Extremely Metal-Poor Representatives Explored by the Subaru Survey (EMPRESS). I. A Successful Machine Learning Selection of Metal-Poor Galaxies and the Discovery of a Galaxy with M*<10^6 M_sun and 0.016 Z_sun

TL;DR

This study developed a machine learning method to efficiently identify extremely metal-poor galaxies in large optical surveys, leading to the discovery of a galaxy with exceptionally low metallicity, advancing understanding of early galaxy formation.

Contribution

We created a deep neural network classifier to select EMPGs from large photometric datasets, successfully identifying and spectroscopically confirming extremely metal-poor galaxies, including the most metal-poor galaxy known.

Findings

Successfully classified EMPGs with 86% completeness and 46% purity.

Discovered a galaxy with metallicity Z/Z_sun=0.016, the lowest reported.

Identified star-forming galaxies with properties similar to early universe galaxies.

Abstract

We have initiated a new survey for local extremely metal-poor galaxies (EMPGs) with Subaru/Hyper Suprime-Cam (HSC) large-area (~500 deg^2) optical images reaching a 5 sigma limit of ~26 magnitude, about 100 times deeper than the Sloan Digital Sky Survey (SDSS). To select Z/Z_sun<0.1 EMPGs from ~40 million sources detected in the Subaru images, we first develop a machine-learning (ML) classifier based on a deep neural network algorithm with a training data set consisting of optical photometry of galaxy, star, and QSO models. We test our ML classifier with SDSS objects having spectroscopic metallicity measurements, and confirm that our ML classifier accomplishes 86%-completeness and 46%-purity EMPG classifications with photometric data. Applying our ML classifier to the photometric data of the Subaru sources as well as faint SDSS objects with no spectroscopic data, we obtain 27 and 86 EMPG candidates from the Subaru and SDSS photometric data, respectively. We conduct optical follow-up spectroscopy for 10 out of our EMPG candidates with Magellan/LDSS-3+MagE, Keck/DEIMOS, and Subaru/FOCAS, and find that the 10 EMPG candidates are star-forming galaxies at z=0.007-0.03 with large H_beta equivalent widths of 104-265 A, stellar masses of log(M*/M_sun)=5.0-7.1, and high specific star-formation rates of ~300 Gyr^{-1}, which are similar to those of early galaxies at z>6 reported recently. We spectroscopically confirm that 3 out of 10 candidates are truly EMPGs with Z/Z_sun<0.1, one of which is HSC J1631+4426, the most metal-poor galaxy with Z/Z_sun=0.016 reported ever.