It Did Not Take Them Long To Blame Paris Floods On Climate Change
By Paul Homewood
From the Guardian:
The Paris floods, that saw extreme rainfall swell the river Seine to its highest level in decades, were made almost twice as likely because of the manmade emissions driving global warming, scientists have found.
A three-day period of heavy rain at the end of May saw tens of thousands of people evacuated across France, and the capital’s normally busy river closed to traffic because the water levels were so high under bridges. As artworks in the Louvre were moved to safety and Paris’s cobbled walkways were submerged, the French president, François Hollande, blamed the floods on climate change.
Now a preliminary analysis by a group of scientists, including the Dutch weather agency and the University of Oxford, has concluded the risk of the flooding event in Paris was almost doubled – multiplied by a factor of 1.8 – by humanity’s influence on the climate.
“Hollande was right to say climate change is playing a role but at the same time it’s important to say that this event could’ve happened without climate change,” said Prof Myles Allen, a climate scientist at the University of Oxford.
“But it means what was a 1 in 200 year event is more like a 1 in 100-something year event. Is that a big change? In terms of weather events, that’s not nearly as big an increase in risk as we’ve seen in heatwave events, where we often come up with a factor of 10. But for precipitation this is kind of what we’re seeing.”
The climate science community is speeding up its efforts to draw the links – the attribution – between extreme weather events and climate change, while such events are fresh in the public and politician’s minds. Previous quick turnaround research has shown flooding in England and heatwaves in Europe were made more likely because of global warming.
“The crucial thing is decisions… they get made in the aftermath of these events, when minds are focused on the impacts,” said Allen. “Getting this information out while people are still thinking about the event is useful. Also, it guards against the risk of over-attribution [overegging climate change’s role in an extreme weather event].”
Richard Black, director of the Energy and Climate Intelligence Unit, an environmental thinktank, said: “The ability to analyse scientifically whether man-made climate change has played a role in specific extreme weather events is advancing at a startling pace. Until very recently, scientists weren’t able to make this sort of judgement, but that’s changing fast.”
The analysis by the World Weather Attribution project published this week used a variety of approaches to look at how much the risk of the rainfall on 29-31 May had been increased by global warming, including statistical analysis of the historical temperature model, trends in climate models and “the results of thousands of simulations of possible weather with a regional climate model”.
All the approaches agreed climate change had made the heavy rains in the Seine and Loire river basins more likely. However, the link between global warming and rainfall in Germany at the same time, which saw intense thunder storms, proved inconclusive.
“The climate science community is speeding up its efforts to draw the links – the attribution – between extreme weather events and climate change, while such events are fresh in the public and politician’s minds.”
This statement sums up everything that is so wrong about this flawed study. There is a rush to blame every bad weather event on global warming for political reasons, and science suffers as a result.
Once you set out with that objective, there is bound to be confirmation bias, along with a determination to ignore contradictory evidence.
Note that the study is all based around “models”. All you need to do is tell the model that global warming leads to heavy rainfall, and Bob’s your uncle!
Meanwhile, back in the real world, we know from Kevin Marshall’s excellent analysis, using official French data, that not only was this year’s flood level lower than many 20thC events, but also that there has been a remarkable lack of significant flooding in the Seine basin in the last 50 years.
The OECD made this last point totally clear in their report “Seine Basin, Île-de-France:Resilience to Major Floods”, published two years ago. They state:
1924 and 1955 also saw major flood events in the Paris region and in the entire Seine basin. Nevertheless, the lack of a significant flood for more than 60 years tends to lessen the memory of risk.
The worst flood in recent times was that of 1910, but even that was not as bad as 1658. By contrast, this year’s flood and that of 1982 don’t come in the same league.
It is true that, in theory, the building of dams upstream in recent decades should help to mitigate flooding in Paris. Unfortunately though, it appears to be the practice to keep reservoirs close to full at this time of year, in anticipation of the dry summer season.
Much is made of the fact that this year’s flooding occurred at the end of May, whereas previous Paris floods tend to be winter events.
There is no evidence presented, or that I am aware of, to suggest that rainfall in the region is increasing at this time of year, and I cannot see how one year proves anything about the trend.
Nevertheless, it does appear that winter floods have become less common in recent decades. I wonder whether Myles Allen will publish a paper explaining how global warming has been responsible for that?
Climate Central have summarised the Myles Allen study, and include this chart:
Note how all of the models come up with big upward trend in extreme rainfall. Yet “Observations” find only an insignificant trend.
To draw conclusions on this basis, and from just one event, is plainly nonsense. But good enough for climate science.
There is one more point to make. When they talk of “increasing the probability of an extreme event like the one that occurred by about 80%”, they do not mean the black or white probability of heavy rainfall occurring or not.
What they mean is the probability of that particular amount of rain.
For instance, it is said that some places received about 4.8 inches of rain over the three days. All the models suggest is that the likelihood of 4.8 inches now is the same as, for the sake of argument, 4.5 inches in the past.
The Guardian headline, “Paris floods made almost twice as likely by climate change”, is simply not true. The models are not saying that this year’s flood may not have happened at all, merely that they may be very marginally worse than they might have been otherwise.