Advancing earthquake ground motion estimation for urban risk mitigation using physics-based simulations

Abstract

As global urbanisation surges, with an estimated 2 billion people expected to move to cities over the next 30 years, many of these new urban centres will arise in regions of high seismic risk. This thesis explores the critical role of advanced geophysical understanding and innovative earthquake risk assessment methods in safeguarding these growing populations, contributing to both the advancement of seismic science and its application in the context of risk assessment for urban centres.

The first part of this thesis focuses on investigating earthquake ground motion in intermontane basins located within active orogenic regions, using the Kathmandu sedimentary basin as a case study domain. These basins, formed by tectonic activity, are subjected to significant seismic hazards, while presenting complexities in ground motion prediction due to their distinct geological and topographic characteristics. While the flat sedimentary layers facilitate infrastructure development, they also amplify earthquake ground motions, heightening the risk.

Using numerical physics-based simulations (PBS) of hypothetical shallow sources, it is shown that the surface topography surrounding the Kathmandu basin acts as a shield for incoming waves, resulting in up to a 40% reduction in peak ground acceleration (PGA). However, this reduction is: (a) frequency-dependent, with higher frequencies being more attenuated; (b) varying with the source azimuth relative to the basin; (c) dependent on source depth; and (d) influenced by the topographic geometry. Additionally, the deeper basin layers play a crucial role in controlling spatial variability in ground motion, emphasising the importance of detailed subsurface information for accurate seismic hazard analysis.

The second part of this thesis addresses the limitations of traditional ground motion models, which rely heavily on extensive earthquake catalogues to derive Ground Motion Prediction Equations (GMPEs). In many developing regions, such comprehensive data is unavailable. To overcome this challenge, physicsbased ground motion simulations are used to develop a simplified deterministic decomposition of ground motion estimation, focusing on regional attenuation and local site amplification parameters, while excluding extended earthquake sourcespecificic information. For a virtual city located in a sedimentary basin setting as a testbed, it is shown that the PGA obtained through the deterministic evaluation method correlates strongly with simulated PGA, revealing significant spatial variability in the city that is primarily driven by near-surface geology. This approach offers a potential solution for hazard estimation in urban centres, which lack detailed seismic event records.

Recognising the urgent need for effective pro-poor earthquake risk reduction strategies, the final part of this thesis showcases a decision-making framework that integrates hazard-related geophysical information with physical and social vulnerabilities in an urban unit. This framework, supported by cost-benefit analysis, provides a practical tool for policymakers to evaluate and implement risk reduction strategies that are both effective and economically viable. This research considers the exposure bias faced by low-income populations, who are often forced to reside in hazard-prone area due to cheaper land availability, such as low-lying floodplains, which often have high ground motion amplification due to sediment deposits. The framework is implemented on a well-established virtual urban testbed and the results show that the policies informed by detailed local geophysical data can efficiently mitigate earthquake risks, particularly for the poor populations living on softer soils.

In summary, this thesis bridges the gap between geophysical research and its practical application for equitable earthquake risk reduction. It offers a comprehensive approach to seismic risk mitigation in rapidly urbanising areas by integrating advanced simulation techniques for detailed ground motion analysis and utilising these to inform decision-making processes.