Equivalent Width Distribution Research Articles

SPECTROSCOPIC OBSERVATION OF Lyα EMITTERS ATz∼ 7.7 AND IMPLICATIONS ON RE-IONIZATION

We present spectroscopic follow-up observations on two bright Ly$\alpha$ emitter (LAE) candidates originally found by Krug et al. (2012) at a redshift of z~7.7 using the Multi-Object Spectrometer for Infra-Red Exploration (MOSFIRE) at Keck. We rule out any line emission at the >5$\sigma$ level for both objects, putting on solid ground a previous null result for one of the objects. The limits inferred from the non-detections rule out the previous claim of no or even reversed evolution between 5.7 < z < 7.7 in the Ly$\alpha$ luminosity function (LF) and suggest a drop in the Ly$\alpha$ luminosity function consistent with that seen in Lyman Break galaxy (LBG) samples. We model the redshift evolution of the LAE LF using the LBG UV continuum LF and the observed rest-frame equivalent width distribution. From the comparison of our empirical model with the observed LAE distribution, we estimate lower limits of the neutral hydrogen fraction to be 50-70% at z~7.7. Together with this, we find a strong evolution in the Ly$\alpha$ optical depth characterized by (1+z)^(2.2 $\pm$ 0.5) beyond z=6 indicative of a strong evolution of the IGM. Finally, we extrapolate the LAE LF to z~9 using our model and show that it is unlikely that large area surveys like UltraVISTA or Euclid pick up LAEs at this redshift assuming the current depths and area.

Low/High Redshift Classification of Emission Line Galaxies in the HETDEX survey

AbstractWe discuss different methods to separate high- from low-redshift galaxies based on a combination of spectroscopic and photometric observations. Our baseline scenario is the Hobby-Eberly Telescope Dark Energy eXperiment (HETDEX) survey, which will observe several hundred thousand Lyman Alpha Emitting (LAE) galaxies at 1.9 &lt; z &lt; 3.5, and for which the main source of contamination is [OII]-emitting galaxies at z &lt; 0.5. Additional information useful for the separation comes from empirical knowledge of LAE and [OII] luminosity functions and equivalent width distributions as a function of redshift. We consider three separation techniques: a simple cut in equivalent width, a Bayesian separation method, and machine learning algorithms, including support vector machines. These methods can be easily applied to other surveys and used on simulated data in the framework of survey planning.

CONTAMINATION OF BROADBAND PHOTOMETRY BY NEBULAR EMISSION IN HIGH-REDSHIFT GALAXIES: INVESTIGATIONS WITH KECK'S MOSFIRE NEAR-INFRARED SPECTROGRAPH

Earlier work has raised the potential importance of nebular emission in the derivation of the physical characteristics of high redshift Lyman break galaxies. Within certain redshift ranges, and especially at z ~ 6-7, such lines may be strong enough to reduce estimates of the stellar masses and ages of galaxies compared those derived assuming broad-band photometry represents stellar light alone. To test this hypothesis at the highest redshifts where such lines can be probed with ground-based facilities, we examine the near-infrared spectra of a representative sample of 20 3.0 < z < 3.8 Lyman break galaxies using the newly-commissioned MOSFIRE near-infrared spectrograph at the Keck I telescope. We use this data to derive the rest-frame equivalent widths (EW) of [O III] emission and show that these are comparable to estimates derived using the SED fitting technique introduced for sources of known redshift by Stark et al (2013). Although our current sample is modest, its [O III] EW distribution is consistent with that inferred for H\alpha\ based on SED fitting of Stark et al's larger sample of 3.8 < z < 5 galaxies. For a subset of survey galaxies, we use the combination of optical and near-infrared spectroscopy to quantify kinematics of outflows in z ~ 3.5 star-forming galaxies, and discuss the implications for reionization measurements. The trends we uncover underline the dangers of relying purely on broad-band photometry to estimate the physical properties of high redshift galaxies and emphasize the important role of diagnostic spectroscopy.

THE SMOOTH Mg II GAS DISTRIBUTION THROUGH THE INTERSTELLAR/EXTRA-PLANAR/HALO INTERFACE

We report the first measurements of MgII absorption systems associated with spectroscopically confirmed z~0.1 star-forming galaxies at projected distances of D<6kpc. We demonstrate the data are consistent with the well known anti-correlation between rest-frame MgII equivalent width, Wr(2796), and impact parameter, D, represented by a single log-linear relation derived by Nielsen et al. (MAGIICAT) that converges to ~2A at D=0kpc. Incorporating MAGIICAT, we find that the halo gas covering fraction is unity below D~25kpc. We also report that our D<6kpc absorbers are consistent with the Wr(2796) distributions of the Milky Way interstellar medium (ISM) and ISM+halo. In addition, quasar sight-lines of intermediate redshift galaxies with 6<D<25kpc have an equivalent width distribution similar to that of the Milky Way halo, implying that beyond ~6kpc, quasar sight-lines are likely probing halo gas and not the ISM. As inferred by the Milky Way and our new data, the gas profiles of galaxies can be fit by a single log-linear Wr(2796)-D relation out to large scales across a variety of gas-phase conditions and is maintained through the halo/extra-planar/ISM interfaces, which is remarkable considering their kinematic complexity. These low redshift, small impact parameter absorption systems are the first steps to bridge the gap between quasar absorption-line studies and HI observations of the CGM.

THE JHU-SDSS METAL ABSORPTION LINE CATALOG: REDSHIFT EVOLUTION AND PROPERTIES OF Mg II ABSORBERS

We present a generic and fully-automatic method aimed at detecting absorption lines in the spectra of astronomical objects. The algorithm estimates the source continuum flux using a dimensionality reduction technique, nonnegative matrix factorization, and then detects and identifies metal absorption lines. We apply it to a sample of ~100,000 quasar spectra from the Sloan Digital Sky Survey and compile a sample of ~40,000 Mg II & Fe II absorber systems, spanning the redshift range 0.4< z < 2.3. The corresponding catalog is publicly available. We study the statistical properties of these absorber systems and find that the rest equivalent width distribution of strong Mg II absorbers follows an exponential distribution at all redshifts, confirming previous studies. Combining our results with recent near-infrared observations of Mg II absorbers we introduce a new parametrization that fully describes the incidence rate of these systems up to z~5. We find the redshift evolution of strong Mg II absorbers to be remarkably similar to the cosmic star formation history over 0.4<z<5.5 (the entire redshift range covered by observations), suggesting a physical link between these two quantities.

KECK SPECTROSCOPY OF 3 &lt;z&lt; 7 FAINT LYMAN BREAK GALAXIES: THE IMPORTANCE OF NEBULAR EMISSION IN UNDERSTANDING THE SPECIFIC STAR FORMATION RATE AND STELLAR MASS DENSITY

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.

A SURVEY OF Mg II ABSORPTION AT 2 &lt;z&lt; 6 WITH MAGELLAN/FIRE. I. SAMPLE AND EVOLUTION OF THE Mg II FREQUENCY

We present initial results from the first systematic survey for MgII quasar absorption lines at z > 2.5. Using infrared spectra of 46 high-redshift quasars, we discovered 111 MgII systems over a path covering 1.9 < z < 6.3. Five systems have z > 5, with a maximum of z = 5.33 - the most distant MgII system now known. The comoving MgII line density for weaker systems (Wr < 1.0A) is statistically consistent with no evolution from z = 0.4 to z = 5.5, while that for stronger systems increases three-fold until z \sim 3 before declining again towards higher redshifts. The equivalent width distribution, which fits an exponential, reflects this evolution by flattening as z approaches 3 before steepening again. The rise and fall of the strong absorbers suggests a connection to the star formation rate density, as though they trace galactic outflows or other byproducts of star formation. The weaker systems' lack of evolution does not fit within this interpretation, but may be reproduced by extrapolating low redshift scaling relations between host galaxy luminosity and absorbing halo radius to earlier epochs. For the weak systems, luminosity-scaled models match the evolution better than similar models based on MgII occupation of evolving CDM halo masses, which greatly underpredict dN/dz at early times unless the absorption efficiency of small haloes is significantly larger in the early universe. Taken together, these observations suggest that the general structure of MgII-bearing haloes was put into place early in the process of galaxy assembly. Except for a transient appearance of stronger systems near the peak epoch of cosmic star formation, the basic properties of MgII absorbers have evolved fairly little even as the (presumably) associated galaxy population grew substantially in stellar mass and half light radius.

Lyα EMISSION FROM HIGH-REDSHIFT SOURCES IN COSMOS

We investigate spectroscopically measured Ly{\alpha} equivalent widths and escape fractions of 244 sources of which 95 are Lyman Break Galaxies (LBGs) and 106 Lyman Alpha Emitters (LAEs) at z~4.2, z~4.8, and z~5.6 selected from intermediate and narrow band observations. The sources were selected from the Cosmic Evolution Survey (COSMOS), and observed with the DEIMOS spectrograph. We find that the distribution of equivalent widths shows no evolution with redshift for both the LBG selected sources and the intermediate/narrowband LAEs. We also find that the Ly{\alpha} escape fraction of intermediate and narrow band LAEs is on average higher and has a larger variation than the escape fraction of LBG selected sources. The escape fraction does not show a dependence with redshift. Similar to what has been found for LAEs at low redshifts, the sources with the highest extinctions show the lowest escape fractions. The range of escape fractions increases with decreasing extinction. This is evidence that the dust extinction is the most important factor affecting the escape of Ly{\alpha} photons, but at low extinctions other factors such as HI covering fraction and gas kinematics can be just as effective at inhibiting the escape of Ly{\alpha} photons.

The distribution of equivalent widths in long GRB afterglow spectra

The extreme brightness of gamma-ray burst (GRB) afterglows and their simple spectral shape make them ideal beacons to study the interstellar medium of their host galaxies through absorption line spectroscopy. Using 69 low-resolution GRB afterglow spectra, we conduct a study of the rest-frame equivalent width (EW) distribution of features with an average rest-frame EW larger than 0.5 A. To compare an individual GRB with the sample, we develop EW diagrams as a graphical tool, and we give a catalogue with diagrams for the 69 spectra. We introduce a line strength parameter (LSP) that allows us to quantify the strength of the absorption features as compared to the sample by a single number. Using the distributions of EWs of single-species features, we derive the distribution of column densities by a curve of growth (CoG) fit. We find correlations between the LSP and the extinction of the GRB, the UV brightness of the host galaxies and the neutral hydrogen column density. However, we see no significant evolution of the LSP with the redshift. There is a weak correlation between the ionisation of the absorbers and the energy of the GRB, indicating that, either the GRB event is responsible for part of the ionisation, or that galaxies with high-ionisation media produce more energetic GRBs. Spectral features in GRB spectra are, on average, 2.5 times stronger than those seen in QSO intervening damped Lyman-alpha (DLA) systems and slightly more ionised. In particular we find larger excess in the EW of CIV1549 relative to QSO DLAs, which could be related to an excess of Wolf-Rayet stars in the environments of GRBs. From the CoG fitting we obtain an average number of components in the absorption features of GRBs of 6.00(-1.25,+1.00). The most extreme ionisation ratios in our sample are found for GRBs with low neutral hydrogen column density, which could be related to ionisation by the GRB emission.

On the use of Lyα emitters as probes of reionization

We use numerical simulations to study the effects of the patchiness of a partly reionized intergalactic medium (IGM) on the observability of Lyα emitters (LAEs) at high redshifts (z ≳ 6). We present a new model that divides the Lyα radiative transfer into a (circum)galactic and an extragalactic (IGM) part, and investigate how the choice of intrinsic line model affects the IGM transmission results. We use our model to study the impact of neutral hydrogen on statistical observables such as the Lyα rest-frame equivalent width (REW) distribution, the LAE luminosity function and the two-point correlation function. We find that if the observed changes in LAE luminosity functions and equivalent width distributions between z ∼ 6 and 7 are to be explained by an increased IGM neutral fraction alone, we require an extremely late and rapid reionization scenario, where the Universe was ∼40 per cent ionized at z = 7, ∼50 per cent ionized at z = 6.5 and ∼100 per cent ionized at z = 6. This is in conflict with other observations, suggesting that intrinsic LAE evolution at z ≳ 6 cannot be completely neglected. We show how the two-point correlation function can provide more robust constraints once future observations obtain larger LAE samples, and provide predictions for the sample sizes needed to tell different reionization scenarios apart.

Detection of dark galaxies and circum-galactic filaments fluorescently illuminated by a quasar at z = 2.4

A deep narrow-band survey for Ly-alpha emission carried out on the VLT-FORS2 has revealed 98 Ly-alpha candidates down to a flux limit of 4.e-18 erg/s/cm^2 in a volume of 5500 comoving Mpc^3 at z=2.4 centered on the hyperluminous quasar HE0109-3518. The properties of the detected sources in terms of their i) equivalent width distribution, ii) luminosity function, and iii) the average luminosity versus projected distance from the quasar, all suggest that a large fraction of these objects have been fluorescently "illuminated" by HE0109-3518. This conclusion is supported by comparison with detailed radiative transfer simulations of the effects of the quasar illumination. 18 objects have a rest-frame Equivalent Width (EW0) larger than 240A, the expected limit for Ly-alpha emission powered by Population II star formation and 12 sources among these do not have any continuum counterpart in a deep V-band imaging of the same field. For these, a stacking analysis indicates EW0>800A, effectively ruling out Ly-alpha powered by internal star formation. These sources are thus the best candidates so far for proto-galactic clouds or "dark" galaxies at high-redshift, whose existence has recently been suggested by several theoretical studies. Assuming they are mostly ionized by the quasar radiation, we estimate that their gas masses would be about 10^9 Msun implying that their star formation efficiencies (SFE) are less than 1.e-11 yr^-1 several times below the SFE of the most gas-rich dwarf galaxies locally, and five hundred times lower than typical massive star-forming galaxies at z~2. We have also discovered extended, filamentary gas, also likely illuminated by the quasar, around some of the brightest continuum-detected sources with EW0>240A. This emission is compatible with the expectations for circum-galactic cold streams but other origins, including tidal stripping, are also possible.

THE OPTICALLY UNBIASED GRB HOST (TOUGH) SURVEY. IV. Lyα EMITTERS

We report the results of a spectroscopic search for Lyman-alpha emission from gamma-ray burst host galaxies. Based on the well-defined TOUGH sample of 69 X-ray selected Swift GRBs, we have targeted the hosts of a subsample of 20 GRBs known from afterglow spectroscopy to be in the redshift range 1.8-4.5. We detect Lya emission from 7 out of the 20 hosts, with the typical limiting 3sigma line flux being 8E-18 erg/cm2/s, corresponding to a Lya luminosity of 6E41 erg/s at z=3. The Lya luminosities for the 7 hosts in which we detect Lya emission are in the range (0.6-2.3)E42 erg/s corresponding to star-formation rates of 0.6-2.1 Msun/yr (not corrected for extinction). The rest-frame Lya equivalent widths (EWs) for the 7 hosts are in the range 9-40A. For 6 of the 13 hosts for which Lya is not detected we place fairly strong 3sigma upper limits on the EW (<20A), while for others the EW is either unconstrained or has a less constraining upper limit. We find that the distribution of Lya EWs is inconsistent with being drawn from the Lya EW distribution of bright Lyman break galaxies at the 98.3% level, in the sense that the TOUGH hosts on average have larger EWs than bright LBGs. We can exclude an early indication, based on a smaller, heterogeneous sample of pre-Swift GRB hosts, that all GRB hosts are Lya emitters. We find that the TOUGH hosts on average have lower EWs than the pre-Swift GRB hosts, but the two samples are only inconsistent at the 92% level. The velocity centroid of the Lya line is redshifted by 200-700 km/s with respect to the systemic velocity, similar to what is seen for LBGs, possibly indicating star-formation driven outflows from the host galaxies. There seems to be a trend between the Lya EW and the optical to X-ray spectral index of the afterglow (beta_OX), hinting that dust plays a role in the observed strength and even presence of Lya emission. [ABRIDGED]

Statistical study of the ISM of GRB hosts

AbstractGamma-ray burst (GRB) afterglows shine, during a brief period of time as the most luminous objects that can be detected in the Universe. They have been observed at almost any redshift, from our nearby environment (the nearest one, at z = 0.08) to the very distant Universe (the current record holder at z = 9.4). Their optical spectra are well reproduced by a clean, simple power law, making them ideal light houses to probe the interstellar medium of their host galaxies at any redshift. We have used the largest sample of GRB afterglow spectra collected to date to perform a statistical study of the interstellar medium in their host galaxies. By analysing the distribution of equivalent widths of the most prominent absorption features we evaluate the different types of environments that host GRBs and study their diversity.

Can galactic outflows explain the properties of Ly α emitters?

We study the properties of Ly-alpha emitters in a cosmological framework by computing the escape of Ly-alpha photons through galactic outflows. We combine the GALFORM semi-analytical model of galaxy formation with a Monte Carlo Ly-alpha radiative transfer code. The properties of Ly-alpha emitters at 0<z<7 are predicted using two outflow geometries: a Shell of neutral gas and a Wind ejecting material, both expanding at constant velocity. We characterise the differences in the Ly-alpha line profiles predicted by the two outflow geometries in terms of their width, asymmetry and shift from the line centre for a set of outflows with different hydrogen column densities, expansion velocities and metallicities. In general, the Ly-alpha line profile of the Shell geometry is broader and more asymmetric, and the Ly-alpha escape fraction is lower than with the Wind geometry for the same set of parameters. In order to implement the outflow geometries in the semi-analytical model GALFORM, a number of free parameters in the outflow model are set by matching the luminosity function of Ly-alpha emitters over the whole observed redshift range. The models are consistent with the observationally inferred Ly-alpha escape fractions, equivalent width distributions and with the shape of the Ly-alpha line from composite spectra. Interestingly, our predicted UV luminosity function of Ly-alpha emitters and the fraction of Ly-alpha emitters in Lyman-break galaxy samples at high redshift are in partial agreement with observations. Attenuation of the Ly-alpha line by the presence of a neutral intergalactic medium at high redshift could be responsible for this disagreement. We predict that Ly-alpha emitters constitute a subset of the galaxy population with lower metallicities, lower instantaneous star formation rates and larger sizes than the overall population at the same UV luminosity.

EVIDENCE FOR ASYMMETRIC DISTRIBUTION OF CIRCUMSTELLAR MATERIAL AROUND TYPE Ia SUPERNOVAE

We study the properties of low-velocity material in the line of sight towards nearby Type Ia Supernovae (SNe Ia) that have measured late phase nebular velocity shifts (v_neb), thought to be an environment-independent observable. We have found that the distribution of equivalent widths of narrow blended Na I D1 & D2 and Ca II H & K absorption lines differs significantly between those SNe Ia with negative and positive v_neb, with generally stronger absorption for SNe Ia with v_neb > 0. A similar result had been found previously for the distribution of colors of SNe Ia, which was interpreted as a dependence of the temperature of the ejecta with viewing angle. Our work suggests that: 1) a significant part of these differences in color should be attributed to extinction, 2) this extinction is caused by an asymmetric distribution of circumstellar material (CSM) and 3) the CSM absorption is generally stronger on the side of the ejecta opposite to where the ignition occurs. Since it is difficult to explain 3) via any known physical processes that occur before explosion, we argue that the asymmetry of the CSM is originated after explosion by a stronger ionizing flux on the side of the ejecta where ignition occurs, probably due to a stronger shock breakout and/or more exposed radioactive material on one side of the ejecta. This result has important implications for both progenitor and explosion models.

Modelling high redshift Lyman α emitters

We present a new model for high redshift Lyman-Alpha Emitters (LAEs) in the cosmological context which takes into account the resonant scattering of Ly-a photons through expanding gas. The GALICS semi-analytic model provides us with the physical properties of a large sample of high redshift galaxies. We implement a gas outflow model for each galaxy based on simple scaling arguments. The coupling with a library of numerical experiments of Ly-a transfer through expanding or static dusty shells of gas allows us to derive the Ly-a escape fractions and profiles. The predicted distribution of Ly-a photons escape fraction shows that galaxies with a low star formation rate have a f_esc of the order of unity, suggesting that, for those objects, Ly-a may be used to trace the star formation rate assuming a given conversion law. In galaxies forming stars intensely, the escape fraction spans the whole range from 0 to 1. The model is able to get a good match to the UV and Ly-a luminosity function (LF) data at 3 < z < 5. We find that we are in good agreement with both the bright Ly-a data and the faint population observed by Rauch et al. (2008) at z=3. Most of the Ly-a profiles of our LAEs are redshifted by the diffusion in the outflow which suppresses IGM absorption. The bulk of the observed Ly-a equivalent width (EW) distribution is recovered by our model, but we fail to obtain the very large values sometimes detected. Predictions for stellar masses and UV LFs of LAEs show a satisfactory agreement with observational estimates. The UV-brightest galaxies are found to show only low Ly-a EWs in our model, as it is reported by many observations of high redshift LAEs. We interpret this effect as the joint consequence of old stellar populations hosted by UV-bright galaxies, and high HI column densities that we predict for these objects, which quench preferentially resonant Ly-a photons via dust extinction.

THE TEMPERATURE STRUCTURE OF Be STAR DISKS IN THE SMALL MAGELLANIC CLOUD

The temperature structure of Be star circumstellar disks at the sub-solar metallicity appropriate to the Small Magellanic Cloud (SMC) is investigated. It is found that for central stars of the same spectral type, Be star disks in the SMC are systematically hotter by several thousand degrees compared to Milky Way (MW) disks with the same density structure. For early spectral types (B0e - B3e), this results in systematically smaller H{\alpha} equivalent widths for Be stars in the SMC. The implication of this result on Be star frequency comparisons between MW and SMC clusters is shown to be a 5 - 10% lowering of the detection efficiency of Be stars in SMC clusters. These calculations are also compared to the known H{\alpha} equivalent width distributions in the MW and SMC. For the MW, reasonable agreement is found; however, for the SMC, the match is not as good and systematically larger Be star disks may be required.

THE FAINT-END SLOPE OF THE REDSHIFT 5.7 Lyα LUMINOSITY FUNCTION,

Using new Keck DEIMOS spectroscopy, we examine the origin of the steep number counts of ultra-faint emission-line galaxies recently reported by Dressler et al. (2011). We confirm six Lyman Alpha emitters (LAEs), three of which have significant asymmetric line profiles with prominent wings extending 300-400 km/s redward of the peak emission. With these six LAEs, we revise our previous estimate of the number of faint LAEs in the Dressler et al. survey. Combining these data with the density of bright LAEs in the Cosmic Origins Survey and Subaru Deep Field provides the best constraints to date on the redshift 5.7 LAE luminosity function (LF). Schechter function parameters, phi^* = 4.5 x 10^{-4} Mpc^{-3}, L^* = 9.1 x 10^{42} erg s^{-1}, and alpha= -1.70, are estimated using a maximum likelihood technique with a model for slit losses. To place this result in the context of the UV-selected galaxy population, we investigate how various parameterizations of the Lyman Alpha equivalent width distribution, along with the measured UV-continuum LF, affect shape and normalization of the Lyman Alpha LF. The nominal model, which uses z~6 equivalent widths from the literature, falls short of the observed space density of LAEs at the bright end, possibly indicating a need for higher equivalent widths. This parameterization of the equivalent width distribution implies that as many as 50% of our faintest LAEs should have M_{UV} > -18.0, rendering them undetectable in even the deepest Hubble Space Telescope surveys at this redshift. Hence, ultra-deep emission-line surveys find some of the faintest galaxies ever observed at the end of the reionization epoch. Such faint galaxies likely enrich the intergalactic medium with metals and maintain its ionized state. Observations of these objects provide a glimpse of the building blocks of present-day galaxies at an early time.

KECK SPECTROSCOPY OF FAINT 3 &lt;z&lt; 8 LYMAN BREAK GALAXIES: EVIDENCE FOR A DECLINING FRACTION OF EMISSION LINE SOURCES IN THE REDSHIFT RANGE 6 &lt;z&lt; 8

Using deep Keck spectroscopy of Lyman break galaxies selected from infrared imaging data taken with WFC3/IR onboard the Hubble Space Telescope, we present new evidence for a reversal in the redshift-dependent fraction of star forming galaxies with detectable Lyman alpha emission in the redshift range 6.3 < z < 8.8. Our earlier surveys with the DEIMOS spectrograph demonstrated a significant increase with redshift in the fraction of line emitting galaxies over the interval 4 < z < 6, particularly for intrinsically faint systems which dominate the luminosity density. Using the longer wavelength sensitivities of LRIS and NIRSPEC, we have targeted 19 Lyman break galaxies selected using recent WFC3/IR data whose photometric redshifts are in the range 6.3 < z < 8.8 and which span a wide range of intrinsic luminosities. Our spectroscopic exposures typically reach a 5-sigma sensitivity of < 50 A for the rest-frame equivalent width (EW) of Lyman alpha emission. Despite the high fraction of emitters seen only a few hundred million years later, we find only 2 convincing and 1 possible line emitter in our more distant sample. Combining with published data on a further 7 sources obtained using FORS2 on the ESO VLT, and assuming continuity in the trends found at lower redshift, we discuss the significance of this apparent reversal in the redshift-dependent Lyman alpha fraction in the context of our range in continuum luminosity. Assuming all the targeted sources are at their photometric redshift and our assumptions about the Lyman alpha EW distribution are correct, we would expect to find so few emitters in less than 1% of the realizations drawn from our lower redshift samples. Our new results provide further support for the suggestion that, at the redshifts now being probed spectroscopically, we are entering the era where the intergalactic medium is partially neutral.

THE H I MASS DENSITY IN GALACTIC HALOS, WINDS, AND COLD ACCRETION AS TRACED BY Mg II ABSORPTION

It is well established that MgII absorption lines detected in background quasar spectra arise from gas structures associated with foreground galaxies. The degree to which galaxy evolution is driven by the gas cycling through halos is highly uncertain because their gas mass density is poorly constrained. Fitting the MgII equivalent width (W) distribution with a Schechter function and applying the N(HI)-W correlation of Menard & Chelouche, we computed Omega(HI)_MgII ~ Omega(HI)_halo =(1.41 +0.75 -0.44)x10^-4 for 0.4<z<1.4. We exclude DLAs from our calculations so that Omega(HI)_halo comprises accreting and/or outflowing halo gas not locked up in cold neutral clouds. We deduce the cosmic HI gas mass density fraction in galactic halos traced by MgII absorption is Omega(HI)_halo/Omega(HI)_DLA=15% and Omega(HI)_halo/Omega_b=0.3%. Citing several lines of evidence, we propose infall/accretion material is sampled by small W whereas outflow/winds are sampled by large W, and find Omega(HI)_infall is consistent with Omega(HI)_outflow for bifurcation at W=1.23^{+0.15}_{-0.28}\AA; cold accretion would then comprise no more than ~7% of of the total HI mass density. We discuss evidence that (1) the total HI mass cycling through halos remains fairly constant with cosmic time and that the accretion of HI gas sustains galaxy winds, and (2) evolution in the cosmic star formation rate depends primarily on the rate at which cool HI gas cycles through halos.