Validating radiation pressure force models for GRACE with SLR

Precise non-gravitational force models are crucial for satellite gravimetry, thermospheric neutral density estimation, and precise orbit estimation (POD). Besides the dominating aerodynamic acceleration acting on low Earth orbit satellites, radiation pressure accelerations impact their motion. The acceleration due to the Earth radiation pressure (ERP) decreases with increasing satellite altitude, whereas the effects of the Solar radiation pressure (SRP) and thermal reradiation pressure (TRP) vary mainly with the satellite’s orientation towards the Sun.

In previous investigations, we extended the existing radiation pressure force models with a focus on a GRACE-like satellite. However, validating the non-gravitational force modelling remained difficult. For GRACE and GRACE-FO, a comparison to measured accelerometer data is possible. However, it is obviously difficult to separate residual effects of the necessary accelerometer calibration procedure from errors in the radiation pressure models.

To overcome these disadvantages, here we perform a validation of modelled non-gravitational accelerations using independent satellite laser ranging (SLR) measurements, which do not require such calibration. This approach is twofold. In a first step, a POD is performed, where kinematic GNSS-orbits and modelled non-gravitational accelerations serve as input together with gravitational state-of-the-art background models. Second, the residuals between the SLR measurements and the derived orbit are computed. Since we assume the SLR observations as the truth, small residuals indicate that the modelled non-gravitational accelerations represent reality well. A comparison against GNSS orbit solutions is not performed here, since kinematic orbits are already part of the first step and the comparison would not be independent.

We choose the year 2008 for our experiments, where the radiation pressure signal is at the same order of magnitude as the aerodynamic signal. Our two-fold approach is performed several times, where we vary the parametrization of the radiation pressure force model, e.g., the consideration of the Earth’s radiation data sets or the satellite’s thermal reradiation. The aerodynamic model is kept fixed during these experiments and a scale factor for this force is co-estimated to account for its mismodelling. Our preliminary results indicate that the smallest residuals occur when considering non-instantaneous TRP acceleration with modified material (diffusion) parameters in addition to our extended ERP and SRP force models. In this case, the annual RMS per pass of the residuals is 2.72cm, which is 2.4cm below the results obtained with standard models and still nearly 4mm less than without modifying the TRP parameters.