Severe potential 2/23-2/25

[buckeye05]

Keysburg, TN area from the Feb 24 event. I'd like a healthy dose of reason as to why this was rated 135 MPH EF2, especially given the context. Look how far the vehicles were thrown!

 

[SilentShadow87]

Yeah I'm relatively conservative with tornado ratings but the NWS in the last four years or so has gone from "reasonable conservativism" to "looking for any excuse to give the tornado the lowest rating possible". That looks like pretty solid EF3 damage.

 

[CheeselandSkies]

Yeah I'm relatively conservative with tornado ratings but the NWS in the last four years or so has gone from "reasonable conservativism" to "looking for any excuse to give the tornado the lowest rating possible". That looks like pretty solid EF3 damage.

That first picture is textbook E/F3 damage, with only some interior walls standing...taking into account the cars thrown into a field, so EF2?...



That said, there is a window of time for ratings to be revised before being made permanent. Not sure what it is off the top of my head, but I recall El Reno '13 being initially EF3, then EF5, then back to EF3 over the course of a few weeks or months.

 

[Equus]

Honestly, for most events that aren't devastating and widespread, 135mph EF2 seems to be the upper limit. I know we've had years of discussions over why EF5 seems to be an impossible standard, but now it seems that even EF3 is a difficult threshold to reach without leveling half a town and killing half a dozen or more.

Obviously, construction quality is the limiting factor and as someone who helps build and renovate I can vouch for the dreadful construction quality of an alarming percentage of American single family homes, but the throwing of cars that far at the very least could make the case for being underrated. My fear, unfounded as it may be, is that surveyors seem to be defaulting to the lowest possible estimate for the DOD in nearly every instance. That's standard procedure for substandard construction, but it seems there is always a technicality that prevents a higher rating in the last half decade or so, even when construction quality is found to be adequate. And I absolutely believe that even at the best of times there is a slight degree of office-to-office bias in ratings. FFC vs BMX, as a notable local example.

This is solid EF3 damage if the house is well constructed, so either there were serious construction deficiencies and the tossing of vehicles that distance is not seen as valid secondary evidence, or EF2 is too low. But unfortunately, I do not make that call.

 

[rolltide_130]

Honestly, for most events that aren't devastating and widespread, 135mph EF2 seems to be the upper limit. I know we've had years of discussions over why EF5 seems to be an impossible standard, but now it seems that even EF3 is a difficult threshold to reach without leveling half a town and killing half a dozen or more.

Obviously, construction quality is the limiting factor and as someone who helps build and renovate I can vouch for the dreadful construction quality of an alarming percentage of American single family homes, but the throwing of cars that far at the very least could make the case for being underrated. My fear, unfounded as it may be, is that surveyors seem to be defaulting to the lowest possible estimate for the DOD in nearly every instance. That's standard procedure for substandard construction, but it seems there is always a technicality that prevents a higher rating in the last half decade or so, even when construction quality is found to be adequate. And I absolutely believe that even at the best of times there is a slight degree of office-to-office bias in ratings. FFC vs BMX, as a notable local example.

This is solid EF3 damage if the house is well constructed, so either there were serious construction deficiencies and the tossing of vehicles that distance is not seen as valid secondary evidence, or EF2 is too low. But unfortunately, I do not make that call.

It's something I've honestly thought about. I've more or less considered "Hit a major metropolitan area" a requirement for EF-5 the last several years.

However, I think it's strictly a WFO issue and not an office issue. The WFO there rated clear low-mid range EF-1 damage as 120 mph, so I think it was just poor surveying.

 

[warneagle]

Maybe the biggest story from this whole event is the flooding. The Ohio River is experiencing its worst flooding in >20 years.

 

[SilentShadow87]

I remember someone jokingly saying "EF4 is the new EF5" back around 2015 on another forum, but that seems to becoming more and more true all the time. I put together a bit of data comparing the average numbers of F4/EF4 tornadoes in 10-year periods over the past 50 years.

1968-1977 - average 11.4 F4 tornadoes per year
1978-1987 - average 6.6 F4 tornadoes per year
1988-1997 - average 7.8 F4 tornadoes per year
1998-2007 - average 5.2 F4/EF4 tornadoes per year
2008-2017 - average 6.8 EF4 tornadoes per year

On the surface there might seem to be no trend past the first 10 years. The 1998-2007 period is pulled down a fair bit by the "La Plata syndrome" period from 2002 until 2007 while the 2008-2017 period is pulled up by the 2010-2011 "flare" (can't really think of a better word) of strong tornadoes. If we take a closer look at 6-year periods we can see the recent drop-off more clearly.

1989-1994 - average 8.8 F4 tornadoes per year
1995-2000 - average 7.1 F4 tornadoes per year
2001-2006 - average 4.7 F4 tornadoes per year
2006-2011 - average 7.8 F4 tornadoes per year
2012-2017 - average 4.4 F4 tornadoes per year

My fear, unfounded as it may be, is that surveyors seem to be defaulting to the lowest possible estimate for the DOD in nearly every instance. That's standard procedure for substandard construction, but it seems there is always a technicality that prevents a higher rating in the last half decade or so, even when construction quality is found to be adequate.

I've honestly been getting that impression as well. In the past four years alone there have been hordes of officially-rated EF2 tornadoes that could have been rated EF3 pretty comfortably, several official EF3 tornadoes which could have been rated EF4, and at least two EF4 tornadoes* which probably could have been rated EF5.

In my opinion the worst example of this was with the 4/27/14 Vilonia, AR tornado. John Robinson's statements basically boil down to "it should be practically impossible to give an EF5 rating". I've heard it said before that if all NWS offices used Robinson's standards of rating then we could expect to see an officially-rated EF5 once every 15 years or less, and from 2014 on EF4 really does seem to be the absolute limit of the scale.

*These two, in my opinion, being Vilonia and Abilene. There's no question both would have been rated F5 in the 90's and probably have been rated EF5 before 2014. Rochelle was borderline but in my opinion better off with a high-end EF4 rating because of some legit contextual issues.

This is solid EF3 damage if the house is well constructed, so either there were serious construction deficiencies and the tossing of vehicles that distance is not seen as valid secondary evidence, or EF2 is too low. But unfortunately, I do not make that call.

I'm no structural engineer, but I know a thing or two and just from the images I can say with some confidence that that looks like a fairly well-built house. Probably relatively new, and doesn't look like the structure was compromised in any obvious way before the tornado. The only explanation for the lack of an EF3 rating that the NWS seems to give is "overall context" which actually appears very much in favor of an EF3 rating.

 

[rolltide_130]

Confirmation bias with warning decisions plays a role as well I believe... Louisville dropped the ball with their lack of an adequate warning. It's very possible they kept it below "significant" to avoid a bad look if they were slow to act on an EF-3+

 

[CheeselandSkies]

Confirmation bias with warning decisions plays a role as well I believe... Louisville dropped the ball with their lack of an adequate warning. It's very possible they kept it below "significant" to avoid a bad look if they were slow to act on an EF-3+

I wouldn't be so sure but if so, that's a pretty ****ty reason to lowball a rating.

Just to pick nits, Grazulis' definition (the man who literally wrote the book on significant tornadoes), a tornado is "significant" if it is at least (E)F2 OR causes at least one fatality. Therefore, Saturday's KY-TN tornado qualifies on both counts.

Any reason to believe the Quapaw, OK/Baxter Springs, KS tornado of 2014 (Same day as Vilonia, AR EF4 if I recall correctly) was underrated for this reason? In a similar situation, it was rated EF2, was a killer, and did a good chunk of its damage before the warning went out.

There seems to be a lot of inter-office variability for warning decision thresholds as well as ratings. My local office (MKX) is staffed by bored (been over 10 years since this CWA had a daylight/non-wrapped EF3+) severe weather geeks who put out a tornado warning practically every time a red pixel shows up next to a green one for one scan. There are others where I'm left wondering if they're waiting for reports of debris falling out of the sky.