Model Output Discussion - Into the new year

[winterof79]

38 minutes ago, Tim Bland said:

GFS & GFS(p) in FI give me a bad feeling....

The short ensembles give me a good feeing. Looking for the end as it begins wil end in FI tears

[feb1991blizzard]

Really not sure how someone can say the GFS trop model and strat model are different, they are not, they are all one model which runs based on laws of physics, its just the boundary layer moves up and down in altitude with different thicknesses, its still all modelled together.

[Met4Cast]

2 minutes ago, Bullseye said:

Should the GFS have the strat warming wrong would the current GFS runs (including last night) be wrong too?

 

The answer is.. yes & no. 

The GFS (the charts we're looking at) have a "lid", they only respond to goings on in the troposphere/oceans & do not factor in what the GFS stratospheric forecasts are showing. So for that reason, it's likely the GFS is wrong in FI because it WONT be factoring in the up-coming SSW. 

So.. yes it's likely the GFS is wrong, but not because it's struggling to resolve the SSW, but because it literally cannot "see' that an SSW is about to happen. Once the SSW is underway and we start to see predictions of downwelling, only then will the GFS (tropospheric charts) being to 'see' possible impacts from the SSW, but we're talking a lead time of perhaps 4 days here, we're well outside of that timeframe at the moment.

I hope that's understandable, I feel like I've just dribbled on a bit of paper and gone "here you go".

[Kirkcaldy Weather]

45 minutes ago, Man With Beard said:

I'm seeing a lot of posts downplaying the set-up in the coming days, some latching onto John Holmes post this morning and I think misinterpreting his intent. For someone seeking cold and snow, there is absolutely no reason to downplay chances in the D4-D8 period, except perhaps for those near the east coast or to the west of very high ground.

The important guiders for me are 1. the mean charts, and 2. don't think you can predict too far out.

Latest GEFS & ECM means for D7 

The key ensemble mean is the ECM really, the GEFS us prone to sudden flips.

When I look at that D7 chart from ECM, it 100% ticks the box. The Atlantic is blocked, and the flow comes from a long way east. I don't care how warm the med is, a flow with 1000 miles over the continent in early January will cool enough to be snow friendly by the time it reaches the UK, except for E coasts on a very slack flow. I do recall a great fail from such an easterly in 2005, but that was late February, where solar heating was already much more significant. In early January, long nights and low solar power allows even mild air from the Greece area to cool by the time it reaches the UK. Also, if it's frozen over Europe, uppers really really aren't as important, far less than on a northerly without very low pressure.

Above this, it must be remembered that these are not usual weather patterns for models, and we must not hold great expectations of what will evolve after such a D7 or D8 chart. Anyone who thinks they can confidently predict a two week freeze even on amazing charts is overstepping the mark. In unusual synoptics, you concern yourself with D1-D6, maybe up to D8/D9 if the mean is really solid and consistent, but anything beyond that is just trends (unlike locked in zonal situations), though some like Caracol, Chinomanic or Snowking have a better handle than others.

I happen to think the D7 means today give plenty of reason to expect sufficient cold/snow may arise. As John Holmes says, not deep cold like Feb 2018 -yet- , and so no North Sea convection, but if frontal activity develops, it may not be needed.

 

This isn’t an argumentative response just want to comment on the highlighted section re North Sea convection, as I posted about at the start of December when the comments about the “lack of cold air” were aplenty and very similar setup last night (still some that continue to misinform about a lack of cold air) check the Scotland thread out Edinburgh got a couple of inches and areas of Fife got good amounts too, even areas further south and west (not all high ground either) all this from 850 hpa temperatures of only -2 to -4, sea temperatures here  around 9-11 degree temp difference and the radar wasn’t far off how it normally looks during a typical cold easterly up here even with such a slack flow, with the outlook showing chances for upper air temps to be double those amounts of the start of Dec and last night (-6s and -8s have been common in recent runs) a bit of simplistic working out suggests chances for convection (given a supportive E flow and troughs etc) could be double the intensity and coverage on radar. A really great time unfolding for us snow lovers and setups that support snowy surprises cropping up pretty much anywhere, exciting times .

Edited December 29, 2020 by Kirkcaldy Weather

[pureasthedriven]

6 minutes ago, Met4Cast said:

The answer is.. yes & no. 

The GFS (the charts we're looking at) have a "lid", they only respond to goings on in the troposphere/oceans & do not factor in what the GFS stratospheric forecasts are showing. So for that reason, it's likely the GFS is wrong in FI because it WONT be factoring in the up-coming SSW. 

So.. yes it's likely the GFS is wrong, but not because it's struggling to resolve the SSW, but because it literally cannot "see' that an SSW is about to happen. Once the SSW is underway and we start to see predictions of downwelling, only then will the GFS (tropospheric charts) being to 'see' possible impacts from the SSW, but we're talking a lead time of perhaps 4 days here, we're well outside of that timeframe at the moment.

I hope that's understandable, I feel like I've just dribbled on a bit of paper and gone "here you go".

Perfectly explained dribbling, thanks. 

[Kiwi]

Apparently GFS and ECMWF  continue to support a cold outlook.

[Paul]

15 minutes ago, Met4Cast said:

The answer is.. yes & no. 

The GFS (the charts we're looking at) have a "lid", they only respond to goings on in the troposphere/oceans & do not factor in what the GFS stratospheric forecasts are showing. So for that reason, it's likely the GFS is wrong in FI because it WONT be factoring in the up-coming SSW. 

So.. yes it's likely the GFS is wrong, but not because it's struggling to resolve the SSW, but because it literally cannot "see' that an SSW is about to happen. Once the SSW is underway and we start to see predictions of downwelling, only then will the GFS (tropospheric charts) being to 'see' possible impacts from the SSW, but we're talking a lead time of perhaps 4 days here, we're well outside of that timeframe at the moment.

I hope that's understandable, I feel like I've just dribbled on a bit of paper and gone "here you go".

As I understand it, It's not that it has a lid so to speak, it does forecast right up into the strat and will therefore being showing the effects of whatever is has forecast in the strat that run. But, with fewer vertical levels, it's essentially less detailed through the vertical layers of the atmosphere than some other models, so may have more of a tendency to bark up the wrong tree. 

Edited December 29, 2020 by Paul

[Paul_1978]

52 minutes ago, feb1991blizzard said:

This has West based -NAO written all over it.

 

That is why the uptick at the end of the eps.

Why are you (and others for that matter) looking for problems 264 hours away and beyond when anything within T+120 is up for grabs, and there a SSW in the offing?

[DEYS(Kent)]

This thread is a stark reminder as to why there are pros and amateurs. As an amateur weather enthusiast I keep my hobby fun and like to keep realistic.

If the charts were not so promising and there was a SSW on the horizon we would be patiently waiting for charts to improve as the downwelling occurred. Let’s give that a try over the next week or so?

[snowice]

6 minutes ago, Paul_1978 said:

Why are you (and others for that matter) looking for problems 264 hours away and beyond when anything within T+120 is up for grabs, and there a SSW in the offing?

Agree it creates unnecessary pessimism however he is only commenting on the output ,but your right it's time for a half glass full approach.

Edited December 29, 2020 by snowice

[winterof79]

1 minute ago, Paul_1978 said:

Why are you (and others for that matter) looking for problems 264 hours away and beyond when anything within T+120 is up for grabs, and there a SSW in the offing?

Praise the lord, maybe some live their lives worrying.

[weathercold]

The models are flipping around with the placement of high pressure at circa 5 days ahead and some are flagging the far reaches of FI for signals of milder conditions.

This is a fluid situation, 5 days is a push for much accuracy let alone 10 days +
Don’t get too hung up on over 5 days ahead, many changes ahead, be it good or bad depending on your preference for cold or milder weather.

[Deep Snow please]

6 minutes ago, Paul_1978 said:

Why are you (and others for that matter) looking for problems 264 hours away and beyond when anything within T+120 is up for grabs, and there a SSW in the offing?

Because they want to discuss the potential evolution of the models in different scenarios as this thread was intended for rather than ignore anything not snowy. 

[Bricriu]

43 minutes ago, Met4Cast said:

If we wouldn't take a day 16 cold GFS chart seriously.. why are we taking a day 16 mild chart seriously? Worth noting the GFS wont be including impacts from the SSW yet, the GFS strat is essentially a different model to the GFS trop and they don't "connect" until downwelling begins to occur. 

Unfortunately the GFS is good at spotting trends. Also, a milder chart is more likely to verify than a cold one. We have seen GFS runs in  the last few days toy with the idea of a west- based nao, so it's not one run in isolation either. I hope you're right that it's an outlier, and that this model is unlikely to have it right due to the effects of a potential SSW , but it would not surprise  me if we end up milder.  We can have all the eye candy background signals in our favour, but it does not guarantee we will get the weather we want unfortunately in this region. Still, at least we have the chance of something much better, unlike this time last year.

Edited December 29, 2020 by Bricriu

[SLEETY]

4 minutes ago, Deep Snow please said:

Because they want to discuss the potential evolution of the models in different scenarios as this thread was intended for rather than ignore anything not snowy. 

But should mention at the timeframe being discussed thats its unlikely to verify anyway,which a lot of posters dont mention ,so it gets confusing!

Since when has the gfs been good at spotting trends lol,its all over the place in fl as it always is,this forum sometimes

Edited December 29, 2020 by SLEETY

[Paul]

26 minutes ago, feb1991blizzard said:

Really not sure how someone can say the GFS trop model and strat model are different, they are not, they are all one model which runs based on laws of physics, its just the boundary layer moves up and down in altitude with different thicknesses, its still all modelled together.

Yep, the GFS forecasts the entire atmosphere in one hit starting at ~0.27hpa at the top and then working down through the remainder of it in 64 levels. Some more info for those who may be interested:

Quote

The atmospheric forecast model used in the GFS is a global spectral model (GSM) with spherical harmonic basis functions. In response to increased computing resources and changing computer architecture at NCEP, the GFS has evolved to higher resolution, both horizontally and vertically, and a more modular code structure. The current operational horizontal resolution is T1534 (T574), or approximately 13 km (34km) at the equator for days 0-10 (days 10-16) forecasts. In the vertical there are 64 hybrid sigma-pressure (Sela, 2009) layers with the top layer centered around 0.27 hPa (approximately 55 km). 

GFS

WWW.EMC.NCEP.NOAA.GOV

---

 

 

[Methuselah]

2 minutes ago, SLEETY said:

But should mention at the timeframe being discussed thats its unlikely to verify anyway,which a lot of posters dont mention ,so it gets confusing!

Touché!

[Sweatyman]

8 minutes ago, weathercold said:

The models are flipping around with the placement of high pressure at circa 5 days ahead and some are flagging the far reaches of FI for signals of milder conditions.

This is a fluid situation, 5 days is a push for much accuracy let alone 10 days +
Don’t get too hung up on over 5 days ahead, many changes ahead, be it good or bad depending on your preference for cold or milder weather.

They cant even sort Wednesday out down south yet!

[Gael_Force]

41 minutes ago, feb1991blizzard said:

Really not sure how someone can say the GFS trop model and strat model are different, they are not, they are all one model which runs based on laws of physics, its just the boundary layer moves up and down in altitude with different thicknesses, its still all modelled together.

Yes, I'm amazed at the reaction to that inaccuracy.

Some models with more layers may be better able to create continuity/accuracy through the levels but to claim they are two separately modelled entities is just plain wrong.

[Man With Beard]

42 minutes ago, Kirkcaldy Weather said:

This isn’t an argumentative response just want to comment on the highlighted section re North Sea convection, as I posted about at the start of December when the comments about the “lack of cold air” were aplenty and very similar setup last night (still some that continue to misinform about a lack of cold air) check the Scotland thread out Edinburgh got a couple of inches and areas of Fife got good amounts too, even areas further south and west (not all high ground either) all this from 850 hpa temperatures of only -2 to -4, sea temperatures here  around 9-11 degree temp difference and the radar wasn’t far off how it normally looks during a typical cold easterly up here even with such a slack flow, with the outlook showing chances for upper air temps to be double those amounts of the start of Dec and last night (-6s and -8s have been common in recent runs) a bit of simplistic working out suggests chances for convection (given a supportive E flow and troughs etc) could be double the intensity and coverage on radar. A really great time unfolding for us snow lovers and setups that support snowy surprises cropping up pretty much anywhere, exciting times .

Agreed! I was thinking of 2018 when I posted, but of course, still good enough for some.

[Paul]

14 minutes ago, Bricriu said:

Unfortunately the GFS is good at spotting trends. 

I think that's a bit of a myth tbh.

All the models pick up on a trend ahead of the others sometimes, but equally they can all also run off in a different direction to the rest only to come running back later on. If I'm honest, since the GFS generally has lower skill (verification) than the likes of the ECMWF and UKMO models, I'd say it's slightly more likely to be wrong when it's out on its own than those other two. That's not to say it always is, and I'm certainly not saying it is for sure this time, as it definitely has its moments. But since we're potentially looking at a strat driven period here, the fact it has a lower vertical resolution than the others adds a bit of weight to the theory that it being out on its own is more likely to be a stroll up the garden path only to return later rather than it setting a new trend that the others will shortly follow. 

We'll see though!

[Biggin]

19 minutes ago, Paul said:

Yep, the GFS forecasts the entire atmosphere in one hit starting at ~0.27hpa at the top and then working down through the remainder of it in 64 levels. Some more info for those who may be interested:

GFS

WWW.EMC.NCEP.NOAA.GOV

---

 

 

Plus other models have more levels. It should be the more levels the better the handle on whats going on.... but..

Edited December 29, 2020 by Biggin

[Met4Cast]

1 hour ago, Met4Cast said:

The answer is.. yes & no. 

The GFS (the charts we're looking at) have a "lid", they only respond to goings on in the troposphere/oceans & do not factor in what the GFS stratospheric forecasts are showing. So for that reason, it's likely the GFS is wrong in FI because it WONT be factoring in the up-coming SSW. 

So.. yes it's likely the GFS is wrong, but not because it's struggling to resolve the SSW, but because it literally cannot "see' that an SSW is about to happen. Once the SSW is underway and we start to see predictions of downwelling, only then will the GFS (tropospheric charts) being to 'see' possible impacts from the SSW, but we're talking a lead time of perhaps 4 days here, we're well outside of that timeframe at the moment.

I hope that's understandable, I feel like I've just dribbled on a bit of paper and gone "here you go".

Turns out this explanation was "dribble on a page" and I had somewhat misunderstood. 

The GFS DOES take every layer of it's forecast into account, however resolution & fewer vertical layers in the model compared with others can give it limited accuracy on modelling tropospheric impacts from SSW's, particularly at the lead times we're currently dealing with. 

ECM/GLOSEA often have a better handle on things.

[Mike Poole]

59 minutes ago, Paul said:

Yep, the GFS forecasts the entire atmosphere in one hit starting at ~0.27hpa at the top and then working down through the remainder of it in 64 levels. Some more info for those who may be interested:

GFS

WWW.EMC.NCEP.NOAA.GOV

---

 

 

According to Meteoceil, the GFS parallel has 127 layers up to 80km - which is one reason why it may have been more consistent re the SSW than the current GFS.

Edited December 29, 2020 by Mike Poole
Deleted random chart

[danm]

3 hours ago, Jason M said:

Some realism has returned this morning I see. The last paragraph of John Holmes post should be pinned to the top of the thread.

For me this has been a really frustrating period of model watching as at face value the output has been amazing, but in reality if you dig a bit deeper it’s simply showing a couple of weeks of rather cold weather. I’m relieved that John has posted because I was starting to feel that I must be going mad especially when other knowledgeable posters were in full ramp mode and I was looking at the charts wondering what the fuss was about. 

whilst that Russian high is pumping warm air up into Europe we are going to really struggle to access the depth of cold needed for a really impressive cold spell. given how static everything is it’s hard to see that changing in the near term so sadly I think we will be chasing marginal wet snow events for the foreseeable. If you live with altitude somewhere though or in the north a good period coming up for you I think.

 

I’d suggest though that some people may not be reading the posts from our more knowledgeable friends properly. Nothing is yet nailed on in terms of deep cold in the reliable time frame, and no one has suggested it is. The talk of something memorable in January is all based on the potential from a range of factors (SSW, EAMT etc) that could deliver something we’ve not seen in many years. The analogy of having a ticket for the raffle has been used. 

So if some people take this to mean that a 62/63 type January is nailed on and they should go and buy a sledge ready for a big freeze next week then more fool them. 

Yes, you are correct that in the reliable (5-7 days) nothing exceptional is yet showing in terms of deep cold, but that isn’t what has been suggested. 

We all need to have a little patience as this is something that could unfold over the next few weeks. 
 

I’d also add that I’m glad we have some supremely knowledgeable people on here sticking their neck out to make a forecast and being brave enough to do so. That’s why we come onto this forum. If all people were interested in was what is nailed on 4/5 days ahead we’d just stick with the Beeb and Met Office and not bother looking further ahead.