Gordo,
I was specifically commenting that averaging such a small data set
with such a broad range had very little meaning.
please note that in fact the weather today is solid heavy rain so far.
whereas in fact the 24 hour forecast doesn't even represent the past
very well:
10:00 sunny, though in fact overcast
13:00 sunny intervals, though in fact totally overcast with heavy
downfalls.
regards
Jonathan Chetwynd
On 25 Sep 2006, at 09:19, Gordon Joly wrote:
At 07:33 +0100 25/9/06, Jonathan Chetwynd wrote:
hedging the weather
barcamplondon featured weather averaging for four major suppliers.
I queried the sense of this given the BBC's approach...
for instance today in London the (day) forecast is "light rain shower"
that's half the rays of sun, plus a dark cloud with one rain drop.
already pretty vague you might surmise...
However when you switch to the 24 hour forecast the 3 hourly
forecast throughout the day is "Sunny" throughout the day, with
clear morning and evening.
given the vast expense of weather forecasting, this cannot be said
to qualify as science, it doesn't even pretend to scientism or
scientistic tendencies, more like pure guesswork of a pretty vague
and less than helpful sort?
cheers
Jonathan Chetwynd
The formulation of some sort of "mean weather" (average weather) from
different datasets is one approach.
http://en.wikipedia.org/wiki/Ensemble_forecasting
http://en.wikipedia.org/wiki/Probabilistic_forecasting
http://en.wikipedia.org/wiki/Spaghetti_Plots
I use Metcheck now: http://www.metcheck.com/
They tell you which model and which run of that model is being
presented.
Model : GFS
Run : 00Z
Updated : 24 September 2006 : 08:01:52 AM
YMMV,
Gordo
--
"Think Feynman"/////////
http://pobox.com/~gordo/
[EMAIL PROTECTED]///
-
Sent via the backstage.bbc.co.uk discussion group. To unsubscribe,
please visit http://backstage.bbc.co.uk/archives/2005/01/
mailing_list.html. Unofficial list archive: http://www.mail-
archive.com/[email protected]/
-
Sent via the backstage.bbc.co.uk discussion group. To unsubscribe, please
visit http://backstage.bbc.co.uk/archives/2005/01/mailing_list.html.
Unofficial list archive: http://www.mail-archive.com/[email protected]/
