Enhanced Fujita Ratings Debate Thread

[Shakespeare2016]

http://204.227.127.200/images/gld/pdf/newsletter0710.pdf#page=13

There are controversy over on how wide this tornado was but if I were to estimate I would say around a mile-wide. Grazulis has it as 1500 yards wide and others have it as a mile to a mile and a half wide.

 

[locomusic01]

Yeah, I've got the Storm Data photos and a few really poor quality newspaper scans. The ones Dean has are high-quality color photos. He posted a couple of them to his website years ago, but it seems they're inaccessible now even with the Wayback Machine. Storm Data does list the names of some of the people who supplied photos, so some day I may see if I can track some of them down.

I checked my archives and I actually do have two color photos:

Edit: Didn't see your second post - nice find! I was hopeful maybe the linked NWS page would have more photos but it's gone now and there don't seem to be any archived copies of it. Bummer. Might contact Goodland and see if they've still got anything, though.

 

[Shakespeare2016]

Yeah, I've got the Storm Data photos and a few really poor quality newspaper scans. The ones Dean has are high-quality color photos. He posted a couple of them to his website years ago, but it seems they're inaccessible now even with the Wayback Machine. Storm Data does list the names of some of the people who supplied photos, so some day I may see if I can track some of them down.

I checked my archives and I actually do have two color photos:

Edit: Didn't see your second post - nice find! I was hopeful maybe the linked NWS page would have more photos but it's gone now and there don't seem to be any archived copies of it. Bummer. Might contact Goodland and see if they've still got anything, though.

Besides this tornado and the Camp Crook tornado there are two other tornadoes that I know of that granulated vehicles and farm equipment and those were Jarrell, Texas on May 27, 1997 and the Chapman/Solomon tornado on May 25, 2016. The Chapman/Solomon tornado ripped up CWR railroad tracks.

 

[warneagle]

Yeah I don't doubt Chapman was EF5 strength at some point (and I think the surveyors said as much)

 

[Shakespeare2016]

Yeah I don't doubt Chapman was EF5 strength at some point (and I think the surveyors said as much)

Here is the survey from the Camp Crook tornado on June 28, 2018. If you also read further a large farm outbuilding was destroyed and it foundation was violently ripped from the ground. While that is not a really strong structure but the foundation being ripped out of the ground sounds quite impressive. Though it also could have been a weak foundation. https://www.weather.gov/unr/2018-06-28

 

[Equus]

I had mentioned on the old TW that the Chapman tornado's damaging of mainline Class One CWR (continuously welded rail) is as far as I know unprecedented in the modern era. Modern mainline rail weighs ~130 lbs per yard and is thermite-welded in lengths of at least hundreds of feet. Older lighter jointed rail in past decades, sure, it reportedly happened occasionally, but never on a track with modern standards and rail weight. Union Pacific repaired it pretty fast so no analysis was done as far as I know, but if the damage was dealt via wind and not impact with big debris at high speed (which is very possible) that absolutely counts as "incredible damage".

 

[locomusic01]

I'm still not really sure what to make of that rail line damage. I've certainly never seen or heard of anything quite like it. As you said, there have been historical accounts of damage to rail lines (the 1896 Sherman tornado was noted to have caused incredible damage to rail lines, and the Tri-State tornado reportedly scoured away several hundred feet of trackbed and ripped up the overlying tracks somewhere between Gorham and Murphysboro), but nothing modern like this.

A large missile impact would definitely make sense, but that's not really what it looks like to me. On the other hand, it's hard to imagine that big a section of rail being displaced so far by the force of the wind alone. Especially when the trackbed isn't scoured out, which might have otherwise exposed it a bit more. My knowledge of rail lines is super limited so I don't really know enough one way or the other, but it's hard for me to wrap my head around. I'd be really curious to know what kind of force it'd actually take to accomplish something like that.

Pretty obvious there were some seriously violent winds there, though. The vegetation damage in that immediate area was really intense as well.

 

[locomusic01]

I forgot to post this earlier. While I was researching some other stuff, I dug up the paper on tree damage in Joplin and Tuscaloosa I vaguely remembered seeing a while back. You can read the full thing here, but the bit that's relevant to this thread regards Tuscaloosa:

Nevertheless, the best-fit vortex is shown with Rmax = 200 m (estimated from aerial photography), Vtan = 36 m s−1, Vr = 76 m s−1, and peak winds near 99 m s−1. The results suggest that the tornado was of EF5 intensity during this stage of its life, despite the EF4 rating assigned by the NWS. The EF4 rating may be attributable to a lack of EF-scale damage indicators in this section of the track, as shown in Fig. 3c. Of interest is that the tornado destroyed a railroad bridge during this period, as noted in Fig. 3c, implying very high wind speeds, but this indicator could not be used in the NWS assessment (K. Laws, NWS Birmingham, 2012, personal communication). Again, strong radial near-surface winds were needed to produce the best-fit tree-fall pattern.

The idea that Tuscaloosa may have reached EF5 intensity at some point is hardly news, obviously, but I thought it was noteworthy that the area of probable EF5 intensity using this method more or less coincides with the prevailing opinion. This area includes, among other things, the steel train trestle over Hurricane Creek Canyon and the coalyard where coal cars were thrown up to ~400 feet. This approach of using treefall patterns as a proxy for intensity seems to be pretty sound, so it's worth consideration I think. The study also suggests that max intensity coincided with clear RFD internal surges, which has been documented in several other tornadoes and is backed up in this case by both treefall damage patterns and radar data.

The structural damage nearest this area was already borderline EF4-5, and there are multiple lines of evidence that suggest it further intensified from there, so at this point I feel pretty comfortable mentally classifying Tuscaloosa as an EF5.

Also, just for convenience, note that 99 m/s is ~222 mph. The max wind speed estimate for Joplin in this study was 104 m/s, which is ~233 mph.

 

[locomusic01]

So I'm not sure how I missed this when it came out - or maybe I didn't and I've just forgotten? I dunno - anyway, Tim Marshall was on WeatherBrains last year and offered a lot more info on the ongoing EF-scale reevaluation process.



If the video doesn't jump to the right time, it starts around 1:10:46.

I also saw a presentation by Jim LaDue somewhere but can't find it right now. He also gave an overview of the process and everything they intend to cover. I've also reviewed some of the papers that've come out on the topic. Anyhow, I'd really recommend watching Tim's section of the video (and checking the other stuff if you can find it, or if I can track it down again and edit in links later), but the summary is that it sounds like they're hitting on pretty much everything many of us have brought up over the years.

Maybe most notably, there are proposals for utilizing direct measurements (mobile radar, mesonet, etc) in ratings in some form or another. They may be used as a basis for ratings or they may be added as sort of a secondary rating when the info's available (e.g. El Reno 2013 might have been given an EF-3-5 rating or somesuch, indicating an EF3 damage-derived rating and an EF5 wind speed). I still prefer the former, personally, but the latter would at least be progress. Either way, they're also working on developing standards for what sorts of measurements are usable, how to handle different measurement heights, etc.

They're also in the process of adding a bunch of new DIs (yay!), changing or eliminating some existing ones, reevaluating the DODs and so on. Some of the specific ones I've seen mentioned are jersey barriers (think Mayflower/Vilonia), grain bins, center-pivot irrigation systems, parking stops (think Joplin) and expanding the types of manufactured homes. They're also working to develop very specific guidance for assessing each DI/DOD and working on improving training so that there's better consistency between WFOs.

Marshall and others have also mentioned that they're taking the vegetation stuff completely out of the engineering side of things and turning it over to arborists and other experts who have more relevant experience. I know there's been talk of trying to create a more comprehensive set of "natural" DIs, so to speak, including different types of vegetation. I'd imagine the hangup there will be coming up with something that accounts for the huge number of variables that might influence vegetation damage without having to run extensive tests for each survey, which would obviously be very impractical. If it can be figured out, that'd be really big. It appears they're also considering ways to incorporate treefall pattern data like I mentioned above, which is pretty exciting. It's a really fascinating approach, and if it's actually as useful as it appears, it'd also be a huge help for rural tornadoes.

Oh, and they're also pushing for all surveys to be done through the Damage Assessment Toolkit, which would be awesome for us. All in all, it sounds like we should be seeing a much-improved EF-scale once this process is finished. Marshall and LaDue have both mentioned 2020 as a tentative timeframe, though of course it then depends on how quickly the recommendations are adopted once released.

 

[SilentShadow87]

I think this is good news. The point of the EF scale should be "how strong was this tornado to the best of our knowledge", nothing more and nothing less. While we don't want to see a return to the 1950s-1970s style of massively overrating some tornadoes (e.g. Tanner II on 4/3/74), I definitely think more DIs should be incorporated into the scale.

 

[locomusic01]

Yeah, I think any and all credible data should be considered. The ultimate goal is to better understand climatology and get a more accurate picture of when and where tornadoes happen and how intense they are. As it currently is, the EF-scale probably tells us more about population density and construction practices than maximum tornado intensity. It'll always be a tough thing to navigate, but I don't know that there's much of a downside to using all possible tools at our disposal (provided they're quality-controlled and reflective of near-surface tornado intensity).

 

[Shakespeare2016]

Yeah, I think any and all credible data should be considered. The ultimate goal is to better understand climatology and get a more accurate picture of when and where tornadoes happen and how intense they are. As it currently is, the EF-scale probably tells us more about population density and construction practices than maximum tornado intensity. It'll always be a tough thing to navigate, but I don't know that there's much of a downside to using all possible tools at our disposal (provided they're quality-controlled and reflective of near-surface tornado intensity).

That is a nice way of putting it.

 

[Peter Griffin]

Figured I would just give my opinion on the Chapman tornado and the damage it did to the railroads. I remember seeing the pics of the tracks and thought that it had to be mostly from debris impact. One of the bends in the track was just too sharp it didn't appear that it could be done by wind alone. I think debris impacted it at this point (possibly others) and resulted in the bending and shifting of the track. I'll see if I can dig up the pic later and point out what I am talking about. There is not doubting the Chapman tornado was violent but I don't think wind alone is responsible for the track damage. I'm no engineer though lol.

 

[Shelby]

I was

Don't forget this and it was far past the tornado's peak.
As the tornado moved across a coal yard in this area, a 35.8-tonne (78,925 lb) coal car was thrown 391 ft (119 m) though the air.
To me, when watching the tornado track video, the 70 or so miles of timber completely flattened in a mile wide swath is more indicative of an EF5 than a tornado that briefly causes a thousand feet of "EF5" damage. Only the strongest of strong tornadoes can pack that kind of wallup for that many miles. The fact that the Hackle-Campbell did that for 132 miles is incomprehensible.

They gave the Joplin tornado an EF5 rating based on the amount of damage which it did a lot of but it didn't do 80 miles of destruction.

JOPLIN, Mo. — A new engineering study of the damage caused by the May 2011 tornado that struck Joplin found no evidence that it was an EF5, as the National Weather Service found, because the city's homes and businesses weren't built to withstand wind speeds that strong, making such a determination impossible.

(MORE: Magnets for Disaster?)

The study by the American Society of Civil Engineers found that more than 83 percent of the damage on May 22, 2011, was caused by winds of 135 mph or less, which is equal to the maximum wind speed of an EF2 tornado, and that about 13 percent of the damage was caused by winds of 138-167 mph, consistent with an EF3 tornado. Only 4 percent of the damage was indicative that it had been an EF4 tornado, which can have winds speeds ranging from 168 to 199 mph, the report said.

The ASCE team also found that while the tornado's maximum wind speed was around 200 mph, there was no evidence of building damage from winds at 200 mph or greater, the minimum threshold for an EF5. The ASCE investigators concluded it was impossible to find evidence of E-5 ratings in the damage because none of the buildings met the high construction quality threshold required for determining that level of wind speed, The Joplin Globe reported Saturday.


The findings of the ASCE damage-assessment team are based on five days of surveying damage in more than 150 buildings in a six-mile segment of the tornado's Joplin path. The total tornado path was 22 miles. More than 7,000 structures were destroyed or badly damaged by the tornado, and 161 people were killed.

The ASCE findings, however, do not change the National Weather Service's classification of the Joplin tornado as an EF5, with peak winds of 200-208 mph.

I put absolutely no stock in that report. I was in Joplin after the tornado and saw numerous damage indicators that were indicative of EF-5 damage. The area is HEAVILY forested and not a single tree was standing in the path. Every tree was debarked and torn to shreds. Foundations were missing or portions were torn up, concrete porches were torn out and thrown away, one hundred pound manhole covers were sucked out of the roads, a rubber hose was impaled in a tree. The nursing home was destroyed and partially swept away with 21 fatalities occurring, Concrete parking stops weighing over 300 pounds and anchored with rebar were torn loose and thrown over fifty yards, A semi-truck was thrown roughly a quarter of a mile, stripped to the steel frame and wrapped around a debarked tree. Grass was scoured and pavement was torn out from roads.

A friend of mine was in the tornado and thrown several blocks and survived. She moved to Europe for several years due to the emotional trauma.

 

[CheeselandSkies]

The ASCE study on Joplin is utter BS. A tornado simply does not kill >150 people in this day and age without being exceptionally violent (barring some freak nightmare scenario like tracking along a gridlocked freeway at rush hour, or hitting a stadium just as a big event lets out). Even Tuscaloosa-BHM "only" had about half that death toll despite hitting similarly populated areas. Lower-bound violent tornadoes (St. Louis '11, Hattiesburg '13 off the top of my head) have struck cities this decade with no fatalities at all.

 

[warneagle]

Given what the EF scale has revealed about home construction practices, I could believe that most of the people killed were in homes that couldn’t withstand winds that were well below the EF5 threshold, but there were enough contextual indicators to justify the rating. I’m not sure I’ve seen the ripped out manhole covers and parking stops happen in any other tornado.

 

[Shelby]

I was


I put absolutely no stock in that report. I was in Joplin after the tornado and saw numerous damage indicators that were indicative of EF-5 damage. The area is HEAVILY forested and not a single tree was standing in the path. Every tree was debarked and torn to shreds. Foundations were missing or portions were torn up, concrete porches were torn out and thrown away, one hundred pound manhole covers were sucked out of the roads, a rubber hose was impaled in a tree. The nursing home was destroyed and partially swept away with 21 fatalities occurring, Concrete parking stops weighing over 300 pounds and anchored with rebar were torn loose and thrown over fifty yards, A semi-truck was thrown roughly a quarter of a mile, stripped to the steel frame and wrapped around a debarked tree. Grass was scoured and pavement was torn out from roads.

A friend of mine was in the tornado and thrown several blocks and survived. She moved to Europe for several years due to the emotional trauma.

After causing five deaths at the Elks Lodge, the tornado continued to rapidly intensify as it devastated medical buildings just west of St. Johns Hospital. Large, steel-anchored parking stops weighing approximately 300lbs were ripped from the ground and hurled more than 50 yards. Partha Sarkar, a wind engineer from Iowa State University, concluded that winds of at least 205mph were required to uproot the parking stops (Sarkar, 2011). Winds of that intensity only inches above the ground are indicative of significantly stronger winds at roof-top level. (Image presented at 2012 AMS Conference)

 

[CheeselandSkies]

Time for the standard post-outbreak bump to this thread. The Pickens County tornado remains an EF2 several days after the rating was initially released, despite it seeming that the level of damage to some of the site-built homes along Settlement Road would support at least low-end EF3 by the guidelines of the scale regardless of construction quality (I.E. home completely demolished in EF4 fashion, but was poorly built/anchored).

 

[SilentShadow87]

Yeah under the rating standards of recent years EF3+ ratings seem extremely difficult to give. It's reminding me of the Elon, VA tornado from 4/13/2018. Yes, there are legit construction quality issues. No, I'm not saying they should have pulled the trigger on an EF4 rating - not even close. But if they're literally going to rate lower than the lowest possible DI on the scale, then what's the point of even having the scale in the first place?

Dixie offices seem to have been lowballing ratings for most of the past decade, honestly. I think one example which doesn't get talked about a lot is the Houston/New Wren, MS tornado on 4/27/2011. You had several homes completely flattened (though I will give them that they weren't very well built), trees debarked to bare trunks, and a 5,000 gallon tank was tossed almost half a mile. It's debatable, sure, but based on descriptions and photos of the damage the Houston tornado seemed closer to low-end EF4 than the official mid-range EF3 rating it got.

 

[Equus]

Looks like the middle Tennessee tornadoes will be finalized very reasonably (though Cookeville would have unquestionably been an F5 by pre La Plata standards) but another rating has irked me a bit over the last month, the killer Marengo County tornado from February 6th in BMX territory-

This tornado was rated 105-110mph EF1. Single wide mobile home destruction can only be rated above EF2 in exceptional circumstances but the destruction of these two - with the death of an occupant in one - was only rated high EF1. The rule of thumb has been, rate mobile home damage F/EF1 if the structure is still recognizably intact and F/EF2 if it's completely destroyed, and I don't see how one can destroy a mobile home steel frame much more thoroughly than this. Even if not strapped down this would seem to be classic EF2 damage, and the only justification to keeping it below that threshold I can think of is if there were walls or floor sections still intact out of the photo, but that seems unlikely. It's rated 105-110 and 105 is below the minimum allowed threshold for DoD 9; the damage survey viewer even confirms that BOTH of these particular damage points were rated DoD 9 which means turning the slider to the absolute minimum of 110 on TWO separate mobile home obliterations to avoid an EF2 rating. My question is... why?

Also, yeah, the deadly Settlement Road tornado in Pickens County from Jan 11th is still worth mentioning again...

In theory, this brick home damage matches up with the high EF2 the tornado received (that survey point is listed as 132mph, the expected value for the exterior walls collapsed DoD 7) but one could make the argument that this is more standard for low EF3 (between DoD 7 and DoD 8 perhaps) - it feels like what used to receive low to mid EF3 now receives high EF2, and requires many more interior walls down before an EF3 is considered. Here's another house also rated ~130mph EF2 from the damage viewer - damage which a few WFOs would probably rate low EF3. High EF2 seems reasonable imo, but I think this would have been easily F3 a couple decades ago.

These are the two killer tornadoes in BMX territory thus far this year, and either could have plausibly been rated a category higher; Marengo especially.