The long-established funnel VS tornado paradigm: flawed?

[Dan Robinson]

It's long been the practice of considering a visible condensation funnel partway to the ground as only a "funnel" and not a tornado, in the eyes of the NWS, spotters, emergency managers and in storm data. But given what we commonly observe in the field - that is, that most visible funnels more than 1/3 of the way to the ground have a damage-capable ground circulation and therefore are, by definition, tornadoes - isn't this a flawed paradigm?

I commonly hear of tornado reports dismissed either by chasers or officials because the condensation did not visibly extend to the ground. I've experienced this myself - a tornado with condensation to the *treeline* (this one here) wasn't counted because ground contact could not be clearly observed below the trees!

We all know that tornadoes are wind, not condensation. The visible funnel is only the *core* of the tornado that has managed to condense if the RH and pressure drop permits, and that the tornadic circulation extends beyond this. Most of us have been close to what more distant observers identify as "funnels" to see that there is indeed a ground circulation. It is kicking up spray/small debris that isn't visible from more than a quarter mile away or from behind trees.

My experience with cold-core "cold air funnels" has revealed that many of these are actually (very weak) tornadoes, with ground circulations that would cause damage if they struck a low-end DI (like a carport or mobile home). There are pictures of such "cold-air funnels" extending more than 2/3 of the way to the ground.

So why hasn't the paradigm shifted to start giving more "funnels" the status of tornado? We know from *direct observation* that they are more likely to be tornadoes than not if the visible funnel extends more than 1/3 of the way down.

 

[calvinkaskey]

Tornado I chased in North Carolina

For much of its estimated 20 minutes it looked like it was off the ground but there was damage even in the very formative stages and it looked to end in a wisp on the ground without hardly any visible connection to the cloud.

It's long been the practice of considering a visible condensation funnel partway to the ground as only a "funnel" and not a tornado, in the eyes of the NWS, spotters, emergency managers and in storm data. But given what we commonly observe in the field - that is, that most visible funnels more than 1/3 of the way to the ground have a damage-capable ground circulation and therefore are, by definition, tornadoes - isn't this a flawed paradigm?

I commonly hear of tornado reports dismissed either by chasers or officials because the condensation did not visibly extend to the ground. I've experienced this myself - a tornado with condensation to the *treeline* (this one here) wasn't counted because ground contact could not be clearly observed below the trees!

We all know that tornadoes are wind, not condensation. The visible funnel is only the *core* of the tornado that has managed to condense if the RH and pressure drop permits, and that the tornadic circulation extends beyond this. Most of us have been close to what more distant observers identify as "funnels" to see that there is indeed a ground circulation. It is kicking up spray/small debris that isn't visible from more than a quarter mile away or from behind trees.

My experience with cold-core "cold air funnels" has revealed that many of these are actually (very weak) tornadoes, with ground circulations that would cause damage if they struck a low-end DI (like a carport or mobile home). There are pictures of such "cold-air funnels" extending more than 2/3 of the way to the ground.

So why hasn't the paradigm shifted to start giving more "funnels" the status of tornado? We know from *direct observation* that they are more likely to be tornadoes than not if the visible funnel extends more than 1/3 of the way down.

 

[Jeff Duda]

We know from *direct observation* that they are more likely to be tornadoes than not if the visible funnel extends more than 1/3 of the way down.

We do? Can you name a study that has compiled such statistics to corroborate your statement?

I'm not saying there aren't issues with that operational/spotter definition of a tornado, but I would guess that threshold is based on dated knowledge and uncertainty. It would certainly help if someone could provide hard numbers to give an indication of how likely a condensation funnel is to be a tornado based on how close to the ground it is.

 

[Shane Adams]

I've always gone with the "if it's more than halfway down it's a tornado" rule, for many of the reasons Dan stated. I've never been interested in pragmatic, scientific, err-to-the-side-of-caution type analysis of this topic, because it doesn't take into account the observations chasers and spotters make. You can tell a lot about a condensation funnel (and its potential to have winds on the ground) from years and years of experience watching them. Size, shape, duration, visible rotation, angular momentum....these are all aspects that only a visual observation can take and make a determination. Is there any way to prove beyond the shadow of a doubt that a funnel is/isn't a tornado without a visible debris cloud? No. But how many full condensation tornadoes have you seen where the water vapor is reaching the ground but zero visible debris/dust is being kicked up? Is full condensation to the ground required for tornadic winds to be present? No.

I lost interest years ago in trying to report and "keep the data accurate" because there have been so many instances of official reports not reflecting what I personally witnessed and reported, or my reports being dismissed altogether. I used to have a "butthurt, they didn't listen to me" kind of attitude about it all, but these days my mind is more like "I know the data will never be 100% accurate because I know tornadoes I have in the can that aren't in the storm records." So beyond a trivial interest for the sake of conversation, I have no stake in any official "rule" or "stance" about what makes a funnel a tornado or not. I make those calls myself, and log the data into my own personal records; I really couldn't care less how it's officially documented.

FWIW, here's a great example of what I'm talking about. Although this funnel never fully condensed to the ground and we never observed debris, I still reported it as a probable tornado. I made this judgment based on the visual observational clues I talked about above. I sent the video to ICT, and they confirmed it based on my video. It is in the NCDC archives.


June 3, 2001 - NW of Attica, KS

 

[Adam Lucio]

There have been quite a few times where I've seen a dinky, not so impressive looking funnel cloud reach the ground. Im a stickler when it comes to my own personal stats, if I only see a funnel halfway down but can't see the ground because of trees or buildings I won't count it unless a damage survey or other chaser observation confirms it for me. Part of it is I want to be accurate in what I report/record and part of it because I want to establish credibility for myself.

An example of this would be this funnel before the Bennington tornado, I wouldn't doubt it had some sort of little ground poof at one point, but I didn't see it, can't confirm it, and have not heard anything else to indicate that it actually touched down, so it remains as just a funnel cloud for me.

Pardon the cheesy quote, its from my facebook banner photo and the only piece of photographic evidence I can reference. Too lazy to upload a different image.

 

[Dan Robinson]

We do? Can you name a study that has compiled such statistics to corroborate your statement?

Absolutely. Personal observation and countless photos/videos. I'm not aware of any official study, but that's why I'm bringing it up. My question is, what study is this paradigm based on? I'm challenging the status quo based on what I see in the field. I believe the "if the condensation isn't on the ground and you can't see debris, it can't be called a tornado" rule is a relic from the old days akin to "open your windows to prevent your house from exploding" that just has never been put under any real scrutiny.

Adam, the closest chaser video to that first Bennington tornado/funnel I know of is Daniel Shaw, who along with another spotter reported it as a tornado: I can't make out much in his video, but here it is: https://www.youtube.com/watch?v=tWwPksvXiHo

 

[JustinPhillips]

I also have a pretty good example of a tornado that doesnt look at all in contact with the ground. The brief wedge near Madelia, Minnesota on May 8 of this year didnt look like it was on the ground from my point of view, or in my photos. But video from Kholby Martin and a couple others shows that vorticies were indeed briefly on the ground. I think I seen a faint dust swirl right as I was taking the shot but wasnt sure. I kinda assumed from the size and motion that it had indeed touch down, but wasnt sure until after when the reports came in.

 

[rdale]

We know from *direct observation* that they are more likely to be tornadoes than not if the visible funnel extends more than 1/3 of the way down.

...is not the same as...

I believe the "if the condensation isn't on the ground and you can't see debris, it can't be called a tornado" rule is a relic from the old days akin to "open your windows to prevent your house from exploding" that just has never been put under any real scrutiny.

We don't use that "rule" now (if ever?) Most consider more than half-way to be a valid tornado report.

But there are MANY funnel clouds that extend 1/3 - 1/2 way down which never cause damage at the surface. Maybe some 2/3rds for all we know. The "open your windows" rule was scientifically invalidated, which is why we don't say that anymore. A handful of observations showing that funnels with 1/3rd coverage can cause damage is not a scientifically valid dataset.

 

[Sean Ramsey]

I've always been on the side of the fence where if I can't see activity on the ground, or if there is any question whether it is or not, then I don't call it a tornado. There are times where I'm too far away, or trees are in the way or something else where I just can't tell, and I have a hard time calling it a tornado if I just don't know. If it's confirmed by someone else after the fact, then I have no problem calling it what it is. Earlier this year near Larned, KS I had this happen, where i had a nice visible funnel about halfway down, but there's no evidence to me that said it reached all the way down, neither by my own eyes or anyone else's report. So it remains a funnel in my books, although if someone wants to talk me into calling it a tornado my 2014 season would look much better

 

[Skip Talbot]

Maybe some 2/3rds for all we know.

Or all the way condensed. If your condensation funnel is associated with a benign circulation that is incapable of doing damage, is that still a tornado? I think there are some examples that get pretty close to this.

On the other hand, many (most?) funnels seem to form from circulations that develop at the surface or are surface based. You could argue that all of these are tornadoes. I've seen a lot more tornadoes than I have funnels not associated with tornadoes, as it seems whenever I see a well defined funnel that isn't a tiny shear funnel on the gust front or an elevated funnel in the midlevels of the updraft tower, it's usually kicking up something on the ground or has a surface based wind field circulating underneath it. I think it might be possible that almost every funnel cloud you see forming under a surface based supercell in the region of tornadogenesis has a connection to the ground. Are they all tornadoes or only the ones that can do damage?

There's a huge grey area. Is this a tornado?

I thought I saw the funnel dip to the treeline so I reported it as one. A closer spotter reported it as a funnel, and there was no damage, but that funnel is indeed most of the way to the ground. I didn't count it as a tornado when I wrote my log up for that event.

How about this?

Looks like a nice little tornado from that angle, but the circulation causing that powerflash is actually quite broad and diffuse, something many would call RFD wrap around rather than a tornadic debris cloud. That's definitely in that fuzzy grey area between sloppy RFD circulations and weak tornadoes.

Same storm, same sized funnel, same location. Without the wet RFD surge at the surface, many would just call this a funnel, even though it might still have a similarly sized and intense wind field at the surface.

This was an obvious tornado, but the circulation at the ground was barely damaging for much of its life. NWS rated this EF0 with a very short track (less than a mile I think), even though the circulation under that funnel traveled for several miles. It was so weak it wasn't doing damage. I think it would still be nearly universal among chasers that it's still a tornado, even if the ground circulation isn't doing EF0 damage, but it's approaching that threshold where you have a nicely defined, ground kissing funnel that might not be intense enough to be called a tornado.

Here's an obvious example of not a tornado. This feature was rotating under a classic supercell, and developed right where a tornado should, and it condensed most of the way to the ground. Some chasers (most at a distance of a few miles) reported this as a funnel. It was, however, rotating at the speed akin to a roll cloud than a tornadic funnel. Definitely too weak to be a tornadic funnel cloud, but it is rotating and it might have a weak surface based circulation. There's a big grey area between these examples.

 

[Jeff Duda]

Absolutely. Personal observation and countless photos/videos. I'm not aware of any official study, but that's why I'm bringing it up. My question is, what study is this paradigm based on? I'm challenging the status quo based on what I see in the field.

It's fine if you want to challenge the status quo, but that generally places the burden of proof on you. Thus you need to come up with some compelling evidence to convince people to make changes. I'm sorry, but "personal observation" really doesn't qualify as "compelling evidence".

Again, I agree with you that there is a problem with the way NWS offices treat inaccurate reports of funnel clouds that should be tornadoes (and vice versa). I think at the heart of the matter is the definition of a tornado - it is subjective and bears uncertainty. I think the relevant phrase in the definition is not the ground contact, but "violently rotating". That is a qualitative phrase. How "violent" is violent enough for a tornado? Capable of doing damage??? If so, to what? A 25 mph gust can break off a 3-mm diameter twig from a tree. Is that not damage?

The use of qualitative language lends people to assuming threshold values. According to the EF-scale, the slowest wind speed for an EF0 is 65 mph, by which I presume to understand that the EF-scale committee apparently thought that a 65-mph wind was the definition of "violent". That's fine. I accept that. We, as scientists, can work with that. However, that still doesn't address the issue of when a funnel cloud/tornado is viewed remotely. Unless there are instruments under the funnel cloud, this issue will pretty much never be resolved, because no one can say with certainty if the pre-tornadic surface vortex is indeed sub-tornadic in strength or not.

Long story short, you can challenge the notion all you want, but I don't think there is a better alternative save for having direct measurements of the air flow near a funnel cloud. One reason in favor of keeping existing guidelines for defining tornadoes is it allows for the bias in the tornado counts to be estimated better than if the thresholds and criteria for a true tornado were always changing. That's kind of a hand waving argument, though. Again, the issue is uncertain.

 

[calvinkaskey]

I tried to get the NWS in Binghamton, N.Y. to count what I knew was a funnel. The funnel showed clear violent rotation, showed them where and when in the video to look and it was in the right area of the thunderstorm, good hook echo and just had a warning 5 min. before. Next year 4 people die near the same area from a tornado. All sighting count toward risk. Best experts in the business said yes it was probably a funnel.

 

[Dan Robinson]

If the original criteria isn't based on an objective study, then I think it's reasonable to question it in light of what is frequently observed. You've got a point in that excluding some bona-fide tornado reports might compensate for bias, although that would be hard to quantify. How many weak, non-fully-condensed tornadoes have we seen in which the debris cloud was present only because dry, unplowed fields allowed it? Were those really tornadoes if the dust was kicked up by a 50mph wind? The same events may very well have been called funnels - and thus not counted - if they had occurred over a rain-soaked field of planted wheat.

I suppose this is the same problem as the EF5-intensity tornado that doesn't strike high-end DIs and ends up as an EF3. If a tree falls in the forest and no one hears it, did it happen?

 

[calvinkaskey]

Looking at numerous damage pictures of trees and structures it seems like one can rate tornadoes confidently on tree damage. I notice there is a big difference between tree damage with and without leaves, but if one knows if the tree had leaves I don't see many problems with this approach.
I suppose this is the same problem as the EF5-intensity tornado that doesn't strike high-end DIs and ends up as an EF3. If a tree falls in the forest and no one hears it, did it happen?[/QUOTE]

 

[Dan Robinson]

Again, someone correct me if there was indeed a scientific study that resulted in the current criteria. I just don't know of one. It just seems incongruent to say one one hand, "tornadoes are much more than their visible condensation funnels" and then "if the condensation doesn't reach the ground and we can't see debris (even from great distances and behind trees/obstacles), then it's not a tornado". Especially since we have ample evidence of partial funnels that have damaging ground circulations/debris clouds. Yes, there are probably a few exceptions as Skip noted, but if it's more common than not, I think there's reason to question the standard.