The 2017 hurricane season is over, and New Orleans dodged all the bullets. Though Hurricane Nate may have prompted some to evacuate, it was Hurricane Harvey in east Texas and western Louisiana, along with Hurricane Maria in Puerto Rico and Hurricane Irma in Florida, that caused the biggest problems.
Communicating the potential danger of any storm has long been a difficult task for forecasters. Like any extreme weather phenomenon, hurricanes can be hard to predict because of the sheer number of convergent variables, many of which develop at the most critical time: as the system nears landfall. That's where the storm meets weather systems such as the jet stream over the mainland. It's also where the bathymetry (topographical conditions of the ocean floor) can rob the storm of its strength — or power it up to destroy communities with surging water.
Meteorologists have long been aware of the limitations of the Saffir-Simpson scale, which ranks from 1 to 5 a hurricane's strength and the danger it poses, based on maximum sustained wind speed. Critics, meteorologists and the public know that hurricane danger is not limited to the speed of the storm's cyclonic winds. Hurricane Katrina — the costliest hurricane to hit the American mainland in modern history, causing $108 billion in damages — made landfall in Mississippi as a Category 3, with maximum sustained winds up to 129 miles per hour. The sheer size of the cyclone, however, caused more than 50 breaches in flood protection systems such as levees and floodwalls, according to a 2007 report by the American Society of Civil Engineers. Allison in 2001 never got beyond tropical storm force, but it flooded metro Houston just as 2017's Hurricane Harvey did.
Dennis Feltgen, public affairs officer for the National Hurricane Center (NHC), said he has not heard meteorologists complain about the scale and emphasized that it does its job — classifying hurricanes according to wind speed.
"It's a wind scale only," he said. "It has nothing to do with rainfall, storm surge [or other factors]. The limitation is in the perception of the scale." Feltgen recounted the damage ($71.4 billion) caused by 2012's Hurricane Sandy, which made landfall as a Category 1 storm but produced a massive storm surge that flooded much of the eastern seaboard.
"They were lulled into a false sense of security," he said. When Gambit asked whether such widespread underestimation indicated a flawed classification system, he said it did not, but that the NHC had instituted storm surge warnings and watches just before the 2017 hurricane season. He suggested the augmented warnings were responsible for the relatively low loss of life incurred throughout the season despite a number of severe storms hitting populated areas.
"Almost 90 percent of fatalities from hurricanes are related to water, and half of those to storm surge," he said. "Only 8 percent are attributable to wind-caused deaths.
"We used (the new system) three times this year: for Harvey, Irma, and Nate. There were no storm surge-related deaths. None." He added that the surge from Hurricane Sandy claimed 41 lives in 2012.
As of now, that system has been built out only in the continental United States — not the Virgin Islands or Puerto Rico, which were devastated by Hurricanes Irma and Maria.
The information public officials receive from meteorologists informs whether they recommend people evacuate or ride out the storm in place. Boarding up homes and leaving town are expensive, but not adequately warning citizens in the path of danger could be lethal — and the Saffir-Simpson hurricane wind scale has been the primary vehicle for that message.
It's not as comprehensive a tool as one might want, and as more people move into areas prone to hurricane devastation, its utility may prove inadequate to express the urgency of the threat.
"The population has grown and so have the population centers near the coast where tropical events are more likely to impact," WWL-TV meteorologist Chris Franklin told Gambit. "More people equals more impacts from severe weather."
The scale only takes cyclonic wind speed into account and doesn't include other factors such as the speed of the overall storm, the physical size of the storm and any likely accompanying surge. But earlier versions of the scale did. Pressure and projected storm surge formed the first incarnation of the scale.
Those factors were dropped in 2009 in an attempt to provide greater clarity to the public, as a hurricane with a relatively small diameter can be rated highly while producing less storm surge than a lower rated hurricane of a larger total size. According to the NHC, 2008's Hurricane Ike made landfall as a Category 2, but the large size of the storm, nearly 250 miles across, including tropical storm force winds, allowed it to produce a storm surge of 20 feet. In contrast, 2004's Hurricane Charley made landfall in Florida as a Category 4, but it was relatively tiny — only 50 miles wide — and its surge peaked at around 7 feet.
While reducing the factors expressed results in a more "scientifically defensible" scale, according to the NHC, it also reduces the total amount of information the public can use for preparation.
"Many meteorologists hope for a new classification scale for tropical storms and hurricanes," Franklin said. "However, the NHC has experimented with other ways of describing the strength of a storm. The [Saffir-Simpson] scale was designed for purposes of expressing the damage potential based on winds. It was known then, as it is now, that there are many other elements responsible for damage and loss of life."
New Orleans native Geoffrey Gauchet decided to design his own scale. A software engineer and stand-up comedian, the Gentilly resident began fleshing out his alternative to the Saffir-Simpson in 2011.
"I call it the Tropical Cyclone Intensity Scale, or TCIS," Gauchet said. "My goal with it is to explain that a storm can be devastating even if it's 'only' a Category 1 or 2 on the [Saffir-Simpson] scale. Right now, the distance-from-land value is whatever the NHC gives as the closest land to the storm, which isn't the best value, for instance, if the storm is by the Lesser Antilles but is heading to New Orleans. It's only giving the intensity for the Lesser Antilles and then, when it gets in the middle of the Gulf, it's not by any land, so its score will drop a tad. I think it'd be best suited to be tailored for a specific area. So a storm that's an 8.1 for Houston would be significantly lower for Mobile (Alabama), since it's pretty far away." (His website, tcis.geoffreygauchet.com, carries this disclaimer: "Other than his interest in tropical weather and being a software engineer, he has no real qualifications to create this type of scale. As such, it should not be used for emergency planning purposes. This is an unofficial scale and has not been overly tested, nor has it been peer reviewed.")
Gauchet doesn't have a degree in meteorology, but he said his work as a software engineer makes it easy for him to work with large, complex data sets. His system factors in cyclonic wind speed, just as the Saffir-Simpson scale does, but includes several other factors, such as size of the storm, pressure, projected storm surge and overall velocity, which he attempted to weigh according to their likely impact on the storm's destructive power.
"So once I ranked what I felt was the importance, I weighted each indicator," Gauchet said. "I don't remember the initial weights, but the first storm I tracked with this was Katia in 2011. I used that to adjust the indicators to get it close to what I considered my 12-point scale would line up with the Saffir-Simpson category scale as a baseline. Then, using reports of the actual damage caused by the storm, I adjusted the weights to fall more in line with what actually happened."
Gauchet has been working with information from about a dozen storms over the last six years, ranging from tropical depressions to Category 5 storms like 2015's Hurricane Patricia in the Pacific Ocean, continually adjusting the weights of his system until they lined up with observed data.
"I went with a 12-point scale because I originally tested 10, but it didn't let the storms' nuances affect their scores properly," he said. "I ran this against some historical storms, and some low-damage, low-intensity storms were too close to very damaging and intense storms like Katrina. There's probably a better way to determine the max score and calculate based on that, but 12 seemed the best case for now."
Franklin has been working with hurricanes since returning to New Orleans shortly before Hurricane Katrina in 2005. Gambit asked Franklin to pick apart Gauchet's system and to look for holes in the methodology.
"Maximum sustained winds are universal," Franklin wrote in an email. "What I mean by this is that winds are not dependent on the location. Building codes can be utilized to estimate potential damage to structures. Storm surge, on the other hand, differs based on coastal geography and bathymetry (the underwater terrain) leading up to the coast. A steep slope of the continental shelf just offshore, i.e. [Southeastern Florida], tends to lead to a smaller storm surge, compared to a gradual slope, such as off the [Louisiana] coast, which would allow the exact same storm to produce a greater surge. Storm surge forecasts are still very difficult and can dramatically change leading up to landfall."
Franklin says Gauchet's idea to include proximity to land would be very confusing due to the same storm having multiple "categories" for multiple cities.
"For example, a storm 50 miles south of New Orleans but headed for Houston would have a greater category value in New Orleans because it is closer, even if the storm is not forecast to make landfall here," Franklin said. He added that, for all its inadequacies, the Saffir-Simpson scale was easy for the public to understand, and that simplicity was valuable for forecasters helping residents make important decisions.
"The NHC toyed with the idea of a category number for each of the storm impacts — wind, surge, rainfall, etc. This would get far too confusing and not help the general public," he wrote. "One of the greatest challenges for meteorologists, especially ones on air, is trying to get the public to understand the danger. This is why we, as well as the [NWS] and Hurricane Center, discuss the potential dangers for all types of weather. ... We, the media, as well as the NWS and NHC, have graphical interpretations to convey all the dangers he listed: storm size, wind speed and distance from land. ... I believe most meteorologists agree that the SS scale is not completely satisfactory since it only addresses the danger and damage caused by wind. Rainfall, storm surge, storm size all are addressed in the hurricane center's discussion, but not given a 'number' corresponding to a value."
Franklin also noted some of the values in Gauchet's proposed scale are famously difficult to correlate with the actual outcome.
"After covering tropical weather in New Orleans before and after Katrina and every other major hurricane in the U.S. since, I don't believe the public thinks of a storm as 'only' a Category 1 or 2," he added. "Just since Nate (in October), questions and comments from the public lead me to believe people now take all tropical weather seriously."
Franklin also noted that, while the essential character of hurricanes hasn't changed, the areas they tend to affect have changed considerably. Population increases in hurricane-prone areas have resulted in more paved surfaces, increasing the risk of flooding. The loss of natural barriers to storm surge and wind, such as cypress forests and marshlands, have increased our exposure to the most damaging effects of these powerful storms. Those population increases have also made evacuating large, densely inhabited areas a formidable logistical challenge — one that a scale couldn't remedy.
"I would welcome a new scale with open arms, if we could find the correct way to do it," Franklin said. "All of the impacts I've mentioned — rainfall, storm surge, storm size — need to be included and are in the discussion. It may be that we never find a way to incorporate these elements in a number classification system, but need to continue including them in the forecast discussion and issue watches, warnings and advisories."
Gauchet said he will continue to make adjustments to his scale, looking for ways to keep it more accurate. Like many in New Orleans and the Gulf Coast, he also will keep his weather eye out for storms that threaten lives.
The story behind the men who came up with the hurricane scale
To learn more about Geoffrey Gauchet and his TCIS hurricane scale, visit his website, www.tcis.geoffreygauchet.com, which has the information he has aggre gated over the last six years, as well as a more thorough explanation of his method- ology. His Twitter handle is @animatedGeoff.
For weather forecasts, as well as storm-related issues and questions, follow Gambit's TV partners at www.wwltv.com/weather or on Twitter, @WWLTVWeather. You also can follow meteorologist Chris Franklin at @CJohnFranklin.