SPECTRAL DECOMPOSITION TECHNIQUE BASED ON STFT AND CWT FOR IDENTIFYING THE HYDROCARBON RESERVOIR

Abstract

The spectral decomposition is one of the advanced interpretation techniques such as seismic inversion, amplitude versus offset analysis, and seismic attribute that helpful in direct interpretative approach in seismic exploration. This technique is a transformation algorithm, thus a signal can be transformed into its varying frequency contained in the seismic signal. There are a variety of spectral decomposition algorithms in the decomposing seismic signal from time domain into frequency domain. These algorithms include Short Time Fourier Transform (STFT) and Continuous Wavelet Transform (CWT). The STFT algorithm is a conventional and simple technique for computing a time-frequency spectrum, which is based on the application of Fourier transform. However, the STFT algorithm has a problem related to the frequency resolution. In its implementation, this algorithm is limited by predefi ned window length. In contrast, the CWT algorithm is believed to be able to overcome the limitation of window length. The CWT threats wavelet at certain window length, which is defi ned by the characteristics of the wavelet. In this study, the comparison between spectral decomposition technique based on STFT and CWT method was performed, particularly in its application to the synthetic and real data set. Each algorithm has its own advantages and disadvantages in decomposing the seismic signal. Further, this analysis can be used as a reference to select one of two algorithms for the specifi c application. The synthetic data set application shows that CWT algorithm produces better frequency resolution compared to STFT algorithm. In addition, the real data set application shows that time frequency section of the seismic line provides a spectral feature, which is useful to identify the hydrocarbon reservoir, which is associated with low-frequency shadow zone.