Fastest Tornado Forward Speed

May 1, 2004
3,403
746
21
Springfield, IL
www.skip.cc
The satellite we observed orbiting and being pulled into the developing Wakefield EF4 on June 16 appeared to be moving at an incredible speed.

At 11:40
Watch video >

How fast though? I decided to try and measure its speed. The (preliminary?) damage survey map posted by OAX does not include this portion of the track, or it was indistinguishable from the Wakefield damage path, so I had to use triangulation to get the tornado's position.

I synced three different videos, the above and these two:
http://youtu.be/rir58pK4n_A?t=12m56s
http://youtu.be/deI_IIrhvEk?t=8m39s

Those were the first two that I found that included a continuous shot of this sequence of that fast moving rope out from a fixed and stable location. I synced the three to this lightning bolt:




I was then able to use Google Maps satellite and street view imagery to triangulate the rope's position at different time points given the location from which the video was shot:




I used a couple different methods to check my numbers and see if they were reproducible with repeatable measurements.

I picked two time points at which the rope appeared to be moving the fastest and where there were some foreground reference points to aid the triangulation. I ran the numbers several times, mainly because I found the result to be really astonishing. My more precise attempt produced a result just over 90 mph! The distance between the two points was about 2200 feet over a 16.5 second duration. A high end estimate would put it in the low 90's and a more conservative, low end estimate given the potential error would still be in the low 80's.



Is anyone familiar with how this might compare with other fast moving tornadoes? The record listed on Wikipedia goes to The Tristate with a forward speed of 73 mph. I assume this is over the duration or a significant portion of the track, and not a more momentary speed. It also has the caveat of a "significant tornado" alluding to smaller, more brief tornadoes going faster. I imagine it might not be unusual for brief, small tornadoes and rope outs to get caught in and carried by larger wind fields and pick up much faster forward speeds. This rope out, as far as I can tell, is the rope out of the EF4 that went through Pilger, the west/north most of the twins. So this was indeed a significant tornado, although in the rope out phase and the measured speed was more momentary than a longer average over its life.

I suspect the Pilger EF4, as it was occluding, moved into the RFD surge of the developing Wakefield EF4, where it was then ejected at great speed and then into the Wakefield circulation like a sling shot. It's really fascinating, almost mind blowing to think about. I hope more thorough studies are done on this storm.
 

Rob H

EF5
Mar 11, 2009
825
6
0
Twin Cities, MN
This is a) awesome b) something I'm glad I could contribute to and c) way more useful than building a tank and driving into a tornado "for science". Not a rip on Sean, or even Reed, but on some of the copycats that claim a lot of things in the name of science.
 
May 16, 2010
26
12
1
Skip, I was wondering about this as well. Specifically, I was curious if the power line towers could be used as markers to help confirm ground speed. If you could find where those towers existed on your map, and then use the time that the rope bisected each tower, you might be able to add another data point to your equation.

Also, is it worth using the lightning flash and subsequent thunder to help gauge distance to the rope?

If you need any higher resolution frame grabs, just let me know.

Phil
 
May 1, 2004
3,403
746
21
Springfield, IL
www.skip.cc
I was curious if the power line towers could be used as markers to help confirm ground speed. If you could find where those towers existed on your map, and then use the time that the rope bisected each tower, you might be able to add another data point to your equation.
The power line towers could definitely be used as reference points. I did use them from some of the angles and shots. I used trees and buildings in other shots. The trick is getting points that are consistent in everybody's shot. The tornado crossed those lines at more of a perpendicular angle, so it's difficult to use them for comparing positions at different times.

Also, is it worth using the lightning flash and subsequent thunder to help gauge distance to the rope?
That's an interesting idea. Not sure if you'd get a more or less accurate estimate though since this method would also introduce a whole array of different variables like atmospheric conditions, origin and duration of the thunder.

If you need any higher resolution frame grabs, just let me know.
Your 5k shot of that tornado would definitely make for a more accurate measurement. I'd love to use if we revisit this or for a future study on the storm. A video angle that varies more from the three used (all three were basically SE of the tornado) such as an angle that's W or N of the tornado might produce better triangulation. Shots that I could find from those angles were at a much greater distance or not continuous though.
 

Jeff Duda

EF6+, PhD
Staff member
Supporter
Oct 7, 2008
3,441
2,382
21
Broomfield, CO
www.meteor.iastate.edu
As far as tornado records go, I don't know if this can legitimately be classified as a separate tornado. Your video shows it's basically coming out of the same mesocyclone as the main tornado, just a ways above cloud base. That makes it seem really more like an El Reno situation in which there was a large tornado cyclone with individual tornadic vortices within it. This was just a case where there was more visual separation between a dominant circulation and a different tornadic circulation, but still within the same meso/TC. Therefore the propagation speed of the "satellite tornado" was probably more related to the rotational velocity of the mesocyclone than to the storm speed. You could probably arguably call that satellite tornado a suction vortex. I doubt you see that speed of propagation with a single tornado in a non-cyclic supercell. Still a cool project and accomplishment to get some data out there, though.
 
May 1, 2004
3,403
746
21
Springfield, IL
www.skip.cc
The damage paths and evolution of the fast moving rope tornado suggest it was the rope out of the Pilger EF4, a tornado which developed from a separate mesocyclone than the Wakefield EF4. I could be wrong on that as I haven't seen a solid shot of this tornado actually forming. It just emerges out of the rain in everyone's video, but it's same region of rain into which the Pilger EF4 disappears, and the damage path lines up, so it looks like it was a continuous circulation.

So in that context, would you suggest that the Pilger EF4 at some point just abruptly stops being a separate tornado and is simply the other tornado? Even while we have continuity of the condensation funnel and debris cloud when there still exists separation? I agree at some point the circulation is aggregated, but also say the tornado is operating more as a distinct satellite than a suction vortex since it's orbiting the parent mesocyclone and is physically separate from Wakefield's parent tornado cyclone, the cyclindrical area above the multiple vortex suction vortices. El Reno's orbiting vortices appeared to be shorter in duration visually and bound to the tornado cyclone.

A wider shot of the two tornadoes from earlier:
 

Rob H

EF5
Mar 11, 2009
825
6
0
Twin Cities, MN
It seems plausible:



This was just a case where there was more visual separation between a dominant circulation and a different tornadic circulation, but still within the same meso/TC.
The Pilger storm had multiple mesos and they traded off rather quickly at times. Can you be certain it was the same meso?
 

Jeff Duda

EF6+, PhD
Staff member
Supporter
Oct 7, 2008
3,441
2,382
21
Broomfield, CO
www.meteor.iastate.edu
That is new information to me. In that case, this seems even more a hybrid type of event where initially separate tornadoes from separate mesocyclones end up so close to each other it becomes difficult to distinguish them as separate tornadoes when one is dying. I was giving my impression based solely off of the portion of your video where you made the measurement of the speed of the tornado. At that point in the video, the satellite indeed appeared to emanate from the same mesocyclone as the current dominant tornado. Whether it started out that way does impact the classification. However, this new information also means that the Pilger tornado appears to have eventually been absorbed into the circulation of the next tornado, which makes a time of death of the Pilger tornado more difficult to classify (sounds like a kind of Hesston tornado situation in which the hand-off occurred so quickly and with the two tornadoes in such proximity so that the damage path masked the end of one and beginning of the other).

I would still argue that given the degree of instability and the scale of rotation going on, the two tornadoes were not truly independent at the moment you estimate the speed of the apparent satellite tornado. Moreover I think the excessive speed of that tornado was helped by the fact that it appeared to be at the most advantageous quadrant of the circulation such that the forward speed of the parent mesocyclone added some propagation speed to the tornado. Had you seen this same tornado survive and rotate around to the other side of the main tornado, I bet you would've seen it propagate much slower.
 
Dec 8, 2003
1,382
389
11
Southeast CO
www.youtube.com
I've been following along with this thread but I believed that in order for me to make an informed comment I would have to go review my video again and spend a bunch of time doing other research. I don't have that time available to me this week. What I will say, though, is that I filmed the rope out of the Pilger tornado, and I never got any impression that it was moving fast. In fact, I think it seemed to be nearly stationary to me from my vantage point a few miles south of Pilger. It just seemed to vanish, quickly, and on the vid I commented "Wow, that was a fast ropeout".

I am about 65-70% sure that "your" tornado was not the Pilger tornado.

Skip, I think you should just wait to see my vid, which (hopefully) I should be getting to you by this weekend.
 
Last edited by a moderator:

Rob H

EF5
Mar 11, 2009
825
6
0
Twin Cities, MN
It would be great if there was video of this from the west or even southwest.

In my mind the hypothesis is: the remnants of the Pilger meso/tc interacted with the Wakefield meso/tc, producing a short lived tornado that was ingested by the Wakefield circulation.

This hypothesis is dependent on a few key factors:
1) The storm in question had multiple simultaneous tcs. Visual observation and velocity scans support this with the presence of 2-3 distinct circulations at various times
2) That tornadoes from separate mesos can cross spatially (not necessarily at the same time). We don't even need to look at other events - Pilger 1 and 2 did this just several minutes before wakefield.

The emergence of the rope certainly lines up with the path the Pilger tornado took.

Bob, I'm wondering if the disparity you mention is because of an assumption that there was only one rope out. It's possible Pilger roped out once when wakefield stole it's inflow, then formed again once in the inflow again.

From my perspective the rope materialized and was fixed in position, then quickly slung into the Wakefield circulation. I've never seen a satellite act like that although I admit it's possible. The theory that the Pilger meso got one last breath of inflow, produced quickly from residual vorticity, then got sacked into the stronger Wakefield meso pulling the attendant tornado with it also seems possible however.

Calling it a separate tornado is dependent on verifying that it was a separate meso. L2 velocity overlaid on L3 CC might provide some insights here.

Apologies for the typos and grammar it's hard typing all of this on a phone.
 

Shane Adams

I know this really isn't the topic, but it's been raised so I'll have a go....

I don't understand why two separate tornadoes from the same meso/tc are considered one. I get the multiple vortex tornado, with several smaller vortices lasting just moments as they rotate (very close to) around the main tornado. But I don't see how the rope tornado in Skip's video can be considered part of the larger tornado, in any capacity. It has a separate debris field from a separate condensation funnel, and the ground contact is what classifies it as a tornado. That two tornadoes share a meso/tc doesn't mean anything. I get calling one tornado a "satellite" if the behavior fits, but satellites are still separate tornadoes.
 
May 1, 2004
3,403
746
21
Springfield, IL
www.skip.cc
It just seemed to vanish, quickly, and on the vid I commented "Wow, that was a fast ropeout".
I'd love to see the video, Bob. Most of the videos I've seen where people claim Pilger is dissipating, it looks to me more like it's just being hidden from view. Not that the tornado is actually gone, the observer just can't see it anymore because it becomes rain wrapped. That was the case for us, where the tornado looked like it just became completely rain wrapped. Even if the condensation funnel quickly dissolved I suspect the tornadic circulation persisted in that rain core. It's amazing to me how much difference a little rain makes in the visibility between two different chasers. A half mile, or a 30 degree difference in viewing angle, and a couple of rain curtains between the two observers and they have radically different views of the storm.

Calling it a separate tornado is dependent on verifying that it was a separate meso.
Do we need separate mesos to have separate tornadoes? I'd say a mesocyclone with a large tornado, that produces a satellite with a distinctly separate debris cloud/damage path, is a separate tornado. Such would be the case for this event too, even if that's not the Pilger EF4 and the Wakefield parent Meso produced a separate satellite. If they were from the same tornado cyclone, then yeah, the damage path and tornadic circulation would probably coincide. The point at which this satellite's circulation moves into the core circulation of the Wakefield EF4 is the point at which I'd say the two merged. That happens later than this sequence though. The debris cloud appears to make it at least around the east side of the Wakefield tornado, heading north, before the funnel and debris rope out to the point of dissipation.

It's one big, dynamic fluid mass though, not distinct objects we're dealing with here. That this rope out was under the influence of a larger wind field shouldn't discount or be a caveat in counting its forward speed. Tornadoes are always under the influence of the larger wind wind fields they are moving through, and the RFD and adjacent tornadic circulations are likely always influencing a tornado's forward speed.
 
May 1, 2004
3,403
746
21
Springfield, IL
www.skip.cc
I should add that there are some good examples of a violent tornado's condensation funnel rapidly dissolving, while leaving the wind field and debris cloud intact at the ground, only to then later have the condensation reform in a nearly stationary and stable form. The '98 Columbus "Crazy Farmer" tornado is a famous example off the top of my head.
 

Rob H

EF5
Mar 11, 2009
825
6
0
Twin Cities, MN
Do we need separate mesos to have separate tornadoes?
I would personally say no, but it seems ambiguous, and arguable. Perhaps "dependent" was too strong a word. If they are from separate mesos (which I believe is the case here) then it makes this conversation a lot easier, however!
 
Apr 23, 2010
146
14
6
The Guin, Alabama tornado was not only one of the worst twisters to hit my state, but one of the fastest as well:

" This was among the most intense tornadoes ever to hit Alabama...Moving at up to 75 mph, the funnel passed through a city lot in two seconds, carrying away everything on the lot." http://www.srh.noaa.gov/hun/?n=toryear1974
 

calvinkaskey

Guest
Feb 17, 2014
384
31
11
During the superoutbreak on April 27,2011 some of the cells and or tornadoes were moving 70 mph or more. Here is one that is you look at how much of the view it covers even before the camera is zoomed in it is about 75 percent or so of the horizontal view in only 51 seconds. https://www.youtube.com/watch?v=zEr2lCnQcow&list=WL&index=22


Given a 2 mile wide base for the mesocyclone [tornado along with stratus] and that it covered basically all the screen in the video and the tornado traveled about 75 percent of the screen width in 51 seconds the forward speed could be about 100 mph. This is dealing with the video I linked above. I'm listing this comment separately because the number seems kind of crazy.
 
Last edited by a moderator:
Very interesting thread, and some great insights/thoughts - as well as Skip's great initial post.

It begs wider thoughts on the way 'we' (as human beings) try to classify things around us, in this case, atmospheric vortices. Just because we want to compartmentalise a violently-rotating column of air associated with deep, moist convection, a 'tornado', is irrelevant to the way the atmosphere works.

There must be (at least) two key larger-scale processes governing the velocity of a tornadic circulation:

1) The forward speed of the parent supercell, and the associated low-level mesocyclone;

2) The velocity of the low-level wind field around the mesocyclone, and associated inflow/gusts fronts.

Number 1, to a greater or lesser degree, dictates the overall velocity of the tornado 'threat' (i.e. how quickly the area of interest/danger/whatever) moves towards a town, or person, or whatever. In some case, a tornado associated with this (especially a large one) may 'move' at a reasonably similar forward velocity.

Number 2, to a greater or lesser degree, dictates the velocity of the tornado at a very local level - for example, we know that sometimes tornadoes rotate around the parent mesocyclone, meaning the forward velocity is a product of both the mesocyclone's forward velocity, and its rotational velocity.

Thus, when we see reference to historic tornadoes having high forward velocities, I suspect in many cases this is due to a high forward velocity of the parent storm, because folks weren't generally looking in fine detail...I suspect many of these are not referring to tornadoes racing around in the mesocyclone (as in the case Skip has highlighted).
 
May 1, 2004
3,403
746
21
Springfield, IL
www.skip.cc
Bumping this thread because there's been new work and revelations on this event. My colleagues Hank Schyma and Anton Seimon revisited the case, and refined some of the earlier estimates. Josh VandenTop's video was included, along with my and Hank's own shots, and Rob Hurke's to produce a precisely measured and timestamped track of the fast moving rope. A summary of the major take-aways:
  • Maximum forward speed was measured at 94.6 mph over a duration of 5.33 seconds
  • Damage scars in aerial imagery indicated a continuous path identifying the fast moving tornado as the Pilger EF4
  • The track is mostly straight line in the minutes leading up to dissipation, not curving around the Wakefield EF4 as its satellite or subvortex, suggesting the vortex was independent of the larger developing tornado
The latter point was the most surprising finding to me, as I long assumed the fast moving rope was orbiting Wakefield, and that its remnants may have even merged into Wakefield. However, the measured track and a closer examination of extended footage show this not to be the case. The dissipating Pilger EF4 appears to track with, run into, and stall on the apex of the storm's RFD gust front. I suspect that an intense surge of RFD is the driving mechanism for the rapid acceleration of the Pilger EF4, and not that it became entrained in the Wakefield circulation.

The work is detailed along with other infamous fast movers in Hank's latest production: