[IGSMAIL-6374] igs08.atx including GLONASS satellite antenna corrections

[Ralf Schmid]

Dear IGS colleagues!

With [IGSMAIL-6355] the transition from igs05.atx to an updated phase 
center correction model called igs08.atx was announced that will be 
consistent with the new IGS reference frame IGS08:
http://igscb.jpl.nasa.gov/pipermail/igsmail/2011/006347.html
This transition is still intended for GPS week 1632 (17 April 2011). So 
far, only a preliminary version of the antenna model without GLONASS 
satellite antenna corrections was available:
ftp://igs.org/igscb/station/general/pcv_archive/igs08_1604_woGLO_final.atx

In the meantime, CODE and ESOC have closed this gap by reprocessing 8 
and 3 years of data, respectively. This step was necessary, as GLONASS 
observations were not considered for the first IGS reprocessing campaign 
repro1. On the other hand, a separate reprocessing provided the 
opportunity to reestimate phase center variations (PCVs) besides the 
offsets in radial direction.

So, additional effort was needed by CODE and ESOC. In general, combined 
GPS+GLONASS solutions were analyzed. Those were aligned to the new 
reference frame IGS08, and the GPS satellite antenna corrections were 
fixed to the values contained in igs08_1604_woGLO_final.atx. In detail, 
the following strategy was pursued:
1) Estimate satellite-specific GLONASS z-offsets in radial direction for 
given fixed zero PCVs
2) Estimate block-specific PCVs for fixed z-offsets from step 1
3) Remove remaining offset-dependent fraction from PCVs
4) Average PCVs from CODE and ESOC
5) Reestimate z-offsets for fixed PCVs from step 4
6) Average z-offsets from CODE and ESOC

Strategy and general results are similar to those presented during the 
last IGS Workshop in Newcastle, but now consistent with IGS08:
ftp://stella.ncl.ac.uk/pub/IGSposters/Dilssner.pdf
There is a nearly perfect agreement between CODE and ESOC for the PCVs, 
but a bias of about 7 cm in the z-offsets. Besides a common PCV curve 
for all GLONASS and GLONASS-M satellites, satellite-specific PCV values 
were estimated for GLONASS-M satellite R714. Moreover, new phase center 
corrections were estimated for the first Block IIF satellite, namely 
G062, as its PCV values were purely based on ESOC results so far.

The new model is available at
ftp://igs.org/igscb/station/general/igs08_1629.atx
or
ftp://igs.org/igscb/station/general/igs08.atx
As the G062 results are further investigated, the model might be updated 
before the official switch to IGS08/igs08.atx.

In detail, the following changed compared to igs08_1604_woGLO_final.atx:
1) Historical GPS Block I satellites (active prior to 1994) were added 
with a block-specific mean z-offset of 195 cm.
2) GLONASS satellites were added with satellite-specific z-offsets and 
block-specific PCVs. The maximum nadir angle could be increased from 14° 
to 15°.
3) Satellite-specific corrections were added for GLONASS-M satellite R714.
4) Historical GLONASS satellites (active between 1998 and 2003) were 
added with a block-specific mean z-offset of 210 cm.
5) The correction values for GPS Block IIF satellite G062 were updated 
with new estimates from CODE and ESOC.
6) Validity period for GPS Block I satellite G010 adapted to igs05.atx.

Many thanks to Rolf Dach, Florian Dilssner and their colleagues from 
CODE and ESOC for their huge effort!

Ralf Schmid
on behalf of the IGS Antenna Working Group