On 09/01/17 17:17, Eskimo Will wrote:
On 08/01/2017 22:02, John Hall wrote:
In message , Bernard Burton
writes
Never mind the medium range, today's forecasts in BBC R4 have been
pretty useless. Unless my hearing is faulty, the only mention of rain
on the 7.55 forecast was some drizzle over Wales. There was already
some precipitation in the SE showing on the radar then, and here we
had intrermittent slight drizzle all night, and this developed into
over an hour of moderate rain and drizzle here, and appeared to be
quite widespread on the radar. The models obviously missed this
development, and a lack of a decent covering of 'real' obs and a
dearth of radio soundings makes this type blunder only to be expected.
I noticed too, that the EGLL TAF has almost as many amends as there
were METARS!. On the 1255 forecast I heard a lot about how cold it
would be later in the week, but no mention or apology for the
downright duff forecasts for today.

The forecast for SE England currently shown on the Met Office website,
and time-stamped at 13:19 says:

This Evening and Tonight:
Often overcast skies tonight with mist and patches of fog. Light winds
at first will freshen later. It should be a dry night for most, but with
a few spots of drizzle in the wind. Minimum Temperature 5 °C.

I happened to be visiting a friend in SE London today, and there was
moderate rain there from at least 18:30 to 19:00. So it seems that
nobody at the Met Office noticed that their forecast was wrong till
later than 13:19, and even then they didn't bother to issue an updated
forecast. I know it's Sunday, and perhaps there aren't mainly people on
duty, but it's not very impressive.

TBH I don't think many care anymore. Morale is still rock-bottom and I
know of several who want to give up operational meteorology as the plan
is still full automation (apart from warnings) by 2020.


Nothing too new regarding automated forecasting. Work on
computer-generated shipping forecasts was begun at least forty years ago.

Also, when I started working on MOS forecasts almost forty years ago,
the problems the Canadian Met service had had with forecast verification
could have been associated with lowered morale or, perhaps, laziness. As
MOS output had improved, so forecasters' accuracy deteriorated.

When I was forecasting, there were a few forecasters I would have been
glad to have seen replaced by a machine. As an example, on one morning I
came on duty to find the night-duty forecaster has drawn a single pencil
line on the 0300 UK chart that was supposed to mark a warm front and
that was the sum total of his work on those charts. It struck me as an
unusual front in that it had no cloud on it whatsoever let alone any
sign of the rain he said it would bring in the afternoon. It came as
little surprise that the forecast he handed over to me bore no
relationship to reality and I was having to issue amendments before he'd
left the building. I think the 'afternoon rain' may have got him before
he got to his car at 0845.

In those days, some forty five years ago, it was well-known that TAFs
issued by LAP for their outstations were nigh on useless and that it
would take much effort to persuade them to correct a really bad TAF.
Automation couldn't come soon enough.


--
Graham P Davis, Bracknell, Berks. [Retd meteorologist/programmer]
Web-site: http://www.scarlet-jade.com/
There are more fools than knaves in the world, else the knaves would
not have enough to live upon. [Samuel Butler]