Download Surface volcanism data since 400 Ma as a constraint of mantle

Surface volcanism data since 400 Ma
as a constraint of mantle dynamics
and Principal Inertia Axes
Marianne Greff and Jean Besse
Institut de Physique du Globe de Paris
Geomagnetic and paleomagnetic teams
I-2) Position of two antipodal upwelling domes
in the lower mantle estimated from the intraplate alkaline volcanism
2
1
0
−1
−2
Shear velocity variation
I-1) Large-scale pattern of mantle dynamics:
contributions due to subducted lithosphere
and to long wavelength upwellings.
II) Principal Inertia Axis and Polar Wander
I-3) Amount of subduction since 400 Ma estimated from the study of the subduction volcanism in the past and plate motion history
0
O180
2
O1
230
End of The
Pal
eo
- Te
C2 280280
th
ya
n
ge
180
C'2
C1
230
230
120
C'1
C3
100
100
300-270Ma
C'3
140
0
O3
280
50
190
Ri
d
0
Surface volcanism data since 400 Ma as a constraint of
mantle dynamics and Principal Inertia Axes
• Simple model for the large-scale pattern of mantle dynamics:
contributions due to subducted lithosphere and to long wavelength
upwellings.
from [Courtillot et al., 2003]
Mantle Downwellings [Ricard et al., 1993]
Vc
• Slabs sinking vertically into the mantle
at the surface velocity
• At 660km depth, they are slowed and
thickened by a factor 4
Ex: Vc ≃ 8 cm/yr ⇒ Vt = 2 cm/yr
⇒ Characteristic time needed for a
plate to sink through the whole mantle
is 120 Myrs.
⇒ Temporal evolution of the mantle density heterogeneities since 280 Myr
is modeled assuming that plates sink
down to the core-mantle boundary since
400 Myr.
Upper Mantle
Vc
νUM
νLM
660 km
Vt=Vc/ln(νLM / νUM)
Lower Mantle
(CMB) 2890 km
Ricard et al. (1993)
⇒ We need the position of the subduction zones and surface velocities
since 400 Myr.
Mantle Upwellings
Two large regions in which the material is thermally or compositionally less dense
than the surrounding mantle
⇒ responsible for the superswells observed at the Earth’s surface [Davaille et al., 1999].
Domes modeled as two hemispheres:
• Pacific dome assumed to be antipodal
◗
• h = 500 km from present-day tomography
◗
◗
◗
α
◗
h
◗
◗
◗
◗
◗
◗
• Stable large zones ⇒ the time taken by the
upwelling flow to cross the mantle is less than
20 Myr, because of pre-existing conduit between the dome and the surface.
• From the traces left at the surface by intraplate volcanism, we assessed the geometrical
parameters of the African dome
Surface volcanism data since 400 Ma
• Extraction from three database [Georoc, Usgs and the GPMDB ver. 4.6 data base [McElhinny & Lock, 1996] ] of the location
and age of volcanic edifices
• Sorting of volcanism into two types:
intra-plate and subductions:
• Reconstruction of their paleopositions in a mantle reference
frame where Africa is fixed or in the paleomagnetic frame from [McElhinny et al., 2003] VPG.
• Interpretations:
→ Position of the small circles above the Africa dome
→ Position of the subduction zones
2
1
0
−1
−2
Shear velocity variation
Paleopositions of intraplate volcanic edifices in the
paleomagnetic frame since 400 Ma
Tomographic image of the deep mantle at the depth=2880 km (Ritsema et al. 1999)
(black: recent time (< 100 Ma); purple: 100 < t < 200 Ma; pink: 200 < t < 300 Ma; white: 300 < t < 400 Ma)
Paleopositions of intraplate volcanic edifices in our fixed
African reference frame since 400 Ma
60˚
• Small circle modelling the trace left
at the surface by intra-plate volcanism
above the dome for each 20 Myrs
interval from -400 Ma to present.
60˚
30˚
30˚
0˚
0˚
−30˚
−30˚
−60˚
−60˚
• Temporal evolution of the center of small circle related to the
motion of Africa in the paleomagnetic
frame.
• The width of the small circle above
the ’African’ dome is around 45 − −50o
Paleopositions of subduction zones in the Africa fixed reference frame
since 400 Ma
380-410Ma
Paleopositions of subduction zones in the Africa fixed reference frame,
390-330 Ma
Paleo- Teth
yan
R
idg
eA
ct
iv
e
390-330Ma
Paleopositions of subduction zones in the Africa fixed reference frame,
340-300 Ma
n
io
idg
e
He
rcy
ni
an
/V
ar
is
ca
nC
ol
li s
Paleo- Teth
yan
R
340-300Ma
Paleopositions of subduction zones in the Africa fixed reference frame,
300-120 Ma
End of The
Pal
eo
- Te
th
ya
n
Ri
d
ge
300-270Ma
Paleopositions of subduction zones in the Africa fixed reference frame,
270-230 Ma
Birth of th
eN
eoT
eth
ya
nR
id
ge
270-230Ma
Paleopositions of subduction zones in the Africa fixed reference frame,
230-120 Ma
Kim
me
ri a
nC
ol
NeoTethya
l is
nR
io
n
idg
~2
e
3
0M
a
230-120
Paleopositions of subduction zones in the Africa fixed reference frame,
since 120 Ma
• For the past 120 Myr, we use a compilation of plates motions by [Lithgow-Bertelloni et
al., 1993] in the hotspot reference frame which we convert into our fixed Africa frame.
Perturbation of the Earth’s inertia tensor
Simple model for the large-scale pattern of mantle dynamics combining contributions
due to subducted lithosphere and to long wavelength upwellings
n X X
nmax
Geoid(θ, ϕ) = a
cnm cos mϕ + snm sin mϕ Pnm (cos θ)
n=2 m=0
Inertia tensor of the Earth:
 c20
3



2
Iij = M a 


− 2c22
−2s22
−c21
−2s22
c20
3
+ 2c22
−s21
−c21
−s21
−2 c20
3








⇒ Diagonalization ⇒ Eigenvalues and Principal Inertia Axes (PIA)
Mantle mass anomalies and geoid
Total geoid (mass + deformation)
Total geoid (mass + deformation)
Geoid due to the mass
surface
surface
Surface topography
Surface topography
The gravity decreases
The gravity decreases
The geoid descends downwardly
Geoid due to the mass
Upper Mantle
The geoid descends downwardly
Upper Mantle
Lower Mantle
Lower Mantle
CMB
CMB
CMB topography
CMB topography
⇒ The geoid depends on the depth and the size of the mass anomaly.
Earth’s model and degree 2 geoid kernel
6000
6000
5000
5000
4000
4000
lithosphere
manteau superieur
manteau inferieur
noyau fluide
-0.1
0
0.1
0.2
Degree 2 geoid kernel
A positive degree 2 mass anomaly
located in the upper part of the
mantle will involve a positive geoid
whereas a positive mass anomaly in
the lower mantle will involve a negative geoid.
1e+20
1e+21
1e+22
Viscosity (Pa.s)
Inertia product perturbations:
I13 (t) = −
4π
5
Z
r4 ker2 (r) σ2c1 (r, t)dr
Inertia Eigenvalues
PIA eigenvalues
4e-05
Three eigenvalues clearly
distinct:
2e-05
0
⇓
since 280 Ma, there is no
inertial interchange.
-2e-05
-4e-05
-280
-240
-200
-160
-120
Time in Ma
-80
-40
0
Principal Inertia Axes
Minimum PIA: relatively stable
130
60˚
0
230
50
60˚
Maximum PIA has driven in a
plane perpendicular to Africa
along a quasi-meridian.
190
30˚
280
230 280
190
280
40 120
30˚
0
100 130
180
0˚
100
0
−30˚
−60˚
0˚
−30˚
−60˚
This meridian seems to be
stable with time, similarly to
large-scale pattern of the subduction around the Pacific.
Polar wander
0
180
0
230
50
280
280
190
180
120
230
230
140
0
Conservation of the angular momentum
⇒ at geological time-scale, the rotational axis
coincides with the maximum PIA.
280
Polar motion since 280 Ma for two cases:
plates alone and both domes and plates.
⇒ the effect of domes does not change
significantly the shape of the curve.
100
100
In red:
Observed Apparent Polar Wander
Path (APWP) of Africa: motion of the magnetic
dipolar axis with respect to Africa
Polar wander
0
180
0
230
50
280
280
190
180
120
230
230
140
0
• Differences in the amplitudes and directions of rotation vs paleomagnetic poles
⇒ Possible motion of our mantle reference frame
upon which Africa plate is fixed with respect to
the paleomagnetic reference frame
280
• Common features:
Cusps around 230 Ma and 180 Ma appear in
apparent polar wander paths of different plates,
as well as in our computed pole.
100
100
⇒ These cusps are essentially due to the
Hercynian (340-300 Ma) and Kimmerian (270230 Ma) continental collisions.
Pal
eo
- Te
th
ya
n
Ri
d
He
rcy
ni
an
/V
ar
is
ca
n
io
End of The
idg
e
ge
nC
ol
li s
Paleo- Teth
yan
R
340-300Ma
300-270Ma
• The stop of the subduction of the Paleo-Tethys plate 300 Ma ago and
the creation of a perpendicular new convergence zone during the stage
300-270 Ma will create a cusp in the pole curve 230 Ma ago.
Without taking into account the subduction of the Tethys plate (during 270-230 Ma ),
this cusp will occur around 200 Ma ago. But the start of the Tethysian subduction moves the
point of inflection C1 toward 230Ma.
Birth of th
eN
eoT
eth
ya
nR
Kim
me
ri a
nC
ol
NeoTethya
l is
nR
io
n
idg
~2
e
3
a
ge
0M
id
270-230Ma
230-120
• The cusp around 180 Ma is essentially due to the stop of the subduction
of the Tethys plate 230 Ma ago.
This plate has an averaged velocity about 10 cm/year and it needs less than 30 Myrs to
reach the depth about 1200 km where the kernel of the degree 2 geoid changes its sign and 95
Myrs to reach the CMB where it is isostatically compensated: the change in the PIA directions
will occur consequently between 30 and 95 Myrs after the collision.
The new convergence zone appearing after 230 Ma has too small velocity to perturb the
PIA at this time-scale.
Polar wander
• Cusps around 230 Ma and 180 Ma appearing in apparent polar
wander paths of different plates, as well as in our computed pole
are essentially due to the Hercynian (340-300 Ma) and Kimmerian
(270-230 Ma) continental collisions.
• The drift direction in the observed African APWP and in the
computed path is quite different.
⇒ Possible motion of our mantle Africa fixed reference frame
with respect to the paleomagnetic reference frame
⇒ Mantle mass anomalies associated with plates do not systematically reach the core mantle boundary ?
The depth down to which the subducted plates present a
significant density contrast with respect to the
surrounding mantle is unknown.
Fukao et al., 2001
Ricard et al., 2005
• Test: If the density contrast beneath South America for depth larger than
1400 km is zero:
60˚
60˚
30˚
30˚
0˚
0˚
−30˚
−30˚
−60˚
−60˚
Temporal evolution of the rotational axis in our Africa fixed reference frame: in green line, all
the plates reach the CMB; in black line, the test
⇒ the pole drift in the same great circle but in opposite direction
⇒ the intermediate PIA and the maximum PIA have been exchanged
Conclusion
1) The peri-pacific subduction is a quite permanent feature of the earth’s
history at least since 400 Ma
⇓
• The minimum PIA is relatively stable and close to the axis passing through the
central Pacific and its antipodal position.
• Because of this stable structure, the rotation axis (Maximum PIA) is in a
plane perpendicular to this axis.
2) The ”Tethyan” subductions having a complex history with successive
continental collisions and episodically rebirth of E-W subduction zones,
they involve cusps (around 230 Ma and 180 Ma) in the polar motion.
Next step: Do the mantle mass anomalies associated with plates systematically
reach the core mantle boundary ?
2
1
0
−1
−2
Shear velocity variation
Paleopositions of subduction volcanic edifices in the paleomagnetic
frame since 400 Ma
Reconstructed subduction volcanism for the last 400 Myr in the paleomagnetic reference frame
Tomographic image of the deep mantle (model S20RTS of Ritsema et al. [1999] at the depth=2880 km).
Example: one slab sinking vertically within the mantle
• At t = 0, the plate starts to sink within
the mantle at the surface velocity of 12
cm/yr
• At t = 115 Myr, its stops to sink.
Associated inertia tensor perturbations
7.5e−06
I11/Io
I22/Io
I33/Io
I12/Io
I13/Io
I23/Io
5e−06
2.5e−06
T1
T2
T3
> T4
T5
< T6
surface
670 km
1200 km
0
−2.5e−06
−5e−06
T1
0
T2
50
T3
T4
T5
100
Time in Myr
150
T6
CMB
Associated PIA
Pole de rotation
VR=40 - manteau homogene
180˚
0˚
0˚
200
24
0˚
100
90˚
270˚
100
0
80
60
60
0˚
40
20
˚
30
Latitude in degree
15
0˚
21
12
Longitude in degree
300
T1
0
T2
50
T3
T4
100
Time in Myr
150
200
33
˚
0˚
30
0˚