With the advancements of three-component seismic instruments, much valuable information about source distributions and subsurface structures can be utilized to improve passive surface-wave imaging with anthropogenic seismic noise in urban environments. Current passive surface-wave methods, however, are mainly concerned with the Rayleigh waves in the vertical (Z) component, often neglecting the useful dispersion information in the radial (R) and transverse (T) components, particularly in higher modes of surface waves. So, we introduced the common-midpoint two-station analysis (CMP-TS) to extract the multimode dispersion curves of Rayleigh and Love waves from three-component noise recordings using multicomponent ambient seismic noise cross-correlations (ZZ, RR, and TT components). Results from synthetic data sets from given models show that the CMP-TS method is able to retrieve higher-mode Rayleigh waves from the RR component and improve the multimode dispersion measurements of Rayleigh waves by the summation of the ZZ and RR spectrograms. Besides, this method can extract multimode dispersion curves of Love waves with high resolution from the TT component. We applied the CMP-TS method to process three-component field data and retrieved the dispersion curves of Rayleigh waves with the fundamental and the first higher modes, as well as Love waves with the fundamental, the first, and second higher modes. The S-wave velocity model is constructed by inverting the fundamental and higher mode data and validated through borehole S-wave velocity measurements.

In noninvasive near-surface investigations, with the emergence of massive seismic datasets, surface wave inversion using deep learning (DL) can efficiently attain the shear-wave velocity (Vs) model. Existed researches on DL inversion, however, cannot handle the inversion nonuniqueness effectively. The geological constraint is only reflected in their training dataset. The input of their neural networks only contains dispersion curves, and thus the density and compressional-wave velocity (Vp) need self-learning. To decrease the nonuniqueness, we propose a multi-information fusion neural network (MFNN) in which we add the Vp, density, and sensitivity as parts of the input. To verify the effectiveness of the MFNN, we used a synthetic test and field work which both contain data from six regions to conduct multi-region simultaneous inversions. We compared the inversion results of the MFNN with the results of a convolutional neural network and the ground truth. In the two experiments, although the calculated dispersion curves based on the inverted Vs model from all the methods match well with the observed ones, only the inverted Vs from the MFNN successfully reflect the actual Vs structure and locate the high-velocity layer and low-velocity layer accurately. The constraints on the Vp, density, and sensitivity thus effectively reduce the nonuniqueness of inversions.

Surface wave methods, commonly applied in diverse fields, encounter challenges in complex subsurface environments due to limitations inherent in traditional inversion techniques. Conventional 1-D inversion (1DI), with its reliance on fixed grids and deterministic linear approaches, often introduces biases, diminishing lateral resolution. Laterally constrained inversion (LCI) improves robustness by addressing lateral coherency but falls short in delineating arbitrary interfaces due to its dependence on fixed grid models. The advent of distributed acoustic sensing (DAS) technology offers extensive seismic data, yet its potential for high-resolution imaging remains underutilized. We introduce a multigrid spatially constrained dispersion curve inversion (MCI) method to overcome these challenges, aiming to harness high-resolution DAS surface wave imaging capabilities. This article details the MCI scheme, evaluates its efficacy through synthetic tests, and applies it to a DAS field study in Imperial Valley, CA, USA. Our findings demonstrate a refined, higher resolution S-wave velocity model, offering new insights into the region’s fault system and emphasizing the necessity of improved spatial resolution in large-scale geophysical studies.

Seismic interferometry (SI) is widely used to retrieve surface waves for linear arrays from anthropogenic noise in urban environments. It is difficult to accurately retrieve the surface-wave dispersion information extracted from cross correlation functions using a linear array because of the spatial and temporal variation of noise sources. It is also challenging to suppress the influence of offline noise sources by stacking noise cross correlation functions (NCFs). Data segment selection is an existing method for removing the segments with offline noise sources and utilizing the segments with in-line noise sources. We propose a two-step selection method to retain data segments that are useful for surface-wave retrieval based on polarization analysis. We retain data segments whose azimuths of noise sources are stable for a linear array in the first step. We then retain data segments whose noise sources are located within the stationary-phase zones (SPZs). The retained data segments are used to retrieve accurate high-frequency surface waves. A synthetic test is conducted to prove the superiority of the proposed method for improving data quality of Rayleigh and Love waves. Two field examples in Hangzhou also demonstrate the feasibility of the proposed method in improving the virtual shot gathers (VSGs) and dispersion images. Furthermore, the proposed method is applicable to noise sources with different azimuths and intensities. Our scheme is an option for improving data quality in urban environments where the distribution of noise sources is complex.

Train traffic has been realized as a powerful seismic source for imaging and monitoring the shallow subsurface. High-speed trains running on viaducts generate seismic waves through bridge piers. The sources at bridge piers with a moving train are correlated in space and time. In this study, we extract Rayleigh, Love, and virtual refraction P waves by applying seismic interferometry to high-speed-train-induced vibrations for near-surface characterization. We use crosscoherence instead of crosscorrelation to eliminate the spurious imprint of source self-correlations. Stationary-phase analysis suggests that it is easier to retrieve direct and refracted waves using a linear array along the railway with a moving train source. In the experiment, we deployed a three-component (3C) “T” shape nodal array consisting of one linear array along the railway and the other one approximately perpendicular to the railway. With totally 9 high-speed train events and by using stationary-phase segment selection, we extract strong multicomponent Rayleigh waves in range of 3-20 Hz. We clearly retrieve the virtual refraction P waves in range of 20-25 Hz using the in-line linear array and 25-35 Hz using the out-of-line array. Love waves in range of 3-20 Hz are also observed. The extracted Rayleigh, Love and virtual refraction waves are used to estimate S and P wave velocities of subsurface down to a depth of about 100 m. This study is the first 3C surface and body wave reconstruction and application for near-surface characterization from high-speed-train-induced vibrations.