Assessing the performance of LTE and NLTE synthetic stellar spectra in a machine learning framework

Abstract

In the current era of stellar spectroscopic surveys, synthetic spectral libraries are the basis for the derivation of stellar parameters and chemical abundances. In this paper, we compare the stellar parameters determined using five popular synthetic spectral grids (INTRIGOSS, FERRE, AMBRE, PHOENIX, and MPIA/1DNLTE) with our convolutional neural network (CNN, $\text{StarNet}$). The stellar parameters are determined for six physical properties (effective temperature, surface gravity, metallicity, [$\alpha$/Fe], radial velocity, and rotational velocity) given the spectral resolution, signal-to-noise, and wavelength range of optical FLAMES-UVES spectra from the Gaia-ESO Survey. Both CNN modelling and epistemic uncertainties are incorporated through training an ensemble of networks. $\text{StarNet}$ training was also adapted to mitigate differences between the synthetic grids and observed spectra by augmenting with realistic observational signatures (i.e. resolution matching, wavelength sampling, Gaussian noise, zeroing flux values, rotational and radial velocities, continuum removal, and masking telluric regions). Using the FLAMES-UVES spectra for FGK type dwarfs and giants as a test set, we quantify the accuracy and precision of the stellar label predictions from $\text{StarNet}$. We find excellent results over a wide range of parameters when $\text{StarNet}$ is trained on the MPIA/1DNLTE synthetic grid, and acceptable results over smaller parameter ranges when trained on the 1DLTE grids. These tests also show that our CNN pipeline is highly adaptable to multiple simulation grids.

20 pages, 17 figures, published by MNRAS. Code for StarNet available at https://github.com/Spiffical/StarNet

References in corpus (28)

• The Radial Velocity Experiment (RAVE): first data release
• 4MOST: Project overview and information for the First Call for Proposals
• Gaia-ESO Survey: The analysis of high-resolution UVES spectra of FGK-type stars
• NLTE abundances of Mn in a sample of metal-poor stars
• Abundance Estimates for 16 Elements in 6 Million Stars from LAMOST DR5 Low-Resolution Spectra
• A non-LTE study of neutral and singly-ionized calcium in late-type stars
• Non-LTE line formation of Fe in late-type stars - III. 3D non-LTE analysis of metal-poor stars
• The Sixth Data Release of the Radial Velocity Experiment (RAVE) -- II: Stellar Atmospheric Parameters, Chemical Abundances and Distances
• Photometric redshift estimation via deep learning
• Automated derivation of stellar atmospheric parameters and chemical abundances: the MATISSE algorithm
• The Gaia FGK Benchmark Stars - High resolution spectral library
• Center-to-Limb Variation of Solar 3-D Hydrodynamical Simulations
• A New Catalog of Globular Clusters in the Milky Way
• Signatures of unresolved binaries in stellar spectra: implications for spectral fitting
• Chemo-dynamics of outer halo dwarf stars, including \textit{Gaia}-Sausage and \textit{Gaia}-Sequoia candidates
• The Pristine Survey -- VI. The first three years of medium-resolution follow-up spectroscopy of Pristine EMP star candidates
• The Gaia-ESO Survey: double, triple and quadruple-line spectroscopic binary candidates
• The Gaia-ESO Survey: calibration strategy
• Red Supergiant Stars as Cosmic Abundance Probes. III. NLTE effects in J-band Magnesium lines
• Pristine IX: CFHT ESPaDOnS Spectroscopic Analysis of 115 Bright Metal-Poor Candidate Stars
• The Gaia-ESO Survey: detection and characterization of single line spectroscopic binaries
• BONNSAI: correlated stellar observables in Bayesian methods
• Testing Stellar Population Fitting Ingredients with Globular Clusters I: Stellar Libraries
• Influence of inelastic collisions with hydrogen atoms on the non-LTE line formation for Fe I and Fe II in the 1D model atmospheres of late-type stars
• Forecasting Chemical Abundance Precision for Extragalactic Stellar Archaeology
• Carbon-enhanced metal-poor stars in the SDSS-APOGEE database
• The Peculiar Globular Cluster Palomar 1 and Persistence in the SDSS-APOGEE Database
• 4MOST Survey Strategy Plan

Cited by in corpus (15)

• The RAdial Velocity Experiment (RAVE): Parameterisation of RAVE spectra based on convolutional neural networks
• Stellar Populations With Optical Spectra: Deep Learning vs. Popular Spectrum Fitting Codes
• Stellar Parameters and Chemical Abundances Estimated from LAMOST-II DR8 MRS based on Cycle-StarNet
• The Gaia-ESO Survey: Preparing the ground for 4MOST & WEAVE galactic surveys. Chemical evolution of lithium with machine learning
• Artificial Intelligence Assisted Inversion (AIAI) of Synthetic Type Ia Supernova Spectra
• Line-depth ratios as indicators of effective temperature and surface gravity
• A Model RRNet for Spectral Information Exploitation and LAMOST Medium-resolution Spectrum Parameter Estimation
• Galaxy Spectra neural Network (GaSNet). II. Using Deep Learning for Spectral Classification and Redshift Predictions
• Reliable stellar abundances of individual stars with the MUSE integral-field spectrograph
• SpectroTranslator: a deep-neural network algorithm to homogenize spectroscopic parameters
• The CARMENES search for exoplanets around M dwarfs -- A deep transfer learning method to determine Teff and [M/H] of target stars
• Parameter Estimation of LAMOST Medium-resolution Stellar Spectra
• StarUnLink: identifying and mitigating signals from communications satellites in stellar spectral surveys
• Interpreting Stellar Spectra with Unsupervised Domain Adaptation
• The SAPP pipeline for the determination of stellar abundances and atmospheric parameters of stars in the core program of the PLATO mission