Recursive adjustment approach for the estimation of physical earth parameters from polar motion

The connection between highly precise time series of Earth orientation parameters (EOP) and geophysical processes in the Earth system can be studied on the basis of analytical or numerical forward models. Such models are dependent on a variety of parameters describing geometrical, physical or rheological properties of the Earth. A sensitivity analysis showed that some weakly determined Earth parameters (e.g. Love numbers) have a large effect on the forward model results. We aim at the improvement of such parameters on the basis of observed EOP. In order to make use of the long EOP time series that cover several decades, a recursive adjustment procedure is being developed. We present the principle of the approach and compare it to a least-squares adjustment in the Gauss–Helmert model. It is shown that both approaches are comparable with respect to the results, but the recursive procedure is superior in terms of computational efficiency. Our study focuses on the pole tide Love number k₂, a specifically critical model parameter that is directly related to period and damping of the modelled Chandler oscillation. In order to simplify the case, the algorithm is developed and tested for an example of a two-dimensional spring mass damper system in which the simulated damped oscillation is equivalent to the Chandler oscillation.