Demodulated synchrosqueezing s-transform and its application and economic value in seismic data analysis

Jun Zhou,Liejun Li,Qingnan He,Wei Liu,Hongzhi Ma
2023-01-01
Journal of seismic exploration
Abstract:The synchrosqueezing transforms based on short-time Fourier transform (FSST) and wavelet transform (SSWT) have been widely used in the analysis of non-stationary signals. In the light of the superiority of S-transform (ST) over short-time Fourier transform (STFT) and wavelet transform (WT) in time-frequency representation (TFR), we propose a novel time-frequency analysis method, termed as demodulate synchrosqueezing S-transform (DSSST), which achieves a highly energy-concentrated TFR by making full use of two operations including demodulation technique and ST-based synchrosqueezing transform (SSST). The formulas for the DSSST and its inverse transform are derived. Synthetic example shows that the DSSST has higher time-frequency resolution compared with the standard ST and SSST. Then we apply the DSSST to perform the spectral decomposition on a real field data including gas-filled sand. The results demonstrate that the DSSST can be utilized to well characterize the spectral anomalies related to hydrocarbon reservoir, which renders that this new approach is promising for seismic data analysis.
What problem does this paper attempt to address?