snowking

Just to put this into context, this means that the entirety of the ECMWF Ens suite will be higher in horizontal resolution than any operational run from any other medium range suite. And we now get access to the lot. There's more to it than just horizontal resolution but in effect we'll have an additional 50 deterministic runs to mull over. Weather nerds assemble.

We have been here so many times before, and for such a tiny landmass it's amazing what a difference just 50 miles north or south can make to the mood in here (geography dependent) At this range, still 5 or so days away, we should still be scouring the ensembles and taking a consensus from the widest possible range of data available to us (within reason...we shouldn't be desperate enough to resort to the NAVGEM ensembles yet). MOGREPS is worth following here for such a reason, but all mainstream ensemble sets should be consulted. If we take the example of MOGREPS here: When we see a spread from the Scottish Borders down to the north of France, we know that we're still a little while away from a resolution. It's probably also worth dropping in that, in my recent memory at least, MOGREPS-G does tend to be a bit of a follower rather than a leader, so if you start to see some continuity in MOGREPS which is consistent with other suites, you know you're probably honing in on a finalised track - and we're not there yet. Through the course the of the weekend, especially once we get to Sunday, we should start to switch our attention further from ensemble suites and more towards the higher resolution runs, which should in theory at least have a better handle on things by this stage - this includes set and control runs as well as higher resolution suites which start to come into focus such as the UKV, but until at least Sunday, such outputs should not be taken entirely at face value. All of this said, we've already seen one of numerous examples from over the years in this very winter of how even at T+0 models can struggle to get a grip on likely placement of such features - back in December, snowfall ended up some 50 miles further north than originally forecast. Historically, certainly prior to the middle of the last decade, I would have said that on balance these systems do tend to end up further south than modelled. However, whether the cause is improved modelling physic, a continuing change in our climate, or a combination of the two, since this point I have seen probably as many of these episodes end up further north as have ended up further south. We probably will not be much the wiser until Tuesday in all reality, and even then we could theorise that the time of year renders some potential for disappointing marginality in events. Until then, try and keep sane. Original post: https://community.netweather.tv/topic/98450-model-output-discussion-a-cold-start-to-spring/?do=findComment&comment=4816138

We have been here so many times before, and for such a tiny landmass it's amazing what a difference just 50 miles north or south can make to the mood in here (geography dependent) At this range, still 5 or so days away, we should still be scouring the ensembles and taking a consensus from the widest possible range of data available to us (within reason...we shouldn't be desperate enough to resort to the NAVGEM ensembles yet). MOGREPS is worth following here for such a reason, but all mainstream ensemble sets should be consulted. If we take the example of MOGREPS here: When we see a spread from the Scottish Borders down to the north of France, we know that we're still a little while away from a resolution. It's probably also worth dropping in that, in my recent memory at least, MOGREPS-G does tend to be a bit of a follower rather than a leader, so if you start to see some continuity in MOGREPS which is consistent with other suites, you know you're probably honing in on a finalised track - and we're not there yet. Through the course the of the weekend, especially once we get to Sunday, we should start to switch our attention further from ensemble suites and more towards the higher resolution runs, which should in theory at least have a better handle on things by this stage - this includes set and control runs as well as higher resolution suites which start to come into focus such as the UKV, but until at least Sunday, such outputs should not be taken entirely at face value. All of this said, we've already seen one of numerous examples from over the years in this very winter of how even at T+0 models can struggle to get a grip on likely placement of such features - back in December, snowfall ended up some 50 miles further north than originally forecast. Historically, certainly prior to the middle of the last decade, I would have said that on balance these systems do tend to end up further south than modelled. However, whether the cause is improved modelling physic, a continuing change in our climate, or a combination of the two, since this point I have seen probably as many of these episodes end up further north as have ended up further south. We probably will not be much the wiser until Tuesday in all reality, and even then we could theorise that the time of year renders some potential for disappointing marginality in events. Until then, try and keep sane.

If you're interested in a really decent cold spell, given the time of the season, the most important thing over the next 10-14 days is to get a really decent cold pool to drop down into Europe to give us something to tap into should the cards fall in our favour - with the stratosphere seemingly imprinting down into the troposphere (arguably a little MJO help is at hand here too) and the vortex being shunted across to the Eastern hemisphere from our perspective that shouldn't be too much to ask for. If we can't overcome that crucial step then we risk ending up with a February 2005 type scenario - it was nice to see some snow falling but something synoptically very promising ended up delivering very little. We've got a shed load of tickets for the lottery, now we just need a little luck.

Always worth remembering when monitoring the stratosphere that temperature is just one element of what we're looking to shift. When looking at heights too: Not quite obliterated, but a huge Wave 1 pushes the vortex off of it's usual axis across to the Siberian sector and gives it a very good stretching too. Battered and bruised at this stage, and we can also see from looking at the wind charts that a technical SSW has taken place (and a pretty significant one at that): All in all very good signs again - vortex moved completely off-axis in a useful direction at 10mb at least and zonal wind speeds significantly reduced (and in fact here reversed). However as always trying to figure out how this translates to conditions in the troposphere is a fun game for the coming weeks. Original post: https://community.netweather.tv/topic/98326-model-output-discussion-into-february/?do=findComment&comment=4801502

Always worth remembering when monitoring the stratosphere that temperature is just one element of what we're looking to shift. When looking at heights too: Not quite obliterated, but a huge Wave 1 pushes the vortex off of it's usual axis across to the Siberian sector and gives it a very good stretching too. Battered and bruised at this stage, and we can also see from looking at the wind charts that a technical SSW has taken place (and a pretty significant one at that): All in all very good signs again - vortex moved completely off-axis in a useful direction at 10mb at least and zonal wind speeds significantly reduced (and in fact here reversed). However as always trying to figure out how this translates to conditions in the troposphere is a fun game for the coming weeks.

What a thing of beauty to watch unfold. Relentless pressure on the stratospheric vortex. Wave 2, followed by a huge Wave 1, whilst flirting with the idea of a return to Wave 2. All still way out there by NWP standards, however if this were to bear any sort of reality then the PV is well beyond just being on the ropes. A fascinating watch upcoming, with something of a wintry flavour in the shorter term to keep us satisfied while we watch. Original post: https://community.netweather.tv/topic/98196-model-output-discussion-new-year-and-beyond/?do=findComment&comment=4788477

What a thing of beauty to watch unfold. Relentless pressure on the stratospheric vortex. Wave 2, followed by a huge Wave 1, whilst flirting with the idea of a return to Wave 2. All still way out there by NWP standards, however if this were to bear any sort of reality then the PV is well beyond just being on the ropes. A fascinating watch upcoming, with something of a wintry flavour in the shorter term to keep us satisfied while we watch.

Always good to check the ensemble mean output at such range (again, crying out here for ECM data on this to become available freely somewhere soon). https://ftp.cpc.ncep.noaa.gov/htlee/StratoMonitorPolar_GEFS_v2.0/TMP_10mb_fcst_nh.html You can see here fairly impressive continuity between the deterministic run and the ensemble mean (note: this data seems to run a couple of days in arrears, hopefully we’ll see the more recent days data shortly) on the location of the warming wave at 10mb. This works its way down to 100mb fairly cleanly too towards days 15 and 16: https://ftp.cpc.ncep.noaa.gov/htlee/StratoMonitorPolar_GEFS_v2.0/TMP_100mb_fcst_nh.html We then see the first signs of some wave 1 forcing coming from the Pacific sector of the NH at this time too, most prominently at 10mb: https://ftp.cpc.ncep.noaa.gov/htlee/StratoMonitorPolar_GEFS_v2.0/HGT_10mb_fcst_nh.html Even at 50mb at this timeframe though we see some imprinting on heights from wave 1, though note at this level the initial pressure seems to be forecast more from Canada than the Pacific https://ftp.cpc.ncep.noaa.gov/htlee/StratoMonitorPolar_GEFS_v2.0/HGT_50mb_fcst_nh.html Which is a good prompt for my point here. Firstly it’s worth throwing in the usual 2 caveats of 1) this is a considerable distance away still, and 2) this is the GEFS suite, confirmation from European suites would be much welcomed. However, assuming any activity from above imprints itself on the troposphere (which is far from a given even at this stage of the winter this year it seems), Wave 1 is always a gamble for the UK. What we can sometimes see is a wave from the pacific side pushing the core of the stratospheric vortex over to the Atlantic sector, which when imprinted into the troposphere results in a net gain in jet stream strength in our part of the northern hemisphere. Always worth reiterating this - such is the very complex nature of stratospheric impacts on the troposphere. However we should be able to take heart that even in the single forecast snapshots from above, just moving from 10mb down to 50mb we see a rotation in the angle of the wave activity from closer to the Aleutians around to Canada. This goes to show that rarely do we see a direct downwards propagation of the heights pattern imprint through the layers of our atmosphere, instead we tend to see some slight rotation of the pattern in its journey downwards. This means that even if we make the massive assumption that the GEFS is modelling this in any way correctly, at this stage it is pretty fruitless to speculate on where any such downwelling wave would end up imprinting itself in the troposphere (which probably makes this entire post pointless - merry Christmas all eh?) So for now very much a watching brief, but interest at least partially piqued stratospherically for me

I must say that whilst this is the one year where if everything goes proverbials up I'm actually okay with it against the background of rising energy costs, and I am always wary of scenarios reliant on Greenland blocking when primarily being peddled most strongly by ECMWF output until we're within a much closer range to T+0, the imagery being produced by NWP right now is simply beautiful. We've seen a lot of dream synoptics already laid out in 2D, how about a peek at what an internal wave break over Greenland does to the tropospheric and lower stratospheric vortex in 3D: Very easy to identify where a cross-polar flow finds its way through there.

Well what a deflating experience this has all been if you are of the cold persuasion. I'm still left mourning for the cold, dry Christmas that looked really quite likely before all of the shenanigans over the last 7 days or so. Any time there is what we in the cold wishing community would call a "failure" (and there's quite a few of those to go around), it's always worth sitting back, taking stock of what happened, and seeing if there's any lessons which can be learned. Here's a few I've noted so far, really interested to see if anyone else has any others to contribute 1. There is absolutely nothing guaranteed about the synoptic setup until we get it to within 72 hours. This most recent episode probably ranks up there with December 2012, although in fairness back then pretty much all modelling was solidly in agreement for a few runs before a slow climbdown model-by-model. Which brings me on to the next point... 2. Wait for consistency. I can't recall a single day in the last 7 days where we have ended the day knowing any more about the likely pattern around the UK by 25th December than we did at the start of the day, owing to fairly large divergence for the timescales involved between different suites. This kept the alarm bells ringing for me, but it's always difficult to try not to seek the coldest route. 3. Unless the EPS and high-res ECMWF are solidly on board, remain cautious. I think those of us who have been on here long enough to remember December 2012 still have PTSD (Promised Torrential Snow Disorder) when it comes to the ECMWF as it was the last one to jump off board the cold train. However, there is a reason both the operational and ensemble suites continue to consistently be the number 1 verifying global forecasting model that we have access to, and we really should ignore them at our peril. 4. Keep monitoring the ensemble suites. When looking beyond 5 days it really isn't good for the health of your sanity to keep riding the frame-by-frame rollercoaster of the operational models trudging out when they could be complete outsiders or even outliers against the ensemble suite. Be more @bluearmy on a Saturday, switch off the alarm and have a bit of a lay in, come back and assess things at 9am once the full set of data from the 0z runs is available. 5. Longwave pattern-affecting dynamics such as the MJO and increasing/decreasing angular momentum within the atmosphere are a global-scale driver, and should be treated as such. We can produce anomaly analogues for various in association with other factors such as ENSO and solar cycle states based on previous analysis, but I think these are often either at least misinterpreted, or at worst completely misused. We need to bear in mind that the landmass of the UK is a little bit under 0.05% of the total surface area of the Earth, and therefore drawing any definitive conclusions from global-scale drivers for the UK weather is fraught with danger. That's not to say they shouldn't be used to give a general view of the long wave pattern potential. It also needs to be remembered that such analogues tend to be anomaly-based - a positive anomaly of a very negative base state still potentially leaves a net negative state, just a slightly weaker one. 5. Don't let the chase for snow be your everything. Certainly when I was younger my mood could be completely affected by either a successful or a failed snow chase. Now I'm an old git I am far more laid back about this stuff. At the end of the day, it's just a bit of weather guys. Some of the sniping, the proclamations of "I told you so" etc. have been quite frankly embarrassing for a forum predominantly full I would assume of fully grown adults, when in all honesty nobody has got anything "right" or "wrong" in this latest episode of model drama. we've all just been taken for a ride. Whilst "The Real Snowchasers of The Netweather Forums" I'm sure would be a thoroughly entertaining reality show (if you're in to that sort of thing), the script I've read so far looks appalling (which to be fair might be because that's how I find all reality TV - and again the grumpy old git part). Finally, I think it's fair to say that with the possible similarity to December 2012, this spell has been pretty unique in the level of uncertainty and the level of naughtiness from the models. Whilst there are probably some interesting discussions to be had as to whether climate change might lead to more of this uncertainty in the future, I hope these sort of occurrences do not become too frequent. I hope this recent spell hasn't dampened spirits too much, and I doubt the above will prevent frustrations again even during the course of the rest of this winter, but I hope for any newer members who might be wondering why nobody seems to have had a clue what will happen this time around these points of reflection offer a little help for the future.

I was simply trying to lighten the mood a little in what has been a tetchy thread in the last few days, whilst pointing out that the phenomena you have referred to in the original tweet would neither impact the UK directly nor the surface conditions for us, so apologies if you felt that I was being particularly dismissive or sarcastic (I am incredibly sarcastic in general as my wife would attest to), but as for " you think stratospheric warming doesn't effect our climate", given I am usually a heavy contributor to the stratosphere thread and have been for many years, I'm not sure this quite holds true. It's also fair to say that what Mika was referring to in his tweet is nothing to do with stratospheric warming, it is simply an interesting phenomenon in the tropopause he was pointing out in tandem with some research from a similar episode in North America back in 2019, something quite separate from both the tropospheric and stratospheric polar vortices. Incidentally, and this applies to the wider thread too, it is perfectly acceptable to not be an expert about something in life in general, never mind something as complicated as atmospheric physics, for which the overwhelming majority of us on here are rank amateurs. If we could all just admit at times that we are not sure about a topic, or even that we got something wrong, and asked for help rather than trying to throw out random terminology in an attempt to sound enlightened about a particular topic, then attacking members who disagree or point out that something might not be correct, maybe the mood in here would be a bit friendlier with a bit less "member slinging" going on than has been witnessed today. Just my opinion and happy for anyone to disagree.

It sure does if you're planning a trip up to the tropopause - at about 30,000ish feet Even Kasim might struggle to see the individual snowflakes up that high

A few of notes on MOGREPS 1. The version of MOGREPS us mere mortals are now fortunate enough to have visibility of is MOGREPS-G - this is the global version of the ensemble suite that the Met Office run. They also run a UK-Specific version of MOGREPS which they often refer to when trying to make forecasts in uncertain situations. There's an enormous difference between the two versions - MOGREPS-G has a horizontal resolution of 20km, compared to the MOGREPS-UK model which has a variable resolution of ~2.2km. Meanwhile the ECMWF EPS have 18km in horizontal resolution and the GEFS are 25km. The difference is even more stark in vertical resolution, with MOGREPS-G I believe currently only offering 70 vertical layers, compared to 137 for the EPS and I believe now that they're using the FV3 core, 127 layers for the GEFS. 2. MOGREPS is just another forecasting suite, and as you can see above, only really comparable to other ensemble suites in terms of horizontal resolution and far weaker in terms of vertical layers. This makes it just as fallible, or perhaps even more so, as all of the other models out there. So to conclude, if the rest of the NWP output is clambering around in the dark trying to find a solution, MOGREPS-G is just as likely to be doing the same, hence the matched wild swings. We should not view MOGREPS-G as superior to any other ensemble suite, in fact quite the opposite really. Nevertheless, I always welcome the addition of new and different data for comparison purposes. Now if we could get our hands on MOGREPS-UK that might be a different matter, but then owing to its higher resolution and purpose, that only goes out to T+120.

@Tamara do you happen to have a link to the GSDM dataset now? The previous one I had bookmarked has long since expired and I have been unable to find where they are hosted now ever since. Cheers!

The thicker blue line is the control run - I don’t think there is an operational run for the MOGREPS suite as it’s designed entirely to be an ensemble suite In this instance that means the Control run is the only run for which the “starting conditions” are not changed in any way - I’ve no idea whether it’s any higher resolution than the other peturbations but I would guess not

Just to throw some more chaos in… MOGREPS sticking with the milder bump during Xmas eve/day, but fairly firmly behind a colder picture thereafter, even in the south in the majority of cases There’s some way to go on this yet

In far too many years of staring at the ECM op, I can only remember 1 infamous occasion where it was wrong vs all the others at t+120…that said, this is still relatively different from previous runs, so keen to get t+120 down to t+72 on Monday before coming out from behind the sofa Great stuff though. Suspicion is that the increase in AAM is slowly being fed in to the modelling of the jet stream coming off of the eastern seaboard, resulting in the gradual slowing of the initial Atlantic low and allowing more time for both our high to the north and the trough to the north-east to eat their share of mince pies in the run up to the big day and bloat over us, allowing the colder air to creep just that little further south. Complete guesstimate here, but in about 80% of cases I can remember, these lows from the west end up going South of the UK, often because initial modelling has over-deepened them vs the eventual reality. Let’s see where this sits in the ensembles, I would guess somewhere towards the bottom of the spread

~15.15 Icon rolls out, doesn't show what any coldie wants it to, gets disregarded as cannon fodder ~16.00 GFS & UKMO rolls out, GFS op shows a full blown Easterly, is declared the "Greatest Forecast System" ever assembled, cold spell declared, Daily Express starts preparing the Caps Lock key for a battering. UKMO not on board but rightfully ignored. ~16.45 GFS ensembles are analysed, nobody as a clue what is going on ~17.00 Jules finds 4 bad ensemble members, has a moan ~18.00 ECM starts to roll out, everyone looks to Sheiky for a "steady as she goes at T+0" post ~18.45 ECM sits somewhere in the middle of the envelope ~20.00 Everyone asks Bluearmy and Mulzy what the clusters are showing ~20.15 One of them dashes all of our dreams as the clusters prove inconclusive ~22.00 GFS 18z rolls out, is completely different to 12z ~04.00 the overnight crew takes over, are usually downbeat. Rinse, repeat.

Interesting Dan. Hopefully a good test of that coming up then!

Yeah that's a good point and we have to be mindful that the ECM in particular has been prone to exaggerating heights near Greenland in the mid term for quite a few years now (to the point that when it gets one right, it's normally just stopped clock syndrome). Then again, certainly when thinking back to Jan/Mar 2013 we also saw underestimation of the resilience of the Griceland wedge in the medium term across most models, so all in all, clear as mud

As noted by @Singularity and others, I too was pretty disappointed to lose what looked like a fairly solid prospect for a cold, potentially frosty high on Christmas Day, and will be particularly irritated if it is to be replaced by chilly, rainy conditions instead. Current operational runs are just throwing out all of the ensemble possibilities run after run, which is to be expected given the current global setup. It's going to be early next week I suspect before the envelope starts to narrow enough for us to make an accurate call - given the fine margins involved here and the myriad results for the synoptic setup over the UK, you really don't want to make a call on this until we're in to at least T+120, given that even at that stage the shortwave energy from the ever-broadening NW European trough can still throw up surprises, both pleasant and unpleasant. Taking that in to account alongside the tendency from most modelling, particularly the GFS over the years, to underestimate the stubbornness of dense, cold air masses and we're in a real mess at the moment. I would re-emphasise that there is very little point in riding the rollercoaster of each op run before we've had a chance to see the accompanying ensemble suite, at least this side of the weekend. From Monday onwards is when you would really be looking to the higher resolution operational/control runs to help resolve the micro detail around the UK's locale. Just fingers crossed we're still in the lottery draw by that stage.

I think the only thing we can say with certainty right now is there is too much uncertainty to be certain. Given the relatively wild swings over the past 24 hours in ensemble suites (and I emphasise the word relatively there - relative to what we would usually expect from ensembles) it would be wise to take a step back and wait for the picture to settle down somewhat - as pointed out already by a few on here, suspicion is the pacific typhoon is wreaking havoc with the longwave pattern projection right now. If we can see some consistency across ensemble suites and models over the next 24-48 hours in setting out this battleground type scenario then we could be in for something really quite exciting. High risk though. are you feeling lucky punk?

Also worth adding to this of course that the EC46 is based off of the 0z run, and so if the 12z suites have sniffed a “spoiler” from a cold perspective, then the EC46 would theoretically already be out of date. Worth re-iterating Nick’s point here to all at this stage that given we’re really looking at Day 8+ even for the building blocks, ensemble suites are the ones to pay attention to, not individual deterministic runs, if anyone would like to hold on to their sanity that is…which is probably contradictory to the traditional sentiment of this thread.

High risk, potentially high reward stuff on the horizon here. Lots of indicators pointing in the direction of something colder than average for the UK - as Catacol has mentioned there is the potential for the upcoming pattern to drive an EAMT event, which in turn feeds back into the stratosphere further down the line to hopefully prevent anything from above coming in with any real strength to spoil the party. So for now, all over to the troposphere to lead things - MJO looking fairly favourable for continued chances of high pressure at the sort of latitudes we want to see it at. I saw somebody mention 2013 earlier in the thread. Feels like a fairly good comparison to me. Wedges ~ Sledges.