Man With Beard

Navgem isn't the only model going silly hot by the way - ARPEGE also gets 35C into E Anglia on Wednesday. I'd say chances of this coming off are low but not impossible.

No real change on 00Z runs so far. Although daytime records look out of reach on this occasion, I was wondering about nighttime ones? The warmest ever night, as far as I can see, was 23.9C in 1990. I'd have thought there was a chance of London falling no lower than 22C on Tuesday night - so not far away. Also just put off interest, the GEM brings 20C uppers over much of southern England and 16C or more up into parts of Scotland. You could add a couple of degrees to expected temperatures if that verifies. Further afield, the GFS continues to push weak ridges into the UK, keeping the southern half fairly warm, but the week nature of the ridges means confidence can't be too high on exactly what will transpire.

I'll just clarify - this is not correct. It's a textbook plume.

Now this is going to be interesting. At T144, the ECM dives another small trough just before it reaches the UK. A further draw of warm (not hot) air from the south - could still be clinging on to 25C in many S/E parts. But I sense there may be more developments to come in this direction...

So a nationwide hot day. Shows how things can change quickly, even 48 hours ago many of us were writing off chances of heat for the north! The GFS really is exceptionally hot - 18C 850s in place for a whole day over southern UK - a slight tweak could see 35C threatened on Wednesday - that would be the absolute ceiling for this situation, and I rather suspect that the GFS 00Z has slightly overdone things anyway. 32C a more likely target.

just to back up my point above, uncertainty over the trough placement by D10 is nicely shown up by the spreads - both GFS and ECM struggling to pin down exactly where the trough to our west will be by D10:

I would love it to be right but I think I recall it showing 40C+ on some of its runs last year actually, certainly there was one model doing this and I think it was NAVGEM. Best chance of getting well into the 30s would be if the exceptional uppers could be delayed by just a few hours on Wednesday morning, like on this GEFS member: The delaying of a breakdown is far from uncommon, but then again the speeding up of one is common too. Unfortunately the best of the uppers look to be coming overnight. Still, just a few days ago next week was supposed to be "back to normal", so plenty for sun lovers to be grateful for! I don't think it will be the last heat burst of the summer, regardless of what the mean charts are telling us - I think we'll see another buckle in the system in the following two weeks and another chance of a plume (*This is a hunch, not a forecast, by the way!) ..... Looking at the GEFS in a bit more detail - two clear alternatives come up by D10. The first option (example below) is a stalled low either over us or just to the east, with high pressure generally building north through the Atlantic. The result will be a northerly influence, cool and less settled, particularly in the east. 11 members follow this theme in some way. The second option (example below) sees the Azores High getting back in ahead of the low (or arguably pushing in more vigorously behind a clear low), and the low stalls to our west at least for a little while longer. This option allows for a more southerly influence and possibly another warm spell of weather. 9 members go this way in some shape or form: Therefore, it's not a good evening to consider the GEFS mean as representative. With such an even split of runs going distinctly one way or another, the GEFS mean simply flattens everything out, which betrays the fact almost all of the ensemble members have much more amplification wither in one part of the north Atlantic or another. Looking three days further ahead (D13) at the two example members I showed above, you can see something else that is interesting: For the unsettled one at D10, the trough has now cleared and high pressure is building back in. But the "settled" member at D10 now finally succumbs to the trough. Lots of the members copy a version of these runs in their evolution. What does this tell me? Probably this. The D8-D14 period may end up "averagish" on the whole. But these ensembles tell me we're not likely to see average weather on each day. Rather, we'll probably see some very good days and some very poor days. Maybe I'm stating the absolute obvious here(!), but I think this is an important point to make, especially as some on here may confuse people's forecasts of "average weather" with "no good/bad weather at all".

Thanks Knocker (and JH). Completely agree. Hopefully I (and maybe others?) will be able to cover more scope in future months, if the community on here think it is genuinely useful.

Having recently been challenged to provide more evidence for my ideas (!), I have done what I do not think anyone on here has done before - a mini-study on the accuracy on the GFS and ECM, specifically relating to the UK. Please have a look at the attached PDF - any comments would be gratefully received, as I would like to extend this study when I get time later in the year. It was done in rather a rush, so feel free to point out any errors I have made. A summary of my findings, which cover the T144 modelling of Central England by the GFS and ECM 00Z runs for June 2016: - Overall, the GFS was slightly more accurate than the ECM - The errors made by the GFS were almost entirely due to overestimating heights. The errors made by ECM were more evenly spread between overestimates and underestimates - The ECM was slightly more likely to make bigger errors than the GFS (insignificant difference, though) - The ECM made no errors at all for “trough” dominated days GFS ECM 00Z Comparison for June 2016.pdf

Wednesday has the potential to be hot too on all main models, but it will all be about timing and local differences - Camborne probably least favoured for heat on that day

GFS 18Z so far has an even better shape to bring some very hot weather closer to the UK for mid-next week - that developing low in Biscay will cause a surge from the south ahead of it: - would also be highly thundery next Wednesday if it came off.

This is an excellent mean chart, Karl. You will struggle to get a much warmer mean that this at T168 even in a heatwave. It's looking increasing like next week will be an extremely different experience of summer at least south of the border. We could be looking at a few days between 25C and 30C. Certainly warm enough both in and out of the sun. Of course some on here will already be plotting the end of this spell before it even begins (and even now there's still an element of doubt over it) - Knocker rightly points out the threat of increasing heights in the Atlantic - but as I said the other day, patterns once established have a habit of repeating so certainly too soon to call it a flash in the pan.

I think he's just seen number 1 - and now he's so excited, he needs a number 1!

The GEFS mean 850s now stay above 10C for many days in the south - and a few runs get close to the magic 20C line http://www.meteociel.fr/cartes_obs/gens_display.php?ext=1&x=306&y=141&ville=Londres All touch and go though as Knocker alludes - could easily be a complete miss too

A fair 18Z so far. Not wall to wall sunshine, not always hot but some fairly decent days between now and next Wednesday for all. I'm drawn to all that heat building to the south by T180 though ... not far away from a pattern which would allow it to seep northwards in the days after...

how one more thought for now, though, everyone seems to agree we aren't seeing a pattern change - but I would argue we ARE looking at a pattern moderation coming up. Compare the present with GEFS D10 mean The current chart shows a westerly upper flow - with fluctuations - but its a very southerly based pattern, so these fluctuations rarely have allowed those lovely summery oranges to get up to us. But by D10 In some senses little has changed - westerly upper flow, some ebbs and flows. But look how much further north that yellow line is. In the context of the chart, that suggests to me that there will be far more input from southerly sources into the pattern than we current have. This pattern moderation may be enough to bring a completely different experience for at least southern UK in ground conditions between D8 and D10 - and seeing that D4-D7 already look pretty useful, I think current modelling would represent not a July 2013 style heatwave, but a very different experience of summer nontheless for the second half of July when compared with the first. Check the D10 mean from 10 days ago, compare it to today's D10 mean and I think you will see where I am coming from.

Having had a bit more time to look at this now, I have to agree this is fair. Studying the GEFS clusters often reveals if the mean is representative or just an average of wildly different scenarios. A good look at the GEFS 06Z members suggests it is, at the moment, representative:: http://www.meteociel.fr/cartes_obs/gens_panel.php?modele=0&mode=1&ech=240 No runs at all with high pressure in charge by D10. Best chances in the south However ... I would caution with this - once a different pattern appears in the short term (such as for this weekend), in my observations it is not uncommon for models to then repeat the same feat afterwards. I'd hold this page before writing off chances for the rest of July.

Your cherry tree looks like bearing fruit, Karl! It's yet another reminder to us about being sucked in by one run (yesterday 00z). It's now looking like the northerly movement of the jet, picked up in so many ensembles in the past week, will be realised. A much better outlook for the rest of July therefore. Or am I also just getting sucked in by this one run??!!

Very good spot Knocker - sometimes the "Atlantic " view doesn't pick up little kinks in the pressure that well - it needs an eagle eye like yours to pick them up

Keep watching, keep watching ... the ECM builds the heat again from the south early next week

Well I mentioned in my post yesterday morning about expecting a bounce-back but I didn't quite expect that! These morning's charts are completely different for the UK (fair play to UKMO, it was a lot closer to this on the 12Z run yesterday - and still waiting for ECM of course). Now looking hot and sunny for most this weekend again. What on earth happened to throw all the models out like that yesterday? Missing data? Nick Sussex, where are you!! (Hibernating for the summer in his French pool, I imagine!)

That's a shocker. We'll need a pretty decent last third just to get above 16C.

Now come on. You have said "no significant change on the NOAA" - did you stop to check yesterday's 6-10 day chart? The influence of the trough has moved hundreds of miles east since yesterday, and as a result has far more influence over the UK. Your post was very misleading and lacked the necessary evidence to support your claim

Knocker, once get a spare couple of hours (which may be in about 10 years time), I will collate my evidence from the model archives. To me, quoting the official NH statistics on model accuracy without applying them to our local situation and actual physical model output is doing exactly the same thing. A simple way for those learning about the models to get useful evidence would be to go back through this thread and count the number of times people posted "The ECM has moved towards the GFS" or vice versa. This is called "observational evidence" and is acceptable in scientific research if done fairly.

It does seem to the naked eye though, at least for the UK, that the GFS is getting most of the big decisions right. Which this morning means a total car crash for the ECM, which ditches its high pressure outlook for the weekend having been so consistent for days. I still think we will see a little rebound back to a slightly better outlook, but I think the idea of yesterday's ECM has gone for good.