Polar motion and non tidal signals in the superconducting gravimeter observations in brussels

Through the analysis of 131⁄2 years of tidal gravity observations the gravity effect of the polar motion is estimated, that is, the amplitude factor pole δ of the pole tide. This data set is extracted from the complete series of 18 years (from 21.04.1982 till 22.09.2000) tidal data obtained in Brussels (Royal Observatory of Belgium) with the GWR T003 Superconducting Gravimeter (SG), from which a non-reliable initial part has been discarded. Very important precondition of success of this work is that the length of 131⁄2 years allows a good separation of the annual components of the data and the Chandler wobble. The analysis has been performed with the computer program VAV/2002 (Venedikov et al., 2001, 2003) for tidal data processing. Some options have been specially developed for this task. The analysis has taken into account the tidal signal, including a rather efficient determination of the LP (long period) tides, seldom well estimated, as well as the effect of the air-pressure. We build then a model of the non tidal part of the signal often called “drift”, directly orientated towards the estimation of pole δ . We include the theoretical pole tide, estimated from the IERS observations, multiplied by pole δ , a piecewise polynomial representation, an annual components at 1cpy (cycle per year) with its harmonics till 6 cpy and a model of the temperature effect. The result obtained for the polar motion tidal factor is 008 . 0 181 . 1 pole . This is higher value than the value δ=1.158 theoretically predicted for periods near one year by a model including mantle inelasticity. The difference is due to the tidal loading effect of the ocean pole tide. It appeared impossible to get a good estimate of pole δ when the water table is included in the drift model due to a strong interference with the polynomials, approximating the instrumental drift. Nevertheless, the effect of the water table, i.e. the corresponding regression coefficient WT c , has been separately estimated, so to say, as a