Hurricane season started on June 1 (despite a hurricane being reported prior to that), and a new analysis predicts that this year—and ones coming forward—could be much, much worse than the tumultuous hurricane season of 2017.
The study, from the Journal of Climate, was conducted by scientists at the National Center for Atmospheric Research (NCAR), a federal agency funded by the National Science Foundation.
Researchers analyzed two pre-existing simulations of 22 hurricanes that occurred between 2001 and 2013. One simulation took place in the current climate; the other took place in the climate predicted for 2100, which experts forecast will be about 9 degrees warmer than current temperatures.
For each of the 22 hurricanes, the team compared changes in factors such as wind speed, rainfall rate, and translation speed (the movement of the overall hurricane) between the current and future iterations.
They found, on average, that future hurricanes would move 9 percent more slowly—which would give them more time to inflict damage—with winds traveling 6 percent faster than today’s hurricanes, and producing 24 percent more rainfall.
In one particularly terrifying example, the future version of Hurricane Ike—which killed 195 people and ravaged the Gulf Coast in 2008—would have 13 percent faster winds, move 17 percent more slowly, and have 34 percent more rainfall.
Of the measured conditions, faster and heavier winds will pose the biggest threat to future populations, the paper’s lead author Ethan Gutmann told The Daily Beast.
“The most dangerous aspect of this depends on where you live and what your exposure is,” Gutmann, a hydrologist at NCAR’s Research Applications Laboratory, said. “What we’ve seen in a lot of hurricanes recently is it’s really the rain that causes the most damage. In some locations, it’s the wind... particularly thinking about offshore infrastructure, be it oil platforms or what have you.”
Jeremy Gregory, executive director of MIT’s Concrete Sustainability Hub is not affiliated with the NCAR study and had a bleaker perspective on which factor would pose the biggest infrastructural threat: “They’re all bad.”
It’s not hard to think of the implications of these findings. Last year’s hurricane season cost the nation $215 billion, according to a National Science Foundation analysis. Just last week, an independent analysis from Harvard showed that Hurricane Maria claimed 4600 lives last September, much more than the 64 lives originally reported, and more than the number of victims from the 9/11 attacks and Hurricane Katrina combined.
Studying simulated hurricanes has proven difficult in the past. Climate models, which track general phenomena over decades or centuries, tend to lack the necessary resolution to isolate a specific hurricane. Models with the requisite high resolution, like the one that generated the simulations used in this study, require astronomical computing capacity, time, and money to study long-term changes.
It took almost a year to run this model on the Yellowstone supercomputer at the NCAR-Wyoming Supercomputing Center in Cheyenne. On a regular, dual-processor computer, the same computations would have taken 40 million hours, or 4,566 years, Gutmann said.
The results are not without limitations. Kerry Emanuel, a professor of atmospheric science at MIT who is not affiliated with the study, cited two preliminary concerns stemming from the fact that the model was designed to conduct more general atmospheric research, and lacked some hurricane-specific meteorological adjustments.
First, Emanuel noted that the model could have overestimated the drag coefficient, or amount of resistance, that a hurricane encounters in its path. Typically, as wind moves faster across the ocean, it faces increasingly more resistance. But, Emanuel notes, there’s a caveat: “Once you get up to about hurricane wind speeds, the drag coefficient stops going up, and may actually go down a little bit,” he said.
Because this drag coefficient plateau was not included in the simulation, he added, the results could have overestimated the air resistance, thereby underestimating the wind speed of future hurricanes. In other words, future hurricanes might actually be more dangerous than what was predicted in this analysis.
Emanuel also posited that the model did not account for the fact that hurricanes draw up colder water as they approach the coastline. Given that hot water powers hurricanes, this influx of cold water can weaken a hurricane’s intensity, sometimes substantially.
Gutmann acknowledged the validity of Emanuel’s concerns, and the model’s inability to simulate either current or future conditions with perfect accuracy. He argued, however, that any flaws in study design would have been replicated in both the current and the future simulations. This means that even if a simulation over- or underestimated the exact speed or rainfall of a hurricane, it would nevertheless accurately capture the degree of change between the current and future versions.
And despite the limitations he cited, Emanuel agreed that the change NCAR discovered holds scientific value.
“The study represents [an] important advance in the science linking climate change and hurricanes,” he said. “[It] is a nice step forward in understanding and predicting how hurricanes will respond to climate change.”