Extrasolar
Planetary
Systems
Forma&on
and
Discovery
of
Phil
Armitage
(Colorado)
Physical principles of planet forma2on orbital stability + encounter outcomes Giant exoplanets
dynamical evolu&on, migra&on Terrestrial exoplanets
coupling to giants, Kepler
Extrasolar Planetary Systems Forma&on and Discovery of
protoplanetary
disks transi4on
disks debris
disks
planets
What can we observe?
Gas
Dust s ~ 1mm Gas – dust
interac4on Dust from collisions
N-‐body dynamics Missing:
• growth mm – km (“planetesimals”)
• planet-‐gas disk interac4ons
• young planetary systems
105 yr 106 yr 107 yr 108 yr 109 yr µm
mm m km
103 km
Gas disk life&me
Dust, coagula4on / fragmenta4on equilibrium
Planetesimals Cores
Observa2ons
Debris disks
β Pic
Theory
Dynamics of growth
a
Dynamics of growth
a
Feeding zone:
€
Δa = C M
p3M
*#
$ % &
' (
1 3
a
mass in zone 4πaΔaΣ increases for reasonable Σ(r), m ~ a
Dynamics of growth
a
Feeding zone:
€
Δa = C M
p3M
*#
$ % &
' (
1 3
a
mass in zone 4πaΔaΣ increases for reasonable Σ(r), m ~ a
Ra4o:
€
vesc
vK ∝ Mp M*
a Rp
if large, sca[ering or ejec4on
if small, collision
Dynamics of growth
Terrestrial planets
vesc < vK: terrestrial planets grow “in place”
High ini4al disk Σ:
• more massive terrestrials
• fewer
Feeding zone narrow: collisions lead to low eccentricity
simula&on: Sean Raymond
1 2 3 r / AU
108
107
106
t / yr e
Dynamics of growth
Works well at leading order for the Solar System – largest discrepancy is over-‐predic4on of mass of Mars…
Dynamics of growth
Giant planets
Require: form > 5 MEarth core before gas is dissipated in ~ few Myr
too li[le mass in
feeding zone sca[er rather than
collide -‐> slow growth
~1-‐3 AU ~10-‐20 AU
Movshovitz et al. 2010
Large uncertain4es due to envelope opacity (cooling) but consistent with Jupiter, Saturn to leading order
Giant Exoplanets
Observa&ons
Winn et al. (2010)
Sky projected angle between stellar spin axis and planetary orbital axis
Require migra4on and
eccentricity excita4on Hot Jupiters are some4mes misaligned or retrograde
Measured via
transit RV spectroscopy
Giant Exoplanets
Observa&ons
Working hypothesis: explained as consequence of
• “standard” giant planet forma4on (core accre4on) -‐ possibly at modestly smaller radii than in Solar System
• evolu2on due to exchange of energy and angular momentum with gas, other planets, binary
companion
Giant Exoplanets E, L exchange processes
Planet-‐gas
disk interac4on Kozai-‐Lidov
interac4on (planet + misaligned binary)
Planet-‐planet
sca[ering Secular chaos
Giant Exoplanets Planet-‐planet scaSering
Moeckel & Armitage (2012)
Planet forma4on + migra4on typically leads to unstable mul4ple planet system as gas dissipates
Eccentricity and hot Jupiters form dynamically
Occurs early, but gas may be negligible to leading order
Ini4al condi4ons:
3 gas giants, circular orbits, forming as close as 1 AU
N-‐body only
Match f(e) distribu4on for giant exoplanets 0.1 AU < a < 1 AU Payne et al. (2014)
Ini4al condi4ons:
3 gas giants, circular orbits, forming as close as 1 AU
N-‐body only
Broad inclina4on distribu4on of planets sca[ered to e ~ 1 and then 4dally circularized (c.f. Nagasawa et al. 08; Beauge & Nesvorny 12)
Sca[ering gives consistent but not unique solu4on to most close-‐in proper4es of giant exoplanets
Dynamics of growth
Large radii
Neptune and extrasolar
planets at “large” radii (50 AU) are also incompa4ble with in situ core accre4on
HR8799 and other directly imaged systems cri4cal constraints
Marois et al. 2008
Dynamics of growth
Large radii
First evidence for a new gravita4onal instability channel for giant planet forma4on?
Predicted to be inevitable for large massive disks, but hard to keep masses below brown dwarf scale…
(Rice et al. 2010; KraSer et al. 2010)
OR – mul4ple cores formed at smaller scales, migrated out, and later accreted gas?
Need more data….
Terrestrial Exoplanets Theory
• “Solar System-‐like”
-‐ slow (~100 Myr), hence gas free -‐ in place
-‐ ~independent of giant planets
• Giant-‐controlled
-‐ substan4ally impacted by violent giant planet dynamics
• Migra4on dominated
-‐ orbital evolu4on among terrestrial precursors
Giant dominated
Assume
planet-‐planet sca[ering
dominant (Raymond
et al. 2011, 12)
Terrestrial Exoplanets Theory
Rich terrestrial planet systems live in systems with near-‐
circular giant planets
Predict currently unobserved popula4on of dynamically
excited terrestrials
Kepler systems
Batalha et al. 2013
2 obvious challenges for theory…
High abundance of planets with radii not represented in Solar System… what are these planets?
Many stars with close-‐in planetary systems, where forma4on 4me is so short (<105 yr) that gas disk effects must be important
evidence for a migra4on dominated mode?
magnetosphere r ~ 0.05 AU
inner edge of dead zone
T ~ 800 K
snow line (Kretke & Lin ‘07)
solid par4cles
solids drin radially inward under aerodynamic drag and encounter
traps in disk (Hasegawa & Pudritz 11)
planetesimal forma4on at traps
M = 10˙ 8M yr 1
αin = 10-‐2, αout = 10-‐3, width w = 2h
c.f. coagula&on models of Drazkowska et al. (2013)
Local pressure maxima trap par4cles of sizes
formed from coagula4on (mm-‐cm) readily, especially in outer disk
If gas disk has local maxima, par4cle
density much higher at these loca4ons
high densi4es lead to planetesimal forma4on via collec4ve instabili4es
2D streaming: Jake Simon
“Least problema4c” route to planetesimal forma4on from small par4cles involves instabili4es in coupled gas / par4cle mixtures (“streaming instability”, Youdin & Goodman 2005)
Require loca4ons in disk where ρpar4cle / ρgas > threshold to form planetesimals… in a disk with traps this will be at the loca4on of the traps
forma4on of planets if planetesimals form at preferred loca4on in inner disk produce packed mul4ple systems for mass fluxes of ~10 MEarth / Myr into traps in inner disk
also form co-‐orbital planets…
not yet clear if orbital proper4es are be[er or worse match to
Kepler systems than in situ models Bruns & Armitage, in prep
Summary
Solar System appears to be a planetary system where the giant planets were only moderately dynamically ac&ve, and the mass in the terrestrial region was low enough that
the Earth & Venus formed aner the gas was gone
More ac4ve giant planets (higher mass, closer together, less damping from Kuiper belt) are common – result in eccentric extrasolar gas giants, hot Jupiters
More mass in (or migra4ng through?) the terrestrial region forms low mass planets earlier – close-‐in Kepler systems?