The MUSE Hubble Ultra Deep Field Survey. XV. The mean rest-UV spectra of Ly_alpha emitters at z>3

2020 
We investigated the ultraviolet (UV) spectral properties of faint Lyman-α emitters (LAEs) in the redshift range 2.9 ≤ z  ≤ 4.6, and we provide material to prepare future observations of the faint Universe. We used data from the MUSE Hubble Ultra Deep Survey to construct mean rest-frame spectra of continuum-faint (median M UV of −18 and down to M UV of −16), low stellar mass (median value of 108.4 M ⊙ and down to 107 M ⊙ ) LAEs at redshift z  ≳ 3. We computed various averaged spectra of LAEs, subsampled on the basis of their observational (e.g., Lyα strength, UV magnitude and spectral slope) and physical (e.g., stellar mass and star-formation rate) properties. We searched for UV spectral features other than Lyα , such as higher ionization nebular emission lines and absorption features. We successfully observed the O III]λ 1666 and [C III]λ 1907+C III]λ 1909 collisionally excited emission lines and the He IIλ 1640 recombination feature, as well as the resonant C IVλ λ 1548,1551 doublet either in emission or P-Cygni. We compared the observed spectral properties of the different mean spectra and find the emission lines to vary with the observational and physical properties of the LAEs. In particular, the mean spectra of LAEs with larger Lyα equivalent widths, fainter UV magnitudes, bluer UV spectral slopes, and lower stellar masses show the strongest nebular emission. The line ratios of these lines are similar to those measured in the spectra of local metal-poor galaxies, while their equivalent widths are weaker compared to the handful of extreme values detected in individual spectra of z  >  2 galaxies. This suggests that weak UV features are likely ubiquitous in high z , low-mass, and faint LAEs. We publicly released the stacked spectra, as they can serve as empirical templates for the design of future observations, such as those with the James Webb Space Telescope and the Extremely Large Telescope.
    • Correction
    • Source
    • Cite
    • Save
    • Machine Reading By IdeaReader
    80
    References
    13
    Citations
    NaN
    KQI
    []