This thesis addresses the efficiency improvement of seismic wave modeling and migration in anisotropic media. This improvement becomes crucial in practice as the process of imaging complex geological structures of the Earth's subsurface requires modeling and migration as building blocks. The challenge comes from two aspects. First, the underlying governing equations for seismic wave propagation in anisotropic media are far more complicated than that in isotropic media which demand higher computational costs to solve. Second, the usage of whole prestack seismic data still remains a burden considering its storage volume and the existing wave equation solvers. In this thesis, I develop two approaches to tackle the challenges.
In the first part, I adopt the concept of prestack exploding reflector model to handle the whole prestack data and bridge the data space directly to image space in a single kernel. I formulate the extrapolation operator in a two-way fashion to remove he restriction on directions that waves propagate. I also develop a generic method for phase velocity evaluation within anisotropic media used in this extrapolation kernel. The proposed method provides a tool for generating prestack images without wavefield cross correlations.
In the second part of this thesis, I approximate the anisotropic models using effective isotropic models. The wave phenomena in these effective models match that in anisotropic models both kinematically and dynamically. I obtain the effective models through equating eikonal equations and transport equations of anisotropic and isotropic models, thereby in the high frequency asymptotic approximation sense. The wavefields extrapolation costs are thus reduced using isotropic wave equation solvers while the anisotropic effects are maintained through this approach. I benchmark the two proposed methods using synthetic datasets. Tests on anisotropic Marmousi model and anisotropic BP2007 model demonstrate the applicability of my approaches.
|Date of Award||May 2014|
|Original language||English (US)|
- Physical Science and Engineering
|Supervisor||Tariq Ali Alkhalifah (Supervisor)|