The relationship between Gas and Galaxies

Abstract

We have investigated the 2D 2-point correlation function, ƐAG, between low column density Lya absorbers and galaxies at a redshift z ~ 1 for the first time. 141 Lya absorbers between redshifts z = 0.68 → 1.51 over a total redshift path length of ∆z = 1.09 were collected from HST STIS E230M absorption spectra towards the quasars HE 1122-1648 (z = 2.4) and PKS 1127-145 (z = 1.187). The column density of the correlated Lya absorbers ranged from 13.18 ≤ log(_10) (N(_HI) (cm(^-2))) ≤ 17.42, with a median column density of log(_10) (N(_HI) (cm(^-2))) 13.99 ± 0.21. A total of 200 galaxy redshifts within the surrounding 6.8' x 5.7' field of view of both quasars were identified in a R magnitude limited survey (21.5 ≤ R(_vega) ≤ 24.5) using the FORS2 spectrograph at the VLT. An upper-limit of Ɛ(_AG) = 18.3 was found when 194 Lya absorber-galaxy pairs were binned in redshift space, in a bin of size ∆σ = 1.0, ∆π= 2.0 h(^-1_70) Mpc along the projected separation and line of sight distances respectively. The upper-limit in the cross-correlation was found to be 3.2σ lower than the central peak in the galaxy auto-correlation, Ɛ’(_GG), which was equal to 3.89 ± 0.65, and » 5σ lower than the galaxy auto-correlation that was measured by the 2dFGRS and VVDS investigations. Thus we measured clustering amongst galaxies to be significantly stronger than the clustering between low column density Lya absorbers and galaxies. We then used GIMIC, a high-resolution hydrodynamical simulation, to re-create the cross-correlation. When binned with ∆σ= 1.0, ∆π = 2.0 h(^-1_73)Mpc in redshift space the simulations were consistent with the observations. No significant correlation exists between galaxies and Lya absorbers with log(_10) (N(_HI) (cm(^-2)))= 13 - 17, and ƐAG had a peak of 2.65 ± 0.78 at a redshift z = 1.0. The simulated Ɛag was observed to only marginally increase when measured with absorbers of an increasing column density between log(_10) (N(_HI) (cm(^-2)))- 13 - 17. Likewise Ɛag increased by < lσ to 3.96 ± 1.21 when the cross-correlation was measured at z = 0.5. Hence in the models there is no significant evolution in Ɛag with redshift. Ɛ GG from the GIMIC simulation was 27.05 ± 4.06 at z = 1.0, so when plotted in redshift space Ɛgg was again significantly greater than the cross-correlation. Thus we reached a conclusion that the galaxies that inhabit the nodes and filaments of the dark matter cosmic web are embedded within, but not necessarily correlated with the low column density Lya absorbers (log(_10) (N(_HI) (cm(^-2)))< 17) that loosely trace this filamentary structure. Hence the Lya absorption fines in quasar spectra are predominantly caused by photons passing through this diffuse medium, and not because the sightline passes through a galaxy halo. In a small side study, 47 C IV and 18 O VI absorption lines located in the UVES spectra of quasars HE 1122-1648 and PKS 1127-145 were used to calculate the oxygen and carbon metallicity of the Lya absorbers. We found no evolution with redshift for either species. The mean carbon solar metallicity between 1.0 < z < 2.35 was [C/H]= (-0.05±0.34)z-1.24±0.58. There was a large scatter that varied from solar abundances to [C/H]-3.41(^+0.12_-0.09). The mean oxygen solar metallicity between 2.0 < z < 2.4 was [0/H]= (1.01 ± 1.54)z - 3.94 ± 3.35, again with a similar large scatter in the oxygen abundance. These results are similar to previous findings by Schaye et al. (2003) and Aguirre et al. (2008), who studying C IV and O VI lines from 19 quasars at similar redshifts also found no significant evolution in the metallicity. All of these results indicate that many of these metals must have been expelled into the IGM at higher redshifts