Modeling monthly mean variation of the solar global irradiation

The monthly mean variation of the solar global reaching the Earth's surface has been characterized at a global level by a regression model. This model considers the monthly variation itself (to different horizons and even the maximum annual variation) as the study variable, and it is applied without using data corresponding to measured meteorological variable. Two explicative variables have been used, the variation of the extraterrestrial irradiation and the variation of the clear sky global horizontal irradiation. The work has been carried out from datasets including average global daily solar irradiation for each month of the year measured on the ground. The model quality has been proven to be very dependent of the temporal variation considered, in such a way that higher variations, that is to say, higher distances between months, lead to an improvement in the model outcomes.