A comprehensive study of H$\hat{I}\pm$ emitters at $z\sim$ 0.62 in the DAWN survey: the need for deep and wide regions

Abstract

We present new estimates of the luminosity function (LF) and star formation rate density (SFRD) for an H$\hat{I}\pm$ selected sample at $z\sim 0.62$ from the Deep And Wide Narrow-band (DAWN) survey. Our results are based on a new H$\hat{I}\pm$ sample in the extended COSMOS region (compared to Coughlin et al. 2018) with the inclusion of flanking fields, resulting in a total area coverage of $\sim$1.5 deg${}^2$. A total of 241 H$\hat{I}\pm$ emitters were selected based on robust selection criteria using spectro-photometric redshifts and broadband color-color classification. We explore the effect of different dust correction prescriptions by calculating the LF and SFRD using a constant dust extinction correction, A{${}_{\text{H}\hat{I}\pm }=1$} mag, a luminosity-dependent correction, and a stellar-mass dependent correction. The resulting H$\hat{I}\pm$ LFs are well fitted using Schechter functions with best-fit parameters: L${}^{\ast }=10^{42.24}$ erg s${}^{-1}$, $\ddot{I}{•}^{\ast }=10^{-2.85}$ Mpc${}^{-3}$, $\hat{I}\pm =-1.62$ for constant dust correction, L${}^{\ast }=10^{42.31}$ erg s${}^{-1}$, $\ddot{I}{•}^{\ast }=10^{-2.8}$ Mpc${}^{-3}$, $\hat{I}\pm =-1.39$ for luminosity-dependent dust correction, and L${}^{\ast }=10^{42.36}$ erg s${}^{-1}$, $\ddot{I}{•}^{\ast }=10^{-2.91}$ Mpc${}^{-3}$, $\hat{I}\pm =-1.48$, for stellar mass-dependent dust correction. The deep and wide nature of the DAWN survey effectively samples H$\hat{I}\pm$ emitters over a wide range of luminosities, thereby providing better constraints on both the faint and bright end of the LF. Also, the SFRD estimates $\ddot{I}{}_{\text{SFR}}=10^{-1.39}$ M${}_{\odot }$yr${}^{-1}$Mpc${}^{-3}$ (constant dust correction), $\ddot{I}{}_{\text{SFR}}=10^{-1.47}$ M${}_{\odot }$yr${}^{-1}$Mpc${}^{-3}$ (luminosity-dependent dust correction), and $\ddot{I}{}_{\text{SFR}}=10^{-1.49}$ M${}_{\odot }$yr${}^{-1}$Mpc${}^{-3}$ (stellar mass-dependent dust correction) are in good agreement with the evolution of SFRD across redshifts ($0<z<2$) seen from previous H$\hat{I}\pm$ surveys.

16 pages, 8 figures, Accepted for publication in ApJ