Enhanced Fujita Ratings Debate Thread

[TH2002]

Huh?

Tell me more about this...I've never heard of it! Probably isn't even in my map either.

EDIT: Is this in reference to the West Texas EF2 on 4/29?

EDIT 2: Also...

That outbreak sure felt like a year, but...

EDIT 3: Also surprised Marion 2004 didn't get the "italics treatment".

Yep, Terrell County 4/29/2009. Too lazy to embed the photos in this post but it was definitely extremely violent. Scoured rocks and gravel leaving bare soil, shredded and debarked low lying vegetation and mesquite trees, and (perhaps most unusually) stripped most of the paint from a large storage tank. NWS article here: https://www.weather.gov/maf/2009_04_29_TerrellCountyTornado

Marion didn't get the "italics treatment" because I'm on the fence about it. Have lots of damage photos and it was most definitely very violent, but haven't seen any slam-dunk F5 damage from it. Of course it doesn't help that the TornadoTalk article is behind a paywall...

 

[buckeye05]

For 5/3/99, don't forget Dover.

Also, thanks for reminding me of Canton Lake. I think it was probably a highest grade tornado based on asphalt scouring.

It’s important to note that this is not a reliable basis. In fact, after many years of looking into the damage from countless tornadoes of various intensities, I’ve come to accept that asphalt scouring is the least reliable “violent” contextual indicator there is. It seems so viscerally impressive and I used to roll my eyes at anyone who dismissed such a seemingly incredible phenomenon, but after too many counter-examples, it’s apparent that it is loosely correlated with violent tornadoes at best. I’ve simply seen it happen too many times with tornadoes that were definitely below the violent threshold, and not just ones with “defaulted” lower ratings due to a lack of DIs.

A few examples I can remember:

-Watonga, OK 1998: Narrow, moderate intensity tornado scoured pavement from a road, while only causing F2 damage otherwise.

-Waupaca County, WI 2004: Completely scoured pavement from a road while at low-end F2 intensity. Worst structural damage was roof loss to homes, and softwood pine trees in the immediate vicinity of the road were not debarked and did not suffer any damage anywhere close to the violent threshold.

-Americus, GA 2008: Caused impressive pavement scouring, but did not produce structural damage exceeding EF3 intensity anywhere along the path. Tree damage wasn’t violent either.

-New Boston, TX 2022: Low-end EF3 tornado scoured pavement from a road. A house at the roadside took a direct hit, and only lost its roof and some walls, confirming that the pavement scouring indeed happened while the tornado was at low-end EF3 strength. In fact, a piece of the asphalt was thrown into the house.

These are just the ones I can remember off the top of my head, and I know there are other examples too. Despite my tangent, I agree that Canton was violent based on the grass scouring, extreme debarking, and mobile radar readings, but the asphalt scouring was not the clincher here.

 

[slenker]

It’s important to note that this is not a reliable basis. In fact, after many years of looking into the damage from countless tornadoes of various intensities, I’ve come to accept that asphalt scouring is the least reliable “violent” contextual indicator there is. It seems so viscerally impressive and I used to roll my eyes at anyone who dismissed such a seemingly incredible phenomenon, but after too many counter-examples, it’s apparent that it is loosely correlated with violent tornadoes at best. I’ve simply seen it happen too many times with tornadoes that were definitely below the violent threshold, and not just ones with “defaulted” lower ratings due to a lack of DIs.

A few examples I can remember:

-Watonga, OK 1998: Narrow, moderate intensity tornado scoured pavement from a road, while only causing F2 damage otherwise.

-Waupaca County, WI 2004: Completely scoured pavement from a road while at low-end F2 intensity. Worst structural damage was roof loss to homes, and softwood pine trees in the immediate vicinity of the road were not debarked and did not suffer any damage anywhere close to the violent threshold.

-Americus, GA 2008: Caused impressive pavement scouring, but did not produce structural damage exceeding EF3 intensity anywhere along the path. Tree damage wasn’t violent either.

-New Boston, TX 2022: Low-end EF3 tornado scoured pavement from a road. A house at the roadside took a direct hit, and only lost its roof and some walls, confirming that the pavement scouring indeed happened while the tornado was at low-end EF3 strength. In fact, a piece of the asphalt was thrown into the house.

These are just the ones I can remember off the top of my head, and I know there are other examples too. Despite my tangent, I agree that Canton was violent based on the grass scouring, extreme debarking, and mobile radar readings, but the asphalt scouring was not the clincher here.

I would guess that a significant portion of asphalt/pavement scouring is the result of debris impacts that initially “dent” the surface, then give way for the wind to undercut it. That may be why the asphalt scouring that occurred in Smithville was “rolled up” and occurred outside of the most violent core (not super far away from it though, just slightly to the left of it). It would also make sense given how inconsistently it seems to appear when violent tornadoes impact urban or suburban areas (outside of Jarrell, but it’s obvious why that tornado inflicted the pavement damage it did). It would make sense for these weaker tornadoes inflicting such damage in that sense, at least to me. I’m unsure about the qualities of concrete, asphalt and pavement as a whole though.

I was struggling for the life of me to remember which non-violent tornadoes inflicted asphalt scouring, so keeping these examples in mind is something I’m gonna do, at least if anyone tries to bring up such damage as a violent indicator.

I’m also inclined to believe the thickness of the asphalt plays a major role, but that’s well outside of my depth on whether or not it could be true. The instance of Henryville’s asphalt scouring seemed exceptionally impressive to me, where the road was torn up in 6-inch thick chunks and blasted hundreds of yards away with little debris to do such damage - whereas Joplins didn’t seem as impressive to me despite its extreme appearance, due to the surroundings said tornado was dealing with. Ultimately, though, it makes sense to omit it as a violent indicator for sure.

 

[buckeye05]

I wouldn’t “omit” it per se, as there is still a loose correlation, but it should not used as the primary determining factor when it comes to rationale behind making an argument for a violent rating.

 

[buckeye05]

See this? This is the essence of everything I can’t stand about current state of the online tornado/tornado damage analysis community. So many people who think they know what they’re looking at but don’t, and end up posting their misidentifications and misconceptions as fact, and people just eat it up. Now at least 428 social media users are under the impression that this house in Bremen had its driveway and slab scoured, when in reality it’s just scoured crop stubble covering part of the driveway and garage slab. Next time there’s another Mayfield rating debate, those pics are going to be used as “proof” even though they don’t really show what people will claim they show. It took me about a decade to analyze damage pics with reasonable accuracy and objectively, and I still get things wrong sometimes. It’s frustrating, but there’s not much that can be done at this stage.

 

[TH2002]

See this? This is the essence of everything I can’t stand about current state of the online tornado/tornado damage analysis community. So many people who think they know what they’re looking at but don’t, and end up posting their misidentifications and misconceptions as fact, and people just eat it up. Now at least 428 social media users are under the impression that this house in Bremen had its driveway and slab scoured, when in reality it’s just scoured crop stubble covering part of the driveway and garage slab. Next time there’s another Mayfield rating debate, those pics are going to be used as “proof” even though they don’t really show what people will claim they show. It took me about at decade to analyze damage pics with reasonable accuracy and objectively, and I still get things wrong sometimes. It’s frustrating, but there’s not much that can be done at this stage.
View attachment 44361

(the following post is satire and a guide of what NOT to do when it comes to damage photo analysis)

What do u mean? I have been studying tornado damag for to hole WEEKS and I know what to look for!

Here are classic F5 hall marks:

Well bilt home swept clean:

This is a scoured foundayshin of a home:

Intense ground scoring:

Trees debarcked:

More violetn scouring this tornado should have been EF5 WORST SURVEY EVER:

And another really bad survey from Nws green bay one of the WORST nws offices. This asfalt scouring got rated F2 WHY???

Another home to have its foundayshin scoured, and this photo was posted by an actual meteorologist so I know they cant be wrong... WHY WAS MAYFIELD RATED EF4 I DONT GET IT!!!

And while your at it check out my Favorite YouTuber Swegle Studios... he is really really great. I get all my information from his videos, and I belive him!

 

[Lake Martin EF4]

Ringgold, Cordova

I just remembered the fact you italicized both of these tornadoes. Previously you were siding with EF4 for the former, and while the latter was nasty contextually I didn't see any clear cut EF5 damage from it. What decided things?

 

[buckeye05]

(the following post is satire and a guide of what NOT to do when it comes to damage photo analysis)

What do u mean? I have been studying tornado damag for to hole WEEKS and I know what to look for!

Here are classic F5 hall marks:

Well bilt home swept clean:

This is a scoured foundayshin of a home:

Intense ground scoring:

Trees debarcked:

More violetn scouring this tornado should have been EF5 WORST SURVEY EVER:

And another really bad survey from Nws green bay one of the WORST nws offices. This asfalt scouring got rated F2 WHY???

Another home to have its foundayshin scoured, and this photo was posted by an actual meteorologist so I know they cant be wrong... WHY WAS MAYFIELD RATED EF4 I DONT GET IT!!!

And while your at it check out my Favorite YouTuber Swegle Studios... he is really really great. I get all my information from his videos, and I belive him!

This post raised my blood pressure even though it’s satire

 

[A Guy]

Randy Zipser posting on Stormtrack again:



That's like... not correct, and actually astonishing considering he worked in the field at the same time.

Unless Zipser picked something up when he was at OKU that never made it into published literature and that no-one else has ever mentioned, the opposite's the case. The original Fujita scale isn't based on modelling - Smith's article Zipser mentions says Fujita used 'little to no modelling' - the opposite of what he claims.

In Fujita's 1971 paper he explains the issues around characterising tornadoes by intensity as opposed to length or area and surmises that it should be possible to make five or six intensity categories. He then explains that categories of a certain size should be easy enough to tell apart to make a rough estimate rather than doing complex calculations on individual bits of damage.

In deriving the F scale it's essentially arbitrary, based on a formula of increasing step sizes of about the right size, to connect the Beaufort Scale to Mach 1. He also proposes that as wind speed increases time required for damage decreases, so the scale was originally based on the fastest quarter mile rather than a single timespan. The actual F scale damage categories he assigns are based off the (scant) engineering estimates that existed at the time, in particular Melaragno 1958, and basic engineering calculations of dynamic forces. That's it, he mentions no other specific modelling. There is one group however, who did rely extensively on modelling, and that's... Joe Minor's group at TTU, of which Tim Marshall is the most most enduring protege.

I'd add that the EF scale isn't based on sophisticated analysis that we can actually see. It's based on the averaged opinions of a panel of six (count 'em... six) experts, who reviewed "all available data" (whatever that means) over a couple of days. One of those experts believes that DoD 10 damage to a house can occur at 130 mph.

The current claims that the EF scale windspeeds aren't correct are based largely either on radar observations or on other empirical methods, and don't rely on modelling more than the EF scale or EF scale affirming analyses. Doppler radar data have their problems, it's not a straightforward matter of simply reading the speed off the screen, but they're the most available in situ measurement we've got. The ongoing use of the EF scale is less to do with proven validity that it is to do with institutional inertia (the original F scale was used for over three decades) and people not wanting their turf trodden on.

 

[buckeye05]

Randy Zipser posting on Stormtrack again:

View attachment 44371

That's like... not correct, and actually astonishing considering he worked in the field at the same time.

Unless Zipser picked something up when he was at OKU that never made it into published literature and that no-one else has ever mentioned, the opposite's the case. The original Fujita scale isn't based on modelling - Smith's article Zipser mentions says Fujita used 'little to no modelling' - the opposite of what he claims.

In Fujita's 1971 paper he explains the issues around characterising tornadoes by intensity as opposed to length or area and surmises that it should be possible to make five or six intensity categories. He then explains that categories of a certain size should be easy enough to tell apart to make a rough estimate rather than doing complex calculations on individual bits of damage.

In deriving the F scale it's essentially arbitrary, based on a formula of increasing step sizes of about the right size, to connect the Beaufort Scale to Mach 1. He also proposes that as wind speed increases time required for damage decreases, so the scale was originally based on the fastest quarter mile rather than a single timespan. The actual F scale damage categories he assigns are based off the (scant) engineering estimates that existed at the time, in particular Melaragno 1958, and basic engineering calculations of dynamic forces. That's it, he mentions no other specific modelling. There is one group however, who did rely extensively on modelling, and that's... Joe Minor's group at TTU, of which Tim Marshall is the most most enduring protege.

I'd add that the EF scale isn't based on sophisticated analysis that we can actually see. It's based on the averaged opinions of a panel of six (count 'em... six) experts, who reviewed "all available data" (whatever that means) over a couple of days. One of those experts believes that DoD 10 damage to a house can occur at 130 mph.

The current claims that the EF scale windspeeds aren't correct are based largely either on radar observations or on other empirical methods, and don't rely on modelling more than the EF scale or EF scale affirming analyses. Doppler radar data have their problems, it's not a straightforward matter of simply reading the speed off the screen, but they're the most available in situ measurement we've got. The ongoing use of the EF scale is less to do with proven validity that it is to do with institutional inertia (the original F scale was used for over three decades) and people not wanting their turf trodden on.

So in a nutshell, false/bad information, or at the very least misconceptions and misunderstandings, are being thrown around by people involved with or adjacent to the EF scale in regards its basis, development, and application? Shouldn’t those involved to such a degree have a rock solid understanding of what proceeded and went into it rationale/research wise down to an extremely detailed and accurate degree? When you see misconceptions, falsehoods, and questionable assertions presented even in the upper echelons of this specific corner of the field, it certainly doesn’t bode well or inspire confidence.

It reminds me of my previous anecdote about my NWS friend who made it apparent that he had never heard of El Reno/Piedmont 2011 while we were having a conversation about tornado damage surveying and intensity, and he was sure I was the one who was mistaken until proven otherwise. He was simply adamant that there was only one historic, violent El Reno tornado and that it was in 2013, and that I must have been confused. Very smart guy, and is directly involved in tornado damage surveying, so it was shocking. The crux of my point is that people can be simultaneously brilliant and completely out of touch or misinformed, and it’s a gut punch when that suddenly becomes apparent.

 

[Shakespeare 2016]

So in a nutshell, false/bad information, or at the very least misconceptions and misunderstandings, are being thrown around as facts by people within the inner circle of those directly involved with the EF scale in regards its basis, development, and application? Shouldn’t the people directly involved know about everything that proceeded and went into it rationale and research wise down to an extremely detailed and accurate degree? When you see such misconceptions and falsehoods among the upper echelons of this field, it doesn’t bode well or inspire confidence.

It reminds me of the story I told about my NWS friend who made it apparent that he had never heard of El Reno/Piedmont 2011 while we were having a conversation about tornado damage surveying and intensity, and he was sure I was mistaken until proven otherwise. He was simply adamant that there was only one historic, violent El Reno tornado and that it was in 2013, and that I was confused. Very smart guy, and is directly involved in tornado damage surveying, so it was shocking. People can be simultaneously brilliant and completely out of touch, and it’s a gut punch when that suddenly becomes apparent.

Yeah, a lot of surveyors seem to be out of touch. I think extreme contextual damage non DI'S should be considered when rating tornadoes.

 

[buckeye05]

Yeah, a lot of surveyors seem to be out of touch. I think extreme contextual damage non DI'S should be considered when rating tornadoes.

Yeah we’re aware. Not really related the topic at hand here.

 

[AJS]

See this? This is the essence of everything I can’t stand about current state of the online tornado/tornado damage analysis community. So many people who think they know what they’re looking at but don’t, and end up posting their misidentifications and misconceptions as fact, and people just eat it up. Now at least 428 social media users are under the impression that this house in Bremen had its driveway and slab scoured, when in reality it’s just scoured crop stubble covering part of the driveway and garage slab. Next time there’s another Mayfield rating debate, those pics are going to be used as “proof” even though they don’t really show what people will claim they show. It took me about a decade to analyze damage pics with reasonable accuracy and objectively, and I still get things wrong sometimes. It’s frustrating, but there’s not much that can be done at this stage.
View attachment 44361

That’s the home that was blasted in Smithville fashion If I recall. A lot of the cinder blocks were torn to bits and the stem wall was obliterated. Looking at this, this tweet just sounds like a clickbait tweet to gather attention and overhype tornado damage. This house was clearly part poured concrete and part CMU. Don’t know why some people are still overhyping tornado damage.

 

[Aaron Rider]

See this? This is the essence of everything I can’t stand about current state of the online tornado/tornado damage analysis community. So many people who think they know what they’re looking at but don’t, and end up posting their misidentifications and misconceptions as fact, and people just eat it up. Now at least 428 social media users are under the impression that this house in Bremen had its driveway and slab scoured, when in reality it’s just scoured crop stubble covering part of the driveway and garage slab. Next time there’s another Mayfield rating debate, those pics are going to be used as “proof” even though they don’t really show what people will claim they show. It took me about a decade to analyze damage pics with reasonable accuracy and objectively, and I still get things wrong sometimes. It’s frustrating, but there’s not much that can be done at this stage.
View attachment 44361

Then it will also be used to discredit those who reasonably believe Mayfield was highest grade and discredit those who'd like some reform.

 

[Aaron Rider]

(the following post is satire and a guide of what NOT to do when it comes to damage photo analysis)

What do u mean? I have been studying tornado damag for to hole WEEKS and I know what to look for!

Here are classic F5 hall marks:

Well bilt home swept clean:

This is a scoured foundayshin of a home:

Intense ground scoring:

Trees debarcked:

More violetn scouring this tornado should have been EF5 WORST SURVEY EVER:

And another really bad survey from Nws green bay one of the WORST nws offices. This asfalt scouring got rated F2 WHY???

Another home to have its foundayshin scoured, and this photo was posted by an actual meteorologist so I know they cant be wrong... WHY WAS MAYFIELD RATED EF4 I DONT GET IT!!!

And while your at it check out my Favorite YouTuber Swegle Studios... he is really really great. I get all my information from his videos, and I belive him!

Haven't watched Swegle. How bad is he?

 

[CheeselandSkies]

Randy Zipser posting on Stormtrack again:

View attachment 44371

That's like... not correct, and actually astonishing considering he worked in the field at the same time.

Unless Zipser picked something up when he was at OKU that never made it into published literature and that no-one else has ever mentioned, the opposite's the case. The original Fujita scale isn't based on modelling - Smith's article Zipser mentions says Fujita used 'little to no modelling' - the opposite of what he claims.

In Fujita's 1971 paper he explains the issues around characterising tornadoes by intensity as opposed to length or area and surmises that it should be possible to make five or six intensity categories. He then explains that categories of a certain size should be easy enough to tell apart to make a rough estimate rather than doing complex calculations on individual bits of damage.

In deriving the F scale it's essentially arbitrary, based on a formula of increasing step sizes of about the right size, to connect the Beaufort Scale to Mach 1. He also proposes that as wind speed increases time required for damage decreases, so the scale was originally based on the fastest quarter mile rather than a single timespan. The actual F scale damage categories he assigns are based off the (scant) engineering estimates that existed at the time, in particular Melaragno 1958, and basic engineering calculations of dynamic forces. That's it, he mentions no other specific modelling. There is one group however, who did rely extensively on modelling, and that's... Joe Minor's group at TTU, of which Tim Marshall is the most most enduring protege.

I'd add that the EF scale isn't based on sophisticated analysis that we can actually see. It's based on the averaged opinions of a panel of six (count 'em... six) experts, who reviewed "all available data" (whatever that means) over a couple of days. One of those experts believes that DoD 10 damage to a house can occur at 130 mph.

The current claims that the EF scale windspeeds aren't correct are based largely either on radar observations or on other empirical methods, and don't rely on modelling more than the EF scale or EF scale affirming analyses. Doppler radar data have their problems, it's not a straightforward matter of simply reading the speed off the screen, but they're the most available in situ measurement we've got. The ongoing use of the EF scale is less to do with proven validity that it is to do with institutional inertia (the original F scale was used for over three decades) and people not wanting their turf trodden on.

I tried one post to reason with him, now I don't even bother opening that thread for fear of doing physical destruction to my computer (as with certain political topics).

 

[A Guy]

I tried one post to reason with him, now I don't even bother opening that thread for fear of doing physical destruction to my computer (as with certain political topics).

I was originally going to tag you, but I took so long to write the post I forgot the reason. It's not the only time I've seen him write something at odds with reality.

When I see that kind of thing from a person, it makes me more wary of any research they've done in the past.

 

[N0mz]

So... what do we think of these tors from last night

 

[Grand Poo Bah]

Randy Zipser posting on Stormtrack again:

View attachment 44371

That's like... not correct, and actually astonishing considering he worked in the field at the same time.

Unless Zipser picked something up when he was at OKU that never made it into published literature and that no-one else has ever mentioned, the opposite's the case. The original Fujita scale isn't based on modelling - Smith's article Zipser mentions says Fujita used 'little to no modelling' - the opposite of what he claims.

In Fujita's 1971 paper he explains the issues around characterising tornadoes by intensity as opposed to length or area and surmises that it should be possible to make five or six intensity categories. He then explains that categories of a certain size should be easy enough to tell apart to make a rough estimate rather than doing complex calculations on individual bits of damage.

In deriving the F scale it's essentially arbitrary, based on a formula of increasing step sizes of about the right size, to connect the Beaufort Scale to Mach 1. He also proposes that as wind speed increases time required for damage decreases, so the scale was originally based on the fastest quarter mile rather than a single timespan. The actual F scale damage categories he assigns are based off the (scant) engineering estimates that existed at the time, in particular Melaragno 1958, and basic engineering calculations of dynamic forces. That's it, he mentions no other specific modelling. There is one group however, who did rely extensively on modelling, and that's... Joe Minor's group at TTU, of which Tim Marshall is the most most enduring protege.

I'd add that the EF scale isn't based on sophisticated analysis that we can actually see. It's based on the averaged opinions of a panel of six (count 'em... six) experts, who reviewed "all available data" (whatever that means) over a couple of days. One of those experts believes that DoD 10 damage to a house can occur at 130 mph.

The current claims that the EF scale windspeeds aren't correct are based largely either on radar observations or on other empirical methods, and don't rely on modelling more than the EF scale or EF scale affirming analyses. Doppler radar data have their problems, it's not a straightforward matter of simply reading the speed off the screen, but they're the most available in situ measurement we've got. The ongoing use of the EF scale is less to do with proven validity that it is to do with institutional inertia (the original F scale was used for over three decades) and people not wanting their turf trodden on.

It's so deflating that these highly respected figureheads are just making $hit up now to defend the hill they're destined to "die on". Not only was the EF scale derived from the opinions of only 6 experts, but two of those experts had such egregiously low wind estimates for each DI (100 mph+ below the widely accepted beliefs at the time) it brought down the whole average. It may have been an even comittee of three engineers and three scientists, but the engineers used dirty tactics to get their way, while the meteorologists used actual reasonable estimates.

It's so hypocritical for him to use the line "garbage in, garbage out" to describe Fujita's methods, when their methods have ruined an entire generation of tornado science that has been based on false pretenses of tornado strength probabilities and wind speeds. The original proposal of the EF scale even shrugged off using models because they were "too expensive" and "time consuming" as if there weren't dozens of qualified scientists enthusiastically up to the task.

All that being said, the last few pages in this thread have really raised my optimism that reasonable, passionate, and extremely well informed people still exist in this field and someone with the same passion will influence some positive change sometime in the future. If anything, Randy Zisper's comments really show how desperate his ilk have become to defend the broken system they created. When the lies are this transparent and obvious, it means they've run out of ground to stand on and are willing to risk their good faith to maintain their influence. I think the tide is finally starting to turn against them.

 

[Grand Poo Bah]

Sorry for all the hyperbole, but man, Zisper just has way too many credentials to actually believe he's just making an honest mistake or misremembering there.