In article ,
John Hall writes:
In article ,
Dawlish writes:
As an extra: the MetO have not picking up on this in their 6-15 day
précis, though to be fair, this was last updated yesterday morning.
It shows how quickly the models can change. Why the MetO
doesn't update this more frequently has always been a mystery to
me.

Presumably they don't see any need to update a forecast of something
that's at least 6 days away any more frequently than once a day. The
longer the range of a forecast, the less frequently it needs to be
updated, but of course it's open to debate where the line should be
drawn.

Having just looked at the UKMO site, I was surprised to see that the
6-15 day forecast was updated at 4am today rather than at 11am-noon,
which I thought was the time of day that it was normally updated. That
means that it will have been produced without benefit of the 0Z model
output. I can understand the forecast only being updated once a day -
after all they don't have unlimited resources. but one would think that,
to tie in with the latest model runs, around either midnight or noon
would be the natural time to do it - probably noon, as if it was at
midnight most people wouldn't see it for at least another eight hours.
Since the 0Z model output becomes available between about 04:30 (GFS)
and 08:30 (ECMWF) - with the UKMO's own model output somewhere in
between those times - that would allow three hours or so for thorough
analysis of all the operational runs and the ensembles before
formulating the forecast.
--
John Hall "He crams with cans of poisoned meat
The subjects of the King,
And when they die by thousands G.K.Chesterton:
Why, he laughs like anything." from "Song Against Grocers"