Talk contents Part A: Introduction and methods - Caltech
←
→
Page content transcription
If your browser does not render page correctly, please read the page content below
Talk contents Part A: Introduction and methods 1. Observational motivation 2. Population synthesis principle 3. Input physics: global models 4. Initial conditions 5. Observational biases Part B: Results and perspectives 1. Compute the population: individual systems 2. Overview of statistical results 3. Comparisons with observations 4. Perspectives and conclusions
Compute the synthetic population
Andrews+2018
Initial Conditions: Probability
Models of individual Global end-to-end form-
distributions of disk properties
processes ation & evolution model Disk gas mass
From
Disk dust mass
Accretion, migration, … Disk properties planet properties observations
Disk lifetime
Draw and compute
synthetic population
New instrumentation
better observational constraints
Apply observational
detection bias
Predictions
(going back to the full
synthetic population)
Observed population Stat.
Comparison:
Observable sub-population Model
- Frequencies solution
No match: change - Orbits, masses, radii, luminosities
- Architecture, multiplicity Match found
parameters, improve - Correlations
model, reject model …..
One learns a lot even if a synthetic population does not match the observed one!
Emsenhuber et al. 2021 1. Compute the population: individual systems
Global model: Example outcome
Numbers are
Bern Generation III model -[Fe/H]
-Mgaz,0 [MSun]
-Msolids,0 [MEarth]
Setup/parameters
-solar-mass stars
-1000 systems (stars)
-100 ini6al embryos per
system
-embryo mass 1 Mluna
-uniform in log out to 40
AU at t=0.
-viscosity α=0.002
-opacity red. factor 0.003
Gas-dominated giant (Menve/Mcore >=1) (“Jovian”) Protoplanet lost during formation
and evolution process (accreted by
Volatile rich planet, with H/He (“Neptunian”)
another more massive protoplanet,
Volatile rich planet, without H/He (“water world”) ejected, collided with the star).
Iron/silicate planet with H/He (“H/He terrestrial”) Black bar: peri- to apoastron
Iron/silicate planet without H/He (“Earth-like”) distance (showing eccentricity)
A global model in action: low solid mass Initial conditions -initial disk gas mass: 0.017 Msun -initial solid mass: 57 MEarth Class 1 architecture
Class 1. The in situ Earths and ice worlds systems
t=5 Gyr
A global model in action: mid solid mass Initial conditions -initial disk gas mass: 0.042 Msun -initial solid mass: 100.1 MEarth Class 2 architecture
Class 2. The migrated sub-Neptune systems
t=5 Gyr
Peas-in-a-pod (Weiss+2018, Millholland+2017): Misra+2021
A global model in action: high solid mass Initial conditions -initial disk gas mass: 0.042 Msun -initial solid mass: 245 MEarth Class 3 architecture
Class 3. The mixed systems
A global model in action: very high solid mass Initial conditions -initial disk gas mass: 0.044 Msun -initial solid mass: 327 MEarth Class 4 architecture
Class 4. The dynamically active giants
Manara et al. 2016 2. Overview of statistical results
Formation of the a-M diagram
Ini6al condi6ons
-solar-mass stars
-1000 systems (stars)
-100 ini6al embryos per
system
-ini6al mass 1 Mluna
Emsenhuber et al. 2021b (NGPPS II,
arXiv: 2007.05562)Nominal population 1 M☉
Distant
super t=5 Gyr Distant
Jupiters
Hot
Cold Jupiters super
Jupiters
Jupiters
Planetary desert Hot Cold Jupiters
Close-in Jupiters
low-mass
? Planetary desert
wet
Close-in
low-mass ice worlds
HZ
dry
Several fundamental features of
the observed popula6on can be
recovered.
The varia6ons of the disk ini6al condi6ons over a range likely occurring in nature leads to
a large diversity of planetary systems, similar as observed.10
Number of system
Number of system
102
Number of syst
Number of syst
10 2
Statistical overview
10 for 1 M☉ 102
2
1
Fundamental 10 demographical results 101
10 1 101
• Synthetic planetary system contain on average 0
8 planets more massive than 1 M⊕
10
including all orbital distances (no obs. constraints yet).
0
5 10 0 10
0 2 4 6
•
10 Fraction of systems
0 with giants
5 planets:
10 18 % (all orbital distances),
0 2 only
4 1.6 %6at >10 AU.
Number of planets Number of planets
anets
• Systems with giantsNumber of average
contain on planets 1.6 giant planets.Number of planets
3 Giant
t 10 Giant
Number of systems
3
10 10 embryos
Number of systems
20 10 embryos
embryos
2 Initial number of
10 50 20
embryosembryos embryos per disk
2
10 50 embryos
100 embryos
1 100 embryos
10
101 The result is independent of the initial nb of embryos
per disk.
0 1 2 3
3 0 1
Number of planets 2 3 Only 5 stars out of 1000 have 3 giant planets.
anets Number of planets
rent types of planets, for the four multi-embryo populations presented in this study. In each panel,
ymofanother to make
•di↵erent
Low-mass them
typesplanets more
of planets, visible. We
for theMfour
(0.3-3 seehabitable
for example
⊕) inmulti-embryo
that for
44the
populations
zone: giant
%presentedplanets,
of stars. outstudy.
inMean
this ofmultiplicity
theIn
1000
each panel,
1.3 (rather
planets. About
tally from an equal
another to makenumber
them (100
moreeach) have
visible. Weone
see or
fortwo giants.that
example Less
for than ten out
the giant of theout
planets, 1000
of the 1000
ynumber
low). Mean
per system
giant planets.
[Fe/H]
Aboutthatan of stars
occurs. with habitable planets -0.11. Different from Solar
equal number (100 each) have one or two giants. Less than ten out of the 1000
System.
highest number per system that occurs.What sets the outcome?
25
The most important ini6al
Number of planets in system with M ≥ 1 Me
4
10 condi6on is the mass of
20 solids ini6ally present in
the disk.
Total system mass [Me]
103
Solar system 15 Grey lines: efficiency of
planetary system
10 forma6on(including H/He)
102 per system
10
We need 2-3 MMSN to
1
1 form the Solar System
10 5 mass.
0.1
100 0
1 2 3
10 10 10
Initial mass in solids [Me]
Another important ini6al condi6on is external disk photoevapora6on,
seRng the disk life6me. It depends on stellar birth environment and
Adapted from Mordasini+ in prep. affects the emerging system architecture (e.g., Winter et al. 2020).Mayor et al. 2011 3. Comparisons with observations
Comparisons with observations
Andrews+2018
Initial Conditions: Probability
Models of individual Global end-to-end form-
distributions of disk properties
processes ation & evolution model Disk gas mass
From
Disk dust mass
Accretion, migration, … Disk properties planet properties observations
Disk lifetime
Draw and compute
synthetic population
New instrumentation
better observational constraints
Apply observational
detection bias
Predictions
(going back to the full
synthetic population)
Observed population Stat.
Comparison:
Observable sub-population Model
- Frequencies solution
No match: change - Orbits, masses, radii, luminosities
- Architecture, multiplicity Match found
parameters, improve - Correlations
model, reject model …..
One learns a lot even if a synthetic population does not match the observed one!Statistical comparison with HARPS survey
Observed
Nb of planets: 161
Mayor et al. 2011 Nb of stars w planets: 102
Lines: detec+on
probabili+es 100% Multiplicity: 1.6
95%
Synthetic biased
[Earth Mass]
80%
1000.
60% Nb of planets: 317
HARPS: high 40% Nb of stars w planets: 204
accuracy radial 20%
Multiplicity: 1.6
velocity planet 10%
searcher. 100.0 5%
2% • Agreement: similar global
structure: relative distribution
(concentrations, voids)
M2sini
10.0 • Agreement: Mean multiplicity
GTO Survey: 822 system architecture.
solar-like stars.
• Disagreement: Factor 2 in
Known bias.
absolute number. Poss.
1.0 explanations: Initial conditions?
See also recent 10+0 10+1 10+2 10+3 10+4 Cluster environment (cf. Winter
results of California et al. 2020)?
Legacy Survey Period [days] • Disagreement: Hot Jupiters.
(Rosenthal+2021, Poss. explanation: Kozai plus
Fulton+2021) tidal circularisation channel
missing in model.
Adapted from Emsenhuber, Mordasini, Udry, Mayor, Marmier + in prep. (NGPPS VII)Quantitative comparison mass distribution
80 Mass distribution detected planets
0.8
80
70 • Agreement: Fundamental bimodal structure
Synthetic
Scaled synth(normalised)
Synthetic
70 HARPS
Scaled
HARPSsynth • Agreement: Change in regime at ~20 M⊕:
60
frequency
HARPS smoking gun of core accre6on: runaway gas
0.6
60
50
accre6on Mcore~Menve~10 M⊕ (but see also
number
Number
50 Bennet et al. 2021).
40
Number
0.4
40
Normalised
30 • Disagreement: Giant planets factor 2-3 too massive
Raw
30
20 • Disagreement: Too few intermediate mass planets by
0.2
20 factor ~2-3 (planetary desert, Ida & Lin 2004).
10
100
100 101 102 103 104
too fast gas accre6on (cf. Nayakshin et al. 2019)
00 0 Msini [M ]
10 101 102 103 104
Msini [M ]
Similar for gas accre6on rate derived from several 3D hydrodynamic models (Machida et al. 2010, D'Angelo et
al. 2010, Bodenheimer et al. 2013)
Possible explana6ons: low viscosity disks (Ginzburg & Chiang 2019a), magne6c regula6on (Batygin 2018, Cridland
2018), angular momentum barrier (Takata & Stevenson 1996), 3D circula6on (Szulagyi et al. 2014), ….
Popula6on synthesis makes it possible to quan%fy discrepancies between theory and observa6ons.Statistical comparison with Kepler survey
c) d)
NASA Kepler space telescope (transits) Synthetic
t=5 Gyr
e)
Observed
Mulders et al. (2019)
Figure 5. Planet occurrence rates of the planet population synthesis model (abcd) compared to those of Kepler (e). Bins span
equal areas in logaritmic area units (d log R d log P ) and roughly correspond to hot Jupiters, warm giants, super-earths, and
• Agreement: rela6ve occurrence, except for (sub)-
mini-Neptunes. The models systematically underestimate the occurrence of mini-Neptunes compared to the observed rates.
Numbers: Bias corrected occurrence rates
(Nb of planets in area / Nb of stars x 100) Neptunes (2-4 R⊕)
• Disagreement: absolute occurrence: too high again (x 5)
For a direct comparison, the end-to-end • Radius valley not clearly visible for high number of ini6al
model should include the long term evolution embryos: Impact stripping? Water-rich planets? Enriched
/ internal structure calculation including
atmospheric escape -> R and L/mag.
envelopes?Statistical comparison with direct imaging
SPHERE@VLT SHINE GTO Actual detections & Synthetic population &
survey (Vigan et al. 2020) sensitivity maps sensitivity maps
150 stars Plot by Clemence Fontanive
cf. Nielsen et al. 2019 GPI
SPHERE
Very Large Telescope VLT
+4.7
Observed: 5.8−2.8 %
Frac6on of FGK stars w. planets
+0.5
(M=1-75 MJ, a=5-300 AU) Synthe6c: 3.4−0.5 %
• Agreements: overall frequency, mass-luminosity rela6on (β Pic b)
• Distant giants in synthesis: Single, massive, eccentric planets from
Probes very different kind of scanering events (see Marleau+2019b), mean eccentricity: 0.39
planets and a different • Disagreement: No HR 8799-like systems: 4 distant massive giants on
observable (luminosity). rather circular orbits
• Structured disks? Forma6on by gravita6onal instability?4. Perspectives and conclusions
Comparisons with observations
Andrews+2018
Initial Conditions: Probability
Models of individual Global end-to-end form-
distributions of disk properties
processes ation & evolution model Disk gas mass
From
Disk dust mass
Accretion, migration, … Disk properties planet properties observations
Disk lifetime
Draw and compute
synthetic population
New instrumentation
better observational constraints
Apply observational
detection bias
Predictions
(going back to the full
synthetic population)
Observed population Stat.
Comparison:
Observable sub-population Model
- Frequencies solution
No match: change - Orbits, masses, radii, luminosities
- Architecture, multiplicity Match found
parameters, improve - Correlations
model, reject model …..
One learns a lot even if a synthetic population does not match the observed one!6
?
Perspectives and conclusions
4
20
2
0 0
2 4 6 8 10 20 40 60 80 100 2 4 6 8 10 20
Companion Mass [MJup ] Companion Mass
100 pc GAIA satellite Nancy Grace Roman satellite
50 pc Detectable
25 pc GAIA
USE AM PLOT TO SHOW GAIA Data release ~2025 Data release ~2030
AND ROMAN
Astrometric technique Microlensing
Explore uncharted territory
Expected yield: 20’000 technique
say nb of giant planets! to 70’000 giant Expected yield:
Detectable
goals planets (!) 2000-3000 cold low-
Roman mass planets
Use: old map -> new map with america
PLATO: transits ARIEL: atmospheric
spectroscopy
Observa6onally driven field: consider coming missions
The future is bright regarding new sta6s6cal
observa6onal constraints.
Data release ~2030 Launch ~2029
Exquisite knowledge of planet mass distribu6on and
demographics in giant and low-mass regime. Ideal to Transit technique Atmo. spectroscopy
inves6gate mechanisms of gas and solid accre6on. Hab. zone planets Sta6s6cs of atmo.
Blue lines: 5 σ detec+on limits for GAIA (Courtesy D. Segransan, Geneva Obs.) Temporal evolu6on composi6onsTerra incognita as litmus test
2021 ~2030
Cross-checking the same theoretical model with population synthesis in many different and especially
unexplored parameter spaces:
Key to understand whether theory really captures the governing underlying physics and is not merely
a sophisticate fit tweaked to explain already known observations.
Much to do on the theoretical side: initial conditions & early phases, disk models (beyond α-models),
hybrid pebble-planetesimal models, link formation-atmospheric composition, gas accretion,…
Observing planet formation as it happens as a new direct constraint on planet formationConclusions •Population synthesis is a tool to compare theory and observation to improve understanding of planet formation • use full wealth of observational constraints • put detailed models to the test • see global statistical consequences: which processes are key? •Observational constraints on many processes • solid and gas accretion rate • N-body dynamics • orbital migration rate •See link between disk and planetary properties •Predict yield of future instruments/space missions •Continuously improving models • population syntheses depend on progress of formation theory as a whole • many new theoretical developments to test, many new obs. constraints to come
Resources
Population synthesis review papers
-Benz et al., Protostars & Planets VI, 691, 2014
-Mordasini et al., IJA, 201, 2015
-Mordasini, Handbook of Exoplanets, 143, 2018
DACE data base: Bern population synthesis models
https://dace.unige.ch/evolution/index
All NGPPS data publicly available
via dedicated interac6ve online
tool on DACE website
Freely available toy population synthesis model
http://nexsci.caltech.edu/workshop/2015/#handsonYou can also read
