Transcript Document

Testing Planet Migration Theories
by Observations of Transiting
Exoplanetary Systems
University of Tokyo
Norio Narita
1/39
Contents


Introduction (15 min)

Diversity of Extrasolar Planets

Planet Migration Theories
Motivation (10 min)



Transits and the Rossiter-McLaughlin Effect
Recent Results (15 min)

Simultaneous Subaru / MAGNUM Observations

Analysis and Results
Conclusion and Future Prospects (3 min)

Significance of Our Results

New Targets and Prospects
2/39
Discovery of extrasolar planets
The first extrasolar planet 51 Peg. b
was discovered by radial velocity
measurements in 1995.
More than 200 extrasolae planets have been discovered so far.
We can discuss statistics of their distribution.
3/39
Diversity of extrasolar planets
Jupiter
Semi-major axis – Planet minimum mass Distribution
4/39
Diversity of extrasolar planets
hot Jupiters
1 AU
(Close-up of the distribution)
5/39
Diversity of extrasolar planets
Eccentric Planets
Jupiter
Semi-major axis – Eccentricity Distribution
6/39
How do they form?


Giant planets lie at ～0.1AU

should originally form at larger orbital distances

planetary migration to inner orbits
Eccentric planets are common

would have mechanisms of eccentricity excitation
 How can we explain these features?

gravitational interactions with other bodies in protoplanetary disk
7/39
Planet migration theories
1.
disk-planet interaction

“Type I & II migration”

resultant planets would not have large eccentricity
2.
planet-planet interaction

“jumping Jupiter model”

have possibilities to produce large eccentricity
3.
planet-binary companion interaction

“Kozai oscillation” in binary planetary systems

also have possibilities to produce large eccentricity

explain HD 80606 system (e=0.927, Wu & Murray 2003)
8/39
Type I & II migration

planetary cores form beyond the snow line

the cores interact with the surrounding disk


planets migrate inward due to torque exchange with the disk

Type I migration： less than ～10ME

Type II migration： more than ～10ME
damping eccentricities and also inclination
Type II migration
Type I migration
(Leiden Observatory Group)
9/39
Jumping Jupiter model

giant planets interact with each other in multi-planet systems

leads to orbital instability

one planet is thrown into close-in orbit

the planet obtains eccentricity and inclination*
Note *: this inclination is relative to the initial orbital plane
10/39
Jumping Jupiter model
90% of samples
have inclination of
more than 10 deg
inclination
eccentricity
produce large
eccentricities
semi-major axis
periastron
periastron distance
finally become
semi-major axis by
tidal evolution in
hot region
Marzari & Weidenschilling (2002)
11/39
Note: Tidal evolution

time scale for planetary orbit circularization

time scale for stellar spin/planetary orbit coplanarization
s: star, p: planet, adopting values for HD 209458b as a typical case
P: orbital/rotation period, k: tidal Love number,
Q: tidal quality factor (cf. 6×104 < QJup < 2×106)
Typically τcopl is much longer than τcirc
Mardling (2007), Winn et al. (2005)
12/39
Kozai mechanism

distant binary companion perturbs a planetary orbit

leads to “Kozai oscillation”

due to conservation of angular momentum

the planetary orbit oscillates high/low eccentricity/inclination

the planet migrates by tidal evolution
orbit 1: high eccentricity and inclination
orbit 2: low eccentricity and inclination (at least 40 deg)
star
binary orbital plane
companion
13/39
Kozai migration
eccentricity
periastron
inclination
Wu & Murray (2003)
14/39
Differences in outcomes
1.
disk-planet interaction

negligible eccentricity and inclination

mainstream of migration theories

but cannot explain eccentric planets
2.
planet-planet interaction

possible large eccentricity and inclination

subsequent tidal evolution damps eccentricity

would explain distribution of eccentric planets
3.
planet-binary companion interaction

large eccentricity and inclination
15/39
Motivation

How can we test these theories by observations?

eccentricity and inclination are possible clues

but eccentricity may be damped within planets’ age


Stellar spin axis would preserve initial orbital axis*


inclination (angle between initial and final orbital plane) would be a good
diagnostic
the inclination is equal to the stellar spin axis and the planetary orbital
axis (spin-orbit alignment)
But can we observe/constraint spin-orbit alignments of
exoplanetary systems?
Note *: Assumption
16/39
Transiting extrasolar planets
Planets pass in front of their host star.
Charbonneau et al. (2000)
periodic dimming in photometry
The first transiting planet HD 209458b was reported in 2000.
17/39
What can we learn from transiting planets?



Radial Velocity

semi-major axis a, minimum mass Mp sin i

Period P, eccentricity e
Transit Photometry

orbital inclination* iorb、radius ratio Rp/Rs

by combining spectroscopy: radius Rp, density ρ
Secondary Eclipse


thermal emission of planetary surface
Transmission Spectroscopy

search for atmospheric components

Na, H, C, O, H2O, SiO detections were reported in HD 209458b

(Subaru observations for HD 189733b tomorrow)
Note *: this inclination is relative to the sky plane
18/39
Radial Velocity during Transit
Transiting planet hides stellar rotation.
star
planet
hide approaching side
→ appear to be receding
planet
hide receding side
→ appear to be approaching
Radial velocity would have anomalous excursion
during transit.
19/39
The Rossiter-McLaughlin effect
This effect was originally reported in eclipsing binary systems.
β Lyrae： Rossiter 1924, ApJ, 60, 15
Algol: McLaughlin 1924, ApJ, 60, 22
20/39
RM effect in transiting exoplanetary system
ELODIE on 193cm telescope
Queloz et al. (2000)
The RM effect was detected in HD 209458b in 2000.
21/39
RV anomaly
What can we learn from the RM effect?
examples of trajectory
time
Ohta, Taruya & Suto (2005)
Radial velocity anomaly reflects planet’ trajectory.
22/39
Definition of λ
λ： sky-projected angle between
the stellar spin axis and the planetary orbital axis
23/39
Planetary trajectories and λ
We can measure λ by observations of the RM effect.
Gaudi & Winn (2007)
24/39
Summary of introduction and motivation
A)
There are several different planet migration theories.
B)
Each theory has different distributions of eccentricity and
inclination.
C)
We can observe the RM effect in transiting exoplanetary
systems.
D)
We can measure λ(sky-projected spin-orbit alignment) via the
RM effect.
E)
λ is an useful diagnostic for testing planet migration theories.
A
B
E
C
D
25/39
Our recent observations
Brief summary
Target : TrES-1 (V=11.8) → the faintest target so far
Observation : Simultaneous Subaru/MAGNUM observations
Challenge : the first RM observation for Subaru & MAGNUM
Result : succeeded in detection of the RM effect and placed a
constraint on λ
Significance 1: extended targets of the RM observations to
fainter systems
Significance 2: discovery of a possible misaligned system
26/39
Backgrounds of the RM observations
History of discoveries of target systems before 2005
HD 209458 : 2000, V=7.65
TrES-1
: 2004, V=11.8
HD 149026 : 2005, V=8.15
HD 189733 : 2005, V=7.67
…
The RM observations were conducted
for brighter targets with Keck/HIRES
HD 209458 : Winn et al. 2005
HD 189733 : Winn et al. 2006
(HD 149026 : Wolf et al. 200?)
27/39
Possible targets of the RM observations
Possible targets → Transiting systems brighter than V～12
(for which we can detect the RM effect with Subaru/HDS)
Our target : TrES-1, V=11.8
The first challenge for a fainter (V～12) target
(also the first RM observation for Subaru/HDS)
28/39
TrES-1

Discovered with 10cm telescope (Alonso et al. 2004)

V=11.8、K0V、V sin Is = 1.08 ± 0.30 km/s）

Poor radial velocity measurements due to its faintness.

The star has several spots.
Upper：TrES-1
Lower：HD 209458
※Charbonneau et al. (2007)
29/39
Simultaneous Subaru/MAGNUM observations
TrES-1 observations with 2 telescopes in Hawaii (UT 2006/6/21)
Photometry with
MAGNUM at Haleakala
Radial velocity
measurement with
Subaru/HDS
30/39
RV measurements with Subaru/HDS
Radial velocities obtained with
Subaru/HDS

20 samples

R : 45000

Exposure time : 15 min

Seeing : ～1.0 arcsec

S/N : ～ 60 (with iodine cell)

Radial velocity analysis by Sato
et al. (2002)

RV precision : 10 ～ 15 m/s
31/39
Photometry with MAGNUM

184 samples

Band : V

Exposure time : 40 or 60 sec

No spot event

Photo. precision : 2 mmag

Timing precision : ～30 sec
V band transit light curve obtained
with MAGNUM
32/39
RV model and parameters



incorporating published data

Keck 12 ( 7 + 5 ) RV samples

FLWO 1149 (3 transits) photometric samples
RM modeling with Ohta, Taruya, & Suto formula(2005)

Simultaneous fitting of radial velocity and photometry

including the RM effect
15 free parameters

K, VsinIs, λ : for radial velocity

iorb, uV, uz, Rs, Rp/Rs : for photometry

v1, v2, v3 : offsets for radial velocity datasets

Tc(234), Tc(235), Tc(236), Tc(238) ： time of transit center
33/39
Note: Constraints on VsinIs

External constraint on VsinIs for TrES-1


VsinIs = 1.08 ± 0.30 km/s (Laughlin et al. 2005)
Fitting with (a) / without (b) considering the constraint
(a)
(b)

χ2 minimization with AMOEBA (Numerical Recipes)
34/39
Results of RV fitting
-0.5
0
0
orbital phase
transit phase
0.05
a : with, b : without
35/39
Constraints on VsinIs and λ
(a) ： VsinIs = 1.3 ± 0.3 [km/s], λ= 30 ± 21 [deg]
(b) ： VsinIs = 2.5 ± 0.8 [km/s], λ= 48 ± 17 [deg]
Contours : ⊿χ2=1,00, ⊿χ2=2.30, ⊿χ2=4.00, ⊿χ2=6.17
36/39
Summary of Our Recent Results

We detected the RM effect in TrES-1 (V～12)

TrES-1 is the faintest target so far


We confirm that similar observations are possible for other
faint systems
We put a constraint on λ in TrES-1 for the first time

large uncertainty, but at least we confirmed that the planet
orbits in a prograde manner

possible misaligned (over 10 deg) system

additional RM observations would pin down λ

the first candidate of the jumping Jupiter model
37/39
What’s next?
New targets were discovered in 2006 & 2007

4 ground-based transit survey teams (XO, TrES, HAT, WASP) succeeded
in detecting new transiting systems

all transit survey teams target V less than ～12

also ESA’s satellite mission (CoRoT) started in 2007

2006 : XO-1, TrES-2, HAT-P-1, WASP-1, WASP-2

2007 : CoRoT-1, TrES-3, XO-2, XO-3, HAT-P-2, GJ 436

(recent news) : XO-4, TrES-4, HAT-P-3, HAT-P-4, more to come!
observational / statistical studies have become possible
38/39
Future Prospects

We can measure the RM effect of new transiting systems

By measuring the distribution of spin-orbit alignment,
we can test planet migration theories


already we have

possible misaligned target TrES-1 → further constraint on λ

at least 15 new targets
We can present observational / statistical distribution of spin-orbit
alignment within several years
39/39