How Many Iterations Are Necessary for Obtaining Proper Depth Velocity Model?

Prestack Kirchhoff Depth Migration (PSDM) is a robust method for obtaining proper image of complex subsurface. In areas of rapid velocity changes, energy is dispersed in such a way that conventional stacking velocities can not provide desired resolution of the data.

Building a proper velocity model for PSDM, can bring significant improvement of seismic data quality (both resolution and continuity) and proper spatial structural positioning. For these reasons, velocity estimation process has great influence in depth imaging.

Construction of a velocity model for PSDM was conducted in two phases: preparation of initial model and its updating. The initial velocity model was derived by RMS stacking velocities. A good initial model limits ambiguity and increases convergence of tomographic solutions. Therefore, an initial model plays key role in a process of migration velocity analysis for complex structures. In the case of tomographic inversion, a wrong provisional velocity distribution makes it impossible to obtain convergent results.

Usually two to five iterations of tomographic inversion are necessary to obtain the final result. Relatively small number of iterations is needed, mainly due to definition of proper velocity gradients. During iterative tomographic updates, the resolution and plausibility of the model increase and small velocity changes can be defined.