Cold Winter Night

A day-to-day comparison of yesterday's UKMO 168h to today's 144h. The trough is quite a bit further Northeast, not West of Ireland, but centred over Scotland. The Greenland High is more tilted, and the Alaskan High is weaker. Less WAA push for Euro heights.

Indeed. The EPS 12z box and whisker plot for De Bilt shows a more defined warmup after next weekend, but at the end of the 15 days (Christmas) that too is slightly colder than the 0z ensemble. The further West you go (in this case Ireland vs Holland), the better the position you might be in, with a Greenland High/Atlantic ridge that is perhaps more West based.

UKMO and GEM are not miles apart at 168h. With the way GEM finishes, that's good to see. GEM does need another push Southeast for IMBY cold, but good enough with all these mild threats appearing.

With FI still at 96h, max 120h if you're stretching it, we can be sure of cold until the end of the reliable, because the cold is here now. Maybe not as cold as you hoped, maybe not with snow for everyone, but it can't be denied that it's here. Perhaps comparisons to 2010 or the like are always a bit unfair. Many Decembers of the past have seen cold episodes like the one we're in right now and for some they end up being more special or memorable than the Big Ones. A face value evaluation of the ensemble output suggests we're going milder, especially the further South and East you are, without a return to real cold before Christmas. The De Bilt 2m plume is ugly in that regard. The KNMI (Dutch Met Office) speak of a 70-80% chance of milder and unsettled later on. A return of the cold would take a big swing in the ensemble. However, in the 96h+ timeframes much depends on the handling of the Atlantic. Many posters have discussed this in great detail, but whether the block is stronger, or better placed than models show, makes all the difference both in that mid term and in what follows longer term. To many observers it may seem like many in here who expect a better setup than the models show are being silly, or even in denial, but tendencies to underestimate blocking and cold are not without precedent. Even comparing the EC 0z T+144h of 6 days ago to what it actually looks like today at T+0h reveals a lot: - The Greenland block is stronger, in better shape and more East-based than EC Op 'thought.' - That infamous tropical storm low is much smaller and less deep (location is spot on though). - There is an Alaskan ridge and associated trough along the US/Canadian Pacific coast, which ECM did not foresee 6 days back. - The Russia/Kazakhstan High has been pushed slightly more SE than foreseen. For De Bilt that results in the next few days being colder than forecasted based on the 5 December 0z EC/EPS. So, my point is, there's still a lot of interest in the models, because of the NH pattern, we may see surprising shifts in the output. Maybe already in the next few hours of 12z's, maybe over the upcoming week, or maybe not! Isn't that the fun of model watching? Original post: https://community.netweather.tv/topic/98087-model-output-discussion-the-cold-has-arrived/?do=findComment&comment=4762561

With FI still at 96h, max 120h if you're stretching it, we can be sure of cold until the end of the reliable, because the cold is here now. Maybe not as cold as you hoped, maybe not with snow for everyone, but it can't be denied that it's here. Perhaps comparisons to 2010 or the like are always a bit unfair. Many Decembers of the past have seen cold episodes like the one we're in right now and for some they end up being more special or memorable than the Big Ones. A face value evaluation of the ensemble output suggests we're going milder, especially the further South and East you are, without a return to real cold before Christmas. The De Bilt 2m plume is ugly in that regard. The KNMI (Dutch Met Office) speak of a 70-80% chance of milder and unsettled later on. A return of the cold would take a big swing in the ensemble. However, in the 96h+ timeframes much depends on the handling of the Atlantic. Many posters have discussed this in great detail, but whether the block is stronger, or better placed than models show, makes all the difference both in that mid term and in what follows longer term. To many observers it may seem like many in here who expect a better setup than the models show are being silly, or even in denial, but tendencies to underestimate blocking and cold are not without precedent. Even comparing the EC 0z T+144h of 6 days ago to what it actually looks like today at T+0h reveals a lot: - The Greenland block is stronger, in better shape and more East-based than EC Op 'thought.' - That infamous tropical storm low is much smaller and less deep (location is spot on though). - There is an Alaskan ridge and associated trough along the US/Canadian Pacific coast, which ECM did not foresee 6 days back. - The Russia/Kazakhstan High has been pushed slightly more SE than foreseen. For De Bilt that results in the next few days being colder than forecasted based on the 5 December 0z EC/EPS. So, my point is, there's still a lot of interest in the models, because of the NH pattern, we may see surprising shifts in the output. Maybe already in the next few hours of 12z's, maybe over the upcoming week, or maybe not! Isn't that the fun of model watching?

Didn't see the EPS clusters come along tonight, I think. Up to 240h not much different from 0z, and in the extended they're not too bad. 35 out of 51 members are in clusters that build positive height anomalies to the NE towards Christmas Day, and even the most Atlantic cluster (1) retains some heights in N-Scandinavia. On the De Bilt plume, the very end is ever so slightly colder than this morning, so we got something to work with, even if we may have to go through a milder (near average) interlude.

A lot of uncertainty about the longevity of the cold this morning ( and yesterday already). The short term looks good for cold. It's here, courtesy of a Greenland High in the here and now. In the longer term and extended, the Northern Hemisphere charts show continued high anomalies to our Northwest and Northeast, which would bode well for continued cold. But at the same time these Highs are, as it stands, modelled just too far away the further South and East you go. An honest, plain analysis of the De Bilt 2m plume shows a return to average, or perhaps slightly above average temperatures around the 21st of December. It is encouraging though to see that the models do not let go of the anomalous NH patterns. Nothing currently suggests a return of the TPV in its familiar NE-Canada/Greenland location. I agree with many here that a rebuild of either Scandinavian or Greenland heights is not unlikely. If the ensemble picks that up too, an ensemble like the De Bilt 2m plume could suddenly switch back to cold. The ensemble lingers on the NAO- side of the regime chart, tending towards the central circle of 'no regime'. Patience is asked once more, but that's a lot easier with cold in place, with snow for some, but at least a seasonal feel. Even without snow, the cold can bring a special experience if you have an eye for it. Yesterday I was taking a walk in nature, as a wall of fog overtook me. In the words of Charles Dickens : "A clammy and intensely cold mist, it made its slow way through the air in ripples that visibly followed and overspread one another, as the waves of an unwholesome sea might do."

If that's what it takes to get a Scandi High for Christmas, then fine!

Keep in mind that the EC46, although we get to see it only after the 12z EC/EPS has come out, is actually run as an extension of the 0z ensemble. Yesterday morning, the 0z ensemble looked pretty decent in the 10-15 day range, so it's not a complete surprise that the first weeks after that looked good on EC46 last night. However, this morning the ensemble 10-15 days out was not as positive, like you said, Here's the 0z clusters for 264-360h, that came out late today. It's all over the place. Some still good with blocking, but many clusters place it a bit too far for us to profit from, exposing us to SW-lies. Cluster 3 could be an inversion High over the UK. It's not inconceivable that a EC46 run today would have looked diferent as well, although the clusters do contain blocking that would fit in with the anomalies on EC46. We'll just never know. Sometime before summer 2023, a big IFS update is scheduled, with a full ensemble on high resolution and a EC46 forecast each day, but that won't help us much today!

EC46 says blocking will be here for a while... Except for some annoying heights over Europe around Christmas, it keeps blocking to our North and near Greenland well into January. Week 2 Week 3 Week 4 Week 5

New data are gathered (weather stations, satellites, airplanes, etc) every time at 0z, 12z etc. though some data are older or not available every run. Then from all those data, the 0h starting state of the atmosphere (and oceans) is calculated. So the starting point is not a blank canvas. From there, the equations representing the physical process of the atmosphere, oceans, and ice produce every next step. If you would like to learn more about the ECMWF IFS model (EC/EPS), this is recommended reading: 1 Introduction - Forecast User Guide - ECMWF Confluence Wiki CONFLUENCE.ECMWF.INT Original post: https://community.netweather.tv/topic/98087-model-output-discussion-the-cold-has-arrived/?do=findComment&comment=4757332

New data are gathered (weather stations, satellites, airplanes, etc) every time at 0z, 12z etc. though some data are older or not available every run. Then from all those data, the 0h starting state of the atmosphere (and oceans) is calculated. So the starting point is not a blank canvas. From there, the equations representing the physical process of the atmosphere, oceans, and ice produce every next step. If you would like to learn more about the ECMWF IFS model (EC/EPS), this is recommended reading: 1 Introduction - Forecast User Guide - ECMWF Confluence Wiki CONFLUENCE.ECMWF.INT

That same GEM takes many of you from -12c on Friday morning to + 12c Sunday afternoon. At least it's not boring!

GEM will have to spend the evening in the naughty chair, that's for sure.

The EPS 12z output is much colder out to Day 10. The Operational is not a cold outlier, but has solid support now from its ensemble. Excellent news. Here De Bilt, Holland: And London, England: There must be some good clusters in there.

A development that was expected, GFS is making more of renewed Scandi heights. Serious attempt this time at Day 9. Polar Vortex fully gone from the Western hemisphere, cold in situ, Scandi High building: Classic winter developments.

As I looked at the morning runs before coming to this place, I thought: "There will be LOTS of pages to plow through." And indeed. Good times though! It's good that especially the EPS has swung back to what it showed yesterday morning. In fact, today's De Bilt 0z plume is nearly identical to yesterday's 0z. Light frost during the nights and slightly above zero temperatures at daytime at least until Friday the 16th. Possible developments later on have been thoroughly discussed here already, so not much to add to that, but if the cold stays for longer and snow chances remain/return, the sea surface temperature output may be of interest too for all of us near the coasts. ECMWF has made the output of the control run available, and it shows the effect of the cold wave on the waters around us. The North Sea and Norwegian sea are cooling rapidly, reducing positive anomalies and even introducing cold anomalies near the coasts, especially in estuaries and other river outlets. At T+0h and T+360h:

Something for everyone in there, I'd say. All over the place. The majority still has solid high latitude blocking though.

Same in The Bilt. So as far West as Ireland, and as far East as Holland, the uncertainty in the 7-15 day period has only grown, with that big spread. Better sit back and wait.

This morning, the EC HRES was in a smaller cluster up to 168h and in the smallest cluster 192-240h Tonight it has shown us a run from another, larger cluster (cluster 2) in both time frames from the 0z ensemble. There is nothing "new" in this operational run. An operational run that is not the best, but that is still better than most runs from most winters, because it has us still very much in the game for more opportunities, as @CreweCold just mentioned too. Anyway, let's wait until we've seen the 12z ensemble. In these tricky setups the high resolution may very well be picking up and enlarging insignificant detail.

GEM at 240h with a December 2010/1962 inspired chart, with Europe already snow covered:

GFS and UKMO at 144h are very different in the Atlantic, but very similar with regards to Scandi heights rebuilding.

If it were to happen as GFS 18z showed, the 384h setup would even be the starting point of another cycle of retrogression, with the Siberian CAA moving towards cold Europe and WAA setting up in the Atlantic. You could add another 10-14 days of intense cold and snow. What a pub run.

ECM and GEM at 192h show remarkable similarities, considering all that is going on.

True. I wouldn't be staking anything on any solution we see at the moment!