A study of ancient flood records shows that recent extreme floods in Europe are not historically unique.
Recent “unprecedented” floods are not exceptional when viewed in the context of the distant past, new research shows.
Led by the University of Exeter, a team of scientists analyzed geological palaeo-flood records to investigate extreme flooding events in Western Europe over several thousand years.
The study reveals that many historical floods were more severe than those seen in recent decades. The findings underscore the importance of incorporating palaeo-flood data, rather than relying solely on modern river gauge records, which typically cover only the past 100 years or less, when assessing flood risk and planning for the future.
Rethinking the Causes of Flood Extremes
The researchers challenge the idea that recent floods can be attributed solely to greenhouse gas emissions, but they warn that the combination of natural extremes and global warming could lead to truly extraordinary floods.
“In recent years, floods around the world – including in Pakistan, Spain, and Germany – have killed thousands of people and caused enormous damage,” said Professor Stephan Harrison, from the Department of Earth and Environmental Sciences at Exeter’s Penryn Campus in Cornwall.
“Such floods are seen as ‘unprecedented’ – but if you look back over the last few thousand years, that’s not the case. In fact, floods we call unprecedented may be nowhere near the most extreme that have happened in the past.”
Unlocking the Past with Palaeo-Flood Science
Palaeo-flood records use a range of evidence including floodplain sediments, dating sand grains, and past movement of boulders to identify past extremes.
Professor Harrison added: “You need that knowledge of the past if you’re going to understand the present and make predictions about the future. Coupling evidence of past extremes with the extra pressure now being added by human-caused global warming – which causes more extreme weather – you see a risk of genuinely unprecedented floods emerging.”
Implications for Infrastructure and Flood Policy
Projects such as housing and infrastructure are built to be resilient to extreme floods – based on assumptions such as a “one-in-200 year” or “one-in-400 year” flood event.
“If we rely on relatively short-term records, we can’t say what a ‘one-in-200 year’ flood is – and therefore our resilient infrastructure may not be so resilient after all,” said Professor Mark Macklin. “This has profound implications for flood planning and climate adaptation policy.”
The study examined palaeo-flood records for the Lower Rhine (Germany and Netherlands), the Upper Severn (UK), and rivers around Valencia (Spain).
In the Rhine, records for about 8,000 years show at least 12 floods that are likely to have exceeded modern peaks.
The Severn analysis shows that floods in the last 72 years of monitoring are not exceptional in the context of palaeo-flood records of the last 4,000 years.
The largest flood in the Upper Severn occurred in about 250 BCE and is estimated to have had a peak discharge 50% larger than the damaging floods in the year 2000.
Reference: “Robust climate attribution of modern floods needs palaeoflood science” by Stephan Harrison, Mark G. Macklin, Willem H. J. Toonen, Gerardo Benito and Kim M. Cohen, 26 March 2025, Climatic Change.
DOI: 10.1007/s10584-025-03904-9
Never miss a breakthrough: Join the SciTechDaily newsletter.
Follow us on Google and Google News.
4 Comments
“While recent flooding is worsened by global warming, …”
Where in the article can the evidence be found to support that assertion? Indeed, the essence of the research is that the commonly assumed meme is wrong: “The researchers challenge the idea that recent floods can be attributed solely to greenhouse gas emissions, …”
The statement, “Coupling evidence of past extremes with the extra pressure now being added by human-caused global warming – which causes more extreme weather – you see a risk of genuinely unprecedented floods emerging” is contradictory. An unsupported opinion is expressed that warming causes more extreme weather. The very core of their research is that our recent weather (at least precipitation and consequent flooding) is NOT more extreme than in the past, thus it is not becoming more extreme. They are denying their own work!
This is an example of how political correctness is undermining the Scientific Method. The basic assumption of the Scientific Method is that the researcher(s) is/are “disinterested” (unbiased) observers. Their remarks demonstrate that they are actually biased towards a belief that their own work doesn’t support.
To wit, from the abstract of their paper:
“… attribution studies are presently unable to claim that human-created greenhouse gas emissions have increased flood magnitude. We show that flood magnitude was significantly HIGHER before the 20th century, despite there being a negligible greenhouse gas contribution from humans, which means that natural variability might be significantly higher than assumed by climate modellers.”
It would appear that perhaps the person writing the press release was influenced by their personal beliefs, and/or the authors were under political pressure to soften the impact of their conclusions in order to get published. Whatever the reason(s), science is not being served when claims are not supported by the actual research.
I don’t disagree with Mr. Spencer’s interpretations and I suspect that a motivating factor in the written product is the need to appeal to financial opportunities that are tied to a certain predetermined/premeditated purpose or outcome. Some scientists now days believe their own success(es) are measured by the grants or revenue stream they generate to continue their research. I’m not stating that is the case here – just that it is an observable set of circumstances in this socio-climate. Those succumbing to financial metrics inevitably fall victim to thinking (either conscientiously or not) that they may “owe” something in return for the financial offering. If implemented properly, the scientific method typically leads one to the truth, whether the results support the original premise or not. To dilute the scientific method is to taint not only the expedition immediately in front of the scientist, but the trail it might lead to, the follow-on investigation(s) are also tainted.
My short time on this plant of 66 years is full of evidence of climate change – it’s all around us because the planet is dynamic, its processess are dynamic. My structural geology knowledge tells me there is truly a history of events we can hardly comprehend, either in magnitude, or duration or both. We read the clues, surmise the driving factors and deduce a cause-affect relationship to better understand the evidence. We mustn’t attach bias to that relationship unless we premeditatively want to ‘direct’ the outcome.
Clearly the climate has changed during our lifetimes. Evidence for it is as mundane as the changes in the planting schedules for crops and ornamentals. I’ll even go so far as to acknowledge that greenhouse gases can and do absorb IR.
However, as the article points out, there was no supporting evidence found in the study that supports increased frequency or severity of flooding or droughts, let alone evidence for a significant human influence on the weather. That is probably because the dynamic weather system is dominated by negative feedback loops as exemplified by Le Chatelier’s Principle. Thus, most changes tend to regress about the the arithmetic mean, slowing down changes and tending to oscillate over long time periods. However, the fact that climate does continually change, rather than getting stuck in a permanent state like a ‘hot house’ or ‘ice house,’ strongly suggests that the concept of a ‘Tipping Point’ is flawed.
Predictive models (General Climate Models) tend to run warm compared to actual empirical measurements, but most climatologists seem to give more credence to the models than reality. They are basically claiming that they are more clever than Mother Nature.
I appreciate your viewpoint. I too am a geologist by academic background.