Physics-based monitoring of subsurface processes using seismic ambient noise

Abstract

Understanding subsurface processes is essential for resource management, hazard mitigation, and environmental protection. Passive seismic methods using ambient seismic noise have emerged as a promising tool for subsurface monitoring and imaging. These methods offer non-invasive, large-scale, and continuous measurements, providing valuable insights into geological structures, geophysical properties, and subsurface processes. However, these methods are often based on empirical relationships between seismic velocity variations and dynamic subsurface properties, and the physical mechanisms underlying these relationships have not yet been fully exploited. This doctoral research contributes to advancing the understanding of seismic velocity changes and leverages this knowledge to enhance the monitoring of subsurface dynamics. It addresses the physical mechanisms behind seismic velocity variations, offering a fundamental understanding that goes beyond empirical relationships. Physics-based approaches allow for more accurate and comprehensive interpretations, predictive modeling, and technological advancements in subsurface monitoring and imaging. The research begins by constructing a physics-based model that connects seismic velocity changes to variations in pore pressure and vertical stress, related to fluctuations in groundwater level. The model utilizes established relationships between seismic velocities and induced stress, coupled with wave propagation theory, basic hydrology and geomechanics, to establish direct links between seismic velocity variations and specific dynamic subsurface properties, namely fluctuations in pore pressure and vertical compressional stress. For pore pressure fluctuations, the model’s validity is confirmed using passive image interferometry on seismic ambient noise data from Groningen, the Netherlands, demonstrating its ability to explain surface-wave phase-velocity variations caused by pore pressure fluctuations. The direct link between seismic velocity variations and pore pressure is subsequently exploited for four-dimensional space-time pore pressure monitoring using surface-wave phase-velocity changes. As such, I introduce pore pressure sensitivity kernels as a direct connection between depth-dependent pore pressure variations and frequency-dependent changes in surface-wave phase velocities, showcasing their utility in inferring pore pressure variations in the Groningen subsurface. The inferred pore pressure models align closely with independent pressure head measurements, highlighting the potential for quantitative pore pressure inferences. The sensitivity of surface-wave phase velocities to pore pressure changes was found to decrease much faster with depth than the sensitivity to changes in elastic parameters, limiting the monitoring approach in the Groningen subsurface to the shallowest 200 m for natural pore pressure variations. In contrast, pore pressure variations caused by human activities are significantly larger. Therefore, this research continues with a feasibility assessment of utilizing surface-wave phase-velocity changes to monitor anthropogenic pore pressure developments in deeper reservoirs, particularly in the Harlingen and Groningen gas reservoirs. It expands pore pressure sensitivity kernels to deeper depths and models surface-wave phase-velocity changes in response to hypothetical production scenarios. While monitoring the shallow (∼ 1 km) Harlingen reservoir appears feasible, monitoring the deeper (∼ 3 km) Groningen reservoir presents substantial challenges, as the existing measurement uncertainties for velocity changes must be significantly reduced. Finally, the research delves into the physics of temperature-induced seismic velocity changes in the shallow unconsolidated subsurface. It reconciles field and laboratory experiments by considering intrinsic temperature dependencies, thermally induced stress, and thermally induced strain. The study predicts seasonal temperature-induced seismic velocity variations and their implications for site amplification. The site-specific material properties determine whether site amplification is more pronounced during summer or winter. Overall, this research enhances our understanding of seismic velocity changes and benefits their utility in subsurface monitoring, thereby contributing to resource management, hazard assessment, and a sustainable environment. The physics-based approach facilitates amore comprehensive understanding of subsurface dynamics and supports the development of innovative monitoring and imaging techniques.