Gedday,
There were all sorts of interesting aspects of the sequence over the weekend into Monday-Tuesday, the highlite of which was probably the wind storm over Victoria. On my web page I have placed a few maptool plots, showing the expanse of high winds.
There
were 20 - 30 kt sustained surface winds over most of the state from daybreak
on Sunday (0000 UTC 15th).
By the afternoon (06
UTC) they were at 25 to 40 kts across the State and by 10 at night
(12UTC) they were 30 to 40 kts
everywhere. This kept up all night until the front went through central
victoria at around 21 UTC
I was up at the VRO chart
discussion yesterday (Tuesday) morning; and Ken Dickinson had a photograph
of the anemograph trace for Mount Bulla, showing a peak gust of 52 m/s,
which must be around 190 or so km/hr.... it was darn windy up there.
I actually can't recall a situation at any time of year where we have had
such a sustained period of greater strong winds in Melbourne..... any comments
on how unusual the situation was?
A short summary of the sequence is on my web-page in the form of the 00 UTC surface map for each day from Friday morning (13th) through Monday morning (16th). You can see that the key event was the cyclogenesis of a surface closed- low just southwest of WA and its subsequent movement across the Bight. The low level pressure gradient over South East Oz was totally driven by this, with the pressures over eastern Victoria hardly changing as the winds developed (say between 00 on the 14th and 00 15th)
Now, for the interesting aspects of the event:
a) IPV maps.
I have never got much out of isentropic Potential Vorticity maps. Every academic university type who has ever wandered through has urged that we naive, non-mathematical weather Bureau types should be looking at the potential vorticity, and they go off sniggering to their colleagues... ah..those weather bureau guys don't have much insight... snigger, you'll never believe this... snigger... they don't even look at the potential vorticity.
So... we followed their advice and Jim Fraser and the NMOC team dutifully put hemispheric IPV charts up on the web, and we looked and we looked and we looked and we never learned very much.
One of the things we were looking for was wave-breaking on the tropopause. This idea seems to have come from the famous Hoskins, McIntyre and Robertson 1985 Quarterly Journal paper. In that I (at least) saw for the first time polar stereographic maps of IPV showing the mass of cyclonic IPV centred on the pole and representing the polar vortex. The edge of the mass is the tropopause where it intersects that isentropic surface; and cyclone developments show up as waves curling over and filaments breaking off (wave-breaking). There is a review of this, with some interesting looking diagrams in a recent Bulletin of the American Met Society (Neilsen-Gammon, BAMS, June 2001); I just sent it off to the printer; but haven't actually read it yet.
As I said, we have looked and looked and never seen anything much. Part of the problem is that we don't get very many clean and clear-cut cases of mid-latitude development occurring in our longitudes.
Well.... as shown on the mslp sequence on my web-page, the development of a 989 mb surface low near WA at 00 UTC on Saturday (14th) was as clear a development as you'll ever see...... So.... I went and looked at the IPV sequence from the NMOC page.... and lo and behold... we get a snake of cyclonic IPV poking out of the high latitudes towards south-eastern WA. On my web-page I have placed the IPV sequences at both 320K and 350K for the last few days. You can see the stratospheric/cyclonic air undergo a change of shape so that there is a "trough" poking up into Oz longitudes and a ridge poking down on either side; and this wavy-trough corresponds to the development. I haven't read the papers (e.g. Nielsen-Gammon referenced above) at any length so any interpretation on my part will be naive. Unfazed, however, I make two comments:
a) At 320K the sequence over southern WA, spectacular as it is, is little different to the events downstream over the Coral Sea, where we did not get any surface (or even much of an upper level) development.
b) At 350K (around 200 hPa) the WA region DOES stand out as being the major development, and the deformation of the tropopause is quite spectacular. An interesting aspect, though, is that the real action occurred between about 12UTC 15th and 12UTC 16th, which is actually AFTER the surface and mid-tropospheric development of the closed low had taken place.
320K
sequence
350K Sequence
b) Severe Weather:
At NMOC chart discussion
on Monday we looked at the coldies diagnostics for that afternoon and there
were "cold-season tornadoes" forecast for that afternoon for western
NSW, which when you look at the progs would have been in the warm air,
well ahead of the deep cold pool. Did these eventuate? Any
reports of coldies or anything similar? Synoptically, and in terms
of low-level CAPE, does anyone understand why there would be coldies in
the pre-frontal air?
c) Predictability.
In general terms this seems to have been a highly predictable event. We were talking about it on synoptic_discussion from the preceding Monday. Watching it through the week (last week), all models (EC, GASP, US, JMA etc) were getting the same sequence, though with slight differences in timing and structure. In contrast, if you look at the current set of progs, all the models diverge at about day four. Clearly some sequences are more predictable than others.
I had a look at the ensemble
spaghetti diagrams for the event on Michael Naughton's great archive (http://gale.ho.bom.gov.au/bm/internal/daas/ensemble2/mv_spaghetti/index.html)
I'll have to look
at it more carefully, but from a first glance the high predictability (all
ensembles showing the same behavior) doesn't stand out. This is possibly
a function of the choice of the contour that the spaghettis is based on.
My feeling though is that at the current state of the science, consistency
between the models is a better guide to predictability than is spread of
the ensemble for an individual model. I get into trouble from the
ensembles boys (the assimilation mob) whenever I speak on these matters...
but perhaps they have some comment??
Also, have they/we considered spaghetti diagrams and such based on the ensemble of different models that re available? I realise this would require a bit of work to set up. Since I am exposing myself to the ensemble hitmen, I may as well go on.... It seems intuitive that a high amplitude/spectacular situation involving tropopause folds and surf zones would be more predictable than the weaker , more wishy-washy structure prognosed for the next few days. Does any one know of any work quantifying this?
d) Summer fronts and winter fronts:
A theme of mine has been
the differences between summertime and winter-time cold fronts in our region.
As we discussed last summer, the summertime fronts have much more theta
structure, whereas the winter fronts seem to have more thetae structure.
This and the previous major front of a week earlier, reminded me of summertime
or spring time fronts, especially with the roaring northerlies ahead of
them. If there is anyone still reading this far down in the message,
I am sure they are getting tired, so I will save you from the analysis
of the theta and thetae diagrams. However, an interesting aspect
of the the weekend's front is the way it continued across in to Bass
and so we had a stretch of southerlies behind it passing up the New-South
Wales coast (see the meso-laps sequence on my web-page)
This is very similar to
the large scale environment for a Southerly-Buster, as modelled and studied
by Kathy McInnes and myself a decade ago (Monthly Weather Review, July
1993). Was there any Southerly-Buster behaviour observed with this
front???
, 
Samantha Taylor
From: John McBride <jmb@BoM.GOV.AU>
To: synoptic_discussion@BoM.GOV.AU
Subject: [synoptic_discussion]
Model Ensembles (fwd)
>From Sam Taylor... very slighted edited.. to protect the guilty...
Have a read of this... her multi-model ensemble display is terrific!!!
JMcB
---------- Forwarded message
----------
Date: Wed, 18 Sep 2002 16:33:28
+1000
From: Samantha Taylor <Samantha.Taylor@BoM.GOV.AU>
To: J.Mcbride@BoM.GOV.AU
Subject: Model Ensembles
Dear John,
Jeff Callaghan forwarded
on to me your discussion of the recent cold front events that have affected
such a vast area of southern and south eastern Australia over the past
few days as we have been watching this event here
ourselves (with little else
happening in our part of the world at the moment). Anyway, in this discussion,
you make the following comment about model ensembles:
>Also, have they/we considered
spaghetti diagrams and such based on the
>ensemble of different models
that re available? I realise this would
>require a bit of work to
set up.
Well, today must be your
lucky day because about 2 weeks ago I created most of a McIDAS based web
page (I can programme in McIDAS, but the web is another matter, so our
SMSD, Matt Saunderson, finished it off for me) here
in Queensland that plots
the 540, 560 and 580 dm 500 hPa geopotential height lines. (The numbers
were chosen to meet Queensland requirements.) It does this for all of the
EC, UK, US, LAPS, GASP and JMA and they are all on
the one diagram. You can
check these out at:
http://serva.qld.bom.gov.au/~mcuser/images/index.html
The ensembles are the first
options available in the menu on the left hand side. You can click on any
of the days (say the analysis), then scroll up or down to see the rest
or you can click from the menu. The current 12Z
runs are up and running
as well as the previous 12Z run. The idea to put the 00Z run there was
not mine and not part of my original McIDAS script, so I am not sure if
that is running properly or not as yet. (At any rate, I
do not look at it because
there is no 00Z run of the EC and that is by far our favoured model in
this part of the world.)
[cut]
And, as expected, the point
at which the models diverge does vary, sometimes they stick together for
a few days, other times they can't even agree on the analysis (eg yesterday's
run of the UK, which certainly
had its problems, compared
with the rest of the models).
[cut]
It is the only ensemble
that I am aware of and was actually created because one of our SPOCs, Manfred
Greitschus, liked the spaghetti page you refer to so much, he wanted one
that compared all the models rather than
various runs of just the
one model.
Another one of our bits of
web guidance that you might find of interest is an experimental product
in warm air advection that seems to be useful for thunderstorms. It can
be found on the same McIDAS web page menu bar as
referenced above, but under
the EC heading, then choose Precip Guide and one of the 8 times on offer.
Basically, it is an attempt to reduce the graphic on the left hand side
of the screen to a number, and if you are
technically minded, all
we did to get a number quantifying the warm air advection was take the
shear in the 850-500 hPa layer (simple vector subtraction using the EC
model output values of the wind at 850 hPa and 500
hPa) as a vector and dotted
it with the 700 hPa wind vector in order to maximise the usage of the available
data, though whether is necessarily scientifically perfect is another matter.
Once again, it has a Queensland
perspective on the world,
though large slabs of New South Wales are visible. Roughly speaking, a
value of 6 and some moisture (i.e. in an area within the >=70% RH at 700
hPa) seems to be needed for warm air
thunderstorms. Coldies are
a different matter - we get so few of them up here that we really don't
have any sort of feel for whether their needs are adequately expressed
on this diagram or not and if so, what sorts of
numbers might be worth watching
for.
If you are interested, keep
an eye to this web page, as I have plans for producing a couple of more
additions in the not too distant future, including an oceanic extension
to the warm air advection plots for the
cyclone season (not to mention
Coral Sea/northern Tasman cyclogenesis) and also an areal plot of storm
relative helicity values (for both left and right moving thunderstorms)
derived from LAPS data. Depending on what
Manfred thinks of his spaghetti
diagrams, there may be more of these too.
Samantha Taylor
Trainee Meteorologist
Severe Weather Section
Bureau of Meteorology, Brisbane.
Andrew
Burton
At 02:03 AM 18/09/2002 +0000,
John McBride wrote:
>b) Severe Weather:
>
>At NMOC chart discussion
on Monday we looked at the coldies diagnostics
>for that afternoon and
there were "cold-season tornadoes" forecast for
>that afternoon for western
NSW, which when you look at the progs would
>have been in the warm air,
well ahead of the deep cold pool. Did these
>eventuate? Any reports
of coldies or anything similar? Synoptically,
>and in terms of low-level
CAPE, does anyone understand why there would be
>coldies in the pre-frontal
air?
John - if you consider what
the coldies diagnostic represents it becomes
apparent that the diagnostic,
although named "coldies" doesn't care whether
the air is pre or post frontal
("warm" or "cold" air). We often see the
strongest coldies signal
in the prefrontal airmass - (and we often issue
warnings for "coldies" in
"warm" ie. prefrontal air). The coldies signal
represents the coincidence
of instability (<-1 700hPa LI for strong
signal), upmotion (0.9500
sigma <-15x10-3 s-1 or 700hPa <-15hPa hr-1) and
low level shear (850hPa
- 10m > 11ms-1). The salient characteristic
of most "coldies"
producing environments is an atmosphere characterised by
relatively low CAPE together
with high shear in the low levels. These occur
most often in a winter environment
- but I don't think the atmosphere
particularly cares about
"warm" or "cold" air (forgive my
anthropomorphising).
Often the shear is stronger ahead of the front.
Regards
---------------------------------------------------------------------
Andrew Burton
Regional Manager Severe
Weather Section
Bureau of Meteorology, Perth, Western Australia
John McBride
Thanks.... I wasn't
really concerned whether the coldies occurred in cold
air or warm air... I was
wondering more why there should be much
instability in that
situation..... You have partly answered it by
pointing out that
they usually ocur with "relatively low CAPE"
John Mcb
Elly Spark
Re: Coldies in NSW on Monday.
We did have thunderstorms in the Upper Western
as evidenced by GPATS. No
reports that I know of to confirm tornadoes or
otherwise.
Elly Spark, NSW