Analysis of HY2A precise orbit determination using DORIS

HY2A is the first Chinese marine dynamic environment satellite. The payloads include a radar altimeter to measure the sea surface height in combination with a high precision orbit to be determined from tracking data. Onboard satellite tracking includes GPS, SLR, and the DORIS DGXX receiver which delivers phase and pseudo-range measurements. CNES releases raw phase and pseudo-range measurements with RINEX DORIS 3.0 format and pre-processed Doppler range-rate with DORIS 2.2 data format. However, the VMSI software package developed by Van Martin Systems, Inc which is used to estimate HY2A DORIS orbits can only process Doppler range-rate but not the DORIS phase data which are available with much shorter latency. We have proposed a method of constructing the phase increment data, which are similar to range-rate data, from RINEX DORIS 3.0 phase data. We compute the HY2A orbits from June, 2013 to August, 2013 using the POD strategy described in this paper based on DORIS 2.2 range-rate data and our reconstructed phase increment data. The estimated orbits are evaluated by comparing with the CNES precise orbits and SLR residuals. Our DORIS-only orbits agree with the precise GPS + SLR + DORIS CNES orbits radially at 1-cm and about 3-cm in the other two directions. SLR test with the 50 degrees cutoff elevation shows that the CNES orbit can achieve about 1.1-cm accuracy in radial direction and our DORIS-only POD solutions are slightly worse. In addition, other HY2A DORIS POD concerns are discussed in this paper. Firstly, we discuss the frequency offset values provided with the RINEX data and find that orbit accuracy for the case when the frequency offset is applied is worse than when it is not applied. Secondly, HY2A DORIS antenna z-offsets are estimated using two kinds of measurements from June, 2013 to August, 2013. The results show that the measurement errors contribute a total of about 2-cm difference of estimated z-offset. Finally, we estimate HY2A orbits selecting 3 days with severe geomagnetic storm activity and SLR residuals suggest that estimating a drag coefficient every 6 h without any constraint is sufficient for maintaining orbit accuracy. (C) 2014 COSPAR. Published by Elsevier Ltd. All rights reserved.