https://www.theweatheroutlook.com/charts/gfs/06_174_ukdp.png 

This is why I can never understand when people get touchy with those who say it isn't going to happen. We know from experience that with proper cold spells the models almost always fail. Otherwise why (IMBY) would I have gone several years with no lying snow of note?

despite loads of predicted cold spells?

I think it’s that people ‘get touchy’ mostly when someone deliberately misinterprets the charts or gives undue emphasis to a particular outcome that isn’t supported by the model output. 🙂

I get touchy with them because it's lazy and requires no thought at all - there's seldom any analysis, it's just "oh, it never happens so it won't happen this time". Irritating.

The fun of our hobby is the chase, of course. It's spotting the signs early, watching it get closer and then - finally - watching the snow come down. We all know full well that a covering of snow is now a twice a decade event, if you're lucky, down here, which is why there's always fascination when it might be shown. I don't doubt that if we were facing our usual zonal SW'lies this place would be a heck of a lot quieter - it's the out of the ordinary which draws us all to this hobby.

We're seeing the same thing amongst some on here recently. There have been plenty of those lazy posts, and there's been precious little analysis from them either - barely better than one-liners, in fact, and it shows a lack of skill. It's damned easy to "predict" the form horse, after all!

I'd far rather have discussion about what the charts actually show (which is roughly a 1-in-3 chance of a notable cold spell at present), with discussion about what could go wrong - things like whether the alignment and strength of the jet might be mis-modelled, whether heights over Iberia may end up higher than forecast, even whether a "spoiler" low might pop up and stop the advection of the deep cold air just before it reaches us.

At the moment we still have a decent chance of some serious cold down here. Will it happen? Quite probably not, but you can't rule it out. And that's the fun part for me!

I'll leave you with this morning's "fax" chart.

https://www.meteociel.fr/ukmo/fax/fax0.gif?01-18 

We may not have won the jackpot right now, but even so... we have a 1048hPa Scandinavian High, and we're still (just) in the heart of winter, something I define as mid-December to mid-February. It's been a long, long time since we last saw these synoptics, it's just a shame it's not a couple of degrees colder.

https://www.wetterzentrale.de/en/show_diagrams.php?geoid=50893&model=ecm&var=2&run=0&lid=ENS&bw=1 

https://images.meteociel.fr/im/50/29232/gfs_0_216qfx4.png 

Here's what the 6z GFS had for today, when it was 216 hours out:

https://images.meteociel.fr/im/60/23006/gfs_0_216mpt3.png 

...and here's the reality. I think you'll get my point!

https://images.meteociel.fr/im/57/1240/gfs_0_6bbf4.png 

The crual charts themselves are not too exciting, as they have HP more or less over the UK, but they would be cold! Eg look at P4, P16...

But two weeks out the mean charts are definitely spelling Game Over: https://modeles16.meteociel.fr/modeles/gens/runs/2025021006/gens-31-1-348.png

The irony is that January has been the coldest in a number of years yet the south in particular saw very little if any snow (Moors did ok) but lowland southern England did poor as well as for frosts as well. As I type is +3c with more cold drizzly rain - if this set up occurred 20 years ago we could have been looking at -2c with snow grains.

No, 20 years ago it'd have been just over 2C with cold drizzly rain. It's warmed around 1.5C since the 1961-90 period, and of course 20 years ago is still in the most recent climate period - 91-20. That's one consolation with the current spell - even in the golden days it still wouldn't have quite been cold enough.

A couple of other things as a reply to Rob's post. One is that the mean charts saying "game over" in 2 weeks is always going to be the case - even in 2010 the GEFS persistently went back to or above normal by the end of its runs, each and every day. And two, the 6z GEFS does indeed still have plenty of colder options. Are they the majority? Nope. Can we write off any chance of deep cold as a result? Nope again!

Several really good posts, Darren, thanks. 🙂

As for the models tending to default to the climatological norm, I think that’s true in the absence of any strong signal for something atypical. I tend to think back to previous cold spells where the models have predicted, often with good ensemble support, that a breakdown was coming, but closer to the time the output changed.

I'm starting to think the ensembles are incorrectly spread at different ranges, at this particular range I think they are underspread. The pattern is very close to being good so I wouldn't worry too much about this for now. The upgrades have been modest in terms of pattern changes in the last 24 hours, but that's been enough to create radically different conditions at the surface.

Hold the faith (at least until the 12Z has run).

I think there’s a tendency to overstate the significance of the ensembles.   In essence they’re simply testing the stability of the evolution of the synoptic pattern based on the opening position.  Suppose the 06z Op is correct and we get a sub -10c 850hPa airmass over the UK. That would make the ensemble output for beyond a week today completely wrong, because there would be a very different starting point.

The frigid air to our NE is tantalisingly close and it wouldn't take much to advect it Westwards. 

So close but yet so far. 

One other spanner is a deep trough in the arctic ocean between the arctic and greenland high at T+60 (see Nhem charts). That trough is an issue because it weakens the connection between the scandi and Greenland and aids collapse. Also in earlier runs from a few days ago that trough ended up over Europe and contributed to the cold surge. We need it slightly further east or weaker.

Oh btw everyone check out P8, its one of the best charts of the winter.

True, but I find looking at the isobar values, even if they are in that typical average pattern, is a good indicator whether is is more likely to be unsettled (below about 1010 hPa) or less so (above about 1015 hPa). 

https://images.meteociel.fr/im/4/19190/156_101UKuar5.GIF 

I'll post a link to an article discussing this in the UIA.

Maybe if we just look at the 850 hPA ensembles but they exist for a lot of other parameters. But what do we have that is better? People grasp at individual runs but we know that they swing the most in either direction and they are also there to see anyway amongst the other ensemble members. When there is good agreement. which is still generally no more than 5 days, the mean of them will always be the best bet in my opinion, by definition. After that a lot is still chaos effect most of the time.