Jump to content
Sign in to follow this  
Captain Shortwave

Model Output Discussion - The final stretch to Christmas

Recommended Posts

Good evening, I don't post often do but follow this thread with interest and admire all who are brave enough to make predictions at the moment. This is indeed a volatile period of model watching and I for one have no idea how it is going to pan out, nor do I think most people have, although there have been some very good posts regarding the larger / longer term perspective rather than what each individual model run shows.

 

Just a curious question. Why not look back to what a model was showing last week or 4 days ago and see if it has verified this week or 4 days hence? It just seems logical to me to try to see what was predicted (fair enough there will be various different scenarios) and what actually occurred in order to gain some idea of how accurate those various predictions were. I know there are verification stats available for each model but I rarely see anyone posting / confirming what has happened against what was predicted.

 

I think when you have so many different models, run at numerous times, it can become extremely confusing for people trying to learn when totally contrasting posts are made regarding the outcome based on maybe just one particular run. I for one would find it easier if a more balanced perspective were taken using a consensus of all information.

 

Shoot me if I'm talking b*****ks, but I am just trying to educate myself on what is overall a fantastic forum.

 

Valid point Cap jack

 

As alluded to by RJBW above the GFS P seems as though it has been the most accurate assuming the low comes off on saturday close to the UK. RJBW has counted as many as 16 runs out of 20 for the GFSP which is brilliant (if it comes off).

 

Also of interest was the fact that a few of those 16 actually went for an Easterly afterwards..We will see.

 

The latest 3 runs of GFSP have moved the bombing low more easterly, so it appears to have lost the signal at short range. If it is correct now then it means it will have failed on 16 out of 20!!!!

 

There again ALL models are obviously been struggling to define some new trend which wants to set up in the medium term.. Hence their variability.    

 

The ECM however only showed it on 2 or 3 charts of the low in the North sea or UK  at the same time..

 

Interestingly nearly all the models have switched away from this bomb around the North Sea (though still some showing impacts), but in the last 24 hours the Euros (MO and ECM) have turned on to the idea of the low coming straight down the UK.

 

Don't know what this  means in terms of stats, but it must  be seen as a feather  in the cap of the new GFSP, with its better resolution post 180hrs if it has picked it and stayed with it.

 

MIA

Share this post


Link to post
Share on other sites

Thank you UKMO you've really made us Kentish folk have something to cheer about looks like we would fair best from that chart on Sunday! I just hope it comes true:-) all I've heard on here for weeks is Scotland and northern England forecasted possible snow/wintry showers etc at range- with no hope for the heavily populated south-so to see a chart like that at 144 hours is atleast a possible solution:-)

(Having said this it will probably be dropped by the morning!)

Edited by Kentspur

Share this post


Link to post
Share on other sites

Valid point Cap jack

 

As alluded to by RJBW above the GFS P seems as though it has been the most accurate assuming the low comes off on saturday close to the UK. RJBW has counted as many as 16 runs out of 20 for the GFSP which is brilliant (if it comes off).

 

Also of interest was the fact that a few of those 16 actually went for an Easterly afterwards..We will see.

 

The latest 3 runs of GFSP have moved the bombing low more easterly, so it appears to have lost the signal at short range. If it is correct now then it means it will have failed on 16 out of 20!!!!

 

There again ALL models are obviously been struggling to define some new trend which wants to set up in the medium term.. Hence their variability.    

 

The ECM however only showed it on 2 or 3 charts of the low in the North sea or UK  at the same time..

 

Interestingly nearly all the models have switched away from this bomb around the North Sea (though still some showing impacts), but in the last 24 hours the Euros (MO and ECM) have turned on to the idea of the low coming straight down the UK.

 

Don't know what this  means in terms of stats, but it must  be seen as a feather  in the cap of the new GFSP, with its better resolution post 180hrs if it has picked it and stayed with it.

 

MIA

Thank you for your reply, I missed that post from RJBW, must pay more attention! 

Share this post


Link to post
Share on other sites

Welcome, the models tend to trade off spells of doing relatively well but this coming period is a more complex challenge than they often face, so how they've done in past week or so would be generally inconclusive. Also this zonal pattern has been easier for them and errors were much smaller than what we're likely to find looking back from the 27th. Experienced model watchers develop ideas about what models do in certain situations and what their strong or weak points might be. GFS used to be a bit notorious for over-deepening lows at 5-7 days then backing off towards actual time. GEM is often thought to be too prone to over-amplification. European is regarded as most likely to verify when there's a conflict, but not by a wide margin. This coming situation is low on analogues and high on uncertainty even today. A good guess might be made by considering which way each model has trended past three days and seeing if those trends point to a convergence point. The best forecast at any given point is probably the accumulation of least error after weighting the models by preference. At the moment, you could assume that a moderately deep low will cross northern Scotland around the 27th and head towards southern Norway or Denmark by the 28th but possibly factor in that it might head more south than southeast. What actually verifies could be different from what any model shows today. There have been situations in the past where no guidance was all that close at day 6, but this is rare nowadays.

Thank you for your reply Roger. I respect your opinion and advice and so basically at the moment due to the current chaotic predictions the past cannot predict the future!

Share this post


Link to post
Share on other sites

I live on the North East coast .would I be right in saying that there is also a spring tide due at the same time as this deep low on the 27th onwards.if so I know people who have just got there property repaired after last December's storm surge. Sorry if I've posted in the wrong thread wasn't sure where to ask with me been new on here . Regards Steve from Bridlington.

Share this post


Link to post
Share on other sites

Thank you UKMO you've really made us Kentish folk have something to cheer about looks like we would fair best from that chart on Sunday! I just hope it comes true:-) all I've heard on here for weeks is Scotland and northern England forecasted possible snow/wintry showers etc at range- with no hope for the heavily populated south-so to see a chart like that at 144 hours is atleast a possible solution:-)

(Having said this it will probably be dropped by the morning!)

Yes it would be nice for UKMO to come off (from IMBY perspective)but so much uncertainty atm with the models I am not holding my breath. One good thing that the models have shown in last couple of days is that it is getting proper cold in Central/Eastern Europe in next few days and I think as I alluded to last night that the UK will benefit as a whole from early January and hopefully tap into the cold air. Although not a consistent feature, some models are showing Easterly components in not too distant feature. Be good if that becomes a trend but don't think we will know Medium Term future until this weekend.Interesting video by Gavin P today which will give coldies some heart for January. :cold:  :good:

Share this post


Link to post
Share on other sites

Thank you UKMO you've really made us Kentish folk have something to cheer about looks like we would fair best from that chart on Sunday! I just hope it comes true:-) all I've heard on here for weeks is Scotland and northern England forecasted possible snow/wintry showers etc at range- with no hope for the heavily populated south-so to see a chart like that at 144 hours is atleast a possible solution:-)

(Having said this it will probably be dropped by the morning!)

Yes we've been digging ourselves out for the last week due to the no snow events we've seen this winter as have many other Northerners. Its good to see  some of our Southern contingent  having something to cheer about if it verifies that is, but least we forget many others from all other areas  have seen no wintry conditions for the best part of 18 months.

Edited by Hocus Pocus

Share this post


Link to post
Share on other sites

The ECM operational run is towards the coldest cluster of solutions:

 

http://www.weerplaza.nl/15daagseverwachting/?r=midden&type=eps_pluim

 

The ECM postage stamps still show a large split in placement of the low at T120hrs:

 

post-1206-0-92953700-1419282389_thumb.gi

 

Amongst those you'll see the best will have the low digging se, the ones you don't want to verify are something like member 17, member 4 looks interesting.

 

I still think we need one more run to be sure of the split, you'll see member 17 keeps a train of energy running sw/ne to the nw.

 

Unbelievably after looking at those ensembles I think FI begins after T96hrs!

 

 

 

 

Share this post


Link to post
Share on other sites

The ECM operational run is towards the coldest cluster of solutions:

 

http://www.weerplaza.nl/15daagseverwachting/?r=midden&type=eps_pluim

 

The ECM postage stamps still show a large split in placement of the low at T120hrs:

 

attachicon.gif120.gif

 

Amongst those you'll see the best will have the low digging se, the ones you don't want to verify are something like member 17, member 4 looks interesting.

 

I still think we need one more run to be sure of the split, you'll see member 17 keeps a train of energy running sw/ne to the nw.

 

Unbelievably after looking at those ensembles I think FI begins after T96hrs!

96hrs. Crazy situation. Never known the models to be struggling like they are. We are no clearer into what happens with the low come end of the week/weekend. Hopefully we can manage to scrape some back edge snowfall for many from it, just to help those who are getting withdrawal symptoms from the phenomenon snow!

Share this post


Link to post
Share on other sites

96hrs. Crazy situation. Never known the models to be struggling like they are. We are no clearer into what happens with the low come end of the week/weekend. Hopefully we can manage to scrape some back edge snowfall for many from it, just to help those who are getting withdrawal symptoms from the phenomenon snow!

In normal circumstances the higher resolution of the operational run you'd normally factor in much more at T120hrs and would expect the ensembles to start moving across however given events of the last few days I won't feel comfortable until we see a few more outputs.

 

You can see from those ECM postage stamps that some are ugly for cold, some much better and a few very nice!

Share this post


Link to post
Share on other sites

Unsurprisingly,not much confidence from NOAA in their discussions  for the 6-10 and 8-14 day outlook.

 

FORECAST CONFIDENCE FOR THE 6-10 DAY PERIOD: BELOW AVERAGE, 2 OUT OF 5, DUE TO
POOR MODEL CONTINUITY AND LARGE DIFFERENCES AMONG THE ENSEMBLE MEANS ACROSS THE
EASTERN CONUS.

 

Full discussion here.

 

 http://www.cpc.ncep.noaa.gov/products/predictions/610day/fxus06.html

Share this post


Link to post
Share on other sites

In normal circumstances the higher resolution of the operational run you'd normally factor in much more at T120hrs and would expect the ensembles to start moving across however given events of the last few days I won't feel comfortable until we see a few more outputs.

You can see from those ECM postage stamps that some are ugly for cold, some much better and a few very nice!

Sums it up really a mixture of solutions, no strong signal. Its the saga that keeps rolling

on!

The NOAA discussion sums it up that cloud has posted above.

Edited by Mark N

Share this post


Link to post
Share on other sites

Reliable timeframe remains Boxing Day. Roughly around 50% of all model runs going mild thereafter, 50% colder. Met Office going for the colder solution with heights building over a cold pool, despite uppers being quite mild, surface temps cold for the time of year.

 

I suspect this will be a case of nowcasting and it will be Boxing Day before anyone can confidently call events for the end of the month.

 

A good time to take a break from the models - they will be prone to high margins of error over next 2-3 days and no doubt continue to frustratingly confuse.

Share this post


Link to post
Share on other sites

The level of uncertainty for 120 hrs shows up nicely on the ECM ensemble spreads when comparing with 96 hrs.

 

96..  120..

 

 

Anyway,the pub run is churning out now,so no doubt more drama.

Edited by Cloud 10

Share this post


Link to post
Share on other sites

It was tongue in cheek :-) albeit I wouldn't discount it!

Ironically it would probably feel quite cold at ground level even though it's mild aloft.

Valid , and a fair comment

 

Cheers

Share this post


Link to post
Share on other sites

The level of uncertainty for 120 hrs shows up nicely on the ECM ensemble spreads when comparing with 96 hrs.

 

96..attachicon.gifEEM1-96.gif  120..attachicon.gifEEM1-120.gif

 

Wow! Says it all in very graphic terms!! If ever there was little argument for FI begins at D4, this is surely it.

 

'bombing' lows as per the one forcast for then are always a completely nightmare to nail. Often it comes down to nowcasting as the exact track can be up for grabs just hours before, let alone 5 days. I wonder what are the possibilities of the pattern being pulled west 300 miles or so? Aided by increased upstream amplification has to be at least an outside chance...Now that could be a game changer!

Share this post


Link to post
Share on other sites

Ahhh.... yes. The 850's aren't a good guide (always use with caution anyway in this sort of winter set-up). EC 12z offers anywhere from 2-10cm snow from Cotswolds down across into central-southern England / northern Home Counties. MOGREPS, meanwhile, focuses further to E/SE and across NW as colder 'dig' focuses around W/NW flank. But none of this trustworthy, of course, given noticeably broad EPS spread at this range on how the evolution takes shape. Variance in track/depth/phasing quite stark in latest postage stamps, some (minority) even taking the low centre to exit from SW England!  Winds, however, a wider issue, whether in north, south or east. Still a way to go in resolving this one and history teaches us that potential explosive cyclogenesis may not resolve confidently (re track, depth and regionalised impacts) until short lead time. I'm not expecting this one to be much different (as model flux at present amply shows), so no point getting exercised about it for a while yet!

enigmatic as ever ian!

if you were asked to present that summary on tv, how would you put it?

Share this post


Link to post
Share on other sites

Well better heights on the gfs op 18z with a negatively tilted shortwave that may slide under the block at t126

Share this post


Link to post
Share on other sites

 

 

Just a curious question. Why not look back to what a model was showing last week or 4 days ago and see if it has verified this week or 4 days hence? It just seems logical to me to try to see what was predicted (fair enough there will be various different scenarios) and what actually occurred in order to gain some idea of how accurate those various predictions were. I know there are verification stats available for each model but I rarely see anyone posting / confirming what has happened against what was predicted.

 

I think when you have so many different models, run at numerous times, it can become extremely confusing for people trying to learn when totally contrasting posts are made regarding the outcome based on maybe just one particular run. I for one would find it easier if a more balanced perspective were taken using a consensus of all information.

 

Shoot me if I'm talking b*****ks, but I am just trying to educate myself on what is overall a fantastic forum.

not talking that at all Jack-a very good idea, it can sometimes give a btter handle on the latest synoptic outputs

Edited by johnholmes

Share this post


Link to post
Share on other sites

Missed the update on this but only forgot to record the 18z gefs spread from last night

Anyway, newest

post-6981-0-16586600-1419286269_thumb.jp

post-6981-0-44201500-1419286258_thumb.jp

post-6981-0-80137100-1419286248_thumb.jp

post-6981-0-26192700-1419286240_thumb.jp

post-6981-0-07826100-1419286230_thumb.jp

post-6981-0-65332000-1419286220_thumb.jp

Spreads upstream not as large in depth but still pretty evident where the issues are which impacts the weekend low and upstream ridge

Edited by bluearmy

Share this post


Link to post
Share on other sites
Guest
This topic is now closed to further replies.
Sign in to follow this  

×
×
  • Create New...