I know it's all been said before but you can see why the Met Office doesn't
seriously forecast to the public more than 5 days in advance. I don't know
if it costs any more for the GFS model to be run past 7 days but if it does
it is a waste of money. There can be little value in a model that can
oscillate member runs by 14C (850 hPa values) every six hours other than to
conclude "anything can happen" -because we know that don't we!

Dave

Dave.C wrote:
I know it's all been said before but you can see why the Met Office doesn't
seriously forecast to the public more than 5 days in advance. I don't know
if it costs any more for the GFS model to be run past 7 days but if it does
it is a waste of money. There can be little value in a model that can
oscillate member runs by 14C (850 hPa values) every six hours other than to
conclude "anything can happen" -because we know that don't we!

Dave

But it is for reasons of hinting at coming trends that they run these
models - I'd be interested to see the statistics on how correct the GFS
has been for anything past plus240z over the years.

You'd think we'd all learned our lessons by now about the rogue GFS -
but like lemmings we go back time and again...

In article .com,
Scott W writes:
But it is for reasons of hinting at coming trends that they run these
models - I'd be interested to see the statistics on how correct the GFS
has been for anything past plus240z over the years.

You'd think we'd all learned our lessons by now about the rogue GFS -
but like lemmings we go back time and again...

Yep. And for looking for those hints about coming trends it's much
better to look at the ensembles, such as

http://217.160.75.104/pics/MT8_London_ens.png

rather than just the operational run.
--
John Hall
"Honest criticism is hard to take,
particularly from a relative, a friend,
an acquaintance, or a stranger." Franklin P Jones

Scott W wrote:

But it is for reasons of hinting at coming trends that they run these
models - I'd be interested to see the statistics on how correct the GFS
has been for anything past plus240z over the years.

Are there any data on the accuracy of each 6-hour/12-hour output from
the models?

--
Jonathan Stott
Canterbury Weather: http://www.canterburyweather.co.uk/
Reverse my e-mail address to reply by e-mail

"Jonathan Stott" wrote in message
...

Are there any data on the accuracy of each 6-hour/12-hour output from
the models?


I refer to this http://wwwt.emc.ncep.noaa.gov/gmb/ST.../aczhist6.html
occasionally to gauge how the models are performing

Originally posted by Martin.

Jon.

In article ,
Jonathan Stott writes:
Scott W wrote:

But it is for reasons of hinting at coming trends that they run these
models - I'd be interested to see the statistics on how correct the GFS
has been for anything past plus240z over the years.

Are there any data on the accuracy of each 6-hour/12-hour output from
the models?

http://www.emc.ncep.noaa.gov/modelperf/

But if you can understand it you're a better man than me, Gunga Din.
--
John Hall
"Honest criticism is hard to take,
particularly from a relative, a friend,
an acquaintance, or a stranger." Franklin P Jones

John Hall wrote:

http://www.emc.ncep.noaa.gov/modelperf/

But if you can understand it you're a better man than me, Gunga Din.

I think this is what I am looking for:
http://wwwt.emc.ncep.noaa.gov/gmb/ST...rmsmezdie.html

Something that shows how each much confidence can be placed on average
on each forecast day but that covers the full range of the GFS output
would be better.

--
Jonathan Stott
Canterbury Weather: http://www.canterburyweather.co.uk/
Reverse my e-mail address to reply by e-mail

I do tend to only look at the ensembles these days but frankly, past 7 days
they're not much use either, especially at the moment.

Dave