Identification of clear days from solar irradiance observations using a new method based on the wavelet transform

A new method using the wavelet transform properties is developed to determine clear days of solar irradiance. These days are needed to model the solar radiation and to compare the existing empirical models. We use this method to process four years of global solar irradiation data collected at the Research Unit of Applied Renewable Energies at Ghardaïa city in Algeria. We also determine clear days from this data set using a standard method based on the clearness index criteria. The results show that the two methods give different numbers of clear days. The effect of this difference is analyzed by computing the Global Solar Radiation (GSR) with the Iqbal C model but also by the estimation of turbidity parameters using for that a innovative approach. We find that some significant differences are observed in the GSR modeling leading to bad estimation of turbidity parameters. We conclude that using our method is therefore more efficient since it is not dependent of the site and observations.