Introduction

In SEP-73, Cole and Claerbout 1992 presented a 3-D data interpolation scheme. For each output location, a number of neighboring traces were combined along a best trajectory picked from an estimate of event coherency as a function of 3-D dip.

A somewhat similar scheme is used by Reiter et al. 1993 in 2-D - not as an interpolation tool, but to perform noise suppression and moveout-dependent filtering. Two modifications suggested by this 2-D technique can be used to improve the noise suppression and moveout-dependent filtering capabilities of the 3-D interpolation scheme.

First, the output data are weighted by the semblance (measured along the best dip) as a function of time. While the stacking of the interpolation attenuates noise and events that do not lie along the best-dip direction, typically only a few neighboring traces are used, and the amount of noise suppression is small. The semblance weighting greatly enhances the ability to attenuate energy other than the event that gives rise to the high coherency value.

Second, a median filter is applied along the best-dip direction, rather than the mean of conventional stacking. Because it is less sensitive to outliers, the median will remain unaffected by strong noise bursts on isolated traces.

In this note, I show examples of the noise suppression and velocity filtering power of the new 3-D interpolator. The examples include 2-D synthetics and a real 3-D example, a quarry blast recording from the SEP passive seismic experiment.