Volume 643, November 2020
|Number of page(s)||25|
|Section||Cosmology (including clusters of galaxies)|
|Published online||20 November 2020|
The search for galaxy cluster members with deep learning of panchromatic HST imaging and extensive spectroscopy⋆
Department of Physics and Earth Science of the University of Ferrara, Via Saragat 1, 44122 Ferrara, Italy
2 INAF – Astronomical Observatory of Bologna, via Gobetti 93/3, 40129 Bologna, Italy
3 INAF – Astronomical Observatory of Capodimonte, Salita Moiariello 16, 80131 Napoli, Italy
4 Department of Physics of the University of Milano, via Celoria 16, 20133 Milano, Italy
5 Dark Cosmology Centre, Niels Bohr Institute, University of Copenhagen, Lyngbyvej 2, 2100 Copenhagen, Denmark
6 INAF – IASF Milano, Via A. Corti 12, 20133 Milano, Italy
7 Kapteyn Astronomical Institute, University of Groningen, Postbus 800, 9700 AV Groningen, The Netherlands
8 INAF – Astronomical Observatory of Trieste, via G. B. Tiepolo 11, 34131 Trieste, Italy
9 INFN, Sezione di Ferrara, Via Saragat 1, 44122 Ferrara, Italy
Accepted: 16 September 2020
Context. The next generation of extensive and data-intensive surveys are bound to produce a vast amount of data, which can be efficiently dealt with using machine-learning and deep-learning methods to explore possible correlations within the multi-dimensional parameter space.
Aims. We explore the classification capabilities of convolution neural networks (CNNs) to identify galaxy cluster members (CLMs) by using Hubble Space Telescope (HST) images of fifteen galaxy clusters at redshift 0.19 ≲ z ≲ 0.60, observed as part of the CLASH and Hubble Frontier Field programmes.
Methods. We used extensive spectroscopic information, based on the CLASH-VLT VIMOS programme combined with MUSE observations, to define the knowledge base. We performed various tests to quantify how well CNNs can identify cluster members on ht basis of imaging information only. Furthermore, we investigated the CNN capability to predict source memberships outside the training coverage, in particular, by identifying CLMs at the faint end of the magnitude distributions.
Results. We find that the CNNs achieve a purity-completeness rate ≳90%, demonstrating stable behaviour across the luminosity and colour of cluster galaxies, along with a remarkable generalisation capability with respect to cluster redshifts. We concluded that if extensive spectroscopic information is available as a training base, the proposed approach is a valid alternative to catalogue-based methods because it has the advantage of avoiding photometric measurements, which are particularly challenging and time-consuming in crowded cluster cores. As a byproduct, we identified 372 photometric cluster members, with mag(F814) < 25, to complete the sample of 812 spectroscopic members in four galaxy clusters RX J2248-4431, MACS J0416-2403, MACS J1206-0847 and MACS J1149+2223.
Conclusions. When this technique is applied to the data that are expected to become available from forthcoming surveys, it will be an efficient tool for a variety of studies requiring CLM selection, such as galaxy number densities, luminosity functions, and lensing mass reconstruction.
Key words: Galaxy: general / galaxies: photometry / galaxies: distances and redshifts / techniques: image processing / methods: data analysis
The spectroscopic training set and the members identified in the four clusters RX J2248-4431, MACS J0416-2403, MACS J1206-0847 and MACS J1149+2223 are only available at the CDS via anonymous ftp to cdsarc.u-strasbg.fr (184.108.40.206) or via http://cdsarc.u-strasbg.fr/viz-bin/cat/J/A+A/643/A177
© ESO 2020
Current usage metrics show cumulative count of Article Views (full-text article views including HTML views, PDF and ePub downloads, according to the available data) and Abstracts Views on Vision4Press platform.
Data correspond to usage on the plateform after 2015. The current usage metrics is available 48-96 hours after online publication and is updated daily on week days.
Initial download of the metrics may take a while.