Evolving the Saffir-Simpson Hurricane Categories
When Hurricane Irma leveled the small Caribbean island of Barbuda with 185 mph winds on September 6, 2017, it was obvious that Irma was no ordinary Category 5 storm. People exclaimed, “There’s nothing like this!” And “It’s a Category 6!” Others countered with “There’s no such thing as a Category 6!” Or “There’s no need for a Category 6 because Category 5 already destroys everything!” Or, in the case of the current director of the National Hurricane Center, “I’ll fight ’em tooth and nail under my regime!” And the Scientific American asked, “Are Category 6 Hurricanes Coming Soon?” Of course, it is true that there is no Category 6 on the Saffir-Simpson hurricane scale.
The Richter scale once had a similar problem. It only went up to magnitude 7 because seismographs of the time couldn’t measure anything worse. But there were quakes where magnitude 7 was clearly too small. So scientists developed better seismographs, extended the Richter scale, and continued to work on better earthquake models. Oddly, publically used hurricane scales haven’t similarly evolved. In contrast to the original Richter scale, the Saffir-Simpson scale was actually dumbed down in 2009 to eliminate information about storm surges and is now officially known as the Saffir-Simpson Hurricane Wind Scale (SSHWS). Moreover, additional categories have not been added as especially devastating hurricanes came along. Of course, even the dumbed down Saffir-Simpson wind scale still has great utility in meteorology and climate science despite these limitations.
The Saffir-Simpson scales measure local severity, meaning severity as it affects people at the time they are struck by the storm. By contrast, global measures of severity are good for retrospective classifications of hurricanes. They attempt to capture a hurricane’s total energy, taking into account the size, strength, and duration of the storm. Here we will be talking only about local measures.
Three main factors determine the local severity of a hurricane: wind speed, storm surge, and rainfall. Storm surge is the rising of the sea as a result of various factors such as wind speed, atmospheric pressure changes, the size of the storm, and the shape and slope of the local continental shelf. The original Saffir-Simpson scale developed in 1971 did include storm surge levels. When they were eliminated the National Hurricane Center gave a couple of reasons: The techniques they had been using to measure wind surge were embarrassingly inaccurate, and they wanted a very simple metric that was easy to explain.
As with the original Richter scale, the Saffir-Simpson scales could easily be extended to accommodate more extreme events. The average increment in wind speed for the seven original Saffir-Simpson categories is 25, a nice round number that can be used as the increment for new Categories. Additionally, I have restored the storm surge levels which are tied to overall categories by allowing a hurricane to achieve a given category either because of wind speed or because of storm surge. Restoring the storm surge metric makes sense, as the storm surge numbers could easily be based on the computerized model the National Hurricane Center is now using for predicting storm surges, the SLOSH model developed in 1992. The resulting extended Saffir-Simpson scale looks like this:
If we’ve done it right, the most severe hurricanes should now be Category 6 (or in the case of Hurricane Allen, Category 7). In the last 100 years, there have been six such hurricanes (per Wikipedia). The most expensive hurricane was Katrina with $108 billion in damages, strongly suggesting the need for more attention to storm surge and resulting flood damage.
The fact that only six hurricanes in the last hundred years get reclassified means that historical discussions based on the original Saffir-Simpson scale will still be valid without the need for significant revision.
With either the original or the dumbed down versions of Saffir-Simpson, it makes sense to use categories for structural requirements, e.g., to require that a structure withstand Category 4 hurricanes. But such a requirement wouldn’t have made sense for Category 5 hurricanes because the Category 5 severity was unbounded. With the extended scale, it now makes sense to specify structures that can withstand Categories 5, 6, or 7 as well. Of course, a Category 7 requirement is probably unreasonable given how few Category 7 hurricanes there are, currently.
One doesn’t have to look far to see that the barriers to using a better hurricane scale are as much social as technical. The need to base storm severity not only on wind but also flooding (i.e. storm surge and rainfall) is advocated by meteorologists such as Marshall Shepherd, who is a past president of the American Meteorological Society. Tim Heller, chief meteorologist of the ABC affiliate in Houston has argued that even going back to the original Saffir-Simpson scale based on a maximum of wind and storm surge categories would be a significant improvement. A key question is whether Americans just too dumb to use a metric based on both wind and water. It’s a legitimate question deserving of informed public discussion. And it is clearly not a purely meteorological question that should be left to meteorologists or the National Hurricane Center.
Ideally, wind categories should be based not on wind speed but on wind load, the force of wind on a sufficiently large static structure, as wind load is more relatable to structural damage. Speed and load are related by the equation
where q is the pressure in pounds per square foot and V is the velocity in mph. The main point of this equation is that with increasing categories, the wind speeds should be going up quadratically rather than linearly as they do in the above extended Saffir-Simpson scale. With care, it would be possible to harmonize a new wind load scale with Saffir-Simpson so as to minimize discrepancies for average strength hurricanes — the principle of least surprise. An analogous harmonization took place when seismologists developed the new moment magnitude scale to more reliably measure severe earthquakes. The two scales match fairly well for mid sized earthquakes despite being based on different technologies.
This article has ignored global measures of hurricane severity not because they’re unimportant but because they serve a different purpose. Significant progress on global measures of severity is found, for example, in the TIKE model developed at Florida State University.
The intent of this article has been to stimulate discussion of immediately available improvements to the Saffir-Simpson wind hurricane scale. In the longer term, there is a great deal of research that should eventually produce much better metrics that could lead to better public awareness of storm threats. A glimpse into what’s going on can be seen in Mark Powell’s presentation “A New Metric for Hurricane Destructive Potential.”