A new technique for qualifying input data for building statistical models for satellite-derived hourly irradiances

摘要

The knowledge and quantification of the solar radiation coming to the earth surface is an important issue for many applications. However, there is no ground data available for every potential location of interest. In the best case, measurement stations are located about 200 km far from each other. In this regard, the use of geostationary satellites for deriving solar irradiances appears as a good method to obtain continuous information of the state of the atmosphere and on the cloud cover with a spatial resolution of a few kilometers. The different models available for finding the solar irradiance from satellite images can be grouped into two different approaches: statistical and physical (Noia et al., 1993). Statistical methods are based mainly on statistical regressions between satellite count and the corresponding measure at earth's surface. The physical methods are based exclusively on physical considerations on how interact the radiant energy with the earth's atmosphere. The work presented here is focused on statistical models for satellite-derived irradiances.