From RFerland at NRCan.gc.ca Thu Oct 9 10:57:55 2003
From: RFerland at NRCan.gc.ca (=?iso-8859-1?Q?=22Ferland=2C_R=E9mi=22?=)
Date: Thu, 9 Oct 2003 13:57:55 -0400
Subject: [IGSMAIL-4642]: IGS00 (v2)
Message-ID: <5ACEEE96F4F9D41190790002A528B51E78E7A3@s0-ott-x11.nrn.nrcan.gc.ca>
******************************************************************************
IGS Electronic Mail 09 Oct 11:01:47 PDT 2003 Message Number 4642
******************************************************************************
Author: R?mi Ferland
Dear colleagues,
You will find below a proposed update of the IGS00 realization of
ITRF2000. It includes the rationale as well as details of the
analysis/comparisons/tests. A similar version of the IGS00(v2)
was distributed within the RFWG two weeks ago. A few
comments/suggestions were received and the proposed IGS00(v2) was
updated.
The ACs are encouraged to do your own independent checks, tests,
comparisons, etc. and share them.
If there are any comments, concerns, problems etc. I will be
happy to address them and update the proposed IGS00(v2), if
necessary.
I suggest that two weeks should be sufficient to do your own
tests and report any problem. If no problems are encountered
after that period, then the proposed realization will be
considered ready for implementation.
As in the past, it will be up to the AC coordinator, Gerd,
and the ACs to agree on the implementation.
Four files are available:
README - the rest of this e-mail
IGS00_v2_7.xls - An excel file used to summarize some statistics
on several stations (some info may be out of
date).
IGS03P33_RS99.snx - A SINEX file containing the stations retained
in this second iteration.
Original_54.tiff - Stations distribution in the current IGS
realization.
proposed_99.tiff - Station distribution in the proposed IGS
realization.
You can find them at:
ftp macs.geod.emr.ca
cd /pub/requests/sinex/rfwg
==================================
1) Introduction
After ITRF2000 was made available in early 2001, a
realization of the frame was prepared using the cumulative
solution from GPS week 1131 "IGS01P37.snx". It is referred
below as IGS2000(v1). An important reason for doing this was
to improve internal consistency. More about this can be found
in:
Kouba, J., J. Ray and M.M. Watkins, IGS Reference Frame
Realization, 1998 IGS Analysis Center Workshop Proceedings,
European Space Operations Centre, Darmstadt, Germany.
For IGS2000(v1), a subset of "high quality" stations was
agreed to within the reference frame working group. It
originally included 54 stations and was implemented on GPS
week 1143 (01/12/02). Since then, the number of usable
stations for the reference frame realization has gradually
decreased to about 45 stations.
Suggestions were made last spring to improve
the existing realization (namely from J.Ray, M. Heflin and
G. Gendt) by increasing the number of stations. Those
suggestions were combined. Several of those suggested
were located in Europe and North America. Those two regions
have already a good coverage. In an attempt to improve the
global distribution, the selection effort I made, was mainly
concentrated in locations with relatively sparse coverage.
Altogether, close to 200 stations were considered, about half
of them have passed all the selection process steps. The
criterions used for the selections generally followed the in
Kouba et al. (reference above). The selection criterions were
also reiterated and expanded in:
Ray, J.R., Reinforcing and securing the IGS Reference
Tracking Network, ....
As was the case in the past, few stations met equally
well all the criterions listed in the above references.
During the iterations, some subjective decision making took
place. If you find the selection of some stations unreasonable
(excluding politics), the suggested set can certainly be adjusted.
It was suggested that for some stations which did
experience discontinuities, to use separately, the appropriate
portion of the solution for the alignment and for the
realization. This suggestion was implemented. In those cases,
the velocity estimate for the segments was forced to be unique.
This will cause some temporary side effects on the alignment
of the cumulative solution which can easily be handled.
The idea was also extended to split the stations contributing
to the alignment to ITRF2000 and the stations forming the
realization. Although most stations are common, this allowed
to retain some remote stations that would otherwise have been
rejected. Some stations had limited data available when
ITRF2000 was made available about 2.5 years ago. Also, the
most recent data used in ITRF2000 to estimate the stations
coordinates/velocity estimates is at least 3 years old.
In some cases, this caused some occasional differences.
After the alignment and realization processes, the
magnitude of the discontinuity between the current and
proposed realizations should be small, since both subsets
were extracted from consistent solutions.
2) Review of some stations IGS2000(v1)
BRMU Discontinuity last spring (Mar. 22).
Suggest to drop it.
DGAR Station performance is currently deficient.
It is located in an isolated location.
Suggest to keep it in.
KERG Station has had occasional data problems.
Suggest to keep it in.
KWJ1 Station down since 02/11. No sign of change.
Suggest to Drop it.
LHAS Station has had some problems recently.
Suggest to keep it in based on past performance.
OHIG Off. Use OHI2.
SHAO Station down since 02/10. No sign of change.
Suggest to Drop it.
THU1 Off. Use THU3
ZWEN No data since 03/03. No sign of change.
Suggest to Drop it.
3) Suggested stations in the new realization IGS2000(v2)
The proposed set currently includes 99 stations:
ALGO ALIC ARTU ASC1 AUCK BAHR BILI BOR1 BRUS CAS1
CEDU CHAT CHUR COCO CRO1 DARW DAV1 DGAR DRAO DUBO
EISL FAIR FLIN FORT GLSV GODE GOLD GOUG GRAS GRAZ
GUAM HOB2 HOFN HRAO IISC IRKT JAB1 JOZE KARR KELY
KERG KIT3 KOKB KOUR KSTU LAMA LHAS LPGS MAC1 MAG0
MALI MANA MAS1 MATE MAW1 MCM4 MDO1 MKEA NICO NKLG
NLIB NOUM NRC1 NTUS NYAL OHI2 ONSA PERT PETP PIE1
POL2 POTS RBAY RIOG SANT SCH2 SFER STJO SYOG THTI
THU3 TIDB TIXI TOW2 TRAB TRO1 TSKB UNSA URUM VESL
VILL WES2 WSRT WTZR YAKT YAR1 YELL YSSK ZIMM
8 stations from IGS2000(v1) have not been retained in
the proposed IGS2000(v2), Those are:
BRMU KWJ1 OHIG SHAO THU1 TROM WSRT ZWEN
Some solutions (OHIG -> OHI2 , THU1 -> THU3 , TROM -> TRO1 )
had alternative solution at the same site.
There is 47 common stations between IGS2000(v1) and
IGS2000(v2). They are:
ALGO ASC1 AUCK BAHR CAS1 CEDU CHAT DAV1 DGAR DRAO
FAIR FORT GODE GOLD GRAZ GUAM HOB2 HRAO IRKT KERG
KIT3 KOKB KOUR LHAS LPGS MAC1 MALI MAS1 MATE MCM4
MDO1 NLIB NYAL ONSA PERT PIE1 POTS RIOG SANT TIDB
TSKB VILL WES2 WSRT WTZR YAR1 YELL
52 new stations are in the proposed in IGS2000(v2):
ALIC ARTU BILI BOR1 BRUS CHUR COCO CRO1 DARW DUBO
EISL FLIN GLSV GOUG GRAS HOFN IISC JAB1 JOZE KARR
KELY KSTU LAMA MAG0 MANA MAW1 MKEA NICO NKLG NOUM
NRC1 NTUS OHI2 PETP POL2 RBAY SCH2 SFER STJO SYOG
THTI THU3 TIXI TOW2 TRAB TRO1 UNSA URUM VESL YAKT
YSSK ZIMM
4) Sites with different alignment & realization points
Six sites used in the reference frame realization had
experienced significant discontinuity since the original
IGS realization and since ITRF2000 became official. In the
IGS cumulative solution (updated weekly), at stations with
identified discontinuities, the position/velocity were
estimated independently. For the alignment/realization,
the independent velocity estimates at a station were
combined to provide a unique velocity per site. The sites
with estimated discontinuity are:
BILI COCO FAIR GUAM HOFN KOKB
The cumulative solution for GPS week 1231 (IGS03P33.snx)
was used for the proposed IGS2000(v2).
5) Alignment to ITRF2000
After forcing unique velocity estimate at selected sites on
the cumulative solution IGS03P33.snx; the 71 common stations
between ITRF2000 and the IGS2000(v2) realization (listed
below) were used for the alignment:
ALGO ASC1 AUCK BAHR BOR1 BRUS CAS1 CHAT CHUR COCO
DAV1 DGAR DRAO DUBO EISL FAIR FLIN FORT GLSV GODE
GOLD GRAS GRAZ HOB2 HOFN HRAO IISC IRKT KARR KELY
KERG KIT3 KOUR KSTU LAMA LHAS LPGS MAC1 MAS1 MATE
MAW1 MCM4 MDO1 MKEA NLIB NOUM NRC1 NTUS NYAL ONSA
PERT PIE1 POL2 POTS RIOG SANT SCH2 STJO SYOG TIDB
TIXI TOW2 TROM TSKB VILL WES2 WSRT WTZR YAR1 YELL
ZIMM
The other stations were not used in the alignment either
because:
1) There was significant difference in position
and/or velocity between ITRF and its IGS proposed
realization.
- or -
2) The station was not available in ITRF2000.
Those 28 "orphan" stations are:
ALIC ARTU BILI CEDU CRO1 DARW GOUG GUAM JAB1 JOZE
KOKB MAG0 MALI MANA NICO NKLG OHI2 PETP RBAY SFER
THTI THU3 TRAB TRO1 UNSA URUM VESL YAKT YSSK
Note: unit weights were used for the alignment process.
5.1) ITRF2000 - IGS03P33_RS99.snx
The residuals (below) between ITRF2000 and the 70
stations used in the alignment. The RMS of the residuals is
sub-mm in the horizontal position and velocity, and 2.6mm &
1.5mm/y in vertical position and velocity.
Code Pt Soln d. Lat d. Lon d. Hei d.VLat d.VLon d.VHei
(mm) (mm) (mm) (mm/y) (mm/y) (mm/y)
ALGO A ---- -.1 -.1 2.0 -.3 -.2 -1.9
ASC1 A ---- .1 -.8 2.2 .3 .3 -1.6
AUCK A ---- .8 .3 -.6 -.3 .9 1.7
BAHR A ---- -.5 .7 1.1 -.4 -2.0 -1.0
BOR1 A ---- -.2 -.4 2.2 -.1 -.4 -.2
BRUS A ---- -.1 -.4 .2 .3 .0 -.6
CAS1 A ---- .4 .5 -4.3 .3 .1 .8
CHAT A ---- .9 -.1 -.6 -.2 .3 1.5
CHUR A ---- -.3 -.3 -.6 -.4 -.1 2.1
COCO A 1 -.5 2.3 .0 .4 -2.3 .0
DAV1 A ---- -.7 .2 -1.2 .8 .1 1.7
DGAR A ---- -.4 .0 -.8 .3 -.1 .6
DRAO A ---- -.2 -.7 -.1 -.5 .0 -.4
DUBO A ---- -1.0 -.3 1.7 .1 .3 -2.4
EISL A ---- .0 1.1 -2.5 1.1 -1.4 2.0
FAIR A 1 -.5 -.3 .0 -.8 -.6 3.1
FLIN A ---- -.7 -.1 -2.0 -1.2 .5 1.5
FORT A ---- .3 -.1 4.3 1.3 -.4 1.0
GLSV A ---- -.2 .5 -1.3 -.5 -.5 1.7
GODE A ---- .3 .2 -2.1 -.4 .0 .1
GOLD A ---- .2 -1.2 -.6 .2 2.9 -3.0
GRAS A ---- -.8 -.5 -2.6 -.2 -.1 -.6
GRAZ A ---- -.4 .1 .0 .2 .7 -2.1
HOB2 A ---- .8 .0 -.2 .1 .6 -3.2
HOFN A 1 .6 -.3 -5.5 -.6 .9 .6
HRAO A ---- .6 -1.4 -3.0 .5 .1 .4
IISC A ---- -.5 .5 -1.1 .4 .2 1.8
IRKT A ---- -.5 -.1 .6 .0 .6 -1.2
KARR A ---- -.9 .6 -6.7 .8 -.3 1.6
KELY A ---- -.6 -.4 -1.3 -.5 .1 1.3
KERG A ---- -.5 .6 3.5 .2 .3 .6
KIT3 A ---- -.2 .3 1.1 .4 .3 1.3
KOUR A ---- .0 .0 4.0 1.0 -.1 -1.0
KSTU A ---- -.1 .7 4.3 -1.5 .0 -2.6
LAMA A ---- -.5 .7 -2.3 .9 .6 2.4
LHAS A ---- -1.1 .4 1.5 .2 1.3 1.3
LPGS A ---- .5 1.5 2.3 1.0 -.6 -.8
MAC1 A ---- .3 -.3 .8 -.8 .9 -1.0
MAS1 A ---- -.1 -.7 .3 -.3 .4 .2
MATE A ---- -.1 -.3 1.8 .3 -.8 -1.7
MAW1 A ---- -.4 -.8 -3.8 .6 .5 1.8
MCM4 A ---- .0 .5 3.2 .2 .8 -1.0
MDO1 A ---- .0 -.9 -1.7 .0 .3 -.3
MKEA A ---- .2 -1.6 2.4 -.3 .8 -2.2
NLIB A ---- .0 -.5 .7 .2 .4 -2.3
NOUM A ---- 2.9 1.5 -.9 -.5 -.3 -.6
NRC1 A ---- .0 -.6 -.3 .0 .1 .1
NTUS A ---- -2.1 .2 2.2 3.0 .7 -.1
NYAL A ---- .1 -1.0 2.9 .0 -.3 -.9
ONSA A ---- -.3 -.3 -.1 -.3 .1 2.0
PERT A ---- -.1 .1 3.0 .1 .3 2.4
PIE1 A ---- -.7 -1.2 -4.8 -.5 -.8 -2.0
POL2 A ---- -.8 .5 2.5 -.5 .3 .3
POTS A ---- -.2 -.3 -.9 .1 .1 -.2
RIOG A ---- -1.4 1.0 -4.3 1.2 .0 .7
SANT A ---- .7 .9 2.2 1.4 -1.9 1.9
SCH2 A ---- -.4 -.1 -3.2 -.6 -.3 .5
STJO A ---- .1 -.2 -.4 -.4 -1.0 -.6
SYOG A ---- 1.8 .0 4.0 -.6 -.5 -2.5
TIDB A ---- .6 -.4 -.2 .1 .7 -2.0
TIXI A ---- .5 2.7 2.0 -.6 -1.3 .2
TOW2 A ---- -.1 -.8 -4.4 .5 1.1 2.2
TROM A ---- .0 -2.2 3.3 -.1 -1.3 1.4
TSKB A ---- -.9 -1.2 4.7 -1.6 -.7 -.6
VILL A ---- -.5 -.2 2.1 .2 -.3 -1.1
WES2 A ---- -.9 2.2 .6 .2 .7 -1.8
WSRT A ---- .0 -.1 -.8 -.1 .3 .5
WTZR A ---- -.2 -.4 1.5 -.1 -.2 -1.2
YAR1 A ---- -.3 .1 -1.0 .2 .2 -.1
YELL A ---- -.2 -.4 .5 .0 -.1 -.5
ZIMM A ---- -1.3 .3 -6.5 .5 -.7 2.2
5.2) NNR-NUVEL1A - IGS03P33_RS99.snx
Comparison with the plate motion model provides some "quality"
control over the estimated velocity field in the cumulative
solution. Since the current plate motion model is limited to
the horizontal components, the estimated vertical velocity is
compared to a null (0.0mm/y) velocity. On most of the Earth
surface, vertical velocity is expected to be small, except in
active zones.
The vertical velocity for the stations CHUR & SCH2 is consistent
with expected post-glacial rebound. Stations with large horizontal
residuals are generally located close to tectonic plate
boundaries (E.g.: GUAM, LHAS, SANT).
Excluding stations CHUR, EISL, GUAM, IISC, LHAS, MAC1, NTUS,
PETP, SANT, SCH2 and URUM, the RMS between NNR-NUVEL1A
and the proposed realization is 3.0mm/y,2.3mm/y and 2.9mm/y
in the north, east and vertical velocity components.
Code Pt Soln d.VLat d.VLon d.VHei
(mm/y) (mm/y) (mm/y)
ALGO A ---- 1.5 -.6 -4.3
ALIC A ---- 1.4 .9 -3.7
ARTU A ---- -2.1 -.4 .7
ASC1 A ---- 1.4 -.5 1.0
AUCK A ---- -1.3 1.8 -1.4
BAHR A ---- 6.7 -2.8 -1.5
BILI A ---- .1 2.0 -1.6
BOR1 A ---- -1.0 -.1 1.0
BRUS A ---- .7 .4 -1.1
CAS1 A ---- 1.2 -.6 -2.9
CEDU A ---- .9 -.4 -2.1
CHAT A ---- -.8 1.2 -1.1
CHUR A ---- 1.1 -1.5 -10.0
COCO A 2 4.5 -.1 -1.9
CRO1 A ---- -1.7 -6.2 5.8
DARW A ---- 3.5 1.9 -1.9
DAV1 A ---- 2.8 -.7 -2.6
DGAR A ---- 9.5 -1.8 -2.1
DRAO A ---- -1.6 -1.8 -1.7
DUBO A ---- .3 -.7 2.9
EISL A ---- -.3 11.6 1.8
FAIR A 2 1.6 -2.1 2.7
FLIN A ---- 1.3 -.9 -1.6
FORT A ---- .8 -1.6 .1
GLSV A ---- -1.1 .6 .7
GODE A ---- .4 -.5 2.0
GOLD A ---- -6.3 7.7 -1.7
GOUG A ---- .5 4.3 5.2
GRAS A ---- -.5 -.3 .3
GRAZ A ---- -1.2 -.4 .2
GUAM A 2 -2.5 -29.9 -1.8
HOB2 A ---- -.9 -.4 -4.5
HOFN A 2 1.0 -.9 -6.4
HRAO A ---- 2.7 2.7 .0
IISC A ---- 8.0 -.4 .0
IRKT A ---- -1.1 -1.3 -1.1
JAB1 A ---- 3.1 1.1 -1.5
JOZE A ---- -.8 -.4 1.9
KARR A ---- 3.1 1.3 -4.5
KELY B ---- 2.0 -1.4 .8
KERG A ---- 2.0 .7 -4.3
KIT3 A ---- -3.4 -1.6 2.8
KOKB A 2 -.3 4.5 -2.7
KOUR A ---- .1 -1.4 -3.6
KSTU A ---- -1.5 -.7 .5
LAMA A ---- -.7 1.6 4.3
LHAS A ---- 28.9 -7.1 1.3
LPGS A ---- .3 -.8 -1.9
MAC1 A ---- -5.9 -9.0 -1.1
MAG0 A ---- 4.3 2.2 2.8
MALI A ---- 3.5 -1.2 -1.9
MANA A ---- -5.4 .0 -3.1
MAS1 A ---- .6 .8 -1.2
MATE A ---- -4.9 -2.4 -.7
MAW1 A ---- 3.5 .8 -.9
MCM4 A ---- .2 -1.3 -1.7
MDO1 A ---- .1 .5 .3
MKEA A ---- -.8 5.3 .9
NICO A ---- -4.7 5.3 -1.0
NKLG A ---- 2.4 3.2 1.1
NLIB A ---- .3 -.3 .6
NOUM A ---- -1.9 2.7 -.7
NRC1 A ---- 1.3 -.5 -3.5
NTUS A ---- -2.3 -9.4 -1.5
NYAL A ---- -.4 2.2 -7.3
OHI2 A ---- 3.7 -.6 -8.4
ONSA A ---- -.3 1.4 -.6
PERT A ---- 1.8 -.7 2.4
PETP A ---- -12.0 12.7 .7
PIE1 A ---- .6 .2 -1.6
POL2 A ---- -5.3 -1.6 .3
POTS A ---- -.7 .7 1.1
RBAY A ---- 4.0 2.5 -2.8
RIOG A ---- -.7 -1.6 -4.5
SANT A ---- -5.5 -21.7 -2.8
SCH2 A ---- .2 -.7 -10.8
SFER A ---- 1.4 2.1 -1.4
STJO A ---- 1.0 -.3 -.8
SYOG A ---- 3.1 1.4 -4.6
THTI A ---- -.1 4.0 2.3
THU3 A ---- .0 .0 4.4
TIDB A ---- -1.1 .0 -3.4
TIXI A ---- -.5 .8 .9
TOW2 A ---- .7 1.8 -3.3
TRAB A ---- -4.2 -.9 -.4
TRO1 A ---- -3.1 .1 -2.3
TSKB A ---- 15.0 .4 .0
UNSA A ---- -.4 -8.1 -.1
URUM A ---- -11.5 -3.9 -2.7
VESL A ---- 2.1 3.6 -4.2
VILL A ---- .1 -.8 .4
WES2 A ---- 1.5 .2 -.8
WSRT A ---- -.8 .6 -.2
WTZR A ---- -1.0 -.1 -.3
YAKT A ---- -1.2 -3.7 -7.1
YAR1 A ---- 3.4 .1 -.4
YELL A ---- .7 -1.3 -5.5
YSSK A ---- -1.7 5.0 -2.8
ZIMM A ---- -.3 -1.1 2.7
6) Transformation IGS2000(v2) -> IGS2000(v1)
IGS2000(v1) was originally aligned to ITRF2000 using
54 stations. It included data up to fall 2001.
IGS2000(v2) was also aligned to ITRF2000 using 70
stations. It includes data until a few weeks ago.
A 14 parameter transformation was estimated using the
following 44 stations:
ALGO ASC1 AUCK BAHR CAS1 CHAT DAV1 DGAR DRAO FAIR
FORT GODE GOLD GRAZ HOB2 HRAO IRKT KERG KIT3 KOUR
LHAS LPGS MAC1 MALI MAS1 MATE MCM4 MDO1 NLIB NYAL
ONSA PERT PIE1 POTS RIOG SANT TIDB TROM TSKB VILL
WES2 WTZR YAR1 YELL
6.1) Estimated 14 parameters:
R X (mas) : .00069
R Y (mas) : .00380
R Z (mas) : -.00598
T X (m) : -.00011
T Y (m) : .00016
T Z (m) : .00015
SCL (ppb) : .11586
d R X (mas/y) : -.00405
d R Y (mas/y) : -.00506
d R Z (mas/y) : .00008
d T X (m/y) : -.00004
d T Y (m/y) : -.00022
d T Z (m/y) : -.00012
d SCL (ppb/y) : -.05650
At reference epoch 1998.0
In a perfect world, the transformation parameters would be null.
However due to data & network selection effects, there are
some small transformations.
Note: unit weights were used for the alignment process.
6.2) Residuals IGS01P37_RS54.snx - IGS03P33_RS99.snx(transformed):
Code Pt Soln d. Lat d. Lon d. Hei d.VLat d.VLon d.VHei
(mm) (mm) (mm) (mm/y) (mm/y) (mm/y)
ALGO A ---- .1 .1 -.1 .0 .3 .6
ASC1 A ---- -.1 -.1 .9 -.4 .2 -1.6
AUCK A ---- .0 .1 -.5 -.2 .3 .1
BAHR A ---- .0 .7 .1 -.1 -1.0 .4
CAS1 A ---- .1 -.3 -.1 .1 .4 -.2
CHAT A ---- -.1 .0 -.2 -.3 -.3 -.5
DAV1 A ---- .1 -.2 -.9 .0 .3 1.0
DGAR A ---- -.2 .2 -1.0 .4 -.1 .7
DRAO A ---- .2 .1 .1 -.2 -.4 -.4
FAIR A 1 .1 .2 .0 -.4 -.1 -.9
FORT A ---- .1 -.3 .1 -.1 -.8 .3
GODE A ---- .1 .1 -.5 -.1 .9 .6
GOLD A ---- .3 .1 .4 -.2 -.2 -1.5
GRAZ A ---- -.1 .0 .0 .1 .6 -.1
HOB2 A ---- -.1 -.3 .1 .4 .3 -.3
HRAO A ---- -.8 -.4 -1.4 .7 .3 .5
IRKT A ---- .0 .1 .2 .0 .1 .1
KERG A ---- -.1 .2 .0 -.1 .2 1.0
KIT3 A ---- .0 .0 -.2 .3 -.1 .5
KOUR A ---- .1 .0 .1 -.4 -.5 -.3
LHAS A ---- -.1 .3 .4 .3 .3 -1.2
LPGS A ---- .2 .1 .1 -.1 .2 .3
MAC1 A ---- .1 -.3 -.3 -.5 .5 .0
MALI A ---- .0 .3 -.2 1.1 -2.2 -1.2
MAS1 A ---- -.1 -.1 -.8 .0 .3 .9
MATE A ---- -.1 -.1 -.2 .0 .2 .5
MCM4 A ---- -.1 .0 .2 .0 .5 1.7
MDO1 A ---- .3 .0 .4 -.2 -.1 -1.3
NLIB A ---- .2 .0 .2 .0 .1 -.6
NYAL A ---- .3 -.1 .3 1.1 .5 .2
ONSA A ---- .0 -.1 .3 .3 .2 -.6
PERT A ---- -.1 -.5 .6 .3 .4 1.5
PIE1 A ---- .3 .1 .3 -.5 .0 -.6
POTS A ---- .0 -.2 -.1 .1 .2 .2
RIOG A ---- -1.5 .6 1.2 .7 -.4 -.5
SANT A ---- .4 .1 .3 -.1 -.2 .6
TIDB A ---- .0 -.4 -.2 .7 .6 -.2
TROM A ---- .0 -.2 .4 .0 .3 1.0
TSKB A ---- -.2 .3 .2 .1 -.9 .4
VILL A ---- -.1 .0 -.4 .0 .2 .2
WES2 A ---- .3 .2 -.2 .4 .7 .6
WSRT A ---- .0 -.1 -.8 -.1 .3 .5
WTZR A ---- -.1 -.2 -.2 .0 .1 .5
YAR1 A ---- -.2 -.4 .7 .1 .8 -2.4
YELL A ---- .1 .2 -.1 -.1 -.3 .1
R.M.S. .3 .3 .5 .4 .5 .8
These residuals show the effect of adding about two
years of weekly solutions to the cumulative solution IGS01P37.snx.
7) ERP's
The effect of the new IGS realization of ITRF2000 on the ERP's
can be estimated with the above estimate of the transformation
parameters using formula:
(R X) = (R X)o + (d R X)*dt
(R Y) = (R Y)o + (d R Y)*dt
With the proposed alignment/realization procedure, there
are some small approximations because the realization
includes a somewhat different stations set than the
transformation parameters estimation uses. As suggested
above, these effects are expected to be small.
To verify the magnitude of the approximation, eleven
weeks (1220-1230) were used to compare the effect of the
actual realization changes and the estimated changes (with
the 14 parameters transformation) on the pole position.
The GPS weeks were reprocessed using the proposed IGS2000(v2)
realization. The estimated daily ERP's were compared with
the ERP's aligned using the current realization. The
difference of the two realizations (IGS00(v1) - IGS00(v2))
on the ERP's are:
dXp = -0.033 +- 0.005 mas
dYp = -0.017 +- 0.007 mas
Using the above formula, the estimated rotations at GPS week
1225.5 is:
(R X) = -0.024 mas
(R Y) = -0.023 mas
The two estimates agree at the 0.01mas:
(R X) - dXp = 0.011 mas
(R Y) - dYp = -0.006 mas
This validates that the estimated transformation parameters
are valid, and precise to about 0.01mas. This also indicates
that network effects, other than those predicted by the
usual 14 transformation parameters are very small on the
ERP's. The small rotational differences (0.011, -0.006mas)
can probably be attributed to the additional 52 stations.
This is an indication that the station coordinates are
highly consistent. The additional stations in the proposed
realization probably contribute to the noise reduction in
the ERP's. The sigmas on the differences (0.005, 0.007 mas)
are probably an indication of the maximum level of that
reduction.
8) Apparent Geocenter
Similarly, the effect of the proposed IGS realization of
ITRF2000 on the apparent geocenter can be estimated with
the above transformation parameters using formula:
(Xgc) = (T X)o + (d T X)*dt
(Ygc) = (T Y)o + (d T Y)*dt
(Zgc) = (T Z)o + (d T Z)*dt
The approximations discussed above also apply here. The
reprocessed weeks can also be used here to verify the
effect of the proposed realization on the apparent
geocenter. The difference of the two realizations
(igs00(v1)-igs00(v2)) on the apparent geocenter are:
dXGC = -0.4 +- 0.5 mm
dYGC = -1.2 +- 0.5 mm
dZGC = 0.0 +- 0.6 mm
Using the above formula, the estimated effect on the
apparent geocenter for GPS week 1225.5 is:
(T X) = -0.3 mm
(T Y) = -1.1 mm
(T Z) = -0.5 mm
Their differences are:
(T X) - dXGC = 0.1 mm
(T Y) - dYGC = -0.1 mm
(T Z) - dZGC = 0.5 mm
The two estimates are almost identical (at 0.1mm) in
the X&Y components. Although, the level of agreement
is slightly less in the Z component, it is still
within the noise level.
This is a check of the estimated transformation
parameters. The discussion above on the ERP's comparison,
can certainly be extended to the translation/geocenter
test.
9) Misc.
The residuals between the proposed realization and the
weekly (1220-1230) solutions were also checked. Except
for one isolated case, the residuals were all within 3
sigmas.
Adding/removing a few stations would unlikely have a
significant impact on the above results.
Once the final set is agreed, the latest cumulative
solution will be used to generate the final set.
==================================
Best Regards,
Remi,
R?mi Ferland
rferland at NRCan.gc.ca, 613-995-4002, fascimile/t?l?copieur 613-995-3215
Natural Resources Canada, 456-615 Booth Street, Ottawa, Ontario K1A 0E9
Ressources naturelles Canada, 456-615 rue Booth, Ottawa (Ontario) K1A 0E9
Government of Canada / Gouvernement du Canada