Enhanced Fujita scale confusion

[Tim Shane]

Hey guys,

It's my first post here. I'd just briefly like some input from the experts with respect to a question I have about the EF scale. I'm not a professional met, but it seems looking at some damage photos from EF-3 rated tornadoes, the damage *looks* like EF-4 or EF-5 to my untrained eye. Here are some examples

From the recent Jackson, MS tornado:
http://www.srh.noaa.gov/images/jan/Weather_Events/2011/0415severe/hinds/100_1165.JPG

From the 2005 Stoughton, WI tornado
http://www.solarnavigator.net/geogr..._Stoughton_roof_brick_building_demolished.jpg


Now, obviously, if trained professionals agree that these are not EF-5, I will defer to their expertise. But why not? To my eye, these houses both appear to be solid brick. The first one is wiped clean of its foundation, the second one is completely flattened (EF-5 and EF-4, respectively).

 

[Andy Wehrle]

Agreed the damage in the first link *looks* higher than EF3.

As for Stoughton, which actually took place in 2005 and was rated under the original Fujita scale, most of the "brick" houses were actually brick veneer (the bricks did not provide any structural support). It was still a very intense tornado, especially by Wisconsin standards, and an F4 rating was considered.

I would be surprised if we DIDN'T see at least one EF4 when all the surveys are done and reviewed for the April 15-16 event.

Somewhere I heard that local WFO survey teams cannot independently assign a violent category rating, and must wait for the damage to be surveyed and the rating validated by a NOAA "Quick Response Team."

 

[Jeff Snyder]

Somewhere I heard that local WFO survey teams cannot independently assign a violent category rating, and must wait for the damage to be surveyed and the rating validated by a NOAA "Quick Response Team."

That *used* to be the case. However, it is my understanding, from different sources, that the National QRT no longer exists. It used to consist of selected damage experts, with the goal of supporting damage assessments for high-end tornadoes. However, there may still be separate QRTs organized by each NWS region, and they may or may not be called for any particular event. Why this resource (the national QRT) went away, I do not know. Now, it may be a budget issue, but how much does it really cost to bring in a couple of experts (e.g. Tim Marshall, NSSL/WDTB damage scientists, etc.) for a few events every year? It's neither here nor there now, regardless.

 

[Andy Wehrle]

I would be surprised if we DIDN'T see at least one EF4 when all the surveys are done and reviewed for the April 15-16 event.

So much for that, as 20 minutes before I posted the Mobile WFO confirmed that the "high-end EF3" will remain just that:

PUBLIC INFORMATION STATEMENT
NATIONAL WEATHER SERVICE MOBILE AL
805 PM CDT TUE APR 19 2011

...ADDITIONAL SURVEYS COMPLETED WITH TORNADO TRACKS IDENTIFIED IN
STATE LINE...YARBO...AND CITRONELLE...

...LEAKESVILLE MS TO DEER PARK AL TORNADO CONFIRMED AS A HIGH END
EF-3 INTENSITY TORNADO WITH WINDS OF 160 MPH...

I am surprised, given some of the radar signatures passing through populated areas I would have expected multiple EF4s much like last June's Minnesota outbreak. Just shows you what do I know.

 

[Tim Shane]

Yet another example. This picture is from the 2008 Super Tuesday outbreak. As you can see, the house in the foreground has been completely swept clean, a classic indicator of EF5 damage. The house in the background also appears to be swept clean, with just a few short columns of brick standing:

http://en.wikipedia.org/wiki/File:Lawrence_County,_AL_tornado_damage.jpg

Yet this tornado was only rated EF4. I wish the individual NWS offices would have a more detailed explanation of their rating process, especially high-end events.

 

[mwenzbauer]

I've been wondering the same thing. I thought the pictures from the Askewville/Colerain tornado in NC looked at least EF4, but then again I don't know the quality of the construction there. Several pictures seemed to show just slabs, debris thrown pretty fair distances and large trees torn apart. Just seems odd with how many EF4's we had last year, it seems this year can't seem to get one! Then again, I know tornadoes like Enterprise, AL 2007 were initially rated EF3 and later bumped to EF4, I guess time will tell.

 

[Jason Foster]

Construction variables account for A LOT of the differences. While many photos may appear to show similar damage, a really good understanding of construction methods and materials is vital. One of the biggest, single most important words I try to relay when people ask me the difference: REINFORCED. Brick walls, CMU (aka cinderblock) walls are not by themselves strong. While maybe a light bit better than stick frame & vinyl siding, it is not going to survive a tornado (even a weak one) unless it is reinforced. 90% of homes are not currently built (brick or not) to withstand the forces. But things like cast concrete or reinforced brick/cmu walls are more than capable.

EF scaling has gotten better in regards to assessing winds to damage. However, with the above noted, you may find at times that structures seem like they were destroyed by higher winds when they were not. ALSO, keep in mind quantity of debris. In places like NC, there was a much higher density of homes, thereby creating a higher amount of building and vegetation debris. This in turn probably caused significantly more damage than those sometimes seen in rural areas.

In the last photo exampled (from wiki) the GLARING problem I see there is the lack of good connection between the bottom plate and the rest of the exterior walls. It is a similar case to what I personally saw in Snow Hill (see 4/16/2011 reports thread) where homes were taken off their foundation as a whole and moved 50 to 100 feet away, and remained intact to some level. That was a brand new home too...as it isn't even on the google aerial view (satellite in google). In general, that neighborhood has significant lack of quality construction. I don't know if they were in violation of codes, or the codes are just piss poor there, but it certainly meant that higher EF ratings seemed somewhat appropriate, but has been surveyed to be EF3.

 

[Stan Rose]

What Jason said.
Complete destruction of a residential building is completely consistent with EF-4 damage. Ive seen plenty of houses with nothing but foundations left that were the result of tornadoes rated F3 under the old scale. Absent any indication of reinforced construction, you'd probably have to see damage to a larger/sturdier structure (like a motel or office building) to get an EF-5 rating.

 

[J Allen]

But thats assuming that the Enhanced Fujita scale is actually reflective of the damage done by a tornado, perhaps more of a case that it was produced based on not field measurements by people who understand tornadic behaviour, but engineers who believe that scale wind tunnels actually correspond to near realistic. Theres an excellent paper on the problems and unrealistic nature of this scale by Brooks and Doswell. Too much politics has also crept into ratings, now funding the repair bill depends on the rating and those doing the surveys (or rather those having the final says) are reluctant to rate tornadoes above EF3 as they are being attributed some special symbol rather than reflecting damage as they are meant to.

 

[Mike Bennett]

I had the privilege of taking the NWS on a damage tour the day after the Carrier,OK tornado in April 1999. I was connected with the county Emergency Management at the time. It was really interesting and I learned from them on how they assessed damage in determining the then "F" scale. This one experience doesn't by any means intend to imply I know anything about it, however. But here's some things they told me they look at.

The best example comes from a brick framed house that took a direct hit a couple of miles west of Carrier. The only thing left standing were most of the inside walls and the south outside wall. The roof, and outside walls were gone. At first impression one would think an upper end F3 or lower end F4. However, the west side of the house had a covered patio. The patio cover was tied into the main roof and was part of the original structure. The damage assessors told me when the tornadic winds got under the patio cover (the weakest part of the structure), it had no where to go and could easily lift the patio cover taking the main roof with it, pealing it open like a can. They told me a weaker than F3 tornado could do that given the structural design, thus when the main roof is gone the rest of the structure is greatly weakened of course. There were other examples of things they at, some of them I can't remember.

I guess the main thing is, when the experts determine EF scale damage, there is a lot that goes into it. Obviously, they don't determine it at first glance, but look at weak points of structures, structure quality etc....

 

[A Luttig]

Is it not possible that those pictures were isolated cases of EF-4 damage, yet the rest of the destruction caused by the tornado was more consistent with EF-3 damage? It was my understanding ratings weren't given based on maximum damage caused. These photos may be the anomalies, not the norms. Kind of hard to base a tornado's strength on one picture.

 

[mwenzbauer]

I believe the rating is based on maximum damage (someone can correct me if I'm wrong). I thought I read where Greensburg's EF5 rating came from just 2 or 3 areas, and most of the damage was EF3-EF4. It's interesting to note in the last decade (2000-2009) how many fewer F4/F5 rated tornadoes there have been than previous decades. I have a hard time believing tornadoes are getting weaker, and suspect we're just slowly getting better and more accurate at rating them.

 

[J Allen]

Mwenz,
You are correct on the rating based on maximum damage, but the rating system is unfortunately compromised by politics rather than being solidly related to the damage that occured. The large number of damage indicators used can be both a help and a hinderance, even within tree species there is enormous variation in breaking/uprooting/debarking depending on a whole host of factors including soil moisture etc. outside of the Hardwood/Softwood categorisations used in the EF-scale which can't be covered by engineering lab tests. The other major problem with any rating system is the damage to windspeed relation. Realistically the EF-scale is not related to any real wind speed, its based on scaling assumptions and a small set of tests, rather than being of the actual winds experienced by a structure: the tests behind the scale are generally based on force winds rather than vortice related (suction vortices can have a very different effect to the overall damage) and this can have a marked effect on the behaviour of a house. The final problem you might see with the EF-scale is the disruption of the climatology, realistically comparison between supposed equal categorisations is not possible (for instance, rate Greensburg, not in the same ballpark as Moore or Jarrell). We are left with a scale which doesn't really achieve anything, is far more complex in its application, and in the end achieves nothing as it is a post rating scale of damage sustained. That doesn't really help the people who had their lives devastated by the tornado.

 

[John Wetter]

The thing that most strikes me with the Jackson, MS tornado:
http://www.srh.noaa.gov/images/jan/Weather_Events/2011/0415severe/hinds/100_1165.JPG

is look at the trees around the blocks. The vegetation saw nowhere near EF-4 as has been thrown around here in conversation. If you've ever seen Tim Marshall speak, you'll have seen his slides from La Plata, MD which is the poster child for over-rating a tornado.... This house looks like it was just tipped off of the bock crawlspace. Looking at the trees around it, they have broken limbs, etc. but nothing that looks even close to EF-4. It's much more important to take in the surrounding area as well and look at the construction types used. I don't see anything in this photo that says anything more than an EF-3. I think the political point is a bit over-stated also. I've never felt pressure to see a change in rating. I can't say it has never happened, but I think in the field right now, there is much more pressure to 'get it right' than there is to hit a certain grade, or intensity, etc.

We've become very efficient counters. Is the Enhanced Fujita scale without fault? No. I do believe it goes much further in accuracy and consistency than has ever previously been done though. That and using more technology and understanding of storms along with GIS has helped to produce more accurate information from surveys as well. When you start looking at houses impacted by tornadoes, it is amazing how ill-prepared some of the structures are for having lateral forced applied to them.

 

[Greg Stumpf]

There is an overall reluctance to rate a tornado highly any more because of the scrutiny it attracts, mainly as a result of the additional funding options that are opened when a tornado is of EF4-EF5 intensity.

What additional "funding options"?

 

[rdale]

The tendency is for this to pull ratings down, as once you get multiple ratings you become concerned that your skills will be questioned and try not to attract scrutiny, but also there can be political pressure from local officials to rate a storm in a particular way.

I'm just curious how you came to this conclusion based on your location? I've never sensed "political pressure" to downgrade EF scale numbers in any scale I've taken part in. If anything, citizens get upset because they want the higher numbers. Do you have any specific cases where this happened?

 

[STurner]

I know what you mean. The EF-scale although not as much as the F-scale has subjectivity. The recent Askewville, NC. tornado seemed to arguably have some low-end EF4 damage. I sent Tim Marshall to see what he thinks and clear up what I may or may not be seeing. As far Stoughton goes it arguably could have been rated F4 and even the Westminster twister from 2006 may have more than an argument for an F4 rating. The man who did the 122 mile monster in Arkansas is more conservative than most surveyors as to why it may not have gotten an EF5 rating. The most recent St. Louis damage in subdivisions did not seem any worse than some of the damage I observed in North Carolina. Also construction can be very hard to tell and even I know little about it.

 

[J Allen]

Don't disagree with the ratings in Jackson, haven't looked at the resources available (the biggest problem here is getting structural information) to give a rating but from my limited viewing the rating given doesnt seem unreasonable.

Perhaps I have not been clear in what I am trying to communicate here, the problem is that we have a completely subjective scale system which relies on all buildings conforming to the building codes commonly found within the US, the people given the ratings all having the same experience (picture a person giving a rating to a high end EF5, and then giving a rating to a low end tornado of a similar damage scale the following week or year, based on the intense damage seen in the first case and the open end of the scale he may be more likely to rate what might be EF5 to EF4 based on this as is suggested in the post above). What we have is a subjective opinion based scale(like the Mercalli Intensity Scale for instance), when what we want is an objective scale (ie. a Richter scale). Doswell's (2009) assertion that the lack of consistent damage indicators which are able to provide indication of EF4 and 5 damage is a real issue given the small are affected by tornadoes in the first place. When we add the complication of money involved, people's emotions we have a massive quandry where this scale is subject to a number of failures. To unravel a whole kettle of worms the 2002 La Plata tornado was a fairly good example of the bias associated with subjectivity. And we still have no real analysis of the windspeeds commonly associated with a particular scale of tornado (missing damage indicators for instance), the categorisations in the EF-scale are based on "expert" opinion rather than real measurements.

The discussion does seem to produce somewhat complex arguments to assess whether it is effective or not. If a Quick Response Team type response by more expert NWS rating staff for high end events was applied there would probably be more objectivity (how many raters locally have been exposed to EF5s in all honesty? The application of the scale is in as much experience as knowledge of the indicators, it also assumes sufficient knowledge in structural engineering which is not usually pre-requisite for those who rate). But then there are funding limitations to this application.

Anyway, happy to leave it there, was just trying to raise some of the issues of the scale beyond the simple rating...its far more complex than just damage indicators.

 

[rdale]

I guess I'm confused about a few things... Having worked in Federal funding after tornadoes, I can guarantee you that the EF number has -no- bearing on that process. It's entirely based on the amount of uninsured damages.

And there is no policy that tornadoes are ranked EF3 by default as a max. I just saw news that yesterday's STL tornado was rated 4.

Yes it is still subjective. It always will be. But to say that tornado researchers are deliberately ranking tornadoes lower because of politics is simply not warranted given the evidence you have shown (none). I am certainly open to facts, but so far the only facts are pointing away from your theory.

 

[mwenzbauer]

The man who did the 122 mile monster in Arkansas is more conservative than most surveyors as to why it may not have gotten an EF5 rating.

I assume you're referring to the 2/5/08 Atkins/Clinton F4? I've always thought that one seemed a possible EF5 candidate, along with the 5/4/03 Franklin tornado. I'll be curious to see if Askewville stays EF3 or goes to EF4, that one from what I've been able to find seems to have the most extreme damage from that outbreak.

 

[J Allen]

Going to leave this debate there and take it to PM. Mods feel free to delete if necessary as this has taken the discussion off the intended topic.

 

[Greg Stumpf]

An EF4-5 rating provides an event additional attention from the media and opens a number of federal funding options, some by request of the local representatives through federal application (which tend to use the rating as a battering ram), and other by the declaration of disaster zones at a presidential level (national and international funding options open with this declaration, seems to result mainly either by widespread events, or a big publicity highly damaging event).

This is hogwash! I've been involved in official NOAA damage surveys for over two decades now, and there is absolutely no such threshold for response funding based on the tornado rating scale.

 

[STurner]

Yes the Girard/Franklin tornado from 5/4/03 seemed very similar in appearance to the Bridge Creek/Moore/OKC F5 tornado from 5/3/99 as well the Red Rock, OK tornado from 4/26/91. The Red Rock, OK tornado, The Girard/Franklin, Kansas tornado seemed like possible F5 candidates. If you remember the Harper, KS. tornado from 5-12-04 seemed very defenite. In fact it did some of the most extreme damage over a small area to a brick home and farm equipment. It was probably some of the most impressive damage I have ever seen from not only photos but aerial view as well. The Clinton/Atkins, AR. tornado from 2-5-08 at times did borderline EF4/EF5 damage. The Askewville tornado does seem very debatable as well. It seems hard to believe you can get no EF4 or EF5 tornadoes from an extremely massive tornado outbreak but get an EF4 rating from the St. Louis tornado the other day which had much fewer tornadoes.

 

[Drew.Gardonia]

personally I think the whole Enhanced Fujita scale is complete garbage.

What was once an F-0 is now an EF-1, what was an F-1, is now an EF-2, what was once an F-2, is now an EF-3, what was once an F-3, is now an EF-4, and what was once an F-4 is now an EF-5.

the entire scale change is stupid, and it should have never been tampered with to begin with.

 

[jaredleighton]

Construction variables account for A LOT of the differences. While many photos may appear to show similar damage, a really good understanding of construction methods and materials is vital.

^^^What he said.

The thing that most strikes me with the Jackson, MS tornado:
http://www.srh.noaa.gov/images/jan/Weather_Events/2011/0415severe/hinds/100_1165.JPG

is look at the trees around the blocks. The vegetation saw nowhere near EF-4 as has been thrown around here in conversation.

^^^What he said.

On this last point and addressing rdale, by have a default rating of EF3 assigned when damage greater is suspected (fairly sure this is policy) there is an inbuilt reluctance to assign a higher rating. I clarify, you are correct, citizens want higher numbers for the attention it brings, and the potential for improved funding, while federal bodies generally don't want to fund an event as much as they don't have to (the amount of damage loss means that many tornadic events are excluded). So you have citizen/local pressure to go up, institutional/government pressure to go down. My location is somewhat immaterial when it comes to talking to people who rate and reading the material that is out there. Granted, I haven't given ratings myself in a formal sense and therefore can't offer it in personal experience, but I would be surprised if this pattern was uniformly applied from what i've heard, it may vary from state to state.

^^^Not what he said.

Now whether this desire to under-rate was due to the scrutiny issues or due to policy it had an effect on the rating dearth of EF5 tornadoes in the period 2000-2007 (which recorded the zero F5 enigma after the consistency of the 1990s).

I'm not sure it's fair to dismiss the absence of a rare phenomenon such as an EF-5 tornado impacting a structure over an 8 year stretch as a political stunt (at least that's how I read it).

This is hogwash! I've been involved in official NOAA damage surveys for over two decades now, and there is absolutely no such threshold for response funding based on the tornado rating scale.

^^^What he said.

When it comes to rating tornadoes, I have never...ever heard of any political pressure from "above" to rate a given tornado a certain way. In fact most surveys are done by meteorological scientists, the same scientists who conduct a large percentage of the research papers in existence. The main goal of any survey team is to get the rating correct. This is important not only for historical purposes but for research purposes as well, and the last thing any of those meteorologists want is bad data mucking up the database.

If you're interested in how the EF Scale actually works and how to accurately use it you can look at the EF-Scale training modules by the WDTB.