Parameter identification of a leakage aquifer system in Daqing region of China using model of coupled FEM and Kalman filter.

Groundwater modelling is made difficult by the large uncertainties on the different modelling components: structural errors related to the model parameterization, parameter uncertainty, and errors in the forcing terms. This paper proposes an approach to parameter identification of the leakage aquifer using the model of coupled finite element method and Kalman filter. The method considers the uncertainty of groundwater flow system. Therefore, the stochastic and deterministic parameters are identified in numerical simulation processes. Statistical tests, the Student's test and Chi-square test are applied to examining the mean and deviance values between the observed and simulated heads at specified confidence interval. A case study for parameter identification of a leakage aquifer system in Daqing region of China shows that the reliability of parameter estimation reaches 90%. Copyright © 2005 John Wiley & Sons, Ltd. [ABSTRACT FROM AUTHOR]