WYorksWeather

GFS 12z looking slightly above average from the 7th to the 11th, then beyond that, not much of a signal. Further north, only slight difference is that the first push of warm air tonight doesn't make it as far as Aberdeen, so the warm up takes until the 7th. Probably somewhat less strong as well - less of a signal for anything above average, more close to average. The Oslo cold situation looks a bit better than on the 06z - less of a signal to turn things mild there. Models still probably struggling to nail down the intensity of any renewed Atlantic push. I'll post the same for the ECM later on when the full ensemble is available.

Very marginal here. Went to the gym a couple of hours ago as the snow didn't look like settling. Gym is near sea level, still nothing when I came out of the gym. Got back home to find a fresh cm or so of snow, which was a nice surprise. Still coming down heavily here, but not sure how much more will accumulate, and expecting a gradual snow-rain transition overnight, so I think it'll likely all be gone by morning.

Been an extremely grey day here - exceptionally poor visibility. Probably got about a centimetre overnight, and a bit of the snow has melted, though relatively slowly. I think probably as others have said, areas with a bit of elevation will do well, but I think it'll be the wrong side of marginal for me at about 80m ASL. Probably sleet turning increasingly to rain as the night progresses.

Going to take a different tack - have to say I'm actually not liking the 06z as much if cold is what you're looking for. I've been keeping an eye on the Oslo chart, and there's a definite deterioration. Today's 06z: Yesterday's 12z: Definite trend there to bring something milder in more quickly, around day 8 or 9. As a metric, the 850hPa temps rise above -5C by the 11 or 12th on today's 06z mean, which is getting closer to the reliable vs. around the 15th yesterday, which put it well out into FI.

To be fair, even if we look at mean maxima, the anomaly for the year is 1.78C. June was a very warm month for the CET region, average max of 22.6C, and the only months that were near average were July (20.1C, anomaly -0.2C) and August (20.5C, anomaly 0.4C). All other months this were at least 1C above 1961-1990 for max CET. I think it's more the timing - the months that have done the heavy lifting were not the summer months, with the exception of June. The three largest monthly anomalies were June, September and February, which gives the year quite an odd feel. I think if you swapped September with July and February with August, it would feel much more like a year deserving of breaking this record. To be honest though - it is kind of inherent in the way this one works. It rewards a consistently mild year, rather than an exceptionally hot one. You only need one cold month to scupper it. What would feel like more of a travesty would be if this year dethrones last year for the highest mean CET on record, but it's very unlikely as we'd need a 7.6C December.

There is also a weather regime effect. Don't want to derail this thread too much, but hope this is still considered relevant, as it does have implications for e.g. model output showing cold and/or blocked patterns. Just a moment... RMETS.ONLINELIBRARY.WILEY.COM Useful chart: In short, there is a higher bar to clear, even within the -NAO regime which is best for cold in our part of the world, to have it be cold enough to deliver anything wintry. Here's another chart from the same paper which illustrates this effect very well: Original post: https://community.netweather.tv/topic/99584-model-output-discussion-into-winter/?do=findComment&comment=4968267

November similar to October in anomaly terms. September still stands out as the most anomalous month of the year overall. All eyes on December to see where the year finishes - of course the warmest year on record is a formality at this point. Latest from Climate Reanalyzer: Still showing global temperatures running over a month behind. Temperatures around 0.3C above the previous highest for this time of year, and more in line with what you'd expect typically in late October. The renewed rise in the anomaly is driven by a slower than usual temperature decline in the Northern Hemisphere, and particularly we see Arctic temperature anomalies at +7C, which is close to a record high for this time of year.

In any case, continuing to look at model output, and comparing the 18z to the 12z, we don't see much of a difference on the ensembles, except a little less spread later on. The lesson I think is to not get too hung up on the individual OP runs - there's really nothing much to see here from an ensemble perspective. Probably all you can say is an increased chance of near-normal temperatures, with a reduced chance of anything overly cold or mild compared to the 12z. Net result being that the mean is almost unchanged. More runs needed...

Sorry yes was replying quickly - forgot to take account of your location. Based on Coatbridge, your nearest station would be Salsburgh. Average high is 5.4C in Dec (so somewhere 6C ish given we're in early Dec). On 1961-1990 your average high is 5.0C, so there is a small increase. Here's your GEFS plot from the 18z. Mostly in line to a little below the relevant averages beyond the current cold spell, and/or no signal.

There is also a weather regime effect. Don't want to derail this thread too much, but hope this is still considered relevant, as it does have implications for e.g. model output showing cold and/or blocked patterns. Just a moment... RMETS.ONLINELIBRARY.WILEY.COM Useful chart: In short, there is a higher bar to clear, even within the -NAO regime which is best for cold in our part of the world, to have it be cold enough to deliver anything wintry. Here's another chart from the same paper which illustrates this effect very well:

Going to take my GFS chart from the 12z for London, and see what happens vs. the 18z as the ensemble rolls out. So far, out to the 8th, it looks fairly similar. Squinting at it, it's ever so slightly later with the arrival of mild (breaking through the mean early on the 8th rather than late on the 7th), but has a slightly higher peak. Will have to wait a little longer to see where the ensemble goes later in the run.

For my 12 months above 1961-1990 thread, was having a look at CET for the next few days. I think the most likely outcome to the 5th is probably around 0C give or take a degree. Beyond that, uncertainty reigns I think...

Now reviving this thread. I'll be tracking throughout December. Based on the last few years, we have an a priori chance of around 80% for any given month to be above average on 1961-1990. So as a starting point, you'd predict that this enigma is likely to go this year. But, as we all know, December has started with a cold spell. I'm going to update this fairly regularly, and give my assessment of where I think we're at. So, with that being said, our target to beat is 4.5C, so I'll set 4.6C as the minimum required. From my post in the December CET thread, we have the 10-year trailing average (2013-2022) at 5.9C, and the 30-year trailing average (1993-2022) at 4.9C. So, we can therefore calculate using today's average, substitute in the 10-year and 30-year averages, then see what happens. I'll do two versions - a naive calculation, then one based on an outlook for the next few days which should be a little more realistic. Naive version - averages only Today's average: -0.5C (1 day) 10-year average for rest of December: (-0.5 * 1 + 5.9 * 30) / 31 = 5.7C 30-year average: (-0.5 * 1 + 4.9 * 30) / 31 = 4.7C Assumed average to Dec 5th Based on model output, a reasonable guess for the running average to the 5th is probably around 0C, give or take. Substituting that in, we get the following. 10-year average: (0 * 5 + 5.9 * 26) / 31 = 4.9C 30-year average: (0 * 5 + 4.9 * 26) / 31 = 4.1C Conclusion Based on a short term outlook, this looks like it will be a very interesting thing to keep an eye on, with a real chance of a second December miss in a row! All eyes on the modelling beginning to firm up for week 2. My (subjective) chance of the record going as of today is going to be set at 60%.

You're right that more recent averages have risen a bit, but even 'back in the day' if we look at things like the Met Office CET averages for 1961-1990, the average max was still 7.2C in December. And that's for December as a whole, so at this time of year specifically, probably round it up to about 8C.

Not seeing anything overly mild on the ECM charts tonight. I think some people forgetting what normal is for the time of year - temperatures by day in the high single figures and by night in the low single figures are just average. It's only if we start seeing daily means approaching double figures that I'd call it mild weather for the time of year. This looks to me like a reversion to the mean, nothing more. GFS relatively similar, though overall the ensemble is somewhat less keen on hanging onto the colder air further north. There also remains a cross-model consensus to retain the cold into western Scandinavia (using the Oslo chart) at least up to day 7, after which the signal diminishes, but nothing much to be worried about. Here's GFS, GEM, ICON and ECM:

And right on cue - just started snowing here a few minutes ago. Might be enough to tempt me out for a late night walk despite the cold!

Yeah I don't take their forecasts too seriously.

A couple of hours ago got a very small fresh accumulation at my location - just enough to cover e.g. road surfaces. Not settling on house roofs though, which shows how marginal it is. Probably about half a centimetre. BBC forecast has sleet at my location in the next hour or so, turning to snow later this evening and heavier overnight. Will be interesting to see if we have a more significant accumulation by morning.

I've had a look at records for 850hPa, and can't seem to find any good records. However, in a previous discussion on here I am leaning towards saying that 13C at 850hPa must be close to record-breaking. In a discussion in this thread back at the start of October, the consensus was that the latest 850hPa above 15C was around Halloween, or possibly very early in November. This would certainly be comparable to December 2015-like synoptics if it did come off, based on re-analysis charts. Of course, it is an outlier, so I'm not expecting it to happen. It does illustrate what I was saying a few days ago though, that with the very mild SSTs over the Azores, we'd expect exceptionally mild temperatures as a result of that if we do get a full on south-westerly blast at any point. Relevant date records would be 16.8C on the 14th, 16.3C on the 15th, and 17.2C on the 16th. Very unlikely to verify, but still a pretty crazy model run. On to the 12z, which in FI will probably show us plunging into the freezer, or the UK being wiped out by a passing hurricane

Genuinely I think the COP climate process is one of those cases where it's the least bad option. I don't like the optics of it, it sends entirely the wrong message (there's no reason why most of the key meetings can't happen remotely, so far fewer people need to attend). I agree also that attendees should as a matter of setting the right example either use more sustainable forms of transport, or if they're unwilling or unable to do so, fully offset their emissions via an externally-reviewed and audited process. However, it remains the case that the COP is the only setup we have that gets lots of world leaders talking about climate-related issues. I don't really know how you could replace it. In any case this is straying too far from the thread topic so I'll leave it there. On topic, here is the latest global temperature chart from Climate Reanalyzer. Anomalies have fallen back somewhat to 1.0C above 1979-2000. Will be interesting to see how December turns out!

I'm going to post three charts, one of which will depress the mood, one of which is neutral ish, and one of which will lift it, so I'll do them in that order. To get the bad news out of the way, the GFS 18z solution is not very well supported by the ensembles. There are some other runs that are as cold or even colder, but they're only about 20% of the ensemble. That chart was for London, but further north, we see a bit more of a signal for the cold to hang on. I think that unless one of the mildest options verifies there are still likely to be frosty nights and risk of snow further north and at elevation. Better news on Scandinavia - there's no ignal for any relaxation of the cold in Norway (using the chart for Oslo here) out to day 10. Of course the signal does get weaker into deep FI but that is nearly always expected. In the short term, we're losing the signal for cold further south, and snow and ice chances will be much lower next week than they have been this week. Best chance of hanging onto winter weather will be further north. In the longer term though, it is encouraging that the cold temperatures remain very much entrenched over even western parts of Scandinavia, with very few ensembles predicting anything close to a fully zonal mild reset for all of NW Europe.

Data is one thing, but specific criticisms of the models are another. Perhaps I'm arguing semantically here - I'm not even saying that you're wrong! All I'm saying is that a 20 year old page which specifically states that it is no longer being maintained is not a reliable source for anything. The fact that the page hasn't been taken down is irrelevant - there are tons of old web links that still work but contain massively outdated information. There may well still be a notable bias in the UKMET in terms of over-doing the ability of Atlantic weather fronts to push cold air aside. I struggled to find anything more recent for the UKMET, but here's a page for the ECMWF. I imagine that some of these biases will differ from model to model of course, but just as a flavour. Known IFS forecasting issues - Forecast User - ECMWF Confluence Wiki CONFLUENCE.ECMWF.INT I'd be very interested in seeing the equivalent for other models, based on something more recent. I wasn't able to find anything via a quick search though.

I've noticed a lot of disagreement on here between those posters who are tending to follow background drivers and teleconnections, and those who are following the NWP output. I sometimes dabble a bit in both, so don't really take a side on the issue. Really though, the disconnect isn't as big as a lot of people think it is. Background drivers tend to cause shifts in either probability or impact of weather regimes, whereas NWP gives us a short range look at what is happening. We can tot up the factors that we know affect our weather, and come to a lot of statements about probabilities and impacts. For example, northern blocking increases the probability of blocking the Atlantic off and getting us into a northerly or easterly, and colder than average conditions over Scandinavia increases the impact (cold) effect of that event if it happens. Similarly, the general trend towards an expansion of the Azores high over recent decades increases the probability of south-westerlies, and the current warm sea surface temperatures increase the impact (mild). Certain MJO cycles favour colder weather, and others favour mild. But these are only probabilities - we could have a good-looking cold setup with blocking but if the positioning and timing isn't right, we can end up with raging south-westerlies anyway. An SSW can still be a bust. Similarly, despite the record global warmth at the moment, who is to say that Europe can't be the exception to the rule? In short, the two schools of thought should complement one another. The background drivers give an idea of probabilities and impacts, but aren't a guarantee, and should be looked at alongside NWP, and vice versa.

That page was last updated in 2000. All the models have been massively upgrade since then. To give some idea, I found a set of guidance on the ECM model from 2006. There are numerous physics improvements in the years since then which are somewhat harder to quantify, but to give some idea of how much the models have improved, I'll deal with the resolution figures - there are conversion factors for what e.g. T799 means. In 2006, the ECM deterministic model had a horizontal resolution of 25km at the equator. That figure is now 9km. Vertical resolution increased from 91 levels to 137 levels. The ensembles had a horizontal resolution of near 40km, which is now also 9km. Vertical resolution had 62 levels, and topped out at 5hPa, or about 30km, which means that only about half of the stratosphere was modelled, and now we have modelling up to 137 levels, or 40km, much closer to the top of the stratosphere. If you look through the ECM history here (used as an example because it's quite a good website), you can see a lot more. Computing power of the top supercomputers which are used for these model outputs have improved massively, and the physical models are also much better. There are improvements to wave models, modelling of ozone, ocean salinity, precipitation, humidity, soil moisture, and dozens of other parameters. So it's not just a case of more computer power either. History | ECMWF WWW.ECMWF.INT ECMWF was established in 1975. Its purpose – then as now – was to pool Europe's meteorological resources to produce accurate climate data and medium-range forecasts. Similar improvements will be seen on all the other models. Naturally, the models are still flawed, and making weather predictions beyond the common 5-7 day horizon is still difficult. But I don't think it's at all reasonable to use a 20-year old website to talk about issues with the models today.

Pattern looks very variable out to day 10 - another cold shot at day 8-9: But then much milder briefly as we approach day 10: After that, it looks like the warmth clears east again. Of course, long way to go for any of this, and just one run. Early indications probably continuing to illustrate what we've expected for some time - a relatively cold first 5-7 days of December with continued snow chances, then the potential for swings from mild to cold. Probably too early to make much assessment of which way it will land.