Deconvolution of Seismic Data - An Overview

It is common practice to model a seismic trace as a convolution of the reflectivity function of the earth and an energy waveform referred to as the seismic wavelet. The objective of deconvolution is to extract the reflectivity function from the seismic trace. We will describe four techniques which have been and are being used to accomplish this objective. These techniques are predictive deconvolution, homomorphic filtering, Kalman filtering, and deterministic deconvolution. In addition, we shall outline the physical effects governing the bandwidth of the seismic data such as shooting geometry, recording filters, and absorption.