Spatio-temporal dynamic clustering modeling for solar irradiance resource assessment

Nowadays, with the development of international policies and agreements to promote the integration of renewable energy sources, mainly solar and wind, modeling the solar resource by including the spatio-temporal variability is crucial to determine future PV power plant locations and estimate potential power generation performances. However, contributions involving long-term periods and different time windows to explore such potential solar resource variability are generally scarce. Under this framework, the present paper proposes a methodology focused on characterizing and clustering the spatio-temporal solar resource variability through the global horizontal irradiance analysis. Hierarchical clustering technique is firstly used to classify the spatial data. Different time windows — from short-term to long-term data — can be subsequently evaluated by using various sources of information. The Spanish territory is selected as case study, considering 22-year period data (1999–2020) and 1,936,917 observations from online satellite database. Spatial variability and geographical clustering differences are discussed and compared depending on the selected time windows, identifying relevant spatial variations for some specific months. Additionally, some years present more variability as well, in line with the sunspot peak of the solar cycles. The proposed approach gives an alternative comprehensive spatio-temporal clustering and characterization of GHI evolution, providing a suitable methodology to help the current European sustainable energy transition.