Shortly after an intense heatwave struck South Asia earlier this year, World Weather Attribution (WWA) reported that climate change had made the event about three times likelier and up to 1° C warmer than it would have been in a world without climate change.
WWA is an international research group that assesses the impact of climate change on extreme weather events. At the heart of these assessments is the science of extreme event attribution, whose goal is to trace the human fingerprint in extreme weather.
However, many experts have raised questions about how well extreme weather can be pinned on climate change.
Building the evidence
WWA co-founder Friederike Otto said extreme event attribution is powerful at bringing home the reality of climate change and conveying what it actually means.
“Using a combination of computer modelling, long-term weather data, and physical understanding of climate systems, climate model simulations are run first for a counterfactual world with pre-industrial conditions, when carbon dioxide levels [in the atmosphere] were 280 ppm, and then for present-day conditions when they are over 420 ppm,” J. Srinivasan, professor emeritus at the Indian Institute of Science, Bengaluru, and a lead author of two U.N. Intergovernmental Panel on Climate Change (IPCC) reports, said.
“The difference in the magnitude or likelihood of the extreme weather event can then be attributed to human influence, including higher carbon dioxide. To ensure this has not occurred by chance, many simulations and rigorous statistical tests are required.”
Attribution science draws on two intertwined concepts: detection and attribution.
“Detection involves determining whether an event falls outside the range of its normal occurrence due to natural variations in the climate system,” Vimal Mishra, the Bikram Sarabhai Chair Professor in the Department of Civil Engineering at IIT-Gandhinagar and a specialist in climate science, hydrology, and climate extremes, said.
“Attribution involves using long-term observations or probabilistic approaches to determine the influence of human-induced climate change on not just the likelihood, but also the intensity, duration and characteristics of extreme weather events.”
Gaps in evidence
During one extreme event attribution committee hearing at the U.S. National Academies of Sciences, Engineering and Medicine (NASEM) in 2025, Stanford University climate scientist Noah Diffenbaugh said, “I think something that is worthwhile for the committee to grapple with is the broad question of whether single event attribution is even possible at all, from a scientific point of view.”
Similarly, Mark Keenan, a former U.N. and U.K. government scientist, wrote in a July article in American Thinker, “How do we know [a climate model] accurately represents a climate that no one has ever observed? The question is whether modern attribution studies can confidently separate natural variability from human influence to the extraordinary degree claimed.”
NASEM experts have previously noted that it is challenging to ascertain the influence of climate change on extreme weather. As one summary of the committee’s views put it, “It involves consideration of a host of possible natural and anthropogenic factors — such as large-scale circulation, internal modes of climate variability, anthropogenic climate change and aerosol effects — that combine to produce the specific conditions of an event.”
“Extreme weather events can occur even without anthropogenic climate change,” Dr. Srinivasan said.
In a new report published this year, NASEM also cited challenges like structural limitations and biases in models, insufficient understanding of how the dynamic drivers of extreme weather respond to climate change, and methodological hurdles in the attribution of compounding events.
Confidence in extreme event attribution findings therefore varies by event type. In IPCC assessments, for example, “confidence” relates to how strong and consistent the evidence is and how much scientists agree on it. According to the IPCC, confidence is generally higher for large-scale, longer-duration extreme weather events like heat waves and widespread heavy rainfall — and less so for smaller‑scale convective events such as thunderstorms and tornadoes.
In the Sixth Assessment Report, IPCC experts also noted that some local phenomena such as land-sea breezes and mountain wind systems can only be realistically represented by simulations at a resolution of around 10 km or less, which current models do not support. Similarly, short-duration, sub-daily extreme events — such as intense bursts of rain lasting only an hour or so — are harder to model and link to climate change than longer events like day-long heatwaves or several days of heavy rain. This is because most global climate models work with daily averages and cannot capture fast-changing processes.
According to Dr. Mishra, the analysis of localised events is further hampered by limited long-term weather records and computing power.
“For example, without long-term hourly or sub-hourly rainfall observations for a particular place, we cannot confidently mark out the range of natural variability for a heavy rainfall event there,” Dr. Mishra explained. “Uncertainties arising from observations and models can lead to unscientific claims and sweeping generalisations.”
In fact, scientists are also not agreed on what amounts to ‘extreme’.
“For instance, when our group reconstructed droughts over roughly 1,200 years, we found that some kinds of droughts seen in the past no longer occur,” Dr. Mishra said. “An event that appears extreme over a 10-20-year timeline may not look exceptional when analysed with a dataset spanning 100 or 500 years.”
Liability trail
An independent group of scientists and communicators called Climate Central has said attributing extreme events to climate change impinges on urban planning, legal and financial accountability, and environmental policy. So how an attribution’s findings are interpreted and used matters as well.
“Extreme event attribution is about assigning legal liability for emissions, either at the country level or the corporation level,” Raghu Murtugudde, a climate scientist and professor emeritus at the University of Maryland, said. “Studies claim that carbon majors can be defined and certain climate changes can be attributed to certain organisations. The real question is whether these findings can hold up in court.”
‘Carbon majors’ are businesses with particularly high emissions.
In a report this year for the Global Warming Policy Foundation, physicist Ralph Alexander wrote that attribution studies “are deeply flawed, with fundamental logical and methodological errors” because they are “being created for legal and political, not scientific, reasons.”
Dr. Murtugudde also said climate models will never be perfect enough for counterfactual worlds — the worlds that might have existed if not for some fact today, like climate change — to be accepted without question.
“The entire exercise of attribution has to be considered with some kind of global agreement on its exact purpose and whether the resources put into it are justified,” he said.
Instead, Subimal Ghosh, convener of the Interdisciplinary Programme in Climate Studies at IIT-Bombay, said preparedness is more important. “Our priority should be to develop more effective early warning systems and adaptation infrastructure in order to save more people,” he said.
“We may not gain much, scientifically or in terms of climate adaptation, mitigation and design, by holding anyone responsible for climate change.”
Harsh Kabra is an independent journalist and commentator.
