Transcript Planetary formation and the role of water

The first 50 Ma:
Planetary formation
and the role of water
Eagle nebula:
Hester, Scowen
(ASU), HST,
NASA
L. T. Elkins-Tanton, E.M.
Parmentier
Mars Fundamental Research
Program
((NASA/J
The Hadean Earth: Old Version
Outline
•
•
•
•
Planetary accretion
Likelihood of a molten planet
Processes of solidification
Time to clement conditions
Planetary disk
NASA/JPL
Ca-Al inclusions in carbonaceous chondrites: 4567.2±0.6 million years old
Iron meteorites: The cores of failed planetesimals
Accretion simulations: Planets form from differentiated material
Planetary accretion simulation from Raymond et al. (2006), using 1054 initial
planetesimals from 1 to 10 km radius. Earthlike planets are formed, but their orbital
eccentricities are too high.
Outline
•
•
•
•
Planetary accretion
Likelihood of a molten planet
Processes of solidification
Time to clement conditions
Young planets are hot
Factors lead to heating and melting early in a terrestrial
planet’s history
Accretion of large bodies (meters [planetesimals] to hundreds of
kilometers [embryos])
Converts kinetic energy to heat
Radioactive decay of elements
•
26Al
(700,000 y) → 26Mg
•
60Fe
(1,500,000 y)→ 60Co (5.2714 y) → 60Ni
Core formation
Converts potential energy to heat
Far side of the Moon
NASA/Lunar Orbiter
4
Do terrestrial planets experience magma oceans? MOON
MOON:
Lunar anorthositic highlands led to the model of a deep initial magma ocean
Extreme age of the anorthositic highlands
Their positive Eu anomalies and the corresponding negative anomalies in the
mare basalts
Experimental evidence that plagioclase is not present on the liquidus of the
mare basalts
(Wood et al., 1970; Wood, 1972; Töksoz and
Solomon, 1973; Taylor and Jakes, 1974; Wood,
1975)
(Nyquist, 1977; Nyquist and Shih, 1992)
(e.g., Morse, 1982; Ryder, 1982; Haskin et al., 1981)
(Green et al., 1971a, b)
NASA/JPL/Galileo
MARS:
Do terrestrial planets experience magma oceans? MARS
SNC meteorites: Extreme eNd values and presence of 142Nd anomalies
Magnetized crustal provinces are evidence for very ancient crustal formation
Connerney et al., GRL (2001)
(Jones, 1986; Breuer et al., 1993; Harper et al., 1995;
Borg et al., 1997)
(Acuňa et al., 1999; Purucker et al., 2000)
Do terrestrial planets experience magma oceans? EARTH
EARTH:
Difference between 142Nd/ 144Nd anomalies between Earth and chondrites may
require an unsampled reservoir in mantle and may indicate silicate
differentiation >4.53 Ga (Boyet and Carlson, 2005)
BUT isotopic heterogeneity may be
expected
(Ranen and Jacobsen, 2006)
The core formation constraint
Sinking of iron-rich liquids by porous flow through a silicate
mantle requires interconnectivity of pore space.
[Stevenson, 1990].
In an iron-rich Martian mantle high dihedral angles would
prevent percolation of iron-rich core material at pressures
above 3 GPa.
[Terasaki et al., 2005] .
The timing of core formation is therefore an upper bound on
the completion of accretion, melting, and core formation
[e.g. Jacobsen, 2005].
Outline
• Planetary accretion
• Likelihood of a molten planet
• Processes of solidification
– Solidification
– Overturn
– Cooling before convection
• Time to clement conditions
Linked solidification and cooling processes
Elkins-Tanton
Volatiles partition among three reservoirs
Elkins-Tanton
Heat flux through atmosphere allows calculation of cooling rates
Process after Abe and Matsui (1985); Solomatov
(2000)
Emissivity parameterizations from Pujol and
Elkins-
Structure and thermal evolution of an evolving magma ocean
Elkins-Tanton et al. (2003)
Mineralogy and solidus based on data from Longhi et al. (1992)
Settling or entrainment: Calculation of liquid/crystal density inversion
EARTH,
MARS
2.1 GPa
MOON
3.6 GPa
Stolper et al. (1981) and Walker and
Agee (1988): olivine buoyancy
inversion between 7.5 and 9 GPa.
Elkins-Tanton et al. (2003)
Outline
Magma oceans would solidify from the bottom up
Volatiles partition between cumulates and atmosphere
Garnet may sink into a near-monominerallic layer
Shallowest cumulates are denser than deeper
cumulates
• Planetary accretion
• Likelihood of a molten planet
• Processes of solidification
– Solidification
– Overturn
– Cooling before convection
• Time to clement conditions
Idealized overturn
Elkins-Tanton et al. (2003,
2005)
Gravitational overturn: Nonmonotonic density gradients
Elkins-Tanton et al. (2005)
Overturn creates a laterally-heterogeneous mantle
Before overturn
After overturn
Contours of initial depth (proxy for
composition)
Contours of density
Axisymmetic models show: The
majority of overturn complete in <2
Myr; small-scale heterogeneities last a
long time
Models by Sarah Zaranek
Gravitational overturn: Numerical models in spherical coordinates
View movie m3H of numerical experiment of solidstate overturn of gravitationally unstable mantle
cumulates: Cooled surface produces shorterwavelength downwellings
Numerical experiment by Elkins-Tanton
Gravitational overturn: Numerical models in spherical coordinates
View movie m3L of numerical experiment of solidstate overturn of gravitationally unstable mantle
cumulates: Insulated surface produces longerwavelength downwellings
Numerical experiment by Elkins-Tanton
Mantle heterogeneity after gravitational overturn
Elkins-Tanton et al. (in prep, 2007)
Depths of origin of lunar volcanic rocks
Elkins-Tanton et al. (in prep, 2007)
Melting from overturn
Elkins-Tanton et al. (2005)
Outline
Cumulate overturn is fast and efficient
Following early differentiation there will be no simple compositiondepth relationships
Following early differentiation the mantle will have a stable density
stratification
• Planetary accretion
• Likelihood of a molten planet
• Processes of solidification
– Solidification
– Overturn
– Cooling before convection
• Time to clement conditions
Volatile content considerations
• The Moon likely accreted dry, unlike Mars and the Earth
• Volatiles form an insulating atmosphere and partition into
nominally anhydrous mantle cumulates
• No mafic quench crust is likely to form on a magma
ocean, because of density and temperature considerations
• Water in the silicate liquid inhibits the formation of
plagioclase
• If plagioclase forms and floats, it may significantly slow
planetary cooling
Volatile contents of possible planetary building blocks
If planet up to ~1,300 km
radius retains volatiles
(Safronov), bulk Marssize planet contains:
C1 (CI)
C2 (CM)
CR
C3 (CV, CO)
OC (Type 3)
C
H2O
3.5% 20%
2.5
13
1.5
6
0.5
1
0.5
1
John A. Wood (2005) The chondrite types
and their origins.
CO2
0.06%
H2O
1.2%
0.03
0.01
0.36
0.06
Atmospheric growth and planetary solidification: Earth
500-km-deep magma ocean on Earth
Elkins-Tanton et al. (in prep, 2007)
Magma ocean solidification
Parmentier and Elkins-Tanton (in prep,
2007)
Evolving magma ocean liquids
Elkins-Tanton et al. (in prep, 2007)
Depth of first plagioclase crystallization
PLANETARY SURFACE
EARTH (30 km,
0.005r)
Without a plagioclase
flotation crust, planetary
solidification is very fast:
90 vol% in less than 200,000
years
MARS (80 km,
0.02r)
MOON (240 km, 0.14r)
Elkins-Tanton et al. (in prep, 2007)
Dynamic effects of volatiles in cumulates
MAR
S
• Even small amounts of water have a large
effect on solid-state creep rates
• Volatiles are available for later degassing
Elkins-Tanton et al. (in prep, 2007)
Constraints on planetary formation from isotopic systems
Elkins-Tanton et al. (in prep,
Outline
Formation of a plagioclase flotation crust is controlled by volatile
content and planetary size
A plagioclase flotation crust may be the only way to slow solidification
In the absence of plate tectonics, mantle volatile content should be
controlled by processes of early differentiation
•
•
•
•
Planetary accretion
Likelihood of a molten planet
Processes of solidification
Time to clement conditions
MARS: Liquid water within ~15-20 Ma of a magma ocean
Elkins-Tanton et al. (in prep, 2007)
EARTH: Liquid water within ~35-40 Ma of a magma ocean
Elkins-Tanton et al. (in prep, 2007)
Conclusions
Solidification is fast
unless there is a
conductive lid
An initially molten planet
creates a mantle slow to
convect
There were likely multiple
magma ocean events
Valley et al. (2002)