Identification of thin gas reservoir in reflection seismic data by synchrosqueezing S-transform in time-frequency representation

Time-frequency seismic attributes extract important information from the seismic signals and exhibit valuable properties and seismic characteristics of the acquired data. However, parameter identification, bandwidth selection, random noise, and the data analysis approach all affects efficiency of these attributes. In the presented study, we analyze capabilities of high-resolution synchrosqueezing S-transform (SST) on nonstationary seismic signals for information extraction. In the subsequent step, the seismic data is presented in a separate domain known as the time-frequency representation (TFR). This potential enables us to identify and investigate thin gas bearing zones, which are not efficiently illustrated by conventional transformation methods. The presented methodology here consists of extracting related information from time-frequency attributes using the SST method. The methodology is initially applied on synthetic data for analyzing high-resolution time-frequency map. Then it is applied on an offshore gas field data. Selected attributes in this strategy is able to delineate the gas accumulation zones and accurately separate two different gas reservoir formations through increasing resolution in both time and frequency directions. Results on identification and separation of gas zones by the SST method were compared with the S-transform method. Comparing the results showed that the SST method provides efficient spectral attributes and better identifies gas accumulation zones.