T32A-01 10:20h
Mantle \emph{S}-velocity structure of the Mediterranean region using joint tomography of surface and \emph{S}-body waves
We present a new 3-D \emph{S}-velocity model for the upper mantle beneath the Mediterranean region. This new model is based on a joint inversion of surface waveform fits and body-wave arrival time data. We use the waveform fits of Marone et al. (2003), cross correlation-derived relative delay times from teleseismic events recorded by MIDSEA and other seismic stations, and absolute delays from the reprocessed ISC-catalog (Engdahl et al.,1998). We further include all available results from receiver function and reflection/refraction studies to have independent constraints on the depth of the Moho. It is shown that surface wave data and body wave data can be combined to derive a better constrained model of velocity perturbations than is possible by inversion of just a single dataset alone. The model is consistent with both data sets to within four standard deviations. The teleseismic data increase lateral resolution in station-pupolated regions, add resolving power for the uppermost lower mantle, and increase it for the transition zone, resulting in sharper high-velocity features compared to the model derived from regional S and surface wave trains only (Marone et al., 2004). The new model shows high velocities related to past and present subduction beneath the Hellenic trench and southern Italy. High velocities at transition zone depths are observed beneath Greece and Eastern Spain, most likely related to past subduction. The Tyrrhenian Sea shows a high-velocity anomaly at the very bottom of the transition zone.
T32A-02 10:35h
Scattered Wave Imaging of the Lithosphere and the Asthenosphere Beneath Eastern North America
To better understand the properties that define the lithosphere and the asthenosphere beneath continents, we have analyzed P-to-S phases produced by scattering beneath permanent seismic stations HRV, PAL, LMN, LBNH, BINY, and SSPA in eastern North America. Our prior work has shown that a strong (3-11%) drop in shear-wave velocity occurs at depths (90-110 km), consistent with the westward dipping lithosphere-asthenosphere boundary imaged by surface-wave tomography. Inversions of data at LMN and HRV require that this velocity gradient be sharp and occur over less than 11 km in depth. This result indicates that the lithosphere-asthenosphere boundary is not defined by temperature alone and suggests the presence of water or partial melt in the asthenosphere. To investigate the properties of the asthenosphere at greater depths, we are now investigating later scattered arrivals at these stations. A strong phase in the ~~290 km depth range is observed at stations PAL and HRV. Although modeling discontinuities at such depths is often hindered by the interference of reverberations from shallower discontinuities, the phase at PAL appears to be uncontaminated, and we have been able to model the ~~290 km depth phase at HRV by carefully including shallower discontinuities and their reverberations. This ~~290 km depth phase appears on both SH and SV migrated waveforms, and analysis of these particle motions demonstrates that the phase has undergone shear-wave splitting. We plan to further investigate data at HRV and PAL as well as other stations in the region to determine the implications of this phase for the depth distribution of anisotropy and the velocity gradient structure at the base of the asthenosphere.
T32A-03 10:50h
Understanding Low Velocity Zones and Melt Reservoirs
Low velocity zones within the mantle are poorly characterized seismically, especially in the upper mantle. They do, however, play a significant role in understanding mantle convection and the Earth's thermal evolution. In the recent years numerous investigators [Revenaugh and Sipkin, 1994; Thybo and Perchuc, 1997; Vinnik and Farra, 2002; Vinnik et al., 2003; Savage et al. 2003; Song et al., 2004] have reported substantial velocity decreases between the crust and 410 km discontinuity that have been interpreted as sublithospheric partial-melt zones. Encompassing all tectonic settings, these studies have relied upon a varying techniques from receiver functions to waveform modeling in order to identify the velocity decreases. We seek to develop a systematic procedure that is optimized for detecting sublithospheric low velocity zones that may reflect the presence of partial melt. As a first step, we use the reported low velocity zone models noted above, and create synthetic data sets in 1,2,and 3 dimensions to determine the distinguishing characteristics of low velocity zones when probed with a variety of techniques. It is well known that first-arrival travel times are not an appropriate data set for low velocity zone investigations. Rather, wavefield propagation in the horizontal (later-arriving phases from local and regional earthquakes) and the vertical (receiver function and mantle reflectivity profiles from teleseisms) directions provide important and complementary constraints on the thickness, lateral extent as well as the absolute velocity of these regions. This investigation aims to combine these two approaches to provide useful characteristics of low velocity zones so they may be mapped worldwide in conjunction with upper mantle heterogeneity. Further, it has been suggested [Silver et al. 2004] that large, long-lived reservoirs of melt may exist just beneath old cratonic lithosphere, and are hypothesized to source for the flood basalts created during catastrophic diking events. Our initial areas of interest are thus cratonic areas possessing broadband seismic data from regional and teleseismic events, such as the Kaapvaal craton in southern Africa and cratonic South America. Using these data, we test for the presence of melt reservoirs by looking for extensive low velocity zones in the sublithospheric upper mantle.
T32A-04 11:05h
Geophysical Evidences for Chemical Variations in the Australian Continental Mantle
The relative density-to-shear velocity scaling ($\zeta$) provides a diagnostic for the presence of compositional variations in the mantle. We invert shear-wave velocity from the recent 3-D model AUS04-Vs and gravity anomalies EGM96 for radial profiles of $\zeta$ of the uppermost mantle beneath Australia. We performed calculations for the three major tectonic provinces that constitute the continent, and found significant differences between them. The $\zeta$ profile for the Phanerozoic region can be explained by thermal variations alone. In contrast, negative values of $\zeta$ suggest that variations in composition are important between $\sim$75 and $\sim$150 km depth in the Proterozoic continental lithosphere (central Australia). It is likely that chemical variations are also required to explain our inferences for the Archean craton (west Australia), but poor tomographic resolution precludes a definitive conclusion. The scaling factors found are consistent with chemical depletion of deep Precambrian lithosphere, which supports a tectosphere model for the Australian continental roots.
T32A-05 INVITED 11:20h
Seismic Anisotropy and Flow in the Oceanic and Continental Upper Mantle: Inferences from SKS Splitting Observations
Seismic anisotropy provides a direct estimate of flow in the Earth's upper mantle. In this study, we examine measurements of anisotropy inferred from shear-wave splitting at continental and oceanic island stations. In the ocean basins, the splitting fast polarization directions are extremely well fit by a global flow model that is driven by a combination of plate-motion and mantle density heterogeneity. Global mantle flow is determined from instantaneous flow calculations that assume a radially variable, but laterally homogeneous viscosity structure. In general, the misfit between the observed fast polarization direction and the predicted direction of maximum shear in the asthenosphere is very small: on the order of the error in the splitting measurements for both fast- and slow-spreading ocean basins. In particular, a flow model that incorporates active upwelling from the lower mantle associated with the African superplume is required to explain the splitting data in the southern Atlantic and Indian Oceans. However, while this global flow field can account for the anisotropy in the ocean basins, it is a very poor fit in many continental regions. For example, although observed anisotropy in western North America is well fit by the global flow model, the splitting observations in eastern North America are nearly orthogonal to the model predictions. We hypothesize that this poor fit is caused by either the influence of continental roots on the mantle flow field, a component of fossil lithospheric anisotropy, or some combination of these two effects. These competing models will be evaluated by 1) estimating mantle flow from finite element calculations that incorporate radial and lateral variations in viscosity associated with the depth- and temperature-dependence of mantle viscosity, and ocean-continent and age-dependent variations in lithospheric thickness, and 2) examining the continental splitting data for evidence of a fossil lithospheric component of anisotropy.
T32A-06 11:40h
The nature of the Lehmann discontinuity from its seismological Clapeyron slopes
While the existence of the Lehmann discontinuity at 220 km in Earth's mantle has been known for over 40 years from seismology, it is still debated what causes this discontinuity. The nature of the Lehmann discontinuity is of major importance for our understanding of upper mantle composition and flow. Here we report measurements of seismological Clapeyron slopes for the Lehmann discontinuity, which are key to explaining the seismic discontinuity as either a manisfestation of phase transitions or of other mineral physical processes. The Clapeyron slopes are measured by correlating discontinuity depths with local velocity perturbations from a tomographic model, assuming that the velocity perturbations are solely due to temperature variations. We find that in most regions the Lehmann discontinuity is characterised by a regionally varying negative seismological Clapeyron slope. Known phase transitions in the upper mantle above 400 km depth all have positive Clapeyron slopes. In the case of the Lehmann discontinuity the only remaining hypothesis for a negative Clapeyron slope is that it represents the transition in deformation mechanism from dislocation to diffusion creep. This corresponds to a change from anisotropic structure above, to isotropic structure below the discontinuity and thus is independent evidence for the maximum depth extend of upper mantle anisotropy. The depth at which this transition occurs is dependent on water content, grain size, stress level and temperature (Karato, 1992). In a dry upper mantle, the transition will appear at 340-380 km depth, which is too deep to explain our observations of the Lehmann discontinuity from 220 km depth. The transition depth is shallower in a wet mantle, so our observations are also an indicator of the existence of a significant amount of water in the Earth's upper mantle. Alternatively, a smaller grain size or lower stress level would lead to a shallower discontinuity. The regional variability in the size of the Clapeyron slope could thus be due to variations in water content, grain size or stress level.
T32A-07 INVITED 11:55h
Strain and anisotropy in the deep upper mantle and transition zone
Seismic anisotropy has proven to be a powerful tool to image the uppermost mantle deformation. Recent experimental data on flow mechanisms of mantle minerals under high-pressure conditions allow now to relate seismic anisotropy observations and deformation deeper in the mantle. We use forward models based on high-pressure experimental data on olivine, wadsleyite, and ringwoodite to predict the seismic anisotropy produced by plastic strain in the deep upper mantle and in the transition zone. For the upper mantle below 250 km, polycrystal plasticity simulations with dominant [001] slip in olivine produce CPO that result in an extremely low seismic anisotropy characterized by fast directions roughly normal to the flow direction. Transition at high pressure from dominant [100] to [001] glide in olivine may explain the variation with depth in P and S waves anisotropy patterns even if the entire upper mantle deforms coherently with a dominant horizontal shearing component. The present results challenge previous interpretations of the weak seismic anisotropy in upper mantle below 250 km as resulting of deformation by diffusion creep or of poor coherence of the deformation at seismic length-scales. For the transition zone, forward models using experimentally determined [100] and 1/2[111] slip systems for wadsleyite and [110] slip for ringwoodite predict a weak seismic anisotropy for a polycrystal of pyrolitic composition at upper transition zone conditions: ~2% for P and ~1% for S-waves, and a roughly isotropic behavior in the lower transition zone. Analysis of global observations of seismic anisotropy in the transition zone in the light of these models supports dominant horizontal flow also in the uppermost transition zone, suggesting coherent flow of the upper 520 km of the mantle.