T13G-01
Active deformation in Western Turkey: new GPS observations and models
How the continents deform remains a matter of debate. One view postulates that continental deforming zones are comprised of a limited numbers of rigid (elastic) microplates. If true, the surface motion can then be described by the relative rotation of blocks, and strain should be localized along the major faults separating the blocks. An alternative view is that the deformation at depth is distributed over wide areas, can be modelled by a viscous flow responding to boundary conditions applied on it and gravitational potential energy gradients related to variations in topography, and the surface strain simply reflects this deformation. Western Turkey is a region of crustal extension, part of the Nubia/Eurasia plate boundary. Its kinematics is often modelled by the relative motion of a small number of rigid blocks (Nyst & Thatcher, 2005, Reilinger et al., 2006). However, until now, the limited number of GPS velocity vectors available has prevented a detailed examination of which is the more appropriate description. We present a new geodetic velocity field including ~100 sites from the longitude the Central Anatolian plateau to the Aegean coast, derived from a combination of campaigns carried out between 1997 and 2006, and continuous GPS operating since 2003, which we use to test the different models. While the kinematics of the area can be correctly modelled by a block model, a good fit to the velocity field requires blocks with sizes smaller than 100 km and still fails to adequately predict the strain rate observed within blocks . Alternatively, we test an approach where the lithosphere is modelled as a thin viscous sheet, responding to the gravitational potentiel energy contrast between the high plateau of eastern Turkey to the east and the subduction along the Hellenic trench in the southwest. The simplistic model has only one free parameter (the force applied by the subducting oceanic lithosphere on the Aegean ), but provides a good agreement with the observed GPS velocities and correctly predict the increase of extension rate from central Anatolia towards the Aegean sea. This results therefore favours a model where the surface deformation is driven by the flow of the whole lithosphere responding to the forces acting on it.
T13G-02
Computation of 3-D static stress in the lithosphere from topographic loads at the surface and base
We develop the mathematical theory for semi-analytically calculating 3-D stress within an elastic plate from loads at its surface and base. We are primarily concerned with stress from surface topography and Moho topography at intermediate wavelengths: longer than those compensated by an elastic halfspace behavior (≥ 102 km) but shorter than those compensated by thin elastic plate flexure (≤ 103 km). We compute the 3-D stress from real surface and Moho topography to estimate a lower bound on absolute stress magnitude in the lithosphere and examine the regional and depth dependence of that magnitude. We also examine the orientation of the principal stresses induced from topography and find that at the front of mountain ranges, the maximum principal stress direction transitions from vertically oriented to horizontally oriented over a few hundred km. We compare regional calculated stress with patterns of seismicity and stress moment tensors for large events.
T13G-03
Preliminary Results from SAFOD Phase 3: Implications for the state of stress and shear localization in and near the San Andreas Fault at depth in central California
Strain localization along the San Andreas fault system in central California appears to result from both a thermally-weak lower crust and upper mantle (reflecting northward migration of the Mendocino triple junction and its associated slab window) and a fault zone in the upper brittle crust that is distinctly weaker than the surrounding crust. Geophysical logs and cuttings analyses from SAFOD Phase 2 (completed in 2005) revealed the San Andreas Fault Zone at approximately 2.7 km depth to be relatively broad (about 250 m), with several discrete, localized zones only 2-3 m wide with very low P- and S-wave velocities and low resistivity. Since 2005, fault creep at two of these localized zones has deformed the casing and thus demonstrates that these zones are actively creeping faults. During SAFOD Phase 3, continuous cores were obtained across these two actively creeping faults. Another core was obtained near the geologic boundary between the Salinian terrane (Pacific plate) and Great Valley/Franciscan terrane (North American plate). Each set of cores reveal zones of profound strain localization and probable weakening. These include ultracataclasites, highly-foliated shear zones (some containing veined serpentine) and intervals that appear to be cohesionless, compacted fault gouges which are likely composed of minerals with low frictional strength. No evidence of significantly elevated fluid pressure is observed within the fault zone. Information about the state of stress in the fault zone and adjacent crust comes from observations and modeling of wellbore failures, direct measurements of the magnitude of the least principal stress and the direction of stress-induced shear wave velocity anisotropy. Observations made after rotary drilling through the fault in 2005 indicate that the San Andreas is a weak fault imbedded in a strong crust. These observations made within about 100 m of the active fault zone at 2.7 km include i) stress orientations that are nearly perpendicular to the strike of the San Andreas, ii) very small differences in the magnitudes of the three principal stresses, and iii) magnitudes of all three principal stresses that are significantly above lithostatic. In contrast, high stress differential stresses in the crust outside the fault were observed in the SAFOD pilot hole at a distance of 1.8 km from the San Andreas, consistent with Byerlee's law and hydrostatic fluid pressure. At the time of this writing, additional geophysical logs are about to be obtained to extend this information directly into and across the zones of active deformation at depth. http://safod.icdp-online.org
T13G-04 INVITED
Stress evolution and dynamics of the lithosphere from 2-D and 3-D numerical models of long-term tectonic processes.
The processes that generate stress in the lithosphere are incompletely understood. Whereas it is obvious that lithospheric deformation (and topography) is ultimately caused by cooling of the Earth from the time of formation, it is less clear how lithospheric deformation is coupled to mantle flow and how this affect stresses. Part of this is due to the somewhat complicated rheology of the lithosphere, which varies from brittle (elastoplastic) to ductile (viscous). In addition, vertical layering of the lithosphere may give rise to instabilities which affect its dynamics and stress evolution in a non-trivial manner. Obtaining a better insight in these processes thus requires numerical tools that can model the mantle-lithosphere system in a self-consistent manner (i.e. in a single computational domain) including topographic effects (i.e. free surface) and viscoelastoplastic rheologies. I have recently developed 2-D and 3-D numerical tools that incorporate the above mentioned features. Here I focus on a number of case studies to illustrate how differences in rheology and boundary conditions alter the dynamics and in particular the stress evolution of the lithosphere. Instabilities such as bending or buckling of compressed lithosphere reduce the average stress („structural weakening"). Viscoelasticity results in time- dependencies, which are particularly pronounced in highly viscous parts of the lithosphere (e.g. the mantle lithosphere). Strong parts of the lithospere thus don't necessarily have large differential stresses (and earthquakes). The Christmas-tree approximation should therefore be used with care to infer stress levels in the lithosphere. Finally I will illustrate differences in stresses between "kinematically-driven" and "internally-driven" lithospheric- scale deformation models.
T13G-05 INVITED
Role of Plate Coupling and Mantle Wedge Flow in Affecting Stresses in Subduction Zone Upper Plate
We have studied the state of stress in forearc - back arc systems of many subduction zones by examining and, where possible, inverting focal mechanisms of upper plate earthquakes. The study led to two general observations. (1) The margin-normal compressive stress in the forearc region of most subduction zones, excluding the frontal sedimentary prism, is no greater than lithostatic. (2) Wherever data are available, the stress states of the forearc and back arc regions are similar. Because the magnitude of the margin-normal stress as compared to lithostatic stress is controlled by the gravitational force and frictional coupling of the converging plates, the first observation indicates that the shear stress along the plate interface is generally very low. The strength of the subduction fault can be represented by an effective coefficient of friction μ'. Using a model of two converging elastic plates in frictional contact, we estimate the value of μ' to be of the order of 0.03 - 0.05. These μ' values give an average shear stress of about 10-20 MPa along the seismogenic part of the subduction fault. The observed stress drop of a few MPa in subduction earthquakes is therefore a significant fraction of the absolute stress along the fault. However, although subduction faults appear to be weak in general, a large coupling area or unusual roughness of the surface of the subducting plate may induce large compressive stresses in the upper plate. The second observation, that is, the lack of a large stress gradient across the forearc - back arc system, indicates that the basal drag force due to mantle wedge flow is very small. Using a model of slab-driven viscous mantle wedge flow with dislocation-creep olivine rheology, we find that the shear stress in most of the flowing mantle wedge is less than 2 MPa, mainly because of the high temperature. The mantle wedge is thus incapable of exerting significant drag on the base of the overriding lithosphere. If mantle drag is insignificant, the coupling force along the subduction interface must be balanced by far flied forces, most likely at other boundaries of the plate.
T13G-06
Active mantle flow and crustal dynamics in southern California
We present numerical modeling analysis of active upper mantle flow and its role in driving crustal deformation in southern California. The forces driving lithospheric deformation at tectonic plate boundaries can be thought of as the sum from two sources: (1) forces transmitted from the far-field by rigid tectonic plates, and (2) forces created locally at the plate boundary by heterogeneous density distribution. Here we quantify the latter by estimating the stresses acting on the base of the crust caused by density-driven flow of the upper mantle. Anomalous density structure is derived from shear wave velocity models (Yang & Forsyth, 2006) and is used to drive instantaneous incompressible viscous upper mantle flow relative to a fixed crust; this allows isolation of stresses acting on the crust. Comparison of results with the finite element codes Abaqus (commercial) and GALE (community- developed) is good. We find that horizontal tractions range from 0 to ~3 MPa and vertical tractions range between approximately -15 to 15 MPa (negative indicating downward, positive upward); Absolute magnitudes depend on the assumed velocity-density scaling relationship but the overall patterns of flow are more robust. Anomalous density beneath the Transverse Ranges, in particular beneath the San Bernardino Mountains and offshore beneath the Channel Islands, drives convergent horizontal tractions and negative vertical tractions on the base of the crust there. Anomalous buoyancy beneath the southern Walker Lane Belt and anomalous density beneath the southern Great Valley create a small convective cell (the Sierra Nevada "drip"), which promotes extension on the eastern edge of the Sierra Nevada block and subsidence of the Great Valley. Favorable comparison with contemporary crustal thickness, heat flow, and surface strain rate indicates that upper mantle flow plays a very important role in active crustal deformation in southern California and much of the non-ideal behavior of this transform boundary can be attributed to the heterogeneous density distribution-driven upper mantle flow.
T13G-07 INVITED
Mantle-lithosphere coupling and the state of stress of the lithosphere
An understanding of the tectonic stress field is geologically important because it is the agent, which preserves in the crust a memory of past dynamical processes. The ability to model the manifestations of intra-plate stresses, mountain building, rifting, etc., depends on our knowledge of dynamical processes in the Earth's interior over geologically long periods of time. Sources of stress include variations in density and thickness of the lithosphere as well as basal and edge tractions that arise from plate driving forces. Mantle-lithospheric coupling and in particular the transmission of stress from mantle to lithosphere is determined by the rheology of the plates themselves and that of the mantle below. We show past and current results on the relative importance of the difference sources of stress for understanding intraplate stresses. We examine the importance of assumptions about the composition and structure of the lithosphere, including continental roots, and mode of compensation, and present a new global model of lithospheric structure. We also focus on the effects mantle shear tractions that arise from mantle density and viscosity heterogeneity. We compute global mantle flow in the presence of strong lateral viscosity variations using the finite element code CitComS and lithospheric stresses with the finite element package ABAQUS. We find that both mantle and lithospheric heterogeneity are important in determining the state of stress in continents and oceans and the mantle signal is particularly strong in SE Asia and other areas of long- lived subduction. The primary effect of lateral variations in mantle viscosity is to strongly couple continents and deeper mantle, enhancing deformation near continental roots.
T13G-08
Temporal Evolution of Continental Lithospheric Strength in Actively Deforming Regions
It has been agreed for nearly a century that a strong, load-bearing outer layer of the Earth is required to support mountain ranges, transmit stresses to deform active regions, and store elastic strain to generate earthquakes. But the depth and extent of this strong layer remain controversial. We use evidence from post-seismic transient and earthquake cycle (EC) deformation, reservoir loading, glacio-isostatic adjustment (GIA), and lithosphere isostatic adjustment to large surface and subsurface loads (LIA) to infer the distribution of lithospheric strength in the active western US from seismic to steady-state timescales. The nearly perfectly elastic behavior of the Earth's crust and mantle at the timescale of seismic wave propagation evolves to that of a strong ~elastic crust and weak, ductile upper mantle lithosphere at both earthquake cycle (EC, ~1–1000 years) and glacio-isostatic adjustment (GIA, ~1000–10,000 years) timescales. Topography/gravity field correlations indicate lithosphere isostatic adjustment (LIA) on ~1-10 Ma timescales occurs with most lithospheric stress supported by upper crust which overlies a very much weaker ductile substrate. These comparisons suggest the upper mantle lithosphere is weaker than the crust at all timescales longer than seismic. In contrast, the lower crust has a chameleon-like behavior, strong at EC and GIA timescales and weak for LIA and steady-state deformation processes. However, lower crust might take on a third identity in regions of rapid crustal extension or continental collision, where anomalously high temperatures may lead to large-scale ductile flow in a lower crustal layer locally weaker than the upper mantle. Modeling of lithospheric processes in active regions thus cannot be done using a one-size- fits-all prescription of rheological layering (relation between applied stress and deformation as a function of depth) but must be tailored to the timescale and tectonic setting of the process being investigated.