From AC Tue Jan 3 13:28:20 1995 From: AC (AC) Date: Tue, 3 Jan 1995 13:28:20 PST Subject: [IGSMAIL-0824] ITRF92/ITRF93 IGS product change Message-ID: ****************************************************************************** IGS Electronic Mail Tue Jan 3 13:28:20 PST 1995 Message Number 0824 ****************************************************************************** Author: AC Coordinator Subject: ITRF92/ITRF93 IGS product change Dear colleagues, As agreed by all the IGS analysis centres, starting on January 1, 1995, all the individual centre and IGS combined products will be based on ITRF93. This reference frame change is necessary as the new ITRF93 represents a significant consistency improvement with respect to ITRF92 which was used for all the IGS products in 1994. The reference frames in the IGS products are realized by fixing or constraining a set of 13 stations at the ITRF coordinates & velocities, see the IGSMAIL #430 for ITRF92 used in 1994 and the IGSMAIL #819 for ITRF93, to be used in 1995. Unfortunately, on Jan 1/95, this reference frame change will introduce small discontinuities in all the IGS products. Since in 1995 we are still using the same 13 stations to realize ITRF93 we can estimate the transformation between the 1994 and 1995 IGS products more precisely than the nominal transformation and time evolution published in the 1993 IERS Annual Report and the IERS Technical Note #18. Below are the estimated and expected changes between the 1994 and 1995 IGS products: EXPECTED/ESTIMATED DISCONTINUITIES OF IGS PRODUCTS (ORBITS, EOP, SSC, CLOCKS) AT 1995.00 (IGS(1995)-IGS(1994) PRODUCTS T1(cm) T2(cm) T3(cm) D(ppb) R1(mas) R2(mas) R3(mas) Remarks Y-pole X-pole -UT1 (1) IGS AC's(orb, EOP,SSC,clocks)-.9 -1.2 .3 -.6 1.4 1.1 .4 (2),(6) sigma .5 .5 .5 .8 .2 .2 .2 (3),(6) Rates (./year) -.02 -.11 .15 -.13 .16 .30 -.13 (4),(6) IGS Combined (orbits,EOP) -.9 -1.2 .3 -.6 1.66 .61 .4 (2),(5) sigma .5 .5 .5 .8 .5 .5 .2 (3),(5) Rates (./year) -.02 -.11 .15 -.13 .12 .15 .13 (4),(5) -------------------------------------------------------------------- Remarks: (1) The actual changes in EOP have the opposite signs,i.e. the 1995 EOP x- pole and y-pole will be decreased by about 1.1 and 1.4mas w.r.t. 1994. The transformation parameters (T1-3,D,R1-3) are consistent with the 1993 IERS Annual Rep.,( eqn 3, p. II-52); i.e ITRF92 => ITRF93. (2) Applicable only to the SSC of the daily constrained IGS solutions. Clock solutions are insensitive to the rotation parameters. (3) Sigmas represent the consistency of both ITRF(IGS) realizations, the precision of the 1994/95 product discontinuity should be much higher (4) can be used to precisely transform the 1995 IGS products to be consistent with the 1994 and vice versa. (e.g. for plate tectonics, time evolution, etc.) The associated epoch for the parameters and rates above is 1995.0. (5) For the IGS combined products the ITRF-IERS(EOP) misalignments in R1,R2 only as it is used in the IGS combination, i.e. the difference (at 1995.0) computed from the Tables II-3 of the 1993 and 1992 IERS Annual Reports (p.II-19 and II-17, resp.) (6) The transformations & rates based on the following 13 ITRF constraining stations (identity weighting): ALGO, FAIR, GOLD HART, KOKB, KOSG, MADR, SANT, TIDB, TROM, WETB, YAR1, YELL (see IGSMAIL #819, #430) The first set was obtained from the 7 parameter transformation for the 13 station coordinates/velocities in ITRF92 and ITRF93 (see IGSMAIL#430,#819) and should be a good approximation of the expected change for all the IGS AC's. Since individual AC may be constraining more stations, using some different stations, data weighting, etc., the actual changes may vary slightly from Centre to Centre and from day to day. (For more details see the individual AC summary report for GPS Week 782 or later, which may include better estimates of the centre product changes). The transformation above is also quite consistent with the published transformation between ITRF92 and ITRF93, which is based on all ITRF stations, nevertheless, on the average, the above values should be more representative of the actual IGS product changes. The second set of transformation parameters, applicable to the IGS combination was obtained the same way as above, except for the R1,R2 orientation parameters which were computed from the IERS-ITRF92 misalignment(see the IERS 1992 Annual Rep., p.II-17) applied to the IGS combinations in 1994 and the IERS-ITRF93 misalignment (see 1993 IERS Annual Rep, p.II-19) which will be applied to the IGS Combined products in 1995. Also note the ITRF93 velocities are slightly biased w.r.t. NNR NUVEL1A, however the ITRF93 velocity field greatly reduces the drift between IERS(EOP) and ITRF93 frames. The non NNR ITRF93 velocities causes only small orientation changes with practically the same relative precision. The rates for R1, R2, and R3 above, which again are consistent with the differences between NNR NUVELl1 and ITRF93 in the IERS Technical Note #13, p.17) can be used if the past time evolution of IGS products is to be maintained, or when transforming the 1995 IGS products to a NNR reference system. Again, the above values should be more consistent and representative of the IGS products. For more information/comments and suggestions regarding the above, please contact: J. Kouba (kouba at geod.emr.ca) ACKNOWLEDGEMENTS: Cooperation of all the IGS Analysis Centres (AC), Drs Z. Altamimi and C. Boucher of the IERS ITRF Section are greatfully acknowledged. Also received able and rapid assistance of the 'emr' AC team members, Remi Ferland and Pierre Tetreault. [Mailed From: Jan Kouba ]