fergieweather

See http://link.springer.com/article/10.1007/s00382-014-2155-z Huge uncertainty however, re scope of any teleconnection for UK winter 16-17 (models still yet to confidently resolve re ENSO neutral or, as per BOM-POAMA, a weak Modoki La Nina... thus influence on seasonal modelling will vary at this juncture). Whilst we see some current similarities emerging in latest ECMWF, GloSea5 and CFSv2 output for Dec-Feb, we can't yet be too clever on understanding exactly why they are returning the present signals. QBO situation also yet to be resolved with confidence, of course, albeit interesting to note ongoing tentative signs of a possible reversal/E'rly outbreak even as early as November. In short: far too early to accept any of the current seasonal output with any reasonable degree of confidence, albeit interesting signs of a potentially very different winter to the last one (perhaps unsurprisingly, given the very different base state we had this time last year!)

Not true. Rest assured I think all weather aficionados will be delighted when they see the final product and the massive improvements when visualising the weather 'story'. Bear in mind the scope of what we will be showing / mapping: ranging from global, right down to hyperlocal scale; with model products including ECMWF's full suite and UKMO-UKV to 'fuel' all you will see. Keep faith.

Yes, we've done a raft of work on that.

The latest EC Monthly calls for +ve MSLP anomalies building across Scandinavia post-easter period, followed by same into early-mid April in the Arctic. There's some GloSea5 support for this. UKMO thus consider most probable current outcome is for temperatures a bit below average late March and through early April, with likelihood of N-NE flow and continued risk of frosts for some time.

Just a single set of clusters further into medium range in 12z EC-EPS. The latest EC Monthly, meanwhile, retains largely weak or insignificant anomalies for temperatures into this latter part of March but then leans more towards -ve anomalies (esp to NW/W) by end of 1st week April, whilst hinting at +ve MSLP rises by then to NW/N.

Yes, sounds reasonable. There has been a distinction by weekend between models re potential cloud cover & TMax...this will remain the key day to day forecast challenge across the next few days. Take for example the comparative f'cast profiles for 12z Sunday, London, from UKMO-GM (Red), ECMWF (Black) & GFS (Blue). Note lower temperature and moister ascent (at lower levels) in UKMO model versus drier/milder EC & GFS. Hopefully the continental fetch by then will favour more cloud breaks, but still a rather chilly feel in the SE corner whilst springlike in any sunshine as you head NW'wards.

Indeed. You might recall how a few days ago I mentioned how GloSea5 wasn't wholly convincing on whether easterlies would propogate right down; whilst the reversal aloft has been clear-cut (eg see recent GloSea5 strat diagnostics below; I can't show latest version), the tropospheric effects remain less immediately tangible. Latest thinking is that whilst SSW hasn't incepted the HP story emerging by this weekend (= too short lead time), it does probably result in its apparent longevity. That is in itself still an imponderable, as (per previous posts), the ongoing signal for a return to lowering GPH and increasing cyclonicity post-21st *may* be too fast (models always tend to overcome blocking too readily)...but nonetheless, that outcome remains a clear and arguably growing possibility, given repeating ENS signs of such, including 12z EC-EPS.

Yes, tricky. The issue here is that various models generate very little snow, but UKMO were suspicious of that - considering (even their own models) to be "over optimistic", as they termed it, in not generating the risk sufficiently. So, they've run the very clever tools they have at Ops Centre that employ the Wilderspin and Lumb methodology to gauge potential of model rain signal changing to snow, given varied background parameters. This experimentation and tweaking showed just how sensitive the relationship is between intensity and PPN phase within the yellow (rain) warning area, and that's even allowing for the strength of the wind, which normally would create sufficient mixing in the boundary layer to minimise risk of snow getting down much below the WBFL.

No - it's actually exactly in line with UKMO expectations, which are, quoting the Chief Forecaster earlier this afternoon: "Allow for temporary snowfall – maybe even heavy wet snow/sleet at times – within the warning area, mainly in its southern half, with some possible (though low point prob) accumulations of 1-2cm at low-level. This should melt again quickly, especially as ppn readily turns back to rain as intensities ease slightly. Greater risk (and possible slightly higher accumulations, say 2-4cm) on higher ground along this track – Salisbury Plain, Mendips – incl Bristol Airport, Cotswolds and perhaps even the Chilterns. All of this low confidence and again the track of the Low is critical." Tricky, in short.

The later clusters (eg below) illustrate well why the subsequent period is very tricky. Yes, EC Monthly may offer some additional hints. *May*....!

Well, at least it's now just down to a 50-50 to contend with further into medium range in 12z EC at T+240....!! EDIT but more expansive set of cluster solutions thereafter out to T+360, with (yet again) many opting back to a cyclonic, unsettled/changeable regime by 21-22nd (refer to my post above re that)

Yes, the signs are evident between D7-9 as the easterlies propogate down to the troposphere. Initially this manifests as a SSW'ly anticyclonic flow in most ensembles, albeit the potential for this to turn into a colder set-up has been signposted by (as yet) a smaller number of members, but the risk is clearly there. The downward trend in GPH from around 21st signalled in recent runs continues to look potentially suspect (ie spells of blocking could show greater longevity, with any supposed unsettled return perhaps a red herring, as models lean back towards climatology). GloSea5 continues downward propagation of E'lies for a fair part of March, so the potential for colder weather remains latent throughout, albeit not any certainty. As noted a day or so back, MJO (below, latest) remains of little influence within the inner circle for a protracted period and thus offers no forecasting steer.

Particularly post-21st, no ensembles can yet be considered trustworthy/representative guides, for now. No model will yet adequately synthesise SSW impacts (but they will often lean more towards climatology and give a skewed impression). As for exactly what the SSW impacts may be... unknown. But a greater propensity for blocking, certainly (that does NOT necessarily = colder than average), and perhaps greater than ENS presently indicate.

The GFS prognosis of 60+mm is considered low probability. However, 15-30mm in the period Tues eve/night to early Weds has good ensemble support. Equally, there's a clear risk of gales or even severe gales on Weds. We expect the rain to become an issue tail-end Tues, rather than earlier in the day. Further ahead, little faith can be given beyond 20/21st in any model ensembles, given potential for increased blocking from SSW and expected lag response before this is correctly synthesised tropospherically. However, whether this eventually yields colder than average conditions remains in the balance, but is clearly possible. The MJO, meanwhile, will run into weaker territory across cooler waters of E Africa and thus not be of predictive value further ahead this month. So, the SSW outcome looks to be the sole decider for later March-early April, but the threat of any *significant* cold outcome still looks low in UKMO modelling....but with caveat that we can't yet be too clever re how it unfolds (as with any SSW).

Unanimous 12z EC cluster support for the settled anticyclonic story 13-15 March inclusive...at least. There's two clusters thereafter to end of run (20th), the largest of which continues the settled theme beyond 15th until it relents later (below). Interestingly and perhaps instructively, the increasingly cyclonic look evident by 20th on EC12z cluster 1 now tallies with longer range suites, which also go that direction w/c 21st (d15). Thereafter lies the conundrum: whether the UK sees a more blocked pattern eventually set-up that could go 50-50 (either blocked as an anticyclonic regime here, or as a cyclonic one). Whilst the long range products do agree on GPH falling after d15, as noted on Thurs the latest EC Monthly offers no useful signal thereafter re temperatures etc, with GloSea5 adding little additional steer. However, GloSea5 strat diagnostics do continue to show some degree of downward propagation of post-SSW easterlies late March-early April and as the UKMO note, should those reach the lower troposphere (and that's not *yet* a given), then the risk of more meandering jet and blocking/colder than average weather would increase into the turn of month. At present, the outcome later this month is too tricky to call, but at least the likelihood of a pleasant spell into mid-month continues to find growing support.

No, this is expected to be the final warming. Given latest GloSea and EC Monthly output, signals for anything cold are now looking much less bullish later in the month, with little discernable temperature anomaly in either suite (refer to my post last night re EC prognosis). However, expert view from UKMO Seasonal Team is that the SSW (and in part, MJO) continue to favour onset of a more blocked regime later in the month, albeit they stress net result of final warmings can be variable (as can timing of lagged response at SFC). The critical point however, as alluded to in posts yesterday, is that even if such a regime establishes it may *not* actually usher-in colder than average conditions, not least due to the time of year but equally due to uncertain broadscale positioning/flow. Thus, forecast emphasis has shifted to mention chance of drier, settled weather both mid and again later this month, but no strong signal currently for anything in the way of pronounced cold. That doesn't dismiss the possibility: it merely reflects the 'null signal' for temperature anomalies emerging in both of the main longer range products. Meanwhile, the milder phase into mid-month has been well signalled for a few days now and will be a pleasant, spring-like change from the current chillier weather.

Indeed so. Whereas if we look back at a snapshot of GloSea5 strat diagnostics (1st from late Dec; 2nd mid-Jan) we can see how clustering always favoured something late in season & progressively signposted it as late Feb or during March. Hence UKMO never bought into notion of a Jan SSW.

A quite different EC Monthly today, emphasising UKMO caution re depth/longevity of any cold emerging later March. Distinct +ve temp anomalies & settled look signposted through mid-March, as per 12z EC (hurrah: we hope) and albeit still then pointing towards downturn to -ve temp anomalies and tendency to blocking thereafter, the signals in all parameters are increasingly rather indistinct (especially versus Monday’s run) and noticeably so by early April. Given lead time, this is not entirely surprising. So, a chillier (if not genuinely cold) end to month still looks the most probable outcome, but we still can't expect clear model handling of key drivers (SSW/MJO) a fortnight ahead.

GloSea5 MJO forecast for comparison:

....but Cluster 1 from 00z continues to go blocked, following the story of last EC Monthly (update this eve) and current GloSea5 into later March...UKMO expect models to take a while getting to grips with SSW effect:

The point to stress here is the distinction between *seeing* a bit of (mostly hill) snow in the south versus what warrants a yellow warning. The latest output from EC/MOGREPS/UKMO-GM (& it's downstream products) have greatly reduced the risk of anything warnable, including ice risk (front decaying readily; ppn intensity needed for snow dropping; WBFL fairly high). Thus, last night UKMO advised warning could be cancelled and sure enough, it duly was as overnight and 06z modelling emerged.

Yes, overall: but much uncertainty as we get later into the month (after a phase of temperature recovery back up towards average). Whilst GloSea5 forecasts of both MJO and SSW response are indicative of a stronger chance (versus average climatology) of blocked/easterly conditions late March-early April, this does not - even if it happens - necessarily translate to *pronounced* or prolonged cold, given steadily increasing insolation by then. Way back in mid-Dec, I recall discussing the GloSea5 strat diagnostics with a senior UKMO forecaster. He noted that whilst all signs were pointing to late Feb-March SSW, it would, quote "....probably be too late to significantly offset an otherwise markedly mild season, which is doubtless how 99% of population will remember it." Now we are at this stage, the point here is to manage expectations of SSW influence: yes, a higher chance of cold weather later this month, but this shouldn't be interpreted as a de facto risk of *very* cold/snowy weather.

The very polarised distinction between 1st two EC clusters exemplifies why that period in March is proving exceptionally tricky and with low predictability, particularly as models toy with SSW effects further into the month.

It's possible. For sure, operational models will struggle in resolving that for a period now. The 12z GFS 500hPa pattern does lean towards where latest EC Monthly height anomalies (below) take us in the broadscale story late March.

Winter forecasting in the longer medium range/seasonal sense benefits from some areas of greater predictability, e.g. MJO forcing, which becomes indistinct and ultimately of no forecasting value as we move on into the summer, plus various global teleconnections known to have more distinct influence on winter outcome (as we saw very evident this past winter). Clearly however, in terms of forecasting shorter-range SFC conditions, both seasons have a raft of very differing challenges for the UK.