Evolution of protoplanetary discs
and why it is important for planet formation
Bertram Bitsch
Lund Observatory
April 2015
Observations of planets
⇒ How can we explain this diversity?
Data from exoplanet.org
Bertram Bitsch (Lund) Evolution of protoplanetary discs April 2015 2 / 41
Outline
Introduction
Protoplanetary disc structure
Planet growth and migration
Planet formation in evolving
protoplanetary discs
Planetary system evolution
(a) Collapse of interstellar cloud
(b) Formation of protostar with protoplanetary disc made out of dust and gas (c) Small particles stick and
form bigger objects
(c)-(e) Formation of planetesimals and planetary embryos (f) Formation of planets,
clearing of gas in disc in
≈ 10 Myr
Bertram Bitsch (Lund) Evolution of protoplanetary discs April 2015 4 / 41
Protoplanetary discs in the Orion Nebula
Observations of protoplanetary discs
Bertram Bitsch (Lund) Evolution of protoplanetary discs April 2015 6 / 41
Composition of the disc
The inner regions of the disc are hotter than the outer regions:
⇒ Icy particles only in the outer parts of the disc!
The four steps of planet formation
1 Dust to pebbles
µm→ dm: contact forces during collision lead to sticking
2 Pebbles to planetesimals
dm → km: gravitational collapse of pebble clouds form planetesimals
3 Planetesimals to protoplanets
km → 1,000 km: gravity (run-away accretion)
4 Protoplanets to planets
Gas giants: 10 M⊕ core accretes gas (< 107 years) Terrestrial planets: protoplanets collide (107–108 years)
Bertram Bitsch (Lund) Evolution of protoplanetary discs April 2015 8 / 41
Time-scale to build gas giants
Mamajek (2009)
Giant planet formation has to happen within the disc’s lifetime of a few Myr!
Important quantities in the disc
Temperature T
Viscosityν = αH2ΩK with:
I H is the thickness of the disc
I ΩKis the Keplerian frequency, ΩK=pGM?/r3
I α≈ 10−2− 10−4
Gas densityρg and gas surface density Σg: Σg=ρgH√
2π
0.5 1 1.5 2 2.5
r [aJup] -0.3
-0.2 -0.1 0 0.1 0.2 0.3
z in [aJup]
1e-11 2e-11 3e-11 4e-11 5e-11 6e-11 7e-11
ρ in g/cm3
Bertram Bitsch (Lund) Evolution of protoplanetary discs April 2015 10 / 41
Importance of the disc structure
Growth and formation of planets relies on the disc structure:
Growth of dust particles to pebbles (Zsom et al., 2010; Birnstiel et al., 2012)
Movements of pebbles inside the gas disc (Brauer et al., 2008) Formation of planetesimals via streaming instability (Johansen &
Youdin, 2007)
Formation of planetary cores from embryos and planetesimals (Levison et al., 2010) or pebble accretion (Lambrechts & Johansen, 2012) Migration of planetary cores in the disc (Ward, 1997; Paardekooper &
Mellema, 2006; Kley et al., 2009)
Radial drift of dust particles
P
v
Kep(1− ) η
F F
GDisc is hotter and denser close to the star
Radial pressure gradient force mimics decreased gravity ⇒ gas orbits slower than Keplerian:
η = −1 2
H r
2
∂ ln(P)
∂ ln(r )
Particles do not feel the pressure gradient force and want to orbit Keplerian Headwind from sub-Keplerian gas drains angular momentum from particles, so they spiral in through the disc
Particles sublimate when reaching higher temperatures close to the star
Bertram Bitsch (Lund) Evolution of protoplanetary discs April 2015 12 / 41
Streaming instability: planetesimal formation
Gas orbits slightly slower than Keplerian
Particles lose angular momentum due to headwind
Particle clumps locally reduce headwind and are fed by isolated particles
v Kep(1− )η
F FG P
Streaming instabilities feed on velocity difference between two components (gas and particles) at the same location
Interested in details of the S.I.? Ask Anders Johansen or Chao-Chin Yang!
Streaming instability: lots of literature!
This is just a tiny example of literature to the streaming instabilty!
Bertram Bitsch (Lund) Evolution of protoplanetary discs April 2015 14 / 41
Streaming instability: a movie
(Johansen et al. 2011)
Protoplanetary disc structure
Bertram Bitsch (Lund) Evolution of protoplanetary discs April 2015 16 / 41
The Minimum Mass Solar Nebula - MMSN
Spread appropriate mass of solids around the orbit of each planet in the solar system and multiply by 100 (add gas)
Power law through data (Hayashi (1981), Weidenschilling (1977)):
Σg(r ) = 1700 r 1AU
−3/2
g/cm2 The planets accreted all solids (hence
”Minimum”)
The planets formed on their present orbits and did not move
Constraints from Observations
Σ∝ R−γ with γ = 0.4− 1.1
Bertram Bitsch (Lund) Evolution of protoplanetary discs April 2015 18 / 41
Time evolution of the star and the disc
Accretion rate ˙M (∝ Σg) changes with time (Hartmann et al., 1998)
⇒ Accretion rate changes by a factor of 100 in 5Myr!
Star changes luminosity in time (Baraffe et al., 1998)
⇒ Stellar luminosity changes by a factor of 3 in 5Myr!
0 0.5 1 1.5 2
2 5
0.1 1 10
10−9 5× 10−9 10−8 5× 10−8 10−7
LinL⊙
˙ MinM/yr⊙
t in Myr L M˙
⇒ The disc is subject to massive changes in its lifetime!
Disc Model
2D hydrodynamical disc model with viscous heating, radiative cooling and stellar irradiation with S ∝ L?:
1 2 3 4 5 6 7 8 9 r [aJup]
0 0.5 1 1.5 2 2.5 3
z in [aJup]
-13 -12.5 -12 -11.5 -11 -10.5 -10 -9.5 -9
log ( ρ in g/cm3 )
Bitsch et al., 2013
Mass flux through disc: ˙M constant at each r withα viscosity:
M = 3πνΣ = 3παH˙ 2ΩKΣ
Bertram Bitsch (Lund) Evolution of protoplanetary discs April 2015 20 / 41
Influence of opacity on cooling
Cooling of the disc:
F=− λc ρκR∇ER
Grey area marks transition in opacity at the ice line Change of opacity:
⇒ change of cooling Change of cooling:
⇒ change in T (r)
log (κ in cm2/g)
T in K
Transition κR = κP κ* -3
-2.5 -2 -1.5 -1 -0.5 0 0.5 1
10 100 1000
TinK
r [AU]
Transition M = 3.5˙ × 10−8M⊙/yr MMSN 50
200 500
1 10 100
2 3 4 5 20
1 10
Bitsch et al., 2015
Influence of viscosity and ˙ M
Hydrostatic equilibrium:
T = H r
2
GM? r
µ R bump in T : bump in H/r M disc:˙
M = 3πνΣ = 3παH˙ 2ΩKΣ M constant at each r :˙
⇒ dip in Σ
Σing/cm2
r [AU]
H/r
M = 3.5× 10˙ −8M⊙/yr MMSN
50 200 500
10 100 1000
2 3 4 5 10 20
1
M = 3.5× 10˙ −8M⊙/yr 0 MMSN
0.01 0.02 0.03 0.04 0.05 0.06 0.07 0.08
Bitsch et al., 2015
Bertram Bitsch (Lund) Evolution of protoplanetary discs April 2015 22 / 41
Importance for the streaming instability
Streaming instabilities feed on velocity difference between two components (gas and particles) at the same location, caused by a reduction of the effective gravitational force by the radially outwards pointing force of the radial pressure gradient:
∆ =ηvK
cs =−1 2
H r
∂ ln(P)
∂ ln(r ) Reduced ∆ helps the formation of large clumps via streaming instability (Bai & Stone, 2010b)
∆
r [AU]
H/r
M = 3.5˙ × 10−8M⊙/yr 0 MMSN
0.02 0.04 0.06 0.08 0.1 0.12
2 3 4 5 10 20
1
M = 3.5˙ × 10−8M⊙/yr 0 MMSN
0.01 0.02 0.03 0.04 0.05 0.06 0.07 0.08
Bitsch et al., 2015
Change of ˙ M in time
TinK
r [AU]
Transition M = 1.0˙˙ × 10−7 M = 7.0× 10−8 M = 3.5˙ × 10−8 M = 1.75˙ × 10−8 M = 8.75˙ × 10−9 M = 4.375× 10˙ −9
50 200 500
10 100
2 3 4 5 20
1 10
M decreases with decreasing Σ˙ M = 3πνΣ = 3παH˙ 2ΩKΣ Inner disc dominated by viscous heating for high ˙M, dominated by stellar heating for low ˙M
ΣGing/cm2
r [AU]
H/r
M = 1.0× 10˙ −7 M = 7.0× 10˙ −8 M = 3.5× 10˙ −8 M = 1.75× 10˙ −8 M = 8.75× 10˙ −9 M = 4.375× 10˙ −9 1
10 50 100 200 500 1000
2 3 4 5 10 20
1
M = 1.0× 10˙ −7 M = 7.0× 10˙ −8 M = 3.5× 10˙ −8 M = 1.75× 10˙ −8 M = 8.75× 10˙ −9 M = 4.375× 10˙ −9 0
0.01 0.02 0.03 0.04 0.05 0.06
Bitsch et al., 2015
Bertram Bitsch (Lund) Evolution of protoplanetary discs April 2015 24 / 41
Time evolution of the disc
Evolution in time follows Hartmann et al. 1998 equation:
log M˙ M/yr
!
=−8.00 − 1.40 log tdisc+ 105yr 106yr
.
TinK
r [AU]
Transition M = 1.0˙ × 10−7 M = 7.0˙ × 10−8 M = 3.5˙ × 10−8 M = 1.75˙ × 10−8 M = 8.75˙ × 10−9 M = 4.375˙ × 10−9
50 200 500
10 100
2 3 4 5 20
1 10
Planet growth and migration
Bertram Bitsch (Lund) Evolution of protoplanetary discs April 2015 26 / 41
Pebble accretion
Small pebbles (τf < 1) can be easily accreted by planetesimals (Lambrechts & Johansen, 2012)
Stokes number τf and friction time tf: τf = tfΩK= ρ•R
ρGcsΩK = ρ•R ρGH
”Pebbles are weakly enough bound to the gas to feel the gravitational pull from the core, but strongly enough to deposit their kinetic energy through drag forces, when passing the core.”
Lambrechts & Johansen, 2012
Scaling of pebble accretion
Core growth via pebbles
(Lambrechts & Johansen, 2014):
M˙c= 2 τf 0.1
2/3
rHvHΣPeb Growth faster in inner regions of the disc
Red dotsmark pebble isolation mass: Pebble accretion does not continue forever!
Lambrechts & Johansen, 2014
Bertram Bitsch (Lund) Evolution of protoplanetary discs April 2015 28 / 41
Pebble isolation mass
Pebble isolation mass:
Miso= 20 H/r 0.05
3
MEarth
⇒ After pebble isolation mass is reached, gas accretion can start!
Lambrechts et al., 2014
Gas accretion
0 10 20 30 40 50 60 70 80 90 100
1 105 2 105 3 105 4 105 5 105
M [ME]
t [yr]
Mc Menv Mtot
Phase 1: accretion of solid core until pebble isolation (≈ 2 × 105 yr) Phase 2: envelope contraction (≈ 4 × 105 yr, following Piso &
Youdin, 2014), when Mc> Menv
Phase 3: rapid accretion of gas onto core when Mc< Menv (Machida et al. 2010)
Bertram Bitsch (Lund) Evolution of protoplanetary discs April 2015 30 / 41
Planets in discs
Planet embedded in a disc creates a wake
Wake directed forwards in inner disc and backwards in outer disc Planetpushes inner (outer) discinwards (outwards),
inner (outer) discpushes planetoutwards (inwards)
Type-II-migration
High mass planets (MP > MJup) open gaps in discs (Crida et al.
2007):
P = h/q1/3+ 50α/qh2 < 1 with h = H/r Planet is locked in the disc
Planet moves with disc on ”accretion timescale”:
τν = rp2/ν (up to a few Myrs) It is called Type-II-migration
0 50 100 150 200 250 300 350 400 450 500 550
0.5 1 1.5 2 2.5
Σ in g/cm2
r [aJup]
initial 200 Orbits
Bertram Bitsch (Lund) Evolution of protoplanetary discs April 2015 32 / 41
Type-I-migration
Planet fully embedded in the disc, free to migrate with respect to the gas Waves carry energy and angular momentum
Torque calculated through Lindblad Resonances (e.g. Goldreich & Tremaine, 1980)
ILR: positive torque, OLR: negative torque OLR stronger than ILR:
⇒ inward migration( ˙a∝ q)
120 140 160 180 200 220 240 260 280 300 320 340
0.5 1 1.5 2 2.5
Σ in g/cm2
r [aJup]
initial 200 Orbits
Migration map
Paardekooper et al. (2011): Analytic torque estimate of embedded low mass planets using gradients in the disc (Σ∝ r−αΣ; T ∝ r−βT)
γΓtot/Γ0=−0.9 − 3.22αΣ+ 3.92βT Γ0 =q h
2
ΣPrP4Ω2K
Bertram Bitsch (Lund) Evolution of protoplanetary discs April 2015 34 / 41
Planet formation in evolving
protoplanetary discs
Evolution track starting at t
0= 2 Myr
0.0001 0.001 0.01 0.1 1 10 100 1000
5 20 30 40
0.1 1 10
M [ME]
r [AU]
r0 = 4 AU r0 = 10 AU r0 = 20 AU r0 = 30 AU r0 = 40 AU tD=3.0Myr
Bertram Bitsch (Lund) Evolution of protoplanetary discs April 2015 36 / 41
Planet formation at all orbital distances r
0at t
0= 2 Myr
M [ME] rf [AU]
r0 [AU]
Mcore Menv Mtot rf r0
0.1 1 10 100 1000
5 10 15 20 25 30 35 40 45 50
0.1 1 10 100 1000
0.01 0.02 0.03 0.04 0.05 0.06 0.07 0.08
2 5 20
1 10
H/r
r [AU] tdisc= 0.0Myr
tdisc= 0.1Myr tdisc= 0.2Myr tdisc= 0.5Myr tdisc= 1.0Myr tdisc= 2.0Myr tdisc= 3.0Myr
Planet formation at all orbital distances r
0at t
0= 2 Myr
M [ME] rf [AU]
r0 [AU]
Mcore Menv Mtot rf r0
0.1 1 10 100 1000
5 10 15 20 25 30 35 40 45 50
0.1 1 10 100 1000
0.01 0.02 0.03 0.04 0.05 0.06 0.07 0.08
2 5 20
1 10
H/r
r [AU]
tdisc= 0.0Myr tdisc= 0.1Myr tdisc= 0.2Myr tdisc= 0.5Myr tdisc= 1.0Myr tdisc= 2.0Myr tdisc= 3.0Myr
Bertram Bitsch (Lund) Evolution of protoplanetary discs April 2015 37 / 41
Planet formation in evolving disc
Everything below blue line: pebble isolation reached Everything above white line: Mcore> Menv
0.5x106 1.0x106 1.5x106 2.0x106 2.5x106 3.0x106
5 10 15 20 25 30 35 40 45 50 t0 [yr]
r0 [AU]
0.1 1 10 100 1000
MP in ME 0.1 0.5 1.0
5.0 10.
20.
Power law disc: MMSN
0.5x106 1.0x106 1.5x106 2.0x106 2.5x106 3.0x106
5 10 15 20 25 30 35 40 45 50 t0 [yr]
r0 [AU]
0.1 1 10 100 1000
MP in ME
0.1 20.
10.
Bertram Bitsch (Lund) Evolution of protoplanetary discs April 2015 39 / 41
Schematic view of planet formation
Summary
Protoplanetary discs evolve in time and change their properties (Σ, T , H), which matters for all processes inside the disc!
Resulting planetary systems depend crucially on the underlying disc structure
Early planet formation produces mainly gas giants
Giant planets form far out in the disc: no in-situ formation Smaller planets can form in-situ, but form late
Late formation scenario preferred: larger diversity of planetary types as predicted by observations
Bertram Bitsch (Lund) Evolution of protoplanetary discs April 2015 41 / 41