I decided to find a way of quantifying heatwave magnitude and length to make it easier to compare events, so I've designed a very simple scoring method: total max temp degree days above 27C for the duration of a spell with consecutive days of 28C or more. Ie 28C scores 1, 30C scores 3 and so on upwards. 

Weatheronline has some very helpful historical daily data so I've applied this formula to Heathrow to see how events rank. Unfortunately it goes back only to 1981 which means I can't compare with the legendary June-July heatwave of 1976. EDIT: 1975 and 6 data kindly provided below.

1. 23 Jun-8 Jul 1976 (88)


2. 3-13 Aug 2003 (58)


3. 2-8 Aug 1975 (41)


4. 16-22 Jul 2006 (37)


5. 29 Jul-3 Aug 1995 (29)


6. 11-18 Jul 1983 (29) 


7. 1-4 Aug 1990 (28) 


8. 20-25 Jul 1989 (25)


9. 13-19 Jul 2013 (24)


10. 24-28 Jul 2006 (20)


11. 15-22 Aug 1995 (19)


12. 29 Jul-2 Aug 1999 (19)


13. 30 Jun-4 Jul 2006 (18)


14. 8-12 Aug 1997 (14)


15. 21-25 Jul 1994 (12)


16. 22-24 Jul 2013 (11)


17. 13-15 Jul 2003 (10)

Based on GFS 12z low res max temps for the Heathrow area next week, and adding 1C each day to take into account under-prediction, we are on course to score 18, which would put this just outside the top 10 and on a par with the third biggest heatwave of 2006. But if GFS underestimates by 2 rather than 1C, we'd end up in 23 just behind the Jul 2013 heatwave.

If anyone has the 1976 daily numbers for Heathrow (or somewhere comparable) let me know. It would be great to compare with 2003, which is so far out in front of the others here.

http://en.tutiempo.net/m/climate/06-1976/ws-37720.html

Love this heatwave scoring system, superb job you have done here and it will be interesting to see how this heatwave fairs, it actually looks a little disappointing that it might not even be in the top ten given all the hype. 

Meteo France have done something similar for their summer heatwaves.

Those are some interesting results. However any definition you chose for 'heatwave' is inevitably going to be somewhat arbitrary, as I'm sure you realised when you considered July 2006. If only July 23rd had been a little hotter then there would have in fact been a tie for the top spot instead of one heatwave being head and shoulders above the rest. No doubt in 2006 it felt like one long heatwave rather than two shorter ones in very quick succession. But of course you have to define it somehow and apply the same rules to everything.

Interesting thread...

I did think about this, but I wanted to differentiate between a heatwave, which is a defined and relentless period of hot weather, and a hot month or summer. The two are different. If there is a day or two of "relief" before the hot air comes back, then to my mind this is the end of one heatwave and the beginning of another. But as you say, the 2006 example tests the logic more than most.

I also thought about including night minima, but it was difficult to find the right threshold. What I might do, though, is build in double points for weekends and bank holidays. That would hurt our upcoming heatwave.

Latest output still shows this week scoring something around the 18-22 mark, so despite the wrist slitting on the model thread it's still up there.

As of today's 06z GFS, we are looking at London maxima from Tuesday onwards of 29, 34, 30, 33, 31 before it cools down on Sunday. That would mean a creditable but mid-table score of 20.

Meanwhile central-Eastern France (Macon) could well see, starting last Thursday: 28, 30, 28, 30, 31 [EDIT: 32 confirmed], 35, 37, 38, 40, 38, 35, 38, 39, 38, 33 before a moderate cool down on the 10th. That would score it a handy 113 on the index. The 2003 canicule delivered a score of about 150 to the same location. So not far behind.

Hey Thought I would do the same thing for the daily CET series, except this time for every 0.5 degrees over a daily mean of 19.0C or higher, 1 point is awarded (as average daily mean is obviously likely to be between minima and maxima). Since 1772 these are the 25 greatest heatwaves:

1) 23rd June - 6th July 1976 - 144.8               14) 15th to 22nd August 1995 - 49.6

2) 6th - 23rd August 1997 - 102.2                  15)  4th to 11th July 1852 - 49.2

3) 16th to 30th July 2006 - 96.2                     16) 12th to 19th August 1893 - 48.8

4) 5th to 18th July 1983 - 84.4                       17) 12th to 17th July 1876 - 38.2

5) 25th July to 4th August 1995 - 83.6            18) 16th to 20th July 1921 - 37.2

6) 4th to 13th August 2003 - 73.8                   19) 10th to 13th July 1923 - 37.0

7) 26th June to 9th July 1826 - 70.6                20) 2nd to 10th August 2004 - 36.8

8) 12th to 19th July 1808 - 66.2                      21) 1st to 4th August 1990 - 36.2

9) 1st to 9th August 1975 - 65.8                      22) 1st to 6th July 2006 - 36.2

10) 12th to 20th July 1825 - 59.0                     23) 17th to 21st July 1901 - 35.8

11) 13th to 23rd August 1947 - 54.4                 24) 10th to 16th July 1847 - 35.6

12) 26th to 31st July 1948 - 53.4                      25) 26th to 30th August 1930 - 35.4

13) 20th to 26th July 1989 - 50.4

So yup, 1976 is well and untruly king. However a few periods on here have 1 day that dips below the threshold in a heatwave meaning they could have scored higher, August 1893 and 1975 being the best examples.

Some months were intermittently warm throughout meaning they aren't included here (July 1783 most notably), but the difference with Tim's data is that this obviously gives additional weighting to heatwaves with high minima, 1997 being the best example. Although I remember in NE England how exceptional that August was and it still holds the record for the hottest month on record in the NE.

Quite interesting overall and the heatwaves have become far more common since the mid 1970's.

This doesn't include periods outside the summer months but I can only imagine Spetember 1911 standing a chance of making it onto this list.

Fantastic post Yorkshire Lad.  Really insightful.  In almost any way of objective measurement, the 1976 heatwave is simply the king.  It is hard to conceive of a more stable and sustainable event for the UK to generate heat. 

Yes great stuff. Well done for taking the time to pull this together. As well as considering the hottest and most sustained individual heatwaves it is also worth considering individual summers as a whole. Two years appear twice on that list - 1995 and 2006. If you combine the figures for individual years to consider all summer heatwaves the top 3 become:

1976 - 144.8

1995 - 133.2

2006 - 132.4

I remember the latter two years well and they both stick out in the memory. Unfortunately I was only one year old in 1976 so have to rely on the stories of my parents for that year.

Another look at the figures... July 1808 cannot have been far away from August 2003 surely as most heatwaves in the top ten lasted two weeks, this only lasted 7 days! July 1948 also looks to have been impressive and contains the highest every CET daily mean.

Apparently the thunderstorms during that heatwave were responsible for the largest hailstones ever recorded in the UK, as you go further back in time though I can imagine the chances of the temperature being recorded in direct sunlight increase. Therefore the absolute maximums may be a bit too high but apparently temperatures got up to 35C in Hull at the peak of the heatwave and was strongest in Eastern England. Must have been some strange synoptics for that one.

I can't think of any other heatwave that could possibly top 1976, 1707 would have probably appeared high on the list from what I've read and I heard 1540 and 1556 where apparently exceptional summers, but there you go.

July 1783 could easily have been right at the top but this was a very consistent spell of warmth without that much of a standard deviation, could be some unusual weather pattern from Katla erupting.

It seems that really notable heatwaves now seem to occur in clusters every six years. Before the 1970s they were very rare.

July 1923 is another notable one, only 4 days and it still made it into the top 25. 

Here are the top 30 tallied up just for completion, 1995 has the highest heatwave index, surprised to see 2001 so high up.