Optimal Background Model Extraction and Improvement of RTM Illumination on Basis of Born Modeling

Seismic processing methods exist that view the actual velocity configuration of the subsurface as being composed of a reference model plus a perturbation of the reference velocity. Seismic migration (or imaging) requires a velocity model to locate structure at depth. Recent progress in seismic inversion provides an estimated perturbation corresponding to a given background/reference model before imaging. This motivates the study of Born modeling, which requires as input, both a background velocity model and a perturbation in the velocity. In this thesis, I examine the background model used in Born modeling. Different models are tested in numerical experiments in order to find the optimal choice for the background. In this study, when a Born modeling result has a minimum difference with a finite-difference (FD) modeling result, the background model used in Born modeling is called the optimal model. All the background models in my numerical experiments are generated by the Gaussian smoothing of actual models. Please note that the actual models are assumed to be known in all tests, since I am focused on forward modeling in this part of this study. In the second section, I apply Born modeling data for reverse-time migration (RTM), the later being a processing or inversion step. Industry standard RTM is a linear and model-dependent migration method. RTM typically works with a smoothed background model to locate the imaging points using only primaries. RTM has its own technical disadvantages of producing low illumination beneath high velocity structures, such as salt bodies. To improve the illumination beneath salt, I introduce scattering by Born modeling into wave-propagations in an RTM scheme. An improved RTM results from adding scattering into both source and receiver wave-field. I illustrate the effects of improved RTM with tests using a synthetic sub-salt model. The improved imaging result helps with delineating part of the salt boundary and salt-sediment connections. Finally, the merits and shortcomings of both original and improved RTM are analyzed.