3
Report of the Panel on Galaxies Across Cosmic Time

I get wisdom day and night

Turning darkness into light.


—St. Paul Irish Codex, translated by Robin Flower

SUMMARY

The study of galaxies across cosmic time encompasses the main constituents of the universe across 90 percent of its history, from the formation and evolution of structures such as galaxies, clusters of galaxies, and the “cosmic web” of intergalactic matter, to the stars, gas, dust, supermassive black holes, and dark matter of which they are composed. Matter accretes into galaxies, stars form and evolve, black holes grow, supernovae and active galactic nuclei expel matter and energy into the intergalactic medium (IGM), galaxies collide and merge—and what seemed a static world of island universes only a few decades ago turns out to be a lively dance of ever-changing elements. Across all epochs, these processes are coupled in a complicated evolutionary progression, from the relatively smooth, cold universe at high redshift (z > 40 or so) to the highly structured cosmos of galaxies and intergalactic matter today.

The Astro2010 Science Frontiers Panel on Galaxies Across Cosmic Time began its deliberations by reading the extensive set of white papers submitted by the astronomical community to the National Research Council (NRC) at the request of the Committee for a Decadal Survey of Astronomy and Astrophysics. The panel reviewed the substantial advances in the understanding of galaxy and structure evolution that have occurred over the past decade or two. It then identified the four key questions and one discovery area that it believes will form the focus for research in the coming decade:



The National Academies | 500 Fifth St. N.W. | Washington, D.C. 20001
Copyright © National Academy of Sciences. All rights reserved.
Terms of Use and Privacy Statement



Below are the first 10 and last 10 pages of uncorrected machine-read text (when available) of this chapter, followed by the top 30 algorithmically extracted key phrases from the chapter as a whole.
Intended to provide our own search engines and external engines with highly rich, chapter-representative searchable text on the opening pages of each chapter. Because it is UNCORRECTED material, please consider the following text as a useful but insufficient proxy for the authoritative book pages.

Do not use for reproduction, copying, pasting, or reading; exclusively for search engines.

OCR for page 95
3 Report of the Panel on Galaxies Across Cosmic Time I get wisdom day and night Turning darkness into light. —St. Paul Irish Codex, translated by Robin Flower SUMMARY The study of galaxies across cosmic time encompasses the main constituents of the universe across 90 percent of its history, from the formation and evolution of structures such as galaxies, clusters of galaxies, and the “cosmic web” of interga- lactic matter, to the stars, gas, dust, supermassive black holes, and dark matter of which they are composed. Matter accretes into galaxies, stars form and evolve, black holes grow, supernovae and active galactic nuclei expel matter and energy into the intergalactic medium (IGM), galaxies collide and merge—and what seemed a static world of island universes only a few decades ago turns out to be a lively dance of ever-changing elements. Across all epochs, these processes are coupled in a complicated evolutionary progression, from the relatively smooth, cold universe at high redshift (z > 40 or so) to the highly structured cosmos of galaxies and intergalactic matter today. The Astro2010 Science Frontiers Panel on Galaxies Across Cosmic Time be- gan its deliberations by reading the extensive set of white papers submitted by the astronomical community to the National Research Council (NRC) at the request of the Committee for a Decadal Survey of Astronomy and Astrophysics. The panel reviewed the substantial advances in the understanding of galaxy and structure evolution that have occurred over the past decade or two. It then identified the four key questions and one discovery area that it believes will form the focus for research in the coming decade: 95

OCR for page 95
Panel rePorts—new worlds, new HorIzons 96 • How do cosmic structures form and evolve? • How do baryons cycle in and out of galaxies, and what do they do while they are there? • How do black holes grow, radiate, and influence their surroundings? • What were the first objects to light up the universe and when did they do it? • Unusual discovery potential: the epoch of reionization. To maximize progress in addressing these issues, the panel considered the wide array of observational and theoretical programs made possible by current or future facilities. Observational programs were discussed in sufficient detail to allow an understanding of the requirements (numbers of objects, sensitivity, area, spatial resolution, energy resolution, etc.) so that this panel could provide the most use- ful input to the study’s Program Prioritization Panels (PPPs; see the Preface for further information on this process); however, any assessment of the suitability of existing or proposed facilities to the key science issues outlined here is left to the PPPs and the survey committee. This report describes the scientific context for the area “galaxies across cosmic time,” and identifies the key science questions in this area for the next decade and a set of science programs—observational and theoretical—that will answer the most important questions in the field. Some of these programs would require new ob- servational facilities, whereas others could be done with existing facilities, possibly with a reprogramming of resources. In order to provide more useful input to the Astronomy and Astrophysics 2010 (Astro2010) Survey, the top science programs selected by the panel for purposes of this report are identified in three categories: most important, very important, and important. The panel considered many other programs that were eventually excluded from its list but that remain valuable ways to make progress, and it anticipates that significant progress will also come from unexpected directions. This Summary addresses each of the four key questions in turn, listing only the programs ranked “most important,” plus those “very important” activities that rep- resent unique capabilities. The full set of the panel’s top-ranked science programs is summarized in Table 3.1 at the end of this “Summary” section and is presented with rankings and further details in the body of the report. How Do Cosmic Structures Form and Evolve? The answer to this question starts with an understanding of the structure of dark matter halos on all scales. The now-standard lambda cold dark matter— ΛCDM—cosmology provides a detailed foundation on which a theory of galaxy formation and evolution can be built and which in turn can be tested by data.

OCR for page 95
rePort Panel GalaxIes across cosMIc tIMe 97 of tHe on ΛCDM does seem to be validated on the largest scales of the cosmic web and su- perclusters, but some of its predictions seem to deviate seriously from observations on smaller scales, from clusters of galaxies, down to galaxies themselves. Specifically, theory predicts that, even after small clumps of dark matter have merged to form ever-larger structures, many of the small clumps should survive intact, embedded within the merged halos. Yet observations appear to indicate that the dark matter in halos is much less “lumpy” than predicted by the straightforward calculations. Direct constraints on the dark matter distribution can be derived from observa- tions of gravitational lenses, both weak and strong. The panel therefore concluded: • It is most important to obtain Hubble Space Telescope (HST)-like imaging to determine the morphologies, sizes, density profiles, and substructure of dark matter, on scales from galaxies to clusters, by means of weak and strong gravita- tional lensing, in lens samples at least an order-of-magnitude larger than currently available. HST can make an important start on this problem, but to develop large statistical samples will require a much larger field of view or more observing time than HST affords. The best current calculations of cluster formation suggest that gas in the dens- est regions should cool more than is observed, and that more stars should form in cluster cores, especially in the richest clusters. Perhaps the physical processes that affect baryons in clusters need to be better understood, or perhaps extra energy is injected from supernovae, an active nucleus, or some other source. One critical missing piece of information concerns the dynamics of the hot intracluster gas: how turbulent is the gas, how does it flow through the cluster, what is its ionization and velocity structure, and how do these properties depend on cluster richness and cosmic epoch (redshift)? The panel concluded: • High-energy-resolution, high-throughput X-ray spectroscopic studies of groups and clusters to z ~ 2 are most important for understanding the dynamics, ionization and temperature structure, and metallicity of the hot intracluster gas, as well as for studying the growth of structure and the evolution of correlations among cluster properties. Much is still not known about how galaxies were assembled. The well-defined correlations observed among the shapes, sizes, velocity structures, and composi- tions of galaxies, observed mainly in the local universe, are poorly understood. A Sloan Digital Sky Survey (SDSS)-size spectroscopic survey at z ~ 1-3 would pro- vide essential information about the evolution of galaxy correlations and should provide essential clues to the process of galaxy formation and evolution. The panel concluded:

OCR for page 95
Panel rePorts—new worlds, new HorIzons 98 • It is very important to obtain moderate-resolution multi-slit spectroscopy of SDSS-size galaxy samples at z ~ 1-3, in the optical for z < 1.5, and in the near- infrared (IR) for z > 1.5 (with resolution [R] ~ 5,000 to allow effective removal of night skylines in the near-IR). For a representative subset of hundreds of galaxies, high-angular-resolution integral field unit (IFU)1 spectroscopy in the optical or near-IR would help calibrate the slit spectra. To select targets for spectroscopy requires optical/IR pre-imaging over a large area. How Do Baryons Cycle in and out of Galaxies and What Do They Do While They Are There? Along with galaxies, clusters, and dark matter, diffuse baryonic gas is a key part of the cosmic web; indeed, it represents most of the baryonic mass in the universe. The metal enrichment of the gas indicates that a great deal of it was processed through stars in the past, yet little is understood about how galaxies acquire gas across cosmic time, convert it to stars, and eject it back into the IGM. To under- stand this process will require the kind of detailed study of galaxies in the young universe that was done for the local universe with large surveys such as the SDSS and the Two-degree Field Galaxy Redshift Survey. To create a full evolutionary picture for galaxies, study of the following is needed: the star-formation rate, active galactic nucleus (AGN) activity, star- formation history, stellar mass, and stellar and gas-phase metallicity in galaxies at z ~ 1-3, when cosmic star formation and black hole growth rates peaked. By quantifying the correlations of these properties with one another and with the larger-scale environment, astronomers can trace the evolution of galaxies and the baryons within them from the galaxies’ origins to the present day. These detailed galaxy properties are accessible through rest-frame optical spectra that have suffi- cient resolution to measure dynamical and stellar population parameters, sufficient continuum sensitivity to measure absorption lines, and sufficient emission-line sensitivity to measure low levels of star formation (see Figure 3.13 below in this report). The galaxy samples must be large enough to disentangle the covariances among galaxy properties such as luminosity, mass, age, morphology, and metal- licity, over volumes large enough to sample representative galaxy environments. A wide-area survey would trace luminous galaxies, while a smaller-volume survey could probe deeper in order to study the fainter progenitors of typical galaxies today. To develop a complete view of galaxies in the peak epoch of galaxy forma- tion, comparable to the understanding of galaxies in the local universe, the panel concluded: 1 Integral-field units provide spatially resolved spectroscopy, usually across a contiguous field.

OCR for page 95
rePort Panel GalaxIes across cosMIc tIMe 99 of tHe on • It is most important to carry out an SDSS-size near-infrared spectroscopic survey of galaxies at 1 < z < 3 using multiobject spectrographs. This will require near-infrared advance imaging in the J, H, and K bands (at 1, 1.6, and 2 microns) to select targets for spectroscopy. Properly designed, the same large near-IR spec- troscopic survey could serve the first key question as well. To probe baryons when they are in and around and between galaxies, one can use absorption spectra of background sources along lines of sight passing near galaxies. Such techniques probe both the gas distribution and its velocity field and will yield insights into gas accretion, outflows, chemical enrichment, and the over- all cycle of matter between galaxies and the IGM. Theoretical simulations will be critical for connecting such one-dimensional probes to the three-dimensional gas distribution. At z < 1.5, the principal absorption lines of gas outflowing from galax- ies and quasars are in the ultraviolet (UV). UV absorption-line spectroscopy also provides an alternative to X-rays in searching for the “missing baryons” thought to comprise a warm-hot intergalactic medium (WHIM). It may also be possible to image the WHIM directly using IFUs in the UV. The panel therefore concluded: • It is most important to use extremely large optical/infrared telescopes (ELTs) to map metal- and hydrogen-line absorption from circumgalactic and dense fila- mentary intergalactic gas, at moderate resolution toward background galaxies and at higher resolution toward background quasars. • A 4-meter-class, UV-optimized space telescope, equipped with a high- resolution spectrograph and an IFU for spectral mapping, is very important for characterizing outflows from galaxies and AGN at z < 1.5 and for mapping the WHIM. A complete inventory of cold gas in and around galaxies is also crucial for understanding baryon cycling. Molecular gas traced by carbon monoxide (CO), neutral carbon atoms (C I), and higher-density probes provides the raw material for star formation. Neutral atomic gas in the circumgalactic medium likely feeds the growth of galaxy mass. Direct observations of cold gas will make it possible to test theoretical models for complex gas physics and predictions for the evolution of gas content. For the construction of a seamless picture of how gas is processed into stars during the epoch 1 < z < 3, when roughly half of the stellar mass in the universe was formed, a complete inventory of cold gas in and around galaxies is needed. The panel therefore concluded: • It is most important to detect CO emission from a representative sample of typical star-forming galaxies from z ~ 1-3, to develop technology for faster spec-

OCR for page 95
Panel rePorts—new worlds, new HorIzons 100 troscopic follow-up in the (sub)millimeter, and to develop large-collecting-area facilities to study neutral hydrogen (H I) in emission at z ~ 1-3. Accurately characterizing star formation that is obscured by dust is critical for obtaining a complete view of baryon processing during the epoch of galaxy formation. Complementing z ~ 1-3 rest-frame optical spectroscopy with radio and submillimeter imaging, as well as far-IR spectroscopy of the dustiest systems, would provide a complete synthesis. The panel concluded: • It is very important to do sensitive radio and (sub)millimeter continuum mapping over large areas, preferably coincident with a near-IR (rest-frame opti- cal) spectroscopic survey such as the one described above, and to carry out far-IR spectroscopy of luminous dusty galaxies. How Do Black Holes Grow, Radiate, and Influence Their Surroundings? Supermassive black holes (SMBHs), a prediction of Einstein’s general theory of relativity, are ubiquitous within our galaxy and throughout the universe. Observa- tions over the past decade suggest that they play an important role in the evolu- tion of galaxies and clusters. It is still uncertain how and when these black holes form, grow, produce relativistic jets, and feed energy back into the environment. The strong correlation between black hole mass and galaxy mass hints at tightly coupled coevolution and possibly a strong regulatory effect of one on the other. In galaxy clusters, there is equally intriguing evidence that energy liberated by ac- creting black holes—carried by jets or winds—regulates the thermal evolution of the intracluster gas. Gas swirling into SMBHs in luminous AGN apparently forms a nearly Keplerian, thin accretion disk, much as predicted more than 30 years ago. X-rays reflected from the disk are imprinted with spectral signatures that encode the dynamical state of the gas and the relativistic curvature of space-time around the black hole. Coupled with sophisticated computer simulations, these signatures can be used to probe the physics of black holes and accretion disks directly and to determine the spin distribution function of the local SMBH population. The structure of AGN accre- tion disks and jets can also be explored through X-ray polarization measurements. To understand the details of accretion onto supermassive black holes, jet for- mation, and energy dissipation, the panel concluded: • It is most important to have sensitive X-ray spectroscopy of actively accreting black holes (AGN) to probe accretion disk and jet physics close to the black hole as well as to determine the spin distribution function of the local SMBH popula- tion. The effective area should be sufficient to detect the iron Ka emission line on

OCR for page 95
rePort Panel GalaxIes across cosMIc tIMe 101 of tHe on dynamical timescales in a modest sample of the brightest AGN, yielding both spin and mass. To disentangle the effects of absorption in AGN spectra, high resolution (R > 2,000) is required. The same capabilities will yield time-averaged line profiles of more than a hundred AGN with sufficient signal-to-noise ratios to derive the black hole spin distribution. Most of the evidence for black hole feedback into the intracluster medium (ICM) is either morphological or based on low-spectral-resolution temperature measurements. But since such feedback is thought to occur primarily by way of the kinetic energy of jets and winds, kinematic measurements would provide a more direct test. High-throughput, high-resolution X-ray spectroscopy will reveal bulk motions and turbulence in the ICM, allowing the AGN/ICM coupling to be explored. In order to seek evidence of black hole feedback, the panel concluded: • It is most important to measure turbulence and/or bulk flows using X-ray imaging spectroscopy of the ICM of nearby galaxy clusters and groups, with suf- ficient image quality, field of view, energy resolution, and signal to noise to provide ionization and velocity maps on the scale of the interaction between the AGN outflow (e.g., radio source) and the gas. A census of black holes across cosmic time is fundamentally important for un- derstanding when and how black holes formed and grew and for assessing whether the energy liberated is adequate to the feedback task. Various multiwavelength survey techniques have been effective at sampling large fractions of the SMBH population, although no one technique yields a full census. Hard X-rays are most directly connected to the energy-generation mechanism in AGN and penetrate all but the highest line-of-sight column densities; IR observations effectively capture radiation reprocessed by dust; optical narrow-line surveys find faint AGN, even when heavily obscured; and radio surveys are completely insensitive to obscuration and can readily detect jets. The panel concluded: • It is very important to do complementary multiwavelength surveys to track the growth of black holes across cosmic time. A hard X-ray, all-sky survey for AGN is an essential complement to the deep pencil-beam surveys of active galaxies ex- pected from the upcoming NuSTAR Explorer. Long-wavelength IR surveys capture the total energy output, and rest-frame optical spectroscopic surveys allow black hole mass determinations. The next decade offers the prospect of detecting gravitational radiation from merging SMBHs in the 105 to 107 MSun range out to z ~ 10. While the restricted mass range and the possibility of small-number statistics will prevent a detailed

OCR for page 95
Panel rePorts—new worlds, new HorIzons 102 reconstruction of the SMBH merger tree, such observations can discriminate be- tween small- and large-seed scenarios for early SMBH growth and determine the masses and spins of some objects. The panel concluded: • The search for gravitational radiation from merging supermassive black holes, at lower frequencies than are probed with the Laser Interferometer Gravity Observatory, is very important for an understanding of the buildup of supermassive black holes. What Were the First Objects to Light Up the Universe and When Did They Do It? —and Discovery Area: The Epoch of Reionization Concerning the first objects to light up the universe, when and where did these objects form? When did the first galaxies emerge and what were they like? How was the universe reionized? This very early phase of galaxy evolution occurred during the epoch of reionization, which the panel designates as its discovery area because of its great discovery potential. This epoch lies at the frontier of astronomy and astrophysics for the next decade. The first objects to light up the universe could be stars, black holes, galaxies, and/or something less obvious, such as dark matter annihilation. What these ob- jects are and when and where they formed are almost completely unknown. They and subsequent generations provided enough light to reionize the universe by a redshift of z ~ 6, but the topology of the ionization is unconstrained at present. The expectations of astronomers are guided almost entirely by theory. The first stars should have been essentially metal-free and extremely massive (M > 100 MSun), with a radiation field that is very efficient at ionizing hydrogen and helium. For redshifts z < 11, key emission features will appear in the J band. While individual stars will be much too faint (AB ~ 38-40) to be detected directly with the James Webb Space Telescope (JWST) or an ELT, aggregates of stars may be visible in JWST deep fields, especially with the aid of gravitational lensing. Hyper- novae and/or gamma-ray bursts (GRBs), which may be the first individual stellar objects to be observed, can be found through time-domain surveys. GRBs can be used as a probe of the high-redshift intergalactic medium provided that several dozen with z > 8 are detected; this would take several years for a facility with an order-of-magnitude-higher detection rate (which depends on the product of field of view and sensitivity) than is possible with the Swift satellite. To find and characterize the first-generation aggregates of stars, the panel concluded: • It is most important to use JWST to make deep surveys, followed up with near-IR spectroscopy on an ELT.

OCR for page 95
rePort Panel GalaxIes across cosMIc tIMe 103 of tHe on • It is very important to develop a next-generation GRB observatory to search for the first explosions, with an order-of-magnitude-greater GRB detection rate than is possible with Swift, augmented by a rapid follow-up capability for infrared spectroscopy of faint objects. • It is very important to do time-domain surveys to identify the first stars from their supernova or hypernova explosions. One of the most tantalizing probes of the epoch of reionization is the redshifted 21-cm H I line. Ionization pockets in the cold intergalactic gas are expected to cause fluctuations in the 21-cm brightness temperature. Existing experiments (e.g., the Low Frequency Array for radio astronomy [LOFAR], the Murchison Widefield Array [MWA]) may be able to detect these fluctuations to z ~ 10. Ultimately, with future large-area low-frequency radio arrays, it should be possible to map the entire history of reionization by means of an all-sky map of redshifted 21-cm emission. Absorption-line spectroscopy along sightlines toward the first stars, GRBs, or supernovae will allow the detection of the presence of metals and the ionization level throughout the epoch of reionization. Such observations require the collecting area and spectroscopic capability of an ELT. The panel concluded that to explore the discovery area of the epoch of reionization: • It is most important to develop new capabilities to observe redshifted 21-cm H I emission, building on the legacy of current projects and increasing sensitivity and spatial resolution to characterize the topology of the gas at reionization. • It is very important to do near-infrared absorption-line spectroscopy with JWST, ELTs, and 10-m-class telescopes to probe the conditions of the IGM during the epoch of reionization. Although this discussion has so far focused on the first objects, it is very important to find and identify objects residing in the later stages of the epoch of reionization, including radio-loud AGN, quasars, galaxies, supernovae, and GRBs. The panel concluded: • It is very important to do multiwavelength surveys to detect galaxies, quasars, and GRBs residing in the late stages of reionization at 6 < z < 8, including near- infrared surveys for galaxies and quasars, hard X-ray or gamma-ray monitoring for GRBs, and time-variability surveys for supernovae or hypernovae.

OCR for page 95
Panel rePorts—new worlds, new HorIzons 104 Theory and Laboratory Astrophysics in the Next Decade Underlying all of astronomy and astrophysics is critical work in theory and other intellectual infrastructure, such as laboratory astrophysics. Theory is at the heart of astronomical inference, connecting observations to underlying physics within the context of a cohesive physical model. The past decade has seen great advances in theoretical aspects of galaxy formation and black hole astrophysics, particularly in the computational arena, which is driven by technological advances (much as with observations). To understand the universe better, to reap the full value of new observational capabilities as they become available, and to guide the next observations, the panel concludes that investments are needed in the follow- ing theoretical areas: • Cosmological context. Hydrodynamical simulations within a hierarchical structure-formation context, expanding the dynamic range to study detailed galaxy and cluster assembly within a representative volume. • Galactic flows and feedback. Central to galaxy assembly; requires understand- ing of the associated two-phase interfaces and instabilities as gas moves through the inhomogeneous intergalactic medium and of how energy, momentum, and relativistic particles feed back into ambient gas. • Magnetohydrodynamics (MHD) and plasma physics. Studies of how mag- netic fields channel and transport energy over a large dynamic range, including developing a better understanding of magnetic reconnection, particle acceleration, and cosmic-ray transport. • Radiation processes. Coupling radiative transfer models to dynamical galaxy- formation simulations, and incorporating radiation hydrodynamics and nonther- mal processes into models of jets and accretion disks. Summary of the Panel’s Conclusions The panel’s conclusions and top-rated science programs are summarized in Table 3.1. INTRODUCTION Galaxies are complex systems that evolve dramatically across cosmic time (Figure 3.1). Their critical constituents—not only stars, gas, and dust, but also supermassive black holes and dark matter—are strongly coupled to one another. During the past decade scientists have learned that no galaxy is an island: they constantly influence and are influenced by their environments.

OCR for page 95
rePort Panel GalaxIes across cosMIc tIMe 105 of tHe on TABLE 3.1 Summary of Conclusions of the Panel on Galaxies Across Cosmic Time Key Questions How do cosmic How do baryons cycle How do black holes What were the first structures form and in and out of galaxies, grow, radiate, and objects to light up the evolve? and what do they do influence their universe and when did while they are there? surroundings? they do it? Existing or SPT, ACT, VLBA, EVLA, ALMA, EVLA, JWST, Chandra, XMM- HST, JWST, Planck, approved e-Merlin, ALMA, HST, Herschel, 8- to 10-m Newton, Suzaku, LOFAR, MWA, ALMA, facilities and Chandra, Spitzer, OIR telescopes, NuStar, GEMS, EVLA, 8- to 10-m Important JWST, eROSITA- SDSS-like near-IR Astro-H, VLBA, (sub) telescopes, time-domain Observational SRG, Astro-H, (rest-frame optical) millimeter VLBI, surveys Programs HST-like imaging of spectroscopic survey, Fermi, VERITAS, gravitational lenses, CO emission in HESS, MAGIC, HST, OIR MOS and IFU high-z star-forming IR and O III surveys, spectra of galaxies, galaxies, radio time-resolved UV black hole masses continuum survey, spectra with HST with ELT/ radio, long- far-IR spectroscopy COS, multiwavelength baseline radio imaging of luminous dusty surveys for AGN, of gravitational lenses, galaxies, JWST IFU multiwavelength deep multiband cluster spectra spectra blazars/AGN surveys XRSO,a high-resolution New Highly multiplexed XRSO, hard X-ray ELT, sensitive high-z facilities optical imaging of near-IR MOS on 8- survey telescope, 21-cm capability, GRB GWO,c ELT, X-ray needed lenses, OIR MOS and to 10-m, high-z H I observatory, hard X-ray IFUs on 8- to 10-m 21-cm, ELT, 4-m UV polarimeters and gamma-ray and telescopes, ELTb space telescope, far-IR optical time-domain spectra, large (sub) survey telescopes, CMB millimeter telescope, E-mode polarization, new XRSO, ELT IFU ALMA receivers Hydrodynamical simulations of Gpc3 volumes, studies of magnetohydrodynamics and plasma Theory needed physics, studies of gas flows and feedback, calculations incorporating radiative transfer, dynamical simulations within the hierarchical structure-formation context. NOTE: Red, blue, and green are associated, respectively, with the scientific investigations that the Panel on Galaxies Across Cosmic Time deemed most important, very important, and important for addressing the key questions that it identified. Black indicates existing projects or facilities. Italics indicate observational programs. Acronyms are defined in Appendix C. aHigh-throughput high-resolution X-ray spectroscopy observatory. bExtremely large (20- to 40-m) ground-based optical/infrared telescope, with adaptive optics. cGravitational wave observatory sensitive in the 10 –1 t o 10 –3 H z range. The understanding of galaxies and galaxy evolution has changed radically over the past two decades, thanks to major advances in instrumentation across the electromagnetic spectrum, leaps in conceptual and computational theory, and innovative observational programs. These exciting advances can be illustrated with just a few recent examples:

OCR for page 95
Panel rePorts—new worlds, new HorIzons 140 masses in smaller samples are possible with centimeter-wave maser observations with very long baseline arrays supplemented by large aperture dishes, and/or re- verberation mapping of AGN (see the discussion for the first key question). Finally, the bases of jets can be studied by way of multiwavelength observa- tions of nearby radio galaxies and blazars. Exciting discoveries are just beginning to come from multiwavelength campaigns involving the VLBA, (sub)millimeter very long baseline interferometry (VLBI) observations, and the current genera- tion of ground-based and spaceborne gamma-ray observatories. The observations will show which improvements in spatial resolution and sensitivity are needed for further progress. The panel concluded: • Multiwavelength observations of blazars, including ground-based and spaceborne gamma-ray observatories, the VLBA, and millimeter VLBI (and even- tually submillimeter VLBI), are important for determining jet composition and constraining formation models. GCT 4. WHAT WERE THE FIRST OBJECTS TO LIGHT UP THE UNIVERSE, AND WHEN DID THEY DO IT? • Where and when did the first objects form? • When did the first galaxies emerge, and what were they like? • How did these first objects reionize the universe? Progress to Date The epoch between the last scattering of the cosmic microwave background radiation, at z ~ 1,100, and the current high-redshift frontier at z ~ 8, where the most distant quasars, GRBs, and galaxies have been observed, remains completely unexplored. This epoch contains the first stars, first galaxies, and first massive black holes. These objects must produce the ionizing photons that led to the reionization of the neutral hydrogen (formed at recombination, z ~ 1,100) in the universe. The formation of the first astrophysical objects and the subsequent epoch of reioniza- tion are at the frontiers of astrophysical research in the next decade. The study of the first objects and the epoch of reionization (Figure 3.19) is an area of discovery, leading into uncharted territory. Theoretical efforts are the only guide to understanding what the first luminous objects were and how they manifested themselves. Rapid progress is being made, and the interplay of theory and observations promise to answer key questions if, in the next decade, a compre- hensive array of empirical means to study this epoch are developed.

OCR for page 95
rePort Panel GalaxIes across cosMIc tIMe 141 of tHe on FIGURE 3.19 Cosmic timeline showing the forma- tion of structure in the universe from the Dark Ages (left) to the formation of the first stars, galaxies, and black holes that started the reionization of the inter- galactic medium, which is transparent in the present day (right). SOURCE: Courtesy of Abraham Loeb and Jean-Francois Podevin, adapted from Abraham Loeb, The dark ages of the universe, Scientific American 295:46-53, 2006. GCT Discovery Area—The Epoch of Reionization The discovery area identified by the Panel on Galaxies Across Cosmic Time is closely entwined with the panel’s fourth key question, What were the first objects to light up the universe, and when did they do it? Originally the panel intended to separate the two, with the fourth key question addressing that part of the early universe for which there is existing knowledge (e.g., objects discovered out to z ~ 8), and the discovery area referring to the less-well-understood part of the epoch of reionization (e.g., collapse of the first H I structures). However, it proved simpler and more straightforward to combine the two; hence this section includes both the focused questions and the broader inquiry appropriate to a discovery area. Recent years have seen significant progress both in theoretical understanding and in observational probes of this transition epoch in the early universe. Obser- vations of high-redshift quasars have uncovered a number of luminous objects at z ~ 6, the spectra of which imply that the process of reionization was complete by then. There are hints of objects at z ~ 7 and a few candidates at even higher red- shifts that have been detected with the aid of gravitational lensing. GRBs have been discovered out to z ~ 8.2. The recent measurements of the optical depth to electron scattering from the Wilkinson Microwave Anisotropy Probe (WMAP) polarization studies suggest that the first sources of light had already significantly reionized the IGM at z > 11. There is no doubt from these observations that the first galaxies and quasars were well in place by z ~ 6 and may have started to appear as early as z ~ 15-20. The actual history of reionization is still highly uncertain (as illustrated in Figure 3.1); Figure 3.20 summarizes current observational constraints on the inferred neutral hydrogen fraction of the IGM. Theoretical work suggests a general framework for the formation of the first objects: ΛCDM simulations indicate that the first objects formed as early as z ~ 40, out of gas that cooled via molecular hydrogen in the first dark matter halos. Most

OCR for page 95
Panel rePorts—new worlds, new HorIzons 142 FIGURE 3.20 Current constraints on the volume average neutral fraction of the intergalactic medium versus redshift. The red points indicate measurements based on the highest-redshift quasars. The green triangle shows the constraints from cosmic Strömgren spheres and surfaces around the highest-redshift quasars, and the ma- genta point indicates the Wilkinson Microwave Anisotropy Probe constraint. The curves show the expectations for different assumptions about early star formation. SOURCE: Adapted from X. Fan, C.L. Carilli, and B. Keating, Observational constraints on cosmic reionization, Annual Review of Astronomy and Astrophysics 44:415-462, 2006. simulations suggest that the first stars, called Population III, were predominantly very massive; their demise should have created the first black holes. The soft UV and X-rays emitted by these first stars and black holes provided significant feed- back to the chemical and thermal state of the IGM by driving expanding ionized H II regions into it (Figure 3.21). These events may have marked the beginning of reionization. Having dissociated the molecular hydrogen, radiative feedback from the first stars led to a second generation of stars and galaxies in which the gas cooled through radiative transitions in atomic hydrogen and helium. The ionized regions surrounding these primordial galaxies grew and merged until they fully overlapped, marking the end of reionization. It is still very uncertain whether stars were the first significant ionizing sources. An alternative possibility is that a similar fraction of this gas formed massive black holes directly (see Figure 3.21). Moreover, massive metal-free stars should (in most cases) leave behind stellar- mass seed black holes that accrete gas as mini-quasars. Depending on the efficiency with which ionizing photons escape into the IGM, the first halos might blow away their gas after a single episode of star formation and might not form any stars until

OCR for page 95
rePort Panel GalaxIes across cosMIc tIMe 143 of tHe on FIGURE 3.21 Computer simulations of first stars and black holes illustrating the initial steps in the process of reionization. Top: Ionizing radiation emanates from the first massive stars that form inside dark matter mini-halos, creating ionized bubbles (blue) interspersed with regions of high molecular abundance (green), both embedded in the still-neutral cosmic gas. The large residual free-electron fraction inside relic H II regions, left behind after the central star has died, rapidly catalyzes the forma- tion of molecules. Simulations were performed by Bromm et al. (2009), and visualization is courtesy of the Texas Advanced Computing Center. Bottom: Simulation showing the X-rays produced by a black hole (white), created when a first-generation star collapses, and their ionizing effect on nearby gas (blue). SOURCE: Top: Reprinted by permission from Macmillan Publishers Ltd.: Nature, V. Bromm, N. Yoshida, L. Hernquist and C.F. McKee, Formation of the first stars and galaxies, Nature 459:49-54, 2009, copyright 2009. Bottom: Courtesy of KIPAC/SLAC/M. Alvarez, T. Abel, and J. Wise.

OCR for page 95
Panel rePorts—new worlds, new HorIzons 144 deeper potential wells are formed. In general, several feedback processes are likely to be important at these stages, from internal feedback (e.g., due to the presence of supernovae) to external global feedback provided by the UV and X-ray background produced by the first sources. Numerical simulations are the most promising way to address many of these issues; however, accommodating the dynamic range required to resolve the small-scale physics in mini-halos and the large cosmological volumes needed to capture the emergence of rare density peaks is extremely challenging, especially since radiative transfer effects need to be included. Steps for the Next Decade Existing observations provide only a preliminary glimpse into the late stages of the epoch of reionization (Figure 3.22). Because this is an area of discovery, it is difficult to outline specific future science programs, and the panel was guided primarily by uncertain theoretical predictions. The first stars should have been es- sentially metal-free and extremely massive (M > 100 MSun), with a radiation field that is very efficient at ionizing hydrogen and helium. Thus, direct observation of the first stars depends on the detection of the rest-frame UV continuum and line emission, particularly the He II (1,640 Å) and Ly-a emission lines, which for red- shifts z < 11 will appear in the J band. The ability to detect the first stars depends critically on the physics of star formation in a metal-free environment (e.g., initial mass function, star-formation efficiency, etc.) and the extent to which these first objects are clustered. While individual stars will be much too faint to be detected directly with JWST or an ELT, if the first stars formed in larger aggregates then the resulting H II regions are expected to have line fluxes approximately 10−21 erg cm−2 s−1.3 The likeliest way forward is to use JWST deep fields, especially with the aid of gravitational lensing, to find the earliest stars and protogalaxies. Simply finding the first objects is not enough: to probe the astrophysics of the first objects requires follow-up near-IR spectroscopy (R ~ 5,000) and to determine the properties of the IGM requires absorption-line spectroscopy. The first explosions are key probes of the beginning of the epoch of reioniza- tion. Detection of transient objects during this epoch will allow the study of the intervening IGM through absorption spectroscopy. GRBs can be used as probes of the high-redshift universe, provided that several dozen with z > 8 are detected; this would take 2 to 5 years for a facility that detects GRBs at 10 times the rate that Swift does. It is critical to understand stellar astrophysics in metal-free conditions in order to predict the properties of the first supernovae. A Type IIn supernova should 3 See Astronomy and Astrophysics Advisory Committee, GSMT and JWST: Looking Back to the Future of the Universe, available at http://www.nsf.gov/mps/ast/aaac/reports/gsmt-jwst_synergy_com- bined.pdf. Accessed February 2010.

OCR for page 95
rePort Panel GalaxIes across cosMIc tIMe 145 of tHe on FIGURE 3.22 Left: Candidate redshift z ~ 7 galaxies identified by the “dropout” technique whereby each object is detected in the near infrared (J and H bands) but not at optical wavelengths (viz bands). The goal for the next decade is to find galaxies at even higher redshifts, confirm them spectroscopically, and study their chemical composition to identify primordial stellar populations. Top right: Swift ultraviolet/X-ray image (orange, red) of a gamma-ray burst at redshift z ~ 8.2, assuming the extremely red color in the optical/infrared image (bottom right) is due to the Lyman break at rest-frame 912 Å. This is the highest redshift cosmic explosion detected to date. No visible or ultraviolet light (green, blue) was detected at the position of the burst, but near-infrared emission was detected with the United Kingdom Infrared Telescope Facility and (bottom right) the Gemini North telescope. More such events, and more distant ones, will be visible with sensitive wide-field gamma-ray satellites and prompt follow-up observations in the near-infrared and optical. SOURCE: Left: R.J. Bouwens, G.D Illingworth, M. Franx, and H. Ford, z ~ 7-10 galaxies in the HUDF and GOODS fields: UV luminosity functions, Astrophysical Journal 686(1):230-250, 2008, reproduced by permission of the AAS. Top right: Courtesy of NASA/Swift/Stefan Immler. Bottom right: Courtesy of Gemini Observatory/NSF/AURA/D. Fox, A. Cucchiara (Pennsylvania State University), and E. Berger (Harvard University).

OCR for page 95
Panel rePorts—new worlds, new HorIzons 146 have a rest-frame UV magnitude of M ~ -21, giving an apparent magnitude of 27 at z ~ 6. More distant supernovae would be correspondingly fainter. Confirmation of the brighter events at the end of reionization should be possible with the cur- rent generation of 10-meter telescopes, but more distant sources require an ELT. To find and characterize the first-generation aggregates of stars, the panel concluded: • It is most important to use JWST to make deep surveys, followed up with near-IR spectroscopy on an ELT. • It is very important to develop a next-generation GRB observatory to search for the first explosions, with an order-of-magnitude-greater GRB detection rate than is possible with Swift, augmented by a rapid follow-up capability for infrared spectroscopy of faint objects. • It is very important to do time-domain surveys to identify the first stars from their supernova or hypernova explosions. Ionization of the neutral IGM during the epoch of reionization creates bubbles that should correlate with the location of the first sources. These bubbles can be detected by means of fluctuations in the brightness temperature of the 21-cm line of H I; because the universe is nearly transparent at these frequencies, the 21-cm line should be an outstanding probe of the entire history of reionization. At high redshift, the 21-cm emission should display angular as well as frequency structure due to inhomogeneities in the gas density field, H II fraction, and H I spin tempera- ture. This prospect has already motivated the design and construction of arrays of low-frequency radio telescopes (e.g., LOFAR, MWA) that aim to search for this signal from z ~ 6-15, redshifted to wavelengths of approximately 1.5 to 4 meters. Ultimately, it will be possible to map the entire history of reionization by way of an all-sky map of redshifted 21-cm emission. Absorption-line spectroscopy along sight lines toward the first objects—be they stars, GRBs, or supernovae—will allow the detection of the presence of met- als and the ionization level throughout the epoch of reionization. Such observa- tions necessarily require the collecting area and spectroscopic capability of an ELT, which can determine redshifts, stellar masses, and chemical compositions in early galaxies, quasars, and transient events. Detection of ultralow abundances of heavy elements—the definitive indication of the first generation of stars and galaxies— should be feasible within the next decade. It is also desirable to measure the mean fraction of neutral hydrogen with independent probes. The WMAP data were used to derive the first constraints on the redshift range of the epoch of reionization based on E-mode polarization

OCR for page 95
rePort Panel GalaxIes across cosMIc tIMe 147 of tHe on measurements. The Planck satellite will substantially improve on these results. The Planck results might motivate additional measurements with a follow-up mission. The panel concluded that to explore the discovery area of the epoch of reionization: • It is most important to develop new capabilities to observe redshifted 21-cm H I emission, building on the legacy of current projects and increasing sensitivity and spatial resolution to characterize the topology of the gas at reionization. • It is very important to do near-infrared absorption-line spectroscopy with JWST, ELTs, and 10-meter-class telescopes to probe the conditions of the IGM during the epoch of reionization. • It is important to measure the CMB E-mode polarization with Planck and possibly follow-on missions. Although this discussion has so far focused on the first objects, it is very important to find and identify objects residing in the later stages of the epoch of reionization, including radio-loud AGN, quasars, galaxies, supernovae, and GRBs. These populations connect in a directly observable way to the universe today and thus hold the key to a full understanding of its evolution. There is tremendous promise in high-frequency (>30 GHz) searches for CO in the high-redshift uni- verse with both the EVLA and future radio facilities. Cooling by means of the C II and O I fine structure lines might be detectable if metal enrichment occurs early enough. The panel concluded: • It is very important to do multiwavelength surveys to detect galaxies, quasars, and GRBs residing in the late stages of reionization at 6 < z < 8, including near- infrared surveys for galaxies and quasars, hard X-ray or gamma-ray monitors for GRBs, and time-variability surveys for supernovae or hypernovae. • It is important for ALMA to have the capability to search for C II and O I fine structure line emission at redshifts z > 6. THEORY AND LABORATORY ASTROPHYSICS: THE NEXT DECADE Underlying all of astronomy and astrophysics is critical work in theory and other intellectual infrastructure, such as laboratory astrophysics. Theory is at the heart of astronomical inference, connecting observations to underlying physics within the context of a cohesive physical model. The past decade has seen great advances in theoretical aspects of galaxy formation and black hole astrophysics, particularly in the computational arena driven by technological advances. Theory has become more interconnected with data as the fundamental physical processes are better understood and as models are placed within the context of the overall

OCR for page 95
Panel rePorts—new worlds, new HorIzons 148 cosmological paradigm. Many areas are now driven by theoretical goals, such as testing the structure of halos predicted by simulations, observing the cold accretion predicted to be a main channel for supplying galaxies with fuel for star formation, and testing whether jets are powered by black hole spin. To understand our universe better, to reap the full value of new observational capabilities, and to guide the next observations, investments should be made in the following theoretical areas: • Cosmological context. To compare ΛCDM predictions with surveys and understand the connections among galaxies, intergalactic gas, and large-scale struc- ture require hydrodynamical simulations within a hierarchical structure-formation context. The key challenge is to expand the dynamic range of current simulations to study detailed galaxy and cluster assembly within a representative volume, for which developing new algorithms, improving subgrid physics models, and taking advantage of new technologies will be crucial. • Galactic flows and feedback. Studies of galaxy formation require under- standing accretion and feedback processes, which are central to galaxy assembly. A key issue here is to properly understand two-phase interfaces and instabilities that occur during the motion of gas through ambient media of strongly differ- ing temperatures and densities, and how energy is exchanged and released across such interfaces (e.g., mixing processes). Understanding the injection of energy, momentum, and relativistic particles into ambient gas (i.e., feedback) will require simulations of MHD turbulence, with accurate treatment of transport processes such as viscosity and heat conduction. • Magnetohydrodynamics and plasma physics. Studies of accretion disks, jets, and their interactions with ambient plasma require a better understanding of how magnetic fields channel and transport energy over a large dynamic range. Solving some of the key questions will require a better understanding of particle processes such as magnetic reconnection, particle acceleration, and cosmic-ray transport. These processes will have to be incorporated into the next generation of codes in a physically realistic way. • Radiation processes. Calculations incorporating radiative transfer are critical for studying the epoch of reionization, the escape of ionizing and Ly-a radiation from galaxies, and the evolution of the IGM. The challenge arises in coupling such models to dynamical galaxy-formation simulations, to understand fully how ra- diation interacts with the highly clumpy and asymmetric gas distribution. Models of jets and accretion disks must include nonthermal radiation processes such as synchrotron radiation and inverse Compton scattering in order to compute spectra and must include radiation hydrodynamics to capture disk structure accurately.

OCR for page 95
rePort Panel GalaxIes across cosMIc tIMe 149 of tHe on Addressing these questions should be done using a tiered approach, with an emphasis on computational work that takes advantage of rapidly developing technologies. Small teams led by individuals are in a good position to push the frontiers of algorithms and new computing architectures. Medium-size groups of several experts are important to make coherent and concerted progress on key numerical issues outlined above. Large, heroic simulations that push the limits of available technology are critical to drive forefront work; these are best facilitated by large computing centers. All of these areas are informed by analytic theory, which is crucial for providing the basic physical underpinnings for more complex models. Together these aspects are fundamentally important for making progress toward understanding the universe, and the galaxies, clusters, black holes, gas, and dark matter within it. Laboratory astrophysics is clearly important to understanding galaxies, black holes, and clusters across cosmic time. For example, the details of absorption by dust are not understood, even though most cosmic objects are substantially impacted by dust. Similarly, still lacking are important cross sections for hot gas cooler than 4 million degrees, which is significant in clusters and the IGM. Scientists are not sure of recombination rates that determine ionization equilibria. Spectral features at millimeter to infrared wavelengths are especially poorly known and difficult to calculate, as they arise in molecules and clusters of atoms; ALMA will see a forest of lines that, without new laboratory measurements, will be difficult to interpret. Although the lack of laboratory measurements may not today be a limiting factor in the studies outlined earlier in this report, it may well become the limiting factor as the data improve. CONCLUSION We now stand at the threshold of being able to observe the full history of cos- mic structure, across all mass scales, from earliest times to the present day. With a judiciously chosen set of new facilities, instruments, and observing strategies, we can learn how galaxies, clusters, and black holes form, evolve, interact, and influence each other, to a degree of accuracy undreamed of just 20 years ago. This unprecedented wealth of data should be accompanied by a concomitant increase in understanding through the development of new analytic and computational theoretical tools. The past decade has brought dramatic advances in cosmology and an emerging picture of the growth of galaxies and black holes across the immensity of cosmic time. The next decade will open up the entire universe to detailed study and will revolutionize our understanding of the processes by which the observable cosmos came to be.

OCR for page 95