I've now repeated the exercise with the 12Z runs, keeping the forecast time and date as 00:00 on 31st January to allow a direct comparison.

GFS: Op -5.4   Avg. -4.4   >0  3   -5 to 0  14   -10 to -5  12  -15 to -10  3

ECM: Op -3.2  Avg. -3.7   >0  7   -5 to 0  27   -10 to -5  15  -15 to -10  2

Of course for any other date the comparison might turn out somewhat different, but for the 31st if anything the GEFS has got a bit milder, while the ECM ensemble hasn't really changed.

So what is your conclusion, Brian? Both are suggesting that colder weather is probable, with a chance of snow in London, but I’ll take my chances with the AI model, which has the potential for more severe cold  albeit with greater uncertainty.

Feels like modern winters are less about sustained cold and more about volatility  big pattern flips, but they rarely hold for long. Are we also seeing more late-season cold chances now because the Atlantic dominates early winter, and any proper blocking only shows up once we get into late January or February?

....when was there last serious blocking in late January? More like March!

Jan 2013, but yes March is more common in modern winters

Pretty good summary of where desiring cold centric outcomes whilst model watching is at these 1.4 degree warmed times of ours.  I am enjoying this rollercoaster - without huge investment in the outcome, but for now the eye candy charts are fun and a dream. Let’s hope this dream comes true, but it is far from an expectation.

I love your posts Quantum, they are always pregnant with highly plausible hope.  I look forward to you further analysis and lets dream you will be pictured standing next to your sledge with a level foot of fluffy, frozen snow!

So let's see how the models develop... 

LOL

trying to get this easterly is like trying to pass a hardened stool.  

The GFS "AI" model today wavers for the first time in a while - it has precisely two days here with 850s in the -6 to -4 range, then collapses the block and allows the Atlantic back in. To my eyes that looks more reasonable than the 80s-style easterly it's been showing until now! 

There's a bit of a theme in the models of delivering the "round the southern side of the high" cold further north, meaning Scotland cops it more.

It's all moot at this range, of course, but a cold easterly is still far from certain IMO. 

The GEFS, incidentally, now have it as around a 50% chance, having varied between 20 and 50% (mostly 30-40%) over the past couple of days. 

Besides, the output is suggesting to me an element of a Kettley High and then Azores high infiltration cutting off any deep cold before it can begin.  Caution required, Im keeping schtum. 

GFS Op - The above LP deepens 965mb Fri 23rd before filling and moving away to the Med, allowing winds for the whole of Britain to go round to the east Mon 26th. These E-lies are rather weak as the HP itself migrates from the Baltic to Scotland 1035mb Thu 29th and deep cold still staying in Russia. As the centre of HP moves N-wards, something a bit colder manages to reach S Britain on a SE-ly  Sun 1st.

ECM - similar to GFS; at first the E-lies are weaker but strengthen up by Sun 1st but no connection to deep cold which is even further off to the far NE.

AIFS - London, maxima drop from 8C now to an Ice day Sun 25th (but now decreased to only one) recovering to 3C, dry around 25th, small amounts of rain either side. Edinburgh, similar temp profile, rain (more probable than snow) for the week beginning Wed 23rd.

GEFS - mean temp near norm to Mon 26th, then a week with mean well below norm and although ens members spread out from that date, virtually all are below norm. Small amounts of (probably) rain from Tue 20th onwards in most but not all runs, heavier in S & W. Snow row figures from Mon 26th about 30% of possible max in S, about 70% in N..

Colder E-lies are coming, but more like a retriever than a rottweiler.

I was just going to post that the difference in the ens sets for London and Aberdeen rather strongly suggests the north east quadrant of the UK is looking at greatest risk of a more prolonged spell of colder 850s, with MBY closer to LP sitting to the west and moving south or south eastwards in to Europe. That said it's not a particularly warm signal down here and the T2M temps suggest continental air at least as you head through FI. 

I think MOGREPs can be a bit of a mess at times due to the lower number of ens members but they do very nicely show the contrast between the 2 locations this morning:-

Aberdeen:-

London:-

It's a signal that's present on the GEFS and ECM suite too. 

So big picture HP to the east ridging north/westwards, LPs bumping in to the block and generally stalling/filling and/or wandering somewhere south or south east allowing colder air to be channelled into northern England and Scotland, maybe for the south too if the block is robust enough and the LPs weak enough or take a favourable track. Deeper in FI and it remains a case of what the Atlantic does versus whether the block can reposition or intensify, which the ECM just about manages in deep FI bringing in a short fetch but fairly cold easterly. 

Messy, chilly for most and cold at times biased the further North you head. 

Overall I think this morning's GEFS/ECM ens are better than the 12zs from yesterday, particularly the ECM which had more mild members and less colder ones. That's nice to see but we still seem to be wobbling between a more robust HP/weaker Atlantic and vice versa so no consistency yet. 

One of the things I look for is ensemble members which show an ice day down here, as that's a good proxy for an "old school" Continental feed. The ECM ensembles have backed off of those recently, not that they ever had too many, and as you've noted there does seem to be a trend for the deep cold air to hit northern areas first, assuming it makes it across the North Sea in the face of those Atlantic lows.

It's an unusual setup all round (as the traditional easterly would have an upper cold pool being ejected westwards across the south), but at least we're still talking about the possibilities of cold weather - it continues to make a refreshing change from the "will it be 11 or 13C", or "will we see a frost before the spring?" type of thing we usually have this time of year!

Yeah I think that's a fair comment re colder air spreading south with time. I'm being a bit pessimistic at the moment based on the number of milder members (850s in particular which I appreciate doesn't tell the whole story) in the ens suites, particularly the ECM which has generally been less cold IMBY than the GEFS. 

Looking at the spread on T2M temps in the ECM set you get between 5c and 0c for your part of the world (ish) from around 25/01 and 7/8c to 0c (ish again) for London. The majority are probably still around that 5c to 0c range down here but there's enough milder members to keep me on the pessimistic side re cold chances locally for a while yet. 

I may get a bit sulky if we're milder in an easterly set up than NE Scotland and particularly if they get snow and we get rain but as you say initially at least it's not looking like a typical easterly with the strongest/coldest flow across the south although ECM op does remember how easterlies are supposed to work in deep FI at least 😂

I don't know if it was viewer bias and/or whether it's still the case but I was convinced a few years ago that ECM (inc the ens) is a bit pessimistic with temps in the mid to long term range (coldie POV). They always seemed to be higher than GEFS and then correct to a colder solution as the pattern got closer to the reliable range. At the moment that might only be 5c down to 3c and impact how cold any rain is IMBY (being cynical) but if the pattern sort of verifies it'll be interesting to see actual temps versus current modelling. 

That's my conclusion too. I suspect that we may have to wait for another three or four days before we can be fairly sure about whether we are going to get a severely cold spell or not.

Hopefully eventually there will be a lot of ice floes passing under the bridge!

We have seen easterlies fail at 72hrs before.  Ian Brown's train analogy springs to mind