https://www.wetterzentrale.de/maps/GFSOPUK12_84_25.png

Some snowfall on the 12z GFS for these parts and more so west of here. Won't happen like that and is changing every run but shows the possibility is there for somewhere. Due to the complexities of the low pressure I doubt we will know until the day.

It is not dissimilar to the 00z op run and in jerking with the cool “messy” outlook.

Still the chance for some snow in high ground in more western or central parts before then, especially NW UK hills.

Yes GFS is very blah up to 240h. GEM is better though very blocked again by day 10. Will gfs pull something out the fire.

Its no better no worse than the previous runs, lots of potential.  The only difference is the lack of snow over London, which was a bit of a long shot.

This whole period we are entering is more about if it will kick start a proper cold spell

In the short term though, theres chance for snow for anywhere in the UK this week.  But as usual, the higher, more north you are the better.

yes fingers crossed some areas still get the chance for some wintry precipitation. At least get someone posting some decent snow pictures this year 

This setup needs a couple more runs to bake a bit longer, but has potential to be the trigger to all our answers

Sliding low and amplification of the blocks to the north and east.... more of this please

GEFS12z update is a bake from a cold weather perspective. Looks like a clear shift to a more progressive outlook, although with 30 runs to look at it's difficult to take in at a glance.

Edit: not that much change actually on the line graphs

I do wonder how much statistical probabilities are affecting this.   

The models must have a bias in the program where they see a very unusual blocking as a long shot, no idea what the odds are, but if there is a choice of a 1 in 100 rare orientation block, or a 40/100 south westerly, it’s obvious which evolution it is going to trend to most of the runs

In other words even if a scenario appears, there are probably internal checks suggesting ideally show something else

Nice ECM slowly getting colder and colder with a strong block still at day 10. Much better than the GFS.

Would have preferred the strong high over Greenland / Iceland on ECM but not a bad ECM. Just need consistency as no doubt will be different tomorrow

I may be completely off base but I always thought the models don't know or care what is an "unusual" set-up or whether the UK has southwesterlies or whatever. They just take the starting inputs of where the atmosphere is now and crunch the numbers to predict where it will be in the future. They know nothing of Scandi blocks or Greenland highs or statistical likelihoods or whatever. It's pure deterministic maths. But feel free to correct me if I am wrong on this.

This is spot on I think.

As Brian has confirmed, you're absolutely correct. The models are a numerical representation of the atmosphere, which obeys the known laws of physics.

The issues are really still around inaccuracies in the starting position: there's less data than there are data points in the models so the gaps are filled with best approximations. That, of course, is tested via the ensemble suites - and you can see how quickly individual ensemble members deviate from the op if you study them in detail.

The other issue is that the models break the atmosphere into horizontal and vertical grids and although modern computers allow a lot more detail they're still just an approximation of how the atmosphere will behave.

I guess the closest the construction of models gets to reflecting actual behaviour is via the testing of their performance against predictions and the n Cesar you tweaking that must be made to improve performance- but that's not any form of bias.

Yes unlike the gov't the models don't have an option to ignore "the science"

Meanwhile, the pub run goes a bit 1947...

Edit: going to be a special run this, the first corker of the season...