Skip to main content
Wind Speed Alone Isn’t The Best Way To Measure A Weird Hurricane Like Matthew

Hurricane Matthew was inching along the Florida coastline on Friday, so close to land that its eyewall — the region of the hurricane’s strongest winds — is brushing the shore, but not close enough to make an official landfall. That raises a question: When a storm doesn’t necessarily come ashore, but still has the potential to cause massive damage and huge storm surges, how can we tell how serious it is with our current tools?

Hurricanes are fickle beasts that defy easy classification. When there’s one headed toward shore, meteorologists need a better shorthand to describe how bad it might be. I incorporated some existing metrics into a new calculation to offer a possible solution.

More Science & Health

To get one thing clear right from the start, the Saffir-Simpson Hurricane Wind Scale, the official metric by which a hurricane’s strength is rated a category 1 to 5 by the National Weather Service, is far from perfect, and meteorologists have been clamoring for years for something better. That scale considers only one variable: What is the maximum wind speed found at any point within the storm?

At first glance, using wind speed to rate hurricanes isn’t a bad idea. The destructive power of wind increases according to the cube of the wind speed,1 so hurricanes with very high wind speeds, like Matthew has had for most of its life, are much more dangerous than lesser hurricanes. Measuring wind speed in a hurricane is hard, but we’ve gotten pretty good at it. In practice, you need a very good (and safe!) way to measure small areas of strong winds in the core of the storm — like the airborne meteorological laboratories of the Hurricane Hunter planes.

But looking a bit closer, peak wind speed alone isn’t a very good metric to judge a storm’s overall danger to people and property. Water, not wind, is the biggest threat to life and infrastructure in most tropical systems. From 1963 through 2012, only 8 percent of U.S. deaths from tropical storms and hurricanes were directly attributable to wind (for instance, via tree branches or flying debris). Drowning and other deaths related to coastal storm surge were 49 percent, and deaths attributable to inland flooding from rain made up 27 percent. More people died swimming, surfing and boating during a hurricane than from wind. We need a better way to describe the threat of a hurricane like Matthew, a storm whose strongest central winds may not come ashore at all but that still brings a 1-in-500 year coastal flood risk to parts of the Southeast.

Hurricane Sandy was a perfect example of this mismatch between the Saffir-Simpson scale and storm danger: Up until just hours before landfall, Sandy was just a Category 1 hurricane, yet its enormous size and record-breaking storm surge, in proximity to a densely populated region, made it the second-costliest storm in U.S. history in terms of property damage, behind Hurricane Katrina, and the deadliest in the Northeast since 1955. Matthew has a different, but equally problematic attribute: It is forecast to retain hurricane strength as it travels roughly parallel to a long stretch of the Southeast coastline, greatly enhancing its potential damage. An ideal apples-to-apples comparison of hurricanes would consider storm surge, size and track as well as wind speed.

Thankfully, such a metric exists: Integrated Kinetic Energy directly measures the destructive power of a hurricane by calculating the combined total energy, in terajoules, of all its winds above tropical storm force. (For reference, at its peak, Sandy had the largest IKE of any known Atlantic hurricane, about 329 TJ; the atomic bomb the U.S. dropped on Hiroshima released about 63 TJ.) Based on IKE, which takes into account a hurricane’s size, it’s much easier to estimate the amount of water it will push ashore as storm surge. Surge Destructive Potential, based on IKE, is a Saffir-Simpson-style 0-6 scale that gives a quick assessment of how unusual a given hurricane’s storm surge is likely to be. (Sandy topped this scale, too, with a value of 5.8 out of 6.)

2016 Matthew* 120 3 49.2 3.8
2016 Hermine 80 1 32.2 3.1
2005 Wilma 120 3 110.3 5.0
2005 Katrina 75 1 6.8 1.9
2005 Dennis 120 3 55.4 3.9
2004 Jeanne 115 3 74.8 4.3
2004 Ivan 120 3 131.4 5.2
2004 Frances 105 2 78.1 4.4
2004 Charley 150 4 21.0 2.3
1999 Irene 80 1 38.8 3.5
1998 Georges 110 2 77.3 4.4
1998 Earl 80 1 51.9 3.8
1995 Opal 115 3 101.2 4.9
1995 Erin 100 2 10.5 2.0
1992 Andrew 165 5 42.1 3.6
Two measures of hurricane intensity

*Matthew data is as of 11 a.m. EDT on Oct. 7. Matthew has not yet made landfall.

In terms of IKE and SDP, Matthew falls roughly in the middle of the 15 most recent hurricanes to hit Florida. These measures aren’t perfect: 1992’s Hurricane Andrew, which was a small but very intense hurricane, was the fifth-most destructive hurricane in Atlantic basin history in terms of damage, mostly because it hit the highly populated Miami metro area. Andrew was a case where wind was the biggest risk — the storm was too small to produce a very large surge. In contrast, Matthew is larger, but weaker — and ranks higher than Andrew here.

Measures such as IKE and SDP are still based on wind speed, but they are much more comprehensive than just the single most extreme wind speed reading in the entire storm. To take this idea a step further and account for weird storms like Matthew, I factored in the length of coastline affected by hurricane-force winds. I used this to calculate a measure I’m calling “Andrew Units,” which multiplies a hurricane’s Surge Destructive Potential by the miles of coastline a storm affected at hurricane strength (or, in Matthew’s case, is expected to affect) and normalized all of those by the measurement for Hurricane Andrew.2 In short, this metric is designed to rank hurricanes based on the area of coastline they’ve inundated — the deadliest and most damaging thing a hurricane can do.

2016 Matthew 3.8 450 7.4
2016 Hermine 3.1 40 0.5
2005 Wilma 5.0 260 5.6
2005 Katrina 1.9 20 0.2
2005 Dennis 3.9 50 0.8
2004 Jeanne 4.3 120 2.2
2004 Ivan 5.2 165 3.7
2004 Frances 4.4 125 2.4
2004 Charley 2.3 80 0.8
1999 Irene 3.5 75 1.1
1998 Georges 4.4 120 2.3
1998 Earl 3.8 150 2.5
1995 Opal 4.9 150 3.1
1995 Erin 2.0 100 0.9
1992 Andrew 3.6 65 1.0
The ‘Andrew Unit’ for determining hurricane destruction

Matthew data is as of 11 a.m. EDT on Oct. 7.

This is not a wholly original idea — two Florida-based hurricane researchers did something similar last year, but I’m focusing only on the stretch of a hurricane’s lifespan in which it’s affecting land. I’m only considering only the length of coastline here, and not the land speed of the hurricane or the extent to which the specific stretch of coastline is built up and occupied by homes or businesses, so actual damage for two different but equally long stretches of coastline could be quite different. But you can see that for all recent hurricanes to directly affect Florida since Andrew, Hurricane Matthew jumps to the top of the list when you consider its weird track — likely bringing hurricane-force winds from Vero Beach, Florida, to Charleston, S.C. — about 450 miles.

That’s not to say that Matthew is likely to flatten homes with strong winds, like Andrew did, or cause a storm surge that wipes homes from their foundations, like Ivan did. But cumulatively, over a large area, Matthew could be very bad.


  1. Wind power is proportional to velocity x velocity x velocity.
  2. I used Andrew as the baseline because it’s kind of a “storm of record” in Florida, the one people seem to remember and reference.

Eric Holthaus is a meteorologist who writes about weather and climate. His articles have appeared in Slate, Vice, Quartz, the Wall Street Journal, and Rolling Stone.

Filed under Weather 47 posts

Comments Add Comment