Jump to content

Archived

This topic is now archived and is closed to further replies.

Coast

Winter Model Discussion 12Z 09/01/13> The hunt for cold ?

Recommended Posts

OK, here's a brand new thread to discuss the model output from the main sources and how you see that applies to the weather in the next few days - simple eh?

So if you want to rant, rave, ramp or just post something that has nothing to do with the very specific topic here, you may find your post is moved - you may find it's gone altogether.

Please remember these basic things and try to keep your post enjoyable and pertinent for the large audience this thread attracts; then everybody will benefit from it good.gif

Previous thread here:

http://forum.netweat...cold-is-closer/

Share this post


Link to post
Share on other sites

One thing perturbs me. I’ve quite often seen the GFS runs with substantial differences in FI from the previous run only to revert back to it on the following run. The scenario with the substantial difference might then pop up later on odd runs over the next few days and might even become the adopted trend after a while.

But I’ve rarely seen the GFS produce a strikingly different run from the previous, produce it again and then revert to what it was showing.

So at this point in time, though not without hope, I nevertheless have very low confidence in the GFS 12z backtracking from the last two runs and keeping the Atlantic at bay.

I’m not saying this means the dice will fall the way it shows them to, but I do think, if it does persist with it, that Narnia is out of the question and the best we can hope for is a winter of cold shots rather than anything sustained.

In addition I often hear it said that the GFS is always too ready to empower the Atlantic. Well there’s good mathematical reasoning behind this, but as we have seen lately it can show Easterlies when other models are progressing the Atlantic. So I take any claim of Atlantic bias with as much salt as the Atlantic has!

Share this post


Link to post
Share on other sites

The 06z is as useful as a chocolate fire guard, from my observations I believe the 12z be simiiliar to the 12z from yesterday in the short and medium term :)

All models are good apart from the GFS and also the ENS are good, I'm quite positive! :?

Share this post


Link to post
Share on other sites

Model watching driving me mad,one minute snow the next rain and mild,not sure if any of the

models have a clue which way it is going to pan out.

Lets hope for the later.

What rain and mild?

No thanks! blum.gif

One thing perturbs me. I’ve quite often seen the GFS runs with substantial differences in FI from the previous run only to revert back to it on the following run. The scenario with the substantial difference might then pop up later on odd runs over the next few days and might even become the adopted trend after a while.

But I’ve rarely seen the GFS produce a strikingly different run from the previous, produce it again and then revert to what it was showing.

So at this point in time, though not without hope, I nevertheless have very low confidence in the GFS 12z backtracking from the last two runs and keeping the Atlantic at bay.

I’m not saying this means the dice will fall the way it shows them to, but I do think, if it does persist with it, that Narnia is out of the question and the best we can hope for is a winter of cold shots rather than anything sustained.

In addition I often hear it said that the GFS is always too ready to empower the Atlantic. Well there’s good mathematical reasoning behind this, but as we have seen lately it can show Easterlies when other models are progressing the Atlantic. So I take any claim of Atlantic bias with as much salt as the Atlantic has!

I agree with this.

I think it may run with this for the next few runs and then drop it (if it is barking up the wrong tree).

Very nervous about the 12z!

Share this post


Link to post
Share on other sites

Just to give a view to a question which cropped up towards the end of the last thread...

Can somebody tell me, with regards to models, as there is so many ensemble lines, how does the operational or control, or the pretty maps we see actually get chosen? I assumed that the charts we view would be bang on the "mean" of all the models ensembles but this doesnt seem to be the case?

Well, it all really boils down to resolution.

Both the Operational and the Control are run in the highest possible vertical and horizontal resolution, out to around 192hrs. Thereafter, there is a marked drop in resolution; pragmatic, commonsensical decision considering the fact that error rate exponentially grows - therefore you'd simply be wasting supercomputing resource.

The 20 accompanying ensembles are run in much lower resolution - again, vertically and horizontally. Why? Boils down to a couple of things: i) firstly, cost and time: both are at a premium, and you need to drop resolution down in order to satisfy both requirements, ii) secondly, ensemble guidance is a new approach to NWP - so, at the moment, I think it's still being developed.

It may go without saying, but ensemble collation is the future of NWP; the more scenarios you have, the stronger the mean. Personally, I think their real value will show its hand when they're run in resolution consistent with the Operational and Control, otherwise I think you generate something of a false mean - and inherently you'll encounter outliers far more frequently.

So, nothing is 'chosen' as such: the Operational and Control are fixed, as are its accompanying ensembles. The mean is then drawn according to the cumulative output of all 22 runs.

Hope that helps?

SB

Share this post


Link to post
Share on other sites

Here's a quick sum up of how I'd rate the current models for cold and snow, lets see how they score by the end of the 12z runs.

GFS op - 2/10

GFS ensembles - 7/10

ECM - 10/10

ECM ensembles - 8/10

UKMO - 7.5/10

NAEFS - 7/10

GEM - 9/10

NOGAPS - 8/10

Share this post


Link to post
Share on other sites

Seemed to of lost my other post so maybe I didnt put it up properly. Thank you all for your explanations of how snow is forcast, and fingers crossed reading will get some this weekend. Got to wonder, with so many models pointing towards a cold after the weekend , surely odds would be most likely.

Share this post


Link to post
Share on other sites

This SSW malarkey: Well it would seem that there is definitley some strong correalation between this and hemispheric blocking. Now is this an accepted fact by meteorological organisations and if so is it factored into the modelling process once taking place, or is it still the case that the models are reacting to atmospheric changes that are as a result of SSW? So basically are the models proactively instructed by the SSW data or are they merely responding and always trying to catch up to the data?

Sorry to be OT ................not. blum.gif

Share this post


Link to post
Share on other sites

I wonder if there are any cinemas advertising the rolling out of the 12zs on full screen!:)

Share this post


Link to post
Share on other sites

very silly comments here slating the GFS 06z . we saw from the ens at 18Z yesterday this solution having a lot of support .So although the gfs is not the be all and end all at this stage, it does play a significant role.

Share this post


Link to post
Share on other sites

very silly comments here slating the GFS 06z . we saw from the ens at 18Z yesterday this solution having a lot of support .So although the gfs is not the be all and end all at this stage, it does play a significant role.

Rory your clutching at straws mate its one run from the GFS, its on its own all other major models support a sustained cold solution well in to next week. Mark my words that the next run will fall back in to line with previous thinking. Incredible how fickle some people can be after 1 run which is an outier anyway!

Share this post


Link to post
Share on other sites

Compared to the GFS & ECM:

post-12721-0-19505000-1357739165_thumb.jpost-12721-0-80580600-1357739172_thumb.j

at T48, the NAE has the LP further east, much closer to Ireland than the above;

post-12721-0-67043400-1357739219_thumb.j

What effect this could have down the line I don't know!

Share this post


Link to post
Share on other sites

This SSW malarkey: Well it would seem that there is definitley some strong correalation between this and hemispheric blocking. Now is this an accepted fact by meteorological organisations and if so is it factored into the modelling process once taking place, or is it still the case that the models are reacting to atmospheric changes that are as a result of SSW? So basically are the models proactively instructed by the SSW data or are they merely responding and always trying to catch up to the data?

Sorry to be OT ................not. blum.gif

The stratosphere forecasts are produced by the same models.

They have all of this data factored in.

It seems to be a growing misconception that they wait till a SSW occurs and then suddenly feed all this data in, it is there all along!

Share this post


Link to post
Share on other sites

Its actualy find it quite funny how so many people (myself included) hang so much on the GFS. IMO this is only because it churns out 4 X a day and we get to see all of its data for free. Met office and other profesionals often ignore it. With pracrticaly all other models looking good for cold and snow next week im not that bothered by the GFS OP run. I would be more concerned if it showed consistency and churned out the same on the next 2 runs + had some support from its ensembles. GP says hes 70% behind ECM and 30% GFS so thats good enogh for me.

in the reliable time frame we have some interest at the weekend with cross model agreement for some widespread snow. Ive emailed the metoffice to ask if they are going to issue warnings for Sat Evening / sun morning and cited the UKMO / ECM models all looking like a widespread rain to snow event which could wash any salt / grit off the roads first... just wanted to ruffle their feathers lol. i expect the 330pm text forecast to have the word Snow included where as it currently only says rain for Sat.

there is a real risk of Saturdays front stalling somehwere around the Midlands / South East and it could give a solid 12 hrs snowfall if it does... Couple this with a few hours of rain beforehand then it "could" be a recipie for disruption so I expect some early warnings soon.

Share this post


Link to post
Share on other sites

Rory your clutching at straws mate its one run from the GFS, its on its own all other major models support a sustained cold solution well in to next week. Mark my words that the next run will fall back in to line with previous thinking. Incredible how fickle some people can be after 1 run which is an outier anyway!

It is two runs, the 00z was also not brilliant and it is not an outlier. The run has support from the control run and other ensembles, there is a split in the ensembles.

It may be wrong, but it would be silly to bin it when we have seen similar incidences in the past.

Share this post


Link to post
Share on other sites

Just to give a view to a question which cropped up towards the end of the last thread...

Well, it all really boils down to resolution.

Both the Operational and the Control are run in the highest possible vertical and horizontal resolution, out to around 192hrs. Thereafter, there is a marked drop in resolution; pragmatic, commonsensical decision considering the fact that error rate exponentially grows - therefore you'd simply be wasting supercomputing resource.

The 20 accompanying ensembles are run in much lower resolution - again, vertically and horizontally. Why? Boils down to a couple of things: i) firstly, cost and time: both are at a premium, and you need to drop resolution down in order to satisfy both requirements, ii) secondly, ensemble guidance is a new approach to NWP - so, at the moment, I think it's still being developed.

It may go without saying, but ensemble collation is the future of NWP; the more scenarios you have, the stronger the mean. Personally, I think their real value will show its hand when they're run in resolution consistent with the Operational and Control, otherwise I think you generate something of a false mean - and inherently you'll encounter outliers far more frequently.

So, nothing is 'chosen' as such: the Operational and Control are fixed, as are its accompanying ensembles. The mean is then drawn according to the cumulative output of all 22 runs.

Hope that helps?

SB

My (poor) understanding was that the control run is based on the actual data collated at 00z, 06z, 12z, and 18z and pugged into the compuer forecasting model. These data are then pertubated a little and these also run (albeit at a lower resolution i.e. quicker). The forecasters then use their experience to judge which of the pertubations is most likely to be true and this becomes the operational run.

Please correct me if I'm really off the wall with my understanding.

Share this post


Link to post
Share on other sites

Ok, so what is the difference between the (GOOD) ECM and the (BAD) GFS and what do we need to watch out for in the 12z GFS?

Well, the key is the period between midnight Saturday and Midnight Sunday.

In this time frame the ECM transfers the jet power as one large low pressure into Greenland, whilst the GFS splits this into two separate lows over the Atlantic, driving one into Greenland, but another in our direction.

The ECM's pattern allows the low below Iceland to shoot over the UK and keep the cold, whilst the GFS drives the Atlantic power towards us, pushing this 'Icelandic' low much further North-West.

Which is the most accurate?

Well, looking at the Met Fax, the ECM has the low in the closest position and of similar size to the fax (and compared with the GFS) It also has stronger heights in the Atlantic, so it bodes well that the ECM could be more accurate than the GFS on this basis.

See attached for annotations.

post-9222-0-18693500-1357739468_thumb.pn

Paul

Share this post


Link to post
Share on other sites

Ok, so what is the difference between the (GOOD) ECM and the (BAD) GFS and what do we need to watch out for in the 12z GFS?

Well, the key is the period between midnight Saturday and Midnight Sunday.

In this time frame the ECM transfers the jet power as one large low pressure into Greenland, whilst the GFS splits this into two separate lows over the Atlantic, driving one into Greenland, but another in our direction.

The ECM's pattern allows the low below Iceland to shoot over the UK and keep the cold, whilst the GFS drives the Atlantic power towards us, pushing this 'Icelandic' low much further North-West.

Which is the most accurate?

Well, looking at the Met Fax, the ECM has the low in the closest position and of similar size to the fax (and compared with the GFS) It also has stronger heights in the Atlantic, so it bodes well that the ECM could be more accurate than the GFS on this basis.

See attached for annotations.

post-9222-0-18693500-1357739468_thumb.pn

Paul

Excellent post Paul!

Share this post


Link to post
Share on other sites

Just to give a view to a question which cropped up towards the end of the last thread...

Well, it all really boils down to resolution.

Both the Operational and the Control are run in the highest possible vertical and horizontal resolution, out to around 192hrs. Thereafter, there is a marked drop in resolution; pragmatic, commonsensical decision considering the fact that error rate exponentially grows - therefore you'd simply be wasting supercomputing resource.

The 20 accompanying ensembles are run in much lower resolution - again, vertically and horizontally. Why? Boils down to a couple of things: i) firstly, cost and time: both are at a premium, and you need to drop resolution down in order to satisfy both requirements, ii) secondly, ensemble guidance is a new approach to NWP - so, at the moment, I think it's still being developed.

It may go without saying, but ensemble collation is the future of NWP; the more scenarios you have, the stronger the mean. Personally, I think their real value will show its hand when they're run in resolution consistent with the Operational and Control, otherwise I think you generate something of a false mean - and inherently you'll encounter outliers far more frequently.

So, nothing is 'chosen' as such: the Operational and Control are fixed, as are its accompanying ensembles. The mean is then drawn according to the cumulative output of all 22 runs.

Hope that helps?

SB

I always thought that the control was run with the same resolution as the other ensembles (with the same starting conditions as the operational/deterministic run).

Share this post


Link to post
Share on other sites

Currently the best performing model @ day 5 and 6 is the UKMO, just above the ECM.

post-115-0-28384400-1357740124_thumb.gif

post-115-0-17013800-1357740138_thumb.gif

The accuracy is not correlated on who spot's the trends first but who verifies the most on what day.

Share this post


Link to post
Share on other sites

Interestingly, and further on from my post above, the High Res NAE model already has -7c 850hpa temps hitting the east coast on Friday;

post-12721-0-37110300-1357740100_thumb.j

Compared to -3/-4c by the ECM & UKMO;

post-12721-0-52428600-1357740129_thumb.jpost-12721-0-02860400-1357740136_thumb.j

And by the GFS;

post-12721-0-08148200-1357740157_thumb.j

NAE suggestive of a quicker surge of colder air at the weekend?

Share this post


Link to post
Share on other sites

Steady as she goes ......

The updated NAO spread has not waivered since yesterday...still forecasted to go negative from tomorrow till around 13th before rising slightly and then dipping again to a marginal - value.

http://www.cpc.ncep....a/nao.sprd2.gif

The AO spread however displays "Shannon Entropy" after 16th Jan...anything from a +3 value to a -6.

http://www.cpc.ncep....ex/ao.sprd2.gif

Share this post


Link to post
Share on other sites

Rory your clutching at straws mate its one run from the GFS, its on its own all other major models support a sustained cold solution well in to next week. Mark my words that the next run will fall back in to line with previous thinking. Incredible how fickle some people can be after 1 run which is an outier anyway!

1 run? The ens from yesterdays 18z were a major worry with little support for the op. todays OP in both the 00z and 06z have been poor and although split have some support . If this from the model that started this cold ramp has no merit then people may just aswell draw there own charts. wacko.png

Share this post


Link to post
Share on other sites
Guest
This topic is now closed to further replies.

  • Recently Browsing   0 members

    No registered users viewing this page.

×