Common Reflection Point Stacking: A Macro Model Driven Approach to Dip Moveout

摘要

Dip moveout or DMO is known to be a valuable tool in today's seismic data processing schemes. DMO can be seen as an extension to the conventional processing sequence of NMO correction, CMP stacking and poststack migration. Processing with DMO approximates the result of (time) migration before stack. DMO is generally based on the assumption that the medium velocity is constant, although approximate generalizations of DMO in the presence of depth-variable velocities have been proposed. In the thesis a depth-oriented approach to DMO, which the authors call 'common reflection point (CRP) stacking' is presented. In this approach, the CRP operator design is based on a macro model of the subsurface, in which ray tracing is performed. Lateral as well as vertical velocity variations are therefore allowed. To obtain an exact (with respect to travel-times) CRP operator for the inhomogeneous media, offset ray tracing for all offsets should be done. It is shown in the thesis that the multi-offset ray tracing can be approximated by efficient zero-offset ray tracing, in combination with a simple mathematical relation.