ECM D9 Not too bad 

But it returns cold at T168 when the main cold starts anyway. Perhaps more concern might have been the Azores HP moving back in at T240, and although a warm outlier there are several warmer runs then pushing the mean back to -4C in the 12z ensembles. But as the other charts haven't reached this point yet and with the Met Office predicting there will be milder blips during a prolonged cold spell I wouldn't be concerned at this stage.

D10 has the cold over the UK

ECM looks very snowy day 9 and 10. It's similar if not as spectacular as the 0z run but good consistency. 

Strange, I was just thinking how oddly similar the ECM at 240 was to the last run. Not quite as extreme, unsurprisingly, but still showing the jet digging way down into North Africa and flooding Europe with cold. Just hope it doesn’t correct any further eastwards!

It ended in a similar fashion, I was comparing frame on frame, no complaints though , lets see the means 

I can't recall seeing such marked trough disruption before where all the surrounding air is cold.

Like this morning, it is showing heights gradually falling over continental Europe in the last couple of frames to the point of looking portentous. 

Just a pity its way out in la la land as those charts could produce copious amounts of snow.  A fascinating week of model watching awaits. 

Hiding-behind-the-sofa-stuff for the" weak-chinned" amongst us.

The dessert would presumably be Baked Alaska.

Yes its definitely unusual could be a memorable event if we get lucky that is.

Brian - could you explain this chart to me as I find it so confusing with the heavy snow to south and many western parts of country yet heavy rain to the north east? and those temps? Thanks

I know these mean charts have their limitations but the ECM 12z ensemble chart shows that the op run was remarkably close to the mean for the entire duration. This one is for Aberdeen but the others I have checked are similiar:

http://modeles.meteociel.fr/modeles/gens/ecmwf/graphe_ens3.php?mode=0&x=&y=&run=12&lat=57.125506072874494&lon=-2.264150943396224&runpara=0

which is a great sign for a cold spell, definitely going the right way

It looks to me as if one of those two charts must have somehow failed to update from the previous run, even though the timestamps indicate that they both have. Always assuming that one isn't from the operational run and the other from the parallel run.

When I look at the corresponding charts on Wetterzentrale, the temperatures are far lower than shown in the chart posted here, ranging from 1C is south Wales and NW England (and below freezing over much of Scotland) to 7C over parts of East Anglia.

i would say that the first charts is showing a low pressure system over the midlands with a very cold NNW airflow behind turning any showers to sleet and snow. Over the mountains of wales the cold air would be undercutting and turning the back edge of the main front to snow. However 29th is a long way off.

it is correct intermixing of a very cold n/w polar air s/e for n/e

dew about -8 Fh about 600 M

The other answers to your question are all wrong. The reason is simply that the wrong variable was plotted on that chart, the correct values are below. The problem should not occur on the GFS18z which starts in a few minutes.

The positive anomalies for Iceland, Greenland and Spitzbergen suggests that February is expected to be dominated by easterlies rather than northerlies.