Altnaharra No 2 WMO03044 – 30 year climate averages derived from 1960 for site only with readings from 1993. How “Well Correlated” can they really be?

58.287817, -4.442394 Met Office CIMO Assessed Class 4 Temperature readings from 26/1/1993

Altnaharra weather station lies in Sutherland in the Scottish highlands. It is cold in winter here even at its modest 81 metres amsl elevation compared to the surrounding areas. In fact it is often extremely cold – holding jointly with Braemar No 2 the UK’s coldest ever recorded of -27.2°C 30 October 1995. The site has recorded temperatures only since 1993 but, as ever, the Met Office manages to conjure up “climate averages” for this site from 1961. In this post I will evaluate the site and also demonstrate how the claimed long term climate average is simply not plausible by any measures.

Altnaharra is a classic synoptic weather station typical of the latter 20th century. It is highly equipped, securely fenced and specifically intended for the purposes of forecasting. Constraints imposed at the time meant these had to be installed near adequate power supplies hence there are nearly always views of overhead power lines (as below in the distance) in site images.

This site appears well exposed with no major problems at all. The road is close by but not a major issue, in a location like this traffic will be very light and unlikely to cause any temperature elevations.The trees are very remote and wind shading is simply not going to happen here. CIMO Class 4 seems a harsh assessment and Tim Channon previously assessed this as Class 2. Considering the Met Office assess the likes of Heathrow and Chertsey as class 3 it is extremely difficult to often understand their reasoning or interpretation of standards. One thing for certain about this site is that it certainly gets very cold, very often.

Oldbrew recently highlighted the cold extremes here https://talkshop.blog/2025/01/11/uks-coldest-january-night-in-15-years/

The Met Office record listings reveal the following:

Lowest daily minimum temperature record

Country	Temperature	Date	Location
Scotland	-27.2°C	10 January 1982	Braemar (Aberdeenshire)
Scotland	-27.2°C	11 February 1895	Braemar (Aberdeenshire)
Scotland	-27.2°C	30 December 1995	Altnaharra (Highland)
England	-26.1°C	10 January 1982	Newport (Shropshire)
Wales	-23.3°C	21 January 1940	Rhayader (Powys)
Northern Ireland	-18.7°C	24 December 2010	Castlederg (County Tyrone)

The frequency of such cold events is further demonstrated in the TORRO records – Altnaharra holds the coldest day record for 6 days in December alone, whilst Braemar takes 5 days.

What I find most puzzling about this site though is the way the Met Office is divining long term climate records either for this individual location, grid point or any other nearby location. Below are the 30 year averages for Altnaharra from 1961 to 1990 even though the temperature readings do not start until 1993. {I have screen grabbed this complete data for reasons to be explained later in this post.}

Station: Altnaharra No 2

Climate period: 1961-1990

Month	Maximum temperature (°C)	Minimum temperature (°C)	Days of air frost (days)	Sunshine (hours)	Rainfall (mm)	Days of rainfall ≥1 mm (days)	Monthly mean wind speed at 10 m (knots)
January	5.43	-1.68	14.31	–	111.58	18.37	8.74
February	5.86	-1.54	14.45	–	89.11	14.97	7.35
March	6.98	-0.09	11.06	–	109.18	18.13	9.49
April	10.03	1.56	7.50	–	69.40	14.33	8.22
May	13.54	3.69	1.97	–	56.08	12.70	7.64
June	15.71	6.98	0.23	–	55.89	13.33	7.51
July	16.22	9.25	0.03	–	49.27	12.87	6.07
August	17.19	8.94	0.10	–	54.85	14.53	6.49
September	14.76	6.44	0.50	–	109.39	17.37	7.13
October	12.01	4.75	2.40	–	147.47	20.97	8.18
November	7.63	0.94	9.12	–	130.20	18.97	6.79
December	6.16	-0.88	12.52	–	140.48	18.33	7.97
Annual	10.99	3.22	74.19	–	1122.90	194.87

Typically the Met Office claims that such figures are derived from other nearby “Well Correlated sites” but how does that stand up to scrutiny? Below is a 30 km radius circle from Altnaharra. Within this area there are only 2 other currently operational weather stations at Cassley and Achfary. The latter site did not record temperature data until 2011 so would obviously have no useful data to derive an earlier climate average from. Cassley did not start readings until 1974 so at the best only 17 of the the 30 year period can be covered.

Going further back in time back as far as the 1940s I cannot find any other sites within this long distance. Furthermore how likely is even Cassley to be well correlated to Altnaharra? Below are the two sets of readings around the date that Oldbrew’s article covers of 11//1/2025.

In contrast to Cassley

There is very clearly a huge difference between the two sites – suggesting such locations could be well correlated is seriously straining plausibility in an area of such varying topography. Indeed, in response to my post of Tyndrum weather station, reader “Lapogus” offered the following very pertinent and detailed comment that applies across all this northern region.

“Tyndrum is not usually one of the colder sites in the central/southern Highlands, as it’s not that far from the west coast, and less sheltered from milder winds off the Atlantic than some glens.

A key problem is that the Met Office has very few stations in the central / southern highlands / Highland Perthshire area, where temperatures are regularly colder than Tyndrum, and much colder than the other nearest station (Glenalmond just north of lowland Perth), which is routinely 4C warmer than almost everywhere in the Highlands.

It was -7C on Wednesday night, when Met Office models (presumably weighted towards Glenalmond) said it the minimum would be -6C. However, the MO forecast was for it to be 0C by 10am on Thursday morning, when it was still -6C, and their forecast daytime max of +2C at 2pm was also complete bollocks because in reality the warmest it got was -3C.

See https://x.com/lapogus1/status/1862104567845925211

This discrepancy is normal on windless days in winter, because the models do not do topography – so low winter sun does not rise above the hills till much later than in Glenalmond, and any snow fields on higher ground generate colder air ). It’s quite possible that Glen Lyon is one of the coldest low-level places in Scotland, for example it was -24C in February 2011, (comparable with Braemar’s minima in that cold spell) but there’s no Met Office station in or near Glen Lyon so the MO appear to be oblivious to this. This has direct consequence for locals – roads are often icy but are not gritted or gritted sufficiently because councils use Met Office forecasts, and there are consequently often many more cars in ditches on winter mornings than there should be. And despite our much colder climate, cold weather payments for pensioners are rarely triggered. This is a problem for much of Scotland because apart from Braemar, Tulloch Bridge, and Eskdalemuir, the key Met Office stations are at airports and/or coastal locations – Glasgow, Edinburgh, Dyce (Aberdeen), and Tiree and Stornoway, and these stations rarely get the required number of consecutive sub-zero days to trigger the payment.

As for summer maximums, yes, as we are about as far as you can be from the cool Atlantic and cold North Sea, in sheltered valleys and glens, our daytime maxima can be up there with places much further south, and we get some 30-32C days most summers. All in all it’s a good climate, drier than the west (so fewer midges), but the winters can be long, dark and quite hard.“

In the absence of stations within 30 km of Altnaharra what “comparatives” does the Met Office offer?

To suggest any of these sites could be used in any part to derive Altnaharra long term averages is frankly absurd.

Tain Range is both distant, at sea level and on the sheltered Dornoch Firth east coast. This site shows minimal warming over the last 60 years in the region of half that claimed for Altnaharra.

Strathy East by a delightful north Scottish bay bears not the remotest similarity to an inland highland location.

Kinbrace Hatchery may or may not bear any similarity to Altnaharra but again was not reporting temperatures through the 1961 to 1980 period so earlier correlation is again unknown.

Loch Glascarnoch is actually shown as a mean colder site than Altnaharra but unlike superstar performer Braemar and even occasional performer Kinbrace, never appears on the list of national coldest for any day of the year. How well correlated can a consistently cold site, almost 70 kilometres away and not noted for significant Foehn effects, be to such a highly fluctuating site as Altnaharra which has curiously also recorded some of Scotland’s highest daily temperature readings.

As Wikipedia states describing the Altnaharra climate:

“Altnaharra has a Met Office weather station. The village’s northerly latitude and inland location mean that in winter, it is often featured in the daily weather extremes in the United Kingdom. It is unusual in that the coldest month of the year is normally December. On 30 December 1995, the UK’s lowest recorded temperature −27.2 °C (−17.0 °F) was measured there. This matched recordings at Braemar in the Grampians on 11 February 1895 and on 10 January 1982. In a Parliamentary debate on the Spring Statement on 23 March 2022, local MP Jamie Stone said, “The village of Altnaharra in my constituency is the coldest place in the UK every single winter.”

On 20 March 2009, it was recorded as the warmest place in the UK, at 18.5 °C (65.3 °F), which was the station’s warmest recorded March temperature and possibly the first time the station had recorded the warmest UK temperature. The station also reported the equal warmest national temperature of 12.3 °C (54.1 °F), with Tain on 20 January 2020. On 19 June 2020, the station had both the warmest temperature, 22.2 °C (72.0 °F), and the coldest temperature, 7.6 °C (45.7 °F), reported anywhere in the United Kingdom on that day. The March 2009 temperature was beaten on 25 March 2017 when the station recorded 19.7 °C (67.5 °F). Also on 26 May 2017, the station recorded its highest May temperature of 28.0 °C (82.4 °F), beating the previous 27.4 °C (81.3 °F) recorded on 27 May 2012. On 8 January 2010, the temperature dipped to −22.3 °C (−8.1 °F), the coldest temperature recorded in the UK since 1995. On 3 November 2015 the warmest November temperature of 16.1 °C (61.0 °F) was reached, followed by 15.6 °C (60.1 °F) on 17 December 2015 being the warmest December temperature on record at the weather station. On 25 January 2016, the highest January temperature of 13.9 °C (57.0 °F) was reached, being surpassed on 28 January 2024. On 24 September 2020, Altnaharra reported −5.0 °C (23.0 °F) which was the lowest reported September temperature at this station, as well as being the coldest in the UK since 1997.“

It therefore seems improbable that the 1961 to 1990 climate averages (to the second decimal place) can possibly be derived with any form of real accuracy. Furthermore, in the absence of any input stations details being publicly available, all the numbers are not falsifiable and thus meet no scientific integrity tests. I stress this point as I have made it many times before, specifically in relation to Charlwood at the western end of Gatwick airport runway. Following my queries of the veracity of the 1961 to 1990 climate averages predating the site’s existence, the Met Office deleted the data which now looks like this.

Station: Charlwood

Climate period: 1961-1990

Month	Maximum temperature (°C)	Minimum temperature (°C)	Days of air frost (days)	Sunshine (hours)	Rainfall (mm)	Days of rainfall ≥1 mm (days)	Monthly mean wind speed at 10 m (knots)
January	–	–	–	–	79.90	–	–
February	–	–	–	–	51.80	–	–
March	–	–	–	–	61.40	–	–
April	–	–	–	–	55.30	–	–
May	–	–	–	–	56.90	–	–
June	–	–	–	–	56.20	–	–
July	–	–	–	–	47.50	–	–
August	–	–	–	–	56.70	–	–
September	–	–	–	–	68.80	–	–
October	–	–	–	–	75.30	–	–
November	–	–	–	–	78.80	–	–
December	–	–	–	–	81.40	–	–
Annual	–	–	–	–	770.00	–	–

Unfortunately I had not locked down the original data by screen capture and posting here but was able to recall it via the Wayback Machine. I have locked down Altnaharra in case the data meets the same fate. Curiously the Met Office had literally dozens of operational weather stations in the Sussex and surrounding areas of the time from which to compile Charlwood’s data yet still deleted it on questioning – Altnaharra realistically has no such suitable contemporaries.

In conclusion Altnaharra is a perfectly worthy synoptic weather station providing reliable enough data for its primary intended purpose of forecasting. Its readings are equally representative of its small locality. However, for the compilation of climate history to the second decimal place it is only acceptable throughout its recent history and certainly not prior to its installation.

Source: https://talkshop.blog/2026/10/07/altnaharra-no-2-wmo03044-30-year-climate-averages-derived-from-1960-for-site-only-with-readings-from-1993-how-well-correlated-can-they-really-be/