Model Discussion Thread - 15th Jan

[snowking]

Unfortunately this is a little smaller than I would have liked, but this hopefully shows you the striking similarities between the 0z ECM op and GEFS Mean in the short term - also further showing that the GFS op is really quite far out of kilter with even its own ensembles (ECM is on left, GEFS Mean on right). Subtle differences in the longer term though ideally from a cold perspective you would want the pattern to back a little further west than currently shown (perfectly plausible particularly in light of GP's thoughts last night)

http://i.picasion.com/pic48/5dcf3d9bcd14dc32e30b5e575e25ed13.gif

You can view the comparisons manually here:

http://www.netweathe...modelcomp;sess=

ECM pressure spread in chart form highlights the area between Greenland and Norway that the models are really struggling with, around t120/t144.

http://www.meteociel...1-120.GIF?16-12

http://www.meteociel...1-144.GIF?16-12

Regards,

Tom

Really highlights that most of the uncertainty around this longitude is over the UK so many twists still to endure I'm sure!

SK

EDIT: Link working now!

Edited January 16, 2012 by snowking

[Frosty.]

The ECM Ensembles are very much behind a return to a northwest/westerly flow for all by T144hrs, this does not preclude the possibility of a link to Greenland thereafter.

The Gem 00z was my pick of the models beyond T+144 hours with more amplification than the Ecm and miles better than the gfs 00z, up to T+144 the Ukmo 00z looks best to me, stopping milder air from spilling into the uk this weekend.

[Tim Bland]

GFS OP now falling more in line with its ensembles although it is still on the mild side for 21st through to 23rd. Lots of Colder member appearing in FI now as well. Hopefully the 12Z will fall completely in line with the ensembles and be more Akin to the ECM / UKMO and we can begin to see agreement being reached.

I can see GP is lurking here...Whats your thoughts medium to long term??

Edited January 16, 2012 by Tim Bland

[Buzzit]

Thanks OldMetMan that is one of the most informative posts I have seen in the model thread for quite a while...

My personal view is that the model output last wednesday indicated to me that a pattern change was on the way (by way of examples the GFS 00z of the time had almost all of the UK covered in snow which had not been seen up to that date, and it was from the East, in additon there was (and still is) very signfiicant spread in op and ensembles comparing all models), I believe the issue still yet to be resolved is the exact time of change.

With not very many years of model watching under my belt I am convinced that all of the models handle pattern change extremely badly, which is what I would expect given the low level of weather certainty we are able to achieve past 72 hours, and given the extremely difficult nature of modelling so many variables of which we do not completely understand the interaction of... Even though the modelling systems do not have defined bias, I think there is likely to be an inherent bias due to the nature of the inputs and our understanding of.

[Matty M]

My personal opinion is the continued development of a decent height rise around Greenland around T192hrs is the likely scenario from future output.

A weak easterly around T120hr, then the High will retrogress and build a good block in the Atlantic and allow a decent incursion. I would not expect much from round 1 however.

[Zakos]

I think, we are potentially looking at a significant and very snowy cold spell.

A respectable amount of ensembles have us in an extended deep freeze, from the 19th of Jan right through to the end of the run, with no signs of abating.

What does look very interesting to me is the precipitation ensembles. There is forecast to be quite of lot of PPN around....

Looking at the ensembles, I'd say that if the block does form(looks very likely) Id say Its going to be at least a 10 day event.

Otherwise, It could well turn out something like yesterdays CFS, were we in a 3 week very cold and snowy spell.

Edited January 16, 2012 by Zakos

[The PIT]

I'm intrigued by the temps max and min in the GFS being a constant almost outlier in FI yet the 850 temps don't reflect this varying over time. Something wrong there me thinks. Just doesn't seem right.

[nick sussex]

Oh I see the GFS has come up with a different solution on the GFS 06hrs run and finally seems to have woken up!

The higher resolution output is much more in line with the teleconnections, I do hope we've seen the last of that nonsense it's been dishing out over the last few days. It would certainly be nice too see the models agree this evening within the 144hrs timeframe.

In terms of NOAA this morning apparently they're unimpressed with the GFS 00hrs run:

USED THE 00Z/16 ECMWF FOR THE PRELIMINARY FRONTS AND PRESSURES FOR

DAYS 3 THROUGH 7. THE ECMWF AND UKMET HAVE TRENDED TOWARD MORE

ENERGY FILTERING THROUGH THE SOUTHEASTERN STATES...KEEPING HEIGHTS

LOWER OVER THAT REGION THAN THE GFS...WHICH YANKS THE POLAR FRONT

NORTHWARD OVER MUCH OF THE NATION THIS PERIOD. THE ECMWF HAS BEEN

SHOWING SOME MANNER OF SHORTWAVE HEADING INTO THE EASTERN STATES

FOR SEVERAL RUNS NOW...AND HAS HANDLED THE FLOW COMING INTO THE

PACIFIC NORTHWEST WITH MORE CONSISTENCY THAN THE OTHER MODELS.

[shedhead]

24hrs on I must still confess to having very little confidence in a cold outcome from later next week, despite the slow but steady trend towards that evolution by all the models. I'm certainly not buying into

what NOGAPS and GEM show, simply because they offer the best prospects for cold, much as I was retiscent to buy into ECM's view of things a few days ago for later this week. Still another 24, perhaps

even 48hrs required to know if we can look forward a cold last week of Jan, or will it just be pretty much average fayre? The jury remains well and truely out imo.

Edited January 16, 2012 by shedhead

[Frosty.]

In terms of NOAA this morning apparently they're unimpressed with the GFS 00hrs run:

They are not the only ones, the gfs 00z allowed mild air to spill across the uk next weekend with temps recovering to 8-10c, thereafter, the charts looked aimless and listless until a N'ly blast at the end of next week.

[Tim Bland]

At least the NOGAPS ensembles has confidence in its operation output. I expect the cold to be gradualy upgraded for the period 21-23rd on the 12Z GFS output with everything slightly further west.

[OldMetMan]

Thanks OldMetMan that is one of the most informative posts I have seen in the model thread for quite a while...

My personal view is that the model output last wednesday indicated to me that a pattern change was on the way (by way of examples the GFS 00z of the time had almost all of the UK covered in snow which had not been seen up to that date, and it was from the East, in additon there was (and still is) very signfiicant spread in op and ensembles comparing all models), I believe the issue still yet to be resolved is the exact time of change.

With not very many years of model watching under my belt I am convinced that all of the models handle pattern change extremely badly, which is what I would expect given the low level of weather certainty we are able to achieve past 72 hours, and given the extremely difficult nature of modelling so many variables of which we do not completely understand the interaction of... Even though the modelling systems do not have defined bias, I think there is likely to be an inherent bias due to the nature of the inputs and our understanding of.

Thank you Buzzit. Yes I agree. In a way, I am glad that our current level of model sophistication is as it is, I like to see them get caught out from time to time, most notably by major pattern shifts as you mentioned.

I tend to be rather subjective in assessing the models as I will look at it in terms of how I have seen similar evolutions in the past. I know this is by no means a scientific approach exactly, but I have seen so many weather charts over the years, sometimes it works out!

Yes I too believe models, particularly the GFS, revert to what you could call an "average" circulation - a south-westerly flow in other words. There seems also to be particular points in a run sometimes where a model cannot decide between alternate future synoptic developments and then just reverts to the "average". Just an impression, I haven't a clue in mathmatical or physical terms how they work!

[Buzzit]

Going by the today's 6-15 day Met Office outlook they seem to think that the last week of January will be pretty average with not much in the way of anything particularly cold.

My interpretation on the met 6-30 day outlook is that they are representing the model output, which is so inconsistent that the Met is also inconsistent in its long range outlook, even to the point of suggesting that the north and west will have sunny spells interspersed with scattered showers and then also suggesting in the same breath that the north and west will experience spells of persistant rain.

Until the models provide a degree of uniformity , there will be no point looking at the met long range forecast. I dont think the model output will provide this for another few days at best.

[Duane S.]

Thank you Buzzit. Yes I agree. In a way, I am glad that our current level of model sophistication is as it is, I like to see them get caught out from time to time, most notably by major pattern shifts as you mentioned.

I tend to be rather subjective in assessing the models as I will look at it in terms of how I have seen similar evolutions in the past. I know this is by no means a scientific approach exactly, but I have seen so many weather charts over the years, sometimes it works out!

Yes I too believe models, particularly the GFS, revert to what you could call an "average" circulation - a south-westerly flow in other words. There seems also to be particular points in a run sometimes where a model cannot decide between alternate future synoptic developments and then just reverts to the "average". Just an impression, I haven't a clue in mathmatical or physical terms how they work!

Very interesting posts this morning as always from yourself OMM!

I posted earlier on this morning the about longwave pattern/teleconnections surely getting reflected on the models soon. Lo and behold the 06z hints at just this after, IMO, a woefully out of kilter 00z run.

A surge a WAA feeding on a N or NNW angle from the Azores into southern Greenland is my favoured outcome come next week to me. Nothing leading to sudden massive heights but a building block to a a colder and potentially snowier scene.

[Osbourne One-Nil]

If you could all stick to talking about the model output and use the other available threads for anything else, it would give my inbox a rest!

[andy_leics22]

Very interesting Ensembles for London

Quite a lot of members going for below -5 from the 20th of Jan all the way to the end of the run. Some even hitting -10! It is still 50/50 though really, still some milder Ensembles in there but a lot more colder ones now than there were 2 days ago!

[Radiating Dendrite]

Yes everyone seems to think big of the met office. They're no better than anyone else. As long as that Russian high remains, it could at ANY time ridge over into scandi and at least deliver some snow to eastern areas. To completely ignore this possibility is simply wrong IMO.

Im still waiting for the BBQ summer to arrive....

To counter this, as long as the Jet remains active in the Atlantic it could block any westward progression of heights and leave us in mild mush.

The met-office have access to far more model output than we do on here, so to rubbish any forecast they make is rather foolish as it could be correct as they have been all winter so far to be fair.

The best thing is to concentrate on the model output upto 120hrs and just look for little trends and developments. Anything past this is gaga at the moment with the models really struggling to get a grip on what may occur.

[The PIT]

Ah well since everybody likes deep FI nice little storm in the ECM at t240. Sort of thing GFS tends to do. Pity we don't have a T264 as that chart would be interesting once the low passed through.

[Shunter]

^{Favourable Teleconnections response in todays update ( indeed one member reaches -6 on the} AO ^{index )}

^{http://www.cpc.ncep.noaa.gov/products/precip/CWlink/daily_ao_index/}ao^.sprd2.gif

[Frosty.]

Anyway, there is no sign of anything significantly wintry showing up on the models for our part of the world in the next few weeks, as echoed by the experts, the models show a colder spell returning after the mild blip on wednesday, some wintry showers are possible later in the week and maybe into the weekend but then next week could be fairly average for the uk but at least the frosts should return after wednesday as it turns colder.

[Zakos]

AO forecast best so far this winter, solid support for a negative AO

.http://www.cpc.ncep.noaa.gov/products/precip/CWlink/daily_ao_index/ao_index_ensm.shtml

NAO not so great

http://www.cpc.ncep.noaa.gov/products/precip/CWlink/pna/new.nao_index_ensm.html

But, it is looking roughly neutral, and we dont necasrily need the NAO to go negative for cold. It may even give a good battleground scenario...

[nick sussex]

I think the UKMO further outlook 6-15 days looks okay given the teleconnections. Block to the east, troughing near or over the UK and displaced Azores high.

However within that general pattern the weather you get depends on how far west/east this verifies. That outlook suggests a nw/se tilt to the jet and isn't as easy to model as a west/east zonal flow.

On top of this you have the matter of trough disruption, take the GFS 06hrs run at 192hrs as an example:

You can see there energy trying to get forced under the block, now looking at that chart I'd say theres a very good chance that we'd see some disruption with energy heading se, however as the GFS drops into the lower resolution it resolves this by pushing everything east.

The models are poor at handling these types of situations, especially in later timeframes.

So the outlook still remains to a large degree uncertain if those three factors at the top do indeed verify.

[Guest Kent-Weather]

I don't post here ever to be honest but follow avidly, along with the posts of regular members for their knowledge and explaining charts. And for once I'm going to brave the scary waters!

I'm quite surprised how people are not spotting trends in the closer timescale and more so now that we are beginning to reach T+120. The model output (and more notably the stand off between GFS and ECM) has been purely fascinating to watch over the past 2 weeks overall. But now we are scratching the crumbs from T+120 we can start to endeavour our thoughts on a possible outcome.

My view is as that with low pressure suggested on FAX T120 over the North of Scotland (http://nwstatic.co.uk/fax/PPVO89.png), we can begin to focus on this being our main area of interest. What we also would need to know is where it is going of course..

GFS of recent days has promoted the TREND for low pressure to slide SE near to the E coast of the UK, and in recent days we have seen the all important jet stream become more aligned on a NW-SE angle (http://cdn.nwstatic.co.uk/gfsimages/gfs.20120116/06/102/hgt300.png).

Remember that GFS seemed to have an apparent massive error in previous days (hence the NOAA discussion sending out air recon seemingly to eliminate the error in spinup of GFS and thus sending it on a stand off with ECM - as we have seen), and now NOAA are working to resolve this, it's no wonder we are seeing something a little more settled. But until the error was resolved, ECM became the most likely outcome and was consistent in placing low pressure to the E or SE of the UK.

ECM 00z today: T+96

N. Hemisphere:

And again, but T+120

N. Hemisphere:

We still see that LP is likely to the N of the UK before sliding SE at T+120.

Lets factor in some other models/details shall we?

UKMO T+120

A slower evolution I feel but a general theme (TREND) for LP to be situated N of the UK

UKMO T+144

And then lower pressure further SE in a similar track.

Furthermore... as has already been posted.. The ensembles are all pretty good in their agreement in a drop for the 850hPa temperatures and also pressure out to the 20th

So to summarise. ECM was calling for low pressure to the E days while the GFS/ECM stand off was on going. We now see a blend of GFS/ECM and to some degree UKMO + FAX. Factoring in jet stream + ensembles, we can be pretty confident now surely of what's to happen for the rest of this week which may provide some better interest then we have seen in the past couple of weeks! We were looking at flat, zonal and horrible setups. Now we have something much closer to home so to speak. Wednesday we can then focus on snow potential and indeed if there is any potential.

My only advice would be to look for trends in the models themselves. This is one thing I learnt from this forum, the second was that in complete model contrast, a middle ground is the most likely answer. And I think all of the above writing has shown both learnings to be the best advice ever.

[JonMillar]

For those who wanted to compare equivalent runs these are 00z, and 06z runs for the past three days for the 20th, 21st and 22nd.

2012-01-20 Height 500hpa + SLP GFS 00z Model Output.xls

2012-01-21 Height 500hpa + SLP GFS 00z Model Output.xls

2012-01-22 Height 500hpa + SLP GFS 00z Model Output.xls

2012-01-20 Height 500hpa + SLP GFS 06z Model Output.xls

2012-01-21 Height 500hpa + SLP GFS 06z Model Output.xls

2012-01-22 Height 500hpa + SLP GFS 06z Model Output.xls

[Zenarcher]

Review on the weather models this morning,

GFS

Lately it hasn't been performing good with a lot of outliers in the last few days and this morning it continues to struggle it has good support up to 96 hours but after this it completely loses it. All weather models at 120 hours (Saturday) show the rigding further East meanwhile the GFS has it more West. Also the GFS has a low pressure system just entering the Atlantic again no other models show anything like this at all. The GFS long term outlook shows we would have to wait until the end of the month for something cold but does show a blocking occur at the end of its run.

ECM

Still staying fairly consistent and has good support up to 120 hours. At 144 hours is a very important time at the moment because whatever happens here will affect our weather largely afterwards. The ECM like many other models show low pressure hanging about Greenland and Iceland all of them vary on its deepness and positioning. The ECM tries to make a blocking occur but with the low pressure around up there it isn't letting it happen. After this into the long term outlook the ECM shows a cold pool of air over the UK bringing unsettled and wintry weather.

GEM

It has good support up to 120 hours as well but after that like all models this morning takes a different route. It also tries to make a blocking occur at 144 hours and nearly gets there but with low pressure systems moving around the West of Greenland stops it from getting there. Its long range outlook isn't to bad not great for cold weather in the South but Northern parts could see cold weather.

NOGAPS

For the last few runs its been showing excellent charts and sort of agrees with the UKMO. It does try to make a better attempt of blocking although it looks like a toppler however it would give the UK a longer spell of cold weather.

UKMO

Continues to show great charts this morning and I believe its charts are looking the best of the day so far. It shows a good blocking occur with no low pressure systems around to stop it from breaking away. The UKMO continues to be consistent and refuses to back down. All we can hope for here is its right and the others follow.

Overall this morning we see good agreement up to 120 hours apart from the GFS which loses its mind by 96 hours. After this most of the models show a blocking trying to happen but with low pressure systems hanging about they knock it away. However the UKMO and NOGAPS seem confident on a blocking.

Today to Friday: Today staying settled for most and windy across the Western parts also still cold at times. Tuesday staying settled for the South but unsettled in the North. Wednesday will be unsettled for most with only the far South of England escaping the bad weather most of the UK will see wet and windy conditions with severe gales over Western and Northern Scotland. Thursday becoming more settled for all and turning colder across Scotland. Friday windy across Western parts and unsettled in the North with colder weather across Scotland and Northern England.

Saturday to Monday: Saturday will be windy again for Western parts meanwhile the North, East and South will see colder weather. Sunday looking unsettled mainly over the North and Northwest of Scotland where it will be windy. Staying cold over most of the country with Southern Ireland and South West England just missing out on the cold. Monday looks more settled for all and mainly Northern and Eastern parts seeing the best of the cold weather. The South West should start to see colder weather.

Monday and Beyond: Its hard to say what we can expect but the models show two routes. Either it remains cold and wintry until the end of the month and a bit unsettled at times or we go for a more settled period before seeing colder weather at the very end of the month.