The linearized analysis of depth perturbations in ADCIGs obtained by anisotropic migration shows that the RMO function observed when the migration velocity is inaccurate is a function of both the phase aperture angle and the group aperture angle. The synthetic-data examples show that the linearized expression of the RMO function accurately predicts the actual RMO function measured after wavefield migration.
The real data results confirm the accuracy of the theory developed in this paper. The RMO curves predicted by the theory match extremely well the RMO functions observed in the ADCIG migrated according to the assumptions underlying the theory. We observe fairly large differences in RMO functions observed between the ADCIGs computed assuming an isotropic homogeneous half-space hanging from the sea floor and the ADCIGs computed assuming an isotropic heterogeneous model. These discrepancies demonstrate the sensitivity of the RMO analysis to the accuracy with which the background velocity predicts the actual ray bending. This sensitivity is higher for anisotropic media because of the velocity dependence on the propagation angle, suggesting that a tomographic MVA might be even more necessary for anisotropic media than for isotropic ones.
Tomographic updating of the anisotropic parameters from ADCIGs can be based on the same fundamental concepts used to derive the RMO functions presented and tested in this paper. I therefore derive the linearized relationship between depth errors measured in ADCIGs and traveltime errors accumulated along the wavepaths. This relationship should lead to the development of anisotropic MVA methods based on tomographic velocity-updating procedures.
I would like to thank the ExxonMobil Exploration Company for making the Gulf of Mexico data set used in this paper data available to SEP through the generous efforts of Ruth Gonzalez and Joe Reilly. The anisotropic parameters cube were kindly shared with us by Laura Bear and Jerry Krebs, also at ExxonMobil.