On Sep 25, 8:47*am, "Will Hand" wrote:
"Dawlish" wrote in message

...
On Sep 24, 1:46 pm, Dawlish wrote:





On Sep 23, 7:20 pm, "Keith(Southend)"
wrote:

Seems to be consistency in all the models for a northerly plunge next
week as we enter the month of October.

September is turning out another very dry month here in Southend-on-Sea,
with currently only 8.0mm of rain (17% average), as it stands it will be
second driest after 1997 (7.6mm), 1985 (9.7mm, 3rd driest) temperatures
are also standing at 1.0°C above average.
--
Keith (Southend)http://www.southendweather.net
e-mail: kreh at southendweather dot net

There certainly *was* Keith!
I find the turn of events over this last week really interesting. 6
days ago, the models picked up the re-establishment of the high
pressure. The consistency and agreement was enough for me to forecast
at 10 days and barring something very strange, that forecast looks as
if it will be good. However, a short time after my forecast for the
29th, the ECM massively, then, to a lesser extent, the gfs showed a
northerly, which is likely not to occur. This "faux-northerly" stayed
on the charts for a couple of days and then over the last 36 hours has
been erased.

=====================

Paul, thank goodness it is not February, else you would see toys and tipped
over prams everywhere by now!
I would be sobbing into the keyboard and my sledge would be packed away
after wasting endless amounts of grease on the runners. *TWO board would
have gone into meltdown and then despair. Suicides would be on the cards and
grief councillors would be very busy :-)

+++++++++++++++++++++

If the high pressure does dominate on the 29th and I'd be 90% certain
of that at 4 days, my question is this: what happened to allow the
models (gfs in particular) to be correct at forecasting out to 10
days, but then during that same forecasting window, be so incorrect at
forecasting at a week's distance? It's bizarre!

=====================

I don't think it is bizarre just a finely balanced situation. Looking at the
charts at work ISTM that the northerly hinged on how a depression (shape,
warm air distribution etc) developed off the eastern sea board of the
States, which in turn relied on the behaviour of the Pacific jet, small
changes in the Pacific caused knock on effects downstream in the forecast
runs. Put simply - chaos in action! If a piece of information is missing to
all models due to lack of data then the models and ensembles may still show
consistency but all be wrong. This implies of course that in finely balanced
situations it is critical data that are important not consistency. usually
you don't know what that critical data is until after the event. One could
research this case for a long time (mine was just a cursory glance), but
there is no funding to do it in these cash-strapped days. Actually not all
models went for the northerly in this case, some internal MetO models will
turn out to be good in keeping anticyclone close to UK.

Keep up the good work!

Will
--- Hide quoted text -

- Show quoted text -

Probably absolutely true Will. Changes upstream during this last week
tipped the 2 models one way, then back the other. It does illustrate
just how difficult forecasting at 10 days is ATM. I have every
sympathy with the MetO forecasters who have to do this (even though I
joke that they precis the forecast in their tea break).

If this 10-day forecast comes off I'll be the first to admit that it
will be a fortunate one; it'll balance the other "unfortunate" ones
that head over the precipice after 24 hours! My forecast of an Easter
northerly with snow showers to the south coast, 2 years ago, still
haunts me. It fulfilled all my criteria, then 24 hours later, fell off
the edge of the cliff and the Easter weekend was fine and dry with
temperatures of 16C when I'd forecast snow showers. Ouch.