Singularity

Large-scale mechanisms at work behind that settling down and heating up trend later this month. I do also wonder if some positive adjustment may be witnessed later this coming week, as the La Nina-like swing gets toned down. To a slight extent this has already happened but we can't be sure that this trend will continue. Arguably it's being a bit greedy given the impressive longer-term signals, but generally it's best to play for that which is nearer in time and so more dependable.

It's fascinating how this year sees the vast majority of the volume situated on the Atlantic side of the Arctic, in stark contrast to recent years including, most notably, 2012. In theory that means there's a lot more ice in a position exposed to oceanic melt-forcing that typically melts out all of the ice subjected to it by the minimum. On the other hand, this year so far has displayed an impressive ability to keep the region much cooler than usual. So perhaps it will be an atmosphere v. ocean battle. Unfortunately water has vastly more heat capacity than air and the sub-surface heat reserves are relatively large even where surface temps are not that far above the long-term average. Any deep storms that cross the area such as we saw in late winter and early spring will cause some big problems. According to predictive models working from the current situation, only an extreme retention-favourable Jul-Aug such as 2014 can prevent a record minimum in terms of sea ice extent this year.

So ECM more or less follows GEM... Shame it's at such unreliable range but in general it highlights how simply having a low consolidate near the climatological norm position of around Iceland can really help us out during a period of positive summer AO.

Ah, one of those rear-flank secondary lows causing a bit of bother here. Not particularly bad in the S and SE but western residents best hope that shortwave low is a ghostly apparition . After a lengthy hike over uneven, often rocky or boggy terrain in 26*C temperatures and blazing sun, those cool temps aren't looking as disappointing as usual to me right now . This run is being difficult. There must surely be some other factor than GLAAM encouraging a La Nina-like tendency to the model output? I'm away from the PC, without my usual links, so I can't check things out myself. Just getting these charts linked is taking an age . To finish for the day I will abandon this GFS run in favour of GEM, because, well, it's more satisfying to look at . Probably taking the 'room for ridging' idea a bit too far but it does capture the idea of it being finest in the south.

After a nice one-day delay on the 06z, the 12z is back to turning conditions cool with spells of rain for all by Tuesday. - but as GP has also been suggesting (thanks for the input), there's plenty of room in the near-neutral GLAAM setup for the Azores High to improve conditions across the south again before long. GFS 06z managed it by Friday - in fact Thursday if you find dry weather with sunny spells and high teens maximums to be fine enough on occasion. 12z is looking flatter with the jet stream across the N. Atlantic by midweek than the 06z was, but with a major trough approaching Greenland there's still the potential for the Azores High to ridge NE in conjunction with the surge of tropical maritime air heading ENE. Depends if we're seeing another case of a Scandi trough being made too broad in nature. It'd feel reasonable enough in the SW if any sun broke through - but on this run it stays cloudy with high teens. Needless detail at this range I know, but seems realistic enough.

Thanks for the update Tamara I retain the suspicion that the atmospheric state is not far enough from neutral to lock in a retracted Azores ridge for much time so will continue to watch for weak ridges producing a warm/very warm day or two, but must concede that I did not adjust my AAM corrections down fast enough after the recent underestimated ascending period. So next week does look likely to be changeable with temps mostly varying around average or a little below - but still the chance of an appreciably warmer day or two at least in the south. I do however still view the GEFS week 2 drop of AAM as unrealistically large, as barely a day has been seen in the past year in which it wasn't overdone. So too much drive toward establishing or maintaining a cool and unsettled regime. Good thing is, most on here don't take GFS very seriously at that range anyway

We'll see... I find it hard to take the longer-term ECM seriously until if and when the MJO actually does make a jump into phase 2/3 as the model has been predicting for 1-2 days ahead without success for over a week now.

Familiar ground here as the jet stream strength and 'flatness' is reduced nearer the time. So it is that the models are toying with yet another shallow low dropping down west of the UK by Monday. Question is whether we again see a breakaway ridge from the Azores manifest, or have to put up with the low drifting slowly but surely across our lands Tue-Wed. Either way, with the sting taken out of the Atlantic westerlies, there's an increased likelihood that the Azores High will manage to build far enough N that the next push from the west drives an extension of high pressure over to the UK with the jet stream tracking to the north of us. Hopefully this shift will be the start of a trend but I'd not be surprised to see the models remain unsteady in their corrections, just to test some nerves . It's such a shame the way the GLAAM bias causes the majority of GFS runs to try and turn things cool and unsettled by the 12-16 day period, but at least it's keeping people's expectations from becoming unreasonably high .

I should have been clearer sorry; I meant the individual runs. Ensemble-derived products should fare better although the low AAM bias will still tend to restrict the signal for eastward extension of the Azores High a little too much. This week's warmer outcome has been made possible by a reduction of the jet stream strength in our sector as the AAM progression was corrected, with this allowing more NE-ward ridging of the Azores High this past weekend and setting up an environment in which trough disruption and cut off lows often tend to occur. It could well have ended up more unsettled of course hence despite the improved ridging there has never a case for shouting for extensive fine weather this week. For next week, the GFS 06z is the first op run to show something along the lines of what I have been contemplating (given AAM trends against bias) and the Met Office have been hinting at (allowing for some timing offset), while the ECM 00z was a pretty good effort. Not sure how cleanly we'll get there though given some room for small disturbances to get involved this weekend for example.

Something to take into account at the mo is some extreme disagreement among models with respect to the MJO; GFS has a (wacky looking) big phase 1 emerging over the next week or so, yet ECM has a phase 2/3 spike (with big impact on some recent runs; today's 12z the best example yet of phase 2 Scandi trough response), while other models generally show little activity one way or the other. I advise paying little attention to what's currently been shown for beyond next weekend.

Briefly as I'm on mobile (in fact, on holiday... so I'm being a bit sneaky!) - GFS has today trended toward a more effective cutting-off of the Iberian/French low mid-late next week, resulting in more of a ridge to the N/NE with an E flow for E&W with some very warm or even quite hot afternoons Fri-Sat (depending on where you happen to set the thresholds). UKMO is not as enthusiastic, with more eastward motion to the low retained through Friday, shifting the peak warmth away with it although a weak trough undergoing disruption and southward elongation out west of the UK could serve to hold-up the movement of that very warm air, possibly with a breakdown of sorts occurring on Saturday so just a little sooner than GFS (by 12 hours, say). ECM was most progressive this morning with the European low not hanging about at all. Will it at least share GFS' trend, or continue to whisk the low away and bring a return to somewhat changeable westerly flow? I believe we're still seeing a bit too much 'La Nina-ness' being forced on the patterns by the models from as soon as next weekend so I'd not be surprised to see the very warm conditions extended further than the current consensus. Ha - turns out I managed to snatch a bit more time than I expected so there was some reasonable detail in this post after all

Taking GFS 12z and applying increasingly large - with time -eastward correction of the ridges from the Azores to account for likely AAM bias leads to a decent overall outlook for southern parts and not half bad further north too . UKMO looking good for midweek once the weak disturbance has cleared early-mid Tue. ECM unfolding as I type...

17.5*C is my estimate, which is dependent on a decent very warm or at least briefly hot spell later on to offset a fairly typical or perhaps slightly cool first week or so.

GFS (left) with a shallow low moving steadily east to cross the UK overnight into Tuesday. UKMO (right) with the same disturbance fading into oblivion just north of the Azores. That is some impressive disagreement for just 4 days range! While GFS keeps a fairly flat jet across the UK, maintaining the changeable westerly regime as a result, UKMO allows the Azores High to slowly but surely extend a ridge across the UK, thanks to the lack of disturbances in our vicinity. It looks good going forward, with the next Atlantic trough in no position to find a way over to the UK. Glancing at GEM, it continues to side with GFS. Will ECM stay close to UKMO this evening to maintain a standoff between the N. American and European continents? Fascinating to have such markedly different options on the table from such an early point in time - but frustrating if you've got forecasts to make!

I can't believe I missed that for several years! Thanks for the link

Hmm so now there's a sort-of-cut-off low type scenario being put forward by ECM and what may be a fully fledged one by UKMO. Having begun by checking up on GFS this morning, it was a relief to see that it's very mobile solution was not well supported, but ECM does go on to produce an outcome disturbingly similar to what's taken place over the past two days which is highly undesirable. UKMO would probably go a similar way but take 12-24 hours longer to do so. Arguably ECM/UKMO offer more hope than GFS in that the deterioration of conditions is further out in time with more chance of it being modified toward something less troublesome. I can still see some scope for a Scandi High to develop in the 6-10 day range which could bring about a warmer solution with showery weather focused toward the S/SW (or ideally out to sea.... I dare to dream), but this is pure speculation on my part.

Good point snowwman, I have been surprised by the pace of both strengthening and downward propagation with this E QBO. Only recently I was reading an article that suggested the most major contributor to the unusual 1962-63 winter was an exceptionally strong E QBO (if only the Berlin data archive went back that far!). Let the hype begin

You're reading it right, but what I'm doing is adjusting for the usual bias which has continued to be evident even with the tropical Pacific featuring a localised Nina-like signal . Looking at the mean and spread you can see the point at which a slight upward trend (which has already been a little underestimated in recent days) is countered by the bias and then overcome, with a marked step-up in the strength of the bias upon progressing into lower-res (past +192 when it comes to GEFS). It is inevitably an act of presumption on the part of those who apply this method which can fall foul of smaller-timescale events such as tropical cyclones and equatorial Kelvin waves. So we continue to be without a 'crystal ball' for weather prediction (obvious to anyone following any weather forum for any length of time), but something is better than nothing . Truth be told though, there is more 'La Nina-ness' going on than was expected to feature this summer when looking ahead as of mid-late May, so we're looking at a less fortuitous run of events than what once looked likely to be the case, and I still can't see enough reason to anticipate anything particularly noteworthy this side of at least mid-July - but this is of course subject to change.

It seems pretty clear now that the splitting jet idea was most likely a red herring, though it's not entirely a dead option yet (occasionally they spring back but it is rare once inside the 5-day range). So we are indeed left with the underlying theme of a bit more ridging from the Azores into Europe which serves to improve our prospects but not entirely prevent those Atlantic systems from having a bit of influence. GFS (left) seems a bit aggressive with next week's first low plus the strength of the jet stream in the mid-Atlantic, which leads to all sorts of trouble over the following few days. UKMO (right) offers a more reasonable way forward, with the spell of rain Monday night or early Tuesday affecting northern parts most with not a lot down south (GFS does at least show agreement on this aspect), clearing soon enough to allow a pleasant day or afternoon for most except perhaps in Scotland. The timing could well change of course, for better or worse (more likely worse if it does change, as it's pretty good where it stands). Faced with these two runs alone, I'd be thinking okay, we've seen the GLAAM-correction introduce more ridge influence for the first half of the week and are still waiting for it's full effect to become evident in the charts for the second half. BUT there is that shocking ECM 00z run lurking in the darker corners of my mind. It found a way to stall-out next week's first trough in a way that effectively held the door open for the Atlantic troughs to march in regardless of what the Azores High would rather get up to... and the GFS 12z has followed a similar train of thought but with a flatter overall pattern. Now, credit where it's due, GFS does show the Azores High fighting back a bit as it very well should, but the stalling low has done enough to keep things changeable across the UK. So it is that I have been given the impression that next week may well be very forgettable overall as far as the weather is concerned. Unless, that is, the split jet either returns for early-mid next week (low probability) or turns out to be the behaviour exhibited by the next trough in line (higher probability; little support for that at this time, but then there was little for the early-mid week low to be so weak while crossing the mid-Atlantic when it was at over 5 days range). Unless a tropical event upsets the GLAAM trend next week, the overall (I must stress overall; smooth rides are rare in the UK) improving trend should continue into the week after, so there's some hope to be had even then .

As much-maligned as the extent metric is, it's going to be interesting to see whether the predicted return of the Beaufort High and with above average airmass temperatures affecting large parts of the Arctic, starting a few days from now, will manage to send 2017 into uncharted territory for July. It's a relief that the deep low didn't reach the ferocious intensity that was once predicted, but it was still only a few hPa off the June record low pressure, and in combination with a vigorous secondary low over the ESS seems to have managed to draw a deep wedge of relatively warm and moist air across the Pacific side of the Arctic, which has triggered the onset of a lot of melt ponds in a short space of time. Melt ponds cropping up so widely just ahead of a strong Beaufort High arriving is not a good combination at all for ice integrity and duration, but the 12z ECM does at least offer some suggestion that it may not remain strong for long, or even manage to sustain an Arctic Dipole for more than a few days; In fact there are increasingly strong hints of a Scandinavian blocking pattern emerging in early July. This could see the main brunt of attack being borne by the Atlantic side, where unusually cool and cloudy weather has allowed a large lobe of sea ice to persist longer than has been the case since at least 2006 (I think - maybe only 2011?). This side can be attacked much more directly by warm air off the N. Atlantic, as well as undermined by the oceanic currents, so the ice lobe could disappear very rapidly if such a pattern change unfolds. Come to think of it, ECM manages a 'worst of both worlds' solution on this particular run, as a fairly strong and 'warm' high becomes centred over the Pacific side at the same time as the Scandi blocking emerges.This model really seems to have it in for the Arctic this year!

Unfortunately, with no 'GLAAM-adjusted' model output or anything like that, there isn't really any means of doing verification stats. Best that can be done is a qualitative assessment of outlooks constructed using the theory, though I've not come across any freely available that are known to be done that way on anything less than a monthly basis. - - - - - - - Looking at the ECM 12z, it completes the set when it comes to the models bifurcating the jet flow around a UK-centred ridge. A classic setup for fine summer weather here and also of one of those situations where fine details regarding a trough in the mid-Atlantic can open doors where the ensembles struggled to see them; two days ago there was 10% support from GEFS for the bifurcated jet scenario. This evening it looks be around 50, maybe 55%. So going by the operational runs it appears that prerequisite has been met, but the ensemble spread reminds us that complications could still arise that either change the jet configuration (which would really not help us at all) or add an extra disturbance in the flow that interrupts the settled trend early-mid week (this would be irritating but better than the first possibility). Longer-term, GEFS remain heavily influenced by the low-GLAAM bias as evidenced in a majority vote for a La-Nina like pattern with the Azores High a little west of us and a trough over or just east of us. Fair to say, this may undergo substantial modification nearer the time. Oh and a massive kudos to GP for spotting next week's potential before anyone else (yes, even me, who seems to have been searching since the hot spell ended!); June 19th + 15 days = 4th July, and 'toward' that date covers the very period in which things look to be coming together. So this could be an exceptionally good call but of course we can't be sure of that just yet!

As Knocker has highlighted, it's not quite enough for western parts on the 12z GFS as a decaying frontal boundary is able to act as a focus for some showers or longer spells of rain (the latter mostly for Scotland). Still enough sunny spells in between for some fairly warm afternoons though (note that I don't mean in the sense of departure from average); - but as expected, not as fine as the 06z was for midweek. It does try to make up for it with some improved output beyond that time, though. UKMO looks more promising for keeping those frontal boundaries away from the UK entirely although you can never assume all that much when it comes to the weather in this part of the world.

Impressive strength to the ridge from the Azores there. This shift in the model output in the manner that I've been looking out for during the past week serves as demonstration of the effectiveness of studying GLAAM tendencies and trends and using those to draw conclusions as to what the models are most likely to be making some mistakes with. See how the observed GLAAM continues to fight its way slowly but surely closer to the neutral zone, with the mean projection for the next fortnight having to edge upward in accordance. I know it's far from a done deal at this stage, but it's the fact that the models have shifted in the first place that I'm working with; it shows the theory is sound, even if something happens to crop up at shorter notice and make a mess of things .

@knocker my point is that the ECM 00z had the trough disruption, something GFS did not entertain until the 06z. . . . Anyway, the 12z (right) has followed suit from the 06z (left), in fact it's impressively consistent; Looking at Tuesday, I think this run will walk closer to the edge of settled condition midweek, as the jet does not bifurcate as effectively as it did on the 06z (right-most chart).

When the 12z is rolling out we need to watch for whether the jet digs down as sharply as the 06z on the right shows (quite the change from the 00z it has to be said!); this is the sort of behaviour that gives ridges the upper hand in the face of troughs trying to push through, as the latter will tend to disrupt against the ridge. The mechanism for this is too fine for ensembles to pick up very well until shorter notice than the operational runs, so if it was to firm up then the ensembles would show a significant shift in response, probably in the space of just a few runs. I'd not be taking all that much notice of it (this is the 06z after all) if it wasn't for the fact that the ECM 00z also depicted such a change in trough characteristics. Something does seem to be afoot.