Model Of Subduction Zone Research Articles

Stress, strain, and B‐type olivine fabric in the fore‐arc mantle: Sensitivity tests using high‐resolution steady‐state subduction zone models

Shear‐wave splitting observations at several subduction systems show trench‐parallel fast directions in the fore‐arc mantle. The presence of B‐type olivine fabric in the mantle wedge may provide an explanation for this pattern of anisotropy under low‐temperature and hydrated conditions. Sensitivity tests are shown that provide insights into the distribution and magnitude of B‐type fabric using two‐dimensional, high‐resolution, kinematic‐dynamic subduction zone models. These models include a wet olivine rheology and a simulated aseismic creep zone in the fore‐arc mantle with parameterized viscous coupling between the slab and mantle wedge. The calculated thermal structure is in reasonable agreement with heat flow observations and Q tomography for small amounts of coupling and moderate amounts of shear heating in the seismogenic and aseismic creep zones. These small amounts of viscous coupling give rise to a slowly flowing fore‐arc mantle with thermal, stress, and strain conditions suitable for significant B‐type fabric development. The minimum shear stress necessary to drive geologically significant flow in the predicted B‐type region is sensitive to the magnitude of shear heating along the aseismic creep zone and ranges from less than 100 to 300 MPa. The minimum time required to generate sufficient finite strain for fabric development in the B‐type region is around 10 Myr depending on the amount of coupling and shear heating. The predicted distribution of olivine fabric is consistent with a pattern of shear wave anisotropy with trench‐parallel fast polarization directions above the fore arc and a rapid rotation to trench‐perpendicular in the arc and back arc.

Geothermal evidence for fluid flow through the gas hydrate stability field off Central Chile-transient flow related to large subduction zone earthquakes?

Fluids are suspected to play a major role in the nucleation and rupture propagation of earthquakes. In Chile, seismological data were previously interpreted to indicate that fluids captured in the fault zone are released periodically during large underthrust earthquakes, leading to post-seismic fluid flow. In central Chile, heat flow derived from the presence of a bottom simulating reflector (BSR) show a smooth trend across the margin. BSR-derived data are in excellent agreement with thermal subduction zone models. Over the young accretionary prism, both BSR-derived and measured surface heat flow support a common trend. Landwards of the backstop, however, measured heat flow triples over a distance of 20–30 km, producing a profound discrepancy to the BSR-derived data. We suggest that this disparity is related to transient flow of warm fluids through the gas hydrate stability zone possibly caused by fluids released after large underthrust earthquakes. Such flow events may inherently affect the distribution of solid gas hydrates between the seafloor and the BSR.

The thermal structure of subduction zones constrained by seismic imaging: Implications for slab dehydration and wedge flow

Thermal models of subduction zones often base their slab–wedge geometry from seismicity at mantle depths and, consequently, cannot be used to evaluate the relationship between seismicity and structure. Here, high-resolution seismic observations from the recent Broadband Experiment Across the Alaska Range (BEAAR) constrain, in a rare instance, the subducting slab geometry and mantle wedge temperature independent of seismicity. Receiver functions reveal that the subducting crust descends less steeply than the Wadati-Benioff Zone. Attenuation tomography of the mantle wedge reveals a high Q and presumably cold region where the slab is less than 80 km deep. To understand these two observations, we generate thermal models that use the improved wedge geometry from receiver functions and that incorporate temperature- and strain-rate-dependent olivine rheology. These calculations show that seismicity within the subducting crust falls in a narrow belt of pressure–temperature conditions, illuminating an effective Clapeyron slope of 0.1 K/MPa at temperatures of 450–750 °C. These conditions typify the breakdown of high-pressure hydrous minerals such as lawsonite and suggest that a single set of dehydration reactions may trigger intermediate-depth seismicity. The models also require that the upper, cold nose of the mantle wedge be isolated from the main flow in the mantle wedge in order to sustain the cold temperatures inferred from the Q tomography. Possibly, sufficient mechanical decoupling occurs at the top of the downgoing slab along a localized shear zone to 80 km depth, considerably deeper than inferred from thrust zone seismicity.

Subduction zone guided waves and the heterogeneity structure of the subducted plate: Intensity anomalies in northern Japan

The subducting Pacific plate acts an efficient waveguide for high‐frequency signals and often produces anomalously large intensity on the eastern seaboard of northern Japan during deep earthquakes. The waveform records in the region of high intensity show a low‐frequency (f < 0.25 Hz) onset for both P and S waves, followed by large, high‐frequency (f > 2 Hz) later arrivals with a long coda. This behavior is not explained by a simple subduction zone model comprising a high‐velocity plate with low attenuation. From the analysis of observed broadband waveforms and numerical simulation of seismic wave propagation in the Pacific subduction zone we demonstrate that the high‐frequency guided waves traveling in the subducting plate arise from the scattering of seismic waves by heterogeneity in plate structure. Our preferred model of the heterogeneity has elongated scatterers parallel to the plate margin described by a von Karmann function with a downdip correlation length of about 10 km and much shorter correlation length of about 0.5 km in thickness. The standard deviation of wave speed fluctuations from the averaged background model is about 2%. This new heterogeneous plate model generates significant scattering of seismic waves with wavelengths shorter than correlation distance in thickness, but low‐frequency waves, with long wavelengths, can easy tunnel through such lamina structure. The result is frequency‐selective propagation characteristics with a faster low‐frequency phase followed by large and high‐frequency signals with very long coda. A low‐wave speed channel effect from the former oceanic crust at the top of the subducting slab is not necessary to explain the observed dispersed signals and the very long high‐frequency coda. Three‐dimensional simulations, using the Earth simulator supercomputer for modeling of high‐frequency seismic wave propagation in the Pacific subduction zone including plate heterogeneity, clearly demonstrate the scattering waveguide effects for high‐frequency seismic waves traveling in the plate. The region of large intensity for the heterogeneous model migrates away from the hypocenter into northern Japan with an elongated zone along the Pacific coast, almost comparable to the observations from deep events in the Pacific plate.

B-type olivine fabric in the mantle wedge: Insights from high-resolution non-Newtonian subduction zone models

Several hypotheses have been proposed to explain trench-parallel shear wave splitting in the mantle wedge of subduction zones. These include 3-D flow effects, parallel melt filled cracks, and B-type olivine fabric. We predict the distribution of B-type and other fabrics with high-resolution thermal and stress models of subduction zones. A composite viscous rheology is used that incorporates wet diffusion creep, wet dislocation creep, and a stress-limiting approximated Peierls creep law. Rheological parameters for dislocation and diffusion creep are based on deformation experiments. The effects of variable viscous coupling between the slab and mantle are explored using kinematic boundary conditions that change along the top of the slab. Two end-member models are observed, which depend on the depth of partial viscous coupling between the slab and mantle: (1) deep partial coupling which gives rise to cold, high-stress conditions in the forearc mantle and high-temperature, low-stress conditions in the arc and back-arc mantle (2) full viscous coupling at the base of a 40 km conducting lid which is characterized by high temperature and low stress. The case with deep partial coupling, which produces a better match with attenuation tomography and heat flow, shows a large region with suitable conditions for B-type fabric in the forearc mantle and a rapid transition toward the back-arc to conditions that are more suitable for A-, C-, or E-type fabrics. Full viscous coupling gives rise to high-temperature, low-stress conditions unsuitable for B-type fabric. Our modeling predicts that the best candidates for regions with B-type fabric are the forearc mantle and a cold 10–15 km layer above the slab.

Effects of variable thermal conductivity on the mineralogy of subducting slabs and implications for mechanisms of deep earthquakes

Thermal and thermo-kinetic models of subduction zones have been based, in general, on a constant value of thermal conductivity (e.g., 3.138 W/(m K)) for solving the conservation of energy equation. In order to make such models more realistic, we have used variable thermal conductivities ( k) that depend on pressure, temperature, and mineralogy (olivine, wadsleyite, or ringwoodite), as determined recently by Xu et al. [Xu, Y., Shankland, T.J., Linhardt, S., Rubie, D.C., Langenhorst, F., Klasinski, K., 2004. Thermal diffusivity and conductivity of olivine, wadsleyite and ringwoodite to 20 GPa and 1373 K. Phys. Earth Planet. Inter. 143–144, 321–336]. We have used these variable conductivities to model a variety of slabs with thermal parameters (vertical subduction rate × age of lithosphere at the trench) ranging between 3500 and 17,000 km. Our models show that when radiative transfer is included, k could plausibly be increased by as much as 2 W/(m K) relative to a standard constant lithospheric value (3.138 W/(m K); [Stein, C.A., Stein, S., (1992) A model for the global variation in oceanic depth and heat flow with lithospheric age. Nature 359, 123–129.], causing temperatures in the interiors of the slabs to increase by 50–100 °C. In turn, the volume of metastable olivine in the slabs is reduced by as much as 20–30%, with the maximum depth of olivine metastability decreasing by 30–50 km. In general, the maximum depths of olivine metastability are 160–230 km less than the depths of the deepest seismicity. Even in the extreme case of Tonga, metastable olivine falls short of the deepest earthquakes by 140 km in our model. Even taking the various uncertainties into account, these results indicate that deep-focus earthquakes occur in regions already transformed to wadsleyite or ringwoodite and must therefore be caused by a mechanism other than transformational faulting.

A case for hot slab surface temperatures in numerical viscous flow models of subduction zones with an improved fault zone parameterization

Viscous flow models play a fundamental role in our understanding of the dynamics and thermal structure of subduction zones. For example, the relatively cool slabs produced in 2D viscous flow models have been central to the argument against slab and sediment melting at subduction zones except in special cases, such as near slab edges. Because flow models have such an important role in our insight to the subduction process, it is imperative to understand how various assumptions in the models affect their results. The fault zone is a key part of subduction models, but the effects of different parameterizations on the results have been largely ignored. The two main fault parameterizations used in the literature are weak (low viscosity) nodes imposed at the fault surface and an imposed rigid overlying plate. Each formulation has pros and cons, but their only physical justification is that they keep the overlying mechanical lithosphere from deforming in unrealistic ways. Because the assigned thickness of the imposed plate or the dimensions of the weak zone governs the amount of ablation that occurs in the corner and the temperature of the material advected to the subducting slab, different arbitrary implementations can result in significantly different flow patterns and thermal structures. A fault zone parameterization with a rigid plate defined by temperature and strain rate rather than by depth or other geometry could ameliorate the predicament. This formulation removes much of the arbitrariness from the modeling, as it is based on a physical process. This formulation has the advantages of restricting viscous deformation in the overriding plate without limiting processes such as ablation by imposed boundary conditions. Model results with this parameterization show hotter slab surface temperatures than previously shown in the literature. The results also agree as well as or better than other formulations with geophysical observations such as heat flow, seismic velocity, and seismic attenuation. The calculated higher slab surface temperatures could lead to a reappreciation of sediment melting at subduction zones without requiring an atypical tectonic environment.

Serpentine and the subduction zone water cycle

This study explores a chemo-thermo-dynamic subduction zone model that solves for slab dehydration during subduction. We investigate how changes in the incoming plate's hydration and thermal structure may effect the efficiency of sub-arc water release from sediments, crust, and serpentinized mantle. We find that serpentinized lithospheric mantle may not only be an important fluid source to trigger arc melting but is also an efficient ‘transport-lithology’ to recycle chemically bound water into the deeper mantle. In fact, an old slab may remain sufficiently cold during subduction to retain up to 40% of its initial ‘mantle’ water at 8 GPa (∼240-km depth) after serpentine transforms to higher pressure hydrous phase A. Furthermore, deep water recycling at subduction zones is parameterized in terms of slab age and speed. Coupling this parameterization to a parameterized mantle convection evolution model allows us to calculate the mantle-surface geologic water cycle throughout the Earth's history. We find that the present-day Earth mantle may be highly outgassed containing only a small fraction of the Earth's water, which would mostly be recycled water from the exosphere.

Upper-mantle reflectors: modelling of seismic wavefield characteristics and tectonic implications

SUMMARY Deep seismic experiments on continental lithosphere generally reveal marked reflectivity from structures in the crust and a significant decrease in reflectivity from the upper mantle. However, reflected and refracted energy from mantle lithosphere are observed in both near-normal incidence and wide-angle data. The origin of the reflective structures is a matter of debate. Hypotheses include remnant subduction zones, shear zones, fluids and seismic anisotropy. Through analytical and numerical modelling studies, including full wavefield modelling, we investigate seismic characteristic signatures generated from a variety of geologically plausible models. We have found that both upper-mantle shear zones of reduced velocity and density and remnant subduction slabs containing high-density eclogites may contain sufficient seismic impedance contrasts to normal mantle peridotites to generate near-normal incidence reflectivity. Wide-angle energy originates from subduction slabs containing either high- or lowvelocity eclogites, whereas intermediate-velocity eclogites are unlikely to produce significant wide-angle phases. In general, energy of seismic phases originating from upper-mantle zones of anomalous seismic velocities and densities is significantly increased if homogeneous zones are replaced by zones of inhomogeneous petrophysical properties resulting from constructive interference. Maximum wavefield anomalies are generated from sub Moho dipping slabs of incomplete transformation of low-velocity/low-density crustal material to high-velocity/highdensity eclogites. Localized shear zones generated in mantle peridotite generally do not produce significant wide-angle energy. Only if highly inhomogeneous structures containing material of marked (ca 10 per cent) velocity and density reduction are present, may shear zones be observed in wide-angle data. Analyses of two specific deep-seismic data sets (MONA LISA data) from the North Sea and (BABEL data) from the Baltic Sea, show good agreement between observations and modelling results for dipping remnant subduction slabs containing small-scale inhomogeneities associated with incomplete transformation. Our modelling results improve our possibilities of distinguishing between two often contrasting tectonic interpretations for dipping upper-mantle seismic reflectors, the remnant subduction and shear zone models.

Thermal evolution of the Earth as recorded by komatiites

Komatiites are rare ultramafic lavas that were produced most commonly during the Archean and Early Proterozoic and less frequently in the Phanerozoic. These magmas provide a record of the thermal and chemical characteristics of the upper mantle through time. The most widely cited interpretation is that komatiites were produced in a plume environment and record high mantle temperatures and deep melting pressures. The decline in their abundance from the Archean to the Phanerozoic has been interpreted as primary evidence for secular cooling (up to 500°C) of the mantle. In the last decade new evidence from petrology, geochemistry and field investigations has reopened the question of the conditions of mantle melting preserved by komatiites. An alternative proposal has been rekindled: that komatiites are produced by hydrous melting at shallow mantle depths in a subduction environment. This alternative interpretation predicts that the Archean mantle was only slightly (∼100°C) hotter than at present and implicates subduction as a process that operated in the Archean. Many thermal evolution and chemical differentiation models of the young Earth use the plume origin of komatiites as a central theme in their model. Therefore, this controversy over the mechanism of komatiite generation has the potential to modify widely accepted views of the Archean Earth and its subsequent evolution. This paper briefly reviews some of the pros and cons of the plume and subduction zone models and recounts other hypotheses that have been proposed for komatiites. We suggest critical tests that will improve our understanding of komatiites and allow us to better integrate the story recorded in komatiites into our view of early Earth evolution.

Finite element model predictions of static deformation from dislocation sources in a subduction zone: Sensitivities to homogeneous, isotropic, Poisson‐solid, and half‐space assumptions

Dislocation models can simulate static deformation caused by slip along a fault. These models usually take the form of a dislocation embedded in a homogeneous, isotropic, Poisson‐solid half‐space (HIPSHS). However, the widely accepted HIPSHS assumptions poorly approximate subduction zone systems of converging oceanic and continental crust. This study uses three‐dimensional finite element models (FEMs) that allow for any combination (including none) of the HIPSHS assumptions to compute synthetic Green's functions for displacement. Using the1995 Mw = 8.0 Jalisco‐Colima, Mexico, subduction zone earthquake and associated measurements from a nearby GPS array as an example, FEM‐generated synthetic Green's functions are combined with standard linear inverse methods to estimate dislocation distributions along the subduction interface. Loading a forward HIPSHS model with dislocation distributions, estimated from FEMs that sequentially relax the HIPSHS assumptions, yields the sensitivity of predicted displacements to each of the HIPSHS assumptions. For the subduction zone models tested and the specific field situation considered, sensitivities to the individual Poisson‐solid, isotropy, and homogeneity assumptions can be substantially greater than GPS measurement uncertainties. Forward modeling quantifies stress coupling between the Mw = 8.0 earthquake and a nearby Mw = 6.3 earthquake that occurred 63 days later. Coulomb stress changes predicted from static HIPSHS models cannot account for the 63‐day lag time between events. Alternatively, an FEM that includes a poroelastic oceanic crust, which allows for postseismic pore fluid pressure recovery, can account for the lag time. The pore fluid pressure recovery rate puts an upper limit of 10−17 m2 on the bulk permeability of the oceanic crust.

The structure and dynamics of the mantle wedge

A large amount of water is brought into the Earth’s mantle at subduction zones. Upon subduction, water is released from the subducting slab in a series of metamorphic reactions. The resulting flux into the mantle wedge modifies its chemical and physical properties by mineral hydration with associated weakening, flux melting and changes in the dynamics and thermal structure of subduction zones. Water guides the formation of volcanoes, earthquakes, continent formation and the long-term chemical evolution of the Earth’s mantle. Recent observational advances include the better documentation of the role of water in causing melting from minor and trace elements in arc lavas, improved structure of the mantle wedge derived from seismic tomography, and documentation of hydration of the mantle wedge from converted phases. High-pressure experiments allow for a quantification of the role of water on seismic velocities and attenuation and rheological changes, which provide essential input into models of subduction zones. Computational models provide additional evidence for the importance of the mantle wedge in subduction zone dynamics.

Subduction factory 2. Are intermediate‐depth earthquakes in subducting slabs linked to metamorphic dehydration reactions?

New thermal‐petrologic models of subduction zones are used to test the hypothesis that intermediate‐depth intraslab earthquakes are linked to metamorphic dehydration reactions in the subducting oceanic crust and mantle. We show that there is a correlation between the patterns of intermediate‐depth seismicity and the locations of predicted hydrous minerals: Earthquakes occur in subducting slabs where dehydration is expected, and they are absent from parts of slabs predicted to be anhydrous. We propose that a subducting oceanic plate can consist of four petrologically and seismically distinct layers: (1) hydrated, fine‐grained basaltic upper crust dehydrating under equilibrium conditions and producing earthquakes facilitated by dehydration embrittlement; (2) coarse‐grained, locally hydrated gabbroic lower crust that produces some earthquakes during dehydration but transforms chiefly aseismically to eclogite at depths beyond equilibrium; (3) locally hydrated uppermost mantle dehydrating under equilibrium conditions and producing earthquakes; and (4) anhydrous mantle lithosphere transforming sluggishly and aseismically to denser minerals. Fluid generated through dehydration reactions can move via at least three distinct flow paths: percolation through local, transient, reaction‐generated high‐permeability zones; flow through mode I cracks produced by the local stress state; and postseismic flow through fault zones.

High‐resolution models of subduction zones: Implications for mineral dehydration reactions and the transport of water into the deep mantle

Arc volcanism is intimately linked to mineral dehydration reactions in the subducting oceanic mantle, crust, and sediments. The location of slab dehydration reactions depends strongly on the temperature and pressure conditions at the top of the subducting plate and hence on the detailed thermal structure of subduction zones. A particularly important physical property of subduction zone thermal models is the viscosity of the mantle wedge. The introduction of an olivine rheology, with appropriate stress and temperature dependence, focuses flow into the tip of the mantle wedge. This leads to a temperature increase of a few hundred degrees in the wedge and the top of the slab as compared to the isoviscous case. Sensitivity tests show that this conclusion is robust under a variety of subduction zone parameters. The new high‐resolution finite element models are used to reevaluate the thermal structure of the Honshu and Cascadia subduction zones using the more realistic olivine rheology. For Honshu, the model predicts slab–mantle interface temperatures of ∼800°C beneath the volcanic front. Deeper parts of the subducting oceanic crust and upper mantle remain relative cool (e.g., the subducted Moho beneath the volcanic front is ∼400°C) because of the rapid subduction of old Pacific lithosphere. High interface temperatures are consistent with Th and Be data, indicating sediment melting, whereas cooler crustal temperatures are consistent with boron evidence, indicating lower temperature dehydration. The strong temperature gradient at the top of the subducting plate may thus reconcile conflicting estimates of slab temperature in subduction zones that are characterized by rapid subduction of mature oceanic lithosphere. Low temperatures persist to great depth in the shallow upper mantle of the subducting plate, which allows for water transport in the form of hydrous minerals to the deep mantle. The thermal structure of the Cascadia subduction zone is markedly different because of the age of the incoming lithosphere and low subduction speed. In this case, shallow dehydration and melting of the subducting crust and lithosphere is predicted.

Tectonic and economic implications of trace element, 40Ar/ 39Ar and Sm–Nd data from mafic dykes associated with orogenic gold minerals in central Victoria, Australia: reply

The western and southwestern Pacific preserve evidence of Cenozoic collisions that guide our understanding of processes and geometries involved in collisions in ancient orogens, in particular in this case, the Palaeozoic Tasmanides of southeastern Australia. Although several styles of collisions are present in the Pacific, ranging from arc–arc collision to arc–plateau collision, the dominant two are oblique and strike-slip collisions between island arcs and rifted continental fragments, and collisions between forearc lithosphere and continental fragments. The 58 Ma collision along the northern margin of the Australian plate in New Guinea, the 44–34 Ma collision preserved in New Caledonia and the 26–25 Ma collision in the North Island of New Zealand may be parts of a single plate boundary collision that migrated southwards along the plate boundary. They characterize the main style of deformation in which a collision between forearc crust and continental fragment produces subduction flip or rollback, thereby avoiding a classic arc–continent collision. Processes involved in, and geometries that have developed from, SW and W Pacific style collisions have been applied to the interpretation of the evolution of the Delamerian Orogen and Lachlan Orogen in the southern Tasmanides with varying degrees of success. The ophiolite obduction model has been successfully applied to the western Tasmania part of the Delamerian Orogen, although there is discussion about its applicability to the mainland. The best example of an arc accretion, that of the Ordovician Macquarie Arc in the eastern Lachlan Orogen, developed from rare geometry in the western Pacific wherein (with the constraint that no forearc or subduction complex has been identified) the arc lies on the continental plate, above a continental-dipping subduction zone. The multiple subduction zone model of Halmahera has been widely applied to the back arc of the Lachlan Orogen, but evidence for clear subduction zones or arcs is not present.

Three‐dimensional attenuation model of the shallow Hikurangi subduction zone in the Raukumara Peninsula, New Zealand

The material properties of the shallow subduction zone are imaged with data from a passive seismic deployment in the Raukumara Peninsula. Attenuation images are obtained using t* values measured from 2848 spectra of P wave arrivals. The t* values are inverted for frequency‐independent three‐dimensional (3‐D) Qp, using a 3‐D velocity model. Linked nodes in a preliminary 2‐D inversion are used to obtain the most reliable initial model for 3‐D inversions. The results show the benefits of obtaining 3‐D Q in addition to 3‐D Vp and Vp/Vs. Q is related to localized permeability and heterogeneity, and thus 3‐D Q is an important tool for interpreting active tectonic regions. Qp in the shallow overlying plate generally corresponds to the mapped geology. Two high Qp features (>600) represent subducted seamounts or Cretaceous volcanics within the overlying plate. The gradient above the subducting plate is the strongest feature in the Qp model, with the slab having high Qp (600–900). Below 20‐km depth and above the plate interface, there is a zone of 50% decrease in Qp and high Vp/Vs about 10 km thick which represents actively subducting sediment and has distributed microearthquake activity associated with active underplating.

Experimental constraints on the depth of olivine metastability in subducting lithosphere

The hypothesis that metastable olivine persists in some subducting slabs into the transition zone has wide implications for mantle dynamics and rheology. In order to evaluate this possibility we derive new thermo-kinetic subduction zone models to predict the extent of olivine metastability within the stability fields of its high-pressure polymorphs, wadsleyite and ringwoodite. Our updated models improve on previous work by incorporating experimental kinetic data on realistic mantle compositions ((Mg, Fe) 2SiO 4) rather than analogue systems. Furthermore, latent heat due to the transformation is fed back into both the kinetics and the thermal model. We also consider the effects of transformation stress on growth kinetics and the possibility of an intracrystalline transformation mechanism, previously thought to be important only at high shear stresses. Our models predict significantly smaller wedges of metastable olivine than previous work. In the case of Tonga, for example, where high values of lithospheric age (100–140 million years) and convergence rate (∼14 cm per year) are most favorable for metastability, models considering only grain boundary nucleation and interface-controlled growth predict olivine metastability to ∼600 km depth, in contrast to ∼660 km predicted previously by Kirby et al. [Rev. Geophys. 34 (1996) 261]. When intracrystalline transformation is considered, the depth of metastability is further reduced by as much as 100 km, due to the large increase in the density of nucleation sites. Inhibition of growth by transformation stress can increase the depth interval over which the transformation takes place, but is unlikely to be a dominant factor, especially if the intracrystalline mechanism operates. These results indicate that the existence of metastable olivine at depths corresponding to those of the deepest earthquakes (∼680 km) requires subduction of old lithosphere (>100 million years) and a high vertical subduction velocity (≳15 cm per year). Such conditions currently may only be achieved in the very northern part of the Tonga subduction zone, yet earthquakes occur at depths down to 680 km in other subduction zones in which lithospheric age and/or subduction velocity are relatively low (e.g. Indonesia and the Marianas). Therefore, mechanisms other than transformational faulting in metastable olivine must operate to cause the deepest earthquakes.

Homogeneous vs heterogeneous subduction zone models: Coseismic and postseismic deformation

A finite‐element model (FEM) incorporating geologic properties characteristic of a subduction zone is compared with FEMs approximating homogeneous elastic half‐spaces (HEHS)s to investigate the effect of heterogeneity on coseismic and postseismic deformation predictions for the 1995 Colima‐Jalisco Mw =8.0 earthquake. The FEMs are used to compute a coefficient matrix relating displacements at observation points due to unit dislocations of contact‐node pairs on the fault surface. The Green's function responses are used to solve the inverse problem of estimating dislocation distributions from coseismic GPS displacements. Predictions from the FEM with heterogeneous material properties, loaded with either of the HEHS dislocation distributions, significantly overestimate coseismic displacements. Postseismic deformation predictions are also sensitive to the coseismic dislocation distribution, which drives poroelastic and viscoelastic relaxation. FEM‐generated Green's functions, which allow for spatial variations in material properties, are thus preferable to those that assume a simple HEHS because the latter leads to dislocation distributions unsuitable for predicting the postseismic response.

Cs–Rb–Ba systematics in phengite and amphibole: an assessment of fluid mobility at 2.0 GPa in eclogites from Trescolmen, Central Alps

Eclogites from Trescolmen that contain abundant hydrous minerals (phengite, amphibole, paragonite, zoisite, talc, apatite) show petrographic evidence for fluid infiltration under conditions of 2.0 to 1.8 GPa, 650 °C. Large ion lithophile elements (LILE, e.g. Cs, Rb, Ba and Sr) were analysed by in-situ techniques in all eclogite mineral phases in order to characterize the behaviour of fluid-mobile elements at high pressure. In-situ analysis of carefully-chosen metamorphic assemblages circumvents the problem of partial late-stage alteration, which can severely influence the calculated element budgets of whole-rock samples. Phengite is the dominant host for Cs, Rb, and Ba in both eclogite and adjacent garnet mica schist samples, and incorporates >90% of the budgets of these elements in whole rocks. LILE contents of phengites in phengite-rich rocks are likely to record the Cs/Rb and Ba/Rb ratios of their host rock protoliths. The LILE patterns of eclogite are consistent with protoliths derived from basalt that underwent seafloor alteration, whereas those of mica schist are almost identical to average upper continental crust. In contrast, LILE patterns of eclogite samples that lack phengite, but do contain amphibole, are unlike any plausible protolith, but are identical to those of amphibole in phengite-bearing samples. This observation points to homogenization of the LILE in different lithologies, which we correlate with petrographic evidence for fluid infiltration. Because phengite in garnet mica schist has a strong capacity to buffer the fluid with respect to Cs, Rb, and Ba, homogenization of amphiboles is best explained by fluid infiltration from the surrounding metapelites into eclogite bodies, implying at least metre-scale fluid mobility. The amphibole homogenization can be most easily modelled by a pervasive open-system fluid flux through the eclogites, possibly facilitated by ductile deformation during the early stages of uplift. Simple calculations give minimum fluid-rock ratios of ~0.001 to 0.004. Demonstration of the mobility of very small volumes of fluid through eclogite is an important prerequisite of many subduction zone models that try to explain across-arc variations in trace element geochemistry. The low fluid-rock ratios from this study are not in contrast with oxygen isotope heterogeneities reported from other eclogite localities. Fluid mobile elements such as Cs, Rb and Ba are more sensitive indicators of small volume, fluid-rock interaction and are therefore potentially valuable for understanding fluid infiltration processes in systems where oxygen isotope shifts are not large enough to be detectable.

Upper plate contraction north of the migrating Mendocino triple junction, northern California: Implications for partitioning of strain

Geologic measurement of permanent contraction across the Cascadia subduction margin constrains one component of the tectonic deformation along the convergent plate boundary, the component critical for the seismic hazard assessment of crustal faults. A comprehensive survey of active faults in onshore subduction margin rocks at the southern end of the Cascadia subduction zone indicates that these thrust faults accommodate ∼10 mm/yr of convergence oriented 020°–045°. Seismotectonic models of subduction zones typically assign this upper plate strain to the estimate of aseismic slip on the megathrust. Geodetic models include this permanent crustal strain within estimates of elastic strain accumulation on the megathrust. Both types of models underestimate the seismic hazard associated with crustal faults. Subtracting the observed contraction from the plate convergence rate (40–50 mm/yr; directed 040°–055°) leaves 30–40 mm/yr of convergence to be partitioned between slip on the megathrust, contraction within the southern Juan de Fuca plate, and crustal contraction outside the subduction complex rocks. This simple estimate of slip partitioning neglects the discrepancy between the plate convergence and contraction directions in the vicinity of the Mendocino triple junction.The San Andreas and Cascadia limbs of the Mendocino triple junction are not collinear. The eastern edge of the broad San Andreas boundary is ∼85 km east of the Cascadia subduction boundary, and across this zone the Pacific plate converges directly with the North America plate. The skewed orientation of crustal structures just north of the leading edge of the Pacific plate suggests that they are deforming in a hybrid stress field resulting from both Juan de Fuca‐North America motion and Pacific‐North America motion. The composite convergence direction (50 mm/yr; directed 023°) is consistent with the compressive stress axis (020°) inferred from focal mechanisms of crustal earthquakes in the Humboldt region. Deformation in such a hybrid stress field implies that the crustal faults are being loaded from two major tectonic sources. The slip on crustal faults north of the Mendocino triple junction may consume 4–5 mm/yr of Pacific‐Humboldt convergence. The remaining 17–18 mm/yr of convergence may be consumed as distributed shortening expressed in the high rates of uplift in the Cape Mendocino region or as northward translation of the continental margin, north of the triple junction.