Winter Model Discussion 18Z 12/2/13 onwards.

[Simonl]

Well all I can say is the Gfs Ensembles don't seem to support any Arctic outbreak 850's wise , I was surprised When I just looked at the ECM , although I haven't seen any of the other runs for the last couple of days so don't shoot me down, I would suggest the GFS is going for High pressure over the country with 850's around +2 , And ECM is going for an Undercut with pressure rising further North.. The battle starts at +144 so UKMO tonight may also give some indication.

GFS is on its own with that scenario. ECM and UKMO go for cold from what I can see...

[snowking]

Just to illustrate why some of us are not yet too upbeat with even the ECM Op -if looking for a proper cold and snowy spell.I should add this is more likely to be useful to newer members.

I have at random picked this Feb.2005 setup where the High was in the right place and orientated in such a way to advect the coldest air from the NE and the E around the block.

Now compare the 00z ECM at T240hrs.

We can see whilst still quite cold the block is further south and the coldest air misses the UK and goes into France and further east-drier too.

The pattern is still to evolve of course and maybe the block will regress further north west in time but i hope i have shown here why my recent post was perhaps a little less than enthusiastic.

Let`s hope for some improvements in later runs.

Spot on. Whilst the outlook looks okay to me, this is bearing in mind that I don't mind dry and cold!

The ensemble mean heights are varied today, with the ECM suite very keen towards days 8-10 to transfer the heights towards Greenland, with heights lowering to the south of the UK - this ties in the met office mid range outlook of dry and cold, but increasingly unsettled (from the ECM Mean heights, the unsettled part is across southern counties potentially)

Meanwhile, the GEM and GFS mean solutions keep the heights closer to the N of the UK into the 6-10 day period, though significantly both suites send lower heights down through Scandinavia towards the end of the 11-15 day period as per GFS 0z op) with retrogression of the mean heights towards Greenland.

So then, the dry and cold theme remains favourite into the 6-10 day period, then both suites differ on how early they would allow lower heights to undercut - the ECM op, though still a rather more extreme solution, is gaining more and more support from its EPS suite for this to occur towards the second half of next week, and whilst the GFS/GEM suites take longer, they also arrive at a fairly similar conclusion. Given the superiority of the ECM EPS suite over the GEM/GFS suites (higher resolution, better verification, and significantly more runs!) you would feel that into the second half of next week, though greatly uncertain, something even colder and unsettled could arrive from the NE/E, but as ever it's a route fraught with danger.

In the shorter term once again it's a case of dry and increasingly cooler, some frosty nights I imagine returning through the weekend into next week.

Crucially, any runs showing energy going over the top of the block and sinking it south continue to be outlier solutions, and should be taken with little regard for now (without a significant shift in the ensemble mean projections)

SK

[Glacier Point]

Spot on. Whilst the outlook looks okay to me, this is bearing in mind that I don't mind dry and cold!

The ensemble mean heights are varied today, with the ECM suite very keen towards days 8-10 to transfer the heights towards Greenland, with heights lowering to the south of the UK - this ties in the met office mid range outlook of dry and cold, but increasingly unsettled (from the ECM Mean heights, the unsettled part is across southern counties potentially)

SK

This part important variation from what would appear to be a concensus from extended range modelling.

The state of overall angular momentum favouring a phase 4-8 type evolution in the Global Wind Oscillation (see model summary and technical thread for the height anomalies associated with these two phases).

That suggests a lowering of pressure over Europe and in the Atlantic. So perhaps a reasonable expectation that within 8-14 days we'll see an attempt for lows to swing up from the SW, but equally a strong signal for +ve height anomalies over Scandinavia extending across towards Iceland and southern Greenland (Iceland to Scandinavia I think most favoured). Not a signal for above average or even average temperatures and increasing risk of ppn signal the further sw and in time we go.

[weather eater]

Good post SK, cool/cold and dry looks the order of the day for next week which if we can’t get a decent fall of snow suits me just fine, thereafter much uncertainty, hope it then develops into something colder but expect it not to, meanwhile snow here still not amounting to much.

[Mair Snaw]

Spot on. Whilst the outlook looks okay to me, this is bearing in mind that I don't mind dry and cold!

The ensemble mean heights are varied today, with the ECM suite very keen towards days 8-10 to transfer the heights towards Greenland, with heights lowering to the south of the UK - this ties in the met office mid range outlook of dry and cold, but increasingly unsettled (from the ECM Mean heights, the unsettled part is across southern counties potentially)

Meanwhile, the GEM and GFS mean solutions keep the heights closer to the N of the UK into the 6-10 day period, though significantly both suites send lower heights down through Scandinavia towards the end of the 11-15 day period as per GFS 0z op) with retrogression of the mean heights towards Greenland.

So then, the dry and cold theme remains favourite into the 6-10 day period, then both suites differ on how early they would allow lower heights to undercut - the ECM op, though still a rather more extreme solution, is gaining more and more support from its EPS suite for this to occur towards the second half of next week, and whilst the GFS/GEM suites take longer, they also arrive at a fairly similar conclusion. Given the superiority of the ECM EPS suite over the GEM/GFS suites (higher resolution, better verification, and significantly more runs!) you would feel that into the second half of next week, though greatly uncertain, something even colder and unsettled could arrive from the NE/E, but as ever it's a route fraught with danger.

In the shorter term once again it's a case of dry and increasingly cooler, some frosty nights I imagine returning through the weekend into next week.

Crucially, any runs showing energy going over the top of the block and sinking it south continue to be outlier solutions, and should be taken with little regard for now (without a significant shift in the ensemble mean projections)

SK

Great post SK as always full of explanation for us newer members, my basic knowledge makes me think looking at the models today that the gfs as ever is about to come inline with the ecm with slow backtracks over the coming days or week. think the next 5-6 days of model watching is going to be a real roller coaster with many toys and many prams etc. as for todays IMBY been snowing since 10am and currently have 6cm and still heavy snow falling. 4th major snow event of the winter here with many other small snow falls along the way, so the thought of the ecm verifying and bringing the coldest snowiest weather of the winter will have me tuned into this fantastic thread day and night. looking forward to Steve murr boooomage posts and all the other great posts from respected members.

[summer blizzard]

Having observed the models over several days i'm now ready to state that i am backing the GFS. Ensembles have moved into excellent agreement is my primary reason.

Edited February 13, 2013 by summer blizzard

[Cheshire Freeze]

Having observed the models over several days i'm not ready to state that i am backing the GFS. Ensembles have moved into excellent agreement is my primary reason.

I've seen entire ensemble suites (after showing excellent agreement) shift after just one or 2 runs. IMO ensembles can be pretty useless

[SomeLikeItHot]

Having observed the models over several days i'm not ready to state that i am backing the GFS. Ensembles have moved into excellent agreement is my primary reason.

Huh?

If you are not ready why did you do it? Or didn't you , and you reason for not being ready was excellent agreement?

[SomeLikeItHot]

More generally though ensembles are great for showing a solutions sensitivity to initial conditions, but ultimately they are all produced by they same model. If that model has a bias, all ensemble members will have the same bias.

[summer blizzard]

I've seen entire ensemble suites (after showing excellent agreement) shift after just one or 2 runs. IMO ensembles can be pretty useless

It's possible but with not a single stray run it is in my opinion unlikely, i disagree on the whole they are often great indicators.

Huh?

If you are not ready why did you do it? Or didn't you , and you reason for not being ready was excellent agreement?

Meant 'now', have edited.

[Cheshire Freeze]

It's possible but with not a single stray run it is in my opinion unlikely, i disagree on the whole they are often great indicators.

If the ECM persists watch the GEFS mean start to drop.

Seen it time and time again

Edited February 13, 2013 by CreweCold

[summer blizzard]

If the ECM persists watch the GEFS mean start to drop.

Seen it time and time again

There will no doubt be some GFS runs with a higher ensemble spread but in 6-8 days (19th-21st) i will put a virtual pound on the isotherm being above 0C.

[Captain Shortwave]

Having observed the models over several days i'm now ready to state that i am backing the GFS. Ensembles have moved into excellent agreement is my primary reason.

Interesting point, on the other hand ECM pretty much has unanimous support for it's op run up to pretty much 144h which is completely different to the GFS run

op

http://www.meteociel...1-144.GIF?13-12

mean

http://www.meteociel...1-144.GIF?13-12

This is also backed up by UKMO

http://www.meteociel...44-21.GIF?13-06

Sorry this is a classic case of ECM vs GFS and it's various strengths and weaknesses and drawing a definitive conclusion is usually a bad idea at this stage. I and many others are backing ECM mainly due to the many shortcomings GFS has had over this winter. Of course this doesn't mean it's wrong this time.

Edited February 13, 2013 by Captain shortwave

[Cheshire Freeze]

There will no doubt be some GFS runs with a higher ensemble spread

Yes because every ensemble suite is only as good as the run it originated from. 6-12 hours later they're out of date.

Anyway with UKMO and ECM in general agreement in the mid term I can't see the GFS pulling off some amazing coup.

The Metoffice have far more sophisticated ensemble systems than any GFS or ECM and I remember their video discussing Sunday's snow event (at 48 hrs out) and we were in a zone deemed very likely to see >1cm of snow, along with much of N England. This was a hi res ensemble run, 48 hrs out. It was wrong and the snow was further south and not even 1mm lay here.

Ensembles have limited use in more complex set ups. With a model seemingly not able to get energy distribution correct at relatively short time scales, the ensembles have no chance.

Edited February 13, 2013 by CreweCold

[LomondSnowstorm]

Having observed the models over several days i'm now ready to state that i am backing the GFS. Ensembles have moved into excellent agreement is my primary reason.

The GEFS backs the GFS operational, the ECM ensemble suite (51 members no less) would make the GFS a mild outlier, or at least close to it. Either way you've got a pretty major 'stand off' once again.

[summer blizzard]

The GEFS backs the GFS operational, the ECM ensemble suite (51 members no less) would make the GFS a mild outlier, or at least close to it. Either way you've got a pretty major 'stand off' once again.

Indeed.

Do you have the ECWMF ensemble spread? The mean may be in agreement with the operational but there could well be a bigger spread on the up and downside?

[summer blizzard]

Yes because every ensemble suite is only as good as the run it originated from. 6-12 hours later they're out of date.

Anyway with UKMO and ECM in general agreement in the mid term I can't see the GFS pulling off some amazing coup.

The Metoffice have far more sophisticated ensemble systems than any GFS or ECM and I remember their video discussing Sunday's snow event (at 48 hrs out) and we were in a zone deemed very likely to see >1cm of snow, along with much of N England. This was a hi res ensemble run, 48 hrs out. It was wrong and the snow was further south and not even 1mm lay here.

Ensembles have limited use in more complex set ups. With a model seemingly not able to get energy distribution correct at relatively short time scales, the ensembles have no chance.

That's a pretty poor example though. When it comes to the actual timing or location of a front there's a good margin of error and i pretty much ignore most non-met stuff. This is a much more simple "where will the high end up" and whilst complicated to a degree it is'nt nearly as complex as that setup.

[Cheshire Freeze]

it is'nt nearly as complex as that setup.

No but it is for the GFS!

[Captain Shortwave]

Indeed.

Do you have the ECWMF ensemble spread? The mean may be in agreement with the operational but there could well be a bigger spread on the up and downside?

Here you go, also at 144h

http://www.meteociel...1-144.GIF?13-12

Largest spread for the low in the atlantic and the shortwave dropping south through Scandinavia, the two areas I expect there would be a little uncertainty. Heights building to the north look pretty solid to me.

Just for balance here is the spread at the same time for the GFS 06Z

In the end it's just going with gut feeling. Many think ECM is correct you think GFS, lets see how the 12Z go,

Edited February 13, 2013 by Captain shortwave

[LomondSnowstorm]

Indeed.

Do you have the ECWMF ensemble spread? The mean may be in agreement with the operational but there could well be a bigger spread on the up and downside?

I have the 2m temperature ones for London but the 850 hpas would only be for De Bilt:

The 6Z was markedly colder at ground level than the 0Z though for London so hard to make judgements on the 850 spread, especially with HP in charge giving potential for inversions:

[Tom Jarvis]

What's up with the ukmo it missing some charts

[snowmadchrisuk]

850's Ensembles are not really worth looking at at the moment, as It is all to do with where the High pressure will be placed , there both in agreement as to a High pressure spell for the UK ... Hopefully though if the High places it self correctly over the UK for a few day's there will be an injection of Cold uppers to our East , If the High then proceeds North towards GL there would be enough -10 uppers around to make for a Classic end to Winter.

[Mucka]

What's up with the ukmo it missing some charts

If you are looking on Wetter it pushes out a couple of later charts first, the run is still coming out. Out to 96h on meteoceil.

[Panayiotis]

Few more pushes of heights to the north and the UKMO and GFS could be showing something like the ECM - undercut?

Edited February 13, 2013 by Panayiotis

[Mucka]

I wouldn't really mention GFS and UKMO in the same description at the moment, they are very different by 96h.