Deconvolution is an important seismic processing tool for increasing the resolution with which the seismic method can image the earth. There are many techniques for accomplishing the task, each of them using as much information about the data as is known and making certain assumptions about the characteristics of the seismic records or the underlying geology. One of the key assumptions made in most methods is that the data are statistically stationary, that is, the data characteristics are unchanging with time. Since this is often not even approximately true, there is strong motivation to construct a deconvolution method that does not rely on stationarity for its effectiveness. Presented here is a practical implementation of such a method, based on the Gabor transform, in the ProMAX processing package. This chapter describes the step-by-step numerical procedure, provides an explanation of all parameters, and illustrates the use of the algorithm on several examples of real data from field surveys.
View full article as PDF (1.03 Mb)