Met Office’s Hottest Summer Claims Disproved By CET


By Paul Homewood


The Met Office has now officially declared this summer as the joint hottest on record in the UK:



Update: Having further assessed the temperature data for the UK as a whole for summer 2018 the figures are so close that we are declaring it as the joint hottest on record together with 2006, 2003 and 1976.


They also confirm that it was the hottest in England, with a mean temperature of 17.16C, versus 17.01C in 1976.

As I pointed out a few days ago, this all seems very strange, because the CET only ranks this summer as 5th warmest, 0.50C less than in 1976, and not even as hot as 1826:


CET Average Mean Temperatures


Of course, Central England is not the UK, nor even England. Nevertheless, as Scotland, Wales, N Ireland and the north of England were nowhere near record temperatures, there appeared to be something seriously wrong with the Met Office’s figures.

Now that the Met Office has published its data for August, I have concrete evidence of this.

As well as the national numbers, the Met Office also publish temperature data for the districts. One of these is the Midlands, which closely matches the area defined in the CET:

Shows the standard areas (districts) used by the Met Office when generating climatologies.


Just as in the national numbers for England, the Met Office say that this summer was the hottest on record, 17.27C against 17.05C for 1976.

In other words, an even bigger margin than in England as a whole:




To recap, the CET puts 1976 as 0.50C warmer, but the Met Office say this summer was 0.22C hotter. A discrepancy of 0.72C.

It is utterly inconceivable that the two methods could legitimately come up with such a big difference. There has to be something seriously wrong with one or the other. [BTW- it has nothing to do with UHI adjustments, as the CET adjusts all temperatures since 1974 by 0.2C, to allow for UHI – in other words, 1976 and 2018 have been treated in the same way]

And the discrepancy is also well outside any statistical margin of error – the CET give this as between 0.09C and 0.2C. The Met Office don’t give official margins of error, but suggest it is well below 0.1C.

We know that an enormous amount of work has gone into building up, checking, analysing and homogenising the CET series over the years, which rightfully has earned it a worldwide reputation.

Unfortunately, the same cannot be said of the Met Office’s UK dataset beginning in 1910. For instance, it includes all sorts of inappropriate sites, such as Heathrow, a car park in Motherwell, many urban locations, and Faversham, which regularly records the highest temperatures ( as it did again this summer) but has only been operational since 1998 and has been condemned as unreliable.

[The Met Office confirmed to me the locations they use, and they can be seen here]

There also appears to be little transparency as to how the Met Office build up their UK dataset. How, for instance, do they allow for changing mix of stations? Or allow for UHI? Or cater for missing stations?


At the moment, the credibility of their UK dataset is totally shredded. The Met Office should immediately withdraw all claims about “hottest summers”, and explain why the two datasets diverge so wildly.


September 5, 2018 at 01:12PM

Leave a Reply

Fill in your details below or click an icon to log in: Logo

You are commenting using your account. Log Out /  Change )

Google photo

You are commenting using your Google account. Log Out /  Change )

Twitter picture

You are commenting using your Twitter account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )

Connecting to %s

%d bloggers like this: