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

TL;DR

This paper introduces a Bayesian framework and a Python tool for accurately estimating accretion rates in young stellar objects using spectroscopic data, improving uncertainty handling and parameter correlation analysis.

Contribution

We developed and publicly released 'nuts-for-ysos', a Bayesian Python package for inferring YSO accretion rates from spectroscopic data, incorporating uncertainties and parameter correlations.

Findings

The Bayesian method reliably estimates accretion rates.

Comparison validates the approach against emission line methods.

The tool is adaptable to other spectroscopic datasets.

Abstract

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 ($R \sim 800$, 3500-5500\r{A}) 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 datasets. 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 NGC1333 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.