Page 261 - Global Tectonics

246 CHAPTER 8

eastward-dipping fault surface. In the lower crust, the of the San Andreas Fault, a paradox exists in that heat

two components separate, producing two zones of flow observations (Lachenbruch & Sass, 1992) show no
deformation (Fig. 8.24l). These results illustrate how an frictionally generated heat, so that the fault must slip in
evolving thermal structure resulting from asymmetric response to very low shear stresses.
erosion and exhumation stabilizes the lateral and con- One possible explanation of the high-angle stress
vergent components of oblique collision along a single directions in California is that the San Andreas is an
dipping fault. They also suggest that a partitioning of extremely weak fault that locally reorients the regional
deformation onto separate strike-slip and dip-slip faults stresses (Mount & Suppe, 1987; Zoback et al., 1987;
is favored where thermal weakening is absent. Zoback, 2000). In this interpretation, shear stresses far
from the fault are high and contained by the frictional
strength of the crust, but shear stresses on planes paral-
lel to the “weak” faults of the San Andreas system must
8.7 MEASURING THE be quite low. Consequently, the principal stresses
become reoriented so as to minimize shear stresses on
STRENGTH OF planes parallel to the San Andreas Fault. This requires
a rotation such that the direction of maximum horizon-
TRANSFORMS tal compressive stress (σ 1 ) becomes nearly orthogonal
to the fault if the regional compression direction is at
an angle in excess of 45° to the fault, which occurs at
present. However, if this angle is less than 45°, the
Measures of the strength of continental transforms and maximum horizontal compression is rotated into
large strike-slip faults provide a potentially useful means approximate parallelism with the fault. This latter type
of testing models of continental rheology and evaluat- of rotation may have characterized the San Andreas
ing the driving forces of continental deformation Fault at some time in the past when relative plate
(Section 8.5.1). In many intraplate areas, the long-range motions were different than they are now.
(1000–5000 km) uniformity of stress orientations and This model of a weak continental strike-slip fault

their relative magnitudes inferred from measures of offers one explanation of conflicting geologic and geo-
strain or displacement suggest that plate-driving forces physical data in California. However, alternative inter-
provide the largest component of the total stress fi eld pretations involving a strong or an intermediate-strength
(Zoback, 1992). Models of GPS-derived horizontal San Andreas Fault also have been proposed. These latter
velocities in some regions, such as southern California, models are based on frictional theories of faulting,
tend to support this view (McCaffrey, 2005). However, which suggest that σ 1 rotates to ∼45° from the fault
in other areas, such as the Basin and Range Province trace within a ∼20–30-km-wide zone in the Big Bend
(Section 7.3), stresses caused by lateral variations in region (Scholz, 2000). Scholz (2000) interpreted reports
crustal buoyancy (Section 7.6.3) also appear to contrib- of high σ 1 angles in this area as representing local

ute significantly to the horizontal stress field (Sonder & stresses related to folding instead of regional stresses.
Jones, 1999; Bennett et al., 2003). He also concluded that the presence and sense of the

There have been numerous attempts to evaluate the stress rotation fits predictions of a strong fault rather
strength of the San Andreas Fault using various geo- than a weak one. High fluid pressure (Section 8.6.3) is

logic and geophysical indicators (Zoback et al., 1987; a possible mechanism for decreasing the strength of the
Zoback, 2000). For some fault segments (Fig. 8.25), fault and could explain some rotation of the stresses
stress data suggest that the direction of maximum hor- (Rice, 1992). Alternatively, the strength of the fault and
izontal compression (σ 1 , Section 2.10.1) lies at a high the adjacent crust generally could be much lower than
angle (β) to the fault zone. In central California these predicted by considerations of fault mechanics
angles are as high as β = 85°. In southern California (Hardebeck & Michael, 2004).

they are lower at β = 68° (Townend & Zoback, 2004). These conflicting observations and interpretations
These observations are problematic because classical concerning the strength of large strike-slip faults have
theories of faulting (Section 2.10.2) cannot explain com- yet to be resolved. In the case of the San Andreas Fault,
pression at high angles to a strike-slip fault with such a part of the controversy may be related to different
small component of convergence. Moreover, in the case mechanical behaviors of the creeping versus locked