On 4 Feb, 10:07, John Hall wrote:

Both the 18Z and 00Z GFS operational runs went for very cold NEerlies by
the middle of next week, and so far the 06Z seems to be evolving very
similarly. With such cold air aloft and quite a long sea track over the
North Sea, I imagine that eastern regions could see quite a bit of snow.

Given that inter-model comparisons are very useful in this sort of
flip-flop situation, thought I'd draw people's attention to something
that the guys at Weatheronline knocked up on my request - basically a
comparison of the models for the same time - all stuck on the same
panel, along with the ensemble means, too.

http://www.weatheronline.co.uk/cgi-b...&ARCHIV=0&WMO=

They're also working on a chart that shows the same model, verifying
at the same time, but for forecasts going back in time - so T+24, T
+48, T+72 etc.. which I'm sure will be useful for seeing how each
individual model is coping with the current situation and how quickly
it has "come into line". The weatheronline guys have said it should be
online in a few days.

Cheers
Richard