Prestack Stochastic Inversion Based On Geostatistics

The objective of seismic inversion is obtaining the information about the strata configuration and the lithology distribution.There are so many problems limiting the application of the seismic inversion.For example,the inverse problem is usually ill-posed,whose solution procedure is not stable.Solving the inverse problem is usually lack of efficiency,and obtaining the prior model for seismic inversion is hard.To overcome these problems,we conduct research on the pre-stack geostatistical inversion,and the reservoir modeling technology based on multiple-point geostatistics(MPG)and two-point geostatistics(TPG).The forward simulation is made by the Aki-Richards approximation formula.The calculation efficiency and stability are enhanced by the prior information obtained from TPG,and the reservoir lithology model is obtained based on the MPG.The two inversion methods,which belong to linear and non-linear method,respectively,are both based on the geostatistical prior information.The uncertainty of the inversion methods can be expressed by the variance of the inversion results.Many well logs are lack of S-wave velocity,which is crucial for inversion to prepare prior models of S-wave velocity.Therefore,we conduct the prediction of S-wave velocity based on the rock physical model,which depicts the pore form by the consolidation parameter.We can obtain the S-wave velocity by ignoring the influence of pore aspect ratio.The S-wave velocity can be estimated by the consolidation parameter,which can be calculated by the measured P-wave velocities and porosity.We test the inversion methods and S-wave velocity prediction method by theoretical synthetic and actual data.The prior information of lithology distribution is used in the test of S-wave velocity prediction because of the complex lithology.The S-wave velocity prediction shows good results.The test results of the prestack inversion have high resolution and computational efficiency.