Fine-scale spatiotemporal earthquake casualty risk assessment considering building function types

Abstract

Lessons from recent global earthquakes underscore the persistent high risk of seismic casualties. Despite considerable attention to earthquake casualty risk assessment (ECRA), studies adopting fine-scale spatiotemporal ECRA remain limited. We argue that fine-scale ECRA would benefit from considering building function types (BFTs) due to their significant impact on population dynamics and indoor occupancy rates. This study uses an intelligent BFT extraction model to identify BFT clusters with similar population dynamic patterns, and proposes a fine-scale ECRA framework based on the explicit integration of BFTs with detailed nearly real-time population data. Using Beijing as an example, we modeled fatalities under 240 (5 × 24 × 2) earthquake scenarios with intensities from VI to X on weekdays and weekends by hour, with a resolution of 500 m, and quantified the impact of BFTs on ECRA. Findings of seismic intensity X reveal that: (1) the seismic casualty risk of Beijing exhibits notable spatiotemporal differences, with approximately 85 % of fatalities occurring within the 6th Ring Road and approximately 58.06 % more fatalities at 03:00 than at 10:00; (2) by considering BFTs, the accuracy of the ECRA framework improved by up to 47.59 % compared with traditional methods. Moreover, this study demonstrates the applicability of the proposed method by employing it to estimate the potential fatalities in Beijing using the 1679 Sanhe-Pinggu M_S 8.0 earthquake scenario. Our method provides nuanced multi-scenario considerations for earthquake preparedness and response, offers insights into the application of BFTs in ECRA, and highlights the utility of intelligent technologies and big data for maintaining current assessment models.