"Big Data" in Рredicting the Climate Resistance of Building Materials. Actinometric Indices

Abstract

The paper presents the results of analyzing the actinometric indices, in particular, the intensity of solar radiation and ultraviolet radiation of A and B bands. It also considers the distribution of actinometric indices depending on the day, month and year (2015-2022) for moderately continental climate (Saransk). The calculated values of total solar radiation obtained from the automatic control station, as well as the indices given in GOST 16350-80 for the temperate climatic region (Moscow) were compared. Three options of the distribution of solar radiation intensity were analyzed: data only for the day period and civil twilight (option 1); data without taking into account the night period (option 2); all indices recorded with a frequency of 10 minutes. It was found that despite the significant duration of the night and twilight periods, the average annual decrease in total solar radiation is 1.3% (option 1) and 0.8% (option 2). Based on the small values of actinometric indices corrections, taking into account the duration of astronomical and civil twilight periods, it was decided to use all the values of solar radiation and ultraviolet radiation intensities of A and B bands, recorded with a frequency of 10 minutes, in the development of machine learning models.