Going Bayesian on the ages of nearby young stellar systems

I. The expansion rate method

¹ Departamento de Inteligencia Artificial, Universidad Nacional de Educación a Distancia (UNED), c/Juan del Rosal 16, 28040 Madrid, Spain
² Depto. Estadística e Investigación Operativa, Universidad de Cádiz, Avda. República Saharaui s/n, Cádiz, 11510 Puerto Real, Spain
³ Laboratoire d’astrophysique de Bordeaux, Univ. Bordeaux, CNRS, B18N, allée Geoffroy Saint-Hilaire, 33615 Pessac, France

^★ Corresponding author: jolivares@dia.uned.es

Received: 27 February 2025
Accepted: 5 June 2025

Abstract

Context. Determining the ages of young stellar systems is fundamental for testing and validating current star-formation theories. Aims. We aim to develop a Bayesian version of the expansion-rate method that incorporates the a priori knowledge of the stellar system’s age and solves some of the caveats of the traditional frequentist approach.

Methods. We upgraded an existing Bayesian hierarchical model with additional parameter hierarchies that include, amongst others, the system’s age. We propose prior distributions inspired by literature works.

Results. We validate our method on a set of extensive simulations mimicking the properties of real stellar systems. In stellar associations between 10 and 40 Myr and up to 150 pc; the errors are <10%. In star forming regions up to 400 pc, the error can be as large as 80% at 3 Myr, but it rapidly decreases with increasing age.

Conclusions. The Bayesian expansion-rate methodology that we present here offers several advantages over the traditional frequentist version. In particular, the Bayesian age estimator is more robust and credible than the commonly used frequentist ones. This new Bayesian expansion-rate method is made publicly available as a module of the free and open-source code Kalkayotl.