Figure 1 graphically represents the basic idea behind migration velocity analysis. Velocity analysis after migration estimates the velocity along the vertical from the surface to the migrated output point. Of course, because it assumes that the source and receiver are coincident, one cannot base a velocity analysis approach on this kind of concept alone.
Developing a reasonable approach to velocity analysis requires that we have redundant data and that we exploit this redundancy in order to provide velocity estimates. Figure 2 illustrates the geometry in the case of non-zero offset from a ray-based modeling perspective. We see that the recorded data, in yellow, is placed below the surface midpoint. The red and green rays illustrate the paths taken by the illuminating energy. After migration, the yellow apparent event is now places at its true location below the vertical black image ray. Migration has moved the source and receiver locations directly above the correct location. Thus, after migration, source and receiver locations are lost. We can maintain redundancy by migrating common-offset sections and hopefully develop velocity analysis approaches that are applied after migration. Our goal is to provide velocity analysis methodologies that exploit residual redundancy to refine initial velocity estimates.