******************************************************************************** README ******************************************************************************** University of Bern Astronomical Institute (AIUB) Martin Lasser martin.lasser@unibe.ch or: Ulrich Meyer ulrich.meyer@unibe.ch Address: Sidlerstrasse 5 3012 Bern, Switzerland Short notice on AIUB GRACE-FO monthly gravity fields: |------------------------------------------------------------------------------| | Processing strategy | |-------------------------------------|----------------------------------------| | Software | Bernese GNSS Software v5.3 | | GNSS system | GPS | | Arc length | 1.0 days | | Reference system | | | Time | GPS time (GPST) | | Position | Earth-centered Earth-fixed (see below) | | Polar motion and UT1 | CODE final products | | Pole model | IERS 2010 conventions | | Precession/Nutation | IERS 2010 conventions | | GPS orbits | CODE final products | | GPS clocks | CODE final 5s clocks | | GPS measurements | | | Observations | Ionosphere-free linear combination of | | | phase measurements, pseudo-range used | | | for clock synchronization and wide-lane| | | ambiguity resolution | | Sampling | 10s | | Elevation angle cutoff | 0 deg (RL01) and 10 deg (RL02) | | Antenna offset (w.r.t. CoM) | Time-dependent values | | GPS antenna phase center var. | Determined for each antenna from | | | stacking of carrier phase residuals of | | | reduced-dynamic orbit determination | | Observation weighting | Constant | | Antenna phase-center wind-up | Correction applied | | Relativistic corrections | Applied | | Solution strategy | Daily least-squares solution based on | | | data from 00:00:00 to 23:59:59 | | Estimated parameters | | | Receiver position | Every epoch | | Receiver clock correction | Every epoch | | Carrier phase ambiguities | Fixed (using CODE clock and OSB prod.) | |------------------------------------------------------------------------------| General remarks =============== This product contains kinematic orbit positions of the GRACE-FO satellites given in daily batches per single satellite. File naming convention: GFO_RL_.KIN : C or D (specifying the corresponding GRACE-FO satellite) : Release number (see below) : year (last two digits) : day of year Format description ================== File header: Line 1: description of orbit including release number and orbit generation date and time Line 2: dummy line Line 3: information about geodetic datum and date & time of first epoch (GPS time) Line 4: a posteriori standard deviation of L1 GPS carrier phase res. [m] Line 5: description of columns Line 6: dummy line Data records: After the header each line represents one epoch. The columns contain the following quantities: Column 1: LEO name (GRACE-FO C: GRCC, GRACE-FO D: GRCD) Column 2: modified SVN number (GRACE-FO C: L64, GRACE-FO D: L65) Column 3: GPS antenna used for POD (main antenna: POD1, redundant antenna: POD2) Column 4: GPS week Column 5: GPS second within the given GPS week [s] Column 6: position x-component (Earth-centered Earth-fixed) [m] Column 7: position y-component (Earth-centered Earth-fixed) [m] Column 8: position z-component (Earth-centered Earth-fixed) [m] Column 9: quality flag (possible values are K, X, S and G, see below) Column 10: xx-element of 3x3 epoch-wise cofactor matrix Column 11: yy-element of 3x3 epoch-wise cofactor matrix Column 12: zz-element of 3x3 epoch-wise cofactor matrix Column 13: xy-element of 3x3 epoch-wise cofactor matrix Column 14: xz-element of 3x3 epoch-wise cofactor matrix Column 15: yz-element of 3x3 epoch-wise cofactor matrix Meaning of quality flags: K: kinematic orbit position could be properly determined X: kinematic orbit position could not be determined (less than 4 satellites) S: kinematic orbit position could be determined, but less than 5 satellites were available G: kinematic orbit position has been flagged during internal AIUB orbit screening (if screening applied) Please note: - Users who would like to use all available epochs or do their own orbit screening should use all epochs marked with the flags K or G. - Users who would like to rely on the internal AIUB orbit screening should use only epochs marked with flag K. - It is recommended not to use epochs denoted with the flag S. Note on cofactors: Columns 10-15 contain the epoch-wise 3x3 cofactor matrix. In order to compute the epoch-wise covariance matrix for the positions these cofactors need to be multiplied by the square of the a posteri RMS value found in line 4 of the header. Releases ======== RL01: - Orbits for 30-May-2018 (day 18/150) to 31-Dec-2021 (day 21/365). - Based on RL04 GRACE-FO GPS and attitude data. - GPS data not screened for ionosphere-induced artifacts. - An elevation cutoff of 0 deg was applied. - Carrier phase ambiguities are fixed to their integer values based on undifferenced ambiguity resolution using CODE's final clock and Observation Specific Bias (OSB) product. - Reference frames (defined by GNSS products): - 18/150-20/137: IGS14 - 20/138-21/365: IGb14 - Kinematic orbit positions were not screened for outliers. - Notice that the orbits of GRACE-FO C after day 21/082 and for GRACE-FO D after day 21/201 are not of optimal quality as they are based on PCVs not taking into account the lowered elevation cutoffs. - When using these GRACE-FO kinematic orbits, please cite them using the following reference: Arnold, Daniel; Jäggi, Adrian (2020). AIUB GRACE-FO kinematic orbits, release 01. Published by Astronomical Institute, University of Bern. URL: http://www.aiub.unibe.ch/download/LEO_ORBITS/GRACE-FO; DOI: 10.7892/boris.147231 RL02: - Orbits for 23-May-2018 (day 18/143) to now (regularly updated). - Based on RL04 GRACE-FO GPS and attitude data. - GPS data not screened for ionosphere-induced artifacts. - An elevation cutoff of 10 deg was applied. - GNSS products: CODE repro I20 products (http://ftp.aiub.unibe.ch/REPRO_I20) for days 18/143-18/365, operational CODE final products from day 19/001 onwards. - Carrier phase ambiguities are fixed to their integer values based on undifferenced ambiguity resolution using CODE's clock and Observation Specific Bias (OSB) products. - Reference frames (defined by GNSS products): - 18/143-20/137: IGS14 - 20/138-22/330: IGb14 - 22/331-now : IGS20 - Kinematic orbit positions were not screened for outliers. - When using these GRACE-FO kinematic orbits, please cite them using the following reference: Arnold, Daniel; Jäggi, Adrian (2022). AIUB GRACE-FO kinematic orbits, release 02. Published by Astronomical Institute, University of Bern. URL: http://www.aiub.unibe.ch/download/LEO_ORBITS/GRACE-FO; DOI: 10.48350/169040 Internal remarks: GRACE Follow-On gravity field modelling ======================================= Affects campaigns ML_GFO{18..25} ------------------------------------------------------------- The version AIUB-GRACE-FO-operational_2018-{06,07,10..12}.gfc AIUB-GRACE-FO-operational_2019-{01..12}.gfc AIUB-GRACE-FO-operational_2020-{01..12}.gfc AIUB-GRACE-FO-operational_2021-{01..12}.gfc AIUB-GRACE-FO-operational_2022-{01..12}.gfc is outdated (GRACE Follow-On rl01) and exists until Dec. 2022. This version is based on the computations in classic160_I0ifgACT_avcei4, classic160_I1ifgACT_ovce_avcei4 classic160_I0n_avcei4, classic160_I1n_ovce_avcei4 classicSBFfullScaleKin30s_I0n_avcei4 default: GRAVclassic160_I1ifgACT_ovce_.GRV specials: 2018-06 -> GRAVclassic160_I1n_ovce_1806.GRV 2020-{09,11} -> GRAVclassic160_I1ifgACT_ovce_avcei4_.GRV 2021-{03,04} 2022-{03..07, 09,11} 2022-01 -> GRAVclassic160_redoneWithNewXG_I1ifgACT_ovce_avcei4_.GRV 2022-02 -> GRAVclassicSBFfullScaleKin30s_I0n_avcei4_.GRV --------------------------------------------------------- Current version is AIUB-GRACE-FO-rl02op_2018-{06,07,10..12}.gfc AIUB-GRACE-FO-rl02op_2019-{01..12}.gfc AIUB-GRACE-FO-rl02op_2020-{01..12}.gfc AIUB-GRACE-FO-rl02op_2021-{01..12}.gfc AIUB-GRACE-FO-rl02op_2022-{01..12}.gfc AIUB-GRACE-FO-rl02op_2023-{01..06}.gfc This version is based on the computations in empKinKb_wPCA_G3P_i0, empKinKb_wPCA_G3P_i1 and computed with $XG (LM_VERS) and buildMonth_perlTest_opEmpTS_run.pl ($us). Update: From Nov 2023 onwards: buildMonth_perlTest_g3p_rl02op_fromNov2023.pl used ($us) with $XG (LM_VERS). default: empKinKb_wPCA_G3P_i{0,1} (the name is the same for all solution) specials: 2018-06 and 2022-02 computed with JPL-ACTs: -> notes: none - full scale matrix estimated in RSW - KIN taken with 30 s sampling - KIN emp. cov. only for 360 epochs 2023-01 and 2023-02: -> notes: none - full scale matrix estimated in RSW - VCE on stoch. parameters - only emp for K-band (no emp for KIN) - no VCE for K-band 2023-03..2023-05 -> notes: repeat months - full scale matrix estimated in RSW (would not be necessary for 2023-05) - VCE on stoch. parameters - only emp for K-band from average 2018-2022 (no emp for KIN) 2023-06..2024-01 -> notes: for 2023-09 the JPL-ACH product or 5 min PCAs delivers slightly better results (not used) - full scale matrix estimated in RSW (would not really be necessary) - VCE on stoch. parameters - only emp for K-band (no emp for KIN) 2024-02.. 2024-05 -> notes: since 2024 JPL-ACH product delivers slightly better results (clearly better since 2024-03) - JPL-ACH used - full scale matrix estimated in RSW - VCE on stoch. parameters - only emp for K-band (no emp for KIN but would be possible with JPL-ACH for 2024-02 and 2024-03) - i0,i1,i2 with initial solution linearised (i0), i1 for VCE and i2 for VCE and emp 2024-06..2024-11 -> notes: since 2024-06 JPL-ACH is replaced by JPL-ACH2 product (JPL-ACH2 also produces better results for relaxed-pointing months 2023-01, 2023-02, 2023-07..2023-12) - JPL-ACH2 used (note for 2024-06 JPL-ACH is a copy of JPL-ACH2 product) - full scale matrix estimated in RSW - VCE on stoch. parameters - only emp for K-band (no emp for KIN) - i0,i1,i2 with initial solution linearised (i0), i1 for VCE and i2 for VCE and emp 2024-06..2024-11, 2025-01, 2025-02, 2025-03 -> notes: since 2024-06 JPL-ACH is replaced by JPL-ACH2 product (JPL-ACH2 also produces better results for relaxed-pointing months 2023-01, 2023-02, 2023-07..2023-12) - JPL-ACH2 used (note for 2024-06 JPL-ACH is a copy of JPL-ACH2 product) - full scale matrix estimated in RSW - VCE on stoch. parameters - only emp for K-band (no emp for KIN) - $check4outliersInArcwiseVCEasap set to "yes", thus, arcs already checked in i1 (before KBR-VCE) - i0,i1,i2 with initial solution linearised (i0), i1 for VCE and i2 for VCE and emp 2024-12 -> notes: since 2024-06 JPL-ACH is replaced by JPL-ACH2 product (JPL-ACH2 also produces better results for relaxed-pointing months 2023-01, 2023-02, 2023-07..2023-12) - JPL-ACH2 used - full scale matrix estimated in RSW - VCE on stoch. parameters - only emp for K-band (no emp for KIN) - $check4outliersInArcwiseVCEasap set to "yes", thus, arcs already checked in i1 (before KBR-VCE) - i0,i1,i2 with initial solution linearised (i0), i1 for VCE and i2 for VCE and emp # 2024-?? # -> notes: since 2024 JPL-ACH product delivers slightly results # - JPL-ACH used # - full scale matrix estimated in RSW # - VCE on stoch. parameters # - only emp for K-band (no emp for KIN) ??? # - emp for KIN and K-band (no emp for KIN) ??? # - i0,i1,i2 with initial solution linearised (i0), i1 for VCE and i2 for VCE and emp From Jan. 2023 onwards the JPL_ACT (transplant to GF2) data is deprecated and not produced anymore: -> The JPL_ACH data is provided operationally -> Parallel computations: classic_I{0,1}n61, classicSBFfullScaleKin30s_I{0,1}n61 (both 2018-06 until 2022-12) classicSBFfullScaleKin30s_empKinKb_I{0,1}n61 (starting from 2023) ======== Remarks: - c00 in AIUB-GRACE-FO-operational_ (GRACE Follow-On rl01) is not correctly scaled and refers to GM of DE405 (inconsistent with the header). This is fixed for AIUB-GRACE-FO-rl02op (in .gfc files [with the Matlab-Routine]). All coefficients now refer to GM of DE405. This GM is correctly stated in the header.