Neural network clustering of a set of geophysical data in comparison with the distribution of earthquake focuses in the middle basin of Lake Baikal

The high degree of earthquake clustering observed in the central part of the Baikal rift zone necessitates an investigation into the conditions conducive to such a distribution of seismic events. The crust peculiar properties, as characterized by geophysical data, affect the course of the major tectonic and seismic processes. To study the contribution of different types of geophysical data to the formation of seismic event grouping structures in the study area, clustering based on the Kohonen neural network was performed. Clustering was performed on parameters such as the propagation velocities of longitudinal, transverse seismic waves and their ratio, resistivity, and temperature. As a result, clusters outlining zones of increased seismic activity were identified through the combination of temperature and resistivity.