Winter Model Discussion 06Z 10/01/13>

[Yarmy]

At T48 the ECM/GFS/UKMO/GEM are all very similar:

http://www.meteociel.fr/modeles/ecmwf/runs/2013011000/ECM1-48.GIF

Given that these patterns are unlikely to change significantly, why are we confident to see a resolution tonight or even tomorrow night? At this point on, it looks like a bifurcation to me:

http://en.wikipedia.org/wiki/Bifurcation_theory

I'd hazard that we might even have to watch the whole thing unfold in real time, before we know for sure (the NOW model ).

[ITSY]

GEM fantastically cold supporting ECM all day long.

SO TENSE

personally (i have already given reasons why) i am expecting GFS to win but HOPE not

[Kentish Snowman in Yorks]

Well it's all getting rather exciting isn't it?!

In the red (mild corner) we have the GFS and the majority of it's ensembles and number one fan Ian Brown, and in the blue (freezing cold corner) we have the majority of the rest of the output, the METO, Steve M and the entire Netweather Forum!

Who will prevail? There really is only one way to find out......stay tuned to the 12z folks!

On a serious note, I noticed that somebody posted this chart earlier and it got me thinking;

Is what determines the outlook for next week as simple as a race between whether the two highs link up first or the two lows win the race. I.E. if the Azores and Scandi highs link up we get the Atlantic or if the French and Icelandic lows link we get the freezer?

Probably a very simplistic way of looking at things but it was just a thought!

In terms of the comments regarding ECM and GFS outputs, I know some are saying that ECM is heading in the direction of GFS and vice versa, and I would not like to call it at this stage. I would say however that the METO seem extremely bullish about the cold / snow arriving next week, so that would be (tentatively) where I would put my money if pressed.

The other thing that has got me thinking today is GP's comment from yesterday;

Posted by Glacier Point on Yesterday, 11:56 in Forecasting Model Discussion

If I could allay a few fears here..

The 10 hPa and 100hPa operational 06z still looks consistent with overall strat developments, and general signal for ridge in the NE Atlantic locale.

t300 GEFS mean height anomaly very well defined, ridge signal eastern side Greenland and deep trough over Europe including UK.

The operational GFS and some ensemble members maybe picking up a reasonable idea for shortwave development between Iceland and Greenland, which makes next week's forecast that bit more problematic. Clearly the quicker the cold uppers become embedded the better for snow prospects next week. Interestingly NOGAPS which had championed this idea has now binned it 06z. Just demonstrates the massive spreads that currently exist in the North Atlantic in the medium range and large scale inter-run disagreement in NWP.

That doesn't alter the general trend here. Cold to begin this weekend, continued signal for a slight moderation in this before cold becomes more aggresive once more, timed for second and most significant downwell from the SSW. Note how the GEFS and also ECM ensembles begin to re-intensify cold beyond day 10.

Bottom line - don't mistake a tendency to moderate cold next week (and undoubted model variability to expode that out of all proportions) with a failed cold spell.

If we take GP's comments at face value, even if the GFS has the pattern correct in the 72/96h timeframe, we are surely not dead in the water from a cold perspective surely? The updated 6z GEFS whilst not brilliant IMBY do show at least the prospect of conditions that are conducive to a further cold / snowy spell with a couple heading down towards the -15 mark out in the extended range so all may not be lost.

This chart from the end of the GFS 6z op is also not the worst chart in the world and could easily lead to a much colder scenario in the longer term;

Anyway, that's my two pence worth. I could be barking up completely the wrong tree, and I have waffled on far too much, but all I will say is that this afternoon's 12z suite really will make for some fascinating viewing...

[Wilxy]

Surely if the GFS has played this correctly and the other models are playing catch up the GFS would end up being a bit more respected.

thats 5 runs now u can't hide that i can't see us getting prolonged cold.

And the GFS has a bit more support today, as NOGAPS tags along.

Edited January 10, 2013 by TomW

[Guest bjaykent]

There has been a lot of reference today about not being able to ignore the GFS due to a number of consistent runs showing a quick return to mild after the weekend. Could you not use the same argument to support the Gem which has since the shortwave was first modeled consistently moved it south east due to the north east flow around the top of the Atlantic high pressure.

In fact it has been far more consistent than both GFS and ECM which have both flipped alarmingly before settling on the completely different outcomes for next week.

I don't say that the GEM route will be the outcome as there is only a small margin for error in its evolution but at least it has stuck to it consistently..

[Paul]

Seems like some posters seem to feel the need to post every 10 minutes regardless of whether they have something new to add to the discussion. As a guide, if you've made your opinion clear once and nothing has changed, there's really no need to keep banging on the drum with the same lines but in a slightly different order - it's adding nothing new, isn't any more likely to make people listen to you or change their own opinions so why not wait until you have something new to add.

[mulzy]

One US model that firmly backs the European models is the CFS:

108: Shortwave tracks SE across the country:

132: Resulting in...

288: Crazy FI

[mountain shadow]

Remember the GFS Ops also disagrees with around 1/2 of its ens. at day 4, so does,t even agree with itself.

The updated NAE is a big bonus and with latest data from the 12 o clock input surely the GFS Ops will come on board

[Sugarloaf80]

I am a definate newbie but one thing that I have noticed as an argument for the GFS being the correct solution is it's consistency over a number of runs. However, as GFS runs 4 times a day it shows twice as many solutions over the same time period. Therefore, 5 GFS runs is the equivalent of 2 UKMO or ECM runs. From what I noticed the UKMO and ECM have been pretty consistant for at least 3 or 4 runs. This is more consistant than 5 GFS runs. I'm not saying the GFS is wrong but sometimes it is given more wieght as it runs 4 times a day.

[karlos1983]

Well I'm going to miss the 12z due to a meeting, fairly relieved about that. Here's to hoping the GFS comes back in line with ECM/UKMO. I look forward to reading all comments and viewing the Models after. Good luck all!

[stainesbloke]

Had a couple of texts from friends/family - 'why haven't you told me about this' etc....

Told them milder on Wednesday.

I can't believe they can readily disregard the GFS when the ECM could easily have gone that way and I believe will - later.

Agree. The GFS appears to be handling things far better than the other models this winter, just like in December. Any SSW effects will take longer to show up I reckon but for the medium term I'd say it will be a return to milder conditions.

[Timmytour]

Wouldn't it be fantastic to have the frames of the 12z rolling out live on TV with a panel there to discuss the implications? Pre- rollout discussion about where the panellists think it will be going, analysis on what they think it's doing comparing it to the previous run and elaborating on the differences from runs 24h previously, and post rollout analysis on what the implications of the runs will be.

Use the Match of the Day studios....replace Lawrenson and Shearer with Ian Brown and Steve Murr!

Highly anticipating these runs,,,, i imagine this whole scenario will be referred to for many a year afterwards on these forums.

[Weather Boy]

There has been a lot of reference today about not being able to ignore the GFS due to a number of consistent runs showing a quick return to mild after the weekend. Could you not use the same argument to support the Gem which has since the shortwave was first modeled consistently moved it south east due to the north east flow around the top of the Atlantic high pressure.

In fact it has been far more consistent than both GFS and ECM which have both flipped alarmingly before settling on the completely different outcomes for next week.

I don't say that the GEM route will be the outcome as there is only a small margin for error in its evolution but at least it has stuck to it consistently..

I don't believe it is correct to say that the GFS has been consistent. If you look back at John's post #15 on page 1 of this thread: http://forum.netweat...ion-06z-100113/ he quotes Gibby's excellent model summary as being a complete turnaround from the GFS. Although I haven't and don't check every model run, I think that this demonstrates that its consistency has only been since last night.

Even if it has been consistent, the experts tells us that consistency does not equal correctness (although I've said before that I struggle with this notion).

I am going for the ECM, failing which something that is nearer the ECM than the GFS. Why? Well, in one corner we have the GFS with a degree of support from its ensembles and Ian Brown. Whereas in the other corner we have the professionals (the Met office), ECM (Strongly supported by its ensembles), UKMO, GEM, CMA, BOM, GP.....I could go on!

Whilst of course the GFS could be right and win all the other ones round, like a rogue juror after a trial, I think the weight of opinion and evidence against it is immense and I am struggling to understand why anyone would think that the GFS is likely to be right, let alone with the ridiculous degree of certainty that some on here seem to have. The Met office don't. The fact that it was right in December, in my view, carries very little weight at all as we are in a totally different environment now, synoptically and with the SSW.

Edited January 10, 2013 by Weather Boy

[Just Before Dawn]

Stepping aside from the GFS vs ECM/UKMO free-for-all for a moment, and looking back on the 06 GFS Op....

I think it might have been Kold that identified the winter storm progged to cross the country at T+170-T+186 - aside from the storm-force westerlies and heavy rain for the SW (really not what that part of the world needs, given the recent flooding issues), as the system moves out into the North Sea, there's the potential for some petty nasty conditions over the hills of northern England and South central Scotland overnight on the 17th-18th - snowfall over 2000 feet in the Western Lakes, Westmorland, the North Pennines, Cheviots and mountains of central Scotland, with blizzards over the higher routes.

We all know the GFS's tendancy to blow up LPs to biblical proportions out in LaLa land, but should the GFS verify, perhaps something to keep a bit of an eye on.

[Met4Cast]

Agree. The GFS appears to be handling things far better than the other models this winter, just like in December. Any SSW effects will take longer to show up I reckon but for the medium term I'd say it will be a return to milder conditions.

I'd agree with you IF the GFS ensembles were supportive but at the moment they're not, with even high resolution models disagreeing with the GFS now I think it's a pretty safe bet that the GFS has this wrong.

Everyone keeps bringing up December but that doesn't mean anything, this is a completely different situation and setup

[Jason M]

Remember the GFS Ops also disagrees with around 1/2 of its ens. at day 4, so does,t even agree with itself.

The updated NAE is a big bonus and with latest data from the 12 o clock input surely the GFS Ops will come on board

I think you raise an important point which I made very early this morning. GFS doesnt even agree within its own ensemble suite. I looked through them this morning and only three or four agreed with the opp by day five. Now that is crazy and probably the lowest confidence i've seen in an opp run for a very long time. By FI many were heading zonal but it was the crazy sort of flat as a pancake zonal that low res churns out when its clueless.

I made a call this morning for the METO simply because the issues are at 72 hours (less now!) and METO runs at higher res upto 72 hours in my understanding. Pretty hard though when all three main models have completely different set ups at day 5!

Fascinating side show this evening with IB vs Steve Murr. IB called December but Steve made a call re the Feb cold spell last year at 72 hours when all three models were against him (and gave reasons at the time). Exeter have more info than us and were now in range of the short term models that we dont get access to. Media communications are therefore telling at this late stage.

One last point. Fergie made it clear yesterday that Exeter do consider GFS. He essentially said they use all tools available. Very scientific and sensible IMHO. I only repeat it now as some may not have taken on board that message amongst the carnage on here yesterday.

Jason

[BLAST FROM THE PAST]

I think all options remain on the table here and I doubt if even 12z will have this resolved just yet

BFTP

[stainesbloke]

I'd agree with you IF the GFS ensembles were supportive but at the moment they're not, with even high resolution models disagreeing with the GFS now I think it's a pretty safe bet that the GFS has this wrong.

Everyone keeps bringing up December but that doesn't mean anything, this is a completely different situation and setup

Everyone is bringing up Decembers fail because its all happening again. Agree that it's a different setup but a mild outlook looks increasingly likely in the medium term to me with snow potential for this weekend downgrading too. Not what I'd like to see of course! But am keeping my feet firmly planted rather than get carried away.

[Paul]

Latest thoughts and weekly forecast from Michael Fish - more questions than answers for next week at this stage...

[LomondSnowstorm]

I'd agree with you IF the GFS ensembles were supportive but at the moment they're not, with even high resolution models disagreeing with the GFS now I think it's a pretty safe bet that the GFS has this wrong.

Everyone keeps bringing up December but that doesn't mean anything, this is a completely different situation and setup

Agreed, it's all very well saying that the GFS has handled a past event better but it's not even a case of the euros v the GFS, it is every major and minor model bar NOGAPS along with every single (AFAIK) mesoscale model and the vast majority of ECM ensembles +MOGREPS basically modelling things very differently to the GFS upstream.

[nick sussex]

Yes nick just seen this and he seemed confident about cold lasting into next week,

Also the new model you posted from the us, is it a high resolution model?

Yes its the high resolution ensemble system from the NAM, its latest run from 09hrs backs up the earlier 03hrs, thats the latest available data available to us before the evening model runs.

[Frosty.]

Agree. The GFS appears to be handling things far better than the other models this winter, just like in December. Any SSW effects will take longer to show up I reckon but for the medium term I'd say it will be a return to milder conditions.

Well I expect the gfs to do a number of backflips in the next few runs and show something like this, nearly all the other models, the meto and experts like GP are going for cold and snow. It looks like by sunday and monday there will be a lot of snow showers pushing into the eastern side of the uk with significant falls, then more of the same for the rest of next week with lots of frost and ice.

[Met4Cast]

Everyone is bringing up Decembers fail because its all happening again. Agree that it's a different setup but a mild outlook looks increasingly likely in the medium term to me with snow potential for this weekend downgrading too. Not what I'd like to see of course! But am keeping my feet firmly planted rather than get carried away.

Can you post some charts to backup why you think mild is more likely? From where i'm sitting, it's everything against the GFS.

If everything else was showing mild but the GFS was showing cold, would you still favour the GFS?

[Weather Boy]

Everyone is bringing up Decembers fail because its all happening again. Agree that it's a different setup but a mild outlook looks increasingly likely in the medium term to me with snow potential for this weekend downgrading too. Not what I'd like to see of course! But am keeping my feet firmly planted rather than get carried away.

No, that's not right (yet).

In December, the 18z led the way at close range and the others fell into line very quickly. We now have a stand off over a reasonable period.

However, even if you were right, it would not scientifically and evidentially be sound to use that as evidence to suggest that the same thing will happen again. That's the equivalent of saying that because you flipped a coin and it was heads, it's bound to be heads again. Clearly that is not the case, although I accept that the analogy is not perfect as it is not totally random.

[Snowy Liverpool]

Can't ignore the slight backtrack from the ECM. Yesterday the cold comfortably made it, today if a 200 mile shift North happens on the current forecast then all of the UK is mild. It was on the mild side of the ensembles though for the crucial time (t96-t144). UKMO is a strange one, the shortwave moves Northeast, but the high pressure makes it North too. No other model is going for this, got to say it looks great at t144. Would that be a possibility? The other models that have shown the shortwave moving northeast have then gone on to end any hope of cold.

Edited January 10, 2013 by Snowy Liverpool