Évolution spatio-temporelle de la sismicité de la partie Nord-Est de l’Algérie

Abstract

The seismic activity in Northeastern Algeria is characterized by the occurrence of swarms, often without a mainshock, or classic sequences comprising a mainshock followed by aftershocks. These seismic events exhibit strong temporal and spatial concentrations in both cases, lasting from days to months and spanning distances ranging from a few to tens of kilometers. Occasionally, these activities demonstrate a migration pattern over several kilometers across several weeks. While the advent of various seismological methodologies at CRAAG has enhanced seismic studies in Algeria, there remains a notable absence of detailed statistical analysis to characterize and categorize seismic sequences according to their spatiotemporal patterns and physical processes. This thesis aims to address this gap by employing robust statistical models to characterize earthquake sequences, focusing on quantifying their spatio-temporal patterns. It embarks on a comprehensive exploration of spatio-temporal pattern modeling, elucidating both mathematical and physical characteristics. Key aspects covered include determining magnitude completeness (mc), understanding magnitude distribution using the Gutenberg-Richter model, and examining spatiotemporal models like the Omori-Utsu law and the ETAS (Epidemic Type Aftershock Sequence) model. Additionally, it highlights physical models such as Rate-and-state friction, Coulomb failure, and time dependent stress rate. The thesis also scrutinizes various statistical analysis approaches, exploring the effects of declustering on parameter estimation and the influence of misestimating completeness magnitude on ETAS parameters. It further investigates techniques for achieving unbiased parameter estimation and explores the impact of disregarding aseismic forcing transients. Moreover, it compares the isotropic and anisotropic spatial distribution of aftershocks, shedding light on the complexities of seismic activity. Another significant aspect of the study is the detection and characterization of earthquake swarms, utilizing spatio-temporal ETAS models and moment release time series analysis. Employing two methodologies to identify swarm-like behavior, the research considers the tectonic and geological characteristics of the area, offering valuable insights into the nature of seismic swarms. Special emphasis is placed on the Beni-Ilmane earthquake sequence, one of the most important seismic events in Northeastern Algeria, in terms of the number of earthquakes produced and studies conducted. Through the application of statistical tests, the research endeavors to understand the underlying mechanisms driving the heightened seismicity observed in this sequence, drawing comparisons with previous physical and geological studies. In conclusion, this thesis significantly contributes to the understanding of seismicity in Northeastern Algeria. The analyses conducted provide insights into the spatio-temporal characteristics of earthquakes, crucial for predicting and managing seismic risks in the region. The comprehensive categorization of seismic sequences enhances targeted mitigation strategies, making a valuable contribution to the field of seismic research and facilitating informed decision-making for earthquake-prone regions.