Conclusions

We have introduced a method for interpolating 3-D data based on a local mono-planewave assumption. As expected, the algorithm works well when this assumption is met. When conflicting dips are present, only the one dip that gives the best coherence is used for interpolation by the algorithm, as it stands now. Problems can arise when two (or more) different dips are present in the small overlapping time windows used by the program.

It should be possible to extend the method to handle more than one dip. Since we are already doing a global search of a 2-D dip space to find the best dip, we could also pick other dips that give large coherencies, and incorporate these other dips into the interpolation. However, the resolution in dip space is limited, because the operator is localized, so it may not be possible to distinguish different dips unless they are widely separated. Another possibility for handling multiple dips is a plane-fitting scheme.