5/21/24 Greenfield tornado rating

[Matthew Crowther]

I see that the preliminary rating of the Greenfield tornado is EF-3. If that goes down as the final rating, that is as big of an underestimate, in my opinion, as El Reno '13. Doppler winds > 200, Timmer's video looks like Andover '91 on steroids (do they still do photogrammetry?). At the very least it was a 4, but what do I know? I know that they are considering using non-building DI's, IMO that can not come soon enough.

 

[Dave C]

The BEST project that CSWR DOWs and scouts are involved with this year is still is not down to ground level with radar scans where most damage happens. I believe they are sampling at 150-180 feet on this storm from preliminary reports. I am not a fluid dynamics expert but I think there are valid arguments for faster or slower surface winds than doppler samples taken tens of meters above depending on vortices, which side of the tornado is encountered, etc.

I am somewhat of an advocate of rating based upon damage potential because it is silly to measure an EF5 over grassland and call that an EF0 if it hits nothing. However, there are many complexities. EFx Tornadoes are not EFx everywhere. They may fluctuate wildly in strength, ground speed, direction with area and that all affects severity in terms of damage potential. We cannot measure surface winds yet with radar or photogrammetry (far too complex and expensive for sub ~20mph accuracy with both technologies right now). Perhaps when RaXPol synthetic beam steered aperture radar is available, it will be feasible to get down to the ground level with radar.

Since so much politics like building codes, warning policy, insurance payouts, etc. can be affected by reporting severity of events accurately (like when is a disaster declared and when not, when is a payout forthcoming and with what restrictions on rebuild based upon building codes, etc.) it seems unlikely and perhaps even unwise to completely decouple from damage indicators.

Personally I would like to see wind speed considered when relevant to damage, and always registered with records when available. This would give us what we need for science and in some cases damage politics, without totally divorcing from the damage indicators method that still has some merit but definitely needs to be updated to consider more characteristics of wind fields and perhaps refine ratings.

In the end different audiences use the EF scale differently, so it will be interesting to see what comes from the ongoing effort to refine it.

 

[Dan Robinson]

May be a topic for another thread, but does anyone know if there is any new info about the pending EF scale updates? It's been more than a year since we've heard anything. I know these things take time, but surely we're closer to something after a year.

Doswell's original blog post about tornado ratings has been removed. I'm wondering if the consensus in the science community about using mobile radar measurements has changed since then?

 

[Drew Terril]

I did find it interesting that the DOW reading was from (if memory serves) 44m AGL. I know there have been instances in the past (2011 El Reno-Piedmont, for example), where readings were obtained much closer to the ground, so I'd be curious as to what limiting factors were in play, given their positioning fairly close to the storm. My guess would be that terrain played a role; they may not have had an available vantage point that would have allowed them to sample closer to the surface. Either way, I would be very interested to hear what those limiting factors were.

As far as the final rating, most (if not all) NWS offices get input from civil engineers to sign off on an EF-4 or higher rating. This is not in any way intended to be a knock on meteorologists, but the majority have no background in either engineering or construction. So it's absolutely the right thing to call in someone who does have that expertise. Going by a checklist can only do so much. Someone could write out line by line instructions for programming on a computer, but because I don't have that expertise, I'd be SOL the moment there's a hiccup. Just like I could write out instructions for someone on rebuilding a transmission, but without that other person having experience, those instructions would be inadequate to the task. That said, my guess based on what I saw is the finally rating will be somewhere in the 180-185 range. Obviously, there's probably damage that I haven't seen that could possibly skew that upwards.

Anytime the discussion comes up on discord, I tend to nitpick any pictures that are posted because I did work off and on for several years in construction and know how things (particularly as it pertains to single family homes) are supposed to go together. Something I noticed both from Tuesday and from the event in Nebraska a few weeks ago, was a puzzling inconsistency in the application of anchor bolts on single family homes. While many states don't adopt ICC building codes (and I don't believe building codes in Iowa or Nebraska are up to that level), it's still a good guideline of how things are supposed to be. The thing I noticed with the anchors was varying distances between anchor bolts, that I couldn't attribute to the standard of being 4-12" from each end of a sill plate. There were multiple instances that I could see where there wasn't an anchor within 12" of a corner, and others where there was more than 6' between anchors (which is the maximum allowed distance between anchors for a single story home; two story homes require them no more than 4' apart). It just didn't make a lot of sense, as someone who used to do that for a living, to go to the effort of installing anchor bolts and not do it correctly.

I've heard many people over the years say that FEMA payments in particular vary based off the rating of a tornado, and that it's supposedly why some borderline tornadoes haven't gotten the higher rating. I have yet to see any kind of written policy that spells that out, so I wonder how much of that is urban legend. I'm pretty sure as far as insurance payouts, anything EF3 or above is likely to cause a "total loss", where it's easier and cheaper to raze and rebuild than to repair the existing structure. Given that homeowner insurance rates are based in part on the value of the home and cost to rebuild to equivalent spec, I find it difficult to believe that there would be a different insurance payout with the same level of damage over a rating.

 

[Mike Smith]

We would be far better off going back to the original Fujita Scale (it was changed primarily out of envy from what I have heard). The original allowed for wind speed measurements (anemometers, Doppler radar) as well as winds that are more realistic with what are measured in the field.

 

[Mike Smith]

Fox Weather is reporting that it has been rated EF-4.

 

[John Farley]

Mike is correct. EF-4.

https://kamala.cod.edu/offs/KDMX/2405232058.nous43.html

 

[Drew Terril]

What I saw from DMX was that they have it down as 175-185, so my guess based on what I'd seen seems to have been spot on.

 

[Jeff Duda]

The BEST project that CSWR DOWs and scouts are involved with this year is still is not down to ground level with radar scans where most damage happens. I believe they are sampling at 150-180 feet on this storm from preliminary reports. I am not a fluid dynamics expert but I think there are valid arguments for faster or slower surface winds than doppler samples taken tens of meters above depending on vortices, which side of the tornado is encountered, etc.

I am somewhat of an advocate of rating based upon damage potential because it is silly to measure an EF5 over grassland and call that an EF0 if it hits nothing. However, there are many complexities. EFx Tornadoes are not EFx everywhere. They may fluctuate wildly in strength, ground speed, direction with area and that all affects severity in terms of damage potential. We cannot measure surface winds yet with radar or photogrammetry (far too complex and expensive for sub ~20mph accuracy with both technologies right now). Perhaps when RaXPol synthetic beam steered aperture radar is available, it will be feasible to get down to the ground level with radar.

Since so much politics like building codes, warning policy, insurance payouts, etc. can be affected by reporting severity of events accurately (like when is a disaster declared and when not, when is a payout forthcoming and with what restrictions on rebuild based upon building codes, etc.) it seems unlikely and perhaps even unwise to completely decouple from damage indicators.

Personally I would like to see wind speed considered when relevant to damage, and always registered with records when available. This would give us what we need for science and in some cases damage politics, without totally divorcing from the damage indicators method that still has some merit but definitely needs to be updated to consider more characteristics of wind fields and perhaps refine ratings.

In the end different audiences use the EF scale differently, so it will be interesting to see what comes from the ongoing effort to refine it.

There's a ton of seeds for really good and long-running philosophical discussions within these posts, Dave! Glad you brought it to light.

My two cents on the biggest issue is the meteorological vs. non-meteorological factors and how much we weigh them. The (Enhanced) Fujita scale, while intended to be a tornado intensity scale, has always been used as a tornado damage scale instead. That's the reason why NWS refused to rate El Reno 2013 as EF5 despite the robust measurements of wind speeds a little above the ground that were way above the EF5 threshold.

You're also correct that tornadoes are entered into the official Storm Events database using their maximum intensity and maximum path width rather than average values. This is also problematic because, as you've said, even an EF5 tornado is not EF5 along its whole track. Most EF5s only produce a handful of actual EF5 damage instances, and again, that's based on damage. Only in idealized simulations have I seen tornadoes in which the wind speeds exceeded EF4 or EF5 thresholds consistently over a long period of time. And given the fractal nature of the structure of wind velocities in tornadoes, it is difficult to know for sure what the 10-m AGL wind speeds are in tornadoes on a second-by-second basis. This fact will probably forever restrict a fully proper rating system that addresses both human infrastructural and scientific needs.

I agree with you, Dave, that I would also like to see mobile Doppler radar wind speed measurements used to influence, if not corroborate or set, tornado ratings (provided certain location and quality standards are met). Unfortunately, surface-based anemometers will probably never be able to provide in-situ measurements unless we start installing sonic anemometers that are encased in steel mesh cages that are mounted on 1-foot diameter steel poles that anchor 30 feet into the ground (and that happen to actually take a direct hit from a tornado). Nearly all anemometers fail readily in direct tornado hits, usually due to debris striking them. Every once in awhile we get a bit lucky, though (e.g., the TIV2 mounted anemometer reading 175 mph before finally failing in the Lebanon, KS tornado in May 2013). But once in 11 years isn't going to do much for statistics.

 

[Sean Ramsey]

I wonder if there wouldn't be some value in two measurements in one rating, with one of them being top measured wind speed via best measurement method (maybe representing potential for damage) and another with actual realized damage.

So if you have wedge with top end wind speeds that touched down in a field it would be like an E5D0
Or if you have a wedge that caused considerable damage it would be like an E5D5
Or a bird fart that touched down in town would be an E0D0
Or a cone that does moderate damage might be an E2D3.

Ok I've thought about this enough for today.

 

[Dave C]

I wonder if there wouldn't be some value in two measurements in one rating, with one of them being top measured wind speed via best measurement method (maybe representing potential for damage) and another with actual realized damage.

So if you have wedge with top end wind speeds that touched down in a field it would be like an E5D0
Or if you have a wedge that caused considerable damage it would be like an E5D5
Or a bird fart that touched down in town would be an E0D0
Or a cone that does moderate damage might be an E2D3.

Ok I've thought about this enough for today.

I really like that idea actually, as it serves all audiences that use the ratings. Also, they sound like star wars droid names, so there is that.

 

[Diana Nevins]

I wonder if there wouldn't be some value in two measurements in one rating, with one of them being top measured wind speed via best measurement method (maybe representing potential for damage) and another with actual realized damage.

So if you have wedge with top end wind speeds that touched down in a field it would be like an E5D0
Or if you have a wedge that caused considerable damage it would be like an E5D5
Or a bird fart that touched down in town would be an E0D0
Or a cone that does moderate damage might be an E2D3.

Ok I've thought about this enough for today.

I think that's a great idea, but it would also be nice to find a way to add the estimated height of the winds measured by the Doppler radar. A DOW measurement of winds 150 feet above the ground if going to be more significant than a stationary Doppler measurement of wind speeds 2,000 feet above the ground,