Anisotropic Velocity Updating For Converted-Wave Prestack Time Migration

The conventional method to estimate velocities for converted-wave (C-wave) prestack time migration is an awkward procedure, because the P-wave velocity (V-P) comes from P-wave processing, the velocity ratio gamma (V-P/V-S) is estimated from C-wave data, and the S-wave velocity (V-S) is then derived from V-P and gamma. Instead, by using the C-wave velocity (V-C), effective gamma (gamma(eff)), and anisotropic parameter chi(eff), velocity updating becomes straightforward and more reliable. To update V-C for converted-wave time migration, one can carry out hyperbolic moveout analysis on the hyperbolic moveout migrated common mid-point (HMO-MCMP) gathers (Dai and Li (2004). However, the errors in initial gamma(eff) and anisotropic parameter chi(eff) can only be corrected by a trial-and-error method. In this paper we propose to remove the effects of initial gamma(eff) and chi(eff) in the HMO-MCMP gathers by inverting the moveout related to the initial gamma(eff) and chi(eff). This enables a full non-hyperbolic velocity analysis to be conducted in order to update not only V-C but also gamma(eff) and chi(eff). To obtain reliable chi(eff), we also developed a simultaneous PP/PS anisotropic parameter estimation method so that the eta(eff) estimated from P-wave is immediately compared with the eta(eff) derived from chi(eff) by using C-wave data. This provides a better constraint for estimating anisotropic parameters. The method has been tested and has shown consistent improvement in converted wave prestack time migration velocity estimations.