The paper presents the results of applying the Discrete Perfect Set (DPS) topological filtering algorithm to analyze the spatial clustering of seismic epicenters in the Lake Baikal region. The study utilizes earthquake data recorded by the seismic network of the Baikal Branch of the Geophysical Survey of the Russian Academy of Sciences within the latitude range 48°N, 58°N and longitude range of 99°E, 122°E for the period from 1964 to 2018. Clustering characteristics are obtained for (i) the recording period from 1989 to 2018 with varying parameters of the DPS algorithm and four levels of the minimum energy class KР of seismic events and (ii) six non-overlapping time intervals from 1964 to 2018 and seismic events of energy class KР ≥ 8.6 with fixed parameters of the DPS algorithm.
The dynamics of the clustering parameters from 1964 to 2018 may characterise the variability of the seismic regime of the region. Specifically, the decrease in the linear size of the areas of identified epicenter groups from about a thousand km to tens km may indicate a significant change in the seismic regime of the Lake Baikal region at the end of the 1990s and the beginning of the 2000s compared to the period between 1964 and 1997.