On 28/05/2010 22:45, Andy M. wrote:
On May 28, 10:08 am, wrote:

Here's the present MetO forecast for 6-15 days:

UK Outlook for Tuesday 1 Jun 2010 to Thursday 10 Jun 2010:
Tuesday will probably still be rather unsettled with outbreaks of rain
in places, this locally heavy. It will then become increasingly drier
and brighter through Wednesday and Thursday as high pressure builds,
bringing settled conditions to all areas. Mist and fog patches are
possible around the coasts, whilst northwestern parts may remain on
the cloudier side with perhaps some rain, but elsewhere increasing
amounts of sunshine. Temperatures will be near or slightly above
normal initially before becoming warm and rather humid by Thursday.
Fine and warm weather is expected to continue into the weekend with
warm or very warm temperatures, ahead of a likely thundery breakdown
early the following week. There are then indications that the weather
will remain unsettled and cooler towards the end of the period.

Updated: 1224 on Thu 27 May 2010

Unfortunately, two-three weeks ago, the 15-30 day forecast was adamant
that we were headed for cool and wet weather at the start of June, for
aover a week.

I don't really monitor this, I've just remembered this as an example
(and I know full well that memory is a poor tool in that, but did
comment on it a lot at the time and it will be documented on here).
I'm not sure anyone bothers to monitor it and the MetO just allow what
they've said on the 15-30 day forecast to slip by without comment when
the weather is nothing like what they predicted. Of course there is no
seasonal forecast now, for good reason.

My point is; how useful is this to anyone. The MetO said they'd
introduced it because customer surveys showed they wanted it. Do they
really? Is it really of any use to anyone as it's accuracy seems to be
extremely suspect. Anyone with any other info about the monitoring,
I'd be interested.

Is it of any use to anyone?

They should just put the 30 day ensemble output online and let people
make up their own minds!

That's all very well for those who have the expertise to do this, but
only very few of us have. The general public believe that they pay
taxes for the Met Office's services and so they are entitled to an
interpretation of the data - ie, a forecast. After all, if most people
could interpret the data for themselves it would be bad news for the
forecasters!

As I said in a previous discussion on this theme, what is actually
needed is an honest appraisal of what the forecast models can and cannot
do. Once the variance in the ensembles reaches a certain point, the
output is effectively worthless and no meaningful forecast can be
honestly issued past that date.

This might mean that the Met Office would at times have to replace the
16-30 day section with an admission that the computer models cannot
predict the current situation out this far. Even the 6-15 day section
would at times have to have a "cut off" date.

Most people aren't stupid and would rather be told that a reliable
forecast can only be issued for X days than to be provided with an
unqualified forecast which turns out wrong. Unfortunately, what seems
to have happened in this case is that a fairly strong signal for a
settled spell of weather (judging from the confident wording of the
extended outlook) has suddenly evaporated. In fact, this is not that
much of a surprise as examination of the succession of forecast charts
issued on Wednesday and Thursday showed that the models were struggling
with the detail for this weekend so a month ahead was not looking good.

Everyone is now used to "cut off" dates for food and other perishable
items. Why cannot people be expected to understand that the nature of
the beast means that there must be a "cut off" date for weather
forecasts? In fact, I think most people already know this. It would
look better if the meteorologists could be humble enough to accept this
and be honest about forecast reliability.

--
- Yokel -

Yokel posts via a spam-trap account which is not read