Abstract: The non-stationary variation of frequency content of seismic signal with time characterizes information related to the properties of the subsurface reflection stratification. Time-frequency analysis of seismic signal was an important tool for detecting local properties of spectral content. In order to adapt to the seismic signal processing in real application, a generalized S transform (GST) was presented by altering the normal S transform. The seismic sequence identification section with high resolution was achieved by applying GST to analyze and compensate the high frequency content of seismic signal. The real data processing demonstrated that using GST-based time-frequency decomposition, the properties analysis of seismic stratification of sandstone reservoir possess both high resolution and better signal-to-noise ratio. Time-frequency spectral decomposition of seismic signal based on GST can aid in the classification of seismic sequences and serve as an important tool for seismic interpretation.

Key words: seismic sequence, time-frequency analysis, S transform, generalized S transform, spectral decomposition