Stellar labels for hot stars from low-resolution spectra - I. the HotPayne method and results for 330,000 stars from LAMOST DR6

Abstract

We set out to determine stellar labels from low-resolution survey spectra of hot stars, specifically OBA stars with T eff 7500 K. This fills a gap in the scientific analysis of large spectroscopic stellar surveys such as LAMOST, which offers spectra for millions of stars at R ∼ 1800 and covering 3800Å ≤ λ ≤ 9000Å. We first explore the theoretical information content of such spectra for determining stellar labels, via the Cramér-Rao bound. We show that in the limit of perfect model spectra and observed spectra with S/N∼ 50 − 100, precise estimates are possible for a wide range of stellar labels: not only the effective temperature T eff , surface gravity log g , and projected rotation velocity vsini , but also the micro-turbulence velocity v mic , Helium abundance N He /N tot and the elemental abundances [C/H], [N/H], [O/H], [Si/H], [S/H], and [Fe/H]. Our analysis illustrates that the temperature regime of T eff ∼ 9500 K is challenging, as the dominant Balmer and Paschen line strength vary little with T eff . We implement the simultaneous fitting of these 11 stellar labels to LAMOST hot-star spectra using the PAYNE approach, drawing on Kurucz's ATLAS12/SYNTHE LTE spectra as the underlying models. We then obtain stellar parameter estimates for a sample of about 330,000 hot stars with LAMOST spectra, an increase by about two orders of magnitude in sample size. Among them, about 260,000 have S /N > 5 Gaia parallaxes, and their luminosities imply that 95% among them are luminous star, mostly on the main sequence; the rest reflects lower luminosity evolved stars, such as hot subdwarfs and white dwarfs. We show that the fidelity of the results, particularly for the abundance estimates, is limited by the systematics of the underlying models, as they do not account for NLTE effects. Finally, we show the detailed distribution of v sin i of stars with 8000 − 15, 000 K, illustrating that it extends to a sharp cut-off at the critical rotation velocity, v crit , across a wide range of temperatures.

Zeta-Payne: A Fully Automated Spectrum Analysis Algorithm for the Milky Way Mapper Program of the SDSS-V Survey

“…We conclude that as long as the LSF of the instrument is properly accounted for in the analysis of stellar spectra, atmospheric parameters of OBAF-type stars can still be accurately inferred from low-resolution optical spectra, albeit with typically lower precision than from their high-resolution optical spectra. Xiang et al (2021) came to similar conclusions from the analysis of LAMOST spectra of some 330,000 OBA-type stars and validation of their HOTPAYNE algorithm based on medium-to high-resolution spectra in the literature. 5.…”

“…As was briefly mentioned in Section 1, we aim to develop an efficient spectrum analysis algorithm for the MWM survey of SDSS-V, with the primary focus on intermediate-to high-mass stars of spectral types O, B, A, and F. One of the main requirements at this stage is a wide applicability range of the algorithm in terms of the wavelength coverage and resolving power of the obtained spectra, and without the need for human intervention between data reduction pipelines and spectrum analysis itself. Here, we use the heritage of one of the publicly available ML-based methods, namely the (HOT)PAYNE algorithm (Ting et al 2019;Xiang et al 2021), to which we introduce a number of modifications to comply with the aforementioned primary requirement of MWM.…”

“…The Cannon algorithm is often referred to as a "data-driven approach" owing to the limited model dependency that occurs only at the stage of the algorithm training. Another example of, this time, a fully modeldependent ML-based approach is the (HOT)PAYNE (Ting et al 2019;Xiang et al 2021) algorithm that employs a neural network as an efficient predictor of (continuum-normalized) synthetic spectra. When teamed up with an optimization algorithm, the approach offers a fast way to analyze large volumes of spectroscopic data in the parameter space the neural network has been trained in.…”

“…In this paper, we build upon the legacy of the (HOT)PAYNE algorithm (Ting et al 2019;Xiang et al 2021) to develop an ML-based spectrum analysis method for the MWM program of the SDSS-V survey, with a particular focus on intermediateand high-mass stars of spectral types O, B, A, and F, irrespective of whether those are observed with the BOSS low-resolution optical or the APOGEE medium-resolution near-IR instrument. Our algorithm is implemented in the ASTRA software framework and is employed for the analysis of spectra of OBAF-type stars targeted by MWM.…”

Overview of the LAMOST survey in the first decade

“…Using LAMOST DR5, the largest O/B-type star catalog of 16,032 such stars has been established, 88 for which key physical parameters have also been derived. 89 A number of peculiar O-type stars were simultaneously discovered, such as the very rare Oe stars, which present O-type spectra with the emission of H lines but without N III l4634À4640À4642 or He II l4686 emission features. 90 Stars with intrinsic brightness variation.…”