The hunt for cold - Model discussion (late November)

[weirpig]

All if and buts  But if the initial low  dosent blow up like the 6z runs portray it to do,  and is shallow and sinks on a  more southerly  one would think on the latest run that many could well be digging themselves out in a few weeks   a good number of lows after the initial dive on a southerly route just below the Uk   that first low is obviously the key to what we get further down the line 

[snowangel32]

5 hours ago, SLEETY said:

The ukmo never goes as far as the other models so how has it joined the party late exactly.

Its t144 output yesterday evening was not much different from the other models to begin with,amazing how people hang off one model run ,if it shows anything a tiny bit different 

Sleety i like your posts and you know way more than me but if you cant see the diffrences in the UKMO i dont know what to say to you. Except maybe a trip to specsavers lol. I wish i could post pics but phone wont let me for some reason but UKMO has clearly moved towards ECM. If not show me how it hasnt. Anyway have a good day sir and enjoy the weather watever it may be.

[feb1991blizzard]

10 minutes ago, Daniel Smith said:

The clusters are as clear as mud.. 4 clusters at just T144 though all broadly showing the same pattern

 

5 by 240 showing the support for the Atlantic as seen on the EPS means in cluster 1

 

Though all still showing a rather extensive blocking signal but no huge % number on any of them.. So overall, huge amounts of uncertainty, no point worrying about a single 6z run then.

The good thing about those clusters is there isn't any significant signal whatsoever for any strong heights to the south and any true zonality

[Methuselah]

8 minutes ago, That ECM said:

Hi Ed, not sure that they are  all implausible, the right one is probably in there. It’s impossible to know which one is the right one

I'm not so sure about that TE, TBH...I guess that, if any member were to be correct, it'd be the one fed the most accurate starting data; as, as far as my limited understanding of ensembles goes, they really only serve as a guide to how trustworthy the operationals are...?

And in the case of many current model-outputs that's not very. Not beyond 13/12 anyway?

[winterof79]

The winds they are a changing.

Looking forward to plenty graphs like that as we move into winter deep

[Tim Bland]

27 minutes ago, weirpig said:

All if and buts  But if the initial low  dosent blow up like the 6z runs portray it to do,  and is shallow and sinks on a  more southerly  one would think on the latest run that many could well be digging themselves out in a few weeks   a good number of lows after the initial dive on a southerly route just below the Uk   that first low is obviously the key to what we get further down the line 

Good post, hit the nail on the head. It’s all or nothing. If that low blows the block away we are left hunting. We will only get a couple of cold dry days and a few hrs of sleet before the rain and milder air arrives. I’d back the ECM over the GFS but as above IF that initial low is weaker and goes under the block it could get very snowy, high risk but high reward for snow lovers! 

[ICE COLD]

Gone a bit mad in hear this morning lol . Everyone chill out . We all know how the gfs can be very erratic , remember Ian f saying it a while back . The ECM is rock solid . Add to that the  very good METO update . Winter is coming . ECM the pick of the bunch this morning and the UKMO looks a lot better than yesterday 12z . 

Ps - where’s @Spanish Dani gone

Edited December 5, 2018 by ICE COLD

[feb1991blizzard]

One thing i would like to see with clusters is this, instead of having one member representative of each cluster, i would like to see the actual mean of each cluster, like an ensemble mean within the ensemble means if that makes sense.

[bluearmy]

3 minutes ago, feb1991blizzard said:

One thing i would like to see with clusters is this, instead of having one member representative of each cluster, i would like to see the actual mean of each cluster, like an ensemble mean within the ensemble means if that makes sense.

I guess they want to offer charts that may verify rather than a mean - with the mean you need to know the spreads to get a feel for where the more reliable part of the mean is. It’s complicated !

[Methuselah]

4 minutes ago, feb1991blizzard said:

One thing i would like to see with clusters is this, instead of having one member representative of each cluster, i would like to see the actual mean of each cluster, like an ensemble mean within the ensemble means if that makes sense.

I think I know what you might mean, feb...????

I would prefer it if the clusters were projected onto a graph/plane, similar to that used in those 'political compass' diagrams...That would, I think, make for a better representation of how individual clusters might relate to one another?

As things stand, I find that the ensembles make more intuitive sense...

[feb1991blizzard]

Just now, bluearmy said:

I guess they want to offer charts that may verify rather than a mean - with the mean you need to know the spreads to get a feel for where the more reliable part of the mean is. It’s complicated !

Yes true, do you have info about what representative member is selected? is it computer selected? (i guess thats a rhetorical question) and is it usually the one right in the middle of that cluster?

[stodge]

Afternoon all

I think what strikes me about the 06Z OP, Parallel and Control is how they are all very differently modelling the NH profile in FI.

OP tries to restore the PV in its usual place but both Parallel and Control displace the PV away from the Canadian side. As we know, a displaced vortex isn't a guarantee of cold but it's a good start.

I don't think anyone thought the colder outbreak was going to last a length of time and to be fair cold spells lasting more than 7-10 days are unusual.

Looking at the GEFS I'm struck again by the weakness of the PV in mid December - again, as you might expect from the synoptic evolution, the Parallel makes much more of a second assault on the PV in far FI - the OP is quite muted.

[feb1991blizzard]

1 minute ago, Ed Stone said:

I think I know what you might mean, feb...????

I would prefer it if the clusters were projected onto a graph/plane, similar to that used in those 'political compass' diagrams...That would, I think, make for a better representation of how individual clusters might relate to one another?

As things stand, I find that the ensembles make more intuitive sense...

Are you talking something similar to the spaghetti plots?

[Decemberof2010]

As seems to be the case at this time of year. A poorer model run is followed by an upgraded Meto update. (Very good btw). Then the runs upgrade and the meto downgrades. It’s Happened a few times so far 

[Methuselah]

Just now, feb1991blizzard said:

Are you talking something similar to the spaghetti plots?

I'm not sure...I guess I was visualising a kind of Venn diagram with coordinates...

[Frosty.]

Coldies can breathe a sigh of relief, according to the latest models, next week is set to become much colder with frosts, ice and wintry showers with snow in places..much better extended update from exeter too (at least away from western areas)...signs are good for netweather coldies!!

[markw2680]

This place cracks me up man, I mean why are the models going to be right this morning and not last night? It’s like if they show a slightly not as good run it’s gospel and it’s going to happen and then it’s complete meltdown when they show a stonking run. They are no more likely than the last one lol. It’s about middle ground and small changes that add up to be correct in the end not one bloody run!! 

Anything after about 120hrs is like complete hook a duck, nobody knows or any model it’s just ideas of a theme.

oh well we will get snowed in next week if he next gfs shows it

[markw2680]

2 minutes ago, Frosty. said:

Coldies can breathe a sigh of relief, according to the latest models, next week is set to become much colder with frosts, ice and wintry showers with snow in places..much better extended update from exeter too (at least away from western areas)...signs are good for netweather coldies!!

Exactly what I’m seeing too and that’s about all anybody can say at this range. Glad to see u back mate

[captaincroc]

13 minutes ago, Frosty. said:

Coldies can breathe a sigh of relief, according to the latest models, next week is set to become much colder with frosts, ice and wintry showers with snow in places..much better extended update from exeter too (at least away from western areas)...signs are good for netweather coldies!!

What does give me hope is that when the last cold spell was in the pipeline a few weeks ago they didn't budge from the 'Atlantic moving in' line, which gave us the impression that their computers didn't support it. Now it looks like they do have the support of their models which is only a good thing.

 

In regards to the GFS 6z. I remember a few years ago Steve M mentioned that it had an Atlantic Bias. Again, not saying it won't come to pass but it's something to be noted.

[snowangel32]

7 minutes ago, markw2680 said:

Exactly what I’m seeing too and that’s about all anybody can say at this range. Glad to see u back mate

Same here i just be glad for it to feel more seasonal. Snow would just be a bonus. Just one more storm to get through on Friday then we will hopefully see drier and colder days with chance of some wintry showers.

[Methuselah]

12 minutes ago, snowangel32 said:

Same here i just be glad for it to feel more seasonal. Snow would just be a bonus. Just one more storm to get through on Friday then we will hopefully see drier and colder days with chance of some wintry showers.

Aye Snowangel...Snow is on its way!

[Man With Beard]

46 minutes ago, feb1991blizzard said:

One thing i would like to see with clusters is this, instead of having one member representative of each cluster, i would like to see the actual mean of each cluster, like an ensemble mean within the ensemble means if that makes sense.

Feb, this is a calculated guess, but it is my opinion that the coloured anomalies probably show the mean of the cluster. This is because a. the actual 500mb line plot is too precise to be a mean, b. the anomalies do not precisely match the line plot, c. the anomalies look more general than you'd get off a single ensemble member and finally d. what is the point of the anomalies otherwise?

For this reason, I read the anomalies more than the lines these days. I've been taking this approach for a few months - it hasn't done me too badly. But I could be wrong. Anyone fluent in Icelandic out there?

[feb1991blizzard]

8 minutes ago, Man With Beard said:

Feb, this is a calculated guess, but it is my opinion that the coloured anomalies probably show the mean of the cluster. This is because a. the actual 500mb line plot is too precise to be a mean, b. the anomalies do not precisely match the line plot, c. the anomalies look more general than you'd get off a single ensemble member and finally d. what is the point of the anomalies otherwise?

For this reason, I read the anomalies more than the lines these days. I've been taking this approach for a few months - it hasn't done me too badly. But I could be wrong. Anyone fluent in Icelandic out there?

Cheers, i can sort of concur (although my initial reaction is to the contrary) because you get some serious low anomalies to the south but with actual +ve heights if that makes sense. 

Edited December 5, 2018 by feb1991blizzard

[Iceaxecrampon]

Ppn creeping in.....

[damianslaw]

Many a time I say this - but always worth looking at the Jetstream profile forecasts, for idea on where low pressure may decide to position itself. Today's run shows a southerly tracking fairly moderate zonal jet for all of next week, becoming a bit more diffuse and erratic -shredded is the word as we enter the following week (but that is a long way off). The GFS 6Z run doesn't tie in with the profile - as it suggest a SW-NE position, no strong heights to our north - contrary to what the jetstream forecast is showing.

The main player next week looks to be strong heights to our north east, the extent of height development directly to our north is key, there are signs a ridge could extend out from the main feature to our north and head towards Greenland, if this occurs, then expect a much weaker shallower low development feature which will attempt to encroach SW part of the country, but quickly sink into France - thus we will maintain a cold continental SE flow. If the ridge doesn't develop strongly, then the next plausible evolution would be heights to sink slightly SE into eastern europe which would open the door to a more concerted atlantic attack - but equally a set up dispositioned to trough disruption with heights building to our north and possibly NW. Either way, the theme of 2018 looks like continuing - no atlantic zonal onslaught, instead fine lines between something much colder and wintry, or perhaps akin to what we have now as we enter the run up to christmas. Longer term - further atlantic derailment and hitting the buffers looks very likely.