Statistical functions and relevant correlation coefficients of clearness index

This article presents a statistical analysis of the sky conditions, during years from 2010 to 2012, for three different locations: the Joint Research Centre site in Ispra (Italy, European Solar Test Installation – ESTI laboratories), the site of National Renewable Energy Laboratory in Golden (Colorado, USA) and the site of Brookhaven National Laboratories in Upton (New York, USA). The key parameter is the clearness index k T , a dimensionless expression of the global irradiance impinging upon a horizontal surface at a given instant of time. In the first part, the sky conditions are characterized using daily averages, giving a general overview of the three sites. In the second part the analysis is performed using data sets with a short-term resolution of 1 sample per minute, demonstrating remarkable properties of the statistical distributions of the clearness index, reinforced by a proof using fuzzy logic methods. Successively some time-dependent correlations between different meteorological variables are presented in terms of Pearson and Spearman correlation coefficients , and introducing a new one.