Empirical thoughts about tornadogenesis

[Paul Knightley]

How about cold wet shower curtain moving towards you when you have a shower? Imagine this applied to an RFD!

http://www.wired.com/science/discoveries/news/2001/10/47334

 

[Matt Ziebell]

... Note that the vortex lines associated with the low-level cyclonic vortex arch across to the anti-cyclonic vortex at some height (2-3 km AGL?), and is disconnected from the mid-level mesocyclone entirely. There are aspects to this that seem to not blend well with other observations...

It certainly draws from the conceptual supercell schematic, but I do agree the line arching of the fourth vortex, or stage 4 (depending on the interpretation) *may* be overdone from their wind syntheses. When I see that arching I'm led to believe the vortex line has undergone such deformation by that time (along the apex/crest) that there now exists just two tangible vertical components of vorticity along the updraft.

I feel that the lack of phasing of the RFD vortex lines with the mesovortex would certainly be one reason for tornadogenesis failure. This may be the case in weak low level CAPE environments where the lines are just not tilted sufficiently despite a classic RFD and cloud base rotation, or more simply that the RFD is too dense and quickly propagates outward thereby undercutting the circulation. Just throwing some ideas out there, but wished I had the NCSA nearby!

 

[Dan Dawson]

I don't think of shear as horizontal, but vertical. I can see how vertical shear/spin works and happens. I just don't see how you'd get air to spin horizontally. Even if you have strong unidirectional speed shear, I can maybe see things trying to rotate horizontally, but I don't see it becoming organized rotation. Just like supercells and vertical spin, I can see that almost creating vertical rotation, but it's really more like vertical curl as it moves on. I think you need some other factor to get things past what the shear would sort of dictate and into good rotation.

It sounds like a neat idea to have this horizontal rotating air, that then gets tilted upward by a storm/updraft...I just don't buy that. I don't say there's no tendancy towards something rotating horizontally, but I have a hard time seeing an air mass really spinning horizontally.

If you can see how vertical shear/spin works, all you have to do is tilt this concept 90 degrees over on its side and you can see how horizontal spin works. It's no different, just knock it over on its side in your mind.

Also, one thing to keep in mind, and this is key to understanding this whole issue, is that vorticity does not necessarily imply the presence of a vortex! When we are talking about vertical shear of the horizontal wind leading to horizontal vorticity, it is vorticity associated with the shear (that is the gradient in wind speed/direction with height). On a supercell day, the environmental wind profile has this horizontal vorticity associated with it, but you are right in your observation that there aren't any horizontal vorticies or "spin". However, if that vorticity were to be tilted vertically and stretched, the shear would "concentrate" and start to actually spin about an axis, which is what happens in a mesocyclone. You can think of wind shear, whether horizontal or vertical, as being a sort of vortex "wannabe". It has the potential to become a true vortex, but needs to be intensified past a certain point before it "breaks down" and actually starts to rotate. Mathematically and physically however, it is still vorticity, present in the form of shear, rather than curvature of the flow.

(Incidentally, this is what also happens in KH waves. The shear between two layers becomes so strong that the interface breaks down into many vorticies. That vorticity was already present in the shear, it just got transferred and concentrated into the developing KH roll vorticies.)

Mesoscyclones are clearly vorticies, but there are not as intense as tornadoes, so they look a lot "looser" in appearance. Perhaps this is what you mean by your statement "Just like supercells and vertical spin, I can see that almost creating vertical rotation, but it's really more like vertical curl as it moves on.", if I understood you correctly. It is still vertical rotation, it is just vertical rotation at a lesser concentration than what you see in a tornado vortex.

 

[Mike Hollingshead]

If you can see how vertical shear/spin works, all you have to do is tilt this concept 90 degrees over on its side and you can see how horizontal spin works. It's no different, just knock it over on its side in your mind.

Must be harder in my head. I see a vertical object(storm) and how vertical shear could make that want to spin. It's seeing what makes the horizontal object(any air I guess) spin that is my problem. It has to be a good deal different I'd think. When I think of a good vertical shear profile I of course think of increasing speed and changing direction with height. I have a hard time matching that kind of setup in the horizontal...especially if one is talking pre-storm environment too. That's what I mean I guess.

 

[Danny Cheresnick]

Must be harder in my head. I see a vertical object(storm) and how vertical shear could make that want to spin. It's seeing what makes the horizontal object(any air I guess) spin that is my problem. It has to be a good deal different I'd think. When I think of a good vertical shear profile I of course think of increasing speed and changing direction with height. I have a hard time matching that kind of setup in the horizontal...especially if one is talking pre-storm environment too. That's what I mean I guess.

Mike, what do you mean by vertical shear when you say you see a vertical object and how vertical shear would make that want to spin?

When you see a wind profile, the shear you see is only the difference in speed and direction between adjacent layers in the atmosphere. If there are differences in the winds, they will cause a slight rotation to develop between the layers. Directional turning and speed shear creates a horizontal vorticity. This horizonal vorticity is not strong enough to create a tornado-strength horizonal vortex, but it is present over a much larger depth and width. Conceptually though, it is a broad slowly rotating horizonal tube. Thus, when a buoyant air parcel develops (which can be a storm), it rises upward through this vorticity. This causes the conceptual tube to be bent upward with the air parcel (essentially inside the updraft). By being bent upward, this horizontal vortex is turned into a vertical vortex within the updraft of the storm. This is how a supercell develops rotation in the updraft. Here the buoyancy (the CAPE) take ahold and stretchs this conceptual tube, which causes it to tighten up.

And to clarify, all of this is pre-storm. In fact, this horizontal vorticity is present on virtually every day. The difference is that it is rarely strong enough, even when tilted and stretched, to allow a supercell to form, and we don't always have updrafts when it is present.

 

[Jonathan Garner]

I think a review of Doswell's discussion on vorticity could benefit this thread:

http://www.cimms.ou.edu/~doswell/vorticity/vorticity_primer.html

I agree with Mike (and others) that there is alot of uncertainty regarding the role low-level vertical wind shear plays in the tornadogenesis process--but there is very little debate on the importance of horizontal vorticity (produced by the environmental vertical windshear) and the development of the mid-level mesocyclone. The whole process is reviewed nicely by Klemp (1987)...see link below:

http://twister.ou.edu/MM2007/Klemp87Review.pdf

Jonathan

 

[Mike Hollingshead]

All I'm saying is I can see turning with height helping a storm to turn. I have a harder time seeing speed shear causing much of anything to turn in the horizontal(since it's often not that great of change, especially if you aren't over a very wide/tall distance). Seems like a lot more chaos than turning in that orientation. Yeah it's really simple if you take any object or graphic and have an arrow going left to right at a given speed, then taking another one at a set level above that and have it going quicker....and see how there could be spin between the two. That's how that is often shown. I just run into visualization problems when I realize that "gap" is full of moving air as well, that seems like it would not want to spin all that much....at least not as organized as a taller object like a storm, in relation to the whole turning with height aspect. I guess areas in there would have to from time to time, but it seems hard to see how it would be very organized like one gradual turn in height over a big distance would be.

I also realize I'm a complete idiot when it comes to all this too and I would be better off shutting my hole on it, lol.

 

[Jeff Snyder]

A veering vertical wind profile (e.g SE winds at the sfc with SW winds aloft) creates only horizontal vorticity, just as vertical speed shear (e.g. 10kt SW winds at the sfc and 50kt SW winds at 700mb) creates only horizontal vorticity. In both cases, there is no ambient vertical vorticity. Appreciable vertical vorticity only arises when horizontal winds change abruptly over a short distance (e.g. an outflow boundary or other wind shift line, some cases for which vertical vorticity is generated by baroclinicity), or when the horizontal vorticity is tilted into the vertical.

Imagine that you point a fan straight upwards. On the edges of the air being blown upward by the fan, there is a gradient in vertical velocity -- outside the stream of air, there is negligible upward motion, while there is increasingly strong upward motion as you move farther into the "stream" of air being blown by the fan. It is this horizontal gradient in vertical motion that is the "tilting" term in the vertical vorticity equation. You can see this exact thing in practice if you stretch a piece of string or a piece of paper across the boundary of that stream (just half of it on the fan and the other half outside the fan). What happens? The piece of string embedded in the upward-moving air moves upward, while the part of the string that's outside the area of the fan doesn't move (after all, there fan isn't moving the air up there). So, the string tilts into the vertical. Now, imagine that this is a vortex line instead of a piece of string.

Technically, given homogeneous horizontal winds (i.e. winds are the sfc are southerly at 20 kts everywhere and winds at 700mb are southeasterly at 50kts everywhere), there is a vortex sheet (not necessarily individual vortices). There IS usually a very small amount of ambient vertical vorticity (e.g. near a mid-level trough axis, where winds change directions on a constant pressure surface). But, in scope of individual storms, it's unlikely that the ambient vertical vorticity has much of an effect on any particular storm. The tilting of horizontal vorticity into the vertical and the stretching of this new vertical vorticity can lead to a mesocyclone. The stretching in the updraft (which intensifies vertical vorticity) acts much like the typical "figure skater pulling her arms in" example -- conservation of angular momentum indicates that a narrowing of the radius of the vortex leads to an intensification of the rotation.


I guess I should note, for completeness sake, that a veering vertical wind profile does lead to a veering storm-relative wind profile, which produces streamwise vorticity. This streamwise vorticity indicates a correlation between the vertical vorticity and the vertical motion fields. A straight-line hodograph will produce non-deviant storm motion that lies on the hodograph, which in turn yields only crosswise vorticity. Storm-scale processes can lead to a storm split that can "push" storm motion off the hodograph, which then creates a veering (or backing) storm-relative wind profile. This, then, creates streamwise vorticity. For this conversation, though, I'll stay away from talking about streamwise vs. crosswise vorticity further.

 

[Glen Romine]

All I'm saying is I can see turning with height helping a storm to turn.

As Jeff noted (corrected) above, the veering wind profile doesn't lead to storm rotation, but has implications on how horizontal vorticity is tilted into the vertical.

I have a harder time seeing speed shear causing much of anything to turn in the horizontal(since it's often not that great of change, especially if you aren't over a very wide/tall distance).

This is largely true - though vertical shear can be large over small layers (in the real atmosphere, the vertical shear is typically not uniform from the surface to mid-levels, but is concentrated in the more stable layers), for the most part the shear is small - though still much, much larger than the background vertical vorticity, as Jeff again noted. You only need a small amount of the horizontal vorticity tilted upward. Then with a strong updraft the convergence near the surface helps bring together whatever vertical vorticity is there to be had. Then stretching of that vertical vorticity by the strengthening updraft with height (peak updraft is typically above 500 mb) brings vertical vorticity together to respectable values after some time (it can be a slow process). Still, if you look at typical air parcels going into the thunderstorm, they likely will only have a weak helical curl as they ascend through the updraft, not spinning around the mesocyclone several times. It is likely there are small concentrations of fairly intense rotation (not necessarily tornado strength, but still strong) which can have any number of orientations. A computer simulations of such things suggests as much.

... I just run into visualization problems when I realize that "gap" is full of moving air as well, that seems like it would not want to spin all that much....at least not as organized as a taller object like a storm, in relation to the whole turning with height aspect.

This is can tough stuff to grasp even for many meteorologists. If you are more comfortable with the idea of jet streaks, you can think of a small, compact jet streak immediately along the southern edge of the updraft (for a typical plains storm). Here, you would have vertical vorticity north of the jet streak axis (strongest wind in the jet streak core, with weaker winds to the north, so cyclonic vertical vorticity) - but not necessarily closed rotation like a pinwheel (which I think might be what you are expecting to see with vertical vorticity, but you only need differences in wind speed in the horizontal direction).

I also realize I'm a complete idiot when it comes to all this too and I would be better off shutting my hole on it, lol.

Maybe we all are incapable of explaining this well enough for you to get it, but maybe others will benefit from the discussion.

 

[Patrick Marsh]

Mike,

I'll give it another go for you this afternoon / evening. I'll put together a presentation showing how horizontal vorticity is created and how it is tilted into the vertical. I'll include the math for those who can follow. Maybe this will help?

-Patrick

 

[Gabe Garfield]

Regarding tornadogenesis itself, I have a feeling that the tilting (and subsequent stretching) of vorticity does (do) play a role. I can remember several cases of tornadogenesis in which the incipient tornado was initially oriented in the horizontal. This seems to suggest the possibility (if not the likelihood) that the tilting of horiziontal vorticity on the tornado scale is critical to tornadogenesis (rather than the simple stretching of existing vertical vorticity on the tornado scale).

Consider this tornadogenesis sequence near Putnam, OK on 3/20/06:

Gabe

 

[Rich Thompson]

Tornadogenesis appears to be quite complicated in reality, but the basic process can be simplified to stretching of vertical vorticity. So-called nonsupercell tornadoes may be the simple form of this process, where a swirl in the lower atmosphere becomes colocated with an updraft.

Tornadic supercells appear to produce their own vertical vorticity through internal storm processes. The RFD is the primary suspected source of near-ground vertical vorticity in proximity to the mesocyclone. Jeff and Matt have discussed the vortex line arguments, and a new paper on this will appear soon (www.ejssm.org).

Most of our forecast parameters (CAPE, 0-6 km shear, SRH, and LCL height) appear to be related more to the supercell than the tornado. It's my impression that strong near-ground SRH contributes to a stronger mesocyclone closer to the ground, and thus stronger pressure perturbations and associated vertical accelerations closer to the ground. Thus, stronger shear near the ground can contribute to stronger stretching closer to the ground - a condition that would favor rapid amplification of vertical vorticity.

The LCL height (or RH, temp-dewpoint spread, etc.) appears to be related to the thermodynamic characteristics of the RFD in the low levels. The most unusual aspect of seriously tornadic storms is a relatively buoyant *downdraft*. Deep convection removes instability, but the presence of a relatively warm/moist downdraft would seem to be a failure to stabilize the profile. Perhaps the tornado is a way to alleviate this uncommon inefficiency?

Rich T.

 

[Mike Hollingshead]

Thanks for trying to explain.

If you want to Patrick, sure, but please don't do it for just my behalf.

Gabe, what makes you think that tilting isn't mostly just outflow driven?

For what it is worth, I don't have a hard time imagining taking a horizontal rotation(I have a hard time seeing that happening....but if it was just there), tilting it via an updraft, and that turning into a tornado...or at the least that aiding in things. I have a hard time thinking it is that important and detrimental to all situations. But maybe it has to be. Sounds like it does.

 

[Gabe Garfield]

Gabe, what makes you think that tilting isn't mostly just outflow driven?

When a tornado is hit by outflow, it tends to move from a vertical orientation to a more horizontal orientation. In the pics above, the funnel cloud emerged from the wall cloud already in a horizontal orientation. Then, it was re-oriented into the vertical and "descended" until it became a tornado. This is the opposite of what you would expect from an outflow driven circulation.

Gabe

 

[Mike Hollingshead]

When a tornado is hit by outflow, it tends to move from a vertical orientation to a more horizontal orientation. In the pics above, the funnel cloud emerged from the wall cloud already in a horizontal orientation. Then, it was re-oriented into the vertical and "descended" until it became a tornado. This is the opposite of what you would expect from an outflow driven circulation.

Gabe

I could see there being outflow in place first, not allowing it to start vertical, then as that outflow relaxed....it was allowed to go vertical and descend.

Either case seems possible, I don't know how you'd go about proving either. Since one can see outflow cause a tornado to go horizontal and look just like that, I'd figure outflow was more likely the cause of its orientation in this case.

The first image looks like a "juicy" base, not cold looking at all. The base by the time that funnel is there slanted sideways however looks rather cold.

Was there much left of that storm following that tornado?