Probabilistic Seismic Hazard Analysis for China Based on Bayesian Network

Abstract

Bayesian network (BN) has important applications in disaster risk analysis due to its unique causal structure and probabilistic characteristics. This research begins with a detailed introduction to probabilistic seismic hazard analysis (PSHA) for China, and the utilization of BN-based modeling for seismic hazard and risk assessment. Subsequently, a comprehensive theoretical exposition of PSHA for China based on BN is presented. This includes a clear explanation of the three-level subdivision of seismic sources and the employment of the elliptical ground-motion model (GMM) in China. Regarding BN modeling, the values, conditional probabilities, and the impact of subdivisions of the nodes are carefully discussed with the assistance of a specific example from China. The advantages of BN in terms of both holistic and probabilistic computation are then demonstrated through the disaggregation of seismic hazard and various sensitivity analyses. Finally, the article concludes by summarizing its content, highlighting the advantages of BN, and outlining future work.

Aftershock probabilistic seismic hazard analysis based on enhanced Bayesian network considering frequency information

“…Therefore, to study the BN modeling approach for each process of this system is essential. Liu and Lu (2023) conducted a detailed BN-based PSHA for China, modeling it from a holistic perspective and demonstrating the computational benefits through hazard disaggregation and sensitivity analysis. This paper aims to further utilize BN for modeling and analyzing APSHA.…”