"Joe Hunt" wrote in message
...

http://www.wetterzentrale.de/pics/MT...ester_ens.html
However the operational run looks to be more severe than the rest.

I like the consistency (relatively speaking) up to the 28th -- eight
days ahead is quite a long time for this -- but it is noticeable how
it falls away totally after that, and from the 29th the runs are
practically random (within the climatological envelope). It'll be
interesting to see whether future runs continue to identify the
28th as the day the models lose it.

Philip Eden