Estimation of AVO in the presence of noise using prestack migration

Abstract

Amplitude versus offset (AVO) studies are easiest to implement if the raw data is free of noise and forward modeling can be used to make the interpretation. In general these ideal conditions do not exist and a number of processing approaches have been suggested in order to prepare the data for AVO studies. Unfortunately most of the noise reduction techniques affect the signal as well as the noise. A better estimate of AVO on noisy data can be accomplished using an amplitude-preserving prestack time or depth migration on constant offset gathers. The clear advantage to this approach is the ability to emphasize the desired signal along a migrated wavefront at the expense of other coherent and incoherent signals while performing the migration. Different types of noise can be handled with this technique as long as the moveout of the desired signal is distinquishable from the noise. One undesirable side-effect of viewing prestack migrated data is the wavelet stretch associated with the imaging principle used in most migration algorithms. The wavelet stretch is same type of effect observed in RMO stretch. In this presentation an intuitive discussion of migration stretch effects will be discussed along with amplitude preservation during migration. The prestack migration method is applied to an example dataset with strong multiples from the Gulf of Mexico.