Thumbnail Image
Publication

Post-earthquake impact modeling with uncertainty propagation and Bayesian inference

Engler, Davis T.
Citations
Altmetric:
Advisor
Ganesh, Mahadevan
Jaiswal, Kishor S.
Editor
Date
Date Issued
2025
Date Submitted
Keywords
Bayesian methods
 Hamiltonian Monte Carlo
 spatial statistics
 uncertainty propagation
Collections
2025 - Mines Theses & Dissertations
Show all metadata
Files
Thumbnail Image
Engler_mines_0052E_13071.pdf
Adobe PDF, 129.78 MB
Research Projects
Organizational Units
Journal Issue
URI
https://hdl.handle.net/11124/181151
Embargo Expires
Abstract
Following a major earthquake, applying data-driven models is vital for quickly estimating potential damage, losses, and subsequent hazards, such as landslides and liquefaction, while also accounting for uncertainty. This approach is essential for effective post-disaster planning, rescue operations, and recovery efforts. This thesis enhances existing modeling methodologies developed by the U.S. Geological Survey (USGS) for three post-earthquake impact scenarios by efficiently incorporating uncertainties. These scenarios include: (i) ground motion estimates based on the rupture's characteristics and shaking data in the form seismic recordings and intensity observations; (ii) assessments of liquefaction probability; and (iii) fatality estimates for affected regions. We utilize Bayesian statistical methods to refine existing model estimates and investigate uncertainty propagation in ground motion estimation, analyzing its effects on downstream models, including fatality estimates. Our new modeling approaches, detailed in five chapters of the thesis, focus on: (i) partitioning ground motion uncertainty when conditioned on station data; (ii) evaluating the impact of uncertainty in ground motion forecasts for post-earthquake effects; (iii) assessing the influence of uncertain rupture dimensions on observation-conditioned ground motions and subsequent impact estimates; (iv) updating regional-scale geospatial liquefaction assessments with locally available geotechnical data; and (v) presenting an updated hierarchical Bayesian framework for earthquake fatality estimation. We validate all these methodologies using several recorded post-earthquake data sets.
Associated Publications
Rights
Copyright of the original work is retained by the author.
Embedded videos