PowerPointPrsentation

1
CONT02 studying rigorously combined GPS and
VLBI data
Tesmer, V. (1), D. Thaller (2), M. Krügel (1),
R. Schmid (2), M. Rothacher (3), D. Angermann
(1) (1) Deutsches Geodätisches
Forschungsinstitut, (2) Forschungseinrichtung
Satellitengeodäsie / TU Munich, (3)
GeoForschungsZentrum Potsdam
- network data reduction - results (station
positions, EOP, troposphere) - discussion - the
18-18 h problem
2
CONT02 network data reduction
stations - VLBI 8 - GPS 153
data reduction - homogeneous functional
model - stations - EOP - troposphere - same
parameters - VLBI data 18 - 18 h gt 0 - 0 h !!!


software used - VLBI OCCAM 6.1 (LSM)
DOGS-CS - GPS Bernese 5.0, absolute PCV,
orbits not fixed
gt daily unconstrained normal equations
3
Results 1 station repeatabilities
solution setup terrestrial pole and dUT1
combined, nutation
estimated as one offset and rate for two weeks
radial
east
................
................


north
ties not used
.
.
................
improved repeatability combined (80)
worse repeatability combined (20)
gt GPS and VLBI station positions get more stable
4
Results 2a 1 h EOP RMS
solution setup station positions combined and
stacked for two weeks,
nutation fixed to IAU2000
- IERS C04 IERS 2003 subdaily removed - GPS
orbits free


gt subdaily EOP get more stable
5
Results 2b 1 h EOP, handling of dUT1 (1)
GPS orbits are free gt dUT1 drifts away

2

6
Results 2b 1 h EOP, handling of dUT1 (2)
GPS drift removed

2

7
Results 2b 1 h EOP, handling of dUT1 (3)
GPS drift quadratic part removed

2

8
Results 2c 1 h EOP, wavelet analysis of dUT1
GPS dUT1
combined dUT1
VLBI dUT1
data
periods 14 1.5 days


periods 1.5 0.1 days
gt if GPS orbits are not fixed low dUT1
frequency domain from VLBI, high one from GPS
(but terrestrial pole is dominated by GPS)
9
Results 3 tropospheric zenith delay
Onsala tropospheric wet zenith delay
Zenith delay offsets between GPS and VLBI


gt excellent agreement for temporal variations,
combination problematic due to high
correlation between offset and height
(offsets between solution and Saastamoinen model!)
10
Results 4 tropospheric gradients
North gradient of Fairbanks comparing different
local tie setups
single technique solutions
all ties used
all ties used except for FAIRBANKS


gt good agreement, strong correlation with
horizontal station positions (gt very
suitable to detect tie problems)
11
Summary 1 parameters to be combined (options)


pwlf piece wise linear continuous functions
12
Summary 2 remaining questions


13
18-18 h 1
Adding VLBI 18 - 18 h to 0 - 0 h EOP in normal
equations ( setting them equal!) not the same
period of time gt physically not the same effect
gt distortions? this should be the more
problematic, the more the real EOP signal is not
linear


14
18-18 h 2 Correlation between terrestrial pole
and nutation (1)
24 h, terrestrial pole offset and rate (offset
at mean epoch vs. at 0 h epoch)
Corr (terr. offset / terr. rate) 0


Corr (terr. offset / terr. rate) NOT 0
15
18-18 h 2 Correlation between terrestrial pole
and nutation (2)
24 h, nutation offset, terrestrial pole offset
and rate (terr. pole offset only vs. terr. pole
offset and rate)
Corr (nut. offset / terr. offset) 0


Corr (terr. offset / terr. rate) 0 Corr (nut.
offset / terr. rate) NOT 0 Corr (nut. offset /
terr. offset) 0
16
18-18 h 2 Correlation between terrestrial pole
and nutation (3)
24 h, nutation offset and terrestrial
pole, (terr. pole offset at mean epoch vs. at 0 h
epoch)
Corr (terr. offset / terr. rate) 0 Corr (nut.
offset / terr. rate) NOT 0 Corr (nut. offset /
terr. offset) 0


Corr (terr. offset / terr. rate) NOT 0 Corr
(nut. offset / terr. rate) NOT 0 gt Corr (nut.
offset / terr. offset) NOT 0 at all
17
Conclusions
- to combine complementary and redundant
information from different techniques
helps to - complement strengths and
weaknesses of techniques - detect biases
between techniques / technique dependent
problems gt stabilise and improve solutions
- for this, apriori models and
parameterisation must be homogenized
- VLBI 18 - 18 h might turn out to be
problematical combining normal
equations with EOP (especially with EOP rates)
- further unclear / difficult issues are
- weighting between techniques - terrestrial
connection between techniques (local ties) -
offsets in tropospheric parameters w.r.t. theory


18
Acknowledgement
Thanks to MICHAEL SCHMIDT of DGFI for the dUT1
wavelet analysis pics !!!