Reliable earthquake ground-motion time histories are essential for seismic hazard analysis, structural design, and the risk assessment of distributed infrastructure systems. However, observed records of large earthquakes remain sparse, while physics-based simulations are computationally expensive and difficult to deploy at the scale required for uncertainty quantification. This thesis develops generative modeling frameworks for efficient, scenario-specific synthesis of earthquake ground motion, with an emphasis on neural operators, function-space learning, and scalable regional wavefield generation.
\r\n\r\nThe work begins with conditional generative adversarial neural operators for broadband three-component ground-motion synthesis at individual sites, conditioned on earthquake source, path, and site parameters. It then develops broader methodological foundations for functional generative modeling, including operator flow matching for stochastic-process learning and mesh-informed neural operators for domain- and discretization-agnostic generation on complex geometries. Building on these advances, the thesis introduces Ground-Motion Flow, a physics-inspired latent operator flow-matching framework for regional coherent ground-motion synthesis. By generating wavefields in a physics-aligned latent representation and reconstructing full spatiotemporal motions through neural operators, the framework produces spatially coherent earthquake wavefields conditioned on physical parameters with orders-of-magnitude acceleration relative to direct numerical simulation.
\r\n\r\nTogether, these studies establish a progression from single-site waveform synthesis to regional-scale, uncertainty-aware ground-motion wavefield generation. The results demonstrate the potential of functional generative models to complement physics-based simulation and empirical ground-motion modeling, opening a path toward fast, probabilistic, and scalable seismic hazard workflows for resilient infrastructure systems.
", "doi": "10.7907/3a27-v780", "publication_date": "2026", "thesis_type": "phd", "thesis_year": "2026" }, { "id": "thesis:14535", "collection": "thesis", "collection_id": "14535", "cite_using_url": "https://resolver.caltech.edu/CaltechTHESIS:03312022-021127047", "primary_object_url": { "basename": "Peyman_Ayoubi_2022_PhD_Thesis_Submitted_To_Library_v2.pdf", "content": "final", "filesize": 33559127, "license": "other", "mime_type": "application/pdf", "url": "/14535/1/Peyman_Ayoubi_2022_PhD_Thesis_Submitted_To_Library_v2.pdf", "version": "v4.0.0" }, "type": "thesis", "title": "Modeling and Parameterization of Basin Effects for Engineering Design Applications", "author": [ { "family_name": "Ayoubi", "given_name": "Peyman", "orcid": "0000-0001-6795-4923", "clpid": "Ayoubi-Peyman" } ], "thesis_advisor": [ { "family_name": "Asimaki", "given_name": "Domniki", "orcid": "0000-0002-3008-8088", "clpid": "Asimaki-D" } ], "thesis_committee": [ { "family_name": "Avouac", "given_name": "Jean-Philippe", "orcid": "0000-0002-3060-8442", "clpid": "Avouac-J-P" }, { "family_name": "Kohler", "given_name": "Monica D.", "orcid": "0000-0002-4703-190X", "clpid": "Kohler-M-D" }, { "family_name": "Ross", "given_name": "Zachary E.", "orcid": "0000-0002-6343-8400", "clpid": "Ross-Z-E" }, { "family_name": "Asimaki", "given_name": "Domniki", "orcid": "0000-0002-3008-8088", "clpid": "Asimaki-D" } ], "local_group": [ { "literal": "div_eng" } ], "abstract": "The term \"Basin effects\" refers to trapped and reverberating earthquake waves in soft sedimentary deposits overlying convex depressions of the basement bedrock, which significantly alter the frequency content, amplitude, and duration of seismic waves. This has played an important role in shaking duration and intensity in past earthquakes such as the Mw 8.0 1985 Micho\u00e1can, Mexico, Mw 6.9 1995 Kobe, Japan, and Mw 7.8 2015 Gorkha, Nepal earthquakes. While the standard practice is to perform a 1D analysis of a soil column, edge-effect and surface waves are among the key contributors to the surface ground motion within a basin. This thesis studies basin effects in a 2D medium to help better understand the phenomena, better parameterize them, and suggest a path to appropriately incorporate them in ground motion prediction equations and building design codes. After the introduction in Chapter 1, I present the results in three main parts as follows:
\r\n\r\nIn Chapter 2, we perform an extensive parametric study on the characteristics of surface ground motion associated with basin effects. We use an elastic idealized-shaped medium subjected to vertically propagating SV plane waves and examine the effects of basin geometry and material properties. We specifically study the effects of four dimensionless parameters, the width-to-depth (aspect) ratio, the rock-to-soil material contrast, a dimensionless frequency that quantifies the depth of the basin relative to the dominant incident wavelength, and a dimensionless distance that quantifies the distance of the basin edges relative to the dominant wavelength. Our results show that basin effects can be reasonably characterized using at least three independent parameters, each of which can significantly alter the resultant ground motion. To demonstrate the application of dimensional analysis applied here, we investigate the response of the Kathmandu Valley during the 2015 Mw 7.8 Gorkha Earthquake in Nepal using an idealized basin geometry and soil properties. Our results show that a simplified model can capture notable ground motion characteristics associated with basin effects.
\r\n\r\nChapter 3 uses the identified parameters from the previous chapter to estimate surface acceleration time-series given earthquake frequency content, basin geometry and material properties, and location inside a basin. This is of practical use when the amount of available data is limited or the fast estimation of time-series is desirable. For that, we train a neural network to estimate surface ground acceleration time-series across a basin. Three input parameters are needed for the estimation: basin-to-bedrock shear wave velocity ratio, aspect ratio of the basin, and dimensionless location. These parameters define an idealized-shaped basin and the location at which the time-series are to be computed. It will be shown that the model performs with high accuracy in comparison to the result of a full-fidelity Finite Element (FE) simulation (ground truth) and generalizes reasonably well for input parameters outside of the training set. Moreover, we will also use the model for the case of Kathmandu Valley, Nepal during the 2015 Mw 7.8 Gorkha earthquake and compare the results of NN versus recordings of the mainshock, similarly to Chapter 2.
\r\n\r\nOnce we have studied basin behavior in a homogeneous case in previous chapters, we focus on material representation inside a basin in Chapter 4. Here, we study basin effects for the cases where high-frequency response and realistic material representation are desirable. However, the lack of sufficient information about the material properties and stratigraphy of a basin prevents accurate simulation of the phenomena. To do that, we perform a stochastic analysis using the Monte Carlo technique, where a random field represents basin material. Similarly to the previous chapters, we use a 2D FE model with an idealized basin subjected to vertically propagating SV plane waves and investigate the spatial variation of surface ground motion (SGM) associated with basin effects by assuming different realizations of the correlated random field. We then study various correlation lengths, coefficients of variations, and autocorrelation functions to evaluate their contribution to SGM. We show that the coefficient of variation is the most influential parameter on SGM, followed by correlation lengths and type of autocorrelation function. Increasing the coefficient of variation not only affects the mean surface amplification, but also results in a dramatic change in the standard deviation. Correlation lengths and autocorrelation functions, on the other hand, are of less importance for the cases we examine in this study.
", "doi": "10.7907/4e61-q346", "publication_date": "2022", "thesis_type": "phd", "thesis_year": "2022" }, { "id": "thesis:14019", "collection": "thesis", "collection_id": "14019", "cite_using_url": "https://resolver.caltech.edu/CaltechTHESIS:12092020-002934412", "primary_object_url": { "basename": "kusanovic_danilo_2021_thesis.pdf", "content": "final", "filesize": 9676088, "license": "other", "mime_type": "application/pdf", "url": "/14019/1/kusanovic_danilo_2021_thesis.pdf", "version": "v5.0.0" }, "type": "thesis", "title": "Improving Reduced Order Models of Soil-Structure Interaction Using an Ensemble Kalman Inversion Finite Element Model Updating Framework", "author": [ { "family_name": "Kusanovic", "given_name": "Danilo Smiljan", "orcid": "0000-0002-0935-2577", "clpid": "Kusanovic-Danilo-Smiljan" } ], "thesis_advisor": [ { "family_name": "Asimaki", "given_name": "Domniki", "orcid": "0000-0002-3008-8088", "clpid": "Asimaki-D" } ], "thesis_committee": [ { "family_name": "Daraio", "given_name": "Chiara", "orcid": "0000-0001-5296-4440", "clpid": "Daraio-C" }, { "family_name": "Stuart", "given_name": "Andrew M.", "orcid": "0000-0001-9091-7266", "clpid": "Stuart-A-M" }, { "family_name": "Andrade", "given_name": "Jose E.", "clpid": "Andrade-J-E" }, { "family_name": "Asimaki", "given_name": "Domniki", "orcid": "0000-0002-3008-8088", "clpid": "Asimaki-D" } ], "local_group": [ { "literal": "div_eng" } ], "abstract": "In civil engineering, almost all structures are somehow in contact with soil - i.e., have foundations or support elements that either rest on or are embedded in soil. Thus, their seismic response is governed by the interaction between the structure, the non-structural components, the foundation, and the surrounding soil. Predicting such interaction becomes increasingly complex when uncertainties of soil and structural material, ground motion variability, and dissipation mechanisms are considered. The accuracy of numerical models to predict the linear or nonlinear responses of structures depends not only on how well the uncertainties in the material properties and input motion are estimated, but also on how well the various sources of energy dissipation and their interaction are modeled. Therefore, high-fidelity simulation of soil-structure interaction (SSI) problems require advanced models that can capture the nonlinear behavior of soils and structures, and parallel computing capabilities to optimize the cost associated with large scale problems. In spite of this fact, SSI in practice is widely accounted for using fixed-base building and reduced-order-models (ROM) which usually trade accurate solution for fast ones. Unfortunately, if SSI effects are neglected or poorly estimated, then critical response measures of a structure can be over- or under-estimated, which in turn can lead to unsafe or overly\r\nconservative designs.
\r\n\r\nMotivated by the previous challenge, in this thesis work we present a robust and efficient framework for finite element model (FEM) updating based on ensemble-Kalman inversion (EnKI). The EnKI-FEM updating framework is used to obtain suitable parameters to inform a ROM from data generated using high-fidelity FEM simulations. Since high-fidelity SSI simulations call for accurate and computationally efficient capabilities, as a part of this work, we developed Seismo-VLAB, a simple, fast, and extendable C++ finite element software to optimize large-scale simulations of dynamic and nonlinear SSI problems. The EnKI-FEM updating framework is thus integrated in Seismo-VLAB allowing to identify any parameter of the ROM without compromising accuracy. The so-generated ROM are finally employed to propose a new dimensionless frequency mapping to estimate the soil impedance for time domain analysis and to investigate soil-structure-interaction effects at a regional-scale. The presented methodology is general enough and it can be extended to more complex structural and/or geotechnical systems, allowing to construct highly-accurate ROM in a simple manner.
", "doi": "10.7907/m2qj-s182", "publication_date": "2021", "thesis_type": "phd", "thesis_year": "2021" }, { "id": "thesis:13587", "collection": "thesis", "collection_id": "13587", "cite_using_url": "https://resolver.caltech.edu/CaltechTHESIS:11212019-100323260", "type": "thesis", "title": "Application of Path-Independent Integrals to Soil-Structure Interaction", "author": [ { "family_name": "Garc\u00eda Su\u00e1rez", "given_name": "Antonio Joaqu\u00edn", "orcid": "0000-0001-8830-4348", "clpid": "Garc\u00eda-Su\u00e1rez-Antonio-Joaqu\u00edn" } ], "thesis_advisor": [ { "family_name": "Asimaki", "given_name": "Domniki", "orcid": "0000-0002-3008-8088", "clpid": "Asimaki-D" } ], "thesis_committee": [ { "family_name": "Ravichandran", "given_name": "Guruswami", "orcid": "0000-0002-2912-0001", "clpid": "Ravichandran-G" }, { "family_name": "Ortiz", "given_name": "Michael", "orcid": "0000-0001-5877-4824", "clpid": "Ortiz-M" }, { "family_name": "Meiron", "given_name": "Daniel I.", "orcid": "0000-0003-0397-3775", "clpid": "Meiron-D-I" }, { "family_name": "Asimaki", "given_name": "Domniki", "orcid": "0000-0002-3008-8088", "clpid": "Asimaki-D" } ], "local_group": [ { "literal": "GALCIT" }, { "literal": "div_eng" } ], "abstract": "Assessing seismic pressure increment on buried structures is a critical step in the design of infrastructure in earthquake-prone areas. Due to intrinsic complexities derived from the need to match the solution in the far-field to the localized solution around the structure, the near-field, researchers have aimed at finding simplified models focused on engineering variables as the seismic earth thrust. One such model is the so-called Younan-Veletsos model, which pivots on a stringent assumption on the stress tensor.
\r\n\r\nAt the same time, the might of the path-independent integrals of solid mechanics to deal with problems in Geotechnical Engineering at large, and Soil-Structure Interaction in particular, has remained unexplored, despite of a rich landscape of potential applications. The unbridled success of these path-independent integrals in Fracture Mechanics, a discipline which cannot be understood without them currently, may be mirrored in problems in Geotechnical Engineering, since the two fields, despite appearing very detached from each other at first glance, share deep traits: in both cases, the system under consideration can be conceptualized as a domain with simple, easy-to-assess regions (the areas where remote loading is applied and the far-field, respectively) and also with other complex, hard-to-understand regions (the crack tip, the near-field).
\r\n\r\nWe present the first derivation of the exact solution of the Younan-Veletsos problem, which is later analyzed to reveal phenomena not captured by previous approximate solutions. Then, we introduce a novel model which relies on the path-independent Rice\u2019s J-integral, a customary tool in Fracture Mechanics, which is applied here in the Soil-structure Interaction context for the first time. This novel model captures those features of the exact solution that were missed by prior approximations. The capabilities of the J-integral to, first, find an upper bound of the force induced by earthquakes over the walls of underground structures, under some conditions, and, second, to understand the soil-structure kinematic interaction phenomenon are also assessed.
\r\n\r\nAdditionally, the intermediate step of analyzing of the far-field yielded some results concerning Site Response Analysis which are also included in the text.
", "doi": "10.7907/MMWW-B046", "publication_date": "2020", "thesis_type": "phd", "thesis_year": "2020" }, { "id": "thesis:13762", "collection": "thesis", "collection_id": "13762", "cite_using_url": "https://resolver.caltech.edu/CaltechTHESIS:06012020-154218098", "primary_object_url": { "basename": "Kien_Nguyen_2020_PhD_Thesis.pdf", "content": "final", "filesize": 9921236, "license": "other", "mime_type": "application/pdf", "url": "/13762/1/Kien_Nguyen_2020_PhD_Thesis.pdf", "version": "v6.0.0" }, "type": "thesis", "title": "Reduced-Order Model for Dynamic Soil-Pipe Interaction Analysis", "author": [ { "family_name": "Nguyen", "given_name": "Kien Trung", "orcid": "0000-0001-5761-3156", "clpid": "Nguyen-Kien-Trung" } ], "thesis_advisor": [ { "family_name": "Asimaki", "given_name": "Domniki", "orcid": "0000-0002-3008-8088", "clpid": "Asimaki-D" } ], "thesis_committee": [ { "family_name": "Andrade", "given_name": "Jose E.", "clpid": "Andrade-J-E" }, { "family_name": "Daraio", "given_name": "Chiara", "orcid": "0000-0001-5296-4440", "clpid": "Daraio-C" }, { "family_name": "Hall", "given_name": "John F.", "orcid": "0000-0002-7863-5060", "clpid": "Hall-J-F" }, { "family_name": "Asimaki", "given_name": "Domniki", "orcid": "0000-0002-3008-8088", "clpid": "Asimaki-D" } ], "local_group": [ { "literal": "div_eng" } ], "abstract": "Pipelines are very vulnerable infrastructure components to geohazard-induced ground deformation and failure. How soil transmits loads on pipelines and vice versa, known as soil-pipe interaction (SPI), thus is very important for the assessment and design of resilient pipeline systems.
\r\n\r\nIn the first part, this work proposes a simplified macroelement designed to capture SPI in cohesionless soils subjected to arbitrary loading normal to the pipeline axis. We present the development of a uniaxial hysteresis model that can capture the smooth nonlinear reaction force-relative displacement curves (FDCs) of SPI problems. Using the unscented Kalman filter, we derived the model parameter κ that controls the smoothness of the transition zone from linear to plastic using published experimental data. We extended this uniaxial model to biaxial loading effects and showed that the macroelement can capture effects such as pinching and shear-dilation coupling. The model input parameters were calibrated using finite element (FE) analyses validated by experiments. The FDCs of the biaxial model were verified by comparison with FE and smoothed-particle hydrodynamic (SPH) simulations for different loading patterns: cyclic uniaxial, 0-shaped, 8-shaped, and transient loading. Accounting for smooth nonlinearity, hysteresis, pinching, and coupling effects, the proposed biaxial macroelement shows good agreement with FE and SPH analyses, while maintaining the computational efficiency and simplicity of beam-on-nonlinear-Winkler foundation models, as well as a small number of input parameters.
\r\n\r\nNext, this work presents analytical solutions for computing frequency-domain axial and in-plane soil impedance functions (SIFs) for an infinitely long rigid circular structure buried horizontally in homogeneous elastic half-space. Using Hankel— and Bessel—Fourier series expansion, we solved a mixed-boundary-value problem considering a harmonic displacement at the structure boundary and traction-free boundary condition at the half-space free surface. We then verified our analytical solutions using results obtained from FE simulations. The SIFs of a buried structure in a homogeneous elastic half-space calculated by these two approaches are in perfect agreement with each other. In addition, we used analytical solutions and FE simulations to comprehensively investigate factors that affect the SIFs in homogeneous and two-layered half-spaces, respectively. The parametric study shows that SIFs of buried structures in elastic half-space primarily depend on frequency of excitation, shear modulus and Poisson's ratio of the half-space, burial depth and radius of the structure. In a two-layered soil domain, SIFs depend also on material contrast and the distance from the structure location to the interface between soil layers.
\r\n\r\nLastly, it demonstrates how the SIFs obtained previously can be incorporated into a reduced-order model to analyze SPI problems, specifically a straight pipe subjected to Rayleigh surface wave propagating through homogeneous and heterogeneous elastic half-spaces. Calculated displacement time histories at the control points are shown to agree well with those computed by direct two-dimensional FE analyses.
", "doi": "10.7907/mekk-dc25", "publication_date": "2020", "thesis_type": "phd", "thesis_year": "2020" }, { "id": "thesis:11571", "collection": "thesis", "collection_id": "11571", "cite_using_url": "https://resolver.caltech.edu/CaltechTHESIS:05302019-150220368", "type": "thesis", "title": "Improving Site Response Analysis for Earthquake Ground Motion Modeling", "author": [ { "family_name": "Shi", "given_name": "Jian", "orcid": "0000-0002-1969-7579", "clpid": "Shi-Jian" } ], "thesis_advisor": [ { "family_name": "Asimaki", "given_name": "Domniki", "orcid": "0000-0002-3008-8088", "clpid": "Asimaki-D" } ], "thesis_committee": [ { "family_name": "Lapusta", "given_name": "Nadia", "orcid": "0000-0001-6558-0323", "clpid": "Lapusta-N" }, { "family_name": "Andrade", "given_name": "Jose E.", "clpid": "Andrade-J-E" }, { "family_name": "Graves", "given_name": "Robert W.", "clpid": "Graves-R-W" }, { "family_name": "Asimaki", "given_name": "Domniki", "orcid": "0000-0002-3008-8088", "clpid": "Asimaki-D" } ], "local_group": [ { "literal": "div_eng" } ], "abstract": "The modeling of earthquake-induced ground motions plays an important role in the quantification of seismic hazards, which contributes to the ultimate goal of saving lives and reducing economic loss. Site response is a natural phenomenon in which soils in the earth\u2019s shallow crust alter the amplitude, frequency content, and duration of earthquake-induced ground motions. Therefore, improvements in the research of site response directly contribute to ground motion modeling, and eventually to seismic hazard quantification.
\r\n\r\nThis thesis presents two models that advance the current research in site response.
\r\n\r\nThe first model provides a tool to predict near-surface shear-wave velocity profiles from Vs30 (a proxy that represents the general stiffness of a site). This model bridges the gap between the lack of information about near-surface soil properties and the need to model site response on a regional scale (city, county, or above).
\r\n\r\nThe second model is a stress-strain model for describing 1D shearing behaviors of soils. It is capable of capturing both the small-strain and the large-strain behaviors, which makes it suitable for modeling very strong ground motions. More importantly, this model enables seismologists to construct stress-strain curves from only shear-wave velocity information, again improving our ability to model site response on a regional scale. Our validation study shows that this model outperforms the prevalent stress-strain model (namely, the MKZ model) by a considerable margin.
\r\n\r\nLastly, we demonstrate how the two models above can improve earthquake ground motion modeling: we develop an improved version of site factors for the Western United States. These site factors are provided as Fourier spectral ratios, and phase factors are provided for the first time, which enables the time delay of earthquake waves to be modeled. They can be used for incorporating site response in earthquake ground motion simulations, as well as for improving seismic hazard maps for the Western United States.
", "doi": "10.7907/X5NZ-DQ21", "publication_date": "2019", "thesis_type": "phd", "thesis_year": "2019" } ]