Seite - 134 - 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

−1 0 1 One active SCAhead Two active SCAheads −0.1 0.0 0.1 0.00.10.20.30.4 PDF of ∆αˆA [unitless] −0.1 0.0 0.1 0.0 0.1 0.2 0.3 0.4 PDF of ∆αˆB [unitless] GRACE-A GRACE-B Figure 9.4: TLS estimate of improvement of satellite orientation for one active SCA head (left column), and two active heads (right column). The estimate for GRACE-A is plotted in green, that for GRACE-B in orange. Note that the scale is different for the roll angle (top row). respective spacecraft, covering 37.8% of the total timespan for GRACE-A and 37.1% of the total timespan for GRACE-B. The right column gives the statistics for the remaining time periods, where observations from both SCA heads are included in the SCA1B product. Note that in fig. 9.4, the correction for roll is given in µrad, while those for pitch and yaw are given in mrad. For periods where both heads are active, shown in the right column of fig. 9.4, there is no noticeable difference in the distribution of the corrections for the two satellites. For all three angles, the correction is on average smaller when both heads are active, which confirms the observations made in fig. 9.3. When only one SCA head is active, the correction in roll is larger for GRACE-A (green) than it is for GRACE-B (orange). This behaviour is reversed for the pitch and yaw angles, where the correction for GRACE-B (orange) is larger. The overall magnitude of the correction is, on average, rather small. For the complete time series corrections larger in magnitude than 1mrad, as observed in fig. 9.3, are uncommon. Chapter9 Co-Estimation of Orientation Parameters134