Just going back to the Met Office July forecast here (which I'm sure will disappear soon for next month's forecast it would be interesting to know where this month's temperature sits on their forecast of a normally-distributed very average month:

http://www.metoffice.gov.uk/media/pd...s-temp-JAS.pdf

I actually emailed the Met Office to include verification of the previous month/3 month temperatures to see how they're doing.

Richard

On Wednesday, July 24, 2013 11:03:08 PM UTC+1, Richard Dixon wrote:
Just going back to the Met Office July forecast here (which I'm sure will disappear soon for next month's forecast it would be interesting to know where this month's temperature sits on their forecast of a normally-distributed very average month: http://www.metoffice.gov.uk/media/pd...s-temp-JAS.pdf I actually emailed the Met Office to include verification of the previous month/3 month temperatures to see how they're doing. Richard

Oh dear! Just adds to my theory that they need replacing. My old boss, who has at last had a decent haircut (almost short back and sides!!!!) is going to have a field day with this. Any idea on who would be suitable for our roving 'cantankerous old git' position?

"For July below-average UK-mean temperatures are more likely than above-average."
"Although spells of hot weather are still possible, these are less likely than would typically be expected at this time of year."
Overall, the probability that the UK-mean temperature for July-August-September will fall into the warmest of our five categories
is 15% and the probability of falling into the coldest of our five categories is 20% (the 1981-2010 probability for each of these
categories is 20%)."

On Wednesday, July 24, 2013 11:03:08 PM UTC+1, Richard Dixon wrote:
Just going back to the Met Office July forecast here (which I'm sure will disappear soon for next month's forecast it would be interesting to know where this month's temperature sits on their forecast of a normally-distributed very average month:

http://www.metoffice.gov.uk/media/pd...s-temp-JAS.pdf

I actually emailed the Met Office to include verification of the previous month/3 month temperatures to see how they're doing.

Richard

"Overall the forecast signals through this period are very weak......."

What gets me is the implication that there are times when forecast signals are strong - which, again, implies that forecast success rates are significantly higher in some months than others.

I don't see any evidence for that.

On Wednesday, 24 July 2013 23:23:28 UTC+1, Jim Cannon wrote:

My old boss

So sorry to hear your news, shooed out by Piers, what ignominy.

Richard

"Dawlish" wrote in message
...
On Wednesday, July 24, 2013 11:03:08 PM UTC+1, Richard Dixon wrote:
Just going back to the Met Office July forecast here (which I'm sure will
disappear soon for next month's forecast it would be interesting to know
where this month's temperature sits on their forecast of a
normally-distributed very average month:

http://www.metoffice.gov.uk/media/pd...s-temp-JAS.pdf

I actually emailed the Met Office to include verification of the previous
month/3 month temperatures to see how they're doing.

Richard

"Overall the forecast signals through this period are very weak......."

What gets me is the implication that there are times when forecast signals
are strong - which, again, implies that forecast success rates are
significantly higher in some months than others.

Aren't times 'when signals are strong' likely to be those occasions when you
are
prepared to issue a 10 day forecast based on model consistency?
--
Col

Bolton, Lancashire
160m asl
Snow videos:
http://www.youtube.com/channel/UC3QvmL4UWBmHFMKWiwYm_gg

On Thursday, 25 July 2013 17:55:43 UTC+1, Col wrote:

Aren't times 'when signals are strong' likely to be those occasions when you
are
prepared to issue a 10 day forecast based on model consistency?

Beat me to it...!!

Richard

On Thursday, July 25, 2013 5:55:43 PM UTC+1, Col wrote:
"Dawlish" wrote in message

...

On Wednesday, July 24, 2013 11:03:08 PM UTC+1, Richard Dixon wrote:

Just going back to the Met Office July forecast here (which I'm sure will

disappear soon for next month's forecast it would be interesting to know

where this month's temperature sits on their forecast of a

normally-distributed very average month:



http://www.metoffice.gov.uk/media/pd...s-temp-JAS.pdf



I actually emailed the Met Office to include verification of the previous

month/3 month temperatures to see how they're doing.



Richard



"Overall the forecast signals through this period are very weak......."



What gets me is the implication that there are times when forecast signals

are strong - which, again, implies that forecast success rates are

significantly higher in some months than others.



Aren't times 'when signals are strong' likely to be those occasions when you

are

prepared to issue a 10 day forecast based on model consistency?

--

Col



Bolton, Lancashire

160m asl

Snow videos:

http://www.youtube.com/channel/UC3QvmL4UWBmHFMKWiwYm_gg

At 10 days and on occasions. Certainly not at a distance of one month - which this was.

On Wednesday, July 24, 2013 11:03:08 PM UTC+1, Richard Dixon wrote:
Just going back to the Met Office July forecast here (which I'm sure will disappear soon for next month's forecast it would be interesting to know where this month's temperature sits on their forecast of a normally-distributed very average month:



http://www.metoffice.gov.uk/media/pd...s-temp-JAS.pdf

========

Aside from the model verifying badly for July there are a couple of problems with the reasoning in the text forecast: 1. An assumption that SSTs around the British Isles would stay below average; and 2. seemingly a reliance on the ENSO signal (and that alone) slightly favouring cool and unsettled, when even the text said only a weak La Nina (or even neutral ENSO) was in the offing. That's a poor teleconnection for NW Europe. Surely they looked at stronger teleconnections?

Stephen.

Additionally: one gets the impression that the model is heavily leaned on, and reasons sought to justify it. Others of us are fortunate enough to have the freedom to forecast against the models.

Stephen.

On Friday, 26 July 2013 13:40:04 UTC+1, Stephen Davenport wrote:

Aside from the model verifying badly for July there are a couple of problems with the reasoning in the text forecast: 1. An assumption that SSTs around the British Isles would stay below average; and 2. seemingly a reliance on the ENSO signal (and that alone) slightly favouring cool and unsettled, when even the text said only a weak La Nina (or even neutral ENSO) was in the offing. That's a poor teleconnection for NW Europe. Surely they looked at stronger teleconnections?

These aren't statistical forecasts though are they? They're realisations of ensemble members of GloSea5, or that's at least what I thought - so they'll have their own teleconnection patterns within them but any common biases that develop within them that go against reality will cock the lot up.

It's shame we can't get Met Office employees on here to discuss the above forecasts - they're happy to have a twitter account - why not someone on Usenet - probably because Usenet is likely dying a death what with the likes of Farcebook and Twitter.

Richard