Not so sure I agree with this "non-numerical damage labeling" approach. One-word-fits-all labels like those examples above are too subjective and convey practically no useful information without more detailed information (preferably from an independent source) to accompany them. People want to see numbers to understand what kind of beast they were dealing with! The non-meteorological public has heretofore been able to understand "hard" (more precise) numbers and putting these numbers into some sort of numerical context so that these labels (in the present system, "EF-range" categories) can be compared from weakest to strongest has always seemed a logical and acceptable way to handle this. That said, it does not mean that the present rating system is perfect and cannot (or should not) be improved...
Technically, however, you are correct in implying that "damage observed by a qualified professional civil engineer" is not the same thing as the windspeed that appeared to have done that damage (in the absence of actual measurement). For that reason, I've advocated for experienced professional engineers from the private sector (rather than NWS survey teams consisting of meteorologists with random qualifications) to do the field-data collecting and analysis. These engineers are specifically trained at analyzing failed-structures, so their results will likely be detailed and highly-numerical, which is more consistent with a numerical rather than verbal-description classification for the public.
Lastly, there is no easy way to reconcile differences between the Fujita Tornado Intensity Scale and the Saffir-Simpson Hurricane Category Scale. The latter tags each hurricane's intensity based upon the highest windspeed (and more precisely, lowest sea-level pressure) recorded during the cyclone's entire lifetime, which could be measured over a period of one week or more. This frequently occurs earlier over open, deep-tropical-latitude seas, so later on, at landfall (often at higher latitudes), it is usually of lower intensity. On the other hand, local tornado intensity is measured over a period of seconds or minutes at any one location, so the intensity at each location along the path is what the damage appears to indicate, even though the maximum intensity along its entire damage path will be how it is actually rated. Both are rated for maximum intensity, rather than actual intensity, at any given point in time, which contribues to inaccuracy and confusion when trying to describe the damages observed at any particular location by the stricken public. I cannot come up with any good suggestion about how the difference between locally-observed and maximum lifecycle windspeed can be resolved when labelling tornado/hurricane intensity, but it seems that a verbal "damage label" is of little help, either.
I'm not suggesting a non numerical system at all, just consistent and articulate labeling/messaging of what the EF rating system actually
is. It is a 0-5 rating of severity based upon damage assessment
. It is not based upon best estimated wind speeds or else a research doppler with a beam 10-100m above the ground would be allowed in the absence of any damage indication, calibrated anemometers would be allowed if they survive, etc. to at least loosely inform ratings.
Instead, ONLY observed damage indicators on checklists that hopefully correspond well to wind speeds are allowed. There is no factor in the EF scale rating (as far as I know) for a variety of variables such as which side of the tornado hits a structure, debris loading, dwell time, incident angle, and on and on as many others have brought up. It does factor construction quality to the extent the surveyor is capable. EF ratings seem to obfuscate nearly all factors into the observed damage that is then correlated to windspeed by the best available lab experiment or simulation. Only 'done damage' translates to a rating
at all. The system is willing to throw out a 250MPH plus low beam height doppler reading just because a tornado hits nothing, yet we have faith in doppler wind speeds with much worse beam height for warnings... it is just not very clear or consistent once it all tracks back to wind speed, and that is where I think articulate description of the rating system would help and may preclude the need to change the system if the communication was improved. We should stop describing it as a wind speed correlated system out of context.
I don't even think I have the same concerns you have about NWS surveyors following the checklists by themselves vs. a civil engineer licensed as a PE. I think that variable is lost in the noise of the others in many cases. Both parties woud use damage indicators that trace back to lab experiment, as is the heart of most forensic engineering. While I am sure a rating from someone like Tim M./Haag is going to be significantly better than someone just trained still learning about what a well made building is, following the indicator checklist developed by industry and academia is still going to be successful enough to produce an EF rating mostly in family with all the others.
Regarding the 'actual' numbers used in messaging: in my opinion the public probably only needs the simple 0-5 severity ratings, not wind speed. Policy makers, science, etc. are the ones who may need the correlated wind speeds. Few in the public are going to understand what a specific wind speed means for planning purposes.
As to the hurricane rating comparison, perhaps I was unclear, my apologies. I am talking about messaging
only and how the severity ratings are similar (EF0-5, TD-Cat 5) but take almost opposite perspectives: tornado rating is a post-actual damage, hurricane is pre-potential damage, with both approximately correlated to wind speed few really use. While a pre rating for hurricanes is necessaary and prudent, and similar is not possible for tornadoes, at least every hurricane is rated the same way. I've never seen a Cat 5 over open water be called a depression because it hits nothing. I believe any forms of inconsistency can be confusing especially for some in the public, and I am also opposed to badly collected and presented datasets, so for me this entire issue really is about accurate communication of what we are really rating. I don't think it is helpful to call a large tornado an EF0 after the fact if an on air met says 'take shelter folks, we are probably looking at an EF3 or 4 here' (this type of messaging is used by some media). No need to change the rating system if it is the best we can come up with, but it should be talked about precisely for the benefit of public safety and so anyone looking at historical data knows how many holes are in the dataset.
