Dynamics of Lunar mantle evolution: exploring the role of - PowerPoint PPT Presentation

Dynamics of Lunar mantle evolution: exploring the role of compositional buoyancy E.M. Parmentier Brown University September 26, 2018 JAXA/NHK

Global scale characteristics of the Moon 1) hemispheric crustal asymmetry a) mare basalt distribution b) crustal thickness 2) mare basalts: source complementary to anorthositic crust (Eu anomaly) melting at pressure higher than base of crust 3) timing of mare basalt emplacement and evolution from high to low‐Ti 4) mantle structure a) asymmetric distribution of moonquakes b) liquid outer core and weak layer above core/mantle boundary 5) strength and duration of dynamo magnetic field

Clementine Ti‐distribution Mare basalt ages http://curator.jsc.nasa.gov/education/LPETSS/marebasalt.cfm#fig10

Multiple saturation (Ol‐Opx) depth Eu fractionation between anorthosite of lunar ultramafic glasses and mare basalt source High‐Ti Low‐Ti Warren, (2004) Treatise of Geochemistry. Grove and Krawczynski (2009), Elements , 5 , 29–34. Khan and Mosegaard (2001) GRL 28 1791‐1794.

Lunar Prospector Thorium Distribution Thorium detected by gamma ray spectroscopy is concentrically distributed around the Imbrum impace basin (formed ~ 3.85 billion years ago – prior to main mare basalt eruption). Suggests that ilmenite bearing KREEP was concentrated in this area which are the later site for mare basalt eruptions.

Lunar crustal structure post Grail Wieczorek et al. Science 2012.

Interior Structure of the Moon velocity and density models Apollo seismic network Lognonné et al. (2003) Earth Planet. Sci. Lett. 211, 27‐44. Wieczorek (2009) Elements, 5, 35–40. Weber et al. (2011) Science 331:309‐312

Models matching Grail density, moment, and Love number: - fluid outer core with radius of 200–380 km - solid inner core with radius of 0–280 km and mass fraction of 0–1% - deep mantle zone of low seismic shear velocity - mass fraction of the combined inner and outer core is ≤1.5%. Williams, et al. JGR 2014

Schematic structure of solidifying magma ocean Stratigraphy predicted from ideal fractional solidification

Idealized overturn following fractional solidification Ringwood and Kesson (1976) Herbert PLPSC (1980) Spera (1992) Hess and Parmentier (1995)

Model for evolution of overturned mantle Internal structure 300 Myr model time structure (Zhang et al JGR 2013) composition temperature Creep activation energy viscosity ∝ exp � 𝑅 𝑆𝑈� Chemical structure just after overturn Q =100 kJ/mol Q =200 kJ/mol One take home message: Low viscosity allows cooling In dense IBC layer due to small scale convection.

Volume evolution as expressed in surface tectonic features * Absence of global‐scale thrust faulting (as seen on Mercury) limits the amount of contraction that has occurred. Solomon GRL (1978) * Magma ocean solidification would most likely result in expansion of young lunar crust (dikes) * Viscous relaxation of crust would prevent early tectonic features of contraction or expansion from being recorded permanently Elkins‐Tanton & Bercovici (2014) * Heat source distribution and crustal thermal conductivity affect volume evolution 1) Deep heat sources more effectively heat interior. 2) Low thermal conductivity (porous) crust raises the temperature at top of mantle. Zhang et al JGR 2013

GRAIL gravity field to degree and order 300 Andrews‐Hanna et al. Science 2012; science.1231753

Paleointensity of the lunar magnetic field as a function of time * Core dynamo existed on the Moon between at least 4.25 and 3.56 Ga swith surface field intensities reaching ∼ 70μT. * Paleomagnetic data from mare basalts demonstrate that the surface magnetic field had declined precipitously by 3.19 Ga. Tikoo, et al. (2017) Sci. Adv. 3.

Wet layered lunar mantle: affect on temperature/cmb heat flux just after overturn cmb heatflux no water water enriched layer temperature differnce no water 200‐km water enriched layer 500‐km water enriched layer Evans et al. (2014) JGR doi:10.1002/2013JE004494.

How to produce such a strong magnetic field at lunar surface? Core convection predicts magnitudes of only ~1μT For core radius 200–380 km surface dipole field 100–700 times smaller than at cmb Scheinberg, et al., EPSL doi.org/10.1016/j.epsl.2018.04.015 (2018).

KREEP concentrated on near side prior to eruption of hi‐Ti mare basalts Mare basalt ages Lunar Prospector Thorium Distribution Age of Imbrium impact basin Th radially distributed around Imbrium impact http://curator.jsc.nasa.gov/education/LPETSS/marebasalt.cfm#fig10 Haskin JGR (1998)

Mantle evolution with enriched heat sources in/below the PKT crust Equivalent of 10 km of KREEP basalt placed below a 40 km thick crust (blue), in the lower 20 km of the crust (cross hatch), or redistributed over the entire crust (orange). Laneuville et al., JGR 2013

Viscous Rayleigh-Taylor instabilty of a dense fluid layer Gravitational instability of a thin chemically dense (ilmenite‐rich) cumulate created during the fractionation of an anorthositic crust. Long wavelength instability needed to explain the hemispheric asymmetry. Spherical harmonic degree 1 evolves to be the fastest growing wavelength if the viscosity of the dense layer is sufficiently Parmentier, Zhong and Zuber 2002 low relative to that of the deeper mantle.

Dynamic models of the instability of ilmenite‐rich KREEP layer layer viscosity = 10 ‐4 layer viscosity = 10 ‐3 layer thickness = 100 km layer thickness = 100 km Haoyuan Li, et al., submitted JGR (2018)

Overturn during solidifcation solidification rate fast slow Slow solidification Boukare, et al, (2018) EPSL.

Initially molten Moon cumulates cumulate compaction + retained + buoyant melt migration melt ~ 10 4 years ~ 10 8 years Modified from: http://www.psrd.hawaii.edu/Mar09/magmaOceanSolidification.html

The ‘enigmatic’ Mg‐suite Mg‐rich cumulates Hess JGR (1994) Gross et al. ELSL http://dx.doi.org/10.1016/j.epsl.2013.12.006 (2014) Shearer et al. American Mineralogist doi.org/10.2138/am‐2015‐4817 (2015). Prissel et al. American Mineralogist doi.org/10.2138/am‐2016‐5581 (2016).

Melting Mg‐rich cumulates during rapid mantle overturn Modified from Shearer et al., New Views of the Moon, Ch. 4, 2006.

Melting Mg‐rich cumulates during rapid mantle overturn Melting/assimilation of Mg‐rich cumulates during overturn generates Mg‐suite Modified from Shearer et al., New Views of the Moon, Ch. 4, 2006.

Magma ocean evolution controlled by surface heat flux (conductive lid or atmosphere) Initially molten Moon cumulates cumulate compaction + retained + buoyant melt migration melt ~ 10 4 years ~ 10 8 years Modified from: http://www.psrd.hawaii.edu/Mar09/magmaOceanSolidification.html

Cumulate Compaction Fast deposition Slow deposition Shirley, J.Geol. 1986. 𝑙 � 𝜚 � 𝑏 � Both figures at same model times but 1 � 𝜚 � 150 note difference in vertical scale. 𝑊 � � Δ𝜛𝑕𝑙 𝜈𝜚 ⁄ 𝑀 � 𝑙 𝜊 𝜈 over 10 4 years Physical parameters with estimated values 𝑏 cumulate grain size 1 mm compacted 𝜚 melt fraction (at deposition boundary) 50% layer thickness Δ𝜛 cumulate – melt density difference 300 kg/m 3 = 100 ‐ 1000 km 10 17 ‐10 19 Pa‐s 𝜊 compaction viscosity ( = cumulate viscosity/ 𝜚 ) 𝜈 melt viscosity 1 – 10 Pa‐s 𝑙 3 x 10 ‐9 m 2 permeability (calculated) 𝑀 compaction length (calculated) 10‐250 km 𝑊 � melt velocity (calculated) 0.01‐0.1 km/yr

Magma ocean on early Mars? Mantle source composition Topography from Mars Orbiter Laser Altimetry of SNC meteorites Zuber et al. (2000) Science 287, 1788‐1793. 1) Mars displays early developed hemispheric asymmetry 2) Martian basaltic meteorite mantle sources lie on mixing line between lunar mafic cumulates and KREEP