T13B-1330
Investigation of seismic anisotropy in the area of the German Regional Seismic Network (GRSN) using P-to-S converted phases.
Based on SKS and Pn-residual analysis several studies have shown evidence that the crust and/or the upper mantle are seismically anisotropic beneath the area of the German Regional Seismic Network (GRSN). The aim of our study is to estimate the crustal and mantle shares of seismic anisotropy by analyzing the polarization of P-to-S converted waves (receiver functions) generated at the Moho Pms and at the upper mantle discontinuities P410s, P660s. Seismic anisotropy finds its expression in a directional dependence of seismic velocities and in the travel time delay between orthogonally polarized S-waves (S-wave splitting). For the estimation of S-wave splitting we used the energy minimization technique and an azimuthal stacking approach. The P-to-SV converted wave fields generated at the Moho show a frequency variation with azimuth, broading of amplitude peaks and an azimuth periodicity in amplitude strength. These are indications of both a thickness variation of a transition zone and existence of anisotropy. On the SH-component converted energy is observed, too. Its amplitude and polarity periodicity is 360° for most of the GRSN stations. Some stations show less than 360° or unclear periodicity. The greatest change in fast direction of the 3 discontinuities exist between crust and upper mantle. More than a half of the stations show equal directions for the crust Pms compared to SKS/SKKS studies. High delay times δt>0.5s are observed for crustal structures versus small changes in time to deeper strutures. Since about 16 years the GRSN provides continuous seismological data records for most of its stations. Based on an analysis of back azimuth and incidence rotated ray component L we determined an average signal/noise ratio from 4.5 to 22 for these data. Because of the long observation time the azimuth coverage of incidence angles is sufficient for the major part of station (18/25). The data basis for an analysis seismic anisotropy is well established. For the analysis of P-to-S converted waves we selected 2455 events of magnitude mb>5.5 and epicentral distances of 0°-100°. To process the receiver functions - rf, of this data volume we set up an automatic process including the computation of various quality criteria. The processing steps included in the procedure are: (1) rf calculation based on real and theoretical polarization angles, (2) rf selection based of signal-to-noise ratio, (3) rf weighting based on the quality of polarization. The rate of data reduction during this process is approximately between 50 and 70 percent.
T13B-1331
Seismic anisotropy of upper mantle in the northeastern margin of the Tibetan Plateau
By using the polarization analysis of teleseismic SKS waveform data recorded at 30 seismic stations which involved in the digital seismograph network of the northeastern margin of the Tibetan Plateau, we obtained the SKS fast-wave direction and the delay time between fast and slow waves of each station by use of the stacking analysis method, and finally acquired the fine image of upper mantle anisotropy in the northeastern margin of the Tibetan Plateau. From the results, the fast-wave polarization direction is basically NW direction and clockwise rotation trend. The delay times vary from 0.70 to 1.51s in the study region.The fast-wave polarization direction is consistent with the strike of the major faults in the observed region. The characteristics of the fast-wave direction is similar to the trends of the minimum average principal compressional stress directions in the northeast margin of the Tibetan Plateau and to the trends of GPS observation results. Research shows that the regional tectonic stress field resulted in upper mantle material clockwise rotational deformation and flow, and make the alignment of upper mantle peridotite lattice is parallel to the direction of material deformation. The upper mantle is vertically coupled with the crust.
T13B-1332
Linking elastic, mechanical and transport properties in anisotropically cracked rocks
Damage and crack porosity can result in a decrease of the mechanical strength of the rock, the development of elastic and mechanical anisotropy and the enhancement of transport properties. Using Non-Interactive Crack Effective Medium (NIC) theory as a fundamental tool, it is possible to calculate dry and wet elastic properties of cracked rocks in terms of a crack density tensor, average crack aspect ratio and mean crack fabric orientation using the solid grains and fluid elastic properties. Using the same tool, we show that the anisotropy, the shear wave splitting and the dispersion of elastic waves can be derived for anisotropic crack fabrics. Mechanically, the existence of embedded microcrack fabrics in rocks also significantly influences the fracture toughness (KIC) of rocks. We show that KIC can show large amounts of anisotropy as well, the degree and orientation of which being largely constrained once again by the microcrack fabric. NIC can predict relatively well KIC at high crack density, by simply using dimensionless crack densities inverted from velocities. A decrease of 50% for crack densities larger than 1, 80% for crack densities larger than 5 is predicted, in close agreement with our observed experimental variation of KIC. At the microscale, this can be interpreted by the fact that the main fracture is strongly interacting with the pre-existing microcrack fabric. Finally, and above the percolation threshold, macroscopic fluid flow also depends on the porosity, crack density and aspect ratio. Using the permeability model of Guéguen and Dienes (1989) and the crack density and aspect ratio recovered from the elastic wave velocity inversion, we successfully predict the evolution of permeability with pressure for direct comparison with the laboratory measurements. These combined experimental and modelling results illustrate the importance of understanding the details of how rock microstructures change in response to an external stimulus in predicting the simultaneous evolution of rock physical properties.
T13B-1333
B-type olivine fabrics developed during progressive retrogression above subducting slab in the mantle wedge
B-type olivine fabrics occur pervasively within highly depleted dunites in a small Imono peridotite body located in the subduction-type Sanbagawa metamorphic belt in the southwest Japan arc, with various microstructures from relatively coarse granular texture to fine-grained intensely sheared texture. The Mg/(Mg + Fe) atomic ratios (Fo number) of olivine within these dunites are constantly around 0.9 and Cr/(Cr + Al) atomic ratios (Cr number) of chromian spinels are constantly around 0.9. It suggesting that they could have evolved through high depleted magma. This provides strong thermal constraints on the formation of the highly depleted dunites that require hot, hydrous shallow mantle (>1250C at <30km depth) in the mantle wedge. Before these peridotites were finally entrained by the Sanbagawa metamorphic belt during progressive retrogression, they could flow downward along the subducting slab up to 2.8 GPa (>80km) at temperatures of 750-800C. Therefore, B-type olivine fabrics should be developed during downward flow along the subducting slab, possibly in relation with dehydration from the slab. Such occurrences provide a new insight of mantle flow in the fore-arc mantle wedge, where hydrated asthenospheric mantle flow is subsequently altered by retrogressive deformation before/during exhumation with metamorphosed slab materials. The small magnitude of S-wave splitting can be explained by the seismic properties of the B-type peridotites for an approximately 10-km thick anisotropic layer subparallel to the subducting slab, indicating that the B-type layer could be one of the dominant sources of seismic anisotropy in the fore-arc region.
T13B-1334
Layer thinning heterogeneities within an ice sheet inferred using a strain-induced anisotropic flow model.
Interpretation of climatic records preserved in ice-sheet requires an accurate knowledge of the absolute age and the duration of ice layers as a function of depth. This relation between age, duration and depth depends on the flow history experienced by ice since its deposit on the surface. Usually, for low accumulation site like in Antarctica, this dating relation is inferred using simplified ice flow models and leads to a smooth thinning function as a function of depth. In the case of a dome, the model is generally reduced to a one-dimensional flow model assuming axisymmetric flow around the dome. Recent texture measurements of the EPICA Dome C core have shown that some layers have experienced larger deformation than the adjacent layers. This is supported also by the difficulties encountered to match the EPICA Dome C and Dronning Maud Land time scales. These observations indicate that the polar ice deformation presents spacial heterogeneities and consequently the layers thinning function is certainly less smooth than expected. Possible causes for these layer disturbances, and their implication in terms of dating, are studied with a full- Stokes ice flow model which takes into account the strain-induced anisotropic behavior of ice. The proposed method consists in following a layer which presents different initial material properties, in term of viscosity and/or initial fabric. The resulting thinning functions for different initial material properties are discussed.
T13B-1335
Effects of birefringence within ice sheets on bistatic radar measurements
Ice fabric is known to affect radio wave propagation within polar ice sheets. Strong vertical contrasts in the fabric cause reflection, while anisotropy in the fabric integrated along the propagation path causes birefringence. Bistatic radar measurements (i.e. using a transmitter and receiver separated from each other) have more potential to assess fabric than monostatic measurements, because they allow more general geometric relationships between the radio wave and ice fabric. We worked out the effect of fabric on radio waves propagating obliquely through the polar ice sheets in terms of radar frequency, polarization and geometry of the propagation path. The model is applied to fabrics that approximate those found in the polar ice sheets. When ice fabrics are unknown, the geometry of the propagation path relative to the ice fabric is arbitrary. In this case, 1-km- thick ice with a quite gentle and strongly developed fabric causes significant (> 1 dB) birefringence-origin echo variations at radar frequencies higher than 200 MHz and 20 MHz, respectively. An azimuthal pattern of the echo collected with bistatic radar is 90° periodic, not 45° as found in the vertical propagation (monostatic radar sounding). It suggests that bistatic radar can be used to identify principal axes of the ice fabric.
T13B-1336
Observation of the Type-B Crystallographic Preferred Orientation of Olivine in Spinel Peridotite, Baengnyeong Island, South Korea
Crystallographic preferred orientation (CPO)s of olivine and clinopyroxene in spinel peridotite from Baengnyeong Island in South Korea were investigated. Baengnyeong Island is located in the western most part of Korea in Yellow Sea and near the boundary between the Phanerozoic Imjingang belt and Precambrian Nangrim massif in North Korea. The Imjingang belt has been considered to be an eastern extension of the Quiling-Dabie-Sulu collision belt (Ree et al., 1996). The sample spinel peridotite was collected from the Pliocene basalt in Baengnyeong Island. To determine CPOs of olivine and clinopyroxene, we used electron backscattered diffraction (EBSD) and HKL Channel software. We found that olivine [001] axis is subparallel to the lineation and olivine [010] axis normal to the foliation plane, suggesting deformation of olivine by the (010)[001] slip system (olivine type-B fabric). The olivine type-B fabric is of particular interest because it may be used to explain the trench-parallel shear-wave splitting that is often observed at subduction zones. CPO of clinopyroxene showed that [001] axis is subparallel to the lineation and [100] axis normal to the foliation plane, suggesting deformation of clinopyroxene by the (100)[001] slip system. The type-B CPO of olivine was observed experimentally at the conditions of high stress and/or under moderate to high water content (Jung et al, 2006). The sample spinel peridotite is considered to have been deformed at high stress condition, suggesting that upper mantle under Baengnyeong Island had experienced high stress in the past and the Quiling-Dabie-Sulu collision belt was probably extended to the area close to the Baengnyeong Island in South Korea.
T13B-1337
Crustal anisotropy in the Archean Minnesota River Valley Subprovince and its significance
The origin and evolution of the American continental lithosphere is a key question addressed by EarthScope. The Superior Province formed as an amalgamation of Archean/ Proterozoic terranes that subsequently acted as a stabilizing nucleus. This province is characterized by a strong seismic anisotropy (SWS = 1.3 s) of unknown origin. As suggested for the Archean Kaapvaal Craton (South Africa), this could be attributed (1) to current asthenospheric flow, or (2) to fossil lithospheric anisotropy, or (3) to the role of lithospheric keels on modern asthenospheric flow. The first hypothesis is not favored because SWS data for the Superior Province do not fit global mantle flow models. The second hypothesis would be compatible with obliquity between lithospheric mantle and crustal seismic anisotropies, possibly due to oblique docking. The third hypothesis would require asthenospheric flow to be controlled by lithospheric block geometry. The origin of seismic anisotropy and its spatial variations need to be determined to test these hypotheses. The deployment of USArray in the Superior Province in FY10, along with the prospect of deployment of a Flexible Array and the GeoFrame Superior focus area should provide a wealth of seismic data. Yet, the contribution of the Archean-early Proterozoic continental crust to seismic anisotropy is unknown. This study focusses on the Minnesota River Valley (MRV) Subprovince, part of the Superior Province. The MRV Subprovince consists of four juxtaposed blocks (Benson, Montevideo, Morton and Jeffers) of amphibolite to granulite grade migmatites, tonalites, granodiorites, diorites and pelitic rocks interlayered into each other. These blocks are separated by EW-dipping shear zones broadly parallel to SWS observations. In other parts of the world, the crustal seismic anisotropy is generally considered to be modest (SWS = 0.1-0.2 s), although experiments specifically designed to constrain it are scarce. The MRV represents a 200 km-wide, tilted cross-section of Archean-Proterozoic continental crust and is thought to be somewhat representative of the felsic crust at depth. We propose a multiscale methodology to quantify the magnitude/orientation of seismic anisotropy. This approach first establishes the directional relationship between crustal seismic anisotropy and anisotropy of magnetic susceptibility (AMS) measured on oriented specimens. The elastic properties of crustal rocks are calculated by forward modelling using EBSD-LPO measurements and published elastic constants for single crystals. The significance of layering seismic anisotropy can then be determined by in-situ experiments performed on meter-scale parallelepipedic blocks industrially extracted from quarries for dimension stones. The contribution of the continental crust [Sc] to seismic anisotropy can be subtracted from total seismic anisotropy [STOT] to yield the upper mantle only contribution [SUM]. The origin of seismic anisotropy in the continental crust is investigated by imaging the shape preferred orientation (SPO) of leucosomes in migmatites in 3-D. Preliminary data suggest that the leucosomes SPO is coaxial with the AMS fabric which supports the fact that the AMS tracks high-temperature ductile crustal flow. Crustal flow and crustal seismic anisotropy can then be interpreted, with reference to upper mantle anisotropy, either as a coupled or uncoupled crust-mantle system.
T13B-1338
Anisotropic plasticity of single crystals of wet synthetic quartz
Anisotropic plasticity (e.g., crystallographic preferred orientation patterns) for quartz has been reported more than for any other mineral in deformed rocks. In natural quartzites, with increasing deformation temperature, the dominant slip system has been observed to change from basal , to prism , to prism [c]. However, the flow stress at which each slip system is activated and the effect of water content on slip system activity has not been fully investigated. We have undertaken an experimental study to examine the effects of temperature and water content on the slip system activity and strength of single crystals of wet synthetic quartz deformed at 1.5 GPa confining pressure, 10-5/s strain rate, and 700-900oC using a Griggs apparatus. From FTIR measurements, the synthetic quartz shows a broadband absorption around 3400 cm-1 that can be assigned to the vibration of molecular H2O. Samples of synthetic quartz with two different water contents were utilized: ~ 400 and ~ 700 ppm H/Si. Two orientations of axial compression were chosen: O+ (45o to a and c) and c\bot (normal to a and c), and plastic strains up to 40{%} shortening were achieved. Optical microscope observations indicate that samples compressed in the O+ orientation deform by basal single slip at 700oC and by a combination of basal and prism [c] slip at 900oC, and samples compressed in the c\bot orientation deform by prism double slip. The stress-strain curves for both orientations show gradual strain hardening out to ~ 25{%} strain. The difference in water content between 400 and 700 ppm H/Si has no marked effect on the strength of either orientation, indicating that this difference in water content is not important for the dislocation creep strength of synthetic quartz at our experimental conditions. At 700oC, the c\bot orientation (prism slip) is somewhat stronger than the O+ orientation (basal slip). At 900oC, the c\bot orientation has approximately the same strength as the O+ orientation (~ 100 MPa at 40{%} strain). Our experimental results confirm that basal slip is easier at lower temperature (700oC), whereas basal , prism , and prism [c] slip have about same strength at higher temperature (900oC). These results are in partially agreement with previous experimental results conducted at 1.5 GPa confining pressure (e.g., Blacic, 1975) but do not agree with results on synthetic quartz at 300 MPa confining pressure (e.g., Hobbs et al., 1972). Our experimental study provides more accurate mechanical data convincing anisotropic rheology of crustal rocks and the role of fluids.
T13B-1339
Seismicity Patterns and Permeability Reduction Associated With the Anisotropic Propagation of Discrete Compaction Bands in Diemelstadt Sandstone
We report results from triaxial deformation experiments on samples of Diemelstadt sandstone conducted under a confining pressure sufficient to induce compaction bands. Diemelstadt sandstone is a visibly anisotropic rock with an initial porosity of 23% and a mean grain diameter of 0.3mm. We have quantified the void-space anisotropy of this material by: (a) measuring radial elastic S and P waves as a function of azimuth around orthogonally cored samples, and (b) measuring the magnetic susceptibility anisotropy (AMS) of samples saturated with magnetic ferro-fluid. P-wave velocities show anisotropy of 7% and AMS measurements show that the void-space fabric approximates to an oblate spheroid, with the isotropy plane parallel to the bedding plane. Consequently, we have performed experiments on samples cored both normal and parallel to the isotropy (bedding) plane. Previous studies of compactive deformation have concentrated on the growth of a sequence of compaction bands through the volume of the sample. By contrast, we have concentrated on the nucleation and temporal evolution of single compaction bands. We used full wave-form acoustic emission (AE) locations from 10 transducers to record the propagation of individual compaction bands. This allowed us to compute an average band propagation velocity, which was lower than 0.1 mm.s-1. We also computed the seismic b-value from recorded AE events in order to examine any change in the scale of cracking during compaction band growth. We found samples deformed parallel to the bedding plane to be stronger than those deformed normal to it, with the compaction band geometry showing a more tortuous growth in the bedding-normal direction. This pattern of behaviour is also consistent with the very large reductions (3 to 4 orders of magnitude) in permeability observed during initial compaction band formation. A more gradual permeability reduction is observed during compaction band growth normal to the isotropy plane, commensurate with the increased mechanical resistance to compaction band development in this orientation.
T13B-1340
Effect of Fluid on Seismic Anisotropy
We present the effect of fluid on the anisotropic Thomsen's parameters for rocks with transversely isotropic and orthorhombic symmetry. We use the anisotropic form of Gassmann's equation for fluid substitution and consider the following origins of anisotropy in the rock: (a) anisotropy due to aligned fractures, (b) thin layers of isotropic sands with different elastic properties, (c) thin layer anisotropy with alternate layering of anisotropic sand, and (d) stress induced anisotropy. Our modeling results show that fractured rocks exhibit significant changes in anisotropy due to fluid substitution. The predicted change in such situations can reach even five times the initial anisotropy. The magnitude of change in anisotropy arising from nonhydrostatic stress or layering of different facies is less significant. We find that, it is the initial anisotropy of the medium that controls the fluid effect on anisotropy. Higher initial anisotropy causes larger change in anisotropic parameters due to fluid substitution. We present an approximate fluid substitution equation for vertical velocity in anisotropic medium. We express the changes in vertical P-wave velocity due to fluid substitution by isotropic fluid substitution plus a correction term proportional to the Thomsen's parameter 'delta'. This simplification reduces the number of parameters required for fluid substitution in anisotropic rocks. It also helps us to explain the over- or under-prediction of isotropic Gassmann's prediction in anisotropic rock in terms of the magnitude and sign of 'delta'.