Full text loading...
Migration Velocity Analysis is a technique to determine the large-scale velocity structure of the subsurface from the analysis of reflected data. We review here the Differential Semblance Optimization version (DSO-MVA) that estimates the focusing in extended panels. Theoretical arguments ensure a convergence to the global optimal with a local optimization process. In practice, it faces a number of challenges: reflected events should be first extracted from the observed data; the computational cost is currently expensive due to the number of wavefields; a proper amplitude processing is needed before the evaluation of the focusing. The recently developed approximate inverses for replacing the standard adjoint images have played a significant role. In its current shape, DSO-MVA ensures focused images as well as a (linear) data fit at convergence. Recent researches consider a fully non-linear imaging tool, while preserving the DSO-MVA convergence properties.