Eclogite Engine

1
Eclogite Engine

• Don L. Anderson
• Caltech

2
Can we bridge geophysics,geochemistry
geodynamics?

• Continents the missing link
• The lower crust is transient
• It is recycled 6 times faster than upper crust
• Recent arc growth estimates are 5X previous
  estimates!
• Therefore, a huge previously unaccounted for flux

3
THE UPPER MANTLE ECLOGITE CYCLE

• All the components of so-called hotspot or plume
  magmas originate in crust, lithosphere, cumulates
  or mantle wedge
• The Isotope Zoo
• EM1, EM2, HIMU, DUPAL, LONU, High 3He/4He,
  C-, FOZO, Os

4
Mantle is a Top-Down System
delamination
Sources of dissipation
The top boundary condition is not simple
5
Many ways to get big chunks of mafic matter into
upper mantle

• Subduction of seamount chains, aseismic ridges
• Delamination of island arcs
• Delamination of compressional mountain belts

6
Estimating mantle sources sinks, continental
addition loss, through time, depends
critically on when deep subduction, got underway
1 Gyr of oceanic crust subduction70 km of
eclogite
But lots of other things get put into the mantle
OIB 2 Km3/yr
20 km3/yr
Piles up _at_ 70 km/Gyr
David Scholl
7
(No Transcript)
8
Midplate volcanisms arcs can be upper mantle
recycling
9
RIDGE-TRENCH ANNIHILATIONAbandonment of young
oceanic crust and mantle wedge
BAB
Future suture
EM2
LOMU
Trapped crust
FOZO
Mantle wedge
EM1
HI 3HE/U
HIMU
All of the isotope and trace element signatures
of OIB are manufactured in the subduction
factory, wedge crust
10
Delamination cycle

• dense roots
• fall off
• warm up in ambient mantle
• rise
• possible mechanism for Atlantic Indian ocean
  plateaus DUPAL anomaly

11
SHEAR VELOCITY
density
ECLOGITE AT DEPTH HAS LOW SHEAR VELOCITY
12
Non-uniform heat leakage from the top peeling
off of crust
The eclogite engine, as in any engine, involves
changes in volume due to compression, heating and
phase changes.One cannot consider T but not
P. The cycle is cooling, phase change
(eclogite), compression (sinking), heating, phase
change (melting), decompression (rising), heat
exchange The continent may move, rather than
the blob.
Dual eruptions
unstable
heating
13
Delaminated roots warm quickly

• will start to melt before reaching same T as
  surrounding mantle
• already in TBL, so starts off warm
• when 30 melt, garnet mostly gone will start to
  rise

14
Eclogitemelts
Peridotite melts
Eclogite melts
Dry peridotite can only melt in very shallow
mantle hence adiabatic ascent at ridges
Eclogite can melt much deeper, and much more,
even when colder hence, midplate magmatism
15

• The breakup of Gondwana and the uncovering of the
  Indian and Atlantic oceans provides the best
  opportunity for seeing the re-emergence of these
  fertile blobs, after they have been heated by
  ambient mantle. It has been proposed that the
  Indian ocean and the south Atlantic plateaus and
  island chains may be due to these mafic patches
  that were injected into the mantle from the base
  of the Gondwana crust (Anderson, 2005).

16
LIPs are associated with continental breakup

• reconstruction at 30 Ma
• dual volcanism
• on breakup
• 30 Myr later
• oceanic plateaus form 1,000 km offshore
• rising of delaminated root?

17
COUNTERFLOWCHANNEL
MANTLE WEDGE
EROSION
DELAMINATION
NO PLUME NECESSARY
Slow upwelling
Slow upwelling
Broad upwellings
Input into asthenosphere delamination, seamount
chains, broad upwelling, abandoned mantle wedges
18
Fertile blob asthenospheric counterflow model

LVZ
2
4
3
1
--------------------------------------------------
-------------------------
1 delaminated crust, 2 wedge, 3 broad upwelling,
4 young oceanic crust
19
Test of the model lots of scattering in the
upper mantle
There is no mass balance or geochemical reason
for anything to come from below 1000 km
20
Blobs deliver heat magma to surface
Blobs gain heat from mantle
NOTICE! Oceanic crust not involved at all !
21
TRI-CYCLING THROUGH THE EARTH

• RIDGE-TRENCH CYCLE (the escalator)
• SUPERCONTINENT CYCLE (the trolley)
• SLAB-PLUME CYCLE (the tube)
• SUBTERRANEAN CYCLE (delamination, orogenic cycle
  the elevator)

22
Scale matters
gt10-km blobs gravity takes over heating is slow
I 10 cm or 10 km ? I
piclogite
Garnet and clinopyroxene
Olivine and opx
23
Eclogite melts in the mantle react with
peridotite to form pyroxenite

• The olivine of mantle peridotite is consumed by
  reaction with melts derived from recycled crust
  cumulates, to form a secondary pyroxenitic source
• Hotspot basalts such as Hawaii form from
  olivine-free mantle

24
GALILEO THERMOMETER
And bring mantle heat magma to base of plate
Blobs sink to level of neutral buoyancy
But they dont stay there
In mantle, they heat up melt
Not core heat
25
Rates of generation of the continental crust

• The rates at which basaltic magma is added to the
  continental crust have recently been revised
  upwards range to 3.7 km3 yr-1
• The plume affinity of basaltic rocks in
  juvenile crust has been used to support
  deep-seated disturbances in the Earth, as opposed
  to shallow-level processes. But this evidence
  could also be used in support of the delamination
  model.
• The average residence times of the lower crust
  may be at least five times less than the upper
  crust

26
THE ECLOGITE CYCLE

• Mass balance seems o.k.
• Petrology major elements seem o.k.
• Isotopes seem o.k. including stable isotopes
• Dynamics seems o.k.
• Thermodynamics seem fine
• Energetics seem o.k.

27
What caused these?
LIPs CFB SDR NAVP CRB OPB BABB OJP
28
New parameters

• Scale (size)
• Homologous temperature
• Stress
• Buoyancy parameter
• Architecture
• Fertility
• Entrainability

29
MASS BALANCE

• 
  km3/yr
• TOTAL MIDPLATE MAGMA VOLUMES 3-6
  
• Growth rate of arc garnet pyroxenite cumulates
  1.5-6
• Underside erosion delamination of continents
  4-6
• Production rate of magmas in arcs
  4-5
• LOWER CRUSTAL LOSS HOTSPOT MAGMATISM
• CYCLE TIME 40-80 MYR
• 10-30 KM CHUNKS

30
KEY ELEMENTS

• Mantle is source of heat
• Thick basalt piles are the source of material
• True Top-Down system
• Mantle is heterogeneous
• Heterogenity washed out by Central Limit Theorem
  (ridges, tomography), not by mantle convection
• Mantle is gumbo, not fruit cake

31
Unification of geodynamics geochemistry of
mantle

• Delamination of lower continental crust
  subduction of seamount chains fertilize the
  mantle
• These become low meltng point blobs
• It is these, not plumes, that explain anomalies
  such as midplate volcanism, swells and hotspots