Assigning EF Scale Ratings Based on Radar Indicated Wind Speeds

[Hoosier]

With the El Reno, OK tornado being the latest to be rated based on radar wind speeds, I thought I would start this thread to gain some input from knowledgeable folks on this fairly recent trend.

I'm certainly not a professional but to me it gets away from the intent of the old F scale and newer EF scale, which, as we all know, is designed to rate tornadoes based on the damage they produce, and THEN estimate a wind speed from the degree of damage. It's no secret that the scale is biased against tornadoes in rural areas since there are less targets to hit, but this has more or less been an accepted reality. While radar measurements may be a truer reflection of how intense a tornado is, there are a number of issues that I see, perhaps the biggest being that only a very small number of tornadoes are being sampled in the field. Many other tornadoes in the Plains and elsewhere will still be rated just based on damage, which adds additional inconsistencies into a database that already has them.

As someone with interest in local and national tornado climatology, I admit it bothers me to see this new practice of not rating certain tornadoes based on the level of damage. If it's going to continue, I would hope to see some sort of special symbol/notation to indicate which tornadoes were rated based on radar.

What say you?

[Ensō]

I share your concerns about the climatology, especially given that mobile radars are most prevalent (almost exclusively so) in the Plains during peak season. That is obviously going to create a bias toward more intense tornadoes in that time period and region. It's a problem to be sure, but I still favor rating via radar because we should ultimately be aiming for the most accurate record we can obtain. This is a step in the right direction IMO. It creates additional biases, but it at least partially solves others.

[AtticaFanatica]

The purpose has always been to estimate the wind speeds in the tornado. Always. The need for DIs to infer intensity is an unfortunate necessity built upon the inherent small spatial scale and short time scale of the phenomenon.

The intensity database is essentially crap, it might generally be accurate to within +/- 1 EF scale. I don't understand the obsession with willfully keeping it that way in the name of consistency.

Let's put it this way: what scientific question can we better answer by closing our eyes, covering our ears, and telling everybody it was an EF3?

[andyhb]

The purpose has always been to estimate the wind speeds in the tornado. Always. The need for DIs to infer intensity is an unfortunate necessity built upon the inherent small spatial scale and short time scale of the phenomenon.

The intensity database is essentially crap, it might generally be accurate to within +/- 1 EF scale. I don't understand the obsession with willfully keeping it that way in the name of consistency.

Let's put it this way: what scientific question can we better answer by closing our eyes, covering our ears, and telling everybody it was an EF3?

 

The bolded basically nails my thoughts on it, if you do have the data to use, why not use it? It provides a more accurate indicator on tornado intensity than DIs ever will, with credits to the efforts in creating the EF-scale. I know I'm not the only one that thinks that if we did have radar data on some of the violent 4/27/11 tornadoes in Dixie Alley (to use one case as an example), we would've seen some incredible measurements as well.

[Ensō]

The purpose has always been to estimate the wind speeds in the tornado. Always. The need for DIs to infer intensity is an unfortunate necessity built upon the inherent small spatial scale and short time scale of the phenomenon.

The intensity database is essentially crap, it might generally be accurate to within +/- 1 EF scale. I don't understand the obsession with willfully keeping it that way in the name of consistency.

Let's put it this way: what scientific question can we better answer by closing our eyes, covering our ears, and telling everybody it was an EF3?

 

Exactly. I understand the concerns since it will change the climatology that we're all familiar with, but if we're familiar with a pretty worthless climatology then what harm can come from changing it? If nothing else, at least we may have a better idea of what's going on locally intensity-wise.

[Hoosier]

Something I've been wondering that I didn't mention in the original post is do we know how good the measurements are...more specifically, are there times when there may be a big discrepancy between radar indicated wind speeds and what occurs at ground level?

[Buckeye05]

If this practice continues to become more common, I think the amount of EF5 tornadoes discovered using this method will be pretty surprising for some people.

[thewxmann]

If this practice continues to become more common, I think the amount of EF5 tornadoes discovered using this method will be pretty surprising for some people.

 

EF4 and EF5 both. Rozel, Bennington, El Reno would all have been EF3's or lower without Doppler. 

[andyhb]

If this practice continues to become more common, I think the amount of EF5 tornadoes discovered using this method will be pretty surprising for some people.

 

I'm not so sure on this, tornadoes of the intensity of the El Reno tornado the other day are still extraordinarily rare. There are maybe a handful of tornadoes since the EF-scale was introduced (Bowdle comes to mind) that may have been rated as such based on added radar measurements, at least from my point of view. I'd tend to think the Carney tornado from 5/19 is in the same boat.

[The Great Zo]

What makes it a little bit difficult for me to reconcile is that we're using radar estimates to rate a tornado on a damage scale -- there's a fundamental incongruence that does bother me to an extent. The NWS is very clear in its outreach (and tornado survey methodology) that tornadoes are only rated by the damage they leave behind (caveats about available targets, or a lack thereof, aside).

 

Yet, as a scientist, I want to see all potential data sources used to properly estimate the intensity, location, width, and track of any storm. Having to shoehorn mobile radar data into the EF-scale is tricky. It's the best we've got for 99% of tornadoes, but saying the El Reno storm was an "EF5" doesn't match up with to how the EF-scale is used in almost all other cases -- at least with how the system is currently set up to work.

 

Ultimately, those involved in the rating decision made the best possible choice -- ignoring the radar data would simply not have allowed the whole story to be told. In the future, I hope national policies are clarified to allow for this, and in these rare cases, make it known that it's acceptable to augment the EF-rating and wind speed based on mobile radar or other measurements.

[Dunkman]

At the end of the day they just can't ignore the data.  I really think it's that simple.  If it means they have to redo the scale again, then that's what has to happen.

[tornadotony]

I feel like ignoring the mobile radar data would be tantamount to covering our eyes and our ears and going "LA LA LA LA LA" at the top of our lungs.  If we're going to have a rating scale that has both damage AND winds, then the wind part must be taken into account, even if the scale's legacy is to use the damage.  Otherwise, we have no business estimating wind speeds in tornadoes.

[wxmeddler]

I feel like ignoring the mobile radar data would be tantamount to covering our eyes and our ears and going "LA LA LA LA LA" at the top of our lungs.  If we're going to have a rating scale that has both damage AND winds, then the wind part must be taken into account, even if the scale's legacy is to use the damage.  Otherwise, we have no business estimating wind speeds in tornadoes.

 

Perfectly said. There needs to be a merger of engineering and science here. So far we've just been using engineering to account for wind speeds. What needs to happen is take damage, such as from Moore, and compare them to the radar data from TDWR, WSR88D, Phased Array etc. and then see where the engineering calculations for wind and the observations match up and where they don't. My hunch is that the engineering calculations for the damage on the higher end windspeeds are slightly underdone.

[DBriedis]

http://www.tulsaworld.com/article.aspx/El_Reno_tornados_EF5_rating_needs_closer_look_NWS_official/20130606_777_0_ANatio823340

 

SPC seems to want ground damage proof, not just radar evidence, before it will agree that it was an EF5.

[soonertrey]

I'm not so sure on this, tornadoes of the intensity of the El Reno tornado the other day are still extraordinarily rare. There are maybe a handful of tornadoes since the EF-scale was introduced (Bowdle comes to mind) that may have been rated as such based on added radar measurements, at least from my point of view. I'd tend to think the Carney tornado from 5/19 is in the same boat.

 

I was on the Carney tornado.  I was surprised when the rating came out as EF-3.  I've seen quite a few tornadoes and it had some of the most violent motion I have ever seen with my own eyes before it hit Carney.  I passed a parked mobile doppler that was scanning.  It wasn't a DOW, so I assume it was maybe RaXPOL.  Regardless, I haven't seen any documentation about any velocity data that the mobile radar picked up on this tornado.  Has anyone else?  They did have to deal with a hill and some trees in the distance that could have had a negative effect on their lowest scans.  

[blackjack123]

http://www.tulsaworld.com/article.aspx/El_Reno_tornados_EF5_rating_needs_closer_look_NWS_official/20130606_777_0_ANatio823340

 

SPC seems to want ground damage proof, not just radar evidence, before it will agree that it was an EF5.

    I kind of wondered the same thing a lot of you are wondering. I am no expert but I would believe there would be at least some complete tree debarking, vehicles wadded beyond unrecognition, or extreme massive ground scouring associated with this tornado. There was one tornado a couple years ago near El Reno that was upgraded to EF5 based on Doppler radar but it did borderline EF4/EF5 damage to prove it was likely an EF5.

[andyhb]

    I kind of wondered the same thing a lot of you are wondering. I am no expert but I would believe there would be at least some complete tree debarking, vehicles wadded beyond unrecognition, or extreme massive ground scouring associated with this tornado. There was one tornado a couple years ago near El Reno that was upgraded to EF5 based on Doppler radar but it did borderline EF4/EF5 damage to prove it was likely an EF5.

 

That was not borderline EF5 damage that the 2011 tornado caused, it did some truly incredible things.

[andyhb]

http://www.tulsaworld.com/article.aspx/El_Reno_tornados_EF5_rating_needs_closer_look_NWS_official/20130606_777_0_ANatio823340

 

SPC seems to want ground damage proof, not just radar evidence, before it will agree that it was an EF5.

 

Since you can see the scar from high altitude, I'd at least think there is ground scouring to support the EF5 rating, the problem is that there were very few structures where the tornado tracked. It's going to be a bit of a fiasco if this is ultimately downgraded (especially to an EF3) after all of this analysis and the use of radar to rate the Bennington and Rozel, KS tornadoes.

[tornadotony]

Since you can see the scar from high altitude, I'd at least think there is ground scouring to support the EF5 rating, the problem is that there were very few structures where the tornado tracked. It's going to be a bit of a fiasco if this is ultimately downgraded (especially to an EF3) after all of this analysis and the use of radar to rate the Bennington and Rozel, KS tornadoes.

It won't be, not with the wind speed announced publicly. The folks at SPC can think what they want, but in the end, OUN publishes the Storm Data entry on it.

[Ensō]

I think the thing to keep in mind regarding the EF scale being a "damage scale" is that it's a damage scale out of necessity, not choice. The whole idea is to use damage to infer intensity (and therefore wind speed estimates). If our goal is to produce as accurate a record as possible, how does eschewing accurate data for the sake of "consistency" fit in with that goal? I understand the concerns about the climatology being skewed, but all we're doing is taking an already questionable record and improving for some regions. And with the large variation in the density of structures between states, and even different areas within states, it's not as if we're producing a totally consistent and unbiased record using only damage anyway.

[soonertrey]

I think the thing to keep in mind regarding the EF scale being a "damage scale" is that it's a damage scale out of necessity, not choice. The whole idea is to use damage to infer intensity (and therefore wind speed estimates). If our goal is to produce as accurate a record as possible, how does eschewing accurate data for the sake of "consistency" fit in with that goal? I understand the concerns about the climatology being skewed, but all we're doing is taking an already questionable record and improving for some regions. And with the large variation in the density of structures between states, and even different areas within states, it's not as if we're producing a totally consistent and unbiased record using only damage anyway.

 

Very well stated and great points.

[Ed Lizard]

Using damage scale alone leads to what is probably a misleading statistic, that about 1% of the tornadoes causes about 70% of the fatalities.  If a tornado on the open prairie that crosses no roads has a very high bar to meet to be anything higher than an EF-3, than by definition a tornado has to hit well built homes or buildings, ie, population centers, to be rated as one of the 1%.

 

I think someone also mentioned population centers (and I'd suppose forested areas) provide more debris, which will increase the damage.  A 130 mph gust and a 130 mph gust carrying a truck trailer will probably cause different damage to similarly built structures.

 

Similar to the complaining that the satellite era has enhanced storm numbers, and 1933 may have had more actual tropical cyclones than 2005.  If better science distorts climatology, than NHC should ignore European satellite imagery and name nothing East of 50ºW unless it hits a ship.

[f2tornado]

There is a simple remedy for this. Just bust out the former Roger Maris asterisk. EF5*. An extra column could be added to Storm Data for these radar based ratings. The old climatology is preserved while the speed is noted.

[Gino]

It won't be, not with the wind speed announced publicly. The folks at SPC can think what they want, but in the end, OUN publishes the Storm Data entry on it.

 

True, SPC isn't involved directly in the assigning of ratings, however NWS HQ can dictate to the office what it can be rated and Dr U sent out an email to RDs a couple days ago reminding them that directives dictate that the EF-scale ratings be based on damage not radar data. He also re-iterated several times that the EF-5 is preliminary, hinting that the final rating may be based on damage (EF-3) with a note in Storm Data that higher winds were measured. 

[tornadotony]

True, SPC isn't involved directly in the assigning of ratings, however NWS HQ can dictate to the office what it can be rated and Dr U sent out an email to RDs a couple days ago reminding them that directives dictate that the EF-scale ratings be based on damage not radar data. He also re-iterated several times that the EF-5 is preliminary, hinting that the final rating may be based on damage (EF-3) with a note in Storm Data that higher winds were measured.

Well that's a shame, but it is what it is. If there's going to be another La Plata incident, so be it. My concern is the science. I have an extensive philosophical issue rating a tornado EF3 with Doppler-measured winds over 200 MPH if we've been trying to claim that the EF-scale was elicited to be a more accurate representation of the wind-damage relationship in tornadoes, especially with the literature available on tornado wind speed relationships. Now I will admit that the relationship between wind speed and tornado damage is tenuous at best, even in literature. I can easily see the day some years in the future where we find ourselves changing the EF-scale again. If that were the case, I would obviously support changing ratings on tornadoes rated using Doppler wind measurements.

But until that happens, I'm not comfortable, as a scientist, ignoring the data we have simply for the sake of consistency. The mere notion of consistency in the national tornado climatology is a bit of a fallicy. Ratings are not consistent from office to office, and several high-profile tornadoes have already been rated by unconventional, perhaps questionable, means. I don't disagree that this practice would propagate another inconsistency through the climatology, but it's a wash given that it would start filling the known gaps in the climatology, especially on the Plains. To not use it because it can't be applied to all tornadoes would be to say that we should just sit pat with known bad data because we can't fix all the bad data, which again doesn't sit well with me.

[Ensō]

If the issue is the whole idea of rating based on radar, are they going to also reevaluate all of the other radar-based ratings?

[Hoosier]

Thanks for the responses all. It will be interesting to see if the EF5 rating is maintained.

I was thinking, what if max damage was EF3 but measured wind speeds were somewhere around 200 mph for this tornado...then one could've made a case for an EF3, EF4 or EF5 rating.

[Ensō]

Right. I think if the observed velocities were, say, 210 +/- 20 mph or something then it'd make for a tougher decision. Obviously in this case that definitely isn't an issue, but it's something that would need to be worked out. I'd assume in that case they'd default to the lowest rating they were certain of but I don't know.

[tornadotony]

Thanks for the responses all. It will be interesting to see if the EF5 rating is maintained.I was thinking, what if max damage was EF3 but measured wind speeds were somewhere around 200 mph for this tornado...then one could've made a case for an EF3, EF4 or EF5 rating.

I think that would make revisiting the EF-scale wind estimates a more immediately pressing issue.

[AtticaFanatica]

A lot of behind the scenes wrangling going on the last few days that reach far beyond OUN. Would not be surprised if this is downgraded.