Geology Reference
In-Depth Information
collected for several days, the uncertainty can
be reduced considerably more by smoothing
the noise inherent in daily measurements.
Continuously operating, permanent GPS receivers
can reduce the uncertainties still further and can
provide a temporal resolution that is impos-sible
to achieve with occasional reoccupation of
GPS sites.
The choice of whether to acquire GPS data
using permanent or mobile GPS sites is some-
times simply a matter of cost or practicality:
campaign mode is cheaper, faster, and more
portable between distant field sites. But the
nature of the scientific problem being addressed
should also dictate the choice. For example,
if the goal is to delineate short-term crustal
deformation before or after an earthquake,
permanent GPS receivers are far superior to
campaign-style acquisitions. Moreover, some of
the most exciting discoveries in seismology
over the last decade have been prompted by
provocative data from continuous GPS monitoring
(Box 5.1). On the other hand, if the goal is to
define crustal deformation across a broad,
rapidly deforming orogen, data acquired in
campaign mode can commonly define multiple
positions to within a few millimeters after two or
three reoccupations of survey sites. For many
GPS projects in active orogens nowadays, a
compromise is reached: mobile receivers are
used to define the broad deformation pattern,
but a few permanent GPS receivers are embedded
within the study area to provide a robust
reference frame, as well as high temporal
resolution.
Four factors contribute to the calculated GPS
position. First, regional deformation is driven
by large-scale plate tectonic movements and
encompasses the relative velocity between
interacting plates. Second, spatially localized
deformation results from individual seismic
events and aseismic deformation. Third, seasonal
changes in groundwater levels can cause the
Earth's surface to inflate or deflate. Fourth,
measurement errors result from variations in
atmospheric conditions, transmission and
reception of the GPS data, and positioning of
the GPS receiver. Obviously, considerable efforts
are typically expended to reduce the measure-
ment error to a minimum. Extraction of a plate-
motion signal or a seismic signal from the data
is then possible, if the strain related to one or
the other is known. In fact, one check on
calculated seismic displacements is to subtract
independently measured or modeled ground
offsets due to seismic phenomenon from the
total displacement to see whether the expected
rates of plate motion represent the remaining
measured motion in the data set. Alternatively,
by subtracting the expected plate motion,
the resultant seismic displacements can be
compared with measured offsets (Larsen and
Reilinger, 1992).
Over the past decade, numerous GPS
campaigns have been conducted across the
Himalayan-Tibetan orogen. Although details are
still lacking in many areas, a stunning pattern of
crustal deformation has emerged from these
studies (Fig. 5.15A). In the far field, the
Indian subcontinent is converging with Siberia
at
∼
40 mm/yr. As can be seen by comparing the
length of the velocity vectors from north and
south of the Himalaya, about half of this con-
vergence is absorbed by the Himalaya. Both GPS
and leveling-line data (Fig. 5.7) suggest that most
(>70%) interseismic shortening is currently
being absorbed in the Greater Himalaya.
This Greater Himalayan deformation stands in
contrast to the observation that, throughout
the Holocene, shortening at some frontal
faults within the Himalayan foreland has
averaged
∼
20 mm/yr (Lavé and Avouac, 2000).
This mismatch between long- and short-term
deformation (deformation in the Greater
Himalaya versus that in the frontal fault system
in the foreland) suggests that most of the current
deformation represents elastic strain that will be
recovered in large earthquakes that rupture
beneath the Lesser Himalaya and out to the
foreland (Bettinelli
et al
., 2006; Cattin and
Avouac, 2000).
Whereas the orthogonal convergence across
the Himalaya is perhaps predictable, the GPS
data depict extraordinary diversity north of the
range (Fig. 5.15A). In the west, almost half of
the convergence is absorbed in the Tien Shan
(Abdrakhmatov
et al
., 1996), which, given its
location some 1000 km from the collision front,
