Model output discussion - proper cold spell inbound?

[Methuselah]

Let's be honest? GFS's Cackmageddon is still EIGHT DAYS out! I mean, we haven't seen the BOM yet!

[khodds]

4 minutes ago, andymusic said:

gfs likely to be a "model wobble" - just looked into me crystal balls lol

it'll be back on track with more snowiness on the next run - but this run ain't so bad anyway

Yeah. Thank god it’s the pub run. Phew ?

[Snowmadsam]

8 minutes ago, andymusic said:

gfs likely to be a "model wobble" - just looked into me crystal balls lol

it'll be back on track with more snowiness on the next run - but this run ain't so bad anyway

TMI ???

GFS brings in a brief mild sector before cold north westerlies come back. SSW still signalled

[Daniel*]

8 minutes ago, london-snow said:

Just a observation people are moaning about the gfs and how it has shown a consistent trend for the past 2 that’s 2 runs to water down this potential incoming cold spell.

 

Here is a point..

 

Has the ukmo & ecm over the past 12 runs shown this watered down version??

Nope..

Some need to cast their minds a little back when GFS was absolutely not interested in this cold spell at all the UKMO was resolute throughout, even often ECM champion was second to jump. The evening UKMO is good and I must say that is good enough for me, I won’t be losing sleep over one GFS op run. If UKMO does back off I’ll notice until then steady as she goes, let’s see what EC brings..

Edited February 2, 2018 by Daniel*

[supernova]

PS. At least as far as GFS is concerned, FI likely begins at about 96hrs judging by scatter becoming apparent on short ensembles. Can already see that a few members go colder still. Full ENS should be interesting.

[Sky Full]

I am quite prepared to be corrected but as I recall it, the previous two notable cold events this winter were suddenly downgraded by the models at around +96 and all the doommongers were rubbing their hands, saying: 'I told you so', or other gloating phrases.  However, between +48 and +72 the cold conditions were suddenly reinstated, if not upgraded, and everyone else was happy again!  Now this could happen again so be patient, and wait for the really reliable charts to come out.  The models are not creating the weather, they are just trying to predict what will happen, and the general disagreement regarding the details tells me that none of them has it nailed yet.  The various meteorological factors which are coming together to bring us cold conditions are rarely seen, certainly not all at the same time, and this may yet turn out to be an event we will still be talking about in years to come.   Or not.  And that's the fun of model watching, isn't it?  So dont get too hung up on every single run.   

[Dancerwithwings]

Blimey!!!  If this all I have to endear every time I pop into the model thread.....I thought I'd say this "roll on bloody summer"

[Nick F]

38 minutes ago, Catacol said:

Not that I'm in the least bit bothered: I have little regard for NWP op runs on their own and prefer to look at the drivers that are looking positive for cold. Will it snow or not at 144hrs+ is pointless, wasted energy especially when based on the GFS which observationally is less accurate in our small pocked of the world than both UKMO and ECM.

Certainly GFS skill scores for geopotential height have taken a knock over last 5 days or so, EC on top, UKMO in between. Remember GFS was slow to pick up on the blocking developing this weekend into early next week, UKMO the first to pick it up. GFS always seems to be the first to breakdown a cold spell too it seems, so will need to see if it's being too progressive again.

To experience some proper cold down here, i.e. day max temperatures struggling to get much above freezing and some snow swirling around and hopefully settling for a few days but not too disruptive will satisfy me. If lasts all week, which is a good possibility, then even better. 

[Singularity]

The balance of continental versus Atlantic sourced cold air is crucial here with the former better for trough disruption prospects going forward.

UKMO retains just that little bit more continental flow than GFS days 4-6 thanks to LP over Europe being slightly more influential. The ECM 00z had the Euro trough much more influential but that might have been a flight of fancy - we'll soon know. Getting to that requires a small LP to spin up on the southern reaches of Tuesday's frontal boundary as it interacts with the relatively warm Channel sea surface temperatures. That is how a lot of Channel Lows get going so not an unreasonable idea.

The GFS 12z making the Azores High stronger than the preceding few runs had it is surprising given what is theoretically more likely given the AAM changes of late (as summarised excellently by Catacol earlier today). Perhaps this particular run didn't propagate the signal as strongly for some reason but more likely it's a feedback somewhere from a localised tighter thermal gradient or something like that - be that rightly or wrongly the case. 

It's true that we're not in line to score maximum points with the wintry weather potential next week; there is a LP that moves up through western Scandinavia early in the week and takes the deepest and most extensive cold pool with it. Had that stayed put with a ridge over the top... we'd likely have instead found ourselves headed toward one of those more extreme ensemble solutions with something like -16*C 850s knocking on our doorstep. 

Even so, an extensive area of air with 850s of -6 to -10*C or so that becomes near or totally stagnated for a time midweek is a great recipe for developing dense cold surface pooling that works more strongly against the attempts by Atlantic troughs to push frontal systems across. This makes me sceptical of GFS (well, more so than usual) so I'm not too worried about what it shows at this point in time.

[offerman]

3 minutes ago, Daniel* said:

Some need to cast their minds a little back when GFS was absolutely not interested in this cold spell at all the UKMO was resolute throughout, even often ECM champion was second to jump. The evening UKMO is good and I must say that is good enough for me, I won’t be losing sleep over one GFS op run. If UKMO does back off I’ll notice until then steady as she goes, let’s see what EC brings..

Hi Daniel any anyone.

It seems consensous is that the UKMO charts are far more reliable than GFS. What is the point of GFS?

Could some also point me in the direction where i can view MO charts please as i think i`m going to focus more on those too.

Is it a case of MO are better charts for snow and cold predictions and GFS for summer patterns?

Really appreciate if someone could explain to a below novice.

[AWD]

Worth noting, the 12z Arpege brings some quite heavy snow streamers over Essex, Kent way on Monday;

Then Tuesdays front only really brings any notable snowfall to Scotland and NI, dying a death across northern England and the Midlands;

Some cold upper air across England though;

 I

 

[supernova]

4 minutes ago, offerman said:

Hi Daniel any anyone.

It seems consensous is that the UKMO charts are far more reliable than GFS. What is the point of GFS?

Could some also point me in the direction where i can view MO charts please as i think i`m going to focus more on those too.

Is it a case of MO are better charts for snow and cold predictions and GFS for summer patterns?

Really appreciate if someone could explain to a below novice.

All charts easily accessible on here http://www.meteociel.fr/ (GFS & UKMO amongst others). 

And welcome.

Oh and GFS exists purely so everyone on this forum can take the Tut out of it.  TBF it's supposed to be America's best effort at weather prediction but like Trump it mostly lies.

[johnholmes]

I don't normally enter this upgrade downgrade pantomime but just for fun I have had a careful look at the 06 and 12 z outputs. To me there is nothing I would call significantly different in the overall pattern until T+168 h. Tell me how often any model, we have 2 we can see regularly GFS and ECMWF,, that are identical after 120 hours let alone 168 hours?

Okay if you want to have the roller coaster  every 6 hours but try to put some objectivity into the comments please.

Re will it snow or not. Think back to how accurate any model is, or indeed professional forecasters presentations are about when/where/how much rain will fall in the summer, even just 24 at times 12 hours out.believe me predicting rainfall compared to snowfall is a doddle. Wait for 12 hours out, watch the radar, you are far less likely to have a mental breakdown that way.

It is going to get colder and experience tells us that we ALWAYS  try to shift the block of cold air too quickly. It used to be just the same prior to models arriving. Just a time difference, then all we could do was forecast 24, at a pich 48 hours, ahead. Now the models give outlooks days ahead.

Think about how hard a solid lump is to shift with warm water, on a huge scale that is what the atmosphere is trying to do to a cold mass of air. 

sorry to natter on but it is model related.

Edited February 2, 2018 by johnholmes

[karlos1983]

06z v 12z short ensembles (London)

[andymusic]

1 minute ago, johnholmes said:

I don't normally enter this upgrade downgrade pantomime but just for fun I have had a careful look at the 096 and 12 z outputs. To me there is nothing I would call significantly different in the overall pattern until T+168 h. Tell me how often any model, we have 2 we can see regularly GFS and ECMWF, that are identical after 120 hours let alone 168 hours?

Okay if you want to have the roller coaster die every 6 hours but try to put some objectivity into the comments please.

Re will it snow or not. Think back to how accurate any model is, or indeed professional forecasters presentations are about when/where/how much rain will fall in the summer, even just 24 at times 12 hours out. Nelieve me predicting rainfall compared to snowfall is a doddle. Wait for 12 hours out, watch the radar, you are far less likely to have a mental breakdown that way.

It is going to get colder and experience tells us that we ALWAYS  try to shift the block of cold air too quickly. It used to be just the same prior to models arriving. Just a time difference, then all we could do was forecast 24, at a pich 48 hours, ahead. Now the models give outlooks days ahead.

Think about how hard a solid lump is to shift with warm water, on a huge scale that is what the atmosphere is trying to do to a cold mass of air. 

sorry to natter on but it is model related.

absolutely - gfs is constantly caught napping and having to adjust for the freezing uppers because it just doesn't deal with that set up very well - which seems a bit odd considering someone really should put some further "humanistic" programming into the model to help it deal with this phenomenum

[Chesil View]

Just now, johnholmes said:

I don't normally enter this upgrade downgrade pantomime but just for fun I have had a careful look at the 096 and 12 z outputs. To me there is nothing I would call significantly different in the overall pattern until T+168 h. Tell me how often any model, we have 2 we can see regularly GFS and ECMWF, that are identical after 120 hours let alone 168 hours?

Okay if you want to have the roller coaster die every 6 hours but try to put some objectivity into the comments please.

Re will it snow or not. Think back to how accurate any model is, or indeed professional forecasters presentations are about when/where/how much rain will fall in the summer, even just 24 at times 12 hours out. Nelieve me predicting rainfall compared to snowfall is a doddle. Wait for 12 hours out, watch the radar, you are far less likely to have a mental breakdown that way.

It is going to get colder and experience tells us that we ALWAYS  try to shift the block of cold air too quickly. It used to be just the same prior to models arriving. Just a time difference, then all we could do was forecast 24, at a pich 48 hours, ahead. Now the models give outlooks days ahead.

Think about how hard a solid lump is to shift with warm water, on a huge scale that is what the atmosphere is trying to do to a cold mass of air. 

sorry to natter on but it is model related.

Thanks John great post.

Anyone trying to predict amounts and spatial spread of snow for the middle of next week from this range is on a hidingto nothing.

[Summer Sun]

1 minute ago, karlos1983 said:

06z v 12z short ensembles (London)

12z may have gone off on one at the end of the short ens lots getting in a tizz over 1 run which may turn out to be a rouge one

[karlos1983]

7 minutes ago, Summer Sun said:

12z may have gone off on one at the end of the short ens lots getting in a tizz over 1 run which may turn out to be a rouge one

Yes certainly looks that way. Get better clarity on the full set soon. 

And here it is. Does go a bit rogue, not totally unsupported though. That’s not a moan either, just the way it is on this particular run. Lots of cold options still, so nothing lost.

Edited February 2, 2018 by karlos1983

[tight isobar]

5 minutes ago, Summer Sun said:

12z may have gone off on one at the end of the short ens lots getting in a tizz over 1 run which may turn out to be a rouge one

[Summer Sun]

2 minutes ago, karlos1983 said:

Yes certainly looks that way. Get better clarity on the full set soon. 

And here it is. Does go a bit rogue, not totally unsupported though.

 

18 day's at least of the 850 line staying below zero, yes temps will fluctuate at times, but I don't see any very mild weather for the foreseeable future anyone disappointed with tonights GFS run have their expectations too high I've seen far worse in Feb than what is on offer tonight

[frosty ground]

GFS op is not without support but its not in the majority either.

[Daniel*]

33 minutes ago, offerman said:

Hi Daniel any anyone.

It seems consensous is that the UKMO charts are far more reliable than GFS. What is the point of GFS?

Could some also point me in the direction where i can view MO charts please as i think i`m going to focus more on those too.

Is it a case of MO are better charts for snow and cold predictions and GFS for summer patterns?

Really appreciate if someone could explain to a below novice.

Hi offerman, 

the thing is UKMO well publicly only goes out to T+144 whereas GFS goes out to T+384 that’s twice as long as what the UKMO goes out to, therefore, it’s subject to even more ‘chaos’ and experimentation. UKMO is a higher res model I believe the starting data is probably superior to that of GFS, I still think GFS is a good model. There have actually been many cold spells when it has picked it by GFS sometimes even in extreme FI think it did that in 2010 albeit that is a rarity. Of course I want to see it on side with UKMO however one would rather the UKMO on side than GFS. Of late UKMO has been very good. 

The GFS is going to have a huge upgrade planned for 2019 fitting current EC specs.

Edited February 2, 2018 by Daniel*

[Mucka]

Can't help but feel GFS is massively progressive in FI.

Temperatures for NW England are 5 to 7C above the mean by the evening of the 10th and already 2C above by the 9th

Looking through the ensemble suite, that also looks progressive with a breakdown to me but already signs of high pressure rebuilding to the NE and any Atlantic incursion being forced SE.

It is FI and it could be right but I would expect GFS to slowly push back the breakdown to the 11th and perhaps further.

It is very easy to see where the MetO forecast of a settle mid month is coming from and any Atlantic advance West shouldn't be that long lived on current form.

Keeping an eye on De Bilt ensembles from around 12th onward may be quite revealing as it will show weather the Atlantic gets in proper or the cold form the East is holding off and then later fighting back.

Still a little to early to forecast snowfall in the more immediate but Tuesday is looking good still.

[Rayth]

Just signed back up in preparation of tonight’s monumental , pivotal , crucial , edge of your seat ECMWF 12z output , tongue in cheek there,  but I’m loving the Synoptics and various debate going on today ! Shows there’s interest on the horizon and not a westerly juggernaut ! 

[Ice Day]

1 hour ago, Man With Beard said:

I think this would lead to carnage next Friday - very active splitting front inbound, still cold ahead (and behind)

Is there even a route for an Atlantic height rise on the back of it?

In my best Homer Simpson voice, 'Mmmmmmm, carnage'! 

UKMO looks a really good run all the way, and as has already been said, GFS is only really going awry outside of the reliable. We're pretty much nailed on for a 4-5 day very cold spell, so please enjoy it and stop stressing over whether it will snow in your back yard, it will take years off your life!

 So strap in for the ECM and enjoy the ride.

Just 96 hours until this 

Edited February 2, 2018 by Ice Day