Skip to content

Strange Happenings At CET!

July 3, 2018

By Paul Homewood


Yesterday I reported on the CET figures for June, which showed a mean temperature of 16.5C, making it the 10th warmest on record.


It has been pointed out today that the Met Office are now reporting a figure of 16.1C, which puts it into a tie for 18th place.




As the monthly figure was not published when I posted, I simply took the daily figures and averaged.

But the peculiar thing is that the Met Office was already reporting 16.4C on the 29th. The final day of the month was 19.3C, so that figure could only go up:



Figures are always provisional during the month, but I have never known them to be more than 0.1C out.

Whether the original numbers were in error, or whether the latest version is wrong, we will have to wait and see.

But, assuming the latest figures are right, here is the corrected chart:






I have just reworked the average from the daily totals, here, and it still comes to 16.5C!

I have emailed Tim Legg, who maintains the CET series for comment.

  1. Up2snuff permalink
    July 3, 2018 11:06 am

    Must be a bit like GDP figures, then. 😉 Adjusted. Ahem!

    • Mike Jackson permalink
      July 3, 2018 12:53 pm

      Absolutely. There is no point in having too big an anomaly which will need to be adjusted down in five years time to avoid June 2023 being out of whatever kilter it is supposed to be in by then. Better to do the adjustment now when nobody’s looking!

      Except that that pestiferous Homewood is! Curses! Foiled again!

  2. Adrian permalink
    July 3, 2018 11:16 am

    There you go again Paul, using data assuming they are actual measurements. It shows you sad naivety.

    As any climate scientist can tell you actual measurements are merely a starting point and these will need adjustment according to theory, fitting in with research funding, and a range of other non-scientific reasons. 2+2 really does equal 5, or 3 or better still 3.2046 if pre 1980 and 5.811 if post.

    If only you knew these things Paul you would produce a more informative, keep trying tho’.

  3. NeilC permalink
    July 3, 2018 11:45 am

    From my records, I have just looked at Manchester, Birmingham, Bristol and Luton (nearest ones I have to CET) and that confirms what you said at 16.5 Deg C.

    Probably an error by UKMO

  4. Jack Broughton permalink
    July 3, 2018 11:50 am

    An amusing / interesting bit of statistical approach to the CET data is that the mean for each month has an average standard deviation of 1.33 deg K. Thus any anomaly below 2.66 deg K has a 95% probability of occurring naturally, none of the last months exceed that value, so our weather is well within its normal unexciting but benevolent range.

  5. Derek Walton permalink
    July 3, 2018 12:12 pm

    Clearly the have forgotten to “Apply a VERY ARTIFICAL correction for decline”

  6. quaesoveritas permalink
    July 3, 2018 12:56 pm

    There are quite often quite significant differences between the daily figures and the finally published monthly figure.
    I think the monthly figure is almost always lower, because the daily figures only include the first minimum of the day, i,e, the one before midnight. If the one after midnight is lower it reduces the eventually confirmed daily and monthly figures.

    • quaesoveritas permalink
      July 3, 2018 1:50 pm

      On reflection, I think I may have got that the wrong way round.
      I know its something to do with the number of minimum readings, as stated on the MO website:

      “Note that the estimated mean value may not be equal to (max+min)/2 because there is always one more minimum
      value than maximum. This discrepancy is exaggerated during the first part of the month and when the nights have
      suddenly become significantly colder or warmer. “

  7. MrGrimNasty permalink
    July 3, 2018 12:57 pm

    Sorry for the extra work – await Mr Legg’s response with interest.

    My gut says the higher figure is more likely and someone’s finger slipped.

  8. July 3, 2018 2:17 pm

    Reblogged this on Climate Collections.

  9. July 3, 2018 2:29 pm

    The Met Office cannot be wrong because its Quality Management System conforms to BS EN ISO 9001 which would prevent errors creeping in. Everything the Met Office does is of the highest possible scientific standard – its computer models are fully verified and validated. Oops!

    • Joe Public permalink
      July 3, 2018 5:10 pm

      ” … its computer models are fully verified and validated. Oops!”

      Justification methinks, for a newer, more powerful computer?

  10. Glyn Palmer permalink
    July 3, 2018 9:00 pm

    A variation on ‘Mike’s Nature trick’?

  11. styleyem permalink
    July 4, 2018 10:32 am

    Comment from the Netweather monthly CET forum suggesting different stations are used:

    “You do realise that it’s a different set of stations for the actual CET as opposed to the publicised rolling CET for the month? The correction tends to be downwards but there is no way of knowing how much, just guesswork.”

    • quaesoveritas permalink
      July 4, 2018 10:40 am

      I don’t understand.
      AFAIK, there are only 3 stations used for CET and they are the same for the daily figures and the monthly ones.
      As I said above I think it is just the missing minimums for the daily figures which makes the difference.
      Which are the “different set of stations”?

      • quaesoveritas permalink
        July 4, 2018 10:58 am

        Looking at the context of that quote, I think it is referring to Hadley Centre regional temperature for Central England, which I think is a completely different set of temperature data.

  12. Athelstan permalink
    July 4, 2018 10:43 am

    Eye on the ball and figures too, well done Paul.

Comments are closed.

%d bloggers like this: