Inverse Process of Multichannel Analysis of Surface Wave by Using Ensemble Kalman Filter

The Young's modulus is a common and critical soil property. It is difficult to infer the spatial distribution of Young's modulus on a large ground, owing to the uncertainty. In this study, a method using ensemble Kalman filter (EnKF) to solve the inverse problem in the multichannel analysis of surface wave and estimate the spatial distribution of Young's modulus which includes quantified uncertainty is presented. The statistic model derived from other investigation data could be integrated into the inverse process to increase the accuracy of estimate by using sequential Gaussian simulation (sGs) to generate a reasonable initial ensemble. The practical effectiveness of this framework is verified by numerical experiments using synthetic and realistic data from an earth-fill dam. The impact of the initial ensemble and the two update schemes (stochastic and deterministic updates) is discussed through the experiments as well.

Keywords: Data assimilation, Sequential Gaussian simulation, Surface wave method, Ensemble Kalman filter, Inverse problem.