Rifting and subsidence following lithospheric removal in continental back arcs

It has been suggested that post-orogenic lithospheric removal in continental back arcs promotes extension and surface subsidence. However, the surface response of this process and its primary difference from “classical” back-arc opening have remained uncertain. Here, the back-arc extension process with varying continental mantle lithosphere thickness and thermal heterogeneities is studied by using thermomechanical subduction experiments. The experiments illustrate that models with only slab retreat result in minor surface subsidence and extension in the back-arc region. Alternatively, there is notable extension due to the slab retreat and a localized high-temperature zone in the back arc with uniform lithospheric thickness. Models with advecting mantle (after lithospheric removal) in the extending back arc predict rifting (stretching factor b > 2) and surface subsidence (>1.5 km) in the center of the basin. The results of this work suggest that lithospheric removal may be an important trigger for continental back-arc development rather than slab retreat alone causing lithospheric extension and subsidence. The findings help explain rift formation and subsidence in the Aegean Sea–west Anatolia, and possibly other Mediterranean back arcs, such as the Alboran Sea and the Pannonian Basin. INTRODUCTION Retreating ocean-continent subduction systems are associated with thin and hot continental back-arc lithosphere of the overriding plate. It has been recognized that the thinning of the back arc occurs due to the dynamic slab pull exerted in the crust by the oceanic plate that causes trench retreat, surface extension, sedimentary basin formation, and high heat flow (Le Pichon et al., 1981; Faccenna et al., 1996). However, the geodynamics of this extension and the primary factor for the uplift and/ or subsidence in the back arcs are not well understood—for example, the dominant role of induced mantle convection preceding rifting (Toksöz and Hsui, 1978) versus vertical stresses due to the slab-pull forcing that produces dynamic subsidence (Mitrovica et al., 1989; Husson, 2006). In addition to this, there is still controversy about why some back-arc regions are anomalously topographically high, such as the Canadian Cordillera (Hyndman and Currie, 2011), or tectonic subsidence (1.5–2.5 km) has occurred in the Mediterranean back arcs during the past 15–20 m.y. (e.g., Aegean Sea, Pannonian Basin, Alboran Basin) (Le Pichon et al., 1981; Royden et al., 1983; Watts et al., 1993). In this work, a new class of numerical models is presented to account for the back-arc basin subsidence. Many geological observations suggest that back-arc extension may have been initiated or promoted by lithospheric thinning after post-orogenic thickening. Two major geodynamic hypothesis have been suggested as lithospheric removal mechanisms (Göğüs and Pysklywec, 2008): (1) mantle lithosphere delamination (Bird, 1979); and (2) Rayleigh-Taylor type convective removal (Houseman et al., 1981). Conceptual removal models have been put forward to explain rapid subsidence, anomalous heating, exhumation of high-pressure rocks, tectonic mode switching (from compression to extension), and high stretching factors at the Pannonian Basin (Horvarth, 1993; Houseman and Gemmer, 2007), the Tyrrhenian Sea (Channell and Mareschal, 1989; Faccenna et al., 1996), the Alboran Basin (Docherty and Banda, 1995; Platt, 2007), and Aegean Sea– west Anatolia (Dewey, 1988; Seyitoğlu and Scott, 1996). Accordingly, all of these back arcs contain a Neotethyan suture that marks the Alpine continental collision and the shortened, thickened lithosphere. In spite of the potential geodynamic feedback between the lithospheric removal and slab retreat–roll-back driven back-arc extension, proposed models (references given above) have addressed each of these mechanisms in isolation. In this work, thermomechanical numerical models are used to investigate the surface response of hot mantle upwelling (after lithospheric removal) in the retreating subduction zone as a coupled geodynamic process to explain the subsidence and rifting in continental back arcs. A series of computational geodynamic experiments is conducted to test the variation of surface topography and crustal thickness for different back-arc lithospheric thicknesses and thermal fields. The experiments are carried out in two-dimensional procedures for high-resolution purposes, while model predictions are carefully interpreted in the context of three-dimensional geodynamic processes. As a case study, the numerical results are interpreted in the context of the last 20 m.y. of geodynamic evolution of the Aegean Sea–west Anatolia back arc, where the cause of extension is still not well understood (Figs. 1A and 1B). Although the extension is mainly controlled by the *E-mail: [email protected]; [email protected]. GEOLOGY, January 2015; v. 43; no. 1; p. 1–4; Data Repository item 2015014 | doi:10.1130/G36305.1 | Published online XX Month 2014 © 2014 eological Society of A erica. For permission to copy, contact [email protected]. Figure 1. A: Generalized geological map of the central Aegean–west Anatolia region (modified from Okay and Satır, 2000; Jolivet and Brun, 2010). Map from GeoMapApp (www.geomapapp.org). Blue and black arrows (in inset) indicate GPS-derived plate motions (Reilinger et al., 2006). B: Lithospheric-scale cross sections along the Aegean and western Anatolia based on seismological interpretations (Wortel and Spakman, 2000). C: Proposed convective removal–orogenic collapse model (Seyitoğlu and Scott, 1996; Aldanmaz et al., 2000). D: Slab retreat–roll-back model (Meulenkamp et al., 1988; Okay and Satır, 2000) for the Aegean–western Anatolia.