Jump to content
Snow?
Local
Radar
Cold?
IGNORED

Model output discussion - 7th January onwards


Paul
Message added by Paul

Please keep this thread solely to discussing the model output. The model banter thread is available for more loosely model related chat, ramps and moans.

Recommended Posts

Posted
  • Location: Hertfordshire
  • Weather Preferences: SNOW AND FREEZING TEMPS
  • Location: Hertfordshire
4 minutes ago, Buzz said:

I've been pondering that too - surely if it's the most accurate then it's now showing the most favoured outcome?

Or is that only when it shows what people want to see?

GFS P is not the most accurate model ?? And even if it was it is done globally and not just around are shores . 

ECM - 1st 

UKMO- 2nd 

the rest make up the numbers . The meto use ECM , UKMO and in house models and do you no why ? Because they are the best out there .

  • Like 2
Link to comment
Share on other sites

Posted
  • Location: STEVENAGE, HERTS (100M ASL)
  • Location: STEVENAGE, HERTS (100M ASL)
10 minutes ago, Nick L said:

Usual rant from me, these "snow depth charts" are in no way indicative of what the EC is actually showing. The algorithms in them are hyped up purely for clickbait, that's the only explanation I can think of.

Edit: Just noticed that foot and a half of snow near Dublin, yeah alright 

Are you able to share what that run would actually produce? 0.8 inches for East Anglia for example doesn’t seem that hyped up for an easterly?

  • Like 1
Link to comment
Share on other sites

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

Usual rant from me, these "snow depth charts" are in no way indicative of what the EC is actually showing. The algorithms in them are hyped up purely for clickbait, that's the only explanation I can think of.

Edit: Just noticed that foot and a half of snow near Dublin, yeah alright 

Agree, but they did get 1ft last year in parts of that area.

  • Like 1
  • Thanks 1
Link to comment
Share on other sites

Posted
  • Location: Longton, Stoke-on-Trent.
  • Location: Longton, Stoke-on-Trent.

Strange that the GFSP was throwing out the most wintry charts earlier this winter but is now the only one of the main models not interested in significant cold.

You would think that the GFSP overall would verify better than the GFS but it’s too new a model to know whether it handles this sort of setup better.

  • Like 4
Link to comment
Share on other sites

Posted
  • Location: Reading/New York/Chicago
  • Location: Reading/New York/Chicago
8 minutes ago, Buzz said:

I've been pondering that too - surely if it's the most accurate then it's now showing the most favoured outcome?

Or is that only when it shows what people want to see?

It's one solution amongst others. ECM and cluster, current GFS plus ensembles and UKMO (so far as we can see) largely back the colder solution.

It may be that the GFSp has the correct outcome, but we can only go on probabilities dictated by the models we have and, at the moment, it favours the colder outcome.

Link to comment
Share on other sites

3 minutes ago, weirpig said:

I think you are right

Ecm

Ukmo

Para

GFS 

No bad thing going by this mornings ECM 

What we want to see is the charts today carry on through the weekend and then we can tell our friends family and the world.. 

Wontbe suprised to see "downgrades" as the charts this morning are sort of once in 100 year type events 

Link to comment
Share on other sites

Posted
  • Location: uxbridge middlesex(- also Bampton oxfordshire
  • Weather Preferences: blizzard conditions. ice days
  • Location: uxbridge middlesex(- also Bampton oxfordshire

Something to take onboard..

Via prog/stat.

Screenshot_2019-01-17-13-05-04.png

  • Thanks 9
Link to comment
Share on other sites

Posted
  • Location: Liphook
  • Location: Liphook
14 minutes ago, Buzz said:

I've been pondering that too - surely if it's the most accurate then it's now showing the most favoured outcome?

Or is that only when it shows what people want to see?

It has shown NO improvement over the GFS thus far in global verification stats. Both have been running behind UKMO, ECM and even GEM in the last month.

In other words, it still needs tweaks and I've heard over twitter that some usa Mets are secretly happy with the delay as it's not yet upto task. These things often take a little time to iron out problems.

Edited by kold weather
Link to comment
Share on other sites

Posted
  • Location: HILL OF TARA IRELAND
  • Weather Preferences: SNOW ICE
  • Location: HILL OF TARA IRELAND
17 minutes ago, Nick L said:

Usual rant from me, these "snow depth charts" are in no way indicative of what the EC is actually showing. The algorithms in them are hyped up purely for clickbait, that's the only explanation I can think of.

Edit: Just noticed that foot and a half of snow near Dublin, yeah alright 

It's happened in March ,but you are correct just for fun at this Range. 

Link to comment
Share on other sites

Posted
  • Location: Kensington
  • Location: Kensington
3 minutes ago, tight isobar said:

Something to take onboard..

Via prog/stat.

Screenshot_2019-01-17-13-05-04.png

MMM  so by that the ensembles for the gfs  may be under estimating the strength of the Storm out of the US  which in turn may cause some of the perbs to not have substancial ridging?. 

  • Like 1
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, Tim Bland said:

Are you able to share what that run would actually produce? 0.8 inches for East Anglia for example doesn’t seem that hyped up for an easterly?

That's not too farfetched, but a broken clock is right twice a day. But take for example southern Scotland. The ENTIRE EC RUN only has 10-15mm of precipitation. Yet on that chart it's got 15-20 INCHES of snow on the ground! Even assuming all that 10 days of precip falls as snow AND settles AND lasts the 10 days, it's still a tall ask.

Unfortunately I'm not able to post EC charts, we would have to pay a hefty licence fee to do so.

  • Like 3
  • Thanks 2
Link to comment
Share on other sites

Posted
  • Location: Longton, Stoke-on-Trent.
  • Location: Longton, Stoke-on-Trent.
19 minutes ago, Nick L said:

Usual rant from me, these "snow depth charts" are in no way indicative of what the EC is actually showing. The algorithms in them are hyped up purely for clickbait, that's the only explanation I can think of.

Edit: Just noticed that foot and a half of snow near Dublin, yeah alright 

ECM shows anything wintry (Sleet/Very wet snow) accumulating when, in reality, much of it wouldn’t.

Link to comment
Share on other sites

Posted
  • Location: uxbridge middlesex(- also Bampton oxfordshire
  • Weather Preferences: blizzard conditions. ice days
  • Location: uxbridge middlesex(- also Bampton oxfordshire
Just now, weirpig said:

MMM  so by that the ensembles for the gfs  may be under estimating the strength of the Storm out of the US  which in turn may cause some of the perbs to not have substancial ridging?. 

A distinct poss..

Although lets await full 12z..

Then make ref...

  • Like 1
Link to comment
Share on other sites

Posted
  • Location: Hertfordshire
  • Weather Preferences: SNOW AND FREEZING TEMPS
  • Location: Hertfordshire

I know the GFS 6z run was the most wintry run ever and we can’t expect it to be repeated or even land but I remember last feb / March run after run kept getting better and better , everyone was saying how amazing it was as they just kept upgrading run after run . Does anyone think it could happen again this time ??? 

  • Like 4
Link to comment
Share on other sites

Posted
  • Location: Sunderland
  • Weather Preferences: Hot Summer, Snowy winter and thunderstorms all year round!
  • Location: Sunderland
15 minutes ago, ICE COLD said:

GFS P is not the most accurate model ?? And even if it was it is done globally and not just around are shores . 

ECM - 1st 

UKMO- 2nd 

the rest make up the numbers . The meto use ECM , UKMO and in house models and do you no why ? Because they are the best out there .

Sorry IC, but this simply isn't true....The me office use a blend of many models including GFS & JMA....and this is straight from a well known met office source as posted in this thread before 

  • Like 4
  • Thanks 1
Link to comment
Share on other sites

Posted
  • Location: Manchester Deansgate.
  • Weather Preferences: Heavy disruptive snowfall.
  • Location: Manchester Deansgate.
1 minute ago, beerandkebab said:

Sorry IC, but this simply isn't true....The me office use a blend of many models including GFS & JMA....and this is straight from a well known met office source as posted in this thread before 

Yes but the JMA isn't showing that mild westerly, the parallel is on its own, you say many models but how low do they go? the only op showing mild is the GEM apart from the gfs(p)

Link to comment
Share on other sites

Posted
  • Location: Penkridge
  • Weather Preferences: Virgins
  • Location: Penkridge

For what it's worth NOAA model discussion:

Model solutions show only minor differences during the initial 36 hours of the forecast period across the CONUS, which includes the deep layer cyclone offshore the West Coast that lifts gradually north off the West Coast through Thursday, and also the progressive shortwave and associated surface low that lifts from the southern Plains to the OH Valley and lower Great Lakes through Thursday night.

Thus a general model blend should suffice through about 36 hours. However, by around 12Z on Friday, more substantial differences in the model mass fields begin to emerge. In particular the 00Z GFS and 00Z CMC are seen as slow outliers with the next negatively tilted trough and intense surface low offshore of the Pacific Northwest which ultimately advances northeast into British Columbia on Saturday as a trailing cold front crosses the coastal ranges of western WA/OR. The 00Z NAM for its part appears too weak with its surface low evolution.

The 00Z ECMWF has trended a bit weaker from its very strong 12Z solution and has strong support now from the 12Z ECENS. The 00Z UKMET is also very close to ECMWF/ECENS mean camp. Meanwhile, farther east, the guidance continues to struggle with the details of the amplifying shortwave trough that ejects out across the central/southern Plains on Friday and then advances east across the OH/TN Valleys and the southeast U.S. Saturday into Sunday. The 00Z NAM and 00Z GFS are on the faster side of the guidance as the energy advances east of the lower/middle MS Valley, and the 00Z GFS overall is a bit north of all of the guidance with the low track and is also seen as being a bit faster than the 00Z GEFS mean.

The 00Z UKMET/CMC/ECMWF solutions are conversely a bit slower but are noted to have trended a bit more progressive with this latest cycle. Overall, the UKMET remnains the slowest solution. The 12Z ECENS mean generally favors the slower consensus, and is quite close to the deterministic ECMWF. There remains a fair amount of ensemble member spread with the low track when looking into the ensemble plots of the GEFS/Canadian guidance, and to some extent the ECENS suite which suggests a limited degree of confidence in the details of the low track timing and latitudinal placement as it moves into the eastern U.S. this weekend, but there has been at least a modest convergence of the guidance with the 00Z cycle.

Based on the latest available trends and guidance, a solution toward the 12Z ECENS mean will be preferred across the Western U.S. and offshore the West Coast after 36 hours, with a solution toward a compromise of the 00Z ECMWF/12Z ECENS mean and 00Z GEFS mean across the Eastern U.S. after 36 hours. The incorporation of the GEFS mean will be out of respect for the more progressive trends at least seen out of the NAM/GFS solution over the last 24 hours.

  • Like 5
  • Thanks 4
Link to comment
Share on other sites

Posted
  • Location: Sunderland
  • Weather Preferences: Hot Summer, Snowy winter and thunderstorms all year round!
  • Location: Sunderland
1 minute ago, feb1991blizzard said:

Yes but the JMA isn't showing that mild westerly, the parallel is on its own, you say many models but how low do they go? the only op showing mild is the GEM apart from the gfs(p)

not relevant to my post Feb......I was simply putting the record straight as explained to me (and others) by a met office source 

  • Thanks 1
Link to comment
Share on other sites

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

ECM shows anything wintry (Sleet/Very wet snow) accumulating when, in reality, much of it wouldn’t.

It's not the model. Snow depth charts from the EC have to be derived from the data it does run. So any snow depth charts are built in house using their own algorithm. We have our own too. The problem is, theirs is rubbish.

  • Like 6
Link to comment
Share on other sites

Posted
  • Location: Hertfordshire
  • Weather Preferences: SNOW AND FREEZING TEMPS
  • Location: Hertfordshire
6 minutes ago, beerandkebab said:

Sorry IC, but this simply isn't true....The me office use a blend of many models including GFS & JMA....and this is straight from a well known met office source as posted in this thread before 

I know they take them into consideration and use them but they primarily use ECM , UKMO and in house glosea/mogreps . Ian F said it himself they take gfs into consideration along with JMA , that says it all really . . That’s not to say gfs/p hasn’t  got this right , it might well have . Time will tell . 

  • Like 2
Link to comment
Share on other sites

Posted
  • Location: Hertfordshire
  • Weather Preferences: SNOW AND FREEZING TEMPS
  • Location: Hertfordshire

Sorry just to add to the above I should of made it clearer in my first post . Sorry  

Also are the met even considering using the gfs/p at moment as it’s a new one ? 

Edited by ICE COLD
  • Like 1
Link to comment
Share on other sites

Posted
  • Location: St Albans, 95m asl
  • Location: St Albans, 95m asl
30 minutes ago, Nick L said:

That's not too farfetched, but a broken clock is right twice a day. But take for example southern Scotland. The ENTIRE EC RUN only has 10-15mm of precipitation. Yet on that chart it's got 15-20 INCHES of snow on the ground! Even assuming all that 10 days of precip falls as snow AND settles AND lasts the 10 days, it's still a tall ask.

Unfortunately I'm not able to post EC charts, we would have to pay a hefty licence fee to do so.

image.thumb.png.7a37bf1b41b050943b8d72e32f3d42d5.png

Not particularly high resolution, but just to back up what Nick has said there - the scale there is in mm, and running the sequence this looks like total accumulated (with no taking into account of any potential melting etc - as you can see from the snowfall being shown over the sea!) but just to give you an idea

EDIT: I should add, that's rainfall equivalent which has fallen as snow - so you can make your own mind up for mm:cm ratio for snow - depending on the airmass in the UK, anything from 1:2 to 1:10 is possible

Edited by snowking
  • Like 4
Link to comment
Share on other sites

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