Martin
Sorry - can't remember the exact dates. It was probably mid June 1987. The
weather had been clear and settled for the previous week days before we
began thinking about the run home.

I am aware of the Met Office's tendency to forecast based on the poorest
conditions expected in a sea area, and given the lack of time, can see why.

However, in this case a dive boat went down to Rockall and found perfect
conditions for diving, while another yacht had 'enjoyed' (but not sailed)
fine weather in the North of the Minch and over to Orkney, so the settled
conditions seem to have been fairly widespread.
Typical weather conditions at that time of year around NW Scotland tend to
feature lots of (sometimes quite strong) Easterlies, but not that time.

As I mentioned earlier, predicting the behaviour of a stable high pressure
system seemed (on the face of it) to be a less daunting undertaking than
second guessing chaotic events along the polar front (at least to a user),
I was curious to to find out if the underlying modelling process for high
pressure systems had/has a known (or suspected) weakness under some
conditions and keep that in mind when considering how to respond to
forecasts in future.

Many Thanks

Iain



"Martin Rowley" m wrote in
message ...

"Iain Mackay" no_one@here wrote in message
...
Martin
Thanks for your reply.

Do the difficulties you refer to also apply to predictions on a larger
scale (as in running from St Kilda in flat calm when we should have seen
strong winds?). What is the reason(s) for the model failure (If that is
the right word).

... You would need to consider each situation individually, but the
current generation of mesoscale models (on which coastal forecasts would
be based), whilst not perfect, would not explain why you experienced calm
when strong winds are expected. Within the Minches for example, I can
think of examples where, with some wind directions, you might get much
lighter winds than 'generally' forecast, but not on a voyage to/from St.
Kilda which it much more 'open'.

Even several years ago, the model output was pretty good for such
situations - the only problem I can think of offhand is that St. Kilda is
part of 'Hebrides' sea area (in the Shipping Forecast), and a F6/7 *might*
have been forecast for the far west and the peculiar spike between Bailey
and Faeroes, but would not be applicable to the rest of Hebrides. This
was/is a common problem with the larger sea areas - sometimes the
forecaster runs out of words to split the area sufficiently and has to
concentrate on the 'worst-case' scenario. When writing the Shipping
Forecast, by the time you get to Malin, Hebrides etc., you find you're
rapidly running out of words (strictly limited) and have to lump areas
together with no leeway to sub-divide. It's only a guess, but I've known
it happen.

Can you remember the dates? It is possible to look at archive mslp charts
now online and get an idea what sort of situation was in place.

Martin.


--
FAQ & Glossary for uk.sci.weather at:-
http://homepage.ntlworld.com/booty.weather/uswfaqfr.htm
and
http://booty.org.uk/booty.weather/metindex.htm