Solid Earth, 9, 1187–1206, 2018 https://doi.org/10.5194/se-9-1187-2018 © Author(s) 2018. This work is distributed under the Creative Commons Attribution 4.0 License. Oblique rifting: the rule, not the exception Sascha Brune1,2, Simon E. Williams3, and R. Dietmar Müller3,4 1GFZ German Research Centre for Geosciences, 14473 Potsdam, Germany 2Institute of Earth and Environmental Science, University of Potsdam, 14476 Potsdam-Golm, Germany 3EarthByte Group, School of Geosciences, University of Sydney, Sydney, New South Wales 2006, Australia 4Sydney Informatics Hub, University of Sydney, Sydney, New South Wales, Australia Correspondence: Sascha Brune ([email protected]) Received: 3 July 2018 – Discussion started: 16 July 2018 Revised: 12 October 2018 – Accepted: 15 October 2018 – Published: 26 October 2018 Abstract. Movements of tectonic plates often induce oblique 1 Introduction deformation at divergent plate boundaries. This is in strik- ing contrast with traditional conceptual models of rifting and The relative motion of Earth’s tectonic plates often causes rifted margin formation, which often assume 2-D deforma- oblique deformation at divergent plate boundaries. This is tion where the rift velocity is oriented perpendicular to the primarily due to the fact that irregularly shaped plate bound- plate boundary. Here we quantify the validity of this assump- aries generally do not align with small-circles of relative tion by analysing the kinematics of major continent-scale rift plate movement and that changes in plate motion addition- systems in a global plate tectonic reconstruction from the on- ally lead to time-dependent plate boundary obliquity (Díaz- set of Pangea breakup until the present day. We evaluate rift Azpiroz et al., 2016; Philippon and Corti, 2016). Tradi- obliquity by joint examination of relative extension velocity tionally, rift evolution and passive margin formation have and local rift trend using the script-based plate reconstruc- been investigated using 2-D conceptual and numerical mod- tion software pyGPlates. Our results show that the global els assuming an alignment of relative plate motion and mean rift obliquity since 230 Ma amounts to 34◦ with a stan- plate boundary normal. These studies yielded major insights dard deviation of 24◦, using the convention that the angle into first-order subsidence patterns (McKenzie, 1978; White, of obliquity is spanned by extension direction and rift trend 1993), described key phases controlling the architecture of normal. We find that more than ∼ 70 % of all rift segments rifted margins (Lavier and Manatschal, 2006; Huismans and exceeded an obliquity of 20◦ demonstrating that oblique rift- Beaumont, 2011; Brune et al., 2016) and provided insight ing should be considered the rule, not the exception. In many into the fault evolution during rifting (Ranero and Pérez- cases, rift obliquity and extension velocity increase during Gussinyé, 2010; Brune et al., 2014; Bayrakci et al., 2016; rift evolution (e.g. Australia-Antarctica, Gulf of California, Naliboff et al., 2017). The applicability of these concepts and South Atlantic, India-Antarctica), which suggests an under- models is often rooted in the assumption that rifts can be un- lying geodynamic correlation via obliquity-dependent rift derstood via plane strain cross sections orthogonal to the rift strength. Oblique rifting produces 3-D stress and strain fields trend and that the direction of extension aligns with the orien- that cannot be accounted for in simplified 2-D plane strain tation of these cross sections. Many rifts and passive margins, analysis. We therefore highlight the importance of 3-D ap- however, involve segments where the extension direction is proaches in modelling, surveying, and interpretation of most not perpendicular to the rift strike such that oblique, non- rift segments on Earth where oblique rifting is the dominant plane strain configurations occur (Sanderson and Marchini, mode of deformation. 1984; Dewey et al., 1998). Oblique rift segments differ from classical orthogonal ex- amples in several major aspects: 1. In contrast to orthogonal rifts, the initial phase of oblique rifting is characterized by segmented en éche- Published by Copernicus Publications on behalf of the European Geosciences Union. 1188 S. Brune et al.: Oblique rifting lon border faults that strike at an angle to the rift trend. tion, while orthogonal rifts are an expression of purely The orientation of these faults is controlled by the inter- buoyancy-driven (poloidal) flow. play of inherited heterogeneities with far-field stresses, whereas diverse modelling studies independently sug- The impact of rift obliquity on the structural architecture of gest a fault orientation that lies in the middle between continental extensional systems varies between natural cases. the rift trend and the extension-orthogonal direction This is mainly due to rift variability in general, which arises (Withjack and Jamison, 1986; Clifton et al., 2000; Corti, from tectonic inheritance (Manatschal et al., 2015; Morley, 2008; Agostini et al., 2009; Brune and Autin, 2013). 2016; Hodge et al., 2018; Phillips et al., 2018), or from along- These large en échelon faults generate pronounced relay strike changes in rheology, crustal configuration, tempera- structures (Fossen and Rotevatn, 2016) which act as a ture, and rift velocity (e.g. Sippel et al., 2017; Molnar et al., major control on fluvial sediment transport (Gawthorpe 2017; Brune et al., 2017a; Mondy et al., 2017). and Leeder, 2000). Oblique rifting in presently active rifts can be easily de- duced by combining the local rift trend and GNSS-based sur- 2. Seismic cross sections at rifted margins are often taken face velocities (Díaz-Azpiroz et al., 2016). Prominent exam- perpendicular to the rift trend, which in most cases is ples are the Main Ethiopian rift (Corti, 2008), the Levant rift also perpendicular to the coastline. Considering that system including the Dead Sea rift (Mart et al., 2005), the faults in oblique rifts do not strike parallel to the rift Gulf of California rift (Atwater and Stock, 1998; Fletcher et trend, seismic profiles will observe a projection of the al., 2007), the Upper Rhine graben (Bertrand et al., 2005), fault surface that features a smaller dip angle than the and the Cenozoic West Antarctic rift (Rossetti et al., 2003; actual 3-D fault, an error that might cascade further into Vignaroli et al., 2015; Granot and Dyment, 2018). Struc- seismic restorations. ture and kinematics of past rift systems have been studied by surveying obliquely rifted margins (Fournier et al., 2004; 3. Besides these structural implications, oblique rifting ap- Lizarralde et al., 2007; Klimke and Franke, 2016; Phethean pears to be a major factor in governing the geody- et al., 2016) and transform continental margins (Basile, 2015; namic evolution of extensional systems. This has been Mericer de Lépinay et al., 2016; Nemcokˇ et al., 2016). shown via several pieces of evidence. (i) Oblique rift- However, quantifying rift obliquity of a specific rift system ing has been inferred to enhance strain localization by through time is more difficult since the syn-rift velocity evo- enabling the formation of pull-apart basins and large- lution needs to be reconstructed from available geophysical offset strike-slip faults, for instance during the Gulf of and geological data sets. Therefore, a global statistical analy- California opening (van Wijk et al., 2017; Umhoefer sis of rift obliquity through geological time has been missing et al., 2018). (ii) Analytical and numerical modelling so far. suggests that the force required to maintain a given rift Here we strive to fill this gap by deducing the first-order velocity is anticorrelated with the rift’s obliquity. The obliquity history of Earth’s major rifts from the onset of reason for this behaviour is that plastic yielding takes Pangea fragmentation to the present day. We first describe place at smaller tectonic force when the extension is the applied methods and data sets, then we focus on major oblique to the rift trend (Brune et al., 2012). (iii) At individual rift systems that lead to the formation of the At- the same extension velocity, oblique rifts deform with lantic and Indian Ocean basins before we assess rift obliquity a certain rift-parallel shear rate, which is balanced by a evolution and average obliquity on a global scale. lower rift-perpendicular extension velocity. This means that oblique segments of a particular rift accommodate lithospheric and crustal thinning at a lower rate than or- 2 Methods and data thogonal segments of the same rift, a difference that ef- fects the thermal configuration and therefore the struc- 2.1 Rift kinematics tural and magmatic evolution of each segment (Montési and Behn, 2007). We quantify extension velocity using the global kinematic plate reconstruction of Müller et al. (2016). This plate model 4. Oblique rifting holds geodynamic implications on integrates the latest syn-rift reconstructions for the South At- the global scale because of its relation to toroidal lantic (Heine et al., 2013), North Atlantic (Hosseinpour et al., plate motion, i.e. vertical axis rotation components of 2013; Barnett-Moore et al., 2016), Australia-Antarctica sep- plate movements and associated strike-slip deforma- aration and India-Antarctica breakup (Williams et al., 2011; tion. Toroidal motion is enigmatic from the perspec- Whittaker et al., 2013; Gibbons et al., 2015), and Gulf of tive of plate-driving forces, because its purely horizon- California opening (McQuarrie and Wernicke, 2005), among tal motion cannot be directly caused by buoyancy forces others (Fig. 1). in Earth’s interior (Lithgow-Bertelloni et al., 1993; Restoration of the relative position of continents prior to Bercovici, 2003). Oblique rifts feature strike-slip veloc- rifting in the aforementioned regional studies is largely based ity components and therefore contribute to toroidal mo- on deriving the amount of syn-rift extension from present- Solid Earth, 9, 1187–1206, 2018 www.solid-earth.net/9/1187/2018/ S. Brune et al.: Oblique rifting 1189 Figure 1.