Satellite–Meteorological Data Fusion for Enhancing Short-Time Solar Irradiance Prediction

Abstract

Adequate prediction of short-term solar irradiance is necessary to have a reliable contribution of solar energy to power grids, but it is not an easy task since the atmosphere varies rapidly and is mainly influenced by clouds, aerosols, and local weather conditions (Perez et al., 2013; Yang et al., 2018). This paper introduces a satellite-meteorological data fusion system, which is created to improve the short-term prediction of solar irradiance at high time resolution. The suggested solution will combine the geostationary satellite measurements, such as optical properties of the clouds and radiative flux estimates, with ground measurements and reanalysis of meteorological variables, such as temperature, humidity, wind speed, and surface pressure (Schroedter-Homscheidt et al., 2016; Ineichen, 2014). The hybrid model attains data fusion, which involves the use of physical radiative relations alongside data-driven learning algorithms to obtain both the large-scale atmospheric patterns and the local variability (Voyant et al., 2017; Haupt et al., 2018).