Magnetic torquer induced disturbing signals within GRACE accelerometer data

• Published in: Advances in space research Vol. 49; no. 9; pp. 1388 - 1394
• Main Authors: Peterseim, Nadja, Flury, Jakob, Schlicht, Anja
• Format: Journal Article
• Language: English
• Published: Oxford Elsevier Ltd 01.05.2012; Elsevier
• Subjects: Accelerometer; Astronomy; Earth, ocean, space; Exact sciences and technology; External geophysics; GRACE; Magnetic Torquers; Accelerometer; Magnetic Torquers; GRACE; Accelerometers; Gravitational force; Space remote sensing; Time series; Satellite observation; Space research; Electric currents; Spectral analysis; Radiation pressure; Drag; Disturbances; Models; Solar radiation; Gravity
• DOI: 10.1016/j.asr.2012.02.013
• ISSN: 0273-1177

The GRACE (Gravity Recovery And Climate Experiment) gravity field satellite mission was launched in 2002. Although many investigations have been carried out, not all disturbances and perturbations upon satellite instruments and sensors are resolved yet. In this work the issue of acceleration disturbances onboard of GRACE due to magnetic torquers is investigated and discussed. Each of the GRACE satellites is equipped with a three-axes capacitive accelerometer to measure non-gravitational forces acting on the spacecraft. We used 10Hz Level 1a raw accelerometer data in order to determine the impact of electric current changes on the accelerometer. After reducing signals which are induced by highly dominating processes in the low frequency range, such as thermospheric drag and solar radiation pressure, which can easily be done by applying a high-pass filter, disturbing signals from onboard instruments such as thruster firing events or heater switch events need to be removed from the previously filtered data. Afterwards the spikes which are induced by the torquers can be very well observed. Spikes vary in amplitude with respect to an increasing or decreasing current used for magnetic torquers, and can be as large as 20nm/s2. Furthermore, we were able to set up a model for the spikes of each scenario with which we were able to compute model spike time series. With these time series the spikes can successfully be removed from the 10Hz raw accelerometer data. Spectral analysis of the time series reveal that an influence onto gravity field determination due to these effects is very unlikely, but can theoretically not be excluded.