Seismic data processing in transversely isotropic media: a plane wave approach

Occurrences of anisotropy in the seismic data are widespread at all scales. Thus inclusion of these anisotropic effects becomes important for obtaining correct images and target depths. This dissertation addresses some problems pertaining to seismic data processing in transversely isotropic media. I have formulated an interactive traveltime analysis procedure for P-waves in delay-time, slowness domain for wave propagation in the transversely isotropic media with a vertical axis of symmetry (VTI). Using the assumption of weak anisotropy I obtained a simple and physically intuitive two-term expression for vertical slowness, which can be used in direct estimation of interval elliptic velocity and the anisotropic parameter kappa. I have also developed a method to automatically estimate these parameters using a non linear inversion technique called very fast simulated annealing. Conventional ray tracing methods are difficult to apply in the VTI media. Unavailability of vertical P wave velocity restricts us to use the time gridded elliptic velocity and kappa as inputs for traveltime computation in offset-time domain. However I have formulated a ray tracing technique based on the Fermat's principle and perturbation theory. The method uses phase velocities unlike other methods, which use group velocities. Head wave paths are not included in the traveltime computation. Comparison with more exact Finite Difference Eikonal solvers for both 1-D and 2-D models show small residuals. I have used source traveltimes computed using the interval elliptic velocity and kappa models to perform prestack split-step Fourier and Kirchhoff time migration in the VTI media. Migration using parameters estimated from moveout analysis and computed source traveltimes for Gulf of Mexico data show good results. The common image gathers show increased flattening after incorporation of anisotropic effects.