Cloud Variability and Cloudless Percentage Days Calculation Using Cluster Analysis
Abstract
Cloudiness variability across the Brazilian territory was analyzed using cluster analysis based on daily values of the C eff and DARR indices. Cloud reflectance data with a 10-minute resolution, collected between 2020 and 2024, was employed to compute these daily metrics. A subset of the data exhibiting the lowest values for both C eff and DARR was selected to calculate the Cloudless Percentage Days ( CPD ) index, which quantifies the most favorable locations for solar energy utilization across the region. Northwestern Minas Gerais and western Bahia emerged as the top-performing areas for solar energy generation during the study period ( CPD > 0.65). Additionally, the Central region of Brazil recorded moderate to high CPD values during the dry season.
Keywords
Solar resource, Cloud variability, Cloudless days, Cluster analysis