Improving seismic resolution is essential for obtaining more detailed structural and stratigraphic information. Seismic wave propagating in the non-complete elastic underground medium will occur the absorption attenuation, which is related to the frequency and the travel time. So the absorption attenuation influences the resolution of seismic data as well as showing the nonstationary characteristics. In view of this, this paper proposes the nonstationary compensation method of the absorption attenuation of the seismic wave. Firstly, we use the matching pursuit algorithm to decompose the seismic signal into different matching wavelets; then, make the product of the travel time and central frequency be the horizontal coordinate and the corresponding amplitude be the vertical coordinate of every matching wavelet putting them in the coordinate plane, the amplitude attenuation function that is changing with the product is established in the coordinate plane, so amplitudes of the matching wavelets can be compensated through the amplitude attenuation function, respectively; finally, the seismic signal is reconstructed with the matching wavelets after compensation. The analysis of the synthetic seismic signal and the real field data shows that the deep energy of seismogram after nonstationary compensation is effectively restored. What’s more, the resolution of the seismic data in deep and middle is improved greatly.


Duing to the viscoelasticity of the underground medium, the obvious absorption attenuation phenomenon will appear in seismic wave and this leads to the amplitude attenuation and waveform distortion. Besides, it will also reduce the resolution of the seismic data, especially in the middle and deep. With the increase of the difficulty of exploration, exploration targets become more and more deep. So people begin to pay more and more attention to the problems of resolution damage duing to the absorption attenuation. Until now, many studies have been done about the absorption compensation of seismic wave. The representative one is inverse Q filtering method (Wang Y, 2002, 2006). However, at present, most of the inverse Q filtering methods need to extract the quality factor of Q firstly. Gladwin et al (1974) proposed the rising time principle method. Jannsen et al (1985) proposed wavelet simulation method. In 2011, Wang X et al proposed the spectral ratio method to calculate the Q value based on S transform. Although Q absorption compensation method has a great advantage, it will amplify the high frequency noise because that this method may improve the energy compensation with the increase of the frequency. What’s more, because that the quality factor can’t be accurately estimated, this will lead to under compensation or over compensation of seismic data. Liu X et al (2006) proposed absorption compensation method based on S transform. Zhang F et al. (2015) proposed the improved Stockwell transform method to extract layer absorption properties. Based on S transform, He B et al (2012) proposed a method which extracts layer relative absorption coefficient. All of the above methods combine S transform with the study of seismic wave attenuation and don't need to calculate the Q value. Margrave et al (1998,2011) proposed Gabor deconvolution whose main idea is to integrate the nonstationary effects into convolutional model and divide time-frequency spectrum into hyperbolic stripes to deal with the signal. The aim of the final deconvolutional processing is to improve the resolution of the seismic data.

This content is only available via PDF.
You can access this article if you purchase or spend a download.