Integrated orbit determination and gravity field recovery enhanced by integer ambiguity resolution for low earth orbit satellites

Precise Orbit Determination (POD) of Low Earth Orbit (LEO) satellites is critical for Earth and space science applications. Reduced-Dynamic Orbit Determination (RDOD) achieves centimeter-level accuracy by absorbing unmodeled forces through pseudo-stochastic parameters, but sacrifices physical consistency and limits predictive capability. Dynamic Orbit Determination (DOD) explicitly models temporal force variations, enabling consistent orbits and high-resolution gravity field recovery, yet remains less precise due to unresolved Global Positioning System (GPS) carrier-phase ambiguities. This study proposes an Integer Ambiguity Resolution (IAR)-enhanced DOD framework to overcome this limitation. Using GRACE Follow-On data, the method yields dynamic orbits with sub-2 cm accuracy, and IAR further reduces discrepancies to ∼1 cm, comparable to leading RDOD solutions. Independent validations yield 1 cm residuals from Satellite Laser Ranging and 0.3 cm from K-band Ranging systems. Furthermore, 24-h autonomous forward propagation driven by accelerometer data confirms improved navigation feasibility, with the IAR solution reducing Three-Dimensional (3D) Root-Mean-Square (RMS) errors to ∼63 cm compared with ∼82 cm for the Float solution. In monthly gravity field recovery, IAR remains consistent with Float up to degree and order 20, while higher degrees show increasing discrepancies linked to GPS phase residuals, warranting further investigation of systematic errors.