Enhanced Fujita Ratings Debate Thread

[Eoline]

This isn't me saying that I think the rating reasonable or that I agree with their decision to keep it at EF4 180 - but I do understand where they are coming from if their concern lies with such simulations containing too many generous assumptions, which is likely to me.

If that is the case, then the fact that they weren't getting values well exceeding 200 mph means that if there is something wrong with the assumptions they made (and there very likely is, when considering the real world vs. idealized simulations), these numbers cannot be 100% trusted, even with Monte Carlo simulations. They add good credence to the fact that it likely reached EF5 intensity, but it's not as set-in-stone as structural damage calculations would be. Unlike with structural damage indicators, there isn't a standardized way to approach granting a rating based on the simulations. I was hoping Enderlin would get the ball rolling when it comes to this type of thing, but it appears that it has not (at least, not yet).

Also, it is worth mentioning that the numbers they are using are almost explicitly median/mean values and ranges. They almost certainly did get runs below the 200 mph threshold for any one or more of these DIs they tested on, and if the amount of runs pointing to EF4 rather than EF5 intensity is within a single standard deviation, the decision to keep it at EF4 is honestly a good one to me. FWIW, I do still think it reached EF5 intensity, and it should be rated as such. Of course, I'm throwing out a lot of hypotheticals here, and I could be wrong on what the true reasoning is... but their decision to state outright "the tornado very highly likely reached EF5 intensity" and not say "certainly" does some heavy lifting here.

It is almost certainly an issue with the equations being experimental - that has to be factored in every time a study is released that includes treefall analyses and Monte Carlo stuff. There simply just hasn’t been enough done to factor it in as a definitive part of the EF scale in my opinion.

 

[WeathermanLeprechaun]

This isn't me saying that I think the rating reasonable or that I agree with their decision to keep it at EF4 180 - but I do understand where they are coming from if their concern lies with such simulations containing too many generous assumptions, which is likely to me.

If that is the case, then the fact that they weren't getting values well exceeding 200 mph means that if there is something wrong with the assumptions they made (and there very likely is, when considering the real world vs. idealized simulations), these numbers cannot be 100% trusted, even with Monte Carlo simulations. They add good credence to the fact that it likely reached EF5 intensity, but it's not as set-in-stone as structural damage calculations would be. Unlike with structural damage indicators, there isn't a standardized way to approach granting a rating based on the simulations. I was hoping Enderlin would get the ball rolling when it comes to this type of thing, but it appears that it has not (at least, not yet).

Also, it is worth mentioning that the numbers they are using are almost explicitly median/mean values and ranges. They almost certainly did get runs below the 200 mph threshold for any one or more of these DIs they tested on, and if the amount of runs pointing to EF4 rather than EF5 intensity is within a single standard deviation, the decision to keep it at EF4 is honestly a good one to me. FWIW, I do still think it reached EF5 intensity, and it should be rated as such. Of course, I'm throwing out a lot of hypotheticals here, and I could be wrong on what the true reasoning is... but their decision to state outright "the tornado very highly likely reached EF5 intensity" and not say "certainly" does some heavy lifting here.

Yes, wording is key but if that's what we're getting at to justify not going for EF5 or speculating why it hasn't, that shows you there's no great reasons it shouldn't if we're scratching at wording to think why they haven't.

 

[jiharris0220]

This isn't me saying that I think the rating reasonable or that I agree with their decision to keep it at EF4 180 - but I do understand where they are coming from if their concern lies with such simulations containing too many generous assumptions, which is likely to me.

If that is the case, then the fact that they weren't getting values well exceeding 200 mph means that if there is something wrong with the assumptions they made (and there very likely is, when considering the real world vs. idealized simulations), these numbers cannot be 100% trusted, even with Monte Carlo simulations. They add good credence to the fact that it likely reached EF5 intensity, but it's not as set-in-stone as structural damage calculations would be. Unlike with structural damage indicators, there isn't a standardized way to approach granting a rating based on the simulations. I was hoping Enderlin would get the ball rolling when it comes to this type of thing, but it appears that it has not (at least, not yet).

Also, it is worth mentioning that the numbers they are using are almost explicitly median/mean values and ranges. They almost certainly did get runs below the 200 mph threshold for any one or more of these DIs they tested on, and if the amount of runs pointing to EF4 rather than EF5 intensity is within a single standard deviation, the decision to keep it at EF4 is honestly a good one to me. FWIW, I do still think it reached EF5 intensity, and it should be rated as such. Of course, I'm throwing out a lot of hypotheticals here, and I could be wrong on what the true reasoning is... but their decision to state outright "the tornado very highly likely reached EF5 intensity" and not say "certainly" does some heavy lifting here.

I completely disagree, it’s standard practice to use the median value in any scientific evaluation. The fact that all three different studies landed on roughly the same values is a classic example of high confidence.
If the results support EF5 intensity, then it’s EF5, If the median and upper runs both support EF5 intensity, that is a majority range, being conservative because some runs (blindly assuming here) did get values below 200mph, makes little to zero sense in objective analysis.
The term “very highly likely” is as good as you can get in the field of science, which is really just another way of saying a “high degree of certainty”. This wording is really just confirming that this tornado was at EF5 intensity.
If being in the lower side of EF5 deters a proper rating, you might as well move the EF5 bar to 220mph.

Edit: This comment isn’t directed at you, and sorry if it comes off a bit hard, it’s mainly directed in just the overall way we rate high end tornadoes in general these days.

 

[WeathermanLeprechaun]

I completely disagree, it’s standard practice to use the median value in any scientific evaluation. The fact that all three different studies landed on roughly the same values is a classic example of high confidence.
If the results support EF5 intensity, then it’s EF5, If the median and upper runs both support EF5 intensity, that is a majority range, being conservative because some runs (blindly assuming here) did get values below 200mph, makes little to zero sense in objective analysis.
The term “very highly likely” is as good as you can get in the field of science, which is really just another way of saying a “high degree of certainty”. This wording is really just confirming that this tornado was at EF5 intensity.
If being in the lower side of EF5 deters a proper rating, you might as well move the EF5 bar to 220mph.

Thank you regarding the "highly likely" bit. I don't think there's another better way of saying it was a EF5 without saying it. It's probably just cautious language instead of using certain as not to discredit the rating or throw up further discourse

 

[slenker]

I completely disagree, it’s standard practice to use the median value in any scientific evaluation. The fact that all three different studies landed on roughly the same values is a classic example of high confidence.
If the results support EF5 intensity, then it’s EF5, If the median and upper runs both support EF5 intensity, that is a majority range, being conservative because some runs (blindly assuming here) did get values below 200mph, makes little to zero sense in objective analysis.
The term “very highly likely” is as good as you can get in the field of science, which is really just another way of saying a “high degree of certainty”. This wording is really just confirming that this tornado was at EF5 intensity.
If being in the lower side of EF5 deters a proper rating, you might as well move the EF5 bar to 220mph.

Edit: This comment isn’t directed at you, and sorry if it comes off a bit hard, it’s mainly directed in just the overall way we rate high end tornadoes in general these days.

I never claimed it wasn't common practice to use median values, that's reasonable. Utilizing median values actually tends to be better than using the mean in a lot of scientific fields, especially my own. I'm saying that if their issue is the simulations and the no-doubt generous assumptions they gave it, then you can't call that high confidence, even if every result agrees on a number exceeding 200 mph.

You'd be relying on overly generous assumptions in the first place. And also, defining the "majority" or where you'd "cut off" runs of low confidence (with high or low wind calcs) would be defined by a standard deviation. If the amount of runs that support EF4 intensity is within a single standard deviation, then that isn't high confidence, even if you get median values in the 200+ range. That's essentially saying that >35% of your model runs yield sub-EF5 level intensity. In truth, however, I'm making a bunch of assumptions here myself. I could be off about any one of these things, I'm just offering a possible (and reasonable IMO) explanation as to why they kept it at EF4.

It seems as though a lot of this discourse would be solved if there was a much more detailed account on why they kept it at EF4, or even a corresponding paper that would clear it up. I'm assuming the best here, which may be a mistake given how inconsistent damage surveying is.

Also, this isn't too hard at all, you're good. If I am off-base about anything, I expect to be called out or given the other side of the argument. No need to specify that.

 

[buckeye05]

I could be wrong, but the vibe I am getting is that they suspected EF5 from the beginning, did everything in their power to get it upgraded via non traditional means, but didn't get approval from higher ups due to it not being the high confidence slam dunk that Enderlin was. I guess for lofted objects to be counted as EF5 moving forward, the calculations will have to come out well above and beyond the EF5 range threshold? I do understand that if its one standard deviation away from EF4, it might not be quite solid enough from a scientific perspective, but I just hope this isn't the beginning of backpedaling away from using math to determine EF5 intensity.

 

[WeathermanLeprechaun]

Got blasted on Reddit for asking why it wasn't already upgraded given the precedent Enderlin set and I got a bunch of "the rating isn't finalised" and to have patience. The difference is the maths is there already, so by that logic, why isn't it upgraded already? I suppose asking on Reddit isn't the best because, obvious reasons. It's confusing.

 

[buckeye05]

Got blasted on Reddit for asking why it wasn't already upgraded given the precedent Enderlin set and I got a bunch of "the rating isn't finalised" and to have patience. The difference is the maths is there already, so by that logic, why isn't it upgraded already? I suppose asking on Reddit isn't the best because, obvious reasons. It's confusing.

Lol but it is finalized. It’s literally in NCEI, which is where finalized data is entered. I’m not sure how anyone could reach that conclusion.

Reddit unfortunately seems to be where the…less knowledgeable of tornado enthusiasts gather, so it’s not the place to find answers imo.

 

[jiharris0220]

I could be wrong, but the vibe I am getting is that they suspected EF5 from the beginning, did everything in their power to get it upgraded via non traditional means, but didn't get approval from higher ups due to it not being the high confidence slam dunk that Enderlin was. I guess for lofted objects to be counted as EF5 moving forward, the calculations will have to come out well above and beyond the EF5 range threshold? I do understand that if its one standard deviation away from EF4, it might not be quite solid enough from a scientific perspective, but I just hope this isn't the beginning of backpedaling away from using math to determine EF5 intensity.

That’s what I’m getting at too, I simply can’t find a valid reason as to why this was kept the rating it’s at, at the very, very least, one could’ve expected a 195 or 200mph rating, but not even that.
If you blindly assume the study just left out that a good enough majority of the runs were below the EF5 threshold to deter confidence, then that means they’re actively contradicting themselves.
The only possible explanation other than straight egregious underrating from higher ups, is that these three studies were finalized well after the conclusive 180mph ratings was inputted, and we’ll simply have to wait awhile for this tornado to get upgraded again, similar to Enderlin. One can hope.

 

[Grand Poo Bah]

It seems as though a lot of this discourse would be solved if there was a much more detailed account on why they kept it at EF4, or even a corresponding paper that would clear it up. I'm assuming the best here, which may be a mistake given how inconsistent damage surveying is.

This doesn't exist because keeping it EF4 wasn't their choice, and I'd imagine they're pissed about it. They aren't going to do a detailed write-up justifying a rating they disagree with.

This is setting a terrible precedent that all the work and research all these scientists have done has less weight than the opinions of a couple civil engineers. It's huge egos hanging onto the remnants of a flawed system they created. I don't expect it to last much longer, but I fully expect it to get increasingly frustrating before a breaking point is reached and change is forced.

I agree with @jiharris0220 's points, so I don't have anything else to add really. There's a higher probability it required EF5 winds to cause the damage they studied, than EF4 winds. There is nothing "safe" about assigning the less likely rating. In fact, it's less "safe" to do so.

 

[WeathermanLeprechaun]

Lol but it is finalized. It’s literally in NCEI, which is where finalized data is entered. I’m not sure how anyone could reach that conclusion.

Reddit unfortunately seems to be where the…less knowledgeable of tornado enthusiasts gather, so it’s not the place to find answers imo.

It probably wasn't obvious but that's why i put the "obvious reasons". Going to go fry them for a few minutes about being confidently incorrect.

 

[buckeye05]

That’s what I’m getting at too, I simply can’t find a valid reason as to why this was kept the rating it’s at, at the very, very least, one could’ve expected a 195 or 200mph rating, but not even that.
If you blindly assume the study just left out that a good enough majority of the runs were below the EF5 threshold to deter confidence, then that means they’re actively contradicting themselves.
The only possible explanation other than straight egregious underrating from higher ups, is that these three studies were finalized well after the conclusive 180mph ratings was inputted, and we’ll simply have to wait awhile for this tornado to get upgraded again, similar to Enderlin. One can hope.

But I’m pretty sure once something is in NCEI, the proverbial door is closed so to speak. There is nothing in the wording that suggests they are leaving things open to adjustments.

It really sounds like NWS Norman knows this thing was an EF5, and they did everything in their power to prove it, but got shut down by higher ups. Remember how Enderlin required all kinds of up the chain approval? I would bet that Norman tried that pathway but things didn’t go well this time.

 

[WeathermanLeprechaun]

A office that has given a tornado a EF5 rating BASED OFF a non-traditional DI knows their stuff. The most good explanation is like many have said, they got turned down. That is extremely irritating, man. I guess we're still going backwards no matter what.

 

[buckeye05]

A office that has given a tornado a EF5 rating BASED OFF a non-traditional DI knows their stuff. The most good explanation is like many have said, they got turned down. That is extremely irritating, man. I guess we're still going backwards no matter what.

Exactly. Norman is one of the most intuitive, forward thinking offices when it comes to damage surveying and EF scale application techniques, but what can you do if your attempts to move forward are blocked by people farther up the chain? Maddening.

 

[jiharris0220]

But I’m pretty sure once something is in NCEI, the proverbial door is closed so to speak. There is nothing in the wording that suggests they are leaving things open to adjustments.

It really sounds like NWS Norman knows this thing was an EF5, and they did everything in their power to prove it, but got shut down by higher ups. Remember how Enderlin required all kinds of up the chain approval? I would bet that Norman tried that pathway but things didn’t go well this time.

Really does look like Enderlin was a “fluke”, and that the only way we’re ever getting an EF5 is undeniable evidence that a tornado achieved that intensity, in which the damage indicator has to be well into EF5 territory to be accepted.
It seems that we’re still very much in the days of “what evidence is there against an EF5 rating” rather than “what evidence proves this was an EF5”.

 

[WeathermanLeprechaun]

Exactly. Norman is one of the most intuitive, forward thinking offices when it comes to damage surveying and EF scale application techniques, but what can you do if your attempts to move forward are blocked by people farther up the chain? Maddening.

This is definitely another situation where a tornado was robbed of a EF5 rating and it's all because of higher ups and stretching the goalposts and being constantly dismissive. We will never get anywhere with ratings with attitude like this. How are these higher ups even in their position when all they do is ruin historical records

 

[slenker]

By no means is what I said blindly following anything, it’s pretty standard scientific procedure. Calculating the statistics of these model runs (standard deviation, variance, mean, median, etc) would be extremely important for determining how the data should be interpreted. I’m not omitting the possibility that it was from higher ups stonewalling it, I’m just saying that there exists a plausible and reasonable explanation to keep this at EF4, despite the simulations mostly agreeing with an EF5 intensity tornado.

Again, FWIW, I think the evidence presented for it being EF5 would be strong enough on its own, but again, the possible reasons for it being kept at EF4 that I presented are very justified. If you want to argue that the reasons I presented aren’t good enough, go right ahead, I’m all ears.

 

[Kds86z]

Exactly. Norman is one of the most intuitive, forward thinking offices when it comes to damage surveying and EF scale application techniques, but what can you do if your attempts to move forward are blocked by people farther up the chain? Maddening.

100% if Norman NWS could do it they would.

 

[buckeye05]

By no means is what I said blindly following anything, it’s pretty standard scientific procedure. Calculating the statistics of these model runs (standard deviation, variance, mean, median, etc) would be extremely important for determining how the data should be interpreted. I’m not omitting the possibility that it was from higher ups stonewalling it, I’m just saying that there exists a plausible and reasonable explanation to keep this at EF4, despite the simulations mostly agreeing with an EF5 intensity tornado.

Again, FWIW, I think the evidence presented for it being EF5 would be strong enough on its own, but again, the possible reasons for it being kept at EF4 that I presented are very justified. If you want to argue that my reasons aren’t good enough, go right ahead, I’m all ears.

That’s totally fair as well and I’m still very much acknowledging that. I just would want to hear a full breakdown of the rating decision process from the people who made the calls before I call it valid. I’m currently in the camp of this being enough for EF5 when you combine the math evidence with the contextual damage this thing produced, especially when the NCEI entry calls it “very highly likely” or whatever the wording was.

 

[slenker]

That’s totally fair as well and I’m still very much acknowledging that. I just would want to hear a full breakdown of the rating decision process from the people who made the calls before I call it valid. I’m currently in the camp of this being enough for EF5 when you combine the math evidence with the contextual damage this thing produced, especially when the NCEI entry calls it “very highly likely” or whatever the wording was.

And that is entirely reasonable. Nothing in my post was directed towards anyone in particular, my bad if it came off that way.

I fully agree that they should have had more to say about why they didn’t give it the rating, given the numbers they gave here. It’s frustrating because the first question someone with the smallest bit of interest in this topic is going to ask is, “why isn’t it an EF5 if all these studies you just told me about support it being one?” The fact that it isn’t addressed here is frustrating from a science communication standpoint, they should be very clear on why such data is not taken into account in the final rating.

Experience in my own field tells me that researchers and scientists are really bad at communicating ideas to people outside their own fields (and even within their own fields too!), and this just feels like a prime example of that.