Spectral Modeling Deconvolution Based on Improved S Transform and its use in Hydrocarbon Detection

In this paper, we normalized energy for window function of S transform to improve the accuracy of frequency resolution in time-frequency spectrum. Combining improved S transform and spectral modeling method, the spectral modeling deconvolution in time-frequency domain can increase the resolution of seismic section for the following fine reservoir description. Through high resolution seismic data and high resolution time-frequency analysis method, absorption characteristic variation of spectrum can be calculated, which can be introduced to hydrocarbon detection. Both the low-frequency and high-frequency energy information laid a good foundation for oil or gas-bearing reservoir analysis. This hydrocarbon detection method can distinguish strong amplitude abnormal of lithology and abnormal absorb of gas-bearing reservoir, which can reduce the problem of multiple-results when reservoir prediction.