Steven W. Stahler’s research while affiliated with University of California, Berkeley and other places

As a stellar group forms within its parent molecular cloud, new members first appear in the deep interior. These overcrowded stars continually diffuse outward to the cloud boundary, and even beyond. Observations have so far documented only the interior drift. Those stars that actually leave the cloud form an expanding envelope that I call the ‘stellar mantle.’ Simple fluid models for the cloud and mantle illustrate their basic structure. The mantle’s expansion speed is subsonic with respect to the cloud’s dynamical temperature. I describe, in qualitative terms, how the expanding mantle and Galactic tidal radius might together shape the evolution of specific types of stellar groups. The massive stars in OB associations form in clouds that contract before extruding a substantial mantle. In contrast, the more slowly evolving clouds forming open clusters and T associations have extended mantles that encounter a shrinking tidal radius. These clouds are dispersed by internal stellar outflows. If the remaining group of stars is gravitationally bound, it appears as a long-lived open cluster, truncated by the tidal radius. If the group is unbound, it is a late-stage T association that will soon be torn apart by the tidal force. The ‘distributed’ populations of pre-main sequence stars observed in the outskirts of several star-forming regions are too distant to be stellar mantles. Rather, they could be the remnants of especially low-mass T associations.

We present a quantitative, empirically based argument that at least some Class I sources are low-mass, pre-main-sequence stars surrounded by spatially extended envelopes of dusty gas. The source luminosity arises principally from stellar gravitational contraction, as in optically visible pre-main-sequence stars that lack such envelopes. We base our argument on the fact that some Class I sources in Orion and other star-forming regions have been observed by Spitzer to be periodic variables in the mid-infrared, and with periods consistent with T Tauri rotation rates. Using a radiative transfer code, we construct a variety of dust envelopes surrounding rotating, spotted stars, to see whether an envelope that produces a Class I spectral energy distribution at least broadly matches the observed modulations in luminosity. Acceptable envelopes can be either spherical or flattened and may or may not have polar cavities. The key requirement is that they have a modest equatorial optical depth at the Spitzer waveband of 3.6 μ m, typically τ 3.6 ≈ 0.6. The total envelope mass, based on this limited study, is at most about 0.1 M ⊙ , less than a typical stellar mass. Future studies should focus on the dynamics of the envelope, to determine whether material is actually falling onto the circumstellar disk.

We present a quantitative, empirically based argument that at least some Class I sources are low-mass, pre-main-sequence stars surrounded by spatially extended envelopes of dusty gas. The source luminosity arises principally from stellar gravitational contraction, as in optically visible pre-main-sequence stars that lack such envelopes. We base our argument on the fact that some Class I sources in Orion and other star-forming regions have been observed by Spitzer to be periodic variables in the mid-infrared, and with periods consistent with T Tauri rotation rates. Using a radiative transfer code, we construct a variety of dust envelopes surrounding rotating, spotted stars, to see if an envelope that produces a Class I SED at least broadly matches the observed modulations in luminosity. Acceptable envelopes can either be spherical or flattened, and may or may not have polar cavities. The key requirement is that they have a modest equatorial optical depth at the Spitzer waveband of 3.6 {\mu}m, typically {\tau}3.6 {\approx} 0.6. The total envelope mass, based on this limited study, is at most about 0.1 M{\odot}, less than a typical stellar mass. Future studies should focus on the dynamics of the envelope, to determine whether material is actually falling onto the circumstellar disk.

We explore the relationship between young, embedded binaries and their parent cores, using observations within the Perseus Molecular Cloud. We combine recently published VLA observations of young stars with core properties obtained from SCUBA-2 observations at 850 um. Most embedded binary systems are found toward the centres of their parent cores, although several systems have components closer to the core edge. Wide binaries, defined as those systems with physical separations greater than 500 au, show a tendency to be aligned with the long axes of their parent cores, whereas tight binaries show no preferred orientation. We test a number of simple, evolutionary models to account for the observed populations of Class 0 and I sources, both single and binary. In the model that best explains the observations, all stars form initially as wide binaries. These binaries either break up into separate stars or else shrink into tighter orbits. Under the assumption that both stars remain embedded following binary breakup, we find a total star formation rate of 168 Myr^-1. Alternatively, one star may be ejected from the dense core due to binary breakup. This latter assumption results in a star formation rate of 247 Myr^-1. Both production rates are in satisfactory agreement with current estimates from other studies of Perseus. Future observations should be able to distinguish between these two possibilities. If our model continues to provide a good fit to other star-forming regions, then the mass fraction of dense cores that becomes stars is double what is currently believed.

We explore the relationship between young, embedded binaries and their parent cores, using observations within the Perseus Molecular Cloud. We combine recently published VLA observations of young stars with core properties obtained from SCUBA-2 observations at 850 um. Most embedded binary systems are found toward the centres of their parent cores, although several systems have components closer to the core edge. Wide binaries, defined as those systems with physical separations greater than 500 au, show a tendency to be aligned with the long axes of their parent cores, whereas tight binaries show no preferred orientation. We test a number of simple, evolutionary models to account for the observed populations of Class 0 and I sources, both single and binary. In the model that best explains the observations, all stars form initially as wide binaries. These binaries either break up into separate stars or else shrink into tighter orbits. Under the assumption that both stars remain embedded following binary breakup, we find a total star formation rate of 168 Myr^-1. Alternatively, one star may be ejected from the dense core due to binary breakup. This latter assumption results in a star formation rate of 247 Myr^-1. Both production rates are in satisfactory agreement with current estimates from other studies of Perseus. Future observations should be able to distinguish between these two possibilities. If our model continues to provide a good fit to other star-forming regions, then the mass fraction of dense cores that becomes stars is double what is currently believed.

We assess the evolutionary status of EXors. These low-mass, pre-main-sequence stars repeatedly undergo sharp luminosity increases, each a year or so in duration. We place into the HR diagram all EXors that have documented quiescent luminosities and effective temperatures, and thus determine their masses and ages. Two alternate sets of pre-main-sequence tracks are used, and yield similar results. Roughly half of EXors are embedded objects, i.e., they appear observationally as Class I or flat-spectrum infrared sources. We find that these are relatively young and are located close to the stellar birthline in the HR diagram. Optically visible EXors, on the other hand, are situated well below the birthline. They have ages of several Myr, typical of classical T Tauri stars. Judging from the limited data at hand, we find no evidence that binarity companions trigger EXor eruptions; this issue merits further investigation. We draw several general conclusions. First, repetitive luminosity outbursts do not occur in all pre-main-sequence stars, and are not in themselves a sign of extreme youth. They persist, along with other signs of activity, in a relatively small subset of these objects. Second, the very existence of embedded EXors demonstrates that at least some Class I infrared sources are not true protostars, but very young pre-main-sequence objects still enshrouded in dusty gas. Finally, we believe that the embedded pre-main-sequence phase is of observational and theoretical significance, and should be included in a more complete account of early stellar evolution.