Diapositive 1

1
Terrestrial Planet Accretion
Alessandro Morbidelli CNRS, Observatoire de la
Cote dAzur Nice, France
2
OUTLINE

• The dynamics of planet accretion
• Final orbital excitation, accretion timescale
• Dependence on the environment (Jupiters orbit,
  embryos in the asteroid belt.)
• Comparison with radioactive chronometers data
• Open issues
• Final Angular Momentum Deficit
• Mass of Mars
• Accretion of water from the asteroid belt region
• Compatibility with geochemical constraints
• Alternative scenarios

3
Dynamics of planet accretion from embryos and
planetesimals
Obrien, Morbidelli and Levison, 2006 (ObML06)
4
Canup Asphaug (2001)
Giant collisions constitute the growth mode of
terrestrial planets The proto-Lunar collisions
was NOT an exceptional event (Agnor et al.,
1999) No two-stage models please
5
The Final Orbits
AMD
Inclination (degrees)
AMDss -0.0018 AMDsimul
-0.0010 AMDchambers -0.0070 Difference with
previous work the consideration of a large
number of individually small planetesimals
(Dynamical Friction is the Key)
ObML06
Semi major axis (AU)
6
Formation Timescale
Median time for acquiring 90 of final mass 40My
Median time of last giant impact 31 My
Timescales are a factor 3 shorter than
previously obtained
Very good agreement with the pre-2007
interpetation of 182Hf/182W chronology (see also
Nimmo and Agnor, 2006)
ObML06
7
The previous results have been obtained assuming
that Jupiter and Saturn had initially their
current (eccentric) orbits. If we assume that
the orbits of Jupiter and Saturn were
quasi-circular, as expected from models of their
formation/evolution, the results of terrestrial
planets formation simulations are less good
AMDss -0.0018 AMDsimul -0.0030
ObML06
Semi-major axis (AU)
8
The terrestrial planets accretion timescale also
becomes longer (100 My).
Median accretion time
9
The interaction with embryos from the asteroid
belt is the main cause of the long accretion
timescales of the terrestrial planets If
Jupiters orbit is eccentric so all the embryos
have been eliminated quickly, without having a
chance to interact with the growing planets The
situation is analog to that where planetary
embryos are not assumed to have formed in the
asteroid belt (Agnor et al., 1999 Kenyon and
Bromley 2006), which also leads to a 30 My
accretion timescale Longer accretion timescales
lead to larger angular momentum deficit because
there are fewer planetesimals at later times to
exert dynamical friction
10
We need embryos in the asteroid belt to explain
the main properties of the asteroid population
(Wetherill, 1992 Petit et al, 1999 OBrien et
al., 2007)
11
In the Hydrodynamical simulations, Jupiter and
Saturn keep a quasi-circular orbit (Morbidelli
and Crida, 1997)
Sat.
eccentricity
semi major axis (arbitrary units)
Sat.
Jup.
Jup.
Time (orbital period at a1)
Time (orbital period at a1)
We believe that the eccentricities of Jupiter and
Saturn have been acquired during the LHB
(Tsiganis et al., 2005 Gomes et al., 2005)
12

• If the preferred environment is that with embryos
  in the asteroid belt and Jupiter Saturn on
  quasi-circular orbits, we need to understand
• If a 100 My timescale of accretion is consistent
  with chronometers
• How to obtain in the simulations final planets
  with smaller angular momentum deficit

13
THE AGE OF THE EARTH
Difficult to determine because the interpretation
of the chronometers depends on how the Earth
formed ( of giant collisions etc.) and on the
degree of core/mantle re-equilibration during
these impacts
14
Perhaps the best way to date the formation of the
Earth is to date the formation of the Moon

• Why do we think that the Moon formed with the
  last giant impact
• SPH simulations work better (they require a
  smaller projectile/target ratio)
• The stability of the Earth-Moon system might be
  at risk if there are close encounters with other
  massive embryos
• If the Earth accretes a significant amount of
  mass after the Moon-forming event, the Moon
  should have accreted more siderophile elements
  than it actually did
• In the isotope equilibration scenario of Pahlevan
  and Stevenson (2007), the Earth needs to remain
  essentially unchanged after the Moon forming event

15
For the Moon, the H/W chronometer indicates a
formation age of 30-40 My
However, new results from the ETH group shows
that Lunar 182W is cosmogenic. This delays the
formation of the Moon to gt 60 My Other
chronometers also indicate a formation age of
100 My
More in the next talk (Kleine)
16
THE ANGULAR MOMENTUM DEFICIT PROBLEM
Eccentricity/Inclination could be damped by
tides exerted by a tenuous remnant nebula
Kominami and Ida, 2004

• Typically produces too many planets, too small
  and too fast.
• Proposed solutions
• Secular resonance sweeping during gas dissipation
  (Nagasawa et al., 2005 Thommes et al., 2008) but
  require (again) eccentric Jupiter Saturn.
• Turbulence (Ogihara et al., 2007) but timescales
  remain short.

Problems with Type-I migration
17
THE ANGULAR MOMENTUM DEFICIT PROBLEM
Partial regeneration of the planetesimal
population when embryos collide with each other
should allow dynamical friction to remain
effective as long as giant collisions occur (to
be tested with simulations)
Asphaug et al., 2006
Erosional collision
Accretional collision
18
THE PROBLEM OF THE MASS OF MARS
In the simulations the planet at the location of
Mars is systematically too massive. The only
scenario that gives the correct mass to Mars
assumes that Jupiter and Saturn were from the
beginning at their current locations on eccentric
orbits (seems impossible giant planets should
have migrated from a more compact
configuration) Possibly, simulations starting
from smaller embryos could lead to better results
(to be tested)
Chambers, 2001
19
Accretion of Water on Earth from the Asteroid Belt
The eccentricity of Jupiter plays a crucial role
in determining from which source regions the
planets accrete material.
Circular Jupiter
If Jupiter was originally on a quasi-circular
orbit, some 15 of the terrestrial planet
material should have come from water-rich
carbonaceous chondrites from the outer asteroid
belt, as originally proposed in Morbidelli et al.
(2000)
ObML06
Eccentric Jupiter
20
In the circular-Jupiter case, the material from
the outer asteroid belt is accreted among the
latest
Raymond et al., 2006
21
The Earth acquires the water relatively late, but
nevertheless during its own accretion process.
NOT A LATE VENEER
The Late veneer (defined as the material accreted
after the last giant impact with an embryo)
constitutes only 1 of the total Earth mass if
Jupiter had a quasi-circular orbits, or 10 of
the total Earth mass if Jupiter had an eccentric
orbit, and is essentially due to material from
the inner solar system, NOT from the outer
belt. Origins of water and of mantle siderophile
elements are not necessarily related.
22
The Earth and the Moon The Oxygen isotope problem
The Earth and the Moon have indistinguishable
oxygen isotope composition. The material that
constitutes the Moon comes from the proto-Lunar
impactor. All Solar System bodies have different
oxygen isotope composition. HOW IS THIS
POSSIBLE?
23
Even if planets had accreted NO material from the
asteroid belt, it is unlikely that the
proto-Earth and the proto-Lunar impactor could
have the same composition even if they formed
very close to each-other
24
A possible solution EQUILIBRATION! Pahlevan and
Stevenson, 2007
25
ALTERNATIVE SCENARIOS ON THE ORIGIN OF WATER

• Local planetesimals were water-rich, because
  water could be absorbed by grains even inside the
  snowline (Muralidharan, Drake et al., 2008)
• Would water be lost when grains accrete into
  planetesimals?
• Why are the parent bodies of enstatite and
  ordinary chondrites so dry?
• Dust small icy/hydrated planetesimals drifting
  inwards from beyond the snowline due to gas drag
  could have brought water to the terrestrial
  planet region (Lauretta and Ciesla, 2005)
• This mechanism was invoked by Cyr et al. (1999)
  to explain the hydration of C-type asteroids
• The deficiency of water in S/E type asteroids
  suggests that this mechanism was not effective
  inside 2.5-3 AU
• Primitive atmospheres of H could have been
  captured by planetary embryos from the solar
  nebula the reaction of H with the silicate could
  have hydrated the embryos (Genda and Ikoma, 2008)
• This could explain why embryos were hydrated even
  if planetesimals were not
• The water produced by this mechanism would have a
  solar D/H composition. Necessity for a
  fractionation mechanism. Similarity with D/H
  ratio in carbonaceous chondrites would be a
  coincidence.

26
CONCLUSIONS

• Quite satisfactory understanding of terrestrial
  planet formation in the Solar System
• Reasonable final orbital excitation, although
  slightly excessive if Jupiter is initially on a
  circular orbit
• Accretion timescale consistent with new
  chronology of the Moon forming event
• But why is Mars so small?
• Simulations start to be a useful guide for a
  better interpretation of geochemical data and
  constraints.
• The accretion of water from the asteroid belt is
  a likely scenario, but alternatives (e.g.
  reaction of silicates with a H-atmosphere on
  embryos) need to be explored further.