Climate Models Become More Precise Even as Political Attacks Sharpen

As back-to-back heat waves batter the US and Europe, scientific models linking climate change to extreme weather are increasingly important for policymakers and commercial enterprises like insurers. But a growing political backlash aims to undermine their credibility.

Climate scientists can now quickly assess how much worse a given extreme-weather event has been made by human-caused climate change. Attribution science, as it’s known, has existed since the early 2000s, but improved computational resources and scientific understanding in recent years have made the findings more useful to policymakers.

Some studies can now be completed even as a weather event is ongoing. During the heat wave that hit the eastern US earlier this month, researchers with World Weather Attribution found the extreme combination of high temperatures and high humidity would have been “virtually impossible” without the greenhouse gases released by burning fossil fuels.

“This is a very mature science now,” said Mingfang Ting, professor of climate science at Columbia University, speaking on Bloomberg’s Zero podcast. “How likely an event would be to happen in a world without fossil fuels, versus with fossil fuels? We run models with those two conditions and then get a probability of the event happening,” she said.

“The basic scientific approach is that we have to frame a hypothesis and test it. And that requires a counterfactual,” said Noah Diffenbaugh, a climate scientist at Stanford University, who works on attribution science. “We can’t generate that counterfactual by putting the Earth in the lab,” he added, so the next best thing is to use a computer simulation.

As attribution science becomes more precise, some academic experts believe the research will be increasingly used in court to hold companies and governments financially liable for the damage caused by climate change.

A landmark report on extreme-weather attribution science, due to be published on July 16 by the US National Academies of Sciences, Engineering and Medicine, examines how much the field has improved and where further research is needed. There’s already a concerted effort to weaken the credibility of the report by some Republicans in Congress and members of the fossil-fuel industry, according to Politico.

Models are facing greater political scrutiny more generally. In May, for example, US President Donald Trump declared some climate models “wrong,” prompting scientists to rush to defend the utility of their work.

Climate researchers have operated under political duress for decades. Pressure “has been present the entire time, starting when I was in graduate school” in the early 2000s, said Diffenbaugh.

Whether attribution studies can be used in a legal setting remains to be seen, but the science already has a real-world application in the insurance industry. Attribution scientists “use climate models to generate hypotheses about what the climate would be with and without different forcing factors. And then we use observations of the climate to test those hypotheses,” Diffenbaugh said.

That’s the same process that insurance brokers use to predict how global warming is changing weather threats in the present and near future, and to determine premiums as a result. As climate change alters the atmosphere, historical conditions become a less reliable guide to future environmental risks. Insurers use computer simulations to price risk by understanding the probable location and frequency of extreme weather events, as well as how much damage they’ll cause.

“It is similar to these attribution studies in the sense that it is a way to try to get information about what is the difference between now and some period in the past,” said Erik Lindgren, a natural catastrophe specialist at Swiss Re, a Zurich-based insurer. “What happened over the past hundred years is not going to repeat over the next 10 years.”

The models can be misleading as well as useful, Lindgren added. “You need to understand what the limitations are,” he said. “But you can absolutely use climate models to better your understanding of what is happening in the broader climate system.”

Twenty years ago, when Diffenbaugh began working in attribution science, computers were less powerful, satellite records only went back two decades and assessing whether any single event was made worse by climate change was much harder.

The political landscape may not have become any easier, but Diffenbaugh professed confidence in the scientific methods. “Our responsibilities are to pose testable hypotheses, test those hypotheses objectively, quantify the uncertainty and accurately report those results,” he said.

Photo: A meteorologist monitors weather activity in College Park, Maryland. Photographer: Michael A. McCoy/Bloomberg