The method used to conduct an attribution study consists of eight steps, described here. The first step is the selection of an extreme event to study.
After selecting an extreme weather event to study, the first step is to define the event, which provides a framework for the study.
Researchers determine the geographical boundaries of the most impacted area, the best index to quantify the meteorological extreme (eg. maximum temperature, average rainfall, etc), and the duration of the event.
For example, for a study on a heatwave that impacted South Western Europe and Northern Africa, World Weather Attribution researchers focused on the daily maximum land surface temperature averaged over a three-day period from 26-28 April 2023, in a region that included Spain, Portugal, Morocco and Algeria.
By analysing observed real-word weather data, collected from weather stations in the study region, researchers can find out how unlikely the extreme weather event is (the return period) and identify any trends in the intensity of similar events.
Further analysis can then determine how frequently weather events of a similar intensity would have been expected to occur before humans started burning fossil fuels in the late 1800s, which means the researchers can calculate the difference in the probability of an event occurring in the late 1800s, compared to the probability of an event occurring today.
The second part of the study uses computer modelling to try to isolate the role of human-induced climate change in the event.
Researchers use modelling that simulates the likelihood and intensity of the weather event in two ‘worlds’: today’s world with 1.2ºC of human induced global warming, and a hypothetical world without human-caused climate change. Researchers then compare the results to find out if climate change made the event more or less likely and intense.
A wide range of quality checks are completed throughout the study, including checking the models to ensure they adequately depict the weather in the area being studied, which includes key seasonal and spatial patterns.
Finally, the results from the real-world data and models results are combined to provide an overarching statement about how climate change influenced the intensity and likelihood of the weather event.
This video breaks down each step of our recent study on the late April heatwave that impacted Spain, Portugal, Morocco, and Algeria, highlighting how much hotter and more likely climate change made the event.
While the April 2023 heatwave impacted large areas of Europe and Northern Africa, the geographic boundary of the WWA study was determined by the regions experiencing the most extreme heat, which included parts of Spain, Portugal, Morocco and Algeria.
Occasionally, there are methodological challenges which means that quantitative results can not be provided in the findings of a WWA study. These challenges include the availability of reliable observations of the weather and the suitability of models to replicate the event being studied.
If a study does not have a conclusive result because of these challenges, that does not necessarily mean that climate change played no role in the weather event.
For more further information, see Pathways and pitfalls in extreme event attribution and A protocol for probabilistic extreme event attribution analyses.