EXCLUSIVE: New Freedom of Information Request and the UK Met Office has to Rewrite its Temperature Explanations Again

The UK Met Office’s excuses for its invented temperature data from non-existent stations get more fanciful by the day. Explanation after explanation fails to live up to rigorous examination, with statutory Freedom of Information (FOI) requests leading to frequent rewriting and new things to believe before breakfast. Data from non-existent stations are obtained from well-correlated neighbouring stations (WCNS) the Met Office told us, but there were no well-correlated neighbouring stations at Lowestoft, which had itself been closed since 2010. Via a new FOI, the Daily Sceptic sought details about these invisible WCNSs at Lowestoft and has been told, “We do not use well-correlated neighbours in this case.”

Well we kind of knew that all along, due to the fact that citizen sleuth Ray Sanders couldn’t find any in the first place. For closed stations on its historic station database the new FOI explanation is: “We used the gridded value from the closest grid point from our version controlled UK climate data set HADUK-Grid.” The suspicious might note that the HADUK-Grid uses WCNS data derived from the Met Office’s Integrated Data Archive System (MIDAS) in modelling infill data for closed stations. This is said to address gaps caused by stations opening or closing, “ensuring consistent national coverage for monitoring and research”.

In effect the Met Office is no further along in its explanations. It is simply excusing its inability to justify its invented numbers at Lowestoft by passing the buck to a computer programme that uses the ‘well-correlated neighbouring stations’ that it is unable to identify.

The Met Office is not completely stupid. When doubts arose about the figures at Lowestoft and two other closed stations at Nairn and Paisley, it hastily withdrew large chunks of data relating to the periods when the stations did not exist. But it was obvious that this could lead to questions being asked about misleading data cascading through the entire Met Office temperature database. Using the data that have had to be withdrawn from public view as WCNS input would have quickly corrupted estimates elsewhere and required a major cleaning of vast amounts of related data. What if all of the data were found at some point to be corrupted by dubious figures that could no longer be shown to the public? Invented data from invented data is hardly a good look, even in science, media and academic circles desperate for ‘highest evah’ official figures to boost the Net Zero fantasy. Happily this will not be necessary since the publicly-available historic database now states:

The purpose of this webpage is to provide a sample of historical station data across the UK for general interest. It is not used for formal climate monitoring.

What a relief. Curiously, a check from a few weeks back on the Wayback Machine shows the same page but without this particular notification. Also missing from the past edition is a note that the stations shown comprise “only a small fraction of the full UK climate network”. The remainder would be the data and statistics for other unpublicised stations “and associated charges” that can be obtained from the Met Office’s Customer Centre. Meanwhile on the latest page, five stations are now admitted to have been closed, namely Lowestoft, Paisley, Nairn, Southampton and Ringway.

This is not the first time that the Met Office has had to hastily rewrite explanations on its publicly available temperature databases. Earlier Ray Sanders investigations revealed that more than a third – 103 of 302 – stations listed with individual coordinates and elevations and supplying long-term averages did not exist. As a result, the Met Office changed the name of the database called ‘UK climate averages’ to ‘Location-specific long-term averages’. The original suggestion that selecting a climate station can provide a 30-year average from 1991-2020 – a claim available to view on Wayback – was replaced with the explanation that this page “is designed to display locations that provide even geographical coverage of the UK, but it is not reflective of every weather station that has existed or the current Met Office observation network”.

Of course those with a cynical mind jumped to the conclusion that the Met Office solved its problem of inventing data from invented stations by suggesting that the data now arose from “locations” which may or may not bear any relation to stations that once existed, or indeed exist today. Enter Science Feedback into the fray with a feeble ‘fact-check’ last December that found the reports by Ray Sanders and his Tallbloke Talkshop blog along with the Daily Sceptic were “misleading”. The average data presented by the Met Office for specific weather stations that have closed are not “fabricated”, it concluded: “They are estimated using [wait for it] well-correlated neighbouring stations.”

At the time the WCNS argument started to fall apart. Sanders noted that the Cawood station in Yorkshire was a pristine Class 1 site designated by the World Meteorological Organisation as providing uncorrupted data over a large area going back to 1959. But no rolling 30-year average for Cawood is provided. Instead the Met Office flags data for five other sites between two and 27 miles distant. Unlike Cawood, all of these have average data despite the fact that four no longer exist and the fifth, High Mowthorpe, is 27 miles away and at a 163 metres higher elevation. Quite why the Met Office should ignore an excellent set of readings from Cawood and rely on inventions and far distant temperature readings must remain a mystery. For Ray Sanders the term ‘smoke and mirrors’ came to mind.

And of course we hardly need reminding that pristine Class 1 sites are as rare as hen’s teeth in the Met Office’s near 400 temperature station nationwide network. A recent FOI from the Daily Sceptic revealed that over 80% of the stations, constituting the bedrock of all Met Office data and claims, were sited in junk Classes 4 and 5 with ‘uncertainties’ of 2°C and 5°C respectively. Over the course of 18 months the junk percentage had actually risen, while in the same period the number of Class 1 sites fell from 24 to 19.

The Met Office regularly ‘updates’ its explanations for how it compiles temperature data in the UK. As a service to readers we will attempt to keep abreast of all future changes. Watch this space.

Chris Morrison is the Daily Sceptic’s Environment Editor. Follow him on X.