Model output discussion - winter proper underway

[jvenge]

2 minutes ago, Snowy L said:

Sorry but how accurate can these charts be. I find it difficult to believe 95% of the North Hemisphere will have a milder than average winter. The thing that sticks out for me is both polar and mid lattitude-wise being warmer than average. You can't have both.

You can.

This is because you can have somewhere cold one month, then warm the next. It's over a 3 month period. When people look at these things, especially for temperatures, they just assume that each day of the three month period will be above average in a static way. But that's not how averages come to be. 

You then also need to take into account that the globe is warming and it would actually be quite unusual if large parts did show a cold anomaly over a three month period in regards to a previous 30 year base period. 

 

 

[nytram43]

36 minutes ago, johnholmes said:

Not good news for coldies re the overall winter patterns, looking at the latest computer output from UK Met. What the discussions will be in their Ops room trying to marry this output with what we have had and the possibilities according to some on the longer range models we see would be interesting to be able to eavesdrop!

https://www.metoffice.gov.uk/research/climate/seasonal-to-decadal/gpc-outlooks/glob-seas-prob

How does this prediction compare to the previous 3 months predicted with the same tool. Close, better or worse? How accurate do you believe it to be? 

[Frosty.]

The GEFS 6z mean indicates a north / south split next week with high pressure building north across southern / central uk bringing pleasant benign surface conditions with only more northern parts of the uk being affected by more changeable Atlantic conditions. Temperature wise, an upward trend with less chance of frosts. Further ahead, lowering heights from the northwest with increasingly unsettled weather and temperatures dropping slightly too.

Edited December 12, 2017 by Frosty.

[Paul]

As ever - if you think a post is an issue, report it, don't respond to it please.

[Sky Full]

Five new pages of posts since 10.00 pm last night?  And no significant snow in the forecast?  Something must be afoot....  Ah yes, a spell of milder weather awaits us just before Christmas....

This is how the main models currently depict the scene in six days time (18th December):

UKMO      

ECM         

GFS          

GEM         

Reasonably good agreement that milder Atlantic sourced air will be on our doorstep by the 18th.  Only the GEM appears not to be on the same page regarding the extent of the colder uppers pooling over Europe.  This would be a predictable return to normal or should I say average conditions which are common in the UK at this time of year, and it is not unsurprising that the models are taking this direction.  When the next cold spell is going to show, and how we get to it, is yet to be known but given the great start to this winter I am sure it just waiting around the corner, out of sight!  My own feeling is that early in 2018 we are going to see another snowy event - most of the coldest events in UK weather occur in January after all....

In the meantime, for those enjoying wall to wall snow and ice today, spare a thought for the farmers and their livestock who cannot take a day off and still need to find food so that milk and other farm produce can be in the supermarkets when we look for it.....

 

[bobbydog]

@Tamara hints this morning that some members may be starting to view the models without blinkers on...

obviously never any guarantees but using the 'thinking outside of the box' method, i'm expecting to see an improved version of this- 

at around day 10.

if the models have a climatological bias, they will be playing catch-up as predicted background signals unfold.

it should be stressed that to get to this pattern, we have to transition through a more 'zonal' pattern. the risk of course, is that during this transition, 'zonal' overpowers the meridional flow but with a weak vortex, this hopefully won't happen.

[phil nw.]

I see the latest GFS op runs have now waved the white flag and gone completely over to the dark(milder)side for next week after yesterday going against the others and it's own majority of ens in delaying an Atlantic visit.More energy across the north now being modeled which of course tips the balance.

A much tighter grouping around the Op run today supporting the warm-up.

Meanwhile this week still looks rather cold generally as our cold air remains out to the weekend with rain and hill snow coming through over the next 24hrs and then a polar nw to finish the week.

              

Next week suggests a milder turn generally then with our old friend/foe the Azores high settling in for a spell and that Canadian vortex leaking it's energy east  across Greenland to our north as we lose the Atlantic ridging ,at least for a while.

The return of the Alaskan ridge again showing from day 7/8 but the ext ens suites continuing to show a westerly pattern across the Atlantic. 

[Frosty.]

It appears the GEFS 6z mean could have over egged the high pressure Influence next week. It sounds more generally unsettled (zonal) with spells of wet and windy weather, especially further NW...that applies to the christmas / new year period too with a mixture of mild / cold zonality.

[frosty ground]

6 minutes ago, Frosty. said:

It appears the GEFS 6z mean could have over egged the high pressure Influence next week. It sounds more generally unsettled (zonal) with spells of wet and windy weather, especially further NW...that applies to the christmas / new year period too with a mixture of mild / cold zonality.

What are you basing that on Frosty? You are referring to the Met Office forecast I see, I'm  slight confused on what data you are validating one Forecast other the other considering the time period in question has yet to happen? I am even more confused when you consider only a few updates ago the Mets outlook as well as the other general Model output has been for a mixture of Cold/Mild output.

Edited December 12, 2017 by frosty ground

[Nick F]

1 hour ago, johnholmes said:

Not good news for coldies re the overall winter patterns, looking at the latest computer output from UK Met. What the discussions will be in their Ops room trying to marry this output with what we have had and the possibilities according to some on the longer range models we see would be interesting to be able to eavesdrop!

https://www.metoffice.gov.uk/research/climate/seasonal-to-decadal/gpc-outlooks/glob-seas-prob

Oh John, don't make coldies reach for the Prozac. If it wasn't for the MJO heading into 'colder' phases, as Nick Sussex has kindly reminded us through the CPC update, I think I would be waving the white flag for a return of cold the rest of this month after the expected bout of milder more mobile weather indicated to take over next week.

With the MJO already in phase 7 and expected to move through into phase 8, with the associated area of upper divergence moving east across the tropical Pacific will likely induce further poleward wave propagation over the N Pacific and a return of a strongly +PNA / -EPO ridge later this month, as is being highlighted by the 00z EPS.

Couple this with an east Asian mountain torque event, working with the trop ridging to induce wave breaking over the N Pacific, which will continue to keep the strat PV displaced from the pole to other side over toward N Russia. 

For now, we are probably seeing the lagged effects of the MJO movement of phase 6 as we see the models extend the influence east of the Canadian trop PV and enhance the Azores high toward western Europe as the Euro trough declines eastward, a composite of mean monthly 500mb heights for December in phase 6 produces this:

similar look to what's shown by GEFS mean next week:

However, move on to a phase 7 and 8, remember there's a lag of 10-15 days, we could be looking at a pattern change to more blocked look for later this month.

Ridging signal building to the N and NE with 7 and 8, with troughing over eastern N America and the polar front jet extending east on a more southerly track (-NAO) as a result. This could be the way forward, but, in reality the actual patterns often don't play out like these composites, which at the end of the day, are means over several years when the MJO is in certain phases.

So although the model guidance is all pointing to a more mobile, milder and maybe even zonal outlook as we count the days to Christmas, there is room for optimism given the MJO signal for a pattern change upstream. However, there is the caveat that a pattern change away from the likely increasing influence of the trop PV  over Greenland and Iceland may not transpire until after Xmas. 

As with all longer range predictions using drivers such as the MJO and ENSO, they aren't an exact science, with their influence perhaps more muted this side of the northern hemisphere than over N America, so there is the possibility that we may stick with a more mobile and milder pattern to the end of the year if the trop PV to our NW proves resilient to expected re-amplification upstream from the wave propagation over the N Pacific.

 

Edited December 12, 2017 by Nick F

[Daniel*]

I think we can only get to day 5 with some degree of confidence, despite general agreement amongst the models I wouldn’t say this suggests confidence going forward of the pattern, probably models falling back on the climatological norm understandable perhaps if MJO was just in ‘death orbit’. I don’t feel the ensembles will be of any help too, it will be the OP which sniff out new signals first IMO. Is it outlandish we could get a ridge thrown up and gain some latitude in relatively short term probably, but the result of such would be day and night to what we’re currently seeing by NWP. I for one cannot see a protracted spell of very mild Atlantic gunk, I still feel the Xmas period is up for grabs, we shall see as ever. But I don’t feel we’ll have to wait long, before the model output puts out some festive cheer to feast our eyes on.

[Nick B]

1 hour ago, johnholmes said:

Not good news for coldies re the overall winter patterns, looking at the latest computer output from UK Met. What the discussions will be in their Ops room trying to marry this output with what we have had and the possibilities according to some on the longer range models we see would be interesting to be able to eavesdrop!

https://www.metoffice.gov.uk/research/climate/seasonal-to-decadal/gpc-outlooks/glob-seas-prob

Possibly similar discussions in nature to the ones they were having in November 2009, where I guess the background signals maybe didn't agree locally (well, actually for large chunks of Europe?) with their global probability maps?  All finely balanced and nuanced, I guess... and at the end of the day they have to carry the can for what they say publicly.

We're only covering a small area of the globe.  Their model here deals globally, though of course, effectively small changes to the whole can bring big local impacts, meaning much more to those based here!

Edited December 12, 2017 by Nick B

[Catacol]

2 hours ago, johnholmes said:

Not good news for coldies re the overall winter patterns, looking at the latest computer output from UK Met. What the discussions will be in their Ops room trying to marry this output with what we have had and the possibilities according to some on the longer range models we see would be interesting to be able to eavesdrop!

https://www.metoffice.gov.uk/research/climate/seasonal-to-decadal/gpc-outlooks/glob-seas-prob

The 500hpa probability maps look pretty inconclusive to me. Nothing there of note. 2m temp charts I reckon are a waste of time - model predictions of 2m temps are constantly incorrect at 24 hours range let alone 4 months!

Edited December 12, 2017 by Catacol

[Glacier Point]

About those long range model forecasts..

Both CFS and GPC heading for major bust on December (Nov and Dec for GPC / GLOSEA5) in the North Atlantic.

Ridges and troughs in completely the wrong places, particularly CFS for December.

Pertinent as both models have similar forecasts for January-February periods. Maybe not a tropical lead signal ?

[Man With Beard]

2 hours ago, johnholmes said:

Not good news for coldies re the overall winter patterns, looking at the latest computer output from UK Met. What the discussions will be in their Ops room trying to marry this output with what we have had and the possibilities according to some on the longer range models we see would be interesting to be able to eavesdrop!

https://www.metoffice.gov.uk/research/climate/seasonal-to-decadal/gpc-outlooks/glob-seas-prob

Thanks John, and I think this one is worth remembering when looking back at the winter - predicting a good chance of a low pressure dominated winter, especially further north:

[bobbydog]

5 minutes ago, Man With Beard said:

Thanks John, and I think this one is worth remembering when looking back at the winter - predicting a good chance of a low pressure dominated winter, especially further north:

especially if it turns out to be wrong- as apparently, it is already!....

[bluearmy]

Just a correction on the terminology being used (I am also guilty of referring to the next amplification as pos PNA renewal) )

we have been seeing a positive PNA pattern with ridge in western USA/Canada 

once this flattens, we then see a renewal of the ridge under a negative EPO (east Pacific oscialltion) which places the mean ridge a bit further west and into Alaska 

 

Edited December 12, 2017 by bluearmy

[Frosty.]

Just been casting the net out into the outer limits of the GEFS 6z and this caught my eye..looks very cold, hopefully a sign of things to come during late Dec and through January, the prime time of winter!

Edited December 12, 2017 by Frosty.

[johnholmes]

2 hours ago, nytram43 said:

How does this prediction compare to the previous 3 months predicted with the same tool. Close, better or worse? How accurate do you believe it to be? 

about the same as the previous 2 months outputs showed.

They got the overall idea of higher pressure but milder than average last winter, I think!

[Methuselah]

Right then. For all of us looking to see if the models are starting to pick-up on the long range teleconnections and what have you, the next few GFS outpourings ought to start showing 'signs', as the post-Christmas period creeps into view...?

PS: And, no, I don't mean -12C uppers over London before the New Year!

[johnholmes]

re the Met O model seasonal outputs.

I have no axe to grind one way or the other, but I do like to be as constructive as possible about forecasts.

The link below allows anyone to see these predictions back a long time, certainly to this time last year, I've not looked beyond it.

https://www.metoffice.gov.uk/research/climate/seasonal-to-decadal/gpc-outlooks/glob-seas-prob

My professional days are long gone and meteorology and models have moved on way beyond anything I dealt with so I do not feel able to comment.

One thing, mentioned by jvenge, I think, is that average or above average does not mean the same every day. So an above average forecast of 2 m temperatures can be correct over a 3 month period but still allow a cold spell much as we are having. For instance my own weather data shows an T of 0.1 C for the last 4 days, the average for the first 12 days is 4.4 C. I hope helps explain how quite large variations in averages/values over differing periods can occur.

My forecasting is limited to 10-14 days, sometimes I'll go out to 20 days. Nothing specific but using what I consider are the necessary tools to provide an overview in the 5-20 day period in the upper air. For shorter times scales then the 3 main models and my own experience as a forecaster.

Sorry to clutter up the thread but I hope this post helps some of you.

Edited December 12, 2017 by johnholmes
to add link

[nick sussex]

Looking at the latest MJO forecasts, disagreement between the GEFS and ECM, we've been in phase 7 then a brief entry into the COD.

The ECM brings the signal out quickly into phase 7, the GEFS into phase 6 taking a few extra days to go into 7. However the GEFS has a large increase in amplitude in phase 7. The ECM a slight increase.

The EMM which includes climatology has a big increase in amplitude in phase 7.

For those interested we have a further MJO update out this evening from the Global Tropics Hazards outlook which goes into more detail.

For newer members the amplitude of the signal is important:

That's the GEFS so the further the signal is away from that inner circle the stronger it is. There are other factors that can mute out the MJO but you have a much better chance of a NH pattern change with a stronger signal.

PS The COD which you often see referred to stands for circle of death! basically the signal isn't doing anything and won't have any impact.

 

Edited December 12, 2017 by nick sussex

[bluearmy]

Might be an interesting day for ecm to follow icon 

[feb1991blizzard]

7 minutes ago, bluearmy said:

Might be an interesting day for ecm to follow icon 

Its good but I still feel it would flatten out again though every little helps I suppose.

[bluearmy]

9 minutes ago, feb1991blizzard said:

Its good but I still feel it would flatten out again though every little helps I suppose.

From small acorns ............

There were sig clusters showing an upper ridge stretched ne towards Scandi