Model output discussion - cold proper on the way?

[kold weather]

1 minute ago, nick sussex said:

Not sure I’m bothered about what the EPS show past day ten because the same ones showed apparently rock solid support for the alleged upcoming cold spell !

I’ve  said it for years ensembles at that range are just as liable to change as the ops , variables unforeseen at one point pop up then the whole cycle of jam tomorrow continues .

 

I'm not sure the ensembles were ever Rock solid apart from maybe a couple of suites 24-36hrs ago.

Even so, I know what you mean for sure!

[Tristrame]

The highs in 2012/13 were more Russian in origin than scandi iirc incarnated completely differently to whats hinted in recent runs so not comparable in my opinion 

Still a lot of uncertainty on the US northeaster with implications for over here so a case of wait and see tbh?

[snowking]

27 minutes ago, bluearmy said:

I think we just lost another week ...........

I fear Nick is right here I must say. When I made this post on Thursday:

I was starting to fear the worst and a situation we have seen a few times over the years where longer range modelling suggests a clean separation of troughing, only for it to find smaller features which it previously did not model as we get closer to T+0, preventing ridging from spreading its wings.

We can of course hope this is just a wild solution from what have been some very volatile ensemble suites of late - we'll see come the 12z. But as we are into the ranges of day 4-5, you would usually be expecting to start to look to deterministic runs to use their superior resolution to decipher the likely direction.

Anyway, this far from a winter is over post - as Nick says, this would essentially mean we lose a week, and given both the overall direction of travel with the persistent NW Euro trough/higher heights towards the N Atlantic, combined with big spikes in MT and GLAAM right now, an amplified pattern with cold spilling out into lower latitudes should be expected to continue right through into February. Its also worth remembering that if modelling of next week does turn out to have been incorrect, it can just as easily be incorrect with the pattern further out too and swing in our favour.

It would just be a real shame if we couldn't get the party started a little bit earlier.

[Ice Day]

Well the ECM is a punch to the solar plexus, no doubt about that.  From a few fairly consistent runs, the models have started spewing out all sorts of output so the old 'Shannon Entropy' is obviously increasing.  Purely anecdotal and not based on anything scientific, but having spent about 1 million hours following winter model runs over the last 10 years, it's notable how frequently you'll get several day's of 'upgrades' then it goes into reverse and along comes the 'downgrades' only for it to perk up again.  We're now on a knife-edge, but we only need relatively small corrections at T72 to greatly improve T120 and so on.  I agree with Blue, even if this first bite of the cherry does not pan out, we should get further opportunities.  Winter 2019 is testing our resolve, stand firm gang!

[MattStoke]

Hmm. Can’t really pretend that that ECM update isn’t disappointing.

UKMO and GFS still look very good, if not as great as previously (how often do the most spectacular runs verify anyway?).

I’d at least see what this afternoon’s update bring before writing off any easterly next week. Even if this one come later to nothing, the long term outlook is pretty high confidence on things remaining rather cold with more chances at significant cold.

[nick sussex]

Just now, kold weather said:

I'm not sure the ensembles were ever Rock solid apart from maybe a couple of suites 24-36hrs ago.

Even so, I know what you mean for sure!

The EPS prove my point this morning . The period 25 to 27 which had solid support for the cold have switched with more support for the op .

Whether they switch back we’ll know by tonight .

[Nick F]

Morning all, we were promised the finest champagne by ECMWF and instead they end up giving us Babycham!

Although I think we could be drifting away from bitterly cold snowy northeasterly now, don't think the 00z EC det. is necessarily correct with the Atlantic resurgence to the NW after mid-week trough disruption and all the other models will fall in line.

I think the models are struggling with the emergence of an arctic high off eastern Greenland next week, probably a trop response to the easterlies downwelling from the lower strat. 00z UKMO has the high dropping down to Scandinavia, EC did last few days, but this morning EC is drifitng the high towards the Barents Sea instead - which doesn't block the Atlantic resurgence from the west like the UKMO solution.

UKMO T+144

EC T+144

So, although a big upset from EC this morning, don't expect it to be right, even at this short range, because of that arctic high.

 

[feb1991blizzard]

3 minutes ago, nick sussex said:

The EPS prove my point this morning . The period 25 to 27 which had solid support for the cold have switched with more support for the op .

Whether they switch back we’ll know by tonight .

They very rarely do, particularly when they've jumped ship recently, although that said, i havent seen them in any detail, just the mean.

[kold weather]

Having looked through some ensemble members the NW that the means are showing does have some bite, could themselves have a marginsl event especially further NW and higher ground.

Also as others have pointed out, really tasty solutions start popping out, just as good as some of the ones the GFS has been rocking for snowfall.

[Met4Cast]

5 minutes ago, feb1991blizzard said:

They very rarely do, particularly when they've jumped ship recently, although that said, i havent seen them in any detail, just the mean.

Really?

The ECM & it’s entire ensembles flipped twice just a few days ago. 

The GEFS have been just as eratic in the lead up to this cold period as well. The ENS have been just as inconsistent as the Det runs have been. 

My thoughts remain unchanged despite the unfavourable ECM/Shift in the EPS. UKMO wasn’t interested & then it was and it’s yet to drop the signal since picking it up. 

Worth noting the quite frankly huge spread in the EPS post day 4 too  

As for the GFS.. well, i’d expect if it doesn’t flip on the 6z run we can be a little more confident, but if it does? Well, then is probably the time to actually be concerned. 

We’ll know this evening either way. 

Edited January 19, 2019 by Daniel Smith

[feb1991blizzard]

Just now, Daniel Smith said:

Really?

The ECM & it’s entire ensembles flipped twice just a few days ago. 

The GEFS have been just as eratic in the lead up to this cold period as well. The ENS have been just as inconsistent as the Det runs have been. 

My thoughts remain unchanged despite the unfavourable ECM/Shift in the EPS. UKMO wasn’t interested & then it was and it’s yet to drop the signal since picking it up. 

As for the GFS.. well, i’d expect if it doesn’t flip on the 6z run we can be a little more confident, but if it does? Well, then is probably the time to actually be concerned. 

We’ll know this evening either way. 

Really?

[Banbury]

29 minutes ago, CreweCold said:

Ensemble mean of the EC at 144 is fairly conclusive I'd have thought, but like others, I hold total judgement until the 12z suite. 

If this is a fail, into Feb we look! 

So your FI starts on D 11 then ?

[nick sussex]

2 minutes ago, feb1991blizzard said:

They very rarely do, particularly when they've jumped ship recently, although that said, i havent seen them in any detail, just the mean.

True the ECM has had two underwhelming runs . We certainly won’t survive a third as the timeframes involved won’t allow for much change after that .

Across the board we have seen a less favourable trend . Even today’s better outputs aren’t as good as last night but it’s whether we can still salvage something good or whether the others will follow the ECM , European Centre of Misery model !

[knocker]

Quite an eye opener with the EPS as it has a strong east Pacific ridge,  a cross Polar flow to the emerging intense vortex and trough that plunges arctic air way south down eastern North America. This clearly seems to impact downstream.

 

Edited January 19, 2019 by knocker

[bluearmy]

9 minutes ago, nick sussex said:

The EPS prove my point this morning . The period 25 to 27 which had solid support for the cold have switched with more support for the op .

Whether they switch back we’ll know by tonight .

The failure of low euro heights sustaining is the issue but that’s a consequence of other things ..... fwiw, the extended eps have never been supportive of high slp to our nw .......the cold spell has always been reliant on low euro heights and probable wedges to keep the jet south ........now there’s change on the horizon ....... the extended eps now see a decent probability of high slp in the griceland area post day 11/12 ...... it’s a step change and gives me some confidence that should we continue to fall the wrong side of things for the next ten days that we will have a real ally for proper cold come the turn of the month .......

[Timmytour]

20 minutes ago, Nick F said:

Morning all, we were promised the finest champagne by ECMWF and instead they end up giving us Babycham!

Although I think we could be drifting away from bitterly cold snowy northeasterly now, don't think the 00z EC det. is necessarily correct with the Atlantic resurgence to the NW after mid-week trough disruption and all the other models will fall in line.

I think the models are struggling with the emergence of an arctic high off eastern Greenland next week, probably a trop response to the easterlies downwelling from the lower strat. 00z UKMO has the high dropping down to Scandinavia, EC did last few days, but this morning EC is drifitng the high towards the Barents Sea instead - which doesn't block the Atlantic resurgence from the west like the UKMO solution.

UKMO T+144

EC T+144

So, although a big upset from EC this morning, don't expect it to be right, even at this short range, because of that arctic high.

 

Nick, is it possible that the ECM was struggling with the emergence of that arctic high and hence giving us the stellar runs of a couple of days ago, and that the more sober runs of late are down to it starting to come to terms with it?

Edited January 19, 2019 by Timmytour
Typos

[Cloud 10]

3 minutes ago, nick sussex said:

the ECM , European Centre of Misery model !

Not much continuity between the 12z and 00z runs,with the latter being about as cheerful as listening to a Morrissey album in a darkened room!

 

 

 

 

[kold weather]

2 minutes ago, bluearmy said:

The failure of low euro heights sustaining is the issue but that’s a consequence of other things ..... fwiw, the extended eps have never been supportive of high slp to our nw .......the cold spell has always been reliant on low euro heights and probable wedges to keep the jet south ........now there’s change on the horizon ....... the extended eps now see a decent probability of high slp in the griceland area post day 11/12 ...... it’s a step change and gives me some confidence that should we continue to fall the wrong side of things for the next ten days that we will have a real ally for proper cold come the turn of the month .......

Yeah I think your right, knowing our luck this winter so far we'll probably end up with rip roaring southerly jet and mild southerlies. There is an amazing chart from the 70s with a huge 965mbs low spread in the subtropics, large Greenland block...and due to the relative position we stay stuck in the mild air from that low for all but the last 24hrs of that blocks life!

I'd imagine there would be some real sadness if that sort of thing happened next month!!

Also good to note that the Greenland/Azores ridging is a signal that is strong both in the ECM and also in the GFS ensemble suite, though that is masked somewhat by the latter having still quite a few strong Scandi blocks which helps to mask that mean a little.

[snow mad]

Just goes to show how far we’ve come. Which is not very, what with all the ensembles, ops and anomaly charts that show something for days only to flip a few days out is almost comical. My goodness this is frustrating. 

[bluearmy]

The time for analysis of what model was right/wrong and where they were right/wrong hasn’t arrived yet cos we don’t know what will verify ...... and by the time we reach that point we will all be ruminating on more crazy week 2 output rather than scientifically analysing  the old nwp .....

that’s why we rely on the verification stats 

[kold weather]

2 minutes ago, snow mad said:

Just goes to show how far we’ve come. Which is not very, what with all the ensembles, ops and anomaly charts that show something for days only to flip a few days out is almost comical. My goodness this is frustrating. 

Yep, I remember a really shocking one back in 2006 which looked like a monster cold spell. We ended up getting a cold spell but it was much reduced. Same thing with Xmas 2005, and of course who could forget Jan 2012, which also did kinda land 'ish' but nothing like what was previously advertised.

15 years of being on here and the same old problems still come up, nice to see that despite clear improvements, the models can still absolutely suck at times

BA- your so right, by the time we get to the 25th I'm sure we will all be over-analysing the next shot forecasted in deep FI, or maybe not so deep FI...

Edited January 19, 2019 by kold weather

[Rocheydub]

Disappointing but not wholly surprising, you could see the pressure drops rearing it’s ugly head in green/Iceland region with a developing shortwave on last nights 18z as early as 86 hrs. It prevented the earlier link up of Atlantic high and arctic high, but was saved by the stronger arctic high. This sort of earlier model confusion has signposted confusion in difficult and unusual NWP. 

 

I feel there is a lot of resolving to go here and we are not done yet. Strange that ECM has had the most volatile flip though.

 

Finally, a lot if ICON bashing going on here. I think it’s handled itself very well in these volatile situations previously. You’d be very foolish to ignore it!

[nick sussex]

A crumb of hope. 

The ECM was discarded from the short term diagnostic discussion for the west USA for the period upto the 22 January .

The preferred solution was a combination of this mornings UKMO/GFS and NAM.

How much that impacts events later still uncertain . 

[Polar Maritime]

NetWX-MR still shows an Easterly Thursday next week.

[snowspotter]

Quick question , when was the last year we had a blocking anticyclone to our north east in January that lasted a reasonable length of time?  I can't remember one!