Doppler Based Fujita Ratings

Several recent tornadoes have been given official ratings via radar scans, the practice causing a bit of a stir in the meteorological community. The massive El Reno, Oklahoma tornado was classified as an EF-5 based on a Doppler radar indicated wind speed of 296mph. The traditional way of estimating wind speed and intensity is via a damage survey along the tornado path (damage surveys originally rated the tornado at EF-3). Some feel that using radar will negatively impact the consistency of the current methods. Others think that radar should take the lead in future storms. My personal belief is that while damage surveys should still be performed on any tornado, using radar-based wind speeds when available to assign a rating is a good idea. Here is my main argument why.

The Fujita Scale has always been a past-tense measurement. When Dr. Ted Fujita first devised the scale in 1971, there were no Doppler radars to determine wind speed. The only way to estimate wind speed was through evaluating the damage caused by a given tornado. While reliable, the Fujita Scale also had a major drawback in that it assumed buildings of a given type were all constructed to the same standard. In several cases, tornadoes were downgraded when it became apparent that builders took shortcuts like not anchoring roofs or loosely attaching homes to concrete slabs. The more recent Enhanced Fujita Scale uses more thorough analysis of damage and accounts for differing levels of workmanship in construction. Since it is based on the original Fujita Scale, it too is only applicable after a tornado has destroyed something.

Imagine if hurricanes had no ratings and no wind speeds given until after they destroyed a coastline. It would undoubtedly be much harder to prepare for them. This is the problem facing residents and government officials who deal with tornadoes on a regular basis. The Fujita Scale is determined using the opposite method of the hurricane Saffir-Simpson scale. The wind speed dictates what category the storm falls into, giving forecasters, community planners and emergency response teams the information they need to make wise decisions. Of course a primary difference is that a hurricane’s large size and slow progression means wind speeds can be monitored by multiple radars, satellites and airplanes.

The biggest benefit in my view to assigning ratings based on wind speed relates to building construction. The National Weather Service states that only 1% of all tornadoes are at or above the EF-4 level. This is true only when counting tornadoes that were surveyed for damage. The ones that were not surveyed or went unseen by either radar or spotters do not count. With no damage survey, we do not know how powerful they were. A tornado in an open field can easily have EF-4 level winds but only do EF-2 levels of damage to trees and crops. This technical loophole gets repeated year after year and leads to an underestimate of average tornado intensity and has dire effects on building strength and codes.

The devastation in Moore, OK shows how violent tornadoes can be. While not many storms reach that level of potency, the question is how do designers build to provide adequate shelter when faced with winds of 150mph, let alone 250mph. Currently in the most tornado prone parts of the country, the minimum national standard for residential home wind resistance is a 3 second 90mph gust. Individual states and cities may have higher standards but this is still somewhat alarming. This design standard is not even enough to protect against a slow moving EF-1 storm. The 3 second gust ratings also do not include loss of structural integrity due to flying debris impacts. Repeated impacts can open holes in walls, break windows and weaken the structure to the point of failure, even with the sustained winds far below the gust value.

The design standards along the south Atlantic and Gulf coasts on the other hand are much higher, up to 150mph in some areas. This is in part due to the way that hurricanes behave and are monitored. With a hurricane strike, a large number of houses are guaranteed to encounter powerful sustained winds. The historical data backs up this assumption with detailed information on wind field size, intensity and duration. Adhering to a higher standard is therefore not just the moral thing to do, but it makes economic sense. People can survive inside a properly rated structure and repairing a slightly damaged structure is much less expensive than rebuilding completely. With tornadoes, the area affected is far smaller but the intensity is concentrated. By designing to the minimum standard, builders take the gamble that a significant tornado will not strike. Most of the time, this is true. When it is false, it is devastating to these buildings and those seeking shelter within their confines.

The main drawback is the lack of Doppler radar coverage at ground level in many locations. Unless a storm passes fairly close to a stationary radar, it is very hard to get accurate data. There are several mobile Doppler trucks used by research organizations but they are relatively rare (the El Reno storm was scanned by a Doppler On Wheels). These mobile units are the most accurate when it comes to scanning the extreme low levels of the atmosphere. With high risk weather forecasts usually giving at least 2 days notice, positioning a mobile radar into a region where severe weather is likely is possible. However, getting the radar into a clear position to scan a tornado within a given storm in that region is still a challenge.

Getting a better understanding of the life cycle intensity of tornadoes is very important. Without this data, structures will continue to be built to a much lower standard putting people’s lives and property at risk. Accurate wind speeds have helped coastal regions enhance the building standards for their most threatened areas. There is no reason that the residents in Tornado Alley should not have the same protection.


About Christopher Williams

It's easier to lie about being boring than it is to be honest about being extraordinary.
This entry was posted in Technology and tagged , , , , , , , , . Bookmark the permalink.

5 Responses to Doppler Based Fujita Ratings

  1. Doreste says:

    “Currently in the most tornado prone parts of the country … Individual states and cities may have higher standards but this is still somewhat alarming.”
    [You’re right on! This is very alarming, since being able to resist 3 second 90 mph wind gust isn’t equivalent to resilience in an EF-2 (86mph-110pm)tornado if you take into account the base pressure of tornadoes.]

    “Repairing a slightly damaged structure is much less expensive than rebuilding completely”

    “While reliable, the Fujita Scale also had a major drawback in that it assumed buildings of a given type were all constructed to the same standard”
    [Though this statement is true, it is beside the point. We don’t use the Fijita (F) scale anymore, we use the Enhanced Fijita (EF) scale which like you said does take into account the construction type of individually rated structures. In addition, the EF scale uses more damage indicators than did the F scale. To point out this flaw in the F scale while trying to make your point that mobile Doppler radar is a better means of grading tornadoes that the EF scale is a bit misleading.]

    “Imagine if hurricanes had no ratings … Of course a primary difference is that a hurricane’s large size and slow progression means wind speeds can be monitored by multiple radars, satellites and airplanes.”
    [I’m confused to why you included this paragraph comparing the grading of hurricanes to tornadoes, ignoring probably the biggest factor, speed of formation of a tornado, until the end. Hurricanes and tornadoes are not comparable in their rating methods for that same reason. If we did use mobile Doppler radar to rate tornadoes, are you asserting that it would be easier for people to prepare for them. That’s preposterous! Tornadoes form within minutes and after touching down are gone like that! Additionally, they are just too predictable.]

    [So though I agree that mobile Doppler radars would be a better means of rating tornadoes, I do not agree with your method of argumentation and I think you may be overstating the benefits of this technology.]
    Please do not edit this post. Either post the complete comment or do not post it at all, Thank you,


    • Great comments!

      The Fujita and Enhanced Fujita scale were both designed to estimate tornado wind speed. Now that we have the ability to sample winds with radar, we should use it whenever possible to supplement the damage assessments. We should never give up surveying the debris field for the simple fact that there are always clues on how to build safer, stronger structures if we look for them. Additionally since radar can’t follow a long track tornado for its entire life and get uninterrupted coverage, the survey is critical in determining intensity over the entire path. But radar is undoubtedly more precise (accuracy depends on the operator and the conditions) and can tell us things that we previously had to make an educated guess at.

      The comparison between two completely different types of storm systems is on purpose. My reason for bringing up hurricanes is that if they were rated the same way tornadoes have been historically, higher gust standards wouldn’t exist in regions prone to tropical events. We have had the ability to sample hurricane winds for a long time and that data history gives a more accurate picture of the average and maximum winds a region can expect to see within a given time frame (10 years, 50 years, 100 years, etc). Tornado formation is so rapid that having the proper sensors in place to record wind speed has not been practical until recently, and then only in limited situations. My concern is that this has led to an underestimation of wind potential over large regions of the country simply because we don’t know how bad it can be or how often. Builders in hurricane prone regions can’t say “It doesn’t happen here so we don’t need to anchor the roof or reinforce the structural walls”. There is too much solid evidence to go against. In Tornado Alley, all of the data is basically second-hand, obtained by surveys. While important to meteorologists and scientists, the sad reality is that economics don’t put much faith in damage surveys. The “less than 1% of tornadoes are EF4 or greater” argument is probably uttered many times by developers when asked about how safe new homes are from extreme winds.

      My major argument was about structural integrity, not warning time. As you probably know, the average lead time for tornado warnings is 13 minutes. This is a very short time to prepare if one has to search for shelter. Many houses in the southern plains and gulf states do not have basements (I live in north Texas and basements are very rare). This means that people have to either hope for the best in a bathroom or closet, or try to make it to a location that does have adequate shelter. This is why so much attention is given to increasing tornado warning time. If on the other hand a house was built to higher standards and could safely withstand at least EF2 winds (vertical components included), people could confidently shelter in place without having to worry about the place being torn to shreds. Ideally, a dedicated safe room or small underground shelter accessible from inside should be included to provide maximum protection. But knowing the basic structure could stand up to over 80% of all tornadoes (EF0-EF2) would be a massive relief to those who live in those homes.

      In no way am I advocating sending out DOW6 to transmit real time wind data. We’d end up with every OCM in the warning box telling people that the tornado headed for Wichita is an EF1. Citizens would interpret that as “Oh it’s weak so I can go about my business.” Then if the storm intensifies to an EF4 and kills 300 people who should have been taking shelter, you can guess what the ramifications would be.

      • Doreste says:

        “My major argument was about structural integrity, not warning time.” [I’d say you’re right about the importance of structural integrity. I was a bit confused about your statement,] “With high risk weather forecasts usually giving at least 2 days notice, positioning a mobile radar into a region where severe weather is likely is possible.” [But now, when I reread that statement it seems you are saying the 2 days notice would be so the DOW could be station in the right place, not that the people of a particular town would have 2 days to prepare.]
        Right now, I’m still researching to figure out if using mobile Doppler radar is a reliable way to rate tornadoes, especially because of how the recent El Reno tornado and how the May 24, 2011 El Reno tornado was rated. I just happened upon your blog post. If you are interested, here’s a report I’ve been mulling over. It actually documents the use of damage surveys in conjunction with mobile Doppler radars over the span of the tornado’s life.
        Thank you for replying so quickly,


      • That paper brought up a lot of the technical aspects of operating a Doppler near a tornado that most people probably would never consider. The issue of not getting wind speeds below 30 meters is a problem since the wind at the surface is what structural engineers want to know. Of course, if we’re discussing high rise hotels or office buildings, that 30 meter sample elevation will be useful data. It’s clear that using radar is not just a point-n-shoot activity and scientists are required to fully decipher the information. Once they get reports written, hopefully it will be concrete enough evidence to spur developers into strengthening houses in tornado prone areas. Joplin, MO has changed their city building code to require roof clips and foundation anchors on all newly constructed buildings

        The scary part in that article is the damage survey’s analysis that 83% of the winds were EF-2 level or less. We can have all the warning time in the world but if houses are going to disintegrate in 130mph winds, what’s the point of sounding the alarm? Structures in the rest of the country need to be more resilient to wind damage (not just tornado but RFD, derecho, gust front, downburst, etc).

        I’m guessing from your email that you’re one of the Wind Hazard researchers at the University of Florida?

  2. Doreste says:

    Well, I’m a very new addition to the team, but yes.

Go on...I know you want to say it!

Fill in your details below or click an icon to log in: Logo

You are commenting using your account. Log Out / Change )

Twitter picture

You are commenting using your Twitter account. Log Out / Change )

Facebook photo

You are commenting using your Facebook account. Log Out / Change )

Google+ photo

You are commenting using your Google+ account. Log Out / Change )

Connecting to %s