[IGSMAIL-1391] Jun30 IGS products changes
AC
AC
Wed Jul 17 07:54:17 PDT 1996
******************************************************************************
IGS Electronic Mail Wed Jul 17 7:54:17 PDT 1996 Message Number 1391
******************************************************************************
Author: AC coordinator
Subject: Jun30 IGS products changes
Dear Colleagues,
According to the recommendations of the 1996 AC Workshop, held in March 19-21
the following changes/improvements have been implemented for the IGS combined
products starting with Wk 0860 (June 30, 1996):
1. All IGS combined products (orbits, EOP and station solutions) are based on
ITRF94, realized by the published ITRF94 positions/velocities of the same 13
ITRF stations (TROM, MADR, KOSG, WETT, HART, FAIR, KOKB, YELL, ALGO, GOLD, SANT
TIDB, YAR1). The ITRF93/94 change introduces small discontinuities on June
30/96, mainly in orientation (< 1-2 mas) which should be accounted for in
highly precise geodetic and geodynamical applications (see the appendix below
for more details and transformations to and from different IGS realizations of
ITRF used since 1994)
2. The IGS Preliminary (IGP) orbit/clock/EOP combinations became IGS Rapid
(IGR) combinations, produced daily within 24h, starting from June 30/96. There
are no IGP products after this date.
3. The previous IGS Rapid combination products, produced weekly, typically
within 11 days and performed directly in ITRF without any alignment to IERS
EOP, became the IGS Final combination on this day. The current IGS Final
combinations based on the IERS Bull B will be terminated on this date. The
discontinuities should be at or below the noise level (.1-.2 mas). There will
no longer be any need for renaming IGS (Rapid) products (IGS=>IGR) as it was
done up to this date after completion of the Bull B (Final) combination.
4. Since June 30 all IGS solutions employ 1996 IERS diurnal/semidiurnal EOP
model. In particular, the IGS orbits include this effect (orientation changes
with amplitudes up to 1 mas and periods of nearly 12 and 24h). However, the IGS
EOP solutions (for compatibility, simplicity and to facilitate an efficient
interpolation) DO NOT INCLUDE this diurnal/semidiurnal EOP model and it must be
added to the reported EOP values if precise instantaneous EOP's are required.
This convention is also applicable to all contributed AC solutions.
5. Since June 30 all submitted PM rate estimates are combined as weighted means
utilizing the orbit weights and they are included in the IGS erp files along
with corresponding sigmas. This should facilitate an efficient interpolation of
IGS EOP even for a single day (IGR) erp. Currently all but two ACs are
estimating and submitting PM rates.
6. Since January, 1996, the operational 'broadcast' ephemeris are included in
the IGP (since June 30, the IGR) combination summaries (for statistics only;
designation 'BRD') in order to evaluate broadcast orbit precision and the BRD
reference frame stability and to provide daily average (over 24h)
transformations of the broadcast orbits (WGS84) to ITRF94.
FUTURE PLANS:
The following are currently under development or soon to comence:
1. IGS orbit predictions based on the IGS Rapid orbits, to be made available
in near real time, are currently being tested.
2. Implementation of the diurnal/semidiurnal EOP model in the Long Arc
evaluations of IGS combined and individual AC orbits.
3. A combination of UT1-UTC solutions submitted by ACs.
Finally, many thanks are extended to all ACs for their cooperation and the
extra effort connected with most of the above changes.
If there are any questions, or more information is required on the above
changes, please let me know.
With many greetings,
Jan Kouba
--------
APPENDIX: ITRF94/93/92 transformations
Starting on June 30, 1996 all IGS products, including the IGS combined orbits,
are based on ITRF94. This reference frame change will introduce small
discontinuities, mainly in orientation. More specifically the following
changes took effect on June 30/96 (day 0, GPS week 760):
(IGS(ITRF94)=>IGS(ITRF93); epoch 1996.50; transformation parameters consistent
with the IERS conventions)
IGS Orbits T1(cm) T2(cm) T3(cm) D(ppb) R1(mas) R2(mas) R3(mas)
PM x PM y -(UT1-UTC)
Value -2.1 -.1 .1 -.2 -1.27 -.87 -.54
Sigma .1 .1 .1 .2 .05 .04 .05
Rates per year -.27 .00 .20 -.09 -.13 -.20 .04
The above transformation is based on the ITRF94 and ITRF93 position/velocity
sets for the 13 ITRF stations above, while utilizing the corresponding ITRF
sigma weights. The above transformation should be well suited for the IGS
products. The transformations for individual AC solutions may vary slightly
mainly due to observation/station weighting, geometry and modeling. Some ACs
may have already estimated and included separate transformation parameters
which are better suited for their solutions in their summary reports. To assist
IGS users to bridge discontinuities caused by the ITRF changes a simple ftn77
program, transformation files and a description have been prepared and made
available on the following ftp site:
ftp macs.geod.emr.ca
cd pub/requests/itrf93_94
Below is a brief description also put in the above directory.
---
ITRF TRANSFORMATIONS (trnfsp3n.doc)
A simple transformation program was written primarily to transform IGS sp3
orbits between different realizations of ITRF's used by IGS since 1992. The
program (trnfsp3n) can also be used to transform a station coordinate set of
the same epoch to and from different ITRF's. The sp3 orbit transformations are
required for precise processing of data spanning the ITRF changes.
The following ITRF realizations have been employed by IGS since 1994:
ITRF92 Jan 2, 1994(Wk 0730) to Dec 31, 1994 (Wk 0781)
ITRF93 Jan 1, 1995(Wk 0782) to Jun 29, 1996 (Wk 0859)
ITRF94 Jun 30, 1996(Wk 0860) to present
Using the above program, sp3 orbits can be instantaneously
transferred into ITRF92/93/94 and vice versa as required. Since all IGS sp3
orbits contain epochs and are in the ITRF of the epoch, the proper
transformation of the date is utilized.
The ftn77 code source (trnfsp3n.f) has been put onto our ftp directory.
PROGRAM USAGE:
to run the program use:
trnfsp3n inputfile(sp3) outputfile.sp3 transffile
where the 'transffile' is a file containing appropriate transformation
parameters including rates.
The following transformation files ( which are based on Tables 9 and 5 of the
1994 and 1995 Analysis Coordinator reports in 1994 and 1995 IGS Annual Reports,
resp.) have been put onto the same directory:
ITRF94.itr93 transforms itrf93-> itrf94
ITRF93.itr92 transforms itrf92-> itrf93
ITRF92.itr93 transforms itrf93-> itrf92
ITRF93.itr94 transforms itrf94-> itrf93
For example, to transform the orbit file IGS08600.sp3 (in ITRF94(current date))
into IGS08600.itrf93 in ITRF93(current date) simply run:
trnfsp3n IGS08600.sp3 IGS08600.itr93 ITRF93.itr94
or to transform the 13 ITRF93(1995.0) station coordinate set in the
file ITRF93.13 into ITRF94(1995.0) in ITRF94.out simply run:
trnfsp3n ITRF93.13 ITRF94.out ITRF94.itr93
The files ITRF93.13, ITRF94.out, and this write up (trnfsp3n.doc) were also put
into the directory for your convenience.
PROGRAM LIMITATIONS:
All ITRF transformations are based on the published in 1994, 1995 IGS Annual
Reports, which are applicable to the IGS ITRF realizations. Since this is very
close to the nominal (official) ITRF93/93 realizations, the transformation
files can be used for all applications related to ITRF92/93 with little or no
precision compromise.
Only sp3 orbits are tranformed, no corresponding transformation (i.e.
correction is done to erp files), if it is required it is up to users to apply
the corresponding PM/UT1-UTC corrections to the values in the erp file.
There are only a few syntax limitations, and they all pertain to the station
transformations (they are imposed by the sp3 format) namely: The first line
must have an epoch (YYYY MM DD) in a free format starting after col #3, the
station coord (in mm, m, or km) must start with "P " or "V " in col. 1. sta
name of only 3 chars and coordinates must start in, or after col.6.
--
[Mailed From: Jan Kouba <kouba at geod.emr.ca>]
More information about the IGSMail
mailing list