Dalwhinnie no 2 Addendum – Extreme Variations

Following on from my report on Dunstaffnage , I now contrast that coastal site’s overlapping data from manual to automatic operation with that of an inland high altitude station, Dalwhinnie‘s transfer as above.

As in previous reports on the artificial appearance of rising minimum temperatures due to observation protocol change, the above graph clearly shows the older once only daily reading over-recording colder weather events. The scale of variations at Dalwhinnie (as at Cassley) can reasonably be described as quite spectacular with extremes reaching a staggering 14°C. To put this error into context consider this statement from a highly experienced meteorologist.

“Ten days of +5 difference with 80 days near zero difference gives a net offset of +0.5.”

With that scale of distortion potential in mind, consider these numbers.

The X axis is detailing the number of days of discrepancies. We should soon be able to calculate the amount this will move averages, though it is rather obvious that this frequency and scale of differences will move averages very significantly.

How on earth can the Met Office “square up” this level of discrepancy showing averages to the 100th of a degree for “Climate Averages” site?

Looking at the number of differences by month (Y Axis, 1 to 12 refers to January to December) also reveals significant numbers of days where the PRT is recording colder (negative numbers) than than the LIGT . This should never happen unless something is going very seriously wrong.

Comparing the above graph to coastal Dunstaffnage shows the remarkable contrast – what should not happen (the PRT readings being lower than the LIGT) only happens very rarely at Dunstaffnage.

I feel this is further evidence supporting my assertion that opening the screen door to take manual readings is exposing the PRT to the real outside temperature to which it rapidly responds. The Victorian technology screen is not functioning correctly in the low wind speeds prevalent on colder inland sites usually overnight and is insulating daytime warmed entrapped air inside. Conversely the coastal Dunstaffnage site will very rarely encounter such low wind speeds and avoid such distorting effects.

I further assert this effect is well known to the Met Office, however, they are wilfully ignoring this problem. Evidence to support my assertion is their retention of effectively redundant LIGTs (which are solely used for minimum reporting) when a full 1,440 individual minute reading PRT and Data Logger is fitted at manual stations.

Where stations transition to automatic reporting, no minimum LIGT is used as it is not required. However, in such circumstances the screen door is rarely opened to reveal the problem effects associated with that opening. Allowing the PRT to take over all readings in a manual station would (and I argue has already) produced irrefutable evidence of the shortcomings of antiquated screen technology by exposing the problem as demonstrated in these few overlapping stations data. It is very notable that no recent conversions show any overlapping data, but as I showed at Cambridge stations the effect is ongoing.

Additionally, this would negate, in terms of the crude meteorological averaging used, the overheating effects equally associated with such technology. All of these problems can be completely removed by artificially aspirating screens as stipulated by the World Meteorological Organisation – that the Met Office does not do so seems remarkably suspicious.

Source: https://tallbloke.wordpress.com/2025/11/28/dalwhinnie-no-2-addendum-extreme-variations/