Evaluation of multicomponent depth migration

This paper presents an evaluation of multi component depth migration. This type of migration can be carried out by most techniques commonly used for conventional single component acoustic migration. These can be broadly categorized as downward continuation methods and reverse time migration methods. We demonstrate downward continuation by extensions of the phase shift method introduced by Gazdag (1978) and Bolodni et al. (1978), as well as by the generalized phase shift method (Kosloff and Kessler, 1986). We then evaluate elastic reverse time migration (Sun and McMechan, 1986, Chang and McMechan, 1987). This study indicates that although multi component elastic migration can be carried out technically, however from a practical point of view this may be using the migration concept beyond its applicable limits. Specifically we show that when the correct velocity is used reverse time migration is incapable of reproducing correct amplitudes. Furthermore, use of the exact velocity and the free surface boundary condition creates false events in all methods. Consequently, smoothed velocities or a modification of the boundary conditions are needed thus abandoning hope of reproducing correct amplitudes. It may Finally turn out that simple acoustic migration of P and S potentials is the most promising alternative. We present a series of tests on simple synthetic examples which supports this opinion.