On Dec 22, 1:43*pm, "Norman" wrote:

These statistics, of course, are averages covering whole hemispheres. Regional
errors may well cancel each other out. The period shown covers Fri 16th Dec. At
T-5 all the models were predicting a very intense low passing across northern
Scotland on that day, as I'm sure most people in here remember. In the event,
of course, we know that it didn't happen. It was a hugely significant error for
our particular bit of the planet but there isn't a hint of a blip on the
hemispheric 500 mb anomaly statistics.

I'm aware of that Norman - other parts of that website on closer
inspection detail Europe, but in not as much detail as the hemispheric
data. Hemispheric data is of little relevance on the local scale for
statistics ! I have contacted the author to get some of the historical
data as I am interested to see predictability as a function of NAO
index for NW Europe/N Atlantic to get something more meaningful for us
users/digesters of model output.

Richard