T13A-0422
Magnetic anisotropy of the subcontinental mantle: a proxy for seismic anisotropy?
Our knowledge of the petrology and seismic properties of the subcontinental mantle is deduced mostly from three sources: (i) large scale seismic anisotropy investigations; (ii) non-oriented mantle xenoliths and (iii) rare peridotite bodies such as the Ronda massif. These three sources provide information at different scales, from grain size (xenoliths and peridotitic bodies) to tens of kilometers (seismology). Xenoliths exhibit lattice-preferred orientation (LPO) due to plastic flow in the mantle, which, in turns, results in seismic anisotropy. Similarly, seismic waves propagating beneath continents display variations in speed and polarization direction attributed to the mantle deformation history. An implicit correlation is generally assumed between the two anisotropies measured at these very different scales. A new magnetic technique, investigating an intermediate scale, has been developed in the past few years. This technique measures the anisotropy of magnetic susceptibility in high field, using a vibrating sample magnetometer to saturate the ferromagnetic minerals. Therefore it effectively isolates the paramagnetic and diamagnetic susceptibility anisotropy. Magnetic measurements, performed on various high field instruments (vibrating sample magnetometer, torque magnetometer, cantilever magnetometer), both on natural and synthetic samples, confirm the need to conduct measurements in high field in order to isolate the intrinsic paramagnetic and diamagnetic properties of olivines, orthopyroxenes and clinopyroxenes. These measurements also demonstrate the orthorhombic nature of the intrinsic magnetic properties in these silicates, and yield new data concerning the relationship between crystallographic axes, magnetic anisotropy and other physical anisotropies. Seismic anisotropies of a suite of Ronda peridotites are calculated, using a forward modeling approach (Anis2K) and measured LPOs. The same LPOs are used to calculate magnetic anisotropies, using new single crystal magnetic data and a similar approach. Finally, calculated magnetic and seismic anisotropies are compared with measured magnetic anisotropies to evaluate the potential of magnetic fabrics as a proxy for seismic anisotropy in the subcontinental mantle.
T13A-0423
A Natural Example of Olivine LPO Variation With Shear Strain
Understanding mantle rheology, seismic anisotropy, and strain localization requires knowledge of the evolution of olivine lattice preferred orientations (LPOs) as a function of strain and melt content. The Josephine Peridotite in southwest Oregon is ideal for analysis of the interaction of deformation and melt, due to the presence of shear zones with associated melt migration structures. We present results on the variation of olivine LPO with shear strain in these shear zones. The LPO evolution with proximity to the shear zone center provides a link between experimental data and LPO evolution models, which can be applied to the interpretation of mantle seismic anisotropy and to models of melt extraction at mid-ocean ridges. Results for harzburgites from the largest Josephine shear zone, which is 100m wide and contains syn-deformational dunite, indicate that the olivine LPO rotates in the shear zone so that the [100] maxima lies parallel to the shear direction. Outside of the shear zone, the harzburgites have a pre-existing LPO. Shear strain for the Josephine samples is calculated from the deflection angle of the regional pyroxene foliation. By ~100% shear strain, the olivine LPO evolves so that the [100] axis is aligned with the flow direction. These results on olivine LPO evolution during shear agree with the experimental data of Zhang and Karato (1995). They extend observations of olivine LPO to much higher strains and lower stresses. At face value, the Josephine shear zone data also agree with predictions from polycrystal plasticity models that incorporate the effects of dynamic recrystallization (Ribe and Yu, 1991; Wenk and Tomé, 1999). The variation of olivine LPO during simple shear is important for correctly interpreting seismic anisotropy. Our results indicate how much strain is necessary for olivine LPO to align in the flow direction. In conjunction with modelling studies, this information can be used to interpret the kinematics of deformation in regions where mantle anisotropy varies spatially.
T13A-0424
Magnetic Properties of Mantle Xenoliths and Evidence of Localized Modification of the Mantle Beneath the Rio Puerco Volcanic Field, New Mexico
Little is known about the magnetic petrology and processes that affect the magnetization of the upper mantle. Petrologic and geochemical studies of a suite of xenoliths from the Rio Puerco volcanic necks (RPVN), west-central New Mexico, show that pyroxenites (PYX) have a metasomatic origin, as a result of interaction between spinel lherzolites (SL) and basaltic and carbonatitic melt or fluid. This study demonstrates that magnetic properties of these mantle xenoliths can characterize localized mantle modification events and heterogeneity in mantle oxidation states. In situ, oriented PYXs carry a well-defined post-emplacement, cooling-related remanence (typical NRM of 0.23 A/m) defined by progressive thermal and AF demagnetization. Thermal demagnetization of SL and PYX remove >90% of the magnetization by $580°C and IRM acquisition curves reach saturation by 0.3T, indicating a dominance by magnetite in both rock types. SL and PYX have relatively small concentrations (~0.01%) of magnetite (bulk susceptibility of 10$^{-4}$ to 10$^{-5}$ SI vol). SLs generally contain multi-domain magnetite (mean destructive fields of NRM between 20 to 40 mT), whereas PYXs are dominated by single domain magnetite (MDFs between 20 to 70 mT). The magnetic properties of SLs and PYXs are a reflection of phases formed in the mantle and not from basalt-xenolith interaction en route to the surface. In addition, the differences in magnetic properties give insight into how melt infiltration modifies the magnetization of mantle xenoliths. In comparison to other SLs, red-colored SLs found only at Cerro de Santa Rosa, one of the RPVN, contain hematite and relatively low-coercivity magnetite. Complete thermal unblocking of a high coercivity phase occurs at $680°C and a medium to low-coercivity fraction at $580°C. Textural evidence suggests that alteration involved oxidation in the mantle, prior to transport of these xenoliths to the surface in the host basalt. TEM analyses reveal micron-sized needles of amorphous silica and magnetite within olivine, indicating an oxidation reaction at or close to the QFM buffer. However, hematite formation in the mantle implies that the oxidation state reached the HM oxygen buffer. We infer that the unusual oxidation state in the mantle was highly localized, based on the isolated occurrence of the red SL xenoliths. The oxidation agent is interpreted to be a CO2-rich phase, consistent with the conclusion that carbonatitic melt or fluid related to incipient Rio Grande rifting was present beneath the RPVN.
T13A-0425
The Exposed Moho In the Kohistan Arc: Seismic Properties At High Temperature And Pressure
The exposed crustal and mantle section in the Kohistan paleo-island arc in Pakistan is one of the most continuous and fresh example of exposed Moho. The geometrical relationships between mantle rocks and lower crustal intrusives and metamorphites are well preserved, making this section a natural laboratory for geological and geophysical investigation. In this study we investigated the relationships between the petrological and seismic Moho at geological relevant conditions, by measuring the speed of ultrasounds in rocks samples at high pressure and temperature. The compressional wave velocities were measured at confining pressures of up to 0.5 GPa and temperatures of up to 1200°C in an internally heated gas medium apparatus (Paterson rig) in three orthogonal cores per sample, cut parallel or normal to the mineral lineation and the foliation normal. Measurements were done on dunites, pyroxenites and wherlites representative of the upper mantle; on garnetites, amphibolites and gabbros representative of the lower crust. Interestingly, both the pyroxenite and the garnetite developed partial melting at temperatures greater than 1100°C at about 0.5 GPa confining pressure. During the melting we did not observed a remarkable reduction in the seismic velocities, but a remarkable reduction in the seismic amplitude, suggesting a large variation in seismic attenuation. The geological implication of our findings is that at (or above) the Moho, the presence of partially molten rocks might be detected by anomalies in the P-wave attenuation.
T13A-0426
Electrical Anisotropy Across the Kaapvaal Craton Derived From SAMTEX MT Studies Compared With SASE SKS Seismic Anisotropy Results
Data from part of the Southern African Magnetotelluric Experiment (SAMTEX) have been analysed using an impedance tensor decomposition method in order to gain a better understanding of the geoelectric directionality and dimensionality and electrical anisotropy properties of the region. The SW-NE transect from Sutherland to Messina, in South Africa is coincident with part of the Southern Africa Seismic Experiment (SASE) deployment. Shear-wave splitting observations, from SKS analyses, have been conducted on the SASE data, where it was observed that splitting delay times appear to exhibit geological control, with the Neoarchean regions showing greater anisotropy and the older Mesoarchean core in the SE exhibiting no detectable anisotropy. In contrast however to the inherent depth ambiguity with the SKS method, MT has the ability to place relatively accurate bounds on the depth at which the anisotropy is most significant, making for a very complementary study. The electrical anisotropy is derived in terms of approximate depth, rather than frequency. Our results for upper lithospheric mantle depths (40-100 km) indicate very similar northeast-southwest strike directions to those inferred from the SKS shear-wave splitting analyses. However our results for crustal depths (5-40 km) have a distinctly different orientation, thereby further validating the interpretation that the causative region for seismic anisotropy is in the upper lithospheric mantle, i.e., it is a fossil anisotropy, and not at the base of the lithosphere nor in the crust. Not surprisingly, the crustal electrical anisotropy directions appear to be strongly controlled by surface geological structure. As with seismic anisotropy, the electrical lithospheric mantle anisotropy likely indicates Archean deformation that occurred during the formation of the Kaapvaal and Zimbabwe cratons, and has been preserved in the lithospheric roots typical of Archean cratons.
T13A-0427
The Electrical Lithosphere of the Archean: Insights from the Kaapvaal craton and elsewhere.
Deep probing measurements of electrical conductivity, made with the magnetotelluric (MT) method, offer important insights into the structure of the lithosphere, as conductivity is sensitive to variations in thermal structure, composition, and grain size, and the interconnectivity of a conducting phase, such as fluid content, presence of metallic oxides, graphite, and sulphides. These variations map fossil fabrics and structural geometries of the lithosphere, from which formation and deformation processes can be inferred. The Kaapvaal craton is arguably the best studied craton in the world, featuring extensive seismic and geochemical coverage. Over the last two years, as part of a project named SAMTEX (Southern African Magnetotelluric Experiment), we have collected an extensive MT data set across the Kaapvaal craton and its immediate surroundings to complement those other geoscientific studies. The MT data set provides insights into the regional electrical structure to depths of several hundred kilometers, including the depth distribution of anisotropic fabrics. Together with other data sets, these new measurements allow us to understand better how the craton was assembled. Results from the Kaapvaal craton can also be compared to those from other cratons, such as the Slave and Superior cratons of Canada, where similarly extensive and high quality electrical data sets exist. We can also look at differences between Archean and Proterozoic terranes. We present initial results from the Kaapvaal craton and discuss these in the light of other cratons and lithospheric structure in general.
T13A-0428
Electromagnetic Images of the Subcontinental Mantle: Shining a Light on Archean Mantle Processes.
Electromagnetic methods such as those obtained by magnetotellurics provide an image of modern and ancient subduction zones. MT images complement those obtained from earthquake and active-source seismic studies of the crust and mantle at subduction zones due to the inherent ensitivity of EM methods to: saline fluids released from the subducting slab; electrically-conductive sedimentary rocks scraped from the subducting oceanic slab and imbricated in the overlying crust and mantle; and hydrated rocks in the mantle above the subducting slab. A dramatic north-dipping electrical structure is observed penetrating into the mantle lithosphere at a boundary between two Meso-Archean crustal blocks within the western portion of the Archean Superior craton of North America. Taken together with seismic images, and the observation of magmatic rocks similar to those observed in modern subduction settings, the resistivity image provides evidence that old subcontinental lithosphere was affected by subduction processes similar to those occurring in modern plate tectonics. Such processes provide a viable explanation for formation of buoyant continental lithosphere required to explain global preservation of Archean cratons.
T13A-0429
Towards an Integrated Seismic Characterization of the Slave Craton
Archean cratons form the core of the majority of Earth's continents and offer a unique window into the evolution of continents and plate tectonics over geological time. The dynamics that led to the evolution and stabilization of cratons over one billion years ago, however, remain poorly understood. The Archean Slave province, located in the NW Canadian Shield, is an ideal site to study the formation of cratons due to its high degree of preservation and petrological evidence that its lithosphere possesses a distinct stratification resulting from cratonic assembly. The last decade has witnessed an explosion of seismological work in the region, with more than 45 broadband seismic stations deployed over variable lengths of time by the University of British Columbia, the Geological Survey of Canada, the POLARIS consortium and MIT. These data have been subjected to a wide array of seismic analyses: body- and surface-wave tomographic inversions were applied to the complete dataset, whereas receiver functions and shear-wave splitting were applied to subsets of the data. When considered together, these results yield an unprecedented seismic characterization of the Slave province. The Slave's lithosphere is, on average, 200 km thick and displays seismic velocities that are ~2-3% faster than surrounding Proterozoic orogens and ~2% faster than average cratonic values. At smaller scales, a low velocity anomaly centered to the south of the Lac de Gras kimberlite field is observed between 50-300 km depth. The anomaly has a radius of ~100 km, it exhibits a 2.8% slowness contrast with respect to the surrounding mantle, and may represent post-stabilization alteration of the cratonic lithosphere by processes responsible for kimberlite magmatism. Coherent results from shear-wave splitting and surface-wave analyses show evidence for two layers of anisotropy beneath the Slave craton: one in the uppermost lithosphere that may be associated with crustal structure from the last episode of regional deformation; and another one in the mantle lithosphere and possibly the asthenosphere, with a principal axis aligned with the direction of absolute plate-motion. Receiver functions show evidence of finer-scale anisotropic layering throughout the Slave's lithosphere $-$ a possible sign of cratonic assembly by processes of shallow subduction and underplating. The seismic signature of the Slave craton therefore contains information about the entire evolution of the region, from its initial assembly during the Archean to its current deformation in response to plate motion.
T13A-0430
Variations in Lithospheric Thickness Across the Superior Province, Ontario, Canada: Evidence from Tomography and Shear-Wave Splitting
The Superior Province of the Canadian Shield is the largest contiguous region of Archean crust. It is made up for the most part of east-west trending belts of plutonic, metasedimentary, granite-greenstone, and gneissic rocks; these belts are interrupted by the northeast-trending Kapuskasing Structural Zone which cross-cut the subprovinces of the Superior at circa 2.6 Ga, representing a region of substantial uplift. The degree to which the crustal structure of the Superior Province is reflected in the mantle lithosphere is not well understood, as past seismic studies of the Superior were necessarily somewhat piecemeal in scope. With the advent of the FedNor seismic array, which provides broadly-spaced seismometer coverage over much of the Superior Province in Ontario, it is now possible to examine the lithosphere of a large portion of the Superior using passive seismic techniques. We combine datasets from the FedNor and CNSN arrays with data from previous temporary deployments in the region (the APT89, Abitibi and TW\~{}ST experiments), and examine the mantle beneath the Superior using shear-wave splitting and travel-time tomography. The shear-wave splitting shows a consistent difference between the eastern and western Superior: the western Superior exhibits very large SKS splits (averaging 1.4 seconds) with a consistent ENE fast direction, while the eastern Superior exhibits more-variable (E to NE) fast directions, with smaller split times averaging 0.8 second. Travel-time tomography shows an overall pattern of higher velocities in the western Superior, with the anomaly interpreted by Sol et al. (2002) as a remnant slab possibly representing the eastern edge of the high-velocity region. In the eastern region, overall velocities are lower; we observe the linear low-velocity feature previously interpreted by Rondenay et al. (2000) as the track of the Great Meteor hotspot. A larger low-velocity anomaly in the centre of the model remains enigmatic, though its resolution will improve with the inclusion of additional data from new stations. Interpretation of these results is in progress; one possibility is that the mantle lithosphere beneath the western Superior is well-preserved, including remnant imbricated high-velocity slabs, while the lithosphere in the eastern region has been more extensively disrupted, perhaps due to thermo-mechanical erosion by a mantle plume and processes related to the Kapuskasing uplift. The large shear-wave splits in the west would then represent the combined effect of aligned lithospheric and asthenospheric fabric.
T13A-0431
Three-Dimensional Structure in Southern Ontario via Receiver Function Imaging
The Precambrian bedrock in Southern Ontario is part of the Grenville province, a complex orogenic belt of circa 1.1 billion years in age that truncates several older geological provinces. In this study, two approaches (common conversion point [CCP] stacking and scattering tomography) were used to characterize 3-D crustal and upper mantle structure across southern Ontario, in a region of approximately 600 by 500 km, based on teleseismic receiver functions recorded by 29 three-component broadband stations deployed by the POLARIS (Portable Observatories for Lithospheric Analysis and Research Investigating Seismicity) array and CNSN (Canadian National Seismograph Network). Our preliminary results from CCP stacking clearly delineate the crust-mantle boundary and several upper mantle structures as continuous, nearly flat features. We detect the Moho at a depth of around 41-50 km, as well as a sub-crustal negative-polarity arrival at 60-80 km which might be correlated with subduction occurring during orogeny. Higher resolution images of crustal structure were obtained using scattering tomography. The Ottawa-Bonnechere Graben can be identified in the NE corner of the study area; as well, southeast dipping features and major crustal discontinuities, including the Moho, agree in some regions with previous controlled-source studies. The two imaging methods used complemented each other, scattering tomography providing greater crustal resolution, while CCP stacking images greater depth. The CCP and tomographic results disagree in some areas, and do not perfectly match previous studies; interpretation of large-scale 3-D structure in southern Ontario is in progress.
T13A-0432
Probing the Upper Mantle Beneath Northern and Eastern Ontario
Analysis of teleseismic data from the POLARIS/FedNor seismograph network across northern and eastern Ontario is providing valuable new information about the structure and properties of the lithosphere beneath the Superior and southwestern Grenville Provinces. Surface wave data from 34 broadband seismograph stations across the region are used to map out lithospheric structure, using two different methods. A two-station analysis technique has been applied to over 80 station pairs to measure path-averaged Rayleigh wave dispersion. The best coverage is for a set of northwest-southeast paths along the 'Diamond Corridor' of eastern Ontario, with more sparse coverage in central and western Ontario. In addition, an array analysis method is applied to two arrays of 5-7 stations each, in the Abitibi region of northeast Ontario and in southeast Ontario, resulting in an average Rayleigh wave dispersion curve for each array. Upper mantle structure is modelled from the dispersion curves using a Monte-Carlo method, which provides information on the range of mantle structures that satisfy the data sets. The models show significant variation in the thickness and properties of the continental lithosphere across Ontario. Lithospheric thickness varies from under 140km in the southeast to over 220km in the northeast, with intermediate values across northwestern Ontario. It is known from SKS splitting results that the upper mantle beneath much of Ontario is significantly anisotropic, with the greatest anisotropy beneath the western Superior Province. Rayleigh wave dispersion curves from different azimuths also suggest the presence of lithospheric anisotropy, though constraints on its nature will require a longer period of data collection by the POLARIS/FedNor array.
T13A-0433
Mantle Fabric at Multiple Scales Across an Archean-Proterozoic Boundary, Eastern Ontario, Canada
In eastern Ontario, Canada, the Proterozoic Grenville orogen abuts against the Archean Superior province; the complex tectonic history of the region is reflected in the pattern of electrical and seismic anisotropy within the lithosphere, while asthenospheric anisotropy is expected to reflect current patterns of mantle flow. Magnetotelluric and teleseismic data from the POLARIS and FedNor experiments and the Lithoprobe Abitibi-Grenville transect are examined for SKS splitting and geoelectric strike, and receiver functions are generated at selected stations, in order to characterize both vertical and horizontal variations in anisotropy in the eastern Ontario upper mantle. The average shear-wave split direction coincides with the direction of plate motion. Split times splits are found to be strongest in the southern part of the Grenville, where asthenospheric flow is enhanced by the presence of a lithospheric divot; the Ottawa-Bonnechere graben corresponds to a region of altered split direction indicative of a lithospheric component of anisotropy. North of the Grenville Front, there is a reorientation of the split direction from ESE to ENE at approximately 48°N, which is not easily attributable to crustal tectonics, and may represent the northern limit of lithospheric deformation produced by the Grenville orogen. Electrical anisotropy is pervasive in the study area. The magnetotelluric pattern of strikes is more complex than the SKS pattern, though the Groom-Bailey geoelectric strikes correlate fairly well with SKS measurements at nearby stations; the obliquity between SKS and magnetotelluric results shows no consistent orientation. Receiver-function analysis at three selected stations is indicative of a subcrustal anisotropic layer with a consistent SSE fast direction, underlain at station GAC by a sequence of anisotropic layers with varying directions; this sequence is not observed at station SADO. Combining these results, we interpret the strength and direction of anisotropic fabric in the Grenville to vary strongly with depth; the upper part of the lithosphere contains thin anisotropic layers perhaps related to eclogitization and relict slabs, but is insufficient to explain the observed SKS splits. The lower lithosphere is likely to be more ductile and uniformly anisotropic, and may be the main control on mantle electrical anisotropy as well as a significant contributor to SKS splitting. The largest contributor to SKS splitting in this region is interpreted to be asthenospheric anisotropy related to absolute plate motion.
T13A-0434
Variations and anisotropy of the elastic thickness of the lithosphere determined by the wavelet method: Examples from the Canadian Shield.
Different spectral methods have been used to estimate the elastic thickness and the mechanical properties of the lithosphere. We have used a wavelet transform to compute the local variations of the coherence between Bouguer gravity and topography in eastern Canada. The isotropic coherence is calculated by averaging the wavelet spectra from optimally overlapping 2-D Morlet wavelets having an isotropic spectral enveloppe in adjacent directions within 180 °, defining the so-called 'fan' wavelet. The local isotropic wavelet coherence is inverted to obtain local estimates of the elastic thickness ($T_e$) of the lithosphere. We calculate the anisotropic coherence by spatially averaging adjacent local wavelet spectra obtained from the rotation of the Morlet wavelet. The anisotropic direction of maximum observed coherence is diagnostic of the direction of preferred isostatic compensation, or the direction of mechanically weak lithosphere. We have carried out extensive tests on synthetic topography and Bouguer gravity data sets to verify that: (1) the wavelet method can recover $T_e$ for simple models with either homogeneous or spatially variable rigidity patterns; and that:(2) the method can determine azimuthal variations in the 2-D coherence for homogeneous models with anisotropic $T_e$. We have then used real data from the Canadian Shield to infer the variations in $T_e$ and the anisotropy of the coherence. The relative variations in $T_e$ agree remarkably well with our previous studies where we used the maximum entropy method to determine the elastic thickness [{\it Audet & Mareschal}, 2004a]. The wavelet transform gives $T_e$ values between 20 and 90 km. $T_e$ is generally high ($>70$ km) throughout eastern Canada. Lower values (30-50 km) are found around Hudson and James Bay, and near the Abitibi subprovince. High values are found within Hudson Bay, which is consistent with the previous studies. The main difference between this study and the one by {\it Audet & Mareschal} [2004a] is the absence of a low $T_e$ region in the southeastern Churchill Province. While $T_e$ was poorly estimated by both the maximum entropy and multitaper methods in the Appalachians [{\it Audet & Mareschal}, 2004b], the wavelet method yields values ranging from 60 to 80 km. The direction of maximum coherence obtained from the wavelet method is also consistent with our previous results obtained with the multitaper method and shows that the weak mechanical axis is perpendicular to the fast seismic axis where seismic anisotropy has been detected. Audet, P., & Mareschal, J.C., 2004a, Variations in elastic thickness in the Canadian Shield, {\it Earth Planet. Sci. Lett., 226}, 17-31, doi:10.1016/j.epsl.2004.07.035. Audet, P., & Mareschal, J.C., 2004b, Anisotropy of the flexural response of the lithosphere in the Canadian Shield, {\it Geophys. Res. Lett., 31}, L20601, doi:10.1029/2004GL021080.
T13A-0435
Rayleigh-wave Phase Velocity Structure and Azimuthal Anisotropy of the Lower Crust Beneath the East-central US
We use the 2-stations method to compute dispersion curves for Rayleigh-wave phase velocities ({\it C}) along more than 70 paths located in the East-central United States, implying 17 seismic stations from different networks (USNSN, GSN, and FLED). The area sampled by these paths extends from 33°N to 39°N, and from -90°E to -82°E. For each path, the dispersion curve can be interpreted as the average Rayleigh-wave phase velocity (<{\it C}>) along this path as a function of the period of the Rayleigh-wave {\it T}. To map the variations of {\it C} in the lower crust and shallow mantle beneath the East-central United States, and to test the presence of azimuthal anisotropy in this region, we invert our collection of <{\it C}> for isotropic anomalies of the Rayleigh-wave phase velocity (d{\it C}) and anisotropic terms (2-psi and 4-psi), as a function of {\it T}. The model is parameterized with a grid of 33 knots and a grid spacing of 150km. The inversion method is a general LSQR, including damping and smoothing. We report a positive gradient in the isotropic anomalies from SE to NW. Between periods of 10s and 70s, azimuthal anisotropy is clearly present, the fast direction being SW to NE. Interestingly, we do not see indication for azimuthal anisotropy at larger periods. These results suggest that the azimuthal anisotropy beneath East-central United States observed by shear-wave splitting methods is, at least partly, located in the lower crust.
T13A-0436
Lithospheric Structure Beneath Northern California
The coast range of northern California is a tectonically very active region. Many of the tectonic features are directly related to the northward migration of the Mendocino triple junction (MTJ) along the western edge of North America during the past 30 million years. South of the MTJ the Farallon slab has completely subducted beneath the North American plate, leaving a slab window which is thought to be filled by the upwelling asthenospheric materials. Accurate knowledge of the thickness of the lithosphere is crucial to understand the tectonic evolution of northern California. High temperature in the upper mantle estimated from seismic tomography and xenolith data indicates a thin (<100 km) thermal lithosphere. A zone of low shear wave velocity is imaged down to a depth of 300 km by surface wave tomography. Receiver function technique is one of the most powerful tools to detect seismic discontinuities in the crust and upper mantle. Here we jointly use P and S receiver function methods for 24 permanent broadband seismic stations of the Berkley Digital Seismic Network in northern California. The P receiver functions reveal complicated crustal structures. A strong intra-crustal interface at a depth of ~20 km effectively masks the converted signals from the Moho and lithospheric interfaces at many sites. The S receiver function method overcomes this problem, as the direct S-to-P converted waves arrive earlier than the S waves and thus consequently separated from the multiples that arrive later. We calculated S receiver functions from waveform data of S and SKS phases for all the stations. A significant phase with negative amplitudes is observed in the S receiver functions at ~8 s, indicating a low velocity interface in the upper mantle at a depth of ~70 km. We interpret this phase as the S-to-P conversion at the base of the lithosphere. Waveform modeling shows that the lithosphere-asthenosphere boundary here is ~10 km sharp.
T13A-0437
Impact of Deep Mantle Dynamics on the North American Continent
Recent advances in mapping 3-D mantle structure by simultaneously inverting global seismic data and convection-related surface observables (Simmons et al., 2005) yield high resolution models of the mantle convective flow which successfully reconcile both the seismic and geodynamic data sets. We have developed a mantle flow model, based on the most recent joint seismic-geodynamic inference of 3-D mantle structure, which incorporates a depth-dependent viscosity profile derived from the simultaneous inversion of glacial isostatic adjustment (GIA) and convection data (Mitrovica & Forte, 2004). In addition, this mantle flow model explicitly resolves the distinct chemical buoyancy which characterizes the continental roots. Geophysical observations of surface gravity and topography over the North American landmass have been corrected for the effects of present-day post-glacial rebound and crustal heterogeneity, thereby revealing large-amplitude regional variations which are successfully explained by the tomography-based mantle flow model. We find that a large fraction of the corrected surface gravity and topography variations over North America originate from density anomalies below 400 km depth. Similarly, the mean amplitude of maximum horizontal stress, SHmax, exerted at base of the crust by the convective flow ($~ 9$ MPa) also has a large deep-mantle contribution (depth $> 400$ km) which explains 60% of SHmax. Finally, we also find that the horizontal traction field acting on the base of the crust shows a striking convergence (compression) in the central portion of the continental US which may be attributed to the dynamics of the subducting Farallon plate in the deep mantle.
T13A-0438
The Preservation of Crustal Roots in Cratons: Constraints on Lithospheric Strength
Tectonic events and emplacement of large mafic intrusions induce lateral variations in crustal thickness and density which induce deviatoric stresses $\approx$10-30MPa. In the Canadian Shield and the Kaapvaal craton, seismic reflection and refraction studies show that the Moho is flat in regions that have experienced compression. One explanation is that with past crustal heat production higher than present, the lower crust was hot and weak and could not sustain the stress induced by a crustal root. Within the Canadian Shield a few crustal roots have been preserved: (1) the 1.9Ga Kapuskasing structural zone in the Archean Superior Province, (2) the LynnLake region in the 1.8Ga Trans Hudson Orogen, (3) the eastern part of the 1.1Ga Grenville Front, (4) the 1.1Ga Keweenawan rift system beneath Lake Superior. The average of all the heat flow values available in the southern Canadian Shield is 41$mW~m^{-2}$. In all the regions where a crustal root has been maintained, we noted that the heat flow is low: 33$mW~m^{-2}$ at Kapuskasing, 28$mW~m^{-2}$ at Lynn Lake, 28$mW~m^{-2}$ at the Grenville Front east, 30-35$mW~m^{-2}$ beneath Lake Superior. For all these regions, we have back-calculated crustal heat production and heat flow and determined lithospheric temperatures after the time of formation of the crustal root. We show that temperatures in the lower crust remained sufficiently low and that the rheology was sufficiently strong for the crust to sustain the stress due to crustal thickening without flow.
T13A-0439
Anisotropic Seismic Structure Beneath Cratons From Interstation Surface-Wave Dispersion Measurements on GSN and Geofon Data
Seismic structure of stable continental lithosphere offers insight into the origin and evolution of the tectosphere. The structure is still poorly known, especially its anisotropic components. Measurements of interstation dispersion of Rayleigh and Love surface waves can produce accurate, radially anisotropic S-velocity profiles that average between pairs of stations. A recently proposed (Lebedev and Meier, 2005) combination of station-station cross-correlation and event-station multimode waveform inversion allows to maximize the frequency band of the measured dispersion curves and yields improved structural constraints. With a view to applying the new method to large global data sets, here we make the measurements using all pairs of GSN and Geofon stations that are situated on stable continents and separated by sufficiently short interstation distances. If the distance between the two stations exceeds about 1000 km then it is difficult to make measurements at shorter periods---below 20 s or so---and trade-offs between the crustal and mantle structure cannot be resolved: in particular, nothing can be concluded on the nature or even the presence of radial anisotropy in the mantle. Limiting the scope of the study to station pairs separated by a few hundred kilometers, we make dispersion measurements using all suitably situated events and select a few pairs---all within stable tectonic blocks in Eurasia---for which broad-band dispersion curves are constrained with large amounts of data. Inverting the curves for S-velocity profiles, we find that the magnitudes of both the isotropic high-velocity anomaly and radial anisotropy in the mantle lithosphere vary strongly from one location to another. When radial anisotropy is present, it is with SH waves being faster than SV waves. We shall discuss the anisotropic seismic structures in the context of regional tectonic settings.
T13A-0440
The Effects of Melt Depeltion on Mantle Density and the Formation of Subcratonic Lithospheric Mantle
The extraction of partial melt in the Earth's mantle leaves behind a residue depleted in incompatible elements and modified in mineral compositions and mode. This phenomenon, referred to as melt depletion, is geodynamically important because a melt-depleted residuum is expected to be more buoyant than its fertile parent material. This process can lead to the formation of buoyant upper mantle ''tectosphere'' beneath cratons. We examine the effects of melt depletion at pressures from 1-7 GPa, and find mantle densities vary significantly with pressure. At 20% melt removal, residue density changes are -0.42%, -0.46%, -0.90%, -1.14%, -0.95%, -0.66%, and -0.57%, for pressures of 1, 3, 3.5, 4, 6, and 7 GPa, respectively. We note that at adiabatic temperatures, realistic composition upper mantle has a higher thermal expansivity than olivine, ranging from α = 4.91-3.47 x 10$^{-5}$/° between 1-7 GPa. This implies that 1% melt depletion leads to a density change equivalent to the increase in mantle temperature of 3-15° , depending on pressure. Under Archean cratons, where cold melt-depleted mantle is generally considered to have the same density as fertile adiabatic mantle (i.e. subcratonic mantle is isopycnic), we find subcratonic mantle above ~110 km is negatively buoyant with respect to adiabatic mantle. This suggests vertical transport of residues initially formed above 110 km may play a role in the stabilization of subcratonic mantle. Furthermore, the volume of basalt necessary to form isopycnic mantle below 110 km is greater than the total volume of existing cratonic crust, supporting geochemical evidence for the recycling of mafic lower crust . We also find that melt depletion has little, if any, effect on seismic velocities for garnet peridotite residues, and leads to only a 0.5% decrease in $V_P$ with 20% melt removal from fertile spinel peridotite. The major element effects of melt depletion are thus insufficient to produce the high mantle velocities imaged under cratons or cause significant lateral velocity variations below about 40 km depth.
T13A-0441
A Little Theory, a Tomography Model, a List of Hotspots, and GMT
Small-scale, or edge driven convection has been proposed as an alternative mechanism for hotspot volcanism. Using two theoretical results from mantle dynamics: 1) the endothermic spinel to perovskite plus ferro-periclase phase boundary is an effective barrier to short-wavelength convection (Tackley, 1995), and 2) the most unstable mode of convection has a horizontal wavelength $\sqrt{2}$ times the depth of the convecting cell (e.g., Turcotte and Schubert, 1982), I search for hotspots within 600-1000~km of the fast seismic velocity anomalies in the 100-200~km depth range in mantle tomographic models. These fast anomalies form the cores of the stable continents. I group hotspots into a class that has the potential to be explained by small-scale convection and a class that requires an alternate explanation. With the possible exception of Iceland, the list of potential small-scale convection hotstops does not contain any of the hotspots that Courtillot et al. (2003) classify as 'primary.' With the exeption of the Canary Islands and Eifel, the list of potential small-scale convection hotstpots does not contain and of the deep-rooted hotspots identified by Montelli et al. (2003) interpreting their tomographic results. I will discuss the exeptions and the hotspots that do not fit into the deep or small-scale classifications. Courtillot, V., Davaille, A., Besse, J., and Stock, J., 2003, Three distinct types of hotspots in the Earth's mantle, Earth Planet. Sci. Lett., 205, 295-308. Montelli, R., Nolet, G., Dahlen, F. A., Masters, G., Engdahl, E., R., and Hung, S.-H., 2003, Finite-frequency tomography reveals a variety of plumes in the mantle, Science Express, 10.1126/science.1092485 Tackely, P. J., 1995, On the penetration of an endothermic phase transition by upwellings and downwellings, J. Geophys. Res., 100, 15,477-15,488. Turcotte, D.L. and G. Schubert, 1982, Geodynamics: Applications of Continuum Physics to Geological Problems, 450 pp. J. Wiley and Sons, New York.
T13A-0442
Global 1 deg x1 deg Thermal Model TC1 for the Continental Lithosphere and the Growth Rate of the Continents
This study reports a new 1 deg x 1 deg consistent global thermal model TC1 for the continental upper mantle (Artemieva, PEPI, in press) constrained primarily by heat flow data and thus suitable for discrimination between structural/compositional and thermal heterogeneities observed in seismic, gravity, and electromagnetic models. A map of tectono-thermal ages of lithospheric terranes complied for the continents on a 1 deg x1 deg grid forms the basis for the TC1 model. Geotherms for continental terranes of different ages (>3.6 Ga to present) constrained by reliable data on borehole heat flow measurements (Artemieva and Mooney, JGR, 2001, B8) are statistically analyzed as a function of age and show a strong linear correlation with r=0.98. The statistical age relationship of continental geotherms (z=0.04*t+93.6, where z is lithospheric thermal thickness in km and t is age in Ma) is used to estimate lithospheric temperatures in continental regions with no or low-quality heat flow data (ca. 60% of the continents). These data are supplemented by cratonic geotherms based on electromagnetic and xenolith data; the latter indicate the existence of Archean cratons with two characteristic thicknesses, ca. 200 km and >250 km. Statistical analysis of continental geotherms reveals that thick (>250 km) lithosphere is restricted solely to young Archean terranes (3.0-2.6 Ga), while in old Archean cratons (3.6-3.0 Ga) lithospheric roots do not extend deeper than 200-220 km. It is proposed that the former were formed by tectonic stacking and underplating during paleocollisions of continental nuclei and have a limited lateral extent. This conclusion is supported by an analysis of the growth rate of the lithosphere since the Archean, which is constrained by statistical relations between geological ages and lithospheric thermal thickness. Growth models for the lithosphere do not reveal a peak in lithospheric volume at 2.7-2.6 Ga as expected from growth curves for juvenile crust. A pronounced peak in the rate of lithospheric growth (10-18 km3/year) at 2.1-1.7 Ga is the robust feature of the model; the peak correlates with a peak in the growth of juvenile crust and with a consequent global extraction of massif-type anorthosites. It is proposed that large-scale variations in lithospheric thickness at cratonic margins and at paleoterrane boundaries controlled anorogenic magmatism. Emplacement of Proterozoic anorthosites could have been caused by edge-driven convection triggered by a fast growth of the lithospheric mantle at 2.1-1.7 Ga. About 50% of the present continental lithosphere existed by 1.8 Ga.
T13A-0443
Azimuthal and Radial Seismic Anisotropy Beneath the Baltic Shield
The SVEKALAPKO passive seismic array in Finland provides us with an exceptional opportunity to study seismic anisotropy in and below the lithosphere in a shield. The array was composed of almost 150 sensors - out of which 46 were broadband - in a regular 2D grid which facilitated high-quality array analysis. We analyse phase velocities of both Love and Rayleigh waves to constrain radial and azimuthal anisotropy. We invert for the anisotropic parameters $\xi$ and $G_{C}$ on the one hand, and for the percentage of aligned olivine on the other. This latter parametrization of the inverse problem makes it straightforward to quantitatively compare the radial and the azimuthal anisotropies, under the assumption that aligned olivine dominates the anisotropy. The radial anisotropy, for which we have resolution in the lithosphere only, is strong, and can be explained by 40%-60% of the rock being olivine with the a-axis in the horizontal plane, equivalent to values of $\xi$ between 1.09 and 1.14. This radial anisotropy is stronger than observed in shield areas in global models (e.g. Beghein and Trampert, 2004). The azimuthal anisotropy is on the contrary very small in the lithosphere. This indicates that the orientation of the olivine minerals is random within the horizontal plane or that the overall effect across the area is negligible due to different orientations in different domains. Results from body-waves (Plomerová et al., 2005, Vecsey et al., in prep.) would support the latter interpretation. The azimuthal anisotropy as estimated by Rayleigh wave analysis is on the contrary significant below 200-250km depth, and corresponds to approximately 15%-20% of the rock being olivine with the a-axis aligned in direction N20. Xenolith analysis in the area shows that the rheologic lithosphere is at most 250km thick, so we suggest that this observed anisotropy is sub-lithospheric. Interestingly, the fast direction is significantly different from the absolute plate motion of the Baltic Shield, indicating that the lithosphere is not simply coupled to the underlying convecting mantle.
T13A-0444
Deep Source Anisotropy Revealed From Back-Azimuthal Variation of Shear-Wave Splitting in Southwest Ireland
Observation of large splitting in core shear wave phases such as SKS/SKKS are usually attributed to upper mantle anisotropy due to either a fossilized deformation imprint in the lithosphere or the influence of asthenospheric flow, or to a combination of both. Here we report splitting results from southwest Ireland that show a strong back-azimuthal dependence of fast direction which suggests the influence of a deeper source of anisotropy (i.e. sub-lithospheric, e.g. the transition zone or D"?). Shear-wave splitting measurements were made on data from two permanent stations (5-10 year deployments) and from up to 23 mobile stations (up to 2 year deployments), using a selection from 200 shallow and deep focus earthquakes with epicentral distances within the range $90°-$130° and with magnitude Mb above 6.0, with most of the high quality splitting analyses from earthquakes in South America and the East Asia region. Results show an average delay time of 1.2 seconds which is generally considered to require an accumulative anisotropic layer thickness of 80-100 km. Events from East Asia generally show more variation in fast polarization directions but with a dominant trend closely related to the Caledonian/Variscan tectonic fabric indicating that this may be preserved in the mantle lithosphere. However, events from South America give a consistently more northerly fast-polarisation direction, which does not appear to carry the signature of sub-continental lithospheric deformation history. There is no observable alignment of the fast polarization direction with the absolute plate motion direction so there is no direct correlation with mantle flow. Preliminary modelling based on two-layer anisotropy with a contribution from both mantle flow and possible lithospheric mantle deformation does not fit the observed splitting pattern. Hence, a more complex mechanism and origin for the observed anisotropy is required to explain our results.
T13A-0445
Seismic Anisotropy in Southern Tyrrhenian Subduction Zone (Italy) from Shear Wave Splitting.CAT/SCAN Project Preliminary Results.
We present shear wave splitting measurements in the Calabrian arc-Tyrrhenian basin subduction system. The data analyzed consist of several teleseisms and local earthquakes recorded at the CAT/SCAN temporary network (40 stations) and at the INGV national network (15 CESIS and MEDNET networks stations) from 2003 to 2005. We analyzed SKS phases of earthquakes with distance greater than 85° and magnitude greater than 6 and local S phases from slab earthquakes (deeper than 150 km). We used the method of Silver & Chan (1991) to obtain shear wave splitting parameters: fast direction and delay time. The splitting parameters, with delay time up to 2.0 s., reveal the presence of strong seismic anisotropy in the mantle beneath Southern Italy. The SKS show prevalent fast directions oriented NE-SW in the Calabrian. To the north, rotate to NNW-SSE, following the strike of the Southern Apennines mountains.This pattern suggests the presence of strained material in the mantle in and below the slab due to the eastward rollback of the slab. The local S phases, in contrast, show considerably more variability in the fast directions and smaller delay times, suggesting a frequency dependence of the anisotropy and a complex upper mantle deformation in front (i.e. Tyrrhenian side) of the slab.
T13A-0446
Upper Mantle Structure Beneath the Bohemain Massif, Central Europe - Another Baby Plume?
The BOHEMA project (BOhemian Massif HEterogeneity and Mantle Anisotropy) has brought together geophysicists from 10 institutions in the Czech Republic, Germany and France for a joint study of the structure and dynamics of the lithosphere and asthenosphere in the geodynamically active western part of the Bohemian Massif. An array of seismic stations covered a territory of approx. 270x150 km, with its long axis oriented perpendicularly to the strike of major tectonic units and to the Eger Rift. The network consisted of 61 permanent and 92 temporary stations operating between October 2001 and the end of 2003, with a core of recordings in 2002. Three-component short-period stations represent about 1/3 of the network, while broad-band stations constitute the remaining 2/3. Spacing of stations was generally less than 30 km, while in the central part of the array the station spacing was as small as 10-15 km, hence allowing for a lateral spatial resolution of approx. 20km in the upper mantle. The BOHEMA project aims at showing an existence or non-existence of a mantle plume beneath the Eger Rift, similarly to what has been established for several regions belonging to the European Cenozoic rift system (e.g., in Massif Central, Eifel), which may have a common source of volcanism in the mantle. Preliminary results of the P-velocity tomography indicate an asthenospheric upwelling beneath the region of Marianske Lazne. Besides isotropic velocity tomography, intensive research of body wave anisotropy is conducted. Evaluated parameters of seismic anisotropy are inverted jointly to retrieve a 3D self-consistent anisotropic model of the upper mantle, particularly of different mantle lithosphere domains. Both P- and S-wave anisotropy show two different orientations of the large-scale fabric in the Saxothuringian and the Moldanubian with a transitional type in the northern part of the Tepla-Barrandian.
T13A-0447
Shear Wave Splitting on the East European Platform
The method of shear wave splitting (e.g. Silver & Chan, 1991; Savage, 1999) provides a unique possibility to identify seismic anisotropy in seismological observations, i.e. to measure the orientation of fast and slow wave propagation directions. These directions result from the lattice-preferred orientation of anisotropic minerals. Though the depth of the anisotropic layer is less well constrained there is evidence that most of the splitting occurs in the asthenosphere and/or lithosphere. Anisotropy occurring in the upper portions of the lithosphere is often referred to as 'frozen' anisotropy as it displays old tectonic events. On the other hand, asthenospheric anisotropy results from deformation associated with plate motion. This deformation aligns the minerals (in particular olivine) in the direction of relative motion. Here we study seismic anisotropy under 17 broad band stations of the East European Platform (EEP) that is roughly located between Poland and the Ural Mountains, the Black Sea and Arctic Ocean. Some stations operate for more than a decade and thus provide an ideal dataset with a wide range of backazimuths of events. Those stations may therefore allow a study of mantle anisotropy with a resolution beyond a single anisotropic layer. This region is particular due to the presence of a thick lithosphere which reaches deep into the mantle and thus represents a mechanical obstacle to motion of the (Eurasian-) plate relative to the mantle. This mechanical obstacle might also channel the mantle flow as has been suggested for the American Plates. First results as well as previous studies in central Europe indicate fast directions aligning with the margins of the EEP. Though plate motion is slow (~ 20mm/a; Gripp & Gordon, 2002) the motion vector of the European plate correlates well with this study. There is a clear change of fast direction at the western edge of the craton. The data were processed using a novel "SplitLab" workflow, based on the Matlab platform. This graphical user interface provides the entire shear wave splitting process from seismogram request of data centers to the splitting measurements, as well as the post-processing of the results.
T13A-0448
Shear-Wave Splitting Beneath the Arabian Shield and Red Sea
The Red Sea Rift zone is composed of distinct geologic provinces in close proximity to one another resulting from the rifting and rotation of the African plate relative to the Arabian plate. Since the rift zone is a prototype of a newly formed oceanic basin, understanding of its geodynamic framework will provide important constraints on how seafloor spreading initiates and how continental rifting evolves. Our goal is to extend previous studies of this complex tectonic environment to generate a more complete characterization of the lithospheric structure in the Red Sea region. As part of this work, shear-wave splitting analysis, following the method of Silver and Chan (1991), has been employed to measure seismic anisotropy near the Red Sea Rift. This allows us to compare the anisotropic signature obtained with different candidate models of continental rifting to investigate mantle deformation and rifting mechanics. Data for our study comes from both the eight stations of the PASSCAL Saudi Arabia Broadband Array, which operated from November 1995 to March 1997, as well as the 25 broadband stations of the Saudi Arabian National Digital Seismic Network (SANDSN). Data from the SANDSN are uniquely available to us through collaboration with the King Abdulaziz City for Science and Technology. Splitting parameters, including fast polarization directions and delay times, have been determined for S, SKS, and other core refracted phases recorded at the Saudi Arabian stations. Stations along the eastern margin of the Red Sea display little variation with back azimuth and generally indicate a rift-parallel fast polarization direction. This is consistent with a single anisotropic layer model with hexagonal symmetry and a horizontally oriented fast axis. However, stations extending into the central region of the Arabian Peninsula display more pronounced back azimuth dependence. This may be associated with either lateral variations across the study area or with more complicated anisotropic structure, such as dipping or multiple layer anisotropy. These findings have important implications in that they do not support a "passive" rifting model, where the entire lithosphere below the rift extends and forms a rift-perpendicular fast direction. Instead, these results may indicate the presence of more "active" rifting processes, where the lithosphere is thinned through small scale convection, resulting in more complex anisotropy. In addition, the effects of fossilized anisotropy from previous tectonic events or the alignment of magmatic cracks along the rift zone may also play an important role in the observed anisotropic signature. Additional modeling, using an approach similar to Hartog and Schwartz (2000, 2001), will allow us to further examine these variations and resolve the anisotropic structure beneath the Red Sea and the Arabian Shield.
T13A-0449
Lithosperic Structure of the Arabian Peninsula From Joint Inversion of Teleseismic Receiver Functions and Surface Waves
With a goal of improving structural estimates of the Eurasian lithosphere, Lawrence Livermore National Laboratory (LLNL) has collaborative projects with a number of institutions in the Middle East. This has given us a unique opportunity to study Earth's structure below these regions. One of the widely used types of measurements to study lithospheric Earth's structure is surface-wave dispersion, from which 3-D variations in seismic velocities could be obtained. In this study we use a combination of grid search and iterative inversion scheme to fit jointly the surface wave group velocity dispersion (from 7 to 100 seconds for Rayleigh and 20 to 70 seconds for Love waves) and the receiver functions for the broadband stations installed on the Arabian Peninsula. For the grid search we use a database of pre-calculated theoretical receiver functions and dispersion curves, which allows us to significantly reduce the computing time and investigate wide range of structural models. We initially fit receiver functions and shorter periods of the observed dispersion curves with the structure within the crust and immediately under the crust. We then use an additional grid search to characterize the lithospheric lid and low velocity zone in the upper mantle, fitting longer periods of the dispersion curves. Results for several stations located in the Arabian Shield confirm crust thinning near the Red Sea and thickening towards the Arabian interior, which is consistent with previous studies. However our results indicate the presence of polarization anisotropy in the lithospheric upper mantle, even further away from the Red Sea than previously reported. Without invoking anisotropy, it is not possible to jointly explain the observed receiver functions and Love and Rayleigh wave dispersion discrepancy.
T13A-0450
Crustal Structure and Evidence for a Hales Discontinuity Beneath the Seychelles Microcontinent
It is well known that the Seychelles Plateau consists of a sliver of continental crust cast adrift during the formation of the Indian ocean. However the extent of the continental crust beneath the microcontinent and the cause of its isolation is poorly understood. Here we use receiver functions, interstation phase velocities obtained from surface waves, and wide angle reflections from controlled-source seismic data to investigate the lithospheric structure of the region. The $H$-$\kappa$ method is used to calculate depths and Poison's ratio at 26 temporary stations distributed across the plateau and Mascarene basin. The $V_p/V_s$ ratios and depths at stations on the plateau are typical of continental crust. To explain the major features of the RFs a simple two layer crust is proposed for the island of Mahé. The islands of Silhouette and Nord display a more complex crust consistent with the islands volcanic history. Praslin and its satellite islands display a simpler crust but display signs of a deeper discontinuity (~40 km) beneath the Moho which is possible evidence for underplating associated with Deccan age volcanism. Bird Island (Moho~18 km) and Desroche (Moho~23 km) show signs of being situated on islands above the transition from continental to oceanic crust. Alphonse, Coetivy and Platte all show receiver functions expected for oceanic crust, with Moho depths ~10 km. Inter-station phase velocity inversions from surface waves support these results with paths sampling the plateau region showing dispersion curves expected for continental crust, and those travelling between stations off the plateau showing evidence for oceanic crust. A deeper arrival is observed on the plateau stations at ~7 s or ~65 km. This feature is also seen in wide-angle controlled source work and the inter-station phase velocity inversions. Candidate interpretion for this Hales discontinuity include a Precambrian suture assoicated with shallow subduction or a shear-zone assoicated with deformation during breakup. Either feature may have influenced plume-related breakup in the region.
T13A-0451
Seismic constraints on temperature of the Australian uppermost mantle
We derive estimates of temperature of the Australian continental mantle between 80 and 350 km depth from two published S-velocity models. Lithospheric temperatures range over about 1000°C, with a large-scale correlation between temperature and tectonic age. In detail however, variations ranging from 200 and 700°C occur within each tectonic province. At the current seismic resolution, strictly Proterozoic and Archean blocks do not have substantially different temperatures, nor does the Phanerozoic lithosphere east and west of the Tasman line. Temperatures close to an average (moist) MORB source mantle solidus characterize the eastern seaboard and its offshore. Differences between the temperatures derived from the two velocity models illustrate the importance of well-constrained absolute velocities and gradients for physical interpretation. The large range of lithospheric temperatures cannot be explained solely with documented variability in crustal heat production, but requires significant variations in mantle heat flow as well.
T13A-0452
Lithospheric Shear Velocity Models Beneath Continental Margins in Antarctica Inferred From Genetic Algorithm Inversion for Teleseismic Receiver Functions
Seismic shear velocity models of the crust and the uppermost mantle were studied by teleseismic receiver function analyses beneath the permanent stations of the Federation of Digital Seismographic Networks (FDSN) at Antarctic continental margins. In order to eliminate the starting model dependency, a non-linear Genetic Algorithm (GA) was introduced in the time domain inversion of the receiver functions. A plenty of velocity models with an acceptable fit to the receiver function waveforms were generated during the inversion, and a stable model was produced by employing a weighted average of the best 1,000 models encountered in the development of the GA. The shear velocity model beneath the MAW (67.6S, 62.9E) has a sharp Moho boundary at 44 km depth that might have involved in a reworked metamorphic event of adjacent Archaean Napier Complex. A fairly sharp Moho was identified about 28 km depth beneath DRV (66.7S, 140.0E), with a middle grade variation of the crustal velocities that might have been caused by the Early Proterozoic metamorphism. A similar sharp Moho has been found at 40 km beneath SYO (69.0S, 39.6E). Thus Moho depth is consistent with that from refraction / wide-angle reflection surveys around the station. Fairly complicated velocity variations within the crust may have a relationship with lithology of granulite facies metamorphic rocks in the shallow crust associated with Pan-African events. Broadening low velocity zones about 30 km depths with transitional crust-mantle boundary at VNDA (77.5S, 161.9E), might be caused by the rift system besides the Trans Antarctic Mountains. As for the Antarctic Peninsular, very broad Moho was found around 36 km depths around PMSA (64.8S, 64.0W). The evidence of velocity variations within the crust reflects the tectonic histories of each terrain where these permanent stations are located.