The developing asymmetry of rifting and continental breakup to form rifted margins has been much debated, as has the formation, mechanics and role of extensional detachments. Bespoke 3D seismic reflection data across the Galicia margin, west of Spain, image in unprecedented detail an asymmetric detachment (the S reflector). Mapping S in 3D reveals its surface is corrugated, proving that the overlying crustal blocks slipped on S surface during the rifting. Crucially, the 3D data show that the corrugations on S perfectly match the corrugations observed on the present-day block-bounding faults, demonstrating that S is a composite surface, comprising the juxtaposed rotated roots of block-bounding faults as in a rolling hinge system with each new fault propagation moving rifting oceanward; changes in the orientation of the corrugations record the same oceanward migration. However, in contrast to previous rolling hinge models, the slip of the crustal blocks on S occurred at angles as low as ~20°, requiring that S was unusually weak, consistent with the hydration of the underlying mantle

SYNOPSISNine phases of deformation, D1-D9, are recognised in the Lewisian complex of Torridon. D1-D5 pre-date the Scourie dykes and are thus of Scourian age, whilst D6-D9 post-date the dykes and belong to the Laxfordian.The N.-S. to NE.-SW. S3 foliation which is dominant in the north of the area has been affected by NW.-SE. large-scale (F3A) and subsequent small-scale (F4) folds and related axial-planar foliation (S4). S4 was originally regarded as Laxfordian by Sutton (in Sutton and Watson 1951), but has been assigned here to the late Scourian or Inverian. In D5 both large and small scale F5 folds were produced with SE.-plunging axes, affecting sub-horizontal belts of S4. L5 is co-axial with L4.The earliest Laxfordian structure S6 is usually parallel or sub-parallel to dyke margins. In D7 were formed F7 folds, S7 axial-planar foliation and L7 lineation which increase in intensity towards the south. S7 is co-planar with S4 and S6, and L7 is co-axial with L4. The D8 structures are restricted to narrow NW.-SE. belts in which F8 small-scale folds and L8 lineations were formed. The last penetrative structure is the S9 mylonite banding affecting certain NW.-SE. belts and is co-planar with S4 and S7.Correlations are made with other areas in the southern Laxfordian belt. C O N T E N T S Introduction Lithological units Structural relationships of dykes and gneisses Structural history A. The pre-dyke structure (i) The D3 and pre-D3 structures (ii) The D4 structures (iii) The D5 structures B. The post-dyke structure (i) Hot shears in the dykes (ii) The D6 structures in the dykes (iii) The D7 structures in the dykes (iv) The D8 structures (v) The D9 structures C. Geometry of the large-scale structures Conclusions

SummaryContinental hyperextension during magma-poor rifting at the Deep Galicia Margin is characterised by a complex pattern of faulting, thin continental fault blocks, and the serpentinisation, with local exhumation, of mantle peridotites along the S-reflector, interpreted as a detachment surface. In order to understand fully the evolution of these features, it is important to image seismically the structure and to model the velocity structure to the greatest resolution possible. Travel-time tomography models have revealed the longwavelength velocity structure of this hyperextended domain, but are often insufficient to match accurately the short-wavelength structure observed in reflection seismic imaging. Here we demonstrate the application of two-dimensional (2D) time-domain acoustic full-waveform inversion to deep water seismic data collected at the Deep Galicia Margin, in order to attain a

SIRS,Beach, Coward and Graham (1974) in their recent paper on the structure of the 'Laxford Front' have criticised our interpretation (Park & Cresswell 1972, 1973 of the structural relationships between the Scourie dykes and the host gneisses. We feel that the authors have misrepresented our views and have also presented an oversimplified picture of the geological relationships in question.We argued that the lateral variation in geometry of the dykes (eg. from discordance to concordance, thick to thin, regular to irregular, etc.) was only partly due to the effects of subsequent deformation as had formerly been thought and that there was a significant original variation in geometry for which we tried to find an explanation.The criticism of Beach et al. (op. cit) is in two parts: (a) They claim that our statement "concordance and discordance (of dykes) are to a large extent primary and not the result of deformation" is contradicted by the observation that concordant bodies are more deformed than discordant.We would refute this criticism on semantic grounds: what we are saying is that dykes intruded concordantly subsequently become more deformed, in general, than those inintruded discordantly. There is no contradiction here: one could disbelieve our evidence but hardly quarrel with our logic! (b) Their second criticism is more significant: that we "fail, apparently, to see the significance of deformation in the modification and/or obliteration of discordant structures". In our paper, where we present several field sketches and diagrams in addition to a detailed discussion of the structural relationships, it is clear that we do in fact recognise the role of deformation in modifying the dykes although we were mainly concerned with trying to establish the nature and extentof the original variation before deformation. We would admit that there is room for disagreement as to the regional extent or proportion of original as compared with deformational variation, because of the ambiguous nature of much of the evidence. However, in the area in question the evidence in many places is quite adequate to demonstrate pre-Laxfordian variation in trend as well as thickness. Moreover our observations over a large area of the Lewisian have revealed numerous examples of variation in geometry of dykes without any trace of corresponding deformation in the host gneisses or, often, in the dykes themselves. There can be no question therefore that there is considerable primary variation in geometry regionally.We state in our paper: "these figures (9a-c) demonstrate that the dykes in this area (ie. Scourie-Laxford) are post-ISi (Fi of Beach et al. op. cit) and generally discordant although with local concordance accompanied sometimes by abrupt changes in direction Scott.

<p>We investigate the structures of hyper-extended continental crust and the 3D nature of the development of syn-rift fault networks at the Galicia margin, West of Spain, based on observations from a 3D multi-channel seismic reflection dataset acquired in 2013. This seismic volume provides, for the first time, 3D high-resolution imaging of a fault network geometry above a detachment fault (The “S reflector”) in the distal setting of a continental margin. The Galicia margin is sediment-starved, magma-poor and salt-free, thus providing optimal observations of the structures through seismic data.</p> <p>We use the 3D data to observe the geometries of the faults, to analyse the fault heaves at different levels of the litho-stratigraphic sequence (i.e. at the top of the crystalline basement, at the top of the pre-rift/early syn-rift sediments and at the top of the syn-kinematic sediments), and to make a stratigraphic analysis to constrain the dynamics and the kinematics of fault activity within the successive half-grabens.</p> <p>Our 3D interpretations demonstrate that the continental crust thins to zero during the rifting by the simultaneous development of initially individual fault planes, which progressively link with adjacent faults to form a network of active faults. The linked roots of the faults altogether form the surface of the S at depth, and allow the oceanward propagation of the detachment fault during the rifting. The faults throughout the network remained active and progressively rotated with further extension, until their deactivation when they acquired an angle of ~30°. Whereupon, a new network of active, initially isolated, faults developed and linked one step (~10 km) oceanward. The system repeats until the break-up of the continental crust, resulting in the progressive focussing of the locus of the extension toward the ocean, where the continental crust is the thinnest. </p> <p>Given the similitude of the features observed at the Galicia margin with other magma poor continental margins, we expect that most margins worldwide might have formed following similar processes, thus representing a paradigm shift in the global understanding of late fault network development at rifted margins during continental break-up.</p>