AVF attenuation inversion method for viscoelastic media based on bayesian theory
In this paper,a viscoelastic medium inversion method suitable for underground geological model and parameter estimation is established by introducing Bayesian inversion framework,integrating AVF analysis into the inversion mechanism,using geological logging seismic interpretation data as prior information,constructing the likelihood function from viscoelastic parameters to seismic data,and giving the inversion results in the form of a posterior probability Firstly,by simplifying Aki-Richards formula and introducing two new attenuation parameters Qp and Qs in viscoelastic medium,the approximate expres-sion of AVF in viscoelastic medium is derived.Then,starting from the linearized AVF formula in a viscoelastic medium,based on the framework of Bayesian inversion,the prior information in the initial model and the cross-correlation information of the logging curve are added to the data inversion.The posterior probability model is obtained through numerical experiments of theoretical models and practical applications,which show that this method can directly invert the attenuation of vertical and horizontal waves of underground media.The attenuation anomalies can indicate the possible gas-bearing reservoirs,thus provi-ding more information for reservoir prediction and reducing the risk of prediction.