2023 , Volume 28, № 6, p.46-56

Summary. The purpose of this research is to identify climate classes for the Baikal natural territory during various time intervals.

Methods. The study uses climatic characteristics such as surface temperature and atmospheric pressure. A mathematical model of the climate signal was constructed in the form of a quasi-periodic oscillation caused by the global annual cycle of the Earth’s rotation, modulated in amplitude and phase. The carrier oscillation has a period equal to one calendar year. An essential task is to establish patterns of changes in the Earth’s climate system based on the parameters of the amplitudes and phases of climate characteristics. For this purpose, a dynamic clustering algorithm was employed.

Experiment. An array of climate data was generated, which included the following meteorological characteristics: temperature and pressure, measured at weather stations located in the Baikal natural territory (BNT). The purpose of the experiment is to construct a characteristic for classes of BNTs and assess their variability with regard of the amount of spatial structure over time using the developed algorithm.

Results. The presented results have shown that the composition and structure of climate clusters is changed during various time intervals. On average, for all data and periods, it was possible to determine two conditional places of stable formation of clusters.

Conclusions. The advantage of the proposed method is that it allows, based on the determination of typical patterns, identifying unique clusters without any a priori information about the number or other parameters of clusters. The resulting climate classifications can be used to assess changes in regional climate at various time and spatial scales.