Skip to content

England Rainfall Series

December 22, 2020

By Paul Homewood


As promised, more charts on daily rainfall at long running English sites.

The first series plots the highest daily rainfall for each year. I have annotated the highest for each site:










It’s all rather a mixed bag. Interestingly the series highs all fall in different years at each site. Bradford, and to a lesser extent Buxton, are the only stations with noticeably heavier rainfall in the last two decades.

In contrast the southern stations of Oxford, Eastbourne and Yeovil clearly show the opposite.

What we probably can say at every site is that there has been no daily rainfall in the last decade which has been in any way exceptional or unprecedented. Even at Bradford, the event which really stands out after 1986 was back in June 2007.

Increasingly years such as 2000 can and should now be regarded as weather of the past, and not “today’s climate”.


Moving on, the second series of charts plot the number of days each year with 20mm or more:




bux vh






The outliers at Bradford and Yeovilton are in 2012. But overall there seems to be little cause for concern at any site.


European Climate Assessment & Dataset:

  1. jack broughton permalink
    December 22, 2020 11:59 am

    The “scientists and climate experts” keep repeating the mantra that the weather is getting wilder and more damaging. These charts show very clearly that the standard deviation for rainfall is, if anything, slightly reducing in recent years.

    Of course if one relies entirely on the outputs of GIGO models, one will get different conclusions. I would hope that a jury would adopt the history rather than the prognostications of dubious science.

    • Harry Passfield permalink
      December 22, 2020 12:28 pm

      Of course, maybe the GIGO models are using the same SW as the Dominion voting Machines – who’s SW was developed……in Venezuela, of all places.

  2. Christopher Hall permalink
    December 22, 2020 12:20 pm

    It is interesting to see the evidence that there is no upward trend in high rainfall days. However, purely from personal observation, the incidence of very intense rain does seem higher. Huge drops, overwhelming car wipers seem to be more common. Previously I had only seen rain like this in the tropics. Do I worry? As anthropomorphic change skeptic, no, but what do others think?

  3. Beagle permalink
    December 22, 2020 12:41 pm

    I can see the headlines for Thursday morning “Highest rainfall ever yesterday”. Yes there is some heavy rain expected tomorrow and I bet they have their headlines already written.

  4. David Woodcock permalink
    December 22, 2020 3:24 pm

    Despite the Met Office trying over and over again to sell the story to a dumb media that extreme rainfall events are more common now than in the past due to man made climate change, it’s not true, not a word of it. It’s pure hype and nonsense because the data just doesn’t support the rhetoric. This is why they have thrown more emphasis on extreme flooding events. However an increase in flooding is not due to more extreme weather either. Flooding in most cases is clearly down to increased building and development on brownfield sites in flood plains. A practice which has been forced on local authorities due to tight greenbelt restrictions. In the past, the animals grazed in flood plains where the best grassland exists and homes were built on higher ground for obvious reasons. However, today planners have reversed all of that and thus we have an increased incidence of homes and businesses flooding while the cows look down on the carnage from the high ground with impunity. Even the flooding across the Summerset plains was down to a deliberate lack of drainage clearance to allow water levels to rise to increase wetland habitat for birds.

  5. ThinkingScientist permalink
    December 22, 2020 4:43 pm

    Rainfall is likely a random distribution across the country. To detect increases in rainfall or extreme you would need a network of the same locations continuously recording for 100 yrs+. Note the same stations are required. The key are the set of locations need to be fixed.

    I suspect the reason that the Met Office is so fixated on extreme weather events is that you can get a higher frequency of such events simply by adding more stations. Also you could get new records by selecting new stations in areas likely to have high rainfall. It then appears to be a “record” because its not been observed before (note that is NOT the same as it not happening before).

    I suspect rainfall follows a fat-tailed distribution such as Levy Stable. I worked with an expert group about 25 yrs ago who were adapting Levy Stable models for seismic inversion. Their expertise came from hydrology, so these fat-tail distributions are well known in the relevant literature, with very large random values occurring with higher frequency than distributions such as Gaussian.

    • chriskshaw permalink
      December 22, 2020 4:58 pm

      Sounds very much like global temp increases which are similarly assigned. Crazy to think they can add rainfall to the same trickery.

      • ThinkingScientist permalink
        December 22, 2020 8:34 pm

        Rainfall is much more difficult to “adjust” to get trends. Hence the relentless search for “extremes”.

        Back around 2009 – 2011 when we had the run of three bad winters, the Met Office went on record as saying that these winters had a 1/20 chance of occurring and there was no dependence from one year to the next.

        So three in a row is therefore a 1/8000 probability, so must have been an almost unique event during the entire Holocene? The claimed statistics are obvious nonsense and even worse when fabricated through climate models.

        Similarly they used to publish long range seasonal forecasts. The famous one was the “BBQ summer” that was a a washout, but the best was when they gave probabilities of temps:

        30% below average
        35% average
        35% above average

        Which is so close to a uniform distribution that it is not a prediction at all. After all I can “predict” the outcome of tossing a coin as:

        50% heads
        50% tails

        Which is simply random and has no predictive value.

  6. ThinkingScientist permalink
    December 22, 2020 8:47 pm

    Paul H,

    regarding record lengths, I think the Met Office is not providing earlier data even though they have it.

    I went to the KNMI page and checked out the rainfall data for Newton Rigg in Cumbria. The series starts in 1961. At the Met Office the data also start 1961, so that is consistent.

    I have a spreadsheet copy a colleague downloaded in the past (20 years ago? Provenance uncertain) with a continuous monthly rainfall series for Newton Rigg from 1891 to 1999. The numbers are similar but not the same as currently published post-1961.

    Is the Met Office unaware of the earlier data? Lost it? Revised it? Simply don’t publish it?

    • ThinkingScientist permalink
      December 22, 2020 8:53 pm

      Erratum – the set I have commences 1895, not 1891. Where they overlap 1961- 1999, they are correlated (R=0.83) but the more recent Met Office version is 58 mm greater annual rainfall on average.

    • December 22, 2020 9:53 pm

      I think their excuse is that prior to 1961 it was not “digitalised”!

      In other words, just ignore the inconvenient data

    • ThinkingScientist permalink
      December 22, 2020 10:42 pm

      The set I have appears to be from a digital table that’s been pasted from a page.

      The peak years for rainfall in the Newton Rigg series are 1903 and 1928. Here’s the top 25 ranking:

      Rank Yr Annual (mm)
      1 1928 1318.8
      2 1903 1291.9
      3 2015 1222.2
      4 2000 1198.8
      5 2002 1154.3
      6 1954 1133.3
      7 1927 1112.1
      8 1916 1109.7
      9 1923 1108.8
      10 1918 1104.1
      11 2012 1099.6
      12 1934 1096.3
      13 1912 1089
      14 2008 1085.8
      15 2014 1079
      16 1967 1070.5
      17 1924 1053.6
      18 1931 1052.1
      19 2009 1048.8
      20 1899 1047.2
      21 1914 1042.3
      22 1968 1039.7
      23 2006 1037.9
      24 1990 1030.3
      25 1921 1024.5

      Notice something? If you only use post-1961 then the year’s rank as follows:

      1 2015
      2 2000
      3 2002
      4 2012
      5 2008
      6 2014
      7 1967
      8 2009
      9 1968
      10 2006

      With the accompanying headline “the top 6 years rainfall at Newton Rigg were all in the 2000s and 8 of the top 10 years are in the 2000s”.

      Whereas 7 of the original top 10 are pre-1961.

Comments are closed.

%d bloggers like this: