Cloudy with a Chance of Iron

1
Cloudy with a Chance of Iron
Clouds and Weather onBrown Dwarfs

• Adam Burgasser
• UCLA

2
Andy Ackerman Mark MarleyNASA Ames
J. Davy KirkpatrickCaltech/IPAC
Didier SaumonLos Alamos NL
Katharina LoddersWashington University

• Adam Burgasser
• UCLA

3
Summary (i.e., what Ill try to convince you of!)

• Cool brown dwarf atmospheres have the right
  conditions to form condensates or dust.
• Observations support the idea that these
  condensates form cloud structures.
• Cloud structures are probably not uniform, likely
  disrupted by atmospheric turbulence.
• Clouds have significant effects on the spectral
  energy distributions of these objects and
  analogues (e.g., Extra-solar giant planets).

4
What are Brown Dwarfs?
Failed stars objects that form like stars but
have insufficient mass to sustain H
fusion. Super-Jupiters objects with similar
size and atmospheric constituents as giant
planets, but form as stars.
5
Brown Dwarfs
Stellar evolution
(1)
(2)

• Adiabatic contraction (Hayashi tracks)
• Ignition, formation of radiative core, heating
  dynamic equilibrium (Henyey tracks)
• Settle onto Hydrogen main sequence radiative
  equilibrium

(3)
Hayashi (1965)
6
Brown Dwarfs
PPI chain p p ? d e ?e, Tc 3 ? 106 K
Below 0.1 M?, e- degeneracy becomes significant
in interior (Pcore 105 Mbar, Tcore TFermi)
and will inhibit collapse. Below 0.075 M?,
Tcore remains below critical PPI temperature ?
Cannot sustain core H fusion.
Kumar (1963)
7
Brown Dwarfs
With no fusion source, Brown dwarfs rapidly
evolve to lower Teff and lower luminosities.
cool off inexorably like dying embers plucked
from a fire. A. Burrows
8
Some Brown Dwarf Properties

• Interior conditions ?core 10-1000 g/cm3, Tcore
  104-106 K, Pcore 105 Mbar, fully convective,
  largely degenerate (90 of volume),
  predominantly metallic H (exotic?).
• Atmosphere conditions Pphot 1-10 bar, Tphot
  3000 K and lower.
• All evolved brown dwarfs have R 1 RJupiter.
• Age/Mass degeneracy old, massive BDs have same
  Teff, L as young, low-mass BDs.
• Below Teff 1800 K, all objects are substellar.
• NBD N, MBD 0.15 M

9
Why Brown Dwarfs Matter

• Former dark matter candidates - no longer the
  case.
• Important and populous members of the Solar
  Neighborhood.
• End case of star formation, test of formation
  scenarios at/below MJeans.
• Tracers of star formation history and chemical
  evolution in the Galaxy.
• Analogues to Extra-solar Giant Planets (EGPs),
  more easily studied.
• Last source of stars in distant future of
  non-collapsing Universe - Adams Laughlin (RvMP,
  69, 337, 1997).

10
M, L, and T dwarfs
Three spectral classes encompass Brown Dwarfs
M dwarfs (3800-2100 K) Young BDs and low-mass
stars. L dwarfs (2100-1300 K) BDs and very
low-mass, old stars. T dwarfs (lt 1300 K)
All BDs coolest objects known.
11
M, L, and T dwarfs
M dwarfs are dominated by TiO, VO, H2O, CO
absorption plus metal/alkali lines. L dwarfs
replace oxides with hydrides (FeH, CrH, MgH, CaH)
and alkalis are prominent. T dwarfs exhibit
strong CH4 and H2O and extremely broadened Na I
and K I.
12
Condensation in BD Atmospheres

• At the atmospheric temperatures and pressures of
  late-M and L dwarfs, many gaseous species are
  capable of forming condensates.
• e.g.
• TiO ? TiO2(s), CaTiO3(s)
• VO ? VO(s)
• Fe ? Fe(l)
• SiO ? SiO2(s), MgSiO3(s)

Marley et al. (2002)
13
Evidence for Condensation - Spectroscopy

• Relatively weak H2O bands in NIR compared to
  models require additional smooth opacity source.
• The disappearance of TiO and VO from late-M to L
  can be directly attributed to their accumulation
  onto condensate species.

Kirkpatrick et al. (1999)
14
Evidence for Condensation - Photometry
The NIR colors of late-type M and L dwarfs are
progressively redder can only be matched by
models that allow dust formation in their
atmospheres. However, bluer colors of T dwarfs
require a transparent atmosphere dust must be
removed.
Dusty
Gliese 229B
Cond
Chabrier et al. (2000)
15
Evidence for Rainout - Abundances
L
T
Without the rainout of dust species, Na and K
would form Feldspars and atomic species would be
depleted in the late L dwarfs.
Burrows et al. (2002)
16
Evidence for Rainout - Abundances
L
T
With rainout, Na and K persist well into the T
dwarf regime.
Burrows et al. (2002)
17
Evidence for Rainout - Abundances
K I (and Na I) absorption is clearly present in
the T dwarfs ? dust species must be removed from
photosphere.
Burgasser et al. (2002)
18
Cloudy Models for BD Atmospheres

• Condensate clouds dominate visual appearance and
  spectrum of every Solar giant planet likely
  important for brown dwarfs.
• Condensates in planetary atmospheres are
  generally found in cloud structures.
• Requires self-consistent treatment of condensable
  particle formation, growth, and sedimentation.
• Ackerman Marley (2001) Marley et al. (2002)
  Tsuji (2002) Cooper et al. (2003) Helling et
  al. (2001) Woitke Helling (2003)

19
Basics of the Cloudy Model

• Simple treatment assume transport of dust by
  diffusion and gravitational settling.
• Horizontal homogeneity.
• No chemical mixing between clouds.

20
What is frain?

• If L, qc/qt constant, scale height
• frain 0 ? dusty atmosphere.
• frain ? 8 ? clear atmosphere.
• Earth frain 0.5 (stratocumulus) 4 (cumulus).
• Jupiter frain 1-3 (NH3 clouds).

qt(z) q0 exp(- frain qc/qt w/? z)
21
What is frain?
frain determines extent of cloud, particle size
distribution, and hence cloud opacity.
Ackerman Marley (2001)
22
Basics of the Cloudy Model
The cloud layer is generally confined to a narrow
range of temperatures ? for cooler BDs, cloud
will reside below the photosphere.
Ackerman Marley (2001)
23
Basics of the Cloudy Model
L5
Condensate cloud may or may not influence
spectrum of brown dwarf depending on its
temperature explains disappearance of dust in T
dwarfs.
L8
T5
Ackerman Marley (2001)
24
Cloudy Model Results

• Accurately predicts M/L dwarf colors down to
  latest-type L dwarfs.
• Matches turnover in near-infrared colors in T
  dwarfs.
• Cannot explain J-band brightening across L/T
  transition.

dusty
clear
cloudy, frain 3
Burgasser et al. (2002)
25
The Transition L ? T

• Dramatic shift in NIR color (?J-K 2).
• Dramatic change in spectral morphology.
• Loss of condensates from the photosphere.
• Objects brighten at 1 mm.
• Apparently narrow temperature range Gl
  584C (L8) 1300 K
  2MASS 0559 (T5) 1200 K.

26
CondensateClouds
Clouds are not uniform!
27
At 5 ?m, holes in Jupiters NH3 clouds produce
Hot Spots that dominate emergent flux ?
horizontal structure important!
IRTF NSFCam 1995 July 26 c.f., Westphal,
Matthews, Terrile (1974)
28
Evidence for Cloud Disruption - Theory
2D models of dust formation in BD atmospheres
predict patchiness due to turbulence and rapid
accumulation of condensate material.
Number density
Mean particle size
Helling et al. (2001)
29
Evidence for Cloud Disruption -
Variability
Many late-type L and T dwarfs are variable, P
hours, similar to dust formation
rate. Atmospheres too cold to maintain magnetic
spots ? clouds likely. Periods are not generally
stable ? rapid surface evolution.
Enoch, Brown, Burgasser (2003)
30
Evidence for Cloud Disruption -
Spectroscopy
Strengthening of K I higher-order lines around
1?m ? reduced opacity at these wavelengths from
late L to T.
Burgasser et al. (2002)
31
Evidence for Cloud Disruption -
Spectroscopy
Reappearance of condensate species progenitors
(e.g., FeH) ? detected below cloud deck.
Burgasser et al. (2002)
32
Evidence for Cloud Disruption -
Spectroscopy
Presence of CO in Gliese 229Bs atmosphere
16,000x LTE abundance ? upwelling convective
motion.
Oppenheimer et al. (1998)
33
A Partly Cloudy Model for BD Atmospheres

• An exploratory model.
• Linear interpolation of fluxes and P/T profiles
  of cloudy and clear atmospheric models.
• New parameter is cloud coverage percentage
  (0-100).
• Burgasser et al. (2002), ApJ, 571, L151

34
Wavelength Matters!
z
J
K
I
1400 K
FeH
K I
Relative brightening at z and J (1 ?m) can be
explained by holes in the clouds.
35
Success?
Cloud disruption allows transition to brighter T
dwarfs. Requires very rapid rainout at L/T
transition, around 1200 K. Data fits, model is
physically motivated, but is it a unique solution?
Burgasser et al. (2002)
36
Arguments Against the Model

• Small numbers of objects with parallaxes, could
  be a statistical fluke.
• Recent parallaxes for 10-20 late-L/early-T show
  identical trends brightening is real.
• Early T dwarfs could be young, late L dwarfs old.
• Fairly tight trend, some T dwarf companions are
  known to be old, some late L dwarf companions
  known to be young.
• May indicate different sedimentation efficiencies
  in different objects.
• Fit for L dwarfs is excellent for frain 3,
  would require a rapid shift in atmospheric
  dynamics partial clouding is simpler.

37
Extrasolar Planet Weather?

• 3D Hydrodynamic models of hot EGP atmospheres
  produce vertical winds/structure.
• Weak Na I in HD 209458b high clouds?
• Presence of clouds affects detectability of EGPs.

Showman Guillot (2002)
Charbonneau et al. (2002)
38
More Work is Needed!!

• More data across L/T transition needed new
  discoveries (SDSS, 2MASS), distance measurements
  (USNO), better photometry.
• Development of a fully self-consistent model
  convective motions, cloud disruption can be
  drawn from terrestrial/Jovian studies.
• What are the cloud structures - Bands? Spots?
• How do rotation, composition, age influence
  transition?

39
(No Transcript)