Model output discussion - 7th January onwards

[Chesil View]

20 minutes ago, nick sussex said:

 No just telling it like it is ! I don’t do spin . Dads Army you know I’m far too young to have seen that ! I wish !

I know Nick. I always enjoy the reasoning you put into your posts and you are right to point out the pitfalls that could scupper things. It's just that most of your recent posts did seemed to fit into the 'we,re all doomed' category. Lol

[markw2680]

Everybody is so obsessed about 7-10 days away all the time, yes I’d like to no how cold and will it snow too but nobody or no model will have a clue exactly how cold or snowy that far out so it really dosnt matter if it’s slightly better or slightly worse than the last run. We all no this but sometimes reading stuff in here is as if we think it’s nailed regardless.

people who come on here that really haven’t a clue and are just trying to learn must be very confused indeed from one to another 

[Stuie]

UKMO, is it prepping itself to send a ridge up to join the HP party?

 

Edited January 12, 2019 by Stuie W

[knocker]

11 minutes ago, bluearmy said:

ec loses the plot at day 8 in the atlantic

no point in looking at it after day 7

if it’s anywhere near right then the eps and other ec ens tools are a complete waste of time and have been for weeks .......(chuck in all other global forecasting tools ) 

the noon run is important to see what it does at days 7/8 with the next train out of the Canadian vortex station ..... I woud have expected it to follow closer to icon 

that T300 eps mean from yesterday is a little more marked a day later at T276

I’m noticing a trend today for the diving systems to back further west as week 2 progresses

the larger feature (as currently modelled) around day 11/12) could well end up dropping down the western side of the uk and not introduce a ‘warm sector’ at all to much of the UK 

 

Yes that is looking quite important vis sustainability of the Atlantic ridge.The FV3 take on this this morning

[winterof79]

17 minutes ago, northwestsnow said:

GFS 00z para also an absolute snorter !!

Its cold pretty much all the way through ..a snapshot in FI..(a very snowy Fi i might add)

Very very nice.

[Ice Day]

I feel that the mist is starting to clear a little and the pathway of how this upcoming cold spell is going to play out is becoming clearer.  Based on the valuable input from some of our more learned members, I'm anticipating several attacks from the north-west which will deliver some fun and games for those in the north but not much down south.  Around days 10-12 we will get an arctic outbreak with a northerly, veering north-easterly (and maybe eventually easterly).  Some outrageous cherry-picking, but I've picked this perturbation out of the GEFS as to what I think the eventual solution will be for day 11.

What's particularly striking though, is that whatever pans out over the next 10-14 days, this is likely to be just the start of a protracted period of cold weather, possibly lasting right through February.  I'll be very surprised if we don't all see a notable snow event, maybe several, before February is out.

Whatever happens, most on here just love the thrill of the chase and I am confident that the next few days all the models will be showing x-rated charts...... just need to get those down to T0.

 

[Frosty.]

Largely delightful 00z output..even better than this time yesterday!..cold spell continues to firm up with a risk of snow!

[BLAST FROM THE PAST]

This has been and is a slowburner alright and way, way slower than I thought, interesting ECM....I mused on winter 18/19 thread a week ago of possible way forward. Again it’s slower but around 20-24 I suggessted a Scandi HP to develop but to be transitory as the main PV displacement would then dive down over us.

Looking at the 00z varying themes, varying speeds and cold depths.....but ALL heading to cold and very cold.  Not often we can say that. 

If the downwelling result lasts as long as the downwelling itself....then a 20th century great could be matched....all to play for

 

BFTP 

Edited January 12, 2019 by BLAST FROM THE PAST

[Steve C]

32 minutes ago, nick sussex said:

 No just telling it like it is ! I don’t do spin . Dads Army you know I’m far too young to have seen that ! I wish !

And long may it continue! I won't pretend I understand the dynamics of your analyses of upstream patterns but I accept them. 

I like your reasoned postings a lot more than so many of the hyperbolic types that infest this place in the Winter and will seek them out on my visits here. Obviously there are other superb posters here as well mind.

If a person took at face value the number of day 6 etc. changes to / continuations of cold spells that are posted here and those events happened, the Winter of 1963 wouldn't stand a chance!

[Ali1977]

20 minutes ago, Stuie W said:

UKMO, is it prepping itself to send a ridge up to join the HP party?

 

I hope you’re right as that would dramatically change the way forward 

Edited January 12, 2019 by Ali1977

[Methuselah]

00Z for Day 7: It's getting closer!

[WhiteFox]

1 hour ago, karlos1983 said:

Looks like a cold outlier to me Ali

It's not that different from previous day's where the mean was around -3 to -8 towards the end.

I did look at that final frame of ECM this morning and thought it was one for @nicksussex! A threading the eye of the needle easterly with lots of shortwaves!

Looking at the models this morning it's still clear that looking for detail past seven days is futile. ECM and GEFS means still showing averages around -5 from around the 19th with a bit more scatter than yesterday, admittedly, but signals are all still there, and certainly not worth worrying about an uptick in ensemble means at the 14 day range; we're not even in the colder pattern yet!

[SMU]

Nice little T120 upgrade from the ICON 06z

More chance of a back edge snow event by 108-114

[shaky]

2 minutes ago, Steve Murr said:

Nice little T120 upgrade from the ICON 06z

More chance of a back edge snow event by 108-114

Yup and looks more like ukmo at 120 hours if not better!!low looks like.sliding aswell!!

[Nick F]

Increasing spread for the mslp over NW Europe, far N and NW Atlantic towards the end of 00z EC deterministic, which suggests low confidence in this area past day 7. 

The differences between 00z EC and GFS ops in mslp and more reliable z500 flow configuration quite stark in this area next weekend. So a lot to be resolved here and not worth looking in too much depth and getting hung up on details past end of week.

Edited January 12, 2019 by Nick F

[bluearmy]

To those worrying about an uptick in uppers at any point - once into a cold troughy spell, we will generally have weather systems attempting to push the trough away or join the broad trough - either way, uppers would be expected to rise somewhat - that wouldn’t necessarily translate to higher surface temps and could be seen to be representative of moisture injection required to boost snowfall ??

[karlos1983]

4 minutes ago, WhiteFox said:

It's not that different from previous day's where the mean was around -3 to -8 towards the end.

I did look at that final frame of ECM this morning and thought it was one for @nicksussex! A threading the eye of the needle easterly with lots of shortwaves!

Looking at the models this morning it's still clear that looking for detail past seven days is futile. ECM and GEFS means still showing averages around -5 from around the 19th with a bit more scatter than yesterday, admittedly, but signals are all still there, and certainly not worth worrying about an uptick in ensemble means at the 14 day range; we're not even in the colder pattern yet!

Trust me, I’ve been studying the ens religiously 4 times a day since November! This mornings short ens are a disappointment on the past couple of days, but nothing to be worried about yet. The ext ens will show more detail later.

[Duane S.]

All looks good in the hood this morning. Standard Op variances. Absolutely no trend backwards. Just the models trying to resolve an outcome from a particularly complicated set of variables.

The Arctic high may look innocuous, meandering around largely over the other side of the pole, but firstly it’s longevity and behaviour is the single biggest clue at the 500mb Level that something above is not well (the SPV). And secondly, it is the key that will eventually unlock winter 2018/19. We are not there yet, it’s a slow burner for sure but on the flip side, once in place, unlikely to be a flash in the pan. 

I still think we are looking at the US getting a severe spell of snow and cold, possibly record breaking, think early’85. Yes, this could work against us but with this current setup think, rather than fire up the northern jet, more a slower ‘bleed’ NW to SE. Ingredients are certainly here for slider utopia...

It’s always worth going back to see how the models handled impending cold spells as well. Case in point, the Jan 2013 wedge, which of course ended up giving us a very decent spell of winter weather. Very interesting how it handled it in the lead up and during. Similarities for sure.

Edited January 12, 2019 by s4lancia

[WhiteFox]

41 minutes ago, bluearmy said:

ec loses the plot at day 8 in the atlantic

no point in looking at it after day 7

if it’s anywhere near right then the eps and other ec ens tools are a complete waste of time and have been for weeks .......(chuck in all other global forecasting tools ) 

 

I know it's a turn of phrase, and this is not targeted at you (some posters seem to believe it!) but models don't "lose the plot"; they predict outcomes based on imperfect input and run it against hundreds of thousands of equations. The fact that the ECM produces a particular outcome for a given input is not unexpected and that is why we have ensembles with slightly tweaked input.

Of course, as you correctly say, it's not worth focusing on the single run; as we progress further from the initiation date, the greater the level of chaos becomes, especially in situations where we have large scale background drivers about which we have limited knowledge and information. 

Surprised nobody has mentioned Shannon Entropy for a while, but it seems to me that there is greater uncertainty (chaos) right now than there would be in a more familiar setup.

My focus is still on ensemble means and background drivers for the next few days. I think 21st is the date to watch for the colder pattern, but don't expect any reliable synoptic SLP detail until at least 4-5 days before....

[Met4Cast]

Sorry guys, coming on board yesterday ruined everything!! 

This mornings EPS certainly a step away from High Lat blocking, ideally we’d want the Atlantic block much further East. 

That is a strong signal for a European trough however so mild it wont be, below average for the foreseeable with snow risks from slider lows (primarily in the North) 

As Nick alluded to above, a lot of spread across NW Europe so little point in getting too hooked on what models show post day 5 

[Summer Sun]

The 00z ECM mean in the extended range doesn't get the daytime high much above 3c on the London graph

The ENS mean shows 2cm of snow in the extended range on the London graph

[Man With Beard]

ECM ensembles 00Z - by 24th Jan, 32 members clearly cold enough for widespread snow, 12 not cold enough and 7 unclear. So, pretty consistent with yesterday for that timeframe. 

I get the concerns of some but, realistically, it isn't very often such a strong signal is very wrong on the ECM ensembles. Yes it happens, but 1 in 5, maybe even 1 in 10 at best. Been watching them a while now . The most likely error (if it could be called that) is the euro trough stays closer to the south, and only the north gets the freeze. But that's a way off yet. 

It starts to get quite messy by 27th, generally cold still the front runner but potential breakdowns (even if just temporary) look more possible. Of course low pressure coming up against an already cold situation with no conclusive breakthrough = ...? 

(clue: something white! ) 

Edited January 12, 2019 by Man With Beard

[Tim Bland]

3 minutes ago, Daniel Smith said:

Sorry guys, coming on board yesterday ruined everything!! 

This mornings EPS certainly a step away from High Lat blocking, ideally we’d want the Atlantic block much further East. 

That is a strong signal for a European trough however so mild it wont be, below average for the foreseeable with snow risks from slider lows (primarily in the North) 

As Nick alluded to above, a lot of spread across NW Europe so little point in getting too hooked on what models show post day 5 

A euro trough and higher than average pressure to our NW. not sure what’s not to like? It’s not going to be record breaking cold but if the lows continually slide we could be looking at some record snow falls! I’d rather have 0c and snow than -5c and dry.

[winterof79]

32 minutes ago, bluearmy said:

To those worrying about an uptick in uppers at any point - once into a cold troughy spell, we will generally have weather systems attempting to push the trough away or join the broad trough - either way, uppers would be expected to rise somewhat - that wouldn’t necessarily translate to higher surface temps and could be seen to be representative of moisture injection required to boost snowfall ??

I think the output is erring towards us being under a flabby low system mixing with embedded cold air to start this cold spell. Think it happened in 2009 where streamers came off the Irish Sea up here as winds alternated in direction.

February 2009 loads of snow

Edited January 12, 2019 by winterof79

[bluearmy]

17 minutes ago, WhiteFox said:

I know it's a turn of phrase, and this is not targeted at you (some posters seem to believe it!) but models don't "lose the plot"; they predict outcomes based on imperfect input and run it against hundreds of thousands of equations. The fact that the ECM produces a particular outcome for a given input is not unexpected and that is why we have ensembles with slightly tweaked input.

Of course, as you correctly say, it's not worth focusing on the single run; as we progress further from the initiation date, the greater the level of chaos becomes, especially in situations where we have large scale background drivers about which we have limited knowledge and information. 

Surprised nobody has mentioned Shannon Entropy for a while, but it seems to me that there is greater uncertainty (chaos) right now than there would be in a more familiar setup.

My focus is still on ensemble means and background drivers for the next few days. I think 21st is the date to watch for the colder pattern, but don't expect any reliable synoptic SLP detail until at least 4-5 days before....

‘Loses the plot ‘ = goes against all known model guidance outside of its reliable period (both current and over previous days) rather than shows something we aren’t keen to see (as is sometimes posted on here )

i stand by my original post white fox  ......

Edited January 12, 2019 by bluearmy