Jonathan E. Downton, Laurence R. Lines

Traditionally, AVO is performed on NMO-corrected gathers. However, NMO correction is principally a kinematic correction that distorts the amplitude and character in an offset dependent fashion. This leads to bias and error in subsequent AVO reflectivity estimates. This paper develops a simultaneous AVO-NMO inversion methodology thereby avoiding these distortions. Since the inversion problem is ill-conditioned, frequency data and a priori reflectivity constraints are introduced to improve the condition number of the problem.

The methodology is demonstrated on a synthetic and real data example. The synthetic example shows that the AVO-NMO methodology introduces less bias and error into the reflectivity estimates. Further, the frequency bandwidth and signal-to-noise of the AVONMO synthetic results are better than the AVO estimates after NMO correction. The data example also shows that the AVO-NMO reflectivity estimates have higher frequency and better signal-to-noise ratios.