Model output discussion - into 2018

[ACCOUNT DELETED]

9 minutes ago, tinybill said:

just see  a  warning  fro  met office storm  heading winds  now be 80 mph  over  much of the  uk

Strongest winds in northern UK.

They were unsure if the winds will turn westerly which would increase the risk of damaging gusts through the central belt of Scotland that's what I recently heard. Apart from that, nothing's changed including the warnings. 

[Gibby]

Incidentally some dissing the ECM Verification should view the latest 10 day hemispheric ones which see ECM reign supreme over GFS almost incessantly, currently 47.6 ECM to 40.6 GFS taking all 10 day predictions over the last 31 days. Just because output might not show what you want doesn’t make it any lesser a model.

Edited December 30, 2017 by Paul

[bluearmy]

13 minutes ago, Gibby said:

Incidentally some dissing the ECM Verification should view the latest 10 day hemispheric ones which see ECM reign supreme over GFS almost incessantly, currently 47.6 ECM to 40.6 GFS taking all 10 day predictions over the last 31 days. Just because output might not show what you want doesn’t make it any lesser a model.

Pointless posting the stats martin because more than 80% of the time the ecm out verifies all the other models. since its last upgrade I'm not sure how it is days 8/10 as before that, gfs used to be about the same stats (both sub 50%) 

[SMU]

19 minutes ago, Gibby said:

Incidentally some dissing the ECM Verification should view the latest 10 day hemispheric ones which see ECM reign supreme over GFS almost incessantly, currently 47.6 ECM to 40.6 GFS taking all 10 day predictions over the last 31 days. Just because output might not show what you want doesn’t make it any lesser a model.

There’s always someone trawling out statistics.

we don’t live across the whole of the NH - we live in a tiny portion of it - probably about 1000th the area therefor all the stats anyone wants to point out of irrelevant.

Anyone paying close attention to the operational ECM this winter will know how poor it’s been. It was the worst model with the slider & indeed the run up to the slider- The GFS was relentless & ECM followed suit with the blocking - so whilst you can quote numbers across the globe here’s a typical fail from the last cold spell in true ECM progressive style.

Forecast day 7 V actual ...

& for fairness the GFS same run

much much closer- where it matters - in the Atlantic.

So no one is dissing it because it doesn’t show the desired output - it’s being dismissed because it’s performing poorly in our region...

Please understand the method of why comments are being made before throwing  out meaningless statistics.

Edited December 30, 2017 by Steve Murr

[Guest]

Some great output today for only a week or so away, mostly going in the right direction for coldies. I do not have time for one of my cross-model analyses today. With the interest now turning towards an easterly quarter and a "possible" Scandinavian HP setting up residence, I see that there have already been concerns raised regarding just how cold will the easterly be (both surface temperatures and uppers). It's high time that I did one of my Eurasian current and predicted temperature and snow cover reports. My plans for celebrating the New Year tomorrow evening with a couple of friends have just been cancelled (due to them having flu) - so my new plans are to pour myself a couple of glasses of Merlot and write a report tomorrow evening (in a reasonably sober state!).

In the meantime, I see that there has been further discussion about the easterlies that we used to see back in the 1960s to 1980s and comparing them to several of those starting to show up in some of the middle term model output. I used to keep some pretty thorough records from 1960 onwards. Yes, easterlies were far more common back then and we had some which delivered memorable cold spells and some which were very snowy. There were, however, just as many which were bone dry and some which were not much colder than average. I did several posts earlier in December regarding easterly air streams. I do not wish to clog up this thread by repeating long parts of those posts now. Some of you might have missed them and as it seems that they might be becoming much more relevant, here are the links to those posts:

9.12.17 Winter Model Thread - December 11th, "A Possible Easterly Some Time After Christmas - What Can We Expect": 

https://www.netweather.tv/forum/topic/88969-model-output-discussion-winter-proper-underway/?page=17      (it's about two thirds of the way down) 

This was actually one of my long reports into European temperatures but the first part of that post was all about easterlies.

 

 

9.12.17 Winter Model Thread - December 16th, "The Northerlies vs Easterlies Debate": 

https://www.netweather.tv/forum/topic/88969-model-output-discussion-winter-proper-underway/?page=59      (it's about halfway down).

Another point of interest in that post was that I included a number of GEFS FI ensemble charts for T+384, for January 1st, which showed disrupted PVs, Scandi HPs and several easterlies. A few of those charts look quite similar to some of those currently modelled for around January 7th. So, will the GFS get lucky for once - just 6 days adrift with a D16 chart (you don't need to answer that one!). 

 

9.12.17 Winter Model Thread - December 20th, "How Rare is a Split Vortex and What to Expect from Them": 

https://www.netweather.tv/forum/topic/88969-model-output-discussion-winter-proper-underway/?page=83        (again about halfway down).

With some of the recent model runs starting to show some thoroughly disrupted PVs again with some unusual looking Arctic profiles, it reminded me of the post that I did on these 10 days ago. I had conducted some quite detailed research into them which revealed some interesting answers. There are cross links there to a number of specialist papers on the subject and I produced a short summary from each "abstract page" (since that post I've been reading some of those papers in full and I find them totally absorbing). I referred to a number of cold winters which had some very disrupted PVs. My conclusion was that some of the best examples were in winters which saw the PV come under a series of attacks over several weeks (or even longer) and despite the usual seasonal strengthening of the PV, the cold spell eventually developed after a lag time of typically around two weeks. Now, this winter, we've had the initial disruption going on for a while, then the PV showing signs of battling back and now further disruption being modelled. Could what's showing up in some of the output now be at least partly a consequence of this periodic disruption?   

 

Now, I'm not trying to score brownie points or claim that I was on to something 2 to 3 weeks ago (that's not at all my style) and next week's "possible" easterly is a very long way from being nailed on. I simply felt that these three posts are highly relevant to some of the current model output and would interest a few of you. 

 

Edited December 30, 2017 by Guest
Correct typos and check links

[johnholmes]

A degree of truth in what you say Steve but also in what Gibby posted.

That is that large scale wave patterns is really what the main centres are checking models on. We all appreciate to get deep cold over the UK, just as for any other western facing maritime coastal area, the wavelength pattern has to be just right for this to occur. Thus very short wave pattern, less than 45-50 degree in longitude will generally give fairly unsettled weather. Longer wave patterns, if I remember from my forecasting days, UK Met use 60 degrees plus. This is the type of slow moving or even, in very special instances, retrogressing upper air patterns, that give the deep cold so popular on here. In this then as Gibby posted, especially at time scales (this is my memory so it may be a touch faulty) of 144-168h and more then ECMWF regularly outperforms GFS. If one reads the NOA forecast commentaries, (you quite often see Nick Sussex) quoting these in favoutable set ups, they often use ECMWF over GFS.

[Man With Beard]

On the storm front, the northern England cities better take heed of ARGEPE tonight as it develops a tiny New Year's Day storm out of nowhere - 80 mph gusts or more INLAND

[Daniel*]

1 minute ago, Man With Beard said:

On the storm front, the northern England cities better take heed of ARGEPE tonight as it develops a tiny New Year's Day storm out of nowhere - 80 mph gusts or more INLAND

What’s the ARGEGE  the model has a tendency to overdo wind speeds and the inland nature of the strongest winds and a little suspect I’ve not had a proper look but no others models really develop anything. 

[SMU]

12 minutes ago, johnholmes said:

A degree of truth in what you say Steve but also in what Gibby posted.

That is that large scale wave patterns is really what the main centres are checking models on. We all appreciate to get deep cold over the UK, just as for any other western facing maritime coastal area, the wavelength pattern has to be just right for this to occur. Thus very short wave pattern, less than 45-50 degree in longitude will generally give fairly unsettled weather. Longer wave patterns, if I remember from my forecasting days, UK Met use 60 degrees plus. This is the type of slow moving or even, in very special instances, retrogressing upper air patterns, that give the deep cold so popular on here. In this then as Gibby posted, especially at time scales (this is my memory so it may be a touch faulty) of 144-168h and more then ECMWF regularly outperforms GFS. If one reads the NOA forecast commentaries, (you quite often see Nick Sussex) quoting these in favoutable set ups, they often use ECMWF over GFS.

Hi John

in the round the ECM performes the best however last few cold spells very progressive

notice the significant dip V the GFS on the last cold spell 10 Dec 

ECM red GFS black

 

[johnholmes]

Very impressive that Steve. I have to admit that I don't keep anything lie as close an eye on the statistics as I used to do. If I could suggest that yourself and any others might post such charts to add background to any claims being made. It all helps everyone to see why a particular stance is taken.

 

[SMU]

2 minutes ago, johnholmes said:

Very impressive that Steve. I have to admit that I don't keep anything lie as close an eye on the statistics as I used to do. If I could suggest that yourself and any others might post such charts to add background to any claims being made. It all helps everyone to see why a particular stance is taken.

 

No worries -

I thought by now people might trust what I’m saying knowing I’m not making it up

GFS our to 192 - strong NE winds over the UK with the -8c isotherm approaching s

[northwestsnow]

Would like to see some energy in the Atlantic at 180 go south- sharpish!!

[northwestsnow]

Yes, iwould imagine the precip turning wintry across central UK by 186hrs  Nice NE wind...wow,how rare are they!!

snow showers moving across the SE therafter..

Edited December 30, 2017 by northwestsnow

[MR EXTREMES]

1 hour ago, phil nw. said:

Still a fascinating period coming up with the polar jet wavering across the UK.The cold air looks like it is never going to be far away from Scotland through the coming week.

Next week end looks a tricky period to resolve as energy is split off the Atlantic as it comes up against the modest wedge of heights further north.

Currently it looks like the northern arm of the jet will eventually hold sway but as the gef stamps show by day 8 it is a close run thing.

By day 10 ECM mean in line with the it's Op. run as the Atlantic looks too strong this time. 

The persistence of that Canadian vortex is the main driver and it still looks like further incursions of polar maritime air from the nw is the more likely rather than any Scandi.high in week 2 at the moment. 

Well I see shallow heights from iceland to Scandinavia and evolves into a more pronounced block.

The gem has churned out 3 runs on the trot with a much less progressive run.

Get new year out the way and we have more information on what direction and how well the models are performing.

To be fair the phantom ECM Greenland highs is not as bad this year.

Amount of winters with shredded nerves because ECM went on a blocking fetish.

But in my opinion nearly all models suggest something cold even if it is brief.

But I do think a block of sorts into Scandinavia Iceland could deliver our first cold Easterly or northeasterly.

Edited December 30, 2017 by MR EXTREMES

[Daniel*]

Us snow starved SE/EA folk would do well out of this very clean long drawn flow that.. Thames streamer direction but I do not think the uppers are quite cold enough.

Edited December 30, 2017 by Daniel*

[Scandinavian High.]

Cold north easterly with snow showers in to east and south eastern England.

[feb1991blizzard]

Here comes a blizzard.

Didn't materialise again.

Edited December 30, 2017 by feb1991blizzard

[NorthernRab]

11 minutes ago, northwestsnow said:

Yes, iwould imagine the precip turning wintry across central UK by 186hrs  Nice NE wind...wow,how rare are they!!

snow showers moving across the SE therafter..

You might say precipitation has turned very wintry across Scotland by this point.

 

[Scandinavian High.]

3 minutes ago, feb1991blizzard said:

Here comes a blizzard.

Any charts.

[feb1991blizzard]

4 minutes ago, abbie123 said:

Any charts.

Corrected post above - there wasn't actually a chart - I was going to post one as I thought a massive battleground was setting up.

[Scandinavian High.]

When we get with that low coming in from west south west is it case the model is being very progressive and steaming it right through. As we get  nearer to time block may put more of a fight and go under with a snow avent just my thinking I could be way of the mark here..

[Man With Beard]

Clusters T264 highlights the ongoing trouble heights to the east are going to cause forecasters - incoming trough responding in a number of ways, all of which affect the UK in very different ways.

[northwestsnow]

6 minutes ago, Man With Beard said:

Clusters T264 highlights the ongoing trouble heights to the east are going to cause forecasters - incoming trough responding in a number of ways, all of which affect the UK in very different ways.

Cluster 1 shows a cold NW wind with the 528 DAM line across the south..

on closer inspection it looks very cold for NW Britain with scotland in particular very wintry!

LOL, whod have thought it- we were staring down the barrel of zonal misery not long ag-

 

Edited December 30, 2017 by northwestsnow

[bluearmy]

34 minutes ago, Steve Murr said:

No worries -

I thought by now people might trust what I’m saying knowing I’m not making it up

GFS our to 192 - strong NE winds over the UK with the -8c isotherm approaching s

Surprised you didn't post the 12z from the 2nd steve

But then  it wouldn't fit your narrative 

been back and checked very ecm chart from day 9 and two were too progressive without diving/sliding the trough. Gfs was indeed like a dog with a bone but then it overdid the slider too far southwest for several runs within T156. It also overstated the ridging in the n Atlantic. Ecm was never keen on that big area of high heights and that turned out to be much closer to reality. And let's not pretend it was good with the placement of the slider within T84. 

The stat chart posted above was presumably from Tuesday morning ec run where it was way off to the north. 

Neither model is ever right post day 7 and whilst I will agree that ecm hasn't been as reliable as it usually is as an indicator in the 7/10 day period, lauding gfs fo being any better isn't right either. 

[blizzard81]

36 minutes ago, Man With Beard said:

Clusters T264 highlights the ongoing trouble heights to the east are going to cause forecasters - incoming trough responding in a number of ways, all of which affect the UK in very different ways.

Shame the most favourable option for longer term cold is the least favoured. It goes along with my thoughts that this will be at best a very short lived cool/cold easterly. Think easterlies of 87 and 91 as your big fillet steak and this one as your tough rump and only 6oz.