Summary

I proposed an antialiasing filter for reverse-time migration (RTM). It is similar to the traditional antialiasing filter used for Kirchhoff except that it now accounts for both primary and multiple reflection events. The antialiasing filter is applied to the generalized diffraction-stack migration operator, which gives the same migration image as computed by an antialiased RTM. Numerical results verify the effectiveness of this procedure. The main disadvantage in applying this antialiasing filter is the increased computational cost compared to standard RTM, except in a target-oriented mode. An alternative to antialiased RTM is least-squares migration (LSM) or migration deconvolution (MD). It is observed that LSM (Nemeth et al., 1997) and MD (Yu et al., 2006) appears to overcome aliasing problems and, unlike the traditional antialiasing filter, does not low-pass filter the data. Until recently, LSM was computationally expensive, but Dai et al. (2012) shows that a phase encoded multisource formulation makes it no more expensive than standard RTM. Therefore, MD and LSM might be a viable alternative to an explicit low-pass antialiasing filter applied to the data.

Figure: Reverse-time migration (RTM) examples.