Using analogue selection for 30-day periods and pattern
sequences for 5-day periods during July/August, then using
best-fit years, gives us the following for September 2003:

Mean sea-level pressure over the British Isles:
westerly

Mean pressure anomaly pattern over the Br.Is:
west to northwest, anticyclonic in the south

CET +0.2 degC
E&W RR 108%
E&W sss 106%

Verification for August's Not the Long Range Forecast:
Sea-level pressure pattern:
Actual: anticyclonic
Forecast: anticyclonic

Anomaly pattern:
Actual: anticyclonic, E to NE
Forecast: anticyclonic easterly
CET:
Forecast +2.0 degC (max +2.9, min +1.1)
Actual +2.1 degC (max +2.8, min +1.3)

Rainfall:
Forecast 53%
Actual 23%

Sunshine
Forecast 128%
Actual 123%

Philip Eden

Philip looking back at your forecasts from December to August I see that
your CET error compared to the actual figure has varied from 0.1C to 0.8C
with an average of around 0.3C. As a layman I think that's pretty good
going. I trust your form will last until at least February 2004 (providing
of course your CET figures are welcome!). :-)

Victor

Very professional Philip. You should be a forecaster!
Have you tried TWO they need a replacement for the "man the Mett office
fear"