Attribution: You must attribute the work in the manner specified by the author or licensor ( but no in any way that suggests that they endorse you or your use of the work).Noncommercial ‒ you may not use this work for commercial purpose.No Derivative works ‒ You may not alter, transform, or build upon this work.Sharing ‒ Individual scientists are hereby granted permission, without fees or further requests to GSA, to use a single figure, a single table, and/or a brief paragraph of text in other subsequent works and to make unlimited photo copies of items in this journal for noncommercial use in classrooms to further education and science.

For nearly three decades, debate has swirled around low-angle normal faults (dipping < 30°), since they were first proposed to represent a fundamental mode of crustal deformation (Wernicke, 1981). A major issue has been whether or not such faults move in the brittle regime (Wernicke, 1995; Axen, 2004; Collettini and Sibson, 2001): if so, why they are largely absent from the earthquake record, and if not, do such faults exist at all (Anders and Christie-Blick, 1994)?

In extensional environments and near the surface, the maximum principal stress (σ1) should be vertical (Anderson, 1951). The physics of friction should then limit the fault dips so that for coefficients of friction (μ) of 0.6–0.85 thought typical of many rocks, e.g., Byerlee's Law (Byerlee, 1978), optimally oriented normal faults dip 65–70°. Faults do not have to be optimally oriented and can slip at higher shear stress if the fault zone is weaker than surrounding material; in these cases, continued frictional sliding is easier than rupturing a new optimally-oriented fault. However, there are limits (Fig. 1). For a given coefficient of friction, the ratio of the maximum to minimum effective stress (σ1P)/(σ3P), where P is pore pressure and σ3 is the minimum compressive stress, is a simple function of fault orientation (Jaeger and Cook, 1976). This ratio reaches ∞ at finite dips (e.g., Collettini and Sibson, 2001); a relative of that ratio is plotted in Figure 1. Since σ1 is controlled by the overburden weight, only decreasing σ3 can increase the shear stress on faults. Decreasing σ3P to the tensile strength of the fault results in hydrofracture, which sets a maximum differential stress and a limit to fault misorientation. For normal faults obeying Byerlee's Law, that “lockup” constraint should preclude dips less than 31–40°, consistent with most observations (Collettini and Sibson, 2001; Jackson and White, 1989). While one or two counterexamples may show slightly lower dip (Abers et al., 1997; Abers, 2001), perhaps indicating lower μ, this constraint has proven difficult to reconcile with geological evidence for low-angle normal faults.

In Italy, many destructive earthquakes occur in the Apennines where active normal faults lie close to large populations, and several deadly earthquakes have occurred over the last century (Boschi, 2000), most recently on 6 April 2009 near L'Aquila, with a magnitude of 6.3. Nearly all of these earthquakes occur on normal faults dipping at moderate or high angles to the WSW at depths of 10 km or less (Chiaraluce et al. 2007). However, seismic imaging in part of the Apennines shows that these faults are underlain by a ENE-dipping low-angle fault dipping at ~20°, the Alto Tiberina fault (Boncio et al. 2000). Seismicity there is confined to the hanging wall, including several large (magnitude >5.5) earthquakes, with a few micro-earthquakes lying on the fault; most focal mechanisms show fault planes at high angles to the detachment (Chiaraluce et al., 2007).

In a fascinating new discovery, Hreinsdottir and Bennett (p. 683 in this issue of Geology) show from global positioning system (GPS) data that the Alto Tiberina is creeping steadily at 2.4 ± 0.3 mm/a. If true, this is a remarkable finding, because it indicates that low-angle normal faults can move in the presumably brittle regime without exhibiting stick-slip behavior. Such behavior may go some distance to resolving the ongoing debates, in that it should produce geologic features associated with brittle failure along the fault without generating earthquakes.

A similar situation may be occurring in the western Gulf of Corinth, Greece. Extension there approaches 15 mm/a in a north-south direction, localized to an ~15-km-wide basin (Briole et al., 2000). Microseismicity illuminates a surface dipping north ~10° at depths of 6–10 km that has been interpreted as a detachment (Rietbrock et al., 1996). However, surface faults and faults mapped by seismic reflection show steeper dips of 30–50°, and much microseismicity lies above the low-angle surface (Bernard et al., 2006; Bell et al., 2008). The nature of the subhorizontal feature remains controversial; it has been proposed that it is the brittle-ductile transition (Hatzfeld et al., 2000) and that no motion takes place on the low-angle surface (McNeill et al., 2005). While it is not universally interpreted as a fault, the overall geometry and seismicity patterns resemble those in Italy. Like the Alto Tiberina fault, a shallow-dipping surface exists with seismicity above it showing normal faulting along high-angle planes.

How can these observations be reconciled with fault mechanics, and the global observation that earthquakes do not occur on faults dipping 20° or less? One possibility is that large-offset fault zones such as this tend to be relatively weak, with rheologies dominated by materials that steadily creep (velocity-strengthening). Recent experimental studies of clay-rich fault gouges show μ = 0.2–0.3, much lower than predicted by Byerlee's Law (Ikari et al., 2009); these experiments also show the natural gouges tend to be velocity strengthening, and so would favor steady creep over stick-slip behavior. Similarly, recent experiments on serpentinites and talc show low coefficients of friction (Moore and Lockner, 2008). Talc-rich gouge zones may be common in the Apennines and have been found to dominate exposed older analogs of the Alto Tiberina fault, where hydrothermal circulation of silica-rich fluids through dolomites produces talc + calcite (Collettini et al., 2009). Talc-rich serpentinites have also been drilled on the exposed fault zone of the Moresby Seamount fault in the Woodlark Rift (Taylor and Huchon, 2002), with similarly low μ of 0.21–0.3 (Kopf et al., 2003). In the latter case, the fault dips ~32° but generates large (Mw ~6.1) earthquakes (Abers et al., 1997), so the fault is not creeping. It is not clear what the difference may be, but the static friction coefficient is a different physical property than those that control stick-slip behavior (Scholz, 2002), so they are not necessarily correlated.

Overall, these observations suggest that many apparently low-angle normal faults may be active in the brittle regime, but some may be creeping. Regardless, the observation of steady creep along a shallow normal fault (Hreinsdottir and Bennett, 2009) and experimental evidence for low (0.2–0.35) frictional coefficients with velocity-strengthening properties suggest that low-angle faults could slip at relatively shallow dips without requiring highly elevated pore pressure or stress rotation.