The period-age relation of long-period variables

TL;DR

This study investigates the period-age relation of long-period variables using nonlinear pulsation models and observational data, revealing different slopes for O-rich and C-rich stars and emphasizing the importance of including semi-regular variables for population studies.

Contribution

It provides the first theoretical examination of the period-age relation for LPVs, deriving new relations for O-rich and C-rich stars based on combined models and observations.

Findings

Period decreases with age, consistent with observations.

Period distribution at a given age is broad and skewed toward short periods.

Different period-age slopes are predicted for O-rich and C-rich LPVs.

Abstract

Pieces of empirical evidence suggest the existence of a period-age relation for long-period variables (LPVs). Yet, this property has hardly been studied on theoretical grounds thus far. We aim to examine the period-age relation using the results from recent nonlinear pulsation calculations. We combined isochrone models with theoretical periods to simulate the distribution of fundamental mode LPV pulsators, which include Miras, in the period-age plane, and we compared it with observations of LPVs in Galactic and Magellanic Clouds' clusters. In agreement with observations, models predict that the fundamental mode period decreases with increasing age because of the dominant role of mass in shaping stellar structure and evolution. At a given age, the period distribution shows a non-negligible width and is skewed toward short periods, except for young C-rich stars. As a result, the period-age relations of O-rich and C-rich models are predicted to have different slopes. We derived best-fit relations describing age and initial mass as a function of the fundamental mode period for both O- and C-rich models. The study confirms the power of the period-age relations to study populations of LPVs of specific types, either O-rich or C-rich, on statistical grounds. In doing so, it is recommended not to limit a study to Miras, which would make it prone to selection biases, but rather to include semi-regular variables that pulsate predominantly in the fundamental mode. The use of the relations to study individual LPVs, on the other hand, requires more care given the scatter in the period distribution predicted at any given age.