Disentangled Representation Learning for Astronomical Chemical Tagging

Abstract

Modern astronomical surveys are observing spectral data for millions of stars. These spectra contain chemical information that can be used to trace the Galaxy's formation and chemical enrichment history. However, extracting the information from spectra, and making precise and accurate chemical abundance measurements are challenging. Here, we present a data-driven method for isolating the chemical factors of variation in stellar spectra from those of other parameters (i.e. \teff, \logg, \feh). This enables us to build a spectral projection for each star with these parameters removed. We do this with no ab initio knowledge of elemental abundances themselves, and hence bypass the uncertainties and systematics associated with modeling that rely on synthetic stellar spectra. To remove known non-chemical factors of variation, we develop and implement a neural network architecture that learns a disentangled spectral representation. We simulate our recovery of chemically identical stars using the disentangled spectra in a synthetic APOGEE-like dataset. We show that this recovery declines as a function of the signal to noise ratio, but that our neural network architecture outperforms simpler modeling choices. Our work demonstrates the feasibility of data-driven abundance-free chemical tagging.

arXiv commentAccepted by ApJ. code available at github.com/drd13/tagging-package

References in corpus (57)

• Gaia Data Release 2. Summary of the contents and survey properties
• The Apache Point Observatory Galactic Evolution Experiment (APOGEE)
• The New Galaxy: Signatures of its Formation
• The Radial Velocity Experiment (RAVE): first data release
• The GALAH Survey: Scientific Motivation
• Star Clusters Across Cosmic Time
• Chemical abundances of 1111 FGK stars from the HARPS GTO planet search program.Galactic stellar populations and planets
• APOGEE Data and Spectral Analysis from SDSS Data Release 16: Seven Years of Observations Including First Results from APOGEE-South
• Abundances, Stellar Parameters, and Spectra From the SDSS-III/APOGEE Survey
• The Cannon: A data-driven approach to stellar label determination
• Red giant masses and ages derived from carbon and nitrogen abundances
• New ATLAS9 And MARCS Model Atmosphere Grids for the Apache Point Observatory Galactic Evolution Experiment (APOGEE)
• The Payne: self-consistent ab initio fitting of stellar spectra
• Spectroscopic determination of masses (and implied ages) for red giants
• Deep learning of multi-element abundances from high-resolution spectroscopic data
• Accuracy and precision of industrial stellar abundances
• The Chemical Homogeneity of Sun-like Stars in the Solar Neighborhood
• The Variational Fair Autoencoder
• Carbon, nitrogen and $\alpha$-element abundances determine the formation sequence of the Galactic thick and thin disks
• The influence of atomic diffusion on stellar ages and chemical tagging
• Measuring Radial Orbit Migration in the Milky Way Disk
• APOGEE Data Releases 13 and 14: Stellar Parameter and Abundance Comparisons With Independent Analyses
• Prime Focus Spectrograph (PFS) for the Subaru Telescope: Overview, recent progress, and future perspectives
• Chemical Cartography with APOGEE: Multi-element abundance ratios
• The chemical homogeneity of open clusters
• Label Transfer from APOGEE to LAMOST: Precise Stellar Parameters for 450,000 LAMOST Giants
• SDSS-V: Pioneering Panoptic Spectroscopy
• APOGEE Chemical Abundances of Globular Cluster Giants in the Inner Galaxy
• Principal Component Analysis on Chemical Abundances Spaces
• Galactic Doppelganger: The chemical similarity among field stars and among stars with a common birth origin
• Chemical tagging in the SDSS-III/APOGEE survey: new identifications of halo stars with globular cluster origins
• The Gaia-ESO survey: the non-universality of the age-chemical-clocks-metallicity relations in the Galactic disc
• Chemical Abundances of Main-Sequence, Turnoff, Subgiant and red giant Stars from APOGEE spectra II: Atomic Diffusion in M67 Stars
• Early turbulent mixing as the origin of chemical homogeneity in open star clusters
• In the Galactic disk, stellar [Fe/H] and age predict orbits and precise [X/Fe]
• Spatial variations in the Milky Way disc metallicity-age relation
• Tidal interactions between binary stars drives lithium production in low-mass red giants
• The RAVE-on catalog of stellar atmospheric parameters and chemical abundances for chemo-dynamic studies in the Gaia era
• Chemical tagging can work: Identification of stellar phase-space structures purely by chemical-abundance similarity
• Chemical (in)homogeneity and atomic diffusion in the open cluster M67
• The Gaia-ESO Survey: Evidence of atomic diffusion in M67?
• The Cannon 2: A data-driven model of stellar spectra for detailed chemical abundance analyses
• Cosmic phylogeny: reconstructing the chemical history of the solar neighbourhood with an evolutionary tree
• The Fastest Travel Together: Chemical Tagging of the Fastest Stars in Gaia DR2 to the Stellar Halo
• Abundances in the Milky Way across five nucleosynthetic channels from 4 million LAMOST stars
• Strong chemical tagging with APOGEE: 21 candidate star clusters that have dissolved across the Milky Way disc
• Blind chemical tagging with DBSCAN: prospects for spectroscopic surveys
• The dimensionality of stellar chemical space using spectra from the Apache Point Observatory Galactic Evolution Experiment
• Cycle-StarNet: Bridging the gap between theory and data by leveraging large datasets
• APOGEE chemical tagging constraint on the maximum star cluster mass in the $\alpha$-enhanced Galactic disk
• A Dynamical Model for Clustered Star Formation in the Galactic Disk
• From birth associations to field stars: mapping the small-scale orbit distribution in the Galactic disc
• A phylogenetic approach to chemical tagging. Reassembling open cluster stars
• The GALAH survey: Co-orbiting stars and chemical tagging
• Actions are weak stellar age indicators in the Milky Way disk
• An unsupervised method for identifying $X$-enriched stars directly from spectra: Li in LAMOST
• A Cyclically-Trained Adversarial Network for Invariant Representation Learning

Cited by in corpus (7)

• The homogeneity of the star forming environment of the Milky Way disk over time
• A variational encoder-decoder approach to precise spectroscopic age estimation for large Galactic surveys
• A GALAH View of the Chemical Homogeneity and Ages of Stellar Strings Identified in Gaia
• The Astrophysical Variance in Gaia-RVS Spectra
• Disentangling stellar atmospheric parameters in astronomical spectra using Generative Adversarial Neural Networks
• Functional Data Analysis for Extracting the Intrinsic Dimensionality of Spectra: Application to Chemical Homogeneity in the Open Cluster M67
• Towards model-free stellar chemical abundances. Potential applications in the search for chemically peculiar stars in large spectroscopic surveys