Enhanced Fujita Ratings Debate Thread

[TH2002]

I’ve thought about this. Generally it’s thought that EF4+ wind speeds have to occur for this to happen but the current transmission tower DI tops out at 165 MPH, hence the EF3 rating in Kouts this year.

One issue though in terms of calculating it though may stem from the fact that it’s not a “solid” object (not the best description but you know what I’m getting at). Instead, it’s a latticework of interconnected metal bars with lots of space in between for the wind to pass through. I could be wrong, but it might be tough to calculate due to the wind passing through much of the tower during a tornado impact. Another problem is a transmission tower isn’t going to maintain the same shape and weight distribution when being hit by a tornado. It’s going to twist and crumple into different unpredictable shapes while being lofted.

The bridge support that was thrown up the canyon embankment at Hurricane Creek outside of Tuscaloosa also had a similar truss-type structural makeup. I think that one might also be tougher to calculate for that reason too, among others.

Wasn't Bowdle (at least partially) rated EF4 based on damage to transmission towers?

FWIW, Spann wrote in his 4/27 book that he believes the destruction of the railroad bridge over Hurricane Creek was indicative of EF5 intensity. I know he's not a damage survey expert, but he is a well-respected meteorologist in the field, and honestly, that (combined with the AMETSOC paper that rates the tornado EF5 based on treefall patterns near the bridge) is good enough for me.

 

[buckeye05]

Wasn't Bowdle (at least partially) rated EF4 based on damage to transmission towers?

FWIW, Spann wrote in his 4/27 book that he believes the destruction of the railroad bridge over Hurricane Creek was indicative of EF5 intensity. I know he's not a damage survey expert, but he is a well-respected meteorologist in the field, and honestly, that (combined with the AMETSOC paper that rates the tornado EF5 based on treefall patterns near the bridge) is good enough for me.

Oh yeah Bowdle absolutely was. But they had to “break the rules” and go beyond the assigned upper bound for transmission towers, which is something most WFOs are not going to. Hopefully the new upper bound for transmission towers will be extended into the EF4 range once the updated scale is rolled out.

Also I’m not at all doubting that EF5 winds were required to do that to the Hurricane Creek railroad bridge support. In fact, I’m highly confident that EF5 winds had to have occurred there, and the treefall data from that location backs it up even more. I’m just saying getting a precise calculation might be tough there due to the shape/design of the support, along with the fact that the calculation wouldn’t just be as simple as “what windspeed is needed to loft this bridge support?”. The calculation would be, “What windspeed is needed to roll this bridge support up an embankment with a slope of this specific degree, tumbling it uphill but without fully lofting it?” That’s very specific and tricky. But is it impossible? Probably not, but it would definitely take a lot more work and involve far more variables than a lofted fertilizer tank or train car.

The railroad bridge is one of the top two things I want to see calculated, but I’ve been really thinking about it lately, and it’s going to be a lot more complicated than I first thought it would be.

 

[Grand Poo Bah]

Oh yeah Bowdle absolutely was. But they had to “break the rules” and go beyond the assigned upper bound for transmission towers, which is something most WFOs are not going to. Hopefully the new upper bound for transmission towers will be extended into the EF4 range once the updated scale is rolled out.

Also I’m not at all doubting that EF5 winds were required to do that to the Hurricane Creek railroad bridge support. In fact, I’m highly confident that EF5 winds had to have occurred there, and the treefall data from that location backs it up even more. I’m just saying getting a precise calculation might be tough there due to the shape/design of the support, along with the fact that the calculation wouldn’t just be as simple as “what windspeed is needed to loft this bridge support?”. The calculation would be, “What windspeed is needed to roll this bridge support up an embankment with a slope of this specific degree, tumbling it uphill but without fully lofting it?” That’s very specific and tricky. But is it impossible? Probably not, but it would definitely take a lot more work and involve far more variables than a lofted fertilizer tank or train car.

The railroad bridge is one of the top two things I want to see calculated, but I’ve been really thinking about it lately, and it’s going to be a lot more complicated than I first thought it would be.

I'm currently at the realization that it's completely unreasonable to try and attach an exact wind speed to tornado damage. We really need to go back to more broad wind speed ranges for these classifications.

I'm sure the question "Is it probable sub-200 mph winds could cause this damage?" Is much easier to answer mathematically than "what exact wind speed caused this damage?"

 

[buckeye05]

I'm currently at the realization that it's completely unreasonable to try and attach an exact wind speed to tornado damage. We really need to go back to more broad wind speed ranges for these classifications.

I'm sure the question "Is it probable sub-200 mph winds could cause this damage?" Is much easier to answer mathematically than "what exact wind speed caused this damage?"

Yeah I’ve often thought about that. Maybe keeping things more broad and maybe assigning a range of possible wind speeds could solve some of the issues? It’s an interesting idea.

 

[eep]

Interesting notes on the Floyd NM tornado from Anton Seimon, apparently a bird can be used to approximate windspeeds?

 

[WeathermanLeprechaun]

Seconding the Floyd research by Anton here. This has always been my TOTY of 2025 (yes, i hold the hot seat for this one), and this definitely proves the theory some have that maybe we underestimate how truly easy it is for a tornado to breach EF4-EF5 intensity.

Honestly quite extraordinary research, and i know that Anton Seimon worked with Pecos Hank, i believe in 2019 scanning one of the tornadoes on 5/28 that year. Note these findings are VERY preliminary but this proves to me you don't need 400 SRH at all to produce a violent tornado too.

 

[Grand Poo Bah]

It's definitely from Greenfield, from a CAP overflight of the damage path per andyhb's original post: https://talkweather.com/threads/severe-weather-threat-5-19-5-22-2024.2275/page-53#post-121789

Honestly, that would be textbook EF5 damage if the home was well-anchored. I don't think it was, but have never been able to definitively confirm which damage point on the DAT corresponds to that house.

Huge thanks for this!

Yeah I’ve often thought about that. Maybe keeping things more broad and maybe assigning a range of possible wind speeds could solve some of the issues? It’s an interesting idea.

I think a silver lining to the NWS's stubbornness to upgrade tornadoes has been all the good science that has been published in the meantime. If so much of these findings didn't contradict official ratings, they might not have gotten the same recognition and widespread discussion. While we might not get upgrades anytime soon, it's still important to gather the data that will need to be used to determine corrected wind speed ranges.

There's certainly enough methodology now to peg down some pretty accurate wind speeds.

We have solid understandings of:
Lofting of large objects
Tree fall
Cycloidal marks
Strong doppler radar correlations like debris height
Engineering

None of these methods are individually perfect, but together, much more realistic probabilistic analysis can be done. A realistic curve would probably have the maximum EF rating on the lower tail, radar data on the upper tail, then a combination of the rest in the middle. Based on the confidence values associated with the middle analysis you could come up with a pretty realistic most likely wind speed.

 

[Tomcats Forever!]

What is the point of the "exp" on a slabbed house being 200 when there hasn't been a house rated over 190 in 11 years?

 

[UK_EF4]

 

[CheeselandSkies]

Wish I'd known sooner that conference was happening in Madison so I could have made plans to attend.

 

[Grand Poo Bah]

Would you look at those probability curves! It's beautiful.

 

[Grand Poo Bah]

The slides are funny because they keep reiterating the same thing

General agreement in the estimated wind speeds between treefall pattern method and the EF scale outside the violent core of damage.

Basically, they're in agreement everywhere but the most important region of the tornado lmao. The following bullet point says,

Treefall pattern can likely provide a reasonably comparable wind speed estimate to the EF scale when validated with the entire wind field.

It's so weird how experts constantly tip-toe around the EF Scale like it's a person with feelings or something. The scale was never devised to evaluate "general wind speeds" across an entire tornado. The goal has always been to find the maximum damage and maximum wind speeds. This approach is revisionist to the entire history and purpose of both Fujita scales.

I understand Lyza is playing the long game here, and trying not to ruffle any feathers, but it begs the question, whose feathers are getting ruffled over this? Why is there this weird controlling shadow over tornado science holding it back at every turn, when there's seemingly no actual public figures leading this field?

 

[Aaron Rider]

Seconding the Floyd research by Anton here. This has always been my TOTY of 2025 (yes, i hold the hot seat for this one), and this definitely proves the theory some have that maybe we underestimate how truly easy it is for a tornado to breach EF4-EF5 intensity.

Honestly quite extraordinary research, and i know that Anton Seimon worked with Pecos Hank, i believe in 2019 scanning one of the tornadoes on 5/28 that year. Note these findings are VERY preliminary but this proves to me you don't need 400 SRH at all to produce a violent tornado too.

In case anyone was wondering, the only "F4" in New Mexico history was in October (!) of 1908

Post in thread 'Significant Tornado Events'

Jul 26, 2022

One forgotten yet significant tornado outbreak is the October 18, 1908 High Plains tornado outbreak. The Weather Bureau and Grazulis don't list any tornadoes from this outbreak, despite reliable newspaper reports of funnel clouds, debris being carried long distances, and significant tornado damage. At least three significant tornadoes occurred on this day, two of which were likely long-tracked tornado families.

Tornado #1 - Tucumcari NM
This intense tornado (family?) moved from near House to 12 miles NE of Tucumcari, devastating rural areas. Two people were killed on separate farms...

• MNTornadoGuy

• 

 

[slenker]

The slides are funny because they keep reiterating the same thing

Basically, they're in agreement everywhere but the most important region of the tornado lmao. The following bullet point says -img

Essentially, my biggest takeaway becomes that the EF scale vastly improved the F-scale's ability to rate weaker/borderline strong tornadoes, but vastly set us back for the upper end of EF3 and beyond. The F-scale becomes better suited for characterizing more violent outbreaks, AKA the ones that matter the most. If I absolutely had to pick between the two, I really don't know which one I would pick. In a vacuum, I'd pick the EF scale, since it is objectively better when it is applied and used correctly and when effort is put into it.* But there's no denying that surveyors are extremely inconsistent across WFOs and that leads to horrendous misuses of it that we see today across a lot of the surveyors in the country. At least we have Grand Forks, Norman, and Jackson to show how the EF scale can be better when used correctly, unlike the Memphis and Springfield offices.

It's so weird how experts constantly tip-toe around the EF Scale like it's a person with feelings or something. The scale was never devised to evaluate "general wind speeds" across an entire tornado. The goal has always been to find the maximum damage and maximum wind speeds. This approach is revisionist to the entire history and purpose of both Fujita scales.

I understand Lyza is playing the long game here, and trying not to ruffle any feathers, but it begs the question, whose feathers are getting ruffled over this? Why is there this weird controlling shadow over tornado science holding it back at every turn, when there's seemingly no actual public figures leading this field?

This is how professional science is conducted, and they word it like this to be as objective and unbiased as possible in their assessment/interpretation of the data. Individual studies like this will serve as compounding evidence against the current EF scale's use. I don't think anyone's feathers are getting ruffled here, he's saying it like it is in a professional way. Personally, I wouldn't look at it this way at all, and I would imagine that this presentation was received well by the leading experts in tornado science that were present at this conference.

* By "objectively better" I'm not talking about the windspeed estimates attached to the scale, I'm talking about the ratings themselves. EF scale windspeed estimates are very poor, as we all know.

 

[eep]

In case anyone was wondering, the only "F4" in New Mexico history was in October (!) of 1908

Post in thread 'Significant Tornado Events'

Jul 26, 2022

One forgotten yet significant tornado outbreak is the October 18, 1908 High Plains tornado outbreak. The Weather Bureau and Grazulis don't list any tornadoes from this outbreak, despite reliable newspaper reports of funnel clouds, debris being carried long distances, and significant tornado damage. At least three significant tornadoes occurred on this day, two of which were likely long-tracked tornado families.

Tornado #1 - Tucumcari NM
This intense tornado (family?) moved from near House to 12 miles NE of Tucumcari, devastating rural areas. Two people were killed on separate farms...

• MNTornadoGuy

• 

As someone who has periodically lived there, I am like 90% sure the reason is kinda similar to a lot of northern plains outbreaks where there are like, 0 damage indicators (not even trees!) out there, especially far outside of Albuquerque, Roswell, and Santa Fe. I'd wager some of the other NM tornado events are similarly comically underrated.

 

[CheeselandSkies]

Interesting aside, the only other time I (and probably many other severe weather history buffs) have heard of Tucumcari is in reference to the (now defunct) NOAA wind profiler site located there, which is what allowed the SPC to catch the incoming jet streak in time to upgrade the 1630Z outlook to a moderate risk on May 3, 1999.