The purpose of the imaging condition in prestack depth migration is to obtain a reflectivity estimate with high spatial resolution and correct position. This goal can only be reached by perfectly matching an upgoing wavefield with a downgoing wavefield. At each depth of interest, the ratio of the upgoing (reflected) wavefield to the downgoing (incident) wavefield estimates the reflectivity. Usually, the upgoing wavefield is created from downward extrapolated surface data, while the downgoing wavefield is created by downward continuation of the source signature through a velocity model. The upgoing waves were influenced by earth attenuation, anisotropy, multiples, velocity and density of the medium, and etc. All these factors cannot be exactly estimated in the complex geology environment and cannot be fully included in the downgoing wavefield. Thus, the downgoing wavefield is not consistent with the upgoing one at the image position, which reduces the resolution of the final image. An alternative becomes possible if recording is done simultaneously in surface and downhole receivers. Assuming that the downhole receivers have sufficient depth coverage to enable wavefield separation, it then becomes possible to use VSP wavefield separation to estimate the downgoing wave. The estimated downgoing waves are much closer to the real downgoing wavefield than estimated from forward modelling. This work tried to build a downgoing wavefield from synthetic VSP data, and likewise the upgoing wavefield received on the surface, and then created the depth image through the FOCI approach (Margrave et al, 2004). The result shows that the depth image marches the velocity model very well.
View full article as PDF (0.76 Mb)