Abstract: To address the challenges of precise orbit determination (POD) for low Earth orbit (LEO) satellites under various ionospheric conditions, this study proposes an uncombined GNSS data processing method that incorporates spatiotemporal ionospheric constraints. Using onboard GNSS data from the Sentinel-6A satellite, we systematically compared the orbit determination performance of the ionosphere-free (IF) combination and four uncombined strategies with different ionospheric constraints during both geomagnetically quiet and storm periods. Results indicate that during quiet periods, all uncombined strategies outperformed the IF combination in terms of orbit accuracy. With limited observations, the strategy applying priori corrections and temporal ionospheric constraints performed best. Under sufficient observations, the strategy incorporating both priori corrections and spatiotemporal constraints achieved optimal performance, yielding a three-dimensional root mean square (3D RMS) error of 5.11cm for the combined GPS/Galileo solution. During storm periods, all strategies experienced accuracy degradation, while the IF combination remained more stable, with only a 3% average reduction in accuracy and a 3D RMS of 5.64cm for the GPS/Galileo solution. In terms of convergence time, the four uncombined strategies performed similarly and were all faster than the IF combination. Among them, the strategy incorporating priori corrections and spatiotemporal constraints converged in about 215s during quiet periods, increasing to approximately 286 s during storm conditions. These findings provide valuable insights for optimizing POD strategies for LEO satellites under varying ionospheric conditions.

Key words: ionospheric delay, keyword2, geomagnetic storm, low earth orbit satellites, precise orbit determination