We will be seeing lot of sunshine even with northerly or NW the temperatures are not very cool so that will keep indoors warm from the sun.  Hope the low 20's later next week will not be the last but to see once more again in October, not seen a decent warm day in October since 2011.  Rain can stay away as long as possible.

And odd view, given it's vital to life. But, hey, water comes out of taps...

It looks like being dry for the foreseeable - 2/3 weeks. The drought is back...

Or never went away for some, obviously location dependent. 

I rekon. In Essex mostly it seems, going by some data I've seen anyway.

We've had about 100mm of (often steady) rain this month but I bet it's still dust a foot or two down and for many feet further down.

Whatever, normal synoptic service is still absent. Western Europe is dominated by high pressure, again.

Or never even existed for others, depending on location of course.

WX temp summary can't settle to a consistent outlook. Although there's still a trace of warmth on W European coasts in week 1, the cooler air is back and in week 2 all the way down into Spain - any recovery is reserved for the area near the Black Sea. In week 2 even some little patches of <0C (and that's a day-night average) in Iceland and Norway's Hardangervidda. Rain in week 1 heaviest on N Atlantic and also patchily across Europe; in week 2 fairly solidly across W Europe with Britain on the edge of the heaviest.

GFS Op - HP drifting E -wards and decaying by Fri 23rd to be replaced by HP on Atlantic generating a variety of NW/N/NE winds until Wed 5th when it finally moves E to Britain. Strongest N-lies associated with LP E of Shetland 985mb Tue 27th and again 995 mb Mon 3rd. 

ECM - a similar sort of pattern but the HP is closer in at first so N-lies don't start until after Mon 26th and then the LP is less deep so slack NW-lies  developing on Thu 29th

GEFS - ens members agree on recovery of temps to norm briefly around Fri 23rd after which mostly below norm to Wed 5th (less certain in Scotland). Rain on and off from the 23rd (a lot more than yesterday's forecast for S).

Rain will come eventually but not now until clocks go back ward which usually become unusable outside but now need warmth and sunshine for heating stay off and reduce the boring Autumn type.  1995 was the shortest at 1 month only.  Anyway with models showing lot of HP we still see lot of moisture and greenery.  Any water table now wont drop anymore due to weak sun and no heat and easily recovered from rain and snow in winter.

First taste of cold incoming. Looks like the boiler will have to come on

One of the most extreme ensembles, this could bring night time flurries to lowland scotland

This is about as rare as snow during June.

WX temps still see-sawing; while generally cold air continues to occupy Europe (and some notably cold patches over Norway), week 2 now shows a resurgence of warmth from the SW to affect Spain, W France and S England. Rain for N & S Europe week 1, Britain in a damp middle of the sandwich, in week2 a very dry area Biscay and into France with rain around the periphery incl N Scotland.

Jet - nothing much at first, a blow from the N on Mon 26th, dying away, before a more traditional W-ly flow near N Scotland from Fri 30th.

GFS Op - current HP moving E and weakening while HP arises on Atlantic, strong enough to affect Britain until next weekend, after which it retires W-wards and opens the way for N-lies Mon 26th, not lasting as on Thu 29th the HP has shifted SE to France with Britain in a W-ly regime, indeed intensifying 1030mb Britain Tue 4th before moving back S. Fiona still at hurricane strength Labrador Thu 29th, breaking up and remnants running across Iceland to N Norway.

GEFS dropping back to cool for a week or more after a brief interlude at norm on Fri 23rd, though op & control take a different view from other ens members and offer a warm spell from Mon 3rd. Rain on and off from Fri 23rd in many ens members though not in Op. Scotland less consistent - temps generally cool but more up and down, and rain in bursts esp 23rd and 26th.

ECM makes more of Monday's N-ly, pushing a large amount of cold air S-wards to generate LP over France Wed 28th (note difference from WX) with NE-lies for Britain before bringing back HP block to the W of Britain Fri 30th with little sign of W-lies

This is what the T850hpa situation looks like for the coldest September day ever (29th 1915)

Its notably less extreme than many of the ensemble forecasts for the same day this year.

One to watch.

What do people know of Snow to low ground in September. I know its extrodinarily rare, and happens less reliably than in June.

I'd be surprised if that shows the lowest 850s recorded on a September day. I assume you're talking about 2m temps?  

Yeh, you are probably right about that.

But surely it still has potential.

Looking at the 1915 event doesn't show that much special aside from the relatively cold airmass. I imagine some front or something got stuck under it, but some models show that for 2022 too.

If you want decent cold in the south I remember 1975. This from Trevor Harley's site - The southerly tracking low sticks in my memory 

"The hot summer ended abruptly on the 12th with an intrusion of arctic air. It was very wet: September was one of only two wet months this year. The 13th was very wet in the south, with 50 mm of rain widespread, 75 mm on the north Kent coast, and 91 mm at Margate - accompanied by a severe NE gale. The temperature that day only reached 10C in London, and as low as under 9C in parts of the south east - this is the coldest day recorded so early in the second half of the year. The 14th was also a cold day. A tornado destroyed glass in Barnham (Sussex) also on that day. As the winds eased on the 15th, with arctic air and clear skies the temperature widely dipped across the SE to -3 C, and as low as -6.1C at Dalwhinnie. A cold night at Lagganlia in the Highlands: -6.2C on the 27th." 

September 19/20th 1919

NOAA_1_1919092018_1.png (959×770) (wetterzentrale.de)

From Trevor Harley: "There was even snow cover on low ground from northern England north and on higher ground in Wales and the southwest as well as high ground in the Midlands on the night of the 19-20th. The snow was 2" deep at Princetown in Dartmoor. Snow cover lasted on Snowdon for a week. This is probably the earliest snowfall date."

I can recall some very noteworthy cold maxima in the SE in late September 1993. 8's and 9's from memory. 

Will need to check.

In the north, the Peak District hills could be seen from Manchester covered in white with this event; I would estimate snowfall down to about 300m.

Looked at many of the synoptses for these events. The 2022 event has the potential to eclipse all of these.

To enphasise: this almost certainly won't happen but there is huge potential here.

The frigid airmass over greenland combined with the transient GH reminds me of early April 2021 which brought some of the most insanely cold uppers I've ever seen for the time of year.

No reason why we can't get below -5C in Scotland T850 which should be enough for low level snow at night.

Progged maxima on the 27th

11-12C in S ENgland

8-10C N England

7-9C S Scotland

6-9C N Scotland (and obv colder in the highlands)

I thought -5c 850's in the UK were fairly common in September. Nothing out of the ordinary ? 

Outside of the northern isles I don't think they are common at all. In fact they would even be considered rare at the end of October.

P10 is legitimately record breaking.

This would undoubtedly be the coldest September day ever if it came off. Snow would be widespread, at least by night, to lower levels in the north.