Model Output Discussion 1200 hrs 15/11/14

[johnholmes]

just going back to my post last night - see above. The ECM run this morning goes a long way to proving my point about it's over amplification issues. And this was at T144 to T168.

 

I do not doubt you but to show just what you mean can we see the charts that illustrate this, like the EC 12z yesterday and the 00z, both for the same date/time please?

It might also be worth an e mail with this data to send to UK Met for them to pass to their forecast team for comment?

[nick sussex]

Wow the ECM was wrong within T120hrs! we're not even talking problems past T144hrs, you'll see the upstream pattern is different this morning aswell as the blowing up of the shortwave near the UK. Typically the UKMO has now dropped the bloated shortwave and trended towards last nights ECM, the ECM goes in reverse.

 

If you get this phasing then energy will spill over the top and you're more likely to end up with a MLB so just surface cold on offer rather than being able to tap into deeper cold, unless you get a renewed dig of mid Atlantic troughing later on.

 

Looking at the outputs there still seems to be uncertainty over the eastern USA, and this is in relation to how the final pulse of Arctic air works se into the ne USA, this is relevant to how the jet exits that region and also to any high pressure ridge ahead of the next upstream low.

 

These differences upstream hardly inspire much confidence in the outputs:

 below the UKMO at T120hrs:

 

 

ECM

 

 

 

Until the models agree upstream then its best to wait to see what that consensus is, of course this morning more notable cold hopes have taken a knock but thats par for the course in here.

 

Theres still a good chance of it turning colder but today the more surface cold option is the likelier compared to any deep cold from the east. I'm not sure I quite buy this Atlantic attack.

 

Perhaps an overreaction by the models, we'll see...

Edited November 17, 2014 by nick sussex

[Polar Maritime]

Yes the METO look to have called it right and those vin the West especially will be fearing a repeat of last winter.

 

Carrying on from BA, There is absolutely nothing for anyone to 'fear' about regarding this 'coming' Winter. Far far to early for these statements, Of which are not suited in this thread. No more comments on this please, Let's move swiftly on..

 

Back to Model Discussion please.

Edited November 17, 2014 by Polar Maritime

[TEITS]

I feel the safest prediction based on the model output is a stalemate between the blocking to our E and the Atlantic with neither really dominating our weather. So weather wise this would result in SE,lys with E/SE likely to remain dry but extreme W areas might be threatened by rain at times.

 

Looking at the GEFS SLP mean for my location suggests pressure will remain relatively high above 1010mb.

 

https://max.nwstatic.co.uk/ensimages/ens.20141117/00/t850Cambridgeshire.png

 

However this stalemate could easily swing in either direction in the models from a colder E,ly to a full blown attack from the Atlantic.

[Gibby]

HERE IS MY LATEST ANALYSIS FROM THE OUTPUTS OF THE WEATHER COMPUTER PREDICTION MODELS FOR TWO WEEKS FROM 8AM TODAY MONDAY NOVEMBER 17TH 2014.

NEXT UPDATE TUESDAY NOVEMBER 18TH 08:00

THE CURRENT GENERAL SITUATION Slack Low pressure and troughs over Southern England will slip South and fill with a light SE flow developing from tomorrow.

MODELS-2 WEEK HEADLINE  Becoming generally less unsettled for a time with a lot of cloudy but dry weather, Perhaps becoming more unsettled again later from the West.

THE JET STREAM ENSEMBLE FORECAST The GFS Jet Stream Ensemble forecast shows a split flow this side of the Atlantic as the Southern arm is deflected South to the west of the UK while a detached northern arm moves East over the Arcctic Circle above Europe. This complex pattern persists with the later feed taking the Southern arm into a single flow running NE to the West of the UK late in week 2.

GFS OPERATIONAL The GFS operational this morning shows a persistent block of High pressure over Eastern Europe preventing Atlantic Low pressure from running it's normal course across the UK with it's normally attendant wind and rain. instead only Western Britain look like seeing occasional rain as fronts buffer the block to the East with a lot of dry and benign conditions for the East. It is not until late in the run where a more concerted push of intense Low pressure to the NW finally nudges the block further South and allows a windier SW flow to drive rain and strong winds across all areas in average temperatures.

THE GFS PARALLEL RUN The GFS Parallel run shows less dominance from the European High later this week as Atlantic troughs push further across the UK than is shown from the operational meaning all areas will see some rain by next weekend. Pressure then builds from the South for a time with a quieter drier interlude before the Atlantic pushes back again towards the end of the run with more wind and rain late in the period.

THE GFS ENSEMBLES The GFS Ensembles show more bias towards it's parallel run this morning as it builds High pressure North across the UK in a weeks time after a brief wetter spell around the weekend as a trough moves slowly East over the UK. Some frost and fog might occur for a time before the second half of the run sees a rebuild of intense European High pressure with a benign and quiet NE airflow developing for all under a strong ridge over Scotland with any rain restricted to the far South and SE from Low pressure over nearby Europe.

UKMO UKMO today shows the coming weekend becoming more unsettled as Low pressure moves slowly NE across the North carrying troughs of Low pressure over the UK with temperatures close to average.

THE FAX CHARTS The Fax Charts show the current complex Low pressure near Southern Britain giving way to a SE airflow with troughs across Western and SW Britain filling in situ midweek with a new advancement of troughs towards the far West by the weekend.

GEM GEM this morning shows the block of High pressure to the East of Britain maintained over the next 10 days with repetitive attempts of troughs coming up against it from the Atlantic over Western Britain and slowly crossing the UK while weakening. This means the West will see occasional rain early and later in the period possibly extending to other areas at times later in the period as the block slips a little further South and weakens somewhat.

NAVGEM NAVGEM shows similar evolutions to the GFS Ensembles with the stand off continuing between Atlantic Low pressure to the West and High pressure to the East and later South. Generally mild SE winds will carry benign and quiet weather for many in average temperatures through the rest of this week and into the weekend with just the chance of a frostier night or too as the new High from the South moves in just under a week from now.

 

ECM  ECM this morning also shows the UK weather dominated by the stand off between High pressure over Eastern Europe and Low to the West over the Atlantic. The resultant SE flow will deliver benign and quiet weather through the latter stages of this week with a band of rain crossing slowly East towards the weekend decaying as it does. Then a decidedly sharp dip into unsettled, windy and wet weather is shown by Day 10 as deep Atlantic Low pressure slips ESE towards Britain with strong winds and heavy raine extending to all by then.

THE ECM ENSEMBLES The Ensemble run this morning's indicate a trend towards Low pressure gradually becoming more dominant again later next week as Low pressure out to the NW pushes South or SW winds across the UK with troughs delivering occasional rain from troughs in relatively mild conditions.

NOTABLE TREND CHANGES FROM PREVIOUS RUNS The trend continues to show High pressure over Europe maintaining benign and stable weather across the UK in average temperatures but with some inroads of Low pressure from the West likely late in the period.

MY THOUGHTS High pressure over Europe remains the dominating feature of the weather shown by the models over the next few weeks. It's positioning and orientation is such that cold weather in this part of the world is unlikely as the mainland of Western Europe remains on the mild side of average given the time of year and the drag of mild SE winds from Southern latitudes exacerbates this fact through this week. The net result for the UK is a lot of cloudy but relatively dry weather very common though a little rain from decaying troughs from the West hitting the High block is likely at times through the next week. There will be a noticeable SE wind at times too which will prevent any clearer night skies from delivering any widespread frost and fog with the net difference between night and day temperatures being narrow. Then as we move into the second half of this morning's output the trend appears to be for the block to the East to slowly loosen it's grip with the chance of a more concerted push of Atlantic domination developing slowly. ECM goes for a deep Low just to the West by Day 10 with wind and rain for all while many other runs offer something similar if in a more diluted form. What no model shows is any guaranteed shift towards anything majorly cold for the UK with temperatures never far from average if not slightly above. In my own opinion I think the theory of a change to a more Atlantic based wind and rain type pattern to end the month is a strong one and what's more I can eventually see the slack Jet through this week strengthening to the West and North later overriding the block over Europe and pushing it South later next week. While this may break us out of the blocked pattern the danger from this is that High pressure could become High from a belt near the Azores to the Meditteranean with a marked broad and mild SW flow developing across the UK to start December. This is my theory based on weak signals at present but the building blocks to this scenario are shown in more than one FI chart I have seen this morning so it is a possibility.

Edited November 17, 2014 by Gibby

[Smiler1709]

WE STILL ARE NOT IN WINTER

The forecast from the latest models show a cool down with night frost and cooler days, if you ask me that already is a huge step up from last year, the winter pattern going into the new year is so much better than the outlook last year.

So take a deep breath and chill.

 

Edited November 17, 2014 by Polar Maritime
To remove off topic comments.

[MR EXTREMES]

just going back to my post last night - see above. The ECM run this morning goes a long way to proving my point about it's over amplification issues. And this was at T144 to T168.

you could say this but then the other models were pretty close to the ecm in there ideas to....

 

but i dont see the point in the met office outlook being apart of the model thread they change week by week so there only going to call what models show and none of the models show zonal train or blazing beasterly so why would they be that silly,,,

 

its good to see the slowdown of the alantic and you can see the difference from last year.

so what do we need a pattern reset and removal of heights into scandi or rebuild of heights into scandi from the azores see this happen and open the flood gates to cold this is also a possibility then futher on into december retrogession of heights across iceland into greenland.

 

one thing i have not seen for fair few years is alantic heights they seem to have become rare but these also have opened up the cold gates.

 

but right now stale mate and in fi the models really are stuck for ideas.

[IDO]

The ECM ensembles have changed back this morning from last night's Scandi high back to the consensus of Euro high:

 

 

 

That may all change again this evening. The NAO and AO both show splits in the D8+ pattern:

 

 

 

So a more blocked or more Atlantic flow or in-between all remain on the table. An eureka model moment will happen in the next few runs and I am not sure trends will help in this type of setup, so its a case when the music stops, lets hope it favours the colder route. 

 

 

[phil nw.]

We have  removed an off topic post along with a reply to it.

Just a reminder to please keep to views around the Models in this thread.

We know most of you do but one or two continue to use this thread like a chat room with subjective comments which spoil it for everyone else.

 

Thanks all.

[radiohead]

what ive noticed steve is the 12zs seem to be bring the heights and amplification back into play but as soon as the 00zs come out next day then everything falla flat again and we start the process all over again!!we just not getting anywhere with this and also ecm was way further west with the pattern over europe over the past few days and now has gone east again! !! The ecm had such a strong area of high pressure across scandinavia and well into the uk just 24 hours ago for the end of this week but seems to have pushed slightly East like always!!

 

Remember though, the 0Z and 12Z is the exact same software running on the exact same computer sysem. The only difference between them is the time of day that they are run and the updated initial conditions. There is nothing that could make a 0Z run more likely to produce one type of result over another. It's not like comparing the GFS to the GFS (P), as then you'd be comparing two different sets of code, with differernt resolutions and layers of input etc.. 0Z and 12Z are simply the times when the exact same software is run. They don't have seperate "personalities" if you like. It's just not possible.

 

Think about a word processor on your computer. It won't do anything special with a document you are working on if you use the software at 6pm or 6am.

[MR EXTREMES]

The ECM ensembles have changed back this morning from last night's Scandi high back to the consensus of Euro high:

 

EDM1-240-18.GIF  EDM1-240-19.GIF

 

That may all change again this evening. The NAO and AO both show splits in the D8+ pattern:

 

ao.sprd2 (7).gif  nao.sprd2 (1).gif

 

So a more blocked or more Atlantic flow or in-between all remain on the table. An eureka model moment will happen in the next few runs and I am not sure trends will help in this type of setup, so its a case when the music stops, lets hope it favours the colder route. 

yes major split on the nao and ao just showing the massive uncertain outputs.

but ao -4 yes please that would be nice but its not one anyone can call at this stage a very strong possibility of heights settling over the uk and if the ecm say heights over the uk id be inclined to agree rather than ecm heights over greenland and the gem to that has a habit of getting rather excited over northern heights.

 

to be honest if the gfs throws out a HLB then i tend to agree its always a good idea to blend the models together but all in all its far from winter over.

im on the fence though 

[IDO]

To say ECM has not got a clue after D4 is an understatement. The Dutch ensembles show large spread as early as Friday:

 

 

The op and control in the colder solutions from D7.

Edited November 17, 2014 by i'm dreaming of...

[Luke Best]

Remember though, the 0Z and 12Z is the exact same software running on the exact same computer sysem. The only difference between them is the time of day that they are run and the updated initial conditions. There is nothing that could make a 0Z run more likely to produce one type of result over another. It's not like comparing the GFS to the GFS (P), as then you'd be comparing two different sets of code, with differernt resolutions and layers of input etc.. 0Z and 12Z are simply the times when the exact same software is run. They don't have seperate "personalities" if you like. It's just not possible.

Think about a word processor on your computer. It won't do anything special with a document you are working on if you use the software at 6pm or 6am.

I'm not 100% certain but I believe different runs work on slightly different data due to some weather stations not reporting round the clock.

Hopefully somebody with a little more knowledge can elaborate?

Edited November 17, 2014 by Luke Best

[Midlands Ice Age]

 

To say ECM has not got a clue after D4 is an understatement. The Dutch ensembles show large spread as early as Friday:

 

eps_pluim_tt_06260-2.png

 

The op and control in the colder solutions from D7.

 

IDO....

Rather than your position, I take the view that the ECM is reflecting EXACTLY the degree of uncertainty in the models at the moment. It is not clearcut yet which way the dice will fall.

I think that is superbly shown in your mean of the Dutch assenblies above.

MIA

Edited November 17, 2014 by Midlands Ice Age

[IDO]

IDO....Rather than your position, I take the view that the ECM is reflecting EXACTLY the degree of uncertainty in the models at the moment. It is not clearcut yet which way the dice will fall.I think that is superbly shown in your mean of the Dutch assenblies above.MIA

Yes, "has not got a clue" was just hyperbole. The undercurrent from all models is early onset uncertainty.

[Gael_Force]

what ive noticed steve is the 12zs seem to be bring the heights and amplification back into play but as soon as the 00zs come out next day then everything falla flat again and we start the process all over again!!we just not getting anywhere with this and also ecm was way further west with the pattern over europe over the past few days and now has gone east again! !! The ecm had such a strong area of high pressure across scandinavia and well into the uk just 24 hours ago for the end of this week but seems to have pushed slightly East like always!!

 

This very thing was the subject of a WSI blog in early November - it was attributed, at the time, to loss of data coverage in the snow/ice sector: perhaps something more than that? We will see if any change back in the later runs.

 

HEADLINE: Last week’s GFS operational model runs featured some of the largest temperatures swings between 00Z and 12Z runs observed thus far this year. The 00Z runs were predicting a mild medium-range forecast over the U.S., whereas the 12Z runs highlighted a much more colder solution across the U.S. The analysis performed in this blog suggests that last week’s satellite outage may have played a critical role in the 00Z verses 12Z model battle royal, resulting in large initialization errors in Eurasian snow cover/depth that were manifested in the 00Z runs only. Now while the 12Z initialization runs were less impacted by the satellite data outages remains head scratcher,they were clearly the winner in the warm verses cold battle between the 00Z and 12Z run medium-range U,S. temperature forecast.

 

http://www.wsi.com/blog/energy/what-caused-last-weeks-cyclical-model-volatility-in-the-gfs-operational-model/

[gottolovethisweather]

Statements such as no one model (ECM in this instance, according to some posters) not having a clue as to what might occur by say next weekend, are way off the mark I think. Why? Because ALL outputs aren't handling it with their usual expertise right now and here's perhaps another reason why.

 

Two nights ago, during the usual weekend wobbles from all the various NWP outputs, I stated the following, based on what I had heard in the media forecasts.

 

https://forum.netweather.tv/topic/81842-model-output-discussion-1200-hrs-151114/?p=3070476

 

The take out comment from the link above is my quoted section below.

 

• For those wanting to know what might happen in our small patch of the globe in some six to seven days (FI probably as close as t+96 right now I'd suggest) tonight's 10-dayer from the Beeb hinted at two scenarios for next weekend.

 

As for FI or shall we call it, simply the period of cross-model divergence, the comment highlighted in red remains the case, as a four day timeframe where cross-model uncertainty begins. Furthermore , the following quote also backs up the current reality of the current setup.

 

• Right now, the Atlantic is NOT in charge and deep cold weather is also NOT arriving during the next five days.

 

After all is said and done, hopefully over the next couple of days, more especially in conjunction with the more reliable 12z outputs, we MAY start seeing some better consistency across the outputs for the forecast heading deeper into November month end.

 

My hope is that the current general Easterly type in position over the UK, combined with a cooling continent over Central and Eastern Europe will lead to some colder days and nights before the month is out.

Edited November 17, 2014 by gottolovethisweather

[nick sussex]

More changes from the GFS 06hrs run with the favourable jet cut back towards the UK which you can see by the shortwave which moves sw towards Denmark.

 

Generally with the set up shown by the GFS at T186hrs you'd see more trough disruption but the models are reluctant to do this at longer range, with low heights developing over southern Europe you really should be able to lock in the cold but the GFS may well want to lose it in the lower resolution.

[Nick F]

I'm not 100% certain but I believe different runs work on slightly different data due to some weather stations not reporting round the clock.

Hopefully somebody with a little more knowledge can elaborate?

 

If it were true that some stations were not reporting round the clock, as I'm sure some don't, the data they provide is such a tiny proportion of the actual global observations gathered by models. For every 6 hour cycle for GFS runs, there is now such a large wealth of data out there now for models to number crunch from: surface obs, radar obs, VIS/IR/WV imagery, radiosonde ascents, ship/buoy obs, aircraft obs, wind profiler obs and probably millions of satellite obs. In the great scheme of things, some surface obs missing would have a negligible effect in the broadscale pattern, given so much data available now in recent years would mean missing obs would be smoothed out as it were, or if a missing obs - a recent rather than latest reading is used and assimilated over a time period  However, in the mesoscale environment, I guess missing data could have some impacts, such as the deepening of a depression, snow events, convective events, etc.

 

I think it's a myth that 00z/12z GFS runs are any better that 06z/18z runs because they use different obs, at the end of the day the latest run of a model will always be better than the run issued 6 hrs earlier given newer data assimilated by that model. Though it maybe noted that sometimes over a short scale, particular runs may appear less 'skillful' in their verification with other runs, I believe 06z GFS, recently has verified slightly less well than the other three daily runs in the anomaly correlations over the last month, but statisically over the longer term this verification is not significant to say that 06z GFS is the 'worst' run, as they will all have their ups and downs.

Edited November 17, 2014 by Nick F

[knocker]

I'm not 100% certain but I believe different runs work on slightly different data due to some weather stations not reporting round the clock.

Hopefully somebody with a little more knowledge can elaborate?

 

I don't know the breakdown of the initialisation data but I would think weather station data would be a minor player compared to satellites. There are about 4,500 reporting climate stations so the actual number will be in excess of that and virtually all will have 24 hour coverage.

 

EDIT

Belay the above Nick posted at the same time.

Edited November 17, 2014 by knocker

[beng]

More changes from the GFS 06hrs run with the favourable jet cut back towards the UK which you can see by the shortwave which moves sw towards Denmark.

 

Generally with the set up shown by the GFS at T186hrs you'd see more trough disruption but the models are reluctant to do this at longer range, with low heights developing over southern Europe you really should be able to lock in the cold but the GFS may well want to lose it in the lower resolution.

 

Yep as normal it does - very annoying. Hopefully the control run will be similar out to t180 and provide a better overview afterwards.

 

I can only assume that fixing the models to deal with this problem must be extremely complicated as the issue has been very apparent really since 2009 when the winters began to turn noticeably colder (last year the exception obviously).

 

Anyway - until there's any consistency earlier on from run to run and between model to model, it's hard to take anything seriously in terms of the longer term. 

Edited November 17, 2014 by beng

[fergieweather]

IDO....Rather than your position, I take the view that the ECM is reflecting EXACTLY the degree of uncertainty in the models at the moment. It is not clearcut yet which way the dice will fall.I think that is superbly shown in your mean of the Dutch assenblies above.MIA

Agreed. Shannon Entropy is back to high values again; albeit other models have started leaning more towards MOGREPS somewhat faster return to progression, it's very finely balanced.

[gottolovethisweather]

Agreed. Shannon Entropy is back to high values again; albeit other models have started leaning more towards MOGREPS somewhat faster return to progression, it's very finely balanced.

 

Ah, good to know. I slightly rambled on in my last post which may have been lost in the fast pace of this thread, but good to know you pretty much have confirmation that things are messy right now.

 

https://forum.netweather.tv/topic/81842-model-output-discussion-1200-hrs-151114/?p=3071254

 

Just when do these "Shannon Entrophy" uncertainties begin on current outputs. Where, ie. what synoptic feature should one look for in the charts, could be another most pertinent question now too. Time for a reality check all, methinks before second-guessing what lay ahead beyond this next week.

Edited November 17, 2014 by gottolovethisweather

[phil nw.]

Although we see the ups and downs in the operational runs the blocking signal around Scandinavia remains consistent in the Ens Ht forecasts.

I have been monitoring these for a few days now on the ECM suite and the countdown started at day 10 now down to day 7 here -T240 hrs on first then coming closer.

 

Naefs for day 7 also agreeing

 

so no sign of the block disappearing and the Atlantic trough continues to be held just to our west with signs now that E.Europe will get that feed of colder air around the top of the block on it's lee side.

A bit of a stalemate with the UK out of reach of any real cold currently.

To get that we do need to see the jet driving much further south and east towards the Med.Interestingly in the later Naef's there is a hint of this but as ever we need to see more consistency from that far out. 

 

[Nouska]

For all the dataholics, ECMWF have a page devoted to data collection sources for its daily model runs.

 

http://old.ecmwf.int/products/forecasts/d/overview/monitoring/coverage/dcover!AMVs-Infrared!00!pop!od!mixed!w_coverage!latest!obs/