Using the Hubble Space Telescope (HST) Archive, I have performed the first blind survey for Damped Ly-α Systems (DLAs) at low redshift (z < 1.6). In particular, I have used spectra of 767 quasars, observed with the medium resolution ultraviolet spectrographs (i.e., COS, FOS, and STIS) on-board HST, to perform the search for DLAs. Of these 767 quasars, 463 belong to the ‘statistical’ sample, quasars that were observed without *a-priori* knowledge of any absorption along the line-of-sight. The total survey path length of this statistical sample is Δz = 123.3, corresponding to ΔX = 229.7.

Within the above survey path length we find 4 DLAs, implying an incidence per absorption path length, ℓ_{DLA}(X) = 0.017 ± 0.012; at a median survey path redshift of z=0.623. This estimate is lower, but consistent with estimates at z = 0 from 21cm emission studies. Although the data set is too small to properly sample the HI column density distribution function (ƒ(N_{HI},X)) at these redshifts, the shape is consistent with the shape at z = 2. Indeed, little evolution is seen in the shape of ƒ(N_{HI},X) at the redshifts it can be measured. Adopting the z = 2 shape of ƒ(N_{HI},X), we infer an HI mass density at z = 0.6 of ρ_{HI,DLA} = 0.25 ± 0.15 x 10^{8}M_{⊙}Mpc^{-3}.

The neutral hydrogen mass density found is significantly lower than previous estimates, but consistent with the results at lower redshift. We therefore find that the neutral gas density of the Universe has been slowly decreasing over the past 10 Gyrs. Presumably this gas is fueling star formation. The full paper can be found here.