Andreas Schweitzer

THE LIMITING EFFECTS OF DUST IN BROWN DWARF MODEL ATMOSPHERES

We present opacity sampling model atmospheres, synthetic spectra, and colors for brown dwarfs and very low mass stars in the following two limiting cases of dust grain formation: (1) Inefficient gravitational settling (i.e., the dust is distributed according to the chemical equilibrium predictions) and (2) efficient gravitational settling (i.e., the dust forms and depletes refractory elements from the gas, but their opacity does not affect the thermal structure). The models include the formation of over 600 gas-phase species and 1000 liquids and crystals and the opacities of 30 different types of grains including corundum (Al2O3), the magnesium aluminum spinel MgAl2O4, iron, enstatite (MgSiO3), forsterite (Mg2SiO4), amorphous carbon, SiC, and a number of calcium silicates. The models extend from the beginning of the grain formation regime well into the condensation regime of water ice (Teff = 3000-100 K) and encompass the range of log g = 2.5-6.0 at solar metallicity. We find that silicate dust grains can form abundantly in the outer atmospheric layers of red and brown dwarfs with a spectral type later than M8. The greenhouse effects of dust opacities provide a natural explanation for the peculiarly red spectroscopic distribution of the latest M dwarfs and young brown dwarfs. The grainless (cond) models, on the other hand, correspond closely to methane brown dwarfs such as Gliese 229B. We also discover that the λλ5891, 5897 Na I D and λλ7687, 7701 K I resonance doublets play a critical role in T dwarfs, in which their red wings define the pseudocontinuum from the I to the Z bandpass.

Spherically symmetric model atmospheres for low mass pre-main sequence stars with effective temperatures between 2000 and 6800 k

We present a grid of spherically symmetric model atmospheres for young pre-MS stars. This grid spans the parameter range 2000 K ≤ Teff ≤ 6800 K and 2.0 ≤ log g ≤ 3.5 for M = 0.1 M☉, appropriate for low-mass stars and brown dwarfs. A major improvement is the replacement of TiO and H2O line lists with the newer line list, calculated by the NASA-Ames group, for TiO (about 175 million lines of five isotopes) and for H2O (about 350 million lines in two isotopes). We provide the model structures, spectra, and broadband colors in standard filters in electronic form.

Analysis of Keck HIRES Spectra of Early L-Type Dwarfs

We present analyses of high- and medium-resolution spectra of early L dwarfs. We have used our latest set of model atmospheres to reproduce and analyze the observed features. We can model the optical flux and atomic line profiles with the best accuracy to date. The models used to reproduce the observations include dust condensation and dust opacities. Compared with previous studies using older models, we find that our dust treatment is much improved. The derived parameters for the objects are well in the expected range for old, very low mass objects. This is also supported by the absence of Li in most of the objects. For the objects showing Li, we can be almost certain they are brown dwarfs. However, a spectral analysis in general, and this one in particular, can only very roughly determine mass and age.

Timescales of Disk Evolution and Planet Formation: HST, Adaptive Optics, and ISO Observations of Weak-Line and Post-T Tauri Stars*

We present high spatial resolution HST and ground-based adaptive optics observations and high-sensitivity ISO (ISOCAM & ISOPHOT) observations of a sample of X-ray selected weak-line (WTTS) and post? (PTTS) T Tauri stars located in the nearby Chamaeleon T and Scorpius-Centaurus OB associations. HST/NICMOS and adaptive optics observations aimed at identifying substellar companions (young brown dwarfs) at separations ?30 AU from the primary stars. No such objects were found within 300 AU of any of the target stars, and a number of faint objects at larger separations can very likely be attributed to a population of field (background) stars. ISOCAM observations of 5 to 15 Myr old WTTSs and PTTSs in ScoCen reveal infrared excesses which are clearly above photospheric levels and which have a spectral index intermediate between that of younger (1 to 5 Myr) T Tauri stars in Chamaeleon and that of pure stellar photospheres. The difference in the spectral index of the older PTTSs in ScoCen compared with the younger classical and weak-line TTSs in Cha can be attributed to a deficiency of smaller size (0.1 to 1 ?m) dust grains relative to larger size (?5 ?m) dust grains in the disks of the PTTSs. The lack of small dust grains is either due to the environment (effect of nearby O stars and supernova explosions) or due to disk evolution. If the latter is the case, it would hint that circumstellar disks start to become dust depleted at an age between 5 to 15 Myr. Dust depletion is very likely related to the build-up of larger particles (ultimately rocks and planetesimals) and thus an indicator for the onset of the period of planet formation.

Analysis of Keck high-resolution spectra of VB 10

We use a preliminary version of our “NextGen” grid of cool star model atmospheres to compute synthetic line profiles which fit high resolution Keck spectra of the cool M dwarf VB10 satisfactorily well. We show that the parameters derived from the Keck data are consistent with the parameters derived from lower resolution spectra with larger wavelength coverage. We discuss the treatment of van der Waals broadening in cool, molecular (mostly H2) dominated stellar atmospheres. The line profiles are dominated by van der Waals pressure broadening and are a sensitive indicator for the gravity and metallicity. Therefore, the high-resolution Keck spectra are useful for determining the parameters of M dwarfs. There is some ambiguity between the metallicity and gravity. For VB10, we find from the high-resolution spectra that 5.0 < log(g) < 5.5 and 0 < � M H � < +0.5 for an adopted fixed effective temperature of 2700 K (Schweitzer, 1995), which is consistent with recent interior calculations (e.g. Baraffe et al., 1995).

Non-LTE effects of Na I in the atmosphere of HD 209458b

The recent announcement that sodium absorption has been observed in the atmosphere of HD 209458b, the only extrasolar giant planet (EGP) observed to transit its parent star, is the first direct detection of an EGP atmosphere. We explore the possibility that neutral sodium is not in local thermodynamic equilibrium (LTE) in the outer atmosphere of irradiated EGPs and that the sodium concentration may be underestimated by models that make the LTE assumption. Our results indicate that it may not be necessary to invoke excessive photoionization, low metallicity, or even high-altitude clouds to explain the observations.

Effective Temperatures of Late L Dwarfs and the Onset of Methane Signatures

We present a spectral analysis of a sample of late L dwarfs. We use our latest model atmospheres and synthetic spectra and optical and K-band spectra to determine effective temperatures. We derive effective temperatures of 1400-1700 K for L8-L6 dwarfs. The analysis demonstrates that our recent models that rain out the formed dust completely are applicable to optical spectra of late L dwarfs and that more consistent models are needed for intermediate L dwarfs and for infrared spectra. We compare the results for the effective temperatures with the temperatures of the onset of methane formation. Our models predict methane absorption at 3.3 μm to occur at temperatures about 400 K higher than methane absorption at 2.2 μm. This is consistent with our data and previous observations, which show methane absorption at 3.3 μm, but not at 2.2 μm, in late L dwarfs.

Non-LTE Treatment of Molecules in the Photospheres of Cool Stars

We present a technique to treat systems with very many levels, such as molecules, in non-LTE. This method is based on a superlevel formalism coupled with rate operator splitting. Superlevels consist of many individual levels that are assumed to be in LTE relative to each other. The usage of superlevels reduces the dimensionality of the rate equations dramatically and, thereby, makes the problem computationally more easily treatable. Our superlevel formalism retains maximum accuracy by using direct opacity sampling (dOS) when calculating the radiative transitions and the opacities. We developed this method in order to treat molecules in cool dwarf model calculations in non-LTE. Cool dwarfs have low electron densities and radiation fields that are far from blackbody radiation fields; both properties may disqualify them from the common LTE approximation. Therefore, the most important opacity sources, the molecules, need to be treated in non-LTE. As a case study we applied our method to carbon monoxide. We find that our method gives accurate results since the conditions for the superlevel method are very well met for molecules. Because of very high collisional cross sections with hydrogen and the high densities of H2, the population of CO itself shows no significant deviation from LTE.

Highlights of Stellar Modeling with PHOENIX

We briefly describe the current version of the PHOENIX code. We then present some illustrative results from the modeling of Type Ia and Type II supernovae, hot stars, and irradiated giant planets. Good fits to observations can be obtained, when account is taken for spherically symmetric, line-blanketed, static or expanding atmospheres.

Cool Stellar Atmospheres

We give an overview about the state-of-the-art in cool stellar (and sub-stellar) atmosphere simulations. Recent developments in numerical methods and parallel supercomputers, as well as in the quality of input data such as atomic and molecular line lists have led to substantial improvements in the quality of synthetic spectra when compared to multi-wavelength observations. A wide range of objects from M dwarfs and giants down to substellar objects is considered. We discuss effects such as atomic and molecular NLTE (and) line blanketing, external irradiation, and formation and opacities of dust particles and clouds; each of which affects the structure of the atmospheres and their spectra. Current models can simultaneously fit many of the observed features of a given star with a single model atmosphere, however, a number of problems remain unsolved and will have to be addressed in the future, in particular for very low mass stars and substellar objects.