External Calibration of SGG Observations on Accelerometer Level

The satellite mission GOCE (Gravity Field and steady state Ocean Circulation Explorer) uses for the first time a gradiometer for the determination of the earth's global gravity field. The observations suffer stochastic and systematic errors. These errors are caused by rotations and shifts of the accelerometers with respect to their optimal alignments and positions. The main target of this study is to develop methods to deduce the inverse calibration matrix (ICM), which parameterizes these errors, by using least squares adjustment. Beside signal filtering, weighting and rigorous noise analysis, a complete variance-covariance propagation chain is applied. For being able to comply with the given tasks, a simulation environment has been developed. With this environment it is possible to apply data synthesis, data analysis, filtering, calibration and variance-covariance propagation. In the frame of three external calibration approaches, methods for improving the calibration accuracy have been developed and implemented. For an as realistic as possible simulation study, the synthetic data has been defined according to the GOCE project error budget (Cesare et al., 2008). The treatment of accelerometer measurement noise, angular acceleration noise and angular velocity noise is of great importance. By using an adequate band-pass filter, the overall noise level could be strongly reduced. Moreover, variance-covariance information corresponding to the stochastic noise models has been synthesized and introduced into the calibration system. By utilizing filtering and variance-covariance information, the maximal calibration error of the alignment parameters could be reduced to <10^-3. Finally harmonic gravity field coefficients have been estimated from externally calibrated and uncalibrated data. It could be proven that an overall improvement of the signal quality could be achieved. Nevertheless, because of the maximal quality of only 10^-3 of the deduced alignment parameters, larger deviations from the reference gravity field than in the ideal calibration case have to be expected.