Observing Seismic Variations by Earth and Lab Fluids and Fractures

Abstract

Understanding the mechanical dynamics within the Earth's crust is crucial for both environmental and energy sustainability. These dynamics are often tied to mechanical deformations triggered by variations in fluid pressure and stress levels. The primary method of investigation involves seismic monitoring to detect changes in material properties and identify fractures or faults, making the study of the relationship between seismic properties, fluids, and fractures essential. My research utilizes a combination of laboratory experiments, numerical simulations, and field observations to enhance our understanding of the mechanisms behind seismic variations, including changes in wavespeeds, attenuation, and the behavior of laboratory and natural earthquakes. My thesis is divided into two interrelated parts: (1) the influence of fluids and/or deformation on seismoacoustic wavespeed and attenuation, and (2) the effect of fluid properties and pressure on fracture behavior and seismoacoustic signals. In the first part, I develop wavelet-domain techniques for analyzing frequency-dependent velocity changes through coda wave interferometry and delve into the depth sensitivity of these changes using wavefield simulations. I also employ controlled acoustic monitoring and 3D-printed media to examine how various physical conditions, such as consolidation, saturation, and strain deformation, affect frequency-dependent velocity changes and attenuation. These works aim to use these velocity change and attenuation spectra to either determine the depth of perturbations or to understand the underlying physical mechanisms. Moreover, I explore the potential of probing deep volcanic activity through inter-source interferometry, using repeating earthquakes from Mount St. Helens. This part proposes novel methods and insights to tackle the challenges in imaging and understanding of seismic property changes, with applications ranging from subsurface exploration to volcano monitoring. In the second part, I focus on hydrofracturing dynamics, employing advanced imaging and acoustics to study fluid-induced fracturing patterns and their correlation with seismic activities, using the Cascadia region's tectonic tremor swarms as a case study. I also investigate hydromechanics within artificial fault-valve media, specifically fluid migration and its effects on fault instability and permeability. I further explores high-performance seismic processing, developing a new ensemble learning framework to enhance the generalizability of seismic phase pickers. This part aims to advance understanding of fluid-induced deformations and their implications for seismic hazard assessment and resource optimization.