Model Output Discussion:

[nick sussex]

The drama begins again this morning! Although the UKMO is half hearted and we saw what happened the last time that was the case. However the shift in the GEFS ensembles is hard to believe especially as this occurs within 144hrs, I've never seen anything like it!

We'll have to wait to this evening to see if the UKMO comes on board and given the last week it's hard to have too much faith in any outcome until T-48hrs!

Edited February 3, 2010 by nick sussex

[Eugene]

Yes GFS was first with the easterly then dropped it only for UKMO and ECM to go for it then just the UKMO now GFS and most of it's ensembles are going for it but UKMO isn't, ECM is like in the middle, models all over the place currently i have a sneaking feeling cool to cold will win(nothing major yet though) even if we get a brief milder blip.

That mild southerly for sunday/monday is getting squeezed again.

[johnholmes]

...perhaps no model ...output is so random they are using more traditional forecasting methods to reach their solution...although i think as others have said they willl very probably change tomorrow???

I have posted at length about this before but perhaps you were not around then?

They have access to every model output and use every model output. The charts from their own computer which we see are a very small selection of what is available to them.

One of the best ways of picking up on the views of the Principal Forecaster at Exeter is to read the outputs from Ian Fergusson who presents weather forecasts, I believe its on South West(correct me if I'm wrong someone-Ian?). The link up with all weather presenters, both professional Met O forecasters and presenters like Ian is made several times each day-charts are shown to illustrate the view of the senior man and are RARELY not followed. MOS(Model Output Statistics, MOGREPS, one of their own models, the NAE model available at Weather Online, and many more variations of their own Global and various Fine Mesh models are used along with output from ECMWF but as I said at the beginning ALL models are used and often mentioned in their technical outputs.

My knowledge is no longer first hand-I used to have access to this discussion, as an ex senior forecaster, but its no longer available to me.

hope that helps.

ps

I should add that the Fax chart is just about the best possible combination you can get, that is the man-machine mix, so the latest Fax outpus, the furtest is T+120 hours=5 days, is usually the most accurate view on what is going to happen.

Their 6-15 day outlook is, in my view, the best you will find on the web.

For another view on how events may affect the UK then click on the NOAA link for their view on the models over a 1 -6 day and then their 6-14 day outlook. link below

1-6 day

http://www.hpc.ncep.noaa.gov/preepd/preepd.html

6-14 day

http://www.cpc.noaa....day/fxus06.html

Edited February 3, 2010 by johnholmes

[Duane S.]

I can see big upgrades on the way over the next few days. I think height rises over Greenland are going to dictate the upcoming pattern not the Atlantic. ECM and METO are OK this morning, no disaster there at al and the GEFS Ens are a lot better than they were a couple of days ago, and improving.

I admit to struggling to see exactly what role (good or (more likely) bad) the Norway SW will play in all of this though. I never had this down as much of a player a week ago now it seems omnipresent and possibly overplayed by the charts? We shall see.

All to play for though still, fascinating (along with frustrating, annoying, tiring etc etc) watching the models at the moment. We may all moan about the 4 times per day GFS but, we'd sure miss it if it wasn't there. Think of it as a drunk uncle at a party. Most of the time it waffles on incoheriently then suddenly in a moment of inexplicable lucid soberness you are forced to sit up and listen, as its prophetic ramblings actually appear to make sense before, just as suddenly, it's back to knocking over the vase and generally annoying everyone.

Edited February 3, 2010 by s4lancia

[londonsnow]

whilst we are experiencing a winter fairly alien to our shores, in recent times at least, can anyone explain quite why the models are somersaulting around from one run to the next? i can't remember the last time model watching was so random/frustrating/fascinating(!) i can see that overall, despite major blips, the trend is for cooler conditions to return, but it must be a forecasters' biggest nightmare at the moment.

[johnholmes]

Think of it as a drunk uncle at a party. Most of the time it waffles on incoheriently then suddenly in a moment of inexplicable lucid soberness you are forced to sit up and listen, as its prophetic ramblings actually appear to make sense before, just as suddenly, it's back to knocking over the vase and generally annoying everyone.

I just love that s4l-wonderful idea

don't think it'll be human "experience" as we haven't been experiencing winters like this for sometime now - got to be further computer info that we don't have access to that is assisting them with hedging their bets

Afterall they do have super super computer down at Exeter - mega calculations etc - not all of the data it churns out are we gonna be able to access I'm sure

see my other post explaining how they arrive at their view-the classic man/machine mix-the best that can be achieved-see also my comments about NOAA

[Blizzardo]

I like the look of the ECM this morning with a nice northerly block. If things played out this way I can only see any low pressure sliding south east. As usual the GFS seems to be slightly overdoing the energy in the middle part of the run.

[nick sussex]

Just when you least expect it...just what you least expect ! I think the main thing overnight is for the models to show less ridging from the South, thus preventing a SWerley flow. The ECM certainly looks plausible with HP around to the North and parts of the North on the cold side of the PFJ at times.

I'm really quite shocked by the turnaround this morning but more especially by the GFS, however the UKMO doesn't want to know and because of this i still have low confidence. I agree with you regarding the ECM that to me looks more plausible than the GFS. At the moment though nothing would surprise me, the last week has surely been the most volatile since i started looking at the models.

[Eugene]

Here's the GEM 00Z which is excellent and i think could be near to the actual outcome.

http://www.wetterzentrale.de/pics/Rgem1201.gif

http://www.wetterzentrale.de/pics/Rgem1441.gif

http://www.wetterzentrale.de/pics/Rgem1681.gif

http://www.wetterzentrale.de/pics/Rgem1921.gif

http://www.wetterzentrale.de/pics/Rgem2161.gif

http://www.wetterzentrale.de/pics/Rgem2401.gif

TBH just a slight tweek to ECM 00Z OP with a slightly stronger push southwards of HP to our north and it could turn out similar.

[len]

It looks as if the agonising over a possible Easterly flow will continue.

But it might not end there. The deep cold, that had crawled across Poland into Germany last week, has been thoroughly booted back far far away into deepest Russia. http://www.uni-koeln.de/math-nat-fak/geomet/meteo/winfos/euisoTTPPWW.gif

Even if we get the Southeasterly flow being touted it may not be the end of the waiting??

[LomondSnowstorm]

Wow! A big change overnight in the outlook, with more energy heading southeast rather than northeast http://www.wetterzentrale.de/pics/Recm1441.gif

http://www.wetterzentrale.de/pics/Rgem1561.gif

Beware, however, the curse of the UKMO model http://www.wetterzentrale.de/pics/Rukm1441.gif

If it refuses to change I think most of us will struggle to believe this, but with the 12Zs things will become clearer (possibly!)

[Frosty.]

Yes the GEM 00z wins 1st prize for most wintry run so far today with E'ly & N'ly winds, the gfs 00z comes 2nd but is a much drier run with high pressure forming just to the northeast with very little in the way of precipitation beyond the weekend but it would bring rather cold conditions with some frost, the ecm 00z is half way towards the gem with a cold and unsettled run for the north but a bit milder in the south but still very unsettled and then we have the ukmo which is just unsettled with near average to mild temps everywhere. The next few days look milder in the southwest and southern england with 10c in cornwall today and tomorrow and 9c in london on thursday.

[Polar Continental]

Morning everyone,

Well this has been the most incredible couple of weeks of model watching since the advent of the Internet. I have never seen the models so indecisive, in my opinion, no model has shown stability in the outlook beyond 96 hours, they have all struggled with this easterly or no easterly, however you wish to view it.

Now the GFS brings the easterly back at T96 as pressure builds to the east and northeast. The ECM also builds pressure to the east and just about gets the easterly on our door step, where as the UKMO keeps the high further north east and allows the low off the Atlantic to make inroads in to the UK, which one will be correct?

Edited February 3, 2010 by Polar Continental

[snowmadchrisuk]

Morning everyone,

Well this has been the most incredible couple of weeks of model watching since the advent of the Internet. I have never seen the models so indecisive, in my opinion, no model has shown stability in the outlook beyond 96 hours, they have all struggled with this easterly or no easterly, however you wish to view it.

Now the GFS brings the easterly back at T96 as pressure builds to the east and northeast. The ECM also builds pressure to the east and just about gets the easterly on our door step, where as the UKMO keeps the high further north east and allows the low off the Atlantic to make inroads in to the UK, which one will be correct?

I have missed a lot of runs this week as I was hoping that if I left it a few days everything would settle down and become clearer , this is obviously not the case , I don't know if this is an example of how bad the models are preforming at the moment , but the models this morning show rain for today but the Met office / BBC are Insisting it will be Heavy Snow across the Midlands and Wales. Even the NAE model only Shows Snow for the far East . Basically what seems to be happening is every time the warm air looks like it will make inroads it can not make it across the whole country and the Cold air wins. This is what I think will prove to be a big turn around from the Models back towards Cold next week. +96 on the UKMO still gives us hope by the way IMO .

Edited February 3, 2010 by snowmadchrisuk

[cooling climate]

I can see big upgrades on the way over the next few days. I think height rises over Greenland are going to dictate the upcoming pattern not the Atlantic. ECM and METO are OK this morning, no disaster there at al and the GEFS Ens are a lot better than they were a couple of days ago, and improving.

I admit to struggling to see exactly what role (good or (more likely) bad) the Norway SW will play in all of this though. I never had this down as much of a player a week ago now it seems omnipresent and possibly overplayed by the charts? We shall see.

All to play for though still, fascinating (along with frustrating, annoying, tiring etc etc) watching the models at the moment. We may all moan about the 4 times per day GFS but, we'd sure miss it if it wasn't there. Think of it as a drunk uncle at a party. Most of the time it waffles on incoheriently then suddenly in a moment of inexplicable lucid soberness you are forced to sit up and listen, as its prophetic ramblings actually appear to make sense before, just as suddenly, it's back to knocking over the vase and generally annoying everyone.

Excellent stuff, I love the analogy with the uncle.

As far as the models are concerned I was quite optimistic when going to bed last night of

seeing more favorable model output today with regards to a colder outlook. It would be no

surprise to see the lows in the Atlantic take a more southerly track in further model output

and perhaps the energy over Scandinavia is still being modeled incorrectly IE amount of energy

being shown and also the direction this will take.

Plenty to be optimistic about this morning in my opinion.

[Ian Fergusson]

Judging by the latest EC ENS, the midday EC OP output could make for very interesting viewing..... more flip-flopping than a roach on a canal bank of late, but interestingly the SIG PPN signal now enhanced too along with the brrrrr factor .

Watch this space.....

[The Weather Dragon]

Big change on GFS, this weekends low never really makes inroads & then starts spinning back into the Atlantic as pressure rises to the NE - pressure would rise sooner if it wasn't for the norwegian shortwave. Amazing cosidering the amount of energy in that Atlantic low pressure system - this is no ordinary block.

Edited February 3, 2010 by Jonathan Evans

[Frosty.]

Judging by the latest EC ENS, the midday EC OP output could make for very interesting viewing..... more flip-flopping than a roach on a canal bank of late, but interestingly the SIG PPN signal now enhanced too along with the brrrrr factor .

Watch this space.....

Hi Ian,

Does that mean the meto will probably ditch their own output then in favour of the gem or ecm? the only reason I say that is that the ukmo looks very ordinary this morning. In my humble opinion, only the far northeast of the uk looks like being on the cold side with the majority having temps we would expect at this time of year.

[Snowplough33]

Morning everyone,

Well this has been the most incredible couple of weeks of model watching since the advent of the Internet. I have never seen the models so indecisive, in my opinion, no model has shown stability in the outlook beyond 96 hours, they have all struggled with this easterly or no easterly, however you wish to view it.

Now the GFS brings the easterly back at T96 as pressure builds to the east and northeast. The ECM also builds pressure to the east and just about gets the easterly on our door step, where as the UKMO keeps the high further north east and allows the low off the Atlantic to make inroads in to the UK, which one will be correct?

Is it not correct that even though GSF at T96 swings to the east, the feed off the continent is not particularly cold by this point anyway as the block c/w cold air has been pushed north?

[TEITS]

Incredible yet again from the models. However some of us yesterday hinted this may change.

Look at +72 again.

http://www.wetterzentrale.de/pics/Rtavn781.png

Now note rather than the LP moving NE that the models intially suggested this moves SE. Now yesterday myself, Tamara and a few others did say the LP between +120/+144 might start doing the same once it enters a closer timeframe. Look below at the GEFS control run and you will see this being correct.

http://www.wetterzentrale.de/pics/Rtavn781.png

http://91.121.94.83/modeles/gens/run/gens-0-1-138.png?0

Don't ask me why but for some reason when a LP comes up against blocking to the NE/E the models are reluctant to take this SE but rather NE. However when this appears in the +72/96 timeframe the models have a tendancy of moving this SE rather than NE.

The change in the ensembles are simply shocking considering the timeframe!

http://charts.netweather.tv/ensimages/ens.20100203/00/t850Cambridgeshire.png

Rather than making any assumptions im going to just sit on the fence and expect these wild wings to continue. Whatever the outcome this has been fasinating viewing.

[Solar Cycles]

Will it, won't it, will it, won't it! Amazing model watching, this is better than any whodunnit novel! I still maintain, that we will have to wait until at least Saturday, before we get cross model agreement for next week!

[Ian Fergusson]

Hi Ian,

Does that mean the meto will probably ditch their own output then in favour of the gem or ecm? the only reason I say that is that the ukmo looks very ordinary this morning. In my humble opinion, only the far northeast of the uk looks like being on the cold side with the majority having temps we would expect at this time of year.

Dunno, but it must be stressed they're regularly changing the prospects / making tweaks to the GM based on other models .... as you'd rightly expect. I've stressed this before. Fairly often I see briefings explaining tweaks based on better continuity in Arpege, GFS, etc etc (other nations doing likewise based on UKMO; it's all swings and roundabouts). I've never really seen the folks at Ops Centre locked slavishly into some quasi-religious following of their own (longer range) modelling at the expense of sound scientific analysis and due flexibility. They've remained largely bullish about the likelihood of something colder appearing into the 10-15 day period for a while now..... anyway, we'll see further mid-range briefing details after 12Z so I'll await with interest.

[Lewis]

After seeing the 00z it looks promising to say the least, how many times do we see massive low pressures on the GFS for it too in more reliable time frame fill out and not really push in.

In autumn when the atlantic dominates, even then we would see large low pressure systems in the atlantic shown pushing over the U.K with winds of upto 80pm etc, for it only in reliable time to deliver 40mph winds etc.

The GFS is not the most accurate model with regards energy over the Atlantic, and LP systems. But we still have to remember, the GFS is known for being shortwave crazy, which does not help with situations like this, it makes reading the models hard, and there's always uncertainity with the GFS model, especially when there is height building to the North and East with the Atlantic trying to push in.

Todays 00z run i think is the first time it's handled this situation with regards height building. Lets hope we continue to see this trend.

The metoffice are still fairly persistant with their long range outlook hinting colder weather.

Edited February 3, 2010 by Storm Force Lewis

[Nick F]

A look at the HPC prelim discussions this morning, and yet again they seem to be disregarding the GFS, which is understandable with its wild swings:

A BLEND OF 00Z AND 12Z/2 ECMWF RUNS PROVIDES A REASONABLE STARTING

POINT FOR LEANING AWAY FROM THE QUESTIONABLE GFS AND ACCOUNTING

FOR REMAINING DETAIL UNCERTAINTIES DAYS 3-4 SAT-SUN. THE MANUAL

FCST BEGINS TO INCORPORATE MORE GEFS/12Z ECMWF ENSEMBLE MEAN

GUIDANCE DAY 5 MON AND EXCLUSIVELY ENSEMBLE MEANS BY DAYS 6-7 TO

ACCOUNT FOR THE MORE PROGRESSIVE SCENARIO THAT TELECONNECTIONS

SEEM TO FAVOR OVER THE ERN HALF OF NOAM.

The theme is still there for northern blocking looking at the 00z 8-10day GFS/ECM mean height comparisons, with the jet and the troughing across the Atlantic tending to shift south. Though the core of +ve height anomalies centred over the Canadian arctic to Greenland giving that west based -ve NAO signal. Trying to put any detail on the surface flow likely is going to be hard beyond close range though!

[TEITS]

ECM ensembles show the OP was way warmer than the mean. However the control run does follow the OP very closely!

http://www.knmi.nl/exp/pluim/vijftiendaagse/Data/15DAAGSE_06260_NWT.png

The mean does seem lower compared to yesterdays 12Z.

Big upgrade for my location today on the NAE charts.

http://expert.weatheronline.co.uk/daten/proficharts/en/nae/2010/02/03/basis00/ukuk/prty/10020318_0300.gif

http://expert.weatheronline.co.uk/daten/proficharts/en/nae/2010/02/03/basis00/ukuk/prty/10020400_0300.gif

Edited February 3, 2010 by THE EYE IN THE SKY