***************************************************************************= *** dorisreport 06-Jul-2021 17:38:33 Message No 5337 ***************************************************************************= *** Author: "Lemoine, Frank G. (GSFC-61A0)" Subject: Technical Description of gscwd51 solution (1992-DOY0285 (Oct. 11, Technical Description of gscwd51 solution: June 30, 2020 -------------------------------------------------------- F.G. Lemoine (Frank.G.Lemoine@nasa.gov) D.S. Chinn (Douglas.S.Chinn@nasa.gov) N.P. Zelensky (nzelensk@umd.edu) This series represents a complete reprocessing of all DORIS data from November 1992 to December 2020, and represents an update of the gscwd40 SINEX series, which was submitted for inclusion in the IDS Combination developed for ITRF2020. Based on the feeback provided by the IDS Combination Center, we have made several changes to the gscwd40 solution. In addition, we have retuned the Cr's (solar radiation reflectivity coefficients) for SPOT-2, and SPOT-5. This delivery includes 1473 weekly SINEX files (Earth Orientation Parmeters and Station Coordinates) from 1992-DOY0285 (Oct. 11, 1992) to 2020-DOY0362 (Dec. 27, 2020) The following satellites were processed: SPOT-2, SPOT-3, SPOT-4, SPOT-5, TOPEX/Poseidon, Jason-1, Jason-2, Jason-3, Envisat, HY-2A, Cryosat-2, Saral, Sentinel-3A. This series builds on gscwd40 series and introduces the following changes: those items marked by an asterisk (*) were recommended by the IDS Combination Center, based on an analysis of the single-satellite SINEX solutions for the satellites contributing to the IDS analysis center solutions from 2002 onwards. 1.(*) The radial offset on SPOT-2 is adjusted after 2007-DOY143. 2.(*) The radial offset on Jason-1 is adjusted over its entire data span. 3.(*) We add ENVISAT only after 2004-11-04, based on the performance of the single-satellite SINEX solutions. 4.(*) SPOT-4 is eliminated SPOT-4 after 2013-01-11, based on the poor performance of the satellite in the single-satellite solutions after that date. 5. In the gscwd40 series, we had forward modelled the geocenter with the annual model of Ries (2013). However this is inconsistent with the other ACs who had not forward modelled this effect. Thus, we eliminated this forward modelling of geocenter in wd51. 6. Compared to gscwd40, we fixed the data weighting on Jason-3. In gscwd40, the SAA stations were not downweighted by 3X in the POD. In addition, based on tests by Nikita Zelensky, we reverted to using the a priori value of the specular reflectivity (0.06) for the Jason-3 solar arrays, instead of a tuned value. This very slightly reduced the OPR amplitudes, and improved the SLR fit for SLR+DORIS orbits (even in the presence of empirical accelerations). Note that a Cr/arc is still adjusted in a separate POD run and then held fixed in the final solution. 7A. For SPOT-2 we adopt the following Cr's: 1992-10-16 to 1994-12-25: 0.9939847400000D+00 1995-01-01 to 1995-12-31: 0.9891268100000D+00 1996-01-07 to 1999-07-12: 0.9854453200000D+00 7B. For SPOT-5 we adopt the following Cr's: 2002-06-16 to 2003-12-31: 0.9978032200000D+00 2004-01-04 to 2005-12-25: 0.9910192000000D+00 2006-01-01 to 2007-12-30: 0.9895276700000D+00 2008-01-06 to 2008-12-28: 0.9902049100000D+00 2009-01-04 to 2015-11-15: 0.9901176100000D+00 The pitch of the solar is adjusted as specified in the IDS documentation after January 2008. We corrected a glitch that inroduced the incorrect pitch angle for the solar array orientation in several SPOT-5 arcs, including one in January 2008. 8. The Sentinel-3A data are included starting on 2016-03-02. We note that for our GSFC SLR+DORIS Sentinel-3A orbits, we routinely fit the SLR data to between 6-9 mm RMS. We find that the avg. RMS radial, cross-track, and along-track orbit differences between the DORIS-only and the SLR+DORIS orbits for Sentinel-3A are: 0.158 cm, 1.194 cm, 0.723 cm, respectively. For comparison, we find the following average RMS orbit differences between SLR+DORIS and DORIS-only orbits for Jason-2 and Saral to be: Satellite Radial, Cross-track, Along-track Jason-2 0.397 cm 1.724 cm 1.453 cm Saral 0.271 cm 1.168 cm 1.132 cm Sentinel-3A 0.158 cm 1.194 cm 0.723 cm The gscwd51 solution includes the following attributes which were also part of the gscwd40 solution. I. The SPOT-5 radial offset is adjusted weekly over the duration of the SPOT-5 mission. II. We used 3.5day arcs from 1999 to 2003, instead of 7.0 day arcs in the previous series, for SPOT-2, SPOT-4, SPOT-5, and Envisat. In addition, when the solar flux was very high (> 180), we adjusted empirical accelerations for the along and cross track components every 6 hrs and with a correlation time constraint of 6 hrs as well. (This was tested in the series gscwd39). III. The elevation cutoff is 7.0 deg. IV. We apply an elevation-dependent weighting function on the DORIS data: 1/sqrt(sin(elevation)) V. Since the weighting function on average downweights the DORIS data, we have changed the data sigma used in GEODYN from 2 mm/s to 1.25 mm/s for data after 2003. For data prior to 2003, which already had a higher RMS of fit (cf. Fagard, 2006, J. Geodesy, Lemoine et al., 2010, Adv Space Res), we have not changed the GEODYN-applied data sigma. The data sigma in combination with the weighting factor will propagate through to the formal standard deviations for the estimated parameters. Please do not reply directly to this message, but send comments and suggestions to IDS.central.bureau@cls.fr