Transcript New brown dwarfs and giant planets

Brown Dwarfs and Dark Matters
L dwarfs, binaries and the mass function
Neill Reid, STScI
in association with 2MASS Core project:
Davy Kirkpatrick, Jim Liebert, Conard Dahn, Dave Monet,
Adam Burgasser
Outline
• Finding ultracool dwarfs
• The L dwarf sequence
extending calibration to near-infrared wavelengths
• L-dwarf binaries
Separations and mass ratios
• The mass function below the hydrogen-burning limit
brown dwarfs and dark matter
Some results and a conundrum
• Heavy halo white dwarfs?
Cool dwarf evolution (1)
Low-mass stars:
H fusion establishes
equilibrium configuration
Brown dwarfs:
no long-term energy supply
T ~ 2 million K required for
lithium fusion
Lithium test
Late-type dwarfs are fully convective
 everything visits the core
If core temperature > 2 x 10^6 K
 lithium is destroyed
If M < 0.06 M(sun), lithium survives
Cool dwarf evolution (2)
Rapid luminosity evolution
for substellar-mass dwarfs
Cool dwarf evolution (3)
Brown dwarfs evolve through
spectral types M, L and T
L dwarfs encompass stars
and brown dwarfs
Cooling rate decreases with
increasing mass
Finding ultracool dwarfs
Gl 406 = M6 dwarf
(Wolf 359)
Flux distribution peaks
at ~ 1 micron
---> search at near-IR
wavelengths
Finding ultracool dwarfs (2):
Near-infrared sky surveys
1969 - Neugebauer & Leyton - Mt. Wilson TMSS
custom built 60-inch plastic mirror
arc-minute resolution, K < 3rd magnitude
1996 - 2000 DENIS … southern sky
ESO 1.3 metre, IJK to J~15, K~13.5
1997 - present 2MASS all-sky
Mt. Hopkins/CTIO 1.5 metres, JHK
J~16, K~14.5 (10-sigma)
Finding ultracool dwarfs (3)
Search for
sources
with red (J-K)
and either red
optical/IR colours
or A-type colours
Cool dwarf spectra (1)
Early-type M dwarfs
characterised by increasing
TiO absorption
CaOH present for sp > M4
Cool dwarf spectra (2)
Late M dwarfs:
increasing TiO
VO at sp > M7
FeH at sp > M8
Cool dwarf spectra (3)
Spectral class L:
decreasing TiO, VO
- dust depletion
increasing FeH, CrH,
water
lower opacities increasingly strong
alkali absorption
Na, K, Cs, Rb, Li
Cool dwarf spectra (4)
Low opacity leads to high
pressure broadening of
Na D lines
The L/T transition
Methane absorption
T ~ 1200/1300K
(Tsuji, 1964)
Blue JHK colours
Early-type T dwarfs
first identified from
SDSS data Leggett et al (2000)
Unsaturated methane
absorption
NIR Spectral Classification (1)
Kirkpatrick scheme defined
at far-red wavelengths
Most of the flux is emitted
at Near-IR wavelengths
Is the NIR behaviour
consistent?
K, Fe, Na atomic lines
water, CO, methane bands
NIR Spectral classification (2)
J-band: 1 - 1.35 microns
Numerous atomic lines
Na, K, Fe
FeH bands
UKIRT CGS4 spectra:
Leggett et al (2001)
Reid et al (2001)
NIR Spectral Classification (3)
H-band
Few identified
atomic features
NIR Spectral Classification(4)
K-band
Na I at 2.2 microns
CO overtone bands
molecular H_2
(Tokunaga &Kobayashi)
--> H2O proves well
correlated with optical
spectral type
--> with temperature
The HR diagram
Broad Na D lines lead
to increasing (V-I) at
spectral types later than
L3.5/L4
Latest dwarf 2M1507-1627 L5
Astrometry/photometry
courtesy of USNO
(Dahn et al)
The near-infrared HR diagram
Mid- and late-type
L dwarfs can be selected
using 2MASS JHK alone
SDSS riz + 2MASS J
permits identification of
all dwarfs sp > M4
Note small offset
L8  Gl 229B
Searching for brown dwarf binaries
The alternative model
for brown dwarfs
Binary surveys: L dwarfs (2)
Why do we care about L dwarf binaries?
1. Measure dynamical masses  constrain models
2. Star formation and, perhaps, planet formation
HST imaging survey of 160 ultracool dwarfs (>M8) over
cycles 8 & 9 (Reid + 2MASS/SDSS consortium)
Successful WFPC2 observations of 60 targets to date
--> only 11 binaries detected
Binary surveys: L dwarfs (3)
2M0746 (L0.5)
2M1146 (L3)
Binary systems: L dwarfs (4)
2M0920 (L6.5):
I-band
V-band
Binary systems: L dwarfs (5)
2M0850:
I-band
V-band
Binary surveys: L dwarfs (6)
Binary components lie close
to L dwarf sequence:
2M0850B
M(I) ~0.7 mag fainter
than type L8
M(J) ~0.3 mag brighter
than Gl 229B (1000K)
--> dM(bol) ~ 1 mag
similar diameters -->
dT ~ 25% --->
T(L8) ~ 1250K
2M0850A/B
Could 2M0850AB
be an L/T binary?
Probably not -but cf. SDSS
early T dwarfs
L dwarf binary statistics (1)
Approximately 20% of L dwarfs are resolved
• almost all are equal luminosity, therefore equal mass
2M0850AB – mass ratio ~ 0.8
• none have separations > 10 AU
 L dwarf/L dwarf binaries seem to be rarer, and/or have
smaller <a> than M dwarfs
How do these parameters mesh with overall binary statistics?
L dwarf binary statistics (2)
Brown dwarfs
don’t always
have brown
dwarf
companions
L dwarf binary statistics (3)
Known L dwarf binaries
- high q, small <a>
- low q, large <a>
-> lower binding energy
- preferential disruption?
Wide binaries as minimal
moving groups?
The substellar mass function (1)
Brown dwarfs evolve along nearly identical tracks in
the HR diagram, at mass-dependent rates
No single-valued
M/L relation
Model N(mag, sp. Type)  infer underlying Y(M)
Require temperature scale
bolometric corrections
star formation history
The substellar mass function (2)
Major uncertainties:
1. Temperature scale M/L transition --> 2200 to 2000 K
L/T transition --> 1350 to 1200 K
2. Stellar birthrate --> assume constant on average
3. Bolometric corrections:
even with CGS4 data, few cool dwarfs have
observations longward of 3 microns
4. Stellar/brown dwarf models
Bolometric corrections
Given near-IR data
--> infer M(bol)
--> bol correction
little variation in
BC_J from M6 to T
The substellar mass function (3)
Stellar mass function:
dN/dM ~ M^-1
(Salpeter n=2.35)
Extrapolate using
n= 0, 1, 2 powerlaw
Miller-Scalo functions
The substellar mass function (4)
Observational constraints:
from photometric field surveys for ultracool dwarfs
- 2MASS, SDSS
L dwarfs: 17 L dwarfs L0 to L8 within 370 sq deg, J<16 (2MASS)
--> 1900 all sky
T dwarfs: 10 in 5000 sq deg, J < 16 (2MASS)
2 in 400 sq deg, z < 19 (SDSS)
--> 80 to 200 all sky
Predictions: assume L/T transition at 1250 K, M/L at 2000 K
n=1 700 L dwarfs, 100 T dwarfs all sky to J=16
n=2 4600 L dwarfs, 800 T dwarfs all sky to J=16
Substellar Mass function (6)
Predictions vs. observations
10 Gyr-old disk
constant star formation
0<n<2
All L: 14002100 K
>L2 : 14001900K
T : < 1300K
Substellar mass function (7)
Change the age of the
Galactic disk
Younger age --->
larger fraction formed
in last 2 gyrs -->
Flatter power-law
(smaller n)
Substellar Mass Function (8)
Miller-Scalo mass function
--> log-normal
Match observations for
disk age 8 to 10 Gyrs
The substellar mass function (9)
Caveats:
1. Completeness … 2MASS - early L dwarfs
- T dwarfs (JHK)
SDSS - T dwarfs (iz)
2. Temperature limits … M/L transition
3. Age distribution
we only detect young brown dwarfs
In general  observations appear consistent with n ~ 1
 equal numbers of BDs (>0.01 M(sun)) and MS stars
No significant contribution to dark matter……..but….
A kinematic conundrum (1)
Stellar kinematics are correlated with age
 scattering through encounters with molecular clouds
leads to
1. Higher velocity dispersions
2. Lower net rotational velocity, V
e.g. Velocity distributions of dM (inactive, older)
and dMe (active, younger)
A kinematic conundrum (2)
Stellar kinematics are usually modelled as Gaussian distributions
 (s(U), s(V), s(W) )
But disk kinematics are more complex:
 use probability plots
Composite in V
2 Gaussian components in (U, W)
local number ratio high:low ~ 1:10
thick disk and old disk?
A kinematic conundrum (3)
Kinematics of ultracool dwarfs (M7  L0)
Hires data for 35 dwarfs
~50% trig/50% photo parallaxes
Proper motions for all
 (U, V, W) velocities
We expect the sample to be dominated by long-lived
low-mass stars – although there is at least one BD
A kinematic conundrum (4)
Ultracool M dwarfs have kinematic properties
matching M0-M5 dMe dwarfs
 t ~ 2-3 Gyrs
Does this make sense?
M7
 L0
~2600  2100K
Where are the old
V LM stars?
A different kind of dark matter
• Galaxy rotation curves at large radii are not Keplerian
- heavy halos (Ostriker, Peebles & Yahil, 1974)
- Milky Way M ~ 5 x 10^11 solar masses, R < 50 kpc
visible material (disk + stellar halo) ~ 5 x 10^10 solar masses
=> 90% dark matter – particles? compact objects?
• Microlensing surveys – MACHO, EROS, DUO,OGLE
Given timescale, estimated velocity => mass
MACHO: 13-17 events, t ~ 34-230 days, <V> ~ 200-300 km/s
=> can account for ~20% of the missing 90%
<M> = 0.5+/- 0.3 solar masses
Halo white dwarfs?
Heavy halo white dwarfs? I
• We are in the dark halo – local density ~ 10^-2 M_sun/pc^3
=> search for local representatives in proper motion surveys
• Oppenheimer et al. (Science Express, March 23)
Photographic survey of ~12% of the sky near the SGP
- 38 cool, high-velocity white dwarfs – 4 x 10^-4 stars/pc^3
- local mass density of ~3 x 10^-4 M_sun/pc^3
=> could account for 3% of dark matter
if they’re in the heavy halo
But are they?
Heavy halo white dwarfs? II
• The Galactic disk has a complex kinematic structure
- thin/old disk: 300 pc scaleheight,
90% of local stars
- thick/extended disk: 700 pc scaleheight, 10%
• Should we expect any high-velocity disk stars
 consider a volume-complete sample of 514 M dwarfs
(Reid, Hawley & Gizis, 1995)
Heavy halo white dwarfs? III
•
Thick disk stars can have high velocities
- Reid, Hawley & Gizis (1995): PMSU M dwarf survey
4% of the sample would be classed as dark halo by Oppenheimer et al
=> ~2 x 10^-4 white dwarfs / pc^3
• Most of the Oppenheimer et al. white dwarfs are remnants of the first
stars which formed in the thick disk
• White dwarfs from the stellar halo account for the rest
• There is no requirement for a dark matter contribution
What next?
(1)
Better statistics for nearby stars  F(M), Y(M)
A 2MASS NStars survey
(with Kelle Cruz (Upenn), Jim Liebert (UA), John Gizis (Delaware)
Davy Kirkpatrick & Pat Lowrance (IPAC), Adam Burgasser (UCLA))
Aim: find all dwarfs later than M4 within 20 parsecs
1. 2MASS/NLTT cross-referencing: (m(r) – K)  p
2. Deep van Biesbroeck survey for wide cpm companions
3. 2MASS-direct: (J-K)  p
4. 2MASS/POSS II: (I-J)  p
What next?
(2)
If n~1,
 equal numbers of stars
and brown dwarfs
 Numerous cool
(room temp.) BDs
brightest at 5 mm
 accessible to SIRTF
~10 400K BDs /100 sq deg
F>10 mJy at 5 mm
Summary
1. Brown dwarfs are now almost commonplace
2. Near-IR spectra show that the L dwarf sequence
L0…L8 is consistent with near-infrared variations
 probably well correlated with temperature
3. First results from HST L dwarf binary survey
- L dwarf/L dwarf binaries relatively rare
- Maximum separation is correlated with total mass
nature or nurture?
4. Current detection rates are inconsistent with a steep IMF
 brown dwarfs are poor dark matter candidates
4. Neither are cool white dwarfs
Shameless plug….
Now available from
Amazon.com and in
all the best bookstores
Binary surveys: T dwarfs
A digression:
chromospheric activity
is due to acoustic heating,
powered by magnetic field.
H-alpha emission traces
activity in late-type dwarfs.
Binary surveys: T dwarfs
H-alpha activity
declines sharply beyond
spectral type M7
Binary surveys: T dwarfs
..but 2M1237+68, a T dwarf,
has strong H-alpha emission
- no variation observed
July, 1999 - February, 2000
Possible mechanisms:
- Jovian aurorae?
- flares?
- binarity?
2M1237 : a vampire T dwarf
Brown dwarfs are degenerate - increasing R, decreasing M
- ensures continuous Roche lobe overflow
Brown dwarf atmospheres
Non-grey atmospheres
- flux peaks at 1, 5
and 10 microns
- bands and zones?
- “weather”?
Binary surveys: L dwarfs (1)
Several L dwarfs are wide companions of MS stars:
e.g. Gl 584C, G196-3B, GJ1001B (& Gl229B in the past).
What about L-dwarf/L-dwarf systems?
- initial results suggest a higher frequency
>30% for a > 3 AU (Koerner et al, 1999)
- all known systems have equal luminosity
--> implies equal mass
Are binary systems more common amongst L dwarfs?
or are these initial results a selection effects?
Clouds on an L8?
Gl 584C
- r ~ 17 pc
- 2 G dwarf companions
- a ~ 2000 AU
- age ~ 100 Myrs
- Mass ~ 0.045 M(sun)
- M(J) ~ 15.0
Gl 229B M(J) ~ 15.4