Multiple migration of VSP data for velocity analysis

Summary We investigate the new method that combines the migration of VSP data and updating of the velocity model. The method is based on the comparison of subsurface images obtained by using different types of waves: the primary reflections and surface-related multiples. As a measure of similarity of the images we use the functional based on cross-correlation. Estimation of the velocity model parameters is carried out by maximization of this functional. Therefore the resulting velocity model provides the maximal similarity between images obtained by using the primary reflections and surface-related multiples. We develop an iterative procedure to maximize the functional and demonstrate the efficiency of the method using the synthetic walk-away data. The method can be used for very short VSP receiver arrays and allows us to estimate the interval velocities below receivers, which is considered to be a challenge with VSP acquisition geometry.