Deriving structural parameters of semi-resolved star clusters

Narbutis, D.; Semionov, D.; Stonkutė, R.; Meulenaer, P. de; Mineikis, T.; Bridžius, A.; Vansevičius, Vladas

doi:10.1051/0004-6361/201322577

Cited by 4 publications

(3 citation statements)

References 19 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…We define r h as the radius of a circle on the sky enclosing half of a cluster's stars. These values at the assumed M31 distance (785 kpc) roughly correspond to real cluster sizes in M31 (Vansevičius et al 2009;Narbutis et al 2014). The setup was chosen to target low mass semi-resolved star clusters in order to demonstrate the CNN's capabilities in low signal-to-noise ratio conditions.…”

Section: Artificial Clustersmentioning

confidence: 99%

Deriving star cluster parameters with convolutional neural networks

Bialopetravičius

Narbutis

2020

A&A

Self Cite

View full text Add to dashboard Cite

Context. Convolutional neural networks (CNNs) have been established as the go-to method for fast object detection and classification on natural images. This opens the door for astrophysical parameter inference on the exponentially increasing amount of sky survey data. Until now, star cluster analysis was based on integral or resolved stellar photometry, which limits the amount of information that can be extracted from individual pixels of cluster images. Aims. We aim to create a CNN capable of inferring star cluster evolutionary, structural, and environmental parameters from multiband images, as well to demonstrate its capabilities in discriminating genuine clusters from galactic stellar backgrounds. Methods. A CNN based on the deep residual network (ResNet) architecture was created and trained to infer cluster ages, masses, sizes, and extinctions, with respect to the degeneracies between them. Mock clusters placed on M83 Hubble Space Telescope (HST) images utilizing three photometric passbands (F336W, F438W, and F814W) were used. The CNN is also capable of predicting the likelihood of a cluster's presence in an image, as well as quantifying its visibility (signal-to-noise).Results. The CNN was tested on mock images of artificial clusters and has demonstrated reliable inference results for clusters of ages 100 Myr, extinctions A V between 0 and 3 mag, masses between 3 × 10 3 and 3 × 10 5 M , and sizes between 0.04 and 0.4 arcsec at the distance of the M83 galaxy. Real M83 galaxy cluster parameter inference tests were performed with objects taken from previous studies and have demonstrated consistent results.

show abstract

Section: Artificial Clustersmentioning

confidence: 99%

Deriving star cluster parameters with convolutional neural networks

Bialopetravičius

Narbutis

2020

A&A

Self Cite

View full text Add to dashboard Cite

show abstract

“…We define r h as the radius of a circle on the sky enclosing half of the stars of a cluster. These values at the assumed M 31 distance (785 kpc) roughly correspond to real cluster sizes in M 31 (Vansevičius et al 2009;Narbutis et al 2014). The setup was chosen to target low-mass semi-resolved star clusters in order to demonstrate the capabilities of the CNN in low-S/N conditions.…”

Section: Artificial Clustersmentioning

confidence: 99%

Deriving star cluster parameters with convolutional neural networks

Bialopetravičius

Narbutis

Vansevičius

2019

A&A

Self Cite

View full text Add to dashboard Cite

Context. Convolutional neural networks (CNNs) have been proven to perform fast classification and detection on natural images and have the potential to infer astrophysical parameters on the exponentially increasing amount of sky-survey imaging data. The inference pipeline can be trained either from real human-annotated data or simulated mock observations. Until now, star cluster analysis was based on integral or individual resolved stellar photometry. This limits the amount of information that can be extracted from cluster images. Aims. We aim to develop a CNN-based algorithm capable of simultaneously deriving ages, masses, and sizes of star clusters directly from multi-band images. We also aim to demonstrate CNN capabilities on low-mass semi-resolved star clusters in a low-signal-to-noise-ratio regime. Methods. A CNN was constructed based on the deep residual network (ResNet) architecture and trained on simulated images of star clusters with various ages, masses, and sizes. To provide realistic backgrounds, M 31 star fields taken from The Panchromatic Hubble Andromeda Treasury (PHAT) survey were added to the mock cluster images. Results. The proposed CNN was verified on mock images of artificial clusters and has demonstrated high precision and no significant bias for clusters of ages ≲3 Gyr and masses between 250 and 4000 M⊙. The pipeline is end-to-end, starting from input images all the way to the inferred parameters; no hand-coded steps have to be performed: estimates of parameters are provided by the neural network in one inferential step from raw images.

show abstract

“…Other promising methods are based on the fitting of observed integrated star cluster magnitudes and colour indices to stochastic theoretical models (Deveikis et al 2008;Fouesneau & Lançon 2010;Fouesneau et al 2014;de Meulenaer et al 2013de Meulenaer et al , 2014de Meulenaer et al , 2015aKrumholz et al 2015;Bialopetravičius et al 2019). These methods look promising for the study of unresolved star clusters to much larger distances than the Local Group galaxies, which could be mimicked using ground based observations of M 31 clusters (Kodaira et al 2004;Narbutis et al 2007aNarbutis et al , 2008Bridžius et al 2008). However, when stochastic models are used, the accuracy of derived parameters strongly depends on the uncertainties of aperture photometry (Narbutis et al 2007b) and the proper account for projecting background and foreground stars (hereinafter field stars).…”

Section: Introductionmentioning

confidence: 99%

Deriving physical parameters of unresolved star clusters

Kriščiūnas,

Daugevičius,

Stonkutė

et al. 2023

A&A

View full text Add to dashboard Cite

Context. This work is the seventh study in a series dedicated to investigating degeneracies of simultaneous age, mass, extinction, and metallicity determinations of partially resolved or unresolved star clusters with Hubble Space Telescope broadband aperture photometry. In the sixth work (hereafter, Paper I), it was demonstrated that the adaptive aperture photometry, performed to avoid the majority of the projected foreground and background stars falling within the apertures, gives more consistent colour indices for star clusters. Aims. In this study, we aim to supplement the homogeneous multi-colour aperture photometry results published in Paper I and provide a complete M 31 Panchromatic Hubble Andromeda Treasury (PHAT) survey star cluster photometry catalogue for further analysis. Methods. Following Paper I, we used a two-aperture approach for photometry. The first aperture is the standard one used to measure total cluster fluxes. The second (smaller) aperture is introduced to avoid the bright foreground and background stars projecting onto the clusters. We selected the radii of smaller apertures to be larger than the half-light radii of the clusters. Results. We present the second part of the star cluster aperture photometry catalogues for a sample of 1477 star clusters from the M 31 PHAT survey not covered in Paper I. Compared to the M 31 PHAT star cluster aperture photometry catalogue published by Johnson et al., adjustments were made to the cluster centre coordinates, aperture sizes, and sky background levels.

show abstract

Deriving structural parameters of semi-resolved star clusters

Cited by 4 publications

References 19 publications

Deriving star cluster parameters with convolutional neural networks

Deriving star cluster parameters with convolutional neural networks

Deriving star cluster parameters with convolutional neural networks

Deriving physical parameters of unresolved star clusters

Contact Info

Product

Resources

About