We investigate the contribution of star-forming galaxies to the ionizing background at z~3, building on previous work based on narrowband (NB3640) imaging in the SSA22a field. We use new Keck/LRIS spectra of Lyman break galaxies (LBGs) and narrowband
-selected Lya emitters (LAEs) to measure redshifts for 16 LBGs and 87 LAEs at z>3.055, such that our NB3640 imaging probes the Lyman-continuum (LyC) region. When we include the existing set of spectroscopically-confirmed LBGs, our total sample with z>3.055 consists of 41 LBGs and 91 LAEs, of which nine LBGs and 20 LAEs are detected in our NB3640 image. With our combined imaging and spectroscopic data sets, we critically investigate the origin of NB3640 emission for detected LBGs and LAEs. We remove from our samples 3 LBGs and 3 LAEs with spectroscopic evidence of contamination of their NB3640 flux by foreground galaxies, and statistically model the effects of additional, unidentified foreground contaminants. The resulting contamination and LyC-detection rates, respectively, are 62 +/-13% and 8 +/-3% for our LBG sample, and 47 +/-10% and 12 +/-2% for our LAE sample. The corresponding ratios of non-ionizing UV to LyC flux-density, corrected for intergalactic medium (IGM) attenuation, are 18.0 +34.8/-7.4 for LBGs, and 3.7 +2.5/-1.1 for LAEs. We use these ratios to estimate the total contribution of star-forming galaxies to the ionizing background and the hydrogen photoionization rate in the IGM, finding values larger than, but consistent with, those measured in the Lya forest. Finally, the measured UV to LyC flux-density ratios imply model-dependent LyC escape fractions of f_{esc}^{LyC} ~ 5-7% for our LBG sample and f_{esc}^{LyC} ~ 10-30% for our fainter LAE sample.
We present the results of an ultradeep, narrowband imaging survey for Lyman-continuum (LyC) emission at z~3 in the SSA22a field. We employ a custom narrowband filter centered at 3640A (NB3640), which probes the LyC region for galaxies at z>3.06. We a
lso analyze new and archival NB4980 imaging tuned to the wavelength of the Lya emission line at z=3.09, and archival broadband B, V, and R images of the non-ionizing UV continuum. Our NB3640 images contain 26 z>3.06 Lyman Break Galaxies (LBGs) as well as a set of 130 Lya emitters (LAEs), identified by their excess NB4980 flux relative to the BV continuum. Six LBGs and 28 LAEs are detected in the NB3640 image. LBGs appear to span a range of NB3640-R colors, while LAEs appear bimodal in their NB3640-R properties. We estimate average UV to LyC flux density ratios, corrected for foreground contamination and intergalactic medium absorption, finding <F_{UV}/F_{LyC}>^{LBG} = 11.3^{+10.3}_{-5.4}, which implies a LBG LyC escape fraction f_{esc}^{LyC} ~ 0.1, and <F_{UV}/F_{LyC}>^{LAE} = 2.2^{+0.9}_{-0.6}. The strikingly blue LAE flux density ratios defy interpretation in terms of standard stellar population models. Assuming <F_{UV}/F_{LyC}>^{LBG} applies down to L=0.1L*, we estimate a galaxy contribution to the intergalactic hydrogen ionization rate that is consistent with independent estimates based on the Lya forest opacity at z~3. If we assume that <F_{UV}/F_{LyC}>^{LAE} holds at the faintest luminosities, the galaxy contribution significantly exceeds that inferred from the Lya forest. Further follow-up study of these faint LAEs is crucial, given the potentially important contribution similar objects make to the process of reionization. (Abridged)
(Abridged) Star formation-driven outflows are a critical phenomenon in theoretical treatments of galaxy evolution, despite the limited ability of observations to trace them across cosmological timescales. If the strongest MgII absorption-line systems
detected in the spectra of background quasars arise in such outflows, ultra-strong MgII (USMgII) absorbers would identify significant numbers of galactic winds over a huge baseline in cosmic time, in a manner independent of the luminous properties of the galaxy. To this end, we present the first detailed imaging and spectroscopic study of the fields of two USMgII absorber systems culled from a statistical absorber catalog, with the goal of understanding the physical processes leading to the large velocity spreads that define such systems. Each field contains two bright emission-line galaxies at similar redshift (dv < 300 km/s) to that of the absorption. Lower-limits on their instantaneous star formation rates (SFR) from the observed OII and Hb line fluxes, and stellar masses from spectral template fitting indicate specific SFRs among the highest for their masses at z~0.7. Additionally, their 4000A break and Balmer absorption strengths imply they have undergone recent (~0.01 - 1 Gyr) starbursts. The concomitant presence of two rare phenomena - starbursts and USMgII absorbers - strongly implies a causal connection. We consider these data and USMgII absorbers in general in the context of various popular models, and conclude that galactic outflows are generally necessary to account for the velocity extent of the absorption. We favour starburst driven outflows over tidally-stripped gas from a major interaction which triggered the starburst as the energy source for the majority of systems. Finally, we discuss the implications of these results and speculate on the overall contribution of such systems to the global SFR density at z~0.7.
We present observations of CaII, ZnII, and CrII absorption lines in 16 DLAs and 6 subDLAs at 0.6 < z < 1.3, obtained for the dual purposes of: (i) clarifying the relationship between DLAs and absorbers selected via strong CaII lines, and (ii) increas
ing the still limited sample of Zn and Cr abundances in this redshift range. We find only partial overlap between current samples of intermediate-z DLAs (which are drawn from magnitude limited surveys) and strong CaII absorbers: approximately 25% of known DLAs at these redshifts have an associated CaII 3935 line with REW>0.35A, the threshold of the SDSS sample assembled by Wild and her collaborators. The lack of the strongest systems (with REW>0.5A) is consistent with these authors conclusion that such absorbers are often missed in current DLA surveys because they redden/dim the light of the background QSOs. We rule out the suggestion that strong CaII absorption is associated exclusively with the highest-N(HI) DLAs. Furthermore, we find no correlation between the strength of the CaII lines and either the metallicity or depletion, although the strongest CaII absorber in our sample is also the most metal-rich DLA yet discovered, with [Zn/H] ~ solar. We conclude that a complex mix of parameters determine the strengths of the CaII lines, including the density of particles and UV photons in the ISM of the galaxies hosting the DLAs. We find tentative evidence (given the small size of our sample) that strong CaII systems may preferentially sample regions of high gas density, perhaps akin to the DLAs exhibiting molecular hydrogen absorption at redshifts z>2. If this connection is confirmed, strong CaII absorbers would trace possibly metal-rich, H2-bearing columns of cool, dense gas at distances up to tens of kpc from normal galaxies. (abridged)
We report on a survey for narrow (FWHM < 600 km/s) CIV absorption lines in a sample of bright quasars at redshifts $1.8 le z < 2.25$ in the Sloan Digital Sky Survey. Our main goal is to understand the relationship of narrow CIV absorbers to quasar ou
tflows and, more generally, to quasar environments. We determine velocity zero-points using the broad MgII emission line, and then measure the absorbers quasar-frame velocity distribution. We examine the distribution of lines arising in quasar outflows by subtracting model fits to the contributions from cosmologically intervening absorbers and absorption due to the quasar host galaxy or cluster environment. We find a substantial number ($ge 43pm6$ per cent) of absorbers with REW $> 0.3$ AA in the velocity range +750 km/s $la v la $ +12000 km/s are intrinsic to the AGN outflow. This `outflow fraction peaks near $v=+2000$ km/s with a value of $f_{outflow} simeq 0.81 pm 0.13$. At velocities below $v approx +2000$ km/s the incidence of outflowing systems drops, possibly due to geometric effects or to the over-ionization of gas that is nearer the accretion disk. Furthermore, we find that outflow-absorbers are on average broader and stronger than cosmologically-intervening systems. Finally, we find that $sim 14$ per cent of the quasars in our sample exhibit narrow, outflowing CIV absorption with REW $> 0.3$AA, slightly larger than that for broad absorption line systems.
