Stochastic modelling in attitude determination of GRACE-FO satellites for ITSG gravity field solutions

Abstract

The GRACE/GRACE-FO mission has been monitoring monthly variations of the Earth’s gravity field for more than two decades. One source of error in the gravity field estimation process is the uncertainty of the satellites’ attitude. The errors of the attitude determination propagate to the inter-satellite range observations via the antenna offset centre (AOC) correction.Compared to GRACE, GRACE-FO carries an additional third star camera, which, together with the gyroscopes, usually makes attitude errors negligible. However, the current coarse pointing mode, employed to increase mission lifetime, enlarges the propagated attitude errors.As a next step, we aim to incorporate full covariances for the AOC correction into the stochastic model. For this purpose, this work conducts in-house level-1A attitude data processing, including sensor fusion of the star camera and gyroscope data, along with variance component estimation. The stochastic model of the star camera data is extended to account for spatial dependencies across the celestial sky, as well as for temporal correlations. The impact on the post-fit residuals and the quality of the recovered gravity field is analysed as part of the ITSG gravity field processing at the Institute of Geodesy at Graz University of Technology.