APOGEE Net: a unified spectral model

DR17
APOGEE-2
STAR
Dani Sprague, Connor Culhane, Richard Olney, Marina Kounkel, Kevin Covey, Brian Hutchinson, Ryan Lingg, Keivan Stassun, Carlos Román-Zúñiga, Alexandre Roman Lopes, Jura Borissova, Amelia Stutz, Serena Kim, Guy Stringfellow, Karla Pena Ramirez, Valeria Guadalupe Ramírez Preciado
Type:
Alternative Pipeline Catalog

This catalog contains the predicted stellar parameters (effective temperature, surface gravity, metallicity) for the APOGEE spectra. The parameters are derived using APOGEE Net (Olney et al. 2020, Sprague et al., in prep). APOGEE Net is a convolutional neural network, which has been calibrated to operate on all stars, including those for which the parameters have been challenging to derive by other means (e.g., M dwarfs, pre-main sequence stars, and OB stars). The catalog contains a unique entry for each APOGEE ID; in cases where multiple observations of the same source exist in several apStar files, only the one with the highest signal-to-noise ratio (SNR) is reported.