The Study And Application Of AVO Seismic Parameter Inversion Methods

With the deepening of the degree of prospecting, the exploration activity turns from structural hydrocarbon reservoir to lithologic reservoir. As the most effective exploitation method, seismic exploration is playing a more and more important role. Seismic data contain abundant lithology, physical property and fluid information, which can be acquired by various kinds of seismic inversion methods, and therefore offer more abundant materials for reservoir prediction and characterization, especially in complicated reservoir. AVO inversion is an important approach of extracting lithology parameters which are in hiding in seismic information. The classical method of AVO inversion usually estimates parameters only using two-parameter equation. The location of reservoir can be decided by P-wave and S-wave information. However if we obtain the density information, then it can be conductive to identify the reservoir fluid. According to the relational expression of the elastic parameters, we can obtain other dynamic parameters by three given elastic parameters. Three-term AVO simultaneous inversion has a great future.The three-term linearized AVO equation is typically too ill-conditioned to be reliably solved for. The traditional AVO inversion usually makes inversion stable by solving for two parameters rather than three, thus introducing bias into the estimate. The method based on Bayes theorem can incorporate probabilistic constraints with the likelihood function, the three-term inversion can be constrained so as to give stable estimates, but with minimal bias. The three-term AVO inversion establishes parameter covariance matrix, making the parameter covariance matrix change as a function of depth. With the constraint of the prior geological information, the results of inversion will be in a proper area. So it can develop the results and stability of inversion.The three-term AVO inversion based on the Bayes is improving constantly by the secular study. First, this thesis discusses AVO fundamental theory, then studys the three-term AVO inversion based on the Bayes. On the basis of primary methods, this thesis studies and improves the three-term AVO inversion based on the Bayes. It mainly studies the impact of different prior distribution functions on inversion results so as to find corresponding rules. There are many improvements including the correctness of low-frequency trend, the constraint of the least error, the modification of the parameter covariance matrix and high and low trend fit method. Tests on synthetic and real data show that the reliable parameters estimates in AVO inversion are obtained by implementing a series of theoretical improvements in inversion method. And the inversion results are improved well.The improved method of three-term AVO simultaneous pre-stack inversion based on the Bayes can be used in the inversion of actual seismic data. By the comparison with the inversion results of EI, we can know that the results of these two methods are similar. The improved method of three-term AVO simultaneous pre-stack inversion based on the Bayes obtains good effect in the inversion of actual seismic data. It can obtain the P_impedance, S_ impedance and density which can identify the lithology and fluid medium.