2013 El Reno Tornado Damage Survey

Image of vegetation damage near the location where the highest doppler velocities were recorded more than 400ft above the ground.

Image of vegetation damage near the location where the highest doppler velocities were recorded more than 400ft above the ground. The tornado was given an EF5 rating less than 36 hours after dissipating solely due to mobile doppler radar velocities between 290 and 336mph, possibly the strongest ever recorded (AMS, 2013). The tornado was later downgraded to an EF3 due to a lack of EF5 damage indicators.

□ On May 31, 2013, a train of violent supercell thunderstorms erupted in the sky to the west of Oklahoma City. In the city suburb of Moore, rain began to fall over the tangled remains of homes and businesses obliterated less than two weeks earlier by a catastrophic EF5 tornado. Meteorologists at the National Weather Service recognized the tell-tale signs of a rotating mesocyclone on radar, and reports from storm chasers near the community of El Reno quickly verified the presence of a large, nebulous mass of clouds that had spun down to the ground. A rare tornado emergency was issued for the area a second time.

An unusually large number of storm chasers, both amateur and seasoned, drove down the perfect grid of county roads to the south of El Reno to film the ensuing storm. Like many large tornadoes, the El Reno storm began with a series of transient funnels beneath a rapidly rotating mesocyclone. Heavy rain left few good angles to film the storm, so most chasers concentrated near Highway 81 just northeast of the tornado. As the storm progressed slowly to the east-southeast it underwent a period of explosive strengthening. The already large tornado suddenly doubled in size in less than one minute (many say less than 30 seconds) to over two miles in width. Footage from multiple storm chasers showed a sudden increase in surface winds well away from the visible funnel. Near the intersection of Choctaw Avenue and SW 15th Street, two vehicles were engulfed by the storm, killing both drivers. One of the men killed, Richard Henderson, became the first amateur storm chaser fatality in history.

Storm chaser Richard Henderson sent this photograph to a friend several minutes before he was killed. Henderson was on the phone with the same friend

Storm chaser Richard Henderson sent this photograph to a friend several minutes before he was killed by the tornado on SW 15th Street. The same friend was on the phone with Henderson when debris began to strike the chaser’s car. Moments later, the line went dead (Kelly, 2013).

Trees along SW 29th Street were damaged and a piece of timber pierced 11 inches (measured) into the ground.

Trees along SW 29th Street were damaged and a piece of timber pierced 11 inches (measured) into the ground. The distance of visible damage along Highway 81 was 2.5 miles, although extreme straight line winds blurred the distinction. All of the damage visible from Highway 81 was in the EF0 to EF2 range, although wide spacing between buildings and trees left few reliable damage indicators.

A Weather Channel vehicle driving on Highway 81 was impacted by a violent wind feature, causing it to tumble through an adjacent field (initial reports stated the car travelled 200ft while later broadcasts reported 200 yards). All of the vehicle’s passengers were injured to some degree but most were able to walk away from the wreckage. Around this time, the tornado was completing an unexpected turn to the northeast. Highly knowledgable storm chasers were caught off guard by the storm’s size and unpredictability. A motorist due east of the Weather Channel crew was killed by the tornado at the intersection of Alfadale Road and Reno Street. Along SW 10th Street, a white car carrying three professional storm chasers was swept off the road west of Radio Road. Unlike the Weather Channel vehicle, the car driven by Tim Samaras was hurled 650 yards through the air at a high rate of speed (AMS, 2013). The three chasers, likely with cameras in hand, were all killed in the “unsurvivable” wreck. Several minutes later, the tornado swept across the I-40, killing a young mother and her infant son in a car hurled from the freeway. Forty minutes after first forming, the tornado weakened and dissipated.

I surveyed the damage from the El Reno tornado and interviewed local residents on June 4th and 5th. Photographs from my survey are shown below, and my final analysis can be found at bottom.

*In August of 2013, the El Reno tornado was officially downgraded to an EF3 by the National Weather Service (Querry/NWS, 2013).

A home just south of SW 15th Street along highway 81 experienced EF1 damage despite being near the geographic center of the storm. The tornado's multi-vortex nature meant that most of the damage swath encountered winds in the EF1 and EF2 range.

A home just south of SW 15th Street along Highway 81 experienced EF1 damage despite being near the geographic center of the storm. The tornado’s multi-vortex nature meant that most of the damage swath encountered winds in the EF1 and EF2 range. The tornado’s first two fatalities occurred in two vehicles a mile west of Highway 81 along SW 15th Street (KFOR, 2013).

Wheat crop was blown to the ground in swaths approximately 100ft wide along SW 15th Street.

Wheat crop was blown to the ground in swaths approximately 50ft wide a half mile east of Highway 81 on SW 15th Street.

Fence posts adjacent to homes that experienced EF1 and EF2 damage were generally only lightly damaged. In some areas, they were bent to the ground or removed entirely. In some instances, metal stakes were twisted in various directions.

Fence posts adjacent to homes that experienced EF1 and EF2 damage were generally only lightly damaged. In other areas they were bent to the ground or removed entirely. In some instances, metal stakes were twisted due to bursts of winds from various directions.

A home on XX Street suffered severe internal damage but was left largely standing, like most of the homes in the area.

A home on Alfadale Road suffered severe internal damage but was left largely standing, like most of the homes in the area. One survivor who was staying with a relative on SW 29th Street said that the tornado “lasted about five minutes, but the worst of it happened in the first 30 seconds when every window shattered at the same time.”

A home on Reno Road experienced EF2 damage.

A home on Alfadale Road experienced EF2 damage. A motorist in a vehicle at this intersection became the storm’s third fatality.

A blizzard of dry plant materials was blown into standing fences along Reno Road.

A blizzard of dry plant materials was caught by standing fences along Radio Road.

Vegetation just northeast of the intersection of Radio Road and 10th Street, where the most intense winds were recorded by mobile doppler radar.

View northeast at the intersection of Radio Road and 10th Street, where the most intense winds were recorded by mobile doppler radar. While analysis is ongoing, velocities between 290 and 336mph were recorded as a single, exceptionally powerful sub-vortex slingshotted around the south side of the tornado at 177mph. The peak winds occurred on the eastern edge of the vortex where all of the rotational velocities combined. While of record intensity, the peak winds were found 110 yards above the surface and would have impacted a standing structure for only half a second (AMS, 2013).

10th and Radio Road was marked by a sign placed after the tornado to direct local traffic.

10th and Radio Road was marked by a sign placed after the tornado to direct local traffic.

The only tree near the intersection of 10th and Radio Road was 80ft to the east on 10th Street. The tree was stripped of leaves in a fashion consistent with winds in the EF3/EF4 range.

The only tree near the intersection of 10th and Radio Road was stripped of leaves and damaged in a fashion consistent with past EF4 tornadoes.

Vegetation was blown to the ground and strewn with small pieces of debris, including fragments from one white and one red vehicle. Metal fence posts were bent to the east-northeast or absent entirely from the ground.

Vegetation was blown to the ground and strewn with small pieces of debris, including fragments from destroyed vehicles. Metal fence posts were bent to the east-northeast or absent entirely from the ground. The gravel along 10th Street was blown almost completely away in areas affected by sub vortices. This is where a vehicle driven by Tim Samaras was recovered.

A half mile northeast Radio Road and 10th Street a fence was ripped from its posts and left in a tangled mass in a field.

A fence was ripped from its posts and left in a tangled mass a half mile northeast of 10th and Radio Road.

A white truck driven by iconic storm chaser Tim Samaras, his son, Paul Samaras, and chase partner, Carl Young.

A vehicle carrying storm chasers Tim Samaras, Paul Samaras and Carl Young was swept off 10th Street approximately 545 yards west of Radio Road. The car was thrown more than 650 yards, killing the three occupants (AMS, 2013). According to ABC News, Tim Samaras was found in the front seat whereas the other two passengers were ejected from the vehicle and found up to a half mile away. DOW observations at the time indicated that an intense sub-vortice with a forward speed of 177mph impacted the vehicle unexpectedly from the north (AMS, 2013). (Image by Jason Morris)

Area on 10th Street where the chase vehicle containing three occupants was found (note the trees in the background).

Area near 10th Street where the chase vehicle was swept off the road.

Approximate location where the chase vehicle driven by Tim Samaras was found.

Flattened wheat crop just south of 10th Street.

Tree damage near where the chase vehicle was recovered. Trees 100ft to the north on the other side of 10th Street were damaged but not defoliated. Flooding carved deep ditches into areas that were previously crop fields.

Tree damage near where the chase vehicle was overtaken by the storm. Trees 100ft to the north on the other side of 10th Street were damaged but not defoliated. Flooding carved deep ditches into areas that were previously flat fields.

A quarter mile south of the I-40, an RV park was severely damaged in the tornado. A vehicle from the facility was thrown several hundred yards. Most of the powerlines along XX Street were snapped above the ground.

A quarter mile south of the I-40, an RV company was severely damaged in the tornado. A vehicle from the facility was thrown several hundred yards. Most of the powerlines along N2880 Road were snapped just above ground level.

The most severe vegetation damage appeared to be along the I-40, north of where the most intense winds purportedly occurred.

Some of the most severe vegetation damage appeared to be along the I-40, north of where the peak doppler velocities purportedly occurred.

A livestock trailer was blown 300 yards to the south from a complex across the highway and left tangled in the remains of a fence.

A livestock trailer was blown over 300 yards to the south from a complex across the freeway and left tangled in the remains of a fence.

View of the XXX where the livestock trailer originated.

View of the destroyed OKC West Livestock Market where the trailer originated.

Debris in a wind damaged field just south of the I-40.

Debris in a wind-damaged field just south of the I-40.

Tree damage just south of the I-40 near the OKC West Livestock complex.

Tree damage just south of the I-40 near the OKC West Livestock complex. The tornado’s massive size and slow movement would have exposed many areas to tornadic winds for greater than five minutes, though peak velocities occurred only in suction spots.

Extremely heavy rainfall on the day of the tornado led to widespread flooding in the affected areas. Visible here is the OKC West Livestock Market and a pool of water that trapped large pieces of debris, including sections of broken powerlines.

Extremely heavy rainfall on the day of the tornado led to widespread flooding in the affected areas. Visible here is the OKC West Livestock Market and a pool of water that trapped large pieces of debris, including sections of broken powerlines.

Closer view of the destroyed livestock complex. Dozens of large animals were killed throughout the tornado's path, leaving the smell of rotting flesh as the days passed.

Closer view of the destroyed livestock complex. Dozens of large animals were killed throughout the tornado’s path, leaving the smell of rotting flesh several days later.

Heavy tree damage just north of OKC West.

Heavy tree damage just north of OKC West.

Deep impact mark on a hillside just north of the I-40. Several vehicles were swept from the freeway in this area, resulting in three fatalities in two vehicles.

Deep impact mark on a hillside adjacent to the I-40. Several vehicles were swept from the freeway in this area, resulting in two fatalities.

Personal Damage Survey Conclusions:

Due to my belief that the Enhanced Fujita Scale grossly underestimates winds in violent tornadoes, the wind ranges I utilize are based on my research, discussions with wind engineers and comparisons between known surface readings and adjacent damage indicators in past tornadoes. My wind estimates are significantly higher than those employed by the National Weather Service.

Peak Intensity: EF4 (≈220mph)

□ The most intense damage occurred to vegetation in a swath of varying width from an area commencing just west of the intersection of 10th Street and Radio Road and ending at the I-40. Steel fence posts were bent to the ground by winds alone and not debris impacts. Powerlines were sheared just above ground level, vehicles were thrown over 200 yards and surface crops were severely damaged and bent to the east-northeast. The most intense structural damage likely occurred to the OKC West Livestock Market just north of the I-40 and several homes west of Highway 81. The most intense structural damage was consistent with winds in the EF3 range. The scarcity of trees and buildings left few reliable damage indicators, but no EF5-level vegetation damage was noted.

□ While the tornado has been deemed the “second strongest” in recorded history due to extreme doppler velocities, this claim is completely unfounded. Few violent tornadoes are ever tracked by mobile doppler radar, so the available readings are not an objective method of classification.

□ While the tornado is being called the “widest” in history, this claim is also unfounded. The 1999 Mulhall tornado likely left a significantly wider damage path. In terms of violent tornado damage, the damage swath from the 2013 El Reno tornado was narrower than many documented tornadoes, including the 2011 Joplin tornado.

The width of EF0+ damage along Highway 81 was 2.55 miles in width, as measured by the distance between the northernmost and southernmost instances of missing shingles and downed tree branches. The southern margin was 300 yards south of SW 29th Street and northern margin was just north of 10th Street. Straight line wind damage was found in areas unaffected by the tornado, so the exact damage contours were impossible to ascertain.

The vegetation damage from the El Reno tornado was noticeably less intense than the damage caused by the 2013 Moore tornado.

The vegetation damage from the El Reno tornado was noticeably less intense than the damage caused by the 2013 Moore tornado. Few trees were debarked in El Reno, whereas all of the trees in the core damage path of the Moore tornado near County Edge Drive were completely stripped of bark and branches.

75 thoughts on “2013 El Reno Tornado Damage Survey

  1. Good stuff Max, thanks for sharing those photos. I very much agree about the “strongest” and “widest” claims. Even if it’s technically true, it holds no value at all. Would you care to share your personal wind speed ranges?

    Also, just one thing. Tim was actually driving one of the mobile mesonet vehicles, a Chevy Cobalt. Apparently someone else (maybe Tony Laubach?) was in their box truck further northwest.

    • Oh, I knew Tim wasn’t in his normal chase vehicle but was still under the impression it was a truck.

      As for my wind speed ranges, do you mean in regards to my version of the EF-scale? There’s certainly overlap but I’d say EF4 tornado damage is associated with winds between 220 and 260mph, and EF5 damage is associated with winds in excess of 260mph. This wouldn’t be the same with the F-Scale due to the less stringent restrictions imposed before 2000.

      The worst damage from the El Reno tornado was right beneath where the doppler velocities were recorded (I walked a half mile out into the fields) and just south of the OKC West Livestock complex.

      • I am wondering if you can elaborate on why you think that the Enhanced Fujita scale underestimates wind speeds in violent tornadoes. My understanding is that the original Fujita scale gave wind estimates based on straight-line winds, as opposed to cyclonic winds (which would be more appropriate), thus overestimating the intensities of violent tornadoes. The EF scale took cyclonic winds into account and therefore lowered the wind estimates to more accurately reflect the actual wind speeds necessary to cause the damage that is left behind. Many other people have made the same claim that the EF scale is too inaccurate, but the revisions make sense to me. I’ve just never heard what the other side of the argument is.

      • John – I’ll have to seriously condense my argument, but my main reason for believing the EF-Scale underestimates wind speeds in violent tornadoes is that every single surface wind measurement that has been taken within a tornado has been significantly higher than what would be expected from the damage using the EF-Scale. For instance, a gust of 151mph was measured on the ground in the 2011 El Reno tornado near the EF2 damage contour, whereas much, much more intense damage occurred near the center of the storm.

        This year, the TIV2 recorded wind gusts of 175mph before the anemometer failed. The winds appeared to increase significantly following the failure, I’d imagine to a little over 200mph, and remained at that strength for nearly two minutes. Following the Jarrell tornado in 1997, some tried to say that winds of 160mph could have caused the catastrophic damage over the course of three minutes. Yet the winds around the TIV2 didn’t cause any sort of ground scouring and were probably capable of causing low-end EF3 damage, much like the areas in Joplin unaffected by suction vortices.

        In turn, doppler velocities have almost always been more consistent with the wind ranges in the original F-Scale.

        And hurricanes never cause debarking or any of the other “extreme” damage indicators common in intense tornadoes. While the differences in wind direction, acceleration and debris load are significant, the fact that 200mph wind gusts over a long time frame (Hurricane Andrew, Cyclone Zoe) failed to cause any of these incredible phenomena indicate, in my mind, that tornadic winds are significantly stronger.

        If, in fact, 160mph tornadic winds can scour homes, roads and the ground to a depth of 18″ (Jarrell) then it is futile to compare tornadic winds to any other type of high wind event as their unique properties result in damage that could barely be attained by straight-line winds in excess of 300mph. If this is the case, then using the “mph” system should be abolished with tornadoes as it creates a false sense of comparability.

      • I see your point, but I’m still not sure why this matters. The EF-Scale ranks a tornado’s intensity based on the damage it causes. Wind estimates are used to determine the degrees of damage for each damage indicator, but even if the wind estimates were completely off, the methods used are standardized meaning they should be off by the same amount for each tornado analyzed. If I am correct about this then the accuracy of the wind estimates don’t actually matter (as long as they are consistent). If you equate the intensity of a tornado with the damage it inflicts (as the EF-Scale does), then it shouldn’t matter what the actual wind speeds are. Yes, I think there is ample evidence that the wind estimates in the EF-Scale are lower than they should be, but what I am wondering is: why is that an issue? You can argue whether a tornado requires winds of 140mph or 200mph to rip the roof off of a house, but survey teams looking at the house will classify the tornado as an EF3 either way.

        Now, I think your article on the Tuscaloosa tornado is interesting because it clearly shows how two different survey teams can come up with different EF ratings for the same damage. Here is where I think the EF-Scale is still lacking. Degrees of damage for each indicator is still based on one’s perspective on just how extensive the damage is. What exactly constitutes a home of “superior construction”? However, I don’t think the original F-Scale has the upper hand here because some homes are more solid than others. It had to be the case that the F-Scale was overestimating the intensities of some tornadoes. I think the EF-Scale is on the right track, but it’s not perfect. But human error and vague damage indicators aside, I’m still unsure why there is so much mistrust associated with it.

      • Forgot to mention- regarding ground scouring, my understanding is that there is no determined relationship between scouring and wind speed, and that scouring is likely the result of debris sand-blasting the ground. If this is the case, then you wouldn’t expect ground scouring in either the TIV2 tornado intercept or the El Reno 2013 tornado since there was little debris in both events. Also, I wouldn’t think that the EF-Scale is comparable to other high wind events anyway since those are based on precise wind measurements whereas the EF-Scale is primarily based on damage (I am concerned about the El Reno 2013 rating since it was derived by radar measurements- this is where problems with the EF-Scale wind estimates would be applicable).

      • I agree with you when you say the winds “don’t really matter.” I also believe the EF-Scale is a large improvement over the F-scale. In the case of the El Reno tornado, however, it was rated an EF5 solely due to doppler velocities. Until the NWS has a better understanding of exactly how the doppler velocities equate to surface winds, I don’t think they should rate a tornado an EF-5 based on doppler velocities (or if they do, be consistent about it). I’m waiting for the CSWR to release more information on their observations.

        And having surveyed nearly a dozen violent tornadoes, I can say with extreme confidence that ground scouring is not the result of debris impacts. Maybe sandblasting plays a part, but ground scouring occurs regularly in the exact same conditions I photographed in El Reno. Generally the more rural an area the more intense the ground scouring, whereas plenty of houses and buildings slow the surface winds and reduce ground scouring.

      • I have also found ground scouring and tree damage to be the most reliable indicators of tornado intensity other than damage to well-built homes. I mention in another post that ground scouring does seem to happen more readily in some areas (Oklahoma City/El Reno area being especially vulnerable to scouring) but many surveyors, such as Marshall, refuse to give an EF5 rating unless there is scouring/extreme tree damage. I strongly agree.

      • John, regarding what constitutes a home of “superior construction,” It seems surveyors have differing ideas what qualifies as the minimum, but ideally a home would have anchor bolts with adequately wide washers spaced evenly at least every few feet and embedded at least seven inches into the foundation, as well as hurricane straps and that sorta thing. Just as important is the surrounding context — ground scouring, low-lying shrubs debarked/shredded, etc.I don’t think ground scouring by itself is evidence of an EF5 tornado, but it’s very likely an indicator of EF4+.

        Max, while we don’t know exactly how those winds translate, everything I’ve read and been told indicates max velocities aren’t significantly different at 10m than at, say, 50 or 100m. It seems the structure and duration of the maximum winds are more sensitive to height than the max velocity. RaXPol also found velocities of 125+ m/s on all tilts down to 0 deg. I think we need more research, but I think mobile radar velocities are as reliable (and probably moreso) than damage.

      • Shawn – If this tornado had passed over a grid of well-built homes and failed to leave any EF5 indicators (which I think would have been the case, though I think high-end EF4 damage would have occurred) the doppler velocities should come second to the actual damage.

        Like I said, I was surprised to see the El Reno tornado caused less intense damage than the 2013 Moore tornado. Both tornadoes passed over similar vegetation in rural areas.

  2. Amazing article, EP. I’ve been anxiously awaiting your survey pics from the El Reno tornado. I am surprised you didn’t find more intense ground scarring. I imagined the fields would have been completely stripped bare.

    I sent you an email to see if I could see some more photos from your survey.

  3. Fascinating analysis, this and all of your articles. I’ve spent the past two days reviewing all of your work. I’ve subscribed.

  4. Excellent analysis – I’d have to agree that I didn’t see any really intense EF5-level damage even in the most intense part of the tornado; I also heartiy agree that the EF scale vastly underestimates wind speed. Perhaps the original F scale was closer to the correct speeds for higher categories? I’d like to see more tornadoes measured to compare this one with others.

    • I think the original F-Scale was almost spot-on, actually. The doppler velocities recorded in violent tornadoes during the 90’s (Spencer SD, Bridge Creek OK, Dimmitt TX, Mulhall OK) correlated almost perfectly with the original F-scales wind ranges. Oftentimes, they were even a bit higher than would be expected.

  5. This tornado was rated a high-end F3 on the original Fujita scale. http://tornadostormshelters.com/Westminster.htm. Based on the ground scouring associated with this tornado do you think it could have easily been rated at least an F4 on the original F-scale or even EF4 on the EF-scale. The damage photos from this tornado seem to be even more impressive than the ground scouring done in the El Reno tornado.

    • Shane – I talk a bit about the Westminster tornado in another post. Had I surveyed the damage myself I would have gone with an F5 rating based on the severe ground scouring, large homes swept completely away and vehicles mangled beyond recognition and thrown over 200 yards. The Westminster tornado was definitely stronger than the El Reno tornado, IMO, and probably comparable to the 2013 Moore tornado.

      • Max, do you factor in soil type when you consider scouring? Scouring seems to happen more readily in central Oklahoma, for example, where the soil tends to be loamy and often a little more loosely packed. I think it’s generally still an indication of a violent tornado, but it’s got to take less to scour that sort of soil compared to heavy, densely packed soil. If memory serves (it often doesn’t), I think the soil in the blackland prairie area is mostly tough, compacted clay. That’d certainly lend credence to Westminster being a high-end tornado.

      • I don’t think there is any standardized way of factoring in soil type being that the ground composition can vary significantly within a single square mile, and month to month based on precipitation. I have noticed scouring appears to occur more frequently in sections of the Great Plains, but the vegetation and slower forward speed of the tornadoes may be the primary factors.

        Unless someone has done a damage survey and analyzed the soil it’s impossible to compare, for eg., “the Red River Valley” to “northern Alabama.”

  6. Thank You,

    I always believed the Westminster tornado was definitely much stronger than the rating it was given. The ground scouring reminds me not only of the recent Moore tornado but also the ground scouring done in the Bridge Creek/Moore/OKC tornado from 1999. Of course I think the Westminster tornado was on a smaller scale and that was because it was not quite as wide as them two tornadoes and it also went through more rural area. It also seemed definitely stronger than the recent north Texas EF4 tornado in Granbury. I honestly would have went at least with a solid F4 rating for the Westminster tornado and like you said an F5 rating doesn’t seem out of the question.

  7. Sorry to bother you again but I wanted to let you see a youtube video in regards to the Westminster tornado. The reporter is standing on a SLAB OF CONCRETE of which used to be a fairly good sized home. The whole house was completely swept away and even some of the fixtures looked like they were torn off. IMO this tornado was one of the most grossly underrated tornadoes I have ever heard of. Here is the video. http://www.youtube.com/watch?v=_uVqSJoTuIY. If that house doesn’t qualify for at least an F4 rating then I don’t know what does.

  8. This is such excellent reporting and analysis. It was horrifying to read as you unfolded the story, knowing the fatalities.

  9. Max,
    When you look at the vehicle that was carrying Tim Samara’s and the other two storm chasers, what kind of winds would you estimate it would take to do that to a car? The car does look very bad off but it does not appear to be destroyed beyond recognition. I would estimate myself to be at least EF3 but that is only my guess.

  10. I was reading Max how you believe the EF-scale underestimates the winds associated with violent tornadoes and you believe the F-scale is almost right on. I do tend to wonder that myself. Lets just say that EF-scale is correct with wind speeds associated with violent tornadoes. The DOW clocked winds at 295mph at about 400-500ft above the ground. Could it be possible that the winds at the surface may have been much weaker than those clocked several hundreds of feet above the ground? I know the Bridge Creek/Moore/OKC tornado had winds clocked at 318mph some distance above the ground and it did do intense F5 damage in a number of areas. That tornado had winds pretty consistent above the ground as it did at the surface. Is it possible that some tornadoes may have a big difference in winds clocked above the ground than winds at the surface like this El Reno tornado? While some tornadoes may have very little difference in wind speeds above the ground or at the surface. I don’t know much about this being possible but maybe you can answer.

    • There really isn’t enough information to know as surface readings in tornadoes are almost non-existant – particularly since the winds in tornadoes vary widely due to subvortices and other wind features.

      I believe, however, that the winds in tornadoes follow the same general principles as hurricanes. This would mean that for most tornadoes, the winds at the surface are a bit less intense than the peak winds between 100 and 500ft. With especially strong hurricanes, however, the maximum winds reach all the way to the surface, and sometimes even exceed flight level winds. If this were the case in tornadoes, then the surface winds in EF5’s could be every bit as strong as they are higher up.

      But it’s a lot more complex than that. In some tornadoes there might be transient vortices that bring extreme winds to the surface for only a few seconds. Others might have vortices that develop in other parts of the funnel that fail to reach the ground (as may have been the case in El Reno).

  11. I know this an extremely unlikely case scenario and might even be next to impossible. I believe a funnel cloud must make contact with the ground in order for it to be classified as a tornado. If a funnel cloud went over the top of a large skyscraper and did significant structural deformation to the top of the building but failed to make contact with the ground what would it be considered. I know very little about tornado formation and tend to wonder if its even possible.

    • If it caused damage it would be a tornado. And if damaging winds were to get down to within 1,000ft of the surface then there would certainly be some winds on the ground as well.

  12. What about the Rozel and Bennington, Kansas tornadoes? Those two tornadoes were upgraded to EF4 based on the DOW. The one where was another tornado you described where Sean Casey used some kind of instruments that measured 175mph wind gusts before being destroyed. You claimed it may have gotten a little bit stronger than that though. That tornado was never upgraded it remained an EF3. The Rozel tornado did not seem that impressive as far as the damage goes to indicate an EF4 but the Bennington tornado may have indicated EF4 intensity. I could be wrong about the Rozel and Bennington tornado but I don’t see anything that was that impressive damage or ground scouring wise.

    • I haven’t done much research on those tornadoes so I haven’t formed a strong opinion, but I do believe the TIVII probably encountered winds near 200mph – or, assuming the instrument failed at peak wind intensity, 175mph winds for a solid 90 seconds.

      Considering that the vertical component of tornadic winds are not captured by traditional anemometers (as they hit the wind cups diagonally), the 175mph reading may have been closer to 190mph or a bit more at the time of failure.

  13. I don’t believe the NWS is going to downgrade the rating on this tornado. They seem to be way to overly impressed with the size and wind speeds recorded in this tornado. You said this tornado was probably capable of causing solid EF4 damage at most. Although you don’t believe it would have caused EF5 damage, do you think it would have killed a lot of people had it of hit a rush hour traffic bridge and would have went through the OKC metro?

  14. I don’t know if it is me or my lack of knowledge but so many things about this tornado make no sense. Not only did it have a very incredible width, it had some of the most insane radar velocities ever observed and some of the most extreme wind speeds ever clocked inside a tornado. This tornado may have stayed over the same area of land for over 5 minutes and failed to do any kind of damage that was extremely impressive. I would have expected a tornado like this to have done ground scouring very similar to the Jarrell or Harper tornado. If there really is such thing as E/F5 ground scouring the Jarrell and Harper tornado would take the cake. It looks literally like a large blow torch went across the ground. They leave burn marks, scar marks, and spiral marks all over the ground. Do you think the El Reno tornado failed to do this because the suction vortices were not touching the ground or something else that may have prevented it. I don’t know if it is because 2+ mile wide tornadoes may not be the strongest but I am not really sure about that either. The Mulhall tornado on May 3, 1999 was believed to have been possibly 3 to 4 miles wide and capable of doing a mile-wide continuous streak of E/F4+ damage. You have any thoughts on why this tornado failed to do some type of EF5 damage criteria Max.

    • I find it very confusing as well. While some comparisons are drawn between this and the Mulhall tornado, the Mulhall tornado caused EF2 to near-EF4 damage in the town of Mulhall, which was one mile away from the center of the tornado on the weaker left side of the storm. This tornado passed over two dozen homes scattered rather evenly throughout the tornado’s path, even in areas near where the highest winds occurred, and yet only caused high-end EF2 damage in areas east of the 81 and south of the freeway. The worst damage, even more strangely, occurred west of the 81 and north of the freeway, where the tornado was supposed to be weaker.

      I can’t say I understand why this occurred. Part of me believes, like you said, that the most extreme winds in the vortices (greater than 220mph on my scale) failed to reach within 10m of the surface.

  15. I do agree with you that rating a tornado based on DOW readings does seem concerning. What if we still used the original Fujita scale and wind speeds were clocked at 350 mph? Would such a tornado be rated F6? I have heard from others such as NWS employees, that Fujita’s sole purpose was to be able to rate tornadoes based on winds alone instead of damage. Is this true? Another thing why wasn’t tornadoes in the 90s rated F5 based on DOW reading, such as Red Rock, Spencer, and Mulhall? I also like how you believe that extreme ground scouring and complete tree debarking should be used just as much as the construction of a home when determining a tornadoes rating. I do agree that extreme ground scouring and complete tree debarking is an indication of a at least E/F4 intensity. Also what happened in 2005 and 2006 that there were so few tornadoes rated F4? It made no sense. There were a total of 3 F4 rated events from those two years. There are at least 2 tornadoes that deserved at least an F4 rating and that was Stoughton, Wisconsin on August 18, 2005 and Westminster, Texas on May 9, 2006 and possibly a couple others. I don’t know if this had to do with what happened in La Plata or if it was some other reason. I know I am no expert on rating tornadoes, it seems as NWS offices have contradicted themselves when rating a tornado. They’ll use DOW readings to up grade a tornado but they will not use extreme ground scouring or complete tree debarking. Even if home damage is not that great or not superior, extreme ground scouring and complete tree debarking is a very good indication of a violent tornado.

    • I don’t think there ever was an F6 rating, so it probably would not have ever been used. As for whether Fujita wanted the scale to be a wind scale alone, I’m not sure but it makes sense as that’s what was being done with hurricanes. Yet more research over the years has shown that winds in tornadoes are exceptionally complicated and the severity of the damage is impacted by more things than just the windspeed – such as duration, wind acceleration and upward velocity.

      The rating standards became a lot more strict in the early 2000’s right before the implementation of the EF-Scale. As for the method in which the doppler readings are used now – it seems as though the NWS goes with an EF5 rating only if the measured velocity is well into the EF5 category – probably somewhere in the range of 270mph+. As a result, I think the NWS would have still gone with an EF4 rating for the Spencer, Red Rock and Mulhall tornadoes – even though all three (especially Red Rock and Mulhall) were likely capable of causing EF5 damage

    • Many surveyors will not even discuss ground scouring because there is no official DI for it. I have brought it up other places, and most meteorologists will discount it entirely. Irritating, because if anyone were to do several years of intense studying of historic and recent tornadic events, the link between ground scouring and high-end events is obvious.

      With that said, soil type must be taken into account. In central Oklahoma, pronounced scouring seems to begin in the EF4 range, due to the loose and dusty soil type there. In the deep south, scouring seems to only begin well into the EF5 range, due to the dense, clay laden soil there.

    • Interesting, considering I can’t think of a single EF3/F3 event besides Westminster that produced severe ground scouring.

      • While some say ground scouring has no relevance in the rating system, the Philadelphia tornado was given an EF5 rating solely due to ground scouring. Home construction varies quite a bit more than the composition of vegetated fields throughout tornado alley and the South.

        Tim Marshall makes quite a point of examining ground scouring in his article on EF4/EF5 damage. From my personal experience surveying damage paths and examining thousand of damage photographs, I would go ahead and say it is the ultimate indicator of a violent tornado – particularly if the fields are covered in grass or dense vegetation.

        I’ve also noticed that ground scouring is less tied to wind duration than other damage indicators, such as home damage. Ground scouring only begins to occur at extremely high wind speeds (unless the soil is loosely packed or some other variable is at work) and is therefore evidence of intense instantaneous wind gusts.

  16. After the May 3, 1999 Bridge Creek/Moore/OKC tornado there was a lot of debate whether or not it reached F6 status since winds were clocked in the upper range of an F5 which was like 318 mph + or – 20 mph. An F6 rating seems hypothetical because experts believed it would be nearly or next to impossible to distinguish between F5 and F6 damage. F5 Incredible tornado(261-318mph) and F6 Inconceivable tornado(319-379mph). Some experts believed that tornado winds could probably reach F6 status but like I said above would be rated E/F5 because it would be nearly or next to impossible to differentiate between the two categories. The only two tornadoes that may have come close to F6 damage is probably Jarrell, Texas and Sherman, Texas, your first ones on your list pre and post 1970. Those two tornadoes may have had instantaneous wind gusts of well over 300 mph.
    As far as the DOW readings go I believe Red Rock was 286 mph, Spencer, SD was 246-258 mph, and Mulhall was 277-299 mph. Some sources have Red Rock as 286 mph and others at 268 mph. Where did you get the Spencer, SD DOW reading as being 264 mph? I could not find that anywhere on the internet. Even though, those winds fall well into the F4 category and even midway into the F5 category on the original F-scale. Those winds go well into the EF5 category on the EF-scale which lists winds of over 200 mph but like you said I to tend to believe wind speeds are underestimated in violent tornadoes on the EF-scale. Do you also happen to know what winds were clocked in other tornadoes DOW readings such as Dimmitt, Texas, and Kellerville, Texas etc.

    • Shane – I got the Spencer reading of 264mph/118 m/s from “The 30 May 1998 Spencer, South Dakota, Storm. Part II: Comparison of Observed Damage and Radar-Derived Winds in the Tornadoes”

      As for the Red Rock reading, I know it wasn’t 286mph. I’m too lazy to do a fact check at the moment but 268mph sounds like what I remember.

      The Bridge Creek, El Reno ’13 and El Reno ’11 are the only tornadoes that have been recorded at or above 280mph.

      In my notes I have that the Dimmit, Texas, tornado was recorded with an overall peak rotational velocity of 168mph, but the article mentions that unresolved multiple vortex activity indicated a reading of 224mph (100 m/s) was more appropriate.

      buckeye05 – while some conservative surveyors and meteorologists consider ground scouring to be impossible to categorize, a large portion do take it seriously the way it should be. Long time surveyors and researchers like Marshall and Grazulis utilize ground scouring on a regular basis to either verify or disprove a high-end rating.

  17. Thank You Buckeye05. I totally agree with you on the ground scouring. Though I am not a meteorologist or engineer, I have been studying the F-scale and EF-scale very consistently over the past decade. I emailed Gary Woodall the guy who did the Westminster, Texas tornado survey a few years back. He told me that extreme ground scouring and complete debarking of trees does not guarantee a rating higher than E/F3. I highly disagree. Then I mentioned the wadded up Honda and he told me that the Leighton tornado flipped cars with EF1-EF2 winds. Yeah it flipped cars over, but it did not wad them beyond recognition, rip their engine out, and wrap it around a tree. The Westminster tornado even partially debarked or completely low lying shrubs out of the ground.
    I did look a little bit into the Rozel, Kansas tornado and it did at most EF3 home damage. The DOW clocked winds at 165-185 mph at one point and it was upgraded to an EF4. The most intense damage was where a road was scoured and 1000 gallon propane tank was lifted off its foundation and lofted a quarter mile. That sounds pretty impressive but I wonder if the DOW was being used when that damage happened. I probably would agree with the EF4 rating based on that damage but not necessarily the DOW reading.

  18. Like I said about the Rock tornado some articles state list DOW winds as 257-268mph while others list it at 286mph. Ill put up a few of them. http://www.chaseday.com/tornadoredrok.htm
    I am not sure what the official DOW reading that was measured for the Red Rock tornado. For years it was solely listed as 286mph but more recently it has been listed as 257-268mph and some articles still say 286mph. Just check those sites above one will list it as 286mph while the other will list it as 257-268mph. I don’t know which one is correct.

  19. Top 5 (or 10) list of instances of the most intense ground scouring ever, please! Or could either of yall point me to a scouring post? Jarrell, Smithville, Plainfield? I can’t think of the one with the black and white pics from a newspaper where they “calculated” winds of 400-500 mph or whatever it was, but those are crazy marks. Shawn posted pics of asphalt scouring at Pizza by Stout in his excellent Joplin post in Feb. That has to take some serious wind speed at ground level.

    • I think Shawn posted a picture of the 1960 Charles City tornado on talkweather, if that’s what you’re referring to in regards to the old winds estimates over 400mph.

      That would be a good idea for a post in the future. Off the top of my head the most intense examples of ground scouring would be:

      1. Philadelphia 2011
      2. Jarrell 1997
      3. Kellerville 1995
      4. Lawrence County 1998
      5. Bridge Creek 1999
      6. El Reno 2011
      7. Smithville 2011
      8. Beyond that it gets a lot more diffuse.

      • How off were they with those wind estimates on the 1960 Charles City tornado? Would it be considered one of the worst F5’s of all time if its winds were estimated that high? I have heard very little about it.

      • The estimates were off because they did not have a solid understanding of suction vortices and instead did a mathematical equation using the forward speed of the tornado in relation to the scouring marks as if they were caused by a single rolling object.

        The Charles City tornado would probably not have received an EF5 rating today based on the damage photos I have seen. It was quite a bit less intense than the 2008 Parkersburg tornado.

      • Ha, yeah, Paul Waite and Clarence Lamoureaux calculated 528mph by using the basic equation V equals CNS, with C being circumference of the ellipses, N being number of rings per mile, and S being forward speed of the tornado. Needless to say, this particular technique proved rather ineffective.

        Incidentally, it returned an error when I tried to post the equation using the equals sign. Hmm.

      • Awesome. The Philadelphia ground scouring photos are ridiculous. To think that moving air in a storm caused that… Yes, feel free to go ahead and make a full post on scouring. I can’t seem to find Lawrence Co damage survey pics other than the one with the partially debarked tree. And it’s at #4!

  20. The Harper Kansas tornado on May 12, 2004 also had some of the most extreme ground scouring I have ever seen. It also produced some of the most extreme damage I have ever seen. One guy said he was more impressed with the Harper tornado damage than any damage he observed in Bridge Creek, OK on May 3, 1999. The Harper tornado damage was much like the Jarrell tornado damage except it was over a smaller area.

  21. Regarding Charles City, I did find a picture of an asphalt walkway scoured away, and a picture of a house reduced to an empty basement with the center support beam ripped from its supports. Seems like it may have been pretty violent, possibly marginal F5 imo.

  22. If the highest winds recorded were about 400-500 feet above the ground is it possible this tornado could have done EF5 damage to the top of a high skyscraper such as Devon Energy Tower in OKC. Once again I am not the most certain of this but maybe you can explain this Max.

    • Well since winds near 300mph were recorded at those elevations, yes I think it is nearly certain that winds capable of causing EF5 damage were swirling around about 40 stories above the surface.

  23. I don’t know whether I agree or disagree with the downgrade. Too many things about this tornado made absolutely no sense. On the other hand it is too uncertain as whether or not this tornado was capable of inflicting EF5 damage. There are other events where the tornadoes were nearly as wide as this tornado and wondered if they carried near 300mph winds or were possibly capable of inflicting E/F5 damage. It makes me wonder if they were similar in ways to this tornado. The Yazoo City, MS tornado(1.75-2.00 miles wide), the Trousdale, Kansas tornado(2.25 miles wide) and the Hallam, NB tornado(2.50 miles wide). It makes me wonder if they carried very intense suction vortices that were incapable of making contact with the ground or did but it was only for a fraction of a second. Then on the other hand it is suspected the Pampa, TX tornado on June 8, 1995 and the tornado in North Dakota on August 7, 2010 you mentioned were probably capable of E/F5 damage, despite they were narrow or even drill-shaped.

    • I’d agree with the downgrade, simply for the fact that it is uncertain if ground-level winds were actually capable of causing EF5 damage. I’d rather rate a tornado too low than too high. For considerations of intensity, I don’t think size should be given that much consideration.

    • I agree with Nick, size should not be given much consideration.

      I think that extreme winds in tornadoes often occur in small vortices only 100 yards or so across – either as the entire tornado (Pampa) or suction spots within a larger funnel (Joplin). The overall size of a tornado says little about what’s inside of it.

  24. I agree with both of you that the size of a tornado does not matter when it comes to a tornadoes intensity, but you would think with a tornado that large(2+ miles wide) there would be plenty of small suction vortices that would be able to inflict EF5 damage. There were some people thinking just after this tornado in El Reno was downgraded to EF3 that the El Reno tornado from May 24, 2011 was going to be downgraded to EF4. Yeah they used DOW to upgrade the El Reno tornado on May 24, 2011 to EF5 but that tornado did some amazing things like EXTREME ground scouring, wadded several vehicles and made them look unrecognizable, and moving equipment that weighed 2 million pounds. The EF5 rating seemed very appropriate. I think surveyors get way too wrapped up in the construction of the houses and don’t consider other things you describe.

  25. I was wondering how you concluded that this tornado had it of passed over a row of well-built homes it would have caused mid to high-end EF4 damage. Isn’t high-end EF4 damage is when a well-built home has been destroyed in EF5 fashion but most of the debris remains nearby where the home once stood. EF5 I think is where all or nearly all the debris is swept completely away leaving little to no trace.

    • My conclusion certainly was not objective or empirically supported, it was just based on comparisons with EF4 and EF5 damage that I have seen before. The structural damage I saw was all in the EF1/EF2 and low-end EF3 range. The tree damage, at its most intense, reminded me of EF3/low-end EF4 damage that I had seen in the past and in the periphery of the Moore storm a week earlier. The ground vegetation and vehicle damage was reminiscent of EF4 tornado damage. The extreme doppler velocities and the distance vehicles were thrown made me settle on “high-end EF4.”

  26. Never heard about the El Reno tornado being 4.3 miles wide until now. It sounds absurd but even if for some reason it was 4.3 miles wide does that necessarily mean EF5 strength/damage. I swear a lot of people who know about the F-scale or EF-scale still seem to automatically associate an extremely large tornado as an EF5 tornado. I tend to believe though typically F5/EF5 tornadoes are usually pretty large. Although this is typical of F5/EF5 tornadoes, most of the F5/EF5 damage is typically more narrow than the tornado itself. I am not saying a colossus sized tornado would be incapable of producing EF5 damage but it is likely the EF5 damage would still be confined to a pretty small narrow area. What if one of these days a 30 yard wide EF5 tornado ends up drilling 4 or 5 feet into the ground and dislodging homes and their foundations. Lets say the narrow streak damage ends up being stronger than the Jarrell tornado. Are we going to be disappointed this tornado was a small rope instead of a huge wedge?

    • I haven’t seen too much debate regarding size vs intensity, but of course you’re right. While there is a definite relationship between size and intensity, as well as path length vs intensity, both are correlational as opposed to causal.

      The Philadelphia tornado was sort of like the hypothetical storm you are referring too, narrow and extremely intense. The official path width is around a quarter mile, I believe, but the swath of EF1+ damage when the extreme ground scouring occurred was less than 100 yards across. People often mistakenly believe a tornado’s size is indicative of the width of extreme damage. The Smithville tornado shattered windows over an area approximately a quarter mile wide when it ripped through town, but the EF4/EF5 damage was confined to a narrow core only 60 yards wide.

  27. In terms of intensity and how tornado winds work, are large objects (oil tanker etc) more likely to get damaged if they’re sucked into the core of a tornado versus large objects that aren’t sucked in but instead are picked up and skirted/moved outwards away from the funnel and the core?

  28. I agree that the EF-Scale totally doesn’t estimate. They passed it through (reportedly TTU) and NSSL/OU have been trying to get an Enhanced EF scale – LOL. Who did the article?

Leave a Reply

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

WordPress.com Logo

You are commenting using your WordPress.com 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