Research On Q-value Estimation And Attenuation Compensation For Prestack Data

Due to the viscoelasticity of the subsurface medium,seismic waves experience energy attenuation and waveform distortion during propagation,resulting in narrower frequency bands of seismic data and reduced resolution.The absorption attenuation effect of prestack data is affected by raypath differences,causing distortion of AVA reflection characteristics and resulting in AVA interpretation traps.Attenuation compensation is the primary method to eliminate absorption attenuation’s effect and improve seismic data’s fidelity and resolution.Q value,an important parameter to characterize the attenuation characteristics of the subsurface medium,provides the basis for attenuation compensation processing.However,for prestack data,there are significant challenges to performing accurate and stable Q-value estimation and attenuation compensation due to the influence of substantial raypath differences,high noise level,severe large-angle attenuation,and spectral contamination.To this end,this paper proposes the necessary solutions to the problems associated with the Q-value estimation and attenuation compensation methods to recover the AVA response of pre-stack data accurately and stably and improve the resolution.A series of tests were conducted on the model and actual data to verify the feasibility and practicality of the proposed methods.Accurate Q-values are essential for the attenuation compensation of seismic data.However,the commonly used Q-value estimation methods are often applied to poststack data,which are limited in estimation accuracy,applicability,stability,and band sensitivity due to the influence of frequency-independent amplitude disturbance terms and noise.The paper proposes a Q estimation method based on logarithmic spectral area double difference（LSADD）,which eliminates the influence of frequency-independent amplitude interference terms on the estimation accuracy,effectively improves the noise immunity,and reduces the band sensitivity by using the idea of sliding average.Meanwhile,poststack data ignore the influence of the actual raypath,and distort the original seismic waveform and attenuation characteristics due to the averaging effect of stacking,which will somewhat reduce the accuracy of Q estimation.To address the weakness,this paper uses the radial trace（RT）transform to reduce the influence of raypath differences on the Q estimation,and derives the variation of Q versus apparent velocity（QVAV）to achieve the prestack Q estimation.The QVAV is combined with LSADD theory to improve the estimation accuracy while considering the stability,thus realizing the high-precision Q-value estimation of pre-stack data.The low signal-to-noise ratio of prestack data and the variation of attenuation effects with ray path differences are another bottleneck and difficulty in implementing attenuation compensation.The poststack attenuation compensation techniques are more mature,but applying them to prestack data will result in insufficient compensation for deep and large-angle seismic traces.To address these problems,this paper proposes an attenuation compensation method for prestack data,which considers the difference of raypaths when constructing the forward model and implements the attenuation compensation based on the l₁-norm regularization constraint using the inverse idea.The proposed method maintains both compensation accuracy and noise immunity,achieves amplitude compensation and phase correction,raises the main frequency,broadens the frequency band,improves the resolution,and effectively restores the AVA reflection characteristics of the prestack data.Further,this paper considers the differences between P-and S-wave raypaths and Q values in the converted PS-wave data.It extends the PP-wave prestack compensation method to the converted PS-wave data attenuation compensation processing,which provides accurate data support for the prestack multi-wave joint inversion.In addition,the PS-wave poststack data attenuation compensation requires known P-and S-wave velocities and Q-values.There are certain practical limitations to this problem.This paper uses the idea of equivalent Q-value to equivalent process the P-and the S-wave absorption attenuation effect,which reduces the difficulty of preparing P-and S-wave velocity and Q-value data and makes the compensation method more practicable.Finally,the effectiveness of this paper is verified through the attenuation compensation processing of multiple sets of models and actual data.