Winter Model Discussion - The Hunt For Cold - 12th Dec

[Scoutsy]

My first post on here, so please go easy. I'm still learning to read the models, but I think one thing we've (I've) seen here recently is that we must not forget, ever, that the weather dictates the models, and not the other way round. Things can shift on a dime when we least expect it. We can take nothing for granted.

[SnowBallz]

But we already did! That super computer was bought with taxpayers' money. Agree about the rest.

"That super computer" is used to calculate the whole plethora of NWP models which the Met Office run on the platform. Those models are also then licensed to other meteorological agencies, all around the world, which generates a tidy income.

The next upgrade in 2015 is already being scoped.

[Neilsouth]

"That super computer" is used to calculate the whole plethora of NWP models which the Met Office run on the platform. Those models are also then licensed to other meteorological agencies, all around the world, which generates a tidy income.

The next upgrade in 2015 is already being scoped.

" that super computer was programmed by HUMANS " more like, Humans make mistakes thus so will the computer thus another reason why so many weather models are a load of tosh

[LomondSnowstorm]

" that super computer was programmed by HUMANS " more like, Humans make mistakes thus so will the computer thus another reason why so many weather models are a load of tosh

Do you never just think how amazing it is that we were ever able to build computers that could actually model weather pattern at all, never mind reasonably well to 5/6 days and with a broad idea of the pattern possible sometimes for 10-14 days out based on the current conditions? I find it staggering that it should work at all to be honest, it's just such an incredibly dynamic and in the scientific sense chaotic system that these kinds of long range models are bound to be prone to sometimes large errors, but that they have any record of success at all (ECM 32 dayer has been relatively good at picking out trends so far) is impressive in itself.

[WeatherGuru2012]

The pattern is shifting west nearly every run, as long as we see this in future runs on the models, the more colder options will start to appear on the table. Ita pretty simple if I am honest, thats pretty much what we need. The ecm is a brilliant run for coldies, but its only time, until we start to see some wild cold option showing for next week if i am honest. Just got this suspicion. Lets face it, if that ecm run came of between 168 hours to 216 in particular, it would be nothing like we have seen in recent cold winters. It reminds me of the 80s winters, 1978/79 and the very famous 47, with the succesion of low pressure to the south. I wonder if this might be the year we have a winter like that, where as recent ones have been mainly associated with greenland blocking and northerlies outbreaks.

[SnowBallz]

" that super computer was programmed by HUMANS " more like, Humans make mistakes thus so will the computer thus another reason why so many weather models are a load of tosh

No.

It's because of a mathematical principle whereby error - which given the understanding that chaos theory is accepted to be unpredictable - compounds itself, and consequently grows exponentially. It is for this given reason, that the UKMO is run out to a 6-day boundary; it's because each day, after day 5, the error growth - even if the initial data is only flawed very slightly, is doubled. So, a mathematical and scientific acceptance that you must place governance around 'forecasting', and there is - at the moment - a finite limit beyond which there simply isn't the computing power to break past with any confidence.

Computing power is massive; the 5 day forecast today is as accurate as the 1 day one was, 20 years ago. That progress has only been achieved through breakthroughs in super-computing. Moore's Law predicts computing power to double every 2yrs, so it won't be very long at all 'till that UKMO boundary is pushed to 6 and 7 days. And, by 2020, it's conceivable we'll have confidence out to around 10 days - which, at the moment, is an (accepted) pipe dream.

[Iceni]

No.

It's because of a mathematical principle whereby error - which given the understanding that chaos theory is accepted to be unpredictable - compounds itself, and consequently grows exponentially. It is for this given reason, that the UKMO is run out to a 6-day boundary; it's because each day, after day 5, the error growth - even if the initial data is only flawed very slightly, is doubled. So, a mathematical and scientific acceptance that you must place governance around 'forecasting', and there is - at the moment - a finite limit beyond which there simply isn't the computing power to break past with any confidence.

Computing power is massive; the 5 day forecast today is as accurate as the 1 day one was, 20 years ago. That progress has only been achieved through breakthroughs in super-computing. Moore's Law predicts computing power to double every 2yrs, so it won't be very long at all 'till that UKMO boundary is pushed to 6 and 7 days. And, by 2020, it's conceivable we'll have confidence out to around 10 days - which, at the moment, is an (accepted) pipe dream.

So what about Roger Penrose's hypothesis that we'd only achieve the ability to accurately forecast future events which take place in the universe if we had a computer which encompasses all the information and variables present in the universe? Dr. Phill Ball at Nature did a report into the Stock Exchange and share movements. He concluded that share dealing was a chaotic system and no computer or 'system' could ever predict either trading volumes or prices in a way that would guarantee making a profit. Weather is another chaotic system…

Edited December 14, 2012 by Iceni

[swfc]

Pretty rank oz on all fronts.shocking

[Snowmad79]

This morning GFS 00z, the Good and Bad....

Good: If you like plenty of wind and rain.

Bad: If you were hoping to see progress with energy cutting south east under the block.

Time the GFS model came with content warnings :

"Caution: This run may contain some disturbing images" lol

Edited December 14, 2012 by Snowmad79

[swfc]

Well the PV is supposed to be on the move eastwards and weakening??was or is this forecast in the next month! Looks pretty solid to be fair to the west .have I missed a "trend" again! I do realise"the models are struggling"with the so called PV change but the Atlantic low doesn't really look like sliding does it within the usual reliable time frames !as they say tho its "early days"did I just say that?ill get my coat jeez lol

Edited December 14, 2012 by swfc

[AWD]

Well the PV is supposed to be on the move eastwards and weakening??was or is this forecast in the next month! Looks pretty solid to be fair to the west .have I missed a "trend" again! I do realise"the models are struggling"with the so called PV change but the Atlantic low doesn't really look like sliding does it within the usual reliable time frames !as they say tho its "early doors"did I just say that?ill get my coat jeez lol

I think that's well into next month when the Stratospheric warming's currently being forecast over the Christmas & new year period start coming into affect;

http://cdn.nwstatic.co.uk/gfsimages/gfs.20121214/00/384/npst30.png

Hence why this month is a "take what you can but don't expect anything" month.

Knowing me I'm probably wrong, that's just the way I see it. lol.

[swfc]

I think that's well into next month when the Stratospheric warming's currently being forecast over the Christmas & new year period start coming into affect;

http://cdn.nwstatic.co.uk/gfsimages/gfs.20121214/00/384/npst30.png

Hence why this month is a "take what you can but don't expect anything" month.

Knowing me I'm probably wrong, that's just the way I see it. lol.

sound. Such great forecast by the big guns on here and it's a strat warming in jan maybe! Fantastic and know doubt swfc will be top of league by then! Just a snow shower would do ! Lol

[swfc]

ECM at 144 hrs poor! Pressure to the ene looks poor

[cooling climate]

Things do look pretty grim this morning.Yesterday the models offered us

another tease only to wipe it away this morning adding salt to an open

wound.

Although surprisingly the ECM at t144 still offers us some hope.

[AWD]

I wouldn't be so quick to right off the ECM completely yet.

Don't think it's going to be as good as last nights 12z but looking better than its output of the last few days still.

That aside, the only positive for cold weather I can see this morning is the continued Stratosphere warming forecasts.

[AWD]

It's still there. It's going to be very brief and will quickly get pushed aside, but should provide something Christmassy;

http://www.meteociel.fr/modeles/ecmwf.php?ech=216&mode=0&map=1&type=0&archive=0

[swfc]

It's still there. It's going to be very brief and will quickly get pushed aside, but should provide something Christmassy;

http://www.meteociel.fr/modeles/ecmwf.php?ech=216&mode=0&map=1&type=0&archive=0

hmmm it's 9 days away too but take your point .not a bad end to the run in fi but miles of diff from gfs . Edited December 14, 2012 by swfc

[winterof79]

Whilst ever synoptics remain favourable to our east,the door will not have to open greatly to let the cold in.

http://www.meteociel.fr/...&map=1&archive=0

It will be a testing time for the impatient among us though.I will take a bit of that no matter how brief....

[Snowmad79]

I wouldn't be so quick to right off the ECM completely yet.

Don't think it's going to be as good as last nights 12z but looking better than its output of the last few days still.

That aside, the only positive for cold weather I can see this morning is the continued Stratosphere warming forecasts.

The PV gets massacred on the ECM later in the run and the Babuska Block spawns a Satanic doppleganger prompting round two of block vs atlantic

Edited December 14, 2012 by Snowmad79

[Jason M]

I would urge people to look through the charts themselves this morning as it I'd def not a horror show! There are signs of interest in all the models this morning.

ECM looks good and whilst GFS fails to develop anything there are real signs.

Very interesting output IMO and I wouldn't rule out a mid term easterly suddenly appearing in the next couple of days.

Jason

[AWD]

I think at this stage its all about trends. The last 2 ECM runs have shown something a little more seasonal over the Christmas period rather than mild south westerlies.

Don't look in detail whether its going to snow, or the -5c upper line hasn't quite made it to Southampton yet, as its all FI & it might not even happen yet.

I think the sort of evolution the ECM is showing is the best we can hope for until a Stratospheric warming happens and starts messing with the PV, which isn't forecast until the new year yet.

At least the ECM offers us the potential of something more seasonal over the Christmas period compared to its output a couple of days ago. GFS offers nothing cold.

Edited December 14, 2012 by AWD

[Snowmad79]

I would urge people to look through the charts themselves this morning as it I'd def not a horror show! There are signs of interest in all the models this morning.

ECM looks good and whilst GFS fails to develop anything there are real signs.

Very interesting output IMO and I wouldn't rule out a mid term easterly suddenly appearing in the next couple of days.

Jason

What are the signs of interest in the GFS ? I must have missed them. I know all the models are having serious trouble with the NH pattern from T96 but the GFS is awful. More so because I hate seeing the PV heading to Greenland. At least with the ECM it slows that process down whilst also continuing to pull it apart.

[Ian Brown]

What are the signs of interest in the GFS ? I must have missed them. I know all the models are having serious trouble with the NH pattern from T96 but the GFS is awful. More so because I hate seeing the PV heading to Greenland. At least with the ECM it slows that process down whilst also continuing to pull it apart.

If you look at the GFS and ECM at T192-240, then the differences are relatively minor and it wouldn't take much for either to go the other way in terms of where the boundaries are between the Atlantic and the Continental air. It's only later that the GFS brings a dominant PV into play.

Any thoughts this morning GP ? Other than an each way bet to nothing on Melodic Rendezvous in the last at Cheltenham ?!

Edited December 14, 2012 by Ian Brown

[kold weather]

Exactly, the big difference is just powers up the atlantic low that bit more on the GFS and that allows a greater swing NE of its secondary low, so rather than stalling and trying to undercut like the 00z ECM, it brushes past us and sets us up in to the atrlantic pattern.

Large warming on this run once again at 10mb, that really does look like its going to be the thing that terminates the PV, at least for a little while...

[Isolated Frost]

Along with having -4c 850s into Scotland by t192- the EC delivers a pretty big snow event at D10 on it's op, -6c uppers in the east on this chart-

Quite similar solutions apart from wrt jet energy.. the EC solution is not as organised and therefore wintry for our region as yesterdays 12z suite- though that suite was magnificent and an outlier within the EC ENS.

Another interesting wk to come, if the jet energy dives a la EC, this is one to keep an eye on.