Dustaffnage Addendum – The story on the ground.

Regular readers may recall the excellent guest post on Metrology written for the Talkshop by John Marlow. That post is still one of the most widely viewed in the Surface Station Projects and was a most welcome addition to the overall knowledge base and is a continued reference point.

John very kindly offered a review of his time both setting up and establishing Dunstaffnage Weather Station which I reproduce in full below. I am very confident readers will find this both informative and superbly written.

Dunstaffnage Met Station; a reflection after 55 years..

The Scottish Marine Biological Association (SMBA) was originally based at Millport on the Isle of Great Cumbrae, but in the 1960s they were financed by the Natural Environmental Research Council (NERC).   NERC provided the funds to build a new laboratory at Dunstaffnage (a few miles north of OBAN, Argyll), equip it and expand the research teams.  In 1969 I was recruited to form part of a new Physics Department.   Our remit was to provide help and expertise to the marine biological scientists, plus any other diverse research that the Director of the laboratory required. The work was very varied.     During my time there I developed a micro-tag that was attached to the carapace of a free roaming crab sending heart-beat signals to the laboratory: also a large seabed ‘grab’ that extracted a sample of mud (½ x ½ x ½ metre) and bring it to the surface without there being any mud washed-out of the sample bucket.   We were also involved in Hydrograph research of the North Atlantic between Mull and Rockall.

Quite how it came about for the Met Office to realise that the situation of the laboratory, about half way between the existing Met Stations of Glasgow and Stornoway, would expand their coverage of Western Scotland, I forget.   The Met Office provided me with their expertise and encouragement and by the autumn of 1970 we were ready and started recording.

The Stevenson Screen containing the mercury in glass (MiG) thermometers and the rain gauge were situated in the corner of the laboratory grounds as far away from the laboratory buildings as possible.   The land was sand, covered with about 100mm of soil and grass.   It was about two metres above maximum sea level.   The anemometer was situated another 100 metres, or so, further South on a spit of land that extended into Dunstaffnage Bay.

If one zooms in to the map the Met Station and Anemometer mast are marked.   https://www.trove.scot/place/76406

Thermometers.

The Met Office provided me with a parcel of the various Mercury in Glass (MiG) thermometers that were needed.   Having been, in my previous employment, attached to the Quality Assurance section; presented with a set of Slip-Gauges and given the task of calibrating or rejecting all measuring instruments used in the Research and Production Department    I was surprised that there were only Manufacturer’s Calibration Certificates for the two new thermometers, nothing for the rest.    When I asked the Met Office if there were calibration certification for the other Thermometers there were several seconds of silence on the telephone, then the reply, “I will see if I can find any and send them to you”.   After about a week Certificates arrived for about half of the Thermometers.   One was a Manufacturer’s Calibration Certificate dated 1936 – I remember it well as it is the same year as my birth; a 34 year interval since it was last calibrated.

Quality Assurance of Met Office Data.

Fortunately the Laboratory had purchased a HP Digital Crystal Thermometer of sufficient accuracy to calibrate the thermometers.   Usually a crystal is cut in such a direction so that the frequency remains stable over a wide difference in temperatures.   In this instrument the cut is in a different direction and it results in a device that changes the frequency with a change in temperature with superb repeatability.

All the thermometers were calibrated with one thermometer rejected as it was outside the ±0.1°C limits.

At the time it did not strike me that the Met Office had such a cavalier attitude to accuracy to all their data.   They would have been happy to accept the readings of defective thermometers.

Recording scenario.

Every morning at a few minutes before 9am GMT the observer would collect the Met Office notebook and Stevenson Screen padlock key and walk to the Met Station.  S/he would record the Dry Bulb Temperature (Tdry); Wet Bulb Temperature (Twet) so that the humidity could be calculated.   S/he would record and reset the Maximum Temperature (Tmax) and Minimum Temperature (Tmin).    The quantity of rain was recorded.    Back at the lab the Atmospheric Pressure was recorded.   The Wind Speed and Direction chart was annotated with the Date and Time to help the poor unfortunate who would analyse the chart at the end of the month.    Similarly the Sun Energy chart was annotated and, if necessary the clock was adjusted if the attached chart denoted a change in Local Apparent Time was required.   All this data and charts were sent to the Met Office at the end of each month.

I ran the Met Station until I left the Laboratory at the end of 1969: since that date Platinum Resistance Thermometers have been installed.   One used to be able to interrogate the data-logger on-line and extract the temperatures, but access was denied about ten years ago.

Local Apparent Time (LAT).

There are many articles about LAT on the web which should be read to understand this fascinating subject.

Ray’s Half-hour after dawn for Tmin; and should we read Tmin later.

Ray’s hypothesis that the coldest temperature may occur sometime after dawn may apply more so for Dunstaffnage as  the longitude of the Met Station is about 5.5° west of the Greenwich Meridian, which is equivalent to a 22 minute delay in the sun rising.   This means that we were taking our measurements at the equivalent to 08:38 GMT in the morning.  The time for sunrise at the end of the year for Dunstaffnage is …

• Sat 27 Dec 2025
• • First Light

8:08 GMT

• Sunrise

8:56 GMT

Since Tmin is read at 08:38 local sunshine time; we are really reading it 18 minutes earlier than sunrise.

Another factor is that Ben Cruachan (1127 Metres high) is about 16 miles to the east of the Met Station; this may delay the actual sunrise at the Laboratory even more.

Often we can obtain an anomaly if we measure at 09:00GMT.   Imagine a very cold, cloud free night.   Both Tmin and Tdry read 1.0°C, but later in the morning a warm front arrives and the temperature rises.   The next morning we have the situation of Tmin is still reading 1.0°C but Tdry (the actual temperature at that moment) reads 4.0°C.    When it is reset, Tmin will also read 4.0°C.    Since we use (Tmax + Tmin)/2 to calculate the mean temperature for the day, which one should we use: Tmin or Tdry for that day?   This problem often occurs in winter in my experience.

What about Tmax; would that be better to be measured later in the day?

At my own (very rudimentary) met station – so rudimentary that I do not use capital letters – I usually read Tmin and rainfall at about 09:00 – 09:30 CET (GMT + 1), and Tmax later than 16:00 CET in the winter and later than 17:00 CEST (CET + 2) during the summer.   At the same time as reading Tmax, I reset Tmin so that, the following morning, I measure the true minimum temperature during the night. 

Where is my met station?   Strangely, although I am way south of the UK, the Greenwich Meridian runs through the centre of my village.

John Marlow   02/12/2025

Source: https://tallbloke.wordpress.com/2025/12/04/dustaffnage-addendum-the-story-on-the-ground/