Supplementary Figures: Azimuthal seismic anisotropyin the Earth’s upper mantle and the thickness oftectonic plates

A.J. Schaeffer1, S. Lebedev2 and T.W. Becker3

Geophysical Journal International

July 2016

1. University of Ottawa

2. Dublin Institute for Advanced Studies

3. University of Texas at Austin

1

Tab

leS

1:L

evel

ofsi

gnifi

can

ceb

etw

een

the

rad

iall

yav

erag

edco

rrel

atio

nco

effici

ents

conta

ined

inT

able

1fo

rth

ed

epth

ran

ge

50–700

km

,co

mp

ute

dfo

llow

ing

the

met

hod

ofB

ecke

ret

al.

(200

7).

Th

eu

pp

erh

alf

ofth

eta

ble

isfo

r〈r

8〉

and

the

low

erh

alf

isfo

r〈r

20〉.

SL2016sv

A-Y

B13sv

SL2016sv

Ar-YB13sv

SL2016sv

A-D

R2012a

SL2016sv

Ar-DR2012a

SL2016sv

A-3

D2015-0

7Sva

SL2016sv

Ar-3D2015-0

7Sva

YB13sv

-DR2012a

YB13sv

-3D2015-0

7sv

A

SL2016sv

A-Y

B13sv

0.00

0.05

90.

284

0.06

00.2

30

0.4

34

0.3

86

SL2016sv

Ar-YB13sv

0.13

90.

059

0.28

40.

060

0.2

30

0.4

34

0.3

86

SL2016sv

A-D

R2012a

0.31

40.

181

0.22

80.

119

0.1

73

0.3

82

0.3

33

SL2016sv

Ar-DR2012a

0.43

40.

309

0.13

40.

340

0.0

57

0.1

66

0.1

12

SL2016SvA-3

D2015-0

7Sva

0.13

90.

000

0.18

10.

309

0.2

88

0.4

84

0.4

38

SL2016SvAr-3D2015-0

7Sva

0.39

50.

268

0.09

00.

045

0.26

80.2

21

0.1

68

YB13sv

-DR2012a

0.63

50.

534

0.38

30.

260

0.53

40.3

02

0.0

55

YB13sv

-3D2015-0

7Sva

0.57

40.

465

0.30

40.

175

0.46

50.2

19

0.0

87

2

Tab

leS

2:L

evel

ofsi

gnifi

can

ceb

etw

een

the

rad

iall

yav

erag

edco

rrel

atio

nco

effici

ents

conta

ined

inT

able

1fo

rth

ed

epth

ran

ge

50–300

km

,co

mp

ute

dfo

llow

ing

the

met

hod

ofB

ecke

ret

al.

(200

7).

Th

eu

pp

erh

alf

ofth

eta

ble

isfo

r〈r

8〉

and

the

low

erh

alf

isfo

r〈r

20〉.

SL2016sv

A-Y

B13sv

SL2016sv

Ar-YB13sv

SL2016sv

A-D

R2012a

SL2016sv

Ar-DR2012a

SL2016sv

A-3

D2015-0

7Sva

SL2016sv

Ar-3D2015-0

7Sva

YB13sv

-DR2012a

YB13sv

-3D2015-0

7sv

A

SL2016sv

A-Y

B13sv

0.14

50.

467

0.03

80.

580

0.1

14

0.5

28

0.5

28

SL2016sv

Ar-YB13sv

0.50

30.

580

0.18

40.

677

0.2

56

0.4

08

0.4

08

SL2016sv

A-D

R2012a

0.05

80.

548

0.43

60.

144

0.3

69

0.8

21

0.8

21

SL2016sv

Ar-DR2012a

0.22

20.

309

0.27

70.

552

0.0

77

0.5

56

0.5

56

SL2016SvA-3

D2015-0

7Sva

0.34

90.

742

0.29

60.

537

0.4

92

0.8

73

0.8

73

SL2016SvAr-3D2015-0

7Sva

0.05

70.

456

0.11

50.

166

0.40

00.6

12

0.6

12

YB13sv

-DR2012a

0.82

40.

499

0.84

60.

716

0.92

90.8

00

0.0

00

YB13sv

-3D2015-0

7Sva

0.80

40.

461

0.82

80.

689

0.91

90.7

78

0.0

47

3

Tab

leS

3:L

evel

ofsi

gnifi

can

ceb

etw

een

the

rad

iall

yav

erag

edco

rrel

atio

nco

effici

ents

conta

ined

inT

able

1fo

rth

ed

epth

ran

ge

300–700

km

,co

mp

ute

dfo

llow

ing

the

met

hod

ofB

ecke

ret

al.

(200

7).

Th

eu

pp

erh

alf

ofth

eta

ble

isfo

r〈r

8〉

and

the

low

erh

alf

isfo

r〈r

20〉.

SL2016sv

A-Y

B13sv

SL2016sv

Ar-YB13sv

SL2016sv

A-D

R2012a

SL2016sv

Ar-DR2012a

SL2016sv

A-3

D2015-0

7Sva

SL2016sv

Ar-3D2015-0

7Sva

YB13sv

-DR2012a

YB13sv

-3D2015-0

7sv

A

SL2016sv

A-Y

B13sv

0.08

30.

318

0.46

00.

190

0.3

91

0.3

91

0.2

93

SL2016sv

Ar-YB13sv

0.08

80.

393

0.52

70.

270

0.4

62

0.4

62

0.3

69

SL2016sv

A-D

R2012a

0.48

30.

552

0.16

00.

135

0.0

81

0.0

81

0.0

27

SL2016sv

Ar-DR2012a

0.55

00.

613

0.08

50.

290

0.0

81

0.0

81

0.1

87

SL2016SvA-3

D2015-0

7Sva

0.33

60.

414

0.17

00.

252

0.2

13

0.2

13

0.1

08

SL2016SvAr-3D2015-0

7Sva

0.51

70.

583

0.04

30.

043

0.21

10.0

00

0.1

07

YB13sv

-DR2012a

0.51

70.

583

0.04

30.

043

0.21

10.0

00

0.1

07

YB13sv

-3D2015-0

7Sva

0.41

20.

485

0.08

50.

170

0.08

60.1

28

0.1

28

4

Figure S1: Comparison of isotropic component of SL2016svA (top row, SL2013sv; Schaeffer &Lebedev (2013)) with the isotropic component of SL2016svAr (bottom row) at depths of 100,250 and 500 km. Below each pair of panels, the saturation limits in percent are indicated,followed by the difference in RMS perturbations between each set of panels.

5

Figure S2: Synthetic tests illustrating the leakage between the isotropic and anisotropic modelparameters. The results of two tests are illustrated at depths of 100, 250 and 500 km. The leftpanels illustrate the isotropic velocity anomalies that result from the inversion of a syntheticdataset generated from a purely anisotropic model, whereas the right panels show the anisotropicstructure resulting from the inversion of a synthetic dataset derived from an isotropic model.Colour scales are indicated beneath each set of panels. Saturation limits for isotropic velocity aredenoted top left of each map. For both isotropic and anisotropic maps, the RMS perturbationin percent is denoted at bottom right.

6

Figure S3: Angular misfit of SL2016svA compared with five other azimuthal anisotropy models:SL2016svAr, LH2008a (Lebedev & van der Hilst, 2008; Becker et al., 2012), YB13sv (Yuan &Beghein, 2013), 3D2015-07Sva (Debayle et al., 2016) and DR2012a (Debayle & Ricard, 2013), atdepths of 50 km (left column), 100 km (centre column), and 200 km (right column). The colourscale indicates misfit in azimuthal anisotropy orientation between the models, with weightingbased on the amplitude on SL2016svA. Global average angular misfit (〈∆α〉) is indicated at thelower left of each panel, with the continental (〈∆α〉c) and oceanic (〈∆α〉o) averages at lowerright.

7

Figure S4: Average angular misfit 〈∆α〉 of fast-propagation directions in seismic anisotropymodels as a function of depth, with respect to SL2016svAr, LH2008a (Lebedev & van derHilst, 2008; Becker et al., 2012), DR2012a (Debayle & Ricard, 2013), 3D2015-07Sva (Debayleet al., 2016) and YB13sv (Yuan & Beghein, 2013). Solid curves: the global averages; dashed:for oceanic regions only; dash-dotted: continental regions only. Filled diamonds along thebottom axis indicate the depth-averaged global misfit for each model. Light grey line at 〈∆α〉=45◦ indicates a random average orientation between the two models. Note that the coloursrepresenting the different models change between the different panels.

8

Figure S5: Angular misfits between fast-propagation directions in tomographic models LH2008a(Lebedev & van der Hilst, 2008; Becker et al., 2012), YB13sv (Yuan & Beghein, 2013), 3D2015-07Sva (Debayle et al., 2016) and DR2012a (Debayle & Ricard, 2013) with the global sphericalharmonic expansion of SKS splitting distribution from Becker et al. (2012), for depths of 125 km(left column) and 250 km (right column). For both the tomographic models and SKS splitting,anisotropy was expanded in generalized spherical harmonics up to maximum angular degree20. The colour scale indicates misfit in azimuthal anisotropy orientation between the models,with weighting based on the amplitude of the tomography model. Global average angular misfit(〈∆α〉) is indicated at the lower left of each panel, with the continental (〈∆α〉c) and oceanic(〈∆α〉o) averages at the lower right.

9

Figure S6: Comparison of radially averaged inter-model correlations for global, upper-mantleanisotropy models, up to and including degree 8 (〈r8〉). The left panel illustrates correlationsfor the upper mantle at 50-200 km depths; the right panel shows the correlations for the depthrange 200-700 km. In each panel, the models are sorted (left-right, bottom-up) by the decreasingtotal correlation (excluding auto-correlation, the white squares along the diagonal); these valuesare shown along the right vertical axis of each panel. Note that the sorting of the models differsbetween the two depth ranges.

10

References

Becker, T. W., Ekstrom, G., Boschi, L., & Woodhouse, J. H., 2007. Length scales patterns,

and origin of azimuthal seismic anisotropy in the upper mantle as mapped by Rayleigh waves,

Geophys. J. Int., 171(1), 451–462.

Becker, T. W., Lebedev, S., & Long, M. D., 2012. On the relationship between azimuthal

anisotropy from shear wave splitting and surface wave tomography, J. Geophys. Res., 117(B1),

1–17.

Debayle, E. & Ricard, Y., 2013. Seismic observations of large-scale deformation at the bottom

of fast-moving plates, Earth Planet. Sc. Lett., pp. 1–13.

Debayle, E., Kennett, B. L. N., & Priestley, K., 2005. Global azimuthal seismic anisotropy and

the unique plate-motion deformation of Australia., Nature, 433(7025), 509–12.

Debayle, E., Dubuffet, F., & Durand, S., 2016. An automatically updated s-wavemodel of the

upper mantle and the depth extent of azimuthal anisotropy, Geophys. Res. Lett., 43, 1–8.

Lebedev, S. & van der Hilst, R. D., 2008. Global upper-mantle tomography with the automated

multimode inversion of surface and S-wave forms, Geophys. J. Int., 173, 505–518.

Schaeffer, A. J. & Lebedev, S., 2013. Global shear speed structure of the upper mantle and

transition zone, Geophys. J. Int., 194(1), 417–449.

Yuan, K. & Beghein, C., 2013. Seismic anisotropy changes across upper mantle phase transitions,

Earth Planet. Sc. Lett., 374, 132–144.

11

All in-text references underlined in blue are linked to publications on ResearchGate, letting you access and read them immediately.All in-text references underlined in blue are linked to publications on ResearchGate, letting you access and read them immediately.