DI53A-1092 INVITED
Detailed Structure of Upper Mantle Discontinuities in the Tonga and Mariana Subduction Zones
Recordings of deep Tonga earthquakes from two arrays of 12 broadband seismographs each in the Fiji and Tonga islands were stacked and searched for reflections and conversions from upper mantle discontinuities in the Tonga subduction zone. For the 410-km discontinuity, the results show no systematic variations in depth with distance to the cold slab. The 660-km discontinuity varies between 656 and 714 km in depth. For the southern and central parts of the subduction zone, the largest depths occur in the core of the Tonga slab. For the northern part, two separate depressions of the 660 are observed. These anomalies are interpreted as being induced by the active, steeply subducting Tonga deep zone and a subhorizontally lying remnant of subducted lithosphere from the fossil Vityaz trench, respectively. Interpreting the deflections of the 660 in terms of local temperatures implies a thermal anomaly of -800K to -1200K at 660 km depth. Except for the southern region where it may thicken, the width of the depressed 660 region implies that the Tonga slab seems to penetrate the 660 with little deformation. Waveform modeling suggests that the 660-km discontinuity is at most 2 km thick in many parts of the region, and a first-order discontinuity cannot be precluded. The 410-km discontinuity thickness shows somewhat more variability and range from 2 to 10 km ouside the slab and is at most 10 km thick within the slab.\par We use P-to-S converted phases from teleseims recorded at island and ocean bottom stations in Mariana to investigate the upper mantle structure in the region. We find evidence for double seismic discontinuities at the base of the transition zone near the Mariana slab. A shallower discontinuity is imaged at depths of ~650-- 715 km, and a deeper interface lies at ~740--770 km depth. The large lateral extent at near constant depths for both features is consistent with horizontal interfaces rather than small-scale scatterers. The amplitude ratios of the seismic signals suggest that the shear velocity contrast across the two interfaces is comparable. These characteristics support the notion that the discontinuities are the results of phase transformations in olivine (ringwoodite to post-spinel) and non-olivine component (ilminite to perovskite), respectively, for the pyrolite model of mantle composition.
DI53A-1093
Reconciling Global Stacks of PP and SS Waveforms With Mechanically Mixed Mantle Models
The opposite signs of the Clapeyron slopes render temperature dependent depths of the olivine to wadsleyite (at ~410 km) and ringwoodite to perovskite and ferropericlase transitions (at ~660 km). Seismological analysis of phase conversions and reflections demonstrate thicker and thinner transition zones in ‘downwelling' and ‘upwelling' regions, respectively. New thermo-dynamic calculations of phase equillibria in a mechanically mixed mantle (Xu et al., 2007) suggest that an upper mantle with increased basalt fractions renders diminished velocity jumps at the 410-km and 660-km discontinuities without affecting discontinuity depths. Using new global waveform stacks, we explore whether combined measurements of the reflection coefficient and traveltime of underside reflections at phase transitions in the upper mantle (i.e., PP and SS precursors) may enable us to constrain both compositional and ambient temperature variations in the upper mantle.
DI53A-1094
Revelations From S660S: New Interpretations of Topography on the 660 km Discontinuity
Currently, the only phase capable of mapping the 660 km seismic discontinuity globally is \textsl{S660S}, the shear phase that reflects off of the 660 km discontinuity and arrives as a precursor to \textsl{SS}. The topography of the 660 km discontinuity is an essential element in understanding the dynamics of flow between the upper and lower mantle. Both experimental and theoretical mineral physics studies find that the phase transition of γ-spinel to perovskite plus magnesiowustite occurs at approximately the depth of the observed \textsl{S660S} reflections. While the positive Clapeyron slope (~3 MPa/K) of the olivine to β-spinel transition responsible for the 410 km discontinuity is well constrained, the magnitude and even the sign of the γ-spinel to perovskite plus magnesiowüstite transition are still debated. It is generally accepted that the Clapeyron slope of the γ-spinel to perovskite plus magnesiowustite transition is negative with a similar magnitude to that of the olivine transition, although some studies indicate that it may be very small and perhaps barely positive (Bina and Helffrich, 1994). Therefore, the transition zone is predicted to thicken in cold regions such as subducting slabs and to thin in warm regions such as upwelling plumes. The mapping of \textsl{SS} precursors by Houser \textsl{et al.} (2007) reveals that the 410 and 660 km discontinuities are correlated on a global scale. For instance, the data indicate that the thin transition zone under the Pacific (a region of high data quality) is produced by a large depression of the 410 km discontinuity accompanied by a smaller depression of the 660 km discontinuity. If temperature anomalies are vertically continuous across the transition zone, then this correlation implies that both transitions have positive Clapeyron slopes. At temperatures above ~1800°C in aluminum-bearing pyrolite, γ-spinel transforms to majorite within the transition zone. Subsequently, the majorite to perovskite transition has a positive Clapeyron slope and occurs at depths of approximately 660 km (Weidner and Wang, 1998; Hirose, 2002). Furthermore, the phase loop of the majorite-perovskite transition narrows substantially above 1800°C, thus enhancing the sharpness of the transition (Hirose, 2002). Accordingly, our observations of S660S in these regions where it is correlated with \textsl{S410S}, are most readily explained by the majorite to perovskite transition producing the seismic discontinuity near 660 km depth. Thus, the correlation between these discontinuities indicates that transition zone thinning is a consequence of anomalously hot material at depths spanning from 410 to 670 km and aluminum contents that are compatible with those of fertile peridotite.
DI53A-1095
Fine Scale Imaging of Structure at and Near the Mantle Transition Zone Using a Generalized Randon Transform
The transition zone discontinuities, e.g. the '410' and '660', result from mineral phase changes that occur at depths constrained by temperature, pressure, and mineralogy, and detailed images of them can provide information about thermal and chemical variations in the upper mantle. We apply a generalized Radon transform (GRT), modified from underside core-mantle boundary imaging with SKKS (Wang et al., GJI, 2007), to image the transition zone discontinuities with the broadband wavefield (5-75 sec) containing precursors to SS. Previous studies of topography on the transition zone discontinuities mostly use stacks of SS data with lateral resolution of order of about 1000 kilometers. The GRT employs inverse scattering theory to detect and characterize perturbations in mass density and elastic parameters of a medium and can resolve lateral variations in structure at lateral scale lengths of 100 km. We map upper mantle discontinuities beneath the northwest Pacific Ocean with a lateral spatial sampling of one degree and a vertical sampling of five kilometers. We clearly detect interfaces near 410, 520, 660, and, more tentatively, near 800 km depth. In cross section, the '410', '520', and '660' reveal substantial variations in strengths, depth, and width. Moreover, the pulse shapes in the reflectivity profiles are frequency dependent. The '520' has larger topography than '410' and '660' and, locally, appears stronger than the '410'. Split pulses occur locally for '520' and '660'. We also detect (broader) scatterers outside what is traditionally considered to be the transition zone (e.g., near 800 km depth), but this signal shows significant lateral variation and may not represent globally continuous structures.
DI53A-1096
Topography of the 660-km discontinuity beneath northeast China: Implications for a retrograde motion of the subducting Pacific slab
Mapping the topography of the 660-km seismic discontinuity provides complimentary information to tomography images for understanding mantle circulation. Here we use S to P conversion waves from deep focus earthquakes to image fine-scale lateral variations in the depth of the 660 beneath northeast China, one ideal location to study interaction between a subducting slab and the upper and lower mantle boundary. Our method does not require using a reference velocity model of the heterogeneous upper mantle, and subsequently permits accurate estimate of the 660 depth. Multiple array data from eight deep events occurring beneath northeast China and the Japan Sea were processed with a 2nd root slant stacking technique. Clear S660P arrivals can be identified from stacked seismograms with incident angles well consistent with theoretical predictions. We measured the differential travel times between S660P and P with a cross correlation method and used them to compute the depths of the 660 with 3D velocity models. We found a rapid change in the depth of the 660 within a narrow longitudinal range of 130.8{°}E-131.4{°}E, coinciding with the place where the lower boundary of the subducting Pacific slab encounters the 660. East to the rapid transition, the depth of the 660 appears to be flat, with no statistically significant variations. Towards the west, the 660 deepens steadily as it approaches the cold core of the subducting slab. The maximum depression occurs at the west end of the studied region with an amplitude of ~20 km. Our estimate of the depth of the 660-km discontinuity is consistent with previous results of SS precursor studies, indicating that the large-scale depression and low-amplitude topography of the 660 in this area is probably a true feature. Our observations also imply a scenario for a relatively soft slab lying over the 660 progressively from west to east due to a continuous trench retreat associated with the retrograde motion of the subducted Pacific plate.
DI53A-1097
Fine Scale P-wave Structure of the Upper Mantle Discontinuities Beneath Northern Australia
P-wave triplications arise from rapid increases in upper mantle seismic velocity with depth. Waveform analyses of the triplicated wavefield are powerful for detailed characterization of the upper mantle phase transitions, owing to their sensitivity to the size and depth of the discontinuities, as well as the under- and overlying velocity gradients. However, P-wave triplication studies are inherently more difficult than for S-waves, and hence much less common, since P-wave arrivals are typically much closer in time near the triplication crossovers than for S-waves, e.g., <1-2 sec. Here we study the P-wave field at relatively high frequency (~ 1 Hz), using the Warramunga seismic array (WRA) in Australia. WRA has an aperture of about 20 kilometers and is equipped with 20 short- period vertical instruments. We collected 1863 events between 1990 to 1994 from surrounding trenches recorded by WRA, which have an epicentral distance range roughly between 12 and 25 deg. These data mainly sample the upper mantle beneath northern Australia. We employ Nth-root stacking and fk analyses to study the triplicated waves. For each event, a master trace was computed from the 4th-root vespagram, using a sliding window that collects maximum amplitude and associated slowness information. While some master traces exhibit considerable complexity, many show clear evidence for distinctly separate upper mantle triplication phases, and provide evidence for lateral variability in upper mantle discontinuity structure, as well as heterogeneities, beneath northern Australia.
DI53A-1098
Global Imaging of Upper Mantle Seismic Heterogeneities and Discontinuities From Observations of Body-Wave Travel-Times and Rayleigh Wave Dispersion.
The tomographic inversion of 100779 Rayleigh waveforms including higher modes has recently allowed us to constrain the SV-wave heterogeneities in the upper mantle with a lateral resolution of a few hundred kilometers and a vertical resolution of a few tens of kilometers. We are currently working to improve the resolution of our model in the transition zone by increasing the number and the accuracy of the higher modes measurements. In addition to the large-scale constraints provided by surface-waves on the 3D distribution of shear-wave velocities, we have accumulated global observations on the topography of the 410-km and 660-km discontinuities through the measurement of arrival times of converted and reflected body-wave phases. SS precursors phases can efficiently complete the analysis of P-to-s receiver functions to build a global image of the topography of the transition zone seismic discontinuities. We present results obtained from the separate analysis of each type of body-wave dataset. We are currently working on the simultaneous inversion of body waves and surface waves higher modes for the 3D distribution of SV-wave velocities and the depth of seismic discontinuities. We discuss the strategy for the inversion and present our preliminary results.
DI53A-1099
Lehmann discontinuity due to dehydration of phengite
It is known that the Lehmann discontinuity has several unique features. The depth of this seismic discontinuity is around 220 km depth. However, this discontinuity has not been detected everywhere at around 220 km depth. It is observed under continents more than as often as under oceans. An increase in the compressional or shear wave velocity and in seismic reflections has been reported. A regionally varying negative seismological Clapeyron slope has been estimated. A seismic transition from anisotropic to a more isotropic state occurs at depths corresponding to the Lehmann discontinuity. Although several models have been proposed to explain these features of this discontinuity, previous models failed to clear these unique features. Here, we propose a new model to explain the origin of the Lehmann discontinuity. We conducted experiments with hydrous sediment. The temperature was varied between 1073 and 1673 K, while a pressure of 6-15 GPa was applied using the multi-anvil press. The typical pelite composition was used as the sediment [1]. Garnet, clinopyroxene, and silica phases were present in all of the experiments. Three hydrous phases were observed at temperatures below 1573 K. The stable crystalline hydrous minerals consisted of phengite below 8 GPa, topaz-OH from 9-12 GPa, and phase egg above 12 GPa. The breakdown boundaries of topaz-OH and phase egg show a positive Clapeyron slope. In contrast, the breakdown reaction of phengite gave a negative slope at about 7 GPa corresponding to 220 km depth. The upper temperature limit for phengite is greater than 1473 K. This phase is thus likely to be stable within average adiabatic mantle conditions [1]. Above ~220 km depth the deformation mechanism of olivine is dislocation creep, which gives a preferred orientation to the crystals. The seismic anisotropies are likely to result from the preferred orientation of olivine. As pressure increases, the diffusion creep becomes dominant. If sediments migrate by the convective flow in the upper mantle and intersects the dehydration boundary of phengite, the released water migrates into the surrounding mantle rock and the sharp boundary between the anisotropic and the isotropic structure is formed at around 220 km depth, because water enhances the change in the deformation mechanism of olivine. The Lehmann discontinuity is likely to be related with this boundary corresponding to the change in the deformation mechanism. As this exists only in regions of stagnated sediment, the local detection of the discontinuity is consistent with the hypothesis of sediment dehydration. The most characteristic features of the Lehmann discontinuity, such as 220 km depth, negative seismological Clapeyron slope, local detection, seismic velocity jump, and the seismological anisotropic to isotropic transition can be reasonably explained by our new model [2]. [1] Ono (1998) J. Geophys. Res., 103, 18253-18267. [2] Ono (2007) The Open Mineralogy Journal, 1, 1-4.
DI53A-1100
Kinetics of the olivine-wadsleyite transition from time resolved synchrotron XRD
The mechanisms of the olivine transformation to its high-pressure polymorphs have wide implications for mantle dynamics; in particular for the origin of deep-focus earthquakes. In order to study the kinetics of this transition, α-(Mg,Fe)Si2O4 samples have been reacted in the stability field of wadsleyite (β phase), at 12.5-15 GPa and temperatures up to 1100 K, in a Paris-Edinburgh press. Product phases and extent of transformation were monitored as a function of time by synchrotron X-ray diffraction at ESRF (beamline ID27). The sequence of XRD spectra reveals the early formation of a spinel-structured phase, as observed in previous quenched experiments. This spinel phase further transforms to β phase with increasing temperature and/or time, so that it always disappeared before completion of the reaction. Transformation-time curves are constructed from the relative proportions of the α, β and spinel phases determined by Rietveld refinement of XRD patterns. The kinetic data are analyzed in terms of the Avrami's theory of nucleation and growth in order to determine reaction rates. The influence of sample microstructure and differential stress on transformation rates are also investigated, as well as the effects of iron and water contents. In the light of these kinetic data, we discuss the persistence of a metastable wedge of olivine in the cold interior of subducting slabs.
DI53A-1101
The Post-Perovskite Transition in NaMgF3 Measured Under an Ar Medium
Neighborite, NaMgF3, is isostructural to MgSiO3 perovskite and has been shown to transform to the CaIrO3 type (post-perovskite) at much lower pressure and room temperature. We have conducted X-ray diffraction of the post-perovskite transition in NaMgF3 using argon as a pressure medium at the GSECARS sector of APS and beamline 12.2.2 of ALS. We observe the first appearance of post-perovskite occurring between 37 and 42 GPa during compression at room temperature and a mixed phase of perovskite and post-perovskite was observed up to 49 GPa at room temperature, whereas a complete transition to post-perovskite was documented by previous studies at 19-30 GPa and room temperature with a Si-oil, NaCl, or MgO pressure medium. This difference in pressure may indicate that deviatoric stresses are an important factor in explaining the different observations. In situ double-sided laser heating at GSECARS demonstrated that perovskite transforms completely to pure post-perovskite at ~2000 K and ~47 GPa. We did not observe a new high-temperature phase, N-phase, proposed by Martin (2005) at 37 GPa after heating to 2000 K. We also found that the volume decreases by 2±1% across the post-perovskite transition in NaMgF3.
DI53A-1102
Elastic Moduli of Single-Crystal Orthoenstatite From Room Temperature to 1450 K
Orthopyroxene [(Mg,Fe)2Si2O6] is commonly considered to be one of the four major minerals in Earth's upper mantle. Thus, data on the elastic properties of orthopyroxene over wide ranges of temperature and pressure are necessary to develop reliable models of the composition and structure of the upper mantle. New elasticity data are provided on the nine independent adiabatic elastic moduli of orthoenstatite (Mg end-member orthopyroxene) from room temperature to 1450 K at ambient pressure. These data were obtained using the resonant ultrasound spectroscopy (RUS) with a natural single-crystal specimen; they extend by 380 kelvin the temperature range for which the elastic moduli of single-crystal orthoenstatite have previously been reported. Broad agreement in the temperature dependences of the nine Cij's is found when comparing our results with those from a Brillouin spectroscopy study of orthoenstatite up to 1073 K (Jackson et al., PEPI, 161, 1- 12, 2007). An earlier report (Jackson et al., Am. Mineral., 89, 239-245, 2004) identified marked non-linear temperature effects in the C33 and C55 moduli and correlated this nonlinearity to high-temperature mode softening in orthoenstatite at high temperature. Similar non-linear effects in the C33 and C55 are seen in the current RUS experiments and are carefully documented up to 1450 K in intervals of 25 kelvin. The current RUS study also reveals conspicuous non-linear trends in the temperature dependences of two of the three off-diagonal moduli, C13 and C23, above 1000 K. These results are interpreted in terms of the high-temperature isotropic properties of orthoenstatite, and their impact on our understanding of the properties of Earth's upper mantle is discussed.
DI53A-1103
Cation Disorder in Ringwoodite and its Effects on Wave Speeds in the Earth's Transition Zone
The structure and stability of the Mg2SiO4 polymorphs play a key role in the density and impedance contrasts across the boundaries in the Earth's transition zone. Lateral variations in temperature in the transition zone lead to wave speed variations due to structural softening of the minerals. Spinel-structured Mg2SiO4, ringwoodite, also can contain significant amounts of thermally activated Mg-Si disorder. First-principles, density functional theory calculations on Mg2SiO4 -spinel, inverse spinel, and disordered structures provide constraints on the effects of cation disorder on the elasticity and stability of ringwoodite in the Earth's transition zone. Mg-Si disorder can be described with a regular solution parameter of 14 kJ/mol, leading to the prediction of ~4% Mg-Si disorder in the mantle's transition zone, where 200 K lateral temperature variations can cause approximately 1% change in disorder. Because the base of the transition zone is likely hotter than the laboratory synthesis temperature of most ringwoodite samples, such samples likely contain a greater degree of cation ordering than is found in the mantle. Independent of temperature and compression, 1% cation disorder causes a decrease in c11 and c44 of 0.65% and 0.4%, respectively, leading to cation-disorder induced variation in vp and vs with dvs/dvp=1.4. Measurements made below 1000 K on the effects of temperature on elastic properties are made with approximately constant levels of disorder due to the slow kinetics of cation ordering at such temperatures. Neglecting the effects of cation disorder then leads to a 20-50% underestimate of the effect of temperature on elastic constants and wave speeds, and thereby overestimating the impedance contrast across the wadsleyite to ringwoodite transition.
DI53A-1104
Olivine-Ringwoodite Transformation Kinetics Suggest that the Deep Marianas and Pacific Slabs have less than 90 ppm H2O
Hydrogen increases the growth rates of the metastable olivine to ringwoodite transformation. These ringwoodite growth rates determine the likelihood of a metastable olivine wedge persisting into the Earth's mantle transition zone (410 to 660 km depth) in subduction zones. Seismological evidence for metastable olivine coincides with the observed locations of deep earthquake hypocenters in subducted slabs [Kaneshima, et al., 2007; Iidaka and Suetsuga, 1992], and metastable olivine is likely associated with mechanisms causing deep earthquakes. Additionally the presence of a metastable olivine wedge adds negative buoyancy to the slab which must be accounted for in precise determination of subduction forces. Furthermore the presence or absence of a metastable olivine wedge may constrain the water content of subducting lithosphere in the Earth's transition zone. Previous results have indicated that olivine containing as little as 290 wt- ppm D2O (an H2O proxy) will transform too quickly for a metastable wedge of olivine to survive into the Earth's transition zone [Diedrich, et al., 2007]. In this study, we have investigated the olivine-ringwoodite transformation kinetics using olivine hydrated with approximately 90 wt-ppm H2O. At 18 GPa and 900 celcius, this olivine transforms quickly, with a ringwoodite growth rate of 2.4e-9 m/s, which is nearly identical to growth rates for olivine with 290 wt-ppm D2O at the same P and T. The activation enthalpy for ringwoodite growth in these samples (140 kJ/mol) is also approximately the same as in the 290 ppm D2O samples. Based on the thermo-kinetic models in Diedrich et al. [2007], approximately 90 ppm H2O in olivine within old, cold, and fast subducting slabs would eliminate the metastable wedge of olivine. Conversely, these results imply that slabs exhibiting seismological evidence of a metastable olivine wedge, such as the subducted Mariana slab [Kaneshima, et al., 2007] and the subducted Pacific slab [Iidaka and Suetsuga, 1992], have less than 90 ppm H2O at transition zone depths.
DI53A-1105
Kinetics of Olivine Phase Transformation and the Role of Water
The metastability of olivine is still not very clear as yet. Main reason is that experimental kinetic data on olivine are few. Mostly people employ the kinetic data of analogs to understand olivine's metastability. Among them, more attention has been paid to the results of forsterite. Water has great influence on forsterite's metastabiliy. Recently Hosoya et al. modified the growth equation by introducing a pre-exponential term as power-law function on OH content. Consequently, they refined the kinetic parameters, studied the role of water on forsterite's metastability, and reckoned its existence in earth's deep interior. Here we classify available in-situ growth data of Mg2SiO4 olivine with measured water contents into three categories, i.e. the "dry", "wet", and "saturated" in which the average water content are about 800, 3000, and 5000 wt. ppm respectively. Results reveal that the traditional growth theory is useful and the pre-exponential coefficient of the rate equation has little relation with OH content. Refined values are in accordance with the theoretical vision. This fact lends us a robust constraint for determining the kinetic parameters. Accordingly we refine the activation enthalpy of Mg2SiO4 olivine for different OH content, and examine the influence of OH content on metastability of Mg2SiO4 olivine. At last we assess the metastability of the major mantle mineral (Mg0.89Fe0.11)2SiO4 olivine based on a few kinetic data. Results show that there is much difference in metastability between forsterite and olivine and there would exist observable metastable olivine although it might not reach 660-discontinuity.