[NileQueen]

http://www.nsf.gov/n....mc_id=USNSF_51

Have you all read about this?

FYI
Joanne

[Nigel Bolton]

Interesting read Joanne.

However, tornadoes remain sub-grid scale on many supercomputer model simulations, especially where the forecast domain is large, and therefore it is the ability of the forecaster to be able to read the atmosphere conceptually that really adds value to a dangerously developing stuation.

Post processing of individual storms or tornadogenetic fronts on a small domains here in the UK, have produced some excellent research and therefore an enhanced understanding of these phenomena, especially when a passing storm picks up a shed vortex off the local topography and stretches the vorticity into a tornado.

N.

[NileQueen]

Hi Nigel,

I think the shed vortex you mention might be called a wall cloud here.

Yes I don't think they have it all figured out yet.

Joanne :-)

[Bazmundo]

I think they're better off putting money into building techniques (why so square?) until they find that elusive "hardware" to deploy to take in-situ measurements. Then the serious number-crunching can begin.

I agree with Nigel about the post-process modelling in the UK, there are some fantastic results. For the US, the best new modelling techniques I've seen recently are the RR and HRRR grids, though they do confuse themselves once convection has begun like the RUC. Also, this (experimental) CIMSS "nearcast" model, that rolls off the previous hours of GOES data of two-layer theta-E difference (based upon PW and brigtness temp) and RUC winds to derive a low-mid level instability estimate for the evolving weather. It scored quite well in the last outbreak across Dixie.