Waveform inversion for shear velocity and attenuation via the spectral-element adjoint method

Fan V. Wu, Dmitry Borisov, Frederik J. Simons, Paul Williamson

Research output: Contribution to journalConference articlepeer-review

1 Scopus citations

Abstract

Intrinsic attenuation affects both amplitude and phase behavior of the seismic wavefield. Its three-dimensional distribution in the subsurface needs to be robustly estimated to make earth models accurate. Full Waveform Inversion (FWI) has become the technique of choice for recovering high-resolution velocity structure, but its ability to reliably recover attenuation variations remains challenging. Multistage and hierarchical approaches offer promise to control the delicate trade-off between resolving elastic and anelastic structure in inversions for the shear velocity VS and quality factor Qµ, in regimes with strong surface waves prone to cycle skipping. We present a procedure for their joint reconstruction via spectral-element adjoint-based full-waveform modeling. Starting from a checkerboard example, we develop a frequency and offset continuation strategy for onshore active-source acquisition. In the early stages of our algorithm, the inversion targets the refinement of the shear-wave speed model by gradually widening the frequency band and increasing the offset range. The shear quality factor is of major consideration only in the final stage of the inversion. Our procedure is robust, competitive, and practical.

Original languageEnglish (US)
Pages (from-to)697-701
Number of pages5
JournalSEG Technical Program Expanded Abstracts
Volume2021-September
DOIs
StatePublished - 2021
Event1st International Meeting for Applied Geoscience and Energy - Denver, United States
Duration: Sep 26 2021Oct 1 2021

All Science Journal Classification (ASJC) codes

  • Geotechnical Engineering and Engineering Geology
  • Geophysics

Fingerprint

Dive into the research topics of 'Waveform inversion for shear velocity and attenuation via the spectral-element adjoint method'. Together they form a unique fingerprint.

Cite this