The bank for a changing world

How does climate change alter the risk of extreme weather events?

  • Camille Saurel Journalist at Usbek & Rica
  • Paris
  • 16.01.2018

Weather events are growing steadily more extreme. Natural disasters are becoming more frequent and more devastating. Hurricane Harvey, Irma and Maria are just the latest examples, and each time the same heated question is raised: was it simply nature at work or rather the result of climate change caused by Man?

In recent years, science has provided categoric answers to this question, making it possible to determine if a disaster can be attributed only to natural causes or not. Friederike Otto, physician and senior researcher of the Environmental Change Institute at Oxford University, explained this innovative scientific method on 8 December 2017 at the London office of BNP Paribas on Harewood Avenue. 

Storm Detective

The research method employed by Friederike Otto is called attribution science. In the past, weather events were attributed to climate change without any real scientific basis, relying instead on observation and data comparison. The difference with the method developed by Friederike Otto and her colleagues lies in the use of enormous quantities of data to create reliable climate models and provide statistical evidence. These complex methods are used to imagine what today's weather would be like in a world without climate change, and to compare it to the current model, where external factors such as greenhouse gas emissions caused by man have affected certain natural processes.

The next step is to prove the link between a natural disaster and climate change: to observe this link, Friederike Otto uses the metaphor of rolling a die. “Theoretically you have a 1 out of 6 chance of landing on 6 when you roll a die. But if you roll it 100 times and you find that out 70 out of 100 times, you roll a 6, you might conclude that the die is loaded. And that's basically what we do for extreme weather events,” explained Otto. As of 2004, Friederike Otto was thus able to prove that climate change had at least doubled the probability of occurrence of the deadly heat wave that struck France in 2003. Similarly, Otto and her peers recently proved that the likelihood of Hurricane Harvey occurring was closely associated with climate change. 

Every extreme weather event is ultimately unique and it's always a combination of different forces, external forcings and internal noise in the climate system.

Although she was quick to clarify the common misconceptions generally held by the public: "It would be impossible to say that an event could not have occurred without anthropogenic climate change. Every extreme weather event is ultimately unique and it's always a combination of different forces, external forcings and internal noise in the climate system,” she pointed out, “However, what we can say is how the risk of an event occurring has changed due to anthropogenic climate change and calculate the likelihood of occurrence with certainty.” 

Data comparison

This fact-based approach sets a new foundation for discussions on climate change and the resources allocated to addressing this issue. “When extreme events happen, people do ask the question ‘what's the role of climate change?’ and people will ask the question for so long until they get an answer. If there is no scientific answer, then someone with an agenda that is not a scientific agenda but a political agenda will give an answer,” she explained. The methodology of attribution science has been peer-validated, however, and can be rapidly put to work after a disaster. It only took 5 days, for example, for researchers to attribute Storm Desmond to climate change. 

Conversely, Friederike Otto and her research group can also prove when a natural disaster is not attributable to climate change, as was the case for the record-breaking flood that hit Lower Bavaria in 2013. In that instance, climate data showed no material deviation with projected rainfall. In cases like these, researchers have to look for other explanations for the flood, such as river condensation, the lack of flood zones, … a combination of human decisions primarily made for political reasons.

Planetary policy

These results could one day change the way certain major climate-related disputes are handled. Governments could be slapped with lawsuits for failing to meet their CO2 emission reduction targets. 

It is possible to assess an individual country's influence on the process of climate change, but there are political and moral choices you have to make.
It is possible to assess an individual country's influence on the process of climate change,” Otto admitted, “but there are political and moral choices you have to make: do you only calculate the emissions that have occurred in that country, or do you include the ones that have occurred when products have been made in China that have been consumed in another country, for example?” Attribution science is relatively recent: it was invented just over 10 year ago and its methodology has only been refined in the last 5 years.


In Otto's view, one of the major challenges to be met in the future will not only be to establish international protocols for attribution surveys, but also to increase the number of researchers in the field in order to have real operational services at hand. “This is something politicians are asking for,” added Otto. To this end, it is critical for cities and States to work alongside scientists in a variety of fields to prepare action plans. Climate change is already having an impact on populations, and anticipating this change could save more lives.

The project funded by the BNP Paribas Foundation, co-headed by Friederike Otto and focused specifically on South Africa, draws on several areas of expertise, calling on attribution scientists with economic or sociological backgrounds to assess the potential climate risks in the region. “Our model works very well and very quickly for certain types of events such as heavy rainfall, but for some natural disasters and for some regions, we can't do without local expertise,” confirmed Friederike Otto. From local settings to the world at large, attribution science is bringing researchers together to address the challenge of climate change.

Read more

All news
Give us your feedback Complete our survey