fergieweather

We're especially interested to see what cohesion in broader signal appears 2nd half Dec in EC Monthly at 10pm, versus latest GloSea5.

12z ECMWF ENS set against model climatology lean pretty firmly in median 2m temps towards below avg (at times quite markedly) out to at least 9 Dec.

...albeit the E'ly set-up later Dec showing in both latest GloSea and ECMWF has inevitably low confidence given lead time, so (especially given weak signals prior to then into mid-month), these outcomes signposted later should be treated with due caution, at present.

...except GloSea5 has been signalling it even longer :-). The new EC Monthly is especially interesting, because we know pulling -ve temp anomalies (across S'rn UK in this case) out of a hat in longer range models by week 4 (it finishes 25th Dec) is quite uncommon. But it fits our current broad expectations mid-end Dec.

Given that EC's 00z ENS median T2 for central-southern England sits anywhere from 0 to barely +6C from 28/11 to 4/12, there's really not much disagreement on broad bias towards somewhat below avg temperatures in either that model or GFS (EC showing a few markedly colder members too). However, spread in EPS fields for MSLP remains very marked by early Dec, hence caution required re just how we sit re extremes of temperature/ppn/wind into that period. I know it's stating the obvious, but at that range the deterministic runs are not to be read in isolation (for any model) without examination of their respective ensembles/postage stamps/clusters.

Because whilst we remain in contract with UKMO, our obligation remains to reflect the very latest thoughts and briefing we have from Exeter (ie UKMO Seasonal Team/Medium Range desk)

Marked ENS spread again in ECMWF post-29th, in keeping with 00z suite. Magical mystery tour....

Correct. GloSea5 has repeatedly signalled this outcome. ECMWF prefers (weakly) +ve GPH/MSLP build to NW-N, but postage stamps show varied support towards GloSea5. Given the latter's performance strat-wise (=why Exeter were unconvinced re GEFS SSW debacle a few days ago), there's tentative support for it's evolution but we shall see where EC Monthly heads in update tomorrow.

Reason UKMO won't talk about shortwave features at T+168 is because that sort of nuance is utterly unreliable at that range. We have enough trouble accurately forecasting shortwave troughs at T+12, let alone at a range where we're much more focused on ensemble forecasting, not small-scale vagaries of deterministic output run-to-run.

But it's not really zonality in sense of 2015: the majority signals for early Dec look cyclonic, yes (deeply so in various members), but the pattern in NOT very progressive and very distinct from this time last year. Glosea continues to resurrect blocked set-ups mid-Dec on for rest of month, with generally below average temperatures; it favours high pressure to NE whereas the (broadly not dissimilar) ECMWF shows similar evolution albeit HP weakly signalled to NW. The general look of things is very adequately summarised thus by UKMO this morning: "The main theme (from this evolution) is that December could well be very different from last year, with -ve anomalies for both ppn and temperature for much of the country".

There's marked ensemble spread in 500hPa GPH plumes post-29th Nov, with no *firm* evidence - yet - for pronounced cold weather in the 12z extended clusters (= just as many less cold members).

To skimp on the frame-by-frame detail and varied complexities anticipated into early Dec, tonight's EC Monthly ultimately heads in much the same direction as it's predecessor, with +ve GPH anomalies to north and +ve MSLP anomaly to our NW (Iceland) by mid-Dec (run ends 18th). Steady as she goes... etc.

Not wanting to sound like a stuck record.... but to reiterate, not *all* models have gone hammer and tongs re potential SSW. Where there is better agreement is a -ve NAO/AO signal into Dec, but with potential for an unsettled spell to start the month. Whilst longer range suites still favour broad idea of high pressure to NW, low to east, there still remains considerable fluidity on the exact high-low positions and equal uncertainty on resultant temperature and ppn outcomes. Anyway that's all a way off... whereas a sting jet is here and now.

It offers a broad swathe of possible tracks into Monday and this spread is likely one reason for the lack of strong signal re snow probabilities.

The issues Sun into Mon are especially tricky. UKV generated a plastering of snow even at lower elevations across W'rn parts of the W Country (incl adjacent Bristol Channel/M5 corridor up into Wales and SW Midlands) but considered overdone. However 00z EC gives quite a few cm in pretty much same areas, but that model automatically settles all solid phase ppn. MOGREPS currently disinterested. Jury out....

Disagree. If anything, the clustering for HP to NW and cyclonicity to our E with colder outcomes has strengthened in 00z EPS by early Dec, but not yet as a majority. The bimodal direction from about end of next week remains in ensembles.

Given that we have TMin ENS spread by morning of Dec 2nd ranging from +10C to -11C in the op (!!) & +10 to -6C in the new cycle, I think we can reasonably emphasise the uncertainty....! Indeed even early next week, there's a raft of issues to resolve including threat (yet tbc) of significant snow in the W Country and Wales >ca 200m. NB A more coherent drop in T850 now signalled in the new cycle by end of next week... much less pronounced in the operational

Yes, of course low confidence which is why we need caution on extrapolating the here-and-now, but the tipping point will presumably be whatever unfolds with the stratospheric PV in next (circa) fortnight, as that will feed into whether the stronger PV signals emerging post-Dec - as previously signalled - are rapidly then rendered null and void, or were actually on the money. Either way, that's all some time off being resolved and it's a fascinating watch....

What is emerging by around 27-28th is the fine balance between an anticyclonic bias across the UK (with varying degrees of continental influence versus a more W'ly one); or conversely a +ve MSLP to NW and troughing to E, with resultant directly colder N'ly influence. This sort of umm-ahh bimodality has characterised recent extended range and doubtless will for a while. Hence unclear temp signal for us further ahead.

Cyclonicity yes (with W'rly flow bias) in ensembles, but true mobility as per last year's choo-choo train, no. But as per a previous post I made, at the present time we need caution re assuming *all* winter will see a similar paucity of that signal.

...but a marked disjunction between the operational model versus new version, which nose-dives T850 det/CTRL into end of month. Net result, as before with both parallel models, remains a leaning towards below avg temps overall (albeit growing spread continues to afford low confidence on temperature by early Dec, which has been a recurrent theme). The favoured outcome, broadly, remains a more changeable phase early Dec, followed by signs of heights retrogressing to our NW (as keenly signalled towards/into mid-Dec in latest ECMWF monthly and prior to that period, in far reaches of EC 12z ENS) and blocked leanings then characterising that period. The issue remains exactly how/when/where respective block(s) manifest and resultant outcome for the UK. Despite some similar themes in GloSea and EC suites (including their seasonal output too), it's just too early to offer any confidence whatsoever, not least as stratospheric matters have yet to show *complete and unequivocal* outcome across the main models... quite yet.

Yup, don't understand the gloomier posts this evening. ECMWF 12z ENS continue to favour slightly -ve temps versus avg by day, albeit spread in overnight minima is (at least in part) reflective of cloud cover uncertainty. Either way, 850hPa trend has become more aligned to the chillier members rather than the milder ones; I don't see anything to change the broad prognosis of temperatures overall finishing the month a tad below normal, despite some slightly milder phases. It's hardly 2015 revisited....

Yes, all pretty average-looking fare in latest 15d ensembles. No signs of any exceptional mildness (as per 2015) nor pronounced cold (as per 2010). ENS median currently trends generally a touch below average out to 1 Dec.

Yes, there's no signal of anything replicating last December...!!

Longer range suites aren't deterministic runs! They merely give broad guidance on trends. Despite mixed signals, temperatures are still more likely to end-up below average overall, based on latest output. There's a long way to go before we can express any real confidence in the synoptics/circulation by later Nov and early Dec.