General Model Output Discussion

[nick sussex]

please could someone confirm if there is a ukmo 18z. this is really confusing me, i have never heard or seen one?

Yes i provided the link in an earlier post but regret this now as i fear a major influx of calls to the prozac helpline! not withstanding the likelihood that i won't be able to view it in future as it crashes due to the new influx of net weather snow groupies who only appear when theres a hint of snow but don't suffer like the regulars with all the crappy mild dross synoptics of recent winters!

[casparjack]

London GFS 18Z ensembles.

http://www.wetterzen....png?6767676767

Che? Take your pick!

Regards,

Tom.

The Control looks good, I'll take that...

[Cecil Hogmanay Melchett]

I seem to recall saying the GFS ensembles had peaked earlier-

they are now beginning to crash again...

http://www.wzkarten.de/pics/MT8_London_ens.png

People need to stop panicking- we have 2 evolutions here-

1) its nailed, other than boundry of 0c line-

2) risky- but becoming clearer that the beast might be awakening-....

Simples..

S

Steve is there a 18z ukmo model?

On TWO Brian has said the low is much further north?

Is this correct to your knowledge?

[Snowaddict]

I seem to recall saying the GFS ensembles had peaked earlier-

they are now beginning to crash again...

http://www.wzkarten...._London_ens.png

People need to stop panicking- we have 2 evolutions here-

1) its nailed, other than boundry of 0c line-

2) risky- but becoming clearer that the beast might be awakening-....

Simples..

S

Steve, your confidence is very reassuring and I certainly agree about the 18Z ensembles..

However, I think it's vital that we see the 00Z GFS or at the latest the 06Z GFS evolution trend almost exactly towards this evening's UKMO and ECMWF evolutions.. We also need to see UKMO and in particular ECMWF, hold very firm.

M

Edited December 27, 2009 by Snowaddict

[J10]

Yes there is a UKMO 18Z & 06Z but this only runs to +48 or thats all we see anyway.

The mistake some are making with the 18Z UKMO is they are comparing the +48 18z to the 12Z +48. However obviously +48 now makes this Tues 18.00 whereas the 12Z is 6hrs behind. This gives the illusion that the LP is further N.

Indeed, there is a UKMO 18Hz chart.

As for the NAE 18Hz, it looks in a very similar position to the 12Hz run.

[Guest North Sea Snow Convection]

That 18z NAE frame is consistent with the 12z UKMO for the same time. It is six hours later so gives the illusion of the low being further north.

People stop fretting and looking for banana skins !!

[kold weather]

The thing to remember is that the GFS can be way too aggressive in developing mesoscale features and because it does that in this situation it pulls everything further North. It does this many times but of course we are noticing it because of the enormous impact is going to have on our little patch over the coming days.

Almost certainly it is wrong. But we are nearing absolute crunch time, any more confusion and the feature will be at Lands End and we still won't know where it's going.

Longer term, the 18z is as much as a nonsense as the 06z was in terms of its lowering of heights and chaotic pattern.

To be fair to the GFS the UKMO has nearly the exact same solution as the GFS at 48hrs, so its got to have some support now for it.

However you hit the nail on the head with the last part of your post, those heights won't decay nearly as rapidly as the 18z op run suggests, thats unreal how progressive that it is, the GFS ensembles broadly agree as well.

I still think the idea I had the other night will be correct broadly in that we end up with a cold SE airflow, not the big daddy snow event/easterly event but cold and stubborn with the jet tryin to undercut and mean heights to our north/NE. I see no reason to change my punt just because of an overly progressive 18z run.

[snowmadchrisuk]

Incorrect im afraid my friend

There is a (i believe higher resolution) short term version of the UKMO model run known as the NAE model

You can view it up to 48 hours out here:

http://www.weatheronline.co.uk/cgi-bin/expertcharts?LANG=en&MENU=0000000000&CONT=euro&MODELL=nae&MODELLTYP=1&VAR=prec&HH=0&BASE=-&WMO=&ZOOM=0

This outputs 4 runs a day

Kind Regards

SK

Oh , I know about that one , it is part of the UKMO's Unified Model. But it has other models data tied in with it as well.

[cyclonic happiness]

I am not stupid I have been at this game along time . I am just surprised the fronts associated weather fronts are showing up on radar already . Obviously the main low is well behind that .

And to the person who said I am to far North , Every single precipitation chart I have seen on this evenings runs takes the front into the Midlands. Just because the low stalls North of London doesn't mean the precipitation does.

Sorry matey, that was me trying to make humour but failing lol

You are quite right that the front is lying over the M4 corridor is good for the Midlands region as most of the precipitation would be ahead of the warm front, what we don't want is the front pushing anyfurther north than shown on the faxes or this will push the colder uppers back north a long time before the actual front arrives.

The ensembles have put a little more wind in my sails now though and I can go to bed (after this beer) with a smile on my face....(owww errr missus!! matron! put them away!)

[Geordiesnow]

please could someone confirm if there is a ukmo 18z. this is really confusing me, i have never heard or seen one?

They do but it only goes up 2 48 hours and i have no idea how accurate it is but giving it being such a small run, then i would imagine it would pretty accurate.

if the UKMO 18Z is showing the low being further North, then that is not good news and it would make the 23:57 forecast not that interesting but we shall see.

[TEITS]

Steve is there a 18z ukmo model?

On TWO Brian has said the low is much further north?

Is this correct to your knowledge?

Like I said I think Brian isn't taking into account the 18Z is obviously 6hrs ahead of the 12Z and if hes comparing charts then this gives a false illusion.

Speaking of Brian if this cold spell goes pearshaped I know who to blame. If you see his homepage you will see hes putting a curse on this. At least there isn't any polar bears.

[mountain shadow]

Anyone believing the 18z GFS Ops run over everything else needs their heads testing.

Remember that the only reason we look at GFS so much is because it provides vast amounts of information for free. If the ECM or UKMET provided equal amounts of data we wouldn't touch GFS with a bargepole.

You can see why the GFS is free.

[snowmadchrisuk]

Sorry matey, that was me trying to make humour but failing lol :lol:

You are quite right that the front is lying over the M4 corridor is good for the Midlands region as most of the precipitation would be ahead of the warm front, what we don't want is the front pushing anyfurther north than shown on the faxes or this will push the colder uppers back north a long time before the actual front arrives.

The ensembles have put a little more wind in my sails now though and I can go to bed (after this beer) with a smile on my face....(owww errr missus!! matron! put them away!)

Well we await the 00z . This is one situation where for the Midlands I really really can't call it. I wouldn't even like to guess. Expect rain , if it Snow's it is a bonus.

[Snowaddict]

18Z NAE is also very similar to ECMWF precip extent at that timeframe, just a little bit further north...

Look at the low also, it is almost identical... the NAE has just shifted things a tad further noth is all and these variations will occur..

Calm, Calm :lol:

[SMU]

Steve is there a 18z ukmo model?

On TWO Brian has said the low is much further north?

Is this correct to your knowledge?

Yes there is- It runs to 48 ( as I posted - it used to be on the tropical site which was really handy- but its gone now:( )

Here it is at 48- ( which was the 12z 54)

http://expert.weatheronline.co.uk/daten/proficharts/en/nae/2009/12/27/basis18/euro/prec/09122918_2718.gif

Problem is we dont have an UKMO 54- so its VERY difficult to compare- remember the key being UKMO had the front moving North through 48 out to 60-

SO the 60 chart is here-

http://expert.weatheronline.co.uk/daten/proficharts/en/ukmo/2009/12/27/basis12/euro/pslv/09123000_2712.gif

Not to much difference really...

Steve

[casparjack]

Well we await the 00z . This is one situation where for the Midlands I really really can't call it. I wouldn't even like to guess. Expect rain , if it Snow's it is a bonus.

I think that is absolutely spot on and exactly what I am doing for the first part of the set-up. When the front starts to retreat I will begin to look out of the window again...

[pyrotech]

Indeed, there is a UKMO 18Hz chart.

As for the NAE 18Hz, it looks in a very similar position to the 12Hz run.

To try and clear this up for new members the NAE charts are high resolution and used extensively but not exclusivly by bbc forecasts, so the meto seem to take notice of the nae along with other model data when compiling the fax charts. You will notice often the similarity with NAE and the fax which has human intervention. The NAE is a useful tool at short range.

http://www.weatheronline.co.uk/cgi-bin/expertcharts?LANG=en&MENU=0000000000&CONT=ukuk&MODELL=nae&MODELLTYP=1&BASE=200912271800&VAR=prty&HH=48&ZOOM=0&ARCHIV=0&WMO=

Edited December 27, 2009 by pyrotech

[john w]

People

The METO output a model to t+48 at same intervals as GFS (Some tell me the model actually goes to T120)...for some reason I dont know why. That model follows with a textual update to some govt agencies with associated confidence bands.

Remember that T+48 on the 12z will become T+42 on the 18z run...look at T+36 on the 12z NAE output and then T+30 on the 18z...the low is ever so ever so slightly more elongated but virtually in the same position latitude.

Im not sure however if NAE outputs have ensembles??? Certainly never heard of them.

Noticed on the NEWS24 that the forecast showed snow for many more places further south than the M4....cut it very very fine though and this reiterates a similar wobble of uncertainty entering language in meto outputs which was not present so much yesterday.

GFS really is on a limb now but something tells me not to ignore it.

Edited December 27, 2009 by john w

[BLAST FROM THE PAST]

Just a quickie, models are upgrading cold as anticipated...I would like to issue blizzard and severe cold warnings for 1st 3 weeks of January. Jan 87?.........yes very much possible. This is the biggie folks, you know I like to forecast cold but we have the daddy coming...and I'm dead serious. The overall pattern of very cold setup is in place just detail to nail but Blizzard like conditions for some and severe wind chill are to be experienced, a stronger cold wave than December. Bin the GFS as it is way off the mark and stick with the ECM....although it seems OP loves the GFS.

BFTP

Edited December 28, 2009 by BLAST FROM THE PAST

[kold weather]

Like I said I think Brian isn't taking into account the 18Z is obviously 6hrs ahead of the 12Z and if hes comparing charts then this gives a false illusion.

Speaking of Brian if this cold spell goes pearshaped I know who to blame. If you see his homepage you will see hes putting a curse on this. At least there isn't any polar bears.

Its not that much different in terms of the leading low BUT the big difference is the low to the west which is the one that causes issues on the 18z GFS op run, its more developed on the 18z run and also importantly more to the SW and as we saw on the 18z GFS op run if it takes too long to get east then its not going to get far enough east to drag the colder air down.

In that set-up, it then becomes simply about the raw power of the -ve NAO signal shunting the cold air and the jet southwards...

Not the greatest trend as it does really complicate things and probably would postpone any real cold shot for a good 3-5 days...BUT if we get a east based -ve NAO we'd bre mightly unlucky not to eventually see the cold air take over, just may take longer then the intial idea if that were to be the case.

[Optimus Prime]

Just a quickie, models are upgrading cold as anticipated...I would like to issue blizzard and severe cold warnings for 1st 3 weeks of January. Jan 87?.........yes. This is the biggie folks, you know I like to forecast cold but we have the daddy coming...and I'm dead serious. Blizzard like conditions and severe wind chill.

BFTP

Oh dear please get some perspective!

It's looking very much like February 2005 looking at the models.

Edited December 27, 2009 by Optimus Prime

[Duane S.]

Is that +48 chart of the METO 18z showing our LP system trying to spawn another one? If it did get 'promoted' to a proper closed circulation LP then it wouldn't be too dissimilar to what the GFS is showing. My guess is though that it will simply fill before it even has a chance to get going. Hopefully...

[cyclonic happiness]

Oh dear please get some perspective!

It's looking very much like February 2005 looking at the models.

But without the mild biased baggage of that god awful time :lol:

[casparjack]

Yes there is- It runs to 48 ( as I posted - it used to be on the tropical site which was really handy- but its gone now:( )

Here it is at 48- ( which was the 12z 54)

http://expert.weathe...122918_2718.gif

Problem is we dont have an UKMO 54- so its VERY difficult to compare- remember the key being UKMO had the front moving North through 48 out to 60-

SO the 60 chart is here-

http://expert.weathe...123000_2712.gif

Not to much difference really...

Steve

It looks to me like the first one has the low further North and deeper, 980 mb, where the second one has it futher SW and not so deep? As you say without supporting charts either side it's difficult to make a judgement.

Another issue for me is that there has to be a great deal of uncertainty whether it is snow or rain. It will surely begin as rain for anyone South of Birmingham because there isn't the cold air in place. Then in the section from Mid Wales/Brum up to Manchester, assuming the PPN gets that far, it's marginal.

Veterans of the eighties will remember events where snow was forecast and it was rain and vice versa. I'm racking my brain for an event in 84/85 that saw snow fall all day here but not stick until late in the day when dewpoints fell. It was rain around Wolves/Birmingham I believe until the early evening.

Having grown up in Shropshire during the late 70s and early 80s, I know exactly what you mean, I can remember a similar event when wet snow fell in huge flakes all day then only started to settle in the mid afternoon. Wish my memory was better.

[Altostratus]

As I’ve mentioned before, I think the GFS data issue is being overstated. I think this started as a rumour and everyone seems to have jumped on the bandwagon. The recent poor performance is probably more due to the GFS being unable to get to grips with the large amplifications and subtle jet splits. Blocking makes modelling of the jet streams much more complex (splits, large amplifications etc). As such, there is more to model and more that can go wrong; particularly for the UK, where seemingly small adjustments have a large impact on dictating our air mass.

With a run-of-the-mill zonal pattern, it is much more forgiving and the GFS seems to perform much better. Even during these periods however, it has been noted that the GFS has a history of over developing areas of low pressure.

With the GFS being an American model I guess the majority of investments are focused on improving North American modelling. In theory, the top Euros should be better for here - which I recall is supported by the verification data.

If the GFS turns out to be right on this one, serious questions must be asked of the other models.