Model output discussion - Closing in on Christmas

[Kasim Awan]

Yes, amplify the high like icon and loose the space for the low waving and secondary northerly. It would be possible to get both the initial amplification / northerly and the secondary hit under a moderately amplified ridge, and I feel the UKMO is a good example of this..

[Vikos]

16 minutes ago, TSNWK said:

Is ICON the German based model that replaced DWD ? I think so and if it is I wonder if Vikas? From Frankfurt who has posted recently can share a view on its reputation?

 

Well, ICON is DWD made. Fun-Fact: the daily synoptic Outlook from DWD is always based on the EC-Output (EPS) with a bit MosMix ( https://www.dwd.de/DE/leistungen/met_verfahren_mosmix/met_verfahren_mosmix.html ), and ICON is used mostly for correctional purposes.

 

Edited December 18, 2020 by Vikos

[Mucka]

1 minute ago, Kasim Awan said:

Yes, amplify the high like icon and loose the space for the low waving and secondary northerly. It would be possible to get both the initial amplification / northerly and the secondary hit under a moderately amplified ridge, and I feel the UKMO is a good example of this..

Not impossible but unlikely.

The more amplified the initial ridge is and slower the low then the more delayed the phasing with the Sceuro trough becomes which would mean that low takes a more E trajectory then S.

This in turn prevents the W flank of the trough sharpening in our sector which aids the amplification upstream to give us the cold pattern.

You would much more likely see a flatter pattern develop later with UKMO and ICON than GFS and ECM as things stand but there are no guarantees ECM and GFS will go on to produce such a blocked and amplified pattern either, it is just much more likely.

[Updated_Weather]

21 minutes ago, Kasim Awan said:

Iconic Xmas day run from the ICON 

Would produce a white xmas for many

Sadly it is the icon..

Nothing wrong with the ICON!

[Updated_Weather]

Just now, Updated_Weather said:

Nothing wrong with ICON!

 

[Mucka]

5 minutes ago, Updated_Weather said:

Nothing wrong with the ICON!

Debatable, but what can be said with certainty is that Its last run has a 50% chance of being very wrong or very right. 

 

108h GFS and already big difference with that Atlantic low and therefore ridge (GFS sticking with its 00z output so far)

Edited December 18, 2020 by Mucka

[mushymanrob]

4 hours ago, MattStoke said:

I don't think COBRA meetings get called for wet snow over the hills.

Ridiculous exaggerations from posters in here, as usual. There's nothing significantly snowy or cold for the majority of lowland UK in that model run, and it's academic at this range anyway. Nice synoptically, but talk of snowstorms is such nonsense. People never learn and set themselves up for a fall when future model updates show less wintry potential. 

Really?

From the 27th until the 29th at least,
- there are hard frosts predicted and daytime temps struggling to get above zero
- 528 DAM line far south of the UK
- Low pressure domination
- 500hpa temps below -35c
-850hpa temps below -5c
-Dew points below 0c

Thats all on "that" model run you "dissed" , to suggest theres nothing "significantly snow or cold" is not how those figures look to me!

Now you are quite right of course, it is a long way off and may well turn out to be less wintry, but this is the model discussion forum so discussing what the current models show is surely appropriate.

However, all models are singing from a similar hymn sheet, plus the anomaly charts for that range support that general synoptic pattern. So broadly, id have thought that theres a strong chance of it turning cold and unsettled, there will be snow in the forecast.

[Kasim Awan]

4 minutes ago, Updated_Weather said:

Nothing wrong with the ICON!

Seems to always be somewhat isolated in it's output. That's why a gfs/ecm type evolution is somewhat favoured in here etc.

[Updated_Weather]

1 minute ago, Kasim Awan said:

Seems to always be somewhat isolated in it's output. That's why a gfs/ecm type evolution is somewhat favoured in here etc.

Whilst I agree GFS/ECM and UKMO may have better verifying stats but ICON input still can't be ignored!

[Kasim Awan]

Just now, Updated_Weather said:

Whilst I agree GFS/ECM and UKMO may have better verifying stats but ICON input still can't be ignored!

I never said I was ignoring the icon output.

As Mucka said, we are dealing with probabilities here, which are swayed away from the icon solution for reasons discussed .

[Mucka]

3 minutes ago, Updated_Weather said:

Whilst I agree GFS/ECM and UKMO may have better verifying stats but ICON input still can't be ignored!

It can if it is wrong by 96h as it appears to be here, but no, we cannot be certain it is wrong at the moment.

Edited December 18, 2020 by Mucka

[Vikos]

35 minutes ago, Kasim Awan said:

Iconic Xmas day run from the ICON 

Would produce a white xmas for many

Sadly it is the icon..

Get your facts straight here https://ims.gov.il/en/node/1555

Quote

One can see that the temperature forecasts all year long and the relative humidity in the summer and autumn are generally better in the COSMO model, whereas the relative humidity forecasts in the winter and spring are generally better in the ECMWF model. One can see that the best wind forecast is in the ECMWF model with a slight advantage over COSMO. In addition, the forecasts of the ICON model are mostly on the third place, whereas the forecasts of the GFS and UKMO models are less good for all three variables.

Also https://www.dwd.de/EN/research/weatherforecasting/num_modelling/05_verification/verification_node.html

 

Edited December 18, 2020 by Vikos

[Kasim Awan]

2 minutes ago, Vikos said:

Get your facts straight here https://ims.gov.il/en/node/1555

 

This study analyses performance to 72h and not 96-180h which is the time period I am discussing *potential* weaknesses in the icon.

 

 

[Mucka]

We are already seeing how finely balanced our favoured GFS 00z type outcome is at 138h

We have the faster low and flatter ridge but we need that secondary low just behind to be slower and become "trapped" to help upstream amplification. It could also race through which wouldn't be so bad but phasing with the first low around SW Greenland would be the worse outcome as that could extend the ridge E and prevent clean phasing with Sceuro trough as is happening this run

 

Still could go on to be great but it is an extra complication we could do without.

Hopefully it will go on to produce but the pattern will likely be a little further E as a consequence.

I won't mention this again as it is easy to get obsesses with these shortwave lows and phasing and timing but it is worth noting as a potential fly in the ointment.

Edited December 18, 2020 by Mucka

[Daniel*]

It’s really nothing of note but ridge a little sharper on 12z therefore colder.

[Daniel*]

This is only going one way and it is not mild..... fantastic consistency from GFS for once.

[Matt H]

I don’t think I’ve ever been more apprehensive watching a GFS run come out before...

[Jeremy Shockey]

UKMO 12z 24th Dec

A bitterly cold Christmas Eve if the UKMO 12z is correct 

[Mucka]

4 minutes ago, Daniel* said:

This is only going one way and it is not mild..... fantastic consistency from GFS for once.

Yes the upstream amplification is great and the shift E for reasons outlined shouldn't be a hindrance this run.

Fingers crossed ECM will be the same.

UKMO is a halfway house between ICON and GFS

If we can get to day 8 GFS then the cold pattern is pretty much locked in, just variations of a theme.

That's 8 days not 10 or 12. Squeeky bum time.

Edited December 18, 2020 by Mucka

[Vikos]

1 minute ago, Kasim Awan said:

This study analyses performance to 72h and not 96-180h which is the time period I am discussing *potential* weaknesses in the icon.

 

 

ICON was never designed for long-term forecasts, its power is in the 7 day range. It produces 900gb data for a 7-day-output

[Matt H]

The GFS can stay on the Christmas card list this year now!

[mountain shadow]

Here it comes... 2010 redux. Boxing day evening now the marker.

Edited December 18, 2020 by mountain shadow

[sheikhy]

Gfs saying it dont matter what you do shortwave at 144 hours!!i will insert myself into greenland if i want to ❄

[Vikos]

Oh boy oh boy no big changes in a 192h chart ... is GFS ok?

 

06z--->

Edited December 18, 2020 by Vikos

[Weathizard]

It’s simply stunning