Singularity

It means something. Ensembles are all about tweaking conditions and seeing what effect that has, so when a large number end up in a similar place 8-10 days from now, it indicates that there’s a lot of ‘wiggle room’ around the consensus before the end result differs much. This is especially important when dealing with something usually as finely balanced as a ‘heat plume’ event. There’s an abnormally high probability of a very high temperature event affecting southern UK sometime Fri-Sun. Of course it may still not come about but it’s now irresponsible to dismiss the signal being given. I’m hoping we can avoid an extremely hot outcome but, over the past few days, have become increasingly concerned that we might not.

Impressive consistency from ECM for a couple of days now, assuming the actual outcome is close to its predictions. It’s tended to be near the middle of the EPS spread, too. All while GFS battles with flatness syndrome and UKM switches between more amplified and relatively flat outcomes.

GFS is most progressive with a low heading north from the eastern Med. This brings it far north enough soon enough to be picked up by the polar jet stream, which causes the Atlantic trough to change orientation more toward west-east than southwest-northeast. Hence we see a gradual cooling by westerlies instead of recirculating heat before a potential high-end plume event. Other modelling has a slower and/or weaker low from the Med. that stays independent of the polar jet.

18z GFS is shaping up a lot more like the ECM 12z although still with some variations. Temps in much of England are several degrees higher than the 12z for Sun-Mon due to less of a cut-back flow off the North Sea. Overall there’s a continuous ridge of high pressure across from the Azores maintained, rather than a temporarily separate high that attempts to rejoin. Seems to be important going forward.

So GFS has all but struck out alone… it’s going to look either excellent or tragic soon! Just realised the ECM 12z has 20*C 850s approaching from the southwest as of D10… not something I can ever recall seeing before!

In a way GFS & GEM have swapped places - classic! The hyper-flatten outcome does exist in the ensembles but only a small proportion. GFS 12z is toward the extreme end of that, too. With UKM supporting the 12z GEM there’s still a strong vote for heatwave conditions to establish next week.

The 06z does feel a bit 'out there' in terms of flattening the high, but has some educational value: A flatter outcome positions the high further south, which is a location more conducive to sending the very hot airmass over Iberia northward. I sense the charts for 15th are going to be as extreme as any we've seen so far. Edit: Well, they nearly were. The NE movement of the formerly cut-off low was surprisingly rapid (then again, this is GFS, should I be surprised?). ...but here's the thing - we're seeing a lot of different pathways that lead to temps hitting at least the mid-30s on at least one day.

The 12z EPS clusters make for interesting viewing. The deterministic run fits into the cluster with the most developed Scandinavian trough (#3), which contains just 11 of 51 members, about 21.5%. That suggests that the feature is more likely to be smaller and further to our east, allowing the high to extend across the North Sea a bit more. Broadly speaking, that makes the hotter outcomes more likely going forward. Personally, that's not my preferred outcome, considering just how hot it could feasibly become, but it is what it is. Looking further ahead - for what it's worth - the majority cluster brings a gradual breakdown from the west, but it only has one member more than the 2nd largest cluster, which has the pattern that would send very hot air northward across at least southern parts of the UK. It has 35% of the members, which is a lot for such an extreme scenario at 11-15 days lead time. The remaining cluster of 14 runs is tricky to read into. It has that '2 things at once' appearance, akin to the 12z GFS run.

A common feature of both the ECM & GFS runs is a low heading due northward from the eastern Mediterranean in 7-8 days time. This merges with an incoming Atlantic low and enables a broad trough over Scandinavia when otherwise there would be much less of a feature there. The eastern Med export is an unusual event so I’m wary of taking it seriously at the current lead time. If it is to happen, it conflicts with the move to lower heights to the west of Europe that we ought to se attempted around 15th or so (as AAM nears the peak of its climb), so will be tough for the models resolve - in which case we might see a fair few more peculiar-looking runs in the days ahead.

Looks like the answer is ‘no’, although it wasn’t a very convincing outcome in the end, bit too messy - as if it was trying to do two things at once. A broad trough over Scandinavia with a hot high to its south and the UK peripheral to that (still potentially hot in south at times) did have a little support in the 00z EPS, though.

For 12th-14th July the 12z GFS is the closest I’ve seen to the 18z of two days back that produced the chart with widespread 40+ temps in southern UK. Just thought I’d note that in case this run happens to continue heading in a similar direction, which would be interesting to see and give a sense of whether what goes in early next week has much bearing on what takes place toward the weekend.

Curious to see the GFS 06z pick up on the ECM 00z pathway but take it a bit further. I expect it's also from the low end of the ensemble spread for northern UK, but it at least backs up what I and Captain Shortwave said regarding what happens after day ten (not a northerly), even if it is a very messy outcome on that run. When it comes to watching for a potential heatwave in the 8+ day range, it's best practice to stick with ensemble products and see whether there's more than a smattering of support for such conditions. That's certainly the case for London this morning. Birmingham too, when allowing for the heatwave threshold being a degree or two lower: Further north, the spread inevitably lowers and once you reach the likes of Glasgow, it mostly tops out in the mid-20s. This is not surprising considering the origin of the high pressure system and the signals for it to drift toward Central Europe, meaning the core of heat is focused over the western half of the mainland and potentially southern UK. From the expected position early next weekend, a heatwave in Scotland would require the high to drift a bit northward and then stall, but signs are, there will be too much of a zonal flow tied to a +NAO, +AO pattern for that to occur. What raises the heatwave odds for the south lowers them for the north - nationwide heatwaves are rare for a reason. Still, there's a chance of hot weather reaching Scotland for a time, as around half the 00z ECMWF ensemble runs set up a slow-moving low to the west of Europe at some point between 14th and 19th July, leading to a big northward push to the (in a lot of runs very) hot airmass.

The trough off Canada would ensure the ridge returns back east going forward but this run sure does give the north a hard time relative to what’s possible with just a few adjustments. I’ll be surprised if it’s not at the low end of the ensemble spread for temps & SLP in Scotland. The firing of a vigorous low at Scandinavia is pretty extreme.

A strong MJO event is now sending AAM upwards, likely peaking near neutral for a while in mid-July. It seems that as the MJO heads further east in the Pacific, there will be a short period during which the La Niña standing wave and the MJO event have near equal but opposite influences on AAM, causing to stay near neutral for up to a weak before falling away. The big AAM climb is what brings the Azores High our way to begin with. One that reaches neutral should correspond to that high extending toward Central Europe with low pressure taking shape somewhere between the Azores and Iberia (the La Niña forcing will likely prevent a full displacement of the Azores High with a broad Atlantic trough replacing it). That’s where the scope for a heatwave (and possibly very high temps on top of that) comes from. The big question is how quickly AAM reaches that point and just how far eastward the high will manage to extend. By the look of some model runs, there are some quite slow outcomes in the mix. ECM 12z being one of those - it’s still setting up a more plume-like situation via the Bay of Biscay as of D10 but has taken quite a winding road to get there). Once AAM falls, that high should depart either to the east or northwest, but that currently doesn’t look to be until the final week of the month.

Always worth bearing model bias in mind when analysing runs. Since public access to ECM & UKM charts improved dramatically, I’ve noticed that both tend to overdo cloud amounts a bit, resulting in max temps typically being a degree Celsius or two too low generally, occasionally 3-4 Celsius. A similar bias to GFS - but that model has no excuse because it tends to underestimate cloud cover a bit! So I believe the details from ECM & UKM are a little too pessimistic if you’d prefer it brighter and warmer. After all, it’s a strong high overlapping our lands, with impressively warm 850s arriving by Thu. Even with a lot of moisture in the airmass, that’s not a recipe for widespread overcast skies with trailing fronts failing to break up as they move over land. Sure, it’s not going to be a stunning week of weather, but many of us could do a lot worse (the northwest may be an exception; here a lot of cloud seems likely).

It may only be virtual, but we’re witnessing history with first ensemble (one each of the past three days) and now a deterministic run producing temps in the 40s in England. Yes, usual caveats apply at such long lead time… but with such a persistent signal buzzing around for so long now, it’s clear that this is one time we don’t want to see a classic eastward drift of a strong high pressure cell toward Central Europe. The heat reserve that looks to span NW Africa to Iberia as soon as next weekend is just punishing, utilising an exceptionally warm western Mediterranean & Spanish plains that saw unusually little rainfall even by their standards last month. Add widely drier than usual local soils into the mix and it’s a recipe for reaching the big 40 if the pattern evolves in a conducive manner. It doesn’t even have to be precise - the 4 runs showing 40s have featured more than slight variation to the details. Hopefully the high will stay more toward the UK which is the main alternative outcome in the EPS.

If it pans out like that, there won’t be a lot of forcing to move the pattern on again once high pressure is located over NW Europe. Only once AAM begins to fall and reaches back into the negative values can we expect the high to shift back west again. I currently view any model runs that do that before late July - E.g. GFS 12z - as being too quick to proceed.

Todays trend has been to extend the ‘preamble’ to any very hot interlude, during which UK temps range from mid-20s to low 30s depending on the precise position and orientation of high pressure. A pleasant prospect for many. Looks to begin Thu or Fri next week, perhaps the weekend for western coastal areas. Reviewing the expected MJO & AAM trajectories and associated historical cases, I can’t see much reason not to expect the high to head for central Europe at some point - but it might take a while. The models are often too quick with their initial take on a pattern sequence.

I’ve never seen GFS predict over 36*C three runs in a row at a fortnight’s lead time before… truly extraordinary in itself. The heatwave signal is very strong but, as always at the 7+ day lead times, depends on some assumptions regarding tropical forcing & how cleanly weather patterns respond. Of particular importance is the persistence of a +AO and +sNAO. ECM 00z cast a shadow of uncertainty at D10 but I think there’s just about enough eastward momentum to the western N Atlantic trough that the ridge would be refocused eastward again. Still, it’s enough to give pause for thought. The minority cluster in the EPS keeps high pressure focused over or just west of the UK - warm or very warm, maybe hot at times, but not extreme temp-wise.

There’s a decelerating upper flow across the UK which places the surface ridge with a decent overlap of our lands. Please don’t treat 500 hPa anomalies like actual high & low centres. Edit: Just seen your 12z analysis which is closer to the mark. It supports the 12z ECM very well. ——————— Overall in the modelling I see good support for temps to climb above the warm category by next Friday - some patience needed. Beyond that, EPS are almost 3/4 in support of a heatwave pattern for NW Europe including at least southern UK. A continuation of the signal that first appeared in the extended ECM, but brought forward in time a little. While far from guaranteeing such an event, the strength and persistence of the signal should not be taken lightly IMO.

GFS seemingly determined to find complications with the weekend low pressure complex but continuing to look isolated compared to other modelling which still keeps the low’s influence away from SE UK & then begins the build of high pressure in from the west Sun-Mon. GEM 00z cautioning that we might not be comfortably under the eastern flank of that high - only just reaches eastward enough on that run to bring sufficiently stable conditions via warm air aloft for avoiding a very cloudy outcome.

Being under the eastern flank of a high that’s orientated west-east is very different to when the high is very circular or orientated north-south. It puts the UK in an area of decelerating winds where warm tropical maritime air can accumulate. Cloud is likely to counteract that warming effect by day in the northwest, but further south & east, at least some sunny spells are likely and in those, temps should respond well. Provided, of course, the models are right to have a west-east orientated high that overlaps the UK rather than staying more to the west.

Slower arrival of ‘very warm’ category 850s on the 12z runs generally but with that being due to a more amplified high which corresponds to one likely to stick around for longer. As others have noted, a descent trade-off if you like pleasant summer weather with temps mainly low to mid-20s. Cloud not an issue as widely compared to the flatter 00z runs, as the flow is a bit slacker. Still could be a bother in the northwest, mind.

That has been true of GFS, but ECM & UKM have both been slower the whole time, with GFS having adjusted toward them. In fact ECM was even slower until yesterday’s 00z run. Today’s ECM 12z has not supported GFS 12z regarding Sat so it appears that was a rouge run, the prior 06z being a far better guide at this time.

GFS 12z really threw me with Saturday’s low position - much further south & stronger than prior runs. Had to check that I wasn’t looking at cached charts! UKM 12z hasn’t supported that change but it can be a step behind on new developments. ECM 12z will be important. Regardless, the signal for an impressively strong Azores High to ridge across next week remains - and that signal is also impressively strong right out toward mid-July.