A Markov chain approach for earthquake sequencing in the Aegean Graben system of Turkey

Ünal C., Özel G., Eroglu Azak T.

Earth Science Informatics, vol.16, no.2, pp.1227-1239, 2023 (SCI-Expanded) identifier identifier

  • Publication Type: Article / Article
  • Volume: 16 Issue: 2
  • Publication Date: 2023
  • Doi Number: 10.1007/s12145-023-00961-5
  • Journal Name: Earth Science Informatics
  • Journal Indexes: Science Citation Index Expanded (SCI-EXPANDED), Scopus, Academic Search Premier, CAB Abstracts, Geobase, INSPEC
  • Page Numbers: pp.1227-1239
  • Keywords: Markov chain, Stochastic process, Probability, Prediction, Earthquake, K-means cluster analysis
  • Hacettepe University Affiliated: Yes


© 2023, The Author(s), under exclusive licence to Springer-Verlag GmbH Germany, part of Springer Nature.Earthquakes are prominent natural hazards that cause loss of life and property in residential areas in the world. The Aegean Graben system is geologically under the influence of the Western Anatolian Opening Regime. There are many active fault lines that can be a source of earthquake activity in the Aegean Graben system. Many earthquakes have been recorded in the region, which have caused serious material and moral damages from the past to the present. This situation reveals the importance of investigation this region in detail and by this way is possible to reduce the risks and consequences of earthquakes. In the study, earthquake data from three different earthquake catalogs were examined, and the earthquakes recorded between 1970 and 2020 in the region were declustered and homogenized in the scale of magnitude wave (Mw). Then, the data are clustered based on the seismic classification via k-means cluster analysis. By this way, the predictions for sequential transition of earthquakes within time and magnitude of the earthquakes are analyzed with a stochastic model, namely Markov chain. It has been concluded that the row probability values of the transition matrices of the earthquake transition probabilities between the clusters of the Aegean Region are fixed values depending on time. We also found an equilibrium distribution for the region. Hence, the probability of an earthquake in Cluster 1 in the region is 30%, Cluster 2–33% and Cluster 3–37%, respectively. We then obtain mean passage time of the earthquakes in each cluster. The results suggest that all clusters have similar characteristics as regard to earthquake sequencing. For these reasons, we finally study the without clustering to find the magnitude transitions depending on the time with Markov chain analysis.