Systematic errors in the Z-geocenter derived using satellite tracking data: a case study from SPOT-4 DORIS data in 1998

Improving DORIS geocenter time series using an empirical rescaling of solar radiation pressure models

“…However, as our weekly solutions are transformed every week into ITRF2005, using a 7-parameter transformation, if the error were only related to the Z-translation, no difference is to be expected in the vertical results for stations at high latitude as the Z-translation is removed from the solution. In fact, as shown in Willis et al, 2006a, the Z-translation is also correlated with the once per revolution empirical parameters of the orbit estimation and can lead to such artifacts. It is highly possible that a large part of the latitude-dependency of the DORIS coordinate accuracy (Williams and Willis, 2006) may have vanished for these reasons in the new IGN DORIS solution.…”

“…From DORIS observations, while the X and Y components are rather well observed, the Z-component (axis of rotation of the Earth) is much noisier (Meisel et al, 2005;Altamimi et al, 2007) and sometimes affected by large systematic errors (Willis et al, 2006a). This is currently a critical issue for altimetric mission, as any error in Z-geocenter directly maps into the derived mean sea level, as predicted by Willis (2002, 2005), and observed when doing TOPEX/Poseidon data reprocessing with ITRF2005 (Beckley et al, 2007).…”

The International DORIS Service (IDS): Toward maturity

“…Note especially the differences in WRMS when three satellites are available, including SPOT-3, in 1994SPOT-3, in -1996 to the 1993 or 1997-1998 time period when only 2 satellites are available). DORIS/SPOT4 early data were not used by most of the ACs due to a preprocessing error affecting the 1998 data, when phase center corrections are not recomputed (Willis et al, 2006b) The improvement in WRMS from 1999 to 2001 (another period with three satellites: SPOT2, SPOT-4 and TOPEX/Poseidon) shows evidence of the impact of the network management to improve the DORIS station stability (Fagard, 2006).…”

GSFC DORIS contribution to ITRF2008

“…Not all SPOT-4 data cycles in 1998 are affected, as outlined by Willis et al (2006), but we chose to edit all the 1998 data for simplicity. The SPOT-2 data from cycle 083 (1994-02-11 to 1994-02-21) are excluded because the data gave unacceptable and inexplicably large residuals.…”