Reconstruction of sparsely sampled seismic data is critical for maintaining the quality of seismic images when significant numbers of shots and receivers are missing. We present a reconstruction method in the shot-receiver-time (SRT) domain based on a residual U-Net machine learning architecture, for seismic data acquired in a sparse 2-D acquisition and name it SRT2D-ResU-Net. The SRT domain retains a high level of seismic signal connectivity, which is likely the main data feature that the reconstructing algorithms rely on. We develop an ``in situ training and prediction'' workflow by dividing the acquisition area into two nonoverlapping subareas: a training subarea for establishing the network model using regularly sampled data and a testing subarea for reconstructing the sparsely sampled data using the trained model. To establish a reference base for analyzing the changes in data features over the study area, and quantifying the reconstructed seismic data, we devise a baseline reference using a tiny portion of the field data. The baselines are properly spaced and excluded from the training and reconstruction processes. The results on a field marine data set show that the SRT2D-ResU-Net can effectively learn the features of seismic data in the training process, and the average correlation between the reconstructed missing traces and the true answers is over 85%.
We conducted a comparative analysis of the new deformable layer tomography (DLT) with the traditional fixed-in-space cell tomography along three 2-D crustal profiles of P-wave velocities in southern California using first arrivals from local earthquakes and two Los Angeles Regional Seismic Experiment (LARSE) seismic lines. The DLT method inverts for the depth-varying geometry of velocity interfaces, which include major velocity discontinuities near the top and bottom of the crust and boundaries between velocity layers in the middle crust. In regions characterized by large lateral velocity changes, the DLT method may improve the depth resolution of major velocity interfaces compared to traditional cell tomography using fixed-in-space cells or grids. We also tested the use of crustal thickness estimates from a receiver function study for constraining the depth range of the Moho geometry. Tests indicate that, in areas of sufficient ray coverage, the depth resolution of the DLT for major velocity interfaces is usually less than 1 km in the upper crust and less than 3 km in the lower crust. Our 2-D DLT models fit the first arrival data significantly better than three published 3-D cell tomography models along the profiles. The configurations of basin edges and undulating velocity discontinuities are well delineated in the DLT models and indicative of the crustal architecture of the region.