Graham Mills:

G'day MJO fans, (and Your Verbosity)

Reluctant as I am to launch into public e-speaking, Richard Dare of
BMRC has just shown me some results which I think are absolutely
oustanding!

Some of you may recall a hailstorm up the NSW coast last January
16. Apart from some quite spectacular video footage shown on
national TV, there were two LAPS deficiencies brought up in this
discussion page:

(1)     the Severe Thunderstorm Checklist guidance showed potentially
severe thunderstorms in the area, but did not show any significant
SREH, and so did not identify potential supercell areas.

(2)  it was pointed out that there was a spectacular easterly LLJ
"feeding" the storm, and that LAPS had failed to represent this
feature.

I asked a general question "did the easterly jet drive the storm. or
vice-versa?", but got no reply.

Richard has re-run the case incorporating his Bulk Explicit
Microphysics representation of moist processes. His charts show:

(1)     A blob of intense hail potential on the NSW coast at 28-30S at
12Z 6 Jan.

(2)  An easterly "jet" of around 30-35knots (against LAPS 10-
15knots) driving onshore near Tweed Heads at 12Z

(3) A not surprisingly much greater SREH through the Checklist -
supercell environments are now diagnosed.
 

It goes on and on, but points to a reasonable answer to my chicken-
and-egg question, and also makes me look forward to when this
physics is in the operational suite!

WELL DONE RICHARD!!!!

Graham

Harald Richter

I am amazed (and shouldn't be after years of modelling) that a change to a
model's microphysics package can create a 30-35 kt jet out of a 10-15 kt jet.
To understand why it did so, I am wondering: (1) was the change to the
microphysics the only difference between the "old" and "new" LAPS, and (2) what are the main
microphysical differences between "old" and "new?"

BTW,  are we talking about a decent 850 hPa jet advecting decent TM moisture
onshore?

Cheers, Har*ld

Rob Webb

Thanks Graham

Certainly interesting stuff. Was 12Z used for a reason and have you looked
at the output earlier in the day? I'd be interested to see some of the
afternoon info.

Given SREH is dependant on the movement of the cells, we may need to be
spending some more time sharpening our skills in picking which way these
cells are going to go or maybe obtain access to the predicted values which
are used within the decision tree. This may allow us to 'ground-truth'
things more effectively and anticipate problem areas within this part of the
decision tree output. Did you look into how well the predicted motion of the
cell in question matched the actual movement?

I must admit it was one of those many occasions where even after it happened
I was scratching my head a bit re the chicken and egg thing. Any of these
improvements that will allow us to understand the environment better,
earlier, will mean better warning decisions or at least help us anticipate
the range of possibilities a little better and react more quickly.

This may well be one of those areas we don't use the models smartly enough
operationally i.e. using them not only to tell us what will happen but also
why.

Thanks

Robb Webb