Geology Reference
In-Depth Information
techniques, such as radiocarbon with optically
stimulated luminescence, can provide a valuable
check on the consistency of dates. In a trench
along the normal fault along the Wasatch front
(Fig. 6.6), for example, some radiocarbon dates
vary by 400-700 years from their calendar ages.
Although these temporal offsets can be assigned
from the calibration curve for radiocarbon ages
(see http://calib.qub.ac.uk/calib/), it is rea-
ssuring that thermoluminescence dates (Fig. 6.6A)
reinforce the calibrated radiocarbon dates.
In most trenches, the strata that contain
datable material do not consistently coincide
with the strata that record rupture events. In
such cases, dates on strata above and below the
“event horizon” are used to bracket its age. To
reduce the uncertainty in timing as much as
possible, datable material should be sought as
close as possible to the rupture horizon, and the
highest laboratory precision available should be
used to reduce the analytical uncertainty on the
individual ages (Atwater
et al.
, 1991; Sieh
et al.
,
1989). Concerted efforts to provide the best
possible time control are warranted, because
reliable calculations of recurrence intervals are
highly dependent on the quality of the ages
assigned to rupture events.
Studies along the San Andreas Fault clearly
illustrate the importance and utility of high-
precision dating. Because the San Andreas Fault
runs near or through several of the major
population centers in California, and because it
has generated “great earthquakes” (magnitude 8)
in the past, there is great interest in knowing:
(i) how often earthquakes occur along it; (ii) the
length, displacements, and spatial patterns of
past ruptures; and (iii) the recurrence interval
between major earthquakes. During the past
30 years, much effort has gone into paleo-
seismological studies along many parts of this
major plate-bounding fault. Some of the earliest
insights on recurrence intervals of San Andreas
faulting came from trenches dug at Pallett Creek
(55 km from Los Angeles) during Kerry Sieh's
dissertation research (Sieh, 1978). Several
subsequent studies have attempted to improve
on the radiocarbon dating of the faulting
events at this site, where 11 major earthquakes
are recorded. A comparison of the faulting
2000
1800
potential clustering
of earthquakes
1600
1400
1200
1000
Paleoseismic
Chronology
800
600
old date range
new date range
400
1
2
3
4
5
6
7
8
9
10
11
Earthquakes
younger
older
Fig. 6.8
Radiocarbon chronologies of earthquakes
on the San Andreas Fault.
Two generations of radiocarbon dating along the San
Andreas Fault at Pallett Creek, California. As opposed to
the rather regular recurrence intervals suggested by the
1984 chronology, the much higher-precision dates
published subsequently indicate a distinct clustering of
earthquakes (horizontal shaded bands) into groups of
two or three ruptures each that were separated from
each other by a few decades, whereas the clusters
themselves are separated from each other by 150-300
years. The higher-precision dates define a step-like event
versus time curve (dark shading), as opposed to the
smoothly changing curve (no shading) comprising the
less precise dates. Modified after Sieh
et al.
(1989).
chronologies published in 1984 (Sieh, 1984) and
in 1989 using higher-precision dates (Sieh
et al.
,
1989) is illuminating (Fig. 6.8). Based on the
1984 data, the recurrence intervals between
major earthquakes fall into two groups: prior to
about 1100AD, the recurrence interval was
about 100 years, whereas from 1100AD to the
present, the interval lengthened to 160-200
years. Within each grouping, the resolution of
the radiocarbon dates suggests that the
earthquakes may have been approximately
evenly spaced in time. When radiocarbon dates
with higher precision are used to date the
earthquakes (black bars, Fig. 6.8), a rather
different rupture history emerges (Sieh
et al.
,
1989). Rather than being evenly spaced in time,
the earthquakes appear to cluster. Two or three
