Hi,
Thanks for your thoughts.
1) Drawing more air over the sensor, according to research, causes the sensor to become wetter. Just what research says.
2) There is more to it: the stronger the fan the more it influences the environment outside your screen, again according to research. So on still calm days and nights (I think especially nights) it is like you have some local wind blowing around your station. So a cold night will get less cold and a warm day less warm due to mixing of air, like wind.
3) I swear I tried, but I could not hear the Davis fan in my station 3 miles away. I thought it was my ears but I got in the car and pfoooh: it was broken...Nothing wrong with my ears it seems
4) I know you Americans have 3 holidays ever decade, but we Europeans have 30-40 such days per year (payed by our boss). So we go to places where the fan is not 40 feet away....In other words: how are you going to take care of your fan, in the summer (when people tend to enjoy a nice trip to sunny beaches etc) when it is most needed.
Now this seems far fetched, but at least I am pretty anal about trying to measure as carefree and accurate as possible. A fan that breaks down a fan that breaks down a calm atmosphere around the screen and/or a fan that creates a wetbulb is not helping me. But to each their own!
I can tell you that on cold bright nigts the Davis VP2 in my neighborhood measures higher temperatures than I do. Now he is 12 km away....But in the same very open flat countryside. And 2 km to his east is a KNMI Vaisala station and it also gets lower temperatures during the night than he does. In fact that very user is looking into the Barani, I think for that reason.
It might be better to have a screen with solar aspiration. No sun? No aspiration. But the construction of the Davis would block wind big time. So for a VP2 24h aspirated screen it is probably still not an option.
There is no ideal which is nice for curious people like me, who want to see if there is some ideal that is yet to be discovered. The Barani seems like an excellent effort and it is very interesting if indeed radiation errors are the cause of Jarle's disappointing results.
Edit: here is the research. Sadly this research was done in the most snowpoor period ever in The Netherlands still some interesting results that to some extend coroborate Jarles findings I just read! After 1995 until 2013 the snowcover was still below the average from 1901-1988, but a lot better than 1991-1995. But still some remarkable differences are noted during those days.
https://projects.knmi.nl/klimatologie/onderzoeksgegevens/homogeen_260/meulenbrandsma2007b.pdf