Model Output Discussion - moving into Winter

[Trom]

1 hour ago, Met4Cast said:

Perhaps worth applying a bit of context to the models as we look at them (obsess over them)

The GFS has been quite poor recently with the ECM & to some extent the UKMO leading the way, in fact even the GEM has been performing better than the GFS at times. I believe this is based on the now retired operational run and not the newly updated one, but if a chart posted a few days ago re: new GFS accuracy, it's marginally worse (though probably negligible difference over all).

It's worth noting a couple of points regarding verification stats.

1.  They are for the entire northern hemisphere.  The polar areas and the equator are far more consistent over time and as a result the models have very high verification stats for these areas (especially the continent).  The mid latitude areas are far more changeable and the verification stats are far lower.  There's a huge variation in model performance in different areas of the northern hemisphere.  Catacol posted some really good data on this last January.

2. Not only are we at mid latitude we are also on the boundary of the Atlantic and continental Europe which makes our weather even more variable and difficult to forecast.

So verification stats for the UK would be far far lower than those plots suggest.

I found Catacols post from last year and have copied it below (the second image is really enlightening)

Here's the link to Catacol's original post 

 

Edited December 1, 2022 by Trom

[ICE COLD]

Well something must be brewing because even at the end of the gfs 6z run the gfs manages to slide a low and keep its well south ( and not do it’s usual tricks and blow the low up ) . Which then slides off and brings in an even colder northerly. What a run

[Eagle Eye]

29 minutes ago, blizzard81 said:

Whatever you lot say on here, don't ever say that the next set of runs are 'crucial'. To some it will feel like the proverbial red hot poker has been inserted up their rear ends

I hate to say it

But the next set of runs...

Really don't matter unless we see a trend and/or the models converge on one solution.

Remember this every time in the future...

The weather is finely balanced and small margins come up with big differences the further you go on so unless the models are trending towards one or two solutions don't get hung up on what one model may be showing unless that one model is trending towards that solution in which case it may be picking up something early or may just be a coincidence.

Right now, the model trend is the best it's been in December since 2010 (probably not as good as 2010 though) in terms of cold trends so despite the odd on and off model run, we're heading towards the most interesting period (for me at least) since I've joined here.

[bluearmy]

Looking at the gefs day 11.5, approx half of the runs have lifted the cold out or are about to in a not so great way. That’s not dissimilar to the 00z but the overall lw pattern is a tad more suppressed south which is better for the U.K. 

now to see if the suite reverts colder again as the 00z gefs did (not dissimilar to the 06z op I guess) 

tbh, that op was a reasonable representation of the most likely extended direction of travel for now. The Atlantic making an attempt to get in but being forced to undercut by the cold trough sat across our latitude. 
 

if we can’t get the cold trough sat far enough south (which will provide ample opportunity for lowland snowfalls) then the next best option is the battleground high risk/high reward of the Atlantic attempting to get in. 

Edited December 1, 2022 by Blessed Weather
Unnecessary comment removed.

[Methuselah]

8 minutes ago, Eagle Eye said:

I hate to say it

But the next set of runs...

Really don't matter unless we see a trend and/or the models converge on one solution.

Remember this every time in the future...

The weather is finely balanced and small margins come up with big differences the further you go on so unless the models are trending towards one or two solutions don't get hung up on what one model may be showing unless that one model is trending towards that solution in which case it may be picking up something early or may just be a coincidence.

Right now, the model trend is the best it's been in December since 2010 (probably not as good as 2010 though) in terms of cold trends so despite the odd on and off model run, we're heading towards the most interesting period (for me at least) since I've joined here.

True, EE... And, all things said and done, the weather will simply do whatever it 'chooses' to do...

[Connor Bailey Degnan]

Just to calm the nerves, the ecm had the occasional wobble in the run up to Dec 2010. Overall though, it was a clean route to cold. I fully expect the 12z ecm to come back on board. 

[bluearmy]

1 minute ago, Connor Bailey Degnan said:

Just to calm the nerves, the ecm had the occasional wobble in the run up to Dec 2010. Overall though, it was a clean route to cold. I fully expect the 12z ecm to come back on board. 

Totally agree Connor 

I think the ec op had two consecutive wobbles on the BFTE a few years ago. there is no certainty with anything and even if we get close, trying to pin snowfall down is another problem! (Nice one to have though) 

[ItNeverSnowsInWindsor]

Whatever happens this month, I think we are well overdue on special wintry charts actually verifying this winter.

The last snowfall over a couple of inches here in Windsor was back in 2013!!

[minus10]

Apologies if this has been posted already...currently out working but just grabbed look at gfes 6z 850s..

id say tht is a pretty good run if you want cold...of interest is the fairly substantial cluster supporting the op below the mean..

....however finely balanced though..

Hopefully cold block strong enough to resist warm attacks from the south...

[Geordiesnow]

I don't think a northerly flow of some kind is in any huge doubt, its just whether we will get enough cold air far enough southwards for snowfall. My main concern is will the flow be strong enough to help get the coldest air down. Of course there is the threats of a west based NAO but it will be hugely dissapointing if we go through all these wonderful snyotopics and the air just has not been quite cold enough. 

[TSNWK]

9 minutes ago, minus10 said:

Apologies if this has been posted already...currently out working but just grabbed look at gfes 6z 850s..

id say tht is a pretty good run if you want cold...of interest is the fairly substantial cluster supporting the op below the mean..

....however finely balanced though..

Hopefully cold block strong enough to resist warm attacks from the south...

Nice spaghetti plot. I was hoping the 06z was amongst the pack and it is.   

[Methuselah]

It looks like we have decent GEFS temp. ensembles: plenty cold enough for snowfall in Cambridgeshire, but more likely sleety stuff over here by the North Sea, I would imagine... We shall see:

    

[Snowman.]

1 hour ago, danm said:

I was about to say. In 15 years or more of model watching, this is textbook. Until you get into a reliable timeframe, i.e within about 96-120hrs, the models will inevitably churn out variations on the general theme. We can't expect any one model to countdown from t200 to t0 with the same, clean evolution. There will be different outcomes within an envelope that get shown. This morning's ECM was on the less good side of that envelope for cold and snow lovers - but by no means terrible - and this afternoon's may show it swing back again. The only time to worry is when you see less good/poor scenarios repeat over multiple model runs and between different model suites. 

Last Christmas had changes within 60hrs or so. 

When you see it outside your window, thats when you know its reliable.

[johnholmes]

3 hours ago, mushymanrob said:

Glad you posted that John, because it supports my understanding of what these charts show. And ive posted as much - "A Wintry Mix" is the most likely, but with the potential of a proper freeze still there.

I agree with that mushy, thanks, a bit uncertain as to just how cold in the upper air, any minor troughs etc for 6-10 days and beyond. Would love to be in the operations main forecast room by the senior met forecaster over the next few days

[MP-R]

1 hour ago, Downburst said:

There are archive charts out there somewhere, WetterZentral only goes to 1979, someone will know Reanalysis archives (wetterzentrale.de)

 

Search to your heart's content... right back to 1836.

https://www.meteociel.fr/modeles/archives/archives.php

 

Edited December 1, 2022 by MP-R

[SteveB]

35 minutes ago, minus10 said:

Apologies if this has been posted already...currently out working but just grabbed look at gfes 6z 850s..

id say tht is a pretty good run if you want cold...of interest is the fairly substantial cluster supporting the op below the mean..

....however finely balanced though..

Hopefully cold block strong enough to resist warm attacks from the south...

As a few have alluded to. Since the dawn of model watching, we have never had cold in situ and the classic battleground situation with fronts coming in of the Atlantic. In the late 70's and early 80's I was older enough to remember driving to see my grandparents who also lived in Weston- super-mare, and in the short 10min journey, it rained then started to turn sleety. 

When we left my granparents, 30mins later it was snowing heavily.

I suspect we are seeing something similar in the models for next week. This will a  good test of whether we can still get those frontal events bumping into the block, or it simple pushes on through.

[Guest]

It's a win win situation.

The cold and snow arrives which is fun, or it's milder and the bills are lower.

As someone else mentioned the ECM is illustrating a run within its envelope is all, one thing is for certain it won't be the same this evening.

[Paul_1978]

Before the 12's start to trickle through later this afternoon, I think it's worth remembering that even in the best winters things ebb and flow. You're never going to get wall to wall cold for weeks on end. So the charts today might start to look unfavourable, or they might continue the cold theme. Either way it's 1st December today and winter is only just starting, there is a long way to go. You just have to look back at the archive charts and run through some of the so called "classic" winters and you'll always have milder spells in and amongst the cold periods.

Overall things look good at the moment if you are of the cold persuasion, and for those of us who have been on here for years there's no doubt this is classic model watching. Enjoy it and don't take it too seriously

Edited December 1, 2022 by Paul_1978

[That ECM]

People keep saying that there’s a lot of ramping going on. How I fully expect to watch my lad at football next weekend.

[Vikos]

According to background signals, we will have a cold phase for a certain amount of time. So I really don’t worry about single deterministic runs as they like to wobble a lot… even if EC is a mild run tonight, winter and snow is coming. You just can’t await it at your doorstep, when you live in flatlands. 

[Robbie Garrett]

IMHO,  UKMO are very much on the fence till the cold is probably knocking on the door.  Can't imagine being that person who then has to call HMG for a Cobra meeting and distribute the message of yet another crisis.    How far will the cold remain?  Our patterns have been locked in? The rain we've just had started in September, the warm spell before that, about May/June.    What if the cold pattern locks in till February?   The NHS is already overwhelmed, especially in rural locations,  the motorways network isn't in a great condition, the railway isn't what it was in 2010, and the airports are barely breaking even vs (2019) - do they need another crisis of non-ops with SNOCLO?

Tuesday was a COBRA meeting about Winter, and I wouldn't put it past another happening this coming week.

45 minutes ago, johnholmes said:

I agree with that mushy, thanks, a bit uncertain as to just how cold in the upper air, any minor troughs etc for 6-10 days and beyond. Would love to be in the operations main forecast room by the senior met forecaster over the next few days

Edited December 1, 2022 by Robbie Garrett

[blizzard81]

1 hour ago, Eagle Eye said:

I hate to say it

But the next set of runs...

Really don't matter unless we see a trend and/or the models converge on one solution.

Remember this every time in the future...

The weather is finely balanced and small margins come up with big differences the further you go on so unless the models are trending towards one or two solutions don't get hung up on what one model may be showing unless that one model is trending towards that solution in which case it may be picking up something early or may just be a coincidence.

Right now, the model trend is the best it's been in December since 2010 (probably not as good as 2010 though) in terms of cold trends so despite the odd on and off model run, we're heading towards the most interesting period (for me at least) since I've joined here.

Wise words. However, many on here (myself included) see the weather (particularly wintry weather) as not only a hobby but also (like many hobbies) it's a passion. Now if we translate the historical and original meaning of the word 'passion', you will now know why many post like they do on here

[Jordan S]

17 minutes ago, Vikos said:

According to background signals, we will have a cold phase for a certain amount of time. So I really don’t worry about single deterministic runs as they like to wobble a lot… even if EC is a mild run tonight, winter and snow is coming. You just can’t await it at your doorstep, when you live in flatlands. 

ECM wasn't mild though, just a less cold but still wintry alternative. Probably will show an evolution more towards the GFS later though for the second week and what it was showing yesterday with less "mixing" of the uppers with those lows remaining further south, until close to mid month anyway, plenty of showers falling as sleet and snow though by this time, even in the south probably with perhaps more emphasis on sleet/snow falling in any precipitation here from showers/ disturbances too compared with more a rain/snow mix as I thought couple days ago, especially high ground, though not exclusively.  Also of note is more persistent and significant snow chance from weather systems coming up from the southwest now looking more likely to take place over western and southern areas of the UK at first with the Greenland high very stubborn and extending close to Scotland keeping the cold entrenched over the country/ with winds from the northeast, before the lows edge slowly further north taking the snow threat north close to mid month as said previously. Basically again rather similar to previous post but heightened snow risk from change from the southwest across southern areas. Previously thought this may be a high ground thing in the south, due to uppers possibly more likely rising ahead of the fronts but I think that's now less likely currently but that's my view.

[Eagle Eye]

5 minutes ago, blizzard81 said:

Wise words. However, many on here (myself included) see the weather (particularly wintry weather) as not only a hobby but also (like many hobbies) it's a passion. Now if we translate the historical and original meaning of the word 'passion', you will now know why many post like they do on here

I try and mainly focus on the teleconnection patterns as I know that the basic overall pattern is better until we get within the reliable time frame which for me is about 168 hours but I still succumb to the passion sometimes. 

[Liam Burge]

Love how we are calling runs that don't show the -5 or -8c isotherm getting into the country 'mild'