True-amplitude prestack depth migration

摘要

Most current true-amplitude migrations correct only for geometric spreading. We present a new prestack depth-migration method that uses the framework of reverse-time migration to compensate for geometric spreading, intrinsic Q losses, and transmission losses. Geometric spreading is implicitly compensated by full two-way wave propagation. Intrinsic Q losses are handled by including a Q-dependent term in the wave equation. Transmission losses are compensated based on an estimation of angle-dependent reflectivity using a two-pass recursive reverse-time prestack migration. The image condition used is the ratio of receiver/source wavefield amplitudes. Two-dimensional tests using synthetic data for a dipping-layer model and a salt model show that loss-compensating prestack depth migration can produce reliable angle-dependent reflection coefficients at the target. The reflection coefficient curves are fitted to give least-squares estimates of the velocity ratio at the target. The main new result is a procedure for transmission compensation when extrapolating the receiver wavefield. There are still a number of limitations (e.g., we use only scalar extrapolation for illustration), but these limitations are now better defined.