Is the D Region the Source of Mantle Plumes

1
Is the D Region the Source of
Mantle Plumes?
Thorne Lay
University of California, Santa Cruz IGPP/Earth
Sciences Department
With thanks to Collaborators Ed Garnero, Tine
Thomas, Megan Avants, Juliana Rokosky, Alex
Hutko, Michael Thorne, Claudia Flores, and others
2005 Chapman Workshop The Great Plume Debate
2
OUTLINE

• The CMB Earths major internal boundary
• and the D thermo-chemical boundary layer
• SEISMOLOGICAL Perspective of whether
• mantle plumes rise from this region
• General attributes of D structure
• Inferences on boundary layer dynamics
• Answer Put to a vote, probably Yes, but in
• terms of resolution, it is an open issue.

3
Region Beneath the Central Pacific
No clear D low velocity Under Hawaiian plume
Montelli, Nolet, Dahlen,Masters, Engdahl, Hung
Science, 2004
4
PREM
Preliminary Reference Earth Model
Dziewonski and Anderson 1981, PEPI
5

• The D Region is likely to have a Thermal
• Boundary Layer (TBL) due to flux of heat from the
  core
• From 4-15 TW may flux through CMB
• If there are no mid-mantle TBLs, may have
  1000C
• temperature increase across TBL
• May approach eutectic melting curve in hot
  boundary
• Layer
• The D Region is likely to have Chemical
  Heterogeneity
• due to
• Huge density/composition contrast at CMB in a
• chemically stratified planet
• Chemical flux into/out of core over Earth
  history

6
The Standard Model
High Velocity Regions Under Subduction Zones
Large Low Shear Velocity Provinces
Localized Plumes?
Lay, 2005, Plume Volume, in press
7
Large-scale Patterns of Heterogeneity in D
Lay, 2005, P3 Volume, in press
8
Region Beneath the Caribbean and Central America

• Regional finite frequency
• deep mantle tomography
• DATA Differential Travel Times
• 327 S-SKS
• 297 ScS-S
• 98 Scd-S
• 98 ScS-Scd

Hung, Garnero, Chiao, Kuo, Lay 2005
9
Large Low Shear Velocity Provinces (LLSVPs)
Shear velocity decreases of 3-12, almost no P
velocity decrease, sharp sided, 1000 km high?
Wen, 2004
10
LLSVPs Have Sharp Lateral Edges
Ni and Helmberger, 2004
11
Lay, Garnero, Williams 2004, PEPI
12

• Other important deep mantle findings/analyses
• D discontinuity from anomalous
• reflections
• Ultra-low velocity zones from
• reflections, time and amplitude
• anomalies
• D anisotropy from shear wave
• splitting

Lay, Garnero, Williams 2004, PEPI
13
MANTLE-CORE BOUNDARY LAYERING
D discontinuities

• 200-300 km thick
• 0.5-3 velocity increase

14
MANTLE-CORE BOUNDARY LAYERING
Ultra-low velocity zones (ULVZ)

• 0.5-10s km thick
• 10-30 velocity decrease

Thorne and Garnero JGR, 2004
15
Basal Boundary Layer Structure Ultra-low
velocity zones
Thorne and Garnero 2004, JGR
16
Region A Beneath the Caribbean and Central
America
17
0.4-2.0 Anisotropy. Azimuthal anisotropy
required in Pacific and under Caribbean. Radial
anisotropy viable other places.
Garnero, McNamara, Lay in prep, 2005
18
Region A Beneath the Caribbean and Central
America
19
Region A Beneath the Caribbean and Central
America
Data and processing
3D grid of migration reflectors
Latitude -2 to 18 deg Longitude -100 to 80
deg Depth 2200 to 2888 km
1 deg
1 deg
10 km
Thomas, Garnero, Lay JGR, 2004
20
Migration of S waves under Cocos Plate
Hutko et al., Nature, Submitted
21
D Discontinuity is Probably due to a Phase Change
Lay et al., EOS, 2005
22
Phase Change Could Occur Twice in Thermal
Boundary Layer
But, if is Hard to Observe
Flores and Lay, GRL 2005, submitted
23
D Discontinuity Waveform Stacking Beneath the
Central Pacific
Avants, Lay, Russell JGR, 2005, in revison
24
ULVZ and D Velocity Decreases Below the Central
Pacific
25
Central Pacific Process
26
Region Beneath the Southwest Pacific
(Rost, Garnero, Williams, Manga, Science, 2005)
27
Region Beneath the Caribbean and Central America
Garnero,Maupin, Lay, Fouch 2004, in prep
28
Region Beneath the Caribbean and Central America
Garnero,Maupin, Lay, Fouch Science, 2005
29
Lay and Garnero 2004, AUG/IUGG Monograph
30
So, do Plumes come from D? Well.

• The boundary layer has complex structures
• The D discontinuity is being interpreted as a
  phase change to post-perovskite in cool regions
  (subducted slabs?) - could be compositional
  layering otherwise
• LLSVPs appear to be compositionally distinct.
  Not clearly superplumes, but could be at base
  of plume upwellings
• ULVZs indicate melt fraction in D, just above
  the core-mantle boundary

31
(No Transcript)