Using satellite-based methods to predict daylight illuminance for subtropical Hong Kong.

Satellite-based methods are proposed to estimate global illuminance from geostationary satellite data under local climate conditions. The data used in this model are global irradiance and global illuminance collected from the International Daylight Monitoring Programme research class station in The Chinese University of Hong Kong and the visible channel data of the Geostationary Operational Environmental Satellite 9 from May 2003 to May 2005. The proposed methods consist of an indirect method, which derives global irradiance from satellite images first and then converts global irradiance to global illuminance by a luminous efficacy model, and a direct method, which derives global illuminance directly from satellite images. The root mean squared errors of hourly global illuminance are 35% and 32% for the indirect and the direct methods, respectively. [ABSTRACT FROM AUTHOR]