Satellite Precise Orbit Determination Using Real-Time and Near Real-Time GPS Products

Eva Jalabert (CNES, France)


Denis Laurichesse (CNES, France); Sabine Houry (CNES, France); Alexandre Couhert (CNES, France)

Event: 2014 Ocean Surface Topography Science Team Meeting

Session: Precision Orbit Determination

Presentation type: Type Poster

The Precision Orbit Determination group at CNES currently produces Precise Orbit Ephemerides (POE's) and rapid Medium precision Orbit Ephemerides (MOE's). The first one is a reference orbit, computed within a month with a radial accuracy of 1 cm or less. The second one is computed at one day latency and has radial orbit errors of about 2 cm.
At the last OSTST meeting, altimeter analysts suggested POD groups should think of the possibility of improving the accuracy of short latency orbit products (IGDR), making use of all the available tracking data and models. This would indeed enable to delay the delivery of the current GDR orbit products, making them more suitable for long-term analysis. Until now, only DORIS measurements were used to compute the MOE's. Using GPS measurements could improve the quality of the orbit but standard GPS products such as ultra-rapid IGS products do not suit, because their latency is too high. However, two recent products can suit the need for low latency GPS products: SGU (Near Real Time IGN constellation) and RTS (IGS Real Time Service).
The goal of this study is to assess the quality of Jason-2 POD using SGU and RTS products. The set of data that have been used is presented, along with its quality and availability. Then, over a long time span (6 months), MOE orbits are computed on a daily basis, using GPS measurements. Traditionally, on such orbits, several criteria are used to compare the quality of an orbit to the POD final orbit: radial component of the orbit comparison, laser residuals and crossover points. This study shows that the orbits obtained using RTS and SGU are of acceptable quality (comparable to other low-latency products for RTS, and slightly better for SGU) on each of these criteria, although of course not as good as the final POD orbit that is computed a month after with updated and stabilized input data.
Eva Jalabert