[IGSMAIL-6663] IGb08: an update on IGS08

Paul Rebischung paul.rebischung at ign.fr
Mon Sep 24 23:50:36 PDT 2012


Dear IGS colleagues,

Either because they are decommissioned, dormant or have been affected by
discontinuities since 2009.5, a large number of IGS08 stations are now
unusable for operational reference frame alignments. In particular, the
IGS08 core network has dropped to about 50 usable stations recently,
leaving large holes in South America, Africa, and Eastern Asia. This
situation led us to propose an update of IGS08, called IGb08.

IGb08 mainly includes new sets of coordinates for 33 IGS08 stations having
been affected by position discontinuities since 2009.5, and 3 new stations
co-located with decommissioned IGS08 stations. (The various changes from
IGS08 to IGb08 are described in more details below.) All in all, this
update will increase the number of usable Reference Frame stations by
about 36 and the number of usable core stations by about 15.

IGb08.snx is available with some ancillary files at
ftp://igs-rf.ign.fr/pub/IGb08 & ftp://igs-rf.ensg.eu/pub/IGb08 :
* IGb08.snx           : IGb08 SINEX file
* IGb08.ssc           : IGb08.snx without covariance matrix
* IGb08_core.txt      : list of IGb08 core stations
* IGS08-IGb08.xls     : summary table of the IGS08->IGb08 changes
* IGS08-IGb08.dat     : numerical values of the IGS08->IGb08 changes
* IGb08_1701.png      : state map of the IGb08 network on week 1701
* IGb08_core_1701.png : state map of the IGb08 core network on week 1701

IGb08 will officially replace IGS08 starting with products of GPS week
1709 (7 October 2012). This switch aims at stabilizing the reference frame
alignment of the IGS products and should not lead to any particular
inconvenience for users.

Paul Rebischung, on behalf of the IGS Reference Frame Working Group



1) Introduction of post-discontinuity coordinates
-------------------------------------------------

38 IGS08 stations with discontinuities since 2009.5 were first identified
as "repairable". A special cumulative solution was then formed, based on
the igb/igs weekly solutions up to week 1701, in which station annual
motions were estimated together with their positions and velocities.
Empirical offsets from the pre-break to the post-break station positions
were derived from this cumulative solution and used to introduce
post-break coordinates for the 38 stations in IGb08.

As a validation, IGb08 was used to align the igs weekly solutions of weeks
1632-1701. A check of the alignment residuals for the 38 newly introduced
points led us to reject 5 of them from IGb08 (BAN2, DGAR, GUUG, HOLB,
OHI3).

2) Intro. of new stations co-located with decommissioned IGS08 stations
-----------------------------------------------------------------------

3 pairs of decommissioned IGS08 stations co-located with currently active
stations were identified (CHAT/CHTI, MALI/MAL2, NOUM/NRMD). Empirical
"local ties" between these pairs of stations were derived from the same
cumulative solution as in 1) and used to introduce the 3 new stations in
IGb08. All 3 new stations successfully passed the validation test
described in 1).

3) Application of position offsets to account for metadata corrections
----------------------------------------------------------------------

As explained in IGSMAIL-6662, sub-type calibrations will come into use on
week 1709 for the JPSREGANT_DD_E and JPSREGANT_SD_E antennas and replace
the type mean calibrations used so far. To make IGb08 consistent with the
new sub-type calibrations, position offsets were derived from differential
PPP tests and applied to the IGS08 positions of the 5 affected IGS08
stations (CAGZ, CONZ, KHAJ, MDVJ, ZIMJ).

It was recently discovered that a wrong antenna type had been reported in
the sitelogs of 2 IGS08 stations (ALRT, NAIN). To make IGb08 consistent
with the right antenna type (now in use), position offsets were derived
from differential PPP tests and applied to the IGS08 positions ALRT and
NAIN.

4) Discontinuity suppressions
-----------------------------

The discontinuities introduced for ALRT and NAIN at the time when their
sitelogs were corrected will be removed with the switch to IGb08.

Two discontinuities were introduced when the radome of CRO1 was removed
and re-installed in 2011. But no significant jump is in fact detectable in
CRO1's coordinate time series. Both discontinuities will thus be removed
with the switch to IGb08 and CRO1 will become usable again as a Reference
Frame station.

5) Station suppression
----------------------

Due to uncertainty about its radome loss (IGSSTATION-4610), BRMU (Bermuda)
was removed from IGb08 for the period after 2003.




More information about the IGSMail mailing list