An improved stochastic inversion of P- to S-wave velocity ratio and compressional modulus based on fast Fourier transform moving average co-simulation

Abstract

The stochastic seismic amplitude-versus-angle (AVA) inversion is extensively utilized in the estimation of elastic properties and rock physics parameters from seismic data. This method can substantially improve the resolution of reservoir identification in areas where well data are available. Compared with conventional sequential simulation methods, the fast Fourier transform moving average (FFT-MA) simulation exhibits a superior computational efficiency. However, given the limitations of FFT-MA simulation in the simultaneous simulation of multiple elastic properties (such as compressional modulus, VP/VS, and density), the effective application of this method in seismic AVA inversion is challenging. In this study, we propose an FFT-MA co-simulation method for the simultaneous simulation of multiple elastic parameters. This approach establishes possible models constrained by the joint probability distributions of inverted properties, which enhances the accuracy of co-simulated results. Furthermore, we introduce an iterative update method designed to enhance convergence during inversion processes and increase the inversion accuracy. Consequently, an efficient stochastic inversion method is developed through the combination of the FFT-MA Co-simulation method with the iterative update method. The effectiveness of the proposed inversion method is validated through numerical examples and field data application.