Model Output Discussion 29th December - Into mid-Winter.

[Methuselah]

3 minutes ago, D.V.R said:

It's only a slight chance but it's interesting nonetheless.

That's true. I think I can remember, in the 1980s, when 'blowtorch' long-fetch southerlies almost immediately preceded stonking Beasts-from-the-East? That could of course be false memory syndrome?

[blizzard81]

21 hours ago, blizzard81 said:

Also remember that the only easterlies that are modelled at day 9 or 10 are the ones that never verify. They are only picked up CORRECTLY at around day 6, sometimes less than 6 days. They are notoriously difficult for the models to predict. For this reason I feel there is plenty of time for upgrades.

Steady as she goes. Today's 12z runs, especially ecm, are an upgrade on yesterday's. I feel there are more to come.

[nick sussex]

This trend from the ECM is beginning to remind me of some previous instances where we've seen a sharp fall in the zonal winds.I wonder how many more corrections we can squeeze out of this. With luck enough to give UK coldies some snow at least. Even the misery model is slowly inching westwards with more amplitude after being not interested.

[Long haul to mild]

Seen this many times before over the years where the models try to blow away a block to our East but in fact the result is very different. With so much doubt over the eventual outcome it really does surprise me when I see posts claiming that we'll have to wait for mid-Feb before there's any chance of cold or snow. The only justification for that line of thinking is if you take model output at D16 etc at face value which is a huge leap of faith.

[chris55]

Subtle differences between ECM and GFS at 168 (subtle but also significant)

 

The negative tilt and trough disruption from the Atlantic lows on ECM is a world away from GFS.

If you dig a little deeper and make some assumptions based on the significant mid term changes of the last few days, we could assume that ECM could well progress rather more favourably than the 12z run suggests. 

[IDO]

37 minutes ago, Long haul to mild said:

Seen this many times before over the years where the models try to blow away a block to our East but in fact the result is very different. With so much doubt over the eventual outcome it really does surprise me when I see posts claiming that we'll have to wait for mid-Feb before there's any chance of cold or snow. The only justification for that line of thinking is if you take model output at D16 etc at face value which is a huge leap of faith.

The problem for me is that I don't see as much potential in those charts as others do:

Why? Because the GEFS have been keen to draw the trop PV back to NE Canada from D8-16 as the GEFS mean highlights:

That suggests little hope of retrogression of that high far enough west to benefit the UK. In fact it is the worse case scenario with that Pacific forcing coming into play. With the GFS putting back the SSW till early Feb at the earliest we cannot afford the trop long wave pattern to reflect that D16 mean.That block as it sits at D6 is nice if you like settled weather, again in the SE for a few days in a row highs were 6.5c but feeling warmer in the sunshine, whilst the CET keeps moving up, now 1.7c above average for Jan, so this setup is not really cold (though frosts in my region). Maybe others see it different...

Edited January 19, 2017 by IDO

[tight isobar]

Response of output now' evaluating' evolution. 

10hpa fluxing having response via-run to run output. 

Pv as-other output in general having deterministic dramas..ie- scope/expansion/ placement of heights. Ssw looking to be taking signal' down to raw model/models. 

This without mention of the above mentioned HP mid term exactions.

All in all the most any zonality/mobility. .can ve considered is at best a blip.

An overide' of cold-v-cold synoptic s in output is surely all' but inevitable. 

Winter chances are best as back loaded! 

[Paceyboy]

GEFS pointless after 144hrs IDO.

Any trend in the high resolution is forgotten when into low resolution. 

Each run has kept a continental influence longer over the 48 hours, ECM and GFS. 

Shame there isn't a weather cash-out option when the ensembles flip to cold. Each GEFS is here and gone in a flash prediction wise, more so in winter time.

 

[danthetan]

On 15/01/2017 at 09:27, danthetan said:

Where we go after 120hrs for me is anyones guess, it all seems to hinge on what happens to the trough in the Atlantic around 144hrs. ECM pulls all the energy over the top off the high sitting over the UK, UKMO seems to send most of the energy over the top but some underneath and GFS more or less splits it sending enough energy to at least drag colder air in for a time before pushing the high back into Europe, all these options go on to produce very different outcomes in the latter stages of GFS & ECM. With that in mind we know how the models always struggle to gauge the strength of any such blocking high so for me i need to see some consistency from the models around the 144 hr period before knowing which way were heading.

With reference to my post on Sunday about the (blocking high) we can see the models were underestimating the resilience of the high pressure and has extended the time we are being influenced by it. It will be very interesting to see if the block can sustain the Atlantic attempts at breaking through next week and as hinted by ECM tonight, maybe send some energy under the block. Interestingly while looking for my post on Sunday I came across this attached chart from 1991 someone had posted, quite interesting how close it is to GFS 12z chart for next Tuesday theres hope yet.

[Long haul to mild]

14 minutes ago, Paceyboy said:

GEFS pointless after 144hrs IDO.

Any trend in the high resolution is forgotten when into low resolution. 

Each run has kept a continental influence longer over the 48 hours, ECM and GFS. 

Shame there isn't a weather cash-out option when the ensembles flip to cold. Each GEFS is here and gone in a flash prediction wise, more so in winter time.

 

Agree. If the models are underestimating the strength of the block in the reliable, which is quite possible, then the low res ensembles are hardly likely to show anything other than a return to default UK weather in winter.  

It may be my perception but most cold spells I can remember pop up at about T168, often from nowhere. Therefore, although probability suggests that a zonal mean at T384 may end up being correct, it doesn't mean it will necessarily come to pass.

[chris55]

51 minutes ago, IDO said:

The problem for me is that I don't see as much potential in those charts as others do:

Why? Because the GEFS have been keen to draw the trop PV back to NE Canada from D8-16 as the GEFS mean highlights:

That suggests little hope of retrogression of that high far enough west to benefit the UK. In fact it is the worse case scenario with that Pacific forcing coming into play. With the GFS putting back the SSW till early Feb at the earliest we cannot afford the trop long wave pattern to reflect that D16 mean.That block as it sits at D6 is nice if you like settled weather, again in the SE for a few days in a row highs were 6.5c but feeling warmer in the sunshine, whilst the CET keeps moving up, now 1.7c above average for Jan, so this setup is not really cold (though frosts in my region). Maybe others see it different...

Sometimes the op will lead the ens, and this ultimately makes the long term mean futile and meaningless, and for me we have been seeing this in the past few days with regard to the delayed onset of what looked like a nailed on Zonal/milder push from the Atlantic if you believed the ensembles.

Without the 12Z ECM op, id have said the building Scandi ridge (which has been the reason the Atlantic has been held back) was much more likely to topple with so much energy going over the top and little sign of any undercut.

However with the subtle hints from the ECM12Z edging towards some energy disrupting and heading south east i would hold back on making any assumptions based on a day16 mean chart from GFS.

 

 

[bluearmy]

Beginning to see the extended eps pushing lower height anomolies to our south over w France and Iberia - this opens up the possibilities of further trough disruption and the control is a good example of this as systems post the day 10 (which is very close to the op) disrupt against blocking to our ne 

all bets are off !

Edited January 19, 2017 by bluearmy

[tight isobar]

Tbh' anything -144hrs is at best a diluted probability/impossibility. 

There is at present model mismanagement' via strat/trop response. .

Let alone the persistance of the wobbling' euro/uk/scandi block.

And yes that's a crazy evaluation'of mentioned. .much as miss' interpreted model synoptic s....AT PRESENT. 

Edited January 19, 2017 by tight isobar

[winterof79]

 A long long time ago in1991 there was not a lot happening a high slap bang over us on the 13th

Bit like now

 

Atlantic trying to nudge in

Bit like now

Cold for South East bla bla

 

Block said hang on matey I belong over Blighty

And what's more I will bring reinforcements and you will never get in

 

And we all lived happily ever after.......

Who needs an SSW?

Although it may help

 

 

 

Edited January 19, 2017 by winterof79

[Singularity]

To avoid excessive computing power requirements, the models have to rely on a variety of parameterisation techniques, which are essentially ways to represent the overall effect of phenomena that occur on a sub-grid scale (i.e. finer than the model resolution) without having to try and depict those features individually. Under many circumstances, this does a reasonable job out to 4-5 days range, but the equations used to parameterise are not perfect and so errors do arise - ones that we can witness as a 3 hour difference in the timing of a weather front at just a few day's range, for example. Over time, these errors that result from simplification inevitably lead to further errors which in combination amplify the overall distance the model run is from what actually happens.

Sooner or later this tends to lead to small-scale disturbances being completely overlooked. This effect occurs sooner the lower the resolution of the model. So with GEFS it occurs sooner than with GFS as they are at a lower resolution. As far as I know, ECM ens. run at a higher resolution than GEFS so tend to fare better with these things.

A lot of what drives major pattern shifts is in the way large masses of cold or warm air move about, such as the tropical maritime air that currently looks to be moving NE from near Florida in the 6-10 day range, and the Arctic air that appears likely to move from Siberia to NE Canada to some degree or other in the 9-12 day range. Now here's the thing - a small-scale disturbance can make for a less 'clean' transfer of an airmass from one place to another, in turn making the thermal gradients more inconsistent and hence the jet stream more 'broken-up' with more and longer-lasting opportunities for trough disruption. There's still time for such a feature to tip the balance with respect to the TM airmass moving NE in the 6-10 day range, and more time still for one to interfere with the cross-polar movement of Arctic air in the 8-12 day range.

I am always suspicious, much as Nick S is this evening of the 12z ECM, when broad, stretched-out troughs are depicted, because they are often a case of too much simplification leading to shortwave features either failing to close off into secondary lows or not being developed at all, when in reality there is often at least one of these resulting in the trough becoming divided into two or more distinct features, and with that we find a better environment for trough disruption to occur.

Unfortunately, while these trends are common, they're not guaranteed every time, and even then, just how far they will go and whether that will be far enough remains a matter of guesswork really. It does however mean that even good ensemble agreement can rarely be talked about with great confidence at more than a week's range; the errors described above will affect all ensemble members and limit their ability to spot trough disruption on those occasions when it's actually in the pipeline. 

So concludes an educational type post that I suddenly found myself writing while in a particularly pensive mood. It comes from a decade of unhealthily close NWP model-watching... make of it what you will! 

[winterof79]

21 minutes ago, bluearmy said:

Beginning to see the extended eps pushing lower height anomolies to our south over w France and Iberia - this opens up the possibilities of further trough disruption and the control is a good example of this as systems post the day 10 (which is very close to the op) disrupt against blocking to our ne 

all bets are off !

I do like it when you bring good news

[Ice Day]

17 minutes ago, winterof79 said:

 A long long time ago in1991 there was not a lot happening a high slap bang over us on the 13th

Bit like now

Who needs an SSW?

That's a great post Winterof.  I'm not big on pattern matching but it can't be denied some of the similarities are there for all to see! 

[SMU]

So once again another full day of moves towards a 'continental solution' I think for the time being that should be the phrase as opposed to 'Easterlies' as theres a general perception that Easterlies are going to be action packed bitter cold winds bringing snow-

In order to remove the emotion the 'continental solution' should stoke us up enough to believe there maybe enough in the tanks to get to easterly, however we have to move the dial to get there.

As per the update last night the models are 'generally' in a state of catch up at the moment, which for me signifies the point along the timeline in the suites where patterns can be changed with repercussions more especially for the UK, ive tried to highlight before generally at 144 - 192 the models especially the euros have a grip on the longwave pattern & are starting to develop the shortwave ones- well today I think that arena is -24-36 hours so the longwave resolution is sitting around 108-156 & shortwave accuracy sub 120-

As a result, the refinements that each runs are making are a huge difference for the surface temps over the UK, the means over the last 48-72 hours have been woeful, so much so that this is the difference between day 5 & day 9 just days before - I cant get the mean now however the Op was fairly reflective -

Anyone forecasting via 'means' needs to really ask what the point is as every single time there is model change @120-144 the means will be hopeless...

The crucial timeframe for 'upgrades' to any wintry potential via the battle of West meets the ' continental flow' ideally will come from the 138-168 timezone with more energy diving south & a more amplified high- I think the models have overcooked the northern jet & as seen over many days this week will continue to move favourably over the forthcoming runs-

More on the SSW later -

S

PS I think the ECM day 10 pattern is great about 1000 miles east of us with significant troughing, expect that to move west over the next few Op runs 

 

Edited January 19, 2017 by Steve Murr

[bobbydog]

28 minutes ago, winterof79 said:

 A long long time ago in1991 

Who needs an SSW?

 

 

 

Thing is though, there was an SSW. Almost exactly the same as the one being modelled now... 

[winterof79]

1 minute ago, bobbydog said:

Thing is though, there was an SSW. Almost exactly the same as the one being modelled now... 

Yes Bobby dog I have edited my post to suit

[cyclonic happiness]

Wasn't the 1991 cold spell preceded by weeks of anticyclonic gloom?

The longer we stick with the cap we have, the better imo

[damianslaw]

Another day, and well more of the same, models continue to show no real change in the reliable, anticyclonic holding supreme against a rather ineffective atlantic - which is once again shown to only displace the high in about 8 days time - well beyond the reliable, and by the looks of things, it doesn't look like it would herald a sustained potent atlantic attack, rather one low pressure system probably anchoring itself to our NW, quickly being forced away to the NE with the ridge pushing back in from the south.. the story of the winter.

There is a chance of some trough disruption in the reliable timeframe, something that the models will struggle to pick up if it occurs until the nearer timescale.

In the immediate - quite an unusual synoptic at present, a high pressure system with a slow moving front trapped within - that has been static for 3 days running, and very odd in its nature, with a cold front on its western axis and a warm front on its eastern axis, marking the demarcation between cold clear air over southern england, and milder overcast air over northern england, and instead of sinking south or simply fizzling it now wants to move northwards, so cold will arrive into northern england over next 24 hours and over the weekend we have the threat of some sleety snowy flakes from it - all rather strange.

 

Edited January 19, 2017 by damianslaw

[SMU]

16 minutes ago, bobbydog said:

Thing is though, there was an SSW. Almost exactly the same as the one being modelled now... 

http://www.meteo.mcgill.ca/~pmartineau/updated_graphs/svw_anim/all_gif/1991_01_28

The Merra data says it wasnt but its data min was at 4.5M/S @10HPA over 60N so near as dammit...

[SMU]

18z out to 114 has a direct Easterly flow over England & lower heights which again supports building a cold pool over the UK in situ- 

The theta E charts the continental air growing in 'depth' again & heading downwards as opposed to the 12z

18 v 12

[bobbydog]

1 minute ago, Steve Murr said:

http://www.meteo.mcgill.ca/~pmartineau/updated_graphs/svw_anim/all_gif/1991_01_28

The Merra data says it wasnt but its data min was at 4.5M/S @10HPA over 60N so near as dammit...

I posted some comparison charts this morning. The warming that's forecast is almost identical to the one in '91. Even the dates are within a day or two. If that's what a "near as dammit" warming can do, I'll take it!