Poster
T. Boulet
Abstract
Understanding the Milky Way's formation and evolution requires precise stellar age
determination across its components. Recent advances in asteroseismology, spectroscopy,
stellar modeling, and machine learning, along with all-sky surveys, have provided reliable stellar
age estimates. We aim to furnish accurate age assessments for the Main Red Star Sample
within the APOGEE DR17 catalogue. Leveraging asteroseismic age constraints, we employ
machine learning to achieve this goal. We explore optimal non-asteroseismic stellar parameters,
including Teff, L, [CI/N], [Mg/Ce], [α/Fe], U(LSR) velocity, and 'Z' vertical height from the Galactic
plane, to predict ages via categorical gradient boost decision trees. Our model, trained on
merged samples from TESS and APOKASC catalogs, achieves a median fractional age error of
20.8%, with a variance of 4.77%. For stars older than 3 Gyr, the error ranges from 7% to 23%,
for those between 1 and 3 Gyr, it is 26% to 28%, and for stars younger than 1 Gyr, it is 43%.
Applied to 125,445 stars, our analysis confirms the flaring of the young Galactic disc and
reveals an age gradient among the youngest Galactic plane stars. We also identify two groups
of metal-poor ([Fe/H] < -1 dex) and young (Age < 2 Gyr) stars exhibiting similar chemical
abundances and halo kinematics, likely remnants of the predicted third gas infall episode
around 2.7 Gyr ago.
8th TESS/15th Kepler Asteroseismic Science Consortium Workshop
Porto, Portugal
2024 July
