Model Output Discussion:

[johnholmes]

Thanks mate,

So it's safe to say, with such model agreement across the board, obviously your always going to get milder/warmer members, that we are more or less 70/30 favoured for a very cold start to next week. Just a lot of people are saying the models change so quickly etc and not too get excited, i agree thats the best way of looking at it, forget snow, get the cold and pattern/trend first, certainly looks like we have that.

After all, the other week before it all went pear shaped some of the models (not all) was showing an east/north easterly. It dissapeared and now its back on all models, and what a coinsidence, it vanished and it appeared on the same date!

Mouth watering outputs.

in terms of % probs I would say for a week now its been more likely than not and with these odds

60-40 becoming 80-20 over the past 2 days or so in favour of the cold spell actually occurring.

[Lewis]

in terms of % probs I would say for a week now its been more likely than not and with these odds

60-40 becoming 80-20 over the past 2 days or so in favour of the cold spell actually occurring.

Cheers John,

When will you be doing an update on your LRF? Look forward to reading.

[nick sussex]

Good news from the latest MJO info.

It's continuing to move slowly through phase 7.

http://www.cawcr.gov.au/bmrc/clfor/cfstaff/matw/maproom/RMM/phase.Last40days.html

[TEITS]

Hi,

With you seeing yourself as more advanced than others with regards the model outputs, you should get a clear idea just from the charts (experience does not come down too it with regards your own area) it does when it's nowcasting, and your in the time frame.

I disagree.

For instance during the Dec 05 E,ly the models didn'r predict much snowfall for my region and yet I knew these were wrong because of my experiences under similiar synoptics.

Sorry but whilst I will continue posting about the general trend of the models for the whole of the UK. When it comes to details I cannot cover all regions because I will be here all day!

Back to the models and its interesting there is more scatter for London/Kent than say Cambs/Lincs. This clearly suggests LP systems tracking to the S on some of the members. I haven't had time to view each member.

P.S I don't see myself as more advanced. Just an ordinary member sharing the same passions as others.

Edited February 4, 2010 by THE EYE IN THE SKY

[johnholmes]

Cheers John,

When will you be doing an update on your LRF? Look forward to reading.

best you read my blog Lewis I've stopped doing them

[Geordiesnow]

Its nice too see that any shortwaves near Iceland have been removed now, this makes the ridging much safer and more likely to come off.

How all this affects our weather probably remains to be seen, We can easily have snowy runs like yesterday's ECM run and todays 06Z GFS run or we can have something like this morning's ECM which is must drier although there is a period of time where some snow showers will fall in eastern parts.

So whilst another cold shot looks more and more likely, the detail needs to be sorted out and it could easily be a cold dry spell for most or a cold snowy spell for most. Whatever happens, it should be cold.

[cooling climate]

I'm quitetly beginning to gain some confidece now that the runs are in some agreement, could any 1 with more knowledge PLEASE explain what sort of temperatures we could expect by Monday and Tuesday according to the latest runs. Like some 1 mentioned earlier lets get the cold first then the Snow potential if any.

Will this be the extremely cold Easterly that was mentioned a while ago or a weakened Easterly some members have been posting charts showing -15 is that possible in my part of the country East Hertfordshire? Your thoughts please on the model runs at present in terms of temperatures from the ensembles or should I wait for the answer after the next 2 runs, I hate to say this but after last weeks fiasco on the model runs Im still a little cautious but I understand that all of the models are showing some dgree of cold beginning next week. Thankyou.

If the 06z is correct in its idea to bring the very cold pool of air and instability southwest then we in this

neck of the woods along with all the other areas mentioned could see some appreciable falls of snow

next week.

Also many parts of the UK could very well be looking at several ice days (daytime max temperatures of

0c or lower) at least.

Lets see what the afternoon and evenings runs bring.

[Lewis]

best you read my blog Lewis I've stopped doing them

Ohh no, i loved your LRF'S, i hope you do the odd one or two in the future, and don't forget to spend as much time as you can in the Yorkshire and Lincs thread, we will need a forecaster of your experience in the thread next week.

[Duane S.]

If anyones bored and wants a very interesting read heres an excellent link about how the models work, the most interesting part is headed inherent model errors.

http://www.hpc.ncep.noaa.gov/mdd/

Yes I know I should get out more!

Great link Nick,

"In order to capture the TRUE state of the atmosphere, we would require observations horizontally and vertically down to a subatomic scale over the ENTIRE GLOBE. No such observational network exists"

I wonder if that will ever (could ever) happen?? To me, the only likely advance in model accuracy for the future would be to improve the initial state accuaracy. I guess higher resolutions won't always help and could in fact exacerbate any errors

[karyo]

Good to see the MJO progressing east of the dateline http://cawcr.gov.au/bmrc/clfor/cfstaff/matw/maproom/RMM/phase.Last40days.gif

Hopefully it will make it to phase 8 as slowly as possible!

Karyo

[locneverstop]

Great link Nick,

"In order to capture the TRUE state of the atmosphere, we would require observations horizontally and vertically down to a subatomic scale over the ENTIRE GLOBE. No such observational network exists"

I wonder if that will ever (could ever) happen?? To me, the only likely advance in model accuracy for the future would be to improve the initial state accuaracy. I guess higher resolutions won't always help and could in fact exacerbate any errors

Just think how far avionics has advanced. Just 60 years ago you would have been told supersonic flight was impossible, never mind manned space flight.

Maybe...Just maybe... in 30+ years we will have a GFS accurate to within three years.... We can dream!

A slight adjustment back to topic... When comparing models for the overall situation for cold and/or heat, I've been using a technique of watching out for the 528dam lines and 564 lines for cold/heat waves respectively on the HGT 500-1000 charts. Can anyone reccomend a more accurate chart to use as I believe my method may be ever so slightly innacurate...

[Slinky]

Maybe...Just maybe... in 30+ years we will have a GFS accurate to within three years.... We can dream!

That would take all the fun out of it though dont you think?

Sorry off topic :lol:

[Southender]

Was it not the GFS model that picked up on the cold signal first, nearly 14 days ago?

If I remember rightly it predicted a cold outbreak from the E/NE around the 8th Feb?

And that date still stands by the looks of it...

You have to give the model credit really. OK it's been like a rollercoaster since, but all of the models have.

If the pattern verifies (which is looking increasingly likely) then there has to be a big tip of the hat to the GFS in my opinion.

Edited February 4, 2010 by LeighShrimper

[Eugene]

Was it not the GFS model that picked up on the cold signal first, nearly 14 days ago?

If I remember rightly it predicted a cold outbreak from the E/NE around the 8th Feb?

And that date still stands by the looks of it...

You have to give the model credit really. OK it's been like a rollercoaster since, but all of the models have.

If the pattern verifies (which is looking increasingly likely) then there has to be a big tip of the hat to the GFS in my opinion.

Yes you are correct in everything you say, i posted in here yesterday that GFS was the leader and that yes it changes it mind like other models but in the end is the first model to settle on the eventual outcome.

[D KARLSON]

hopefully as we close in on these forecast cold or very cold synoptics ,we can start to discuss snowfall amounts.

As it stands at the moment and this is conjecture, but eastern northern and southern areas could well catch the brunt of disruptive snowfall and it may well be that february saves the best till last.

Just glancing at the METO update which will of course change; the idea of a north or northeasterly airflow with fronts attempting to move into southern counties is the idea put forward, obviously reality will be different but this could well turn out to be near the mark.

These synoptics if they verify would produce heavy snowfall quite widely certainly eastern areas would suffer from heavy and prolonged snow showers ,but southern counties could see their worst snow in many a year if these projected fronts attempt to move up from the south, as some of our worst and snowiest winters were produced from these types of synoptics ,even something like a taste of a 46 47 style cold spell is not out of the question by any means.

I certainly think there will a few happy faces on here in the coming week or so and also the reverse as the country really does grind to a halt as i think we still have not seen the worst that a british winter can throw at us

[sandstorm2]

If anyones bored and wants a very interesting read heres an excellent link about how the models work, the most interesting part is headed inherent model errors.

http://www.hpc.ncep.noaa.gov/mdd/

Yes I know I should get out more! :lol:

After reading this I will never criticise a model output again, and really it is a miracle we have charts and models to debate at all, never mind the accuracy of them.

Thanks for the link Nick

SS2

[Buzzit]

Was it not the GFS model that picked up on the cold signal first, nearly 14 days ago?

If the pattern verifies (which is looking increasingly likely) then there has to be a big tip of the hat to the GFS in my opinion.

I understand your point Leigh, and it is not the first time this winter that a similar occurence has happened with the GFS

However the logic is slightly messed up in that now all the other days it showed something different wont verify....

So is it actually any more accurate than any other model ? Probably Not...

Does it seem to pick up on things other models do not in FI at times ?

Yes but that is not a logical indicator that the GFS is more accurate than other models

My own view is I am glad that we all have the GFS, primarily because the more models we have the better, but I also like

seeing 'Fools Gold' in FI (which GFS seems to do more than any other model at the moment) and then enjoy the occasional pot at the end of the rainbow when it comes off...

Edited February 4, 2010 by Buzzit

[Bobby]

Compare OP mean

http://91.121.94.83/modeles/gens/run/gens-21-0-168.png?6

To parallel mean

http://91.121.94.83/modeles/gensp/run/gens-21-0-168.png?6

Paralell ensembles are more suggesting a channel low set up than the op ensembles, those means show that... more Ely flow than NEly.

Edited February 4, 2010 by Barb-

[Strider]

Was it not the GFS model that picked up on the cold signal first, nearly 14 days ago?

If I remember rightly it predicted a cold outbreak from the E/NE around the 8th Feb?

And that date still stands by the looks of it...

You have to give the model credit really. OK it's been like a rollercoaster since, but all of the models have.

If the pattern verifies (which is looking increasingly likely) then there has to be a big tip of the hat to the GFS in my opinion.

It did, but then by last weekend it went blowtorch again.

It a model keeps blowing hot and cold (sorry) then it is guaranteed to be correct at some point!

A more relevant question is which model consistently predicted the cold spell before the others? UKMO..?

Edited February 4, 2010 by Strider

[Just Below Zero]

Was it not the GFS model that picked up on the cold signal first, nearly 14 days ago?

If I remember rightly it predicted a cold outbreak from the E/NE around the 8th Feb?

And that date still stands by the looks of it...

You have to give the model credit really. OK it's been like a rollercoaster since, but all of the models have.

If the pattern verifies (which is looking increasingly likely) then there has to be a big tip of the hat to the GFS in my opinion.

14 days? Thats 56 model runs all totally different.

Well if there were enough monkeys........

[swilliam]

Was it not the GFS model that picked up on the cold signal first, nearly 14 days ago?

If I remember rightly it predicted a cold outbreak from the E/NE around the 8th Feb?

And that date still stands by the looks of it...

You have to give the model credit really. OK it's been like a rollercoaster since, but all of the models have.

If the pattern verifies (which is looking increasingly likely) then there has to be a big tip of the hat to the GFS in my opinion.

I agree with this - that is the whole point of looking at and discussing FI which some people moan about - identifying possibilities and discussing the likelihood of them coming off. It is easy for the UKMO GM to look good - it only goes out to +144. Yes the pattern has come and gone and come back again but the 8th Feb has always been the favoured date for the start. Let's see if it does actually happen now - If so I would definitely say we would need to give the model some credit - it just needs to be used in the right way - particularity not slavishly following each run.

It did, but then by last weekend it went blowtorch again.

It a model keeps blowing hot and cold (sorry) then it is guaranteed to be correct at some point!

A more relevant question is which model consistently predicted the cold spell before the others? UKMO..?

errrr no - when it finally came into range -first it didn't show it - then it did - then it didn't and now finally it has come back to it again.

14 days? Thats 56 model runs all totally different.

Well if there were enough monkeys........

No - it stuck with it (somewhat intermittently - especially wrt to the start) for several days which is why it was discussed - then 'came into line with Euros' by dropping it for the extended SW/NE LP with SW winds - then cam back to it (albeit after the ECM).

Edited February 4, 2010 by swilliam

[kold weather]

Right to close this thread shortly for a brand new one.

My view is quite simple really, pretty likely we are now going to be on the northern side of the jet and this means cold conditions are pretty much nailed on for probably a good week at the moment, and who knows how long it may extend. The main uncertainty seems to remain with regards to just how potent any shot will be and also whether there will be enough instablity, the GFS is probably the best in this regard though all three probably at times could have spells of snow...