Subsurface Geometry of the San Andreas Fault in Southern and Central
California: A Summary
Abstract
Seismic, potential-field, and earthquake data have been collected and
modeled across the San Andreas fault (SAF) in a number of locations in
southern and central California. In most locations the fault does not
have a vertical or steep dip (>85 deg). In at least 3
locations it has a 2-part dip: below 6- to 9- km depth the fault dips
moderately (between about 40 and 70 deg), and above that depth interval
it is steep. The subsurface geometry of the fault from southeast to
northwest is as follows. In two locations in the Coachella Valley,
spanning a major bend in the SAF, the deeper segment of the SAF dips
moderately northeast and the shallower segment is steep. Through the San
Gorgonio Pass and San Bernardino areas, the SAF continues to dip
northeast (as gently as 37 deg at one location), but the 2-part dip seen
in the Coachella Valley is not resolvable. From Cajon Pass to the
western Mojave Desert, spanning the two LARSE seismic profiles, the SAF
dips steeply and is approximately planar. Around the Big Bend of the
SAF, southwest of Bakerseld, the fault appears to dip moderately
southwest, although this dip is constrained only by potential field
data, and a 2-part dip cannot be resolved. At Parkfield, the fault dips
steeply southwest and is approximately planar. A southwest dip continues
to be observed northwestward into the Santa Cruz Mountains where we
resolve a 2-part dip for the rupture of the 1989 M 6.9 Loma Prieta
earthquake. The deeper segment of the fault dips moderately southwest,
whereas the shallower segment is steep and complex. The change in fault
dip in the depth interval of 6-9 km, that is observed in at least 3
locations along the SAF in southern and central California, may result
in part from a change in rheology in the upper part of the crust.
Rheology changes from elastic at greater depths, where earthquakes
occur, to non-linear at shallower depths, where earthquakes diminish in
number. A steeper dip at shallower depth likely represents a
minimum-work configuration for fault rupture. The approach of the fault
to the free surface may also play a part in this depth-dependent
geometry.