Historic Fraser Noble Sunshine Recorder & Data

Until into the 21st century the Fraser Noble sunshine recorder was part of the national network of solar radiation stations that reported hourly sunshine data to the Met Office and thence to an archive kept by the WMO (World Meteorological Organisation). The recorder sensor was on the roof of the Fraser Noble building (opened by the University's former professor of Natural Philosopy and Nobel Prize winner, G P Thomson, in 1963) and the recorder housed at the top of the lift shaft in the East wing of the building. The head (i.e. the sensor) was noted as being 20.7 m above ground, 34.7 m above sea level at latitude 57.16° North. The horizon was pretty clear of obstructions except for some intrusion below 2° above horizontal West of South and from 1979 a chimney was erected some distance away that at times had a minor blocking effect on sunny winter days. We were one of the longest standing stations on the network, having been in operation since 1966 but the Met Office finally transferred sunshine records to Dyce Airport, which had for a long time been the official local recording station for all other local Met. Office measurements. The data from the whole network is available under the usual arrangements from the Met Office. The details given below are representative of our solar radiation station.

The recording head was a thermopile, housed under a double quartz dome as shown near right. It had to be kept clean on a regular basis of bird crap and muck from the nearby chimney and other places. The data was recorded by an automatic data logger built in-house and shown in the accompanying far-right image. It was programmed to integrate over hourly intervals in local solar time. For many years the data was transcribed from paper roll printout into tables. It was also recordered on chart recorder rolls for visual inspection. Latterly the data was collected by commercial datalogger and then downloaded daily by landline upon commands sent from near Bracknell in the South of England. In later years Met Office staff paid periodic visits to check the calibration. In earlier years we had two Kipp & Zonen CM heads, one of which was sent to the Met Office for calibration while the other was in use. The datalogger accurately converted the signal into the international standard units of kJ h-1 m-2. The monthly data shown in the graphs below are therefore plots of total energy received per hour in kilo Joules per square metre. The annual figures are shown in W h m-2.

Accuracy of calibration is essential when collecting data for comparison with other stations in the country or, indeed, around the world. Long-term stability is ensured by using a sensor of proven design from an agreed supplier and, of course, regular calibration checks. Our head was latterly a type CM11 supplied through the Met Office by Kipp & Zonen, whose detector is especially designed to record uniformly whatever the altitude or azimuth of the sun. Its spectral range is from 305 nm to 2800 nm (the 50% sensitivity points). The device therefore reads the incident solar energy flux from near UV, through the visible and including the near IR. It does not record the longwave radiant energy flux that reaches the Earth's surface from the atmosphere and clouds. This is effectively re-radiated solar energy.

Mk3 solarimeter

 

Below are some examples of monthly data in units of kJ per hour per square metre. Choose your month:

July 2000 August 2000 September 2000 October 2000  November 2000
December 2000 January 2001 February 2001 May 2001  

The accompanying spreadsheet gives monthly statistics for the 22 years from 1966 to 1987. They show the very large change in solar energy received each month as a result of seasonal change at our latitude of about 57° N. The spreadsheet also illustrates the year by year variability in any month. One 'Watt hour' (Wh) is equaivalent to 3.6 kJ.

The energy shown is the flux through a horizontal surface of 1 square metre. Solar panels in the best scenario can be tracked to follow the Sun, though most installations don't do this. However, anyone erecting solar panels at our latitude will not get this amount of electrical energy because the panels are sensitive to a narrower spectral range of energies than the solarimeter sensor and the conversion of radiant energy to electricity is unlikely to be as high as 20% efficient.

Average solar

Some further technical information

For those who would like to know a little more about our detector, it is of the Moll-Gorczynski type. At its heart is a blackened set of manganin-constantan thermocouples connected in series. The junctions at the edge of the surface are in good thermal contact with the base plate. The junctions near the centre warm up in response to the incident solar radiation. The excess temperature they reach, and hence the voltage produced, depends on the loss of excess heat by conduction to the base, by convection through the air and by radiation. This combined heat loss can't be accurately predicted from first principles so the device has to be calibrated. However, once calibrated, the simplicity of design and the structural stability of the materials give the solarimeter an intrinsic long-term performance stability. The double quartz dome isolates the crucial thermopile from the influence of the wind on the heat loss process. Devices like ours that measure radiation from all around are called pyranometers. Devices that measure direct solar radiation only are called pyrheliometers.

Schematic diagram of solarimeter detecting element

The response time of the device to changes produced by passing clouds is about 30 seconds. Electrically, the thermopile produces a voltage of about 5 microvolts for each W m-2 of incident energy. This signal is amplified and converted to a digital reading.

At night, the solarimeter reading is zero, corresponding to no radiation received in the sensitive wavelength range. A negative thermocouple voltage means that the thermopile is a bit cooler than its surroundings, which can happen when it is looking at a cloudless night sky. Exactly the same process is sufficient in winter to turn condensation into frost.

The only picture found so far of the makeshift solarimeter recording room at the top of the lift-shaft is the adjacent photograph taken in 1992 at the installation of a fluxgate magnetometer as part of a student project. The solarimeter equipment is bottom left. The boxes top right contain chart rolls showing the daily solar radiation variation that was recorded in addition to the hourly integrated figures produced on paper. Also present among the solarimeter equipment is a Campbell Scientific CR10 data recorder connected to several meterological instruments that were part of the datalogging project. Personel shown are, from the top, Prof. Ian Munro, Rhona McPherson (Hons student), Roger Clark (Snr Lect.), John Reid (Lect), Doreen Tyre (Sec), Joe Edwards (lect), Penny Robertson (Research student), Sandra Lonie and Alison Dunnet (Hons students).

For anyone wishing a detailed discussion of the amount of sunlight falling on differently angled surfaces over the day and year, see my article at http://homepages.abdn.ac.uk/nph120/meteo/Sunlight.pdf.

Solarimeter Recorder room

 

JSR

(Aug 2016)