Skip to main content
Menu
Why We’re Stuck With An Inadequate Hurricane Rating System

Update (Oct. 10, 2018, 2:15 p.m.): Hurricane Michael strengthened into a Category 4 storm before making landfall around 1:30 p.m. Eastern time on Wednesday.

Hurricane Michael is expected to make landfall on Florida’s Gulf coast Wednesday, likely coming ashore as a Category 3 stormthe strongest the panhandle region has seen in 13 years. But knowing the category rating of a hurricane is not the same as knowing how dangerous the storm is. Hurricanes also bring flooding, caused by both rainfall and the storm surge, and water accounts for more deaths than wind in these storms. Yet we still categorize hurricanes by wind speed alone.

Sometimes, the Saffir-Simpson hurricane rating system — the scale that assigns a number to a hurricane based on its sustained wind speeds — does a decent job of conveying how concerned people in the path of a storm should be, experts said. But sometimes it doesn’t. Take last month’s Hurricane Florence, for example. Though it was “just” a Category 1 storm when it made landfall, it still dumped as much as 34 inches of rain on the state of North Carolina and produced record-breaking storm surges. At least 40 people died, and anecdotes suggest that some Carolinians put off evacuating because the storm’s category made it seem like no big deal.

This problem is no secret to scientists. But, despite its flaws and a nearly annual chorus of articles calling for its replacement, the Saffir-Simpson scale remains our go-to metric for determining hurricane warnings and rankings. Moving on from it is easier said than done. Everyone wants a way to convey the risks of hurricanes to the public, even if that method is flawed. But finding a better system isn’t easy. A hurricane’s risk profile is shaped by the wind and water packed inside it and where it hits and what else has recently happened in that place. Capturing all of that in a single easy-to-understand metric that can be calculated before the storm makes landfall is probably impossible.

There are lots of different ways to define danger because lots of different factors contribute to risk. You can see how that plays out by looking at our table of historic hurricanes, which includes all the storms since 1980 that caused at least $1 billion in damage.1 According to the Saffir-Simpson scale, Hurricane Andrew, a Category 5, was the strongest storm. But Harvey, a Category 4, came with significantly more rain, and Katrina, which made landfall as a Category 3, cost the most.

A hurricane’s category doesn’t always predict its damage

Hurricanes from 1980 through 2017 that hit the continental U.S. and caused at least $1 billion in damage

Year
Name
Highest category over land
max. Winds
max. Rain
avg. Forward speed
Cost
1992 Andrew 5 167mph 14in 15mph $49b
2017 Harvey 4 132 61 128
2017 Irma 4 132 22 51
2004 Charley 4 150 10 17 22
1989 Hugo 4 138 10 14 19
2005 Katrina 3 127 16 10 165
2004 Ivan 3 121 17 13 28
2005 Wilma 3 121 13 10 25
2005 Rita 3 115 16 11 24
2004 Jeanne 3 121 12 8 10
1996 Fran 3 115 16 12 8
1995 Opal 3 115 19 12 8
1983 Alicia 3 115 10 8
2005 Dennis 3 121 13 15 3
1985 Elena 3 115 16 3
1985 Gloria 3 104 10 2
1980 Allen 3 115 20 2
2008 Ike 2 109 18 13 36
2004 Frances 2 104 24 11 13
2016 Matthew 2 98 19 11
1999 Floyd 2 104 24 13 10
1998 Georges 2 104 38 13 9
2003 Isabel 2 104 20 11 8
2008 Gustav 2 104 21 14 7
1991 Bob 2 104 8 14 3
1998 Bonnie 2 109 15 9 2
1995 Erin 2 98 20 13 1
2012 Sandy 1 75 13 14 72
2011 Irene 1 86 22 12 15
1985 Juan 1 86 18 4
2012 Isaac 1 81 27 7 3
2002 Lili 1 92 8 14 2
2008 Dolly 1 86 15 7 2

Cost figures are in 2018 dollars.

Storms that caused hurricane-force winds over land are included even if the center of the storm did not make landfall.

Sources: national oceanic and atmospheric administration, university corporation for atmospheric research

The Saffir-Simpson scale was once used to predict storm surge as a function of wind speed, said Nathan Johnson, a member of the American Meteorological Society who is also director of weather operations for NBC-owned TV stations. But that link turned out to be erroneous. Over time, scientists came to understand that there was no linear relationship between wind and surge. There were just too many other factors involved, many of which had nothing to do with the hurricane itself. “It depends on the water depth, the shape of the coastline, in addition to the wind speed,” said Lakshmi Kantha, professor of aerospace engineering at the University of Colorado, Boulder. So the Saffir-Simpson scale dropped any pretense at predicting storm surge in 2010.

It gets even more complicated when you remember that any new categorization system would have to work for the world, not just the U.S. In some places, flooding isn’t necessarily the top-ranking danger like it is in the continental U.S. In the Caribbean, Johnson told me, a hurricane’s winds are still, probably, more dangerous than the water it brings. Flooding can still claim lives and ruin property, but hurricane winds are generally stronger there, Johnson said, and may be the bigger threat. It’s a great example of how we can’t just throw away wind speed and say it’s not important. Different things matter in different places.

Nevertheless, experts still agree that we could do better at categorizing hurricanes.

Marshall Shepherd, director of the atmospheric sciences program at the University of Georgia, is one of many researchers who believes that the Saffir-Simpson scale, as it currently stands, is an inadequate tool. A good rating system, he said, would tell people how much land a storm will affect at once, how much rainfall it could bring, what the storm surge risks are, where it’s likely to hit, how fast it’s moving and, yes, its wind speed.

Shepherd envisions an improved system that can classify a storm as bigger than a “Category 5” and offer gradations within each category that convey various kinds of risks. A hypothetical 2a storm, where the risk was mostly in the wind, would be very different from a 2c, where heavy rainfall was expected.

This is just one of many alternative methods for categorizing hurricanes that experts have proposed, both informally and in scientific literature. James Done, a project scientist at the National Center for Atmospheric Research, told me he could think of at least 20 different systems that could replace Saffir-Simpson. He came up with one himself: the Cyclone Damage Potential index, which takes a hurricane’s size and forward speed into account in addition to wind speed. Kantha developed three different hurricane rating systems that incorporate combinations of several factors. There are other systems called IKE and ACE, and even a point-based system, the Hurricane Severity Index, where storms are assigned a score based on both size and wind strength.

But Shepherd, Done, Kantha and Johnson all agreed that none of these systems is likely to replace Saffir-Simpson anytime soon. And there are a couple of reasons for that.

First, the Saffir-Simpson scale effectively has a brand. After being exposed to this scale for 46 years and thousands of hours of prime-time broadcasting, Americans have developed an intuitive sense of what the Saffir-Simpson categories mean — or at least what we think the categories mean. Even if that understanding is flawed or dangerous, trying to shift us all over to a completely different system could produce even more confusion and risk. “Basically, hurricanes and Saffir-Simpson are media stars now,” said Susan Jasko, a professor of communication at California University of Pennsylvania who has consulted with the National Weather Service. “And that kind of popular culture plays by its own rules.”

For that reason, when scientists talk about the problems with Saffir-Simpson, they typically frame it as a tool that needs to be rethought so it can convey more information, not as a system in need of complete replacement. In the future, maybe TV meteorologists will assign a hurricane a Saffir-Simpson category and a complementary category that captures some aspect of the storm’s flood potential, Done suggested.

The other problem with replacing Saffir-Simpson is also the reason that there are so many proposed replacements available: Nobody has come up with the perfect solution. And there probably never will be a perfect scale that accurately expresses all the risks of all hurricanes at all times.

There are a huge number of variables that affect the potential impact of a storm, Jasko told me. Some are intrinsic to the storm itself, like wind speed or forward speed. Others are intrinsic to the places the storms hit — how developed that stretch of coastline is, for example, or what shape the underwater continental shelf takes in that area. And some are transient — a hurricane may be more likely to cause floods if the soil is already waterlogged from previous, non-hurricane rainfall.

And some of that stuff is just easier to count after it has happened than to predict before it hits, Johnson said. Wind speed (and the size of the storm) are the easiest variables to measure via satellite while the hurricane is still days away at sea. And while there is value in a better post hoc ranking system — for the purposes of insurance claims, say — it’s predictive categorization that people really want.

Until then, when a hurricane like Michael heads to shore, our best bet is to rank it by wind speed — and then hope that people read beyond the headline for warnings about rain and flooding.

Additional reporting by Ella Koeze and Rachael Dottle.

CORRECTION (Oct. 9, 5:39 p.m.): An earlier version of this story mistakenly referred to Hurricane Matthew when it should have referred to Hurricane Michael.

Footnotes

  1. Dollar amounts adjusted to 2018 dollars.

Maggie Koerth-Baker is a senior science writer for FiveThirtyEight.

Comments