Dunstaffnage DCNN6068 – An unusual “Bellwether” in more ways than one.

56.45039 -5.43974 Met Office CIMO Assessed Class 4 Archived temperature data from 1/1/1972

Dunstaffnage is located a few miles north of Oban very close to the shores of the Dunstaffnage Bay. This is a key weather station for the nautical community with the prestigious Dunstaffnage Marina close by. Although not a WMO reporting site it does record the full range of parameters with wind conditions vital for safety of mariners. This type of site fulfills the founding needs of the Met Office by Vice Admiral Robert Fitzroy back in 1854 to help more accurately forecast weather and protect mariners.

An often overlooked but important point that even the Met Office admits to is that most weather stations were never originally intended for high accuracy historical temperature recording. Immediate forecasting tended to originally be much more preoccupied with wind speed and direction, air pressure, visibility and precipitation. This is how the Met Office used to describe observing sites

1. The site should be representative of an area up to several tens of km from the station
2. The instruments should be installed on level ground
3. There should be no steeply sloping ground in the vicinity and the site should not be in a hollow
4. The site should be well away from trees or any other large obstructions. The distance of any such object should be not less than twice the height of the object, and preferably four times the height
5. The site should be adequately protected to exclude entry by unauthorised persons
6. It must be possible to meet the exposure requirements of the raingauge, sunshine recorder and anemometer. Separate sites may be chosen for these instruments.

The important point to note is summed up by this addition.

“It is unavoidable that some sites do not meet all these requirements, particularly where a station set up for one purpose gradually takes on a different role, for example an airport site originally established for aviation observing may become a key synoptic or climate station while suffering the effects of urbanisation. A few sites are in city centres and may be unsuitably located close to large obstacles or even on the roof of a building.“

It is reasonable to suggest that Dunstaffnage, on the coast (indeed very close to water’s edge) was established primarily for marine uses and falls into the repurposed or more correctly “dual purposed” category to include climate station reporting. It has been in this secondary role that Dunstaffnage has become both an “Historic Station” from its longevity and a “dominant” Climate Averages station as below.

The actual site itself is very difficult to identify from Streetview imagery and the best I can offer is this somewhat blurry shot below. If anyone can locate a better photo I would be very obliged.

Whilst the wind mast is visible, the screen is probably the box to the left of it though I accept I cannot be certain. There are no indications in any of the records of any re-locations and it seems unlikely to ever have been. There most certainly has been extensive development of the area with the expansion of the Scottish Association for Marine Science seen below though this is unlikely to have had any significant material effect on the weather station.

The CIMO regulations insist this should be rated a Class 4 site but given its location type and original intent it seems unduly harsh. I suspect its temperature readings would well reflect the coastal settlements for a large part of this west coast of Scotland in being so similar to most inhabited low lying areas subject to a Gulf Stream/wet westerly dominated climate. What I find difficult to accept is that this site could be in any way “well correlated” to the inland, high altitude sites it is associated with.

The nearest other quoted “Climate Stations” are Tulloch Bridge at 40 miles distant, 22 miles east of Fort William and sitting at 249 metres elevation. If that seems an improbable comparison, consider that Aonach Mor weather station is at 1,130 metres elevation. Did whoever drew up these Climate Averages stations really fully understand where some of them are located?

Aonach Mor

As previously noted Lephinmore will shortly be celebrating the 60th anniversary of its demise which only leaves Colonsay which is an island 40 miles away to the west. The Homefield weather station opened in 1999, closed in 2015 and left an incredible amount to be desired.

Colonsay:Homefield.

Again, as I have already reported at Cassley , Dunstaffnage is one of those stations whose transitional data from manual to automatic recording had an extended overlap period available to view in the CEDA archives. Dave Woolcock has run several comparison graphs on the period readings which are highly illuminating and followed a predicted pattern in contrast to Cassley.

Firstly the straight minimum readings variances caused by the manual readings being solely taken at 09:00 and regularly double recording cold weather events. {n.b. there was a short period of no comparative data hence the gap}

Again the seasonality is demonstrated with later sunrise times in winter highlighting the effect. The extremes are slightly lower than at Cassley due to the much milder conditions of the coastal site – a point seemingly lost on the Met Office “Climate Scientists” in comparing Dunstaffnage with Aonach Mor.

Dave then went on to analyse the number and scale of differences by month in numerical terms. January to December are months 1 to 12 respectively on the Y axis with the temperature discrepancies to the X axis. The seasonal frequency tending towards later sunrise periods (winter and “shoulder months”) is clearly apparent.

The “Distribution of Differences” supplies a realistic representation of the scale of differences (and where the majority lie) rather than overly emphasising the extremes.

Whilst the majority of readings concur, the number of discrepancies are vastly too many to be disregarded and again I stress the stochastic nature of them does not lend itself to “averaging adjustments”. The raw data is thus corrupted.

More to the point, unlike Cassley, the reverse discrepancies are almost non-existent. The automatic PRT/Data Logger readings almost never show lower than the LIGT readings. I have hypothesised (described in the Cassley report) that opening the screen door to take daily manual readings is releasing artificially warmed daytime air that has lain insulated within the screen in low wind conditions. As the Floors Castle observer noted, this opening allows the entry of ambient air and the PRT rapidly responds to this leading to a slightly lower reading than the LIGT. This discrepancy should never happen.

This “Aitken Effect” is well understood and occurs when and where wind speeds drop below 1 metre/second for extended periods. {N.B. many meteorologists claim even higher wind speeds can still be problematic}. Wind speeds are generally known to reduce overnight but equally well known is that coastal regions much less frequently encounter such calm conditions with a gentle breeze remaining. I suggest the incidence of Aitken effect in naturally aspirated coastal sited screens will be hugely less frequent than inland sites and this Dunstaffnage data reinforces my point.

Dunstaffnage is further demonstrating the distorting effect of different reading protocols creating an artificial impression of warming when changing from one regime to another. I further demonstrated this point with currently active but separate sites with the Cambridge Convolution.

Dunstaffnage is also potentially supplying evidence to support my argument of the shortcomings of naturally aspirated screens and Aitken Effect similarly artificially raising recorded temperatures. Throughout many of my almost 400 reports so far I have repeatedly noted wind sheltering/shading effects increasingly affecting screens through either urban encroachment, new development or screen re-location and most notably an appalling lack of maintenance of surrounding vegetation conditions such as at Pershore, Lentran and the even “Mcdonalds” syndrome at Liscombe amongst others. The recording of wind conditions at Dunstaffnage will be further cross analysed regarding this point as its lack of Aitken corruption is being indirectly demonstrated in its cross referenced temperature record. The Met Office themselves could assist me with this line of research but I seriously doubt they will as I am now considered “persona non grata”.

All of the above issues will be further addressed (with some spectacular extremes) in my upcoming addendum report of Dalwhinnie providing a case study that the Met Office must answer to retain any degree of credibility to their data presentation.

Source: https://tallbloke.wordpress.com/2025/11/27/dunstaffnage-dcnn6068-an-unusual-bellwether-in-more-ways-than-one/