Importance of input data and uncertainty associated with tuning satellite to ground solar irradiation

Abstract

High quality satellite solar irradiation data is used throughout the solar industry to perform energy estimates. The uncertainty of the raw satellite data has been shown to be low. Ground data is often used to correct satellite data but determining the uncertainty of the final dataset could be challenging since the traditional statistical uncertainty and error calculation methods have proven to be unrepresentative. In this paper the limitations of traditional statistical methods are explored along with alternative approaches to calculate a more representative uncertainty value for a long term dataset resulting from ground corrected satellite data.