Research to confirm or debunk funnel vs tornado paradigm

[Dan Robinson]

The current paradigm in meteorology is to assume that all observed funnels are not tornadoes unless there is evidence to the contrary (a visible ground circulation/condensed subvortices/debris cloud or damage after the fact). To my knowledge, there is no evidence or research behind this paradigm, it's just always been an accepted part of severe storms meteorology and storm spotting.

I believe there is enough evidence that this paradigm is false in these specific, but very common, categories of funnels:

A. Laminar funnels under the rotating cloud bases of a supercell mesocyclone. By the time a funnel appears in these scenarios, a damage-capable ground circulation has long since been in progress in some form.

B. Persistent vertical to near-vertical laminar funnels extending more than 1/3 of the way to the ground from updraft bases. These are mostly nonsupercell phenomena associated from the stretching of near-surface vorticity from boundaries in environments with sufficient low-level CAPE. This category would include "cold air funnels" that our current operational paradigm states "rarely reach the surface". In the known cases where a chaser or other observer is directly under such a feature, there has been an observable (damage-capable) ground circulation present.

Proposed research methodology: Assemble a data set of funnel events in the scenarios A and B (outlined above) observed from close range where the ground underneath the funnel is clearly visible, in other words, where the presence or lack of a ground circulation can be conclusively determined. From this data set, calculate the likelihood of a funnel being a tornado (having a damage-capable ground circulation) to either confirm or challenge the current paradigm of assuming most funnels in these scenarios are not tornadoes.

Types of funnels not considered in this research:

Shear funnels: Midlevel vortices or horseshoe vortices.

Transient low-level vortices: Short-lived vortices under updraft bases, advancing gust front shelf clouds or other features.

Criteria: Collected events must meet the following criteria::

1.) A funnel cloud in the category of A or B (described above) is present and extending some discernable distance below cloud base, up to 2/3 of the way to the ground.
2.) The ground underneath of the funnel is clearly visible in at least one of the images/videos, allowing the presence or absence of a ground circulation to be conclusively determined.

Event List: Supercell funnels with ground circulations

1.) April 29, 2022 - Elmo, Kansas
High-based funnel descended 1/4 of the way to the ground. A damaging circulation was documented at ground level.
Funnel image 1
Funnel Image 2
Ground circulation video 1

2.) December 1, 2018 - Bluffs, Illinois
Supercell tornado funnel descended 1/5 of the way to the ground.
Funnel image 1
Ground circulation image 1

3.) May 16, 2017 - McLean, Texas
Supercell tornado funnel extending 1/4 to 1/2 of the way to ground with continuous visible ground circulation. The funnel eventually fully condensed to ground in the later stages of the tornado.
Funnel/ground circulation image 1

4.) May 10, 2010 - Medford, Oklahoma
Funnel 1/3 of the way to ground with debris cloud appearing after funnel dissipated.
Funnel image 1
Ground circulation image 1

5.) August 24, 2015 - Dubbo, NSW, Australia
Supercell tornado funnel 1/4 way to ground with debris, audible roar and occasional condensed near-ground subvortices.
Tornado and funnel video 1

Event List: Nonsupercell funnels with ground circulations

1.) June 13, 2021 - Tuttle, Oklahoma
"Cold air funnel" descending at least 1/3 of the way to the ground.
Funnel video 1
Ground circulation video 1

2.) May 23, 2021 - Arriba, Colorado
High-based nonsupercell funnel along linear convective base with ground circulation.
Funnel image 1
Funnel video 1
Ground circulation video 1

Could this rise to the level to qualify for a formal published paper? As an outsider to the science community, I'd be interested to hear from qualified individuals if this is possible and the best way to approach it to make it a useful endeavor.

 

[Mike Smith]

Dan,

I believe you have had an important insight. Thanks so much for posting it!

On April 29, 2022 (hours before the Andover and the Tampa Tornadoes, the latter I observed at very close range), this tornado did minor damage in Kansas. It came from a supercell and no one -- with the existing paradigm -- would guess it was a tornado. I reported it to the NWS as a "funnel cloud."

Keep up the important work!

Mike

 

[Dan Robinson]

Mike, that is another great example. Here is Max Olson's view of the ground circulation of that one:

 

[Dan Robinson]

I updated the original post in this thread with an initial collection of events. I'll update it as more are inventoried. My prediction is that we'll have countless examples of funnels (meeting the criteria in the first post) that are tornadoes, and very few funnels that conclusively aren't tornadoes. For that reason I'm a little more interested in trying to find evidenced instances of the latter, as they are presumably much more rare. I don't have any myself! I have many images of funnels with inconclusive tornado status (the ground under it blocked by trees, buildings, corn or terrain and/or I was just too far away) but none where I know for a fact that there was no ground circulation.

For the purposes of the study, I think limiting this to funnels at any distance below close base up to 2/3 way down would be sufficient to prove/disprove the theory. I think we can all agree that funnels (meeting the criteria in the first post) below 2/3 of the way down are almost always tornadoes without dispute.

 

[Andrew Pritchard]

This is a fascinating topic, Dan! I was very close to you during that December 1st, 2018 ground circulation with tiny funnel aloft. It was actually the rapidly shifting surface winds and eventual ground circulation that caught my attention first driving north on that highway - I only noted the small funnel cloud aloft after pulling over and looking up.

June 29th, 2024 was a memorable, low-end chase day for me that provides a couple examples. This was a 0% tornado risk day, but I picked out a nearby surface convergence zone that looked similar to past days that resulted in surprise landspout or mini-supercell events. I sat near Hoopeston, Illinois and watched several well-defined funnel clouds develop beneath updraft bases, one of which fit into the "laminar, persistent, nearly horizontal" category.

This funnel cloud was too far away to my northeast to confirm ground circulation, but it was pretty darn convincing because of the persistent and laminar nature.





(sorry for the crude YouTube screenshots)

Shortly after, a new, closer funnel cloud developed and made ground contact resulting in a rather long-lived (12-15 minute) landspout tornado near Hoopeston, IL, eventually crossing the border into Indiana.


Did that first funnel cloud have a ground circulation? Maybe? Probably? I'll never know, but I kinda suspect it did.

Link to the full video from this day if anyone is curious.

 

[John Farley]

I have a few examples relevant to this discussion. First one is near August, AR on May 25, 2011. Smallest funnel I have ever seen on a definite tornado:



It does not show up well in this photo, but there was definitely a dust swirl underneath. And like one that Andrew mentions above, I saw the dust swirl before I looked up and saw the funnel. It is not in any official database, even though it was a tornado. I did not have the local NWS phone number so I called it in to 911 and requested it to be reported to the NWS, but apparently it was not. This tornado was definitely supercellular.

This one, from northwest of Mancos, CO on May 19, 2015, is solidly in the "maybe" category. A tornado warning was issued, based on a spotter report (not mine, the warning was issued almost as soon as I noticed the funnel)., and some wind damage was noted in the area - fence blown over and maybe some trees, IIRC. But it was never surveyed or confirmed as a tornado. Here is my picture of this one"



I was not chasing when I saw this; we had just set out on a drive to California and came upon it. I am guessing a landspout-type funnel, but not totally sure.

My third example is a "probably not," but I can't be totally sure. It occurred west of Chromo, CO on September 5, 2015. I was about a half mile away, and ended up having to turn around and retreat because it was moving directly toward me. By the time I had driven a bit and stopped to look again, it was gone:



After the storm passed, I drove back into the area it was over and could find no damage of any kind. Not to prove there wasn't any, but I could not find any. This storm was supercellular, moving on a path to the right of the motion of other storms, and I observed cloud base rotation near the funnel.

Nothing here that would rise to the level of being publishable, but hopefully information relevant to the discussion.

 

[Andy Wehrle]

Not sure if this is related, but in my (relatively limited compared to most chasers of my generation, but considerably expanded within the last two years) experience, the "textbook" interpretation of tornadoes as originating from a preexisting wall cloud is often, if not more often than not, invalid.

On those occasions in which I've arrived in viewing position on the updraft region of a tornado-warned supercell to find a "classic" low, ominous wall cloud such as that near Genoa, NE a week ago yesterday:



...it has not produced a tornado.

On those few occasions in which I have been witness to tornadogenesis, it has originated from a relatively flat, nondescript rain-free base. The RFD cuts in, carving out a compact cyclonic "swirl" on the north end of the "horseshoe," at or near where it meets the inflow. The funnel descends from this, and the wall cloud forms concurrently if not after. This is what occurred with the EF4 at Keota, IA on March 31, 2023.

Early in the tornado's life cycle, note how relatively flat the entire base is apart from the funnel protruding from it, and the sharp vertical edge of the RFD clear slot on screen left:



The tornado briefly displays a classic fully-condensed cone funnel, but note the "tendril" of condensation already wrapping around it on the left:



Within the span of 45-60 seconds, this condensation has wrapped around the entire top of the funnel, which then undergoes a rapid vortex breakdown, brief multiple-vortex phase and reorganization as a much "shorter" dust-filled stovepipe and then wedge now crowned by what could colloquially be termed a dramatic "wall cloud."