Investigation of site-dependent GPS errors and monument stability using a short-baseline network of braced monuments

1
Investigation of site-dependent GPS errors and
monument stability using a short-baseline network
of braced monuments
Emma Hill, Jim Davis, Pedro Elosegui, Brian
Wernicke, Nathan Niemi, and Eric Malikowski
2
Introduction
REPO-REP2 10 meters REPO-REP3 100
meters REPO-REP4 1000 meters
All monuments are deep-braced except REP2
3
East time series
RMS about a model of the seasonal cycle 0.06 to
0.15 mm
Annual cycle amplitudes REP2 0.22 0.01
mm REP4 0.34 0.01 mm
RMS for a zero-length baseline (SLI4-SLID) 0.03
mm
4
North time series
RMS about a model of the seasonal cycle 0.06 to
0.16 mm
Annual cycle amplitudes REPO 0.13 0.01 mm
RMS for a zero-length baseline (SLI4-SLID) 0.03
mm
5
Up time series
RMS about a model of the seasonal cycle 0.14 to
0.51 mm
RMS for a zero-length baseline (SLI4-SLID) 0.08
mm
6
Monument motion
Baseline length difference for REP2-REP3
Maximum velocity estimate for any baseline
component 0.23 0.02 mm/yr
Previously estimated random walk noise for the
BARGEN sites 1 mm/yr1/2 (Williams et al.,
2004)
REP2 (shallow-braced) behaves similarly to REPO
and REP3 (deep-braced)
7
Seasonal cycles
GPS Multipath
Estimation of multipath from SNR for station REPO
(technique from Bilich 2006)
GPS Temperature (scaled)
Comparison of GPS and (scaled) temperature data
at REP2
8
Summary

• The time series are remarkably stable, with RMS
  values at the level of microns (even for the
  shallow-braced monument).
• Our results appear to limit the level of monument
  wander at these sites to lt0.1 mm.
• The time series do have some structure,
  particularly in the form of clear seasonal
  signals. Multipath may be a cause.

9
Thanks!