Model output discussion - January cold spell ending, what next?

[SMU]

1 minute ago, That ECM said:

Still no one has answered my question  

what was that

Edited January 7, 2021 by Steve Murr

[Griff]

9 minutes ago, Battleground Snow said:

Icon 120 v 126. We might squeeze something out in the medium term, but it's the, 192+ plus range which could be the main course.

 

You've inspired me to look at the 120 time frame (way too tired to adjust any timings so approx comparisons) 

Edit: I've come to the conclusion I'm too tired (grrr homeschooling) to make sensible conclusions 

Edited January 7, 2021 by Griff

[damianslaw]

6 minutes ago, Man Without Beard said:

What a dramatic chart for next Wednesday from the ECM op! Shows the height of the 0C line. Snow in the blues, rain in the greens/yellows, huge temperature difference between the two. 

 

Yes this is what I was referring too, cold air wanting to dig down into NE quarter of the UK whilst atlantic wants to move in from the SW, major clash of air masses at any frontal boundary. We used to see many such set ups in years gone by, but they have been largely absent in recent years, these are the set ups than deliver trifle snow amounts to some places with fine margins whether you are south or north of the front.

[Don]

1 hour ago, sebastiaan1973 said:

Lots of northern blocking. Latest EC46.

Looking better than Monday's update.

[That ECM]

8 minutes ago, Steve Murr said:

what was that

On the ECM the heights to the ne of the low at t120 ish seem to be increasing, would this help build heights to the the nw? The fact you enquired makes me feel loved. Seriously though I’m genuinely interested as I think it does. Please feel free to tell me I’m talking rubbish. Just trying to increase my knowledge. TIA. 

Edited January 7, 2021 by That ECM

[IDO]

45 minutes ago, snowking said:

Just as a quick note here - IF, and it’s a huge if, we are going to see continuing upgrades in the short-medium term, then the ensemble suites and means are going to be next to useless beyond T+144 due to their lower resolution - if we are indeed headed to shorter term upgrades then the higher resolution Operational and Control runs will lead the way and drag the ensembles behind them.

I did do some research on ensembles -v- operational on the gefs, until NCEP stopped the data in mid 2019. It suggested that after d1 the ensembles would gradually verify better than the op, and after d5 there would be an increasing rate of better verification. So I am very much a believer in the "wisdom of crowds". Of course, I did not dig deeper as to in what circumstances the op would override that signal. 

I have never been convinced that the op would more likely be better in a chaotic system, as the whole point of ensembles is to change input data to minimise the chaos via different starting data? Probably one for the expert statisticians, though I suspect the answer is, "it depends"?

[phil nw.]

It's certainly frustrating seeing the deep cold pool building to our ne but as yet modeling is keeping it just out of reach.

The vortex displacement unfortunately has not helped as yet.The ssw has stretched the pv across our side of the pole with a chunk being situated over NA so we have enough upstream energy coming east over the top of the hgh.

These at day 8 at mid and lowest level of the Strat.

 

Jet pattern day 5

In the short/medium term out to day 10 say it will be a battle to see any Arctic cold come this far south giving the overall flatness of the upstream flow.The Azores high in the end fails to ridge north and so leaves us out of reach of the cold which seems so close at times.

 

Those in the N.Isles and maybe the far north of the mainland could well have a taste of battleground conditions in the coming week,it would only need a little resistance from a wedge of heights to push wintery conditions a little further south.

Beyond the less cold period though all bets are off.With another Strat warming being modeled,continued weak zonal winds and  +ve Polar ht anomalies we are still in a reasonable place.

 

[feb1991blizzard]

12 minutes ago, That ECM said:

On the ECM the heights to the ne of the low at t120 ish seem to be increasing, would this help build heights to the the nw? The fact you enquired makes me feel loved. Seriously though I’m genuinely interested as I think it does. Please feel free to tell me I’m talking rubbish. Just trying to increase my knowledge. TIA. 

Better heights on both the GFS and Icon current runs, improving on their 12z runs, whether those pesky shortwaves will derail things though i don't know.

[ALL ABOARD]

Almost back in a cold northerly by +96. 

I'd take any output past 4 days with a massive pinch of salt. 

4 days mind... Not often I say that. 

 

[Chertseystreamer79]

17 minutes ago, Don said:

Looking better than Monday's update.

How often does it update can I ask? 

 

Also...saw several posts saying it's a bit squiff with its output atm....is that fair to say. Certainly a complete flip a week or so ago. 

 

Tia

[DanDorsetUK]

Bartlett Highs video he posted is why we need to be looking towards greenland for height rises. And explains exactly why it takes time for the models to pick up on any SSW. Like i said before the models will not run with any SSW factors untill it sees the downwelling. I think many may be forgetting the time lag of 10 to 14 days for any downwelling. IF the downwelling begins and we hope it does there will be a dramatic change in the output once the downwelling is detected. I just dont think we are there yet considering it's only just happened and with further warming to come.

[chionomaniac]

5 minutes ago, IDO said:

I did do some research on ensembles -v- operational on the gefs, until NCEP stopped the data in mid 2019. It suggested that after d1 the ensembles would gradually verify better than the op, and after d5 there would be an increasing rate of better verification. So I am very much a believer in the "wisdom of crowds". Of course, I did not dig deeper as to in what circumstances the op would override that signal. 

I have never been convinced that the op would more likely be better in a chaotic system, as the whole point of ensembles is to change input data to minimise the chaos via different starting data? Probably one for the expert statisticians, though I suspect the answer is, "it depends"?

As ever it will depend on individual synoptic circumstances. We have seem ensembles throw themselves over a cliff like lemmings after following a different path to the operational and vice versa. Luckily we are in the position of both knowing the nuances of certain outputs, and can look at many different models and ensembles. Human input and experience though does have a merit too when taking all these variations into consideration.

[Don]

2 minutes ago, Chertseystreamer79 said:

How often does it update can I ask? 

 

Also...saw several posts saying it's a bit squiff with its output atm....is that fair to say. Certainly a complete flip a week or so ago. 

 

Tia

I believe it updates on Monday and Thursday evenings.  Not as good as last week's updates.

Edited January 7, 2021 by Don

[Tim Bland]

12z Vs 18z.. that’s a fair shift south of the cold on 1 run...at this rate we will building snowmen on Bournemouth beach! 

[TSNWK]

Just now, Don said:

I believe it updates on Monday's and Thursday evenings.  Not as good as last week's updates.

But better than Mondays

[Don]

Just now, TSNWK said:

But better than Mondays

Yes .  Monday's update was dreadful for cold lovers and if believed, this week would have probably been it for cold!

[IDO]

19 minutes ago, That ECM said:

On the ECM the heights to the ne of the low at t120 ish seem to be increasing, would this help build heights to the the nw? The fact you enquired makes me feel loved. Seriously though I’m genuinely interested as I think it does. Please feel free to tell me I’m talking rubbish. Just trying to increase my knowledge. TIA. 

It depends!

It is a wedge of heights so it could meander and warm out lower heights close by, but simply as a wedge, a painstaking process, but its main attribute would be to direct the cold.

If it can get pumped up by some ridge, that is when it can build blocks, or alternatively bump into the Arctic high and collaborate...

[ICE COLD]

Well it’s still cold next Tuesday

[Tim Bland]

1 minute ago, ICE COLD said:

Well it’s still cold next Tuesday

Will be interesting to see if the low coming east later in the week could slide. Probably won’t in this run. but the way things are changing you never know...

[chionomaniac]

Funny enough, I never believed that the low north of the UK would remain stationary or slowly move east. It just didn’t look right, so good to see it dropping somewhat south now.

[Chertseystreamer79]

3 minutes ago, Don said:

Yes .  Monday's update was dreadful for cold lovers and if believed, this week would have probably been it for cold!

Cheers...seems all weather output is to be taken with a deal of bull$**t until we get this ssw  sorted!!! 

[Jason M]

15 minutes ago, IDO said:

I did do some research on ensembles -v- operational on the gefs, until NCEP stopped the data in mid 2019. It suggested that after d1 the ensembles would gradually verify better than the op, and after d5 there would be an increasing rate of better verification. So I am very much a believer in the "wisdom of crowds". Of course, I did not dig deeper as to in what circumstances the op would override that signal. 

I have never been convinced that the op would more likely be better in a chaotic system, as the whole point of ensembles is to change input data to minimise the chaos via different starting data? Probably one for the expert statisticians, though I suspect the answer is, "it depends"?

Logically, the opp and control should verify better than the GEFS where they diverge from the GEFS early in the run. So, where the opp is picking up something the GEFS are not at say day 4 and its fundamental to the longer term outcome. If the divergence is early, I always go with the opp. 

[Griff]

Bit of a squeeze 

[bluearmy]

No surprise to see the 18z less amplified than the 12z.  Like the 6z it’s currently a bit over progressive days 4 to 7. 

[IDO]

1 minute ago, Jason M said:

Logically, the opp and control should verify better than the GEFS where they diverge from the GEFS early in the run. So, where the opp is picking up something the GEFS are not at say day 4 and its fundamental to the longer term outcome. If the divergence is early, I always go with the opp. 

Yes, makes total sense, just musing in the hope someone has seen a paper on it or has some other empirical proof?