Multichannel Seismic Tomography for Subsurface Fault Characterization in Urban Areas

Thesis Abstract

Urban environments often conceal complex subsurface fault systems that pose significant seismic risk to infrastructure and populations. The study addresses the challenge of accurately characterizing subsurface faults beneath densely built areas where traditional geophysical methods are hindered by noise, shallow heterogeneity, and logistical constraints. The aim is to develop and validate a robust multichannel seismic tomography (MST) workflow to delineate fault geometries and assess seismic velocity contrasts that indicate fault zones. Specific objectives include (1) acquiring high-resolution seismic data across a network of 40–60 vibroseis sources and 200–250 receivers distributed along 5–6 urban transects; (2) processing data with a joint inversion approach that couples traveltime tomography with full-waveform inversion to enhance lateral resolution of fault planes; (3) integrating borehole information from 10–15 boreholes, where available, to constrain inversion and ground-truth interpreted fault locations; (4) evaluating the influence of urban noise on imaging and developing robust noise-reduction and data-selection criteria; and (5) delivering a probabilistic fault map with uncertainty quantified at 95% confidence intervals alongside velocity models. The study adopts a mixed-methods research design combining quantitative seismic imaging with qualitative interpretation of geological context. The population comprises urban subsurface environments with known active faulting tendencies in three metropolitan districts. A stratified random sampling of transects ensures representation of shallow sedimentary basins and crystalline bedrock interfaces. Data collection employs active-source multichannel seismic reflection and refraction methods, vibroseis-generated energy, and passive seismic recordings to augment near-surface characterization. Instrumentation includes a 48-channel seismic recording system per transect, 100 Hz to 1 kHz geophones, and downhole/nondestructive borehole sensors where feasible. Data analysis follows a two-stage framework (i) traveltime inversion using damped least squares with cross-hole constraints to produce preliminary velocity models, and (ii) full-waveform inversion (FWI) incorporating time-domain and frequency-domain formulations to recover detailed velocity contrasts across fault zones. The analysis also integrates Bayesian inference to quantify model uncertainty and to update fault plane probabilities as new data are incorporated. The theoretical basis draws on contemporary geophysical theory of velocity-contrast imaging and structural geology, applying concepts from the theory of seismic anisotropy and the earthquake source–medium interaction to interpret fault zone signatures. Expected findings include high-resolution velocity models revealing discontinuities and low-velocity conduits associated with fault zones, a probabilistic fault map highlighting zones of high fault likelihood, and improved capability to distinguish fault-related contrasts from aureole effects due to urban clutter. The study anticipates that joint inversion will substantially outperform single-method imaging in resolving fault geometries at depths of 20 to 500 meters, with uncertainty bounds that facilitate risk-informed decision-making for urban planning and construction. The contribution to knowledge lies in (a) advancing an integrative MST workflow tailored to dense urban environments, (b) delivering a validated methodology for reliable subsurface fault characterization that combines traveltime and waveform information with borehole constraints, and (c) providing a transferable framework for urban seismic hazard assessment applicable to similar metropolitan contexts worldwide. The main conclusion is that multichannel seismic tomography, when coupled with Bayesian uncertainty quantification and cross-validated with borehole data, yields robust fault delineation and velocity contrasts critical for seismic hazard mitigation in cities. Recommendations emphasize the adoption of the MST workflow by urban earthquake engineering programs, expanded borehole databases to tighten model constraints, and the incorporation of real-time data assimilation for rapid fault assessment during seismic events.