| |
You are here: Experts > Science > Weather > Meteorology (Weather) > Tornado Formation
Expert: Ryan Hastings - 11/2/2009
Question
QUESTION: Ryan, I know that tornado formation is not thoroughly understood, but for decades, the most popular theory involves horizontal tubes created by wind shear being reoriented vertically by convective parcels that strike them from below. I have always thought this to be an unlikely scenario. Since this is closer to your field of expertise, maybe you can shed some light on it or perhaps refer me to more recent hypotheses. By the way, my specialty is climatology, especially as related to landform development, but as a professor teaching introductory meteorology, I need to understand a wide range of topics. I also have some comments concerning your last response about Hadley Cells (I did not see a follow up option this time on the web site). However, for the sake of those reading this, because they are interested in the tornado question, I’ll add my other comments at the end. For anyone interested in tornadoes, however, I think you would find the attached photo and explanation interesting even though it is not related the question above.
There are two reasons I find the above theory unlikely. One is that I have watched hundreds of tornado clips and have never seen one that looked like a horizontal tube being reoriented as the theory would suggest. Such a tube might be invisible at first, but as it begin to be repositioned vertically, the resulting condensation (at least for a high precipitation tornado) should show the movement. Ironically, I have seen the opposite – two funnel clouds forming in close proximity whose bottom portions reoriented to a horizontal position and eventually joined each other!
Secondly, it seems highly unlikely to me that a rising parcel or air could have much of an effect on a spinning horizontal tube considering the angular momentum of the latter. I guess a very powerful updraft could move a very weak horizontal tube, but in that case I would envision it just blasting through and disrupting it, not gradually reorienting it to a vertical position. I suspect you have a better math/physics background than I, so maybe what I see as intuitively unlikely seems reasonable to you.
Back to the Hadley Cell. Thanks for your second comment, I will definitely check out the Valiis book (looks like a good addition to my permanent library). As, for my comment about convective clouds of any size being visible with IR because their tops would be near the upper reaches of the troposphere, that was a mistake on my part. I meant to say convective “storms”, not clouds.
Finally, I am attaching a photo of a shelf cloud that I think you will find interesting considering your specialty. I took this photo at the airport in Dead Horse, Alaska. Dead Horse services the North Slope oil fields and is located on extremely flat land on the coast of the Beaufort Sea at a latitude of 70 ºN. To witness this at such an extreme latitude was amazing to me. I asked several of the people who worked there if they had ever seen anything like it, to which they answered “never around here” The storm dropped about an inch of rain (about 10% of that region's annual ppt). I saw no funnel clouds and no hail.
IMAGE: Shelf Cloud
ANSWER: First off, the tilting of ambient vorticity is actually the origin for mid-level rotation in supercells, not for tornadoes. The origin of low-level rotation and tornadogenesis itself is not clear. Some of the vorticity is transported from upper levels by the downdraft. One of the leading theories right now is that vorticity is generated baroclinically at the interface between the cold downdraft and the warmer surrounding air, and these vortex rings are transported downward and tilted (see: http://www.ejssm.org/ojs/index.php/ejssm/article/view/32/34). But yes, tilting is believed to play a role.
It's not just the updraft, but the presence of a horizontal gradient of vertical wind that causes tilting. http://en.wikipedia.org/wiki/Vorticity_equation gives a derivation. Conceptually, maybe think of it more like a solid rotating body. The axis of rotation is what is being tilted.
Also, when the tornado itself is forming, the vorticity is already vertically oriented. The tilting has already happened, so you won't see it; instead you're seeing the concentration of vorticity and its stretching by the updraft. When a tornado turns into the horizontal, it's usually because it is outflow-dominated and roping out. But it's the same principle of vortex tilting--the downdraft is tilting the vertical vorticity into the horizontal.
Nice shelf cloud, by the way!
---------- FOLLOW-UP ----------
QUESTION: Thanks for the clarification, esp. that "tilting of ambient vorticity is actually the origin for mid-level rotation in supercells, not for tornadoes" Can I assume that tornadoes get their rotation from the parent mesocyclone? Also, surely the Coriolis effect must play a fairly significant role considering how long these cells can last and how large they can grow horizontally. Isn't the fact that the vast majority of tornadoes rotate counterclockwise (in the N.H.) the result of the coriolis force? If it were just the tilting, wouldn't the mesocyclone be just as likely to rotate clockwise?
Answer
I think it's safe to say tornadoes generally get their rotation from the parent mesocyclone; even if the vorticity originates elsewhere, like on the rear flank downdraft, it will still be drawn into the mesocyclone, and essentially what a tornado is, is the near-ground circulation of the mesocyclone being concentrated into a narrow region. However, there's a bit of a caveat below.
The Coriolis effect is neglected when studying supercell dynamics. While they do last long compared to ordinary cells, that is still usually only a few hours. They would need to be going for more like six hours and longer. I know that the Coriolis force has to be considered for bow echoes, and is one of the big reasons why the cyclonic (northern) bookend vortex for a bow echo is usually much stronger than the anticyclonic (southern) one.
The reason most supercells and the tornadoes they spawn rotate cyclonically is actually because of the wind profile. It's actually a bit too long to go into here, since the explanation is about a lecture or two's worth of material, but the basic idea has to do with pressure fluctuations caused by the interaction of the environmental wind shear with the horizontal gradient of vertical velocity (linear dynamic pressure perturbation) and by the fact that any rotation, whether cyclonic or anticyclonic, also causes pressure to lower (one of the nonlinear dynamic pressure perturbation terms). If the shear is unidirectional (straight-line hodograph), then the storms will split into two equally strong supercells, with an anticyclonic one that propagates to the left of the mean wind direction, and a cyclonic one that propagates to the right. In fact, on VORTEX2 we were very fortunate to intercept a long-lived and well-organized anticyclonic one--an event far rarer than even an anticyclonic tornado (which is about one in every ten tornadoes). If, however, the wind curves with height, then the pattern of pressure gradients resulting from the pattern of pressure fluctuations caused from the linear and nonlinear dynamic pressure perturbations will tend to suppress the development of one and favor the other. Clockwise hodographs favor right-moving counterclockwise-rotating storms, and it just so happens that the meteorology of the Great Plains tends to result in clockwise hodographs more frequently than counterclockwise (or even straight-line) ones: favorable storm environments are those with southeasterlies at the lower levels, bringing in Gulf moisture, and turning to southwesterlies above, bringing dry air from the desert. This is actually why most mesocyclones are cyclonic.
Also, note that cyclonic mesocyclones *can* produce anticyclonic tornadoes. This is most likely because of the vortex tilting that probably creates the tornado in the first place, which produces characteristic cyclonic-anticyclonic vortex couplets. (We got some great data on that with the Goshen, Wyoming tornado last June!) There's almost always an anticyclonic circulation south of the tornadic circulation along the rear flank gust front. On rare occasions--when for some reason the vertical velocity is sufficient--the anticyclonic circulation can also become a full-blown tornado.
Add to this Answer Ask a Question
|
|
){kind=link}