Transcript debris - Harvard-Smithsonian Center for Astrophysics

Debris Disks around Nearby Stars
David J. Wilner (Harvard-Smithsonian CfA)
• What are Debris Disks?
dust requires replenishment
• Interest in Resolved Morphologies
holes, blobs as planet signatures
• Imaging Examples:
Vega, e Eridani, Fomalhaut, ...
• Future Prospects
Spitzer Space Telescope, SMA, ALMA
collaborators:
M. Holman (CfA), C.D. Dowell (Caltech), M. Kuchner (Princeton)
SUNY Stony Brook, April 14, 2004
Introduction to Debris Disks
• Vega infrared excess discovered
serendipitously during IRAS
calibration (Aumann et al. 1984)
thermal emission from cold dust
optical
• Orbiting dust particles subject to
gravity, wind/radiation pressure (ejection)
and Poynting-Roberston drag (inspiral to star)
• tP-R = (400/b)(Mo/M*)(r/AU)2 yr << stellar age (~350 Myr)
dust particles must be replenished
• other nearby Vega-excess stars found by IRAS include
b Pic, Fomalhaut, e Eri (the “Fantastic Four”)
far-ir
• b Pic disk geometry confirmed by
images of visible scattered light
(Smith & Terrile 1984)
• ISO 60 mm survey finds 14/84 nearby
main-sequence stars (17%) with
excess emission (Habing et al. 2001)
• debris disks are cool (T<100 K), Kuiper Belt size (R>50 AU)
tenuous (L/L* ~ 10-5 to 10-2, M ~ Mmoon), gas poor
• various analyses of IRAS and ISO databases show:
- 100+ candidates resembling Fantastic Four in T, L/L*
- no strong dependence with stellar type (M, L*)
- dust may decline with age (gradually? abruptly?)
few x 100 Myr ~ Solar System heavy bombardment
Stages of Disk Evolution/Planet Formation
lFl
l
1. embedded protostar
104-105 yr
3. transition phase
~107 yr
2. HAe/Be Star
105-106 yr
4. debris disk
>> 107 yr
Malfait et al. 1998
Observational Probes of Disk Structure
scattered light
optical/near-ir
mid-ir
emitted light
far-ir/submm
< spatial resolution <
< temperature dependence <
> contrast with star, dynamic range >
J band Coronagraph+ AO
850 mm: 14 arcsec beam
HST/NICMOS Scattered Light: Gaps and Rings
JCMT 850 mm SCUBA Images
• First moderate resolution
submm images (14 arcsec)
of Fantastic Four show disk
and ring morphologies,
also emission peaks offset
from stellar photospheres
• Submm emission hints
at sculpting by planets:
cleared interior cavities,
persistent dust features
Planet Detection Parameter Space
Kepler mission
What Creates the Dust Blobs?
• background galaxies: unlikely given the source counts
• dust generated in situ by collisions of large planetesimals:
would have to be recent (disperse in ~10 to 100 orbital
periods) and likely rare (massive enough to release Mmoon)
• dust directly associated with orbiting bodies,
e.g. remnants of circumplanetary disks?
• dust spiralling starward trapped in resonances with planet
(cf. zodiacal dust trapped by Earth, Dermott et al. 1994)
Plutinos are in 3:2 Mean Motion
Resonance with Neptune
CfA Minor Planet Center
Jewitt Kuiper Belt Page
Dust in our Solar System from Afar
(Liou & Zook 1999)
• numerical simulations
suggest Solar System
would be recognized to
harbor at least two planets:
Neptune, Jupiter
• note: Solar System dust
emission at 850 mm at
10 pc only ~ 1 mJy
(<< solar photosphere)
Face-on view of the brightness from a numerical simulation of the column density of 23 mm
dust particles from Liou & Zook (1999). The signatures of the planets are (1) deviation from a
monotonic radial brightness profile, (2) ring along Neptune orbit, (3) variation along ring, (4)
relative lack of particles within 10 AU
Trapping by a low M, low e Planet
• for Neptune, Earth:
first order resonances
substantial trapping
• example: 3:2
• each orbit has j=3
longitudes of libration
for trapped particle
(a) Several particle orbits with different s (longitudes of pericenter). (b) Libration centers of
the 3jl-2lo- term for two of these orbits. (c) Locus of all libration centers. (d) The density
wave follows the motion of the planet at the same angular frequency as the planet.
Structure in the Vega System
(Wilner et al., ApJ, 569, L115)
• Vega (a Lyrae): A0V main sequence star, d=7.76 pc
• system viewed nearly pole-on: vsini, reddening
• JCMT 850 mm SCUBA image
(Holland et al. 1998) shows:
- roughly circular boundary
- an offset emission peak
- asymmetry extended NE-SW
- central cavity around the star
• interferometry allows imaging
with factor > 10x higher angular
resolution; need high sensitivity
see Koerner et al. 2001 for OVRO study
IRAM PdBI Observations
• compact D config
baselines 15-80 m
• dry winter weather
• 4 tracks : tint = 23 h
• l=1.3 mm
+ 3.3 mm
simultaneously
• rms: ~0.3 mJy at 1.3 mm, ~0.1 mJy at 3.3 mm
Images of Vega at l=1.3 mm
2.8 x 2.1 arcsec
stellar photosphere
5.3x4.6 arcsec
and dust blobs
(low surface brightness)
Trapping by a high M, high e Planet
• presence of two peaks
• different separations
of peaks from star
• peaks not co-linear
with star
• patterns from different
principle resonances
occur at same longitude,
3:1, 4:1, 5:1, ...
Libration centers of the 3l -lo-o- term. (a) Several particle orbits with different e and . (b)
The libration centers of two of these orbits when the planet is at pericenter. (c) All the libration
centers. (d) Clumps formed by particles trapped in this term appear to rotate at half the
angular frequency of the planet.
Modeling the Millimeter Emission
(left) A representative numerical simulation of 1.3mm dust emission from orbital dynamics that
includes a Jupiter mass planet, radiation pressure, and P-R drag. The dust becomes
temporarily entrained in mean motion resonances associated with the planet, producing a twolobed structure. (right) Simulated observation of the numerical model, taking account the IRAM
PdBI response for the Vega observations, and the IRAM PdBI image after subtraction of the
stellar photosphere.
Vega Summary
Searching for Light from the Planet
(Metchev et al. 2002)
(left) Composite H band mosaic of Vega region obtained with PALAO. Eight point sources are
detected. (right) H band sensitivity of the deep images to faint objects as a function or radial
distance from Vega (analyzed for the east field). Solid points represent individual
measurements; the solid line delineates the azimuthal average. The area between the vertical
dotted lines indicates the locus of the inferred planet.
Searching for Light from the Planet
(Macintosh et al. 2003)
(left) Deep Keck NIRC2 K’ band image of Vega. All candidate companions are in this field. The
dashed circle indicates a radius of 15 arcsec. (right) 5s sensitivity of the image. The dashed
lines indicate the planet masses from the models of Burrows et al. (1997).
Is Vega like the Early Solar System?
• Thommes et al. (1999)
+Thommes et al. (2002)
suggested Neptune
was scattered into a
highly eccentric orbit
• Malhotra (1995)
suggested Neptune
migrated (outwards) by
7 to 8 AU in ~10 Myr;
see Wyatt (2003) for
Vega model varient
aphelion
Neptune?
perihelion
Uranus?
Saturn?
Jupiter?
The temporal evolution of one of Thommes et al.’s simulations of an unstable Jupiter-CoreCore-Saturn system. Shown are the semi-major axes (thick solid) as well as the instantaneous
perihelion and aphelion distances ofthe orbits.
The e Eri Debris Disk
• single K2V star, age 0.5-1.0 Gyr, d=3.22 pc (3rd closest
naked eye star), closest analog to young Solar System
• (controversial) ~ 1 MJup radial velocity/astrometric planet,
a=3.4 AU, e=0.6 (Hatzes et al. 2000)
• far-ir spectrum
fit by r~60 AU
ring-like disk
(Dent et al. 2000,
Li et al. 2003,
Sheret et al. 2003
Moran et al. 2004)
Structure in the e Eri System
• JCMT 850 mm SCUBA image shows nearly face-on ~60 AU
radius ring with azimuthal variations (Greaves et al. 1999)
• Structure due to a
planetary perturber?
Liou et al. 1999,
Ozernoy et al. 2000
Quillen & Thorndike 2002
Inner Peak?
Potential of Inner Dust Imaging
• interaction of
inspiralling dust
with eccentric
planet should
produce two dust
peaks, like the
Vega system
• follow motions of
dust peaks
to independently
characterize planet
350 mm Observations with SHARC II
• SHARC II: Caltech Submillimeter Observatory
facility camera, 12x32 filled array of ‘pop-up’ bolometers,
optimized for 350 mm (9 arcsec beam)
• Observations made in Jan 2003 commissioning run,
• >16 hours on e Eridani in excellent weather (t225<0.037),
image rms ~3 mJy
Image
Imageof
ofeeEri
Eriat
atl=350
l=350mm
mm
350 mm Imaging Results
• confirm basic ~ 60 AU ring structure
• no evidence for central rise in flux density
corresponding to inner “zodiacal” component;
central clearing bolsters planet scenario
• clumpy structure of ring resolved into two (nearly)
symmetric arcuate features, brightest se and nw
• clumps outside ring consistent with background of
high redshift galaxies >10 mJy ~1 arcmin-2
(Smail et al. 2002)
Signpost of Planet Formation
• bright (L ~ 10-4 L*)
narrow (Da/a ~ 0.1)
ring of observed size
explained by
collisional cascade
in planetesimal disk
stirred by recent
formation of bodies
of radius >1000 km
• does not account for
azimuthal variations
(Kenyon & Bromley 2002, 2004)
Sculpting by a Planet?
• characterization of possible unseen planets requires
matching robust features using numerical simulations
0.2 MJup
e=0
2:1, 3:2
w/ high
libration
< 0.3 MJup
e ~ 0.3
5:3, 3:2
w/ phase
segregation
Ozernoy et al. 2000
Quillen & Thorndike 2002
• models that selectively populate particular resonances
are not realistic unless additional factors invoked, e.g.
parent bodies trapped by planet migration, encounter
Comments on Models
• structure depends on many parameters, e.g. planet
mass, eccentricity, semi-major axis, orbital phase,
inclination, dust properties, orbits of parent bodies
• prominent “two-blob” morphology, like Vega,
Jupiter mass planet in eccentric orbit traps dust in
exterior principal mean motion resonances
• models have time dependence that can be tested by
synoptic observations with sufficient sensitivity and
angular resolution (submm interferometry)
The Closest (< 4 pc) F,G,K Stars
Procyon
binary
61 Cyg
binary
Sun
<0.0001 ME
a Cen
multiple
e Eri
0.01 ME
t Ceti
0.0005 ME
e Ind
multiple
Future Prospects
• Spitzer: high sensitivity at far-infrared wavelengths not
accessible from the ground will provide exquisite SEDs
for a large sample and greatly improve statistics
Riecke GTO Projects,
FEPS Legacy Project
M. Meyer
First Results from Spitzer
Fomalhaut
Young Solar Analog Debris Disk
• HD107146, G2V, distance 28.5 pc, age ~100 Myr
• discovered during ground based support for Legacy
Project “Formation and Evolution of Planetary Systems”
undetected by
IRAS at 25 mm
(Williams et al. 2004)
• substantial population of cold disks, see Wyatt et al. (2003)
• Submillimeter Array: a collaborative project of the Smithsonian
Astrophysical Observatory and the Academia Sinica (Taiwan),
eight 6 meter diameter antennas on Mauna Kea for arcsecond
imaging initially for 1300, 850, 450 mm atmospheric windows
• submm interferometry is challenging
• official SMA dedication was November 22, 2003
• look for first call for external proposals in mid-2004
Early SMA Images
Mars Atmosphere CO(2-1) (Gurwell)
• ALMA: large array (64 x 12 m + 12 x 7 m)
North America, Europe, and likely Japan
high sensitivity, high resolution (10 mas)
full operation
in 2012?
• best possible site, Atacama at 5000 m, large bandwidth,
high fidelity imaging, active compensation for atmosphere
Debris Disks around Nearby Stars
• What are Debris Disks?
dust requires replenishment
• Resolved Morphologies
holes, blobs as planet
signatures
• Imaging Examples
dust structure plausibly due
to resonances with planet
• Future Prospects
Spitzer, SMA, ALMA
Dust can outshine Terrestrial Planets
Dust clumps in the zodiacal cloud from 10 pc: (a) Model of the brightest unresolved clump
from collisions in the asteroid belt; the horizontal line indicates the flux from an Earth, the
vertical line represents the beam size; (b) Model of the Earth’s resonant ring (Dermott et al.
1994) at 10 mm with a 0.06 arcsec beam. The Earth’s emission would be at [+0.1,0]
and would be 10 to 20 times brighter than the bright trailing clump in the ring (Wyatt 2001).
Spectral Energy Distributions of Excess
• ISO 25 mm survey of nearby
main-sequence stars shows
that warm disks are rare
(Laureijs et al. 2002).
• [25/60] impies T<120 K
• evacuated inner regions
are common features of
debris disk systems
Zodiacal Light
Clementine 1994
Fomalhaut: a Nearly Uniform Ring
(Holland et al. 2002)
• residuals reveal a “clump” with 5% of total flux
SCUBA 450 mm images of the Fomalhaut disk (from Holland et al. 2002): (a) observation, (b)
axisymmetric smooth disk model, and (c) the residuals, which show that the asymmetry could
be explained by a clump embedded in a smooth disk. All contours are spaced at 1s = 13
mJy/beam. The dashed white oval in (b) shows the inner edge of the mid-plane of the disk, a
125 AU radius ring inclined 20 degrees to the line of sight. The stellar photosphere has been
subtracted.
JCMT 850 mm SCUBA Images
100 AU
19.3 pc
7.7 pc
7.8 pc
3.2 pc