Elastic Parameters Inversion in 3D Anisotropic Media
Abstract
Seismic anisotropy, as an important wave phenomenon, has been incorporated into various geophysical studies due to the enhanced computational power and advanced data acquisition in recent decades. There are many factors which can contribute to seismic anisotropy, such as lattice preferred orientation (LPO), layering during sedimentation and aligned fracture set during linear slip deformation. Alternatively, a better understanding of anisotropy reflects the rock's physical properties and the deformation it experienced. Most anisotropy inversion studies are performed in 2D and focus on obtaining Thomsens parameter, which limits their maximum complexity to lower-order symmetry, in addition, cannot incorporate the effects induced by fracture. Here we develop a new strategy that allows for the determination of the full set of elastic stiffnesses of sections of the 3D structure. The forward portion of the inversion process relies forward modeling of the P-and S- wavefields using a second-order finite-difference rotated staggered grid (RSG), these calculations are developed by adapting existing public domain algorithms on a shared memory cluster using OpenMP. The inversion process minimizes the misfit between the observed and modelled transit times and retrieves optimized parameters effectively using Bayesian Optimization to reduce the run number of the forward modeling. The synthetic experiment aims to validate the feasibility of anisotropy inversion. The first synthetic experiment inverts the geometry of a 3D tilted transverse isotropy (TTI) material using qP-transit time only. The second experiment inverts the five elastic constants of a VTI (vertical transverse isotropy) material with both qP- and qS-transit time. The third experiment searches for the normal and tangential compliances of a prefer-oriented fracture set with a known background. Our ultimate goal is to apply this inversion method to the analysis of borehole DAS and geophone data acquired within the damage zone of the Alpine Fault, South Island, New Zealand.
- Publication:
-
AGU Fall Meeting Abstracts
- Pub Date:
- December 2021
- Bibcode:
- 2021AGUFMNS35C0377Z