Nick B

Indeed, but for me a notable point of interest this morning is in fact the quite unusual similarity in overall shape for the Northern hemisphere which the 6z has with its own 0z run, even at the outer stretches of medium range. If it's wrong on this (which is still likely at that range!) it's been very true to itself in terms of modelling.

Without wishing to pigeonhole your talents into purely the field of meteorology, if you do indeed wish in the future to pursue it as a career (and I think no-one here would have any doubts you've found a potential calling in it), I'll say this only because I haven't seen anyone else point it out yet (and it started bugging me solely because your posts are otherwise very high quality - keep them coming but keep on top of other subjects in class!)... An eddy (small 'e' works just fine, not like the name, referring to a vortex characteristic in a fluid), and eddies, (again small 'e'). I say only because at some point in the future, you may wish to collate your postings as further evidence to submit to a university to which you may wish to apply. So, you really would want to use 'find and replace' to correct them. And about the models (and not in reply to Mr. Eagle Eye!)... I think we're going to see quite a few more days of disagreement in the 96-144 time period yet. Until perhaps another week or so has gone by. They're grappling with some pretty complex conundrums at the outer reaches of their ranges at the moment. Whether to form a Ural high / Aleutian low or not, how that then affects the stratosphere (i.e. severity, type and placement of the likely SSW), and then to extrapolate once again the knock-on effects from that in the troposphere, which in themselves, even with a QTR, are significantly further forward in time (min. 7-10 days beyond). In addition, we have an atmosphere hopefully primed for a quick response to drivers (e.g. MJO) which could well bring about a pattern change to something more favourable (or not!). All in all, my thinking would be, it's still a little early to be hoping for the models to be settling down in these time frames. I mean, they're actually bl**dy good, but we do tend to keep demanding more from them and should keep in mind situations in which it's only natural output will be more chaotic. TL:DR... over-analysis between runs right now might be less productive than it is even normally! I think trends will be more useful, if they can be spotted.

Right now there's a storm heading in the direction of Wiesbaden (from the WSW) which is growing rapidly and seems to have a distinctly supercell-like structure, looking at the satellite imagery. Almost looks like it would contain a hook echo...

NB under a level 2 (lower end) for both today and tomorrow

Estofex.org is looking pretty interesting for C Europe tomorrow. I’ll probably be a bit too far south in Wiesbaden for the more extreme stuff but may be worth keeping an eye…

I wouldn't trust that rate to stay the same... would you? Being on a bus right now with my phone, I'm consequently less able than usual to check on what's known about the historical rates over the last 100-or-so years. I'll bet others aren't hampered to the same degree, though, and could no doubt show an alarmingly growing trend in its rate of loss. Couple that with the reduced albedo due to the loss of sea ice across the entire Arctic area over roughly the same period, I think it's fair to say the balance of probability is pointing towards a less than rosy future in terms of sea levels. Not to mention, it really doesn't need to lose anything close to all of it to create a whole series of problems which are going to beset the grandchildren of our younger generation. As long as it's their problems and not ours, though, eh?

Maybe, but is it in fact trending N-wards or NE (seeming to find a more comfortable home over NW Europe)? The latter would not be in our favour...

It's hard to think of any extended period (i.e. more than a week or two) over the last 13 years or so when the outer reaches, days 8-10-16, would offer so much and then the weather would continue to deliver so little. There comes a timeframe (and it changes depending on pattern changes etc.) where there should be a loss of confidence in the output (e.g. beyond x days is 'FI'). That's what I find the ensembles useful for. Not to pick out individual ones which may be pleasing to the eye, rather to see how the det run matches the control, where it sits within the ensembles, whether there are trends of clusters appearing, and importantly, where the divergence in those ensembles becomes more marked. It's been clear that for some reason, that timeframe this winter has been closer to day 1 than some posters on here have perhaps expected or been used to. Hence the dashed expectations. It might be worthwhile, if people really wish to follow the models without studying the upstream and background drivers, to 1. First, wait until the ensemble runs are there, then check them to see where that point of divergence starts occurring (i.e. where do the lines start really scattering into spaghetti); 2. Then look for trends of clusters in the last three days' output of the same time for the same model, i.e. ECM 0z with 0z from the past two days... where do the runs really start disagreeing with each other and two or three days after that date are there clusters appearing? GFS 12z with GFS 12z from the last couple of days too, same exercise. That may possibly show which new trends are gaining ground; 3. Acknowledge that, beyond two or three days after that scatter point, the usefulness is watered down even more, so that what appears thereafter from Fantasy Island to what could be a new acronym, PD (Pipe Dream)! While noticing there may be attractive scenarios flaunting themselves, don't give them the trust they don't currently deserve (unless you enjoy mental self-flagellation, in which case, have at it!). 4. Absolutely use the method which John Holmes suggests (as well as mushyman too, now) - thank you both! - with the 6-10 and 8-14 charts (understand how they work first)! It really does give a reasonable idea of the direction in which things are probably travelling, and can be used to temper expectations! Bear in mind, if they are changing (i.e. a pattern change is being forecast), then hold back those expectations until they start settling down again. The models have often appeared to want to deliver at ca. 10 days so much in this region this winter, which for those chasing cold has been doubly hard, when the end result of the weather has proved them wrong at that range. Clearly, there is something about the background factors which the models are not equipped to deal with. That's absolutely not the fault of the people coding the modelling, if we are indeed in an era where e.g. the Hadley cell is expanded and its influence is causing the Euro high to be more of a feature than has been the case in the past. It's an area which is likely the subject of ongoing research by climate scientists and the findings will likely support tweaks to the modelling itself to deal with that situation. Who knows whether that is a new normal or just part of the current dance of climate oscillations or indeed a new climatic regime? There's a growing body of literature pointing in that direction, it seems. I'm left with the impression that the current period of the last decade or so hasn't seen as many long midwinter severe cold and snowy weather spells such as I experienced in the 70's - early 90's in the UK. That absolutely does not preclude the chance of a winter with such a spell again at some point, but it may mean our chances are growing slimmer over time. Not much fun to acknowledge for those of us who may still enjoy seeing powdery snow transforming the outside world into a winter wonderland occasionally. If it's any consolation, those of us living in continental NW Europe have seen even less than those of you in the UK and Ireland this winter. It's not over, but it won't be long before I start really looking forward to warm spring sunshine, the smell of blossom, spring birdsong and the lighter colour of early spring leaves budding on the trees.

Well, it must do, if ECM's 240 is being used to write off a whole ten days. Just to note though, I don't see anything particularly of a cold note in that period. However, that's because of the underlying drivers not being conducive, not so much the vagaries of the post-6-day det output.

Which doesn't remove the op run from the range of possibilities, since it's run at a finer resolution and could be the very beginning of an indicator of a pattern change / new trend. If there are a range of clusters all pointing at different scenarios, the mean is far less useful guidance. The building of heights NE is currently not yet a high probability outcome, taking purely the range of possibilities on offer. Its success will be highly dependent on how much the jet energy upstream ends up being routed further south, enabling the LP to establish centred around Italy. Need to keep HP out of W Europe, otherwise everything's riding over the top again. Better to check out the clusters (Catacol and Singularity often post them) and also check out what's going on upstream and with jet energy on the ensemble runs which either support the op or don't, to get an idea of how they are clustered and in which direction any mid-term trends (144 hrs+) may be moving.

Indeed, not to mention the training and experience enabling the preparatory work behind each and every forecast, with consideration of level of technical detail, content and tone, then the delivery itself with the various props available in the day. Michael Fish, a real British institution and what a well-earned retirement! Merry Christmas!

You'd be able to collate data on number of cold episodes. Our gut would tell us the expanded Hadley cell has been a strong factor. However, a 'failed initiation' from what point in time and from what kind of starting point? The various models are simply calculating from a starting point of right now and forecasting out to a future event which hasn't yet occurred, so what has initiated? Also, models have received numerous upgrades (usually but perhaps not always improvements

Pretty good shot for a stump cam, I'd say! Certainly better than any I got from my mobile.

Damn, that was all supposed to be one edited post. Not 4 or 5. Apologies. Bloody mobile and my lack of posting knowledge. EDIT: I've removed the duplicates for you Nick. BW.