The downgrading of the El Reno tornado from the top of the Enhanced Fujita scale to a weaker rating could have lethal consequences, a nationally recognized weather expert warns.
The massive tornado, which killed eight people, was initially measured as an EF-3 tornado based on damage found in the primarily rural area after it struck May 31. But wind-speed measurements collected by mobile Doppler radar detected wind speeds of nearly 300 miles an hour about 500 feet from the surface, prompting the National Weather Service to change the rating to EF-5.
The threshold for an EF-5 is 200 miles an hour. But the weather service recently reverted to a EF-3 rating.
“Despite the radar-measured wind speeds, the survey team did not find damage that would support a rating higher than EF3,” said a statement released by Keli Pirtle, a public affairs specialist with the National Oceanic and Atmospheric Administration in Norman, Okla.
“While the wind measurements from the mobile radars are considered reliable, (National Weather Service) policy for determining EF ratings is based on surveys of ground damage.”
Weather service officials are “exploring whether policy should change to allow the use of experimental radar data in future EF determinations,” the statement reads.
That needs to happen, weather officials and researchers say, so more accurate reflections of tornado strengths are recorded. Relying strictly on damage to measure tornadoes sets the stage for illusions about tornado strength created by the limitations of relying on damage alone.
El Reno widest tornado
Prior to the downgrade, the El Reno tornado was considered one of the strongest on record. Now it doesn’t even top the list of tornadoes that touched down in late May.
“I find it very hard to believe that El Reno was weaker than Rozel and Bennington,” noted weather researcher Jon Davies said in an e-mail response to questions, mentioning two other substantial May tornadoes.
Howard Bluestein, a tornado researcher and meteorology professor at the University of Oklahoma, called the El Reno tornado “an argument for why we need Doppler radar” to provide wind-speed estimates of tornadoes.
Graduate students deploying a mobile radar measured winds of 296 miles an hour in the El Reno tornado, which at one point was 2.6 miles wide – the widest tornado ever recorded.
Bluestein said estimating tornado strength based on the damage left behind is supposed to be “what you do when you have no other information.”
The El Reno scenario is one in which the Enhanced Fujita scale “really breaks down,” he said, because violent tornadoes that pass through primarily rural areas can be rated lower than their actual strength due to the relative lack of damage left behind.
It would be a largely academic debate except for the fact that reduced wind-speed estimates could lead to “under-engineered” bridges and other critical infrastructure such as nuclear power plants, said Mike Smith, senior vice president for AccuWeather Inc., the private forecasting company.
“This downgrading may mislead engineers,” Smith said.
The top category in the Enhanced Fujita scale, EF-5, includes winds of 200 to 250 miles an hour. But if a structure is engineered to withstand 250 mph winds and is hit by a tornado packing winds at or near 300, Smith said, the results could be “catastrophic” depending on what was struck.
“I’m not saying homes should be designed for 290 mile-per-hour winds,” Smith said. “But this is extremely important for critical infrastructure.”
The same basis for lowering the El Reno tornado on the EF scale was used to lower the rating of the Bennington tornado, which churned through rural northern Kansas in May. The huge, slow-moving tornado was initially rated an EF-4 but has since been lowered to an EF-3.
The reduction came despite a wind-burst measurement from the Doppler on Wheels mobile radar of 247 mph. Weather officials included a footnote on the official rating that mentions the mobile radar wind readings, said Chad Omitt, warning coordination meteorologist for the Topeka branch of the weather service.
“The challenge is that the scientific community needs to figure out whether an instantaneous wind measured at 10, 30 or 100 meters above ground can be used as a proxy or is (an) accurate indicator of what surface winds are,” Omitt said in an e-mail response to questions. “At current, we just don’t have enough data to know for sure and that’s why we continue to use the EF scale and damage indicators as the methodology to determine an EF rating without input from other sources.”
Rozel, Wichita tornadoes unchanged
Portable radars also captured wind speeds for the large tornadoes that just missed Rozel on May 18 and Wichita on May 19.
But those ratings — EF-4 for Rozel and EF-2 for the tornado that touched down near Clearwater and lifted just outside Wichita — won’t be lowered because damage caused by the tornadoes matched the rating linked to the wind speeds detected by the mobile radar, weather service officials said.
A portable radar measured winds of up to 185 mph in the wedge tornado that just missed Rozel in west-central Kansas.
Even without the mobile radar data, said Aaron Johnson, science and operations officer for the Dodge City branch of the weather service, “there’s evidence here of stuff that would be EF-4 anyway.”
Among that evidence is a 1,000-gallon steel propane tank that was lifted and carried a quarter-mile, he said.
“That’s impressive,” Johnson said.
Joshua Wurman, a noted tornado researcher who launched the Center for Severe Weather Research in Boulder, Colo., and whose Doppler on Wheels captured wind speeds for the Bennington, Rozel, Wichita and El Reno tornadoes, said relying on damage to estimate tornado strength is “fraught with error.”
“In many, perhaps most, cases, there are not enough strong structures in the path of the tornado to permit an accurate estimate of tornado intensity based just on what it destroyed,” Wurman said in an e-mail response to questions.
El Reno is a classic example of that, he said.
But portable radar measurements of wind speeds also have limitations, Wurman said. They are almost always taken well above the ground, he said, and may not be much of a reflection of what’s happening at ground level.
Wind speeds in tornadoes will typically be lower at ground level in tornadoes because of friction, said Suzanne Fortin, meteorologist-in-charge of the Wichita branch of the weather service.
Yet a tornado that hit the outskirts of Russell on May 25, 2012, was actually stronger at ground level than it was at the elevations measured by the portable radar, Wurman said. The highest wind speeds recorded in the El Reno tornado were actually the product of small, rapidly moving sub-vortices rotating within the larger tornado.
No timetable has been given on what changes, if any, might be made to how tornado strengths are measured. Smith said he hopes the weather service returns to the original scale created by Ted Fujita, which had top wind speeds for F5s of more than 300 miles an hour.
The Enhanced Fujita scale was adopted by the weather service in 2007 and features an extensive list of damage indicators to assist meteorologists in assessing destruction caused by a tornado. The revised scale was an attempt to remove subjectivity from the storm damage surveys.
But Smith said the revised scale’s flaws outweigh its usefulness.
“Fujita foresaw the possibility of measuring wind speeds in different ways,” Smith said. “His upper-end wind speeds are far more realistic than the EF scales.”