Seismic hazard analysis of China’s islands based on Bayesian network

The majority of China’s islands are located at the intersection of the Asian-European and Pacific tectonic plates, which are exposed to intense seismic hazards. The previous research on seismic hazards primarily relied on data on faults, geology, historical earthquakes, etc. A systematic approach using more comprehensive data that considers various factors linked to earthquakes for Seismic Hazard Analysis (SHA) on islands is expected. Based on previous seismic hazard research, we collected and processed essential data for the islands’ SHA, including basic data about islands, fault data, topographic and geomorphological data, seismic zone data, deep tectonic and geophysical field data, and historical seismic data. Then, we established a comprehensive index system comprising 16 indicators to reflect the seismic hazard of islands. Afterwards, considering the uncertain relationships among various indicators, the inherent uncertainty of seismic hazard, and the challenges in integrating multi-source data, we proposed a Bayesian network-based model for islands’ SHA. This model integrated diverse earthquake-related data types and quantified interrelations among the 16 indicators to characterize the islands’ earthquake probabilities. Then, we applied this model to evaluate the seismic hazard levels of 100 representative islands in China’s marine areas. Furthermore, we validated the results by comparing them with the Seismic Ground Motion Parameters Zonation (SGMPZ) of neighboring districts. The method proposed in this study could serve as a reference for the islands’ SHA, which could contribute to earthquake disaster prevention and reduction.