CET temperature trends

UK temperatures essentially show no obvious warming for July over the last  365 years. There has however been a small warming effect in  January since 1970, which may just be related to rapid population growth.


Comparison of CET values for the coldest month (January) and the warmest month (July) each year with a) England’s population growth and b) CO2 increases.

The Central England Temperature (CET) measurements is the longest continuous dataset in the world and covers a 360 year period from 1658 to 2016. Weather stations are concentrated within a roughly triangular area ranging from  Lancashire, Bristol and London. The results are usually presented as annual temperature ‘anomalies’ relative to a 30-year seasonal average as for example by Ed Hawkins. However we can also simply plot the average measured temperatures for each month to identify trends. The resultant figure, as shown above, demonstrates that for CET there is no obvious trend in July temperatures at all. Perhaps after 1980 July temperatures have been slightly warmer than the long term 10y average, but well within annual variability. For the average temperatures  in January the year to year variability is much larger still. Extreme cold winters occur throughout the CET record. In my lifetime I remember 1963(average -2C), 1979 (-0.5C) and 2010(+1C) standing out as being particularly cold, while the record coldest January  is held by 1795 (-3C). The 10 yearly January average shows two recent warm periods lasting firstly from 1890 to 1940 and secondly from 1970 to 2016.

Is there really any  evidence for anthropogenic global warming in CET? I think you would be hard pushed on the basis of this data to claim that CO2 alone is responsible for recent trends. If the second increase in 1970 to 2016 UK winter temperatures does indeed have an anthropogenic cause, then it is just as likely caused by rapid population growth as it is by CO2. Today some 50 million people in England warm their homes every January driving ~30 million vehicles on well lit roads in warm cities. That massive population growth has also mostly been concentrated within the CET triangle.

This entry was posted in climate science, UK Met Office and tagged , , . Bookmark the permalink.

11 Responses to CET temperature trends

  1. erl happ says:

    Hi Clive,
    If there are daily minimum and maximum temperatures its worth plotting them separately. You may well find that most of the fluctuation is in the daily maximum and little in the minimum suggesting that it is cloud cover during the day that is responsible for change.

    You can dismiss the supposed greenhouse effect. The Southern Hemisphere exhibits no warming at all in the month of December over the last seven decades. I don’t think it plausible to suggest that the greenhouse effect takes a holiday in December. I am not aware of any countervailing force that could reverse the impact of the supposed greenhouse effect.
    Here are three ways to look at the data to verify this assertion:
    2. ?w=612
    3. ?w=612

    The last looks at the departures from the whole of period average for each of the months of the year.

    With a mean of plus 3°C I am sure that you will agree that warmer winters are desirable in mid and high latitudes.The increasing trend in winter is consistent with known modes of climate change of natural origin due to change in the planetary winds ie the Arctic Oscillation.

    • Clive Best says:

      Yes any warming does seem to affect minimum temperatures rather than maximum temperatures. However regarding CET, I checked and the average temperatures they publish are exactly equal to (Max-Min)/2. Here is what I get

      However you look at the data there is very little to worry about any possible warming occurring in England !

  2. Cytokinin says:

    Almost as flat line as a dead man. Being an oceanic climate England is probably a good indicator of work tends since 2/3 of earth’s surface is water. Don’t spoil the party though, I think i have figured out how to sequester all that nasty CO2 and it you butter the bubble, I’ll never be able to sell any units.

  3. Javier says:

    I have data for Spain that shows no warming at all and no change in precipitations for the past 20 years. Apparently the US doesn’t show warming either. Global data says most of global warming is taking place over land in the NH. That claim however is not supported by national data. It is so elusive that it seems to be taking place where it is not being measured. Thankfully the adjustments to the data are making it very evident, because otherwise we could miss it and don’t know that the world is warming. Like the proverbial frogs brought to slow boiling.

    • catweazle666 says:

      Anthropogenic Global Warming invariably takes place where there are no thermometers to measure it and can only be revealed by statistical techniques such as Kriging AKA Making Stuff Up.

      That is axiomatic.

      • JonA says:

        Indeed. Regardless of what you think of Tony Heller, (a pet peeve
        of mine is dismissing an argument due to provenance, not content)
        he recently posted some interesting NOAA data for southern Africa
        – where Namibia and Angola had, apparently, experienced record
        temperatures. Turns out there is hardly any thermometers in the
        region at all. IIRC, only one (at the airport, obviously) in Namibia.

  4. Joe Brannan says:


    Your correlation with population is interesting and reminds me of my favourite bugbear in the temperature anomaly debate. Routine weather forecasting throughout the planet never fails to remind us that it will be colder in the countryside than in the towns, especially overnight. Yet professional meteorologists have been unable to detect an Urban Heat Island effect in their data. I cannot square this circle. Have you ever looked into this?

    • Clive Best says:

      The Urban Heat Island effect is clearly visible in measured weather station temperature data. It is common experience that cities are around 2C warmer than the local countryside. However the trick is that temperature ‘anomalies’ hide this effect in a very subtle way because temperature ‘anomalies’ are calculated relative to a reference seasonal monthly average, usually chosen to be relatively recent like 1971-2000. Large urban areas were already far warmer then, so as a result the early ‘anomalies’ are cooler. This skews the land temperature record and partly explains why the late 19th century and early 20th century appear to be colder than now. Probably half the apparent warming on land is due to this artificial renormalisation.

  5. charplum says:


    It has been a while since I looked at CET. The monthly data are very noisy and I was not able to get a cyclic analysis that I thought was worthwhile. However, the yearly data, although noisy still, allowed a picture to emerge.


    The correlation coefficient for the cyclic fit is 0.72. I usually do better but these data are noisy.

    What I do find interesting, and perhaps you will too, is that I did find a cycle of 18.349 years. This is very close to the 18.6 year cycle you have identified. You may recall I found this cycle in the ocean heat content not long ago and that the ocean area encompasses the region where it might well influence the CET either from the ocean or from its influence on the jet stream.

    It does require some judgment but considering how noise these data are I consider it close enough.

    I like it when things work this way. You furnish a physical explanation of things and then through investigation of the raw data we can discern evidence of its presence.

    18.349 years is close. Beauty is in the eye of the beerholder.

  6. charplum says:

    Perhaps, I should have included the map of the region I used for the heat content. I think this easily explains why it might influence CET.


    Maybe that was clear to you but this really makes it clear.

Leave a Reply