Model Output Discussion 26/11/2012 18Z onwards

[johnholmes]

ECM32 updated overnight

SK

thanks for that update-interesting how it seems to deal with the above average heights in the north

People arent bashing the met office itself. its more the computer model they use and the fact for several days in a row it was showing the very cold air reaching the uk by the weekend which doesnt look like happening now. even last nights fax chart was agreeing with the ukmo model and I understand the fax charts are drawn by a human and not a computer! and I presume more than one human has input into what the final fax chart will look like So that tells us that the chief forecaster agreed that the ukmo model was correct last night,which sadly doesnt appear to be the case.

something you have to live with when a professional forecaster

[Gowon]

Well, I'm certainly with you on that one, Ian. Every year, each and every time Snowmageddon fails to materialize, it's the UKMO that takes the flak. Why not aim it that those who continually ramp everything out of all proportion, as well?

No, folks just need to stop acting like spoilt little brats and learn not to take certain members too seriously.

[johnholmes]

are the anomaly charts wrong though john? a few days ago, when the block looked like breaking down but then rebuilding repeatedly, i had in mind these charts you have been most insistent on. my (novice) thoughts were, if at this time of year, blocking is unusual, then repeated blocking, even if not sustained, would show an anomaly, therefore the charts are not necessarily wrong. also with such wild swings in the models, even if i am mis-interpreting how these anomaly charts work, there is still plenty of opportunity for you to be proven right yet.

Until I've looked closely I am not sure. The anomaly values are day to day based on the long term averages so your suggestion is, to me, probably not correct?

[TEITS]

Good start from the 06Z.

Basically the further S the greater chance of the LP in the Atlantic being a slider.

[nick sussex]

Could be a big snow event coming up, the tension builds!

[Methuselah]

No, folks just need to stop acting like spoilt little brats and learn not to take certain members too seriously.

Ideally, you are of course perfectly correct...but folks like to vent their spleens somewhere...But why must it always be at the UKMO?

And models are called models, for a very good reason: they are models. They are not perfect facsimiles of reality!

[offerman]

Looks like a bit of an improvement for cold on these short charts compared to last couple of days runs. Low developing seems further south on this run but hope its the right run.

[UV-RAY]

Do we ignore the 06z or take it at face value now, all this chopping and changing with which model run is the best is tiresome. Can we not just say they are all worthy of our attention.

[johnholmes]

sorry to go back in the midst of the frenzy with the new run but I posted this in the technical thread I think, 2 days ago

just checked it was 3 days ago and they have been consistent since!

So changes but need a day or two to see if showing less cold/change?

It needs a couple of days with similar outputs to be sure that the signal on the ECMWF is maintained and GFS joins it along with NOAA before any real belief could be given to it. But the possibility of the flow turning more westerly at 500mb across the Atlantic cannot be ruled out.

Edited November 27, 2012 by johnholmes

[AGAL]

Oh the models, the models.

Been following this for over a week now and the last few days reminds me of that great meteorologist Gary 'Barometer' Lineker saying, and I quote?

"The Winter Weather watching Game is all about 3-5 models all trying to compete to show the most outrageous fictional cold synoptic spell ever imaginable, but in ze end, ze Atlantic always wins , ya.

[Frosty.]

Hats off to one or two posters on here who had doubts regarding this cold snap, just shows how the weather can make a fool of us all.

It's all down to expectation levels, the models show mega cold charts beyond the reliable but then it fizzles out as we get closer, it's going to turn colder with widespread frosts and at last relief from the flooding, also a risk of wintry showers grows for northern and eastern coastal areas exposed to the N'ly winds.

[beng]

Mystified by the childish UKMO bashing. This is science, not football. They're fairly happy with how the GM sits now:

" 3. Modifications: The GM’s ideas of a slack and cold northerly over the UK during the weekend is supported by other models. The next Atlantic system pushes in during Monday and this will have a snow risk ahead of it. The GM was accepted without modification"

Complete agree (and I'm quite happy to know the Met Office when it's deserved). I think the UKMO model has performed rather well TBH - it's been very consistent at the 120/144 range. People need to remember that it only goes to 144h - so it's only really on today's chart that we get to see the atlantic attack. I won't be surprised either to see that disrupt somewhat, depending on how embedded the cold can get before we get there.

ECM was getting there with the middle, but lost the plot with the closer range low pressure development off Canada. Each model has its strengths and weaknesses.

[bobbydog]

Until I've looked closely I am not sure. The anomaly values are day to day based on the long term averages so your suggestion is, to me, probably not correct?

would be interesting to know what you find. i just thought that repeated blocking in the same place (rather than sustained) might also show up as an anomaly (just a shot in the dark really)

[TEITS]

Do we ignore the 06z or take it at face value now, all this chopping and changing with which model run is the best is tiresome. Can we not just say they are all worthy of our attention.

All models are worth paying attention to if you wish to make a fair, unbiased assessment of the model output.

Plenty more drama to come I feel over the next few days!

[Matty M]

May not deliver this time but very good trends, Another 100-200 mile shift south-west of the shortwave and the UK&IRE could stay on the cold side of this Atlantic attack and await renewed High pressure build to our north in the cold sector.

[AGAL]

People arent bashing the met office itself. its more the computer model they use and the fact for several days in a row it was showing the very cold air reaching the uk by the weekend which doesnt look like happening now. even last nights fax chart was agreeing with the ukmo model and I understand the fax charts are drawn by a human and not a computer! and I presume more than one human has input into what the final fax chart will look like So that tells us that the chief forecaster agreed that the ukmo model was correct last night,which sadly doesnt appear to be the case.

Yeah but its 'their' computer and it cost the tax payer a hell of a lot of money. Mind you it probably spends most of its number crunching time working out AGW scenarios fifty years down the road.

[danm]

Just as the models have flipped in the last 48 hours to show a less cold evolution, they could flip to show a colder solution once again. It is only Tuesday, and nothing is yet set in stone.

[fergieweather]

I can't understand why people blame the models as if they control the weather. What I will add is just how difficult it is to make an accurate long range forecast for a tiny island in the middle of the ocean, not one person, model or mystic meg, can accurately make a forecast for anything over 14 days away.

It's simply not worth a coronary over model disagreement at T+120. I'd have been writing whining emails to NCEP, UKMO, ECMWF every week for years if it was such a major stress!! Yesterday was MONDAY and the complexity of events by SUNDAY are light years away, in the broad scheme of operational forecasting, with the sort of inter-model disagreement at that range seen yesterday neither unique nor any particular surprise. If UKMO majorly messed-up a forecast at T+12 for a disruptive snow event, that might be a different matter. But it's specifically because of their (correct and prudent) uncertainty heading into the late weekend onwards (as I stressed yesterday) that we haven't bothered with forecasts out to that range in any sense other than cursory mention. Indeed, the nuances of how things will develop then remain enigmatic and not least when trying to afford any public forecast focused on regional nuances.

[Weathizard]

High pressure trying to build into the west coast of Greenland at 132? Looks a much better run this, not brilliant but as has been said if the shortwave is a few hundred miles further s/sw and we'd be in business, it doesn't take much

[Paceyboy]

GFS 06z has better blocking over the arctic around 144 compared to the 00z run.

I really dont think its worth looking past 120hrs at the moment.

The position of the greenland low pressure and the north europe trough are constantly being revised south

Good trends and quite possibly good times!

Edited November 27, 2012 by Paceyboy

[UV-RAY]

It's simply not worth a coronary over model disagreement at T+120. I'd have been writing whining emails to NCEP, UKMO, ECMWF every week for years if it was such a major stress!! Yesterday was MONDAY and the complexity of events by SUNDAY are light years away, in the broad scheme of operational forecasting, with the sort of inter-model disagreement at that range seen yesterday neither unique nor any particular surprise. If UKMO majorly messed-up a forecast at T+12 for a disruptive snow event, that might be a different matter. But it's specifically because of their (correct and prudent) uncertainty heading into the late weekend onwards (as I stressed yesterday) that we haven't bothered with forecasts out to that range in any sense other than cursory mention. Indeed, the nuances of how things will develop then remain enigmatic and not least when trying to afford any public forecast focused on regional nuances.

Top post Ian, one which should be stuck at the beginning of every new model discussion thread.

[nick sussex]

Just as the models have flipped in the last 48 hours to show a less cold evolution, they could flip to show a colder solution once again. It is only Tuesday, and nothing is yet set in stone.

Yes its finely balanced, the GFS 06hrs run is a vast improvement on the 00hrs, this type of set up though is a nightmare to forecast.

However its really all to do with the the angle of the jet into the UK and the shortwave, there is often a tipping point, as soon as the ridge manages to back over the top of the incoming low and you keep some decent pressure in eastern Greenland then the low is stopped from making eastwards progress.

For those worried they might be too far west there is alot of changes that could happen.

[phil nw.]

Just a reminder to stay on topic please.

Some posts have gone because they added nothing to model discussions.

Thank you.

[bobbydog]

Well, I'm certainly with you on that one, Ian. Every year, each and every time Snowmageddon fails to materialize, it's the UKMO that takes the flak. Why not aim it that those who continually ramp everything out of all proportion, as well?

Or better still - just calm down and think about it?

always the way! some people still haven't got the fact that the UKMO model is NOT the forecast. it is a tool for the forecasters to use- they don't necessarily agree with it. the FAX charts are nearer the mark as they have forecaster input. at no point has the Met Office forecast 'snowmageddon'. they have adjusted their forecasts according to the overall output and to my mind have kept things accurate and within the realms of reality. some people just need someone to blame when their hopes fail to materialise.

to say the Met Office got it wrong is simply not true.

[lorenzo]

@fergieweather, just ignore the bashing nonsense. Usually ill informed and sadly one of the negative trends in winter weather watching. Would hate to think that discourages you from posting the much appreciated insight into thoughts from UKMO we are lucky to have.

Not looked at EC 32 yet, advised that signal for cold still strong into week 3... Also mid Atlantic ridge still in play. Less than zonal theme.

GP post nails it re: hemispheric pattern, superb potential here.