Yes, it takes a bit of digging to find it but here they are:

https://www.metoffice.gov.uk/climate/uk/summaries/datasets

I've just used the link and it's easy to navigate. Thanks Tim.

I tend to use the NI data from that page, but it seems to have trouble formatting properly when converted to a CSV file.  If you are using EXCEL, a tip, if it comes out all weird upon opening, is to go to the 'Data' tab and click on 'Text to Columns'. Worked for me the last few times by tidying up the data and placing them in their appropriate columns. 

You're right CP- it didn't all align on one line, if that makes sense- I can just about cope with that as sometimes (oddly, not always) the HadCET dataset does that to me as well. But I'm grateful for the tip.

That said, apart from last year, in recent times September and even October have been more akin to summer.

Just very disappointing that can't can't string three summer months together. This August has been extremely nondescript.

Believe me, I feel the same way about winter! 

That the problem when 1 rogue month like August skew your data like when you see very good summery data in May to July then suddenly you see boring data in August.  In other countries if they have a good summer they last all the way through and only the next season can stop it.  Like last year we lost the summer half way.

Aye. Anyone would think we live in a temperate country with a changeable outlook month to month. 

If on average you get a summer month with a >1C CET anomaly 30% of the time (which is generous but we are getting there in recent decades) then the chance of getting all 3 months that warm is only 2.7%, or once every 37 years.

I don’t know what the current probability / recurrence interval of >1C anomaly months is but clearly having 3 successive >2C anomaly months would be way way rarer still.

A rising baseline temperature plus persistence helps make these events somewhat less rare but still pretty exceptional.

Nobody has answered my question, why do some months seem to get caught in a repeated pattern?

I can see a string of consecutive months falling into a pattern such as a run of wet months, a run of cold months etc but its harder to explain why a month caught in such a pattern defies a pattern of surrounding months, a cuckoo in the nest. 

August 2006 within a string of very warm months.

August 1921 within a pattern of dry months

I'm not sure if anyone can honestly answer your question fully, although there have been postings past on other threads that talk of 'cycles' and other factors such as regularly changing synoptic set-ups. But you are right that there do seem to be patterns. Another arguable one is that in the 22 years since 1997 there have only been 3 Mays below average in CET, yet 14 Mays at least 1 full degree C above. But this might get written off to general warming, so it would be more use perhaps to look for runs where temperatures are below the LTA, including an investigation into the lack of August warmth. The truth there is that since 1994 & including this year, (so 25 years) the CET reveals that there have only been 5 below average Augusts- just 20% of the sample, in 2007, 2010, 2011, 2014 and 2017. The coldest was just 1C below average. June also has had 5 below average months during that time. July takes the trophy, with 7 below average months.

By contrast, in the same period, 7 of the Augusts have been at least 1C above average.For June the figure is 9, for July 10. So perhaps it's not so much the cool Augusts as the lack of really warm ones compared to the other summer months- but the difference isn't that great. This is, of course, all going by CET, but it's a different picture as Tim pointed out at the start of this thread, if you look at England & Wales as a whole, and going by what posters further north have said about this August, maybe the UK as a whole would show an even starker contrast.

I think the run of Novembers that were cooler than Decembers you mentioned is notable; really quirky.

If there’s a large scale SST pattern that favours high pressure over us in June and low pressure in August, and that pattern has interannual persistence, then you could get a string of sunny Junes and cloudy Augusts. An example often quoted is El Niño favouring blocked conditions in late winter but not early winter. So a decade with lots of El Ninos might see a tendency to cold February and March, but mild Dec and Jan.

In the early summer months I understand the big shifts include:

- Melting of Eurasian snow

- Onset of the Summer monsoon circulation

- Northward shift of the subtropical jet and Hadley cell over the Med

That is all connected to some extent and partly accounts for switches in July

In late summer and early Autumn we then have:

- Reversal of temperature difference between Atlantic SST and land heating

- Onset of Atlantic hurricane season

This potentially explaining the fact weather often shifts around the August bank holiday.

These shifts, particularly the SST gap, I think also account for the fact Western oceanic areas are generally better in Spring and early summer while Eastern more continental counties do relatively better in late summer.

In a typical year I think the opposite is usually true though - oceanic climates tend to have more seasonal lag so their best weather tends to come later in the season, while continental climates clearly peak around mid July. In climates like ours, our record temperatures mostly fall in August, while in very continental climates, mostly July. This is very evident in more northern Continental climates that respond very strongly to declining day length, and experience a very rapid temperature drop as September progresses, with places like Moscow going from average highs of 22C in August to 15C in September. Indeed, in cities like Moscow, May is noticeably warmer than September, but in maritime climates like London or Paris, the opposite is true.

As I mentioned, for Heathrow airport, August is the sunniest month of the year in percentage terms with around 46% of possible sun. For here, May is indeed the sunniest month of the year but July and August are ahead of June, and in percentage terms September is ahead of June too. If you look at sunshine percentages for a variety of stations in the UK there is a noticeably decline in June for whatever reason (return of the westerlies perhaps).