Raasay:Oscaig DCNN0540 – Very poor recording of a micro climate, more dubious data.

57.368164, -6.079765 Met Office CIMO Assessed Class 4 Temperature Records from 29/02/2024

Raasay is a small island just 14 miles long and 3 miles wide that lies between Skye and the Scottish mainland. It is reached by ferry from Skye at Sconser. There is no ferry service directly to the mainalnd nor is there any official aircraft landing strip. To travel here from the Met Office’s regional headquarters at Aberdeen will take about 8 hours by road and bridge to Skye and then the ferry. The weather staion is not at the island’s main town of Inverarish rather further north at the tiny hamlet of Oscaig. This is a very new site and quite why it is there is somewhat a mystery.

Raasay is a very sparsely populated island of aroound 180 residents. The current main industries on the island are tourism and the island’s distillery. Climatically the island is sheltered from harsher extremes by the Isle of Skye itself under 2 miles to the west and benefits from the warmth of the Gulf Stream calmer waters particularly on this western side of the island. The image below is slightly AI enhanced to bring out the screen location marked by the red pin. The site has been assessed as CIMO Class 4 presumably partly due to the slight slope and also the shading from low angle early morning sun from the inland mountains rising to 444 metres/1457 feet amsl.

A number of problems concern me regarding this new site . Rather than offer conjecture I will start by a basic hard fact – the observations standard here is very poor indeed. Below is an archive extract from the first weeks since opening.

The most obvious first thing are the high numbers of “NA” in columns “I” and “J” – maximum and minimum temperatures respectively. This extract covers 29th February through to 28th April 2024 inclusive and there are 14 days where one or other or both of the readings are not taken. However it is much worse than that when the date file (column A) is examined – perhaps the missing dates from 2nd to 27th March represent a “false start” in the early days. Look deeper though and the 6th, 16th, and 24th April are also missing. Taken all together of the 59 day reporting period there are only 16 days of complete readings. Not a very conscientious nor succesful start.

If the above was potentially “teething” troubles, it certainly did not get any better. Below is an extract from September 2025.

Nothing has really changed – “NA” appears 6 times and whatever happened to 23rd to 25th inclusive not even being registered in the date file. The sad reality is that observations are very much hit and miss with 68 days missing readings out of a potential 306 days in 2024 (a 22% failure rate).The situation did not significantly improve in 2025 with 61 failures (a 17% failure rate).

This ommission of readings is much more significant than just a dropped set of figures. Manually observed stations have a PRT with a Data Logger displaying readings throughout the day as demonstrated in this video from the Floors Castle observer. The Data logger is only used for maximums with minimums still being taken from the LIGT…….. or at least that is what is supposed to happen. This retention of the LIGT minimum is intended to keep a level of continuity with traditional readings at least for minimums and the potential carry over effect of previous days readings is still there unlike automated sites which solely use the PRT readings largely avoiding the problem.

A crucial point of LIGT observations is that the “rider” is reset after readings taken at 09:00GMT. Below is an AI generated image to display the point. Readings are taken from the bottom of the rider which is then (usually magnetically) reset down to the top of the liquid medium in readiness to record up to the following day. {n.b. this is not a Met Office instrument type merely a general indicative image.}

The obvious point arises that if the minimum thermometer is not reset as no readings were taken, then when it eventually is next read it can only show the same or a lower reading, it should not be able to indicate a higher one. So look above, for example, at the minimum reading of 1st October 2026 of 13.8°C – no reading is shown for the 30th September with that of the 29th being only 8.2°C. If the rider had been reset at the 8.2°C reading it woud still be there on the 1st October if the weather had not got colder so it should not be reading higher.

Amongst other possible explanations ares, the rider was reset at some other time or the observer read the thermometer as a timed reading only at 09:00 thus not a true 24 hour indicator or (quite probably) the reading was taken from the PRT Data Logger. Whatever the reality of the situation is, one thing for certain is the readings have a high probability of being either wrong or misleading. Quality control would, in the past, have picked up on this but now seemingly anything goes unchecked.

The Met Office has confirmed they use the PRT readings as a cross check to minimums for record verifications. The big problem is that historic climate records also use “adjustments” for manual minimum once daily readings to account for potential “carry over” effects of this system. Herein lies a concealed problem that if observers effectively change the protocol by using the Data Logger minimum they covertly change the reading protocol and hence adjustments are applied to readings where they should not be.

This may seem trivial in some ways, but a large scale migration of observation protocol of the approximately 140 manually observed sites would be a major issue. I am in the process of trying to establish a step change caused in the past by conversion from manual to automatic reporting at indivual sites. All manual sites now employ PRTs/Data Loggers from 2017 at the latest following a mass conversion programme over a year so any observation change (particularly a covert one) could have major implications.

Going back to Raasay, there are multiple other issues with the readings. Again from the above data extract, the readings of 17th and 18th September 2025 for both maximum and minimum are the same. This is uncommon at most sites but it is not improbable for two consecutive days to have the same readings. Rassay has a generally narrow diurnal temperature range neither getting hot nor especially cold as is typical of these maritime Gulf stream warmed sites increasing the possibility of similar readings. Unfortunately the number of times this has happened over the less than 2 years at this site (4 occasions) rather stretches plausibility given the patchy and dubious observations standards. Many readings, particularly minimums are the same for several days in sequence.

My next issue leads on from this intermittent observation and my general (possibly) conjecture that a lack of observations diligence is likely to be reflected in maintenance standards of the site and equipment. I have frequently commented on how poor a state some screens are allowed to get into such as at Glenlee and Lough Fea .I am currently investigating Met Office inspection frequency and standards. The Met Office makes many claims about their sites that do not stand up to scrutiny when investigated. Only this summer they had to withdraw alleged records from the likes of Wiggonholt, Cardiff and Eastbourne on the grounds that they did not meet adequate standards to declare records from. My response to that is twofold – I could have (and indeed did) tell them that a long while ago and secondly if they are not good or accurate enough to declare national records from they are not good enough for any readings at all. A future report will cover these specific site record redactions.

Raasay may or may not be well maintained but I suspect (and will be checking) that given the arduous, lengthy and expensive journey time to such a remote location that inspection visits may not be that frequent. I will advise if I am allowed to have such “awkward” questions answered.

A final point revolves around whether or not this site is suitable for inclusion to climate reporting data which is, after all, its sole purpose. Only a very small part of Raasay is inhabited in the south western sheltered flatter area. This really is a tiny micro-climate even for this small island let alone the general wider area of up to “tens of kilometres” that the Met Office states in its user guide that its sites should represent. In my report on Sheffield Weston Park I noted that those in colder areas outside the city could unfairly miss out on Cold Weather Payments as the qualifying cold days data was being derived from an inner city urban heat island affected site that was much warmer. Raasay enjoys the reverse effect – since the closure of the Skye site at Lusa, qualifying days are derived from the very much colder recording Achnagart site 28 miles away on the mainland. I drew up a very short period comparison chart below to show how Raasay and indeed all of Skye benefits from this change and also demonstrates the difference in climate over such relatively short distances compared to say central England. (Minimum temperatures in degrees celsius. shown)

Date……………………………………………Raasay………………………...Harlosh (Skye)………………………..Achnagart

1st January 2025…………………………..3.3………………………………2.1……………………………………….2.9

2nd……………………………………………1.1……………………………..0.2………………………………………..- 1.2

3rd…………………………………………….1.0……………………………-0.8………………………………………..-1.6

4th…………………………………………….1.5……………………………-0 .7………………………………………..-0.9

5th………………………………………….no reading……………………..-1.2………………………………………..-2.3

6th………………………………………….no reading……………………..-1.2………………………………………..-6.4

7th …………………………………………-1.5……………………………..-1.7………………………………………..-5.5

8th……………………………………………1.6…………………………….-0.4………………………………………..-0.9

9th………………………………………….no reading…………………….-3.1………………………………………..-7.2

10th ………………………………………no reading……………………. -3,2………………………………………..-9.6

11th……………………………………….no reading……………………..-2.4 ……………………………………….-2.7

12th…………………………………………3.7……………………………..3.7………………………………………….0.2

13th…………………………………………5.9…………………………… 5.8………………………………………….4.5

14th…………………………………………10.3…………………………..10.6…………………………………………9.1

This brief period demonstrates that Rassay is recording (when readings are taken) marginally warmer than Skye and significantly warmer than its mainland neighbour by several degrees. With this micro climate and its unreliable observation frequency and suspicious readings I feel this site is effectively worthless for climate reconstruction purposes. Furthermore that the Met Office will be accepting many of these questionable readings into its homogenisation and infilling systems is far below an acceptable standard for a national Meteorological organisation.

Source: https://talkshop.blog/2026/10/12/raasayoscaig-dcnn0540-very-poor-recording-of-a-micro-climate-more-dubious-data/