Otterman, I work in a company that produces ATIS systems though we dont usually do the instrumentation. Our systems just broadcast whatever the official 'observed' temperature is.
However it is my understanding that the 'official' way of mounting the temperature sensor is in a solar shield (such as the little white louvred Stephenson's Screen) which was about 4 ft above a grass surface. The modern sensors are often in round versions of the old wooden screen.
I do not know, and cant say I have ever noticed, temperature sensors mounted 10 metres or more above ground but if that is happening I would have believed it to be 'wrong'. Possibly a consequence of using isntrumentation packages some of which combine the sensors so they mount everything on the one pole position decided by the height of the anemometer which if I recall correctly is supposed to be mounted 10 metres above ground.
I am hardly suprised that the ATIS temperature measured in a solar screen over a grass surface is somewhat cooler than might be measured over a hard paved surface.
When I trained as a met observer there were well defined rules about positioning of such things but nowadays I see met instruments sticking out of the buildings and in the most unlikely places. Temperatures measured in the tower or on top of a building is just plain poor practice, IMHO.