T11B-1250 0800h
Dynamic Grain Growth in Forsterite Aggregates Experimentally Deformed to High Strain
The dynamics of the outer Earth are largely controlled by olivine rheology. From previous work it has become clear that if olivine rocks are deformed to high strain, substantial weakening may occur before steady state mechanical behaviour is approached. This weakening appears directly related to progressive modification of the grain size distribution through competing effects of dynamic recrystallization and syn-deformational grain growth. However, most of our understanding of these processes in olivine comes from tests on coarse-grained materials that were reduced in grain size during straining by grain size insensitive (dislocation) creep mechanisms. The aim of the present study was to investigate microstructure evolution of fine-grained olivine rocks that coarsen in grain size while deforming by grain size sensitive (GSS) creep. We used fine-grained (~1 $\mu$m) olivine aggregates (i.e., forsterite/Mg$_2$SiO$_4$), containing ~0.5 wt% water and 10 vol% enstatite (MgSiO$_3$). Two types of experiments were carried out: 1) Hot isostatic pressing (HIP) followed by axial compression to varying strains up to a maximum of ~45%, at 600 MPa confining pressure and a temperature of 950°C, 2) HIP treatment without axial deformation. Microstructures were characterized by analyzing full grain size distributions and texture using SEM/EBSD. Our stress-strain curves showed continuous hardening. When samples were temporally unloaded for short time intervals, no difference in flow stress was observed before and after the interruption in straining. Strain rate sensitivity analysis showed a low value of ~1.5 for the stress exponent {\it n}. Measured grain sizes show an increase with strain up to a value twice that of the starting value. HIP-only samples showed only minor increase in grain size. A random LPO combined with the low {\it n} ~1.5 suggests dominant GSS creep controlled by grain boundary sliding. These results indicate that dynamic grain growth occurs in forsterite aggregates deforming by GSS creep, and we relate the continuous strain hardening to this process. A dynamic grain growth model involving an increase in cellular defect fraction seems best applicable to the grain growth observed in this study. We suggest that the employment of this model to fine-grained olivine rocks can further improve our understanding of the microstructural evolution of this material and related rheological behaviour.
T11B-1251 0800h
Silicon diffusion in wadsleyite at high pressure and implications for rheology of the mantle transition zone
Wadsleyite is one of the major constituent minerals in the mantle transition zone. In order to understand the rheological behavior of the mantle transition zone, it is essential to determine the diffusion rate of the slowest diffusing species which may control the plastic deformation in silicate minerals undergoing diffusion creep or dislocation climb. Although Si is the slowest diffusing species in many silicate minerals, diffusion data in wadsleyite are limited to Mg-Fe interdiffusion. In this study, we carried out Si diffusion experiments in wadsleyite to discuss the rheological properties in the mantle transition zone. High-pressure experiments were carried out using a Kawai-type multi-anvil apparatus installed at Tohoku University. Starting material of polycrystalline wadsleyite was synthesized from powdered forsterite at 18 GPa and 2003K. Surface of the polycrystalline wadsleyite was polished and then coated with a $^{29}$SiO$_{2}$ thin film. Diffusion annealing was conducted at 18 GPa and 1703-1903K. After the diffusion annealing, concentration profiles of $^{29}$Si were obtained by the depth profiling method using secondary ion mass spectrometry (SIMS). The obtained diffusion profiles were composed of two regions. Volume diffusion coefficient (Dv) was calculated from the region near the sample surface using the solution of thin film diffusion model, and the solution by LeClaire (1963) was used to obtain the grain-boundary diffusion coefficient ($\delta$Dgb) from the deeper region. Dv and $\delta$Dgb were determined to be Dv= 3.44$\times$10$^{-11}$ [m$^{2}$/s] exp (-299 [kJ/mol]/RT) and $\delta$Dgb= 1.14$\times$10$^{-17}$ [m$^{3}$/s] exp (-248 [kJ/mol]/RT), respectively. Si diffusion rates are about five orders of magnitude slower than Mg-Fe interdiffusion rates in wadsleyite. Despite oxygen diffusion data has never been reported, Si is a good candidate for the slowest diffusing species in wadsleyite as it is in olivine. Assuming that Si is the rate-controlling species in wadsleyite, geophysical model of the viscosity in the mantle transition zone can be explained by diffusion creep in wadsleyite for a grain size of about 0.5-5 mm. The subducting slabs become weaker than surrounding mantle if grain size of spinel is reduced to less than 1 $\mu$m after the olivine-spinel transformation.
T11B-1252 0800h
Dependence of Dunite Viscosity on Oxygen Fugacity
Although the viscosity of olivine single crystals exhibits a clear dependence on oxygen fugacity, published results for polycrystalline, olivine-rich aggregates are more ambiguous. Based on deformation experiments on single crystals of olivine (Bai, Mackwell, and Kohlstedt, 1991), strain rate increases as oxygen fugacity to the -0.03 to 0.4 power, the exact value depending on factors such as crystal orientation and oxide buffer condition. In experiments on \AA heim and Anita Bay dunite (Chopra and Paterson, 1981), samples jacketed in Ni tended to flow at lower differential stress than samples jacketed in Fe. An understanding of the dependence of oxygen fugacity on dunite will allow extrapolation of flow properties for a variety of mantle conditions. In the present study, samples were cored from blocks of \AA heim dunite. The samples were annealed for twelve hours prior to deformation in a one-atmosphere horizontal furnace at an oxygen partial pressure either near the Ni/NiO or the Fe/FeO buffer to drive off water bound in hydrous minerals and water adsorbed at free surfaces. The silica activity is buffered by the presence of orthopyroxene. Samples were then deformed in a gas-medium apparatus at a confining pressure of 300 MPa and a temperature range of 1400 to 1550 K at a series of constant stress conditions. The resulting stress exponent was in the range of 3 to 4 indicating that deformation was dominated by dislocation creep. Samples buffered by Ni/NiO deformed a factor of 5 to 10 faster than samples buffered by Fe/FeO. Analyses of the results yield an oxygen fugacity exponent of approximately 0.2. Therefore, since oxygen fugacity in the mantle is believed to have increased through time, the viscosity of the mantle would likely have decreased.
T11B-1253 0800h
Deformation Experiments on Polycrystalline Mg2SiO4-Ringwoodite Using the Deformation-DIA and Monochromatic X-Ray Diffraction
Deformation experiments under controlled differential stress and strain rates were carried out on ringwoodite (Mg2SiO4) samples using the deformation-DIA (D-DIA) with monochromatic X-ray diffraction. Samples (0.8 mm dia., 1.2 mm long) were sintered polycrystalline rocks, synthesized at 20GPa and 1523K at Geodynamic Research Center (Ehime Univ., Japan), with an average grain size of $<$10$\mu$m and no observable preferred orientation. Each sample was surrounded by a BN sleeve, with two fully densified alumina pistons above and below as deformation pistons. Two Au foils were put between the sample and the pistons as strain markers. Monochromatic X-rays with an energy of 65 keV ($\lambda$=0.19$\AA$) were used. Two-dimensional (2-D) diffraction patterns were repeatedly collected during the deformation process using an X-ray CCD. Differential lattice strains were determined based on the distortion of the Debye rings. The differential stresses were calculated from the lattice strain using single crystal elastic constants. Total sample lengths were determined by radiography. Multiple stress-strain curves (compression and extension) were obtained at various temperatures at pressures from 4 to 10 GPa, with the total axial strain up to 22$%$. Our results show that (1) at room T ringwoodite deforms in the ductile regime, (2) the strength of ringwoodite increases with P and decreases with T, (3) ringwoodite exhibits modest strain hardening (many discrepancies in strength measurements previously reported, where no total strain information could be obtained, can be reconciled by taking into account the strain dependence on strength), and (4) ringwoodite develops strong lattice preferred orientation even at modest strains, which was clearly observed as intensity variation with angles along Debye rings. At 5-10 $%$ axial strain, a strong texture had already developed with the 110 poles parallel to the compression direction. This texture is typical of materials with spinel structure deformed via the $\{$111$\}$$<$10-1$>$ and $\{$110$\}$$<$-110$>$ slip systems.
T11B-1254 0800h
Grain boundary wetness of texturally equilibrated rocks with implications for the seismic properties of the upper mantle
Melt- or fluid-filled pore geometry in the texturally equilibrated aggregates was investigated quantitatively by measuring the grain-boundary wetness, which is defined by the ratio of solid-liquid boundary area over the total area of interphase boundaries. The merit to measure the wetness is that, being combined with the granular model, the bulk mechnical properties of the liquid-bearing aggregates can be predicted quantitatively. Especially, detailed study on the wetness of the partially molten peridotite was performed to clarify the characteristics of the seismic low velocity zones in the upper mantle. The measurments were performed on several systems characterized by various dihedral angles and degree of faceting. The result shows that the grain boundary wetness increases monotonically with increasing liquid volume fraction. For no faceting system and olivine-basalt system with low dihedral angle, the relation between liquid volume fraction (f) and the wetness (y) is quite consistent with the theoretical prediction from the dodecakeidecahedral packing geometry. On the other hand, the wetness obtained for the systems showing strong faceting is generally lower than the theoretical prediction. Partially molten lherzolite shows systematically lower wetness than the simple olivine-basalt system. For all systems the obtained wetness-liquid volume fraction relationship can be fitted well to formulae $\psi=A\phi^{1/2}$ with fitting parameter A, independently of degree of faceting and dihedral angles. This suggests that the 3 dimensional pore shape is not a disk shape but a tubular one. The values of A ranges from 1.4-2.5 (A=2.3 for the olivine-basalt system and A=1.7 for the partially molten lherzolite). The seismic wave velocities of the systems whose pore geometry is characterized by parameter A can be represented by the equivalent aspect ratio of the oblate spheroid model. A=1.4-2.5 corresponds the equivalent aspect ratio of 0.25-0.09. The equivalent aspect ratios of the olivine-basalt system and the partially molten lherzolite are obtained as 0.1 and 0.15, respectively. The result indicates that the dlnVs/dlnVp expected for the partially molten peridotite in textural equilibrium is 1-1.5, which is much smaller than the value ($\sim$ 2) expected for the thin cracks and dikes. The present results agree with the previous results on the basis of isotropic model. The effect of facetting on the seismic wave velocities is not significant in the partially molten peridotite. These results would allow more accurate prediction of liquid content in texturally equilibrated rocks using Vp (or Vs) data and also more realistic interpretation of the pore geometry determined from the combined analysis of Vp and Vs data.
T11B-1255 0800h
Dislocation Microstructures and Dissociation in Deformed Olivine Displaying the A-, B- and C-type Fabrics
Seismic anisotropy in the upper mantle is generally considered to result from olivine fabrics that result from dislocation creep. Jung and Karato, (2001) have shown that the fabrics developed in experimentally deformed olivine are dependent on H$_{2}$O fugacity and stress, suggesting that olivine fabric and the resulting seismic anisotropy can provide a means of probing H$_{2}$O content and stress in the upper mantle. Olivine fabrics, known as B-type and C-type, occur in olivine deformed under high H$_{2}$O fugacity and high stress (B-type) and low stress (C-type) whereas the more common A-type fabric occurs in olivine that is deformed under lower stress and H$_{2}$O fugacity. To better understand the development of B- and C-type fabrics in olivine, we have used TEM and HRTEM to investigate the deformation microstructures and dislocation core structures in experimentally deformed olivine displaying the A-, B-, and C-type fabrics. In a sample displaying the C-type fabric, tilt boundaries along (100) and (001), consisting of edge dislocations with b = [100] and b = [001], respectively, suggest dominance of the (001)[100] and (100)[001] slip systems. HRTEM imaging of the b = [100] and b = [001] edge dislocations along [010] shows no dissociation for b = [100], whereas the b = [001] dislocations are dissociated into 1/2[-101] and 1/2 [101] partial dislocations. In a B-type fabric sample, we find b = [100] and b = [001] dislocations with mixed character in (010), indicating that the (010)[001] and (010)[100] slip systems are active. Tilt boundaries along (001) with b = [001] suggest that the (100)[001] system is also active. HRTEM imaging of dislocations with b = [001] indicate that these dislocations are dissociated in two different ways. The first is the same described above for the C-type fabric whereas the second consists of two 1/2[001] partial dislocations. The A-type fabric sample contains many (100) tilt boundaries consisting of b = [100] edge dislocations, suggesting that the (010)[100] slip system is dominant. This sample also contains many b = [001] with predominantly screw character. HRTEM imaging of b = [001] dislocation cores was not possible, and b = [100] dislocations show no evidence of dissociation. The EBSD fabric analyses of Jung and Karato (2001) are consistent with the dominance of the (100)[001], (010)[001] and (010)[100] slip systems in the C-, B-, and A-type fabrics, respectively. The effect of high H$_{2}$O fugacity on the deformation mechanisms in olivine is to enhance slip along [001] on either the (100) or (010) planes for the C- and B-type fabrics, respectively. Dissociation of b = [001] edge dislocations in C- and B-type fabric samples is unlikely to enhance slip because the partial dislocations are out of the slip plane. However, the presence of hydrogen in olivine may reduce Peierls stresses for b = [001] in both C- and B-type fabrics.
T11B-1256 0800h
New observations of Q quality factors of a few gravest normal modes from the Global Geodynamics Project (GGP)
he high quality of the GGP superconducting gravimeters contributes to the clear observation of seismic normal modes at frequencies lower than 1mHz and offers a good opportunity for studying the behaviour of these modes. The interest of scientists for the gravest normal modes is due to the fact that these modes do contribute to a better knowledge of the density profile in the Earth, helping to constrain Earth's models. These modes have been clearly identified after some large recent events recorded on superconducting gravimeters. The Peruvian earthquake of June 2001 provided us with individual spectra (in a unique station) with a clear splitting of the fundamental mode 0S2 and identification of each of the five individual singlets, with a resolution never obtained from broad-band seismometers records. The Q quality factors have been determined from the apparent decrease of the amplitude of each singlet with time, according to a well suited technique (Roult & Clevede, 2000). The results are compared to the theoretical frequencies and Q quality factors computed in the PREM model, taking into account the rotation and the ellipticity of the Earth. The two datasets (frequencies and Q quality factors) exhibits a small shift between the splitting of the observed values and that of the predicted ones. That seems to point out that the rotation and the ellipticity don't explain the observations and that we have to take into account additional effects. A new dataset of Q quality factors of all singlets of the gravest modes is under construction, including the 0S2 and 0S3 modes, the radial 0S0 mode and the 2S1 mode recently identified by Rosat et al. (2003).
T11B-1257 0800h
Effect of Water and Iron Content on the Rheological Behavior of Olivine
We have undertaken an experimental investigation of the effect of water and iron content on the viscosity of aggregates of Fe-Mg olivine in order to provide a basis for conparing convection models for the mantle of Earth with those for the more iron-rich mantle of Mars. Our study builds on three experimental observations: (i) At a given temperature, the viscosity of single crystals of San Carlos olivine [Fo$_{90}$ = (Fe$_{0.1}$Mg$_{0.9}$)$_{2}$SiO$_{4}$] is signficantly higher than that of crystals of fayalite, Fo$_{0}$, (ii) the viscosity of San Carlos olivine decreases with increasing water concentration, and (iii) the solubility of water in olivine increases with increasing iron concentration. To extend deformation experiments to polycrystalline samples of olivine of higher Fe content, powders of Fo$_{50}$ and Fo$_{70}$ were fabricated from mixtures of natural olivine, Fo$_{90}$, and synthetic fayalite, Fo$_{0}$. The resulting materials were ground into fine ($<$10 $\mu$m) powders, cold-pressed into Ni capsules, and then hot-pressed at 300 MPa and 1533 K for 2 to 12 h. For experiments under hydrous conditions, two drops of water, each $\sim$0.03 ml, were added before sealing a sample within telescoping Ni cans for deformation. The average grain size of the resultant hot-pressed samples were between 20 and 57 $\mu$m . In the samples deformed under hydrous conditions, water bubbles were present both within olivine grains and along grain boundaries, demonstrating that the samples were water-saturated. High-temperature, high-pressure compressive creep experiments in both the diffusion and the dislocation creep regimes were carried out using a gas-medium apparatus at temperatures of 1223 to 1473 K and a confining pressure of 300 MPa. Under both anhydrous and hydrous conditions, the viscosity of samples of Fo$_{50}$ is a factor of $>$10 lower than the viscosity of samples of Fo$_{70}$, which is a factor of $>$10 lower than the viscosity of samples of Fo$_{90}$. The viscosity of a sample of a specific Fe:Mg ratio deformed under hydrous conditions is a factor of $\sim$10 lower than its counterpart deformed under anhydrous conditions. Therefore, at the same thermodynamic conditions (e.g., P, T, water fugacity), the viscosity of the more Fe-rich mantle ($\sim$18 wt % FeO) of Mars will be a factor of $\sim$3 lower than the mantle ($\sim$8 wt % FeO) of Earth.
T11B-1258 0800h
Oceanic Upper Mantle Rheology as Constrained by Combined Geodynamic and Seismic Modeling of Plate-Mantle Interaction
The rheological parameters are the most important parameters in the dynamics of the upper mantle. A comparison of seismic and numerical modeling studies of the oceanic upper mantle dynamics provides important, independent constraints on the rheological parameters. Recent surface wave tomography results show significant reheating or reduced cooling between 70 Ma and 110 Ma for the Pacific lithosphere. This correlates with the well-known observations of reduced topography and enhanced heat flux at old seafloor relative to the half-space cooling model. Sublithospheric small scale convection (SSC) has been proposed to explain the observed topography and heatflow. The process of SSC is largely controlled by gravitational instability of the bottom part of lithosphere, which is determined by the rheology of lithosphere and mantle. We formulated 3D numerical models to simulate the SSC process. By comparing the thermal structure of lithosphere above SSC to the cooling halfspace model, we determine a 'lithospheric thermal age' that may deviate from the real plate age due to SSC. Comparison of numerical models of the SSC with a seismic surface wave tomography model shows striking agreement, clearly supporting the existence of SSC. This can also provide independent constraints on the rheological parameters of the upper mantle. Lower activation energy will lead to larger temperature variation associated with SSC, which gives mechanical erosion of a larger part of the lithosphere, an thus more thinning. This is again reflected in the thermal age. Furthermore, dislocation creep calculations show a much stronger thermal erosion than diffusion creep does for the same activation energy, and therefore a much larger effect on the thermal state of the lithosphere. For diffusion creep, an activation energy of 120 kJ/mol or lower shows a good fit to the seismic data. For dislocation creep, much larger activation energy values, which are more in agreement with laboratory measurements, fit the seismic data best. This suggests that dislocation creep is the dominant mechanism in upper mantle dynamics, and confirms previous similar conclusions based of laboratory experiments or seismic anisotropy.
T11B-1259 0800h
Glacial Isostatic Adjustment and Relative Sea Level Changes: the Role of Lithospheric and Upper Mantle Lateral Viscosity Variations
The Earth response to the melting of the late--Pleistocene ice--sheets has been mainly studied by spherically layered models, based on well established analytical methods. In the last few years new approaches have flourished, taking advantage of numerical techniques and massive computer resources. These methods allow to evaluate the effects of non-Newtonian rheologies and lateral viscosity variations on glacial isostatic adjustment (GIA) and relative sea level (RSL) changes. In this framework, we used a 3D FE code to include laterally varying structures both in the elastic lithosphere and in the Newtonian viscoelastic upper mantle. Our spherical models reproduce the global structure of the cratons and account for a low-viscosity layer beneath the oceanic lithosphere. We compare our model results to the RSL trends in the last 6000 years showing that, when all of the data available are jointly considered, homogeneous and laterally heterogeneous models perform in a similar way. This suggests that the effect of lateral viscosity variations cancel out globally, so that the common radially stratified models are a useful tool for GIA predictions on a large spatial scale. When particular subsets of the global RSL data set are considered, belonging to regions which have experienced directly the glacial melting (such as North America and Northern Europe), we find that the RSL can be better, albeit marginally, reproduced by models which include heterogeneous structures. This suggests that a further refinement of the laterally varying features on a regional scale may lead to a better agreement between RSL observations and model predictions.
T11B-1260 0800h
Insight Into 3-D Earth Structure From a New Generation of Glacial Isostatic Adjustment Models
Predictions of glacial isostatic adjustment (henceforth GIA), and analyses of related geophysical observables, have generally been based on normal mode approaches valid for spherically symmetric Earth models. We explore the impact of lateral variations in elastic plate strength, including the presence of plate boundaries and variations across the ocean-continent interface, and mantle viscosity, on predictions of 3-D crustal velocities and long-wavelength gravity anomalies. Our calculations adopt a realistic, global model of Late Pleistocene deglaciation and meltwater redistribution and they are based on a new, finite-volume formulation of the (Maxwell) viscoelastic GIA problem. We find that the introduction of realistic mantle and lithosphere structure perturbs GIA-induced crustal velocities at levels which can significantly exceed the current observational uncertainties of space-geodetic methods (e.g., GPS). Predictions of secular variations in long-wavelength harmonics of the geopotential are, in contrast, primarily sensitive to deep mantle viscosity structure. We compare the sensitivities evident in our suite of forward models with global and site-specific geodetic data to assess the prospects for using GIA observations to constrain 3-D Earth structure.
T11B-1261 0800h
Remote Triggering of Deep Earthquakes - Indication for Criticality of Subducted Slabs
Recent findings by Tibi et al. (2003) show that deep-focus earthquakes can dynamicly trigger further events in a subduction zone below 400 km depth at remote distances. The detailed analysis of 19 August 2002 Tonga deep earthquake sequences reveals evidence for both static and dynamic triggering. Based on their data, Tibi et al. propose that the triggered events are caused by transient effects in regions near criticality, where earthquakes have otherwise - without external influences - difficulty to nucleate. Here, we investigate the observed remote triggering of deep-focus earthquakes at the transition zone with a visco-elastic block slider model for ductile instabilities (Riedel, 2003). It shows, that the proposed triggering effect may be explained with a divergence of the correlation length of the stress transfer function near to "criticality", i.e. near to a critical unstable state of the subducting slab, where nucleation of ductile instabilities is prone to occur. The existence of this critical state in the deforming subducted slab is either caused by large bending forces in the mantle transition zone or, alternatively, is largely promoted by an anomaly in the slab viscosity structure resulting from a metastable delay of the olivine-spinel phase transformation (Karato et al., 2001). References S. Karato, M. R. Riedel, D. A. Yuen, "Rheological Structure and Deformation of Subducted Slabs in the Mantle Transition Zone: Implications for Mantle Circulation and Deep Earthquakes", Phys. Earth Planet. Inter. 127 (2001) 83-108 M. R. Riedel, "Modeling shear instabilities with block sliders: brittle and ductile", AGU fall meeting, San Francisco, USA (2003) R. Tibi, D. A. Wiens and H. Inoue, "Remote triggering of deep earthquakes in the 2002 Tonga sequences", Nature 424 (2003) 921-925
T11B-1262 0800h
Time Dependent Geoid Constraints Upon Mantle Viscosity Stratification
The global measurement of the time dependence of geoid height that is being provided by the GRACE satellite system that is now in space will eventually provide the basis for considerably more accurate inversions for mantle viscosity structure than are now possible. However, existing data on the time dependence of geoid height based upon the results of satellite laser ranging already provide very strong constraints upon the effective viscosity of the of the deepest mantle, especially when these are conbined with observations of the spectrum of relaxation times that characterize the process of glacial isostatic adjustment (GIA). Such data, by themselves, very tightly constrain the viscosity structure in the upper mantle and transition zone. We will describe a series of new analyses of the expected global pattern of geoid height time dependence based upon the recently published refined model of the GIA process denoted ICE-5G(VM2), a model based upon a significant refinement of the ICE-4G(VM2) precursor ( see W.R. Peltier, Ann. Rev. Earth and Planet. Sci., 32, 111-149, 2004). The impact of the new model of surface loading upon the mantle viscosity inverse problem turns out to be both interesting and significant.
T11B-1263 0800h
A New Elastic-Viscous Model to Constrain Holocene Relative Sea Level Along the Northern Margin Gulf of Mexico
A variety of mid-to-late Holocene coastal deposits and landforms in the southern hemisphere, and along the US Gulf of Mexico coast, have been interpreted to represent one or more episodes of relative sealevel highstand, perhaps related to glacio-eustasy. For the US Gulf Coast, the sedimentary evidence for this highstand is highly variable, and contradicts high-resolution data from the Mississippi delta region, which shows continual submergence. However, possible alongshore variations in tectonic contributions to this record remain unclear. Global viscoelastic earth models exist to explain the vertical, crustal response to iceload changes and associated water loading (e.g., Peltier, 1974; Lambeck al., 1990). However, in the Gulf of Mexico case, a local model is needed, because of the presence of the Mississippi delta load anomaly, which is not considered in averaged earth global models. We develop a semi-analytical Green function approach for a new elastic-viscous flexure model consisting of an elastic plate over a viscous substratum. The solution is expressed as the Hankel transform of the load and lithosphere-mantle kernel. Over time this model converges to the equivalent response of an elastic plate over an inviscid substratum, which is in agreement with the elastic behavior of the lithosphere for long periods of time ($>10^6$yr). Predicted subsidence rates vary with distance to the load, and decrease over time. By incorporating the local sedimentary loading history, we make a first attempt to reconcile contradictory interpretations of mid-to-late Holocene relative sealevels, throughout the US Gulf of Mexico coast.
T11B-1264 0800h
A Potential Role for Slab/Lithosphere Decoupling and Edge Driven Convection in the Tectonic Evolution of the Western US
About 30 mya, shortly after the foundering of the once horizontally subducting Farallon Plate, extensive volcanism and extensional tectonics began throughout the Rocky Mountain region. It is accepted by many that the foundering of the plate drove a significant portion of the subsequent tectonics in the western US including the activation of the Rio Grande Rift. Here, we present 2D numerical experiments aimed at studying the delamination of the Farallon plate from beneath the western US. In these experiments, the lithosphere and mantle are modeled as a visco-elasto-plastic medium. The brittle parts of the lithosphere are modeled as a frictional and cohesional material. The ductile lithosphere is modeled as a non-Newtonian Maxwell visco-elastic material. Faults in the brittle parts of the model are formed by locally decreasing the cohesion and friction as a function of plastic strain. The rheological structure of the model is controlled by the initial temperature distribution and the temperature boundary conditions. The experiments focus on the conditions that would lead to the delamination of a flat slab from beneath a simple representation of the western US. This simple representation assumes that the slab and the lithosphere are at thermal equilibrium and that the initial topography is flat. We find that in order to generate the instability necessary for the foundering of the plate, the presence of a low viscosity wedge between the remnant continental lithosphere and the subducting plate is necessary. The wedge serves as a proxy for possible mantle hydration and shear heating at the boundary between the subducting plate and the over-riding lithosphere. It effectively decouples the two plates, allowing a lithospheric instability to propagate into the total, or near total, removal of the slab. Depending on the geometry, density, and the viscosity of the wedge, we observe two basic modes for the delamination of the slab. In the first mode, the foundering plate is entrained in an edge-driven convection cell which conveys it laterally eastward and then down beneath the cratonic region of the lithosphere. In the second mode, the plate is decoupled such that it deforms and "ponds" to the west before finally detaching and sinking. We explore various geometries and rheological properties for the wedge with the goal of generating a final thermal and chemical configuration similar to that inferred via seismic tomography for the presumed remnants of the Farallon slab beneath the western US. Furthermore, we examine the predicted evolution of topography in the simulations for consistency with existing tectonic models for the evolution of the RGR and the Colorado Plateau. We posit that the existence of a low viscosity wedge generated the instability leading to the foundering of the Farallon plate. Moreover, the extent and the physical properties of the wedge control the delamination pattern.
T11B-1265 0800h
Effects of the Lithospheric Strength on the Generation, Longevity, and Structure of the Subduction
Rheology of the lithosphere is the most important property to determine tectonics of the terrestrial planets. Because subduction of the plate is a fundamental character of the plate tectonics, we here focus on influences of the strength of the plate interior and boundary on the generation of the subduction. We develop self-consistent dynamical models of the plate incorporated in the mantle convection system, which reproduce the growth of the subducted plate from the initiation to a deep slab. We use 2-D viscous fluid in a rectangular box, and no external forces are applied to the surface boundary and the plate. Rheology depending on the hysteresis of the stress is utilized to produce a thrust fault at the plate boundary (Honda et al., 2000). The rheology includes Arrhenius-type temperature and pressure dependence and maximum yield strength as large as those reported by the experimental studies. We also introduce heterogeneity of the surface region, i.e., "oceanic" and "continental" region, which is treated as a horizontal heterogeneity of the yield stress and/or the composition at the surface. In our models, Earth-like one-sided successfully reproduced. Systematic survey for the strength of the plate (maximum strength of the yielding, a friction coefficient of brittle fracture of the plate interior, and a friction coefficient at the plate boundary fault) shows that generation of the subduction is sensitive to the friction the plate boundary but not to the maximum strength of the plate interior. The subducted slab is able to be formed when the friction coefficient of the fault zone at the plate boundary is small enough ($<$ 0.01), and then the subduction successfully occurs even at the highest value in the model (600 MPa). Our models with long-time integration (100 to 200 Myr) show that the subduction of the plate stably continues and maintains high temperature in the wedge mantle above the subducted plate for long duration. Our results show that fine adjustment for the strength of the lithosphere as previous studies reported is not necessary to generate Earth-like subduction when there exist mechanisms to weaken plate boundary. We have also examined effects of the rheology in the plate and slab on the structure of the subducted slab. We consider difference of the tensional and compressional strength of the lithosphere and/or weakening induced by grain size reduction by the phase transition. When both the effects are introduced into the model with a thin overriding plate, the slab migrates ocean-ward, and that causes shallow dip angle of the slab.