Model Output Discussion:

[robert]

I find the Titan model gives a much bigger picture of this :wacko:

[damianslaw]

I usually look at the ECM first then the UKMO and very rarely GFS as it chops and changes on to regular a basis. Only three days ago it was showing a massive easterly and now a steamrolling south westerly. As John Holmes said, read the senior forecasters threads looking at the longer term upstream signals, they give you a much better insight. In fact I find it quite annoying seeing commentds about the latest GFS, perhaps we should set up a seperate model thread dedicated just to the GFS, its usually in association with the GFS that we see the most ridiculous and outrageous comments.

[Nick F]

One does have to question where the lows are getting all their umph from though. As has been already said, the sea temperatures are below average. I would also like to know what is causing the Atlantic to be so active in general as isn't February supposed to be quite quiet?

Very cold upper air pouring off NE Canada along polar front and a strong jet streak blasting across the N Atlantic is giving the fuel for those lows programmed by GFS, the end of the line of the jet streak across the Atlantic appears to be NW Europe with weak upper ridging building N across Europe, where the end of the line for the Atlantic is seems to be variable between runs, GFS has the 'end of the line' further east so more bullish to bring lows across the UK, as seen again on the 18z. Though I don't think we should be denial that the GFS is right, just because it shows mild SW'erlies and lows tracking NE across the UK.

[swfc]

So given the confusion and output anyone care to make a forecast for the next ten days.???.to be honest id have more chance of taking beyonce from behind while watching swfc win the champions lg than making any sense of that lot.Fax charts to me look the best route. :wacko:

[SP1986]

Whatever anyone thinks about the GFS it's still the best model at introducing trends. Furthermore if the GFS re-introduced a raging easterly tomorrow it would become everyones favourite model again.

I don't think it's the models here that have the problem, it's probably the assessment of the models in the near time frame, which given the current lack of certainty is not surprising.

No one can be blamed for being confused, but for me GFS is still as good a model as the ECM/UKMO, every model has it's ups and downs from time to time.

In a situations like this it would be unreasonable to say a models wrong when they're all finding it difficult to deal with the situation.

[smich]

I often don't bother looking at it. I will look at the ECM and UKMO first and then the GEM and afterwards, sometimes, see how the GFS compares to them all.

The only pattern that it is reliable with is tramline zonal south westerlies (as in corridor of death south westerlies not this particular pattern) Even then, it doesn't know when to stop and generates low bomb after low bomb and only generally comes to a screeching halt after the euros have picked up a pattern change about 2 days before!

Of course I am being a bit tongue in cheek and I will check the other data and stratosphere etc - but I am not a GFS fan I'm afraid!

This provokes an interesting question: Does anyone out there ignore the GFS? I admit it's the first one I turn to.........because it's there! 4 runs, ensembles, etc etc

But if you just checked the UKMO and ECM I wonder what your verification scores might be?

smich

[Bobby]

People pay attention to the GFS becausethe quantity of the data (because it's all freely available), not because of it's quality. ECM and UKMO are consistently more accurate than it.

The stats prove that.

[bluearmy]

I usually look at the ECM first then the UKMO and very rarely GFS as it chops and changes on to regular a basis. Only three days ago it was showing a massive easterly and now a steamrolling south westerly. As John Holmes said, read the senior forecasters threads looking at the longer term upstream signals, they give you a much better insight. In fact I find it quite annoying seeing commentds about the latest GFS, perhaps we should set up a seperate model thread dedicated just to the GFS, its usually in association with the GFS that we see the most ridiculous and outrageous comments.

GFS fi, normally something best ignored has thrown that ridge up again mid next week off the eastern seaboard. at this time we do have a propensity for raised heights over greenland. i recall either the mid december or early jan cold spell being spotted just beyond T240 by some gfs ops in a similar way. not saying that there is anything in it but worth looking for on ECM T240/GEM T240 tomorrow. GP has posted on the tech thread that there is a sign for a pressure rise somewhere in the north atlantic within the next couple of weeks.

[cyclonic happiness]

Whatever anyone thinks about the GFS it's still the best model at introducing trends. Furthermore if the GFS re-introduced a raging easterly tomorrow it would become everyones favourite model again.

I don't think it's the models here that have the problem, it's probably the assessment of the models in the near time frame, which given the current lack of certainty is not surprising.

No one can be blamed for being confused, but for me GFS is still as good a model as the ECM/UKMO, every model has it's ups and downs from time to time.

In a situations like this it would be unreasonable to say a models wrong when they're all finding it difficult to deal with the situation.

I had a look through the technical thread and from what I can gather without all the experience of you guys, is that there may be a cold spell but you still aren't sure? and I'm not being facetious, it seem as if you were saying that all the signals are go but the train is still in the station or am i reading it all wrong?

[Secret Society]

People pay attention to the GFS becausethe quantity of the data (because it's all freely available), not because of it's quality. ECM and UKMO are consistently more accurate than it.

The stats prove that.

yep at the end of that day, quality not quantity is what works.

[Guest North Sea Snow Convection]

Whatever anyone thinks about the GFS it's still the best model at introducing trends. Furthermore if the GFS re-introduced a raging easterly tomorrow it would become everyones favourite model again.

I don't think it's the models here that have the problem, it's probably the assessment of the models in the near time frame, which given the current lack of certainty is not surprising.

No one can be blamed for being confused, but for me GFS is still as good a model as the ECM/UKMO, every model has it's ups and downs from time to time.

In a situations like this it would be unreasonable to say a models wrong when they're all finding it difficult to deal with the situation.

No Stephen, it wouldn't neccesarily. At least speaking personally - although obviously I would like what it showed!

But by illustration of not always going along with it because it shows what I want it to - last week when the GFS persisted with showing a perishing easterly (you remember the first saga over the easterly?) I was continually voicing doubt over it and pointed to the euros who had dropped the idea. I was also sceptical about all the ultra OTT cold that the model was churning out only a few days ago that lots of people were speculating about.

I have only seen the easterly as most likely, strangely enough, in the Llast couple of days or so!

Edited February 2, 2010 by North Sea Snow Convection

[Secret Society]

Whatever anyone thinks about the GFS it's still the best model at introducing trends. Furthermore if the GFS re-introduced a raging easterly tomorrow it would become everyones favourite model again.

I don't think it's the models here that have the problem, it's probably the assessment of the models in the near time frame, which given the current lack of certainty is not surprising.

No one can be blamed for being confused, but for me GFS is still as good a model as the ECM/UKMO, every model has it's ups and downs from time to time.

In a situations like this it would be unreasonable to say a models wrong when they're all finding it difficult to deal with the situation.

i think you will find if it only came out twice a day, it would look even more laughable.

is it any wonder the U.S hardly use it during winter and refer to ecm.

[nick sussex]

This provokes an interesting question: Does anyone out there ignore the GFS? I admit it's the first one I turn to.........because it's there! 4 runs, ensembles, etc etc

But if you just checked the UKMO and ECM I wonder what your verification scores might be?

smich

Generally I check the ECM and UKMO first and then if they've got a cold outlook I don't really care what the GFS shows as that will normally cave in anyway!

To me the GFS is like Iceland supermarkets, lots of quantity and offers but poor quality and crappy old stodge that might fill you up but give you indigestion! The ECM is like Waitrose, you can't afford much but what you get is good quality, the UKMO is like Sainsburys , good but with the occasional irritating Jamie Oliver moment!

Apologies to Iceland fans but any company that employs that irritating Coleen Nolan deserves to go out of business!

Edited February 2, 2010 by nick sussex

[SP1986]

No Stephen, it wouldn't neccesarily. At least speaking personally. Last week when the GFS persisted with showing a perishing easterly (you remember the first saga over the easterly?) I was continually voicing doubt over it and pointed to the euros who had dropped the idea. I was also sceptical about all the ultra cold that the model was churning out only a few days ago that lots of people were speculating about.

I have only seen the easterly as most likely, strangely enough, in the couple of days or so!

Hi Tamara, I agree you and a few others would still retain some sense of balance, although some comments in here today lend me to believe that the GFS would become some people favourite depending on the outlook.

To be honest I'm still a little sceptical because in my eyes the scenario isn't too much different to last week. I do certainly admit an easterly could re-occur but past experience during the 'Atlantic years' have showed me that easterlies have trouble breaking through the Atlantic, although I suppose we should be thankful those lows arent penetrating to eastern Europe.

To be honest though it's literally so close between a southwesterly, westerly, easterly and god forbid for all the cold lovers, a Bartlett.

As the risk of being lynched for not complying to the popular route, I hope the GFS verifies because it will mix things up a bit and get people thinking. It would just be too handy to have an easterly occur.

Given my lack of preference for whatever happens, I am seeing this like a neutral supporter in a football team in that I want the favourite to lose, in a preferential sense.

[dave79]

Hi Guys, interesting reading as usual and a quick novice question to ask.

Do the BBC, Met and other Media Outlets use different models and charts etc to predict whats going to happen with the weather to what users on here use?

Edited February 2, 2010 by dave79

[LomondSnowstorm]

GFS fi, normally something best ignored has thrown that ridge up again mid next week off the eastern seaboard. at this time we do have a propensity for raised heights over greenland. i recall either the mid december or early jan cold spell being spotted just beyond T240 by some gfs ops in a similar way. not saying that there is anything in it but worth looking for on ECM T240/GEM T240 tomorrow. GP has posted on the tech thread that there is a sign for a pressure rise somewhere in the north atlantic within the next couple of weeks.

I have to be fair here and agree with yourself and Stephen: it is the best at picking up patterns, but it does have a tendency to exaggerate things and to go with the wrong pattern for too long. Maybe it has picked something up here, and the pub run has a knack of doing this, e.g. in December, and I can still recall that height rise from central France at +240 hours on the 18Z and the posting of archive charts showing that this is one of the most common, perhaps the most common, route to a significant cold spell.

Maybe it's onto something, maybe not, but the reliable doesn't give me all that much hope anyway!

[keithlucky]

Hi Guys, interesting reading as usual and a quick novice question to ask.

Do the BBC, Met and other Media Outlets use different models and charts etc to predict whats going to happen with the weather to what users on here use?

I wondered what charts they use because they still insist sunday turning cooler with a easterley developing.

[Secret Society]

I wondered what charts they use because they still insist sunday turning cooler with a easterley developing.

well early evening forecasts could have used 00Z and were too late to include in 5pm/6pm broadcasts with the ones on 12z.

but latest late evening forecast ive seen now mention nothing about weekend and wont go past tomorrow.

Edited February 2, 2010 by Secret Society

[dave79]

I wondered what charts they use because they still insist sunday turning cooler with a easterley developing.

Well say for example there was a major weather event on the way, would they want users on the net reading the same info they have. This could cause havoc in the press and on the streets etc. So surely they have other more programmed models which they use to recieve there information regarding the outcome in the weather.

[nick sussex]

I have to be fair here and agree with yourself and Stephen: it is the best at picking up patterns, but it does have a tendency to exaggerate things and to go with the wrong pattern for too long. Maybe it has picked something up here, and the pub run has a knack of doing this, e.g. in December, and I can still recall that height rise from central France at +240 hours on the 18Z and the posting of archive charts showing that this is one of the most common, perhaps the most common, route to a significant cold spell.

Maybe it's onto something, maybe not, but the reliable doesn't give me all that much hope anyway!

It was actually the GEM model that first went with the Azores high retrogression that delivered that short easterly before the northerly. The GFS is good for flat zonal patterns but people give it credit for picking a trend when in fact with 4 runs a day its bound to get something right eventually! And also even if it stumbles across something it often ditches it and comes back to it later by which time the ECM takes over whilst the GFS is still veering from one scenario to another before finally waking up.

It's great for free info but I've lost alot of faith in its output, although as we've seen recently all the models have had problems but when the dust settles the ECM will still be the best global model.

[keithlucky]

well early evening forecasts could have used 00Z and were latest one were too late to include in 5pm/6pm broadcasts with the ones on 12z.

but latest late evening forecast ive seen now mention nothing about weekend and wont go past tomorrow.

jaye wynee 9.25 bbc news24 turning dry friday gradually getting colder sunday hp building from the n east.

[Guest North Sea Snow Convection]

jaye wynee 9.25 bbc news24 turning dry friday gradually getting colder sunday hp building from the n east.

That actually is reasonably in line with the latest ECM as far as Sunday anyway. Certainly doesn't concur with the GFS.

[johnwirral]

Normally just lurk around on here but there seems to be a lot of winging about the inconstancy of the models

In my opinion they all do a good job considering the complexity of the task assigned to them, modelling the weather two weeks in advance across the surface of the globe!!!

Many of the more experienced and professional posters have mentioned that we are in a knife edge situation, this must mean that small changes of input data into the models will produce dramatic changes further downstream.

At least the GFS make all their data freely available, warts and all, something that the met office are probably unwilling to do, as people seem to read long range forecasts as gospel and are disappointed when the weather they are hoping for does not arrive. I can imagine the press every time the met office predicted extreme weather and it did not turn up..

As to what will happen next, I've not got the skills to say, but I would not be upset if a Siberian easterly was not on the cards, from past experience in my part of the world they tend to deliver clear sunny weather with biting winds. I would rather have the sort of Northerly setup that delivered the last lot of snow.

Just a hunch but this winter has delivered, and from all the signals there may be a sting in the tail.

Edited February 2, 2010 by johnwirral

[weather eater]

For those that haven’t worked it out yet the best way to avoid disappointment in regards model output is to realise that all model output past zero hour is subject to an increasing amount of revision and try to take in a boarder range of the evidence available rather than just getting caught up with one run or one model. There are some very knowledgeable people who post in this thread, yet for 80% of posters that information seems to go in one ear and out the other, for a member like JH it must be very frustrating, like having a lone voice in the wilderness.

[Tom D]

The 18z control run is a good example of why things are in so much disagreement. It's vastly different the the operational after about t+120.