10.4122/1.1000000227
Konecny, Franz
Franz
Konecny
franz.konecny@boku.ac.at
Konecny, Franz
Franz
Konecny
franz.konecny@boku.ac.at
On the Kalman smoother approach to the stochastic inverse problem of groundwater hydrology
XVI International Conference on Computational Methods in Water Resources
2006
2006
Key words: Inverse problems, distributed parameter systems, infinite dimensional
Kalman smoother, EM algorihm.
In this presentation we are concerned with the estimation of the hydraulic
conductivity profile on the basis of noisy observations of hydraulic heads. An
augmented Kalman smoother is applied to the stochastic flow equation. The
implementation of the smoother requires the solution of a Riccati type partial
differential equation for the smoother covariance and a stochastic partial
differential equation to obtain the optimal estimate of the augmented state. We
discuss finite dimensional approximations of the smoothing problem and propose an
EM algorithm for the estimation of the conductivity parameters.