ALBERT

Description: We used analogue models to study the fault evolution produced by extension through a heterogeneous crust. In the experiments, the heterogeneous crust consisted of a gently dipping silicone layer surrounded by brittle material. The viscous silicone level simulates a weak, upper crustal nappe stack that formed during a previous phase of shortening. X-ray scanner facilities allowed us to acquire 3D images of the experimental models at regular time invervals and hence to study the fault pattern development and the location of the main depocenters during rifting. The experimental results show that the inherited weak nappe stack acts as a decollement and localizes deformation. In the early stages of extension a system of conjugate high-angle normal faults initiates close to the upper tip of the gently dipping silicone layer near the free surface and propagates upwards, resulting in an initial symmetrical graben configuration. Further extension results in (1) a progressive asymmetry of the rifted zone, due to migration of its right margin down the nappe, (2) a shift of the main depocentre downward along the decollement, and (3) the simultaneous activity of several normal faults within the rifted zone. When the pre-existing silicone layer is oblique to the extension, the normal faults develop in an en echelon array, with a strike intermediate between the azimuth of the gently dipping silicone layer and the extension direction. The experiments also show how rheological differences between areas with potential intracrustal weak layers and adjacent domains without decollement level can lead to significant differences in fault pattern, dimension and orientation of the rifted zone. Complete asymmetry of a rift and switches in fault dip direction between adjacent domains can be explained by the presence of pre-existing upper crustal heterogeneities.

Description: The Gulf of Corinth is a young (1 Ma) active rift currently extending N00, which displays significant contrasts in structural style along strike. A possible explanation for these variations is the presence of the Phyllades nappe in the basement of the western part of the Gulf. Previous 2D thermo-mechanical models have shown that a strong strength contrast between this metamorphic unit and the rest of the basement can explain the kinematics and the spacing of the faults in the western part. The rift, however, displays a wide variety of 3D features (e.g., en echelon faulting, N30 transverse normal faults) that cannot be taken into account using 2D modelling. To obtain 3D insights into the role of an inherited dipping weakness zone, analogue (sand and PDMS) experiments based on the results of the 2D numerical thermo-mechanical model have been performed. The analogue models show that a 30{degrees} discrepancy between the dipping direction of the weak nappe and the direction of extension leads to the formation of en echelon and N30 striking normal faults as observed in the Gulf of Corinth. However, fault spacing and graben width completely misfit both the data and the results of the thermo-mechanical models on which the analogue experiments were based. In order to understand those differences, numerical mechanical benchmarks of the analogue experiments have been run to test different factors (3D lateral displacements, values of the elastic parameters and bottom boundary conditions) that could have affected the dynamics of the analogue model. This approach highlights, for our case study, that the misfits are mostly related to the lack of isostatic compensation at the base of the analogue experiments.

Description: Initiation, geometry and mechanics of brittle faulting in exhuming metamorphic rocks are discussed on the basis of a synthesis of field observations and tectonic studies carried out over the last decade in the northern Cycladic islands. The investigated rocks have been exhumed in metamorphic domes partly thanks to extensional detachments that can be nicely observed in Andros, Tinos and Mykonos. The ductile to brittle transition of the rocks from the footwall of the detachments during Aegean post-orogenic extension was accompanied by the development of asymmetric sets of meso-scale low-angle normal faults (LANFs) depending on the distance to the detachments and the degree of strain localization, then by conjugate sets of high-angle normal faults. This suggests that rocks became progressively stiffer and isotropic and deformation more and more coaxial during exhumation and localization of regional shearing onto the more brittle detachments. Most low-angle normal faults result from the reactivation of precursory ductile or semi-brittle shear zones; like their precursors, they often initiate between or at the tips of boudins of metabasites or marbles embedded within weaker metapelites, emphasizing the role of boudinage as an efficient localizing factor. Some LANFs are however newly formed, which questions the underlying mechanics, and more generally rupture mechanisms in anisotropic rocks. The kinematics and the mechanics of the brittle detachments are also discussed in the light of recent field and modeling studies, with reference to the significance of paleostress reconstructions in anisotropic metamorphic rocks.

Description: After giving a complete analytical solution for the strain softening model associated to Mohr-Coulomb non associated elasto-plastic flow rule (MC-model), the paper demonstrates that this rheology possesses a finite limit load which allows solving for strength drop as a boundary value problem. The MC-model produces a non-dimensional strength drop, which depends on three parameters: the orientation of the shear band versus the least principal stress axis outside the band α 0 , the peak friction angle and the dilatation angle . The maximum reduction of strength obtained with that strain-softening model is on the order of the confining stress p 0 . For this weakest regime, the effective friction of the shear band drops from μ ini = 0.85 at peak to μ ss = 0.64 at the end of the softening phase. In this model, which considers thick shear bands, the weakest regime is not obtained for an orientation corresponding to the exact Coulomb orientation. Instead, the orientation of the weakest shear zone systematically deviates from the coulomb orientation by an angle, which rises with its internal friction angle. The characteristic shear strain needed to achieve steady state is quantified semi analytically and in the range of parameters valid for Earth, this strain is found to be of the order of 7–8%. These numbers are typical of what is observed in the laboratory, which give us confidence on that MC-model is a good and probably the simplest model to localize strain in numerical codes aimed at modeling the brittle part of the Earth.

Description: We analyse Bouguer anomaly data and previously published Moho depths estimated from receiver functions in order to determine the amount of isostatic compensation or uncompensation of the Rif topography in northern Morocco. We use Moho depth variations extracted from receiver function analyses to predict synthetic Bouguer anomalies that are then compared to observed Bouguer anomaly. We find that Moho depth variations due to isostatic compensation of topographic and/or intracrustal loads do not match Moho depth estimates obtained from receiver function analyses. The isostatic misfit map evidences excess crustal root as large as 10 km in the western part of the study area, whereas a ‘missing’ crustal root of ~5 km appears east of 4.3°E. This excess root/missing topography correlates with the presence of a dense mantle lid, the noticeable southwestward drift of the Western Rif area, and with a current surface uplift. We propose that a delaminated mantle lid progressively detaching westward or southwestward from the overlying crust is responsible for viscous flow of the ductile lower crust beneath the Rif area. This gives rise to isostatic uplift and westward drift due to viscous coupling at the upper/lower crust boundary. At the same time, the presence of this dense sinking mantle lid causes a negative dynamic topography, which explains why the observed topography is too low compared to the crustal thickness.

Description: Geodetic observations across most of the major strike-slip faults show an interseismic strain rate, which presents a sharp localization of elastic shear strain in the fault vicinity (20–60 km). The screw dislocation model of Savage & Burford is commonly used to fit these geodetic data and to retrieve the far field velocity and the locking depth. This model is very popular because of its inherent simplicity to derive fault slip rates, and mostly because it predicts locking depths, which are of the same order of magnitude as the base of the seismogenic zone (5–20 km). A first issue with the screw dislocation model is that localization is paradoxically introduced by imposing a step function in the velocity field at a depth where the crust is otherwise recognized to behave following a viscous rheology. A second issue with this model is that it is not consistent with the rheological model of the crust that is valid for both post-seismic (1–10 yr), interseismic (several thousand years) and long-term geodynamic (several thousand to several million years) timescales. Here we use numerical models to study how alternative and more geologically realistic boundary conditions and rheological structures can lead to the localization of elastic strain at the Earth's surface during the interseismic period. We find that simple elastic models resembling the Savage & Burford model but driven by far field plate velocity are inefficient at localizing strain unless this driving velocity is transmitted by a rigid indentor. We also find that models including a weak viscous heterogeneity beneath the fault zone are able to produce appropriate localization of the deformation near the fault. This alternative class of model is shown to be pertinent in regards to the boundary conditions and geological observations along exhumed ductile strike-slip shear zone.

Description: We investigate the putative Pliocene–Quaternary removal of mantle lithosphere from beneath the southern Sierra Nevada region using a synthesis of subsidence data from the Great Valley, and geomorphic relations across the Sierra Nevada. These findings are used to test the results and predictions of thermomechanical modeling of the lithosphere removal process that is specific to the Sierra Nevada, as presented in an accompanying paper referenced here as Part I. Our most successful thermomechanical model and the observational data that it explains are further bundled into an integrated physiographic evolution–geodynamic model for the three-dimensional epeirogenic deformation field that has affected mainly the southern Sierra Nevada–San Joaquin Basin region as a result of underlying mantle lithosphere removal. The coupled Sierra Nevada mountain range and Great Valley basin are recognized as a relatively rigid block (Sierra Nevada microplate) moving within the San Andreas–Walker Lane dextral plate juncture system. Our analysis recognizes that the Sierra Nevada possessed kilometer-scale local and regional paleotopographic relief, and that the Great Valley forearc basin possessed comparable structural relief on its principal stratigraphic horizons, both dating back to the end of Cretaceous time. Such ancient paleorelief must be accounted for in considering late Cenozoic components of uplift and subsidence across the microplate. We further recognize that Cenozoic rock and surface uplift must be considered from the perspectives of both local epeirogeny driven by mantle lithosphere removal, and regional far-field–forced epeirogeny driven by plate tectonics and regional upper-mantle buoyancy structure. Stratigraphic relations of Upper Cretaceous and lower Cenozoic marine strata lying on northern and southern Sierra Nevada basement provide evidence for near kilometer-scale rock uplift in the Cenozoic. Such uplift is likely to have possessed positive, and then superposed negative (subsidence) stages of relief generation, rendering net regional rock and surface uplift. Accounting for ancient paleorelief and far-field–driven regional uplift leaves a residual pattern whereby ~1200 m of southeastern Sierra crest rock and similar surface uplift, and ~700 m of spatially and temporally linked tectonic subsidence in the southern Great Valley were required in the late Cenozoic by mantle lithosphere removal. These values are close to the predictions of our modeling, but application of the model results to the observed geology is complicated by spatial and temporal variations in the regional tectonics that probably instigated mantle lithosphere removal, as well as spatial and temporal variations in the observed uplift and subsidence patterns. Considerable focus is given to these spatial-temporal variation patterns, which are interpreted to reflect a complex three-dimensional pattern resulting from the progressive removal of mantle lithosphere from beneath the region, as well as its epeirogenic expressions. The most significant factor is strong evidence that mantle lithosphere removal was first driven by an east-to-west pattern of delamination in late Miocene–Pliocene time, and then rapidly transitioned to a south-to-north pattern of delamination in the Quaternary.

Description: The putative Pliocene–Quaternary removal of mantle lithosphere from beneath the southern Sierra Nevada region (California, USA) is investigated by the iteration of thermal-mechanical models that incorporate and are tested against a range of data that are geologically observable, including rock uplift and basin subsidence data, structural and compositional data on crustal architecture, and a synthesis of seismic data that image lower crust–upper mantle structure of the region. The primary focus is testing model results with rock uplift and basin subsidence data. The initial state of our models recognizes that (1) the sub–Sierra Nevada batholith mantle lithosphere, including a substantial thickness of eclogitic cumulates that were produced during high magma flux arc activity, termed arclogite, was cooled to a conductive geotherm by amagmatic flat slab subduction at the end of the Cretaceous; and (2) the gravitationally metastable mantle lithosphere was thermally mobilized from beneath in the Neogene by the opening of an underlying slab window. Based on a detailed synthesis of appropriate rheologies of the multiphase system, a preferred class of models correctly predicts (1) the ca. 10 Ma inception of the Sierra Nevada microplate due to a lithospheric separation event along the eastern Sierra Nevada region as a result of the mobilization of the mantle lithosphere as a Rayleigh-Taylor instability; and (2) the subsequent delamination of the arclogite root of the Sierra Nevada batholith that appears to be in progress. Our preferred model also predicts focused rock uplift and basin subsidence resulting from delamination, both of which are anomalous to uplift and subsidence patterns of all other regions of the microplate. The rheology of the Great Valley crust is found to control rock uplift patterns across the Sierra Nevada, and tectonic subsidence in the Tulare Basin of the Great Valley. The Tulare Basin is uniquely situated over the region where the principal residual arclogite root remains attached to batholithic crust. The anomalous rock uplift and tectonic subsidence data are best satisfied by modeling a bulk rheology for the Great Valley crust that is similar to that of the Sierra Nevada batholith. These results are consistent with a recent synthesis of basement core and geophysical data showing that much of the Great Valley basement consists of the western Early Cretaceous zones of the Sierra Nevada batholith. The existence of this batholithic domain within the Great Valley subsurface is also in agreement with recent seismic data that resolve additional residual arclogite root materials along the base of the crust of this region.

Description: Along the western border of the Sierra Nevada microplate, the San Andreas fault (California, United States) is comprised of three segments. Two (north and south segments) are locked and support large earthquakes (e.g., the M 7.7 1906 San Francisco and the M 7.8 1857 Fort Tejon earthquakes), while the central segment, from Parkfield to San Juan Bautista, is creeping. Based on mechanical models, we show that the late Pliocene–Quaternary convective removal (delamination) of the southern Sierra Nevada mantle lithosphere and associated uplift of the Sierra Nevada Mountains causes the Great Valley upper crust to deform by flexure and buckling. Additional three-dimensional flexural models imply that the local flexural bulge overlaps with the creeping segment of the fault system, while geological observations indicate that the local weakening of the San Andreas fault started at the same time that the Sierra Nevada started its recent phase of uplift. We argue that bending stresses promote lithostatic pore pressure to occur in the depth range of 7–15 km, causing the effective strength of the fault to vanish, and locally favoring creep. Our results suggest for the first time that earthquake cycles along a major plate boundary may be influenced by convective instabilities in the adjacent upper mantle.