Processing of Two-Dimensional Surface Wave Seismic Data Using a Python-Based Library

Surface waves are a geophysical method that uses shear-wave velocity to model heterogeneity in the near-surface layer. Today, many geological and engineering companies use the surface-wave method because it doesn’t require high computational power and processes data quickly. The high demand for seismic surface waves to detect heterogeneity creates a need for robust processing software to analyze surface-wave models. Surface waves are dispersive, meaning their velocity depends on frequency. The velocity of a surface wave can be calculated using phase velocity, which is the velocity at which the wave's phase changes. Velocity can be calculated using different methods. Generally, these methods focus on finding smooth lateral variation or sharp velocity contrasts. The method used here is tomography-derived linear inversion, which focuses on finding 2 dimensional model of phase velocity. The algorithm estimates the wavenumber from phase differences between traces within a window. Before calculation, the data must be filtered in the FK-domain dispersion curve to remove high-mode energy. The dispersion curve is calculated using the Fourier transform to get high spectral density. The higher mode is a constructive-interference energy wave with high velocity. The fundamental mode is a constructive-interference wave with low velocity that needs to be kept. The algorithm selects only the fundamental mode and extracts the high-spectral-energy component within it. The algorithm uses the fundamental-mode high-spectral-energy area to calculate phase differences. The phase difference is used to find the wavenumber, which is then used to compute shear-wave velocity. We use real surface-wave data from seismic acquisition to test whether the tomography-derived linear inversion works for finding the heterogeneity. We use the Dinver package to invert for shear-wave velocity with depth and check the algorithm's phase-velocity model results. The Dinver package uses a neighborhood algorithm to find the shear-wave velocity. The Dinver package requires input of the area's physical parameters and the phase-velocity results from the dispersion-curve calculation. The Dinver results show that shear-wave velocity increases linearly with depth from 0 to 30 meters. The shear-wave velocity from the model is between 200 m/s and 1000 m/s. The shear-wave model shows that at distances between 380 meters and 680 meters, the highest value is around 950 m/s. This area needs to be checked because it slightly higher than the lateral layer at that depth. The refraction seismic method is also used to compare the results with the phase velocity model. The results are shown in the separation between lines 1, 2, and 3; the separation is based on the area of interest from the phase velocity model and shear wave model. Results from line 3, which shows a shear-wave anomaly in the Dinver Package, show no lateral change in compressional velocity values. The high shear velocity value occurred because of an artifact in the phase velocity from the superposition of the dispersion curve spectrum. The shear-wave inversion model shows low misfit, meaning the shear-wave model matches the observed phase velocity data. The refraction results mainly show that vertical heterogeneity is the same for both phase velocity and vertical velocity