Model Output Discussion - The Start Of Winter

[bluearmy]

  MattHugo81

For those discounting earlier tweet regarding all 51 EC ENS members find attached graphic for the 25th... http://t.co/rCG4umaTj8

16/12/2014 21:18

 

no idea what the chart matt has posted is but it surely cant be a huge cluster, given the mean.  very odd

[Freezing-Point]

I think there is an issue with the Netweather site.  I thought I had pressed the forecasting model discussion, but seem to have connected to the matt Hugo  tweet discussion.  I am going to restart my PC and see if the issue is fixed.

[chionomaniac]

no idea what the chart matt has posted is but it surely cant be a huge cluster, given the mean.  very odd

Strangely, it is representative of the 51 ENS according to the ECM - Is the clustering computer driven or human inputted?

[bluearmy]

Strangely, it is representative of the 51 ENS according to the ECM - Is the clustering computer driven or human inputted?

 

 

i think there is a misunderstanding somewhere ed. 42 runs clustered like that would not give this mean. it maybe a cluster but it cant be 42/51 members. perhaps member 42 of 51 (shown) is representative of cluster no 1 ???

Edited December 16, 2014 by bluearmy

[SMU]

Evening All-

 

So we shuffle another day towards Christmas, the models have moved forwards around a day & a half today- meaning that the proposed pattern evolution has now edged into the late part of Christmas day.

 

Ive seen a lot of bouncing around regarding is it a pattern change, or is it not?- well yes it is a pattern change 100%.

 

This is the current running December Anomaly chart from 01 - 14th-

 

now here is the Ensemble Mean from the GFS / ECM & GEM @ day 10

 

This December the core of the lowest heights have been around Iceland- however at day 10 there is nominal ridging into this area- particularly to the west & southern tip of Greenland.

The low heights have moved away into Scandi-    Certainly a more favourable position for cold in terms of the UK, with the caveat that due to ENSO pushing positive more energy trough the subtropical jet there will always be systems coming out of the south of the states trying to slide east under any blocking -

 

So the pattern change looks to have gathered momentum today, its the sustainability of the -EPO high & Atlantic ridge though that possibly supports the longer term change, as it stands the positive heights are not quite full on into the pole through days 11-14 so a temporary distorted jet in favour of cold could be the order of the day rather than a long standing hemispheric change-

This however is yet to be resolved & much dependent on how far south any trough digs in the states.

 

As said the overall pattern is good but not 'classic' the core of the positive heights are being pumped up through the extreme eastern tip of Canada

http://modeles.meteociel.fr/modeles/gfs/runs/2014121612/gfsnh-0-192.png?12

 

& because of the nature of the jet in its undulating mode there will be plenty of CAA moving SE past Iceland into Scotland & Ireland, however we *COULD* see a nose of high pressure off the near continent ( as the jet is trying to return North ) feeding in mild air to the South of the UK.

There may be a 50/50 split.

& also some snow along the boundary - Which is currently right at the tail of Xmas day.

 

This battle/ boundary could have some distance to go in terms of movement possibly North or South- however more likely is a push south & east as the models pick up the amplified pattern more & more - & the emphasis on the jet getting into france is crucial if the whole of the UK wants to get in on the COLD action.

It could take a second system out of the STJ to force some more cold south as we get a second bite of the cherry.

 

As it stands, things finely balanced for a cold snap at Christmas, will *potential* for something more sustained in the mix & a hint of the white stuff now showing all but tentatively in the models.....

 

S

 

PS on the topic of the ECM the ensemble suite is clustered to flow evolution. its very rare to get a 1 cluster suite, usually at day 10 there are 2-4 clusters representing the possible outcomes.

The only reason I can see that Matt has chosen the PTB 42 is its the cluster centroid - IE the mean of the cluster. ( Ie the most representative of all the runs)

 

I haven't really looked at these however I wouldn't be to concerned, I doubt there is NO blocking over Greenland, the focus of the Cluster centroid will more than likely gravitate to the low height location....

Edited December 16, 2014 by Steve Murr

[KyleHenry]

The increased ridging of the East European HP and a movement east with the Scandinavian trough?

Trend of High latitude blocking NW Canada/ E Greenland.

Hints of an East based -NAO.

[Downpour]

Desperate measures by Ian Brown perhaps?

   

 

Just logged back in in desperation following earlier misery to check the pub run and saw the hilarity re: Hugo's tweet. 

[chionomaniac]

get_legacy_plot-web248-20141216211243-15422-0845.gif

 

i think there is a misunderstanding somewhere ed. 42 runs clustered like that would not give this mean. it maybe a cluster but it cant be 42/51 members. perhaps member 42 of 51 (shown) is representative of cluster no 1 ???

No.  At that timeframe there are 51 members that are put in clusters by the ECM and they use one of the representatives to give the best idea of that cluster. In this case for XMAS day there is only 1 cluster with all 51 represenatives of which number 42 supposedly gives the best idea...

 

It looks like a classic #ECMclusterfail as I said, because it does not look representative at all...so no wonder Matt said what he did because on the face of it the clustering suggests that this is the case.

 

I bet that if you ask Nick L about the 23rd he may tell you more about the clustering for that date....

[knocker]

Tonight's extended EPS at T360

 

Azores ridge just to west of UK. Weak trough just east of Scandinavia extending south into Europe. Intensive ridge Alaska and a weak trough N. Canada extending SE into Atlantic with neutral heights Greenland.

 

Edited December 16, 2014 by knocker

[bluearmy]

Steve, those ens mean charts are not ensemble means but the mean of ops days 6/10. JH likes to use them to back up the CPC charts to see if the ops are in line with the ens. I don't like them because if the ops are out of sync with general guidance post day 6, they are useless.

[SMU]

Steve, those ens mean charts are not ensemble means but the mean of ops days 6/10. JH likes to use them to back up the CPC charts to see if the ops are in line with the ens. I don't like them because if the ops are out of sync with general guidance post day 6, they are useless.

No worries - good job the OPs are inline with the ensembles tonight- they look a fair representation...

S

[chionomaniac]

The increased ridging of the East European HP and a movement east with the Scandinavian trough?

Meant to come back to you.

 

No, I can't see any increased East European ridging, the trend as far as I could see was for heights to reduce with the ridge to our south (good) whilst the trough on the eastern side of NA increases in intensity (another good - as it increases amplitude)

[johnholmes]

Steve, those ens mean charts are not ensemble means but the mean of ops days 6/10. JH likes to use them to back up the CPC charts to see if the ops are in line with the ens. I don't like them because if the ops are out of sync with general guidance post day 6, they are useless.

 

hardly useless ba, as with any model output one has to use anything carefully, if one does that then not infallible by any means but a great deal more consistent in terms of accuracy in the 6-15 day time scales than anything else I have come across and tried to use in the 10 years or so on this forum.

Of course any of us on Net Wx can use whatever we feel is best and be judged on the results, that is what makes this forum such a good mix of ideas so long as we keep out the personal digs that some use. I am not accusing you of doing that by the way.

[shaky]

Not a bad 18z gfs and parallel run! ! Pretty similar to 12z up to 190 hours!! Atlantic Ridge and northerly incoming! !

[Barry12]

Not a bad 18z gfs and parallel run! ! Pretty similar to 12z up to 190 hours!! Atlantic Ridge and northerly incoming! !

 

They look poor to me to be honest.

[lorenzo]

Guess it's a night for model mayhem, Don't shoot the messenger...

 

[swfc]

Not a bad 18z gfs and parallel run! ! Pretty similar to 12z up to 190 hours!! Atlantic Ridge and northerly incoming! !

never really get any northerly tbh!!looks so complex on the 18z and so many hurdles to clear

[Mucka]

Not a bad 18z gfs and parallel run! ! Pretty similar to 12z up to 190 hours!! Atlantic Ridge and northerly incoming! !

 

Yeah not bad run to run consistency, it just delays the low to the South and develops it a little less but the would be nice for the South as it drags in cold air from its Eastern flank and mixes in less warm air. A bit silly putting that much detail on FI charts but hey...

 

[SMU]

Nice GFSp Chanrt for the UK

 

 

Very stagnant cold air over the UK with Snow in the NW & SE.!!

http://modeles.meteociel.fr/modeles/gfsp/runs/2014121618/gfs-0-264.png?18

S

[Jason M]

18Z P throws up an interesting little scenario with a low pressure in the channel and entrenched cold air

http://www.meteociel.fr/modeles/gfse_cartes.php?ech=252&code=0&mode=0&runpara=1

Liking this run a lot :-)

Edited December 16, 2014 by Jason M

[Gavin Hannah]

Well thats another run with some decent looking winter weather (Temperature wise).  A slight variation on the 12Z upto around t.126-ish.  So, I would have to say, atm, unless something drastic happens, Xmas looks like being cold for most at least. Snow fall on the other hand, well that depends on how the pattern evolves over the next few days as we edge towards the weekend.

 

 

 Woah....

Edited December 16, 2014 by IBringTheHammer

[Mucka]

Well thats another run with some decent looking winter weather (Temperature wise).  A slight variation on the 12Z upto around t.126-ish.  So, I would have to say, atm, unless something drastic happens, Xmas looks like being cold for most at least. Snow fall on the other hand, well that depends on how the pattern evolves over the next few days as we edge towards the weekend.

 

We are binning GFS 18z op then?

Good.

[Frosty.]

At the very least, the GFS18z operation run shows a much colder spell later next week with arctic air bringing wintry showers and widespread frosts which is a damn sight better than mild swly mush. This run tinkered with a pattern change but then lost the plot towards the end.

[Jason M]

We are binning GFS 18z op then?

Good.

Yes, but banking the parallel

[Snowman.]

 

Ahh a channelish low.

 

Snow for the south huzzah!

 

Going for Scandi heights here.

 

Edited December 16, 2014 by SN0WM4N