Novel Surface Wave Imaging Methods

  • Zhaolun Liu

Student thesis: Doctoral Thesis


I develop four novel surface-wave inversion and migration methods for reconstructing the low- and high-wavenumber components of the near-surface S-wave velocity models. 1. 3D Wave Equation Dispersion Inversion. To invert for the 3D background S-wave velocity model (low-wavenumber component), I first propose the 3D wave-equation dispersion inversion (WD) of surface waves. The results from the synthetic and field data examples show a noticeable improvement in the accuracy of the 3D tomogram compared to 2D tomographic inversion if there are significant 3D lateral velocity variations. 2. 3D Wave Equation Dispersion Inversion for Data Recorded on Rough Topography. Ignoring topography in the 3D WD method can lead to significant errors in the inverted model. To mitigate these problems, I present a 3D topographic WD (TWD) method that takes into account the topographic effects in surface-wave propagation modeled by a 3D spectral element solver. Numerical tests on both synthetic and field data demonstrate that 3D TWD can accurately invert for the S-velocity model from surface-wave data recorded on irregular topography. 3. Multiscale and layer-stripping WD. The iterative WD method can suffer from the local minimum problem when inverting seismic data from complex Earth models. To mitigate this problem, I develop a multiscale, layer-stripping method to improve the robustness and convergence rate of WD. I verify the efficacy of our new method using field Rayleigh-wave data. 4. Natural Migration of SurfaceWaves. The reflectivity images (high-wavenumber component) of the S-wave velocity model can be calculated by the natural migration (NM) method. However, its effectiveness is demonstrated only with ambient noise data. I now explore its application to data generated by controlled sources. Results with synthetic data and field data recorded over known faults validate the effectiveness of this method. Migrating the surface waves in recorded 2D and 3D data sets accurately reveals the locations of known faults.
Date of AwardSep 2019
Original languageEnglish (US)
Awarding Institution
  • Physical Sciences and Engineering
SupervisorGerard Schuster (Supervisor)

Cite this