The Washington PostDemocracy Dies in Darkness

AccuWeather developed a hurricane category scale, and it worries some meteorologists

Volunteers help rescue residents from their flooded homes during Hurricane Florence in New Bern, N.C. Hurricane Florence made landfall in North Carolina as a Category 1 storm, but flooding forced hundreds of people to call for emergency rescues. (Chip Somodevilla/Getty Images)

There is consensus that something needs to change about the way forecasters categorize hurricanes. Category 2 storms are inflicting Category 5 damage. If you’re only looking at that one number, you might not think you need to evacuate your coastal home, for example, like the 1,300 people who were rescued in New Bern, N.C., when Hurricane Florence made landfall in 2018.

“A lot of people did stay because the storm went down from a Category 4 to a Category 2, and they think of that number as being the end-all,” county public information officer Amber Parker said at the time.

Since Florence, and Hurricane Harvey before that, and even Hurricane Sandy before that, which technically wasn't even a hurricane when it made landfall in New Jersey in 2012, the meteorological community has been considering whether the Saffir-Simpson scale, which only quantifies wind speed, is good enough to convey a storm's threat.

On Thursday, the private weather company AccuWeather announced it was launching its own hurricane scale this year, 1 to 5, just like the Saffir-Simpson. Joel Myers, AccuWeather’s founder and chief executive, says the RealImpact Scale for Hurricanes, or RI, takes into account wind, flooding rain, storm surge and economic damage. He said he had seen the outcry for a better categorization system and that no one else stepped up.

“The current scale does not reflect the real impact of hurricanes,” Myers said. AccuWeather’s RI is a forecast that tells people what’s going to happen over land, Myers said. “It’s just a larger scale [than Saffir-Simpson] that is more comprehensive.”

Myers wouldn’t share with The Washington Post how the algorithm works (“We’re not going to reveal that, but there is a system”), but he did say AccuWeather tested it on storms dating to 2000 and that it performed well, in the company’s opinion. According to the news release, Hurricane Florence would have been an RI4 at landfall, versus a Category 1 on the Saffir-Simpson scale. Hurricane Harvey would have been an RI5, and so would Sandy.

Meanwhile, as AccuWeather was developing its algorithm, a small group of scientists and communicators were meeting almost weekly to discuss how to improve the Saffir-Simpson scale. Hurricane Florence tipped the scales for them, too.

“My family has property close to New Bern, so I was paying attention” to Hurricane Florence, said Ricky Rood, a professor of meteorology at the University of Michigan. After that storm, he reached out to a few other scientists to see if they were interested in collaborating to change the categorization system.

Ultimately, Rood said, “we decided it was premature.” Several members of the group interviewed people about their interpretation and use of the Saffir-Simpson scale and the warnings, and “they felt like we needed a whole lot more information on how emergency managers used the current information ... before we started to propose new scales.”

"We also felt strongly that we needed to coordinate with the Hurricane Center to be effective," Rood added.

James Franklin worked at the National Hurricane Center for nearly two decades before he retired in 2017. He knows the shortcomings with the Saffir-Simpson scale, which, other than wind, “doesn’t do anything particularly well," he said.

“Rather than rework the scale, it seems to me that we really want to focus on each hazard separately,” Franklin said. “The Hurricane Center has tried — maybe not with a great deal of success — but has tried to de-emphasize the scale and emphasize the hazards instead. We understand the limits of this one scale. And now you’re going to have another one,” he said of AccuWeather’s RealImpact announcement.

The history of the Saffir-Simpson scale is punctuated with revisions. When it was developed by engineer Herbert Saffir and meteorologist Bob Simpson in 1971, it wrapped in all of the impacts: wind, surge, flooding. In 2009, after seeing the scale perform less than adequately in conveying particular threats, hurricane specialists at the NHC decided to reduce the scale to only wind. Since then, they have added storm surge warnings to their portfolio and tweaked the wind speed categories here and there. Generally, though, it appears the NHC would also like the scale to disappear.

“If I could wave a wand and make it go away, I would,” Bill Read, then the NHC director, said in 2009.

Myers says that’s kind of the point. No one is filling this need, so he and his business are stepping in, whether the rest of the community likes it or not. And he’s no stranger to criticism, whether it be for 45-day forecasts, 90-day forecasts or issuing non-government tornado warnings to clients.

“We’re going to continue to innovate,” Myers said, “and as all innovators, we’re going to continue to be criticized.”

But there could be bigger issues than industry acceptance when two scales are being used to categorize a potential disaster: Confusion. Anxiety. Panic. Even AccuWeather’s announcement listed several examples of where the RealImpact category would be significantly different from the Saffir-Simpson. In practice, having the NHC say a storm is a Category 2 while AccuWeather predicts it will be a RealImpact 5 might not go well.

“By itself, this scale has merits,” said Gina Eosco, a weather communications specialist with Cherokee Nation Strategic Programs. “The concern is that we’re asking people to attend to the information more carefully by asking them to understand the underlying mechanism of both [scales], and that can be difficult in a high-pressure situation.”

Rood made it clear he didn’t want to say the public is ill-informed, “but” — he took a long pause — “there are many levels of being informed," he said. “Adding more complexity is likely to create more confusion. That was one of the reasons we felt we should not start creating new scales.”

Myers totally rejects the premise that his new scale will be a problem.

“The Weather Channel for years has been naming storms, and in my opinion, that’s much more misleading,” Myers said. And, he noted, there’s confusion every day in the media, “between Fox News and MSNBC and the New York Times and The Washington Post,” so he finds the accusations weak.

“We’re going to save tens of thousands of lives," Myers said, “and we’re proud of it.”

Read more:

Trump taps AccuWeather CEO Barry Myers to head NOAA

The Weather Service is open 24/7, forecasters are working without pay, and it’s taking a toll

IBM announces high-resolution global weather forecast model