The Sdudy On The Seismic Wave Energy Compensation Methods Based On The Time-frequency Analysis

When seismic waves propagate through the earth subsurface, the anelasticity andinhomogeneity of the subsurface media will cause the dissipation of seismic waveenergy and the velocity dispersion, which will lead to the reduction of seismic dataresolution. With the exploration precision demand becoming higher and higher, inorder to raise the resolution of seismic data, the compensation of seismic wave energybecomes an indispensable part of the seismic data processing.The explanatory theory of seismic wave energy attenuation can be divided intothree kinds: the first theory includes the Biot theory and the Squirt Flow theory whichconsider the earth as a two-phase medium; the second theory is scattering theorywhich consider the earth as inhomogeneous medium; the third theory is setting upviscoelastic medium models. The paper introduces the influencing factors of seismicwave energy attenuation in underground medium including frequency, pressure,temperature, porosity, lithology and pore fluid saturation.The paper also introducesthe parameters of characterizing the stratigraphic attenuation including the qualityfactor Q, the absorption coefficient,the attenuation coefficient and thelogarithmic decrement.The quality factor Q is used to describe the viscoelastic characteristics of stratummedium, which can quantify the seismic wave energy attenuation and the frequencydispersion caused by the viscoelastic properties of the medium. The thesis introducestwo methods of estimating Q in frequency domain including the spectral ratio methodand the spectral matching method. We also verify the accuracy and limitation of thesemethods using simple models. The paper proposes a new method of estimating Qvalue: the Q estimating method based on the Gabor deconvolution. The Gabor deconvolution is a method of improving the resolution of nonstationary seismic signal.As the influence of Q is added into the Gabor deconvolution, so we can get qualityfactor Q by using the results of Gabor deconvolution. We firstly test the effectivenessof the Gabor deconvolution in models and real seismic data processing, then derivethe formula of calculating Q.In order to restore the deep seismic wave energy and improve the resolution ofseismic files, we must compensate the losing energy that caused by the anelasticityand inhomogeneity of the subsurface media. This thesis introduces the Hale inverse Qfiltering and the regularization inverse Q filtering. Then we propose the attenuationcompensation of seismic energy in time-frequency domain based on the continuouswavelet transform. In this part, we firstly specify the time-frequency analysis methodsthat are commonly used in the seismic data processing including the short time fouriertransform, Gabor transform, S transform, generalized S transform and the time–frequency continuous wavelet transform. We conduct the time-frequency analysis ofchirp signal and synthetic seismic signal using these methods to compare theadvantages and disadvantages of these methods. At last,we choose to compensate theenergy in the time-frequency domain based on the TFCWT,compared with in thewavelet domain, we get better results.