Platinum Resistance Thermometers – Are they suitable for weather readings?

Image: The Platinum Resistance Thermometer (PRT) that recorded the 2019 Met Office claimed all-time temperature high at CIMO Class 5 Cambridge Botanical Gardens.

Dr Eric Huxter has provided me with assistance and support in the past regarding data research for the Surface Stations Project. He is equally sceptical of much of Met Office claims and extensively researches noted daily “extremes” readings against weather stations CIMO ratings and known instrumental errors. By way of a “Guest Post” I reprint in full his recent posting on his own blog Frayedends as it is highly relevant to many issues raised in the Surface Stations Project.

https://frayedendsblog.wordpress.com/2025/05/31/spikey-are-prts-actually-fit-for-purpose/

The replacement of LIGTs by PRTs since the 1990s has coincided with an increase in recorded temperature, consistent with the narrative of Global Warming aka Climate Change.

It is becoming apparent that that this recorded increase is an artefact of the change in observation method rather than a real change in temperature.

It has been known for a long time that the Stevenson Screen is the least worst method of recording actual air temperature at 1.5m above the ground, being the temperature in the shade, but the record is subject to error induced by the Screen, especially the Aitken Effect (Harrison 2024). Under certain weather conditions Aitken proved the screen was found to over-record temperatures often by 1°C and in particular extremes as much as 2°C.  Burt (2022) quotes up to 3°C. This error has been enhanced by the use of PRTs, whose transient response bias is well known.

Burt (2022) has shown that for 99.99% of the time Stevenson Screens are not sufficiently ventilated

‘current assumptions of ventilation speeds within Stevenson screens are too optimistic; the ISO 17714 assumption of 1 m s−1 airflow (International Organization for Standardization, 2007) was attained for only 0.01 % of the 3-month experimental period.’ (Burt 2022)

and that the 10m wind speed to ensure minimum ventilation would have to be 51.4kph, not a common wind speed even in the UK.

Although current Meteorologists claim that there is no issue with the Stevenson Screen nor PRTs (Harrison 2024, Allot 1999), quoting small percentages of significant diversion, it should be noted that it is in theses extremes that Records and Trends lie. Harrison (2024) quotes diversions from actual temperature, due to the Aitken effect, of >1°C as 1% and Allot 3% measurements LIGT vs PRT differing by >1°C (promising a follow up paper as to why the difference, which I have yet to find).

Since the PRT data conforms to the Meteorological paradigm of rising temperatures there is little incentive to scientifically investigate the reality behind it, which seems to be the spikiness of PRTs given the Aitken effect and Transient Response bias ( plus the louvre bias of trapping rising warmer air).

This spikiness has already been noted on the 1st May data from Kew gardens, setting a new new UK Daily maximum Temperature Record and can be illustrated by the data from Heathrow. Since I started tracking these data on the 28th April Heathrow has recorded 5 daily UK maxima, which differ significantly from the hourly I minute data. An aside, but it is irritating that there is frequent data drop out from weather stations, around 1300 usually, which makes tracking these data difficult.

The graph shows the hourly data from Heathrow and the daily maxima, showing the spikey nature of these records.

Tracking the data over the past 35 days shows the following:

The mean difference of the 35 days to date is 0.87°C but what is the hourly variation recorded by PRTs? The Meteorological Office uses 1 minute averages, which will make the data more sensitive to transient temperature changes but how much does air temperature change naturally over 60 minutes?

Given the blocking high in Mid May, subjecting the UK to fairly uniform heating and cooling conditions, 25 stations, covering the record setting stations and those in proximity) were selected for analysis using a week’s data from weatherobs.com. An average change per hour (rising and falling temperatures) was calculated for each station and then averaged per CIMO Class (1:3 Stations, 2:1 Station, 3:2 Stations, 4:7 Stations, 5 12 Stations), with the following result:

This suggests that CIMO Class affects the response of the PRTs, with poorer siting producing spikier results, illustrating the impact of the local environment over the regional. It is interesting that one of the CIMO 3 stations is Heathrow, whose 1.48°C/hour change would place it as an extreme CIMO 5 station. Who knew?

Clearly the uncritical adoption of PRT technology by the Meteorological establishment has had an impact on temperature recording and raises questions as to the comparability of older LIGT data sets with the more recent PRTs. The reliance on 1 minute data is also open to question, although as shown by Kew, the use of these data to reinforce the Climate Change narrative is undoubtedly of use to those wedded to alarmism.

References

Allott, T.P. 1999. A preliminary comparison of conventional and electronic air temperature thermometry. Climate Studies Memorandum, No. 7 ,  Met Office (Unpubl)

Allott, T.P. 1999. Comparison of Conventional and Electronic Air Temperature Thermometry. Climate Studies Memorandum, No. 10 ,  Met Office (Unpubl)

Burt, S.D. 2022 Measurements of natural airflow within a Stevenson screen. Geoscientific Instrumentation, Methods and Data Systems 11:2, pp 263–277

Harrison, G. 2024. Victorian technology for measuring the weather is still remarkably accurate – new research. The Conversation August 16

Source: https://tallbloke.wordpress.com/2025/06/02/platinum-resistance-thermometers-are-they-suitable-for-weather-readings/