Seite - 133 - in Contributions to GRACE Gravity Field Recovery - Improvements in Dynamic Orbit Integration, Stochastic Modelling of the Antenna Offset Correction, and Co-Estimation of Satellite Orientations

−3 0 3 6 GRACE-A GRACE-B −2 −1 0 1 2 05:00 06:00 07:00 08:00 −4 −2 0 2 4 05:00 06:00 07:00 08:00 Sensor Fusion TLS Estimate Difference Figure 9.3: TLS estimate of satellite orientation (in pink), orientation from sensor fusion (in brown), and difference (in blue), for 3 hours on June 11, 2010. The roll, pitch, and yaw angles are given w.r.t. the nominal attitude. Darker shaded areas show times where only one star camera head was active on the respective spacecraft. Note the different scales for each angle. diverges. This is striking, as it suggests that the noise floor of the KBR instrument in the frequency band affecting the satellite orientation is located exactly between the noise levels achievable with only one or with two SCA heads. The deviation of the TLS estimate from the a priori orientation is still mostly small, though. For the whole month of June 2010, the largest corrections observed are 0.7mrad for GRACE-A and 1.9mrad for GRACE-B. In general, it can be observed that the corrections are larger for GRACE-B. Figure 9.4 gives the distributions of the estimated improvements for the satellite orientations ∆αˆs for the complete processed GRACE time series. The left column is limited to a statistic of the periods where only one SCA head was active on the 9.3 Results 133