Spectral Decomposition Made Simple

Abstract

Spectral decomposition enables the resolution of seismic data to be improved significantly yielding a new possibility to map thin layers such channel sands and any other stratigraphic features. It has also been used in reservoir characterization. There are three methods for implementing spectral decomposition i.e., The Short Time Fourier Transform, the Continuous Wavelet Transform and the Matching Pursuit Decomposition. Among three of them, the Matching Pursuit Decomposition seems to be the most sophisticated one. It gives the best resolution among them. A simple and logical approach for explaining the spectral decomposition methods together with real data examples are presented in this paper by avoiding complex mathematical formulation.