Catherine Espaillat

Resolved Images of Large Cavities in Protoplanetary Transition Disks

Circumstellar disks are thought to experience a rapid transition phase in their evolution that can have a considerable impact on the formation and early development of planetary systems. We present new and archival high angular resolution (03 40-75 AU) Submillimeter Array (SMA) observations of the 880 μm (340 GHz) dust continuum emission from 12 such transition disks in nearby star-forming regions. In each case, we directly resolve a dust-depleted disk cavity around the central star. Using two-dimensional Monte Carlo radiative transfer calculations, we interpret these dust disk structures in a homogeneous, parametric model framework by reproducing their SMA continuum visibilities and spectral energy distributions. The cavities in these disks are large (R cav = 15-73 AU) and substantially depleted of small (~μm-sized) dust grains, although their mass contents are still uncertain. The structures of the remnant material at larger radii are comparable to normal disks. We demonstrate that these large cavities are relatively common among the millimeter-bright disk population, comprising at least 1 in 5 (20%) of the disks in the bright half (and ≥26% of the upper quartile) of the millimeter luminosity (disk mass) distribution. Utilizing these results, we assess some of the physical mechanisms proposed to account for transition disk structures. As has been shown before, photoevaporation models do not produce the large cavity sizes, accretion rates, and disk masses representative of this sample. A sufficient decrease of the dust optical depths in these cavities by particle growth would be difficult to achieve: substantial growth (to meter sizes or beyond) must occur in large (tens of AU) regions of low turbulence without also producing an abundance of small particles. Given those challenges, we suggest instead that the observations are most commensurate with dynamical clearing due to tidal interactions with low-mass companions—very young (~1 Myr) brown dwarfs or giant planets on long-period orbits.

THE DISK POPULATION OF THE TAURUS STAR-FORMING REGION*

We have analyzed nearly all images of the Taurus star-forming region at 3.6, 4.5, 5.8, 8.0, and 24??m that were obtained during the cryogenic mission of the Spitzer Space Telescope (46 deg2) and have measured photometry for all known members of the region that are within these data, corresponding to 348 sources, or 99% of the known stellar population. By combining these measurements with previous observations with the Spitzer Infrared Spectrograph and other facilities, we have classified the members of Taurus according to whether they show evidence of circumstellar disks and envelopes (classes I, II, and III). Through these classifications, we find that the disk fraction in Taurus, N(II)/N(II+III), is ~75% for solar-mass stars and declines to ~45% for low-mass stars and brown dwarfs (0.01-0.3 M ?). This dependence on stellar mass is similar to that measured for Chamaeleon I, although the disk fraction in Taurus is slightly higher overall, probably because of its younger age (1?Myr versus 2-3?Myr). In comparison, the disk fraction for solar-mass stars is much lower (~20%) in IC 348 and ? Ori, which are denser than Taurus and Chamaeleon I and are roughly coeval with the latter. These data indicate that disk lifetimes for solar-mass stars are longer in star-forming regions that have lower stellar densities. Through an analysis of multiple epochs of Spitzer photometry that are available for ~200 Taurus members, we find that stars with disks exhibit significantly greater mid-infrared (mid-IR) variability than diskless stars, which agrees with the results of similar variability measurements for a smaller sample of stars in Chamaeleon I. The variability fraction for stars with disks is higher in Taurus than in Chamaeleon I, indicating that the IR variability of disks decreases with age. Finally, we have used our data in Taurus to refine the observational criteria for primordial, evolved, and transitional disks. The ratio of the number of evolved and transitional disks to the number of primordial disks in Taurus is 15/98 for spectral types of K5-M5, indicating a timescale of 0.15 ? ?primordial ~ 0.45?Myr for the clearing of the inner regions of optically thick disks. After applying the same criteria to older clusters and associations (2-10?Myr) that have been observed with Spitzer, we find that the proportions of evolved and transitional disks in those populations are consistent with the measurements in Taurus when their star formation histories are properly taken into account.

TRANSITIONAL AND PRE-TRANSITIONAL DISKS: GAP OPENING BY MULTIPLE PLANETS?

We use two-dimensional hydrodynamic simulations of viscous disks to examine whether dynamically interacting multiple giant planets can explain the large gaps (spanning over one order of magnitude in radius) inferred for the transitional and pre-transitional disks around T Tauri stars. In the absence of inner disk dust depletion, we find that it requires three to four giant planets to open up large enough gaps to be consistent with inferences from spectral energy distributions, because the gap width is limited by the tendency of the planets to be driven together into 2:1 resonances. With very strong tidal torques and/or rapid planetary accretion, fewer planets can also generate a large cavity interior to the locally formed gap(s) by preventing outer disk material from moving in. In these cases, however, the reduction of surface density produces a corresponding reduction in the inner disk accretion rate onto the star; this makes it difficult to explain the observed accretion rates of the pre-transitional/transitional disks. We find that even with four planets in disks, additional substantial dust depletion is required to explain observed disk gaps/holes. Substantial dust settling and growth, with consequent significant reductions in optical depths, is inferred for typical T Tauri disks in any case, and an earlier history of dust growth is consistent with the hypothesis that pre-transitional/transitional disks are explained by the presence of giant planets. We conclude that the depths and widths of gaps and disk accretion rates in pre-transitional/transitional disks cannot be reproduced by a planet-induced gap opening scenario alone. Significant dust depletion is also required within the gaps/holes. Order-of-magnitude estimates suggest that the mass of small dust particles (1 μm) relative to the gas must be depleted to 10 −5 to 10 −2 of the interstellar medium value, implying a very efficient mechanism of small dust removal or dust growth.

UNVEILING THE STRUCTURE OF PRE-TRANSITIONAL DISKS

In the past few years, several disks with inner holes that are relatively empty of small dust grains have been detected and are known as transitional disks. Recently, Spitzer has identified a new class of pre-transitional disks with gaps based on near-infrared photometry and mid-infrared spectra; these objects have an optically thick inner disk separated from an optically thick outer disk by an optically thin disk gap. A near-infrared spectrum provided the first confirmation of a gap in the pre-transitional disk of LkCa 15 by verifying that the near-infrared excess emission in this object was due to an optically thick inner disk. Here, we investigate the difference between the nature of the inner regions of transitional and pre-transitional disks using the same veiling-based technique to extract the near-infrared excess emission above the stellar photosphere. However, in this work we use detailed disk models to fit the excess continua as opposed to the simple blackbody fits previously used. We show that the near-infrared excess emission of the previously identified pre-transitional disks of LkCa 15 and UX Tau A in the Taurus cloud as well as the newly identified pre-transitional disk of ROX 44 in Ophiuchus can be fit with an inner disk wall located at the dust destruction radius. We also present detailed modeling of the broadband spectral energy distributions of these objects, taking into account the effect of shadowing by the inner disk on the outer disk, but considering the finite size of the star, unlike other recent treatments. The near-infrared excess continua of these three pre-transitional disks, which can be explained by optically thick inner disks, are significantly different from that of the transitional disks of GM Aur, whose near-infrared excess continuum can be reproduced by emission from sub-micron-sized optically thin dust, and DM Tau, whose near-infrared spectrum is consistent with a disk hole that is relatively free of small dust. The structure of pre-transitional disks may be a sign of young planets forming in these disks and future studies of pre-transitional disks will provide constraints to aid in theoretical modeling of planet formation.

ON THE DIVERSITY OF THE TAURUS TRANSITIONAL DISKS : UX TAURI A AND LkCa 15

The recently recognized class of transitional disk systems consists of young stars with optically thick outer disks but inner disks which are mostly devoid of small dust grains. Here we introduce a further class of pre-transitional disks with significant near-infrared excesses which indicate the presence of an optically thick inner disk separated from an optically thick outer disk; thus, the spectral energy distributions of pre-transitional disks suggest the incipient development of disk gaps rather than inner holes. In UX Tau A, our analysis of the Spitzer IRS spectrum finds that the near-infrared excess is produced by an inner optically thick disk and that a gap of ~56 AU is present. The Spitzer IRS spectrum of LkCa 15 is suggestive of a gap of ~46 AU, confirming previous millimeter imaging. In addition, UX Tau A contains crystalline silicates in its disk at radii 56 AU which poses a challenge to our understanding of the production of this crystalline material. In contrast, LkCa 15s silicates are amorphous and pristine. UX Tau A and LkCa 15 increase our knowledge of the diversity of dust clearing in low-mass star formation.

A Spatially Resolved Inner Hole in the Disk Around GM Aurigae

We present 0. 3 resolution observations of the disk around GM Aurigae with the Submillimeter Array (SMA) at a wavelength of 860 μm and with the Plateau de Bure Interferometer at a wavelength of 1.3 mm. These observations probe the distribution of disk material on spatial scales commensurate with the size of the inner hole predicted by models of the spectral energy distribution (SED). The data clearly indicate a sharp decrease in millimeter optical depth at the disk center, consistent with a deficit of material at distances less than ∼20 AU from the star. We refine the accretion disk model of Calvet et al. based on the unresolved SED and demonstrate that it reproduces well the spatially resolved millimeter continuum data at both available wavelengths. We also present complementary SMA observations of CO J = 3–2 and J = 2–1 emission from the disk at 2 �� resolution. The observed CO morphology is consistent with the continuum model prediction, with two significant deviations: (1) the emission displays a larger CO J = 3–2/J = 2–1 line ratio than predicted, which may indicate additional heating of gas in the upper disk layers; and (2) the position angle of the kinematic rotation pattern differs by 11 ◦ ± 2 ◦ from that measured at smaller scales from the dust continuum, which may indicate the presence of a warp. We note that photoevaporation, grain growth, and binarity are unlikely mechanisms for inducing the observed sharp decrease in opacity or surface density at the disk center. The inner hole plausibly results from the dynamical influence of a planet on the disk material. Warping induced by a planet could also potentially explain the difference in position angle between the continuum and CO data sets.

A SPITZER IRS STUDY OF INFRARED VARIABILITY IN TRANSITIONAL AND PRE-TRANSITIONAL DISKS AROUND T TAURI STARS

We present a Spitzer IRS study of variability in 14 T Tauri stars in the Taurus and Chamaeleon star-forming regions. The sample is composed of transitional and pre-transitional objects which contain holes and gaps in their disks. We detect variability between 5 and 38 μm in all but two of our objects on timescales of 2-3 years. Most of the variability observed can be classified as seesaw behavior, whereby the emission at shorter wavelengths varies inversely with the emission at longer wavelengths. For many of the objects we can reasonably reproduce the observed variability using irradiated disk models, particularly by changing the height of the inner disk wall by ~20%. When the inner wall is taller, the emission at the shorter wavelengths is higher since the inner wall dominates the emission at 2-8 μm. The taller inner wall casts a larger shadow on the outer disk wall, leading to less emission at wavelengths beyond 20 μm where the outer wall dominates. We discuss how the possible presence of planets in these disks could lead to warps that cause changes in the height of the inner wall. We also find that crystalline silicates are common in the outer disks of our objects and that in the four disks in the sample with the most crystalline silicates, variability on timescales of 1 week is present. In addition to explaining the infrared variability described above, planets can create shocks and collisions which can crystallize the dust and lead to short timescale variability.

Accretion Rates for T Tauri Stars Using Nearly Simultaneous Ultraviolet and Optical Spectra

We analyze the accretion properties of 21 low-mass T Tauri stars using a data set of contemporaneous near-UV (NUV) through optical observations obtained with the Hubble Space Telescope Imaging Spectrograph and the ground-based Small and Medium Aperture Research Telescope System, a unique data set because of the nearly simultaneous broad wavelength coverage. Our data set includes accreting T Tauri stars in Taurus, Chamaeleon I, η Chamaeleon, and the TW Hydra Association. For each source we calculate the accretion rate (Ṁ) by fitting the NUV and optical excesses above the photosphere, produced in the accretion shock, introducing multiple accretion components characterized by a range in energy flux (or density) for the first time. This treatment is motivated by models of the magnetospheric geometry and accretion footprints, which predict that high-density, low filling factor accretion spots coexist with low-density, high filling factor spots. By fitting the UV and optical spectra with multiple accretion components, we can explain excesses which have been observed in the near-IR. Comparing our estimates of Ṁ to previous estimates, we find some discrepancies; however, they may be accounted for when considering assumptions for the amount of extinction and variability in optical spectra. Therefore, we confirm many previous estimates of the accretion rate. Finally, we measure emission line luminosities from the same spectra used for the Ṁ estimates, to produce correlations between accretion indicators (Hβ, Ca II K, C II], and Mg II) and accretion properties obtained simultaneously.

THE SPITZER INFRARED SPECTROGRAPH SURVEY OF T TAURI STARS IN TAURUS

We present 161 Spitzer Infrared Spectrograph (IRS) spectra of T Tauri stars and young brown dwarfs in the Taurus star-forming region. All of the targets were selected based on their infrared excess and are therefore surrounded by protoplanetary disks; they form the complete sample of all available IRS spectra of T Tauri stars with infrared excesses in Taurus. We also present the IRS spectra of seven Class 0/I objects in Taurus to complete the sample of available IRS spectra of protostars in Taurus. We use spectral indices that are not significantly affected by extinction to distinguish between envelope- and disk-dominated objects. Together with data from the literature, we construct spectral energy distributions for all objects in our sample. With spectral indices derived from the IRS spectra we infer disk properties such as dust settling and the presence of inner disk holes and gaps. We find a transitional disk frequency, which is based on objects with unusually large 13–31 μm spectral indices indicative of a wall surrounding an inner disk hole, of about 3%, and a frequency of about 20% for objects with unusually large 10 μm features, which could indicate disk gaps. The shape and strength of the 10 μm silicate emission feature suggests weaker 10 μm emission and more processed dust for very low mass objects and brown dwarfs (spectral types M6–M9). These objects also display weaker infrared excess emission from their disks, but do not appear to have more settled disks than their higher-mass counterparts. We find no difference for the spectral indices and properties of the dust between single and multiple systems.

GRB 021004: A Possible Shell Nebula around a Wolf-Rayet Star Gamma-Ray Burst Progenitor*

The rapid localization of GRB 021004 by the HETE-2 satellite allowed nearly continuous monitoring of its early optical afterglow decay, as well as high-quality optical spectra that determined a redshift of z3 = 2.328 for its host galaxy, an active starburst galaxy with strong Lyα emission and several absorption lines. Spectral observations show multiple absorbers at z3A = 2.323, z3B = 2.317, and z3C = 2.293 blueshifted by ~450, ~990, and ~3155 km s-1, respectively, relative to the host galaxy Lyα emission. We argue that these correspond to a fragmented shell nebula that has been radiatively accelerated by the gamma-ray burst (GRB) afterglow at a distance 0.3 pc from a Wolf-Rayet star GRB progenitor. The chemical abundance ratios indicate that the nebula is overabundant in carbon and silicon. The high level of carbon and silicon is consistent with a swept-up shell nebula gradually enriched by a carbon-rich late-type Wolf-Rayet progenitor wind over the lifetime of the nebula prior to the GRB onset. The detection of statistically significant fluctuations and color changes about the jetlike optical decay further supports this interpretation, since fluctuations must be present at some level as a result of irregularities in a clumpy stellar wind medium or if the progenitor has undergone massive ejection prior to the GRB onset. This evidence suggests that the mass-loss process in a Wolf-Rayet star might lead naturally to an iron core collapse with sufficient angular momentum that could serve as a suitable GRB progenitor. Even though we cannot rule out definitely the alternatives of a dormant QSO, large-scale superwinds, or a several hundred year old supernova remnant responsible for the blueshifted absorbers, these findings point to the likelihood of a signature for a massive-star GRB progenitor.