TORNADO DAMAGE SURVEY: COLLIN/GRAYSON COUNTIES TEXAS

[Glen Romine]

Doswell may have found it... Too damn cheap to fly in the survey team.
[/b]

Pride could have as much to do with it as money. It would be embarrassing to call in the QRT and have them berate you for wasting there time with damage below the criteria for F4/5.

Likely? Likely much higher. Let me see. Hurricane Charley had 145 mph sustained winds and the homes in Punta Gorda didn't get flattened like those in TX.
[/b]

The construction practices in hurricane prone FL are hardly comparable to those of rural north Texas. Hurricane winds are also largely horizontal (there are believed to be considerable vertical velocities near the ground in tornadoes) - strongest in a very small swath, and almost never actually measured but inferred based on merging the observations with models, and based on an assumed surface of WATER and not the much rougher land surface that considerably slows the winds. So - it is non-trivial to compare wind damage from a hurricane with those from a tornado to infer wind speeds for a number of factors.

I liked Chuck's essay as well - frank as usual, and agree also that the EF scale really blows (pun intended) in terms of getting meaningful information from the tornado database, despite the well intentions of those who are promoting it.

Glen

 

[Jeff Snyder]

It's obvious that we are trying to move to the idea of classifying tornado based on wind speed. Well, perhaps I should say that using more damage indicators and more degrees of damage helps to provide many more categories of damage to which wind speeds are assigned. Fujita's original scale made no mention of gas station canopies, or the majority of other structures, natural or artificial, that are much more widespread than 'well-built homes'. It would seem that, for objectivity, to more damage indicators and degrees of damage, the better. Of course, there's a point at which we can't confidently say that this-or-that damage was from 5mph higher wind speeds than that-or-this damage. In regards to more detailed damage indicators, and, thus, a refinement of wind speeds available for assigning to tornado segments -- The new EF scale tries to do this (i.e. radar data may be used to support a given EF rating), and F-scale ratings the past few years seem to have done this as well. Should we keep our blinders on and continue to shunt recent engineering work by rating tornadoes based on the brief description given by Fujita (part of which Alex quoted above)? Isn't science about advancement? Isn't it about using new research in the search for truth (in this case, the actual intensity of the tornado)? Are you personally offended when EXPERTS, people like Woodall and Marshall, don't agree with your personal observation? Sure, they can actually look at how the house was built, how the walls were attached to the foundation, but surely we, from our computer hundreds of miles away, can do just as well by looking at a few relatively low-resolution pictures taken by media and other folks. Heck, what's the point of damage assessment if we can just survey everything by online pictures?

As Gabe noted, 15 years ago that tornado probably would have been given F4 (or even F5 with a cursory look at the pics). Perhaps it makes people feel good to know that there was an F4 or F5... I know people tend to want to think they lived through something as bad as it can be (there has been phsychological research into this in regards to hurricanes -- people tend to want to think they lived through something that's as bad as mother nature can dish out... I mean, how many people really experience Cat 5 winds in Andrew?) But, we're moving on, trying to use some engineering realities to remove one more complication from damage rating assessments. We have enough problems with the fact that we can only really judge tornado intensity when structures get hit (even this is being diminished in the EF-scale, with the use of different damage indicators like hardwood and softwood trees). In the past, we've also had a problem with the fact that the EXACT same tornado could be rated entirely different depending upon what type of structure was hit (and, more importantly, how well the hit structure was built). I'm not talking about if the tornado hits an outhouse or a big house, I'm talking about if it hits a house in one development versus another development. Of course most houses today are going to look "well-built", but that's obviously note the case. If we are seeking to rate a tornado as objectively as possible, trying to assess a rating based on the tornado REAL strength, isn't it in the best interest to remove as many complications as possible?

EDIT: Re-reading this, it comes across a bit harsh in areas. No offense intended, obviously!

 

[Gabe Garfield]

Pride could have as much to do with it as money. It would be embarrassing to call in the QRT and have them berate you for wasting there time with damage below the criteria for F4/5.

Glen
[/b]

I think you hit the nail on the head, Glen. After the La Plata, MD snafu, I think a lot of have been a bit more reluctant to give the violent tornado rating. During that case, the press reported that the event was a rare F5 tornado, only to discover that the "F5" tornado damage was only indicative of strong F2. Talk about embarrassing! Was this the case for the survey of Tuesday's tornado in north Texas? I don't think so. But, I imagine there have been other damage surveys where this played at least a minor role in determining the tornado rating (F3 vs. F4, etc.).

Gabe

 

[Glen Romine]

Isn't science about advancement? Isn't it about using new research in the search for truth (in this case, the actual intensity of the tornado)?
[/b]

I don't think you'll find much disagreement with that - but there is good reason not to flush the current tornado database down the toilet in the name of advancing science when it isn't entirely needed. Why not adjust the wind speeds down for the old F scale categories, instead of adjusting the damage indicators toward lower EF categories? Why not have a clean swept, well engineered home still be classified as F5, and simply open the door for F6 were something more subtantially engineered damaged as well? Seems there is a desire to keep the 6 scale levels (0-5) but limit the rating that can be assigned based on the traditional damage indicators - and this in essence means the entire tornado database pre-dating these new assessment techniques are now all suspect. Sure - they were already suspect before - but at least there was some consistency in how the database was formed. Currently a rather random assortment of assessment techniques are in play at various offices - some using the new indicators, some the old, some reluctant to call in the QRTs, all leading to a database that is probably even less useful than what we had before. Sure - in 30 years, this new rating method could result in more relaible estimates of the frequency of extreme wind events, but why does the historical database need to be tossed out to get there? It doesn't imo if we simply adjust the winds down to match the old F-scale system damage indicators and open up the top end.

Glen

 

[Jeff Snyder]

Glen,

That would be fine, except the "old" F-scale still says that a well-built home that is wiped clean is F5. Period. However, it's been shown that a home can be wiped clean (thus be F5 damage) with a tornado that has winds of "only" 150mph. Or 300mph. So now what? If an F5 tornado could be either 150mph or 300mph, that tells us little about the intensity of the tornado. Of course, there is ambiguity in assessing exactly what constitutes a "well-built" house, and this has been more apparent in the past 5 years (which partially explains why we haven't seen an F5 since 1999 IMO).

Maybe we could shift the old F-scale winds down substantially. However, that still leaves relatively few damage indicators. How do NWSFOs know what tree damage, or gas station canopy damage, was likely produced by what intensity of winds? The F-scale description says very little about this. Without these other damage indicators outlined, I'd think that there would be a lot of ambiguity. So, let's add more damage indicators, and consult 'experts' as to what range of wind speeds would be most likely to produce certain levels of damage to certain objects. This way, an NWS employee can take a sheet, go out to a damage area, and begin to assess non-standard damage (non-standard being non-'well-built' homes, trees, large warehouses, etc). Because the intensity and type of damage is well outlined (compared to the previous, F-scale), objectivity should be easier to attain. Well, we've basically just arrived at the EF-scale. An outline of new types of damage, with distinct levels of damage, assigned to probable wind speeds.

The tornado climatology database is very important to sustain, as you note. Noting trends in tornado intensities is difficult if the scale on which the tornadoes are rated is changed. We saw this in the 1970s, when a bunch of grad students and others used newspaper clippings and other evidence to rate tornadoes from the 50s and 60s. Looking at trends, you can see curious changes that occurred in number and apparent intensity of tornadoes after the 1970s. Of course, it's unlikely that the actual numbers and intensities have that changed much in the past 40 years... So yes, climo does have importance. However, do we stay course with a scale that isn't entirely descriptive? Or do we transition to a scale which leaves less room for ambiguity in regards to the actual intensity of a particular tornado? What's the best course of action in the long-term? I maintain that the more detailed assessment scale will produce the most realistic determination of tornado intensities, and thus is in the best interest of the scientific community in the long term. Yes, again, the climo database will likely see a discontinuity next year when the EF-scale goes into practice (actually, it's only a discontinuity if EF-rated tornadoes are compared in number to F-rated tornadoes), but I have to think that's all in the best interest of the community when the goal is to accurately and consistently rate tornadoes as close to their actual intensities as is currently possible.

Do we continue with an imperfect scale in the name of climatology, or do we refine the scale to more accurately reflect reality, but reduce the validity of the climatology? I vote for the later. IMO, scientific accuracy in the form of an enhanced, more detailed tornado intensity rating system is more important then maintaining a flawed system in the name of climo. The EF scale is not perfect, but I think it's benefits outweight it's negative consequences.

 

[Justin Turcotte]

I hear your point Jeff. If a tornado that levels a city is only going to be F3 or EF3 than fine, it is just a number. There may be some scientific value there but the number is meaningless when comparing the new tornado climate record from the one from the historic one. This is my primary concern and Doswell's rant takes the same view. What real value is being added to the climatology with the increased scrutiny of damaged homes and a new rating scale anyway? F4 and F5 tornadoes were already rare and now they are nearly extinct. We are losing the bench mark of reasonable sample size that sets some tornadoes apart from others. Is this what we really want?

Some global warming advocates have proposed that tornadoes are becoming more widespread and damaging without factoring in that we have much better detection rate than we did just 20 years ago and we have more tornado fodder out there. Remember Al Gore's rant after May 3, 1999? I've also read some of the the extreme opposite twisting of data where the non believers in global warming have charts of decreasing numbers of F2+ tornadoes which is partially true, but, they fail to point out that 1970s F2 tornadoes were generally overrated. Now we will have decreasing numbers of F4 tornadoes and no F5 tornadoes. Hey, tornadoes are getting weaker! Go tell Congress, rest assured, that we can keep pumping CO2 in the atmosphere becuase storms are getting weaker.

We are making an already imperfect climatology even more imperfect. I'll try and leave this topic alone now. It gets me fired up as my thesis deals with the F-scale.

 

[Jeff Snyder]

Alex,

It is of less scientific relevance to base every tornado on the level of damage without any regard to whether a set of wind speeds were really required to produce that damage. If that's the case, then just say that any structure that is wiped clean is F5. A small barn wiped clean? F5. A portapotty wiped clean? F5. If F5 represents that whatever was standing is no longer present, than use that. You'll notice, however, that the scale would then serve almost no scientific value at all.

We currently imply that the F-scale should be used to rate the strength of the tornado. Yes, we know that there have been violent tornadoes in fields that don't hit anything. In the past, per the F-scale, these tornadoes technically should not have received anything but F0, since no damage was produced. The new EF-scale at least provides guidance as to damage to trees, signs, etc. You are correct, we will never really know what the EXACT wind speeds in a particular tornado are. However, isn't it better to try and explain that certain damage to hardwood trees would be most likely to have been produced by winds speeds between xxx and yyy? Or that, unlike what we've thought in the past, 300 mph winds are NOT needed to completely wipe a house clean. In the past, the La Plata tornado would have been F5 (and it was preliminarily rated that). However, we know that wind speeds in that tornado probably were NOT >260mph (F5-level), at least as evidenced by the damage produced. So, you are saying that we should have still rated that tornado as F5? What scientific value does that serve? It tells us nothing about the actual intensity of the tornado.

We know the old wind speeds in the F-scale were too high in some cases. Should we not use this information to better the intensity rating scale used for tornadoes? Should we continue with what we have, ignoring research and studies that have shown some wind speeds are grossly overestimated? Or should we just throw out all wind speed ratings attached to the damage scale, thereby making the scale less scientifically relevant in regards to assessing the intensity of tornadoes?

LOL I'll stop posting here, since I don't want to hog the thread.

 

[mikegeukes]

In the past, so many tornadoes were rated by looking at photographs
of damages, the Super Outbreak if you ever read Storm Data, had
no F-Scale Ratings assigned to them. Remember the F-Scale was
devised in 1971. Around 1980 is when the NWS started to verify warnings,
some offices did a better job than others in surveying tornado damage.
Hopefully the new EF scale, will be more accurate than the current F-Scale.

1950-1959 13 F-5 Tornadoes
1960-1969 11 F-5 Tornadoes
1970-1979 14 F-5 Tornadoes
1980-1989 03 F-5 Tornadoes
1990-1999 10 F-5 Tornadoes
2000-2005 00 F-5 Tornadoes

1950-1979 38 F-5 Tornadoes
1980-2005 13 F-5 Tornadoes

Mike

 

[Greg Stumpf]

As I stated before on the other EF thread:

I'm part of the team helping develop NWS training (via WDTB) for the new EF scale. In a recent meeting with the lead scientists who helped develop the EF scale, we discussed the possibility of adding new damage indicators (DI), and higher degrees of damage (DOD) for some of the existing DIs. This includes the possibility of adding a DOD for a well-built well-anchored single/double family residence being swept off a foundation (with the usual caveats - no "sliders", duration of wind, debris from other structures, etc). The discussion is in light of some of the concerns being discussed in this forum.

 

[Tony Lyza]

I wonder if a part of this may be the influence of insurance companies on some members of the QRT. Someone here mentioned that Tim Marshall of Haag Engineering assisted in the survey of the Westminster tornado. Well, I heard a story on CNN awhile back, and I found its transcript:

http://edition.cnn.com/TRANSCRIPTS/0604/17/lt.01.html

Haag Engineering works for State Farm Insurance in disaster assessment. I wonder if lower F-scale ratings are used by insurance companies as an 'out' on covering what they are obligated to cover, i.e. the companies would refuse to fully pay for completely leveling by and F3 because a lower rating on a house completely flattened would indicate poor construction. Here is another article on Haag and the OKC tornado:

http://www.normantranscript.com/localnews/..._106014529.html

This sort of situation also came up in Hurricane Katrina. An article on that situation follows:

http://news.yahoo.com/s/ap/katrina_insurance

It's also interesting, IMO, how Katrina was later downgraded to cat 3, with a pressure 10 mb lower than any other cat 3 on record. It just makes me wonder if the ratings for both tornadoes and hurricanes are used by the insurance companies when determining amount of payment, and, if so, they are contaminating, via the QRT/NWS, the rating of tornadoes today.

 

[rdale]

"I wonder if lower F-scale ratings are used by insurance companies as an 'out' on covering what they are obligated to cover, i.e. the companies would refuse to fully pay for completely leveling by and F3 because a lower rating on a house completely flattened would indicate poor construction."

Where did you even hear a rumor of such a policy? That's clearly not in my paperwork ;>

 

[Tony Lyza]

"I wonder if lower F-scale ratings are used by insurance companies as an 'out' on covering what they are obligated to cover, i.e. the companies would refuse to fully pay for completely leveling by and F3 because a lower rating on a house completely flattened would indicate poor construction."

Where did you even hear a rumor of such a policy? That's clearly not in my paperwork ;>
[/b]

Insurance companies go to great lengths to shirk their responsibilities.

 

[rdale]

I understand, but my policy says nothing about the F-scale being utilized in payouts. Have you actually heard they did it, or are you making a subtle accusation against a well-known tornado researcher with no backing?

 

[Donald Van Dyke]

I pretty much agree with Jeff Snyder's opinon on the EF scale. On the issue of Doswell's essay, I can understand why he is upset with a few things though. In particular, I do agree with the part where he cites an example of how an old F-5 does not equate to a new EF-5. But, as Greg Stumpf said earlier in the thread, that is being looked at currently.

On this issue about the QRT teams, I just cannot believe that there is any conspiracy or even reluctance to call the QRT team. Although Doswell's comments are interesting and do provide some level of evidence to the contrary since he is actually a QRT member, I would also consider these remarks posted on another forum from an NWS Memphis forecaster. http://www.easternuswx.com/bb/index.php?s=...dpost&p=1174620

Personally, I tend to believe that this NWS forecaster's account of events is far more representative of what actually goes on during these surveys of more significant tornadoes than Doswell's implication of some sort of reluctance to use the QRT, or the suggestion by some others that there is some "conspiracy" going on. I just don't understand this "conspiracy" talk at all. It boggles my mind (to put it nicely). Doesn't the NWS Memphis forecaster's account of events seem much more reasonable as to what is likely really going on with regards to most of these events?

 

[Greg Stumpf]

I am a national QRT member as well - if there is a conspiracy, they haven't shared that with me or any of the other QRT members that I have frequent contact with. With regard to Doswell's comments on why the national QRT resource hasn't been utilized as much as we had envisioned, I can agree with most of his comments (but not all). Part of the reason is related to the fact that there just was a change in leadership at NWS (a new National WCM), and that new person is being brought up to speed on the entire process. But as that forecaster from the Memphis WFO states as well, there is no conspiracy! Needless to say, some of the QRT members are having an offline dialog right now discussing all these concerns.

What Tim Marshall does with Haag is try to determine what percentage of insured structure damage is due to faulty construction versus actual wind. If he finds structures that were built below code specifications (and thus, breaking the law) or determines that the damage was not storm-related, then the insurance companies or owners can sue the builders to help compensate for the losses. Tim Marshall is one of the most revered storm damage survey experts, and it would be ridiculous to insinuate that he is part of some conspiracy to favor some secret faction of government that is in cahoots with the insurance companies to prevent F4+ tornadoes. Tim has been a damage survey mentor to many, including myself.