fergieweather

NB: whilst UKMO recognise a greater chance v average of blocked conditions later March (due to SSW/MJO), they're not yet convinced re latest EC Monthly prognosis: suspicion is that EC's later output may be biased by lagged response to previous (and now less well defined) modelling of strong mid-Atlantic ridging further away in medium range. Signals are becoming more mixed by mid-month in latest other ensembles and thus weak mobility may yet come into play, but conversely it could all go down the blocked route...at least periodically. A tricky period, as UKMO note! Fascinating times.

SFC flow charts remain from a N'ly quarter throughout, but with inevitably weakening signal by later March. Bear in mind it's not offering the day to day variance.

Remains through phase 8 to d5; to 1 by d10 and round to 3 then 4 days 15-20.

Tonight's EC Monthly retains the colder than average, effectively northerly-dominated regime all the way through March. So, continuity with the last run.

Heavy wintry showers likely. Extent of any settling at lower elevations will be very variable.

Euro4 has a known systemic issue with modelling WBFL too low. Treat PPN phase output with due caution. 12z EC for Friday generates snow Midlands/Wales/Cotswolds (ie close to 12z UKMO-GM idea), but also extending ESE across M4 towards Thames Valley etc. But given that the critical nuance element of this evolution (emerging from US) for Friday won't happen & be assimilated until Weds, expect continuing run to run variation in how models handle things until mid-week. Hopefully we will see better coherence after Weds.

Not really, i.e. not the operational ones. The longer range suites are/have been assimilating it, but with outcome for UK still uncertain. The SSW is expected to be potentially major (in UKMO assessment at least) but not the final seasonal warming. The net possible result is exhibited in e.g. EC's colder look throughout March and also how GloSea5 has a noticeably blocked, colder start to April, too. Tonight's EC Monthly will offer a more informed steer, given where we now stand on the SSW modelling (versus where we were on that last Thurs). Either way, outcome is not expected to be severe/exceptional cold as seen in the March 2013 set-up and longevity of colder, blocked weather still remains uncertain further through March: likely becoming more periodic in nature, with periods of mobility allowing an erratic trend upwards in temperatures back towards average, but the month still being below average overall (and perhaps ditto early April).

The major issues (for those who are keen on snow) forecast-wise are multi-fold this week, hence treating current snow prognosis charts with due caution. Aside from inevitable shower distribution uncertainty (eg Weds), Fri is complex given varied model treatment of shape of the low; exact positioning; and tendency for occlusion debris circulating in flow to periodically raise WBFL, making exact delineation of snow v mixed phase/rain very difficult. Latest UKMO-GM WBFL/PPN phase (below, obviously NOT to be taken literally!) exemplifies how such margins might be fairly narrowly defined on Fri, but a growing threat certainly of snow on northern flank of the system, even to low levels. In short: a mare.

No, to be clear, I said strong ENS support for the *broadscale set-up*. That's quite a different matter to "strong support for snow". What I wrote was that if the 12z yesterday verified 'as is', it had snow in aforementioned areas. The 00z has a very different areal distribution. And inevitably, that sort of signal will continue to change run to run. Whilst the sub-528DAM, deeply unstable airmass of eg Weds and again later week will quite evidently threaten some areas with heavy, thundery showers and potential snow even to lower elevation in places, the nuance regional detail will remain elusive until short lead time.

No strong support for outcome a) in 12z ECMWF clustering. All 51 ENS agree on the broad deterministic outcome by next weekend (with risk of snow across Wales/much of W Country, *if* it were to verify 'as is'). The dichotomy by 10th onwards is some growing split between members maintaining +ve MSLP to W or building such up to NW, versus an increasing number doing so across Scandinavia. Either way, there's presently no compelling evidence for Atlantic mobility being re-asserted out to mid-month.

Latest GloSea output is consistent and still favours +ve MSLP rises in Arctic later March-early April. So, *overall* below average temperatures look likely across March. Somewhat nearer term, uncertainty days 10-13 re how far Atlantic ridge eases east to offer more settled flavour for a while, esp to NW. The general theme as March progresses is for colder than average conditions to be quite marked initially, but these to become more masked by wider diurnal swings through the month, with a gradual, erratic trend of temps creeping gradually back towards normal. With periods of renewed W/NW mobility running into colder air, snow is possible at times (northern uplands in particular). However, as many have repeatedly indicated, the expected situation is not comparable to the very extreme conditions that characterised March 2013...not that it's prevented some newspapers from much hyperbole re upcoming easter.

The notion of troughing to E & heights to W was repeatedly signalled for March from as far back as Dec in GloSea5. What's noteworthy is how the model remains steadfast in maintaining northern blocking during early April too, with a conspicuous theme of below average temperatures continuing. Hopefully the subsequent signs of temperature recovery thereafter (incl in EC Seasonal Model, which has interesting dry anomalies then later spring/early summer) will become reality. Of course next winter we'll be doing this cold-or-not gig all again, but perhaps against a La Nina backdrop and easterly QBO...!

The largest cluster from 00z EC gives a fair representation of expected outcome towards mid-March. From around then into late March, northern blocking becomes extensive in longer range suites (conspicuously so in yesterday's EC Monthly) and similarly reaches on into (at least early) April in GloSea5, maintaining the colder than normal theme. Note the risk of this outcome was previously suggested in the Feb issue of UKMO 3-monthly probabilistic assessment, before the EC Monthly support emerged.

Effectively all of March in latest EC Monthly is signalled as chilly under northerly-sourced flow, with pronounced +ve MSLP anomalies growing widely to N (esp NW) after mid-month. All in all, it looks like we will see disappointingly depressed temperatures for the start of Spring, with protracted risk of frost.

The ridge looks quite persistent in longer range products. I note Joe B, Weather Company-WSI et al getting on this trend for W Europe via twitter (so what could possibly go wrong?) ;-)

Precisely the dichotomy on offer! So, examination of EPS products offers a rather clearer vision into early March. Strong agreement on the SFC/upper high bolted in mid-Atlantic; N-NW'ly PFJ on its forward side and consequently repeated trough disruptions across western Europe and Mediterranean. Result is periods of colder Pm or occasionally Am outbreaks for the UK. Signals remain strong for this sort of rinse-repeat pattern to establish from later next week, with some EPS members offering threat of deeper cyclonicity at times (severe gales/rain/snow). Currently, GloSea5 and EC Monthly (update due tonight) retain the colder than average *overall* signal well into March but with all signals gradually weakening with time and thus becoming of little use. As I've stressed before, 'colder than average' at this time of year should not be interpreted as big freeze/snow narnia.

Nuances aside, we already know that ECMWF is the top-verifying model and all forecasters (not least in US) wearily await the replacement for GFS planned by NCEP. GFS has been a very competent and useful model in shorter medium range (to T+144) but not a tool used preferentially beyond that. The plans by NCEP for its replacement look very interesting. What's noteworthy here is how the new EC parallel model has moved the bar a notch higher. As Ryan notes in his tweet, it's exciting times and we look forward to the new 9km version becoming operational soon. But of course, there will still be those 'bad days'!

The bottom (9km) is the parallel version due to replace current 16km model. It has been verifying very well. I recall it's due to become operational in April, but will check.

Spot different directions again between 00z EC Operational and EC-P.... (Reading shown)

Well, take the two 12z clusters here (to illustrate where EC goes directly beyond the T+240 DET frames cited by others earlier on the thread)... a general rinse-repeat of forecast shenanigans this weekend. And more could follow.

The snow issue is peripheral as far as I'm concerned. The key point of interest here is the colder look into early March: anyone seeing 'mildness' stamped over output is misreading it. I think one issue here is perhaps folk looking solely at EC DET runs, which need proper examination alongside the ensembles. Note for example the 12z DET for Reading... then compare with the EC parallel (9km) run, which is verifying highest of any model. Note the EC-P deterministic running virtually through ENS median (contra operational 16k model). Clear signal for turning chillier again later next week. As for snow, well, I'll let others stress about that...!

Indeed. As do EC clusters out to 10th March. The expected pattern into early March heralds prospect of repeated bouts of Am incursions, much akin to the current one, with risk of deeper cyclonicity passing to S/SE of UK at times.

Colder than average out to 15d... in a nutshell.

? They all seem complete for 00z and 12z runs and are there in their entirety (just checked). How odd...

The potential spoiler (if you want snow) is how the WBFL rises >300m from W with the ppn. Tricky one this.