We present The Payne, a general method for the precise and simultaneous determination of numerous stellar labels from observed spectra, based on fitting physical spectral models. The Payne combines a ...number of important methodological aspects: it exploits the information from much of the available spectral range; it fits all labels (stellar parameters and elemental abundances) simultaneously; it uses spectral models, where the structure of the atmosphere and the radiative transport are consistently calculated to reflect the stellar labels. At its core The Payne has an approach to accurate and precise interpolation and prediction of the spectrum in high-dimensional label space that is flexible and robust, yet based on only a moderate number of ab initio models ( for 25 labels). With a simple neural-net-like functional form and a suitable choice of training labels, this interpolation yields a spectral flux prediction good to 10−3 rms across a wide range of Teff and (including dwarfs and giants). We illustrate the power of this approach by applying it to the APOGEE DR14 data set, drawing on Kurucz models with recently improved line lists: without recalibration, we obtain physically sensible stellar parameters as well as 15 elemental abundances that appear to be more precise than the published APOGEE DR14 values. In short, The Payne is an approach that for the first time combines all these key ingredients, necessary for progress toward optimal modeling of survey spectra; and it leads to both precise and accurate estimates of stellar labels, based on physical models and without "recalibration." Both the codes and catalog are made publicly available online.
In the ΛCDM paradigm, the Galactic stellar halo is predicted to harbor the accreted debris of smaller systems. To identify these systems, the H3 Spectroscopic Survey, combined with Gaia, is gathering ...6D phase-space and chemical information in the distant Galaxy. Here we present a comprehensive inventory of structure within 50 kpc from the Galactic center using a sample of 5684 giants at and . We identify known structures including the high- disk, the in situ halo (disk stars heated to eccentric orbits), Sagittarius (Sgr), Gaia-Sausage-Enceladus (GSE), the Helmi Streams, Sequoia, and Thamnos. Additionally, we identify the following new structures: (i) Aleph (Fe/H = −0.5), a low-eccentricity structure that rises a surprising 10 kpc off the plane, (ii) and (iii) Arjuna (Fe/H = −1.2) and I'itoi (Fe/H < −2), which comprise the high-energy retrograde halo along with Sequoia, and (iv) Wukong (Fe/H = −1.6), a prograde phase-space overdensity chemically distinct from GSE. For each structure, we provide Fe/H, /Fe, and orbital parameters. Stars born within the Galaxy are a major component at ( 60%), but their relative fraction declines sharply to 5% past 15 kpc. Beyond 15 kpc, >80% of the halo is built by two massive (M ∼ 108-109M ) accreted dwarfs: GSE (Fe/H = −1.2) within 25 kpc and Sgr (Fe/H = −1.0) beyond 25 kpc. This explains the relatively high overall metallicity of the halo (Fe/H −1.2). We attribute 95% of the sample to one of the listed structures, pointing to a halo built entirely from accreted dwarfs and heating of the disk.
We develop and apply a model to quantify the global efficiency of radial orbit migration among stars in the Milky Way disk. This model parameterizes the possible star formation and enrichment ...histories and radial birth profiles, and combines them with a migration model that relates present-day orbital radii to birth radii through a Gaussian probability, broadening with age τ as . Guided by observations, we assume that stars are born with an initially tight age-metallicity relation at given radius, which becomes subsequently scrambled by radial orbit migration, thereby providing a direct observational constraint on radial orbit migration strength . We fit this model with Markov Chain Monte Carlo sampling of the observed age-metallicity distribution of low- red clump stars with Galactocentric radii between 5 and 14 kpc from APOGEE DR12, sidestepping the complex spatial selection function and accounting for the considerable age uncertainties. This simple model reproduces the observed data well, and we find a global (in radius and time) radial orbit migration efficiency in the Milky Way of = 3.6 0.1 kpc when marginalizing over all other aspects of the model. This shows that radial orbit migration in the Milky Way's main disk is indeed rather strong, in line with theoretical expectations: stars migrate by about a half-mass radius over the age of the disk. The model finds the Sun's birth radius at ∼5.2 kpc. If such strong radial orbit migration is typical, this mechanism indeed plays an important role in setting the structural regularity of disk galaxies.
A star in the Milky Way's disk can now be at a Galactocentric radius quite distant from its birth radius for two reasons: either its orbit has become eccentric through radial heating, which increases ...its radial action JR ("blurring"), or merely its angular momentum Lz has changed and thereby its guiding radius ("churning"). We know that radial orbit migration is strong in the Galactic low- disk and set out to quantify the relative importance of these two effects, by devising and applying a parameterized model ( ) for the distribution in the stellar disk. This model describes the orbit evolution for stars of age τ and metallicity , presuming that coeval stars were initially born on (near-)circular orbits, and with a unique at a given birth angular momentum and age. We fit this model to APOGEE red clump stars, accounting for the complex selection function of the survey. The best-fit model implies changes of angular momentum of and changes of radial action as at 8 kpc. This suggests that the secular orbit evolution of the disk is dominated by diffusion in angular momentum, with radial heating being an order of magnitude lower.
It has long been known that the vertical motions of Galactic disk stars increase with stellar age, commonly interpreted as vertical heating through orbit scattering. Here we map the vertical actions ...of disk stars as a function of age ( ) and across a large range of Galactocentric radii, , drawing on APOGEE and Gaia data. We fit as a combination of the vertical action at birth, , and the subsequent heating , which scales as . The inferred birth temperature, is for , consistent with the ISM velocity dispersion, but it rapidly rises outward, to for , likely reflecting the stars' birth in a warped or flared gas disk. We find the heating rate to be modest and nearly constant across all radii, . The stellar age dependence γ gently grows with Galactocentric radius, from for to at . The observed relation at all radii is considerably steeper ( ) than the time dependence theoretically expected from orbit scattering, . We illustrate how this conundrum can be resolved if we also account for the fact that at earlier epochs, the scatterers were more common, and the restoring force from the stellar disk surface mass density was low. Our analysis may reinstate gradual orbital scattering as a plausible and viable mechanism to explain the age-dependent vertical motions of disk stars.
ABSTRACT
Parameter estimation with non-Gaussian stochastic fields is a common challenge in astrophysics and cosmology. In this paper, we advocate performing this task using the scattering transform, ...a statistical tool sharing ideas with convolutional neural networks (CNNs) but requiring neither training nor tuning. It generates a compact set of coefficients, which can be used as robust summary statistics for non-Gaussian information. It is especially suited for fields presenting localized structures and hierarchical clustering, such as the cosmological density field. To demonstrate its power, we apply this estimator to a cosmological parameter inference problem in the context of weak lensing. On simulated convergence maps with realistic noise, the scattering transform outperforms classic estimators and is on a par with the state-of-the-art CNN. It retains advantages of traditional statistical descriptors, has provable stability properties, allows to check for systematics, and importantly, the scattering coefficients are interpretable. It is a powerful and attractive estimator for observational cosmology and the study of physical fields in general.
The Inside-out Growth of the Galactic Disk Frankel, Neige; Sanders, Jason; Rix, Hans-Walter ...
Astrophysical journal/The Astrophysical journal,
10/2019, Letnik:
884, Številka:
2
Journal Article
Recenzirano
Odprti dostop
We quantify the inside-out growth of the Milky Way's low- stellar disk, modeling the ages, metallicities, and Galactocentric radii of APOGEE red clump stars with 6 kpc < R < 13 kpc. The current ...stellar distribution differs significantly from that expected from the star formation history due to the redistribution of stars through radial orbit mixing. We propose and fit a global model for the Milky Way disk, specified by an inside-out star formation history, radial orbit mixing, and an empirical, parametric model for its chemical evolution. We account for the spatially complex survey selection function, and find that the model fits all data well. We find distinct inside-out growth of the Milky Way disk; the best-fit model implies that the half-mass radius of the Milky Way disk has grown by 43% over the last 7 Gyr. Yet, such inside-out growth still results in a present-day age gradient weaker than 0.1 Gyr kpc−1. Our model predicts the half-mass and half-light sizes of the Galactic disk at earlier epochs, which can be compared to the observed redshift-size relations of disk galaxies. We show that radial orbit migration can reconcile the distinct disk-size evolution with redshift, also expected from cosmological simulations, with the modest present-day age gradients seen in the Milky Way and other galaxies.
We measure carbon and nitrogen abundances to a precision of dex for 450,000 giant stars from their low-resolution ( ) LAMOST DR2 survey spectra. We use these and measurements, together with empirical ...relations based on the APOKASC sample, to infer stellar masses and implied ages for 230,000 of these objects to 0.08 dex and 0.2 dex respectively. We use The Cannon, a data-driven approach to spectral modeling, to construct a predictive model for LAMOST spectra. Our reference set comprises 8125 stars observed in common between the APOGEE and LAMOST surveys, taking seven APOGEE DR12 labels (parameters) as ground truth: , , , , , , and . We add seven colors to the Cannon model, based on the g, r, i, J, H, K, W1, W2 magnitudes from APASS, 2MASS, and WISE, which improves our constraints on and by up to 20% and on by up to 70%. Cross-validation of the model demonstrates that, for high- objects, our inferred labels agree with the APOGEE values to within 50 K in temperature, 0.04 mag in , and dex in , , , , and . We apply the model to 450,000 giants in LAMOST DR2 that have not been observed by APOGEE. This demonstrates that precise individual abundances can be measured from low-resolution spectra and represents the largest catalog to date of homogeneous stellar , , masses, and ages. As a result, we greatly increase the number and sky coverage of stars with mass and age estimates.
Core helium-burning red clump (RC) stars are excellent standard candles in the Milky Way. These stars may have more precise distance estimates from spectrophotometry than from Gaia parallaxes beyond ...3 kpc. However, RC stars have values of Teff and that are very similar to some red giant branch (RGB) stars. Especially for low-resolution spectroscopic studies where Teff, , and Fe/H can only be estimated with limited precision, separating RC stars from RGB through established methods can incur ∼20% contamination. Recently, Hawkins et al. demonstrated that the additional information in single-epoch spectra, such as the C/N ratio, can be exploited to cleanly differentiate RC and RGB stars. In this second paper of the series, we establish a data-driven mapping from spectral flux space to independently determined asteroseismic parameters, the frequency and the period spacing. From this, we identify 210,371 RC stars from the publicly available LAMOST DR3 and APOGEE DR14 data, with ∼9% of contamination. We provide an RC sample of 92249 stars with a contamination of only ∼3%, by restricting the combined analysis to LAMOST stars with S/Npix ≥ 75. This demonstrates that high-signal-to-noise ratio (S/N), low-resolution spectra covering a broad wavelength range can identify RC samples at least as pristine as their high-resolution counterparts. As coming and ongoing surveys such as TESS, DESI, and LAMOST will continue to improve the overlapping training spectroscopic-asteroseismic sample, the method presented in this study provides an efficient and straightforward way to derive a vast yet pristine sample of RC stars to reveal the three-dimensional (3D) structure of the Milky Way.
How Many Elements Matter? Ting, Yuan-Sen; Weinberg, David H.
Astrophysical journal/The Astrophysical journal,
03/2022, Letnik:
927, Številka:
2
Journal Article
Recenzirano
Odprti dostop
Abstract
Some studies of stars’ multielement abundance distributions suggest at least 5–7 significant dimensions, but others show that many elemental abundances can be predicted to high accuracy from ...Fe/H and Mg/Fe (or Fe/H and age) alone. We show that both propositions can be, and are, simultaneously true. We adopt a machine-learning technique known as normalizing flow to reconstruct the probability distribution of Milky Way disk stars in the space of 15 elemental abundances measured by APOGEE. Conditioning on
T
eff
and
log
g
minimizes the differential systematics. After further conditioning on Fe/H and Mg/Fe, the residual scatter for most abundances is
σ
X
/H
≲ 0.02 dex, consistent with APOGEE’s reported statistical uncertainties of ∼0.01–0.015 dex and intrinsic scatter of 0.01–0.02 dex. Despite the small scatter, residual abundances display clear correlations between elements, which we show are too large to be explained by measurement uncertainties or by the finite sampling noise. We must condition on at least seven elements to reduce the correlations to a level consistent with the observational uncertainties. Our results demonstrate that cross-element correlations are a much more sensitive probe of a hidden structure than dispersion, and they can be measured precisely in a large sample even if the star-by-star measurement noise is comparable to the intrinsic scatter. We conclude that many elements have an independent story to tell, even for the
mundane
disk stars and elements produced by the core-collapse and Type Ia supernovae. The only way to learn these lessons is to measure the abundances directly, and not merely infer them.
PDF
