Data Accuracy versus Data Integrity – Four ongoing Freedom on Information Requests

I currently have four Freedom of Information requests in various stages with the Met Office regarding straightforward numerical integrity of archived items. That the recording of temperatures at any given time should require such absurd levels of investigation is likely beyond most reasonable people’s rational belief. In the case of the Met Office, such innocuous numbers seem to be ultra secretively held but is this an issue of accuracy or integrity? Given that the Met Office has a habit of changing their websites and data surreptitiously when challenged on issues, I detail my enquiries for the record lest things are covertly changed.

To clarify the points “Accuracy” relates to precision and how close to a set standard something is. “Integrity” relates to whether something is actually correct or indeed truthful. An inaccurate number can be truthful if that was to the best precision available. An incredibly accurate number, however, can be completely lacking in integrity if it was simply made up. It would be a very strong allegation indeed for me to suggest that the Met Office may have both made up some numbers and accepted known wrong numbers and archived them. However, that is exactly what I feel has appeared to have happened and thus I have asked for clarification on certain numbers and events. Quite why such apparent trivial matters have required me to take them not only to FOI request but also for the Met Office to delay answering due to the alleged “complexity” of one of them (so far – others may follow) is almost beyond parody. These are only temperature readings after all surely. Or perhaps there are more aspects to these issues that must be protected – I do not know, so I will simply record my case for consideration.

Motherwell: Strathclyde Park

Many will remember this infamous Scottish “record breaking” site proclaimed by the BBC.

“On Thursday 28 June, a temperature of 33.2C degrees was measured in Motherwell, North Lanarkshire.

It exceeds the 32.9C recorded in August 2003, at Greycrook in the Borders.”

Only something was not quite right.

The reality was an ice cream van had been parked alongside a truly atrociously sited weather station which was still rated as “Unsatisfactory” by the Met Office themselves when I reviewed it in 2024 – plus ça change.

It would be reasonable to assume that the 33.2 °C reading in being acknowledged as inaccurate would be expunged from the records with presumably no reading for that day archived – a classic “NA” that frequently appears. This is what actually appears in the archives. {note: with 24 hour daily recording the 09:00 maximum reading in column “I” relates to the day before i.e. reading date 29/6/2018 is for 28/6/2018}

So my simple question was….where did the reading of 31 °C come from? Was this an “estimate” derived from somewhere else, if so why bother for just one day’s readings – there are literally millions of blank days readings (NA) in the archives for manual stations, one more is insignificant. {Note, I have a personal “theory” where this number came from but I am not divulging that until the Met Office eventually provides an answer to FOI IMS0292291.}

Floors Castle

The Scottish all time record high temperature saga continued. Scottish Television (STV) this time proclaimed this momentous event.

{Worth noting that point about this Floors Castle manual reporting station beating Charterhall’s figure of 34.8 °C which is discussed below} Again this proclamation was somewhat premature and a correction had to be made.

This article offered the bizarre explanation ” A previous high of 35.1C, recorded at Floors Castle, was not officially recognised because the site is not compliant with Met Office standards.” Floors Castle is a fully compliant, official Met Office weather station of longer standing than Charterhall. Why it was ruled out is anyone’s guess it seems – I cannot find any official definitive reason. A cynic might suggest that the Floor Castle site is publicly visible in a walled kitchen garden and people might start to question why the Met Office opts to take climate readings in a deliberately engineered warmer micro climate. Charterhall conversely is somewhat off the beaten track and unlikely to come under close scrutiny.

On the basis that the Floors Castle reading should again be expunged and deemed inaccurate I checked the archives.

On this occasion the “incorrect” or “inaccurate” figure has NOT been removed and it has also been archived for posterity. I therefore asked the Met Office for an explanation why a claimed inaccurate (thus knowingly false) figure had been committed to the archives. Whilst errors are always possible with routine numbers, It seems highly unlikely that a potential national record high should be so carelessly regarded. I am awaiting an answer to FOI IMS0292131. This leads on to a further mystery regarding the alleged official Scottish record.

Charterhall. My initial concerns regarding this site revolved around a noted screen relocation within its own compound. It seemed very odd for the Met Office to record such a tiny change given that in so many cases very large re-locations measured in hundreds of metres are not openly recorded. Even the exact date, 6/12/2022, was recorded. A cynic might infer there was something significantly wrong with the previous site (that it was even over grass was extremely questionable) and it had hurriedly been moved to correct the problem.

I looked at the readings around that record setting date of 19/7/2022. Charterhall is a long term automatic WMO reporting site that supplies temperature readings by the minute. Data from such sites is conventionally archived twice daily at 09:00 and 21:00. In these archives the 21:00 reading is the high point of the day it is recorded on unlike daily (24 hour) reporting Floors Castle and Motherwell:Strathclyde Park which show highs a day in arrears at 09:00.

Despite reporting on a 12 hourly cycle for over 30 years, on the 29th June 2022 Charterhall suddenly, and with no explanation, changed to 24 hour (once daily reporting) suggesting possibly being manually read. The last 24 hour reading was logged for 1/7/2022 when ALL READINGS STOPPED for 17 consecutive days.

The first reading to be resumed from Charterhall was back to 12 hourly reporting at 21:00 19th July 2022 and was the all time Scottish record of 34.8 °C. Given the coincidences here I have asked the Met office to explain this remarkably odd situation. Freedom of Information request IMS0292133 refers.

Waddington I have previously posted twice on this weather station where I found records from RAF Scampton being transferred over to RAF Waddington’s archive. Records for Waddington in general are patchy to say the least however, the main issue here is that the UK all time record high of 40.3°C allegedly set at RAF Coningsby (Typhoon take offs and all) was also recorded for Waddington.

Back in the days when the Met office enquiries desk still felt it acceptable to respond to my information requests I was told

“Dear Mr Sanders,

The 39.9C value on the map in the link you provided is not Waddington but it Cranwell or Scampton, both of whom recorder maximum temperatures of 39.9C that day.

At that time Waddington was having its temperature data marked as suspect as the grass under the screen had been treated with weed killer.

Kind regards,

Weather Desk”

Trying to follow up on the point by requesting why the 40.3 °C was passed into the archives if readings were known to be suspect met with a wall of silence. I therefore raised an FOI IMS0286812 on the 23/4/2025 requesting an explanation of why known inaccurate data was passed into the archives. Again error seems implausible with such a spectacular (but highly disputed) set of figures going into the historic record. Despite a normal 20 working day period time to respond, the Met Office (Legal Department) have advised they are extending their time due to the ” complexity” of my inquiry out to the 20/6/2025.

All these may superficially seem trivial issues but not only is the integrity of archived numbers called into question there are also suggestions that inaccuracy is prevalent. There are other implications which may be revealed dependent upon the Met Office answers. I will report further when more information is available. In the meantime the above data has been locked into the internet archives.

Source: https://tallbloke.wordpress.com/2025/05/22/data-accuracy-versus-data-integrity-four-ongoing-freedom-on-information-requests/