“The extreme events we observe today will be even stronger tomorrow.”

40°C is forecast on Saturday, June 18th in France. a “absolute record” For June, the media and scientists no longer hesitate to attribute it to global warming. Because over the past few years, every extreme and unexpected weather event has raised this question: Is global warming directly responsible? This question opens up a new field in science: the science of attribution. This aims to determine whether or not there are links between extreme weather events and climate change. Interview with Robert Vautard, CNRS Researcher and Member of Refer the weather in the world.

What is attribution?

Robert Vattard : It is a fairly new science that makes it possible to understand the impact of climate change, through concrete and tangible representations, by examining the link between the extreme events that we experience on a daily basis, human activities and the way in which this is done. Climate modification.

It is a scientific field that has developed greatly in recent years. He made a name for himself, with the conceptual and methodological frameworks being gradually put in place. Since about 2015, the number of attribution studies has become so overwhelming, because there is a thirst for understanding how extreme events are related to climate change or not.

Global warming is part of a long time scale, but it is the extreme events that focus the public’s attention. How do you determine whether it is caused by global warming or not?

The emergence of climate change will lead to the proliferation of extreme events, but also to their intensification. The peculiarity of this type of event is that it can occur without climate change, without human action. It is therefore difficult to determine their origin, because there is a great deal of randomness in the formation of these events.

To establish a relationship between an extreme climate event and global warming, British researchers, Myles Allen and Peter Stott, had an idea in 2003-2004 to place this link not at the level of the extreme climate event itself. , but the possibility. its occurrence. The problem is whether extreme weather events are likely to be caused by human activities.

It took us several years to create an effective methodology. Only in the years 2014-2015 did we think that it would be interesting to apply these methods not only in a classic search environment, but also as an operational activity, that is, for calculations and real-time results. We are now more familiar with this type of exercise.

France is witnessing today and tomorrow a severe and rare heat wave for the month of June. How are you going to study it?

For the heat wave that will arrive today and tomorrow, we have to go faster: the event hasn’t happened yet! So we must also be careful: if we make highly uncertain predictions and events don’t happen, we run the risk of telling us that studies on global warming are not reliable.

In this case, I think the forecast is reliable enough to worry about. I had initially estimated that this heat wave would be a roughly ten-year event, but in the end it is much stronger than the data obtained at the start of the week would suggest. The event is much more intense and, therefore, inevitably, will be very rare…records will likely be smashed in the South. In the Paris region, it is not quite certain yet, because very high values ​​were observed in 2011, 2017, 2019 … but also in 1947! It is the backlog that has occurred in recent years that is cause for concern.

What is the protocol for attribution, which makes it possible to determine whether or not an extreme weather event is caused by global warming?

The first step in our protocol is to collect data that is already available. Observations should be evaluated with on-site scientific colleagues who can then confirm the quality of the data. Once we prove it, we can analyze the data and identify trends.

For example, during heat waves in India, we were in contact with a Red Cross team that gave us information on the ground, and told us about disasters in certain places, or a drop in grain production. The fact that wheat production was affected means that the event occurred during the wheat growing season, i.e. between March and April. So we were able to notice that the temperatures in March and April were really out of the ordinary, and that’s over a long period, compared to the simulations available.

These observations then allow us to identify trends, but they do not allow us to give the causes of the event. To determine the causes, it is necessary to define the event, the parameters of the analysis, to have models whose input parameters we control. In most of our models, we know exactly the levels of greenhouse gases from the beginning of the 20th century until today. By comparing the levels of greenhouse gases in India, which are due to human activity, it was possible to show that there was a change in probabilities, which made it possible to establish a link with global warming.

Finally, we try to take into account the vulnerability factor. We work with teams that know the vulnerabilities of economies. An extreme event will not have the same significance, the same impact in two regions of the world with different economies. All the studies we do are published on the . website Refer the weather in the worldin the form of reports, and is submitted to scientific journals.

With the number of extreme events increasing, we have a feeling that every disaster is caused by global warming. Conversely, have you recently conducted studies that prove that an event has nothing to do with climate change?

Yes, during the droughts that Madagascar was able to experience in the south of the island last year, we were able to demonstrate that it was not really climate change that was in question, but other factors such as a very unstable economy, and above. Everything, the Covid context. This has led to the collapse of supply chains and increased prices. The population has moved resulting in increased vulnerability to drought.

Are there events that are still far from the science of attribution?

There are many topics about which we lack information: hurricanes, sandstorms … all phenomena related to storms, such as hail or lightning, for example. As for infrastructure, manufacturers often tell us: “We want to know the impact of climate change on lightning? Is there more or less?” However, we don’t really know how to determine this, because we don’t have enough notes, or they don’t go far back in time. Therefore, the models are not reliable enough to calculate the intensity of lightning.

What was the first extreme event that required the science of attribution in the end?

The event that launched the first true referral study was the 2003 heat wave. It was carried out by Peter Stott in Great Britain and resulted in the publication, “Human contribution to the 2003 European heat wave‘, published in the magazine temper nature in 2004.

Then there were some articles, especially about the 2010 rainfall in Great Britain. It was the English who launched the first attribution studies, even if on our part, especially with my colleague Pascal Yeo, we were also interested in the subject. In 2006 we had an autumn that wasn’t cold at all, with amazing temperatures compared to the average. It wasn’t particularly great. We told ourselves it was simple, but when we looked closely at the measurements, we realized the big difference. Then a study was published in 2007 on this topic. There have been studies of this kind abroad, and they have all been grouped around the specialty of attribution.

There have been many initiatives. In France, we had a project dedicated to more national events, funded by the Ministry of Environmental Transformation and Solidarity: the Extremoscope programme. He was particularly interested in the rains that caused the Seine to flood in June 2016. It was established that this event was not entirely due to global warming, but that the probability of its occurrence increased twofold.

Behind the organism are two scientists, Frederic Otto, of Imperial College, and Geert Jan van Oldenburg, of KNMI in the Netherlands. Refer the weather in the world. They got funding from the NGO central climate in the United States of America. But today we are an informal network whose activity is hardly financed. We exist because we want it to work and because our research organizations, like CNRS to me, allow us to be part of this union.

It appears that these extreme weather events are becoming more frequent. Is attribution science a scientific field destined to become more and more important?

With each case study, we learn things: about our models, the physics of these phenomena, or sometimes about the sociology of the phenomena. Knowing that there is a link between an event and global warming also gives us more confidence in our future predictions.

However, we can’t go into all the extremes and miss a lot of them, because there are more and more of them. At the moment, we are trying to cover the issue of drought in the Sahel regions at the same time as the floods in Brazil, which are disastrous. All this in parallel with heat waves in both the US and Europe… To tell the truth, we are short on staff. We will need the help of the media, and of the communicators in general, to try not only to get the population’s interest in the extreme events we are witnessing in France, but also in the major disasters that are taking place all over the world, far from us.

The study we conducted in India, for example, had a great deal of resonance in the Anglo-Saxon world, but very little in France. These are phenomena that, for us, are absolutely unimaginable! Imagine two months in a row where the temperature is over 45 degrees. We must imagine what this means in the daily lives of individuals.

Current weather events give us a glimpse into the future without realizing their severity. Everything we observe today is what we will observe tomorrow, but it is much more powerful. This gives a tangible idea of ​​the future… but it’s not too painful yet.