Model Output Discussion

[cyclonic happiness]

http://www.wetterzentrale.de/pics/Rtavn181.png O dear gfs 18z at 18h is going the way of the ecmwf!! :o :o :o :o

Are you being serious? it's only just started? why do you think that?

[Snow Queen one]

Guys i am assuming that sebastian1973 means the ECM when he mentions the EC. Could it be possible that we have been worrying over a model that is missing data as suggested which is why it looks so wrong. Can someone try to confirm this for us in the depths of dispair on here sebastians post seems to be ignored but may well give us a bit more info on the outputs.

Thanks

[Mark Bayley]

http://www.wetterzen...cs/Rtavn181.png O dear gfs 18z at 18h is going the way of the ecmwf!!

Its only out to T48!

Hopefully the GFS will stick to its guns, preferably nearer the parallel solution

[Optimus Prime]

What a load of tosh Badboy, most consitant, we must have a different defination of the word, the ECM has been on and off the train more times then I've had hot dinners!

The GFS paralell and the UKMO is the one that is really consitant.

I agree Kold and don't worry there are a few on here that don't really seem to want to see the charts for what they are. A record breaking winter forecast by Ian would be impossible if the GFS comes in on this one but possible under the ECM evolution.

The discussion we had RE about the marginality about precip has been sorted out, doesn't look as marginal as what I first saw it as.

Anyway the 18Z is popping out and may answer a few unanswered questions with regards to the ECM.

With regards to the ECM...we have to be honest it's done us in before! lol

Edited December 10, 2009 by Optimus Prime

[Jason H]

http://www.wetterzen...cs/Rtavn181.png O dear gfs 18z at 18h is going the way of the ecmwf!! :o :o :o

That's a wind up!

That is how the whole evolution begins. With WAA up past Iceland!

[Rob K]

http://www.wzkarten....cs/Rtavn001.png

Well it's coming out now ofcause no change YET

Well, for those that think the ECM is infallible, let's have a look at what its 168hr chart looked like... 168hrs ago (well, 174 actually)

Erm... look much like the current analysis chart to you, because it sure doesn't to me!

[snowmadchrisuk]

TOO ALL THOSE THAT THINK ECM IS RELIABLE AND THE BEST MODEL

http://www.ecmwf.int/products/forecasts/d/getchart/catalog/products/forecasts/medium/deterministic/msl_uv850_z500!Wind%20850%20and%20mslp!168!Europe!pop!od!oper!public_plots!2009120312!!chart.gif

Check how wrong it was for todays forecast , according to ecm we should be under a southwesterly and wet

But that was at +168 Neil , the problem is with todays ECM is that it all goes wrong at +72-96 . On the up side though there is comments all over the web from all different countries that there is a problem with the ECM tonight , missing data is one of the things being blamed , but it looks like some mistake has been made according to the dutch weather service . There is also stuff that says the ECM has been jumpy for the last few days and there is problems. For example MTSAT, the geostationary satellite for East Asia has not been included,

Edited December 10, 2009 by snowmadchrisuk

[Snow Queen one]

But that was at +168 Neil , the problem is with todays ECM is that it all goes wrong at +72-96 . On the up side though there is comments all over the web from all different countries that there is a problem with the ECM tonight , missing data is one of the things being blamed , but it looks like some mistake has been made according to the dutch weather service . There is also stuff that says the ECM has been jumpy for the last few days and there is problems.

AT LAST SOMEONE LISTENED TO WHAT WAS BEING SAID. No model is infalable and it seems on one of the most important times of the winter the ECM has fallen foul to human error or something. Ther is hope yet.

Into =72 now and it looks as if the cold will arrive a bit earlier than the last run. Not too much has changed in the reliable time frame yet.

Edited December 10, 2009 by Snow Queen one

[cyclonic happiness]

out to 60 hrs and the colder upper are a bit further west which could be a good sign for mondays snow (maybe).

http://www.meteociel.fr/modeles/gfse_cartes.php?&ech=66&mode=3

[Thundery wintry showers]

Regardless of what the ECMWF says, I can see the GFS 18Z bringing in that cold easterly flow for Monday/Tuesday again. It will probably appear mainly over southern areas like on the UKMO.

[kold weather]

Yep the problems all esculate from about 72hrs out your right, we are coming really close to the cruical time in the 18z run.

As for the ECM missing data, I'm sure other countries will lose the cold if the ECM is correct as well.

[Matt Jones]

Pretty much the same as 12Z so far. Whats dissapointing though is GFS is now struggling to bring in the -10 uppers a few days ago -10--15 uppers covered the UK on a few runs, I suppose best case scenario now is the the GFS wins the battle over the ECM.

[snowmadchrisuk]

AT LAST SOMEONE LISTENED TO WHAT WAS BEING SAID. No model is infalable and it seems on one of the most important times of the winter the ECM has fallen foul to human error or something. Ther is hope yet.

I hope that is what has happened , it is rather hard trying to sort out translation for all the different websites or I would have posted some links .

18z is looking great and hasn't followed ECM as 1 premiture comment said .

Edited December 10, 2009 by snowmadchrisuk

[weather eater]

hai weatherfanatics, my name is from the Netherlands

I really like to read your forum, so I joined this forum and this is my first attribution.

I would like to say to you, that a Dutch meteorologist just told at weerwoord.be that the EC of tonight is probably containing a fault.

He wrote

Dit is een grote sprong tussen twee ensembles, vrij ongewoon. Er bestaat kans dat het gehele ensemble, en ook de ECMWF operationele run, beïnvloed is door een relatief grotere fout in de analyse. Zo zie ik dat er geen gegevens van de MTSAT, de geostationaire satelliet voor Oost-Azië, zijn meegenomen. Dat is de afgelopen drie maanden niet eerder gebeurd.

THere is a giant gap between the two ENS, which is very unsual. There is a chance that the whole ENS, and also EC operational run is influenced by a relative bigger mistake in the analyses. E.g. I there are no data of MTSAT used, the sattelite for East Asia, this wasn´t case for 3 months.

Another helpful link is this one

http://www.weerplaza.nl/15daagseverwachting/default_pluim.asp?r=midden

it contains the ENS for several areas in the Netherlands

E.g. given southeast of the Netherlands, see that wonderfull control !

Hi Sebastiaan given that our Dutch is much poorer than your English can you give us any more information as to what was said.

[Mr Sleet]

Guys i am assuming that sebastian1973 means the ECM when he mentions the EC. Could it be possible that we have been worrying over a model that is missing data as suggested which is why it looks so wrong. Can someone try to confirm this for us in the depths of dispair on here sebastians post seems to be ignored but may well give us a bit more info on the outputs.

Thanks

Hi SQ1. Sebastian translates it for us in fact and he is saying that dutch meteorolgists are questioning the data for the ECM model and that it may contain some mistakes.My advice is that we should all calm down a bit, if it doesn't come off think of the savings on your heating bill - however I'm fully expecting to wake up to an inch or two of snow on Monday morning.

[wysiwyg]

If you look at the mean plot for the current ECMWF ensembles you can see a distinct "sawtooth" pattern that amplifies consitently as the run progresses. Now fluid dynamics modelling is not my direct area of expertise but I have worked with it in the past. That type of behaviour is a classic sign of incomplete data that has been filled in with "fake" data that has been derived using inappropriate interpolation techniques - This fits with Sebastian reporting that data was missing from this run.

I'm not saying my analysis should be taken as gospel as I don't know the inside track on what data goes into the ECM model and how it is pre-processed. However, on this occasion I believe that taking this purely empirical explaination and couple it with the underlying teleconnection signals - I genuinely believe the latest ECM run to be flawed.

FWIW - this does not make it a bad model and it's output cannot just be consigned to the bin in disgust just because it doesn't show the evolution you are yearning for. Ironically - someone posted the WMC (?) moscow model - that was clearly showing a major glitch that had relocated the north pole in my garden. Again this model would more than likely be suffering from incomplete data - In that case it had probably been hurridly and simply replaced with zeros giving that rather bizarre output.

They are computer models - not experienced forecasters!

[Iceberg]

GFS has decided not to follow ECM (you can tell at 60Z), very similar to the 12Z, but brings the cold in slightly sooner for the weekend.

[Neilsouth]

http://91.121.94.83/modeles/gfs/run/gfs-0-66.png?18

http://91.121.94.83/modeles/gfs/run/gfs-1-66.png?18

Still looking ok at first for east,south east sunday - monday - tuesday ?

[Snowyowl9]

http://www.wetterzentrale.de/pics/Rtavn181.png O dear gfs 18z at 18h is going the way of the ecmwf!! :o :o :o

I don`t think so,thats only friday,fax chart shows a chart like that.

http://www.wzkarten.de/pics/Rtavn661.png

ECM may well be having problems,but as Steve said theres more energy in the northern arm so I`ll have to check those charts out.

[kold weather]

Regardless of what the ECMWF says, I can see the GFS 18Z bringing in that cold easterly flow for Monday/Tuesday again. It will probably appear mainly over southern areas like on the UKMO.

I think if there are going to be any changes from the GFS I suspect it'll be on the 0z run but who knows, its at so close range that the ECM could change again...given it has flipped flopped all over the place recently.

[Dartmoor_Matt]

Hi Sebastiaan given that our Dutch is much poorer than your English can you give us any more information as to what was said.

The English translation is under the Dutch I think.

[SMU]

I don`t think so,thats only friday,fax chart shows a chart like that.

http://www.wzkarten.de/pics/Rtavn661.png

ECM may well be having problems,but as Steve said theres more energy in the northern arm so I`ll have to check those charts out.

TBH this thread needs cleaning up & the some of the nonsence posted deleted- that should remove about 10 pages.... ( Not yours ^^ I add)

I am 99.9% the ECM is wrong based on the ensemble mean( GEFS) the ECM ensembles ( Especially the control), the rest of the model suite- ic the UKMO & of course the modelling of this whole evolution from the start-

S

[Snowaddict]

This information sounds very interesting regarding ECMWF... I only hope that it's correct because although synoptically up to T+168 ECMWF is not bad this evening, it's far from what we have seen advertised on the other models & is a VERY significant shift compared with alternative guidance..

I cannot understand the disparity between ECMWF & GFS at such a close timeframe:

For example:

18Z GFS Operational @ T+78

850's of -4 approaching Ireland, all across the UK with -8's approaching SE England..

If this does go the way of ECMWF as modelled on today's 12Z run then it's a complete failure for GFS, GFS Parallel and it's ensemble suite..

@ T+84 there is a very cold easterly setting up across the British Isles on the 18Z GFS run..

VERY strange disagreement as such a close timeframe...

But history tells us that the ECMWF evolution will be favoured unfortunately.

[weather eater]

18z out to 84hrs on meteociel now, I dont see a problem.

[Iceberg]

18Z shows the snow further west on Monday and again brings a few cms to Kent, Essex and maybe a covering as far as Hampshire and Dorset in the south.

Lincolnshire looks good on the east coast as well for Monday.