SILVERRUSH X: Machine Learning-aided Selection of 9318 LAEs at z = 2.2, 3.3, 4.9, 5.7, 6.6, and 7.0 from the HSC SSP and CHORUS Survey Data

Abstract

We present a new catalog of 9318 Lyα emitter (LAE) candidates at z = 2.2, 3.3, 4.9, 5.7, 6.6, and 7.0 that are photometrically selected by the SILVERRUSH program with a machine learning technique from large area (up to 25.0 deg2) imaging data with six narrowband filters taken by the Subaru Strategic Program with Hyper Suprime-Cam and a Subaru intensive program, Cosmic HydrOgen Reionization Unveiled with Subaru. We construct a convolutional neural network that distinguishes between real LAEs and contaminants with a completeness of 94% and a contamination rate of 1%, enabling us to efficiently remove contaminants from the photometrically selected LAE candidates. We confirm that our LAE catalogs include 177 LAEs that have been spectroscopically identified in our SILVERRUSH programs and previous studies, ensuring the validity of our machine learning selection. In addition, we find that the object-matching rates between our LAE catalogs and our previous results are ≃80%–100% at bright NB magnitudes of ≲24 mag. We also confirm that the surface number densities of our LAE candidates are consistent with previous results. Our LAE catalogs will be made public on our project webpage.

SILVERRUSH. XI. Intensity Mapping for Lya Emission Extending over $100-1000$ comoving kpc around $z\sim2-7$ LAEs with Subaru HSC-SSP and CHORUS Data

“…Although the completeness of the LAEs is as high as 90 % at m NB 24.5 in the CNN of Ono et al (2021), faint LAEs may be missed in the observations and selection. To ensure completeness, we use only LAEs whose L Lyα values are larger than the modes (peaks) of the L Lyα histograms, which are represented as L min Lyα in Table 2.…”

“…We use the LAE catalog constructed by Ono et al (2021) as a part of the Systematic Identification of LAEs for Visible Exploration and Reionization Research Using Subaru HSC (SILVERRUSH) project (Ouchi et al 2018, see also Shibuya et al 2018a,b;Konno et al 2018;Harikane et al 2018;Inoue et al 2018;Higuchi et al 2019;Harikane et al 2019;Kakuma et al 2021). Ono et al (2021) selected LAE candidates based on color and removed contaminants by a convolutional neural network (CNN) and visual inspection. Their final catalog includes (542,959,395,150) LAEs at z = (2.2, 3.3, 5.7, 6.6) in the UD-COSMOS field, and (560, 75) LAEs at z = (5.7, 6.6) in the UD-SXDS field.…”

“…8 − 1. 1 (Ono et al 2021). The NB387 magnitudes are corrected for the systematic zero-point offset by 0.45 mag, following the recommendation by the HSC-SSP team.…”

“…A major update of our data compared to Kakuma et al ( 2021) is that we add new 1051 LAEs at z = 2.2 and 3.3 using the CHORUS data. The catalog of z = 5.7 and 6.6 LAEs are also updated in that we use the latest catalog constructed by Ono et al (2021) based on HSC-SSP S18A images, while Kakuma et al (2021) used the S16A catalog taken from Shibuya et al (2018a). Although the number of the LAEs increased owing to the improved limiting magnitudes, we excluded faint LAEs from those included in Ono et al (2021), which resulting in comparable numbers of the LAEs between Kakuma et al ( 2021)'s and our all samples.…”

CHORUS IV: Mapping the Spatially Inhomogeneous Cosmic Reionization with Subaru HSC

“…Given that we are using the same z −NB921 color cut as theirs, we expect the contamination rates in our sample are similar, though different criteria for non-detection of g and r bands may increase the contamination rate. Ono et al (2021) construct an LAE sample from the same dataset. They use z − NB921 > 1.0 color cut.…”