Shanshan study spotlights science linking warming to extreme weather

On Aug. 29, when a typhoon feared to morph into “the biggest storm ever” made landfall and battered wide areas of Japan with wind gusts, torrential rain and high waves, a burning question was on the minds of many: Is this because of climate change?
An answer came surprisingly quickly.
A day later, while Typhoon Shanshan was still crawling through Japan having turned east in the Kyushu region, a team of scientists at Imperial College London declared that climate change had “supercharged” the typhoon.
Not only that, they pinpointed the role of climate change in numbers: Shanshan’s devastating winds were made 26% more likely and 7.5% more intense because of warming.
Such a rapid assessment from scientists on the influence of climate change on a particular weather event was unthinkable a little more than a decade ago.
Friederike Otto, a senior lecturer at Imperial College London’s Grantham Institute — Climate Change and the Environment who was not directly involved in the Typhoon Shanshan study by her colleagues, is a key person behind this drastic change in the climate science community.
In 2014, Otto co-founded World Weather Attribution (WWA), comprised of an international team of climate scientists, to expedite attempts to study the link between specific extreme weather events, be it heat waves, landslides, tropical storms or wildfires, and human-induced climate change — a field known as “event attribution.”

While the history of attribution science itself dates back to the 1990s, it used to take months or years for scientists to examine and publish the results of their analysis. WWA releases them days or weeks after the event — while the impacts of extreme weather are still fresh in the minds of the public and policymakers — so it can better inform discussions about climate change, the group says.
“One of the big problems with climate change and particularly with acting on climate change was that … while (the planet’s warming) is very clear and very straightforward, it’s not linked to people’s experiences, so it’s not something that they feel is important for their own life,” Otto says.
“The reason why we thought it’s really important to do these studies quickly and to develop the methodology that allows us to do that quickly is to be able to connect people’s experience with this relatively abstract science of climate change.”
But how does it work?
Yukiko Imada, associate professor at the University of Tokyo’s Atmosphere and Ocean Research Institute and Japan’s foremost event attribution scientist, explains a common method by comparing it to the study of the connection between smoking and lung cancer.
Imagine you have one heavy smoker who has developed lung cancer. It is not possible to establish whether heavy smoking led to lung cancer by looking at that person only, because for that person there may have been other factors — such as genetic or other lifestyle issues — that contributed to the disease.
“So you gather 100 heavy smokers and study how many of them develop lung cancer,” Imada says. “You also gather 100 others who do not smoke and count how many of them develop it. Of course, some of the nonsmokers would develop lung cancer, but if you compare the two groups and see that overwhelmingly more people from the smoker group get lung cancer, you can prove that tobacco smoking affects the chance of getting it.”

The same method is used in weather attribution, though in the case of extreme weather, researchers gather lots of Earth models instead of people.
In reality, of course, there’s only one Earth, but attribution scientists create many Earth models on computers by inputting weather data, including ocean conditions and the amount of carbon dioxide and other greenhouse gases in the atmosphere.
“What happens in reality is a coincidental event out of many possibilities, but through the simulations of 100 Earth models, we can create many other coincidental events,” she says. “At the same time, we also create 100 other Earth models without warming factors such as carbon dioxide and aerosols (including sulfur dioxide and black carbon), while reflecting all the conditions that existed before the Industrial Revolution.”
When you compare the 200 Earth models with and without warming effects, a pattern emerges — with the models incorporating human-induced climate change much likelier to have extremely hot days and more days with heavy rainfall, she says.
(Other teams may use a different number of models or another method altogether.)
Indeed, rather than talking about events in terms of being “caused” by climate change, scientists instead speak of warming’s effect on likelihood and intensity, given the multiple factors at play.
Imada credits advances in supercomputing and attribution methods over the last decade for the speedier analysis seen today.
Imada was on a panel of researchers who recently announced that Japan’s record heat in July was “almost impossible” without global warming. The science ministry and the Meteorological Research Institute, the research arm of the Meteorological Agency, commissioned the study.
In that analysis, the scientists used what they call a “predictive probabilistic” method. To shorten the time needed for simulations, the researchers used data for ocean surface temperatures and the amount of sea ice in polar regions from three-month weather forecasts released by the Meteorological Agency without waiting to receive actual observation data from the agency, she says.
But are the results of such expedited analysis reliable? And how can those not involved in the research vet them? After all, while the methods of weather attribution have been published in academic journals following the normal peer-review process, results following specific extreme weather events are not when they are first made public.
“Obviously, it can’t be peer-reviewed when we first publish it, because the peer-review process doesn’t happen in a day,” Otto says. “All our data is publicly available, and we have also done some training for journalists to see how to access some data and to apply some of the methods.”

WWA states on its website that its scientists eventually publish their attribution studies in scientific journals, where the results are largely unchanged from what they first make public. Otto also denies that the speed of the analysis sacrifices quality.
“We have sometimes delayed studies because we had problems with the data or we were not sure,” she says. “There’s always more that you can do. So we definitely limit what kind of analysis we do in a single study, but you also do that when you do a peer-review study.”
Still, Otto acknowledges that when she started rapid analysis a decade ago, the team caused a stir in the scientific community by releasing their findings ahead of peer review. Many people were “not happy” about it, she recalls.
“The scientific community is quite conservative, especially the climate science community,” she says, noting that doing attribution studies rapidly “is a very political act” even though the content of the research is not political.
“We felt that it is irresponsible of scientists to not talk about these things, where we can actually answer questions that come up. And by not speaking up as a scientist, we left the discussion really up to people who only have political opinions.”
Criticism of Otto’s initiative has receded over time.
“We have no problem at all now,” she says. “I think the scientific community has really changed a lot in the last 10 years. Especially with emissions continuing to rise with fossil fuel burning … and therefore the impact being so much more visible, I think most scientists now would say, ‘Yeah, we can’t just quietly sit there and write peer-reviewed papers. We have to make sure they are actually read and part of the public discourse.’”
Imada also says event attribution studies have gained international recognition, with the Sixth Assessment Report of the United Nations Intergovernmental Panel on Climate Change prominently featuring attribution studies as one of its highlights. Otto was one of the lead authors of a working group that contributed to the physical evidence aspect of that report, published in 2021, as well as one of the lead authors of the final report, released in 2023.
WWA has conducted over 400 studies so far, and based on that research, the group can now confidently say that every heat wave in the world is made stronger and more likely to happen because of climate change.
Other extreme events are not as clear-cut, as they are affected by local weather conditions and human factors. For example, the highest confidence that floods caused by extreme rainfall are more common and more intense because of climate change is in northern Europe and central North America, but varying degrees of uncertainty remain in the other areas. The impact of extreme rain also depends on land usage, the quality of water management and flood defenses, the group says.

Going forward, Otto points to the scarcity of observational data and weather and climate models in the developing world as one of the biggest challenges for attribution scientists.
“We understand the very basics there, like more greenhouse gases and more heat waves. But how that plays out locally in droughts and how that, for example, interacts with natural variability systems like ENSO is something that, in some parts of the world, we don’t really have good tools to answer.”
ENSO, short for El Nino-Southern Oscillation, is a recurrent climate pattern that causes changes in the temperatures of waters in the central and eastern tropical Pacific Ocean over a period of three to seven years. The phenomenon played a role in the extreme heat seen around the world this year and last.
Imada, meanwhile, cites the need for more robust international collaboration among scientists to further develop the field. She is also eager to take on studies that highlight what weather would be like on an even warmer planet.
“Event attribution itself is a message about how global warming from past to present has contributed (to extreme weather events), but I would also like to show how horrible the situation would become if temperatures increase by 1.5 degrees or 2 degrees,” she says.
Would she even paint the grim picture of what a global temperature rise of 3 degrees Celsius or more would look like?
“Yes,” she says. “That’s part of our plan.”