Brown Dwarfs : Up Close and PhysicalIn 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

Luminosity History of Low Mass Objects

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.

Burrows et al

Minimum Stellar Temperature

History of Substellar Sizes

Burrows et al.

Core Temperature depends on Age and Mass

Lithium & Hydrogen burning limit

Deuterium burning limit

The Lithium Test

Basri 1997

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

Spectra of L Dwarfs

Geballe et al. 2002

The New Cool Spectral Types : T

DwarfsThe infrared spectrum shows methane in preference to carbon monoxide; the optical spectrum is dominated by resonance line wings of alkali metals. 1300K – 700K

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

Spectra of T Dwarfs

Geballe et al. 2002

Spectral Typing by IR Molecular Indices

Burgasser et al. 2002

Dust formation

The opacities and atmospheric chemistry in brown dwarfs becomes increasingly tied to the physics of condensates.

Tsuji 2002

Basri 1997

Allard 1999

Atmospheric Structure Changes

Marley et al. 2002

Tsuji 1300K

No condensation

condensation

Cloud Formation in Brown DwarfsThe 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.

Photometric Evidence of Rotation and Weather

4.5 hr

7.5 hr

Several days – not periodic:weather (dust clouds)?

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).

Gelino et al. 2002

The Weather Report for Brown Dwarfs

Clear

Cloudy

Dusty

Partly Cloudy

Burgasseret al. 2002

Effect of Clouds on Spectra

Models for T=500K, 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

Atmospheric Changes with Spectral Type

T

L/T

L

Y?

M

“Fine Analysis” vs Structural Effective Temperatures

Smith et al. 2003 ApJ

“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 don’t 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.

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?

Hfalls at the bottom of the

Main Sequence

Gizis et al. 2001

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?

Activity in L dwarfs is very minimal; almost none have detected H or X-rays.All objects cooler than about L3 are brown dwarfs (and significant fraction above).

Rotation in very cool objects

BrownDwarfs

Stellar andSubstellar Objects

VeryLow MassStars

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.

BasriMohanty2000,2003

Accretion and Activity in Young BDs:

H Strength vs Width

These are late M types (5.5-9.5)

Going to very early ages, activity is generally stronger (the objects are warmer), and some of them show accretion from disks. The width of H can be used as a direct accretion diagnostic in high-resolution spectra. Jayawardhana, Mohanty, Basri 2003

Rotation vs H strength in Young BDs

Evidence for disk-locking?

Accretion line

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:1) Find an effective temperature from a spectroscopic

diagnostic that is largely temperature-dependent2) Find a surface gravity from a pressure-sensitive line3) Get the radius from the luminosity (which obtains from the

observed brightness, coupled with a known distance to the region) and derived temperature

4) Find the mass from the radius and surface gravity5) 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.

Sensitivities to Teff and log(g)

TiO is sensitive primarily to temperature.

NaI is sensitive to both temperature and gravity.

Mohanty et al. 2004

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 T=200K. There is only one set of parameters that works for both. This is further confirmed by checking the TiO region surrounding NaI.

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:1) Find an effective temperature from a spectroscopic

diagnostic that is largely temperature-dependent2) Find a surface gravity from a pressure-sensitive line3) Get the radius from the luminosity (which obtains from

the observed brightness, coupled with a known distance to the region) and derived temperature

4) Find the mass from the radius and surface gravity5) 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.

Mass-Luminosity Relation

We confirm that the lowest free-floating objects being found may be below the D-burning limit!

GG Tau Bb

GG Tau Ba

(!)

Radius and

Temperature

vs Mass

GG Bb

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 don’t form).

Model Gravities and Ages

Mohanty et al. 2004

GG Tau Ba

GG Tau Bb

Evolutionary Model UncertaintiesA somewhat arbitrary starting point is used (without accounting

for accretion effects): >30 jupiter start at D ignition; <30 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 (>20 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 D’Antona/Mazzitelli models.

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.

Thank You!