Model Output Discussion - 18z - 25th Nov onward

[Cheese Rice]

I feel like we should bin the ECM output, to a certain extent the GFS output and stick to the UKMO.

From past experience the models always underestimate high pressure when it is to the NW/N/NE of the UK.

I'm expecting something like this at T144

I've edited this chart myself, derived from the UKMO T144 chart.

We are far from a clear solution, a quite messy solution looks quite apparent with shortwave galore in the latter timeframe. Which is why I wouldn't get hung-up over specifics such as 850's as these will change. I wouldn't rule out a full blown easterly post T144, if we get shortwaves aligned in the right place we could see anything from a Northerly (not straight from the Arctic), North Easterly, Easterly or even a cold South Easterly.

There are a variety of options available so I wouldn't be worried at all at this stage. We will see colder weather by the end of next week.

Edited November 25, 2012 by Cheese Rice

[nick sussex]

Yes I did see them Nick (great find by the way!). Possibly better news than the 18z tonight as we have the European based ECM vs (amongst many other things) charts drawn up by a Meteorologist(s) with local knowledge. I know who my money's on!

Thanks those charts are really handy when we're dealing with Greenland blocking because they give us an idea of the forecasters thoughts for the eastern USA which is always critical for that region.

We just have to hope for the best tomorrow. I think whats clouding the issue is how the models handle an injection of energy from the north into that developing low.

The ECM doesn't do that and hence just sends a weak feature across the Atlantic.

[SMU]

http://modeles.meteociel.fr/modeles/gens/run/gensnh-21-1-90.png?18

T90 ensemble mean steps back west a couple of hundred miles...

all good news-

S

[bluearmy]

Teits - i think mogreps 15 is the meto ens model

[Jonan92]

Just 4 hours and a half for 00z run of the GFS, 5 hours for UKMO's run and 7 hours for ECM's run LOL

Edited November 25, 2012 by Jonan92

[cold snap]

Can't remember such a exhausting fun day of model watching

Think it will be an idea to get some sleep because in the morning it's going to be hectic

[Matty M]

We have backing from the GFS Control run around T120hrs if it is of interest to youse

The rollercoaster goes round and round.

We end with probably the best suite of GFS Ensembles to date with many cracking runs at T120hrs.

[Liam J]

Teits - i think mogreps 15 is the meto ens model

Yep your right, heres a bit of info - http://www.metoffice.gov.uk/research/areas/data-assimilation-and-ensembles/ensemble-forecasting/MOGREPS

[lorenzo]

00z run will tell us if the blocking happens or not. GFS, UKMO, Nogaps and Brazilian say that it happens, on the other hand, ECM, GEM, JMA say that it brokes down

Given there was a post either from Kold or Nick earlier today, sorry I cannot link it, that stated the misrepresentation of the blocking anomaly was huge it is likely that these 'lesser' models will be cast aside and leave us with a straight examination of what GFS / UKMO / ECM are doing.

Given that ECM is the winning model - I would be inclined to side with output here - however it was wobbled and Saturdays 00z imaginary torque event got me wondering. Matching ECm with CPTEC, BOM et al and all the exotic models is just rude to the ECM. Yeah it wobbles but how many times has there been this synoptic.

The GFS 18z whilst a good counter to the 12 z ECM catastrophe / apocalpyse run whatever you make of it well it is unfortunately the GFS 18z.

Makes me want to shout..

HAVE WE A RAGING POSITIVE AO FOR 80 DAYS >>>>>>

Leaves is with the UKMO - Solid, non spiking, non deviating, non trending, non deviating. Reliable.

But, and this is a huge but - this is still out of hours at 96 it is all to play for and shortwave carnage comes in.

ECM 12z killed hope for many

00z may do again

If that gets you down go and look at ECM 30 hpa when FU Berlin updates

[watchingweather]

I have no idea what everyone is on about but it makes for fascinating reading followed by endless googling of what certain terms mean. Keep it up guys!

[Matty M]

12/20 members = very good for cold at T156hrs

3/20 members = following ECM OP

5/20 = Mixed

[Weathizard]

Anyone got a link to the ECM ensembles for this evening? How are the 18's looking?

[Zenarcher]

Only 3 perturbations disagree with the 18z GFS all the other 17 give the run good support on the low and it's track.

[Jonan92]

Interesting

Expexting a stronger Greenland High?

[johnholmes]

Im slightly confused because I always thought the MOGREPS was mainly used for short range forecasts only upto 2 days.

taken from Met O site

While the main focus of MOGREPS is on producing short-range ensemble forecasts, we also run a version of the global ensemble to 15 days together with the ensemble system from ECMWF to produce uncertainty information for medium-range forecasts.

Last Updated: 9 January 2012

[summer blizzard]

FAX looks fine to me..

[Eskimo]

FAX looks fine to me..

Nice to see that front getting squeezed and the 528Dam line flirting the Anglian coast.

[Catacol]

http://modeles.meteo...-21-1-90.png?18

T90 ensemble mean steps back west a couple of hundred miles...

all good news-

S

Yes... for once today we had an ensemble mean from ECM that wasnt really worth looking at from the moment it fired up the vortex and lowered polar heights. Just shows that an operational run is thin evidence of anything, and a single ensemble run is not a whole lot thicker. Trend spotting is about a sequence of ensemble runs coming together... just as JH and GP tout.

[phil nw.]

Whilst the operational runs continue to tease the ens.mean ht anomalies continue to be rock solid.

Here are those mean outputs at T144hrs from the 12z ECM and GFS runs.

and the 12z NAEF`s which is the GFS/GEM comb mean hts

Showing the Atlantic block linkup towards Svalbard with low hts over N Europe and a North Easterly flow for the UK by the weekend.

Edited November 25, 2012 by phil n.warks.

[Mark Bayley]

Whilst the operational runs continue to tease the ens.mean ht anomalies continue to be rock solid.

Here are those mean outputs at T144hrs from the 12z ECM and GFS runs.

and the 12z NAEF`s which is the GFS/GEM comb mean hts

Showing the Atlantic block linkup towards Svalbard with low hts over N Europe and a North Easterly flow for the UK by the weekend.

Hi Phil, could it be possible to have a link to the first set of charts please, i don't seem to have them in my bookmarks!

[Zakos]

I think ,IMO, we can bin the ECM output. little support for its rapid breakdown.

It does however look as if there is a signal for the Atlantic to attempt to attack the block.

I think the ECM has majorly overreacted to this signal, and its rapid breakdown looks unrealistic to me.

Also, to those that say the ECM is terrible... well yes it certainly is at first glance. but look at the hemispheric pattern-

This chart, although fairly poor for the UK, is far, far better than any run in December 2011.

I firmly believe that even if the ECM is somehow correct, it will simply be a delay of the cold conditions, rather than a breakdown of the Arctic high pressure

[phil nw.]

Hi Phil, could it be possible to have a link to the first set of charts please, i don't seem to have them in my bookmarks!

Here you go Mark.

http://raleighwx.americanwx.com/models.html

[ronan]

people on twitter saying sleet is falling in ireland just now

I could well believe it! Its certainly cold enough. In the north we have been cold for the past 3 days. The fog never lifted all day today and temps the past 3 days haven't got above 3 degrees with today hovering around 2/3 degrees. In my time on Netweather i don't think that i have ever seen so much in depth analysis of the models! Its brilliant to see but i just can't wait to seen the back of November! I think the signals for cold was first picked up in October but it seems like we have been chasing this for months. Hopefully we can nail things in the next few days!

[weather eater]

Someone asked earlier if the ECM has ever been this wrong at this timeframe. Well the answer is yes it has and I shall use the Dec 2010 as an example.

The ECM made a right pigs ear out of modelling this because I remember vividly as it caused many arguments. The ECM consistently progged that LP to be much further NW into the Atlantic and was actually predicting mild SW,lys! I remember because myself and Nick S were saying the further SE the better and this led to being accused of SE bias.

A good point and I will add to that, in 2010 the GFS was extremely consistent on the lead up to that spell ,the ECM wobbled about all over the place, so yes it is sometimes wrong,

I would also add that it was Fridays 06z GFS run that first picked up the movement of the LPs system across the Atlantic between 60 and 102 hrs all the other models have followed this evolution. We all know the ECMs record is better than the GFS, but it’s not written in stone.

[phil nw.]

Looking at the ens.graphs a pretty tight clustering until the weekend by both models for the cold pattern to establish.

We can see the ECM operational very much on the mild side of the members later on.