Winter Model Discussion 18Z 8/1/13. The hunt for cold (is closer).

[Weathizard]

nogaps looks good

So can anyone find a model that agrees with regards to the SW and what Nick Sussex highlights? Even the NOGAPS has now backed away from the SW drama? If the GFS is right here I think it would be fair to say well done. But I'd still go 80-20 in favour of ECM as its supported strong by all over models.

[johnholmes]

I was wondering if someonme can help me. Ive been watching this forum for a couple of years now and one thing I cant understand is how people will know it will be very snowy from friday night until Sunday? From going by what alot of you have said over time, snow is very hard to predict and normally can only be predicted in a 24h time scale. Many thanks in advance , Alan

Its called hopecasting Alan-your assumption is correct

[Wivenswold]

Zakos re "the met have clearly binned the GFS 06z". How do you know that?

The Met Office haven't clearly binned it, as per Ian F's comments earlier, they don't consider it in their forecasts, regardless of what it throws up. They look at 00z and 12z only.

[alan776600]

Its called hopecasting Alan-your assumption is correct

Thank you John

[MR EXTREMES]

So can anyone find a model that agrees with regards to the SW and what Nick Sussex highlights? Even the NOGAPS has now backed away from the SW drama? If the GFS is right here I think it would be fair to say well done. But I'd still go 80-20 in favour of ECM as its supported strong by all over models.

certainly noticed the gfs likes its sw.

but really until tomorrow were know for sure whats happening with the gfs sw.

i think it looks odd personally but i would not rule anything out right now.

but i think phil is right there without any question things are cooling off nicely.

[johnholmes]

70% ECM 30% GFS

I would agree with those numbers-shame some of you do not look as phi has tried to show you, ignore run to run variablity and take the average or even check the 00 with the last 00 or the 12 with the last 12. How does each model compare to its last 2 or 3 runs. I really find some of the reactions to individual model runs as pretty silly unless you are a newcomer and still learning, those who have been on here for at least 12 months should know better. One does not have to check a model run just drop in here and do a quick check on the mood!

[SnowBallz]

I was wondering if someonme can help me. Ive been watching this forum for a couple of years now and one thing I cant understand is how people will know it will be very snowy from friday night until Sunday? From going by what alot of you have said over time, snow is very hard to predict and normally can only be predicted in a 24h time scale. Many thanks in advance , Alan

Hi Alan,

I think the answer lays within the micro and macro. On a macro level, it's entirely possible to gather from the models the atmospheric conditions, which would be conducive to snow (850s, wind direction, dew points, and so on...) So, on a broad scale, it's possible to state from a longer range than 24hrs that the conditions are favourable to snow.

However, on the micro level, it's a lot more difficult and this is where specifics - which often people request - are best left until the 24-36hr window, when such events are better scoped within the mesoscale models (NAE, UKV, etc...)

What is important to remember - in both instances - is that, unfortunately, nothing is guaranteed. Often there can be, for instance, 'mixing' which occurs that can disrupt/spoil even what looks like the most obvious snow event.

[Zakos]

Zakos re "the met have clearly binned the GFS 06z". How do you know that?

The Met Office haven't clearly binned it, as per Ian F's comments earlier, they don't consider it in their forecasts, regardless of what it throws up. They look at 00z and 12z only.

I see, I did not realise this. my apologies. Does that mean that they believe the 06z/18z to be less reliable?

[johnholmes]

Zakos re "the met have clearly binned the GFS 06z". How do you know that?

The Met Office haven't clearly binned it, as per Ian F's comments earlier, they don't consider it in their forecasts, regardless of what it throws up. They look at 00z and 12z only.

There have been enough comments from Ian F over the last few weeks that they hardly ever take notice of the GFS and certainly not in its further reaches

[craigore]

what time can we expect nxt set of models,charts?

[Ian Brown]

Surely the 06z isn't an outlier? Yes it's at the top end of the ensembles, but there are other ensembles that support it:

Obviously a massive split though

Certainly not an outlier, but the mean there is largely useless - it's one way or the other.

[Robbie Garrett]

This is the MetOffice response regarding variability and confidence. Very helpful guys!

" The variability of the outputs tends to drive our uncertainty, not just on one set of models runs, but how 

they are consistent, or not over a period of 1 to 2 days - Do they flip from one to the other."

Zakos re "the met have clearly binned the GFS 06z". How do you know that?

The Met Office haven't clearly binned it, as per Ian F's comments earlier, they don't consider it in their forecasts, regardless of what it throws up. They look at 00z and 12z only.

Edited January 9, 2013 by Robbie Garrett

[Chris D]

Hello,

Further to my enquiries. The MetOffice have said the following.

"Looking at the output this morning we see that the ECMWF has very definitely switched to a much colder

solution, whilst the GFS has gone milder, which was a big surprise. As a result we are likely to favour the colder ECMWF solution, although perhaps not as extreme, but it does not make the uncertainty any better."

Put simply, basically not a clue but they slightly favour a colder solution!

Summary is, that a pretty potent cold spell is coming up this weekend, with high potential for significant snow in some areas. Thereafter, it's either a bit of Atlantic stuff then some cold later in the month, or - the rest of the month blocked and cold.

[Yarmy]

Posted A minute ago

alan776600, on 09 January 2013 - 12:20 , said:

I was wondering if someonme can help me. Ive been watching this forum for a couple of years now and one thing I cant understand is how people will know it will be very snowy from friday night until Sunday? From going by what alot of you have said over time, snow is very hard to predict and normally can only be predicted in a 24h time scale. Many thanks in advance , Alan

Its called hopecasting Alan-your assumption is correct

I'm currently reading a book on weather which says the first ever computerised precipitation forecast was produced in 1967 by the Met. Just goes to show that predicting ppn is several orders of magnitude more difficult than pressure patterns (which as recent events show is hard enough!).

[The Northern Ramper]

I would agree with those numbers-shame some of you do not look as phi has tried to show you, ignore run to run variablity and take the average or even check the 00 with the last 00 or the 12 with the last 12. How does each model compare to its last 2 or 3 runs. I really find some of the reactions to individual model runs as pretty silly unless you are a newcomer and still learning, those who have been on here for at least 12 months should know better. One does not have to check a model run just drop in here and do a quick check on the mood!

Certain posters follow the models emotionally and struggle therefore to be objective and realistic, it doesn't matter how many times you bang this drum John it won't sink in.

[Alexis]

I was wondering if someonme can help me. Ive been watching this forum for a couple of years now and one thing I cant understand is how people will know it will be very snowy from friday night until Sunday? From going by what alot of you have said over time, snow is very hard to predict and normally can only be predicted in a 24h time scale. Many thanks in advance , Alan

Any precipitation is difficult to predict at range. The models can clearly show that the conditions would be highly favourable for snow, but a couple of days later they could be showing zero precipitation, or conditions have changed which reduce the chance of snow. Sometimes the models show that any precipitation would reach the ground as snow at sea level - these are the charts we want to see unchanged 24 hours before!

It's still difficult to predict rain in any one location in 5 days time in July.

Edited January 9, 2013 by Alexis

[garbagebags]

Can somebody tell me, with regards to models, as there is so many ensemble lines, how does the operational or control, or the pretty maps we see actually get chosen? I assumed that the charts we view would be bang on the "mean" of all the models ensembles but this doesnt seem to be the case?

[davacr123]

what time can we expect nxt set of models,charts?

Can be found in this link

[alan776600]

Hi Alan,

I think the answer lays within the micro and macro. On a macro level, it's entirely possible to gather from the models the atmospheric conditions, which would be conducive to snow (850s, wind direction, dew points, and so on...) So, on a broad scale, it's possible to state from a longer range than 24hrs that the conditions are favourable to snow.

However, on the micro level, it's a lot more difficult and this is where specifics - which often people request - are best left until the 24-36hr window, when such events are better scoped within the mesoscale models (NAE, UKV, etc...)

What is important to remember - in both instances - is that, unfortunately, nothing is guaranteed. Often there can be, for instance, 'mixing' which occurs that can disrupt/spoil even what looks like the most obvious snow event.

Thanks Snowballz, I love watching this forum although alot does confuse me and go over my head, but I can pick up the basics so a bit dumb in some respects to the people who have studied it for years. Im hoping on the short term we do get the snow as like all you coldies, im a big fan as the next person in here . Im guessing over the next few years I will learn more so im then more able to put posts on lol

[Robbie Garrett]

They favour the ECM more than the GFS in most cases.

Put simply, basically not a clue but they slightly favour a colder solution!

Summary is, that a pretty potent cold spell is coming up this weekend, with high potential for significant snow in some areas. Thereafter, it's either a bit of Atlantic stuff then some cold later in the month, or - the rest of the month blocked and cold.

[johnholmes]

Thanks Snowballz, I love watching this forum although alot does confuse me and go over my head, but I can pick up the basics so a bit dumb in some respects to the people who have studied it for years. Im hoping on the short term we do get the snow as like all you coldies, im a big fan as the next person in here . Im guessing over the next few years I will learn more so im then more able to put posts on lol

Alan to help you try looking in the Guides section, lots of info in there that will make model watching easier and more understnadable

Put simply, basically not a clue but they slightly favour a colder solution!

Summary is, that a pretty potent cold spell is coming up this weekend, with high potential for significant snow in some areas. Thereafter, it's either a bit of Atlantic stuff then some cold later in the month, or - the rest of the month blocked and cold.

Being unsure of the outcome is NOT the same as not having a clue

[Mr Sleet]

This looks very much like the climbdown we saw in December, with all the main models progging a sustained cold spell until a few days before the event, when the GFS suddenly had a strong south west flow blasting away any cold after a couple of days,and was proved correct.I suspect the same thing will happen here.

Edited January 9, 2013 by Mr Sleet

[Chris D]

They favour the ECM more than the GFS in most cases.

Well yeh I know they do, they think the GFS is rubbish, but I haven't seen any consistency in any model over the past several days or so. Therefore I think they have got to be still on the fence wrt next week and beyond.

Being unsure of the outcome is NOT the same as not having a clue

Well yes, I apologise they obviously do, sorry! Meant they are unsure, sorry I didnt make that clear

Edited January 9, 2013 by Chris D

[Gary_R_Walker]

So can anyone find a model that agrees with regards to the SW and what Nick Sussex highlights? Even the NOGAPS has now backed away from the SW drama? If the GFS is right here I think it would be fair to say well done. But I'd still go 80-20 in favour of ECM as its supported strong by all over models.

So this post by you earlier where you state, "normally I'd back the ECM over any other model by a long, long distance but I'm not so sure here. Especially as this run has 6hrs more data. Still very much 50-50 I'd say either way." Does that mean in less than two hours you have gone from 50-50 to 80-20 in favour of cold?

[MKN]

I would agree with those numbers-shame some of you do not look as phi has tried to show you, ignore run to run variablity and take the average or even check the 00 with the last 00 or the 12 with the last 12. How does each model compare to its last 2 or 3 runs. I really find some of the reactions to individual model runs as pretty silly unless you are a newcomer and still learning, those who have been on here for at least 12 months should know better. One does not have to check a model run just drop in here and do a quick check on the mood!

Its understandable though given what happened in December. Just as all models agreed by 1030pm that an easterly of some magnitude was on the cards the 00z gfs run puts a shortwave into the mix. And like you just said many people were saying just one run etc etc. Now im not saying im taking the 06z as gospel far from it but I can understand some who are a little bit more cautious now.