Model output discussion - heading into 2018

[BLAST FROM THE PAST]

In keeping with my lrf thoughts and ramblings in winter hopes thread.

A block of sorts to be to our NE is possibly in the mix...  Building off the back of what I believe will be the next countrywide winter event.  (Countrywide as localised wintry weather remains in place).  Period 5/7-12th Jan wild jetstream action to see a displaced piece of PV plunge over the UK as it then moves SSE direction (almost S) bringing strong and very cold N, then NE and then E winds. (Unusual angle could be due to forming ridge/block to NE )

I feel however it won’t develop to a protracted beast from the east.   Cold to ease mid month until last week of Jan (maybe last few days) when from around 28th to 12th Feb for winter’s coldest shot, through similar scenario. 

How I’ve been seeing and still am seeing it and with definitely hints now from models as this is in mid FI range. Been a cracking first 3rd of winter so far imho

 

BFTP

[phil nw.]

Not bad output run for coldies so far.

In spite of the early minor differences with the UK model GFS redeems itself and heads along the ens route of digging a trough se into w.Europe from around day 8.This brings the cold uppers back south and would support the lower temperatures showing in the ens for week 2 .

Snowfall possibilities appear again as Atlantic systems move in as we have seen already this month.Not the ideal deep cold setup but considering the low heights that seem hard to shift  to the north west,not a bad setup for Winter lovers with the angle of the jet from the north west then heading south east.

[feb1991blizzard]

GEFS suite a real mess with no coherent FI signal so confidence low, BUT - its packed with runners, sliders, sinking trough lee Easterlies, a few convective Easterlies (or likely ones) and not a lot of runs that stay mild throughout 192-384.

[SMU]

An interesting debate - in which FWIW I would add my points-

The Forum has has been around for 13 years - & with that teleconnections have really come to the forefront of the discussion for the last 5 years - They we’re always there however as time has progressed so more & more measures / drivers have been identified.

The most crucial part of the understanding how it’s all morphed together ( which is still in its infacy ) in terms of

* What teleconnection is a driver / cause & what is an effect - 

* Then once established what drivers carry the most weighting with respect to not only the global pattern but also the more ‘local’ pattern impacting the UK.

What I see from the posts relating to all the  teleconnections is that there is a Plethera of available data that perfectly illustrates the starting point of where the globes at- a T0 if you like of all metrics -

There after though every reference point I see for what ‘may’ happen is derived from analogues & phases of what happened before - & how that ‘may’ manifest in the next 2 weeks or so- & how this marries up with the forecast outputs for the chosen uteleconnection ( MJO etc )

There are many inherent problems & assumptions with this -

- The assumption that what happened before is likely to occur again

- That the Model data & output for that metric is stable rather than volatile 

- & the trickiest relationship of them all which is how the tropospheric metrics interact with the biggest factor the stratosphere.

As a result what I see is forecasts highlighting & punts based on current data blended with analogues & model data- which is on very shaky ground because the models are poor at all levels & the analogues are probably not valid anymore.

So to invest a lot of time putting a detailed post together is really just a bigger more in-depth explanation of the global starting point..

What it boils down to is as highlighted so many times

* We can’t forecast change. We’re awful at it-Just the same as the models 

* The analogues are out of date due to changing paradigms 

* There are to many unknown unknowns - ( think of an example - failure of -QBO last winter )

As a result of this - I don’t believe there will be any accurate forecasts in the next 20 years - ( for cold anyway ) - but of course some successful mild ones for El Niño + winters.

For me the only way forward is to continue to monitor the stratosphere + strength of vortex & to investigate the teleconnections that are changing at the moment including existing ones - as the sample size is to Small

I admire the persistence in believing that it’s a credible forecasting method - however sadly - there are just to many variables in the way of creating a manageable frame work in which to deliver consistency & Accuracy...

S

Edited December 29, 2017 by Steve Murr

[Summer Sun]

JMA 2m temps for the next month warmer than average across most of the UK and Europe

Chart Ben Noll

Edited December 29, 2017 by Summer Sun

[Chesil View]

Just now, Steve Murr said:

An interesting debate - in which FWIW I would add my points-

The Forum has has been around for 13 years - & with that teleconnections have really come to the forefront of the discussion for the last 5 years - They we’re always there however as time has progressed so more & more measures / drivers have been identified.

The most crucial part of the understanding how it’s all morphed together ( which is still in its infacy ) in terms of

* What teleconnection is a driver / cause & what is an effect - 

* Then once established what drivers carry the most weighting with respect to not only the global pattern but also the more ‘local’ pattern impacting the UK.

What I see from the posts relating to all the  teleconnections is that there is a Plethera of available data that perfectly illustrates the starting point of where the globes at- a T0 if you like of all metrics -

There after though every reference point I see for what ‘may’ happen is derived from analogues & phases of what happened before - & how that ‘may’ manifest in the next 2 weeks or so- & how this marries up with the forecast outputs for the chosen uteleconnection ( MJO etc )

There are many inherent problems & assumptions with this -

- The assumption that what happened before is likely to occur again

- That the Model data & output for that metric is stable rather than volatile 

- & the trickiest relationship of them all which is how the tropospheric metrics interact with the biggest factor the stratosphere.

As a result what I see is forecasts highlighting & punts based on current data blended with analogues & model data- which is on very shaky ground because the models are poor at all levels & the analogues are probably not valid anymore.

So to invest a lot of time putting a detailed post together is really just a bigger more in-depth explanation of the global starting point..

What it boils down to is as highlighted so many times

* We can’t forecast change. We’re awful at it-Just the same as the models 

* The analogues are out of date due to changing paradigms 

* There are to many unknown unknowns - ( think of an example - failure of -QBO last winter )

As a result of this - I don’t believe there will be any accurate forecasts in the next 20 years - ( for cold anyway ) - but of course some successful mild ones for El Niño + winters.

For me the only way forward is to continue to monitor the stratosphere + strength of vortex & to investigate the teleconnections that are changing at the moment including existing ones - as the sample size is to Small

I admire the persistence in believing that it’s a credible forecasting method - however sadly - there are just to many variables in the way of creating a manageable frame work in which to deliver consistency & Accracy....

S

Hi Steve.

As ever a very interesting post from yourself. I think you hit the nail on the head that so much of this study is in its infancy. I'm sure the likes of GP and Tamara would be the first to admit this. However that does not mean that continued study of the subject is not a vital part of the way forward. As you implied Nwp is an incredibly blunt tool and is of course driven by the algorithms abilities or lack thereof in mimicking the effects of the atmopheric and oceanic drivers.  To not continue the study in the teleconnections field would be like saying well we 've already invented a pretty poor square wheel why on earth should we try and  work out how to make a round one.

[Bruegelian]

It is a worry that seemingly all of the mid-long-range models are going for a largely mild & wet January with high pressure to south,that is JMA, CFS, EC, presumably Glosea as well. However today Gav P did a video saying that he thinks that possibly it will not pan out like this, because of the jet stream wanting to keep south of us.

Does anyone know why exactly, the jet stream has been behaving like this - is it the pressure pattern, or perhaps even lower-than-average temps in Africa reducing the thermal gradient, or something else? Thanks.

[Chesil View]

Just now, Bruegelian said:

It is a worry that seemingly all of the mid-long-range models are going for a largely mild & wet January with high pressure to south,that is JMA, CFS, EC, presumably Glosea as well. However today Gav P did a video saying that he thinks that possibly it will not pan out like this, because of the jet stream wanting to keep south of us.

Does anyone know why exactly, the jet stream has been behaving like this - is it the pressure pattern, or perhaps even lower-than-average temps in Africa reducing the thermal gradient, or something else? Thanks.

I think they were also going for a mild December and  look how that turned out and as Fergie twetted earlier Ec now going for below average first threes weeks of Jan and Glosea wavering all over the place.

Edited December 29, 2017 by Broadmayne blizzard

[Matthew.]

6 minutes ago, Bruegelian said:

It is a worry that seemingly all of the mid-long-range models are going for a largely mild & wet January with high pressure to south,that is JMA, CFS, EC, presumably Glosea as well. However today Gav P did a video saying that he thinks that possibly it will not pan out like this, because of the jet stream wanting to keep south of us.

Does anyone know why exactly, the jet stream has been behaving like this - is it the pressure pattern, or perhaps even lower-than-average temps in Africa reducing the thermal gradient, or something else? Thanks.

As a guess I would say pressure hasn’t been as high in Southern Europe this winter, since winter 2012/13 which is helping it drop more south plus positive pressure has been stacking up to the west of us

Edited December 29, 2017 by Matthew.

[Snowjokes92]

Just to add briedly to teleconnections. I personally see them as a whole host of different forecast and models combined together to predict the weather for weeks and month ahead. And add to that we all know really, to predict the weather that far out is impossible  more times than not! I think teleconnections interfere with each other too much maybe and they could be why we see a lot of variance changes. And the big thing is theres a lot of changes in the very short term model output, resulting in changes to midterm model patterns,which would make the output beyond midterm even more conflicting. No one can guarantee what will happen weeks ahead, when there is so many changes on models just a couple days ahead. This more so in winter imo.

Edited December 29, 2017 by Snowjokes92

[nick sussex]

My view re the teleconnections debate is the issue is with our expectations.

In this age of super computers and advanced technology it’s easy to think they have all the answers so I think it’s more about expectations management.

Is what we expect fair or are we expecting too much when weather is fluid and continually evolves .

A forecast say made in November for the whole winter can’t take into account how those initial teleconnections might change.

We should bear in mind though that science needs to know where it went wrong to make advances.

So how we should view teleconnections is perhaps a continuous longer term experiment. 

Eventually we might get a few more answers but I still think LRFs will suffer as there’s simply too much chaos in the atmosphere.

 

Edited December 29, 2017 by nick sussex

[MildCarlilse]

6 minutes ago, Bruegelian said:

It is a worry that seemingly all of the mid-long-range models are going for a largely mild & wet January with high pressure to south,that is JMA, CFS, EC, presumably Glosea as well. However today Gav P did a video saying that he thinks that possibly it will not pan out like this, because of the jet stream wanting to keep south of us.

Does anyone know why exactly, the jet stream has been behaving like this - is it the pressure pattern, or perhaps even lower-than-average temps in Africa reducing the thermal gradient, or something else? Thanks.

Yes the Jet stream,cause or effect? I would love to see models,jet stream and synoptic charts from the winters mid 1700's.What we could learn from them.

[Bruegelian]

Thanks for the responses - Also I think Steve Murr's idea that AGW is changing the landscape faster than computer models can keep up with, may have some credence. But getting into pretty deep water now....

[SMU]

Just to add to the model debate-

ive posted a couple of times that the GFS stratospheric ensemble data had to 2 key dates in the means

1) New Years - peak of the current increase in zonal winds

2) 6th/7th the date at which the ‘dip’ in zonal winds aligns to the long term mean- 

In my world What this translates to is for a short while across that second timeline the stratospheric vortex whilst still strong isn’t the defining feature within the atmosphere allowing this winters signals to become a little more prevalent again-

The Pacific -EPO ridge which has almost been quasi stationary will be kicking about creating the stark east west split across the states - 

Alongside this the Atlantic ridge trying to move N/NE out of the Azores....

It does look like that the 6-7th has become the focal point of the output as there is the window for a shallow high to ridge NE & settle towards Scandi-

Its clear already that the chances are it won’t be a Proper scandi High - but rather a surface high ...

Which ‘could’ be as useful as the real thing as it can deflect the jet SE as well as advert cold air into the UK from the East -

Optimal dates where this works spring to mind - Jan 13 - Feb 96

So let’s see how that develops...

Post that there’s huge divergence between what we are seeing in the NWP V what the GFS stratospheric data is showing which are polar opposites !!! 

Much to be resolved.. but a very nice T2M ensemble trend for Debilt today...

S

[Methuselah]

4 minutes ago, nick sussex said:

My view re the teleconnections debate is the issue is with our expectations.

In this age of super computers and advanced technology it’s easy to think they have all the answers so I think it’s more about expectations management.

Is what we expect fair or are we expecting too much when weather is fluid and continually evolves .

A forecast say made in November for the whole winter can’t take into account how those initial teleconnections might change.

We should bear in mind though that science needs to know where it went wrong to make advances.

So how we should view teleconnections is perhaps a continuous longer term experiment. 

Eventually we might get a few more answers but I still think LRFs will suffer as there’s simply too much chaos in the atmosphere.

 

I think that we, as in a sizeable chunk of us, take Tamaras/GPs musings about QBOs and MJOs as if they are precise, nailed-on forecasts for cold (or for communal wrist-slitting) while they are really intended for guidance. Helpful guidance at that!

[Daniel*]

The ‘trouble’ is the more knowledge we acquire, the more questions are uncovered one thing leads to another. Forecasting is very complex hats off to those that do it professionally, it’s not like it was decades ago. Times have changed for the good really, no doubt we’re making advancements even while subtle ‘jumps’, over time once impenetrable barriers will be overcome. That could be in a hundred years time maybe much sooner, but I see leaps and bounds made in terms of improving accuracy in long range forecasting I doubt it will ever be perfect but more accurate. 

[Polar Maritime]

Severe gales for some over the next day or two.

[northwestsnow]

EC is pretty much a carbon copy of UKMO 144.

One would assume the low will dig SE and pressure will rise across Scandinavia thereafter...

Edited December 29, 2017 by northwestsnow

[Allseasons-Si]

Oh!!!! yes! yes! yes!

Edited December 29, 2017 by Allseasons-si

[Rocheydub]

3 minutes ago, Allseasons-si said:

Oh!!!! yes! yes! yes!

We could see a link up of the Russian Highland the Icelandic high if that rolls on... I am interested in NWP suddenly again!

[SMU]

192 ECM

 

expect a a very big slider scenario @ 216 !

[Rocheydub]

1 minute ago, Steve Murr said:

192 ECM

 

expect a a very big slider scenario @ 216 !

Beat me to it. Cracking chart!

Edited December 29, 2017 by Rocheydub

[SMU]

A slightly underwhelming ECM 216

JMA is the Carlsberg run this eve

https://www.meteociel.fr/modeles/jma.php?ech=24&mode=1&nh=1&archive=0

S

[Captain Shortwave]

The good news is that the models are starting to pick up weak heights developing to our north around the day 6 mark.

     

The GFS not so keen tonight (But has also hinted at something similar on previous runs), but the pattern doesn't look like a normal zonal pattern with low pressure taking a more southerly track. There certainly seems a real possiblity of wintry conditions developing in the first week of January, especially later on that week though details look rather uncertain for now. All I can say is that the outcome will probably not pan out like the ECM with low pressure stalling over SW England.

Edited December 29, 2017 by Captain Shortwave

[comet]

Surprised the ecm 216 chart was as good as that ( which is not saying much). Quite underwhelmed this afternoon and evening by the model runs from a coldies point of view. Yes they are a lot better than last year but that does not make them great. Of course there will be changes to come that's for sure.