A Bayesian Approach to Inferring Accretion Signatures in Young Stellar Objects: A Case Study with VIRUS

The mass accretion rates of young stellar objects (YSOs) are key to understanding how stars form, how their circumstellar disks evolve, and even how planets form. We develop a Bayesian framework to determine the accretion rates of a sample of 15 YSOs using archival data from the VIRUS spectrograph (R ∼ 800, 3500–5500 Å) on the Hobby–Eberly Telescope. We are publicly releasing our developed tool, dubbed nuts-for-ysos, as a Python package, which can also be applied to other spectroscopic data sets (https://github.com/laurenwillett/nuts-for-ysos). The nuts-for-ysos code fits a simple accretion model to the near-UV and optical continuum of each VIRUS spectrum. Our Bayesian approach aims to identify correlations between model parameters using the No U-Turn Sampler (NUTS). Moreover, this approach self-consistently incorporates all parameter uncertainties, allowing for a thorough estimation of the probability distribution for accretion rate not accomplished in previous works. Using nuts-for-ysos, we derive accretion rates of each YSO. We then verify the reliability of our method by comparing to results separately derived from only the spectral emission lines, and to results from earlier studies of the Lupus, Chamaeleon I, and NGC 1333 regions. Finally, we discuss what qualitative trends, covariances, and degeneracies were found among model parameters. The technique developed in this paper is a useful improvement that can be applied in the future to larger samples of YSOs observed by VIRUS or other spectrographs.