To do that, researchers run computer programs known as climate models that simulate weather patterns over time, not unlike those used for a local seven-day forecast. But they re-create the weather over decades or centuries, rather than hours or days.

“The really cool thing about climate models is that you have a world in a computer, and you can do experiments on it,” said Andrew Pershing, vice president of science at the research nonprofit Climate Central. “And you can literally do the experiment of what would this world look like if global warming never happened.”

Connecting a changing climate to human activity dates back to the work of two Nobel Prize winners for physics, Syukuro Manabe and Klaus Hasselmann, who pioneered the development of climate models starting in the 1960s. Climate models help us understand how the climate has changed in the past and may change in the future. They solve mathematical equations that describe how energy and matter interact in different parts of the ocean, atmosphere and land.

Attribution scientists use climate models to replay the past few hundred years on Earth, removing all greenhouse gas emissions made by humans. By contrasting this fictional world with our own, they can see whether extreme events such as floods, droughts or cold spells look any different and what effect those emissions have had on our weather.

For instance, an attribution study in July found that the heat waves in North America and Europe would have been “virtually impossible” in a world without climate change.

If humans had not warmed the planet by burning fossil fuels, these heat waves would still be considered rare. But in reality, we can expect them to come around every 15 years in North America and every 10 years in Europe, according to attribution scientists.

They also warn that, if humans continue to produce emissions at today’s rate, it will speed up to every two to five years starting in the mid-2030s.

“That is really important information for water resource managers, urban planners and policymakers regarding climate adaptation and resilience,” said Kevin A. Reed, professor of marine and atmospheric sciences at Stony Brook University.

Until recently, scientists largely avoided connecting any individual event with climate change, with the idea that weather is by its very nature unpredictable and has no single cause. But in 2004, what is considered the first extreme event study determined that climate change “at least doubled the risk” of the previous year’s heat wave in Europe that killed over 70,000 people.

Almost any weather event might occur by chance, but the authors argued that climate models could be used to tease out the role that humanity played in making such intense heat more likely. They simulated the climate with and without human emissions thousands of times, counting how many times a heat wave as extreme as the 2003 one popped up. While the event was rare in both cases, it occurred twice as often in the world with human emissions.

Since that pivotal first study, scientists have investigated more than 500 weather-related disasters across the globe, with 71 percent of them found to be made more likely or more severe based on human-caused climate change. With the help of faster computers and more precise climate models, researchers can now perform these analyses within days instead of months.

The World Weather Attribution (WWA) initiative, formed in 2015 by an international team of climate scientists, has carried out more than 50 attribution studies, most in the aftermath of events or while they are still happening. The WWA was responsible for the analysis of July’s extreme heat around the world, which took only five days to compile.

Many researchers see extreme event attribution as a communication tool that has the power to connect climate change with people’s everyday experiences. “Doing attribution in real time is what will be the most useful for people to get informed, while there is focus on an extreme event,” said WWA team member Robert Vautard, director of the Institut Pierre-Simon Laplace.

The WWA has pointed to global warming as the main driver for droughts in East Africa since 2020, a 2022 heat wave in South America and 2022 flooding in Pakistan.

But in the case of a 2019 drought in Madagascar, WWA found that reduced rainfall was mostly due to natural climate variation, despite the United Nations claiming otherwise.

Climate Central has adapted the methods developed by WWA to create the Climate Shift Index (CSI), a metric that reveals how much daily weather conditions have been altered by climate change. Pershing and his colleagues average results from 22 climate models, calculating the likelihood of local, daily temperatures with and without historical greenhouse gas emissions.

Their recent analysis that climate change made July hotter for more than 6.5 billion people, or 81 percent of Earth’s population, had looked at data for 4,700 cities and 200 countries. “Virtually no place on Earth escaped the influence of climate change” in July, Pershing said.

Reed’s laboratory specializes in event attribution studies that look specifically at the effect on hurricanes. For each study, he runs 40 simulations over the past 150 years, each with a slightly altered sea surface, atmospheric temperature and atmospheric humidity to add the element of chance in influencing weather conditions. He also includes a “preindustrial control run” that captures the climate of 1850, or a time before the rise in human emissions.

After zooming into a particular time and region to capture a target hurricane, Reed runs seven-day weather forecasts. To identify the human fingerprint, they compare hurricane features over that week, such as rain rates, accumulation amounts, intensity and size.

His study of the 2020 North Atlantic hurricane season, one of the most active on record, discovered that climate change increased rainfall rates by 11 percent and rainfall amounts by 8 percent.