2006 Mini-Workshop: The End of the Dark Ages
From First Light to Reionization
Presentations

The first stars are believed to have formed at redshift z>~20 inminihalos of mass ~10^6 solar masses. The energy released by these first stars had a strong feedback effect on the surrounding gas. One recent study suggests that the ionizing radiation from these first stars promoted subsequent star formation inside nearby minihalos, when the ionized gas there recombined, formed H2 molecules and radiatively cooled (O'shea et al. 2005). Another study, however, finds that the ionizing radiation from the first stars fails to penetrate the neighboring minihalos and, therefore, does not lead to the dense, fully ionized halo gas which previous work required to stimulate H2 and star formation there (Alvarez, Bromm, and Shapiro 2005). We have attempted to resolve this issue by radiation-hydrodynamical simulations of the impact of the ionizing and dissociating radiation from the first stars on the gas in nearby minihalos, during the lifetime of the star and beyond. Our results indicate that the ionization front from the first star is, indeed, trapped by the dense gas inside the nearby minihalos, but that this I-front is preceded by a partially ionized shell of gas in which H2 formation is possible, despite the presence of dissociating UV radiation from the external star. This H2 shell can shield the inner minihalo gas from dissociating photons and expedite the gravitational collapse of H2-cooled gas in the core, even while the neighboring star continues to shine. This result has implications for the widely-held view that the UV background from the first stars easily dissociates the H2 in other minihalos and thereby suppresses further star formation.

Recent observations indicate that the epoch of reionization ended at redshift z ~ 6, motivating direct searches for the star forming galaxies that are its the most likely cause. Star-forming galaxies at redshifts z > 7 may be detectable and mappable with present-day technology. I will describe an ongoing narrow-band search with Gemini for z=8.2 galaxies in the near infrared, comparing it to other techniques for mapping First Light. I will also explore and compare the capabilities of future facilities such as TMT and JWST for finding and studying these galaxies.

I will summarize what we know about z~6 galaxies from the observations.

I will discuss how the universe might have been reionized, how one might construct a physical picture that is consistent with several different, seemingly contradictory observations, how the consequence of reionization may be related to star formation processes at the high redshift universe, when confronted with upcoming 3-yr WMAP data.

With the advent in the near future of the new generation of low frequency radio telescopes a new window on the high-redshift universe will be opened. In particular, it will be possible, for the first time, to observe the 21cm signal from the diffuse Intergalactic Medium (IGM) prior to its complete reionization and thus probe the "dark ages". In this talk I will present results of simulations of IGM reionization and I will discuss the observability of the 21cm signal from the diffuse IGM.

I will discuss the evolution, in the range z = 5 - 6.5, of potential neutral intergalactic medium tracers, such as OI/CII absorption lines and dark gaps in the quasar spectra, and compare with the evolution of ionized tracers, such as Lyman alpha emitting galaxies and the transmission spikes in the Lyman forest region of the quasar spectra. We might hope to see dramatic evolution in the relative behaviour of these tracers at the reionization boundary but as yet there is no strong evidence of such a change.

I will provide an update of the growing evidence that the assembled stellar mass at z~5-6 implies an earlier era of significant star formation, activity that has yet to be directly detected by extant searches. This puzzle provides a strong motivation for new ground-based infrared initiatives probing the redshift range 7-20 over the next decade, both in blind and targeted searches done in conjunction with HST, Spitzer and eventually JWST. I will review issues of strategy as well as upcoming instruments on existing telescopes, concluding briefly with the status of the Thirty Meter Telescope and its likely impact in this area.

To be determined.

Recent observations paint a complex picture of reionization, implying that the ionizing sources and the intergalactic medium interact through a rich set of physical processes. Unfortunately, disentangling these processes requires new windows into this epoch. I will discuss an analytic model for the growth of ionized bubbles during reionization analogous to models for the dark matter halo mass function. This model has a number of implications for the process, including identifying the most salient properties of the first sources, the relevant length scales, and the role of recombinations. I will also discuss the implications of this model for several potential observational probes of reionization, including quasar absorption lines, Lyman-alpha emitters, and redshifted 21 cm emission.

The scientific capabilities of the James Webb Space Telescope (JWST) fall into four themes. The End of the Dark Ages: First Light and Reionization theme seeks to identify the first luminous sources to form and to determine the ionization history of the universe. The Assembly of Galaxies theme seeks to determine how galaxies and the dark matter, gas, stars, metals, morphological structures, and active nuclei within them evolved from the epoch of reionization to the present. The Birth of Stars and Protoplanetary Systems theme seeks to unravel the birth and early evolution of stars, from infall onto dust-enshrouded protostars, to the genesis of planetary systems. Planetary Systems and the Origins of Life theme seeks to determine the physical and chemical properties of planetary systems around nearby stars and of our own, and investigate the potential for life in those systems. To enable these for science themes, JWST will be a large (6.5m) cold (50K) telescope launched to the second Earth-Sun Lagrange point early in the next decade. It is the successor to the Hubble Space Telescope, and is a partnership of NASA, ESA and CSA. JWST will have four instruments: The Near-Infrared Camera, the Near-Infrared multi-object Spectrograph, and the Tunable Filter Imager will cover the wavelength range 0.6 to 5 microns, while the Mid-Infrared Instrument will do both imaging and spectroscopy from 5 to 27 microns.

Reionization is a complex process, many features of which can only be modelled by detailed numerical radiative transfer simulations. I will review the current status of the field. In particular, I will discuss the basic requirements for such simulations in order for them to faithfully describe the different aspects of reionization and the scales relevant to the future observations. I will review the variety of approaches that are currently used and will show results from a Cosmological Radiative Transfer Code Comparison project, currently in progress, which involves most of the community and a wide variety of codes. This project aims to asses the reliability and accuracy of the different approximations employed, and will serve as a benchmark for future code development. Finally, I will briefly discuss recent simulation results and their implications for the observability of this important epoch.

We report an extensive search for Lyman alpha emitters at z=6.5 (z6p5LAEs) in the Subaru Deep Field. Subsequent spectroscopy with Subaru/Keck identified eight more z6p5LAEs, giving us a total of 17 spectroscopically confirmed z6p5LAEs. Based on these 17 spectroscopic sample complemented by the 58 photometric sample, we derived a more accurate Ly-alpha luminosity function of z6p5LAEs, that reveals an apparent deficit at the bright end, of ~0.75 mag. fainter L^*, compared with that of z=5.7. The difference of LAE luminosity functions between z=5.7 and z=6.5 has a 3-sigma significance, which is reduced to 2-sigma when taking into account the cosmic variance. This result could be an implication that the reionization of the universe has not been completed at z=6.5. The spatial distribution of z6p5LAEs was found to be homogeneous over the field. We discuss some implications to the reionization of the universe.

Candidate AGN at redshifts ~ 6 - 7 can be selected by means of their lack of optical detections, as compared with their relatively bright X-ray flux. The discovery of Extreme X-ray / Optical sources (`EXO's) has provided samples of such sources, which may also include a number of intermediate-redshift evolved galaxies. I will describe recent progress on distinguishing between intermediate-redshift and z ~ 6 - 7 sources, by combining deep optical, near-IR and Spitzer data to obtain constraints on the spectral energy distributions of these sources, combined with X-ray data from the Chandra Deep Fields. The resulting samples of candidate z ~ 6 - 7 AGN are used to examine the evolution of the faint end of the AGN luminosity function at high redshift, with corresponding implications for the co-evolution of galaxies and their central black holes.

Semi-analytical models play an important role in helping us to understand the effects of basic physical processes accompanying reionization. Furlanetto, Zaldarriaga and Hernquist (2004, ApJ, 613, 1) developed an elegant, excursion-set-based approach to constructing the HII region mass function, and from that the ionization history of the Universe. Their model, like all analytic models to date, did not include any of the feedback mechanisms suspected to play a role in regulating this process. We have extended their formalism to take into account suppression of baryonic structure formation in small dark matter haloes that form within already-ionized regions. While this is a simple effect to model if haloes are uniformly distributed, it is considerably more difficult to understand feedback in the more realistic case of highly-clustered halo formation. I will present results showing that clustering greatly enhances the importance of feedback during the early stages of reionization.

The neutral hydrogen and ionized helium absorption in the spectra of quasars probe the structure of intergalactic gas in the early universe. At redshifts above 3.5, the universe is opaque to He II absorption. At lower redshifts, this opacity decreases as the extragalactic background radiation gradually ionizes the He II gas. The epoch of He II reionization has several indicators that currently make it more precisely determined than that of H I. Observations of He II Lyman-alpha absorption may then serve as a qualitative analog of the observational appearance of H I Lyman-alpha at the epoch of H I reionization. We will present recent FUSE observations of the He II Lyman-alpha forest along the lines of sight to two different quasars, HE2347--4342 and HS1700+64, and HST observations of higher-redshift quasars spanning the redshift range of He II reionization.

The visibility of Lyman-alpha emission from galaxies should be suppressed completely or partially at redshifts beyond reionization. By comparing luminosity function of Lyman-alpha emitters at z=5.7 and z=6.5, we conclude a factor of three attenuation of the lyman-alpha line is ruled out and that the IGM is at least 50% ionized at z=6.5. By counting the number density of such emitters, one can also infer that at least 20-50% of the IGM by volume is ionized at redshift 6.5. This method is scalable, can be applied at other redshifts and is capable of yielding reionization history of the universe.

A number of radio interferometers are currently being planned or constructed to observe 21 cm emission from reionization. Not only will such measurements provide a detailed view of the Epoch of Reionization, but, since the 21 cm emission also traces the distribution of matter in the Universe, this signal can be used to constrain cosmological parameters at redshifts 6 ~< z ~< 20. The sensitivity of an interferometer to the cosmological information in the signal may depend on how precisely the angular dependence of the 21 cm 3-D power spectrum can be measured. Here, we quantify all the effects that break the spherical symmetry of the 3-D 21 cm power spectrum and produce physically motivated predictions for this power spectrum, utilizing an analytic model for reionization. We find that upcoming observatories will be sensitive to the 21 cm signal over a wide range of scales, from larger than 100 comoving Mpc to as small as 1 comoving Mpc. We discuss how cosmological parameters can be measured, and show that the first generation of 21 cm observations should moderately improve existing constraints on cosmological parameters for certain low-redshift reionization scenarios. A two year observation with the second generation interferometer MWA5000 in combination with the CMB telescope Planck can improve constraints on Omega_w (to +- 0.019, a 30% improvement over Planck alone), Omega_m h^2 (+- 0.0011, 50%), Omega_b h^2 (+- 0.00013, 30%), Omega_nu (+-0.003, 300%), n_s (+- 0.0034, 30%), and alpha_s (+- 0.004, 100%). Larger interferometers, such as SKA, have the potential to do even better. If the Universe is substantially ionized by z ~ 12 or if spin temperature fluctuations are important, we show that it will be difficult to place competitive constraints on cosmological parameters with the 21 cm signal.

I present results of the search for a new population of extremely massive and evolved galaxies at z > 5.

We use the cosmological adaptive mesh refinement code Enzo to do a suite of high-resolution numerical simulations of the formation of Population III protostellar clouds in a cosmological context. These simulations cover a range of box sizes and distributions of large-scale structure, and in each calculation we follow the evolution of the most massive halo in the simulation volume until the collapse of the first primordial protostellar core. We find that these cores have a wide variety of accretion rates - varying by over two orders of magnitude - which may have significant implications for the IMF of Population III stars. We discuss these results, as well as the need for future calculations with more physics (e.g. radiation transport and MHD).

I will present our results and on-going efforts of clustering analysis for Ly-a emitters and dropout galaxies at z=3-7. We have carried out very deep wide-field optical broad- and narrow-band imaging down to i'=27.5 for the contiguous 1 deg^2 field of the Subaru/XMM Deep Field with Subaru/Suprime-Cam. We have obtained sky distributions of 21,114 dropout galaxies at z=3-6 and 356, 101, and 401 Ly-a emitters (LAEs) at z=3.1, 3.7 and 5.7, respectively. After the completion of this survey, we will obtain ~200 LAEs at z=6.6. Based on the large sample found in this wide-field (~150Mpc x 150Mpc at z=3-7 in projected comoving scale), we derive angular correlation function of these LAEs and dropout galaxies. Our preliminary results show that clustering of LAEs is stronger at z=5.7 than z=3-4 by a factor of ~2-3 in bias factor for the samples with the same number density. On the other hand, clustering of dropout galaxies does not show a significant evolution from z=6 to 3. This difference of clustering evolution provides an interesting insight in the relation between LAEs and the neutral fraction of IGM toward the end of reionization era. Our coming results of z=6.6 Ly-a emitter will probably give constraints on the neutral fraction as well as the size of ionized bubbles. Finally, I will introduce our discovery of filamentary large-scale structures made of LAEs at z=5.7 and two concentration of LAEs at z=5.67 and 5.69. These LAE clumps show about ~100 times higher Ly-a flux densities in a sphere with a diameter of 1 Mpc. I will discuss the connection between these LAE clumps and the reionization of the universe.

Reionization is a process that depends on the ultraviolet (UV) output of galaxies integrated over the time interval when they are active UV sources. This fact is particularly relevant for the case of Balmer-break galaxies (BBG) because, when these galaxies are discovered, they have already completed their cycle of intense star formation. We discuss here the implications that the recently discovered population of BBGs at z>5 (Wiklind et al 2006) have for the reionization of the Inter-Galactic Medium (IGM) in the early phases of the Universe evolution.

I discuss several possible sources sources of reionization such as globular clusters, population III stars and accreting black holes and how reionization may happen in these different scenarios.

When the universe was reionized, small-scale cosmic structure exerted a strong feedback effect which left its imprint on all scales from the smallest to the largest. When the first stars formed inside minihalos of mass ~ 106 solar masses at redshifts z > 20, they released ionizing and dissociating UV radiation which heated and expelled the gas inside their host minihalos and leaked out into the surrounding universe, to create intergalactic H II regions and begin cosmic reionization. As the H II regions grew, the ionization fronts which led their expansion encountered other minihalos, which blocked their path and trapped them, causing this minihalo gas, too, to be expelled in a photoevaporative wind. Under some conditions, this ionizing radiation may actually have stimulated the dense, central regions of these minihalos to form enough hydrogen molecules to shield themselves from photodissociating radiation, cool radiatively and form more stars. Eventually, hierarchical clustering led to the formation of more and larger dwarf galaxy sources of ionizing radiation in excess of 108 solar masses, in which atomic cooling was effective, and their intergalactic H II regions grew and merged to become 10's of comoving Mpc's in size. Inside these photoheated H II regions, low-mass dwarf galaxies were inhibited from forming by gravitational instability, so the minimum mass of newly-formed galaxy sources rose to more than a billion solar masses. The process of reionization continued until the intergalactic H II regions finally overlapped everywhere, to fill space with highly ionized gas. We have studied this process by a variety of techniques, on a hierarchy of mass- and length-scales. The latter span the range from the interiors of minihalos, to the giant H II regions produced by the clustered formation of galaxies in space and in time, to the large-scale structure of the patchy distribution of neutral and ionized gas in the universe during the epoch of reionization. Our studies include analytical and semi-analytical treatments of ionization fronts, numerical radiation-hydrodynamical simulations which "zoom-in" on the photoevaporation of individual halos, and the first, truly large-scale radiative transfer simulations of global reionization. We have made extensive predictions of the redshifted cosmic 21-cm background from the dark ages and the epoch of reionization for comparison with observations by future radio arrays. I will summarize these theoretical developments in this talk.

I will discuss observations of galaxies at extreme redshifts (z>5) and what these tell us about the history of star formation and the assembly of stellar of stellar mass. I will describe how HST imaging with ACS (for example the Ultra Deep Field) has uncoverred star forming "i-band dropout" galaxies, which have not been spectroscopically confirmed. Recent Spitzer images tell us about the rest-frame optical light, and indicate that some of these star forming galaxies at redshift 6 also have an older stellar population, which may have formed at redshifts beyond 10. I will discuss the implications for the reionization of the Universe.

I will briefly discuss the Hubble Ultra Deep Field and its present and future followup observations and their potential to investigate the Universe at redshift greater than 7. I will then discuss the strategies that the James Webb Space Telescope will adopt in order to study the evolution of galaxies at the end of the Cosmic Dark Ages, i.e. in the era between reionization and first light.

We have traced the star formation history of the NICMOS region of the Hubble Ultra Deep Field (HUDF) to a redshift of 6 and have compared it to similar results in the Northern Hubble Deep Field (NHDF). The HUDF offers and independent measure of the star formation history in a region that is completely uncorrelated with the NHDF. We find that the star formation rate (SFR) is roughly constant between redshifts of 1 to 6, averaging a few tenths of a solar mass per yer per comoving cubic megaparsec. We also investigate the nature of the star forming galaxies and find that 90% of the star formation occurs in 10-20% of the galaxies and that star formation is concentrated in high SFR regions. We also address the reality of a near infrared background excess and find little evidence for an excess. We elaborate the severe constraints that the NICMOS NHDF and HUDF observations place on any excess.

The gamma-ray burst (GRB) 050904 at z = 6.3 provides the first opportunity of probing the intergalactic medium (IGM) by GRBs at the epoch of the reionization. Here we present a spectral modeling analysis of the optical afterglow spectrum taken by the Subaru Telescope, aiming to constrain the reionization history. The spectrum shows a clear damping wing at wavelengths redward of the Lyman break, and the wing shape can be fit either by a damped Ly alpha system with a column density of log N_HI ~ 21.6 at a redshift close to the detected metal absorption lines (z_{metal} = 6.295), or by almost neutral IGM extending to a slightly higher redshift of z_{IGM,u} ~ 6.36. In the latter case, the difference from z_{metal} may be explained by acceleration of metal absorbing shells by the activities of the GRB or its progenitor. However, we exclude this possibility by using the light transmission feature around the Ly beta resonance, leading to a firm upper limit of z_{IGM,u} < 6.314. We then show an evidence that the IGM was largely ionized already at z=6.3, with the best-fit neutral fraction of IGM, x_HI = 0.00 +- 0.17, and an upper limit of x_HI < 0.60 (95% C.L.). This is the first direct and quantitative upper limit on x_HI at z > 6. Various systematic uncertainties are examined, but none of them appears large enough to change this conclusion. Prospects of future GRB data as the reionization probe will be discussed.

The most distant quasars act as beacons for the study of the gas in the intergalactic medium when the universe was less than one billion years old. We argue that there is strong evidence for a change in the ionized fraction of hydrogen at z=6, which appears to mark the end of the epoch of reionization. However, this claim remains controversial. I will describe the current status of our work and will review some of the arguments for and against the view that we are observing evidence of reionization at z=6.

Applying a selection technique based on near- and mid-infrared photometric data and targeting the Balmer break feature, it is possible to define a sample of galaxies that are old (>200 Myr), massive (~10^11 Msun) and situated at redshifts z > 5. We have used this technique on a K-selected sample of galaxies from the GOODS South field and studied the candidates using population synthesis models. We find 18 candidates with ages ranging from 200-800 Myr and redshifts z=5-7. The average stellar mass is 2 10^11 Msun and the formation redshifts of these systems range from z=10-26. These galaxies are currently quiescent concerning star formation. We compare the number density of these so called Balmer break Galaxies (BBGs) with expectations from Lamda CDM models and with the density of massive systems at lower redshifts.

The Square Kilometer Array (SKA) will be one of a suite of new, large telescopes for the 21st century probing fundamental physics, the origin and evolution of the Universe, the structure of the Milky Way Galaxy, and the formation and distribution of planets. The epoch of reionization (EoR) sets a fundamental benchmark in cosmic structure formation, corresponding to the formation of the first luminous objects and the end of the Dark Ages. Recent observations at near-IR and radio wavelengths imply that we are finally probing this key epoch of galaxy formation at z > 6. The SKA will provide critical insight into the EoR (for a review, see Carilli et al. 2004). First, the ability of the SKA to image the neutral IGM in 21cm emission is a truly unique probe of reionization and is the next necessary and fundamental step in our study of the evolution of large scale structure and cosmic reionization. Second, study of HI 21 cm absorption toward the first radio loud objects probes small to intermediate scale structure in the neutral IGM. And third, the incomparable sensitivity of the SKA allows for the study of the molecular gas, dust, and star formation in the first galaxies, as well as the radio continuum emission from the first accreting massive black holes. Such objects will be obscured at optical wavelengths due to absorption by the neutral IGM.

We present results from a large volume simulation of Hydrogen reionization. We combine 3D radiative transfer calculations and an N-body simulation, describing structure formation in the intergalactic medium (IGM), to detail the growth of HII regions around high redshift galaxies. Our N-body simulation tracks 1024^{3} dark matter particles, in a cubical box of co-moving side length L_{ m box} = 65.6 Mpc h^{-1}. This large volume allows us to accurately characterize the size distribution of HII regions throughout most of the reionization process. At the same time, our simulation resolves most of the small galaxies likely responsible for reionization. We compare this simulation with a fast, Monte-Carlo calculation that associates HII regions with large-scale overdensities and is based on the excursion set formalism. We find that the analytic calculation reproduces the simulated size distribution of HII regions, the power spectrum of the ionization field, and the 21cm power spectrum remarkably well. The simulated ionization field, however, has more small scale structure than in the analytic calculation, a fact we attribute to Poisson scatter in the abundance of galaxies on small scales. We propose and validate a simple scheme to incorporate this scatter into our Monte-Carlo calculations. Our results suggest that analytic calculations are sufficiently accurate to aid in predicting and interpereting the results of future 21cm surveys, especially in conjunction with detailed simulations.

We use the cosmological adaptive mesh refinement code Enzo to do high-resolution numerical simulations of Population III supernovae in a cosmological context. In our simulations, we find a halo in which a Population III star would form at high redshift (z ~ 20) and artificially place a supernova at the halo's center, using both analytical (Sedov) supernova models as well as cutting-edge asymmetric supernova models from Fryer, Rockefeller & Heger. We then advance the simulations for roughly a Hubble time, following the evolution of the supernova remnant until it reaches pressure equilibrium with the pre-galactic medium. The ejecta from even the smallest Type II supernovae can reach distances several hundred proper parsecs away, implying that many neighboring halos can become contaminated with metals. We examine the effects of this metal contamination on nearby halos and discuss the formation of a second generation of smaller, metal-enriched stars.

For objects at high redshift we estimate the effects on the reddening and on the optical depth of dust present in absorption systems along the line of sight. In our modeling we allow for different number densities and dust-to-gas ratios of Damped Lyman Alpha absorbers, taking into account their chemical evolution. We characterize the probability distribution of the mean optical depth to a given redshift and the shape of the effective mean extinction law by means of analytical estimates and Monte Carlo simulations. We present our results in a format useful for applications to present samples of high redshift galaxies and discuss the implications for observations of first light objects with the James Webb Space Telescope.