Deformation During Continental BreakUp

At the very slow rates of geological processes, rocks are brittle if they are cold, but they can flow if they are warm in the same way that glass flows to the bottom of ancient window panes. Despite the development of sophisticated computer and analogue models [Hopper and Buck (1996); Brun and Beslier (1996)], the consequences of such flow for large-scale extension have remained controversial. Computer models have to make many idealizing assumptions, and analogue models, which can elegantly illustrate the consequences of particular flow laws, are limited by an inability to reproduce temperature-dependent changes in the flow behaviour during extension. A further complication is that the composition and, in particular, the water content of the lower crust and upper mantle beneath the continents are

Strength (MPa) 0 200 400 600 0 200 400 600

n ill III.

Strength (MPa) 0 200 400 600 0 200 400 600

n ill III.

Fig. 4 Strength profiles for the continental lithosphere at different stages of extension [Perez-Gussinye and Reston (2001)]. Profiles assume that the crust is initially 32 km thick, the temperature at the base of the crust is 515°C, the upper crust deforms like wet quartz, and the mantle deforms like dry olivine. Solid and dashed profiles indicate a range of possible behaviours for the lower crust. Brittle strength increases linearly with depth, while in the lower crust strength decays rapidly with depth due to increasing temperatures — the 'jelly sandwich' model [Jackson (2002)]. (a) Initial conditions. (b) Strength profile when the whole crust is brittle, which occurs when the crust has stretched by a factor of 3.6 for the solid profile and 6.1 for the dashed profile.

Fig. 4 Strength profiles for the continental lithosphere at different stages of extension [Perez-Gussinye and Reston (2001)]. Profiles assume that the crust is initially 32 km thick, the temperature at the base of the crust is 515°C, the upper crust deforms like wet quartz, and the mantle deforms like dry olivine. Solid and dashed profiles indicate a range of possible behaviours for the lower crust. Brittle strength increases linearly with depth, while in the lower crust strength decays rapidly with depth due to increasing temperatures — the 'jelly sandwich' model [Jackson (2002)]. (a) Initial conditions. (b) Strength profile when the whole crust is brittle, which occurs when the crust has stretched by a factor of 3.6 for the solid profile and 6.1 for the dashed profile.

poorly known. A key issue is the extent to which extension in the upper crust is decoupled from deeper deformation by a weak layer in the lower crust. Such a weak layer is expected on the basis of extrapolation of flow laws based on short-time-scale laboratory measurements to geological time-scales (Fig. 4). These laws predict that, for typical temperatures at the base of the crust of ca. 500°C, the upper crust and uppermost mantle are brittle but the lower crust deforms by flow, and that faults will dip steeply in the upper crust and flatten at depth. However, a fierce debate has been raging over the past few years regarding the validity of this picture [Jackson (2002)].

Whether or not a weak layer is present in the lower crust, flow laws predict that once the crust has been extended by a factor of 3-5, the entire crust becomes brittle (Fig. 4). Once this happens, the behaviour should be more predictable, since the strength of rocks varies little with composition when they are cool enough to be brittle. However, even in the absence of magmatism, a further complication arises in the last stages of continental break-up. If the mantle temperature is not unusually high and the rifting is not unusually rapid, by this time the crust lies beneath ca. 2 km of water and the temperature at the base of the crust has cooled significantly below

500° C. Sea water penetrates the entire crust through faults and fissures and comes into contact with mantle rocks. Olivine, the predominant mineral in the mantle, reacts with water at temperatures below 500° C to produce a weak mineral called serpentine. Laboratory studies have shown that once 10-15% serpentine is present, the strength of mantle rocks drops abruptly, so we might expect to see evidence for a weak layer at the top of the mantle under these conditions.

Seismic studies of highly extended crust at magma-poor rifted margins have imaged faults that flatten significantly with depth and merge with 'detachment' faults with very shallow dips [Reston et al. (1996)]. These detachment faults play a key role in the deformation immediately preceding continental break-up (Fig. 5). In the region of deep drilling west of Iberia, restoration of motion along such faults results in a crustal section only 7-10 km thick. Structures resulting from the extension which reduced the crustal thickness to 7-10 km from the initial thickness of ca. 30 km generally are not resolved [Whitmarsh et al. (2001)]. Seismic observations of such regions rarely resolve basement structures less than a few hundred metres across. Observations from fragments of rifted margins which have been lifted onshore by mountain-building processes, such as in the

Fig. 5 (a) An interpretation of a seismic-reflection profile from the final stages of continental break-up west of Iberia [Whitmarsh et al. (2000); Manatschal et al. (2001)]. LD and HD are detachment faults which flatten at depth; HHD is a 'rolling-hinge' fault. (b) A similar detachment fault now exhumed above sea level in the eastern Swiss Alps. The Err detachment is highlighted by the snow cover and separates granite (G) below from schist and gneiss (S), overlain by dolomite (D), a sedimentary rock deposited before the main phase of rifting began [Manatschal and Nievergelt (1997)].

Fig. 5 (a) An interpretation of a seismic-reflection profile from the final stages of continental break-up west of Iberia [Whitmarsh et al. (2000); Manatschal et al. (2001)]. LD and HD are detachment faults which flatten at depth; HHD is a 'rolling-hinge' fault. (b) A similar detachment fault now exhumed above sea level in the eastern Swiss Alps. The Err detachment is highlighted by the snow cover and separates granite (G) below from schist and gneiss (S), overlain by dolomite (D), a sedimentary rock deposited before the main phase of rifting began [Manatschal and Nievergelt (1997)].

eastern Swiss Alps, where structures can be mapped on scales from a few centimetres to a few kilometres, fill in an important gap in horizontal scale between seismic and borehole observations. Similar styles of faulting have been observed within these fragments [Whitmarsh et al. (2001); Manatschal and Nievergelt (1997); Fig. 5), though unraveling the subsequent compres-sional deformation can be challenging.

A puzzling observation from drilling off west Iberia was that, where mantle rocks formed the basement, they were generally overlain by a layer of fractured rocks called breccia. Also, in the Swiss Alps, large near-horizontal expanses of exposed mantle rocks appear to represent fault surfaces, and this observation has led to the suggestion that regions of flat basement observed west of Iberia may also represent fault surfaces (Fig. 5). Computer models have shown that if faults weaken as motion along them proceeds and they form by extension of a relatively thin brittle layer, faults with an initial steep dip may rotate close to horizontal and acquire almost unlimited offset at these low angles [Lavier et al. (1999)]. These conditions are satisfied in the last stages of continental break-up, where a thin brittle layer overlies hot mantle rocks and faults are lubricated by serpentine minerals. Fault rotation allows large expanses of mantle rocks to be exposed with very low relief, giving a neat match with a variety of observations.

Was this article helpful?

0 0

Post a comment