Better height rises over Eastern Canada by day 7 on the 12z GFS op run (right panel) compared to the 06z (left).

That brings it more in line with the 00z ECM op and the 12z GEM op.

It leads to a glancing northerly of sorts (below left), before LP development to the NW cuts that off and blasts us with some westerlies for a day or two - but as the strat. vortex continues shifting to Siberia, GFS plays nice and keeps those height rises coming across the U.S., allowing a second northerly to end the run which has a better upstream pattern for prolonging the flow from the Arctic to the UK (below right).

I'm pleased to see such a run from GFS, but the GFSP 12z run reverts to a much flatter flow very shortly into lower-res (starts at +240 hours for that model), despite a lot of similarities with the overall pattern. Not only that, but despite the strat. vortex being on the move just as GFS shows, the GFSP run takes some energy from the Pacific and develops a new Greenland/Canada polar vortex lobe from it for the final few days of the run.

That occurs because of a shallow low escaping the Pacific at +180 hours, which is not present on the GFS run. The same feature is present on the ECM 00z run, but that model has a strong ridge of high pressure that could prevent it from traversing to the Greenland area... but really, the 12z GFS op run's idea of having so such low at all is a much safer route toward a decent ridge to Greenland.

It's funny really , you don't see many  Summer is over posts in May , yet Winter seems to be different

Oh trust me, there's just as many idiotic posts about summer in May as there is winter. 

To be fair I've not seen anyone of worth actually say winter is over? Equally we've had to delete some just as stupid posts today telling everyone Xmas blizzards are on the way. Takes all sorts I suppose. 

i wouldn't bother... The latest super-extended CFS shows 2015/16 as a mild washout... Roll-on 2016/17.

One of the most direct westerlies I have seen. It really explains why the air on it is unusually cold; during the last winter constantly rpm conditions only once or twice was it this cold.

This chart -144 hours away - represents an important "crossing of the Rubicon".  The best I can say is that for the first time in a long long time most of northern Europe is flooded with sub 0 upper air.  We might just start to shed off the retained warmth of the 3rd warmest autumn in the UK temperature record, if this can keep going for a bit.

http://www.meteociel.fr/modeles/ecmwf/runs/2014113012/ECH0-192.GIF?30-0

Some chillier spreading across most of Europe

Similar to last year in that we are seeing a very rPm dominated theme:

is the Christmas blizzard off Matty? I've backed the bloody thing now, only 250000-1 at Hills

Brian, Four has won the prize 

With the horror of British winter weather, always back the model showing crud!  It is the only way to stay sane. 

Hey! I didn't say winter is over, just that it is unlikely to start. That is not the same thing at all ;-)

Just when I was starting to take the establishment of persistent low heights to Siberia as a solid trend in the model output, along comes the ECM 12z op run with a totally different angle 

In this scenario, the pacific tropical cyclone hasn't managed to bring amplification to the U.S. pattern, and the low heights are actually moving out of Siberia days 8-10 with a ridge building in behind.

I suppose this really represents more of a disintegration of the polar vortex, which is something GEM shows a more extreme version of.

Based on this latest output, I can only conclude that the models are in fact clueless about what will happen as the polar vortex becomes more unstable from around a week today.

This is why I usually refrain from looking too much into charts beyond about a week's range - my recent analysis of what routes we might take depended on the Siberian vortex being a reliable element of the model output despite it being at over a week's range. I must admit that was foolish of me - when I think back to the likes of 2009/10, the models were usually hopeless at handling an unstable polar vortex until around 4-6 days range.

Let me guess? You've given up on model watching for winter weather because the models keep dishing out warm runs for the next month

So you've decided to try reverse psychology on the weather... Is a good job LRF are to quote G: JFF  Just joshing MM...

Interesting developments with the PV on some of today's GFS runs. 18z GFS & GFSP both show this:

http://www.theweatheroutlook.com/twodata/datmdlout.aspx

Still stuck in the twilight zone for the UK though

http://modeles.meteociel.fr/modeles/gfs/runs/2014113018/gfsnh-0-240.png?18

Didn't realise the Express posted on these forums.

One could be forgiven for thinking that they post here sometimes, given the melodramatic posts we get on this thread in the winter.

The 18z GFS runs have made it clearer what causes ECM to head down a different path from day 7 onward; the jet is just a little more amplified, which sends a shortwave low from Canada on a trajectory towards the UK instead of towards the Siberian lobe of the polar vortex.

I've put this up to illustrate the significance of the low off the East. U.S. Coast becoming a cut-off feature - it enhances the ridge in the western North Atlantic and leads to the amplification. The further south that low is, the stronger the ridge will be and the more the trough near to the UK will dig south. GEM shows the low about as far south as possible and the day 10 outcome is a bit nuts.

Here's a link if you can't help but take a peek: http://modeles.meteociel.fr/modeles/gem/archives/2014113012/gemnh-0-240.png

This low/ridge/shortwave interaction is potentially a game changer, as the GFS runs only manage to maintain the Siberian vortex lobe as a result of the Atlantic low(s) phasing with it (the GFSP 18z op has multiple small lows streaming into it in a rather unrealistic fashion).

ECM has clearly been tending to over-amplify charts beyond 5 days out of late, and the behaviour of that low pressure system is not likely to be captured very well by any model at the moment, so there's a fair chance that the amplification won't quite be enough to take us down the ECM path... but I have seen that low/ridge interaction often enough in times past to know that it can have quite an impact.

Looking back at how that ECM 12z op run ended, the momentum of that vicious low south of Greenland and its position relative to the downstream trough suggests that a day 11 chart would have shown a truly extreme storm for much of the UK. It actually looks like a storm that has transitioned from tropical to ex-tropical, which is not out of the question if the low develops as far SW as ECM has it.

I'm glad to have located the explanation for the sudden split in longer-term outcomes and as a result a new period of particular interest to watch - for a time this evening I felt a bit lost!

Nothing new - just our usual dreadful "neither here or there" climate doing it's thing.  we'll get the proper cold eventually. Just it takes a lot patience on these islands. I'm sure this Winter will deliver.

Well, to be serious for a change, I really do think that for the UK weather you have to look at the overall pattern. Last winter, it was clear that by mid-December we were stuck in a zonal pattern and once that gets established, it takes an earthquake to shift it (or at the least a very major strat warming). In a winter like that, you are lucky if you see any decent cold spells until February, if then, since once the zonal train is in motion, it never seems to come off the rails in January.

So, let's look at the weather in another two weeks - if the zonal train is blasting across the Atlantic, I would say forget about any proper winter weather for six weeks at least.

Ive mentioned this before - I don't buy into that. The weather pattern would, of course, eventually change in periods of zonality. Whether that took 6 days or 6 weeks is not set in stone. There absolutely no reason whatsoever why it should take weeks. Of course it could do, but the notion that once it's in, it's definitely in for weeks or months is inaccurate.