Transcript Slide 1

Slab Imaging in Continental
Subduction under the Northern
Apennines
Bianchi I.1,2, J.J. Park3, T. Mayeau3, N.
Piana Agostinetti2, V. Levin4
1 Universita’ “La Sapienza” di Roma
2 INGV Roma
3 Yale University, CT
4 Rutgers University, NJ
Thanks to members of the RETREAT project seismology tea
APENNINES MARK A SUBDUCTION ZONE
Italian peninsula is almost parallel to the
convergence direction between Africa and
Eurasia.
There is considerable earthquake activity along the
Apennines BUT Earthquakes are concentrated
beneath the mountains and descend to the
depth of ~ 100 km.
Apennines Slab
(Piana Agostinetti et al, 2008)
Much more slab is
imaged and caused by
30 My long subduction.
Lucente et al., 1999
An eastward retreating slab has been
invoked as principal driving process in
this area, able to describe the present
position of the orogen and to justify the
juxtaposing of extension over compression.
(Malinverno & Ryan 1986)
Adriatic Sea
(Adria microplate)
Adapted from Faccenna et al., 2004
Tyrrhenian Sea
(back-arc basin)
Conceptual Tectonic Model for RETREAT (Brandon, Willett, Pazzaglia, et al)
Important Aspects: Crust thickens and thins while passing through orogen
Rigid mantle block beneath orogen to match surface tectonics in geodynamic model
GRAZIE
ADRIA Plate
(Africa)
TYRRHENIAN
Plate (Europe)
SLAB Retreat
Things to look for with Seismology:
Crustal Thickness
Top of the Downgoing Lithosphere
Crustal Wavespeeds (Vp &
Vs)
Crustal Models
Finetti et al., 2001
Test for Crustal Thickness
I:
Vertical motion
surface-wave phase
velocity
3-station wavelet-based crosscorrelation, obtains both
apparent phase velocity and
propagation azimuth
RETREAT
Seismic Array
Correlation
measure
(white is
good)
Phase
velocity
(km/sec)
Wavelet
transform
Propagation
azimuth
Phase velocity (km/sec)
Match the locations of station triangles to a coarse grid of RETREAT Field Area
Compute median phase-velocity, discarding values >6 km/sec or <1.5 km/sec
GRAZIE
Surface Wave period (seconds)
Faster phasevelocity at ~20s
period in Tuscany,
suggestive of
thinner crust there
GRAZIE
Retreat seismic array
Test of Crustal Thickness II:
Receiver Function Transect
Closely-spaced
station transect
Epicentral distribution
of teleseisms M > 5.5
12659 teleseismic
seismic records used
Receiver Functions overview
Three components:
DV(t) = I(t)*S(t)*EV(t)
DR(t) = I(t)*S(t)*ER(t)
DT(t) = I(t)*S(t)*ET(t)
ER(ω)=DR(ω)/DV(ω)
ET(ω)=DT(ω)/DV(ω)
(Ammon, 1997)
Earthquakes
From NE
Tuscan Moho
Adriatic
Moho?
Earthquakes
From SW
Tuscan Moho
Wrong
Polarity!
Piana Agostinetti et al, EPSL in press 2008
Back-azimuth RF Dependence implies both Radial &
Transverse RF amplitude
and implies interface dip or anisotropy
Energy only on
R component
Energy on both R and T components
How to evaluate both radial &
transverse RFs?
Enhance Ps pulses that can be
explained in terms of dipping interface
or anisotropy
π/2 Shift
0
By Katsuhiko SHIOM
Analysis Procedure:
1- Definee each profile from spots
10 km
apart
2- Finding the piercing point at
40 km
depth of every RF collected.
3- Every spot grabs RFs for which 40-km
piercing points are in
Immediately surrounding area.
4- Applying the Harmonic expansion and
migration focusing at 40-km depth.
AA’: spots grab piercing points into a 20x120 km area.
BB’: spots grab piercing points into a 20x80 km area.
CC’: spots grab piercing points into a 20x120 km area.
DD’: spots grab piercing points into a 20x40 km area.
EE’: spots grab piercing points into a 20x120 km area.
RESULTS
0
130 km
(at least 300 RFs to build one wiggle)
0
240 km
0
210 km
Attempt to fit observations with simple dipping isotropic
interface:
The continuity of dipping plane is not clear under the
Tyrrhenian side
Or the interface continues but seismic properties change to
give a weaker pulse
DATA
Modeling:
Synthetics
Observed
3 Model segments to represent BB’ profile:
-Tyrrhenian Model: Moho at 25 km depth for
120 km
-Overlapping Model: Tyrrhenian Moho,
anisotropic layer and
Adriatic dipping Moho for 70km
-Adriatic Model: dipping Moho for 50 km
Discussion
•
•
•
•
Tyrrhenian & Adriatic crusts have different thicknesses, but no confirmation of central
thickening at Apennines crest
A south-dipping subcrustal anisotropic layer appears to lie beneath under the Northern
Apennines orogen, but with limited extent along strike
anisotropic texture aligns with interface plunge
No evidence for semi-rigid mantle body at base of Apennines orogen
POSSIBLE INTERPRETATIONS:
- Adriatic lower crust subducting with lithosphere?
- Subducted lower crust
coming back to shallower
depths as for UHP rocks
(Beaumont et al. 2001)
- Suction of subcrustal
material from Tyrrhenian
toward Adria as Apennines
Slab retreats
Test Analysis for a single station:
Cutigliano (CUTR) verifies the
direction-dependent subcrustal
interface  Anisotropy needed
=
N10ºE
CUTR Synthetics
CUTR Observed