-
While Stormtrack has discontinued its hosting of SpotterNetwork support on the forums, keep in mind that support for SpotterNetwork issues is available by emailing [email protected].
El Reno Oklahoma tornado downgraded to EF3
- Thread starter Kent Heiderich
- Start date
Shawn Schuman
EF3
- Joined
- Mar 3, 2012
- Messages
- 273
Right, which is why I think the concern over biasing the data is invalid. The data is already biased in a number of ways. At least this bias would be skewing the data in favor of enhanced accuracy, if only for a relatively small geographical area.
Jeff Snyder
EF5
David Hoadley said:
David -- We provide data to the NWS if it's useful, but the purpose of collecting high-resolution radar data with a mobile radar (at least with Howie's group, with which I'm most familiar) isn't to rate tornadoes for Storm Data. There are far broader questions that we're trying to answer than "what EF-scale rating should this tornado get?" With the radars I've used in my graduate school tenure, I've focused on examining attenuation correction techniques available to polarimetric radars, polarimetric signatures associated with supercells seen by X-band mobile radars (low-reflectivity ribbons, LoRBs, ZDR columns, KDP columsn, etc.), and other such things. Not only do the data have enough resolution to resolve fine-scale structures of the mesocyclone, RFD, and tornadoes, but we often have much increased temporal resolution that gives us an opportunity to study extremely rapidly-changing structures and processes that are not possible to study with conventional WSR88D data (not just in radial velocity but also in hydrometeor type and distribution using polarimetric data, etc.).
We have very high-resolution data (15 meter range gates, tornado only 3-4 km range, etc.) with new sweeps every 2 seconds during the El Reno tornado. We can track individual subvortices moving 150-180 mph around the large center of circulation. From our location at Banner Rd and I40, I scanned at an elevation angle below 0 degrees to ensure that we got data as much to the ground as we possible could (beam center ranged from ~-50 m to 600+ m from ground level). These things are impossible to study in such extreme detail with any existing fixed-location radar (TDWR, WSR88D, etc.).
The benefits of such data extend well beyond supplemental use by the NWS.
There *are* some very important things that need to be worked out so that radar data (and other observational data!) can be used within the EF scale. The two primary questions that need to be worked out in order to compare fairly radar data and EF-scale winds, in my opinion, are the following: "how does the wind speed at 10 m AGL compare to 50 m or 100 m AGL?", and "how do the nearly-instantaneous winds measured by a radar compare to the 3 second wind gust used by the EF scale?". It's also important to remember that a single radar can only measure the radial component of the wind, so there's two other components (cross-radial/azimuthal and vertical) that aren't being measured at all.
Last edited by a moderator:
Shawn Schuman
EF3
- Joined
- Mar 3, 2012
- Messages
- 273
Which reminds me - looking beyond the EF-scale debate for a moment, I'd imagine you couldn't pick a much better tornado for which to get such abundant, high-quality data. I'm sure there will be a great deal to learn from it.
Slight digression from the topic at hand, but is it clear that the log-wind profile is not applicable in tornadoes?
Let me know if I'm paraphrasing your second question correctly: How do the forces applied by tornadic winds to a DI differ between an instantaneous exposure and a 3-second exposure? As an example, does an infinitesimally short gust of wind at 200 mph do the same degree of damage as a 3-second burst of 144 mph winds (just throwing out reasonable numbers)?
While I'm seeing a number of clever ideas, this thread is making me realize why the U.S. Tax Code is so ridiculously complicated.
KISS - Keep It Simple, Stupid
The Fujita scale is quite simple. It connects the Beaufort and Mach scales. 'nuff said. From everything I know or can find about the Fujita scale, I doubt anywhere near this much consideration was given to the scale by Dr. Fujita. Over the passage of time, we have come to realize that accuracy suffered due to the simplicity, especially given the focus of ratings towards the degrees of damage done to single family frame houses. Hence the expansion into numerous damage indicators and some actual research into the wind speeds required to do certain degrees of damage - the Enhanced Fujita scale.
The above isn't really even what this debate regards. Thus it seems quite simple to me that it's merely a matter of bureaucracy that we're even having this discussion...over ONE SENTENCE in a 14-page document (last sentence of the first full paragraph on the page labeled as 14: http://www.depts.ttu.edu/nwi/Pubs/FScale/EFScale.pdf)! The only thing I see wrong with the science (the EF-scale) is that there is still too much uncertainty in the estimates of wind speeds for each DI and DoD. Again, 'nuff said.
This is a government issue, not a scientific one. As I've already said, any change to the bias or "consistency" (however you choose to quantify it) of tornado ratings will be minimal compared to the inconsistency and inaccuracy that already pervades the database of tornado ratings due to inherent uncertainties and subjectivities in rating tornadoes. In a way, you could draw upon Occam's Razor here, and declare that if you have any data supporting a rating, then why not apply it to the tornado regardless of the source rather than rating based off some arbitrary flowchart of decisions based on the various sources of information used to determine tornado strength? What is so special about damage that absolutely no other source of information on tornado wind speeds can contribute to a "consistent" rating? I see a reply of, "it's the only source that samples winds near the ground." What about DIs #18, 19, 20(?), 24(?), and 25? Most examples of those would be able to sample winds well above 10 m. What if a tornado only hit those DIs? If that is your argument, then such a tornado would have to be assigned EF0, despite any damage supporting a higher rating. That's just senseless and arbitrary (not unlike the downgrade decision, IMO).
Final comment: people are likely more interested in the wind speeds in tornadoes rather than the amount of damage they are capable of (with wind speed information stripped from that information). That was the intent of the Fujita scale. Damage is merely a proxy - a means to an end. If we can obtain the end directly through measurements, then why shouldn't we?
KISS - Keep It Simple, Stupid
The Fujita scale is quite simple. It connects the Beaufort and Mach scales. 'nuff said. From everything I know or can find about the Fujita scale, I doubt anywhere near this much consideration was given to the scale by Dr. Fujita. Over the passage of time, we have come to realize that accuracy suffered due to the simplicity, especially given the focus of ratings towards the degrees of damage done to single family frame houses. Hence the expansion into numerous damage indicators and some actual research into the wind speeds required to do certain degrees of damage - the Enhanced Fujita scale.
The above isn't really even what this debate regards. Thus it seems quite simple to me that it's merely a matter of bureaucracy that we're even having this discussion...over ONE SENTENCE in a 14-page document (last sentence of the first full paragraph on the page labeled as 14: http://www.depts.ttu.edu/nwi/Pubs/FScale/EFScale.pdf)! The only thing I see wrong with the science (the EF-scale) is that there is still too much uncertainty in the estimates of wind speeds for each DI and DoD. Again, 'nuff said.
This is a government issue, not a scientific one. As I've already said, any change to the bias or "consistency" (however you choose to quantify it) of tornado ratings will be minimal compared to the inconsistency and inaccuracy that already pervades the database of tornado ratings due to inherent uncertainties and subjectivities in rating tornadoes. In a way, you could draw upon Occam's Razor here, and declare that if you have any data supporting a rating, then why not apply it to the tornado regardless of the source rather than rating based off some arbitrary flowchart of decisions based on the various sources of information used to determine tornado strength? What is so special about damage that absolutely no other source of information on tornado wind speeds can contribute to a "consistent" rating? I see a reply of, "it's the only source that samples winds near the ground." What about DIs #18, 19, 20(?), 24(?), and 25? Most examples of those would be able to sample winds well above 10 m. What if a tornado only hit those DIs? If that is your argument, then such a tornado would have to be assigned EF0, despite any damage supporting a higher rating. That's just senseless and arbitrary (not unlike the downgrade decision, IMO).
Final comment: people are likely more interested in the wind speeds in tornadoes rather than the amount of damage they are capable of (with wind speed information stripped from that information). That was the intent of the Fujita scale. Damage is merely a proxy - a means to an end. If we can obtain the end directly through measurements, then why shouldn't we?
MClarkson
EF5
Oh, they do. Remember the debate that raged over whether Andrew was a 4 or a 5, and the ongoing conversation of including or not including storm surge in the Saffir-Simpson scale.
Robert Edmonds
EF5
Just wanted to add my thoughts on this...
If we consider boundary layer growth from constant flow over a flat wall, the boundary layer is initially laminar starting near the edge and follows something close to a Blasius like solution. Further downstream the flow eventually becomes turbulent, and empirically it appears that the time *averaged* flow from the wall begins to adhere to something like a log profile. I've always kind of gotten the impression, perhaps I'm wrong, this is where atmospheric scientists got the idea to try a log profile when considering the wind inside a PBL, and hey it seems to work ok. If this is indeed the case, I would initially consider it dubious to apply this to a tornado. Rarely would I consider the flow near a tornado approximating something constant, and usually you *average* over the (vortex like) eddies.
Last edited by a moderator:
Mike Smith
Jeff, I agree this would be an 'academic' issue but for the fact that FEMA -- and others -- urge engineers to use the EF scale when designing structures. By rating El Reno down to EF-3, we are telling architects, engineers, and others (who will not be aware of these discussions) the wind in that tornado was 165 mph or less. The actual winds were, of course, 294. That is a tremendous difference in the amount of wind loading.
I do not advocate building a bridge along I-80 in Winnemucca, NV to EF-5 standards because the cost v. likelihood isn't there. In the Plains it is a different matter By rating El Reno a "3," two things occur:
- We make the one higher measurement (1999 Moore) look like an extreme outlier possibly causing it to be disregarded.
- We make the spatial and temporal frequency of EF-5's less than it is in the real world.
rdale
EF5
Mike - all the FEMA wind maps I see list "250mph" in them. I see no reference where they suggest building to "EF5" or "F5" strength?
http://www.fema.gov/media-library-data/20130726-1508-20490-5853/fema361_chap_3.pdf
http://www.fema.gov/media-library-data/20130726-1508-20490-5853/fema361_chap_3.pdf
Rob H
EF5
The architects and engineers don't care about what tornado was where and what rating it had - they care about what standard they need to build to, and what kind of damage is associated with that damage. This is actually a good reason for not using measurements to drive EF-scale, because if an engineer is instructed to build something to withstand an EF-5, they're not going to have relevant damage to look at with El Reno.
Moore still was an extreme outlier regardless of El Reno being rated EF-5. It doesn't matter if the ratio is 1/55,000 (tornadoes in Storm Data) or 2/55,000, both are really small numbers. Anyone working with the data should be experienced enough to recognize that El Reno and Moore were similar in strength. Many AMS journal articles mention a few biases and statistical anomalies that they had to contend with during their research. I'd *hope* that OKDOT, or whatever they're called, when building a bridge would get in touch with meteorologists aware of these issues when asking what the probability is of an EF-5 striking a specific spot, and not just go to the Wikipedia page for "List of EF-5 tornadoes".
We unfortunately do have an unrepresentative count of EF-5 tornadoes in the real world. Adding mobile radar measurements to the proceed will help improve that (possibly only for OK/KS tornadoes in prime chase season), but the core issue will likely remain for many years.
Shawn Schuman
EF3
- Joined
- Mar 3, 2012
- Messages
- 273
I agree with your larger point(s), but I'm not sure this is a valid concern. The El Reno measurement itself doesn't appear to be in question at all, regardless of the final rating. That information's still out there, even if it isn't being "officially" used. We also have the ~125 m/s measured in the 2011 El Reno tornado, and even before the mobile radar measurements it seems the general consensus was that maximum tornado wind speeds were probably somewhere in the 300 mph range. I don't know whether the Bridge Creek measurement is being used as a design standard anyway, but if it is, it doesn't seem likely to be disregarded because of this. 'Course, this is also a topic I know very little about so who knows.
As far as wind resistance standards go, I think the bigger problem is probably the fact I mentioned earlier -- we don't really know how common violent tornadoes are.
Mike Smith
I'm not talking about FEMA's wind maps, I'm talking about their design advice.
The screen capture is from a presentation about safe rooms. I sat through one of their presentations in ATL in April where they talked about design of critical infrastructure and they keep referring to a top wind of 250 mph (the original top of the EF scale's EF-5 category) which is stated to be "max for all hazards." We know that is not the top wind in tornadoes. The wind loading difference between 250 mph and 300 mph is huge because wind force is logarithmic.
The idea that engineers are scouring the internet looking for meteorologists' conversations about this is not consistent with the way things are done in the real world. As I discussed before, for liability reasons they want to follow "standards" of which FEMA's is one. If they are told 250 is the top speed, why would they design a much more expensive structure for a wind speed that (they think based on FEMA's advice) doesn't exist in the real world?
As I have said before, if this only mattered to meteorologists I wouldn't care very much. But, under-designing critical infrastructure might -- some day -- have serous real-world implications.
rdale
EF5
The 300mph radar obs are 3 second gusts? I thought they were instantaneous?
f.vanalstine
Do the math understanding that the energy of motion equals mass times velocity squared.
Thus all other things being equal, a 200 mph wind is four times as damaging as a 100 mph wind.
Looking at Mike Smith's comments, 250 squared is 62500. 300 squared is 90000. That "downgraded" 300 mph wind is doing 50 percent more damage, not about 20 percent more as the raw number indicate.
Frank VA
Thus all other things being equal, a 200 mph wind is four times as damaging as a 100 mph wind.
Looking at Mike Smith's comments, 250 squared is 62500. 300 squared is 90000. That "downgraded" 300 mph wind is doing 50 percent more damage, not about 20 percent more as the raw number indicate.
Frank VA
