In this thread I will make observations regarding the use of this screen (which may be valid for other screens as well).
My first observation is based on the charts screenshot you see below (Weerkast === Stevenson Screen). The Temperature is the minute value of the main sensor and Temp Weerkast is the temperature measurement inside the Stevenson Screen.
The (interesting) observation is that the Temp inside the Stevenson Screen is lower when rising and higher when going down compared to the main Temperature measurement. This cannot be a device measurement error : that should have a constant deviation. So I explain this by the screen itself absorbing and giving off heat and because of the mass of the screen it is lagging behind. On a 'normal' descent rate this appears to be around 0.3 °C but on the steep rise in the morning it apparently can be up to 1.5°C, quickly catching up when the rise slows down.
I currently don't have the time to analyse the effect on the average, but will do (much) later.
Will take some top and bottom planks out to open up and have more vertical circulation and see what that effect will be.
(A man with one thermometer knows what temperature it is; A man with two thermometers is never quite sure. )