The Gaia 500 pc white dwarf sample: machine-learning and observational spectral classification

Not scheduled
2m
Poster Posters

Speaker

Enrique Miguel García Zamora (Universitat Politècnica de Catalunya)

Description

The third data release of Gaia has provided the astronomical community with astrometric data for 1.8 billion sources and optical spectra for more than 220 million objects, including about 100 000 white dwarfs. White dwarfs are the final evolutionary stage of the vast majority of stars and serve as key tracers of stellar evolution, galactic structure, and the star-formation history of the Milky Way. Their spectra are divided into different classes according to their atmospheric composition and spectral features, which reflect their physical properties and evolutionary pathways. However, the sheer volume of Gaia data makes visual classification impractical, motivating the use of machine-learning and artificial-intelligence techniques.

In this work, we present the results of applying a Random Forest algorithm to the Gaia spectra of the 500-pc white dwarf sample. We provide spectral classifications for 79 000 objects and compare them with other automatic classification methods, as well as with our own follow-up spectroscopy of 255 objects obtained with the Gran Telescopio Canarias, in order to evaluate the performance and robustness of the method. We also discuss the reliability of the classifications and the main sources of disagreement between the different approaches.

Author

Enrique Miguel García Zamora (Universitat Politècnica de Catalunya)

Presentation materials

There are no materials yet.