Martin Rowley wrote:

http://www.metoffice.gov.uk/research...l/gpc-outlooks

Not sure what stage we're in with these ... I assume that the last couple of
iterations at least (issues August & September) have the benefit of the
'high-top', i.e., enhanced stratospheric modelling, and thus the output
indicated is most interesting.

Both August and September issues have fairly strong signals for below-average
temperatures over the 3-month periods NDJ & DJF (where available from August
run): looking at the ensemble mean output, the indication are for departures
from the mean of at least -0.5degC/-1.0degC, and 'fringing' the SW (i.e.,
across the SW Approaches & western Ireland) and also in the broad area
Iceland - Faeroes - Viking, the departure is circa -1.5degC. [Not sure what
the climatology period is ... might be 'model' climatology, i.e., based on
hindcast runs from 1989 - 2002 rather than a 'standard' dataset].

The usual caveats apply of course - high probabilities doesn't mean it's
going to happen - but two consecutive runs with a similar (and reasonably
strong) signal, and with the assimilitation of the consistently anomalous
SSTAs that we've seen as commented upon elsewhere, these products are either
going to be wildly wrong and risible, or are in fact offering us a good steer
on the season to come.

Martin.

In contrast, the current products from the NCEP Climate Forecast System predict
below average temperatures for the British Isles area in October and a very
cold March but for the months in between (Nov-Feb) they show temperatures close
to average. Their averaging period is currently 1981-2006.

All of these products are best taken with a great big pinch of salt, though :-)

--
Norman Lynagh
Tideswell, Derbyshire
303m a.s.l.