Real data results

In order to implement and test Stolt residual migration for converted waves, I use a 2D line from a 3D cube of the OBC data set of the Alba oil field. The 2D line is extracted from the 3D portion after geometry regularization ().

The velocity models were provided by Chevron. I start with these velocity models. I first handle the PP section, perform Stolt residual migration on this section only and update the velocity model with a selection of combined values of $\rho_p$. After the P velocity model is updated, I perform the PS migration and residual migration; however, I do not scan over $\rho_p$ values. I scan only over $\rho_s$ and $\gamma$ values.

The velocities vary both horizontally and vertically. Therefore, there will not be a single set of parameters to correct the entire section. The $\gamma_0$ value will be choosen at Alba's depth at the center of the section.