Mark Smithy

I'm pointing out that the model outputs are not anything like as categoric and clearcut as you are being. This is on a knife-edge. You're right that I do pay attention to downstream signals but it's rather dismissive to describe this as micro analysis. History and many painful lessons aka mistakes in forecasting have taught me that just because a butterfly flaps its wings on Sulawesi it doesn't mean snowfall in Swindon. There are too many downstream variables in the UK that affect, though don't necessarily knock out, the macro upstream signalling. The reason that our model watching so frequently goes pear shaped is that we think what should happen, will do. 3000 miles of Atlantic, warm seas, a North Atlantic Drift, and that other pesky contributor of climate change, too too frequently throw things awry. At the moment the model outputs are not clear cut about an ongoing cold spell beyond end of next week and there is no certainty of further height rise over Scandinavia. Many of the ensembles are showing alternative outcomes as evidenced by the ongoing scatter at relatively short range. Clearly though if in the depths of winter you repeatedly stick a tail on a donkey's derriere, eventually you'll hit the spot. Ignoring the 99 times you didn't. Blocking of Atlantic systems. There are 3 or 4 principal high pressure blocks that can cause UK cold: Greenland High (north-west), Svalbard High (far north), mid-Atlantic High, or Scandinavian High. The latter usually means an easterly but its exact position is important. Sink too far south and it leaves us in mild south-westerlies. Too far east and the same.

Not being critical but you're being a tad subjective. The models are on a knife edge and this could go either way. The form horse is in fact for the mild air to win out because that's generally what happens, especially in recent years. A Scandinavian high pressure is most certainly not a foregone conclusion. The ensembles and yo-yo operationals are evidence of how finely balanced this currently is.

I think the ECM is important here. It has played with both scenarios recently. As we know, the last two GFS operationals have seen a breakdown of the cold although always getting pushed back, so now not until end of next week. The main thing for me is that the GEFS ensembles show huge scatter. Some of the scatter is eye-popping e.g. this yesterday: And now the last two runs (0z not out yet): It's evidently a battleground situation with the UK in the firing line. Historically this has often produced the heaviest snowfalls. But it could just as easily fizzle out into mild mush again. Game on.

Well ECM goes with UKMO ... for now! But the GFS ensembles show the split. A majority (just) favour the milder outcomes: It's knife edge stuff, which usually only ends up one way in this country but let's see. Maybe this will be the 1 time in 20 when it actually comes together?

If the 0z ECM goes with the 0z GFS then it's almost game over. This GFS isn't an aberration. It has been likely from the ensembles for a couple of days with more and more members trending milder from the 14th. It's not nailed on that the cold spell will come to an end next week, not least because the UKMO still favours it sticking with a channel low, but the fact that the two biggest players, GFS and ECM, have both shown southerlies returning next week, and a return to much milder uppers, means that this really is the likeliest scenario.

Well I've been on the verge of posting that this might be the 1 in 20 cold model outputs that verifies. And you'd think looking at the 06z GFS that we're heading into an epic, right? Unfortunately this is far from nailed on. I really do not like these 06z ensembles: Not only is the operational now one of only a few going cold, there are an increasingly large number of members heading milder. I don't like this. If it's a trend, it's trouble.

I think we have to accept that this is not a blocked setup: This is not a pattern change, nor anything out of the ordinary. The Atlantic is resurgant on it and both the ECM and GFS have now picked up on this. However, it's nearly 10 days away and this is not in the reliable timeframe so things may change. It's the absence of a block that should be the biggest concern.

In what way? The 18z and 0z don't inspire me that there's a lot of significance in this at the moment. A colder start to next week for sure but there are certainly signs that the Atlantic is coming back into action and there's a lot of mild air flooding back into Europe. Take this for instance: That's nothing like the old days of embedded cold on the continent which became so important for UK wintry weather. The jetstream is still very much in play too: The ECM was the first to pick up on the potential breakdown with it's almost infamous 12z of two days ago. Now some of the GFS ensembles are picking up the same theme. When this happens it usually means that's the way it will go. So a cold spell of weather to last another week then a breakdown. Not sure many people will see snow but there may be some wintriness especially over the hills in the north. It will feel colder than it actually is in the wind.

Just back from a lovely cold 2 hour hike. These 06z ensembles are absolutely nuts. Huge spread, all over the place. It's obviously cold right now and it will be in the short term, especially early next week. After that? The jury is still out.

There really isn't full support from the GFS. This is a cause for wariness. To be sure there's a tight colder cluster, but the operational is very much on the coldest fringe. What concerns me are the significant number of alternative mild scenarios appearing, which is what the 12z ECM detected and went with in its operational. Anyone dissing this isn't being meteorological. We might get the coldest scenarios verifying but it remains in the balance and historically 19 times out of 20 in the UK that means it will go the way of the mild. Just saying.

What tends to happen is that people pick and choose whichever model, or whatever individual run, suits their preference

I haven't been following this thread but I'm sure that's not implying that those wary of the coldest scenarios are being criticised. There's a lot of historical precedent why we should be wary. And in the last few hours it's not just the 12z ECM that illustrated the dangers but the scatter on the GFS ensembles, with so many mild members appearing. I bet this goes pear-shaped. It nearly always does.

Although the 0z ECM operational has for now ditched the fast exit from cold shown on the 12z yesterday, these GFS ensembles illustrate the reason for the concern: 19 times out of 20 in this country when this starts happening it's not a good outcome (for cold lovers). Maybe this will be the 1 in 20?

Yes it's all gone a bit flacid. The easterly is no longer the potent force that was showing up on the charts a few days ago and the north-easterly afterwards is still pretty uncertain. It's basically going to be cool to cold but fairly slack for a week. After that, as you say we're losing the Greenland and mid-Atlantic block.

Wow

The Atlantic is never dead for the UK. Dangerous words ... it is always "prowling around like a roaring lion, seeking whom it may devour" We can but hope but please let's not tempt fate.

A continued word of caution, for which I apologise, but as has been pointed out by a few others this is not really nailed on yet. At least, not in terms of intensity of cold. I'm still concerned about the ensembles. Whilst there's possibly a move towards more members in the colder section I don't like this much scatter at such short range. It tells me that this is still by no means a done deal:

Looking at the 18z and 0z GFS ensembles I would still urge caution. There are a lot of members not playing ball with this: For as long as there is this kind of scatter and such a pronounced split past T168 I am not on board.

It's perhaps worth noting though that the "this is better than previous years" is a line which gets repeated by cold lovers on every forum every single winter. I suspect this is more a case of us hoping it to be true rather than what the models are showing. There really isn't anything unusual or out of the ordinary so far this autumn (and it is still autumn). Yes there's a bit of blocking out in Siberia, which may or may not build into Scandinavia, but that's pretty normal. Also, and I mean this politely not rudely so please don't take it that way, yes we are "heading colder". It's been a mild autumn and December is about to start, with it the winter season. So, yes, it's going to get a little colder as we come into the winter season. Yesterday afternoon / evening saw a deluge in the west country with wild winds. Very much still an Atlantic onslaught. The 18z and then the 0z GFS suggest a flacid easterly which then gives way to something more like a southerly (again). It will cool down but perhaps the most notable feature of both runs is that the Atlantic onslaught might finally be abating. Although I do note that this has potential for considerable rain around the 5th/6th December. A cooler, possibly cold, spell is still quite likely but no guarantee of much of an easterly. And it may pep up again on the GFS. UKMO 0z still thinks so. But generally it's best to take charts out past T168 with a big pinch of salt. Maybe that's why UKMO stop there. Sensible people.

The general GEFS ensemble trend is cooler, but you'd expect that coming into December from a mild autumn: Cooler yes. Cold? Possibly but at the moment there's nothing particularly exceptional in these ensembles. Often we see really whacky cold ones starting to appear in F.I. Interestingly, this is one of the few GFS operationals that hasn't yo-yo'd either within its own run or against previous outputs. Whether that's a sign that this operational is more credible I'm not sure but it's certainly a general pattern in the short and medium term that looks plausible.

Yep there's a bit of a danger that we're hopping from one model to another to find the one which best suits our preference. The latest pile-on being the GFS parallel. The fact is that not all the models are sharing in the excitement and even when one does, not all the ensemble members are on board. When you add that to the fact that all of this is way out past any kind of reliable timeframe (T168 at the outmost) this really should be treated with considerable caution. Been here many times before and the FI charts are no more exceptional than any other charts seen in FI year after year after year. They tantalise, they tease, and invariably they let us down. Let's see if this comes inside a 5-day timeframe. Then by all means we can sit up.

An even better sign would be "Please take everything past T168 with a pinch of salt" I don't know how many times it happens but people still don't learn. 19 times out of 20 cold set ups in F.I. don't come to pass in the UK. It behoves seasoned model watchers, of whatever weather tastes, to show restraint until charts come inside a 5 day timeframe (T120). Then by all means get excited.

Agreed. But this is still very tentative and it's best not to place too much confidence in it until all 3 main models come consistently on board inside T168 and preferably shorter range than that. We all remember John Kettley's infamous Beast from the East. The 0z GFS shows exactly why we need to be cautious: it certainly throws a spanner in the works. However, I do wonder whether anything is really gained by the GFS having runs every 6 hours rather 12. It seems to throw up the sort of flip-flopping we have seen from it recently: the contrast between the 18z (stupendous) and 0z (meh) being another good example. Cautious excitement the order of the day.

Good morning all. In fact it wasn't. The 18z operational was less out of kilter than the recent flip-flopping GFS operationals. On the milder side but by no means an outlier: Excitement this morning I see about the 0z GFS but we're talking about charts far out into Fantasy Island and it'll be interesting to see where they sit in the ensembles. As Star Wars puts it: 'in a galaxy far, far, away.'

Well this should give good cheer to cold lovers: However, please just remember my caution (which others have also made many times) to treat anything really cold past T168 with a pinch of salt. Most of us who have studied UK winters know from all-too bitter experience that the only result of placing confidence, hope even, in cold scenarios past T168 is repeated heartbreak. If it comes inside reliable timeframe then we sit up.