Defend Truth


Have floods really quadrupled since 1980?

Ivo Vegter is a columnist and the author of Extreme Environment, a book on environmental exaggeration and how it harms emerging economies. He writes on this and many other matters, from the perspective of individual liberty and free markets.

When the national science academies of the EU member states issue a report, one might think it would be hard to dispute. But any amateur can poke big holes in the report, suggesting that the purpose is alarmist climate propaganda, rather than an honest assessment of extreme weather risks.

Last week, the European Academies Science Advisory Council (EASAC), comprising representatives of the 27 national science academies of member states, presented new data in an update to a report on extreme weather first compiled in 2013.

The press release on the update claims that “new data confirms increased frequency of extreme weather events”.

Apparently, global flood frequency has quadrupled since 1980, and has doubled since 2004. Other extreme climatological and meteorological events have doubled in frequency, they claim.

The original warnings issued by the Intergovernmental Panel on Climate Change, which every few years tries to summarise the state of the climate, involve predicting climate far into the future. Importantly, its 2012 Special Report on Extreme Events concluded:

Long-term trends in economic disaster losses adjusted for wealth and population increases have not been attributed to climate change, but a role for climate change has not been excluded.”

Despite this inability to attribute extreme weather losses to climate change, it has become popular among politicians, activists, and some scientists, to point to current weather events as evidence that climate change is already causing great harm.

Of course, there is never a shortage of storms, floods, droughts and other weather events to try to fit to that narrative. Whenever a 100-year storm or a 500-year drought arrives, we do not hear that there was a similar storm 100 years ago, or a similar drought 500 years ago; instead, we hear that this is the result of modern climate change. Even fairly routine weather is roped in to sustain the narrative that we’re experiencing more frequent and more intense extreme weather events than ever before.

When there is a record 12 years between hurricane landfalls in the US, which only happens once every few centuries, they don’t attribute it to climate change, of course. Instead, they try to explain it away. Did you know that if we defined hurricanes to be weaker storms with lower wind speeds, there would be more hurricanes? The geniuses at the Washington Post do. Worrying, isn’t it? Imagine if we defined hurricanes to be any wind at all. We’d have them every day! That would sure change the climate!

Considerable research has been invested to try to place the narrative of more frequent and more severe extreme weather on a sound scientific footing. This effort has had little success to date, however. One only has to read the EASAC reports on extreme weather to appreciate just how fragile these claims are. Even an amateur can poke huge holes in these reports.

Let’s first consider the main attraction, labelled as Figure 1 in the update, and Figure 2.1 in the original report. It purports to show the frequency of geological, meteorological, hydrological and climatological events. Even geological events have increased significantly in frequency since 1980, which is puzzling, since climate change does not cause more frequent earthquakes, tsunamis or volcanic eruptions.

But the zinger is this: according to the updated data, hydrological events such as floods have increased by a factor of four since 1980.

To evaluate this claim, we need to look at a few things. Note first that the addition of only six years of data between the original report and this update has changed the increase of “climatological events” from almost 350% since 1980, to only about 220% since 1980. Clearly, the variability in the data overwhelms the trend over such a short measurement period, which suggests it is not possible to conclude anything from climate trends over just three or four decades.

The report update notes that the apparent increase in the frequency of extreme weather events could be attributable to improved reporting. It might be accounted for by better recording of smaller events, especially since the advent of the internet.

In essence, they admit that their data does not measure the same thing in the 2010s that it measured in the 1980s. The authors simply gloss over this, as if it doesn’t matter. They do not even try to evaluate the magnitude of this confounding factor. That means improved reporting could account for some, most, or all of the observed changes. It is possible that none of the changes in the data are real.

So contrary to their claim, the report’s “new data” confirms nothing, other than that the data isn’t good enough to draw any conclusions about trends in extreme weather frequency.

But it gets worse.

The report tries to find a corroborating factor for its claim by falling back on that hoary old chestnut, the trend in annual loss amounts. This, they say, won’t be much affected by better reporting of small events, since it is dominated by large events. Therefore, an increase in losses ought to confirm an increase in extreme weather frequency.

However, the loss trend is very much affected by growing population size and an increase in property values in the path of natural disasters. The EASAC report admits that there is extremely little data about the effect of development on the values of property in affected areas, which could be used to normalise the loss trend data.

In fact, they could only find “two examples” of data sets that adjust for property values, shown in the report as Figure 2. One of those, normalised North American thunderstorm-related losses, has doubled since 1980, but the other shows the European flood loss trend is “near-static”. This is completely inconsistent with a supposed quadrupling of floods.

The report explains this away by pointing to improved flood defences that supposedly nullified the impact of the great increase in flood frequency. In support of this claim, it appears to cite an academic source.

In fact, the citation refers to a report by the re-insurance giant Munich Re, from which it blatantly copied both charts, without attribution. The Munich Re report does say Europe improved its flood defences after particularly damaging floods in 2002, which no doubt contributed to the fairly flat loss trend.

However, the EASAC report includes no flood loss data from anywhere else in the world that could confirm the claim of quadrupled flood frequency. It is also highly revealing that Munich Re couldn’t find any other weather-related loss charts that showed an upward trend.

Indeed, the EASAC authors neglected to plagiarise from the Munich Re report a set of charts that show that although the number of loss events is rising, because of growing populations and property development, annual global losses, once normalised for these variables, are surprisingly flat. There is almost no perceptible increase in normalised losses between 1980 and 2015. This totally contradicts the EASAC report’s conclusions that extreme weather is on the rise.

On the normalised losses due to thunderstorms in North America, the EASAC report says:

There are meteorological reasons for the increase in the normalised losses from severe thunderstorms.”

The Munich Re report simply says that the increase is consistent with meteorological observations about increasing intensity of thunderstorms.

This goes back to previous claims made by Munich Re, that “climate change is resulting in intensifying storms in the United States”.

The paper which was intended to prove this was Sander et al, published in the journal of the American Meteorological Society in 2013. That paper found increased variability in thunderstorm-related losses, but concludes:

No final attribution of the climatic variability identified in thunderstorm forcing and losses – either to natural climate variability or to anthropogenic climate change – can be conclusively arrived at in this study.”

Munich Re was roundly criticised at the time for making claims about climate change that contradicted the paper that supposedly proved it, and that paper, in turn, was criticised for drawing general, continent-wide conclusions from a trend change that was visible only in one small area (“tornado alley”), as well as for positing a causal mechanism for an increase in thunderstorm severity in “tornado alley”, when that cause actually did not occur.

So the only two cases in which EASAC has the feeling that there might have been an increase in weather-related losses – thunderstorms in North America and floods in Europe – turn out to be badly supported by a report from an insurance company. Why the combined science academies of 27 countries couldn’t find anything better to base their claims upon than a report by an insurance company is beyond me.

That same insurance company report shows that globally, there is no significant increase in weather-related insurance losses, which completely undermines the EASAC report’s conclusions.

Towards the end of the update, the authors do something quite remarkable. They say there are so many studies where no influence of climate change on extreme weather is found, that it is impossible to list them all. But it is possible to list a few that do find climate change increases the probability of some extreme weather events.

This is the very definition of cherry-picking. To give credence to results that confirm a position, while ignoring results that might contradict it, ought to be beneath the assembled national science academies of Europe. It’s a first-year mistake. If qualified scientists still do this, they are deliberately introducing bias into their conclusions. They are trying to deceive. They are writing propaganda.

And that is all this report is. It has a political purpose, to cause sufficient popular alarm to “urge further action” to increase public spending on “the adaptability of Europe’s infrastructure and social systems to a changing climate”.

It is based on dubious evidence, and might be entirely false. It shouldn’t take an amateur to tell the national science academies of Europe not to exaggerate their conclusions to create alarmist propaganda. It certainly shouldn’t take an amateur to tell them to base those conclusions on actual science in the first place.

So, to answer the question in the headline: no, there is no evidence that global flood frequency has increased fourfold since 1980. That claim is exaggerated at best, and possibly entirely false. Clearly, we cannot trust the national science academies of EU member countries to give accurate, measured and politically neutral advice on climate science. DM


Please peer review 3 community comments before your comment can be posted