there seems to rumors that these are compiled from ouput that is 24hrs old?



UK Outlook for Monday 1 Feb 2010 to Wednesday 10 Feb 2010:
Cold with overnight frost and ice throughout, but a good deal of dry and
bright weather for many parts of the UK by day. However, wintry showers
are likely for exposed northeastern parts throughout the period, with a
risk of more persistent sleet or snow for some eastern areas at times on
Monday. Rain, maybe hill snow, might reach the northwest on Tuesday.
Wintry showers may spread further inland towards the end of the week as
winds become more northeasterly. The mainly dry weather with frost, ice
and possibly fog, is likely to continue for much of the rest of the
period, with temperatures mostly below normal. However, wintry showers
or more persistent snow could affect eastern and southern areas towards
the end of the period.

Updated: 1141 on Wed 27 Jan 2010

In article ,
TT writes:
there seems to rumors that these are compiled from ouput that is 24hrs old?


They are released around noon, which suggests that they will be using
output from the ECM and UKMO 00Z runs, which would become available
sometime around 7am. For the GFS, the 06Z run doesn't complete until
about 10:30, which suggests that there might be time for a quick look at
that. Remember that they'll be looking at the ensembles as well as the
operational runs. (Is there a UKMO ensemble, which we don't get to see,
or just the operational run?)

They may give less weight to the GFS than to the other two models, since
the UKMO isn't a stakeholder in it.


UK Outlook for Monday 1 Feb 2010 to Wednesday 10 Feb 2010:
Cold with overnight frost and ice throughout, but a good deal of dry
and bright weather for many parts of the UK by day. However, wintry
showers are likely for exposed northeastern parts throughout the
period, with a risk of more persistent sleet or snow for some eastern
areas at times on Monday. Rain, maybe hill snow, might reach the
northwest on Tuesday. Wintry showers may spread further inland towards
the end of the week as winds become more northeasterly. The mainly dry
weather with frost, ice and possibly fog, is likely to continue for
much of the rest of the period, with temperatures mostly below normal.
However, wintry showers or more persistent snow could affect eastern
and southern areas towards the end of the period.

Updated: 1141 on Wed 27 Jan 2010

That doesn't seem unreasonable to me, based on the 00Z/06Z runs (or even
going by the subsequent 12Z runs come to that, which haven't greatly
changed things).
--
John Hall
"Acting is merely the art of keeping a large group of people
from coughing."
Sir Ralph Richardson (1902-83)

On Wed, 27 Jan 2010 19:33:05 +0000, TT wrote:

there seems to rumors that these are compiled from ouput that is 24hrs
old?
snip

Somebody on TWO (username "snowhope") emailed them today about this very
issue, and the response was clear: it's the 00z runs that are used. The
full response is posted near the bottom of this page, in blue:

http://theweatheroutlook.com/twocomm...x?PageIndex=17

--
Bewdley, Worcs. ~90m asl.

"John Hall" wrote in message
...
In article ,
TT writes:
there seems to rumors that these are compiled from ouput that is 24hrs
old?


They are released around noon, which suggests that they will be using
output from the ECM and UKMO 00Z runs, which would become available
sometime around 7am. For the GFS, the 06Z run doesn't complete until
about 10:30, which suggests that there might be time for a quick look at
that. Remember that they'll be looking at the ensembles as well as the
operational runs. (Is there a UKMO ensemble, which we don't get to see,
or just the operational run?)

They may give less weight to the GFS than to the other two models, since
the UKMO isn't a stakeholder in it.


UK Outlook for Monday 1 Feb 2010 to Wednesday 10 Feb 2010:
Cold with overnight frost and ice throughout, but a good deal of dry
and bright weather for many parts of the UK by day. However, wintry
showers are likely for exposed northeastern parts throughout the
period, with a risk of more persistent sleet or snow for some eastern
areas at times on Monday. Rain, maybe hill snow, might reach the
northwest on Tuesday. Wintry showers may spread further inland towards
the end of the week as winds become more northeasterly. The mainly dry
weather with frost, ice and possibly fog, is likely to continue for
much of the rest of the period, with temperatures mostly below normal.
However, wintry showers or more persistent snow could affect eastern
and southern areas towards the end of the period.

Updated: 1141 on Wed 27 Jan 2010

That doesn't seem unreasonable to me, based on the 00Z/06Z runs (or even
going by the subsequent 12Z runs come to that, which haven't greatly
changed things).
--
John Hall
"Acting is merely the art of keeping a large group of people
from coughing."
Sir Ralph Richardson (1902-83)

yes there is a UK ensemble we don't get to see, I believe they give slightly
higher weighting to each of operational runs. They (the forecasters) I
believe then look at anything which may have affected the model runs e.g.
bad obs getting in.

Jon or Will should be able to give more detail, or correct me.

"Stan Kellett" wrote in message
...


"John Hall" wrote in message
...
In article ,
TT writes:
there seems to rumors that these are compiled from ouput that is 24hrs
old?


They are released around noon, which suggests that they will be using
output from the ECM and UKMO 00Z runs, which would become available
sometime around 7am. For the GFS, the 06Z run doesn't complete until
about 10:30, which suggests that there might be time for a quick look at
that. Remember that they'll be looking at the ensembles as well as the
operational runs. (Is there a UKMO ensemble, which we don't get to see,
or just the operational run?)

They may give less weight to the GFS than to the other two models, since
the UKMO isn't a stakeholder in it.


UK Outlook for Monday 1 Feb 2010 to Wednesday 10 Feb 2010:
Cold with overnight frost and ice throughout, but a good deal of dry
and bright weather for many parts of the UK by day. However, wintry
showers are likely for exposed northeastern parts throughout the
period, with a risk of more persistent sleet or snow for some eastern
areas at times on Monday. Rain, maybe hill snow, might reach the
northwest on Tuesday. Wintry showers may spread further inland towards
the end of the week as winds become more northeasterly. The mainly dry
weather with frost, ice and possibly fog, is likely to continue for
much of the rest of the period, with temperatures mostly below normal.
However, wintry showers or more persistent snow could affect eastern
and southern areas towards the end of the period.

Updated: 1141 on Wed 27 Jan 2010

That doesn't seem unreasonable to me, based on the 00Z/06Z runs (or even
going by the subsequent 12Z runs come to that, which haven't greatly
changed things).
--
John Hall
"Acting is merely the art of keeping a large group of people
from coughing."
Sir Ralph Richardson (1902-83)

yes there is a UK ensemble we don't get to see, I believe they give
slightly higher weighting to each of operational runs. They (the
forecasters) I believe then look at anything which may have affected the
model runs e.g. bad obs getting in.

Jon or Will should be able to give more detail, or correct me.


They rely on MetO and ECM and GFS ensembles backed up by their own ideas on
how things are likely to develop. The entropy measure of the ensembles
(uncertainty) is used for confidence indicators. Time-lagged ensembles also
come into play as well. OP runs are put into the mix to see how they fit in
with other ideas.

Will
--

On Jan 28, 7:38*am, "Will Hand" wrote:
"Stan Kellett" wrote in message

...







"John Hall" wrote in message
.. .
In article ,
TT writes:
there seems to rumors that these are compiled from ouput that is 24hrs
old?

They are released around noon, which suggests that they will be using
output from the ECM and UKMO 00Z runs, which would become available
sometime around 7am. For the GFS, the 06Z run doesn't complete until
about 10:30, which suggests that there might be time for a quick look at
that. Remember that they'll be looking at the ensembles as well as the
operational runs. (Is there a UKMO ensemble, which we don't get to see,
or just the operational run?)

They may give less weight to the GFS than to the other two models, since
the UKMO isn't a stakeholder in it.

UK Outlook for Monday 1 Feb 2010 to Wednesday 10 Feb 2010:
Cold with overnight frost and ice throughout, but a good deal of dry
and bright weather for many parts of the UK by day. However, wintry
showers are likely for exposed northeastern parts throughout the
period, with a risk of more persistent sleet or snow for some eastern
areas at times on Monday. Rain, maybe hill snow, might reach the
northwest on Tuesday. Wintry showers may spread further inland towards
the end of the week as winds become more northeasterly. The mainly dry
weather with frost, ice and possibly fog, is likely to continue for
much of the rest of the period, with temperatures mostly below normal.
However, wintry showers or more persistent snow could affect eastern
and southern areas towards the end of the period.

Updated: 1141 on Wed 27 Jan 2010

That doesn't seem unreasonable to me, based on the 00Z/06Z runs (or even
going by the subsequent 12Z runs come to that, which haven't greatly
changed things).
--
John Hall
* * * * * "Acting is merely the art of keeping a large group of people
* * * * * *from coughing."
* * * * * * * * * * * * * * *Sir Ralph Richardson (1902-83)

yes there is a UK ensemble we don't get to see, I believe they give
slightly higher weighting to each of operational runs. They (the
forecasters) I believe then look at anything which may have affected the
model runs e.g. bad obs getting in.

Jon or Will should be able to give more detail, or correct me.

They rely on MetO and ECM and GFS ensembles backed up by their own ideas on
how things are likely to develop. The entropy measure of the ensembles
(uncertainty) is used for confidence indicators. Time-lagged ensembles also
come into play as well. OP runs are put into the mix to see how they fit in
with other ideas.

Will
--- Hide quoted text -

- Show quoted text -

The trouble is that with all this "mix", which sounds enormously
complex and scientifically careful and precise, the outcomes when you
get to 10 days and beyond are so poor that the forecasters lump the
whole possible gamut of weather into this 150 word 1984-style exercise
in double-plus-ungood ex language reduction for the public!

It actually tells you next to nothing and we are often reduced to
guesswork as to what the forecast actually means. Each daily precis is
then quietly forgotten and the only statistics that are ever produced
for forecasts at this distance are so vague as to be next to useless.
I think the public deserves either better forecasting - which
presently is not possible at 10 days plus; increases in accuracy at
forecasting the UK weather at 10-days plus has stalled IMO, unlike
accuracy at 3-5 days, which has shown significant improvement - or far
better information and explanation as to why what they are seeing may
well not prove correct.

Finding any accuracy figures on the MetO site is not easy. This is an
example: Try typing "forecast accuracy" into the MetO site search
engine. The matches lead you to several pages in the "about us"
section. You can find Brier Skill Scores for rainfall forecast
accuracy, but try finding at what distance out the accuracy scores are
for. 3 days? 1 day? 10 days? All rainfall forecasts?? Dunno???

http://www.metoffice.gov.uk/corporat.../rainfall.html

Same for temperature, Looks like the MetO are achieving targets
recently - but again, at what distance out?? How can anyone judge how
well the MetO is doing from that?

http://www.metoffice.gov.uk/corporat...tion/temp.html

There are other pages, showing Brier Skill Scores over time, but
finding them on the site is hellishly difficult (I gave up; I'm sure
someone will lead me to them, with the implied hubris that they are
there so why didn't you find them?) When you do. it gives you no
indication whatsoever of the accuracy of this particular 6-15 day
forecast. I think the public deserves better. The reader is forced
into their own research, which is interesting but does not pinpoint
MetO 10 day accuracy at all. For example:

http://www.nws.noaa.gov/ost/climate/...huq_032509.htm

http://www.bom.gov.au/amm/docs/2009/wu2.pdf

The NOAA figures for 5/6 day accuracy are well known and the MetO does
OK compared to other models.

http://www.emc.ncep.noaa.gov/gmb/STATS/STATS.html

The conclusions that I'm drawn to are the same ones as are being
raised in the present debate over seasonal forecasting.

1. Acknowledge the difficulties of forecasting at 6 days+ and explain
them fully on the site before issuing these forecasts. The MetO is
painfully *useless* at doing this. The "expert", "we know best", MetO
mentality is one that they could do with ameliorating immediately to
forge a better relationship with their public.
2. Allow the public (who pay for the MetO, remember), access to *all*
the data sources the forecasters use and be far more clear about why a
particular 6-15 day forecast has been phrased the way it has and what
the chances are of it coming to pass and why.
3. Admit to the present shortcomings (not the MetO's fault, really;
presently NWP only allows so much) and abandon forecasting to the
public, beyond a particular distance, as the accuracy is so low.

There is much debate about whether the seasonal forecasts should be
continued. I think the same debate about whether this 6-15 day
forecast should be continued is justified.

At the moment there is a bald forecast: "This is what we think this
will happen at 6-15 days". Unfortunately there is nothing to tell the
public that anything above 10 days is actually less likely to come
true than it it is likely to actually come true..........apart from an
advert for boiler repairs for British Gas. *))

If we all lived in Siberia and were under a 1050 high that was not
moving, a 15 day forcast might be do-able. Alas a butterflies fart can
change the wx in the UK, as we all know. Buckets full of rain expected
here this weekend and a snowstorm "up north".

Michael.
Orange Park, FL.