HR: 0830h
AN: S11F-0360    [PDF]
TI: Optimum Time-Frequency Decomposition For Seismic Data Using Continuous Wavelet Transform
AU: * Sinha, S
EM: satish@ou.edu
AF: University of Oklahoma, 100 E. Boyd St. 810 SEC,, Norman, OK 73019 United States
AU: Routh, P
EM: routh@cgiss2.boisestate.edu
AF: Boise State University, 1910 University Drive, Boise, ID 83725 United States
AU: Anno, P
EM: Phil.D.Anno@conocophillips.com
AF: ConocoPhillips, 600 North Dairy Ashford P.O. Box 2197, Houston, TX 77252 United States
AB: Time-frequency decomposition is an important signal analysis tool for seismic data. The overall frequency content of a signal can be obtained from the Fourier transform. However, for a non-stationary signal, such as seismic signal, whose frequency content varies with time, 1D transformation in frequency is not sufficient. Traditionally, 2D representation in time and frequency space for a 1D signal is achieved by taking Fourier transform over a short-time window. This method is commonly known as short-time Fourier transform (STFT). Time-frequency resolution in STFT is limited by the choice of a window length. Windowing problem in time-frequency analysis is absent in the continuous wavelet transform (CWT) method. CWT utilizes the property of dilation and translation of a wavelet and produces time-scale map where scale, defined in terms of length of time support of a wavelet, represents a frequency band. However, scale can be converted to frequency and the time-frequency map thus produced is called TFCWT. It is produced from CWT in two steps: 1. producing time-scale map, and 2. converting the scale to frequency. TFCWT has optimum time-frequency resolution, i.e. higher frequency resolution at lower frequencies and higher time resolution at higher frequencies, which makes it useful for seismic data interpretation. Visualization and interpretation of seismic sections in frequency space using single frequencies from the TFCWT spectra can be utilized to enhance low frequency shadows caused by hydrocarbon reservoirs. This idea can also be extended in interpreting time slices from 3D seismic data in frequency space. This method has been used to identify thin beds below tuning thickness.
DE: 0902 Computational methods, seismic
DE: 0910 Data processing
DE: 3299 General or miscellaneous
DE: 7299 General or miscellaneous
SC: Seismology [S]
MN: 2003 Fall Meeting