Machine Learning-Assisted Multi-Hazard Fragility Surface Dataset for Earthquake-Tsunami Resilience Analysis

This project presents a data-driven machine learning (ML) methodology for generating multi-hazard earthquake–tsunami fragility surfaces, enabling improved community resilience analysis. The ML model synthesizes 3D fragility surfaces from randomly selected 2D fragility curves for earthquakes and tsunamis, facilitating probabilistic damage assessment and loss estimation in seismic and coastal hazard scenarios.

Key features include:

- Physics-informed ML integration: Enhances the reliability of fragility surfaces for earthquake-tsunami hazard interactions.
- Computational efficiency: The ML algorithm significantly reduces computational time compared to traditional physics-based fragility generation.
- Structural retrofitting analysis: Fragility surfaces can be adjusted to evaluate the effectiveness of mitigation strategies.
- Applicability to community resilience modeling: The framework can be integrated into multi-hazard disaster risk reduction and urban planning strategies.

Reusability:

- The ML-based fragility model can be adapted to other multi-hazard scenarios, including sequential hazards such as hurricanes, floods, and wildfires.
- The dataset provides benchmark fragility functions for structural vulnerability modeling in performance-based seismic and tsunami engineering.
- The methodology supports Monte Carlo-based risk assessment frameworks and can be implemented in IN-CORE and other resilience modeling platforms.

References:

Harati, M., & van de Lindt, J. W. (2024). "Data-Driven Machine Learning for Multi-Hazard Fragility Surfaces in Seismic Resilience Analysis." Computer-Aided Civil and Infrastructure Engineering. DOI: 10.1111/mice.13356.

Harati, M., & van de Lindt, J. W. (2025). "Fragility Function Development of RC Building Portfolio for Use in Earthquake-Tsunami Community Resilience Studies." Journal of Performance of Constructed Facilities, ASCE (Accepted - In Production).

Harati, M., & van de Lindt, J. W. (2024). "Mainshock-Aftershock Building Fragility Methodology for Community Resilience Modeling." Journal of Structures. DOI: 10.1016/j.istruc.2024.117700.

Harati, M., & van de Lindt, J. W. (2024). "Community-Level Resilience Analysis Using Earthquake-Tsunami Fragility Surfaces." Journal of Resilient Cities and Structures, 3(2), 101–115. DOI: 10.1016/j.rcns.2024.07.006.

Harati, M., & van de Lindt, J. W. (2024). "Methodology to Generate Earthquake-Tsunami Fragility Surfaces for Community Resilience Modeling." Engineering Structures, 305(15), 1–15. DOI: 10.1016/j.engstruct.2024.117700.