The South Pole Telescope has discovered 100 gravitationally lensed, high-redshift, dusty, star-forming galaxies (DSFGs). We present 0."5 resolution 870 mu m. Atacama Large Millimeter/submillimeter Array imaging of a sample of 47 DSFGs spanning z = 1.9-5.7, and construct gravitational lens models of these sources. Our visibility-based lens modeling incorporates several sources of residual interferometric calibration uncertainty, allowing us to properly account for noise in the observations. At least 70% of the sources are strongly lensed by foreground galaxies (mu(870) (mu m) > 2), with a median magnification of mu(870) (mu m) = 6.3, extending to mu(870) (mu m) > 30. We compare the intrinsic size distribution of the strongly lensed sources to a similar number of unlensed DSFGs and find no significant differences in spite of a bias between the magnification and intrinsic source size. This may indicate that the true size distribution of DSFGs is relatively narrow. We use the source sizes to constrain the wavelength at which the dust optical depth is unity and find this wavelength to be correlated with the dust temperature. This correlation leads to discrepancies in dust mass estimates of a factor of two compared to estimates using a single value for this wavelength. We investigate the relationship between the [C II] line and the far-infrared luminosity and find that the same correlation between the [C II]/L-FIR. ratio and Sigma(FIR). found for low-redshift star-forming galaxies applies to high-redshift galaxies and extends at least two orders of magnitude higher in SFIR. This lends further credence to the claim that the compactness of the IR-emitting region is the controlling parameter in establishing the "[C II] deficit."

We observed the [C II] line in 15 lensed galaxies at redshifts 1 < z <. 3 using HIFI on the Herschel Space Observatory and detected 14/15 galaxies at 3 sigma or better. High magnifications enable even modestly luminous galaxies to be detected in [C II] with Herschel. The [C II] luminosity in this sample ranges from 8x10(7) L-circle dot to 3.7x10(9) L-circle dot (after correcting for magnification), confirming that [C II] is a strong tracer of the ISM at high redshifts. The ratio of the [C II] line to the total far-infrared (FIR) luminosity serves as a measure of the ratio of gas to dust cooling and thus the efficiency of the grain photoelectric heating process. It varies between 3.3% and 0.09%. We compare the [C II]/FIR ratio to that of galaxies at z = 0 and at high redshifts and find that they follow similar trends. The [C II]/FIR ratio is lower for galaxies with higher dust temperatures. This is best explained if increased UV intensity leads to higher FIR luminosity and dust temperatures, but gas heating does not rise due to lower photoelectric heating efficiency. The [C II]/FIR ratio shows weaker correlation with FIR luminosity. At low redshifts highly luminous galaxies tend to have warm dust, so the effects of dust temperature and luminosity are degenerate. Luminous galaxies at high redshifts show a range of dust temperatures, showing that [C II]/FIR correlates most strongly with dust temperature. The [C II] to mid-IR ratio for the HELLO sample is similar to the values seen for low-redshift galaxies, indicating that small grains and PAHs dominate the heating in the neutral ISM, although some of the high [CII]/FIR ratios may be due to turbulent heating.

We present the Magellan/FIRE detection of highly ionized C IV lambda 1550 and O III]lambda 1666 in a deep infrared spectrum of the z = 6.11 gravitationally lensed low-mass galaxy RXC J2248.7-4431-ID3, which has previously known Ly alpha. No corresponding emission is detected at the expected location of He II lambda 1640. The upper limit on He II, paired with detection of O III] and C IV, constrains possible ionization scenarios. Production of C IV and O III] requires ionizing photons of 2.5-3.5 Ryd, but once in that state their multiplet emission is powered by collisional excitation at lower energies (similar to 0.5 Ryd). As a pure recombination line, He II emission is powered by 4 Ryd ionizing photons. The data therefore require a spectrum with significant power at 3.5 Ryd but a rapid drop toward 4.0 Ryd. This hard spectrum with a steep drop is characteristic of low-metallicity stellar populations, and less consistent with soft AGN excitation, which features more 4 Ryd photons and hence higher He II flux. The conclusions based on ratios of metal line detections to helium non-detection are strengthened if the gas metallicity is low. RXJ2248-ID3 adds to the growing handful of reionization-era galaxies with UV emission line ratios distinct from the general z = 2-3 population in a way that suggests hard ionizing spectra that do not necessarily originate in AGNs.

Modern cosmology predicts that a galaxy overdensity (e.g., protocluster) will be associated with a large intergalactic medium gas reservoir, which can be traced by Ly alpha forest absorption. We have undertaken a systematic study of the relation between Coherently Strong intergalactic Lya Absorption systems (CoSLAs), which have the highest optical depth (tau) in the tau distribution, and mass overdensities on the scales of similar to 10-20 h(-1) comoving Mpc. On such large scales, our cosmological simulations show a strong correlation between the effective optical depth (tau(eff)) of the CoSLAs and the three-dimensional mass overdensity. In spectra with moderate signal-to-noise ratio, however, the profiles of CoSLAs can be confused with individual high column density absorbers. For z > 2.6, where the corresponding Ly beta is redshifted to the optical, we have developed a selection technique to distinguish between these two alternatives. We have applied this technique to similar to 6000 sight lines provided by Sloan Digital Sky Survey III quasar survey at z = 2.6-3.3 with a continuum-to-noise ratio greater than 8, and we present a sample of five CoSLA candidates with tau(eff) on 15 h(-1) Mpc greater than 4.5x the mean optical depth. At lower redshifts of z < 2.6, where the background quasar density is higher, the overdensity can be traced by intergalactic absorption groups using multiple sight lines with small angular separations. Our overdensity searches fully use the current and next generation of Ly alpha forest surveys, which cover a survey volume of > 1 (h(-1) Gpc)(3). Systems traced by CoSLAs will yield a uniform sample of the most massive overdensities at z > 2 to provide stringent constraints to models of structure formation.

We present the discovery of nine quasars at z similar to 6 identified in the Sloan Digital Sky Survey (SDSS) imaging data. This completes our survey of z similar to 6 quasars in the SDSS footprint. Our final sample consists of 52 quasars at 5.7 < z << 6.4, including 29 quasars with z(AB) <= 20 mag selected from 11,240 deg(2) of the SDSS single-epoch imaging survey (the main survey), 10 quasars with 20 <= z(AB) <= 20.5 selected from 4223 deg2 of the SDSS overlap regions (regions with two or more imaging scans), and 13 quasars down to z(AB) approximate to 22 mag from the 277 deg2 in Stripe 82. They span a wide luminosity range of -29.0 <= M-1450 <= -24.5. This well-defined sample is used to derive the quasar luminosity function (QLF) at z similar to 6. After combining our SDSS sample with two faint (M-1450 >= -23 mag) quasars from the literature, we obtain the parameters for a double power-law fit to the QLF. The bright-end slope beta of the QLF is well constrained to be beta = -2.8 +/- 0.2. Due to the small number of low-luminosity quasars, the faint-end slope a and the characteristic magnitude M*(1450) are less well constrained, with alpha = -1.90(-0.44)(+0.58) and M* = -25.2(-3.8)(+1.2) mag. The spatial density of luminous quasars, parametrized as rho(M-1450 < -26, z) = rho(z = 6)10(k(z-6)), drops rapidly from z similar to 5 to 6, with k = -0.72 +/- 0.11. Based on our fitted QLF and assuming an intergalactic medium (IGM) clumping factor of C = 3, we find that the observed quasar population cannot provide enough photons to ionize the z similar to 6 IGM at similar to 90% confidence. Quasars may still provide a significant fraction of the required photons, although much larger samples of faint quasars are needed for more stringent constraints on the quasar contribution to reionization.

Environment is known to have a significant impact on the evolution of galaxies. This is most evident in the local Universe, where the oldest and most massive galaxies are found at the of massive galaxy clusters. Current theory predicts that galaxies will form earlier and evolve more rapidly in the densest regions of the Universe. What is not clear is how rapidly the of environment start to have an impact on galaxies, at what stage can we detect physical differences between galaxies in dense regions and those in the field? By the time galaxies are assembled in virialised clusters the effects are clear, but at higher redshift (z ≳ 2), in the unvirialised progenitors of clusters (protoclusters) the effects are harder to detect. In this thesis I study the impact of environment in a z =3.1 protocluster in the SSA22 field. I consider the fraction of mergers in the protocluster, comparing it to the fraction of mergers in field at a similar redshift. My classification is based on the morphology of Lyman break (LBGs), using HST ACS/F814W imaging, which probes the rest frame UV. I find a marginal enhancement of the merger fraction, 48±10 per cent for LBGs in the protocluster compared 30±6 per cent in the field, suggesting that galaxy-galaxy mergers are one of the key driving accelerated star formation and AGN growth in protocluster environments. Having considered the fraction of mergers in the protocluster I then turn my attention to the physical properties of LBGs. I use multiwavelength data and spectral energy distribution fitting to determine the mass of LBGs in the protocluster and in the field. I find no statistical evidence for an enhancement of mass in the protocluster, suggesting that the protocluster environment has not impacted the average mass of LBGs at this redshift. It is possible that the protocluster LBG population may become more massive than LBGs in the field at lower redshift, or the galaxies may cease to be detectable by the Lyman break method before a mass difference between the protocluster galaxies and field is observable. Finally I consider the Lyman-α blobs (LABs) within the protocluster. These are large (~10- 100kpc) scale regions of diffuse Lyman-α emission, thought to be associated with overdense regions. 35 LABs have been detected in the SSA22 protocluster, indicating the presence of large clouds of gas in the circumgalactic medium. A debate has arisen regarding the powering mechanism of the LABs, particularly between star forming processes (e.g. Lyman-α escaping from a star forming galaxy or photoionizing radiation escaping from a star forming galaxy or active galactic nuclei) and a cold accretion model. The latter involves gas gravitationally cooling as it falls into the centre of a dark matter halo to feed a central galaxy. Some of this energy heats the cold gas, which then emits Lyman-α as it cools. The cold gas accretion theory gained popularity because some LABs appear not to contain a luminous galaxy or AGN which could explain the observed emission. One suggestion is that the central galaxy could be hidden by dust and that this could explain the lack of a detection in UV or optical. I therefore use SCUBA2 850μm imaging to search for submm sources (dusty star forming galaxies) in the LABs. I detect submm sources in only two of the LABs at 3.5δ, however, stacking all the LABs gives an average flux density of S850 = 0.6±0.2mJy. This suggests that on average the LABs do contain a submm source which could be a dusty galaxy. However, stacking the LABs by size indicates that only the largest third (area ≳ 1800kpc²) have a mean detection, at 4.5δ, with S850 = 1.4±0.3mJy, suggesting that different mechanisms may dominate the larger and smaller LAB populations. I explore two possible mechanisms for powering the LABs, cold accretion and central star forming galaxies. I find that central star formation is more likely to be the dominant source of emission, with cold accretion playing a secondary role.

In this thesis, I explore both obscured and unobscured star formation over a large fraction of cosmic time. I use the HAWK-I Y -band science verification data over GOODS-South, in conjunction with optical and infrared data to search for Lyman-break galaxies at z >∼ 6.5 (i.e. within the first billion years of the Universe). I find four possible (two robust) z′-drop candidates (z >∼ 6.5) and four possible (but no robust) Y -drop candidates (z >∼ 7). I use my results to place constraints on the luminosity function at z ∼ 6.5 and find significant evolution in the population of Lyman-break galaxies between 3 < z <∼ 6.5. I also explore obscured star formation with a population of 70μm selected galaxies over the COSMOS field. I use AAT spectroscopy in conjunction with other available spectroscopic redshifts for my sample, and photometric redshifts otherwise, to calculate the total infrared luminosity of each galaxy. Two libraries of spectral energy distributions are considered; Siebenmorgen & Krügel (2007) templates and Chary & Elbaz (2001) models. We have supplemented our data with that of Huynh et al. (2007) collected over the GOODS-North field and adapted it to directly compare with the results of this work. The far-infrared luminosity function is then determined using the 1/Vmax technique. A double power law parameterisation is found to provide the best fit to the data. The far-infrared luminosity function was fitted for all parameters and the evolution was measured out to z ∼ 1. Three different types of evolution were allowed, pure luminosity, pure density and luminosity dependent density evolution. In all cases strong positive evolution was evident with the best-fit case being pure luminosity evolution where p = 2.4+0.6 −0.7. Due to the larger volume surveyed compared to previous studies, this work provides better constraints on the bright end of the far-infrared luminosity function displaying a shallower bright end slope (α2 ∼ −1.6) than previously determined, implying a higher number density of the most luminous objects and thereby a greater contribution from these objects to the total infrared energy density. However the shallower slope determined here can be reconciled with other work if the Chary & Elbaz (2001) models are used instead of the Siebenmorgen & Krügel templates; demonstrating that spectral energy distribution model selection is a key component in determining luminosity functions at far-infrared wavelengths. The far-infrared–radio correlation (FIRC; qIR) was determined for the sample of 70μm selected star-forming galaxies using 1.4GHz radio data over the COSMOS field, and no evolution was found out to z ∼ 2. The 70μm monochromatic evolution in the FIRC was also examined (q70) and no evolution was found in this parameter with redshift.

In my thesis I investigate the luminous z ~ 3 Lyman break galaxies in deep wide field surveys. In the first part of the thesis, I use the LBT/LUCIFER to observe a lensed high-redshift star-forming galaxy (J0900+2234) at z = 2.03. With the high S/N near-IR spectroscopic observations, I reveal the detailed physical properties of this high-redshift galaxy, including SFR, metallicity, dust extinction, dynamical mass, and electron number density. In the second part of the thesis, I select a large sample of LBGs at z ~ 3 from our new LBT Bootes field survey, and study the bright end luminosity function (LF), stellar mass function (SMF) and clustering properties of bright LBGs (1L* < L < 2.5L*). Together with other LF and SMF measurements, the evolution of LF and SMF can be well described by continuously rising star formation history model. Using the clustering measurements in this work and other works, a tight relation between the average host galaxy halo mass and the galaxy star formation rate is found, which can be interpreted as arising from cold flow accretion. The relation also suggests that the cosmic star formation efficiency is about 5%-20% of the total cold flow mass. This cosmic star formation efficiency does not evolve with redshift (from z ~ 5 to z ~ 3), hosting dark matter halo mass (10¹¹-10¹³ M⊙), or galaxy luminosity (from 0.3L* to 3L*).In the third and fourth parts, with the spectroscopic follow-up observations of the bright LBGs, I establish a sample of spectroscopically-confirmed ultra-luminous LBGs (ULBGs) in NOAO Bootes field. With this new ULBG sample, the rest-frame UV LF of LBG at M(1700Å) = -23.0 was measured for the first time. I find that the ULBGs have larger outflow velocity, broader Lyα emission and ISM absorption line profiles, and more prominent CIV P-Cygni profile. This profile may imply a top-heavy IMF in these ULBGs. The ULBGs have larger stellar mass and SFR, but smaller dust extinction than the typical L* LBGs at z ~ 2 - 3. We proposed two evolutionary scenarios, pre-burst and post-burst. The properties of the ULBGs, especially the morphologies, prefer the pre-starburst scenario. Further high spatial resolution HST imaging and IFU spectroscopic observations will allow us to distinguish these two scenarios.

This thesis describes a new approach to the study of high-redshift star-formation and its environments that can be applied to large high-redshift surveys. Instead of relying on spectroscopy or narrow-band photometry to study galaxy line emission in detail, the properties of large emission line galaxy (ELG)populations are estimated from broadband photometry by measuring colour-residuals against colours drawn from a set of line-free stochastic burst models-based on (Bruzual & Charlot, 2003). Simulated star-formation histories drawn from semi-analytic and adaptive-mesh-refinement codes were converted into mock galaxy colours, but neither could-span the range of observed galaxy colours at high redshift. Instead, an existing set of exponentially declining star-formation models with stochastic bursts was used, because it closely spanned the range in observed galaxy colours in the bandsthat were line-free at each redshift. Small colour offsets were measured between the models and the observations, corresponding to the equivalent widths (EWs)of Hα, [OIII] and [OII]. In this way, I measure the rest-frame Equivalent Widths of the Hα, [OIII]and [OII] emission lines as they are redshifted through all filters from CANDELS(near-continuous U to 4.5μm coverage) for a large sample of galaxies from z=0.1up to z=5. This approach relies solely on the line-free models, a set of existing reliable photometric redshifts, and a colour cut (B−K < 2 or equivalent) to select only the dust-free young objects (the majority of identified emission-line galaxies). Once correctly identified, I apply this method to the CANDELS-UDS photometry to characterise the properties of Emission-Line Galaxies (ELGs) through these lines. I find that in this sample the Hα and [OIII] ELG fraction with EW > 150Årises from < 5% at z < 1 up to 40% at z > 2. The co-moving ELG density rises from 5 to 30 ×10 −4 /Mpc −3 at z=2.3. The evolution of median Hα EW with redshift is consistent with results from HiZELS and 3D-HST yielding median EW ∼ M 0.25 (1+z) 1.75 up to z=2.3, from which it departs to values of 450Å atz=4.3. [OIII] remains weaker than Hα for z < 3 and matches its values above that redshift. [OIII] also displays a larger fraction of extreme EWs than Hα. [OII], while correctly identified, never becomes as extreme as the other two lines lines, even when corrected for the evolving continuum. This is evidence of an increasing [OIII]/[OII] ratio with increasing z through-out this sample. While these results agree with spectroscopic and narrow-band surveys, the use of the deeper broadband filter coverage enables a systematic measurement of the increasingly prevalent high EWs ( > 500Å) in galaxies at every redshift spanning the 10 8 to 10 10.5 M range. Subsequently, this method was applied to all the other CANDELS fields (GOODS-South and North, COSMOS and EGS) and further corroborates these results. These results further show that EW dependence on mass is steeper for [OIII] than for Hα. Line EWs are then converted into luminosities for the three lines and fitting formulas are obtained, displaying L Hα ∼(1+z) 3.2 M 0.45−0.6log(1+z), with similar results for the other lines. L Hα is converted into star-formation rate and specific star-formation rate (sSFR). sSFR at low-z aligns approximately with the main sequence (with a steeper dependence in mass), but at high-redshift sSFR remains above the main sequence by a factor of 2 and rising towards medians SFR=100/Gyr around log(M/M )=9, showing a departure of the main sequence of star formation at lower masses log(M/M ) < 9.5. The SFRD of ELGs is 1% at low redshift, but rises to 30% at z=4.5. The L [OIII] /L Hα ratio is used to estimate L [OIII] /L Hβ and the ionization parameter q, for which the median atz > 0.5 stays approximately constant at 10 8 cm/s, and increases with mass. Using the L [OIII] /L [OII] ratio and q, median metallicity is shown to be sub-solar, and can be tentatively estimated for z > 0.5 to be Z/Z ∼0.3. The errors are large, but this could also mean a large range in metallicity from Z to 0.1Z . L [OIII] /L [OII] rises with sSFR as shown in the literature. This method shows great potential to survey emission-line-derived physical quantities for large galaxy populations with a low computational footprint, which could be particularly useful for pixel-by-pixel EW imaging. It is also flexibile, which allows it to be applied to any future deep multi-broadband fields.

Magíster en Ciencias, Mención Astronomía / La observación de emisión Lyman alpha (Lyα) proveniente de galaxias es una técnica ampliamente utilizada para el estudio del universo a alto redshift (e.g. Shapley et al. 2003, Ouchi et al. 2008, Stark et al. 2010, Blanc et al. 2011). Sin embargo, a pesar de su importancia, los estudios estadísticos de esta emisión no abordan posibles sesgos al elaborar las muestras de objetos, induciendo discrepancias. Diferentes metodologías inducen selecciones en masa estelar (M∗), tasa de formación estelar (SFR) y metalicidad. Los perjuicios asociados a la forma actual de estudiar la emisión Lyα no solamente se restringen a sesgos en los resultados obtenidos, sino que también limitan nuestro entendimiento de este proceso radiativo en la población de galaxias. En consecuencia, nuestra comprensión de la emisión Lyα en dicha población es actualmente altamente especulativo. Todas estas razones justifican la necesidad de llevar a cabo un estudio de emisión Lyα en la población de galaxias, lo cual es el objetivo de esta tesis. El presente trabajo detalla los resultados de un exhaustivo análisis de emisión Lyα a alto redshift. Para ello, primero diseñamos una muestra de 629 galaxias observadas por CANDELS (Koekemoer et al. 2011, Grogin et al. 2011). Dicha muestra se encuentra especialmente di- señada para abarcar el rango de masa estelar 7.6 < log (M∗[M⊙]) < 10.6 y rango de redshift 3 < z < 4.6. Tras esta selección, realizamos espectroscopía de los objetos con el Michigan/- Magellan Fiber System (M2FS; Mateo et al. 2012). De esta forma, contamos con mediciones de flujo de Lyα, además de la fotometría CANDELS, la cual permite obtener masa estelar, luminosidad, tasa de formación estelar, y extinción para cada galaxia. Observamos que el Ancho Equivalente y Fracción de escape de Lyα decrecen con la masa estelar, luminosidad en el UV, tasa de formación estelar, y extinción de las galaxias. Además, introducimos aplica- ciones de estadística Bayesiana en este campo de la astrofísica. Esto nos permite concluir que un modelo de distribución exponencial es el más apropiado para reproducir nuestras distri- buciones de Ancho Equivalente de Lyα, además de caracterizar la completitud y significancia de las correlaciones que observamos. También estudiamos los impactos que tienen diferentes técnicas de selección de galaxias en las estadísticas observadas de Lyα. Lyman Break Galaxies y estudios de Lyα emitters seleccionan preferentemente galaxias de baja masa estelar y poco polvo interestelar. Finalmente, realizamos la primera predicción semi-analítica de la fracción de Lyα emitters hasta redshift 7, la cual se puede utilizar para restringir observacionalmente la época de reionización. Todos estos resultados no solamente contribuyen a comprender de mejor manera este tipo de radiación, sino que también establecen un nuevo marco para el análisis estadístico de este trazador del universo temprano.