ERS ORBIT DETERMINATION AND GRAVITY FIELD MODEL TAILORING : RECENT DEVELOPMENTS

In the last ten years significant advances have been made in gravity field modelling, culminating in general-purpose models like JGM-3, TEG-3 and EGM96. Significant defects remain in the quality of the models when applied to orbit determination of some altimeter satellites, leading to large orbit errors, characterised by the geographically-correlated nature of the altimeter crossover height difference. However, we have used these data effectively in the adjustment of EGM96 to a model tailored to ERS-1 and ERS-2. Many gravity field coefficients are “invisible” to altimetry, but still cause radial orbit errors through misfits with the basic tracking data: SLR and PRARE. We have incorporated these observations in the tailoring process, and the adjustment of gravity field coefficients is carried out up to degree and order 70. This led to significant improvements to the EGM96 model, reducing the radial orbit error from 7 to 4.5 cm and cutting the tracking data residuals by 3 cm in root-sum-square sense. During the development of the latest tailored models, we have analysed altimeter crossovers over the ice-covered Arctic, providing valuable information on the behaviour of gravity field induced orbit errors over this remote area. Including PRARE and crossover data in the ERS-2 orbit computation results in significantly more precise and stable orbits. The estimated resulting radial orbit accuracy is below 5 cm, and seems to approach the TOPEX/Poseidon orbit accuracy level. The improved orbits result also in reductions in dynamic sea surface topography mismodelling.