Mantle Plumes As Presently Imaged by Seismic Tomography Barbara

Mantle plumes as presently imaged by seismic tomography

Barbara Romanowicz1,2

1Collège de France, Paris 2Univ. of California, Berkeley

Contributors: V. Lekic, S. French, S. Cottaar, Kaiqing Yuan Collège de France, June 26th , 2018 Hot spots and mantle plumes

76 Ma 61

50 43 30 20 10 5 0

T. Wilso n , 1 9 6 3 Morgan, 1971 Isotopic composition of trace elements (Nd-Sr)

Hotspot volcanoes

Wo rkma n , 2 0 0 5 Griffiths and Campbell, 1990 Depth (km)

Temper atur e ( oC) Courtesy of Cinzia Farnetani • Mantle Plumes exist!!

• As currently resolved by global seismic tomography: – Rooted at the core-mantle boundary – Clustered within the large low shear velocity provinces of the lower mantle – Not purely thermal • Thermo-chemical or • Partially molten and/or involve a non-Newtonian rheology for the lower mantle

• Their morphology highlights the more vigorous convection in the upper 1000 km of the mantle and a very sluggish lower mantle Imaging mantle plumes: Challenge nr 1

Principle of travel time tomography

Hawaii

mantlecore Imaging mantle plumes: Challenge nr 2

Low velocity regions of relatively small size are hidden from view when considering only first arriving waves

2 1

1 Low velocity 2 body

1 Iceland Hotspot and Plume

Absolute w/r 4.5 km/s Relative velocities R. Allen, Nolet, Morgan et al., 2002 P wave travel time tomography

Depth (km)

Montelli , Nolet, Dahlen et al., 2004, Science P wave travel time tomography

Nolet, Allen and Zhao., 2005 P wave travel time tomography Hawaiian “plume”

(1) (1)

(2) (2)

mantlecore

mantle

Wolfe, Solomon et al., 2011 Long wavelength global shear velocity models

“Voting map”

Lekic et al., 2012 Shear velocity Depth = 2800 km

“LLSVP”, or… “Superplumes”

S362ANI SAW24B16 S20RTS

Paleo-pole locations (Besse and Courtillot, 2002) “Degree 2” pattern in the mantle Hotspot distribution

Shear attenuation in the transition zone

Shear velocity at 2800 km depth

Dziewonski, Lekic and Romanowicz., 2010 S40RTS

S40RTS: Ritsema, Deuss et al, 2011 Full Waveform Tomography

Data

Synthetics

3D Synthetics: seismic wavefield in complex 3D Earth models : now can be computed accurately using the Spectral Element Method (SEM) and directly compared to observed “full” seismograms

- Challenges: computational time increases as ω3 UC B e r k e l e y - Thin slow layers in crust - Several hundred of events, non-linear (iterations) S40RTS

S40RTS: Ritsema et al, 2011 Full waveform tomography SEMUCB_WM1

S40RTS

French and Romanowicz, Nature, 2015 Whole mantle model SEMUCB_WM1

Pacific superswell region

Model: SEMUCB_WM1 French and Romanowicz, 2015 Atlantic Plumes

French and Romanowicz, 2015 Rickers et al., 2012 These broad plumes are found under major hotspots that lie over the LLSVPs

SEMUCB-WM1 at 2800 km depth

f r e Other hotspots according to French and Romanowicz, 2015 n Steinberger (2000) c These “plumes” are broad Canary Cape Verde

MacDonald St Helena -> Quasi-vertical from the CMB to ~ 1000 km depth -> Often laterally offset from corresponding hotspots Canary Cape Verde

Depth= 1000 km

MacDonald St Helena - Denser basal layer - Viscosity varies with depth and temperature: - factor of 10 jump at 660 km - hot upwellings 2 orders of magnitude less viscous than background Slabs trapped at 1000 km

107 115

western Java 106 114

410 410 660 ±1.2% 1000 105 ±1.2% 113

104 112 eastern Java

111 103

Fukao and Obayashi (2013) ->The plumes are rooted in (mostly) isolated patches of very low shear velocity Canary Cape Verde

St Helena MacDonald SEMUCB_WM1 Depth = 2800 km Hawaian plume viewed from South East Hawaii

SEMUCB_WM1 ULVZ: Height: ~20-25km Diameter:~910 km Velocity reduction:~20% Cottaar and Romanowicz, 2012 Detection of an “ultra low velocity zone” Filter: 10-20 s at the base of the mantle Observation Prediction

SHdiff, 100-110 degrees azimuth (dg)

time around Sdiff (s)

ULVZ: Height: ~25km Diameter:~800 km Velocity reduction:~20% Cottaar and Romanowicz, 2012, EPSL ULVZ at the base of the Iceland Plume

Yuan and Romanowicz, 2017, Science (A) Event 1, 12.0km, 94−96° (B) (C) (D) −20 x2e−06 m/s SH Data −20 x3.6e−06 m/s SH 3D −20 x3.6e−06 m/s SH SEMUCB

−25 −25 −25

−30 −30 −30

−35 −35 −35 Azimuth [° Azimuth ]

−40 −40 −40

−45 −45 −45 −20 0 20 40 60 80 −20 0 20 40 60 80 −20 0 20 40 60 80 (A) Event 1, 12.0km, 94(A)−96°Event 1, 12.0km,(B) 94−96° (B) (C) (C)(D)∆t [s] ∆t (D)[s] ∆t [s] −20 x2e−06 m/s SH Data−20−20x2e−x3.6e06 m/s−06 m/s SH DataSH 3D−20−20x3.6ex3.6e−06 m/s−06 m/s SH SHSEMUCB 3D −20 x3.6e−06 m/s SH SEMUCB

−25 −25−25 −25−25 −25

−30 −30−30 −30−30 −30

−35 −35−35 −35−35 −35 Azimuth [° Azimuth ] Azimuth [° Azimuth ]

−40 −40−40 −40−40 −40

−45 −45−45 −45−45 −45 −20 0 20 40 60 80−20−20 0 0 20 20 40 40 60 60 80 80−20−20 0 0 20 20 40 40 60 60 80 80−20 0 20 40 60 80 ∆t [s] ∆t [s]∆t [s] ∆t [s]∆t [s] ∆t [s]

Yuan and Romanowicz, 2017 (A) Event 1, 12.0km, 94−96° (B) (C) (D) −20 x2e−06 m/s SH Data −20 x3.6e−06 m/s SH 3D −20 x3.6e−06 m/s SH SEMUCB

−25 −25 −25

−30 −30 ULVZ: −30 Circular base −35 −35 Height: −~15km35 Azimuth [° Azimuth ] Diameter:~800 km −40 −40 Velocity − reduction:~25%40

−45 −45 −45 −20 0 20 40 60 80 −20 0 20 40 60 80 −20 0 20 40 60 80 (A) Event 1, 12.0km, 94(A)−96°Event 1, 12.0km,(B) 94−96° (A) Event(B) (C)1, 12.0km, 94−96° (C)(D)∆t [s](B) ∆t (D)[s] (C) ∆t [s] (D) −20 −20−20x2e−06 m/s −20−20x3.6e−06 m/s −20 x3.6e−06 m/s x2e−06 m/s SH Data x3.6e−06 m/s SH DataSH 3D −20x3.6e−x2e06− m/s06 m/sSH SHSEMUCB 3DSH Data −20 x3.6e−06SH m/s SEMUCB SH 3D −20 x3.6e−06 m/s SH SEMUCB

−25 −25−25 −25−25−25 −25 −25 −25

−30 −30−30 −30−30−30 −30 −30 −30

−35 −35−35 −35−35−35 −35 −35 −35 Azimuth [° Azimuth ] Azimuth [° Azimuth ] Azimuth [° Azimuth ]

−40 −40−40 −40−40−40 −40 −40 −40

−45 −45−45 −45−45−45 −45 −45 −45 −20 0 20 40 60 80−20−20 0 0 20 20 40 40 60 60 80 80−20−20 0−20 0 20 0 20 40 20 40 60 40 60 80 60 80− 8020 0−20 20 0 40 20 60 40 80 60 80 −20 0 20 40 60 80 ∆t [s] ∆t [s]∆t [s] ∆t [s]∆t [s]∆t [s] ∆t [s] ∆t [s] ∆t [s]

Yuan and Romanowicz, 2017 => Roots of “fat” plumes likely contain partial melt Yellowstone

SEMUCB_WM1 Nelson and Grand,2018

French and Romanowicz, 2015 Conclusions u Mantle plumes do exist !! u The major upwelling flow is in the form of ~ 25 broad low shear velocity that are located over the large low shear velocity provinces (LLSVPs)

u They extend quasi-vertically from the CMB to ~1000 km depth -> very sluggish circulation in the lower mantle u Horizontally deflected around ~1000 km depth and likely entrained into secondary scale, more vigorous, upper mantle circulation u The roots of at least some of these “fat plumes” contain large, axi-symmetric ULVZ’s , likely dense and containing partial melt u They may be the manifestation of strong topography at the top of an otherwise very thin core-mantle interaction zone of dense partial melt

u LLSVP’s may not be piles extending to mid-lower mantle depths, but rather bundles of fat plumes in a halo of hotter than average background.