The IGS disseminates all data from the global network, globally
consistent GPS orbits, and precise site coordinates. This has
freed regional investigators from having to deploy fiducial
networks or compute their own orbits, and has relieved concern
about changing reference frames as fiducial sites change.
Heflin et al [1994] now report daily global site position
repeatabilities of about 14 mm (
2 parts in 10
). The published
global reference frame is accurate absolutely to about 1 cm. There
is still room for reference frame improvement as GPS moves towards
sub-centimeter global geodesy. Analysts have long noted that orbit
accuracy degrades when the satellites pass through the earth's
shadow and have suspected mis-modeling of GPS thermal radiation
[ Blewitt, 1993]. Bar-Sever et al [1995] uncovered a
possibly root cause: unpredictable yaw variations when the
satellite loses its sun sensor. They proposed inserting a 0.5 deg
yaw bias in the satellite attitude control subsystem, which was
adopted and
instituted by the Air Force in June of 1994. Early results show a
clear improvement in orbit accuracy during eclipse periods.
Monument instability threatens the integrity of the reference frame
and the authenticity of perceived geophysical signals. While of
only modest concern for global measurements, site stability is a
top priority for the SCIGN array in the Los Angeles Basin, which
will feature 5--10 km site spacing and could detect strain rates of
0.2 mm/yr within one year. The SCIGN science team has set a
provisional goal of 0.5 mm for long term monument stability. Such
local effects as soil expansion from rainfall and slow churning can
in principle perturb even robust monuments by millimeters over
months or years [ D. Agnew, personal communication]. Land
subsidence due to intensive water pumping can be on the order of
centimeters per year. Even relatively benign areas may exhibit
nontectonic subsidence at the millimeter level over a period of
years from fluid extraction or as a result of drought [
Sylvester, 1992; Van Hasselt, 1992]. A focused effort is
needed to probe site stability at the submillimeter level required
by dense networks, and GPS may be an effective tool for the job.
Elosegui et al [1993] report measurement precisions of about
0.1 mm on baselines of less than 100 m with just a few hours of
data, consistent with results from JPL in the late 1980s and from
Genrich and Bock [1992]. Long term GPS measurements on
closely spaced monuments could answer some of the questions now
being raised for the Los Angeles array.