Back-arc basins are enigmatic geological features, which form in the overriding plate in overall convergent plate tectonic settings on the concave side of the arc. Back-arc basins occur in two different tectonic regimes: ocean-ocean subduction and ocean-continent subduction. In the former regime, the back-arc basin is commonly found in between a remnant volcanic arc and the active volcanic arc, which were separated during back-arc opening. In ocean-ocean subduction settings extension is often localised along narrow rift segments or along clear defined spreading ridges. Some examples include the Mariana Arc, the Izu-Bonin Arc, the Tonga Arc, the Kermadec Arc, the New Hebrides Arc, the Scotia Arc and the Lesser Antilles Arc. In the latter regime, the back-arc basin is located on the continental lithosphere. This often results in more diffuse extension over a large area. Examples include the Kuril Arc, the Japan Arc, the Ryukyu Arc, the Banda Arc, the Hellenic Arc, the Calabrian Arc, the Betic-Rif Arc and the Carpathian Arc.
Back-arc extension involves regressive (oceanward) hinge-line (or trench axis) migration of the subducting plate (i.e. rollback). Whether this regressive movement is the cause or the effect of back-arc extension remains a debate, but presently the majority of geoscientists favour the former view. In this sense, rollback is driven by the negative buoyancy of the subducting slab compared to the asthenosphere [Elsasser 1971; Molnar and Atwater 1978]. Hence, rollback provides space along the plate boundary. This space is filled by the overriding plate, which passively follows the retreating hinge of the subducting slab.
Slab rollback is primarily driven by its negative buoyancy compared to the asthenosphere, which is often related to age of the oceanic lithosphere (i.e. the older, the colder, the denser). However, the actual retreat velocity depends on a variety of factors, including negative buoyancy of the slab, tearability of the slab, ability of asthenosphere material underneath the slab to migrate towards the slab corner and viscosity of the asthenosphere.
Figure 1. Tectonic reconstruction of the New Hebrides - Tonga region
Tectonic reconstruction of the New Hebrides - Tonga region (modified and interpreted from Auzende et al. [1988], Pelletier et al. [1993], Hathway [1993] and Schellart et al.(2002a)) at (a) ~ 13 Ma, (b) ~ 9 Ma, (c) 5 Ma and (d) Present. The Indo-Australian plate is fixed. DER = d'Entrcasteaux Ridge, HFZ = Hunter Fracture Zone, NHT = New Hebrides Trench, TT = Tonga Trench, WTP = West Torres Plateau. Arrows indicate direction of arc migration. During opening of the North Fiji Basin, the New Hebrides block has rotated some 40-50° clockwise [Musgrave and Firth 1999], while the Fiji Plateau has rotated some 70-115° anticlockwise [Malahoff et al. 1982]. During opening of the Lau Basin, the Tonga Ridge has rotated ~ 20° clockwise [Sager et al. 1994]. (Click for enlargement)
From geological, structural, geomorphological, magnetic lineation and paleomagnetic data it is clear that different arc - back-arc systems evolve quite differently (Schellart et al. 2002b). For instance, several arcs seems to have formed by radial spreading, resulting in symmetrical arc development with arc-parallel and arc-perpendicular extension in the overriding plate and equal amounts of block rotation on each side of the arc. Examples could be the Hellenic Arc and Aegean Sea, the Ryukyu Arc and Okinawa Trough, and the Carpathian Arc and Pannonian Basin. Another group of arcs seems to have developed due to asymmetrical spreading. Two examples for this type of spreading are the New Hebrides Arc (Schellart et al. 2002a) and the Tonga Arc, located in the southwest Pacific (Figure 1). For each of these two arcs, its back-arc basin has opened up in a wedge shaped manner, with an increasing opening rate from one side of the arc to the other. Development of such arc systems involves asymmetrical back-arc spreading with similar magnitude and sense of rotation along the arc and shearing (with possible related block rotations) at the side of the arc (Figure 2). Other examples of this arc configuration could include the Kuril Arc and Sea of Okhotsk + Kuril Basin, the Japan Arc and Japan Sea, the Calabria – Apennines Arc and Thyrrenian Sea, and rotation of the Corsica-Sardinia block and opening of the Ligurian Sea. Also, the late stage development of both the Hellenic Arc and the Ryukyu Arc could have resulted from asymmetrical spreading.
Figure 2. Development of asymmetrical arcs during asymmetrical rollback
Schematic sketches of development of asymmetrical arc during asymmetrical rollback of subducting slab in plan view (top) and 3D view (bottom). (a) Initial state with rectilinear subduction zone; (b) initiation of rollback with formation of pin points or cusps (points where subducting slab is resisting to roll back); (c) one pin point evolves into transform plate boundary due to formation of vertical tear in subducting slab, while other becomes a hinge point along which the arc rotates; (d) progressive growth of tear and continued rotation around hinge point. Black arrows indicate direction of hinge-line migration. Thick arrows in (d) indicate direction of asthenosphere flow to accommodate slab retreat.
In this paper we will focus on arc - back-arc systems which form by asymmetrical spreading. We will investigate this type of asymmetrical deformation by means of small-scale analogue experiments. In these experiments the model apparatus configuration remains unaltered, but the rheology of the overriding plate is varied systematically for different experiments. We will compare these results briefly with some natural examples of arc - back-arc systems, which display structural features, which could testify to a development resulting from asymmetrical spreading. Finally, the results of these experiments will be discussed and placed in a tectonic context of rollback as a driving mechanism for large-scale extensional features, including Cenozoic extension along the East Asian margin.