Zusammenfassung
Summary
This contribution presents a procedure that allows an accurate parameter estimation using the Gauss-Markov model when the observations are transformed to an alternative observation space. For that, a solution based on a Gauss-Newton iteration is applied, which gets by with Jacobi matrices. In each iteration step the stochastic model is adapted so that also in the alternative observation space non-normally distributed observations can be correctly evaluated. The contribution shows the practical procedure using two numerical examples and analyses their results.