Geology Reference
In-Depth Information
A
120°
118°
Precise leveling
0
100
Although trilateration arrays can successfully
depict horizontal velocity fields, they have not
been used extensively to examine vertical
deformation at regional scales. Slow vertical
movements can be defined using precise leveling
surveys. The conceptual basis is similar to that
for a trilateration survey. Established benchmarks
and newly defined sites are resurveyed over the
course of several years to decades in order to
define the magnitude and rate of deformation in
the intervening period. Because the absolute
height of a benchmark can be poorly known
with respect to the global reference frame,
relative heights are used to calculate rates of
vertical change in most leveling surveys.
Naturally, the largest deformation signal will
usually be recorded with the longest measure-
ment interval. Thus, precise surveys of railway
lines across the Alps in the first half of the 20th
century provide a baseline against which more
recent deformation can be calculated. On the
other hand, the precision and accuracy of meas-
urement techniques typically improve over time.
Hence, a trade-off exists between the range
of time spanned by a survey and the evolution
of our instrumentation. Greater uncertainties
commonly result from reliance on less accurate,
older surveys. But, because the total signal is
larger for longer intervals of measurement, these
older surveys can sometimes provide excellent
deformation histories, despite the imprecision
of individual measurements.
As with any survey, the precision and accuracy
of the measurements need to be evaluated.
Slope-dependent errors can be assessed by
plotting the tilt (the spatial derivative of the
height changes) versus surface slope (the spatial
derivative of the topography) ( Jackson
et al
.,
1992). Because leveling lines typically comprise
an extensive succession of surveyed sites, errors
associated with each measurement accumulate
and propagate through the entire survey
(Fig. 5.7). With long survey lines, the accumulated
uncertainty due to the imprecision of measure-
ments can become considerably larger than the
measured differential uplift. In such cases, it is
sometimes possible to consider smaller subsets
km
San Andreas
Fault
35°
Trilateration
Array
JPL
34°
Transverse
Ranges
Network
Los Angeles
10
B
Distributed
Strain
0
SW
N
E
-10
Transverse
Ranges
Network
-20
-60
-40
-20
0
2
40
0
Distance perpendicular to San Andreas Fault (km)
Fig. 5.6
San Andreas trilateration array in southern
California.
A. Triangulation network in the vicinity of the
Transverse Ranges of southern California. B. The
Transverse Ranges data (1973-89) show no differential
displacement across the San Andreas Fault, which can
be interpreted as being “locked” in this region. Instead,
strain is occurring across the entire surveyed zone.
A suggestion of flattening at the ends of profiles could
be interpreted as representing the expected sigmoidal
shape of a deformation profile across a locked fault
(see Fig. 5.3). Note that less than 25 mm/yr of relative
plate motion is accommodated by displacements along
the profile. This rate indicates that about half of the
Pacific-North American relative plate motion occurs
well beyond the San Andreas Fault zone. Modified after
Lisowski
et al
. (1991).
emphasizes the potential for stored elastic
energy within this array, whereas the fact that
less than 50% of the relative plate motion occurs
within the array itself indicates that considerable
plate motion must occur along faults and folds
beyond the surveyed area.
