Halloween Heatwaves

The temperature reached 22.9C in Kew Gardens on 29th October 2022, the highest in the UK so late in the year except for 23.6C on 31st October 2014. Around Halloween, what were previously once in many decades record high UK temperatures have, with global warming, occurred several times in a decade. For October overall, a mean Central England Temperature (CET) as high as that recorded in 2022 has (empirically) become around 30 times more likely since the 19th century.

Wearing my normal autumn attire of a T-shirt as well as a shirt I was hot and sweaty at around 11 in the morning on Saturday, 29th October, as I took my usual walk from Acton to Ealing. It felt more like a normal summer day than one in late autumn.

So it was no surprise to see on the Met Office extreme observations page that it reached 22.9C at the weather station in Kew Gardens, walking distance away, across the Thames. As often seems to be the case, West London was the hottest place in the UK.

It turns out that 22.9C on 29th October is not quite a “date record”, the term I’ve coined for the hottest (or coldest) temperature so late or early in the year. But only because 31st October 2014 reached 23.6C, also at Kew Gardens, as noted in a New Scientist article lamenting the changing of autumn.

And that’s not even all. 22C had not been exceeded after 26th October at the time of publication (2007) of The Wrong Kind of Snow (TWS). Since then temperatures over 22C have been recorded 4 times later in the year:

29th October 2022Kew Gardens22.9C
31st October 2015Kew Gardens23.6C
1st November 2014Trawsgoed (Ceredigion)22.4C
2nd November 2014Trawsgoed (Ceredigion)22.3C
Fig 1 Recent extreme UK daily maximum temperatures around Halloween (personal notes from Met Office data, corroborated by TORRO website)

This is crazy. There’s a limited amount of data, of course, but the sorts of temperatures that used to occur at this time of year only once a century are now being experienced more than once a decade, maybe 20 times as often. We can only guess what the current once in a century Halloween period event would be now: 25C? I speculate, but judge for yourself how much more often UK Halloween heatwaves are occurring:

DateRecord in TWS (C)YearLocationNew daily record (if any) (C)YearLocation
23rd October21.41996Red Wharf Bay
24th October21.71996Prestatyn
25th October20.21949Jersey
26th October22.81969Southsea
27th October20.31888Old Street21.32022Kew Gardens
28th October20.61958Huddersfield
29th October19.21984Coltishall22.92022Kew Gardens
30th October19.4*1927Reading
31st October19.41968Margate23.62014Kew Gardens
1st November19.71971Hawarden Bridge22.42015Trawsgoed, Ceredigion
2nd November19.41948Prestatyn22.32015Trawsgoed, Ceredigion
3rd November20.61946Aber
4th November21.71946Prestatyn
5th November21.11938Chelmsford
Fig 2 Daily maximum temperature records from 23rd October to 5th November, until 2007 (left-hand columns) and since 2007 (right-hand columns). Date records (hottest so late in year) in bold (the 22.8 on 26th October 1969 has been superceded by 23.6 on 31st October 2015, of course). *Oddly, the TORRO website notes higher temperatures on 30th October in 2001 (21.0C at L’Ancresse, Guernsey), 2005 (20.8C at Writtle, Essex) and 2009 (20.5C at Tregarth, Gwynedd)!

The old daily record of 19.4C for Halloween itself, 31st October, was also exceeded in 2015, by 0.1C (19.5C was recorded at Hampton Water Works, according to my records).

Warmer autumns are, of course, disruptive for ecosystems, as pointed out by the Guardian this weekend. “October’s summery temperatures are ‘confusing’ plants and throwing off fragile ecosystems”, they write. Indeed. The New Scientist article from 2015 (mentioned already) suggested we are seeing nothing less than the “[f]all of fall” and that “(t)he mellow season of autumn is changing forever”.

Of course, one hot day doesn’t have a huge effect. The UK heatwave of autumn 2022 has been quite prolonged, though:

DateMaximum temperature (C)Location
18th October20.0Wiggonholt
19th October21.0Merryfield & Yeovilton
20th October19.3Chivenor
21st October19.6Shoeburyness
22nd October19.4Writtle
23rd October19.9Wiggonholt
24th October18.7Heathrow
25th October18.4Northolt
26th October20.5St James Park
27th October21.3Kew Gardens
28th October19.9Santon Downham
29th October22.9Kew Gardens
30th October18.1Kinlochewe
31st October18.0Northolt
Fig 3 Maximum daily UK temperatures from 18th-30th October 2022, from Met Office Twitter feed. Earlier dates not checked. Daily records highlighted in bold.

Overall, October 2022 has been nearly 2C warmer than average:

Fig 4 CET mean temperatures, year to date, Met Office

By my reckoning October 2022 is on course to be around the 13th hottest October in the 350+ years of the CET record:

RankYearCET mean October temperature (C)
13=202212.4 (provisional) [See note at the end of the post]
Fig 5 Octobers with mean CET greater than or equal to 12 C, from Met Office data [See note at the end of the post]

There are different ways of looking at the list in Fig 5, of course, but we could observe that only one October as warm as that in 2022 occurred in the 19th century (1831). Such an event became 5 times more likely in the 20th century and, given no fewer than 8 in less than a quarter of the 21st century is now (empirically) around 30 times more likely than in the 19th century.

And note that the only October with a CET mean temperature of 12C or more in the 140 years for which there are reliable records before the 19th century was 1731, which was not quite as warm as 2022 (provisionally).

Here’s a graph showing how much warmer Octobers have become over the last century or so:

Fig 6 October CET mean temperatures since 1660, with 5, 11 and 21 year running means (graph by the author from Met Office data)

Not a week to deny the existence of climate change on national TV, you’d have thought.


Note (2/11/22): It continued to be warm in the UK until the end of October, so, rather than the 12.4C discussed in this post, the final mean CET temperature for the month was even warmer, 12.8C. For complicated reasons I won’t go into right now, this will probably eventually need to be adjusted down by 0.1 or 0.2C, but, even if that happens, we’ve just experienced one of the top 10 warmest Octobers in the CET, not the =13th warmest as suggested in Fig 5 (which I did tell you was provisional!).