Using a time-frequency distribution to identify buried channels in reflection seismic data

Geological events such as thin beds and channels cannot be easily revealed on seismic sections due to the interference of reflections from the top and bottom of the layer. Buried channel is one of the hydrocarbon traps, which is important in oil and gas exploration. Spectral decomposition can be used to indicate subtle changes in channel thickness. In using the Fourier transform only the frequency content of the changes is displayed; hence, the exact onset of the changes is missed. Time-frequency distributions are suitable approaches to display and interpret information embedded in non-stationary signals such as seismic signals. Spectral decomposition can be used for seismic attributes calculation, which is used for imaging of thin beds. The conventional spectral decompositions such as Short Time Fourier Transform (STFT) and Wigner–Ville Distribution (WVD) have limitations in terms of Heisenberg uncertainty principle and cross-terms artefacts, respectively. In this paper, we used the Reduced Interference Distribution (RID) for buried channel identification to overcome the mentioned limitations. We compared the obtained results of RID with those of Smoothed Pseudo WVD. The real seismic data is selected from one of the western oil fields of Iran. Our results indicate that a better resolution is achieved by RID in both vertical and lateral stratigraphic features.