- Copyright © 2002 Society of Exploration Geophysicists
The convergence of advanced computing technologies and innovative algorithm designs has made it possible to provide 3D wave equation depth migration methods to many explorationists. It is well known that the wave equation method can provide the full wavefield solution to most imaging problems. On the other hand, the Kirchhoff technique, which has become the industry's standard prestack depth migration (preSDM) algorithm, has some inherent limitations, such as diverging raypaths in steeply dipping areas where high lateral velocity contrasts exist. More accurate imaging in such challenging geologic conditions can often be addressed by a wave equation (WE) depth migration.
Numerous investigators have compared the results of the two methodologies, as applied to the 3D SEG/EAGE salt model. This paper, however, presents a case study in the U.S. Gulf Coast where the 3D Kirchhoff preSDM method was initially used to enhance the images of a salt body and surrounding sediments. This was followed by application of WE to yield finer details of the subsurface, especially subsalt structures.
The wave-equation method in this study is a shot-based approach that employs the phase-shift and split-step Fourier plus interpolation (SSFPI) algorithms for wavefield downward extrapolation. These two approaches are adaptively implemented according to the velocity model structure. Phase shift is used in constant velocity areas (water and salt bodies), and SSFPI in sedimentary strata with varying velocity. The separation of salt (including water in offshore areas) from the sedimentary background greatly reduces the number of reference velocities required in the SSFPI algorithm. In comparison to other extrapolation methods (such as the Fourier finite difference algorithm) phase-shift and SSFPI do not suffer from numerical dispersion and anisotropy errors and are accurate at wide angles. High angle imaging is further improved by employing a modified imaging condition that compensates uneven energy distribution due to velocity variation …