Extreme event attribution, also called attribution science, looks at how much each cause contributes to an event and gives a level of confidence based on statistics. This science is most often used to study how climate change caused by humans affects the frequency, strength, length, and effects of extreme weather events. The goal of attribution science is to find out how much these events are linked to human-caused global warming, rather than being caused by natural changes in the climate or typical weather patterns.

History

Before statistical methods for attribution science were developed, scientists believed that human activities could explain overall climate trends, but they did not think it was possible to link specific weather events to human influence. The Intergovernmental Panel on Climate Change (IPCC) first discussed attribution in its 1990 First Assessment Report, stating that it was "almost impossible to prove a cause-and-effect relationship with high confidence by studying a single factor." However, the IPCC noted that confidence in linking human activities to climate changes increased when predictions matched real-world observations. It described the "fingerprint method," which looks for signs unique to human-caused global warming. The IPCC's 2001 Synthesis Report reported that studies had found evidence of human influence in climate records from the past 35 to 50 years. It also stated that most studies showed the warming caused by rising greenhouse gas levels over the past 50 years was similar to or greater than the warming observed in the real world.

Fraction attributable risk (FAR) was first used in a 2004 study of the 2003 European heatwave, showing a direct connection between human-caused climate change and a specific extreme weather event. In 2011, the American Meteorological Society described attribution science in a report, stating that human-caused climate change played a role in five out of six extreme weather events studied. Improved computer power in the 2000s allowed scientists to simulate weather events many times. Major scientific advances in the early to mid-2010s helped develop attribution science further. In 2016, the American Meteorological Society said, "The science has now advanced to the point where we can detect the effects of climate change on some events with high confidence."

Purpose

Climate Central explained that attribution studies find out if and how much climate change influenced the strength, how often, or the effects of extreme events such as wildfires, droughts, heavy rain, and typhoons. German climatologist Friederike Otto said that attribution science tries to answer the question, "did climate change play a role" in specific extreme events that happened within two weeks of the event being studied. Results from these studies help scientists and journalists say things like, "this weather event was at least n times more likely because of human-caused climate change" or "this heatwave was m degrees hotter than it would have been in a world without global warming" or "this event would not have happened without climate change."

Another purpose of attribution science is to examine how extreme events affect people's weaknesses and how well society can adapt. This science helps people understand more clearly how climate change affects their environment, gives scientific proof for public discussions, and allows communities to prepare better for expected increases in certain types of extreme events. Individuals, groups, and businesses can use information about their specific area to predict future extreme events that might happen there.

Methods

In a paper published in 2004, scientists studied how climate change might influence the likelihood of extreme weather events. Instead of asking whether climate change caused an event, they focused on how much it might have increased the risk of the event happening. This is not a simple yes or no question.

Extreme events can be described in different ways. For example, they might be compared to historical events using percentages, measured as how far they are from normal conditions, or described by how often they are expected to occur (such as a 1% chance each year) or how frequently they are likely to happen (like a 100-year drought).

Scientists use two main methods to study these events. One method uses real-world data to track changes in the frequency or strength of events over time. The second method compares two simulated worlds: one that includes human-caused greenhouse gas emissions and one that does not. Differences between these two worlds are linked to human influence on climate.

Friederike Otto, a German climatologist, explained that scientists use climate models to simulate events under current climate conditions and calculate how likely they are to occur. They then simulate the climate of the preindustrial era by removing human-caused warming from the models or using data from the late 19th and early 20th centuries. By comparing results from these two simulations, scientists determine how climate change may have changed the intensity or frequency of an event. These methods have been reviewed by experts, allowing new studies to be published quickly.

Xubin Zeng, an atmospheric scientist, described a four-step process for attribution studies: (1) measuring the size and frequency of an event using real data, (2) using computer models to compare and confirm these measurements, (3) running the same models on a climate without human-caused changes, and (4) using statistics to compare results from steps 2 and 3.

Different studies may produce varying results because they define "extreme" events differently, use different time periods for historical records, or rely on climate models with different levels of detail. Model results can also differ depending on which aspects of an event are studied, such as peak wind speed versus average wind speed, or total rainfall versus how long the rain lasts.

Findings

In November 2024, Climate Central reviewed over 600 climate attribution studies that examined nearly 750 extreme weather events and trends. The review found that climate change increased the likelihood or severity of 74% of these events, while it decreased the likelihood or severity of about 9% of them.

Climate change affects the intensity and frequency of extreme weather events in different ways. For example, the 2010 Russia heat wave was much more likely to occur because of climate change, but its intensity did not increase. This shows that some studies focus on different aspects of climate change’s effects.

Better understanding of factors like how the atmosphere moves, soil moisture, and land cover improves predictions of heatwaves. Heatwaves are easier to model than other extreme events because long-term temperature data is more reliable. Rainfall is also relatively simple to predict, but droughts, snowstorms, tropical storms, and wildfires are harder to model due to less reliable data.

Applications and influence

Attribution science has been used as evidence in legal cases about climate change. These cases involve companies that caused climate change and governments that did not take action to address it.

Extreme event attribution (EEA) was used to study 213 heatwaves that happened between 2000 and 2023. This method helps determine which human activities contributed to these events. Research found that "carbon majors," which are companies that produce fossil fuels and cement, made 16 to 53 heatwaves more likely. These heatwaves would have been almost impossible to occur in a climate before the industrial era.

In late 2025, the Copernicus Climate Change Service announced the creation of a team that will work continuously to study climate events. This team will provide two reports each month, and each report will be completed within a week of an extreme weather event. These reports can help governments create policies, support insurance companies in assessing risks, and provide information for legal cases about climate change.

Examples

Various groups study how human activities affect extreme weather events. The Sabin Center for Climate Change Law at Columbia University has databases that people can search. These databases include categories such as climate change attribution, extreme event attribution, impact attribution, source attribution, and attribution in the courts. Carbon Brief has an interactive map showing more than 600 studies that cover almost 750 extreme weather events and trends. The American Meteorological Society publishes special reports each year called "Explaining Extreme Events of [year] from a Climate Perspective."

Examples include:

• 2003 European heatwave: "It is very likely (over 90% certain) that human influence has at least doubled the risk of a heatwave exceeding this threshold magnitude."
• 2004–2024: Human activities made the ten deadliest extreme weather events during that time period worse, leading to at least 570,000 deaths.
• October 2012: Hurricane Sandy flooded an area 27 square miles (about 70 square kilometers) larger.
• August 2016: Louisiana floods were 40% more likely.
• February 2017: A U.S. winter heat wave was three times more likely.
• August 2017: Hurricane Harvey was three times more likely and 15% more intense.
• 2017–2018 Tasman Sea marine heatwave: The overall intensity was "virtually impossible" without human activities.
• 1950–2020: Human-caused heat extremes reduced the global population of tropical birds by 25–38%.
• 2019–2020 Australian bushfire season: At least 30% more likely, and heat extremes were more likely by at least a factor of 2.
• 2020 Siberia heatwave: More than 500 times as likely as similar events in 1900; or 600 times ("minimum") to 99,000 times ("best estimate").
• 2021 Western North America heat wave: 150 times more likely.
• 2022 United Kingdom heatwave: A 2-day average temperature event that would normally happen once every 100 years became at least 10 times more likely. A single-day maximum temperature event that would normally happen once every 1,000 years occurred in one region of the UK.
• 2022: A global review found that tens of thousands of deaths were directly linked to heat extremes.
• 2022: A review of rainfall events in the North Atlantic basin found that these events caused $500 billion in damages.
• May 2023 through May 2024: Days with extreme heat (with vs. without global warming): Suriname (182 vs. 24 days), Ecuador (180 vs. 10 days), Guyana (174 vs. 33 days), El Salvador (163 vs. 15 days), Panama (149 vs. 12 days).
• 2023 and 2024 (various heat events): Probability ratios (PRs) in Marshall Islands and Micronesia (PR=35). Indonesia and Philippines (each 24 and 29). Central America (13). Spain and Portugal (8). Southern South America (7).
• 2000–2023: Of 213 historical heatwaves studied, fossil fuel and cement producers made 16–53 heatwaves that would have been virtually impossible in a preindustrial climate.
• 2024: Climate change caused 44% of the damage from Category 4 Hurricane Helene and 45% of the damage from Category 5 Hurricane Milton in Florida.
• 2026: Climate change made the 2026 Chilean wildfires three times more likely (2.5 times more likely in Patagonia).