We analyze seismograms recorded by four arrays (B1-B4) with 100-m station spacing and apertures of 4-8 km that cross the surface rupture of the 2019 Mw7.1 Ridgecrest earthquake. The arrays extend from B1 in the northwest to B4 in the southeast of the surface rupture. Delay times between P-wave arrivals associated with ∼1200 local earthquakes and four teleseismic events are used to estimate local velocity variations beneath the arrays. Both teleseismic and local P waves travel faster on the northeast than the southwest side of the fault for ~4.6% and ~7.5% beneath arrays B1 and B4, but the velocity contrast is less significant at arrays B2 and B3. We identify several 1- to 2-km-wide low-velocity zones with more intensely damaged inner cores beneath each array. The damage zone at array B4 generates fault-zone head, reflected, and trapped waves. An automated detector, based on peak ground velocities and durations of high-amplitude waves, identifies candidate fault-zone trapped waves (FZTWs) in a localized zone for ~600 earthquakes. Synthetic waveform modeling of averaged FZTWs, generated by ~30 events with high-quality signals, indicate that the trapping structure at array B4 has a width of ∼300 m, depth of 3-5 km, S-wave velocity reduction of ∼20% with respect to the surrounding rock, Q-value of ∼30, and S-wave velocity contrast of ~4% across the fault (faster on the northeast side). The results show complex fault-zone internal structures that vary along fault strike, in agreement the surface geology (alternating playa and igneous rocks).

Temporal changes of S-wave velocities at shallow depth on Mars are derived using seismic data from the InSight mission. Autocorrelation functions are computed for three-component seismic recordings to retrieve zero-offset reflection seismograms. Observed S-wave reflection phase with two-way travel time of ~1.2 s and its multiples indicate an interface at ~200 m depth. Daily relative travel time changes (dt/t) with ~5% variations are correlated well with the surface temperature. A top ~1m-thick regolith layer produces a delay of about one Martian day between the dt/t and surface temperature. Assuming the travel time changes are produced primarily in the top ~18 m sand layer, the daily velocity variations in that layer are ~40%. The dominant mechanisms driving the changes are thermoelastic strain in the shallow structure generating the time delays and possible material failures in the regolith layer.

Machine learning algorithm is applied to shear wave velocity (Vs) inversion in surface wave tomography, where a set of 1-D Vs profiles and the corresponding synthetic dispersion curves are used in network training. Previous studies showed that performances of a trained network depend on the input training dataset with limited diversity and therefore lack generalizability. Here, we present an improved semi-supervised algorithm-based network that takes both model-generated and observed surface wave dispersion data in the training process. The algorithm is termed Wasserstein cycle-consistent generative adversarial networks (Wcycle-GAN). Different from conventional supervised approaches, the GAN architecture extracts feature from the observed surface wave dispersion data that can compensate the limited diversity of the training dataset generated synthetically. The cycle-consistency enforces the reconstruction ability of input data from predicted model using a separate data generating network, while Wasserstein metric provides improved training stability and enhanced spatial smoothness of the output Vs model. We demonstrate improvements by applying the Wcycle-GAN method to 4076 pairs of fundamental mode Rayleigh wave phase and group velocity dispersion curves obtained in Southern California. The final 3-D Vs model from the best trained network shows large-scale features that are consistent with the surface geology. Our Vs model has smaller data misfits, yields better spatial smoothing, and provides sharper images of structures near faults in the top 15 km, suggesting the proposed Wcycle-GAN algorithm has stronger training stability and generalization abilities compared to conventional machine learning methods.

Current machine learning based shear wave velocity (Vs) inversion using surface wave dispersion measurements utilizes synthetic dispersion curves calculated from existing 3-D velocity models as training datasets. It is shown in the previous studies that the performances of the resulting networks are dependent on the diversity of the training data. We present an improved semi-supervised deep learning algorithm-based method that incorporates both observed and synthetic surface wave dispersion curves in the network training process. The algorithm is termed Wasserstein cycle-consistent generative adversarial networks (Wcycle-GAN), which combines the architecture of cycle-consistent GAN with Wasserstein loss metrics in optimization. Different from conventional supervised deep learning approaches, the GAN architecture also extracts structural information from the observed surface wave dispersion data in the training process that may improve generalization of the resulting network. The cycle-consistent loss addresses soft constraints on the trained neural networks to be reversible and thus reduces the variance of the trained networks. The Wasserstein metric provides weaker topology for convergence and improves spatial continuity of the predicted shear velocity (Vs) models. We demonstrate these improvements by applying the Wcycle-GAN to 4066 fundamental mode Rayleigh wave phase and group dispersion curves obtained in Southern California (SC). In general, the 3-D Vs model predicted by the best training Wcycle-GAN is consistent with previous surface wave tomography studies of SC in the overlapping area, but with smaller data misfit, yields better spatial smoothing, and provides improved images of structures near faults and in the top 5 km. Our results indicate that the proposed Wcycle-GAN algorithm has strong training stability and generalization abilities.