Model output discussion - is the beast awakening?

[Mark wheeler]

Navgem 6z for the southeast , Squeaky bum time .

[danm]

It does make me laugh that the yo yo-ing of emotions with every model run never stops on here. Just reading through the thread from this morning we've gone from elation at the 0z to some semi-despair from a few at the 6z. 

Until we get into the 72-96hrs time frame, the advice has always been to not get too excited at the eye candy and not too downbeat at a few downgrades, unless they snowball into a new theme of consistent downgrades. There will continue being variations on the theme for the next few days until we zero in on the outcome. 

Great runs on the 0z, but we're still 6 days away from the start of the action, plenty of time for things to go wrong, especially when it comes to easterlies, and also plenty of time for wobbles and rogue runs. The trick is not getting too excited when runs this far out show cold nirvana and not too downbeat if we see run to run variation. If we zero in on the cold/snowy outcome at 72-96hrs then start getting excited and by all means become as despondent as you like if it all gets snatched away from us at that timeframe. Until then, sit back and expect a rollercoaster ride, but don't overreact to any runs - good or bad. 

Edited February 1, 2021 by danm

[cidersnow]

5 minutes ago, PolarWarsaw said:

I'd just like to make a very obvious point, but one that's useful.

When reading this thread, especially this week - it's important to check the location of the user and what they are posting. 

The 6z GFS is a perfectly good run for 90% of the country, it just gets more marginal further south and takes longer for the coldest air to get down to the very far S/SE - you will note the negativity mostly stems from those users in this thread. Much like, on the 0z I'd rather take my chances on an Atlantic incursion, because, for most of the W and NW - it's bone dry and very boring next week, whilst the SE gets in on some streamer action. It's all so fickle and the IMBY nature of things.

As for the set up, it's knife edge, perilous, could go wrong with even one false step and the margins are even finer than virtually ANY other cold set up.

Unlike some cold spells, There's a good 3/4 days to go before this is nailed down, the modelling of the low will be in 'current time' - when a feature such as that is so decisive in giving or taking a cold spell away - FI is down to 72 hours MAX. This could all go wrong or right on Wednesday evening. 

Can people please, not, under any circumstance declare a freeze 'nailed' at Day 7. For 280th time this winter as it just leads to more toys out of the pram and an unreadable thread a few hours later when a computer generated forecast that's never actually come to fruition takes away some snowflakes at the window. 

Morning all,

Of course this post makes good sense but for me misses the essence of this forum and particularly the MOD thread when we are in sight (or think we are) of exceptional cold and snow. That is the amusement that can be gained form watching the irrationality of it all unfolding before your eyes. Intelligent people dealing with a the most analytical aspects of a complex and often chaotic subject. The clash of science and emotions. Head and heart. I love it. 

Our willingness to commit to the models in unrealistic timeframes. Chopping and changing between loving and hating them is what this game is all about.

Russ

[Blessed Weather]

33 minutes ago, TSNWK said:

I've never taken notice of Parallel.  Simply because it's in test and we don't to my knowledge have visibility of what testing conditions it's running ubder. For example it could be a tweaked ensemble member?

I think we can safely assume that NOAA/NCEP would only put GFS Version 16 out into the public domain if they thought it better than the current version of GFS. Here's what they say:

"......the next GFS (v16) upgrade, which will include doubled vertical resolution (64 to 127 layers), more advanced physics, data assimilation system upgrades, and coupling to a NCEP's Global Wave Model using the Unified Forecast System (UFS) community model. Implementation of GFSv16 is targeted for the early winter of 2021."

Source: https://vlab.ncep.noaa.gov/web/environmental-modeling-center/unified-forecast-system

[offerman]

Just now, danm said:

It does make me laugh that the yo yo-ing of emotions with every model run never stops on here. Just reading through the thread from this morning we've gone from elation at the 0z to some semi-despair from a few at the 6z. 

Until we get into the 72-96hrs time frame, the advice has always been to not get too excited at the eye candy and not too downbeat at a few downgrades, unless they snowball into a new theme of consistent downgrades. There will continue being variations on the theme for the next few days until we zero in on the outcome. 

Great runs on the 0z, but we're still 5 days away from the start of the action, plenty of time for things to go wrong, especially when it comes to easterlies, and also plenty of time for wobbles and rogue runs. The trick is not getting too excited when runs this far out show cold nirvana and not too downbeat if we see run to run variation. If we zero in on the cold/snowy outcome at 72-96hrs then start getting excited and by all means become as despondent as you like if it all gets snatched away from us at that timeframe. Until then, sit back and expect a rollercoaster ride, but don't overreact to any runs - good or bad. 

Best post of the week. 
The charts are so inconsistent and especially further out where more often than not they downgrade cold potential . 

just focus on the within five days for more reliability even with this the charts can be flippant . 
 

good luck everyone who wants cold snow s as I do. 

[MATTWOLVES 3]

4 minutes ago, PolarWarsaw said:

I'd just like to make a very obvious point, but one that's useful.

When reading this thread, especially this week - it's important to check the location of the user and what they are posting. 

The 6z GFS is a perfectly good run for 90% of the country, it just gets more marginal further south and takes longer for the coldest air to get down to the very far S/SE - you will note the negativity mostly stems from those users in this thread. Much like, on the 0z I'd rather take my chances on an Atlantic incursion, because, for most of the W and NW - it's bone dry and very boring next week, whilst the SE gets in on some streamer action. It's all so fickle and the IMBY nature of things.

As for the set up, it's knife edge, perilous, could go wrong with even one false step and the margins are even finer than virtually ANY other cold set up.

Unlike some cold spells, There's a good 3/4 days to go before this is nailed down, the modelling of the low will be in 'current time' - when a feature such as that is so decisive in giving or taking a cold spell away - FI is down to 72 hours MAX. This could all go wrong or right on Wednesday evening. 

Can people please, not, under any circumstance declare a freeze 'nailed' at Day 7. For 280th time this winter as it just leads to more toys out of the pram and an unreadable thread a few hours later when a computer generated forecast that's never actually come to fruition takes away some snowflakes at the window. 

Mate you really have hit the nail on the head with this post. People will call a cold snap a bust unless its delivering snow to there area..thats a fact! We see it in the Summer thread when large areas of the UK are cloudy and 18C,and we here the moans about how awful it is.Yet it can be 25C and sunny in the far South and South East,and these guys will be telling us how decent the conditions are. This Estly can make plenty of convection for Eastern parts and perhaps the SE with a NE tilt to the wind,some of this convection could spread inland,how far is always up for debate..could be a disturbance in the flow also at some point. But we need too play a high risk high reward game here,namely the fact that we need Atlantic systems approaching from the West or South West to inject some moisture into the cold air,then we really could be looking at significant snow events for many. This is a possibility moving forward,but if this does occur there's always a chance the milder air will dislodge the colder air,on the other hand we could see any fronts grinding to a halt and not really making much progress inwards..Alot to think about here.

[SLEETY]

Too much overreaction to one gfs run as usual which is the worst performing models in the verification stats. 

Always stick to the Euros in the reliable timeframe, so unless they follow the gfs, which is doubtful, then this gfs run won't verify as shown, its at the top of the pack as well, or we do just ignore that, like some are. 

[offerman]

15 minutes ago, PolarWarsaw said:

I'd just like to make a very obvious point, but one that's useful.

When reading this thread, especially this week - it's important to check the location of the user and what they are posting. 

The 6z GFS is a perfectly good run for 90% of the country, it just gets more marginal further south and takes longer for the coldest air to get down to the very far S/SE - you will note the negativity mostly stems from those users in this thread. Much like, on the 0z I'd rather take my chances on an Atlantic incursion, because, for most of the W and NW - it's bone dry and very boring next week, whilst the SE gets in on some streamer action. It's all so fickle and the IMBY nature of things effects what kind of opinion people have.

As for the entire set up, it's knife edge, perilous, could go wrong with even one false step and the margins are even finer than virtually ANY other cold set up. It's still 60-40 AGAINST something like that 18z coming off. Still only 55-45 in favour of cold uppers next weekend. How is this nailed? 

Unlike some cold spells, There's a good 3/4 days to go before this is nailed down, the modelling of the low will be in 'current time' - when a feature such as that is so decisive in giving or taking a cold spell away - FI is down to 72 hours MAX. This could all go wrong or right on Wednesday evening. 

Can people please, not, under any circumstance declare a freeze 'nailed' at Day 7, incorrectly for the 280th time this winter. It just leads to more toys out of the pram and an unreadable thread a few hours later when a computer generated forecast that's never actually come to fruition takes away some snowflakes at the window. 

It's painful and brings 'that side' of the population out, you know the role your eyes at gang. 

Fantastic post Warsaw, 

Especially your 2nd to last paragraph . 
just no point is too far out .

 

 You’re bang on about the 280th time this winter . 
 

more often than not the charts downgrade when cold snow potential  is involved . 

[TSNWK]

7 minutes ago, Blessed Weather said:

I think we can safely assume that NOAA/NCEP would only put GFS Version 16 out into the public domain if they thought it better than the current version of GFS. Here's what they say:

"......the next GFS (v16) upgrade, which will include doubled vertical resolution (64 to 127 layers), more advanced physics, data assimilation system upgrades, and coupling to a NCEP's Global Wave Model using the Unified Forecast System (UFS) community model. Implementation of GFSv16 is targeted for the early winter of 2021."

Source: https://vlab.ncep.noaa.gov/web/environmental-modeling-center/unified-forecast-system

Thank you.  That helps  

[Tim Bland]

I’m not sure the disappointment at the 06z is just from those in the south. I can’t see much, if any snow away from Scotland on that run Surely it’s a poor run for the whole of England and Wales compared to other runs perhaps a dusting in the far north ...

[Notty]

25 minutes ago, Blessed Weather said:

I think we can safely assume that NOAA/NCEP would only put GFS Version 16 out into the public domain if they thought it better than the current version of GFS. Here's what they say:

"......the next GFS (v16) upgrade, which will include doubled vertical resolution (64 to 127 layers), more advanced physics, data assimilation system upgrades, and coupling to a NCEP's Global Wave Model using the Unified Forecast System (UFS) community model. Implementation of GFSv16 is targeted for the early winter of 2021."

Source: https://vlab.ncep.noaa.gov/web/environmental-modeling-center/unified-forecast-system

It was originally scheduled for implementation this spring so has been pushed back to later this year unless I am mistaken.

[IDO]

25 minutes ago, Blessed Weather said:

I think we can safely assume that NOAA/NCEP would only put GFS Version 16 out into the public domain if they thought it better than the current version of GFS. Here's what they say:

"......the next GFS (v16) upgrade, which will include doubled vertical resolution (64 to 127 layers), more advanced physics, data assimilation system upgrades, and coupling to a NCEP's Global Wave Model using the Unified Forecast System (UFS) community model. Implementation of GFSv16 is targeted for the early winter of 2021."

Source: https://vlab.ncep.noaa.gov/web/environmental-modeling-center/unified-forecast-system

I said last month after some really poor verification runs that the then time scale for implementation, which was Feb, was unlikely. You cannot have statistical failures that the gfsP was showing and it clearly needs some further programming. 

It is def more accurate in benign conditions due to its improvements in resolution and data. That is taken as given and should not influence when it arrives, it is whether it performs outside standard deviation and how far below it falls? It clearly has issues with modelling heights and hopefully they can resolve these by the end of the year!

[frosty ground]

5 minutes ago, Tim Bland said:

I’m not sure the disappointment at the 06z is just from those in the south. I can’t see much, if any snow away from Scotland on that run Surely it’s a poor run for the whole of England and Wales compared to other runs perhaps a dusting in the far north ...

Fantasy snow depth charts thou

[Froze were the Days]

15 minutes ago, SLEETY said:

Too much overreaction to one gfs run as usual which is the worst performing models in the verification stats. 

Always stick to the Euros in the reliable timeframe, so unless they follow the gfs, which is doubtful, then this gfs run won't verify as shown, its at the top of the pack as well, or we do just ignore that, like some are. 

Could see a slight 'coming together' of the scenarios between GFS/ECM which could be a probable outcome and at a guess low pressure close to the south east near day 6/7 and maybe clearing eastwards a bit like the UKMO. Who knows? or that GFS 06z could be a rogue run, but it does have a habit of being consistent bringing stella op runs and then dropping them to a degree i.e. Greenland High episode before mid-January...FI in terms of an easterly particularly if you live towards the south is really about 72z where you see the key differences between the 0z and 06z (that nuisance low needs to be further west of the UK and slightly shallower than what the 06z shows).

[Trom]

1 hour ago, D.V.R said:

The Gfsp looks like it's slightly better guys but I'm going get my toys ready and warm up my shoulder joint incase the afternoon runs flop like the GFS 6z

The reality is that even the good runs have milder air not too far from the South.  So I expect it will be a yoyo ride for the SE run to run depending on the speed and track of the low pressure feature.  Further north and this is less of an issue.

[Stu_London]

My biggest concern is that the 2018 BFTE cold spell was cut short in the south with a similar feature. When first modelled nearly everyone on here discounted it and some even said that the evolution was impossible.

Now we have a scenario where the cold spell is being stopped before it has begun with something very similar. It is not without support in the ensembles. Might be a blip this time, but if it does gain any traction its curtains I would guess, for the south at least.

Edited February 1, 2021 by Stu_London

[PolarWarsaw]

7 minutes ago, Tim Bland said:

I’m not sure the disappointment at the 06z is just from those in the south. I can’t see much, if any snow away from Scotland on that run Surely it’s a poor run for the whole of England and Wales compared to other runs perhaps a dusting in the far north ...

Only disappointing if you took make believe computer generated images with crayon like colourings on them at Day 7/8+ onwards as a 1/100 banker, or even anything like that really. Screenshot them and see how they compare to every single set of 12z. 

The point still stands though regarding IMBYism and it being a massive issue at the moment. It will continue to be so in virtually every cold event. 

If you asked 100 people in this thread how winter went, 60 may say 'good' - 40 south of the M4 would say it's been horrific, the ones that moan will always be seen 'louder' though which would be a proper 'skew' against what, for the larger majority of the country has been a snowy winter so far. 

 

[Tim Bland]

Just now, frosty ground said:

Fantasy snow depth charts thou

I don’t think the run is correct. My point is that some were saying it’s not a bad run for the north and it’s just the southerners that are moaning. Fact is this run showed no snow for anyone away from Scotland where as the 00z run the UK was covered from top to toe.  I’m sure the 06z GFS is nutmeat a wobble and we will be back on track this evening. Problem is we have so many runs to view there is always going to be one or two that wobble 

[northwestsnow]

Excellent 00z runs.

Now we are approaching the nitty gritty ..

Too complicated for where it will snow as we hit weekend, let's get this weeble of a shortwave dropping south first.. 

[SLEETY]

gfs always does this pick out an easterly in fl then when it starts getting into the reliable timeframe starts to wobble,then gets back on track.

The euros are normally the safest bet upto 144 hours so,hopefully they improve on this morning runs.

[Valleyboy]

1 hour ago, Battleground Snow said:

This is still a seriously good cluster for London.

There are going to be ebbs and flows over the next 24 hours which always happens in these situations. 

I fully expect the met update today to add very cold back into the text.

The macro picture looks set over the next 7 days, micro details are going to change.

I missed this earlier in the excitement, but that is some deep cold over the states and Canada next week.

That could fire up the Atlantic, which might be bad news for the UK.

[northwestsnow]

2 minutes ago, SLEETY said:

gfs always does this pick out an easterly in fl then when it starts getting into the reliable timeframe starts to wobble,then gets back on track.

The euros are normally the safest bet upto 144 hours so,hopefully they improve on this morning runs.

Yes, as I said above we are getting to the nitty gritty now with this shortwave although I'm not entirely sure the drama is quite over with that little critter.

As time ticks on we should get a clearer picture so hopefully the southern brigade will do OK.

Certainly there is scope for the SE to get pounded with snow.

 

[Stephen W]

5 minutes ago, Tim Bland said:

I don’t think the run is correct. My point is that some were saying it’s not a bad run for the north and it’s just the southerners that are moaning. Fact is this run showed no snow for anyone away from Scotland where as the 00z run the UK was covered from top to toe.  I’m sure the 06z GFS is nutmeat a wobble and we will be back on track this evening. Problem is we have so many runs to view there is always going to be one or two that wobble 

I don’t think this can be discounted at all.  To me the whole ensemble suite looks less favourable than the 00z.  It’s the low pressure that’s causing the issue and it’s failure to sink that’s preventing the cold from spilling south.  The ECM has been showing this scenario within its op for sometime so to me it could be that the GFS is finally picking up this signal too.  That isn’t good a good sign IMO.  Obviously only one run but not without support.

[Man With Beard]

Time to remember how the GFS often works.

D6-D10: Stick resolutely to a solution, while other models fluster

D6: Other models start to move a little to the GFS and firm up

D5: GFS jumps ship

D3: GFS returns 50% of the way back. We now have the final solution.

[Duane S.]

I'm still laughing at comments on Saturday (as I did at the time) saying it should all be resolved by Sunday!

I would say though, don't discount the 06z OP. I don't think it's implausable at all, just, taking everything into consideration, it's pretty unlikely IMO. Certainly sat right on the bad end of a, still, very wide envelope.

Woth remember that when the GFS does turn out to be correct in these sort of situations (as in the earlier, much better, output), it VERY rarely doesn't have a big wobble around the D5 period. And I do find it often handles these nearby cut off lows better than the ECM for some reason. Keep the faith.

As if to emphasise this ongoing uncertainty, Para 06z goes off on yet another route.