AT&G
- Spotila's active tectonics and geomorphology research
San
Andreas fault transpression
Scope: Although we know quite a lot about
earthquakes and faulting, some extremely simple questions are still unanswered;
for example, how strong are faults? This question has focused on the
debate over the San
Andreas fault's strength over the past decades, inspired by observations of
a lack of a heat-flow anomaly along the fault and nearly fault-normal
stresses. One school of thought holds that the fault is quite weak,
either due to pore pressure or low coefficient of friction (Byerlee's law)
associated with weak fault zone materials. Another school holds that the
fault is strong, but that because it is transpressional (oblique to plate
motion direction), the strain ellipsoid is rotated and this causes the stress
rotations. One way of addressing this issue is by looking at the patterns
of deformation associated with the fault where it is transpressional. In
central California, the obliquity to plate motion is 5 degrees, so that the
strain is partitioned into a minor component of folding and faulting. In southern California,
however, the Big Bend of the San Andreas fault is oblique to plate motion by 27
degrees (wrench-dominated) and should have the long-axis of the strain
ellipsoid oriented vertically in the near-field according to the strain
partitioning model. The stress partitioning model, which considers the
fault weak, instead predicts the fault zone will experience only simple-shear
while the convergence should be confined to the farfield.
To examine this we investigated whether the core of the fault
zone in southern California has experienced major vertical motion.
Because this core is crystalline, we addressed this by looking for recent
exhumation using (U-Th)/He dating. We have found crustal slivers, such as
the Yucaipa Ridge block, trapped within strands of the San Andreas fault zone
in the San Bernardino
Mountains (Spotila et al., 1998; Spotila and Sieh, 2000) that have
experienced exhumation rates over the past million year or so of up to 5 mm/yr
(Spotila et al., 2001). These rates are possible because of the narrow
block width, which translates to an inability to maintain steep relief due to
rapid erosion. These ridges likely maintain a steady-state topography,
and from the photo below it is easy to see how rugged they are as a
result. A location map, age-elevation plot, and slope distribution plot
for Yucaipa Ridge can be found here. These observations suggest that there
is significant vertical strain in the San Andreas fault zone, consistent with
the strain partitioning model. There is a chance that this deformation is
also locally driven, however, given the proximity of a small left-step in the
fault zone at San
Gorgonio Pass. Distinguishing between these possibilities is
important for understanding how faults work.
Our early investigations of the San Bernardino Mountains
isolated some key questions regarding transpressive behavior of the San Andreas
fault in general. Ongoing work has
focused on these, by new field and thermochronologic investigations along
several sections of the fault, including the northernmost San Gabriel Mountains
(J. Buscher's work), the San Emigdio Mountains (N. Niemi's work), and the
Temblor Range (R. Brady's work).
This new work is funded by NSF and should provide valuable new
constraints on the structure and exhumation of ranges along the San Andreas
fault. Stay tuned for results!
Personnel: Ph.D. candidate Jamie Buscher;
collaboration with Ken
Farley (Caltech), Pete
Reiners (Yale), and Doug
Yule (Cal State Northridge), Nathan
Niemi (Caltech), Martha
House (Caltech), Rob Brady
(U. Calgary), Mike Oskin (UNC Chapel
Hill).
Funding: supported by NSF Tectonics program,
award EAR02229628 (11/02-12/05).
Links of interest include Earthscope, a major initiative of NSF
that will go to Congress for a line-item vote in FY2003. This project
includes the deep drilling of the San Andreas fault (SAFOD) which will
address many of the questions above, as well as the Plate Boundary Observatory
(PBO) that will heavily instrument the San Andreas fault. Check out John Hole here at VT,
who's interested in the same problems of the San Andreas fault, but addresses
them from the perspective of seismic imaging. He has a major ongoing
project at Parkfield.
Photo of Yucaipa Ridge in the San Bernardino Mountains,
looking south from San Gorgonio Peak. The ridge is just over 3 km high at
its peak. The Mill Creek strand of the San Andreas fault runs along the
facing side of the ridge (north), and the San Bernardino strand of the San
Andreas fault runs along the southern flank of the ridge. We sampled a
vertical transect along the northern flank of this ridge (facing) for
radiogenic helium dating, the results of which can be found here. The
ages are very young, implying rapid exhumation (~5 mm/yr) for a several hundred
thousand year period during the mid-Pleistocene. This is evidence for
considerable slip partitioning within the San Andreas fault zone due to
transpression.
Here is another photo of Yucaipa Ridge, with the apatite
helium ages illustrated (data from Spotila et al., 2001).
Narrow bedrock slivers occur elsewhere along the highly
oblique section of the San Andreas fault in southern California, including
Portal Ridge (shown below). We
identified this as a potential site for transpressively-driven exhumation,
based on observations at Yucaipa Ridge.
However, the morphology of Portal Ridge is clearly more subdued, and
exhumation rates based on recently determined apatite helium ages are
significantly slower than at Yucaipa Ridge. This implies that the rapid exhumation of crustal slivers
seen in the San Bernardino Mountains does not occur everywhere along the
"Big Bend" of the San Andreas fault, and may be in part driven
locally by the restraining bend at San Gorgonio Pass.
Although Portal Ridge may not be experiencing rapid vertical
strain, Jamie Buscher's Ph.D. research in the Northern San Gabriel Mountains is
finding interesting patterns of deformation associated with San Andreas fault
transpression that have not previously been recognized. The 3-d perspective image below shows
the main crystalline block of the northern San Gabriel Mountains (the white
crest), with the San Andreas fault running along its backside and the Mojave
Desert beyond (to the northeast).
This block has experienced relatively rapid tilting (to the northwest)
on yet-to-be identified structures, based on new apatite helium ages.
We have also synthesized transpression, topography, and
denudation along the entire San Andreas fault. Below is a plot of the obliquity to plate motion of the San
Andreas fault. The convergently
oblique (+), transpressive regime on the south is what is called the "Big
Bend". Note that the majority
of the San Andreas fault is transpressive, and almost none of the fault's
length is actually parallel to relative plate motion. (Figure from Spotila et
al., in review - a paper submitted to a GSA Special Paper on transpression -
this is a copyrighted figure and may not be used without permission.)
Topography seems to mimic this distribution of plate motion
obliquity. Shown below is a graph
of the average elevation along the trace of the San Andreas fault, for both a
20-km and 80-km wide swaths centered on the fault trace. Above this is the graph of plate motion
obliquity vs. fault length, same as shown above. The higher topography along the fault clearly occurs where
the plate motion is most oblique, in southern California. (Figure from Spotila
et al., in review - a paper submitted to a GSA Special Paper on transpression -
this is a copyrighted figure and may not be used without permission.)
The plot below shows the average slope along an 80-km-wide
swath centered on the San Andreas fault trace, binned in 10-km lengths for the
entire fault, versus the plate motion obliquity at each location (positive =
transpressive). There is a very
rough increase in average slope and convergent obliquity, consistent with the
idea that obliquity translates to mountain building and more rugged topography.
(Figure from Spotila et al., in review - a paper submitted to a GSA Special
Paper on transpression - this is a copyrighted figure and may not be used
without permission.)
Exhumation also relates to obliquity. This is a graph of exhumation rate
along the San Andreas fault (within 40 km of the trace) versus plate motion
obliquity. The exhumation rates
are based on individual age determinations for low temperature
thermochronometry all along the fault (based on many studies). Although slow rates of exhumation occur
at all degrees of plate motion obliquity, there is a rough increase in
exhumation rate with increasing plate motion convergence. The King Range is an exception, but
enhanced exhumation there may be driven by a thermal bulge associated with
migration of the triple junction. (Figure from Spotila et al., in review - a
paper submitted to a GSA Special Paper on transpression - this is a copyrighted
figure and may not be used without permission.)
This plot shows exhumation rate along the fault versus
distance from the fault trace, for the entire length of the San Andreas fault.
Again, slow exhumation occurs everywhere, but the highest rates of exhumation
appear to be concentrated in the near-field (i.e. within 10 km) of the San
Andreas fault trace. This is
consistent with models of transpression in which wrench-like vertical
deformation occurs within the fault zone due to transpression, rather than
models that invoke decoupling between simple shear within the fault zone and pure
shear within the adjacent borderlands. (Figure from Spotila et al., in review -
a paper submitted to a GSA Special Paper on transpression - this is a
copyrighted figure and may not be used without permission.)
Comments to: spotila@vt.edu