The High Cadence Transit Survey (HiTS): Compilation and Characterization of Light-curve Catalogs

Jorge Martínez-Palomera, Francisco Förster, Pavlos Protopapas, Juan Carlos Maureira, Paulina Lira, Guillermo Cabrera-Vives, Pablo Huijse, Lluis Galbany, Thomas De Jaeger, Santiago González-Gaitán, Gustavo Medina, Giuliano Pignata, Jaime San Martín, Mario Hamuy, Ricardo R. Muñoz

Resultado de la investigación: Article

1 Cita (Scopus)

Resumen

The High Cadence Transient Survey (HiTS) aims to discover and study transient objects with characteristic timescales between hours and days, such as pulsating, eclipsing, and exploding stars. This survey represents a unique laboratory to explore large etendue observations from cadences of about 0.1 days and test new computational tools for the analysis of large data. This work follows a fully data science approach, from the raw data to the analysis and classification of variable sources. We compile a catalog of ∼15 million object detections and a catalog of ∼2.5 million light curves classified by variability. The typical depth of the survey is 24.2, 24.3, 24.1, and 23.8 in the u, g, r, and i bands, respectively. We classified all point-like nonmoving sources by first extracting features from their light curves and then applying a random forest classifier. For the classification, we used a training set constructed using a combination of cross-matched catalogs, visual inspection, transfer/active learning, and data augmentation. The classification model consists of several random forest classifiers organized in a hierarchical scheme. The classifier accuracy estimated on a test set is approximately 97%. In the unlabeled data, 3485 sources were classified as variables, of which 1321 were classified as periodic. Among the periodic classes, we discovered with high confidence one δ Scuti, 39 eclipsing binaries, 48 rotational variables, and 90 RR Lyrae, and for the nonperiodic classes, we discovered one cataclysmic variable, 630 QSOs, and one supernova candidate. The first data release can be accessed in the project archive of HiTS (http://astro.cmm.uchile.cl/HiTS/).

Idioma originalEnglish
Número de artículo186
PublicaciónAstronomical Journal
Volumen156
N.º5
DOI
EstadoPublished - 1 nov 2018

Huella dactilar

transit
light curve
catalogs
classifiers
cataclysmic variables
quasars
learning
supernovae
inspection
confidence
education
stars
augmentation
timescale

ASJC Scopus subject areas

  • Astronomy and Astrophysics
  • Space and Planetary Science

Citar esto

Martínez-Palomera, J., Förster, F., Protopapas, P., Maureira, J. C., Lira, P., Cabrera-Vives, G., ... Muñoz, R. R. (2018). The High Cadence Transit Survey (HiTS): Compilation and Characterization of Light-curve Catalogs. Astronomical Journal, 156(5), [186]. https://doi.org/10.3847/1538-3881/aadfd8
Martínez-Palomera, Jorge ; Förster, Francisco ; Protopapas, Pavlos ; Maureira, Juan Carlos ; Lira, Paulina ; Cabrera-Vives, Guillermo ; Huijse, Pablo ; Galbany, Lluis ; Jaeger, Thomas De ; González-Gaitán, Santiago ; Medina, Gustavo ; Pignata, Giuliano ; Martín, Jaime San ; Hamuy, Mario ; Muñoz, Ricardo R. / The High Cadence Transit Survey (HiTS) : Compilation and Characterization of Light-curve Catalogs. En: Astronomical Journal. 2018 ; Vol. 156, N.º 5.
@article{6c781ced5b384d05947604499a953c6d,
title = "The High Cadence Transit Survey (HiTS): Compilation and Characterization of Light-curve Catalogs",
abstract = "The High Cadence Transient Survey (HiTS) aims to discover and study transient objects with characteristic timescales between hours and days, such as pulsating, eclipsing, and exploding stars. This survey represents a unique laboratory to explore large etendue observations from cadences of about 0.1 days and test new computational tools for the analysis of large data. This work follows a fully data science approach, from the raw data to the analysis and classification of variable sources. We compile a catalog of ∼15 million object detections and a catalog of ∼2.5 million light curves classified by variability. The typical depth of the survey is 24.2, 24.3, 24.1, and 23.8 in the u, g, r, and i bands, respectively. We classified all point-like nonmoving sources by first extracting features from their light curves and then applying a random forest classifier. For the classification, we used a training set constructed using a combination of cross-matched catalogs, visual inspection, transfer/active learning, and data augmentation. The classification model consists of several random forest classifiers organized in a hierarchical scheme. The classifier accuracy estimated on a test set is approximately 97{\%}. In the unlabeled data, 3485 sources were classified as variables, of which 1321 were classified as periodic. Among the periodic classes, we discovered with high confidence one δ Scuti, 39 eclipsing binaries, 48 rotational variables, and 90 RR Lyrae, and for the nonperiodic classes, we discovered one cataclysmic variable, 630 QSOs, and one supernova candidate. The first data release can be accessed in the project archive of HiTS (http://astro.cmm.uchile.cl/HiTS/).",
keywords = "catalogs, methods: data analysis, stars: variables: general, surveys, techniques: photometric",
author = "Jorge Mart{\'i}nez-Palomera and Francisco F{\"o}rster and Pavlos Protopapas and Maureira, {Juan Carlos} and Paulina Lira and Guillermo Cabrera-Vives and Pablo Huijse and Lluis Galbany and Jaeger, {Thomas De} and Santiago Gonz{\'a}lez-Gait{\'a}n and Gustavo Medina and Giuliano Pignata and Mart{\'i}n, {Jaime San} and Mario Hamuy and Mu{\~n}oz, {Ricardo R.}",
year = "2018",
month = "11",
day = "1",
doi = "10.3847/1538-3881/aadfd8",
language = "English",
volume = "156",
journal = "Astronomical Journal",
issn = "0004-6256",
publisher = "IOP Publishing Ltd.",
number = "5",

}

Martínez-Palomera, J, Förster, F, Protopapas, P, Maureira, JC, Lira, P, Cabrera-Vives, G, Huijse, P, Galbany, L, Jaeger, TD, González-Gaitán, S, Medina, G, Pignata, G, Martín, JS, Hamuy, M & Muñoz, RR 2018, 'The High Cadence Transit Survey (HiTS): Compilation and Characterization of Light-curve Catalogs', Astronomical Journal, vol. 156, n.º 5, 186. https://doi.org/10.3847/1538-3881/aadfd8

The High Cadence Transit Survey (HiTS) : Compilation and Characterization of Light-curve Catalogs. / Martínez-Palomera, Jorge; Förster, Francisco; Protopapas, Pavlos; Maureira, Juan Carlos; Lira, Paulina; Cabrera-Vives, Guillermo; Huijse, Pablo; Galbany, Lluis; Jaeger, Thomas De; González-Gaitán, Santiago; Medina, Gustavo; Pignata, Giuliano; Martín, Jaime San; Hamuy, Mario; Muñoz, Ricardo R.

En: Astronomical Journal, Vol. 156, N.º 5, 186, 01.11.2018.

Resultado de la investigación: Article

TY - JOUR

T1 - The High Cadence Transit Survey (HiTS)

T2 - Compilation and Characterization of Light-curve Catalogs

AU - Martínez-Palomera, Jorge

AU - Förster, Francisco

AU - Protopapas, Pavlos

AU - Maureira, Juan Carlos

AU - Lira, Paulina

AU - Cabrera-Vives, Guillermo

AU - Huijse, Pablo

AU - Galbany, Lluis

AU - Jaeger, Thomas De

AU - González-Gaitán, Santiago

AU - Medina, Gustavo

AU - Pignata, Giuliano

AU - Martín, Jaime San

AU - Hamuy, Mario

AU - Muñoz, Ricardo R.

PY - 2018/11/1

Y1 - 2018/11/1

N2 - The High Cadence Transient Survey (HiTS) aims to discover and study transient objects with characteristic timescales between hours and days, such as pulsating, eclipsing, and exploding stars. This survey represents a unique laboratory to explore large etendue observations from cadences of about 0.1 days and test new computational tools for the analysis of large data. This work follows a fully data science approach, from the raw data to the analysis and classification of variable sources. We compile a catalog of ∼15 million object detections and a catalog of ∼2.5 million light curves classified by variability. The typical depth of the survey is 24.2, 24.3, 24.1, and 23.8 in the u, g, r, and i bands, respectively. We classified all point-like nonmoving sources by first extracting features from their light curves and then applying a random forest classifier. For the classification, we used a training set constructed using a combination of cross-matched catalogs, visual inspection, transfer/active learning, and data augmentation. The classification model consists of several random forest classifiers organized in a hierarchical scheme. The classifier accuracy estimated on a test set is approximately 97%. In the unlabeled data, 3485 sources were classified as variables, of which 1321 were classified as periodic. Among the periodic classes, we discovered with high confidence one δ Scuti, 39 eclipsing binaries, 48 rotational variables, and 90 RR Lyrae, and for the nonperiodic classes, we discovered one cataclysmic variable, 630 QSOs, and one supernova candidate. The first data release can be accessed in the project archive of HiTS (http://astro.cmm.uchile.cl/HiTS/).

AB - The High Cadence Transient Survey (HiTS) aims to discover and study transient objects with characteristic timescales between hours and days, such as pulsating, eclipsing, and exploding stars. This survey represents a unique laboratory to explore large etendue observations from cadences of about 0.1 days and test new computational tools for the analysis of large data. This work follows a fully data science approach, from the raw data to the analysis and classification of variable sources. We compile a catalog of ∼15 million object detections and a catalog of ∼2.5 million light curves classified by variability. The typical depth of the survey is 24.2, 24.3, 24.1, and 23.8 in the u, g, r, and i bands, respectively. We classified all point-like nonmoving sources by first extracting features from their light curves and then applying a random forest classifier. For the classification, we used a training set constructed using a combination of cross-matched catalogs, visual inspection, transfer/active learning, and data augmentation. The classification model consists of several random forest classifiers organized in a hierarchical scheme. The classifier accuracy estimated on a test set is approximately 97%. In the unlabeled data, 3485 sources were classified as variables, of which 1321 were classified as periodic. Among the periodic classes, we discovered with high confidence one δ Scuti, 39 eclipsing binaries, 48 rotational variables, and 90 RR Lyrae, and for the nonperiodic classes, we discovered one cataclysmic variable, 630 QSOs, and one supernova candidate. The first data release can be accessed in the project archive of HiTS (http://astro.cmm.uchile.cl/HiTS/).

KW - catalogs

KW - methods: data analysis

KW - stars: variables: general

KW - surveys

KW - techniques: photometric

UR - http://www.scopus.com/inward/record.url?scp=85056645379&partnerID=8YFLogxK

U2 - 10.3847/1538-3881/aadfd8

DO - 10.3847/1538-3881/aadfd8

M3 - Article

AN - SCOPUS:85056645379

VL - 156

JO - Astronomical Journal

JF - Astronomical Journal

SN - 0004-6256

IS - 5

M1 - 186

ER -

Martínez-Palomera J, Förster F, Protopapas P, Maureira JC, Lira P, Cabrera-Vives G y otros. The High Cadence Transit Survey (HiTS): Compilation and Characterization of Light-curve Catalogs. Astronomical Journal. 2018 nov 1;156(5). 186. https://doi.org/10.3847/1538-3881/aadfd8