Spatiotemporal properties of the 2020 – 2021 Petrinja (Croatia) earthquake sequence

Abstract

The Petrinja earthquake sequence started on December 28, 2020, with a destructive M_L 6.2 mainshock occurring in the area, preceded by a M_L 5.08 foreshock, following a long period of relative seismic quiescence. Over the first six months of the Petrinja earthquake sequence, almost 14,000 events were recorded. In the present work, we separated seismic events based on their spatial concentration using a density-based clustering algorithm, DBSCAN. We identified four main clusters and analyzed their spatiotemporal properties using the notions of Non-Extensive Statistical Physics (NESP). This framework, which relies on Tsallis entropy (S_q), describes the scaling behavior of complex systems. In this frame, we investigated the inter-event time (T) and distance (D) distributions, providing the q_T and q_D entropic parameters, respectively. Additionally, we studied the frequency–magnitude distributions in terms of the fragment–asperity model, leading to the determination of the non-extensive parameter q_M. The results of the analysis suggest that the statistical properties of the Petrinja earthquake sequence can be effectively reproduced utilizing NESP. Furthermore, the coseismic static Coulomb stress changes were estimated, indicating that the clusters’ seismic events may have resulted from a complex fault system's (re)activation. In addition, the effective static stress drop was estimated for each spatial cluster. Lastly, the temporal patterns of the earthquake evolution are discussed using the superstatistics approach, indicating that the temporal progression of the Petrinja earthquake clusters is governed by a very low number of degrees of freedom, highlighting the spatiotemporal organization of each cluster.