Search for: All records

Abstract Numerical simulations of seismic wave propagation are crucial for investigating velocity structures and improving seismic hazard assessment. However, standard methods such as finite difference or finite element are computationally expensive. Recent studies have shown that a new class of machine learning models, called neural operators, can solve the elastodynamic wave equation orders of magnitude faster than conventional methods. Full waveform inversion is a prime beneficiary of the accelerated simulations. Neural operators, as end‐to‐end differentiable operators, combined with automatic differentiation, provide an alternative approach to the adjoint‐state method. State‐of‐the‐art optimization techniques built into PyTorch provide neural operators with greater flexibility to improve the optimization dynamics of full waveform inversion, thereby mitigating cycle‐skipping problems. In this study, we demonstrate the first application of neural operators for full waveform inversion on a real seismic data set, which consists of several nodal transects collected across the San Gabriel, Chino, and San Bernardino basins in the Los Angeles metropolitan area. 

Magma supply rates from the mantle to Hawaiian volcanoes serve as an important control on eruptive behavior at the surface. The Pa ̄hala Sill Complex, a collection of magma-bearing, seismogenic structures at 40 km depth beneath Hawai‘i, presents an opportunity to elucidate interactions between stress and magma transport processes in the mantle. We invert for full moment tensors of sill earthquakes and identify predominantly shear mechanisms with persistent tensile faulting components. Slip occurs in-plane with the sill structures. Pressure axes are radially oriented about a point near Mauna Loa, consistent with a stress field generated by a flexural load. Together, these observations suggest that magma flux through the sill structures generates seismicity by increasing pore pressure and promoting slip. Our results suggest that stress changes in mantle structures may enable fluctuations in magma supply rates to the surface over short timescales. 

Abstract Gas extraction from the Groningen gas reservoir, located in the northeastern Netherlands, has led to a drop in pressure and drove compaction and induced seismicity. Stress-based models have shown success in forecasting induced seismicity in this particular context and elsewhere, but they generally assume that earthquake clustering is negligible. To assess earthquake clustering at Groningen, we generate an enhanced seismicity catalog using a deep-learning-based workflow. We identify and locate 1369 events between 2015 and 2022, including 660 newly detected events not previously identified by the standard catalog from the Royal Netherlands Meteorological Institute. Using the nearest-neighbor distance approach, we find that 72% of events are background independent events, whereas the remaining 28% belong to clusters. The 55% of the clustered events are swarm-like, whereas the rest are aftershock-like. Among the swarms include five newly identified sequences propagating at high velocities between 3 and 50 km/day along directions that do not follow mapped faults or existing structures and frequently exhibit a sharp turn in the middle of the sequence. The swarms occurred around the time of the maximum compaction rate between November 2016 and May 2017 in the Zechstein layer, above the anhydrite caprock, and well-above the directly induced earthquakes that occur within the reservoir and caprock. We suggest that these swarms are related to the aseismic deformation within the salt formation rather than fluids. This study suggests that the propagating swarms do not always signify fluid migration. 

Abstract Distributed Acoustic Sensing (DAS) is an emerging technology for earthquake monitoring and subsurface imaging. However, its distinct characteristics, such as unknown ground coupling and high noise level, pose challenges to signal processing. Existing machine learning models optimized for conventional seismic data struggle with DAS data due to its ultra-dense spatial sampling and limited manual labels. We introduce a semi-supervised learning approach to address the phase-picking task of DAS data. We use the pre-trained PhaseNet model to generate noisy labels of P/S arrivals in DAS data and apply the Gaussian mixture model phase association (GaMMA) method to refine these noisy labels and build training datasets. We develop PhaseNet-DAS, a deep learning model designed to process 2D spatio-temporal DAS data to achieve accurate phase picking and efficient earthquake detection. Our study demonstrates a method to develop deep learning models for DAS data, unlocking the potential of integrating DAS in enhancing earthquake monitoring. 

SUMMARY The spatio-temporal properties of seismicity give us incisive insight into the stress state evolution and fault structures of the crust. Empirical models based on self-exciting point processes continue to provide an important tool for analysing seismicity, given the epistemic uncertainty associated with physical models. In particular, the epidemic-type aftershock sequence (ETAS) model acts as a reference model for studying seismicity catalogues. The traditional ETAS model uses simple parametric definitions for the background rate of triggering-independent seismicity. This reduces the effectiveness of the basic ETAS model in modelling the temporally complex seismicity patterns seen in seismic swarms that are dominated by aseismic tectonic processes such as fluid injection rather than aftershock triggering. In order to robustly capture time-varying seismicity rates, we introduce a deep Gaussian process (GP) formulation for the background rate as an extension to ETAS. GPs are a robust non-parametric model for function spaces with covariance structure. By conditioning the length-scale structure of a GP with another GP, we have a deep-GP: a probabilistic, hierarchical model that automatically tunes its structure to match data constraints. We show how the deep-GP-ETAS model can be efficiently sampled by making use of a Metropolis-within-Gibbs scheme, taking advantage of the branching process formulation of ETAS and a stochastic partial differential equation (SPDE) approximation for Matérn GPs. We illustrate our method using synthetic examples, and show that the deep-GP-ETAS model successfully captures multiscale temporal behaviour in the background forcing rate of seismicity. We then apply the results to two real-data catalogues: the Ridgecrest, CA 2019 July 5 Mw 7.1 event catalogue, showing that deep-GP-ETAS can successfully characterize a classical aftershock sequence; and the 2016–2019 Cahuilla, CA earthquake swarm, which shows two distinct phases of aseismic forcing concordant with a fluid injection-driven initial sequence, arrest of the fluid along a physical barrier and release following the largest Mw 4.4 event of the sequence. 

Interacting magmatic sills at about 40-kilometer depth under the island of Hawai‘i feed into multiple different volcanoes. 

Abstract Fault geometry affects the initiation, propagation, and cessation of earthquake rupture, as well as, potentially, the statistical behavior of earthquake sequences. We analyze 18,250 (−0.27 < M < 4.4) earthquakes of the 2016–2019 Cahuilla, California, swarm and, for the first time, use these high-resolution earthquake locations to map, in detail, the roughness across an active fault surface at depth. We find that the strike-slip fault is 50% rougher in the slip-perpendicular direction than parallel to slip. 3D mapping of fault roughness at seismogenic depths suggests that roughness varies by a factor of 8 for length scales of 1 km. We observe that the largest earthquake (M 4.4) occurred where there is significant fault complexity and the highest measured roughness. We also find that b-values are weakly positively correlated with fault roughness. Following the largest earthquake, we observe a distinct population of earthquakes with comparatively low b-values occurring in an area of high roughness within the rupture area of the M 4.4 earthquake. Finally, we measure roughness at multiple scales and find that the fault is self-affine with a Hurst exponent of 0.52, consistent with a Brownian surface. 

SUMMARY The spatial geometry of seismicity encodes information about loading and failure processes, as well as properties of the underlying fault structure. Traditional approaches to characterizing geometrical attributes of seismicity rely on assumed locations and geometry of fault surfaces, particularly at depth, where resolution is overall quite poor. In this study, we develop an alternative approach to quantifying geometrical properties of seismicity using techniques from anisotropic point process theory. Our approach does not require prior knowledge about the underlying fault properties. We characterize the geometrical attributes of 32 distinct seismicity regions in California and introduce a simple four class classification scheme that covers the range of geometrical properties observed. Most of the regions classified as having localized seismicity are within northern California, while nearly all of the regions classified as having distributed seismicity are within southern California. In addition, we find that roughly 1 out of 4 regions exhibit orthogonal seismicity structures. The results of this study provide a foundation for future analyses of geometrical properties of seismicity and new observables to compare with numerical modelling studies.