Brown Dwarfs : Up Close and Physical

1
Brown Dwarfs Up Close and Physical
In the mass range intermediate between stars and
planets are the substellar objects known as brown
dwarfs. The first BDs were discovered in 1995.
The first confirmations were based in one case on
interior properties (the lithium test), and in
the other case on external temperature (the
methane test). I will concentrate on what we
have learned about their physical properties. We
are only beginning to directly test masses and
evolutionary models, but are learning about
temperatures and atmospheric properties. I also
touch on rotation, magnetic and accretion
activity, in young and old BDs over their entire
mass range. I will NOT cover many topics,
including search techniques and results, the mass
function or space density, binarity, or formation
mechanisms.
Gl 229
2
Luminosity History of Low Mass Objects
Burrows et al
Minimum Stellar Temperature
You could view brown dwarfs as stars which only
have a deuterium main sequence (which is short).
Regular stars also have hydrogen and helium main
sequences, and massive stars have additional
burning phases for heavier elements.
3
History of Substellar Sizes

Burrows et al.
4
Core Temperature depends on Age and Mass
Lithium Hydrogen burning limit
Deuterium burning limit
5
The Lithium Test
Basri 1997
6
The New Cool Spectral Types L Dwarfs
L and T have been added to cover the changes
in spectra at very cool temperatures. The L
dwarfs are marked by a change from domination by
oxide to hydride molecular species. Refractory
metals condense out. This has big ramifications
in the optical.
2200K 1400K
Kirkpatrick et al. 1999
7
Spectra of L Dwarfs
Geballe et al. 2002
8
The New Cool Spectral Types T Dwarfs
The infrared spectrum shows methane in preference
to carbon monoxide the optical spectrum is
dominated by resonance line wings of alkali
metals. 1300K 700K
9
Alkalai Resonance Lines in the Optical
In very cool objects, the lines of sodium and
potassium dominate the optical opacity. This
yields a magenta color for brown dwarfs. The
first measurement on an extrasolar planet shows
sodium
Burrows
10
Spectra of T Dwarfs
Geballe et al. 2002
11
Spectral Typing by IR Molecular Indices
Burgasser et al. 2002
12
Dust formation
Allard 1999
The opacities and atmospheric chemistry in brown
dwarfs becomes increasingly tied to the physics
of condensates.
Tsuji 2002
Basri 1997
13
Atmospheric Structure Changes
Tsuji 1300K
condensation
No condensation
Marley et al. 2002
14
Cloud Formation in Brown Dwarfs
The formation of clouds is poorly understood (not
that great even here on Earth). Particle size
distributions, saturation regimes, horizontal
inhomogeneities, global and turbulent currents
are all crucial to how optically thick the clouds
will be, what their height of formation,
thickness, and covering fraction is, and knowing
when precipitation will occur. Observed spectra
may reflect a blend of different heights and
compositions.
15
Photometric Evidence of Rotation and Weather
The vsini of BDs implies that the rotation
periods should be hours. Direct evidence for this
has been found. Some vary on longer timescales
this could be due to condensation features
(clouds).
4.5 hr
Several days not periodic weather (dust
clouds)?
Gelino et al. 2002
7.5 hr
16
The Weather Report for Brown Dwarfs
Burgasser et al. 2002
Dusty
Clear
Partly Cloudy
Cloudy
17
Effect of Clouds on Spectra
Models for T500K, 1000K, 1500K. Flat spectra
result if dust clouds are optically thick
spectral features are for clear atmospheres.
Marley et al. 2002
There is evidence for cloud formation and then
clearing in the behavior of FeH near the L/T
transition. The FeH should disappear as liquid
iron droplets form, but it reappears even as the
temperature cools further, likely due to breaks
in the clouds that expose hotter interior
regions. Burgasser et al. 2002
18
Atmospheric Changes with Spectral Type
L
M
L/T
T
Y?
19
Fine Analysis vs Structural Effective
Temperatures
Structural temperatures are defined by
measuring the luminosity (from photometry
adjusted with a bolometric correction), combined
with the parallax, then using theory to define a
relation between bolometric luminosity and
radius. High resolution spectroscopy yields
results that dont quite seem to agree with
theoretical models (problems may be bolometric
corrections, atmospheric models). The cooler
objects are inferred to be smaller by
spectroscopy than in the models.
Smith et al. 2003 ApJ
20
Rotation and Magnetic Activity on Brown Dwarfs
Solar-type stars form with a variety of
rotations, perhaps due to disk-locking. They
initially show signs of accretion and outflow.
They are active in their youth because of
relatively rapid spin. The fields carry off
angular momentum, and the stars spin down and
become less active. Does this story apply all the
way down the main sequence into brown dwarfs?
Does this story even extend below the brown dwarf
mass limit?
21
Ha falls at the bottom of the Main Sequence
There is a dramatic fall-off in activity at the
cool end of the M spectral type. Is this due to
rotation, or something else?
Gizis et al. 2001
Activity in L dwarfs is very minimal almost none
have detected Ha or X-rays. All objects cooler
than about L3 are brown dwarfs (and significant
fraction above).
22
Rotation in very cool objects
Basri Mohanty 2000,2003
Stellar and Substellar Objects
Very Low Mass Stars
Brown Dwarfs
The decrease in activity is clearly NOT due to
slow rotation! Rather, the increase in spindown
times is due to temperature. The atmospheres are
becoming very neutral, and cannot couple to the
magnetic field to remove angular momentum.
23
Accretion and Activity in Young BDsHa Strength
vs Width
Going to very early ages, activity is generally
stronger (the objects are warmer), and some of
them show accretion from disks. The width of Ha
can be used as a direct accretion diagnostic in
high-resolution spectra.
Jayawardhana, Mohanty, Basri 2003
These are late M types (5.5-9.5)
24
Rotation vs Ha strength in Young BDs
Evidence for disk-locking?
Accretion line
25
Deriving Fundamental Physical Parameters

• For objects in a star-forming region, one might
  hope to get their fundamental stellar parameters
  (testing the untested evolutionary calculations
  for low masses and young ages).
• The procedure is
• Find an effective temperature from a
  spectroscopic diagnostic that is largely
  temperature-dependent
• Find a surface gravity from a pressure-sensitive
  line
• Get the radius from the luminosity (which obtains
  from the observed brightness, coupled with a
  known distance to the region) and derived
  temperature
• Find the mass from the radius and surface gravity
• Assume all objects are coeval and check with
  isochrones

Note there have been no fundamental mass
determinations for visible substellar objects,
nor has there been confirmation of the claims
that some of these are below the fusion boundary.
26
Sensitivities to Teff and log(g)
TiO is sensitive primarily to temperature. NaI
is sensitive to both temperature and gravity.
Mohanty et al. 2004
27
Breaking the Degeneracy
One can get good fits for different combinations
of T and g in both TiO and NaI. For NaI an
increase of log(g)0.5 can be offset by an
increase of T200K. There is only one set of
parameters that works for both. This is further
confirmed by checking the TiO region surrounding
NaI.
28
Deriving Fundamental Physical Parameters

• For objects in a star-forming region, one might
  hope to get their fundamental stellar parameters
  (testing the untested evolutionary calculations
  for low masses and young ages).
• The procedure is
• Find an effective temperature from a
  spectroscopic diagnostic that is largely
  temperature-dependent
• Find a surface gravity from a pressure-sensitive
  line
• Get the radius from the luminosity (which obtains
  from the observed brightness, coupled with a
  known distance to the region) and derived
  temperature
• Find the mass from the radius and surface gravity
• Assume all objects are coeval and check with
  isochrones

Note there have been no fundamental mass
determinations for visible substellar objects,
nor has there been confirmation of the claims
that some of these are below the fusion boundary.
29
Mass-Luminosity Relation
GG Tau Ba
(!)
GG Tau Bb
We confirm that the lowest free-floating objects
being found may be below the D-burning limit!
30
Radius and Temperature vs Mass
GG Ba
GG Bb
Once again we find a problem between temperatures
found by high resolution spectroscopy and models.
The slope of the M-T relation is wrong, and the
radii of very low-mass objects are too small in
the evolutionary models. It is amazing that the
model spectra can fit so well in detail if the
model atmospheres are wrong (and clouds dont
form).
GG Bb
31
Model Gravities and Ages
GG Tau Ba
GG Tau Bb
Mohanty et al. 2004
32
Evolutionary Model Uncertainties
A somewhat arbitrary starting point is used
(without accounting for accretion effects) gt30
jupiter start at D ignition lt30 jupiter start
at log(g)3.5 (higher than what we
measure). While D burning is occurring, the
collapse of the object is slowed, so this can
cause objects to remain at lower gravity and
larger radius. These initial conditions will
cause very low mass objects to appear too young
for the first 1.5 Myr. This problem should be
gone, however, by the age of Upper Sco If D
burning really starts at lower gravity (3.25) for
the lowest mass objects, they take a very long
time to complete it (gt20 Myr), so they could be
hung up in the state we find them (while 30
jupiter objects would be done by 5 Myr). The
gravity at which D-burning starts has decreased
by 40 in the last 10 years in the
DAntona/Mazzitelli models.
33
Conclusions

• We have learned a lot about substellar objects
  in 8 years!
• We have seen a large range of masses,
  temperatures, and ages for substellar objects,
  down to the substellar mass limit.
• Model atmospheres do amazingly well at
  reproducing spectra, but there is still cause for
  refinements (especially in the infrared).
• Dust and cloud formation, along with
  precipitation and meteorology, are key to
  understanding the appearance of some objects, but
  are very complicated and much work remains in
  this area.
• The magnetic and angular momentum history of
  these objects is very different from all but the
  lowest mass stars.
• Evolutionary models have many good features, but
  we cannot consider them well-tested yet.

34
Thank You!