Abstract's details
Hybrid DORIS+GPS measurements processing using the REGINA and DORIS networks and Sentinel-3A : Ground beacon and onboard clock corrections
Event: 2020 Ocean Surface Topography Science Team Meeting (virtual)
Session: Precision Orbit Determination
Presentation type: Forum only
The required precision for the orbits of the altimeter missions depends on the accuracy of dynamic and measurement modelling. That is the reason why empirical accelerations and corrections are estimated to mitigate those mismodelling errors, and thus it is crucial to increase the observability or/and the accuracy of these corrections. This study mainly focuses on the clock corrections used to model the USO for the time tagging of the DORIS measurements.
The DORIS-derived clock corrections are currently estimated in a two-step process. In the time-tagging step based on the pseudo-range measurements, the USO behavior of each time beacon is modeled by a linear expression derived from the coefficients provided in the header of the DORIS/RINEX, whose offset is adjusted. On the satellite side, the clock correction is a degree-3 polynomials. Additionally, while determining the orbit solution from the Doppler measurements, a frequency offset (df/f) is determined per pass for all stations.
We propose in this analysis to use GPS-derived clock modelling for both ground and satellite sides. Indeed, for Sentinel-3A and B, the DORIS and GPS receivers are connected to the same USO. On the ground side, some stations of the GNSS REGINA (REseau GNSS pour l'IGS et la Navigation – global GNSS station network operated by CNES and IGN) and DORIS networks are in a similar configuration where the output of the DORIS beacon’s USO is used by the GNSS receiver. Three main points have been investigated: the effect of the estimation of the df/f parameters on the orbit determination, the accuracy of the modelling of the DORIS measurements with the GPS-derived corrections, and the possibility to use them to substitute the DORIS-derived corrections, especially the estimation of df/f.
The first results on Sentinel-3A show that the GPS-derived corrections perform equally well to model the DORIS measurements. In addition, despite the limited number of REGINA/DORIS stations with respect to the whole network, they manage to replace the estimation of df/f by passes, and lead to a locally improvement in the error observation and the orbit performance in the along-track direction. The expected improvement in observability with more DORIS/REGINA coupled stations could reach up to 50 % in the along-track direction, and remove the residual signature due to clock effects on the orbit.
The DORIS-derived clock corrections are currently estimated in a two-step process. In the time-tagging step based on the pseudo-range measurements, the USO behavior of each time beacon is modeled by a linear expression derived from the coefficients provided in the header of the DORIS/RINEX, whose offset is adjusted. On the satellite side, the clock correction is a degree-3 polynomials. Additionally, while determining the orbit solution from the Doppler measurements, a frequency offset (df/f) is determined per pass for all stations.
We propose in this analysis to use GPS-derived clock modelling for both ground and satellite sides. Indeed, for Sentinel-3A and B, the DORIS and GPS receivers are connected to the same USO. On the ground side, some stations of the GNSS REGINA (REseau GNSS pour l'IGS et la Navigation – global GNSS station network operated by CNES and IGN) and DORIS networks are in a similar configuration where the output of the DORIS beacon’s USO is used by the GNSS receiver. Three main points have been investigated: the effect of the estimation of the df/f parameters on the orbit determination, the accuracy of the modelling of the DORIS measurements with the GPS-derived corrections, and the possibility to use them to substitute the DORIS-derived corrections, especially the estimation of df/f.
The first results on Sentinel-3A show that the GPS-derived corrections perform equally well to model the DORIS measurements. In addition, despite the limited number of REGINA/DORIS stations with respect to the whole network, they manage to replace the estimation of df/f by passes, and lead to a locally improvement in the error observation and the orbit performance in the along-track direction. The expected improvement in observability with more DORIS/REGINA coupled stations could reach up to 50 % in the along-track direction, and remove the residual signature due to clock effects on the orbit.
Contribution: Presentation_v2.pptx (pptx, 2386 ko)
Back to the list of abstract