Least-square mean effect: Application to the Analysis of SLR Time Series

In this paper, we evidence an artifact due to the least square estimation method and, in particular, to the current modeling used to derive station position time series from spacegeodetic measurements. Indeed, to compute such series, we in fact estimate constant (typically over one week) updates of station positions with respect to a priori models (ITRF2000, solid Earth tides, polar tide and oceanic loading effects). Thus, these estimations must underline the physical models which were not taken into account in the a priori modeling (atmospheric and hydrologic loading effects and even unknown signals, in our case).