Seismic tomography using body or surface wave data is a powerful tool to explore the structure of Earth’s interior structure. In recent decades, joint inversion of seismic body and surface wave data has been widely employed to investigate seismic velocities of the Earth’s lithosphere and asthenosphere. Benefited from the complementary sensitivities of different datasets, seismic velocities determined by joint inversion generally exhibit higher resolution and accuracy. Regular mesh (cell or grid) is commonly used in seismic tomography. As data distribution is uneven in most cases, regularization techniques are implemented in regular mesh seismic tomography method to stabilize ill-posed problems. Despite the selection of appropriate regularization parameters, it is also challenging to achieve multiscale resolution in regular mesh joint inversion method. In this study, we developed a joint inversion method using adaptive irregular mesh according to the real data distribution based on Poisson-Voronoi cells. Synthetic tests show that the newly developed method can better resolve multi-scale structures without regularizations. We applied this method to a dataset with seismic arrays in different scales. The newly determined multiscale velocity model reveals distinct features particularly in areas with dense data distribution.
Abstract Ill‐posed seismic inverse problems are often solved using Tikhonov‐type regularization, that is, incorporation of damping and smoothing to obtain stable results. This typically results in overly smooth models, poor amplitude resolution, and a difficult choice between plausible models. Recognizing that the average of parameters can be better constrained than individual parameters, we propose a seismic tomography method that stabilizes the inverse problem by projecting the original high‐dimension model space onto random low‐dimension subspaces and then infers the high‐dimensional solution from combinations of such subspaces. The subspaces are formed by functions constant in Poisson Voronoi cells, which can be viewed as the mean of parameters near a certain location. The low‐dimensional problems are better constrained, and image reconstruction of the subspaces does not require explicit regularization. Moreover, the low‐dimension subspaces can be recovered by subsets of the whole dataset, which increases efficiency and offers opportunities to mitigate uneven sampling of the model space. The final (high‐dimension) model is then obtained from the low‐dimension images in different subspaces either by solving another normal equation or simply by averaging the low‐dimension images. Importantly, model uncertainty can be obtained directly from images in different subspaces. Synthetic tests show that our method outperforms conventional methods both in terms of geometry and amplitude recovery. The application to southern California plate boundary region also validates the robustness of our method by imaging geologically consistent features as well as strong along‐strike variations of San Jacinto fault that are not clearly seen using conventional methods.
Abstract Cross-correlating continuous seismic data is a commonly employed technique to extract coherent signals to image and monitor the subsurface. However, due largely to site effects and poorly characterized noise sources in oceanic environments, its application to ocean-bottom seismometer (OBS) recordings often requires additional processing. In this contribution, we propose a method to improve the quality of the retrieved surface waves from OBS data and characterize the noise sources. We first cluster the pre-stack noise cross-correlation functions (NCFs) based on a sequencing algorithm, followed by selectively stacking those consisting of coherent and stable signals that are consistent with predicted surface-wave arrival times. Synthetic tests show that the sequenced NCFs can be used to recover the spatial and temporal distribution of noise sources. Applying the method to an OBS array offshore California increases the signal-to-noise ratios of the obtained Rayleigh waves. In addition, we find that the annual temporal distribution of selected NCFs with frequencies ranging from 0.04 to 0.1 Hz is nearly homogeneous during the recording period. In contrast, many NCFs excluded for stacking are temporally clustered. This method has the potential to be applied to other OBS recordings or possibly onland deployments, thus helping to obtain high-quality surface waves and to analyze temporal noise source characteristics.
Abstract Constraining the mechanism of earthquakes in subduction zones requires adequate estimates of source location and near‐source elastic properties. In this study, we propose a wave autocorrelation‐based method to extract depth phase energy from teleseismic earthquakes with moment magnitude down to 4.0. We apply the method to improve location estimates of intermediate‐depth earthquakes in the Japan and northern Chile subduction zones, which represent so‐called cold and warm slabs, respectively, and which are both marked by double seismic zones. A positive correlation of slab age and double‐seismic‐zone width validates a thermally controlled model of slab morphology. The negative to normal differential times (and, thus, low or normal Vp/Vs) of the deep parts of the double seismic zones suggest that the intermediate‐depth earthquakes considered here are not due to dehydration.
'/%3e%3cuse xlink:href='%23a' transform='rotate(23.036 .093 25.536)'/%3e%3cuse xlink:href='%23a' transform='rotate(45.87 1.273 16.18)'/%3e%3cuse xlink:href='%23a' transform='rotate(69.945 .996 12.078)'/%3e%3cuse xlink:href='%23a' transform='rotate(20.66 -19.689 31.932)'/%3e%3c/svg%3e)
'/%3e%3cuse xlink:href='%23a' transform='rotate(60)'/%3e%3ccircle r='17' fill='%23000095'/%3e%3ccircle r='15' fill='%23fff'/%3e%3c/svg%3e)