-
Searching for changing-state AGNs in massive datasets -- I: applying deep learning and anomaly detection techniques to find AGNs with anomalous variability behaviours
Authors:
P. Sánchez-Sáez,
H. Lira,
L. Martí,
N. Sánchez-Pi,
J. Arredondo,
F. E. Bauer,
A. Bayo,
G. Cabrera-Vives,
C. Donoso-Oliva,
P. A. Estévez,
S. Eyheramendy,
F. Förster,
L. Hernández-García,
A. M. Muñoz Arancibia,
M. Pérez-Carrasco,
M. Sepúlveda,
J. R. Vergara
Abstract:
The classic classification scheme for Active Galactic Nuclei (AGNs) was recently challenged by the discovery of the so-called changing-state (changing-look) AGNs (CSAGNs). The physical mechanism behind this phenomenon is still a matter of open debate and the samples are too small and of serendipitous nature to provide robust answers. In order to tackle this problem, we need to design methods that…
▽ More
The classic classification scheme for Active Galactic Nuclei (AGNs) was recently challenged by the discovery of the so-called changing-state (changing-look) AGNs (CSAGNs). The physical mechanism behind this phenomenon is still a matter of open debate and the samples are too small and of serendipitous nature to provide robust answers. In order to tackle this problem, we need to design methods that are able to detect AGN right in the act of changing-state. Here we present an anomaly detection (AD) technique designed to identify AGN light curves with anomalous behaviors in massive datasets. The main aim of this technique is to identify CSAGN at different stages of the transition, but it can also be used for more general purposes, such as cleaning massive datasets for AGN variability analyses. We used light curves from the Zwicky Transient Facility data release 5 (ZTF DR5), containing a sample of 230,451 AGNs of different classes. The ZTF DR5 light curves were modeled with a Variational Recurrent Autoencoder (VRAE) architecture, that allowed us to obtain a set of attributes from the VRAE latent space that describes the general behaviour of our sample. These attributes were then used as features for an Isolation Forest (IF) algorithm, that is an anomaly detector for a "one class" kind of problem. We used the VRAE reconstruction errors and the IF anomaly score to select a sample of 8,809 anomalies. These anomalies are dominated by bogus candidates, but we were able to identify 75 promising CSAGN candidates.
△ Less
Submitted 12 July, 2021; v1 submitted 14 June, 2021;
originally announced June 2021.
-
The R-Process Alliance: Spectroscopic Follow-up of Low-Metallicity Star Candidates from the Best & Brightest Survey
Authors:
Vinicius M. Placco,
Rafael M. Santucci,
Timothy C. Beers,
Julio Chaname,
Maria Paz Sepulveda,
Johanna Coronado,
Silvia Rossi,
Young Sun Lee,
Else Starkenburg,
Kris Youakim,
Manuel Barrientos,
Rana Ezzeddine,
Anna Frebel,
Terese T. Hansen,
Erika M. Holmbeck,
Alexander P. Ji,
Kaitlin C. Rasmussen,
Ian U. Roederer,
Charli M. Sakari,
Devin D. Whitten
Abstract:
We present results from an observing campaign to identify low-metallicity stars in the Best & Brightest Survey. From medium-resolution (R ~ 1, 200 - 2, 000) spectroscopy of 857 candidates, we estimate the stellar atmospheric parameters (Teff, log g, and [Fe/H]), as well as carbon and alpha-element abundances. We find that 69% of the observed stars have [Fe/H] <= -1.0, 39% have [Fe/H] <= -2.0, and…
▽ More
We present results from an observing campaign to identify low-metallicity stars in the Best & Brightest Survey. From medium-resolution (R ~ 1, 200 - 2, 000) spectroscopy of 857 candidates, we estimate the stellar atmospheric parameters (Teff, log g, and [Fe/H]), as well as carbon and alpha-element abundances. We find that 69% of the observed stars have [Fe/H] <= -1.0, 39% have [Fe/H] <= -2.0, and 2% have [Fe/H] <= -3.0. There are also 133 carbon-enhanced metal-poor (CEMP) stars in this sample, with 97 CEMP Group I and 36 CEMP Group II stars identified in the A(C) versus [Fe/H] diagram. A subset of the confirmed low-metallicity stars were followed-up with high-resolution spectroscopy, as part of the R-process Alliance, with the goal of identifying new highly and moderately r-process-enhanced stars. Comparison between the stellar atmospheric parameters estimated in this work and from high-resolution spectroscopy exhibit good agreement, confirming our expectation that medium-resolution observing campaigns are an effective way of selecting interesting stars for further, more targeted, efforts.
△ Less
Submitted 26 November, 2018;
originally announced November 2018.
-
Spectroscopic Validation of Low-Metallicity Stars from RAVE
Authors:
Vinicius M. Placco,
Timothy C. Beers,
Rafael M. Santucci,
Julio Chaname,
Maria Paz Sepulveda,
Johanna Coronado,
Sean D. Points,
Catherine C. Kaleida,
Silvia Rossi,
Georges Kordopatis,
Young Sun Lee,
Gal Matijevic,
Anna Frebel,
Terese T. Hansen,
Erika M. Holmbeck,
Kaitlin C. Rasmussen,
Ian U. Roederer,
Charli M. Sakari,
Devin D. Whitten
Abstract:
We present results from a medium-resolution (R ~ 2, 000) spectroscopic follow-up campaign of 1,694 bright (V < 13.5), very metal-poor star candidates from the RAdial Velocity Experiment (RAVE). Initial selection of the low-metallicity targets was based on the stellar parameters published in RAVE Data Releases 4 and 5. Follow-up was accomplished with the Gemini-N and Gemini-S, the ESO/NTT, the KPNO…
▽ More
We present results from a medium-resolution (R ~ 2, 000) spectroscopic follow-up campaign of 1,694 bright (V < 13.5), very metal-poor star candidates from the RAdial Velocity Experiment (RAVE). Initial selection of the low-metallicity targets was based on the stellar parameters published in RAVE Data Releases 4 and 5. Follow-up was accomplished with the Gemini-N and Gemini-S, the ESO/NTT, the KPNO/Mayall, and the SOAR telescopes. The wavelength coverage for most of the observed spectra allows for the determination of carbon and α-element abundances, which are crucial for con- sidering the nature and frequency of the carbon-enhanced metal-poor (CEMP) stars in this sample. We find that 88% of the observed stars have [Fe/H] <= -1.0, 61% have [Fe/H] <= -2.0, and 3% have [Fe/H] <= -3.0 (with four stars at [Fe/H] <= -3.5). There are 306 CEMP star candidates in this sample, and we identify 169 CEMP Group I, 131 CEMP Group II, and 6 CEMP Group III stars from the A(C) vs. [Fe/H] diagram. Inspection of the [alpha/C] abundance ratios reveals that five of the CEMP Group II stars can be classified as "mono-enriched second-generation" stars. Gaia DR1 matches were found for 734 stars, and we show that transverse velocities can be used as a confirmatory selection criteria for low-metallicity candidates. Selected stars from our validated list are being followed-up with high-resolution spectroscopy, to reveal their full chemical abundance patterns for further studies.
△ Less
Submitted 1 May, 2018;
originally announced May 2018.
-
A Distant Sample of Halo Wide Binaries from SDSS
Authors:
Johanna Coronado,
María Paz Sepúlveda,
Andrew Gould,
Julio Chanamé
Abstract:
Samples of reliably identified halo wide binaries are scarce. If reasonably free from selection effects and with a small degree of contamination by chance alignments, these wide binaries become a powerful dynamical tool, having provided one of the very few experiments capable of constraining the nature of dark matter in the Milky Way halo. Currently, however, the best available sample of halo wide…
▽ More
Samples of reliably identified halo wide binaries are scarce. If reasonably free from selection effects and with a small degree of contamination by chance alignments, these wide binaries become a powerful dynamical tool, having provided one of the very few experiments capable of constraining the nature of dark matter in the Milky Way halo. Currently, however, the best available sample of halo wide binaries is confined to the solar neighborhood, and is plagued by small number statistics at the widest separations. We present the results of a program aimed to probe the wide binary population of the Galactic halo at significantly larger distances, and which informs future searches that could improve the statistics by orders of magnitude. Halo stars were taken from the Sloan Digital Sky Survey after analysing the Galactic orbits of stars in the reduced proper motion diagram. We then select candidate binaries by searching for pairs with small differences in proper motion and small projected separation on the sky. Using medium-resolution spectroscopy, a subsample of candidates is validated via radial velocities, finding that about 68\% of candidate pairs up to 20$^{\prime\prime}$ separation can be considered genuine halo wide binaries, to the limits of the available data. Precise parallaxes from \textit{Gaia} confirm that most of our selected pairs have their components at the same distances, independently confirming the robustness of our selection method. These results should prove valuable to guide the optimal assembly of larger catalogs of halo wide binaries from upcoming large databases, e.g., \textit{Gaia} and LSST.
△ Less
Submitted 16 August, 2018; v1 submitted 18 April, 2018;
originally announced April 2018.
