Page 203 - Global Tectonics

CONTINENTAL RIFTS AND RIFTED MARGINS 189

model composed of an upper crustal layer and a lower This formulation and a 20-km-thick upper crust lying
mantle layer (Fig. 7.28a). These authors incorporated a above a 40-km-thick lower crust allowed them to inves-

strain-rate softening rheology to model brittle behavior tigate how a mechanically stratified crust infl uenced
and the development of fault-like shear zones. Ductile fault spacing and the distribution of strain during exten-
deformation was modeled using temperature- sion. They found that the ratio of the integrated strength
dependent flow laws that describe dislocation creep in of the upper and lower crust governs the degree of

the crust and mantle. Variations in the strength (effec- strain localization on fault zones. When this ratio is
tive viscosity) of the crust at any given temperature small, such that the lower crust is relatively strong,

and strain rate are defined by material parameters that extension results in widely distributed, densely spaced
are derived from rock physics experiments. The use of faults with a limited amount of slip on each fault. By

several flow laws for rocks with different mineralogies contrast, a large strength ratio between the upper and
and water contents allowed the authors to classify the lower crust, such that the lower crust is very weak,
rheologies as either weak, intermediate, or strong. Vari- causes extension to localize onto relatively few faults
ations in crustal thickness and thermal structure were that accommodate large displacements. In this latter
added to a series of models to examine the interplay case, the large-offset faults dissect the upper crust and
among these parameters and the different rheologies. exhume the lower crust, leading to the formation of
The results show that when crustal thickness is small, metamorphic core complexes (Section 7.3). Wijns et al.
so that no ductile layer develops in the lower crust, (2005) also concluded that secondary factors, such as
deformation occurs mostly in the mantle and the width fault zone weakening and the relative thicknesses of the
of the rift is controlled primarily by the vertical geother- upper and lower crust (Section 7.6.5), determine the
mal gradient (Fig. 7.28b,f). By contrast, when the crustal exact value of the critical ratio that controls the transi-
thickness is large the stress accumulation in the upper tion between localized and delocalized extension.
crust becomes much greater than the stress accumula- The results of Wijns et al. (2005), like those obtained
tion in the upper mantle (Fig. 7.28c,d). In these cases by Behn et al. (2002), suggest that a weak lower crust
the deformation becomes crust-dominated and the promotes the localization of strain into narrow zones
width of the rift is a function of both crustal rheology composed of relatively few faults. This localizing behav-
and the vertical geothermal gradient (Fig. 7.28e,f). ior reflects the ability of a weak lower crust to fl ow and

Figure 7.28e illustrates the effects of the strong, inter- transfer stress into the upper crust, which may control
mediate and weak crustal rheologies on rift morphology the number of fault zones that are allowed to develop.

(half-width). The models predict the same rift half-width This interpretation is consistent with field studies of
for mantle-dominated deformation. However, the transi- deformation and rheology contrasts in ancient lower
tion between mantle- and crust-dominated deformation crust exposed in metamorphic core complexes (e.g.
begins at a slightly larger crustal thickness for the strong Klepeis et al. 2007). It is also consistent with the results
rheology than for the intermediate or weak rheologies. of Montési & Zuber (2003), who showed that for a
In addition, the strong crustal rheology results in a rift brittle layer with strain localizing properties overlying a
half-width for the crust-dominated regime that is ∼1.5 viscous layer, the viscosity of the ductile layer controls
times greater than the value predicted by the intermedi- fault spacing. In addition, a weak lower crust allows
ate rheology and ∼4 times greater than that predicted by fault blocks in the upper crust to rotate, which can
the weak rheology. Figure 7.28f summarizes the com- facilitate the dissection and dismemberment of the
bined effects of crustal thickness, crustal rheology, and upper crust by faulting.
a vertical geothermal gradient on rift half-width. These Lastly, a third numerical model of rifting illustrates
results illustrate that the evolution of strain patterns how the interplay among strain-induced weakening,
during lithospheric stretching is highly sensitive to the layer thickness, and rheological contrasts can infl uence
choice of crustal rheology, especially in situations where deformation patterns in a four-layer model of the litho-
the crust is relatively thick. sphere. Nagel & Buck (2004) constructed a model that
A similar sensitivity to crustal rheology was observed consisted of a 12-km-thick brittle upper crust, a rela-
by Wijns et al. (2005). These authors used a simple two- tively strong 10-km-thick lower crust, a thin (3 km) weak
layer crustal model where a plastic yield law controlled mid-crustal layer, and a 45-km-thick upper mantle (Fig.
brittle behavior below a certain temperature and the 7.29a). The model incorporates temperature-dependent
choice of temperature gradient controlled the transi- power law rheologies that determine viscous behavior
tion from a brittle upper crust into a ductile lower crust. in the crust and mantle. The mantle and upper and