H i content, metallicities and spin temperatures of damped and sub-damped Lyα systems in the redshift desert (0.6 < zabs < 1.7)★

Abstract

The H i 21‐cm absorption optical depth and the N(H i) column density derived from Lyα absorption can be combined to yield the spin temperature (Ts) of intervening damped Lyα systems (DLAs). Although spin temperatures have been determined for samples of DLAs with zabs < 0.6 and zabs > 1.7, the intermediate regime currently contains only two Ts measurements, leading to a ‘redshift desert’ that spans 4 Gyr of cosmic time. To connect the low‐ and high‐redshift regimes, we present observations of the Lyα line of six 0.6 < zabs < 1.7 H i 21‐cm absorbers. The data set is complemented by both low‐frequency Very Long Baseline Array observations (to derive the absorber covering factor, f) and optical echelle spectra from which metal abundances are determined. Our data set therefore not only offers the largest statistical study of H i 21‐cm absorbers to date, and bridges the redshift desert, but is also the first to use a fully f‐corrected data set to look for metallicity‐based trends. The metallicities of H i 21‐cm absorbers are consistent with those of optically selected DLAs at the same redshift. In agreement with trends found in Galactic sightlines, we find that the lowest column density absorbers tend to be dominated by warm gas. In the DLA regime, spin temperatures show a wider range of values than Galactic data, as may be expected in a heterogeneous galactic population. However, we find that low‐metallicity DLAs are dominated by small cold gas fractions and only absorbers with relatively high metallicities exhibit significant fractions of cold gas. Using a compilation of H i 21‐cm absorbers which are selected to have covering‐factor‐corrected spin temperatures, we confirm an anticorrelation between metallicity and Ts at 3.4σ significance. Finally, one of the DLAs in our sample is a newly discovered H i 21‐cm absorber (at zabs= 0.602 towards J1431+3952), which we find to have the lowest f‐corrected spin temperature yet reported in the literature: K. The observed distribution of Ts and metallicities in DLAs and the implications for understanding the characteristics of the interstellar medium in high‐redshift galaxies are discussed.

New H i 21-cm absorbers at low and intermediate redshifts

“…The quality of our data and the Ellison et al (2012) data is comparable: they report an rms noise of 4.5 mJy per 0.83 km s −1 channel, while we measure a noise level of 5.0 mJy. We measure a peak optical depth of τ = 0.22 in the spectrum of J1431+3952.…”

“…21-cm absorption studies are potentially a good means of increasing the sample of low redshift DLAs as these studies are relatively successful: there are currently ∼ 20 detections of intervening H I 21-cm absorption at redshifts z < 1 Gupta et al 2013;Ellison et al 2012, and references therein). A more complete sampling of low-redshift H I absorbers will eventually help in accurately determining the evolution of the cosmic mass density of neutral hydrogen ΩHI over half the age of the universe, especially when the results of large scale 21-cm absorption line surveys from the Square Kilometre Array and its pathfinders will become available.…”

Directly imaging damped Ly α galaxies at z > 2 – III. The star formation rates of neutral gas reservoirs at z ∼ 2.7

“…Similarly, the paucity of cold gas and molecules in high-redshift DLAs (see e.g. Carilli et al 1996;Curran & Webb 2006;Jorgenson et al 2006Kanekar et al 2006Kanekar et al , 2014Noterdaeme et al 2008;Ellison et al 2012;Srianand et al 2012) points towards an environment that is atypical for dusty, gas rich, galaxies. Finally, we note that, although the exact relationship between the observed cooling rate and the UV photon flux may be subject of future revision, the presence of high-cool DLAs disfavour a scenarios in which the UV radiation field is completely suppressed by dust.…”

On the mass–metallicity relation, velocity dispersion, and gravitational well depth of GRB host galaxies

“…Observing at low resolution causes a smearing of the absorption lines which will in turn lead to a measurement of ∆v90 which is larger than the true ∆v90. The magnitude of the Ellison et al (2012); (2) Rafelski et al (2012); (3) Peroux et al (2008); (4) Pettini et al (2000); (5) effect depends on both the true ∆v90 and the resolution of the spectrum. If the true width is large then even very low resolution spectra may still be used, if the true width is very small then the width information may have been lost completely even in a medium resolution spectrum and hence only an upper limit for the velocity width can be derived.…”