QR Radius-Basis-function Finite Difference Method with Irregular Topography and Its Application to Elastic Reverse Time Migration

Duan PeiRan,Gu BingLuo,Li ZhenChun,Li QingYang
DOI: https://doi.org/10.6038/cjg2023q0713
2024-01-01
Abstract:Elastic Reverse Time Migration (ERTM) is one of the most accurate migration methods because it is free of dip angle and migration aperture and images complex underground structures with high accuracy. Theoretically, it is one of the most accurate migration imaging methods for multi-wave and multi-component data. It is well understood that wavefield extrapolation is the core of reverse time migration. Because the traditional wavefield extrapolation method is based on the flatten-surface assumption, carries out regular grids based on fixed sampling step size, and uses step approximation to deal with the surface topography, it often brings a loss of precision when dealing with the seismic data acquired in the complex structure with surface topography. We analyze the dispersion relation and stability conditions of Radial Basis Function Finite Difference Method (RBF-FDM) in elastic wave simulation based on mesh-free node distribution. On this basis, we develop the QR-RBF-FDM, design an optimized adaptive node generation method for the model with irregular topography in a flexible and efficient manner, derive the corresponding free boundary condition and hybrid absorbing boundary conditions in elastic modeling, and finally form a new elastic wave simulation method for irregular topography. In addition, the proposed QR-RBF-FDM is used to extrapolate seismic wavefield in mesh-free-based ERTM, where the QR-RBF-FDM is also applied to the P- and S-wave vector decomposition formula. Finally, we validate the proposed method with the Gauss Hill model, the Sag model, and the Marmousi-2 model, and the numerical results demonstrate the feasibility and effectiveness of the proposed method.
What problem does this paper attempt to address?