Infrared luminosity functions based on 18 mid-infrared bands: revealing cosmic star formation history with AKARI and Hyper Suprime-Cam*
Momose, Rieko; Hashimoto, Tetsuya; Goto, Tomotsugu; Kim, Seong Jin; Burgarella, Denis; Nakagawa, Takao; Ohyama, Youichi; Pearson, Chris; Barrufet, Laia; Im, Myungshin; Jeong, Woong-Seob; Matsuhara, Hideo; Sedgwick, Chris; Malkan, Matthew; Oi, Nagisa; Toba, Yoshiki; Ikeda, Hiroyuki; Utsumi, Yousuke; Takagi, Toshinobu; Miyaji, Takamitsu; Buat, Veronique; Chiang, Chia-Ying; Kilerci Eser, Ece; Kim, Helen
Taiwan, Japan, United States, Mexico, South Korea, United Kingdom, Spain, France
Abstract
Much of star formation is obscured by dust. For a complete understanding of the cosmic star formation history (CSFH), infrared (IR) census is indispensable. AKARI carried out deep mid-infrared observations using its continuous nine-band filters in the North Ecliptic Pole (NEP) field (5.4 deg2). This took a significant amount of the satellite's lifetime, ∼10% of the entire pointed observations. By combining archival Spitzer (five bands) and WISE (four bands) mid-IR photometry, we have, in total, 18-band mid-IR photometry, which is the most comprehensive photometric coverage in the mid-IR for thousands of galaxies. However, we only had shallow optical imaging (∼25.9 AB magnitude) in a small area, 1.0 deg2. As a result, thousands of AKARI's infrared sources remained undetected in the optical. Using the new Hyper Suprime-Cam on the Subaru telescope, we obtained deep enough optical images of the entire AKARI NEP field in five broad bands (g ∼ 27.5 mag). These provided photometric redshift, and thereby IR luminosity, for the previously undetected faint AKARI IR sources. Combined with the accurate mid-IR luminosity measurement, we constructed mid-IR luminosity functions (LFs), and thereby performed a census of dust-obscured CSFH in the entire AKARI NEP field. We have measured rest-frame 8 μm and 12 μm LFs, and estimated total infrared LFs at 0.35 < z < 2.2. Our results are consistent with our previous work, but with much reduced statistical errors thanks to the large-area coverage of the new data. We have possibly witnessed the turnover of CSFH at z ∼ 2.