We present Keck/MOSFIRE observations of UV metal lines in four bright gravitationally-lensed z~6-8 galaxies behind the cluster Abell 1703. The spectrum of A1703-zd6, a highly-magnified star forming galaxy with a Lyman-alpha redshift of z=7.045, revea
ls a confident detection of the nebular CIV emission line (unresolved with FWHM < 125 km/s). UV metal lines are not detected in the three other galaxies. At z~2-3, nebular CIV emission is observed in just 1% of UV-selected galaxies. The presence of strong CIV emission in one of the small sample of galaxies targeted in this paper may indicate hard ionizing spectra are more common at z~7. The total estimated equivalent width of the CIV doublet (38 A) and CIV/Lyman-alpha flux ratio (0.3) are comparable to measurements of narrow-lined AGNs. Photoionization models show that the nebular CIV line can also be reproduced by a young stellar population, with very hot metal poor stars dominating the photon flux responsible for triply ionizing carbon. Regardless of the origin of the CIV, we show that the ionizing spectrum of A1703-zd6 is different from that of typical galaxies at z~2, producing more H ionizing photons per unit 1500A luminosity and a larger flux density at 30-50 eV. If such extreme radiation fields are typical in UV-selected systems at z>7, it would indicate that reionization-era galaxies are more efficient ionizing agents than previously thought. Alternatively, we suggest that the small sample of Lyman-alpha emitters at z>7 may trace a rare population with intense radiation fields capable of ionizing their surrounding hydrogen distribution. Additional constraints on high ionization emission lines in galaxies with and without Lyman-alpha detections will help clarify whether hard ionizing spectra are common in the reionization era.
Deep spectroscopic observations of z~6.5 galaxies have revealed a marked decline with increasing redshift in the detectability of Lyman-alpha emission. While this may offer valuable insight into the end of the reionisation process, it presents a fund
amental challenge to the detailed spectroscopic study of the many hundreds of photometrically-selected distant sources now being found via deep HST imaging, and particularly those bright sources viewed through foreground lensing clusters. In this paper we demonstrate the validity of a new way forward via the convincing detection of an alternative diagnostic line, CIII]1909, seen in spectroscopic exposures of two star forming galaxies at z=6.029 and 7.213. The former detection is based on a 3.5 hour X-shooter spectrum of a bright (J=25.2) gravitationally-lensed galaxy behind the cluster Abell 383. The latter detection is based on a 4.2 hour MOSFIRE spectra of one of the most distant spectroscopically confirmed galaxies, GN-108036, with J=25.2. Both targets were chosen for their continuum brightness and previously-known redshift (based on Lyman-alpha), ensuring that any CIII] emission would be located in a favorable portion of the near-infrared sky spectrum. We compare our CIII] and Lyman-alpha equivalent widths in the context of those found at z~2 from earlier work and discuss the motivation for using lines other than Lyman-alpha to study galaxies in the reionisation era.
We present deep spectroscopy of 17 very low mass (M* ~ 2.0x10^6 Msun to 1.4x10^9 Msun) and low luminosity (M_UV ~ -13.7 to -19.9) gravitationally lensed galaxies in the redshift range z~1.5-3.0. Deep rest-frame ultraviolet spectra reveal large equiva
lent width emission from numerous lines (NIV], OIII], CIV, Si III], CIII]) which are rarely seen in individual spectra of more massive star forming galaxies. CIII] is detected in 16 of 17 low mass star forming systems with rest-frame equivalent widths as large as 13.5 Angstroms. Nebular CIV emission is present in the most extreme CIII] emitters, requiring an ionizing source capable of producing a substantial component of photons with energies in excess of 47.9 eV. Photoionization models support a picture whereby the large equivalent widths are driven by the increased electron temperature and enhanced ionizing output arising from metal poor gas and stars, young stellar populations, and large ionization parameters. The young ages implied by the emission lines and continuum SEDs indicate that the extreme line emitters in our sample are in the midst of a significant upturn in their star formation activity. The low stellar masses, blue UV colors, and large sSFRs of our sample are similar to those of typical z>6 galaxies. Given the strong attenuation of Ly-alpha in z>6 galaxies we suggest that CIII] is likely to provide our best probe of early star forming galaxies with ground-based spectrographs and one of the most efficient means of confirming z>10 galaxies with the James Webb Space Telescope.
Bright gravitationally lensed galaxies provide our most detailed view of galaxies at high redshift. Yet as a result of the small number of ultra-bright z~2 lensed systems with confirmed redshifts, most detailed spectroscopic studies have been limited
in their scope. With the goal of increasing the number of bright lensed galaxies available for detailed follow-up, we have undertaken a spectroscopic campaign targeting wide separation (>3 arcsec) galaxy-galaxy lens candidates within the Sloan Digital Sky Survey (SDSS). Building on the earlier efforts of our CASSOWARY survey, we target a large sample of candidate galaxy-galaxy lens systems in SDSS using a well-established search algorithm which identifies blue arc-like structures situated around luminous red galaxies. In this paper, we present a new redshift catalog containing 25 lensed sources in SDSS confirmed through spectroscopic follow-up of candidate galaxy-galaxy lens systems. Included in this new sample are two of the brightest galaxies (r=19.6 and 19.7) galaxies known at z~2, a low metallicity (12 + log (O/H)~8.0) extreme nebular line emitting galaxy at z=1.43, and numerous systems for which detailed follow-up will be possible. The source redshifts span 0.9<z<2.5 (median redshift of 1.9), and their optical magnitudes are in the range 19.6<r<22.3. We present a brief source-by-source discussion of the spectroscopic properties extracted from our confirmatory spectra and discuss some initial science results. With more than 50 gravitationally lensed z>1 galaxies now confirmed within SDSS, it will soon be possible for the first time to develop generalized conclusions from detailed spectroscopic studies of the brightest lensed systems at high redshift.
The physical properties inferred from the SEDs of z>3 galaxies have been influential in shaping our understanding of early galaxy formation and the role galaxies may play in cosmic reionization. Of particular importance is the stellar mass density at
early times which represents the integral of earlier star formation. An important puzzle arising from the measurements so far reported is that the specific star formation rates (sSFR) evolve far less rapidly than expected in most theoretical models. Yet the observations underpinning these results remain very uncertain, owing in part to the possible contamination of rest-optical broadband light from strong nebular emission lines. To quantify the contribution of nebular emission to broad-band fluxes, we investigate the SEDs of 92 spectroscopically-confirmed galaxies in the redshift range 3.8<z<5.0 chosen because the H-alpha line lies within the Spitzer/IRAC 3.6 um filter. We demonstrate that the 3.6 um flux is systematically in excess of that expected from stellar continuum, which we derive by fitting the SED with population synthesis models. No such excess is seen in a control sample at 3.1<z<3.6 in which there is no nebular contamination in the IRAC filters. From the distribution of our 3.6 um flux excesses, we derive an H-alpha equivalent width (EW) distribution. The mean rest-frame H-alpha EW we infer at 3.8<z<5.0 (270 A) indicates that nebular emission contributes at least 30% of the 3.6 um flux. Via our empirically-derived EW distribution we correct the available stellar mass densities and show that the sSFR evolves more rapidly at z>4 than previously thought, supporting up to a 5x increase between z~2 and 7. Such a trend is much closer to theoretical expectations. Given our findings, we discuss the prospects for verifying quantitatively the nebular emission line strengths prior to the launch of the James Webb Space Telescope.
As Lyman-alpha photons are scattered by neutral hydrogen, a change with redshift in the Lyman-alpha equivalent width distribution of distant galaxies offers a promising probe of the degree of ionization in the intergalactic medium and hence when cosm
ic reionization ended. This simple test is complicated by the fact that Lyman-alpha emission can also be affected by the evolving astrophysical details of the host galaxies. In the first paper in this series, we demonstrated both a luminosity and redshift dependent trend in the fraction of Lyman-alpha emitters seen within color-selected Lyman-break galaxies (LBGs) over the range 3<z<6; lower luminosity galaxies and those at higher redshift show an increased likelihood of strong emission. Here we present the results from much deeper 12.5 hour exposures with the Keck DEIMOS spectrograph focused primarily on LBGs at z~6 which enable us to confirm the redshift dependence of line emission more robustly and to higher redshift than was hitherto possible. We find 54+/-11% of faint z~6 LBGs show strong (W_0>25 A) emission, an increase of 1.6x from a similar sample observed at z~4. With a total sample of 74 z~6 LBGs, we determine the luminosity-dependent Lyman-alpha equivalent width distribution. Assuming continuity in these trends to the new population of z~7 sources located with the Hubble WFC3/IR camera, we predict that unless the neutral fraction rises in the intervening 200 Myr, the success rate for spectroscopic confirmation using Lyman-alpha emission should be high.
We present results from a new Keck spectroscopic survey of UV-faint LBGs in the redshift range 3<z<7. Combined with earlier Keck and published ESO VLT data, our sample contains more than 600 dropouts, offering new insight into the nature of sub-L* so
urces typical of those likely to dominate the cosmic reionisation process. Here we use this sample to characterise the fraction of strong Lya emitters within the continuum-selected dropouts. By quantifying how the Lya fraction varies with redshift, we seek to constrain changes in Lya transmission associated with reionisation. In order to distinguish the effects of reionisation from other factors which affect the Lya fraction (e.g. dust, ISM kinematics), we study the luminosity and redshift-dependence of the Lya fraction over 3<z<6, when the IGM is known to be ionised. These results reveal that low luminosity galaxies show strong Lya emission much more frequently than luminous systems, and that at fixed luminosity, the prevalence of strong Lya emission increases moderately with redshift over 3 < z < 6. Based on the correlation between blue UV slopes and strong Lya emitting galaxies in our dataset, we argue that the Lya fraction trends are governed by redshift and luminosity-dependent variations in the dust obscuration, with likely additional contributions from trends in the kinematics and covering fraction of neutral hydrogen. We find a tentative decrease in the Lya fraction at z~7 based on the limited IR spectra for candidate z~7 lensed LBGs, a result which, if confirmed with future surveys, would suggest an increase in the neutral fraction by this epoch. Given the supply of z and Y-drops now available from Hubble WFC3/IR surveys, we show it will soon be possible to significantly improve estimates of the Lya fraction using optical and near-IR spectrographs, thereby extending the study conducted in this paper to 7<z<8.
We present new measurements of the evolution in the Lyman break galaxy (LBG) population between z~4 and z~6. By utilizing the extensive multiwavelength datasets available in the GOODS fields, we identify 2443 B, 506 V, and 137 i-band dropout galaxies
likely to be at z~4, 5, and 6. With the goal of understanding the duration of typical star formation episodes in galaxies at z>4, we examine the distribution of stellar masses and ages as a function of cosmic time. We find that at a fixed rest-UV luminosity, the average stellar masses and ages of galaxies do not increase significantly between z~6 and 4. In order to maintain this near equilibrium in the average properties of high redshift LBGs, we argue that there must be a steady flux of young, newly-luminous objects at each successive redshift. When considered along with the short duty cycles inferred from clustering measurements, these results may suggest that galaxies are undergoing star formation episodes lasting only several hundred million years. In contrast to the unchanging relationship between the average stellar mass and rest-UV luminosity, we find that the number density of massive galaxies increases considerably with time over 4<z<6. Given this rapid increase of UV luminous massive galaxies, we explore the possibility that a significant fraction of massive (M*>1e11 Msun) z~2-3 distant red galaxies (DRGs) were in part assembled in an LBG phase at earlier times. Integrating the growth in the stellar mass function of actively forming LBGs over 4<z<6 down to z~2, we find that z>3 LBGs could have contributed significantly to the quiescent DRG population, indicating that the intense star-forming systems probed by current sub-millimeter observations are not the only route toward the assembly of DRGs at z~2.
Recent studies of galaxies ~2-3 Gyr after the Big Bang have revealed large, turbulent rotating systems. The existence of well-ordered rotation in galaxies during this peak epoch of cosmic star formation may suggest that gas accretion through cold str
eams is likely to be the dominant mode by which most star-forming galaxies at high redshift since major mergers can completely disrupt the observed velocity fields. However poor spatial resolution and sensitivity have hampered this interpretation, limiting the study to the largest and most luminous galaxies, which may have fundamentally different modes of assembly than more typical star forming galaxies. Here we report observations of a typical star forming galaxy at z=3.07 with a linear resolution of ~100 parsec. This spatial sampling is made possible by the combination of gravitational lensing and laser guide star adaptive optics. We find a well-ordered compact source in which molecular gas is being converted efficiently into stars, likely assembling a spheroidal bulge and disk similar to those seen in spiral galaxies at the present day.
We present the results of a systematic search for gravitationally-lensed continuum Lyman break `drop-outs beyond a redshift 7 conducted via very deep imaging through six foreground clusters undertaken with the Hubble and Spitzer Space Telescopes. The
survey has yielded 10 z-band and 2 J-band drop-out candidates to photometric limits of J_110~=26.2 AB (5sigma). Taking into account the magnifications afforded by our clusters (1-4 magnitudes), we probe the presence of z>7 sources to unlensed limits of J_{110}~=30 AB, fainter than those charted in the Hubble Ultradeep Field. To verify the fidelity of our candidates we conduct a number of tests for instrumental effects which would lead to spurious detections, and carefully evaluate the likelihood of foreground contamination by considering photometric uncertainties in the drop-out signature, the upper limits from stacked IRAC data and the statistics of multiply-imaged sources. Overall, we conclude that we can expect about half of our sample of z-band drop-outs are likely to be at high redshift. An ambitious infrared spectroscopic campaign undertaken with the NIRSPEC spectrograph at the WM Keck Observatory for seven of the most promising candidates failed to detect any Lyman-alpha emission highlighting the challenge of making further progress in this field. While the volume density of high redshift sources will likely remain uncertain until more powerful facilities are available, our data provides the first potentially interesting constraints on the UV luminosity function at z~=7.5 at intrinsically faint limits. We discuss the implications of our results in the context of the hypothesis that the bulk of the reionizing photons in the era 7<z<12 arise in low luminosity galaxies undetected by conventional surveys.
