Assessment of the orbit-related sea level error budget for the TOPEX/Poseidon altimetry mission

Transcription

1 Assessment of the orbit-related sea level error budget for the TOPEX/Poseidon altimetry mission Sergei Rudenko (1,2), Saskia Esselborn (1), Tilo Schöne (1) (1) GFZ German Research Centre for Geosciences, Potsdam, 14473, Germany (2) German Geodetic Research Institute of the Technical University of Munich, Munich, 80333, Germany Sea Level CCI Phase II Annual Review #3 and Final Meeting, February 27-28, 2017, ESA ESRIN

2 Objectives Assessment of the radial orbit error for the TOPEX/Poseidon mission ( ) Upper bound orbit errors from gridded radial orbit differences for global mean and regional sea level We use three state-of-the-art orbits derived in ITRF2008 and four test orbits Reference: GFZ (CCI19) = REF External orbits: GSFC std1504 and GRGS Impact of the tracking data: SLR-only (CCI22), DORIS-only (CCI23) Impact of the terrestrial reference frame: ITRF2014 (CCI24, ITRF14 orbit) Impact of Earth's time variable gravity field model: EIGEN-6S2 (CCI25, Geoid orbit) Time scales: seasonal = annual ( days), interannual (5 years), decadal (here: 11 years)

3 The main models used for the GFZ CCI19, GSFC std1504 and GRGS orbits Parameter GFZ CCI19 (REF) GSFC std1504 GRGS Terrestrial reference frame Static Earth s gravity field model Time-variable Earth s gravity field model Solid Earth tide model ITRF2008, SLRF2008, DPOD2008 ITRF2008, SLRF2008, DPOD2008 ITRF2008, SLRF2008, DPOD2008 EIGEN-6S4 GOCO2S (> n=m=5) EIGEN-6S2 EIGEN-6S4 up to n=m=80 IERS Conventions (2010) Updated harmonic piece-wise fit to 5x5 weekly solutions IERS Conventions (2003) EIGEN-6S2 up to n=m=50 Ocean tide model EOT11a GOT4.10 FES2012 IERS Conventions (2010) Non-tidal atmospheric and oceanic gravity GFZ AOD1B RL05 based on ECMWF 6-h fields up to n=m=100 ECMWF 6-h fields up to n=m=50 3-h ERA-interim/ECMWF up to n=m=50 / TUGO R12 up to n=m=50 Non-tidal atmospheric loading effect on stations Based on ECMWF ERA-Interim data none none Ocean loading effect on stations Annual geocenter motion on stations FES2004 GOT4.10 FES2012 Not explicitly modelled Ries (2013) none

4 TOPEX/Poseidon orbit accuracy SLR observations RMS fits DORIS observations RMS fits => The smallest SLR RMS fits (1.59 cm), among GFZ orbits, are obtained for the orbit derived using SLR-only observations. Weights: SLR 3 cm, DORIS: 0.05 cm/s. For SLR+DORIS GFZ orbits, the smallest SLR RMS fits (1.959 cm) are obtained for the Geoid (EIGEN-6S2) orbit, followed by the reference (EIGEN-6S4) orbit (1.963 cm) and ITRF2014 orbit (1.969 cm). => Minor impact on the DORIS observation residuals

5 TOPEX/Poseidon orbit accuracy Radial arc 2-day overlaps (SLR, DORIS, SLR+DORIS orbits) Radial arc 2-day overlaps (EIGEN-6S4 + ITRF2008, EIGEN-6S2 + ITRF2008, EIGEN-6S4 + ITRF2014) Left: the smallest radial arc overlaps (0.88 cm) have been obtained for the DORIS orbit, followed by the SLR+DORIS orbit (0.90 cm). Right: EIGEN-6S2 performs better than EIGEN-6S4 for TOPEX radial overlaps; ITRF2014 brings an improvement, as compared to ITRF2008.

6 Global mean RMS of gridded radial orbit differences per cycle over the oceans for different orbit configurations (REF test) The smallest changes (~2 mm) are related to the ITRF14 orbit. The RMS time series for Geoid, GSFC and GRGS orbits exhibit a seasonal cycle. The GSFC, GRGS, SLR and DORIS orbits are dominated by high frequency variability.

7 5-year running trends (ascending, descending, merged) for the global mean radial orbit differences over the oceans The decadal trend of the SLR-only, DORISonly and GRGS orbits has an opposite sign, when using only ascending or descending tracks. The interannual variability is increased by ~5 times, when analysing the ascending and descending tracks separately.

8 Global mean errors over the oceans related to the POD configuration (ascending, descending) SLR DORIS ITRF14 GEOID GSFC GRGS RMS [mm] Interannual trend [mm/year] Decadal trend [mm/year] 0.04 (0.3,0.3) 0.01 (-0.07,0.07) 0.10 (0.5, 0.4) 0.05 (0.2,-0.3) (0.4,0.3) (0.3,-0.3) The global mean RMS radial orbit errors are of the order of 7 mm. The global mean annual (seasonal) component of the radial error can be neglected. The orbit-related errors of the decadal trends are less than 0.05 mm/year, the interannual trend is maximum 0.1 mm/year at the 5-year time scale.

9 Annual amplitude of the radial orbit differences (REF test) Annual signals are most prominent for the GRGS (5.6 mm), GSFC (5.4 mm), and Geoid (3.2 mm) orbits, when compared to the GFZ reference orbit, with the amplitudes given in the parentheses. The plausible sources: differences in the Earth s time variable models, atmospheric loading corrections and geocenter motion modeling.

10 RMS of 5-year running trend differences (REF test) for all orbit configurations tested (April 1993 June 2004) The trends of the interannual regional variability reach 1.2 and 0.9 mm/year for the GSFC and GRGS orbits, respectively, when compared to the GFZ reference orbit, with the maxima in the regions around South America and Australia. Similar features are found for the Geoid orbit, but with smaller amplitude.

11 Decadal trend differences (April 1993 June 2004) The largest regional changes in the decadal trend are observed for the differences between the REF and GSFC orbits (up to 1.0 mm/yr) followed by the differences between the REF and GRGS orbits (0.7 mm/yr). The differences between the REF and Geoid orbits reach 0.4 mm/yr and those between the REF and ITRF14 orbits 0.2 mm/yr.

12 Regional upper bound errors related to the POD configuration SLR DORIS ITRF14 GEOID GSFC GRGS RMS [mm] Annual amplitude [mm] Interannual trend [mm/year] Decadal trend [mm/year] The RMS of the regional upper bound radial orbit error is more than 10 mm. The annual amplitude of this error is up to 6 mm. The interannual trend reaches 1.2 mm/year at the 5-year time scale. The orbit-related regional errors of the decadal trends reach 1.0 mm/year.

13 Conclusions I Global mean radial orbit errors over the ocean: RMS: up to ~ 7 mm, seasonal negligible, interannual < 0.1mm/year, decadal < 0.05 mm/year Regional upper bound radial orbit errors: RMS <11 mm, strong sub-seasonal signal; Seasonal: up to 6 mm, sources: Earth s time variable gravity field, stability of tracking station sub-networks, AOD gravity modelling, geocenter motion correction; Interannual (5 years): up to 1.2 mm/year, sources: z- component of the reference system, Earth s time variable gravity field Decadal: up to 1.0 mm/year (~interannual variabilty), sources: Earth s time variable gravity field model, reference systems only secondary

14 Conclusions II Potential aliasing of radial orbit errors to large scale oceanic signals: no significant artificial signals in global mean sea level from POD on annual to decadal timescales regional radial orbit errors still reach more than 1 mm/year on interannual to decadal scales. Therefore, further improvement of the orbit quality for altimetry missions is necessary to meet the requirements of the sea level product users!