General summer chat

[B87]

 danm The warmest day at Heathrow in 1991 was 1st September with 29.7c.

July and August were pretty average, so it is surprising that 30c wasn't seen once! Though you get months like July 2014 with an almost 26c average high, only hitting 30.3c once.

[Greyhound81]

 reefIf the 1973 summer would be much warmer now than it was then, what would the summer of 1976 have been like if it happened in the last 10 years?

[B87]

 Greyhound81 July and August I assume would be like those of 2022?

[CryoraptorA303]

51 minutes ago, danm said:

I think the last Summer here not to hit 30c was 2007 or 2008 and that was an anomaly

Those two still managed to reach 30°C further northeast in northern East Anglia, which is almost guarenteed to have not been the case pre-1990. They're the kind of summers that would've seen maxima in the region of 27-28°C, which sounds absolutely bonkers by todays standards as such temps are barely heatwave criteria anymore.

2012 is even more of a shocker for me considering how cool it was. It still managed 32.4°C at Cavendish in mid-August! 2021 there's a case to be made that pre-1990 it could've just barely scraped 30°C at Heathrow but it wouldn't be guarenteed.

43 minutes ago, WYorksWeather said:

1923

The 1923 figure is from the bull Camden Square station:

A quick check on Starling's Roost shows that this location could not have possibly reached 35.6°C on this day. I get the impression that Camden Square was in a location not quite up to today's standards and was unreasonably influenced by the urban heat island. It has a lot of these disparate figures so I tend to discount the station entirely.

With today's coverage a few 34s may have been seen in this heatwave but certainly not anything above 35°C.

While Trevor Harley is a decent starting point, I do have to question the validity of his data as there are some very unsubstantiated things on there. For instance, the claim that we went 14 consecutive days above 30°C in August 1947 - This appears to be a fake record as I can find no other reference to it on the internet, and a check of Paris' data shows that it is highly improbable that we could've seen a run like this in the UK during this heatwave. It had an impressive heatwave for the time (and is a very rare example of us having both a significant May/June heatwave and an August heatwave in the same year) and was also the driest year ever recorded in our history, at least since 1884, but it didn't quite reach this goalpost.

To get back to the original premise though, I calculated 2013-2023's average annual maximum to be around 34.4°C, so this is roughly the goalpost we can now expect half of all years to surpass. Vs something in the region of 32.5°C averaged over each decade before the 90s, you can now see how a sub-30°C maximum was once a roughly 90th percentile figure, but is now more in the region of the 99th percentile, which means we can now expect a sub-30°C annual maximum about only once every hundred years or so. This also gives you some context as to how rare 2007 and 2008 summers are, when they occured they were something like 1 in 40 year summers, now they'd be closer to around 1 in 80. By 2023-2033 we can expect that the average will by then be above 35°C, perhaps even knocking on the door of 36°C, if things continue as they are, in which case sub-30°C years would now statistically be more like once every 500 year summers and something like the 07/08 siblings now around 1/100 occurences. On the inverse 38°C would be around the 30th percentile which would mean we could expect a maximum of 38°C or higher every three years or so. Right now I'd say that's still more around one in five years, which is still much, much higher than the one in infinity it was pre-1990.

To put it simply, the chances of seeing a summer without reaching 30°C again in our lifetimes is very, very low, barring a volcanic winter or some other freak occurence.

 Greyhound81 Absolute hell, to put it simply. As it stands this country would completely and utterly fall apart if we got a modern version of 1976.

[CryoraptorA303]

 B87 Not really. I doubt we'd see 40°C as the real thing didn't break the 36.7°C record of the day and "only" achieved the third-hottest day on record at the time, but the average maximum would be extremely high. We'd probably be sitting at 34°C+ for close to two consecutive weeks with numerous jumps above 36°C, and the very hottest few days would reach around 38°C. Assuming an exactly timed repeat, August would also be an extremely hot month this time around as well, likely seeing another heatwave above 30°C for an extended period. It wouldn't reach anywhere near the same extremes but we're probably talking constant 30-32°C for over a week at least and maybe 33-34°C on the hottest day or two.

Give it another decade or two of warming though and those 38°Cs become 40°Cs and it's more like a consecutive month above 30°C with perhaps nearly three weeks worth of 34+°C days and a week of 38+°C days.

[B87]

 CryoraptorA303 I was thinking more of the averages. July and August 2022 both had average highs of 27c+ here, the only summer to do so. 1995 came close, but July just missed out. Every month in summer 1976 was above 25c.

London was above 31c for 16 consecutive days in 1976. I imagine these days we could probably get 3 weeks out of it.

Edited March 8 by B87

[cheeky_monkey]

models are calling summers over ..better luck  next year ..see you all June 2025 

[CryoraptorA303]

 B87 Oh right, I would guess both June and August would achieve a 28-29°C average and July around 26-27°C. Ironically assuming an exactly timed repeat July would see the very highest temps but both June and August would surpass it in the averages.

I would guess the best candidate to give us our next 40°C day is an August 2003 repeat. 21 years have now passed and we have likely already surpassed the threshold where an exact repeat would deliver 40°C. I calculated 41.7°C assuming 1990 and 2022 as analogues but this may be a bit of a liberal estimate and assumes completely ideal conditions and an exact repeat where Faversham is the record breaker again. Overall I don't see it not beating Coningsby when it does eventually happen, assuming Coningsby is still the record.

 cheeky_monkey Well this time last year they were seeing plumes left right and centre...

Edited March 8 by CryoraptorA303

[Daniel*]

If not for a notably warm June last year, second warmest at Heathrow after 1976, it would have been possible to have not seen 30C, neither July or August achieved it was the first summer since 2008 this happened in midst of hottest year on record on Earth. 

[B87]

 Daniel* I think we would have still seen 30c in September

[Metwatch]

If we don't manage a 30C somehow in the summer, more likely than not it will happen in September!

[B87]

Summer 2021 managed 30c, I think we will be alright.

[Methuselah]

 CryoraptorA303 What are you on about? All-time temperature-records are what they are. . . Your 'calculations' are neither here nor there! 

[raz.org.rain]

 cheeky_monkey I very highly doubt that'll happen.

[Daniel*]

10 minutes ago, B87 said:

think we would have still seen 30c in September

Well we did I’m just speaking for the summer. September was the most exceptional month of year.

 

[danm]

 Daniel* whilst not technically in Summer, the 7 consecutive days above 30c last September would discount that, even if June hadn’t reached 30c. 

18 minutes ago, B87 said:

Summer 2021 managed 30c, I think we will be alright.

Four consecutive days above 30c as it happens. Pretty good for such a poor Summer in the south. 

[CryoraptorA303]

 Methuselah I took the difference as a percentage between 1990 and 2022 and applied that to 2003. Not perfect methodology but it's as good as anyone can get in terms of predicting what a repeat will do now.

[WYorksWeather]

 CryoraptorA303 Yep, Trevor Harley's list is certainly not infallible. The further back you go in any case, the more uncertainty there is - slight differences in measurement, station siting etc. I would prefer however to err on the side of accepting some dubious older records, mainly just to avoid any accusations of 'cooking the books' or 'cooling the past', which often gets circulated.

And to be fair, Trevor Harley does state this in his disclaimer:
 

Quote

* Higher temperatures were reported in July 1911 but these are now considered unreliable; the recording conditions of all early records, inclusing those of 191 an 1932, do not compare with today's and might not be accurate. The August 1990 figure for Barbourne has been subject to scrutiny and might be unreliable - no other location that day reported a temperature over 36C. The 6 August Brent Pelham reading was out of line with neary stations and might not be accurate. Enfield recorded 36.9C on 9 August 2003, but the site was a rooftop and non-standard. The 2003 Faversham figure has been disputed, but it was definitely over 38 in the SE (38.1 Kew).

You're right about the 14 consecutive days over 30C in 1947 though as far as I can tell - can't find a source for that figure anywhere, so I assume that one is a mix up. Maybe an extra 1 got inserted - 4 consecutive days maybe?

In terms of averages - that 34.4C pretty much confirms what I thought - i.e. a warmer than average summer would now have to have an annual max of 35C.

For the odds of failing to reach 30C, I expect we could still do it, as it's not physically impossible, but it would require some ridiculously unlikely synoptics. You would need a summer almost entirely devoid of more than the briefest incidence of any wind direction between E and SSW in the peak warm period from around mid-June to early September, and probably very wet so that if we did get a couple of days of such influence, temperatures would be suppressed by the wet ground. It is possible, but it's so unlikely I don't expect to see it in my lifetime.

In terms of the likelihood of reaching a given annual maximum, with increasing mean annual maximum, the probability of reaching any maximum follows a logistic curve, which until passing 50%, grows exponentially. This explains the quite remarkable increase in some of the more extreme temperatures. 35C was once a 1 in 10 year event, and is now nearer 1 in 2.

For 38C, I'm not sure if I'd agree that it's 1 in 5, we have a distinct lack of data problem on that one as we've only had three such years. We won't necessarily get another period analogous to 2007-2012, but we can't necessarily rule it out. Even those summers brought forward by 15-20 years wouldn't have a prayer of achieving 38C. I think there is a lot of uncertainty on it. Highly unlikely to be lower than 1 in 10 I would say, and probably not higher than 1 in 3. The period up to 2030 will give us a good idea of which end we're at.

At the optimistic end, if it's nearer 1 in 10 at the moment, then I'd expect 38C to become a 1 in 2 chance by 2050 (the way we look at a 35C summer at the moment). At the pessimistic end, if the data up to 2030 shows that 38C is already 1 in 3, then we'd probably be looking at more like 3 in 4 chance of 38C by 2050 (probably the way we currently look at a summer reaching 33C, like 2023).

All predictions about temperatures above 35-36C are subject to a lot more uncertainty though, there aren't enough historic values to look at. We can make pretty clear conclusions about those, they're now close to 1 in 2 whereas before they were at most 1 in 10.

It's interesting to note the differences between UK (and indeed wider mid-latitude) temperature extremes and what a lot of models suggest. There are some papers on this that I've seen, so I think we will see modelling for mid to late century in the UK look much more aggressive when the climate models get updated with new observations. To give some context, the modelling states that reaching 40.3C was a 1 in 300-year event in the current climate - that seems absolutely hopelessly optimistic, and the modelling at the regional level on this is hopelessly out of whack with observations of other extreme heatwaves as well (the 2021 PNW heatwave in Canada was estimated at 1 in 1000 years using similar methodology). Essentially either those models are massively underestimating or we've just seen two extraordinary freak events within a couple of years of each other.

 cheeky_monkey Given the reliability of the long-range last year, I'll pencil in a Coningsby repeat!

[Addicks Fan 1981]

 Daniel* guess with this impending summer it'll hinge on how quickly la niña develops.   To be fair i dont believe these CFS charts that the other poster alluded us to, as they are always subject to change as you and i well know.   Reckon the consolation prize will be for the summer a west QBO if it materialises, also the solar cycle peak will help our summer a bit this year too.    

[CryoraptorA303]

1 minute ago, Addicks Fan 1981 said:

also the solar cycle peak will help our summer a bit this year too.

That has no impact on Earth temperatures and never has. It's an electromagnetic cycle anyway, the actual luminosity of the Sun changes so little that it's completely negligible to meteorology.

[WYorksWeather]

 CryoraptorA303 No impact is probably not quite true. But 'small compared to other factors' is - I think the amplitude is about 0.1-0.2C. About half of ENSO. There is some (admittedly very speculative) evidence that it may have some impact on regional weather that is somewhat larger than that, but the evidence is very weak and contested.

I wouldn't throw it out as complete junk, but on a list of factors affecting our summers, it'd be a fair way down the list.

[Addicks Fan 1981]

 WYorksWeather I'm looking forward to see what la niña does to our summer this year.   I can quite believe world climate service as they don't mince their words with the weather.   

Edited March 8 by Addicks Fan 1981

[Frigid]

 Daniel* Beat me to it, the July-August period last year was completely devoid of heat spikes and warm spells we've been so accustomed to. As with 2011, all the warmth that year were in the months you wouldn't expect them to be. 

[Metwatch]

Just now, Frigid said:

Beat me to it, the July-August period last year was completely devoid of heat spikes

July still scraped 30C on the 7th in Surrey. August I think reached 28C one of the days as well in Warwickshire might have been the 9th or 10th.

[raz.org.rain]

 Frigid 2023 was an anomaly in that respect, it's not something I'd expect a repeat of for a while. It's funny how we can have a string of decent summers and then July 2023 happens and people expect it to be more likely to happen every year.