Non-stationary Seismic Resolution Enhancement Using Gabor Domain Spectral Decomposition

Importance of recording the full range of frequencies is widely accepted for high-resolution imaging. Spectral decomposition is the novel technology which analyses the seismic elements in time-frequency domain. Non stationary spectral analysis of seismic components aims to improve the seismic data quality and behaviour of wavelet for stratigraphy interpretation and reservoir characterization. In this study we use matching pursuit algorithm to find optimum scale and time-delay for decomposition of the data in time-frequency domain. Fine localization of spectral components in time-frequency domain shows the exponential energy lost in media caused by the geological property of layers. We do inverse attenuation filtering using extracted attenuation function in Gabor domain to compensate energy lost in media. Improved seismic data after decomposition makes the reservoir area clear for figuring out role of chimneys for hydrocarbon migration.