Dip picking

The most important step in the interpolation is the dip picking step, because if the dips picked are wrong, the remaining interpolation steps have no meaning. Many techniques, such as the slant stack method, the eigen-structure method (Biondi and Kostov, 1989) and the plane wave destructor method (Claerbout, 1991), can be used to find dips. Among these methods, I have chosen the slant stack method because it can be applied to any kind of missing traces problem such as randomly missing and irregularly sampled traces.

After $\tau - p$ transform, a dip spectrum is made by picking maximum absolute value on the $\tau$ axis for each ray parameter. The dip spectrum usually looks too spiky and needs some smoothing for picking dominant dips. The criteria for picking dips from the dip spectrum vary depending on the data quality and the purposes of interpolation. In the approach using PEF, the number of coefficients of the filter determines the number of dips to pick. In the dip spectrum, one can pick several dominant dips like the PEF approach, or any number of dips according to a given criterion. Thus the dip picking from the dip spectra is more flexible in deciding the dips to interpolate.