in the last years.
I hear around me that we now live in a more extreme world. In stead of just believing this story line, I wanted to see for myself. I came to realize that this data exist. We call extremes in temperatures “heatwaves” and “coldwaves”. Okay, heatwaves increase, but the problem started when I asked myself: “How much?”. The answer was not exactly what I expected it to be…
First things first: looking for the dataset. The Belgian dataset seemed to be very hard to find and was really confusing. I found bits and pieces of it and nothing seem to match. I found that their definition of “heatwave” changed over time. That could well be the reason for these bits and pieces. The Belgian temperature dataset was not freely available either, so reconstructing it would not be possible. This wasn’t going to be easy.
But our neighbors in the North kept their records as well and, more important, they share their data with the world. It was very easy to find the list of heatwaves in De Bilt from 1900. It is not exactly what I searching for, but the situation from the Netherlands should be relatively similar. De Bilt is only a couple 100 km from Brussels. With the same definition, there will be probably somewhat more heatwaves recorded in Brussels (it is more to the South), but considering that in the average in the GISS dataset are made with stations until 1,200 km from each other, this should not be a problem.
This is what the Dutch call a heatwave (it is the current Belgian definition also):
A heatwave is a succession of minimum five summer days (maximum temperature of 25.0 °C or higher) from which minimum three tropical days (maximum temperature of 30.0 °C or higher).
I thought the heatwave data would represent extremes well. It is the threshold above which temperatures are recorded, so only the extremes are visible. The more data comes above this threshold, the more extreme the temperatures.
So I fired up Calc and loaded the KNMI data in a spreadsheet. Then I plotted it as a graph. The result was surprising. I expected, well, a steadily increase. Maybe not as steep as people suppose it is, but yet increasing. But that is not really what I saw.
Heatwaves De Bilt 1900 until 2013
It looks more like a cyclical event in which temperatures didn’t get above the threshold before 1911 and also between 1951 and 1975 (smack in the middle of the period of the rise and fall of the ice age scare).
Also, our well feared heatwaves in the 1990s and 2000s seem to be very bleak against the 1940s. I knew that the 1930s-1940s were warmer globally, but I didn’t expect to see it that clearly in the data closer to home.
The heatwave data consists of summer days (maximum temperature ≥ 25.0) and tropical days (≥ 30.0 °C or higher). The ratio between the two are about the same. One would expect with “global warming” that the tropical days increase against the summer days. It isn’t.
But, but, why do those people say that there are more heatwaves than before? Do they use other data? No, just look at just a very small selection of quotes about more heatwaves than before.
One from Belgium (translated from Dutch):
[…] Between 1981 and 2010, 10 percent of the world’s surface had to deal with extreme heat waves. That is 50 to 100 times more than the 0.1 to 0.2 percent during the period from 1951 to 1980. […]
One from The Netherlands (translated from Dutch):
[…] There were more heat waves in the last thirty years than in the fifties, sixties and seventies of the last century. […]
Not only in our little country, but also from the States:
[…] A new study examining six decades of global temperature data concludes that a sharp increase in the frequency of extremely hot summers can only be the result of human-caused global warming. […]
See a pattern emerging? They all talk about the last 60 years, the period from the 1950s, etcetera. They seem to forget there was life before 1950. If you take this period, where does that takes us in the graph? Let’s look at what they compare with:
Heatwaves De Bilt 1951 until 2013
This makes some things clear. The different time frame gives the (false) impression that:
- this warming is unprecedented (it isn’t, there was a period with similar or even more heatwaves in the past, but when one only looks from the 1950s this could not be seen)
- the intensity increased in the late 1990s till the half of 2000s (it did, but this happened before).
- there were no heatwaves at the beginning of this period and then from the 1970s heatwaves start popping up (that’s true, but there was a period with a similar, even hotter temperatures, before the 1950s)
- heatwaves are ever increasing since the 1970s (if you don’t look at the period before 1950 it is, otherwise it isn’t)
- in 2006 the intensity increased from 1 to 2, the speed is increasing! (it was increasing, certainly when one saw this in 2006-2007, but there were identical or even bigger increases in the past).
This seems not to be much different than many things in climate communication: there is a core of truth in it, but it lacks balance. Although it is an excellent opportunity for those who want to dig into it (I learned a thing or two about climate/weather by figuring things out after being surprised by some claims), for the public this is misleading. If one only look at the data from the 1950s on, there is a reason for alarm. But that scare has more to do with a lack of perspective.