Dynamic Modelling of the Hipparcos Attitude

We present a new method for a high-accuracy reconstruction of the attitude for a slowly spinning satellite. This method, referred to as the fully-dynamic approach, explores the possibility to describe the satellite's attitude as that of a rigid body subject to continuous external torques. The method is tried out on the Hipparcos data and is shown to reduce the noise for the along-scan attitude reconstruction for that mission by about a factor two to three. The dynamic modelling is expected to give a more accurate representation of the satellite's attitude than was obtained with a pure mathematical modelling. As such, it decreases the degrees of freedom in the a posteriori reconstruction. Some of the decrease is obtained through accumulating and subsequently implementing information on high frequency components in the solar radiation torques, which show to be systematic and predictable. This could be expected, as they are primarily linked to the external geometry and optical properties of the satellite. In the context of an astrometric mission, the methods presented here can only be applied as a final iteration step: the star positions that are used to reconstruct the attitude are also part of the scientific objectives of the mission. An estimate for the potential of a re-reduction of the Hipparcos data using the fully-dynamic model for the attitude reconstruction was obtained from test reductions of the first 24 months of mission data. Improvement of the accuracies of the astrometric parameters for all stars brighter than Hp=9.0 appears possible. The noise on the astrometric parameters for these stars was affected significantly by the along-scan attitude noise, which dominated for stars brighter than Hp=4.5. The possible improvement for stars brighter than about Hp=4.5 may, after iterations, be as much as a factor three. The reduced noise levels also allow a more accurate calibration and monitoring of instrument parameters, leading potentially to a better understanding of the instrument and the scientific data obtained with it. This revised version was published online in August 2006 with corrections to the Cover Date.