EDMOND DUSHI, BESIAN RAMA, KLAJDI QOSHI
KEYWORDS : moment magnitude, calibration, earthquake catalogues, seismic hazard assessment, moment tensor inversion.
ABSTRACT :
This study focuses on calibrating the moment magnitude-local magnitude (Mw-ML) relation using seismic data from the Albanian Seismic Network, spanning the last three years. Our approach includes back-processing specific events to expand the dataset, aiming for a comprehensive analysis. Utilizing GISOLA real-time moment tensor inversion software within the Seiscomp (v.4) seismic monitoring system, we compute Mw values. Concurrently, ML values are derived using SEISAN (v.12), following the methodology of Hutton and Boore (1987), with manual amplitude checks for validation. The findings reveal a strong correlation between Mw values computed through GISOLA, utilizing full waveform data for moment tensor inversion, and ML values obtained through routine location in SEISAN. Notably, exceptional efforts in back-processing specific events contribute to enriching the dataset, enhancing the reliability of the calibration. The best correlation scores are observed for moment tensor solutions of quality A and B, corresponding to the highest variance reduction. These results are represented by high correlation coefficients and low correlation standard deviations, indicating the robustness of the calibrated Mw-ML relation. Our study contributes to the development of comprehensive earthquake catalogues for Albania, crucial for seismic hazard assessment and risk mitigation. By refining the Mw-ML relation, we enhance the accuracy of magnitude estimation, facilitating more informed assessments of seismic hazard.