“Towards a comprehensive evolution model for
the outer solar system”
or...
the “Nice II” model
In collaboration with:
A. Morbidelli, H. Levison, R. Gomes
&
R. Brasser, D. Nesvorny
Overview
Why did we need a new view/model for the OSS
The Nice model (v1.0, 2005)
What were the main problems of this model?
The v1.5 Nice model (2007)
Planet-disc interaction revisited → Nice II model...
Conclusions
Why did we need a new model for the OSS?
in situ formation of U and N is highly unlikely → planet
migration...
smooth migration cannot explain the observed (e,I)'s
not high (as for EPS) but much higher
than formation models give...
Why did we need a new model for the OSS?
The strongly depleted and excited Kuiper belt
asteroids, Trojans, ...
Why did we need a new model for the OSS?
The origin of the Late Heavy Bombardment (LHB)
•Cataclysmic event triggered 3,9 Gy ago, ~600My after
terrestrial planet formation
•Global event: traces found on Mercury, Venus, Earth,
Mars, Vesta…., possibly on giant planets satellites
•20.000x the current bombardment rate: 1 km object
impacting the Earth every 20 years!
•Duration: 50-150 My
The Nice model (v1.0 - 2005)
Planets more closely packed
than before → resonances come
to play...
an initially stable configuration
→ planet-disk interactions force
smooth migration
resonance crossing excites
eccentricities → global instability
Instability suppressed by
dispersing the disk → mass
depletion
1J=2S resonance crossing
The Nice model (v1.0 - 2005)
… and the LHB main characteristics
Several “success” stories since:
trojans, irregular satellites, KBOs,
primitive main-belt asteroids, ...
reproduces the orbits of the planets...
What were the problems with the Nice model?
Initial conditions for
the planets were
simply made-up...
●
… they should
represent the “final”
conditions of a
previous phase of
gas-driven migration
Critical (although
monotonic)
dependence on tcrit
with disk parameters
●
The v1.5 Nice model (2007)
J-S migration in a gas disk
(first ~3 My) can lead to
capture in a 3:2 MMR (Masset &
3:2 res
Snellgrove 2001, Pierens & Nelson 2008)
J-S can stop migrating after
that, for a range of reasonable
disk parameters (Morbidelli & Crida
2007)
… all four outer planets can be
trapped in a multiple (Laplace-type)
resonance (Morbidelli et al., 2007)
→ “proper” initial conditions
(see also Batygin & Brown 2010)
Planetesimals can then extract
the planets from their resonance
and...
… force them to migrate, cross
MMRs (e.g. 5:3)
→ evolution similar to Nice I
GREAT initial conditions
WORSE for having a delayed
instability
→ the planets are “pealing off”
the inner edge of the disk but
without moving in a
→ needs extreme fine tunning
of the disk to get tcrit ~ 700 My
Planet-disk interactions revisited
What is missing in all these simulations ?
→ the disk's self-gravity → particle-particle velocity stirring
Gives secular energy exchange between the disk and the planets!
move sunwards ...
… Nice II model
Eccentricity of Uranus
increases (due to 2:3 with S) ...
… until a critical value (~0.07) for
which the critical argument of the
3:4 with Neptune starts circulating
→ U-N unlocked from MMR
→ the system becomes unstable!
… Nice II model
t50%= 730 My
Planetesimal
encounters
Systems go unstable preferentially LATE – 9/21 in 1 Gy – t50% = 730 My
Not a sensitive dependence on the inner edge of the disk!
Both “Proper” initial conditions and GREAT timing for the LHB.
Conclusions
We are converging (slowly) towards a “Nice II” model
Basic ingredients: Nice v1.5 initial conditions + self-gravitating
disc
→ consistent with previous phase of gas-driven migration
→ late instabilities naturally provided by a slow planet-disk energy
exchange + break-up of a resonance (needs more work...)
(+) Point: it should work similarly for all configurations found by
Morbdelli et al (2007) and Batygin & Brown (2010)
(-) Point: not all configurations have the same “success” rate...
→ we're looking into this....
a closer look to some ISS constraints...
Divergent planet migration leads to
variation of the secular frequencies
→ resonances between g5 and gi
(i=2,4) are possible
If migration was slow (~ 10
My) or resonances were
approached slowly
→ possibly devastating effect
on terrestrial planets and
asteroids
a closer look to some ISS constraints...
This favors “fast”
planet migration
→ dominated by
planet-planet
encounters
→ a subset of Nicemodel runs
The terrestrial planets (and
asteroids) are not significantly
affected ...