A question. To what extent are the NAO forecasts independent forecasts or are they merely a compilation of whatever the model outputs are showing? If it's the later, then presumably they aren't really any more illuminating in terms of forecasts than simply looking at the ensemble means or whatever. 

Absolutely right. We have the same discussion every winter. The NAO forecasts are just another way of looking at the same data.

The GFS has often been showing high pressure right over the pole, but with a lobe of the vortex between the pole and the UK, leaving us in westerlies, so it's perfectly possible to have a highly negative AO and still be mild in the UK.

According to Wikipedia, the most negative January AO index was in 1977, which coincided with a cold January for the UK but not an exceptionally cold one (CET 2.8C). By contrast it was extremely cold in the eastern USA.

More here:

https://en.wikipedia.org/wiki/Arctic_oscillation

12z ICON at T+168 to kick things off this afternoon:

Yes, thank you, I read that. I guess there is something in after all it but I haven't noticed much variability in  the ensembles other than the creeping upwards of the 850 means. Nevertheless there are always some posters saying how bad GFS is when it doesn't show what they want and I think it is often unfairly targeted for the reasons I suggest

Seen worse

Not as much energy undercutting vs. it's 00z run (and ECM), but it's a complete contrast to GFS 00z/06z Op at T+168.

Um you missed the point. Never mind I do not want to derail the thread any further.

This time perhaps different but normally I would agree with that.

But to get sustained locked in and true c old you want the NAO and the AO to both go negative like what we had during the feast from the east! Are you saying not to take much notice of the NAO and AO?

What's interesting is that despite period of some -NAO we have had this winter - we still havent had blocking or anything cold so it is a waste really? SO I guess you're right it's pointless.

I think the point Tally is that those NAO forecasts are driven from one model (GFS). 

Icon is quite a cold run and full of eastern promise at the end, tying in with my thoughts.

Yes quite so it's the NAO forecast that reflects the model output not the model output that reflects the NAO FORECAST?

And not buying the EC deep low to the south solution with as you say much deeper cold poised across the water.

iCON ONLY GOES UP TO 180 HOURS? Not really checked this chart much as it's low res:

Looks like a slider low at 75hrs

Bingo

The NAO is simply a measure of pressure at low latitude versus pressure at high latitude. If you have blocking to the north (or deep low pressure to the south) then you have a negative NAO. It's a measure of pressure patterns, not a cause of them.

Back to the models and the GFS shows "snow on snow" for higher parts of the north, with lying snow still present from Friday night when the next band arrives on Monday night.

I don't think the initial slider has been the problem. The issue is the following low emerging from Canada, with most GFS ens members on both the 0z and 6z runs ploughing it eastwards, flattening the northwards-ridging mid-North Atlantic block.

You do realise that the AO (which the post you were replying to was talking about) and the NAO are different things? One is the Arctic Oscillation and the other the North Atlantic Oscillation. 

Someone's come into work unpaid and fixed the GFS. 

Well that looks OK.

prefer the angle on this run, agreed it is what comes out of Canada on if the vortex can split enough to direct where we need