This work evaluates objectively the consistency and quality of a 9 year dataset based on 5 minute average values of downward longwave atmospheric (LW) emission, shortwave radiation, temperature and relative humidity. All these parameters were observed simultaneously and continuously from 1997 to 2006 in the IAG micrometeorological platform, located at the top of the IAG‐USP building. The pyrgeometer dome emission effect was removed using neural network technique reducing the downward long wave atmospheric emission error to 3.5%. The comparison, between the monthly average values of LW emission observed in São Paulo and satellite estimates from SRB‐NASA project, indicated a very good agreement. Furthermore, this work investigates the performance of 10 empirical expressions to estimate the LW emission at the surface. The comparison between the models indicates that Brunt’s one presents the better results, with smallest “MBE,” “RMSE” and biggest “d” index of agreement, therefore Brunt is the most indicated model to estimate LW emission under clear sky conditions in the city of São Paulo.

This content is only available via PDF.
You do not currently have access to this content.