In which case, how do the models contrive such a thing? One of the outcomes is likely to happen (at least in the majority of its relevant characteristics).

It would be interesting to see a % analysis of which of the perts. came closest over a period of time- and have them ranked. Are they already in some sort of probability line, does anybody know?

Midlands northwards, arguably. Marginal down here, with colder uppers nudged out of the way before the precipitation intensifies. Nice lines though. Q would have that undercutting, too (easterly component straight edge rule)

Thats just such a ridiculous statement the jet hasnt changed ...where is your proof ??? 

Amazing chart.  That would almost certainly bring a full on blizzard to much of IRE too, which is why it almost certainly won't happen.

Good, I'm glad this has rolled out:

http://www.theweatheroutlook.com/twodata/chart.aspx?chart=/charts/gefs/gefsens850london0.png?cb=190

Clustering even for London is polarised around the -5C 850 line; the ENS at least is restored to its prettiness from a day or so ago.

Notice how the warmer options mid-way are thinly-spread. The mean says a lot.

It is rather a long way ahead CP.

That said, the 850s for the 372 have -8 across almost all of the Isle, but no real precipitation until the very end of the run at 384 when the uppers have upped.

a

Can you educate me if you excuse my ignorance? What's with the snow row nd the number? I doubt it's Inches or percentage? - Well percentage is more believable 7% chance of snow ..! haha! But what is it?

We wont know how ECM has verified within this mornings charts for 10 days but I think its the best for verification and the reason why I use ECM over GFS for my forecasting. UKMO is also constantly 2nd best over GFS in verifying and never far behind ECM. Discount ECM at your peril but here is how its performed against the other models over the last 10 days with this pattern mirrored pretty much throughout December.

http://www.emc.ncep.noaa.gov/gmb/STATS_vsdb/allmodel/daily/cor/cor_day1_PMSL_MSL_G2NHX.png

http://www.emc.ncep.noaa.gov/gmb/STATS_vsdb/allmodel/daily/cor/cor_day3_PMSL_MSL_G2NHX.png

http://www.emc.ncep.noaa.gov/gmb/STATS_vsdb/allmodel/daily/cor/cor_day5_PMSL_MSL_G2NHX.png

http://www.emc.ncep.noaa.gov/gmb/STATS_vsdb/allmodel/daily/cor/cor_day8_PMSL_MSL_G2NHX.png

http://www.emc.ncep.noaa.gov/gmb/STATS_vsdb/allmodel/daily/cor/cor_day10_PMSL_MSL_G2NHX.png

It's the number of runs (out of 20, or rather out of 22 if the op and control are included) that show snow at that given time.

So the highest percentage on that chart is late on Jan 8 which has 7/22 or just over 30% of runs showing snow.

Thanks for those stats. One thing the last link does is highlights the folly of putting any weight in individual data in isolation at 240hrs. Useful for trends rather than in isolation (and I realise that you do that already).

I still think the ECM 00z 240hr chart is decent though.

The jet has changed, it has got weaker, more meandering and also further north, on average. So yes you could argue that we are more likely to be on the south side of the jet, on average, but also with a less powerful and more meandering jet we should see more cold plunges, which does indeed seem to be the case around the world, with the Middle East, Japan, USA all seeing lots of unusual cold quite far south. For whatever reason we haven't been so lucky here. 

As for ECM15 "painting a grim picture", I would say it paints a very uncertain picture, with extremely wide scatter. By January 9 some runs have ice days, some runs have 15C maxes. 

I see Rob's answered your question tallyho, but you might also find it interesting to click through the locations listed below when you have the ENS chart up, and watching the snow row change; always interesting to see Rome's chances, and some of the northern European cities too.

For those who think that there is a simple relationship between cold conditions in Canada correlating to stormy conditions in the UK. It aint that simple.

For those who are interested here is a paper describing and analysing the many factors affecting development and maintenance of patterns throughout the polar night and within the polar night jet.

Interesting area for further research, especially with respect to Ebdon tropical stratospheric persistence affecting the pacific variable jet at its interaction with the North Atlantic polar jet through western ducting!

http://journals.ametsoc.org/doi/full/10.1175/2011JCLI4203.1

Or maybe it's all cobblers and we just need to get on and enjoy the weather.

For the record, I enjoy cold and snowy weather too :-)

Regards

Steve

It's entirely possible, indeed probable, that none of the perturbations will show exactly the synoptic pattern of a specific time and date in the future.

The essence of the computer models is that (a) they start with an incomplete picture of the atmosphere at T+0 and (b) however good the modelling, it is still only a mathematical representation of how the atmosphere behaves and not - and cannot be - 100% accurate.

If you think how many data points there are to populate the starting position you can understand that there can never be enough pieces of actual data (readings, observations) to fill them all.  The gaps are filled with estimates and that introduces the first area of uncertainty.

From these two factors you can understand the need for running the model with slight variations on the opening data. My understanding is that they vary some of the estimated data and focus more on areas that are perceived to be most sensitive to how the atmosphere behaves.

From this I think it's safe to say that there cannot be a ranking of each perturbation.

I think it's worth emphasising that even the operational includes estimated data in the opening position. Whilst it's the forecasters' best judgement it does mean that the Op may be little better than any of the perturbations.

Also worth saying that the ensemble suite cannot always pick up a change that isn't discernible from the starting data. There are endless permutations and the ensembles only test a small proportion. Even if they tested against may more how would you know which was the correct outcome?  So, I think it's fair to say that even the ensembles don't really offer much beyond a week or so upon which you would want to place any reliance.

Many thanks

Thank you for your answer Peter; which suggests to me that although there is no guarantee that the outcome at any time frame will necessarily have been modelled even closely by any of the Ps, there is no reason to expect that any P outcome is less likely, from and at a given starting point in the synoptics, than another when we are talking about, say 300 odd hours ahead (recognising that this changes as time passes and each successive outcome realises).

That's a decent look , no one should be moaning about the current output, look at the drop from 30th / 31st to the 1st . there is a slight rise from the 2nd to approx. the 6th , which could easily flatten out - then a cold agreement through to 10th , after that some quite cold options also. What rubbish were we looking at last year ?

but that's Aberdeen - for us down here in the extreme south and southwest we have more in common weather wise with bordeaux then Aberdeen.

Yes indeed but its still a cold set on this small tiny island of ours , I take your point for the extreme south and west though

I fully concur Marcus; even on the south coast that drop is prominent and the cold clustering is encouraging. Nice to have a cold mix for the whole nation. Here's Brighton:

http://www.theweatheroutlook.com/twodata/chart.aspx?chart=/charts/gefs/gefsens850brighton0.png?cb=341

Ensembles point to a couple of topplers, but nothing sustained by any means. We may have to wait until later in January for snow not confined the far north & eastern coastal fringes.
Lots to cherry pick from in the output but generally temps slightly below average will be welcomed after a largely mild & dull December.

One thing is for certain it's going to feel very, very chilly after the New Year...

Some strong gusts on the coasts as well to go with the frequent snow showers. Quite a shock to the system 

As the ensembles suggest, it is likely to remain cold thereafter. Just how cold is anyone's guess but the run is in the ride while we enjoy the initial northerly. Anyone suggesting anything between snowmageddon and a 'return to zonality' is speculating and nothing more 

Well it's good to see some cold at last' I'm currently in Disney land Paris have been since Monday. Every night I've been checking the models back home, it's been pretty good here with temps below and good frost. So nice to come from clod to cold.