Enhanced solar radiation pressure modeling for GPS and GLONASS satellites considering radiator impact

Within the IGS (International GNSS Service), precise orbit and clock products of GPS and GLONASS satellites are routinely generated for the purpose of serving scientific and engineering applications. Solar radiation pressure is the dominant non-gravitational perturbation for GNSS satellite orbit determination. Without precise surface models, the empirical CODE orbit models (ECOM, ECOM2) are mostly used. Based on previous studies, an appropriate box-wing model improves satellite orbit significantly if the satellite body is elongated or the attitude law is not yaw-steering. The main goal of this contribution is, therefore, to set up an a priori model and jointly used with the ECOM and ECOM2 models.

GLONASS-M satellite consists of a cylindrical container and a box-like antenna part, and employs specific attitude law in the eclipse. We find that the ECOM model performs better than the ECOM2 model outside the eclipse while it performs two time worse in the eclipse. When applying the conventional box-wing model we observe only very little improvement. Further investigation on the ECOM y-bias estimates demonstrates that there are potential radiators on the -x surface of GLONASS satellites. With this finding, we adjust one additional radiator parameter together with the optical parameters of each satellite as part of precise orbit determination. The adjusted parameters are introduced into a new box-wing model and jointly used with ECOM and ECOM2 models. The resulting GLONASS orbits are evaluated by orbit misclosure between consecutive arcs, accuracy of 24-h predictions and Satellite Laser Ranging (SLR) residuals. Results show that combined with the new box-wing model, both ECOM and ECOM2 models show better performance than without priori model. In particular, the improvement of orbit misclosure and orbit prediction in the eclipse, and the SLR residuals is about 30 %, 50 % and 20 % respectively for the ECOM model. In addition, the spurious pattern of the SLR residuals as a function of the Sun elevation and Sun-satellite argument of latitude is greatly reduced.

The same adjustment of surface optical properties is repeated for GPS satellites, including GPS block IIIA as well. We are expecting to see some improvements in the ERP and Geocenter estimates in the upcoming 3^rd IGS reprocessing when using the new box-wing model.