The results of an experiment designed to test the viability and relative contribution of rooftop weather stations (~1.5 m) on mesonet network observations is investigated. Summertime rooftop temperatures are monitored in Tallahassee, Florida using fan-aspirated wireless solar powered automated weather stations with radiation shields. The stations are also operated without radiation shield and/or fan-aspiration to compare the effects of two important sources of error on daytime and maximum temperatures. In addition, comparisons are made with another similar station much farther off the rooftop (10 m) and a nearby ASOS station.

It is clearly demonstrated that a well-sited, ventilated instrument is capable of producing very high quality temperature reports even relatively close to a strongly radiating roof surface, in particular for daytime temperatures. This reliability is not expected based on conversations with operational meteorologists and climatologists, but is quite robust. In addition, rooftop minimum temperatures are found to still contain a bias that remains, but could potentially be correctable.

For mesonet operators or those who use mesonet observations from networks such as the Citizen Weather Observer Program, or AWS networks, it is advisable to consider siting, instrumentation, and bias corrections before using these data for analysis. In addition, valuable tools available from NOAA's MADIS program provide essential quality control information that should be considered, to help counter effects of different elevations and temperature sensors by reporting potential temperature quality control statistics, for time periods from hourly to monthly.