Jump to content
Snow?
Local
Radar
Cold?
IGNORED

Model Output Discussion - cold spell to end Autumn


Paul
Message added by Paul

Please only post model discussion in this thread. 

For more general chat and banter, or moans and ramps loosely around the models, please head to the banter thread:
https://www.netweather.tv/forum/topic/86721-model-moans-ramps-and-banter/

For general weather chat including about the snow/cold chances around the country, please go to the regional threads:
https://www.netweather.tv/forum/forum/142-regional-discussions/

Thank you!

Recommended Posts

Posted
  • Location: Manchester Deansgate.
  • Weather Preferences: Heavy disruptive snowfall.
  • Location: Manchester Deansgate.
2 minutes ago, Nick L said:

I suspect due to the huge amount of computing power required to do so.

Would it not be better to have only 2 GFS suites per day and use the extra time created to run the computers for longer so that more calculations can be made, so that there is effectively 12 hours complete the suite, and thus improving resolution, and that is definitely true of the CFS, there is absolutely no need for 4 runs a day of a 9 monthly model.

  • Like 1
Link to comment
Share on other sites

Posted
  • Location: NR Worthing SE Coast
  • Location: NR Worthing SE Coast

So what the point of using means to determine the weather if all members apart from the op and control run at a lower res?.

 

they are going to be less accurate anyway.

Link to comment
Share on other sites

Posted
  • Location: Cheddington, Buckinghamshire
  • Weather Preferences: Winter: Cold & Snowy, Summer: Just not hot
  • Location: Cheddington, Buckinghamshire
3 minutes ago, feb1991blizzard said:

Would it not be better to have only 2 GFS suites per day and use the extra time created to run the computers for longer so that more calculations can be made, so that there is effectively 12 hours complete the suite, and thus improving resolution, and that is definitely true of the CFS, there is absolutely no need for 4 runs a day of a 9 monthly model.

Well the EC does this but not sure if that's necessarily the reason why the EC is far better than GFS.

Link to comment
Share on other sites

Posted
  • Location: just south of Doncaster, Sth Yorks
  • Location: just south of Doncaster, Sth Yorks
2 minutes ago, feb1991blizzard said:

Would it not be better to have only 2 GFS suites per day and use the extra time created to run the computers for longer so that more calculations can be made, so that there is effectively 12 hours complete the suite, and thus improving resolution, and that is definitely true of the CFS, there is absolutely no need for 4 runs a day of a 9 monthly model.

I think if you read the notes on the Met O page about computers, the time for each run is quite small. The runs are governed, not by the computers, but by the WMO main observing times of 00, 06, 12, and 18 Zulu.

WMO=World Meteorological Organisation

  • Thanks 1
Link to comment
Share on other sites

Posted
  • Location: Gillingham, Kent
  • Weather Preferences: Snow, Thunderstorms,
  • Location: Gillingham, Kent
2 minutes ago, SLEETY said:

So what the point of using means to determine the weather if all members apart from the op and control run at a lower res?.

 

they are going to be less accurate anyway.

They can still give a general idea, they're not run at a drastically lower resolution so that they're completely useless, but the number of ensembles with slightly different starting data give a better idea of future prospects than having just a single op run. 

Having said that, in situations where very minor changes of energy lead to massive downstream changes (like we're currently dealing with) the ensembles are basically useless.

Link to comment
Share on other sites

Posted
  • Location: Wakefield - God's County
  • Location: Wakefield - God's County

 

Latest CFS model brings the high pressure over southern & central UK next week stopping the cold returning and instead bringing milder weather across the whole of the UK & Ireland. 

All a case of wait and see, although it's very likely the weather will turn less cold this weekend but we lose the sunshine and wintry showers and replace it with a lot of cloud and some patchy rain from weak weather fronts. 

image.png

OR...

I think most posters on here know where our weather is heading from the middle of next week and that is into the freezer. It could be for quite some time as well.

should be an excellent afternoon and evenings model watching.

 

Make your minds up peeps....!! ....who is right...????....who knows !!!

 

  • Like 1
Link to comment
Share on other sites

Posted
  • Location: just south of Doncaster, Sth Yorks
  • Location: just south of Doncaster, Sth Yorks
4 minutes ago, Nick L said:

Well the EC does this but not sure if that's necessarily the reason why the EC is far better than GFS.

it has to do with how the data sets are run, perhaps some indication on the free part of the ECMWF site, or that on the Met O site may give information on the differences between the 3 major models

Edited by johnholmes
Link to comment
Share on other sites

Posted
  • Location: just south of Doncaster, Sth Yorks
  • Location: just south of Doncaster, Sth Yorks
1 minute ago, Daniel Smith said:

They can still give a general idea, they're not run at a drastically lower resolution so that they're completely useless, but the number of ensembles with slightly different starting data give a better idea of future prospects than having just a single op run. 

Having said that, in situations where very minor changes of energy lead to massive downstream changes (like we're currently dealing with) the ensembles are basically useless.

I would not agree with ensembles being as you suggest. They are the main tool used in some professional organisations but they are used carefully and without bias. By bias I am referring to the weather bias most on here have, be it for cold or heat. No dig that we all know that almost everyone on a weather forum has a bias. As an ex professional I can assure you professionals are without bias but it can be very hard to remain so at times.

Link to comment
Share on other sites

Posted
  • Location: Castle Black, the Wall, the North
  • Weather Preferences: Spanish Plumes, Blizzards, Severe Frosts :-)
  • Location: Castle Black, the Wall, the North
2 minutes ago, K9 said:

 

Latest CFS model brings the high pressure over southern & central UK next week stopping the cold returning and instead bringing milder weather across the whole of the UK & Ireland. 

All a case of wait and see, although it's very likely the weather will turn less cold this weekend but we lose the sunshine and wintry showers and replace it with a lot of cloud and some patchy rain from weak weather fronts. 

image.png

OR...

I think most posters on here know where our weather is heading from the middle of next week and that is into the freezer. It could be for quite some time as well.

should be an excellent afternoon and evenings model watching.

 

Make your minds up peeps....!! ....who is right...????....who knows !!!

 

I believe the opposite of what the CFS shows is usually correct..that goes for next week too.:santa-emoji::D

  • Like 4
Link to comment
Share on other sites

Posted
  • Location: just south of Doncaster, Sth Yorks
  • Location: just south of Doncaster, Sth Yorks
18 minutes ago, Dacyfo said:

This may seem a daft question (I'm still but a novice at this) but why aren't all members just run at high Resolution.

The time constraint is the main reason, that time is the need for the many organisations requiring forecasts at set times. What the models spew out then has to be run through another interface to produce whatever the customer wants, or that was how it was done. Again check with the UK Met site for the latest explanation, I may be out of date.

  • Thanks 2
Link to comment
Share on other sites

Posted
  • Location: Gillingham, Kent
  • Weather Preferences: Snow, Thunderstorms,
  • Location: Gillingham, Kent
2 minutes ago, johnholmes said:

I would not agree with ensembles being as you suggest. They are the main tool used in some professional organisations but they are used carefully and without bias. By bias I am referring to the weather bias most on here have, be it for cold or heat. No dig that we all know that almost everyone on a weather forum has a bias. As an ex professional I can assure you professionals are without bias but it can be very hard to remain so at times.

No I agree with that, not sure where the bias talk came from.

I was merely stating that in situations where very minor changes lead to massively different outcomes (be it mild, cold or whatever) the ensembles are useless because they can all go a completely different way.

Link to comment
Share on other sites

Posted
  • Location: Manchester Deansgate.
  • Weather Preferences: Heavy disruptive snowfall.
  • Location: Manchester Deansgate.
5 minutes ago, johnholmes said:

I think if you read the notes on the Met O page about computers, the time for each run is quite small. The runs are governed, not by the computers, but by the WMO main observing times of 00, 06, 12, and 18 Zulu.

WMO=World Meteorological Organisation

So are you saying that the 12z, for instance, has to be initialised at 1530?, surely the WMO have no power over government funded organisations from individual countries?

Link to comment
Share on other sites

Posted
  • Location: DL
  • Location: DL
12 minutes ago, johnholmes said:

it has to do with how the data sets are run, perhaps some indication on the free part of the ECMWF site, or that on the Met O site may give information on the differences between the 3 major models

Last I heard ECMWF uses spectral methods whereas GFS doesn't.  I don't know about UKMO but I don't think they do either.  I'd imagine solving all those basis expansions is hugely computationally expensive, but at a theoretical level it is a huge difference.

Link to comment
Share on other sites

Posted
  • Location: Manchester Deansgate.
  • Weather Preferences: Heavy disruptive snowfall.
  • Location: Manchester Deansgate.
9 minutes ago, Daniel Smith said:

No I agree with that, not sure where the bias talk came from.

I was merely stating that in situations where very minor changes lead to massively different outcomes (be it mild, cold or whatever) the ensembles are useless because they can all go a completely different way.

But that's why the starting data is changed and they are run many times, I will categorically assure you now that if the op runs started throwing out insanely cold outliers now in the mid range, that they would not be 'leading the way' and they almost certainly wouldn't verify.

Link to comment
Share on other sites

Posted
  • Location: Gillingham, Kent
  • Weather Preferences: Snow, Thunderstorms,
  • Location: Gillingham, Kent

Wasn't there an experiment where the ECM computers were fed GFS data and it turned out to be more accurate than both of the models? I might be miss-remembering.

Anyway I fear we're getting a bit off topic, lets hope for a good 12z run!

  • Like 1
Link to comment
Share on other sites

Posted
  • Location: Gillingham, Kent
  • Weather Preferences: Snow, Thunderstorms,
  • Location: Gillingham, Kent
Just now, feb1991blizzard said:

But that's why the starting data is changed and they are run many times, I will categorically assure you now that if the op runs started throwing out insanely cold outliers now in the mid range, that they would not be 'leading the way' and they almost certainly wouldn't verify.

They did in 2010.. and in 2013, and in any other "notable" cold spell in the past few years

Link to comment
Share on other sites

Posted
  • Location: Aviemore
  • Location: Aviemore
13 minutes ago, Daniel Smith said:

I was merely stating that in situations where very minor changes lead to massively different outcomes (be it mild, cold or whatever) the ensembles are useless because they can all go a completely different way.

That doesn't make them useless though, as one of the key uses of ensembles is to give an idea of the confidence you can have in the current output. So if members are wildly different to eachtother, or run to run, or from the operational suite, then you can assume confidence in any outcome is low at that stage. 

  • Like 6
Link to comment
Share on other sites

Posted
  • Location: Manchester Deansgate.
  • Weather Preferences: Heavy disruptive snowfall.
  • Location: Manchester Deansgate.
Just now, Daniel Smith said:

They did in 2010.. and in 2013, and in any other "notable" cold spell in the past few years

But not in the 144-168 range which is what we have been discussing in the last few days regarding shortwave modelling, and for every time the ops picked up a cold pattern in the extended range first, I bet there is umpteen times where a few ensembles  dipped first.

Link to comment
Share on other sites

Posted
  • Location: Netherlands close to the coast
  • Location: Netherlands close to the coast
1 hour ago, Luke Best said:

I don't believe that this is correct.

As I understand it, the op is run at a high resolution and the control at a lower resolution.

The 'GEFS members' are run with slightly adjusted starting data (and also at a lower resolution).

No that's ecmwf, GEFS runs at the same resolution as gfs but with slightly different starting data 

Link to comment
Share on other sites

Posted
  • Location: Aviemore
  • Location: Aviemore
14 minutes ago, ArHu3 said:

No that's ecmwf, GEFS runs at the same resolution as gfs but with slightly different starting data 

That's incorrect - the GFS is run at 0.25 degrees, GEFS at 1 degree. 

Link to comment
Share on other sites

Posted
  • Location: Warsaw, Poland. Formerly London.
  • Weather Preferences: Four true seasons. Hot summers and cold winters.
  • Location: Warsaw, Poland. Formerly London.
13 minutes ago, captaincroc said:

Want to post this before the 12z GFS comes out but this is the run with the most data (along with the 0z) if my memory serves me correct. So let's hope it's a good start to the 12z suite.

Yes, some truth to that, but conversely the 'uselessness' of the 06z and 18z is massively overstated on this forum at times, usually when it doesn't show what people want.

  • Like 1
Link to comment
Share on other sites

Posted
  • Location: Netherlands close to the coast
  • Location: Netherlands close to the coast
5 minutes ago, Paul said:

That's incorrect - the GFS is run at 0.25 degrees, GEFS at 1 degree. 

Then the info onmeteociel is incorrect (according to meteociel all run 1 degree? and there is another high resolution gfs, which only does certain areas 

Edited by ArHu3
Link to comment
Share on other sites

Posted
  • Location: Scouthead Oldham 295mASL
  • Location: Scouthead Oldham 295mASL

Well early days but im wondering if this is not going to be as good as the previous runs, not sure the high is as far north, early days tho hopefully i'm being too negative..

Link to comment
Share on other sites

Guest
This topic is now closed to further replies.
×
×
  • Create New...