Enhanced Fujita Ratings Debate Thread

[slenker]

Interestingly, the 300+ mph (I don’t remember the exact number) reading that was captured was actually, if I remember correctly, over pretty standard EF2 and EF3 damage southwest of town. That could be just because they stopped taking scans as it went northeast into town (I haven’t read the study fully) but it’s interesting nonetheless.

Edit: NASA has a webpage on it and apparently DOW7 did continue scanning it in town; specific numbers aren’t attached to each pixel so it’s impossible to see where the highest winds were.

You’re absolutely correct and I genuinely forgot about this little tidbit of info. It is important and definitely points to its unreliability (or maybe something else, like said 300+ winds not having reached to ground level)

There was another user on here who made a very intriguing read on windspeed radar measurements coincident with damage - I can’t recall who but the post discussed Mulhall and Bridge Creek. IIRC Bridge Creek’s 300+ reading was well entrenched in very, very high end damage. Then you have cases like El Reno 2013 that did extreme vehicle damage, but no ground scarring or super intense vegetation damage coincident with the 300+ readings, so it’s a bit of an oddity for sure.

I will say I’m very skeptical of El Reno 2013 having reached a “true” EF5 intensity myself. I wouldn’t support an upgrade to EF5 for that tornado at all, but that’s another conversation. And that mindset is somewhat contradictory to my statements about Greenfield’s 300+ reading too, so I will back down on that point. But, the parking stops are pretty legitimate in my eyes regardless.

 

[Central Ohio Wx]

You’re absolutely correct and I genuinely forgot about this little tidbit of info. It is important and definitely points to its unreliability (or maybe something else, like said 300+ winds not having reached to ground level)

There was another user on here who made a very intriguing read on windspeed radar measurements coincident with damage - I can’t recall who but the post discussed Mulhall and Bridge Creek. IIRC Bridge Creek’s 300+ reading was well entrenched in very, very high end damage. Then you have cases like El Reno 2013 that did extreme vehicle damage, but no ground scarring or super intense vegetation damage coincident with the 300+ readings, so it’s a bit of an oddity for sure.

I’m really glad we’ve been able to develop tools like DOW since the 80s, but of course there will always be kinks that need to be ironed out. It does make you wonder what tornadoes like Hollister would look like on DOW.

 

[Grand Poo Bah]

Worth noting the El Reno 2013 tornado was scanned 100 meters (330 feet) above ground level, the Greenfield tornado was 30-50 meters (98-164 feet), and Bridge Creek was 32 meters (102 feet). It's probably a safe assumption that the closer to ground level the scan is, the more likely it is to be representative of ground level wind speeds. The Nasa article being referenced does claim the 300 mph winds were sustained throughout the whole town I believe. I'll see if I can find it.

I also always forget the EF scale is about 3 second wind gusts. It's funny because many tornadoes (especially Greenfield) aren't even over the same spot for 3 seconds. The damage is done in less time than that. Greenfield passed through the whole town in less than 60 seconds.

Edit:

Here's the article. It's the very first graphic.

https://www.earthdata.nasa.gov/dashboard/data-catalog/tornadoes-2024-dow

 

[Grand Poo Bah]

I agree with radar data being unreliable on its own, but a 300 mph wind gust is certainly something that can be taken into account because it’s extremely rare to get that by itself, and it’s clearly coincident with very high end damage that possibly could justify an EF5 rating. Enderlin took radar data into account in its rating as well. However, it’s a completely different method of analysis and definitely better than NEXRAD because it utilized Vrot calculations with (I believe) WSR 88-D data, which have been shown to be pretty well correlated to higher tornadic intensities for higher Vrot values in some studies. I’d like to see that methodology applied to the Greenfield tornado to see what it reveals.

I used to be firmly in the “Greenfield isn’t an EF5” camp, but I’ve since stepped away from that viewpoint and it’s mostly due to the same contextual indicators that kept Joplin’s EF5 rating. Pretty sure Greenfield had a manhole cover that was sucked up as well, but I never saw any imagery of that damage, so I don’t know its validity.

EDIT: I also do want to add that if they keep Greenfield’s 185 EF4, I wouldn’t mind either. It’s a fine rating. But the parking stops in my eyes are very valid EF5 indicators if other experts come to the same conclusion, and if more detailed analysis is done on it that supports these findings, I fully support an upgrade to EF5.

Appreciate you bringing some fresh arguments to the table, I think we were starting to debate in circles a bit. Great points.

The only area I personally disagree is "being fine with the 185 mph rating." Based on all the information we have it's actually more likely the tornado had 200 mph+ winds than 185 mph. It's weird that we consider the 185 mph rating the "safer bet" just because it's lower. Conservative doesn't automatically mean accurate or logical. It'd be one thing if the ratings were presented as the "minimum", or if Greenfield was rated "at least 185 mph", but they aren't. Tornado ratings are recorded in the NWS database as "Est. Peak Winds". 185 mph is the maximum rating, not the minimum.

 

[Grand Poo Bah]

The original Fujita scale didn't present ratings this way. If a tornado was F4 it meant the wind speeds were at least 207 mph, and up to 260 mph. It wasn't until the EF scale that we started trying to attach exact numbers, and presenting them as "peak winds". Surveyors constantly reference the unknown for conservative ratings, but then attach extremely precise wind speeds to those ratings. It's counter-intuitive and contradictory.

 

[slenker]

The only area I personally disagree is "being fine with the 185 mph rating." Based on all the information we have it's actually more likely the tornado had 200 mph+ winds than 185 mph. It's weird that we consider the 185 mph rating the "safer bet" just because it's lower. Conservative doesn't automatically mean accurate or logical. It'd be one thing if the ratings were presented as the "minimum", or if Greenfield was rated "at least 185 mph", but they aren't. Tornado ratings are recorded in the NWS database as "Est. Peak Winds". 185 mph is the maximum rating, not the minimum.

The original Fujita scale didn't present ratings this way. If a tornado was F4 it meant the wind speeds were at least 207 mph, and up to 260 mph. It wasn't until the EF scale that we started trying to attach exact numbers, and presenting them as "peak winds". Surveyors constantly reference the unknown for conservative ratings, but then attach extremely precise wind speeds to those ratings. It's counter-intuitive and contradictory.

I'm actually not entirely sure that the minds behind the EF scale were trying to insinuate that the maximum winds were 185 mph for a tornado like this, or for any given rating of xxx mph - and that it is a misconception that stems entirely from the verbiage used to within a given rating to describe the tornado in the confines of the scale, with terms like "most likely maximum possible windspeed." It's hard to put this point into words but I'm going to try anyways. I think what is meant is that that is the minimum required windspeed to inflict such damage for any given DI, but since you cannot prove that it went any higher than that, it is sort of a "min-maxed" estimate for the winds of the tornado. So it's not saying that the tornado had maximum winds of 185 mph - it's saying that based on the confines of the scale, we can only prove winds of at least 185 mph occurred here, due to the nature of this specific DI. And, that is the highest estimate they can get to, so that is the "maximum possible windspeed" that they could prove. That doesn't mean the winds weren't higher than 185 mph.

The reason why I agree with the 185 mph rating is because of this. There isn't any damage outside of the parking stops you can mathematically show exceeding 200 mph without a ridiculous amount of assumptions. The 185 mph rated home seems like a fine application of the scale to me, and that stays within the scale's confines. But I want and hope for them to go outside of the confines of the scale, both because they've done it in the past and also because of what they just did with Enderlin. But the 185 mph windspeed is something that surveyors can say with absolute certainty occurred, and it's hard to argue against that logic. It's a very scientific way of looking at it. At the end of the day, though, it is far from perfect, and the new scale needs to fix the overwhelming amount of issues with the EF scale right now.

Also, keep in mind that my first paragraph could be the wrong way of thinking about it - it makes sense to me, especially with the amount of surveyors and mets who routinely state that the winds were likely or definitely higher than what an EF rating suggests for many different tornadoes. Surveyors and mets are fully aware of the fact that a swept slab of subpar construction could have occurred from a 170 EF4 or a 205 EF5, and the damage will only differ by contextuals and granulation (which needs to badly be taken into account).

 

[Grand Poo Bah]

I'm actually not entirely sure that the minds behind the EF scale were trying to insinuate that the maximum winds were 185 mph for a tornado like this, or for any given rating of xxx mph - and that it is a misconception that stems entirely from the verbiage used to within a given rating to describe the tornado in the confines of the scale, with terms like "most likely maximum possible windspeed." It's hard to put this point into words but I'm going to try anyways. I think what is meant is that that is the minimum required windspeed to inflict such damage for any given DI, but since you cannot prove that it went any higher than that, it is sort of a "min-maxed" estimate for the winds of the tornado. So it's not saying that the tornado had maximum winds of 185 mph - it's saying that based on the confines of the scale, we can only prove winds of at least 185 mph occurred here, due to the nature of this specific DI. And, that is the highest estimate they can get to, so that is the "maximum possible windspeed" that they could prove. That doesn't mean the winds weren't higher than 185 mph.

Nothing is insinuated, though. It is said outright. No one can argue in good faith that "peak winds" actually means "minimum winds". They're exact opposites. I have absolutely no doubt that you're exactly right about the mindset behind the ratings, though. It's very clear surveyors are trying to discern the minimum winds. Would you agree then that this verbiage needs to be corrected in public records, so we're not misinforming the press and public with contradictory information?

This minimum seeking approach still isn't very useful, though. All it's doing is making surveyors assign the lower bound wind speed to DIs because it's the safest option. The problem is the safest option isn't always the most accurate. In this case, it's actually leading to extreme inaccuracies. I would argue that the best option is actually to use established arithmetic, published studies, and the EF scale as it was originally intended to try and get as close as possible to accurate wind estimates. Close is better than safe.

 

[slenker]

Nothing is insinuated, though. It is said outright. No one can argue in good faith that "peak winds" actually means "minimum winds". They're exact opposites. I have absolutely no doubt that you're exactly right about the mindset behind the ratings, though. It's very clear surveyors are trying to discern the minimum winds. Would you agree then that this verbiage needs to be corrected in public records, so we're not misinforming the press and public with contradictory information?

Absolutely, that's what I meant to insinuate in that paragraph. I forgot to state it outright so that's my bad. It's very easy for laymen to get confused on what these numbers mean and I see it repeatedly across the internet.

This minimum seeking approach still isn't very useful, though. All it's doing is making surveyors assign the lower bound wind speed to DIs because it's the safest option. The problem is the safest option isn't always the most accurate. In this case, it's actually leading to extreme inaccuracies. I would argue that the best option is actually to use established arithmetic, published studies, and the EF scale as it was originally intended to try and get as close as possible to accurate wind estimates. Close is better than safe.

This is a great argument.

The only part I disagree with is the bolded point, but not because I don't think that a closer estimate is better than a "safe" estimate, it's because we don't know what the true values are in the first place, so how will we know if >200 mph is closer? If we don't know what the true value is at some given instant (or for 3 seconds, in the case of the EF scale parameters), then we have to resort to what we really know, and what we really know is there is direct ground evidence of winds of at least 185 mph in Greenfield beyond a reasonable doubt because of that home damage. And yes, you could make the argument that the true values are indeed the NEXRAD measurements. But are they really? We've seen plenty of examples of tornadoes having very extreme wind measurements that most certainly weren't of EF5 intensity. It's not that I don't trust NEXRAD, it is that I think the winds that are collected by that measurement device are, for some strange reason, embellishing of a tornado's true intensity at ground level. Whether that is from a fundamental inaccuracy of how it collects this data (which is extremely unlikely) or simply from the fact that it always collects data from higher within the vortex, or something else entirely, I have no idea. I'm simply of the opinion that the EF scale, despite its flaws, is still good science if applied correctly - it just needs to continue to be improved with time, which will no doubt happen.

Not to bring a completely unrelated topic out of the blue here, but the mismatch between damage survey wind speeds vs. radar wind measurements being so far off from each other reminds me of the Hubble Tension, which is a peculiar observation in cosmology that basically boils down to "measuring the expansion of the universe one way leads to a different value for its expansion than the other way." It's an entirely different thing, of course, and there's plenty of differences you could pick apart from these two situations. But it's really hard to find a middle ground when there's such a strong divergence in values between the two different methodologies, for both the Hubble Tension and this. The middle ground might not be correct, one or the other may be "more correct."

 

[buckeye05]

Nothing is insinuated, though. It is said outright. No one can argue in good faith that "peak winds" actually means "minimum winds". They're exact opposites. I have absolutely no doubt that you're exactly right about the mindset behind the ratings, though. It's very clear surveyors are trying to discern the minimum winds. Would you agree then that this verbiage needs to be corrected in public records, so we're not misinforming the press and public with contradictory information?

This minimum seeking approach still isn't very useful, though. All it's doing is making surveyors assign the lower bound wind speed to DIs because it's the safest option. The problem is the safest option isn't always the most accurate. In this case, it's actually leading to extreme inaccuracies. I would argue that the best option is actually to use established arithmetic, published studies, and the EF scale as it was originally intended to try and get as close as possible to accurate wind estimates. Close is better than safe.

Your second point is something I have been thinking about and trying to verbalize for years. In a nutshell, the wind speed at which a specific type of damage could potentially begin to happen, versus the wind speed at which a specific type of damage typically does happen, are two very different wind speeds.

So the question becomes, is it a better idea to go with an EF scale application approach that plays it safe and nearly eliminates any potential for intensity overestimation, but results in an inherent potential for underestimation? Or is it better to go with an EF scale application approach that is more likely to reflect the true intensity of the tornado in question, but also results in an inherent potential for intensity overestimation?

It’s a tricky dichotomy.

 

[Grand Poo Bah]

But are they really? We've seen plenty of examples of tornadoes having very extreme wind measurements that most certainly weren't of EF5 intensity. It's not that I don't trust NEXRAD, it is that I think the winds that are collected by that measurement device are, for some strange reason, embellishing of a tornado's true intensity at ground level. Whether that is from a fundamental inaccuracy of how it collects this data (which is extremely unlikely) or simply from the fact that it always collects data from higher within the vortex, or something else entirely, I have no idea. I'm simply of the opinion that the EF scale, despite its flaws, is still good science if applied correctly - it just needs to continue to be improved with time, which will no doubt happen.

The most interesting aspect of this bolded point is that when you compare the DOW windspeed measurements to the original F scale the discrepancies mostly dissapear.

Your second point is something I have been thinking about and trying to verbalize for years. In a nutshell, the wind speed at which a specific type of damage could potentially begin to happen, versus the wind speed at which a specific type of damage typically does happen, are two very different wind speeds.

So the question becomes, is it a better idea to go with an EF scale application approach that plays it safe and nearly eliminates any potential for intensity overestimation, but results in an inherent potential for underestimation? Or is it better to go with an EF scale application approach that is more likely to reflect the true intensity of the tornado in question, but also results in an inherent potential for intensity overestimation?

It’s a tricky dichotomy.

This graph from the recent compact object study that preceded Enderlin's upgrade really illustrates the point perfectly.

Here's a link because this site is still broken.


There's a 3% chance the combine can be thrown at sub 200 mph winds, and a 97% chance it requires winds over 200 mph. How is a minimum rating (EF4) with only a 3% chance of being correct, safer than the rating with a 97% likelihood (EF5)? You can ask this same question with so many other EF3s and 4s. The minimum possible windspeed (1%) and the more likely windspeed (51%) are so far apart they can't even be considered the same type of measurement.

 

[Grand Poo Bah]

Here's a better illustration, where the highlighted sections represent probability, and It perfectly illustrates what I've been talking about. When you're talking about a rating system that uses wind speed ranges like the EF scale does, it makes no sense to pick the rating with close to zero or zero possibility. EF4 isn't a rating of 166+, it's 166-200. If there's evidence the wind speeds are above 200 mph surveyors have a responsibility to use it, because otherwise the data is incorrect. It may feel good to justify it by saying it's a "minimum", but that's literally not the system we're using here.

 

[Lake Martin EF4]

On paper I would agree......but I just can't get over that (roughly) 5 mph forward speed. Though the tornado was likely capable of causing F5 damage regardless, I do feel that the low forward speed amplified the damage a bit. It's still in my top 10, but barely, at #10.

Now if it were moving at a normal speed and did that sort of damage, THEN it would be a different story.

EDIT: And since I mentioned it, here's my top 10 lol. Still a WIP but my #1-4 and #9-10 are set in stone:

#1 - Piedmont 2011
#2 - Tri-State 1925
#3 - Bridge Creek 1999
#4 - Matador 2023
#5 - Bakersfield Valley 1990/Stratton 1990 (tie!)
#6 - Greensburg 2007/Parkersburg 2008 (tie!)
#7 - New Richmond 1899/Moore 2013 (tie!)
#8 - San Justo 1973
#9 - Smithville 2011
#10 - Jarrell 1997

Updoot

#1 - Piedmont, OK (5/24/2011)
#2 - Tri-State, MO/IL/IN (3/18/1925)
#3 - Bridge Creek, OK (5/3/1999)
#4 - Jarrell, TX (5/27/1997)
#5 - Bakersfield Valley, TX and Stratton, NE (6/1/1990 and 6/15/1990)
#6 - Matador, TX (6/21/2023)
#7 - Greensburg, KS and Parkersburg, IA (5/4/2007 and 5/25/2008)
#8 - New Richmond, WI and Moore, OK (6/18/1899 and 5/20/2013)
#9 - Catania, IT (10/7/1884)
#10 - Smithville, MS (4/27/2011)

Jarrell and Bridge Creek are in an all out brawl for #3. Catania pushed San Justo out of the Top 10. Bakersfield/Stratton jumped ahead of Matador. Little change otherwise.

Edit: Also can't rule out the possibility of a potentially major upward shift in Smithville's ranking.

 

[Lake Martin EF4]

Updoot

#1 - Piedmont, OK (5/24/2011)
#2 - Tri-State, MO/IL/IN (3/18/1925)
#3 - Bridge Creek, OK (5/3/1999)
#4 - Jarrell, TX (5/27/1997)
#5 - Bakersfield Valley, TX and Stratton, NE (6/1/1990 and 6/15/1990)
#6 - Matador, TX (6/21/2023)
#7 - Greensburg, KS and Parkersburg, IA (5/4/2007 and 5/25/2008)
#8 - New Richmond, WI and Moore, OK (6/18/1899 and 5/20/2013)
#9 - Catania, IT (10/7/1884)
#10 - Smithville, MS (4/27/2011)

Jarrell and Bridge Creek are in an all out brawl for #3. Catania pushed San Justo out of the Top 10. Bakersfield/Stratton jumped ahead of Matador. Little change otherwise.

Edit: Also can't rule out the possibility of a potentially major upward shift in Smithville's ranking.

Update:

Jarrell won the #3 battle. @Central Ohio Wx made a claim that Smithville was so strong it nearly ripped itself apart - if that's the case then that's gonna result in the upward shift I mentioned. It's not going to be above Jarrell though.

 

[Central Ohio Wx]

Jarrell won the #3 battle. @Central Ohio Wx made a claim that Smithville was so strong it nearly ripped itself apart - if that's the case then that's gonna result in the upward shift I mentioned. It's not going to be above Jarrell though.

I've changed my stance on Smithville - it isn't top 5 and barely top 10.

 

[Lake Martin EF4]

I've changed my stance on Smithville - it isn't top 5 and barely top 10.

Uh..wow. What caused that drop?

Maybe I was justified in putting it low on my list...

 

[Central Ohio Wx]

Uh..wow. What caused that drop?

Maybe I was justified in putting it low on my list...

Motion doesn't perfectly equate to intensity, and while the instantaneous winds were in the 300-400 range the structural and even contextual damage (3s) isn't as impressive as it's set out to be.

 

[Grand Poo Bah]

Motion doesn't perfectly equate to intensity, and while the instantaneous winds were in the 300-400 range the structural and even contextual damage (3s) isn't as impressive as it's set out to be.

That's certainly an opinion lmao

 

[ColdFront]

That's certainly an opinion lmao

Once you realize the majority of EF scale debate is essentially ESPN radio with Stephen A. Smith-esque hot takes just for the sake of having them, it makes sense.

Same thing with power ranking tornados.

And again, this isn’t directed at anyone, more of a criticism of the entire zeitgeist at the moment. We’re reaching levels of “Michael Jordan was only good because he played in the 90s” levels of discourse.

 

[tornado examiner]

Uhhhhhmm. Why are we now suddenly downplaying the intensity of a tornado that turned large brick structures to “dust”, launched large vehicles over half a mile, trenches the earth. Made an entire semi truck disappear along with its cargo. Embedded another semi truck, and an RV into the ground, turned low lying shrubbery into powder, along with uplifting numerous large trees and reducing them to wood chips, upheaved parts of a home’s foundation slab, and jumped from EF0 to EF5 intensity in just 6 seconds. All while moving at over 60MPH in the single most violent tornadic atmosphere in modern history?

Just curious why suddenly several people think this tornado isn’t top 3 anymore?
Idk how else to say all this so I'm sorry if it comes off wrong.

 

[tornado examiner]

That's certainly an opinion lmao

Yea like HUH? The photo’s speak for themselves why are they suddenly not extremely top of the spectrum impressive?