Variable factor S-transform seismic data analysis

Todor I. Todorov, Gary F. Margrave

Most of today's geophysical data processing and analysis methods are based on the assumption that the seismic signal is stationary and employ an extensive use of Fourier analysis. However due to various attenuation mechanisms of the Earth, the seismic signal is not stationary. The short-time Fourier transform (Gabor transform) has been developed to deal with nonstationary signals. The variable factor S-transform is an extension of the Gabor transform and provides better time-frequency decomposition of a nonstationary signal for all frequencies. An S-transform deconvolution method is developed as an extension of the nonstationary Gabor deconvolution reported in the literature. The new method is tested on a constant Q synthetic data and shows superior results over the traditional stationary Wiener deconvolution and an improvement over the nonstationary Gabor deconvolution. In a separate application an f-t-x CDP noise attenuation method is developed. A synthetic example proves the effectiveness of the f-t-x noise attenuation for both, high-amplitude linear noise and random noise.