Model Output Discussion 29th December - Into mid-Winter.

[A.J]

1 minute ago, TEITS said:

Not just mine. I have never heard anyone else suggest this in any other weather forum for many years.

The GEFS ensembles are a good example of why its futile to compare with model runs over 24hrs ago. The 18Z GEFS control is identical to the ECM & 18Z GFS and yet the rest of the ensembles are a right mess between W,lys and E,lys. Now obviously something has changed and it's finely poised hence why the Ops/Control runs differ to some of the ensembles. So due to this and the chaotic nature of these forecasting models then data that is 24hrs old is useless.

that surprises me, as I've often seen it.....in the grand scheme of things everyone's entitled to their opinions......I do agree with you on the ensemble data as quite often they'll give you just about any physically possible solution amongst the postage stamps...lol..but personally I do like compare the op/control runs from 12/24 hours previous to try and get a handle of where we're heading and why

[Allseasons-Si]

What i forgot to add to my earlier post is this

gfs out in la! la! land a few days ago showed this at 384hrs

only to show this at 312hrs tonight,although watered down a tad the trend is very similar

so the signal is there,just need to get it down to a reliable time frame now.

 

Edited January 6, 2017 by Allseasons-si

[Mucka]

10 minutes ago, karlos1983 said:

What I was trying to say before my post disappeared, is looking at the ens, anyone that can predict an outcome in the extended is either lucky if they get it right, or has a crystal ball. 

The future is uncertain to say the least

Quite an interesting split developing there around the 15th.

I expect the colder ones build a HP across us and give us surface cold and maybe a few others manage an easterly type/continental flow from an Atlantic ridge pushing into Scandi as with the Op

Edited January 6, 2017 by Mucka

[TEITS]

Just now, Bristle boy said:

Maybe the model 'discrepancies' will turn out to be a good test of the relatively new weather teleconnections science. Just pondering......and watching.

I think you know my view on teleconnections!

Been meaning to say this and now its my opportunity. I do not agree with forecasts based on teleconnections. I feel its our lack of understanding them and how they interact with each other that makes me dismissive. However my opinion is based on the teleconnections and not a personal snipe at the members who use them.

Finally the models already take these into account and supercomputers are far more capable than humans in my opinion. As someone on here once said "You cannot solve the unsolvable". that is perfectly apt.

[Purga]

Majority of ENS trending cold with the Op & Control at the cold side of the pack.

Always an interesting sign when the Op & Control concur.

[nick sussex]

8 minutes ago, ajpoolshark said:

to be fair, this is in your opinion...Over my years on the forum I have seen numerous references from many members advising not reading too much i(ie over analysing) into consecutive runs especially for more than 4 days out,but rather look at the corresponding runs from 12 & 24 hours ago, and with John's many years as a met office forecaster, I wouldn't dismiss anything he advises IMHO

I support Daves opinion and can't see the logic in viewing just a particular run with the same run 24hours later run given that you're dealing with data 24 hours old.

[kumquat]

6 minutes ago, ajpoolshark said:

that surprises me, as I've often seen it.....in the grand scheme of things everyone's entitled to their opinions......I do agree with you on the ensemble data as quite often they'll give you just about any physically possible solution amongst the postage stamps...lol..but personally I do like compare the op/control runs from 12/24 hours previous to try and get a handle of where we're heading and why

It is all we've got, after all.

[TEITS]

2 minutes ago, Purga said:

Majority of ENS trending cold with the Op & Control at the cold side of the pack.

Always an interesting sign when the Op & Control concur.

Have to laugh at those ensembles. If anyone posts the GEFS mean tonight I shall scream.

[Snowman.]

GEFS mean

 

sorry

[Met4Cast]

Remember last week when all the ensembles were pointing towards cold weather and within 24 hours they went from cold, to less cold to mild. 

In situations like this, the ensembles are useless. If the ops can't get a hold on something you can forget the ensembles giving a clear idea. 

I'm not going to be getting excited about tonights charts until they're at +72. We've been here more times this winter than I care to count

Edited January 7, 2017 by Daniel Smith

[nick sussex]

2 minutes ago, TEITS said:

I think you know my view on teleconnections!

Been meaning to say this and now its my opportunity. I do not agree with forecasts based on teleconnections. I feel its our lack of understanding them and how they interact with each other that makes me dismissive. However my opinion is based on the teleconnections and not a personal snipe at the members who use them.

Finally the models already take these into account and supercomputers are far more capable than humans in my opinion. As someone on here once said "You cannot solve the unsolvable". that is perfectly apt.

That was probably me! lol I think the issue is its difficult to have say a controlled experiment with the weather as you do in the laboratory. We're never quite sure which variable tipped the balance to a certain direction. Of course science needs to move forward but with such chaos in the atmosphere even super computers will have a hard time in terms of longer range forecasting.

[karlos1983]

5 minutes ago, Mucka said:

Quite an interesting split developing there around the 15th.

I expect the colder ones build a HP across us and give us surface cold and maybe a few others manage an easterly type/continental flow from an Atlantic ridge pushing into Scandi as with the Op

Around the 15th is still the timeframe for me of interest. Could be a long haul to cold if it's the mild members that win out mid month. Interesting times.., said that a few times this winter 

[tight isobar]

Lol..

The age old argument/disagree. .

Ops against the variable ie ens etc.

Its quite amazing' the full frontal....then reverse psychology of "perception"!  When you see what tickles your fancy..over indulge in your fascinations. .then when minor adjustments cross your path implode- and/or follow thought of others. 

A simple tip would be to evaluate and dissect your own analysis. .

Without the guilt of who thinks what!

[kumquat]

5 minutes ago, karlos1983 said:

Around the 15th is still the timeframe for me of interest. Could be a long haul to cold if it's the mild members that win out mid month. Interesting times.., said that a few times this winter 

Indeed, and I'm still intrigued that Fred's forecast brought us here.

Major pattern change. Massive swing into SE by a potentially horrible mass of low pressure cold, sleet, hail and rain.

Probably exactly what one would need to push out High Pressure in the south.

I can't help but feel we have more chance of snow after 7 days and into the second half of winter.

[A.J]

11 minutes ago, nick sussex said:

I support Daves opinion and can't see the logic in viewing just a particular run with the same run 24hours later run given that you're dealing with data 24 hours old.

twice I've mentioned 12 & 24 hour data and not just 24 hour old data, have a re-read......whilst this is model related, it's not current model output discussion so I'll leave it there suffice to say many of have differing opinions, it's what makes this thread what it is........you agree with TEITS, I agree with John Holmes in this particular debate

[TEITS]

13 minutes ago, nick sussex said:

That was probably me! lol I think the issue is its difficult to have say a controlled experiment with the weather as you do in the laboratory. We're never quite sure which variable tipped the balance to a certain direction. Of course science needs to move forward but with such chaos in the atmosphere even super computers will have a hard time in terms of longer range forecasting.

Im sure if the ECM is right we shall here how the teleconnections suddenly became more favourable. More likely is the amount of sprouts I had over xmas has caused a stratospheric warming.

[chris55]

A nice 18z op, a decent ECM and hint from the 18z GEFS that things may be heading in a more favourable direction.

However I think we need to focus on the first cold shot from the northwest/north before we can have any kind of confidence in the possibility of the easterly following afterwards.

if the 00z and 12z suits tomorrow can hang on to the easterly idea then that will give us a little more clarity . As of tonight all I say is

"you only know the train is coming, when you can see the train"  and I can't see it yet ha. In fact it's still in bloody Gloucester!

 

[Jason M]

Well, GEFS have certainly changed. Much less zonal than earlier in the day. Maybe the ECM not such an outlier after all! Personally though, I'm still very reluctant to buy this. Lets see what the morning runs bring.

[Allseasons-Si]

35 minutes ago, blizzard81 said:

So that's ecm op, ecm control, gfs op, gfs control - all going for the easterly. What could possibly go wrong

Looking at the charts instead of text,then yeh,pretty good going

gfs/gefs/ecm and gem at 240hrs,gem the least likely on this run though or would it be least likely on the others and gem would be correct?

gefs mean and ecm mean at 240 hrs

gefs mean wants to stay with the NW pm flow of some sorts with a weaker signal to push height's up into Scandi and the ecm mean builds a ridge from the SW up into Scandi with trough disruption into Europe.

 

 

[nick sussex]

2 minutes ago, ajpoolshark said:

twice I've mentioned 12 & 24 hour data and not just 24 hour old data, have a re-read......whilst this is model related, it's not current model output discussion so I'll leave it there suffice to say many of have differing opinions, it's what makes this thread what it is........you agree with TEITS, I agree with John Holmes in this particular debate

Fine but JH's view is only compare run with same run so using that logic we shouldn't compare the 00hrs with the 12hrs. But yes we all have differing opinions which does make for interesting debate.

Taking this back to the current easterly drama this type of situation is very much effected by some smallish changes early on which either tip the balance for or against. The models seem to be having problems with the trough set up over Europe which is effected by the Russian high. It was evident from T120hrs on the ECM that it would be a better run as those positive heights were exerting more forcing on the troughing near the UK and the upstream pattern was better than the GFS 12hrs run.

Even though the GFS has had upgrades over the years it still is often the flattest solution over in the USA and this normally comes to the fore when the eastern seaboard of the USA is at risk from one of those infamous nor easters. Of course the biggest fail in recent years was that major storm which hit the east just before the 2012 Presidential election which led NCEP to really look at what was going wrong with the GFS. Inspite of upgrades NCEP still deride their own model for its flat bias both in the GFS op and the GEFS.

I think there was a degree of embarrassment on the part of US forecasters which had to rely on the ECM rather than their own home grown model. Putting the obvious deficiencies of the GFS aside we must at least applaud its generosity with data available to the public compared to the stinginess of the UKMO which guards its data like the Crown Jewels!

[Scandinavian High.]

More twists and turns to come I'm not getting my hopes up just yet..

[Richard Fisher]

4 minutes ago, abbie123 said:

More twists and turns to come I'm not getting my hopes up just yet..

Story of this winter confidence "low"

[Geordiesnow]

I think the chances of a cold NW'ly is getting more and more likely now, the questions will be just how cold it will be and how just how clean and strong the flow will be, past experiences tell me shortwaves will make their horrid present and we end up with quite messy charts and cold air being modified at times as hinted by the ECM run.

There is hints of some amplification but no sign of a Greenland high and I would say any easterly risk is very minimal at this stage and I expect the charts to alter again as you would expect at that range.

There is a chance the upcoming NW'ly could well be quite a cold one and with low thicknesses, then there should be plenty of instability and the chances of snow falling at lower levels being quite high but details regarding that is very uncertain at this stage and as I say, we will probably end up more of a messy flow with shortwaves embedded which will mix out the cold air somewhat, hope i'm wrong on that one though.

[Daniel*]

I think my reverse psychology worked! Fantastic GFS 18z pretty much a replica of ECM 12z. Allseasons is 100 per cent right the GFS OP on consecutive runs had that evolution which was later dropped - it seems it has resurfaced. We do know the GFS is good at picking up things! While the teleconnections are supposedly considered to have no influence on Blighty a tweet by mr fergie? It's interesting how we're sneaking increased amplification by the models, maybe the weather gods want to make up for our suffering. Hopefully the only one suffering in the future, will be that pesky squirrel, frozen stone cold! Don't call RSPCA on me.  

Would be typical of the winter for it all to crash down with 00z runs I suspect not. This Siberian high is an absolute monster EC T240 mean, surely it must be leaving a 'stamp' on the NH atmospheric circulation?

[Ali1977]

And GFS backtracks, what a surprise. Fri and Sat still look cold in a Northerly though . Lets see what the ECM does, I'll hazard a guess that it too will backtrack from a prolonged cold spell with an Easterly ending