a few neat things about tornadoes

I just returned from the Severe Local Storms conference in Savannah, Georgia, put on by the American Meteorological Society. Lots of neat stuff there--derechos and bow echos, new radar and satellite techniques for exploring storms, advances in forecasting and model development--but I'm mainly there for the tornado research.

So, a few neat non-technical things that may be of interest:

1. The intensity and size of the mesocyclone (mid-level circulation that causes the familiar "hook echo" and results in a tornado warning being called on a storm) seems to have no statistical link whatsoever to whether or not a tornado will form, nor its size and strength. Nor is there any statistically demonstrable link between the size and intensity of the tornado itself. Nor is it true that the weaker the tornado, the more likely it is to happen--there seems to be a roughly Gaussian distribution of tornado intensities, centered on about the EF2. Nor is there much evidence that the tornado either starts from the ground and develops upward, nor does it start from the cloud base and descend downward. Basically the lowest kilometer of rotating air (from the mesocyclone) suddenly all at once contracts into a tight tornado vortex. At least, that's the way it looks now.

2. For awhile, we've been able to divide the tornado into a few different regions. There's the core, usually visible as a condensation funnel. There's the corner, where the core hits the ground. And there's the inflow, which is the area near the ground where the air flows directly into the tornado, with not much rotation. One thing that is becoming evident, as we are able to get measurements from close to the tornado, is that this inflow region is very shallow. Possibly under a few meters. Also, there's some neat work on simulating corner flow going on--that's another mystery. There's no way for us to see it with a radar--we're going to have to get a direct hit with the TIV or the StickNet or something.

I could go on and on, but I think that'll do for now. Next summer there's a huge field project that will allow us to collect all kinds of great data (weather permitting).

Originally posted by convect
I just returned from the Severe Local Storms conference in Savannah, Georgia, put on by the American Meteorological Society. Lots of neat stuff there--derechos and bow echos, new radar and satellite techniques for exploring storms, advances in forecasting and model development--but I'm mainly there for the tornado research.

So, a few neat non-technical t ...[text shortened]... roject that will allow us to collect all kinds of great data (weather permitting).

Can't you inject chaff into the tornado and track it then? Maybe the chaff would disappear too quickly, not sure. So you would get one blip then nothing?

Originally posted by sonhouse
Can't you inject chaff into the tornado and track it then? Maybe the chaff would disappear too quickly, not sure. So you would get one blip then nothing?

Well, as long as it's a part of the storm that can be hit by radar, the hydrometeors themselves give us the information we need about the three-dimensional wind fields. In fact, with these new polarimetric radars that are now being deployed we can get an idea of the size and shape of the hydrometeors themselves--distinguishing big drops from small ones from froze hailstones to melting ones--and we don't want other debris getting in the way.

The size and shapes of the drops in the rear flank downdraft (RFD) may turn out to be very important for understanding tornadogenesis. Some good but preliminary work has shown a slight skewing towards larger drop sizes in non-tornadic RFDs, which suggests evaporative cooling might be playing an important role. It's pretty well established that non-tornadic RFDs are substantially cooler than the updraft, while tornadic ones are not. But we don't know why this difference exists--what exactly the mechanisms are--among a whole lot of other mysteries.

Originally posted by convect
Well, as long as it's a part of the storm that can be hit by radar, the hydrometeors themselves give us the information we need about the three-dimensional wind fields. In fact, with these new polarimetric radars that are now being deployed we can get an idea of the size and shape of the hydrometeors themselves--distinguishing big drops from small ones from ...[text shortened]... is difference exists--what exactly the mechanisms are--among a whole lot of other mysteries.

A lot of subtle things going on to suss out for sure. What do you see as the main benefit of knowing all that, supposing it was totally figured out, what good would it do? Better prediction, my guess, but not something you could use to stop a tornado, I would think.

Originally posted by sonhouse
A lot of subtle things going on to suss out for sure. What do you see as the main benefit of knowing all that, supposing it was totally figured out, what good would it do? Better prediction, my guess, but not something you could use to stop a tornado, I would think.

Better prediction, mostly. There's just so much energy moving through a supercell that nothing short of a nuclear device could disrupt one already in progress. If we really truly knew everything about these storms--well, some interesting modelling work has shown that the course and intensity of a storm is affected by the soil and vegetation type. It's conceivable that some real smart engineer could figure out how to do some serious landscaping near cities in order to deflect or weaken approaching storms. But that's way far away in science fiction land.