Well I did save the GFS of 1st Dec for Thursday 21st Dec and then looked at
the same chart on 21st Dec. Surprise, surprise - it didn't bear any
resemblance whatsoever.
Dave

What we need is an archive of operational runs at, say, 5 and 10 days
out - to compare with the Wetterzentrale analysis archive. Then, we
could back-test our theories; does anyone know of such an archive? If
not, perhaps we should start our own?

--
Dave