Model Output Discussion

[Iceberg]

As long as people look at GFS FI, enjoy it for the pretty colours and then throw it in the bin, they will keep a bit of mental health. But to confirm the Ensembles don't support it, as much as people might like it to the ECM 00Z doesn't support it and rememeber it was only 24 hrs ago that people said that GFS didn't have a clue (at least when it showed mild).

[Pennine Ten Foot Drifts]

...expect one of the next few GFS op runs to be a bad one, expect to see a battle between the models suggesting something close to Bartlett high and some going for a Scandi high...wouldn't shock me to see both come into plat in December...

Exactly, and when it comes down to a battle of probabilities between high pressure to our north east and high pressure to our south east, we all know who the winner is more often than not.

Or maybe the gulf stream is shutting down..............

[nick sussex]

Exactly, and when it comes down to a battle of probabilities between high pressure to our north east and high pressure to our south east, we all know who the winner is more often than not.

Or maybe the gulf stream is shutting down..............

Theres probably a case to be made for having both, euro high edging north, arctic high edging sw, with all the forcing caused by that deep trough i would be surprised if we dodn't see the azores high nudging ne.

[Paul]

Quick preview for all you model watchers, gem is now available on nw - it's an unfinished page currently with just a few charts but we'll add to these and finish it all off in the next few days.

With this model, the 00z run goes out to 240 hours using a handful of data types, with the other data types going out to 120 hours. On the 12z all data types go out to 120 hours only.

View it here:

http://www.netweathe...ction=gem;sess=

For anyone who hasn't used them as yet, the GFS charts are here:

http://www.netweathe...tion=nwdc;sess=

And from there you can also use the buttons to view the ecmwf, uk met, fax, ensembles and also the model comparison page (which will soon have gem on there too)

[Big Innes Madori]

Im probably jumping the gun here but if an E,ly does arrive i.e 06Z then this would deliver far more convective snowfall than if it occured in Feb.

The simple reason is due to the SSTS in the N sea. This is another reason why im so excited with the current F.I charts because -15C upper temps screaming across the N Sea would bring huge amounts of snow.

The same happens around the Great lakes of the US. You will often find lake effect snow fades into the latter half of the winter due to the great lakes freezing up. This is why in some respects I don't like frequent cold N,ly outbreaks during Nov, Dec.

Ah Dave I love reading your posts mate!

While many on here are full time realists and many are doom merchants, you Dave are and have always been the eternal optimist.

We all get to a point in life where we are told that there is no Santa Claus, but yet we all still get presents at Christmas

"-15 Uppers" dare I suggest, you may be referring to the BEAST

Big Innes

Edited December 2, 2009 by Big Innes Madori

[Stu_London]

As long as people look at GFS FI, enjoy it for the pretty colours and then throw it in the bin, they will keep a bit of mental health. But to confirm the Ensembles don't support it, as much as people might like it to the ECM 00Z doesn't support it and rememeber it was only 24 hrs ago that people said that GFS didn't have a clue (at least when it showed mild).

Agree regarding the reliability of FI.

Disagree with the comments on ensembles - 4 or 5 members are showing some sort of coldish easterly, albeit not with the same intensity of the operational run. This far out, with so much that could go wrong with a set up like this, I think that is the best you can hope for.

If the event, or something similar to it were to come off, I would be expecting an additional 1 or 2 members to get on board per day until most were there at around +144.

I would caveat that by saying we have seen similar set ups (especially easterlies) evaporate at shorter notice than that with almost 100% cross model and ensemble support, so I wouldn't be saying anything would be guaranteed until about +36

[Thundery wintry showers]

I have no doubt that the east coast (even the east coasts of E Scotland and NE England) would get a significant snow cover from the charts at the end of the GFS 06Z run. The Tyne & Wear coast just managed it in the case of 2 February 2009, and also (more surprisingly) on 3 January 2008. Further back I remember lying snow right out to the sea front during the easterly outbreaks of 20-22 November 1993, 14-16 & 21-22 February 1994, 26/27 January 1996, 30/31 December 1996, and again on 28/29 December 2005.

850's of -10C or below are almost a sure guarantee that coastal areas end up on the cold side of marginal. In addition with cold upper air, plenty of heavy snow showers would be very likely.

Now for the "down to earth" part- I've seen similar easterly outbreaks progged by GFS to about T+120, a couple in January 2006 and another for late February 2007. In the 2007 instance, the ECMWF latched onto the easterly for just 1 run. As well as the main stumbling block that Steve Murr described (reasonable heights around the Svalbard area encouraging the high and cold continental air to head W rather than S) there's also the stumbling block of the Icelandic Low. In both of the aforementioned cases we did get the high pressure to the northeast but the cyclogenesis to the northwest stopped the cold air from ever reaching the British Isles, and eventually that high slipped southwards from Scandinavia into central Europe.

What the GFS runs do show is that the current pattern is a possible evolution to a cold easterly spell. It is, however, also a possible evolution to a southerly month with the UK spent in a cloudy damp no-man's land.

[Frosty.]

PS: I can sense the imminent return of SATSIGS and Commander Stratos Ferric. Sir, your public needs you!

Yeah where is he....and maybe mr crazy snowman will return if the gfs 12z keeps the trend going for 4 runs in a row!! surely that would be a first

Anyway, back to reality and there is no end in sight to the very unsettled weather with often strong to gale force sse'ly winds and periods of rain, heavy at times and a continiuing risk of localised flooding with a low pressure fueld atlantic throwing bad weather our way for the next 7-10 days at least, maybe hints then of pressure building closer to the east or northeast and more settled conditions but that's beyond the meteorological horizon.

[Just Before Dawn]

Agree regarding the reliability of FI.

Disagree with the comments on ensembles - 4 or 5 members are showing some sort of coldish easterly, albeit not with the same intensity of the operational run. This far out, with so much that could go wrong with a set up like this, I think that is the best you can hope for.

If the event, or something similar to it were to come off, I would be expecting an additional 1 or 2 members to get on board per day until most were there at around +144.

I would caveat that by saying we have seen similar set ups (especially easterlies) evaporate at shorter notice than that with almost 100% cross model and ensemble support, so I wouldn't be saying anything would be guaranteed until about +36

Very good point Stu, particularly the last paragraph.

What we tend to forget is that these are programmes designed to predict highly complex systems, so accuracy, by the nature of the beast, declines from hour +1 and the complexity of the system will have an impact upon how quickly the results of the model depart from reality. As you say, where deeply complex systems with high levels of unpredictable interactions are concerned, accuracy decreases and 'FI' creeps closer and closer to +1z, and these admittedly smile-inducing charts are miles out in never-never land. Personally, my view is that FI is at t+1 hour and that there are degrees of FI - from highly plausible to 'Madder than a box of frogs'.

That said, they are nice to see and, provided that you take the TETIS approach to enjoying them for what they are - outputs of a theoretical computer modelling programme, then have a good time, because those charts past t+264 are about as good as it gets for heavy snowfall, certainly in Eastern Britain, very cold air, warm North Sea SST and an airflow well east of north can deliver big convective snow.

Edited December 2, 2009 by Just Before Dawn

[bluearmy]

cant copy/paste for some reason but its worth looking at the NH plots on the NAEFS ens on meteociel. the anomolies and mean pressures T192-T240 are astounding. somewhere will have a frigid xmas - more likely eastern europe but looking at the positive anomoly over svaalbard and west from there, you've gotta believe the impossible could just be possible.

[BLAST FROM THE PAST]

A messy fax chart at T120hrs.

Shows the disrupted trough over uk with plenty of low centres and fronts.

A wet and windy period for the weekend with some slightly milder air from the South.

High pressure to the North and East that showed earlier in the week now pushed well away to the East for now.

TThat chart highlights very vividly the concerns of mine for a very stormy, very wet period in dec where ffurther flooding is very likely. Just had a deluge here in Redhill...more to come for sure.

BFTP

[cooling climate]

For the last few days I have tended to view the GFS runs on meteociel n/hemisphere charts.

This way you can view the vortex disruption. With heights over Alaska cutting through the

split vortices's and helping to build heights to our north and northeast, I think this puts us 60%

on the way towards a memorable cold spell of weather.

One thing we don't want to see is a piece of energy from the vortex left trailing over our side

of the pole which would scupper any height rises to the north and northeast.

Personally I am optimistic of a very cold spell of weather on the way with the wind flow being

more east, northeast rather than north or northeast.

[weather eater]

As long as people look at GFS FI, enjoy it for the pretty colours and then throw it in the bin, they will keep a bit of mental health. But to confirm the Ensembles don't support it, as much as people might like it to the ECM 00Z doesn't support it and rememeber it was only 24 hrs ago that people said that GFS didn't have a clue (at least when it showed mild).

I have to say about the ensembles, they are as suspect as just going on the operational run, they chop and change as frequently. Leaving aside where this is all going and the wonderful charts, and the fact that most regular members have seen it all before, I don’t think in 5 years of model watching that I've ever seen the GFS come up with three straight runs of this type of evolution that are pretty near identical. Having seen then the 00z follow the 18z I was astonished to see the 06z follow suit. The ECM had other ideas, a cry in the wilderness or the model of sanity, we shall see. There’s going to be a lot of members on board for the GFS 12z. Also Dave’s comments about not in December fails to convince me that FI charts in December have a greater reliance on coming off, all in all its brilliant when model runs get this thread jumping, but an awful long way to go.

[MR EXTREMES]

Wouldn't want to see an inversion with those uppers!

lol dont lie yes you would lol.

im not going to get to involved with the latest candy but russia looks like getting rather chilly.

[karyo]

There are some interesting posts in the stratospheric temperatures thread. Quite a lot to take in but what stands out from what I've read, is Brickfielder's post who mentions that the vortex split is forecasted to occur in a place that is not favourable for the cold to reach the UK. Sounds typical!

Karyo

[Eugene]

Poor runs today looking mild, wet and windy with winds mostly from the southwest, round 1 and maybe 2 to ian brown but the winter is like a 12 round boxing match so a long way to go yet

GFS extreme FI i think is just winding us up, someone from NOAA likes to play games with us i feel

[nick sussex]

There are some interesting posts in the stratospheric temperatures thread. Quite a lot to take in but what stands out from what I've read, is Brickfielder's post who mentions that the vortex split is forecasted to occur in a place that is not favourable for the cold to reach the UK. Sounds typical!

Karyo

Isn't that what happened last February when the forecast extension to the cold spell imploded because a piece of the PV decided to set up shop near Greenland!

However these thing are variable, i don't think we'll know whats happening with the PV till nearer the time, what it does show though is that having a deeply negative AO and split PV simply suggests that somewhere in the northern latitudes of the NH will get some unseasonable cold temps, its to a large degree a lottery which the uk hasn't won for years!

[Stu_London]

However these thing are variable, i don't think we'll know whats happening with the PV till nearer the time, what it does show though is that having a deeply negative AO and split PV simply suggests that somewhere in the northern latitudes of the NH will get some unseasonable cold temps, its to a large degree a lottery which the uk hasn't won for years!

At least we have bought a ticket this time - more than can be said for many of the winters from the late 80s onwards.

[chionomaniac]

Isn't that what happened last February when the forecast extension to the cold spell imploded because a piece of the PV decided to set up shop near Greenland!

However these thing are variable, i don't think we'll know whats happening with the PV till nearer the time, what it does show though is that having a deeply negative AO and split PV simply suggests that somewhere in the northern latitudes of the NH will get some unseasonable cold temps, its to a large degree a lottery which the uk hasn't won for years!

This is what happened last year.

What we should hope for to prevent this reoccurring is that any negative EPO, that allows the PV to drop into America, actually increases in severity as the event progresses which should cause retrogression of the USA trough.

[Tn9]

Ah, but think how much duller life (and this forum) would be if all we had were model runs that verified the vast majority of the time.

Perhaps more scientific, but a lot of the colour and joy (or irrational exuberance, if you prefer) would be lost.

Besides, FI does pick up on trends from time to time and practice makes perfect.

Be terrible wouldn't it,you would flick on to it and go"oh right" and flick off again,no debate,no arguing,no upgrades,downgrades,no FI,no ramping,no whinging,no irony,no sarcasm,no backtracking,no need to keep a beady eye on any oscillating of this and that,one model,one outcome 100 percent right every time,lets hope this never happens,the surprise and fun would go for ever,guess a lot of other people would be well happy though

Least its early in season,even if the wheels fall off its not like a last ditch attempt at the end of march

[Duncan McAlister]

The polar vortex does look to be shifting out of Greenland in the more reliable timeframe (about 144-168 on GFS) which is certainly a good sign and could bode well for some cold later in the month.

Hope the 12z is as good as the 06z, at least the charts are nice to look at even if the eventual weather we get is disappointing

[nick sussex]

This is what happened last year.

What we should hope for to prevent this reoccurring is that any negative EPO, that allows the PV to drop into America, actually increases in severity as the event progresses which should cause retrogression of the USA trough.

Hi C

The latest forecast for the EPO does go deeply negative but then heads back up later.

http://www.esrl.noaa.gov/psd/forecasts/teleconn/images/compare.we.png

Perhaps if we can get the block in place before the pattern flattens upstream.

[Duncan McAlister]

Be terrible wouldn't it,you would flick on to it and go"oh right" and flick off again,no debate,no arguing,no upgrades,downgrades,no FI,no ramping,no whinging,no irony,no sarcasm,no backtracking,no need to keep a beady eye on any oscillating of this and that,one model,one outcome 100 percent right every time,lets hope this never happens,the surprise and fun would go for ever,guess a lot of other people would be well happy though

Least its early in season,even if the wheels fall off its not like a last ditch attempt at the end of march

Agree with you and Polar Gael, it would be winter on here and on TWO without all the speculation, the highs and the lows.

[Thundery wintry showers]

Personally I don't think the GFS past T+180 or T+192 should be removed from its outputs. For example when doing month-ahead forecasts I find the use of the GEFS ensemble maps very helpful to give a good visualisation of the possible synoptic evolutions that may follow the upcoming spell of weather. On occasion the GFS FI can be a very good trend-spotter. For example, the northerlies in late November 2005, early March 2006, and Easter 2008 were consistently projected by GFS right out to T+264.

The problem is that too many people don't see the latter stages of GFS runs for what they are and get far too excited when they see possible cold spells. I used to be like that when I first started- it is so easy to get lured into checking out every run to see if that all-too-elusive cold spell is still there and getting devastated when the little shortwave appears and lands Britain in south-westerlies. A bit of light-heartedness and excitement is all well and good but when it gets out of hand it rather spoils it for those who would like some kind of discussion rather than hysterical despair.

The general consensus, as far as I'm aware, is that GFS out in FI is as good as any other model, but at around T+96-T+144 the ECMWF has the edge, and closer to hand the UKMO may have the edge (at least via the FAX outputs which include human interpretation).

Teleconnections are often explored by the likes of Glacier Point on these threads and they, too, are good at trend spotting- I think their track record is probably better at long timeframes than the GFS and ECMWF, but again, if just one aspect of the atmospheric circulation doesn't play ball then the teleconnection forecast can be thrown at T+240.

Edited December 2, 2009 by Thundery wintry showers

[SP1986]

Ian, I think the graphical representation of the operational output should be removed perhaps although I do realise that it can be 'fun' to look at but shockingly I think it's taken way too seriously (in terms of a few people expecting something to happen baed on it). I agree though that T192+ should remain in the emsemble outputs as the collective nature of these is far more valuable than one single run of course.