GFS in my experience has always been a good model for spotting trends/pattern changes, albeit it's output can be a bit erratic run to run.
The model has been pretty consistent with it's output recently in terms of the general pattern however. The trouble with the set up though it is predicting, is that very small changes to the output, can make a massive difference to the weather on the ground on any day.
If GFS has the general pattern correct for Christmas, you're absolutely right, it is a weather forecasters nightmare. The difference between dry/wet/snow could vary dramatically hour to hour and across close locations.
Originally Posted by: Arbroath 1320