I personally don't care how easy or hard EF5 is to achieve as long as we're getting the wind speeds right. I think the real reason it seems like bare subfloors would make EF5s too easy is because we've been seeing a dramatic uptrend in violent tornadoes.
We used to average 1-2 F5s per year even when subfloors were being rated as such. If we surveyed this year like we did in the 90s there would've easily been 6 F5s this year. Which is pretty wild when you consider what it means.
A little late on my end as far as this response goes, but I wanted to add that I do still believe that achieving an EF5 rating should be very rare and reserved for the highest end events - something that they really, really need to change in the newer EF scale is the windspeed estimates themselves at each rating, as they are straight-up inaccurate a lot of the time. But I think the way EF5s were rated from 2007-2013 was honestly
okay - not perfect by any means (Goldsby and Chickasha, but I have a reason why these could have potentially been underrated outside of strict surveying as well) but definitely better than what occurred before or after that era of surveying. If anyone gave Elkhorn an EF5 rating because it slabbed a subfloor slider, I simply will never support that notion. Elkhorn was not, in any way, comparable to a typical EF5 tornado in terms of its damage, or even a high end EF4 for that matter. Low end EF4 is the correct call to me, and if it is confirmed that we found that winds exceeding 200 mph are required to inflict such damage, then the bar for EF5s in terms of windspeed should be raised to accommodate this, and windspeed estimates for many tornadoes would need to be revised.
Were bare subfloors being rated F5 a lot in the 90s? If that is the case that is very interesting, because I was under the impression that the 90s tended to be a more violent decade than the others, characterized by many high end events occurring during that era specifically (BCM, Loyal Valley, Jarrell, Andover, Red Rock, Oakfield, Pampa, etc). I believed that in the past, tornado ratings tended to be more liberal in general, especially the further back you go - but as people have shown me on this forum, that isn't entirely true because you can take a look at the 1965 Palm Sunday outbreak and see that they were
very conservative for the ratings in that outbreak. There were at least 3 tornadoes deserving of an F5 there, and many of the other F4s probably did reach F5 intensity. Admittedly, though, I cannot point to another outbreak where they definitely underrated the storms in the past, so it's very possible that it was a one-time thing, in a sense.
If anything, I feel like the amount of violent tornadoes per year has remained pretty constant outside of outlier years like 2011. It is so difficult to tell because of the extreme inconsistency in surveying, though, and to me that is a
major, major issue in the way we are tracking tornadoes and their power. EF ratings are simply not accurate to determine how strong tornadoes are much of the time, as unfortunate as that is. Think of all the violent tornadoes we have missed simply because they didn't hit anything. The first thing that comes to my mind is the Nebraska Sand Hills outbreak and the Plevna supercell from this year -
without a single doubt in my mind, many violent tornadoes from these two events alone were missed. Not that I think that the tornadoes here deserved higher ratings based on the confines of the scale, I'm just pointing out that it shows how much more that there is to learn and apply when it comes to this science.
EDIT: As for Chickasha/Goldsby, I do genuinely believe that they snubbed EF5 ratings for these two tornadoes partially because of a desire to nitpick based on the fact that there already was many rated EF5s that year. But that's baseless conjecture, and I'm only saying that because it seems like something that would make sense based on human psychology. This is just what I believe. If those two tornadoes occurred in a more typical tornado year, I think EF5 ratings would have been granted.
