Met Office temperature records – what do we monitor and how far do they go back?

May saw more than its normal share of fine sunny days and notably high temperatures, reaching a peak of 28.7 Celsius at Northolt on 7th May to coincide with the early May bank holiday. That was just one place on one day. To get a broader picture we also look at the average temperatures across the whole month and all of the UK. The average daily maximum temperature in May was 17.2 Celsius. The persistent sun and warmth meant that this was the highest average maximum temperature on record for May.

But what does “highest maximum temperature for May” mean and how is it calculated? What types of temperature do we monitor, and how do they differ?

Daily Max Temp May

Average of daily maximum temperatures for May.  Setting a new record high for the national series.


Maximum, Minimum and Mean temperatures

All our standard temperature observations are made from the UK’s weather station network. Modern automated technology means we can actually track changes in temperature minute-by-minute. It hasn’t always been this way though, and historically it was not very practical to manually record the weather every minute of every day. Thankfully, British scientist James Six invented a maximum-minimum thermometer in the late 18th Century that could record the highest and lowest temperature recorded over a period of time. This invention allowed early weather observers to summarise the temperature of the day based on these two extremes, and the average of the maximum and minimum is widely used as an estimate of the mean temperature. This practice continues to this day, providing the UK with some very long records of the highs and lows of temperature across the country.

Daily Min Temp May

Average of daily minimum temperatures for May.  Ranked joint 10th warmest in the national series.

Station Records

The Met Office has routinely provided summaries of the monthly average of the maximum, minimum and mean temperature for the weather stations located throughout the UK since 1884. For a few stations we hold records going back much further. These station records provide an accurate picture of the month-to-month and year-to-year variations in temperature at particular locations. Our present-day observing network is comprised of over 400 stations.

Weather Observation network

Map showing the present-day weather observation network of automatic (blue) and manual (red) stations.

Find a selection of our historic station series here.

Central England Temperature records

The Central England Temperature (CET) series is the remarkable achievement of the scientist Gordon Manley. In the mid 20th Century Gordon Manley undertook painstaking research to uncover old weather records and diaries stretching back to the mid 17th Century and from these created the Central England Temperature series, providing an estimate of the monthly mean temperature of a region broadly representative of central England.

This work was continued by the Met Office to maintain a Central England Temperature record to this day, providing monthly mean temperature from 1659, daily mean temperatures from 1772, and separate maximum and minimum temperatures from 1878. The different start dates reflect the historical availability of the different types of observations.

The current CET series is, for historical consistency, comprised of observations from just three weather stations, forming a triangle encompassing roughly Lancashire, London and Bristol. This dataset is the longest of its kind in the world, and is an invaluable resource for investigating variations in our climate across several centuries.

However, the CET does not cover the whole of the UK or take full advantage of the complete network of several hundred weather stations currently monitoring weather across the whole country.  The CET series is available here.

National records

For most of the 20th and 21st Century we have had hundreds of observing stations across the UK recording maximum, minimum and mean temperatures. We could take a simple average of all of these observations to estimate a national average temperature. However, the precise number and distribution of stations across the country has changed over time, so a simple average is not the best way to estimate the national record. Instead we take a slightly more sophisticated approach. All available observations from a given date or month are interpolated to a uniform grid of points 5km by 5km covering all land areas of the UK. This gridding method also accounts for factors such as terrain height (mountain tops are cooler on average than lowlands), coasts (the sea can keep coastal areas cooler by day and warmer by night), and cities (the urban heat island effect means large cities are warmer than their rural surroundings). An average value is then calculated from the resulting grid of points. This approach provides the best estimate we have of national records.

There are currently sufficient temperature data on our computer archives for this gridding method back to 1910, providing over 100 years of national records for the UK, Scotland, England, Wales and Northern Ireland. So these records provide our official UK national climate series – the records that you will see quoted when the Met Office releases statistics.

The national and regional series from 1910 available to download here.


So what is the best measure of temperature for monitoring climate?

The answer is all of the above. The different types of data provide complementary information, all of which are routinely used to monitor UK climate and are published through the Met Office website each month.

Each of the maximum, minimum and mean temperature are all important in their own right for monitoring our climate and its extremes, and have all been observed for well over 100 years.

The national series derived from our gridded datasets from 1910 provide the most reliable estimates of UK-wide temperature, being based on all available station data. The CET series provide an invaluable and much longer climate series, providing a perspective of our climate for over 350 years. It is however based on a much more limited set of locations and stations. None of these records would be possible without the existence of high quality and long running historical station series. These station records provide observations for the site they are located in, but they are the firm foundations of our weather and climate observation heritage.



This entry was posted in Met Office News. Bookmark the permalink.

9 Responses to Met Office temperature records – what do we monitor and how far do they go back?

  1. xmetman says:

    Why do you say “There are currently sufficient temperature data on our computer archives for this gridding method back to 1910” when you make available Daily Weather Reports that started in 1860?

    Surely the DWR data is just a subset of all the paper climate records that you have archived for the period from 1860 to 1910 and would be griddable. Why has the data before 1910 never been digitised, as early adopters of computers in the UK since the 1960’s what exactly have the Met Office done about this in the last 50 years?

    • Hello Bruce, thanks for your comment

      We do indeed have a huge amount of data on paper documents. Many of these, like the Daily Weather Reports have been scanned and are now being provided through our digital archive ( Not all of the observational data contained within them have yet been keyed and are therefore not yet part of our data archives. The process of recovering all these observations is a considerable undertaking. Various activities are ongoing in this regard which are being used to extend the national series back into the 19th Century.

      We’re happy for any help we can get to do this, and anyone can join in the project to help recover daily weather reports with WeatherRescue:


      • xmetman says:

        Thanks Mark for your reply. It’s good to know that you’re giving this work the priority it deserves after so many years of inactivity. To speed this process up there are companies in India who specialise in digitising data from written records, but that of course this would cost the Met Office money, which it appears it would rather spend on it’s next supercomputer.

  2. xmetman says:

    It would also be good to see Gordon Manley’s name mentioned when it comes to the CET data series, because without him there probably wouldn’t be one!

  3. I used to run a putting greet at Prestwick Promenade. We were obviously very dependent on the weather … sun, rain and wind. Am I correct that there used to be a weather station adjacent to Prestwick Prom, as well as at the Airport? Do you have weather records, as folk kept telling me the weather used to be so much better for putting?

    • Hi John
      Thanks for the comment. We’re checking with our colleagues and will let you know what they say

      • Hi John
        We have had a number of sites around Prestwick over the years, none particularly close to the prom though. Our current sites are at Prestwick Airport has been open since 01/01/1957 and at Gannet since 1993. We also have rain information from a site on Pow Burn but that’s only been open since 2014

Comments are closed.