Since 2007, there have been 11 tornadoes in the United States rated as 5s, the highest level in categorizing tornadoes. The 11 5-rated tornadoes before that year occurred between 1985 and 1999. What could have caused such a sharp increase in significant tornadoes?
A spike in instances of severe weather has been a common factor in the climate change debate, and the more frequent number of highest-level tornadoes has been an argument for evidence in favor of global warming. However, that claim may not be taking an important factor into account: changes to the classification system for tornadoes, which are often open to interpretation and leave room for inconstancies.
In 1971, Dr. T. Theodore Fujita created a scale for rating tornadoes based on the damage that winds caused. Called the Fujita Scale, this system ranks tornadoes on a scale of F-0 (weakest) to F-5 (strongest). For decades, this was the method used by the National Weather Service to rate tornadoes, but they point out that there are flaws with this system.
"Tornado wind speeds are still largely unknown, and the wind speeds on the original F-scale have never been scientifically tested and proven," according to the Storm Prediction Center's Roger Edwards. "Different winds may be needed to cause the same damage depending on how well-built a structure is, wind direction, wind duration, battering by flying debris, and a bunch of other factors. Also, the process of rating the damage itself is largely a judgment call--quite inconsistent and arbitrary."
Because of this, the NWS switched from the Fujita Scale to the Enhanced Fujita Scale in February 2007.
"The idea is that a 'one size fits all' approach just doesn't work in rating tornado damage, and a tornado scale needs to take into account the typical strengths and weaknesses of different types of construction," states Edwards.
The EF scale uses 28 damage classifications to rate the strength of a tornado. The newer system requires lower wind speeds to consider a tornado to be 5-rated. The minimum wind speed estimate needed to be an F-5 was 261 mph. The minimum estimated speed to be an EF5 is 200 mph. The May 2007 Greensburg, Kan., tornado was labeled an EF-5 with estimated winds of 205 mph. Had this tornado occurred a few month earlier, it would have only been an F-3. Adding to the confusion is that EF-classified tornadoes may still be written as a F-level storms.
AccuWeather's Adrienne Green explains the EF scale.
While a more consistant and measurable scale than the original, the Enhanced Fujita still uses wind estimates to determine the strength of a tornado based on surveying damage after a tornado has occurred.
"The problem with only using damage to assess a tornado's strength is that tornadoes that don't cause damage may be misclassified," said AccuWeather Senior Expert Meteorologist Bernie Rayno. "What happens when a tornado touches down in an open field? What if there is no structural damage to study?"
It is possible to use portable Doppler radar to classify the tornado's strength. In A Recommendation for an ENHANCED FUJITA SCALE (EF-Scale), the Wind Science and Engineering Center at Texas Tech University writes:
"The technology of portable Doppler radar should also be a part of the EF Scale process, either as a direct measurement, when available, or as a means of validating the wind speeds estimated by the experts."
This statement has been considered controversial in the weather community, however.
"It specifies 'portable' radar, and there aren't as many of those available," Rayno explained. "So not every tornado has radar that can go along with it."
Rayno is concerned that using this technology for some storms and not others creates too many discrepancies between storms.
"If you're going to use it, use it consistently," he said.
Mike Smith, Meteorologist and Sr. Vice President/Chief Innovation Executive of AccuWeather Enterprise Solutions, explained how these inconsistencies occur when radar is only sometimes used, citing the June 18 tornado that touched down near Denver International Airport as an example. Based on the lack of damage the tornado produced, it could have been rated as an F-0. However, according to Smith, this would have been misleading to future architectural projects for the area.
"Engineers and architects use the index data to design structures. If we omit radar measurements, critical structures could be mis-designed," he said. "Based on the METAR wind gust of 97 mph, the tornado should [have been] rated at least F-1, if we were to base the ratings on all reliable data. The instrument failed after that gust, implying the tornado may have been stronger."
"The way I read [the Doppler image], the red pixel inside the green is 150 mph. That would make it F-3. So, if this tornado had hit the terminal building, there would have been a lot of damage. If you use the image, the dark blue east-west line at the bottom is I-70. The airport property line is light gray," he said.
The National Weather Service gave the tornado a preliminary F-1 rating. Had they relied solely on surveys of damage and rated it an F-0, Smith believes that it would have been misleadingly low for future design projects.
"I favor using all reliable data in rating tornadoes," he said.
The Storm Prediction Center's Don Burgess explained the reason why only portable radar is being used when rating storms, stating that fixed radars aren't quite equipped to properly read tornado wind speeds.
"In general, to estimate wind speeds you need to be quite close," he said.
Burgess acknowledges that this can lead to discrepancies if only some storms are rated with radar being taken into consideration and said that there are meetings and discussions to try to find a solution to this problem.
"The first issue is that since portable radars don't collect data all the time and can only do it in open plains with no hills or trees, people are concerned that it could skew the database. It's a point that is being considered," he said.
He explained that the second issue is trying to figure out how to factor in this additional information.
"EF ratings are based on 3-second gusts at 10 meters high," he said. "Radar can pick up speeds quicker than that and higher than 10 meters."
The issue requires calculations and adjustments, but at this point, there is no standard policy for how this is done. It can typically take three months for a rating to be finalized once all data is put together. The El Reno, Okla., tornado from May 31 of this year is still waiting a finalized rating. There were high winds recorded in that tornado, which had multiple vortices. Considerations for how to include all of this information are still taking place.
Burgess believes that the best course of action may be to preserve damage ratings, but to add another column in the database to attribute additional information when it is available, be it from radar, anemometers or other measurements.
What the solution to these discrepancies will be remains to be seen at this time, but the concerns are being addressed.