Analysis of Depth and Magnitude Heterogeneity in Cluster-Based Segmentation of Andaman and Nicobar Earthquake Data

Abstract

The dimension of space (depth from ground) and magnitude of each earthquake play a crucial role in explaining its characteristics. Our study examines seismic characteristics in the Andaman and Nicobar region using power spectral density (PSD) analysis and probability density function (PDF). The dataset is segmented into clusters using the K-means clustering algorithm, determined through the silhouette score method. Analysis of depth shows dominant frequencies concentrated in low-frequency ranges across clusters 1, 2, and 4, indicative of interactions at plate boundaries or subduction zones. Clusters 3 and 5 show heterogeneity in dominant frequencies, likely linked to diverse geological structures and seismic sources. Conversely, analysis of magnitude demonstrates a concentration of seismic energy at lower magnitudes in Clusters 1 and 2, with wider magnitude ranges observed in Clusters 3 to 5. PDF analysis identifies higher probabilities of seismic occurrences, particularly at depths of 29 to 30 km across Clusters 2 to 5. Additionally, Cluster 1 displays localized seismic activity around 12 km depth, potentially associated with unique geological features. Furthermore, the clusters show varying probabilities of earthquake occurrence, indicating distinct seismic hazards across the region. These findings carry significant implications for earthquake hazard assessments and mitigation strategies, offering valuable insights for policymakers, seismologists, and disaster management authorities in enhancing earthquake preparedness and infrastructure resilience in the Andaman and Nicobar region.