In 1984 you published a paper with Frank Shu and Pat Cassen on the collapse of cores of slowly rotating isothermal clouds, which has had a major impact on our understanding of protostellar collapse. Please describe the genesis of this paper.
I remember that Berkeley Astronomy was a tremendously vibrant environment in the late 1970's. In the background was the women's movement; it was a big event that my graduate class had three women, thus doubling the number of female graduate students. Computers were rapidly enabling new types of science. Radio astronomy was strong. For several years I did in fact pursue two different thesis topics, theory with Frank Shu, and radio astronomy with Jack Welch and Paul Ho. At the time Jack was building the first millimeter interferometer at Hat Creek Observatory, and some early targets were massive stars in compact HII regions. Frank first involved me in a project involving density wave theory. However, I did not like the project and was fortunate that Frank made another suggestion, to try including rotation with his 1977 collapse model. From the beginning it seemed interesting, and I remember scouring the literature because I felt it was important to make the initial conditions as realistic as possible. There were suggestions at the time that GMC's did not show global collapse motions, meaning that star formation was not an efficient process.
To form a sunlike star we made some key choices; we focused on low-mass stars because they seemed likely to preserve clues to their birth environments, rotation was assumed slow to match observed upper limits, clouds were cold and the collapse started from a marginal equilibrium state. We did not include magnetic fields, which meant angular momentum was conserved during the collapse. After first finding a solution for an initial state with rotation, we then applied perturbation techniques to solve for the collapse motions in a semi-analytic solution. It was a wonderful experience to work with Frank, and I learned an incredible amount from him. Around then, Cassen & Moosman (1981) were studying accretion onto a disk, which led to our collaboration with Pat. Observation of the first bipolar outflow in L1551 in 1980 was intriguing but puzzling. I remember that observations of newly detected ammonia cores by Myers & Benson in 1983 were particularly interesting, because these seemed to be the likely progenitors of low-mass star formation. For the 1984 journal article the images were displayed on a computer screen, and I took pictures of the screen with my 35mm film camera to send to the journal, thus publishing some early computer graphics. Our 1984 results were elegant, in that the collapse was fully described by only three parameters, the sound speed as, angular rotation rate Ω, and time t, now more commonly expressed as Ṁ, M*, and rdisk. By adding rotation, we were able to relate disk radius to the initial conditions in the cloud, and follow the collapse from 10,000 au all the way down to the protostar, which purely numerical simulations had struggled to do. From the beginning, the theoretical framework that we constructed seemed very promising.
Not long after, you did a study on a completely different subject, namely the size spectrum of molecular clouds in the Outer Galaxy.
I know that it might seem an abrupt change, but I had a very fun summer internship with Butler Burton at NRAO in Charlottesville where I picked up an interest in radio astronomy and galactic structure. I did several observing projects in graduate school including this one with fellow graduate student Mike Fich and post-docs Leo Blitz and Christian Henkel. Reports in the literature suggested large numbers of molecular clouds in the outer Galaxy. Our survey did not confirm the large numbers -- the clouds just weren't there, but by using statistical methods we showed that the molecular cloud size spectrum was consistent between the inner and outer Galaxy.
In a continuation of your interest in the Outer Galaxy you did a study of how infrared cirrus and neutral atomic hydrogen correlate. What did you find?
At the time, results from the IRAS all-sky survey were showing many new phenomena. In 1986 Mike Fich and I showed that there was a linear correlation between large scale far-infrared cirrus and neutral atomic hydrogen, and further that this could be explained by thermal dust emission from dusty gas that was cold (~20 K) and fairly uniform. We also ruled out several other emission mechanisms such as line emission.
Back in 1989 you used the Owens Valley Millimeter Interferometer to study the prevalence of outflows around protostars. What was the context at that time?
This paper is another favorite of mine, because of how nicely things aligned between theory, observation, and good company. At the time, IRAS infrared sources had been associated with NH3 dense cores, and single dish telescopes suggested that CO outflows were common (35%). However, due to low spatial resolution they did not necessarily peak at the same place, making comparison difficult. Millimeter interferometers could do better but required lots of time to image even one source. To show how much things have changed, the Hat Creek Interferometer run by Berkeley and Owens Valley Interferometer run by Caltech had only three antennas each then. Stuart Vogel and I were post-docs at the Owens Valley Interferometer, and Stuart had the clever idea of doing a snapshot survey. That allowed a statistically large enough sample of 25 objects to search for high-velocity CO gas, which we supplemented by full imaging of a few sources at higher sensitivity. Putting together a good sample was important, and Phil Myers knew the sources well. I was hoping to find some targets that were candidates for infall-only gas motions; instead we detected many more outflows (64-100), and established that infall and outflow must coexist over much of the protostar phase. Those who read this and subsequent papers will notice the descriptive phrase “very young low-mass star” instead of protostar. I settled on this phrase because referees objected to using the term protostar for the sources that we observed, and it was a decade or two before they were accepted as bonafide protostars.
You have also been interested in water masers as a probe of circumstellar environments.
Getting sub-arcsecond spatial resolution to study circumstellar structure has been a long standing challenge. In our 1992 paper we used the VLA to look for water masers towards low-mass protostars. We achieved 0.5’’, or roughly 100 au resolution, and detected maser emission towards 10% of sources. The high velocities of the masers showed they were associated with outflows, and their structure provided information on jet/outflow collimation at small scales.
In 1993 you published, together with Claire Chandler and Philippe André, a study of the relative contributions of disks and envelopes to the mm continuum of protostars. What did you learn?
Around that time millimeter continuum surveys were blossoming, due to improved detector technology. Surveys of T Tauri stars detected 1.3 mm emission that was being interpreted as circumstellar disk emission. We wanted to extend the results to protostars and imaged 10 sources at 2.7 mm using the OVRO Interferometer, and combined it with Philippe's 1.3 mm IRAM data. To understand the relative contribution of envelope versus disk emission we did a rigorous investigation using the TSC1984 collapse models. There is a lot packed into this paper, but some results were that the predicted fluxes would be larger in larger beams for envelopes, and that envelopes would dominate at shorter wavelengths too, such as for the IRAM 1.3 mm surveys with 11’’ resolution. Our interferometer images at longer wavelength (2.7 mm) and smaller beams (8’’) showed unresolved point sources that we concluded were protostellar disks with typical disk masses around 0.02 M⊙. The similarity in mass to the T Tauri disks showed that massive disks were rare during the protostar phase.
You have studied the embedded jet source TMC-1 with HST. What were the results?
With HST we achieved sub-arcsecond spatial resolution (0.15’’ = 21 au) at 1.6 μm in deep imaging of the Taurus protostar TMC-1. In this paper we explored the idea that radiative transfer modeling, combined with high quality imaging, could determine important parameters of the protostar ranging from source inclination to mass and age. We suggested that TMC-1 has low-mass, 0.1 -0.2 M⊙ , which would resolve any luminosity problem. The source itself is quite beautiful, it shows a narrow [Fe II] jet bisecting a wide-angle outflow that is illuminated by scattered light.
More recently you have analyzed Spitzer data of the very low-luminosity embedded source L1521F-IRS in Taurus. What is special about this source?
The source L1521F-IRS is a VeLLO that was discovered in the Taurus Spitzer Legacy Survey data. Such low-luminosity objects are extremely interesting to characterize, to see if they are caught in a youthful phase and are likely to continue accreting mass. Updating the techniques from our 1986 IRAS - HI paper, we analyzed the 160 μm dust emission, and estimated that L1521F-IRS sits in a cold cloud core that is 2-5 M⊙, and therefore has a mass reservoir that will allow it to grow. As part of the analysis we determined dust properties that are important but difficult to measure, finding an extinction ratio A160/AK = 0.010, and also determining an empirical dust opacity value at 160 microns.
What have you been doing recently?
The high spatial resolution ALMA image of gaps in the HL Tau disk is truly stunning. It was the inspiration to dust off the TSC1984 collapse models to study disk formation during the protostar phase. Together with colleagues Karen Willacy, Neal Turner, and Andrea Isella we added astrochemistry, radiative transfer, and collapse to make a software package that we call RadChemT. Our first paper with talented student Liz Flores-Rivera (2021) shows that RadChemT is able to reproduce ALMA and CARMA data for the L1527 protostar. Hopefully there will be many such papers to come.