Huaiyu Yuan 123 Scott French 4 and Barbara A Romanowicz 456 1 CCFS EPS Macquarie University Australia 2 CET University of Western Australia Australia ID: 573974
Download Presentation The PPT/PDF document "Azimuthal anisotropy layering in the Pac..." is the property of its rightful owner. Permission is granted to download and print the materials on this web site for personal, non-commercial use only, and to display it on your personal computer provided you do not modify the materials and that you retain all copyright notices contained in the materials. By downloading content from our website, you accept the terms of this agreement.
Slide1
Azimuthal anisotropy layering in the Pacific upper mantle
Huaiyu
Yuan
1,2,3
, Scott
French
4
and
Barbara A.
Romanowicz
4,5,6
1
CCFS, EPS, Macquarie University, Australia2 CET, University of Western Australia, Australia3 Geological Survey of Western Australia, Australia4 Berkeley Seismological Laboratory, UC Berkeley5 Collège de France, Paris, France6 Institut de Physique du Globe, Paris, France
DI21C-07Slide2
Lithosphere Layering and the LAB Lithosphere-asthenosphere-boundary (LAB): one of “Grand Challenges” of modern seismology (Lay 2009)“Elusive” in continents due to smooth transition in seismic velocity (Eaton 2009; Romanowicz 2009; Fischer et al. 2010)
Most extensively deformed “first-order structural discontinuity” (Eaton 2009) in the global tectonics due to strong plate-asthenosphere couplingDifferential motion accommodated by rock deformation across/at the LAB Slide3
Depth dependent azimuthal anisotropy Craton-wide layering in the upper mantleDomains of anisotropy possessing distinct fast axis directions
Yuan and Romanowicz 2010Layering and the LAB
Fast axis
directions
Fast axis directions
- APM
APM: Absolute Plate Motion directionSlide4
Shallow depleted chemical layer (Layer 1; suture-strike parallel)Sub-thermal layer (Layer 2; high angle to APM)Asthenosphere (current APM parallel)
Yuan and Romanowicz 2010
Lithosphere Layering and the LAB
Cooper and Conrad 2009Slide5
Shallow depleted chemical layer (Layer 1; suture-strike parallel)Sub-thermal layer (Layer 2; high angle to APM)Asthenosphere (current APM parallel)Anisotropic LAB (180-240 km under NA craton)
Yuan and Romanowicz 2010
Lithosphere Layering and the LAB
LAB DepthSlide6
The Oceanic LABOceanic lithosphere covers much larger surface area than continentsShort life span: e.g., Pacific < 200Ma compared with >3Ga of continents
Simple tectonic history; perturbed by input of deep seated thermal upwellings (e.g., French 2013).Slide7
The Oceanic LABSimple tectonic history; perturbed by input of deep seated thermal upwellings (e.g., French 2013).
See French 2013 for details
Poster: DI21A-2267 this afternoon
Talk S21E-07 now Slide8
The Oceanic LABThickness of oceanic lithosphere dictated by thermal history Follows 1 half-space cooling model; 2 plate model;
3 or not (Rychert et al. 2013; Schmerr 2012)Thickness in the range ~ 100 km in the oldest (Schmerr 2012) Paleo-APM at shallow and current APM at depth picked up by azimuthal anisotropy (Montagner 2002; Smith et al. 2004; Debayle et al. 2005; Maggi et al. 2006; Debayle and Ricard 2013)Slide9
The ObjectivesCan we define a Pacific-wide LAB?Oceanic upper mantle layering and its tectonic implications.Slide10
Pacific Inversion
Full waveform inversion using 2D finite-sensitivity kernels (NACT; Li and Romanowicz 1995)Based on an earlier (2007) Berkeley global model that honors true oceanic Moho (Crust2.0);Modified linear crustal correction (Lekic and Romanowicz 2009) applied;5° lateral model spacing; 20-30 km vertical spacing B-splines down to 300 km, then 50-150 down to 1000 km.Isotropic Vs and radial anisotropy ξ; then azimuthal anisotropy strength (G) and fast
axis directionsSlide11
Pacific Inversion60-s global fundamental and overtone waveforms (French et al. 2013); 60- and 40-s Pacific regional network data; OBS waveforms Overall 1.3 million paths; or ~25 million
waveform points handled by NACT for forward and inverse problems Model error handled by bootstrap resampling (Efron and Tibshitani 1986): standard deviation of the 100 “new” modelsNo SKS dataset due to the sparse coverage Slide12
Anisotropy map viewsLayered North America confirmed; although weak but AMP-parallel in the asthenosphere Depth dependent anisotropy domains also found in the PacificModified linear crustal correction (
Lekic and Romanowicz 2009) applied;Full waveform inversion using 2D finite-sensitivity kernels (NACT; Li and Romanowicz 1995)5° lateral model spacing ; 20-30 km vertical spacing B-splines down to 300 km, then 50-150 down to 1000 km.Isotropic Vs and radial anisotropy ξ; then azimuthal anisotropy strength (G) and fast axis
Lithosphere depth
Asthenosphere depth
Below AsthenosphereSlide13
Depth cross-section
Isotropic
Vs
Radial Anisotropy
Azimuthal Anisotropy Strength (%)Fast-axisdirectionSlide14
Resolution Tests
Azimuthal
Anisotropy
Strength (%)Fast-axisdirection
Azimuthal Anisotropy Strength (%)Fast-axisdirection
Input
OutputSlide15
Age Profiles: Vs and LABPlate-model likePunctuated between 70-100 MaRadial Anisotropy
Ritzwoller et al. 2004
Vs Profile
Priestley & McKenzie
2006 Vs Profile
Isotropic Vs Age ProfileSlide16
Age Profiles: Vs and LABDefined by maximum Vs depth gradientSeismic LAB: 30-100 km depth rangeConsistent with S-receiver functions and SS-precursor studies
LAB from
negative Vs gradientSlide17
Age Profiles: Azimuthal AnisotropyCurrent APM direction > 100-140 kmHigh angle to the APM at shallow depthTwo domains of anisotropy
Azimuthal anisotropy
Strength (%)
Fast axis directions
with APM removed
Current Plate Motion
Paleo-plate motionSlide18
Age Profiles:
Azimuthal AnisotropyOceanic lithosphere LAB: 50-140 kmDefined by azimuthal anisotropy
LAB from
Azimuthal
anisotorpySlide19
Two LAB horizonsShallower Vs LABDeeper anisotropy LAB
Velocity LAB
Azimuthal anisotropy LABSlide20
Shallower Vs LAB
Deeper Anisotropy LAB
LAB Age Profiles
LAB from
anisotropy
LAB from VsSlide21
Resembles the continent layer model
Chemical oceanic layer?
Cooper and Conrad 2009
LAB from
anisotropy
LAB from VsSlide22
Depleted
Layer in the oceanic lithosphere (
Hirth
&
Kohlstedt 1996; Lee et al. 2005); thickness
dependent on ambient mantle temperature
Growth of thermal sub-layer when thermal effect out-weights chemical effect
Depleted chemical oceanic layer?
Lee et al. 2005
LAB from Vs
LAB from
anisotropySlide23
Correlation of deep Anisotropic LAB with Large Igneous Provinces?
Role of LIPs
Green area:
Large
Igneous Provinces
(LIPs)
Velocity LAB
Azimuthal anisotropy LABSlide24
ConclusionsLayered oceanic upper mantleUniform fossil lithosphere structureAPM parallel asthenosphereVs LAB shallower than anisotropy LABChemical depleted oceanic layerSlide25Slide26Slide27
Age Profiles and Error EstimatesBased on an early Berkeley global model that honors true oceanic Moho (Crust2.0);Modified linear crustal correction (Lekic and Romanowicz 2009) applied;Full waveform inversion using 2D finite-sensitivity kernels (NACT; Li and Romanowicz 1995)
5° lateral model spacing ; 20-30 km vertical spacing B-splines down to 300 km, then 50-150 down to 1000 km.Isotropic Vs and radial anisotropy ξ; then azimuthal anisotropy strength (G) and fast axisSlide28
Pacific Inversion
Model error handled by bootstrap resampling (Efron and Tibshitani 1986): standard deviation of the 100 “new” models60-s global fundamental and overtone waveforms (French et al. 2013) and 60- and 40-s Pacific regional network dataOverall 1.3 million paths (~30 million waveform points) handled by NACT forward and inversion Good azimuthal coverage for Pacific Slide29Slide30Slide31Slide32
Age Profiles: Vs & ζModel points sorted by age and averaged for model parametersIsotropic VsRadial AnisotropySlide33
Age Profiles: Azimuthal AnisotropyModel points sorted by age and averaged for model parametersIsotropic VsRadial AnisotropySlide34
Resolution TestsModel error handled by bootstrap resampling (Efron and Tibshitani 1986): standard deviation of the 100 “new” models60-s global fundamental and overtone waveforms (French et al. 2013) and 60- and 40-s Pacific regional network data
Overall 1.3 million paths (~30 million waveform points) handled by NACT forward and inversion Good azimuthal coverage for Pacific Slide35
Resolution TestsModel error handled by bootstrap resampling (Efron and Tibshitani 1986): standard deviation of the 100 “new” models60-s global fundamental and overtone waveforms (French et al. 2013) and 60- and 40-s Pacific regional network data
Overall 1.3 million paths (~30 million waveform points) handled by NACT forward and inversion Good azimuthal coverage for Pacific