Transcript Presentation by group E.

Paradigms versus paradoxes:
Developing a new paradigm for
the mantle
Attreyee Ghosh, Ricardo Arevalo Jr., Ved
Lekic, and Victor Tsai
with
Adam Dziewonski, Barbara Romanowicz,
Louise Kellogg, and Wendy Panero
(names arranged in alphabetical order by first name)
Geochemistry likes a layered
mantle
• DMM cannot account for the planet’s
budget of:
– Incompatible elements
• UDMM + UCC ≠ UBSE
– Radiogenic heat production
• HDMM + HCC ≠ HBSE
– Noble gas abundances
•
40Ar
DMM +
40Ar
CC +
40Ar
atm
≠ 40ArBSE
Seismology finds a variety of behaviors
for slabs in the transition zone
• Tomographic images
illustrate mass flux
across the 660 km
discontinuity
et al.et(2008)
Van derLiHilst
al. (1998)
Li et al. (2008)
Geodynamics like well-mixed
mantle reservoirs
• Mantle layering
difficult to maintain
for multiple Ga
without significant
mixing
Naliboff and Kellogg (2007)
2 layers  traditional
limited exchange
2 layers  isolated upper
and lower mantle reservoirs
1 layer  wholemantle convection
Hybrid  limited
exchange
Hofmann (1997)
“Marble-cake mantle”
Sobolev et al. (2005)
Morgan & Morgan (1999)
“Plum-pudding mantle”
Becker et al. (1999)
“Blob mantle”
What about this ‘D”’?
• Several geochemical
studies have called
upon an early,
differentiated
reservoir that has
remained “hidden” at
the core-mantle
boundary
Boyet and Carlson (2006)
What about this ‘D”’?
• Seismology and mineral physics observations
indicate a heterogeneous layer at the coremantle boundary
Power Spectra
Lee et al., (2007)
1500
3000
5
10
15
20
25
150 km
Zone of neutrally
or negatively
buoyant melt
Transition Zone
Pressure (GPa)
1000
Temperature oC
2500
2000
410’
660’
Lee and Luffi
Tolstikhin and Hofmann (2005)
What about the role of
thermochemical piles/superplumes?
• Increasing the volume of a deep mantle
reservoir (e.g., including superplumes)
dilutes the required incompatible/
radioactive element budget of this
reservoir
2800 km depth seismic profiles
Kustowski (2006)
Romanowicz and Gung (2002)
Upper mantle:Q - lower mantle: Vsh
Degree 2 only
Romanowicz and Gung (2002)
Defining the volume of a
superplume
superplumes in S362ANI (1% slow anomaly)
Defining the volume of a
superplume
Area (sq km)
Depth (km)
Depth (km)
Area (sq km)
S362ANI (0.6% contours)
SAW24B16 (1% contours)
Defining the volume of a
superplume
Conservative estimates  only consider depths >1000 km
Geochemical implications
• If we know the composition of the
Continental Crust (CC; e.g., Rudnick and
Gao, 2003), the Depleted MORB Mantle
(DMM; e.g., Su, 2002) and the Bulk
Silicate Earth (BSE; McDonough and
Sun, 1995)…
– The size of the DMM dictates the required
composition of a deep, Enriched Mantle
Reservoir (EMR)
Geochemical implications
*Concentration range calculated from uncertainties in
compositional models of CC and DMM
Thermal implications
Table 8 from Van Schmus (1995)
Element
Isotope
Isotopic Abundance (wt%)
Decay Constant, λ (yr-1)
Total Decay Energy (MeV/decay)
Beta Decay Energy (MeV/decay)
Beta Energy Lost as Neutrinos (MeV/decay)
Total Energy Retained in Earth (MeV/decay)
Specific Isotope Heat Production (cal/g-year)
Specific Isotope Heat Production (μW/kg)
Specific Elemental Heat Production (cal/g-year)
Specific Elemental Heat Production (μW/kg)
Potassium
40K
0.0119
5.54E-10
1.34
1.181
0.65
0.69
0.22
29.17
2.60E-05
3.45E-03
Thorium
232Th
100
4.95E-11
42.66
3.5
2.3
40.4
0.199
26.38
0.199
26.38
Uranium
235U
0.71
9.85E-10
46.40
3
2
44.4
4.29
568.7
238U
99.28
1.55E-10
51.70
6.3
4.2
47.5
0.714
94.65
0.74
98.1
Thermal implications

Maintaining neutral buoyancy
Thermal
•Solve: Chemical  Thermal  0
ln vsChemical  ln vsThermal  ln vsSeismic

• Assume:

– Fe is most important chemical variant
– Fe has no effect on modulus or
thermal expansion
– Thermal and chemical effects are
linear wrt velocity
– Fe partitioning between mw and pv
– Fe has a linear effect on density:

(Mg ,Fe)O (xFe )  xFeFeO  (1 xFe )MgO
Chemical
Stixrude & LithgowBertelloni, 2005
What does this mean?
• Uncertainty in partitioning behavior has a first
order effect
• Velocity drop at base of the mantle is >2.5%
– Additional 1.5% Fe (reasonable)
– Excess temperature of 450-700 K
• Velocity drop in mid-mantle is ~1%
– Additional 0.5% Fe
– Excess temperature of 180-275 K
• Super piles are neither on constant adiabat or
isochemical if they are neutrally buoyant
Future questions to address
• How stationary are these superplumes?
– Do surface tectonics dictate the large scale
flow in the mantle, or vice-versa?
– Slab reconstructions (over the last 200 Ma)
and degree-2 signals are well correlated
Slab model of L-B & R (1998)
Vs model S362ANI
(Mid-mantle depths)
Slab model of LithgowBertelloni & Richards (1998)
Vs model S362ANI
• Slab model of
Lithgow-Bertelloni &
Richards (1998)
• Vs model S362ANI
• The degree-2 velocity anomalies at the CMB are extremely well
correlated with the integrated slab signal: the sum of all the slabs
deposited during the last 200 Ma.
Future questions to address
• How long could such a thermochemical
reservoir be dynamically stable for?
– “Bottom-up” dynamical test
• Starting conditions: 2 rigid conical masses
attached to the CMB - representative size of
superplumes
• The transition zone must be able to arrest, at least
temporarily, sinking subducted materials
• The convection experiments, spanning a
sufficiently large parameter space would give us
insight into lower mantle mixing and return flow
Some typical snapshots at t ~ 4.55Ga
H=150 km
B=2
H=500 km
B=0.7
H=1000 km
B=1
H=1600 km
B=0.7
Kellogg and Ferrachat
Dynamic criteria: stability over several Ga, topography
of the interface, net density, and magnetic field
Future work/questions
• What is the mass flux of material into
the lower mantle? Reaching D”?
– How much becomes incorporated in our deep
reservoir?
Fukao et al. (2001)
A new paradigm
• We propose that the lowermost mantle pattern of
the two chemically and thermally distinct superplumes dictates the planform of mantle
dynamics for at least the last 200 Ma.
• The superplumes may have stable locations for
at long periods of time, anchoring mantle plumes
and influencing the paths of Wilson cycles.
• The transition zone plays an important role in
the interaction between subducted slabs and the
superplumes
A new paradigm
• Transition zone may be a “leaky” boundary layer
– Subducted slabs pond in the transition zone, with
sufficient residence time for some oceanic crust to be
re-circulated in the upper mantle
– Ponded material breaks through the 660 km
discontinuity in avalanche-like events and is
deposited around the upwellings giving rise to the ring
of fast velocities girdling the Pacific
– Low-pass filter removes high wavenumber features
from slab signal
– Temperature contrast sufficient to produce plumes at
the 660 km discontinuity
Questions?
Generic Hawaii
72 km
362 km
652 km
942 km
1377 km
2102 km
2827 km
Slab integration model of
Lithgow-Bertelloni and
Richards (1998)
Further Required Assumptions
• Fe partitioning between MgO and
MgSiO3
Andrault, 2001
Assume D=5
Andrault D
Badro D (HS->LS)
EMR = Enriched
Mantle Reservoir
CC = Continental
Crust
•
40Ar
produced by decay of 40K (t1/2 = 11.93 Gyr)
– Too heavy to be lost from atm
– >99.9% Ar is 40Ar
• We know:
– 280 ppm K in equals >150 Eg (1018 g) of 40Ar produced over 4.5 Ga
– 66 Eg in atm, 10-20 Eg may be in crust, the rest must reside in the mantle
– 40ArBSE = 40Aratm + 40ArCC + 40ArDMM + 40ArEMR – 40Ardegassed