Model Output Discussion; into 2015

[bluearmy]

Another interesting end to a Gfs op - this time the gfsp.

Make sure you view the hemispheric option

[Stu_London]

Rare post from me.

 

Don't see a significant cold spell evolving in the first half of January.

 

Bit of a cool zonal set up - nothing blow torch and decent enough for the Scottish Ski industry (although looks quite blowy up there a lot of the time).

 

I'd be surprised if there is much in the way of lowland white stuff in that timeframe.

[Daniel*]

GFS 12z is a lot better IMHO ridge more elongated trying to build towards Scandinavia, I hope this trend continues. Whereas GFSP is very flat. Also +ve heading towards the pole positive steps...As I've said big things have small beginnings, downwelling of the warming also possibly not factored yet. There will be implications the question is if it will result in more favourable pattern for cold in the UK. Just trying to add some balance, much talk about January being a write off - zonal start yes but latter part who knows.

6Z

12Z

Edited December 29, 2014 by Daniel*

[Mucka]

Another interesting end to a Gfs op - this time the gfsp.

Make sure you view the hemispheric option

 

It does seem that we are relegated to deepest FI for signs of a pattern change unfortunately. (current pattern change to mobile set up accepted)

 

That said the strat warming is encouraging as is general view within the strat thread that feedback into the top could be relatively quick.

That still puts any significant developments over 10 days away but I have a feeling that Chinos post might me quite prophetic when he considers the idea of an Atlantic ridge popping up as the new anomaly later into Jan. 

If that coincides with GFSp's idea of an Arctic high moving toward our side of the hemisphere then so much the better.

 

I guess other than hoping for some surprise development in the mid term with more amplification than currently modelled it is a case of looking into FI for a signal that conditions around Greenland will more conducive for high pressure development.

If we do get a significant reduction in the Canadian/Greenland PV then I can foresee the jet being displaced further South  West with increased chances of WAA toward Greenland but little sign of that being modelled as yet. However that will be the area I will be looking at within the the ensembles for any hint of change.

Edited December 29, 2014 by Mucka

[Jason M]

Yep Daniel. Will be interesting to see the ensembles at the same timescale and what they make of it. 

[EML Network]

The 12z GFS Parallel is one mighty storm fest, if (like me), you enjoy extreme's of weather this run is quite exciting. 

 

A continuation of previous runs, once again were looking at 7th - 11th January time frames ...here's the pick of the bunch of Max Wind Speed gusts from the 12z Parallel

 

The whole country would feel this one if it came out as modeled

 

 

and once again the 7th January comes up as a day for high winds, this the day after the chart above !!!

try this for size on the 9th, just 2 days later, this is about as high as wind speeds get for the UK...

 

 

again just 2 days later....

 

 

To balance it out though the GFS Op has backed off a little from it's early doomsday type scenario's and has the strongest winds reserved where you'd expect to see them (I.E NW Ireland and NW Scotland) will be interesting to see how the next 4 or 5 days runs come out, I'm not overly convinced the Parallel is performing better then the Op, are there any statistics on how the Op is performing against the Parallel ?

 

EDIT **

 

Hows this for a MONSTER low pressure system !!

 

Edited December 29, 2014 by EML Network

[mulzy]

About 30% of the GEFS has the ridge that's indicated in the UKMO 144 chart.  Pretty much all of them get flattened out subsequently with the cold air filtering into the Balkans.  However with a little bit more amplification we could be onto something...

 

Clutching at straws maybe but there is a glimmer of hope.

 

Edit: Daniel is alluding to the same issue above!

Edited December 29, 2014 by mulzy

[weatherguy]

Hello everyone, rest assured I've been lurking all winter but due to university commitments I've not really had the spare time to contribute - Hope everyone has had a wonderful Christmas, now it's time to bore you with my 2 cents

 

 

Looking at the last high-res frame of the 12z GFS(P), it's nice to see the PV really looking ragged as the strat warmings appear to take their toll somewhat:

 

 

 

The strat appears to split at 30hpa by T+114 so this is in the (reasonably) reliable timeframe, so it's certainly interesting to see if the effects on the trop continue to show.  The warming looks set to continue and another attack looks set to begin on the Atlantic side, so this may give us some decent synoptics at long last.  No fireworks I'm afraid, but with potential zonal mean winds dropping over the atlantic, any ridges that do form stand a real chance of holding if the HP moves into a favourable position.

 

In the mean time, it looks a wet and windy picture for most with the possibility of some transient snowfall for those in the north and on high ground.

Edited December 29, 2014 by weatherguy

[johnholmes]

Zonal is just the model default pattern, everything goes to default before a pattern change...ie SSW!!!! So no need to panic

 

There is no reality to that remark John, they have no default, other than what the data fed into them and then worked out within the model ends up with what some call default=Atlantic. Default I suppose in that perhaps 70% or more of our weather is Atlantic governed but no model has any default built into it.

[kumquat]

There's still that interesting period coming into hi-res territory where the Arctic High and the Azores look likely to couple up. Currently showing as a kind of col over Iceland at T126. If that link makes it, we are into a different ball park.

 

[Paul_1978]

I'm guessing that as I type you are putting together a post to explain why he might be wrong. Just to start things off and help you along the way, see below for a selection of charts from GFS FI.

 

http://www.meteociel.fr/modeles/gfse_cartes.php?&ech=120&mode=0

 

http://www.meteociel.fr/modeles/gfse_cartes.php?&ech=168&mode=0

 

http://www.meteociel.fr/modeles/gfse_cartes.php?&ech=240&mode=0

 

http://www.meteociel.fr/modeles/gfse_cartes.php?&ech=300&mode=0

 

http://www.meteociel.fr/modeles/gfse_cartes.php?&ech=384&mode=0

 

oh, just to help you a bit further, see the chart below from the GFS(P)

 

http://www.meteociel.fr/modeles/gfse_cartes.php?&ech=300&mode=0&runpara=1

Cheers for that. No further posts planned - I was just surprised in the level of confidence from some posters for a whole two weeks ahead. Anyway, I guess we shall see, strat signals considering and all that! I'll sit tight and keep observing the output ;-) Cheers!

[TEITS]

There is no reality to that remark John, they have no default, other than what the data fed into them and then worked out within the model ends up with what some call default=Atlantic. Default I suppose in that perhaps 70% or more of our weather is Atlantic governed but no model has any default built into it.

That isn't strictly true.

 

If the models work by simply inputting data then the output & performance of the likes of ECM, GEM, GFS would be identical. So why is it the ECM is more accurate than say the JMA or NAVGEM? The answer is how the supercomputer has been programmed to use the data. Like all hardware you need software to run it.

 

So when members say a model is biased then in some respects this is true because the quirks of each model are as a direct result of the programming.

[kumquat]

That isn't strictly true.

 

If the models work by simply inputting data then the output & performance of the likes of ECM, GEM, GFS would be identical. So why is it the ECM is more accurate than say the JMA or NAVGEM? The answer is how the supercomputer has been programmed to use the data. Like all hardware you need software to run it.

 

So when members say a model is biased then in some respects this is true because the quirks of each model are as a direct result of the programming.

Initial start data is crucial. After that it's down to which pressure, temperature and energy equations are used for the iterations over each square patch.

[Kentish Snowman in Yorks]

A rare post from me in this thread.

The GFS does seem to be receiving alot of attention at the moment for it's zonal outlook over the next couple of weeks, with seemingly little evidence to support anything other than a generally Westerly flow.

However, I cannot see any reason why the UKMO is not onto something with its last two runs bringing in a slightly more amplified and subsequently settled pattern around the 120 / 144 mark.

Yes this may get flattened out and a return to a Westerly flow may be expected to follow but that is two consecutive runs the UKMO has thrown out now showing such a pattern (the 0z run at 144 started the shift to a more settled outlook).

I hope I don't get shot down for putting forward an alternative view to the 'two weeks of Westerlys' projected by the GFS but there is nothing to suggest that the UKMO will not verify as it has been fairly consistent so far this winter.

Fascinating stuff that's for sure and who knows if we can get the high pressure in situ over the UK who knows what might develop from there?

Eyes down for the ECM...you never know it might even show some support for its European neighbour?!

[TEITS]

Initial start data is crucial. After that it's down to which pressure, temperature and energy equations are used for the iterations over each square patch.

Yes absolutely but my point is these supercomputers must use software to run the program and it is the software and how it has been programmed. These models pretty much use the same starting data but its the programming that makes the ECM much more reliable than other models.

[johnholmes]

That isn't strictly true.

 

If the models work by simply inputting data then the output & performance of the likes of ECM, GEM, GFS would be identical. So why is it the ECM is more accurate than say the JMA or NAVGEM? The answer is how the supercomputer has been programmed to use the data. Like all hardware you need software to run it.

 

So when members say a model is biased then in some respects this is true because the quirks of each model are as a direct result of the programming.

 

I said no default Dave

[kumquat]

Yes absolutely but my point is these supercomputers must use software to run the program and it is the software and how it has been programmed. These models pretty much use the same starting data but its the programming that makes the ECM much more reliable than other models.

Indeed. And let's not forget that there are brilliant minds at work interpreting and tweeking the programming. The mathematical models behind all current NWP output is pretty much standard fare.

[Kentish Snowman in Yorks]

Further to my earlier ramblings...is it not at all feasible that the Biscay / Iberian high on the chart below gets sucked northwards to join up with the fledgling heights to the East of Greenland to form a more solid block?

Edited December 29, 2014 by Kentish Snowman

[kumquat]

Further to my earlier ramblings...is it not at all feasible that the UK high on the chart below gets sucked northwards to join up with the fledgling heights to the East of Greenland to form a more solid block?

Very much so and an area of high interest.

[bluearmy]

Further to my earlier ramblings...is it not at all feasible that the Biscay / Iberian high on the chart below gets sucked northwards to join up with the fledgling heights to the East of Greenland to form a more solid block?

Only if all the recent x model ens and operationas are wrong!

We have seen ukmo pull a patten change out of the bag a couple of years ago when it dived the trough and split the jet against all other and previous modelling. It was right. Don't see it this time though.

[Jason M]

Cheers for that. No further posts planned - I was just surprised in the level of confidence from some posters for a whole two weeks ahead. Anyway, I guess we shall see, strat signals considering and all that! I'll sit tight and keep observing the output ;-) Cheers!

 

Anytime

 

I don't think its unreasonable to be confident for the next two weeks tbh. We obviously can't say what the weather will be doing a week next Tuesday, but the overall pattern in our part of the hemisphere based upon pretty much all available data is likely to be zonal. That said, confidence and end results don't always match up, so we will see. In addition, I actually hope my own confidence does turn out to be misplaced on this occasion as its a waste of prime winter time!!!

[Frosty.]

I found some wintry GEFS 12z perturbations in the T+168 hours timeframe so potential for snow next week can't be ruled out. :-)

[Long haul to mild]

There is no reality to that remark John, they have no default, other than what the data fed into them and then worked out within the model ends up with what some call default=Atlantic. Default I suppose in that perhaps 70% or more of our weather is Atlantic governed but no model has any default built into it.

Fair enough, however the GFS does have a history of blowing away robust blocks in low resolution by assuming that the Atlantic will conquer all. This tendency was repeatedly proved incorrect in the past (2013 springs to mind as an example?) with breakdowns to mild weather being massively delayed.

[bluearmy]

Models have bias'. NOAA and ECMWF readily accept this. The reasons probably too complex to go into.

12z GEFS - nothing of any note within the next fortnight from a coldie perspective.

However, deep lows and wind /rain + chance of some snow via runners still evident.

Ecm rolling out and I suspect most informed eyes will be well above the trop looking ahead to the second half of jan. The six day chart will carry some interest re the ukmo GM.

[weatherguy]

Further to my earlier ramblings...is it not at all feasible that the Biscay / Iberian high on the chart below gets sucked northwards to join up with the fledgling heights to the East of Greenland to form a more solid block?

 

Is indeed possible, but would take a pretty major change off the east coast - Simply too much energy going over the top there