Two ways to read the story
- Quick Read
- Deep Read ( 5 Min. )
Earlier this month, as hurricane Florence approached the Carolina coast, a team of researchers put out a forecast in advance of landfall – a first – using models to quantify the likely impact of climate change on the storm’s rainfall, size, and intensity. Among other things, it concluded that the rainfall forecasts were increased by more than 50 percent due to global warming. It will take months for researchers to conduct a postmortem of the forecasts made ahead of Florence to confirm or refute their validity. But subsequent analysis of real-time attribution efforts during the 2017 hurricane season, including a study released Thursday in the journal Science, suggests that these models are on the right track. “Increasingly, these questions – did climate change impact this event – are what the public and the media are asking,” says Kevin Reed, an atmospheric scientist at Stony Brook University in New York and a coauthor of a Florence forecast. Researchers hope that a sharper understanding of the role global warming is playing in the intensity and impact of storms like these could improve forecasting and inform mitigation and preparedness planning.
With floodwaters continuing to rise nearly two weeks after hurricane Florence thrashed North and South Carolina, many are wondering how much of the deluge can be attributed to climate change.
That’s a frequent question, and one without an easy answer. Unlike some other extreme weather events, like heat waves or unusually heavy rainfall, hurricanes have been particularly difficult to pin solidly on global warming.
They’re relatively rare, for one thing, and they’re caused by complex meteorological conditions. And some skeptics note that there hasn’t been a consistent detectable long-term trend.
But as extreme-event attribution matures, climate scientists are increasingly willing to connect the dots between certain aspects of hurricanes and climate change. Researchers hope that a sharper understanding of the role global warming is playing in the intensity and impact of storms like these could inform mitigation and preparedness while answering some of the most common questions people have.
“There’s been an explosion of research over the last decade,” says Noah Diffenbaugh, a climate scientist at Stanford University who is an expert on extreme-event attribution. Among the innovations in the past few years, he says, are efforts to use real-time conditions and data as events take place. Complementary research doing longer-term studies, on the other hand, can help quantify the odds of such events occurring and the odds of the physical conditions that create those events. The two types of research “answer different sets of questions, and are both important,” says Professor Diffenbaugh.
In just the past few weeks, analysis of real-time conditions have started to connect the dots linking still-active storm systems to climate change.
Earlier this month, as hurricane Florence approached the Carolina coast, a team of researchers put out a forecast in advance of landfall – a first – using models to quantify the likely impact of climate change on the storm’s rainfall, size, and intensity. Among other things, they concluded that the rainfall forecasts were more than 50 percent higher than they would have been in a world without climate change.
The numbers are preliminary, and it will take months for researchers to conduct a postmortem of the forecasts made ahead of Florence to confirm – or refute – their validity.
Subsequent analysis of real-time attribution efforts during the 2017 hurricane season has also been finding strong links between those storms and climate change.
This Thursday, a study published in the journal Science concluded that warm sea-surface conditions in the tropical North Atlantic were a driving force in the unusually active 2017 Atlantic hurricane season, which included Harvey, Irma, Maria, and three other major hurricanes in the North Atlantic.
Hiroyuki Murakami, an atmospheric scientist at Princeton University and the lead author of the study, says they looked at La Niña as one possible contributor, but their simulations found that it wasn’t a major factor. Instead, the unusual warming of the tropical Atlantic relative to global ocean temperatures was the primary driver.
“Climate models consistently project a warmer tropical Atlantic in the future,” says Dr. Murakami. If it continues to warm faster than other oceans, that could increase the average number of major hurricanes each year from three to five, or even more, by the end of this century, he says.
“It’s still very challenging to attribute extreme hurricane events,” adds Murakami.
This fall, we got the first taste of more immediate attribution when scientists from Stony Brook University, the Lawrence Berkeley National Laboratory, and the National Center for Atmospheric Research released a forecast, and maps, just before Florence was scheduled to make landfall.
One of the biggest factors they attributed to human-caused climate change: the extreme rainfall which, in the end, was responsible for the devastating flooding – similar to Harvey last year in Houston.
Indeed, one preliminary report put Florence just behind Harvey as the second-wettest storm in 70 years. An average of 17.5 inches fell over a region of 14,000 square miles. In Harvey, 25.6 inches fell over that same area.
Several peer-reviewed studies in the past year, using a range of methods, all found that climate change significantly increased the amount of rain that flooded Houston.
While the advance attribution for Florence was unusual, since it was looking at predicted forecasts rather than the actual event, it seems to fit in that same context.
The forecast was part of an effort to answer the questions that people most want to know, says Kevin Reed, an atmospheric scientist at Stony Brook University in New York and a coauthor of the report.
“Increasingly, these questions – did climate change impact this event – are what the public and the media are asking,” says Dr. Reed. “Scientists have devolved into probabilities and trends, and aren’t always answering the questions the public is asking.”
Reed and his colleagues used peer-reviewed models and detailed data about conditions to compare present-day forecasts with forecasts that removed the climate-change elements, or “signals,” from conditions like sea-surface temperatures and atmospheric moisture.
“We’re not just decreasing the temperature by one degree everywhere,” says Reed, explaining the complexity of the models used.
Now, he says, the challenge is to work on an in-depth analysis that’s peer-reviewed and see how well their initial conclusions hold up. “We can let the community decide if these kinds of things are useful or scientifically validated,” he says.
No crystal ball
Despite all the advances in attribution science, there are still limitations, and it remains hard to tease out the role climate change plays in complex weather events like tornadoes, wildfire, and hurricanes.
Some of the elements from an event like Florence or Harvey have a lot of consensus: Storm-surge flooding is higher with higher sea levels. The increased water vapor in the atmosphere as the atmosphere has warmed is “a robust long-term trend,” says Diffenbaugh. “For a given land-falling tropical cyclone … it’s likely to produce a higher amount of precipitation.”
Other aspects of hurricanes – like their frequency and intensity – are more controversial. “We can’t say a hurricane is due to global warming,” says Adam Sobel, a professor of applied physics and applied mathematics at Columbia University, and a leader in event attribution. “We have a lot of confidence they should produce more rain, and that they should produce more wind. The controversy is over whether those changes are already evident.”
But Dr. Sobel also thinks that the standard climate scientists use – trying to disprove the hypothesis that there’s no climate-change signal with very high certainty – is arguably too conservative.
Imagine a credible terrorist attack in a crowded room, he says, and CIA agents trying to hear what the suspects are saying, but the noise of the room drowning out much of their conversation. “Would you say, ‘Can we disprove what they’re saying?’ You’d say, no, we can’t make out exactly what these guys are saying, but that’s not a reason to be complacent.”
Separating out the climate-change influence on hurricanes from the natural variability and limitations of the observational record is analogous to distinguishing those quiet voices from all the background noise, says Sobel.
In other domains where the outcome is important, “you don’t apply this extreme standard, he says. “It’s not the rational standard to use for a public-policy debate about risk.”