Indeed it does, and as Brian has already reminded us - there has to be cross-model agreement to engage with any forecast. I will concede however that the momentum towards colder conditions has shifted a bit today. The EC op looks to be in the milder group of its members so there’ll be some impressive options in there this morning. 

GFS Op 00z - HP as well forecast moving up from the SW and covering Britain to Sat 28th (though expected to be cloudy and mild). On Sun 29th LP 965mb Norwegian Sea gives Britain a brief shot of N-lies before it combines with LP moving in from Atlantic Tue 31st 990mb E Scotland generates N/NW gales (maybe even a named storm). Then a re-load 990mb Ireland Fri 3rd with colder NE-lies; not lasting and by Tue 7th Tropical Maritime air is back with HP building from the SW. So one cold week, which does not a winter make.

ECM - This week's HP hangs on rather longer, in fact until Wed 1st when a large deep LP 970mb N Norway projects 2 or 3 days of N-lies over Britain, blocked from entering the Mediterranean by HP which has retreated there.

GEM - closer to GFS but the re-load Fri 3rd looks less likely with winds going back NW-ly.

GEFS - mild dropping back to cool Sun 29th with good ens agreement. Then mean temp on the cool side to Thu 9th, most runs colder than the mean which is buoyed up by a few very mild runs. Substantial pptn for 4 or 5 days from the 29th (chance of snow better than 50% in Scotland, decreasing S-wards), dry before that and no more than damp after. 

I'm away for Christmas at my daughter's; reviews should continue but at times to suit her household arrangements.

Well put.  A casual response to this might be that NWP models are deterministic, tipping a jar of marbles isn’t.  But the jar of marbles is, to all intents and purposes.  (That is to say, marbles are surely more than large enough for their behaviour to be subject to the laws of classical dynamics, without any quantum effects coming into play.)  But the conditions you describe will never be met, in practice, just as two games of Ker-Plunk will never turn out the same ….

If an NWP models produced different results given the same starting point, I presume it would be because of a bug (e.g. a hard-to-find timing problem in areas of code that does work in parallel …. I’ve encountered that with flawed synchronisation in GPU code).

(With apologies for going hopelessly off topic.  Ho ho ho!)

Yes, clearly still a lot of uncertainty.  I’ve been watching the GEFS line graph for Inverness (I realise it’s not actually Inverness) for several days now; it has consistently been showing promise for a cold snap with snow in the Highlands.  The GEFS mean this morning is probably now as low as it has been, at or below -5° 850hPa for a good week or so from the 29th.

But there has been considerable variation from run to run.  For example, of the 0z Op, the 528dam contour temporarily pushes a long way north (at 12z going just past Inverness) and, of course, it’s impossible to know what will happen in nine days’ time with any real precision.

ECM and GFS 0z are both showing a rather nice Greenland high at 144, for instance, but the result of the picture is quite different, as you say.  It will be interesting to see which (if any) turns out to be reasonably close to reality.

On the ensemble page of Meteociel you can touch a point on the map, so if you know the gird reference then with luck you can pinpoint exactly the right location. I do that for my location.

Yes me too (give or take a mile or so!). The trend is becoming increasingly obvious on the GEFS data although it could of course still be wrong!

https://www.meteociel.fr/cartes_obs/gens_display.php?ext=1&lat=57.2&lon=-2.25 

The run is only out to +120 hr at time of writing?

I like looking at the ENS data for Aberdeen. It makes me feel all inclusive 😉

https://modeles16.meteociel.fr/modeles/gfs/run/gfs-0-156.png?6 

Personally I prefer the relevance of the Hampshire ones for Aberdeen, but each to their own.

Edit: Incidentally in case you missed the context, the message/conversation I was replying to directly was about finding the ensemble data for your own exact coordinates.

It’s almost identical to the previous run at this stage!

It reminds me of what cold spells were like down here back in the day! Suffice to say that GEFS output for Aberdeen ticks all the boxes - a low mean of around -5C for over a week, plenty of activity below -10, the odd -15... for "London" that would be the best in many years.

It's encouraging to see, though, as it means it's more likely than not, I'd say, to lead to a blanket of snow somewhere in Scotland, and not just at the top of Ben Nevis! And at this time of year, it's exciting... as the lack of insolation means some seriously low minima are possible in e.g. the glens, should there be a cover of snow, clear skies and light winds.

Hopefully we'll at least see another spell of some -10s for inland Scotland.

I listen to Eric Snodgrass' podcasts regularly and he is very impressed with the ECM AI version. He claims it has preformed better than even the ensembles at picking out trends over the last few months (for the States at least)

My guess would be the lower resolution of the control run gives a slightly different output to the operational run’s higher resolution.

Yes, I think that is it. There will be fewer grid points in the control than the op. It's also worth remembering that the resolution of the data sets which are downloaded and displayed is not necessarily related to that which the model is run at. For example, the GFS data sets are available at 0.25 deg, 0.50 deg and 1 deg. The 2.5 deg ones were discontinued a number of years ago. They all show the "same" output, but if you set a lat - lon for a non-existent grid point, interpolation will be used. Therefore, it's possible to get slightly differnt values for the same coordinate even when using the same run. The same idea will apply to control runs etc.

I think that lowland areas of northern England might get some snow under a strong N/NW flow along with Scotland, but that would be less likely to deliver snow to southern England. I think that could be a possibility according to some of the model output there is at the moment.

Mind you, I think some parts of England away from the north had snow just over a month ago when we had the week-long cold spell prior to storm Bert, so maybe all is not lost.

Yes, he's posted the same on NW - all based on the MJO, which personally I find very strange. It's effectively saying "The GFS rain forecast for these bits of the Indian Ocean isn't looking good, therefore there won't be much winter in the UK"... as the MJO charts are just another way of looking at where the GFS is showing rain. Now it may well be the case that when it rains more in certain places down there it implies a long-wave pattern which favours the UK, but as we've seen so many times local factors outweigh that massively. I've lost count of how many times in recent winters there have been feverous, "jam tomorrow" posts over on NW, all excited about the MJO going into 7-8-1, but then nothing much happens here and the mood sours.

At best it's just a very weak factor in our weather, and at worst it's just random noise. 20 years ago we saw much the same with the NAO or AO, which are of course just another way of viewing pressure charts.

I think it's fair to say I'm not a teleconnections fan!

IMO the best guidance we have remains the ensembles, and more importantly looking for inter-model and intra-run consistency. Get both of those in place and more often than not you'll have a good idea of what's coming.

Ironically the M4 snow shield is breached, but even so - the M2 and M20 put up some stiff resistance! (I'd tolerate the wind for some snow cover, even if fleeting, but it doesn't reach the far south (of Kent) where you are. 🙁

https://www.theweatheroutlook.com/charts/gfs/12_222_uksnowdepth.png