Below is the uncorrected machine-read text of this chapter, intended to provide our own search engines and external engines with highly rich, chapter-representative searchable text of each book. Because it is UNCORRECTED material, please consider the following text as a useful but insufficient proxy for the authoritative book pages.
4 Galaxies and Stellar Systems Many decades of observations have revealed that our home galaxy, the Milky Way, is a system of a hundred billion stars and is one of many hundreds of billions of similar systems distributed throughout space. The Milky Way is a huge "ecosystem" of stars and gas that circulates and mixes the heavier elements-enriched by each succeeding generation of stars from which our world and we ourselves are made. Moreover, just as stars are the building blocks of galaxies, so galaxies are the building blocks of the universe. As such, their distribution on the sky can be used to map the large-scale distribution of matter in the universe. Galaxies contain interstellar gas and dust, the raw materials out of which new generations of stars are born. As these stars age and eventually die, they eject some of their mass back into space, where it mixes with ambient gas and later forms the next generation of stars. This ejected gas is enriched in elements from carbon to uranium formed as a consequence of nuclear reactions in stars. As a result, each succeeding generation of stars will have a chemical composition different from that of its predecessors. The internal structure of a star depends on its chemical composition, and so do its other properties, including how long it will live. Moreover, the chemical composition of the material in dense clouds- the sites of new star formation determines the cloud's transpar- ency, and the energy that is trapped influences the conditions under which new stars are born. An intricate variety of complex feed-back mechanisms operate in galaxies. Still another complication stems from interactions between galaxies and their environment. Neighboring galaxies in groups and clusters interact with each other, raising tides of material by their mutual gravitational attraction. As galaxies move through an ambient hot gaseous medium, as is commonly found in groups and clusters, the interstellar gas in the galaxies collides with the intracluster gas and may be stripped away. Such interactions can drive both the evolution of the galaxies and the evolution of their environment. There are still many fundamental aspects of galaxies about which astronomers have only a sketchy under- standing. A prime example of such a mystery is the nature, amount, and distribution of dark matter in galaxies. Because we know almost nothing about the dominant constituent of the mass, we have only a phenomenological understanding of how the gravitating mass is traced by its luminous part. This in turn limits our understanding of galactic structure and the large-scale distribution of galaxies. Another example of the limitations of astronomers' current knowledge is evident in the exotic phenomena called quasi-stellar objects, or quasars, which for over 30 years have withstood attempts to incorporate them into current understanding of the origins of galaxies and stars. 35
36 A NEW SCIENCE STRATEGY FOR SPACE ASTRONOMY AND ASTROPHYSICS KEY THEMES A common thread linking much of the research on distant and nearby galaxies is the delineation of evolution- ary paths from the earliest epochs to the present. The various ideas discussed in this chapter can be summarized in the following short list of key themes: Development of present-day structures; Chemical composition of the universe; · Dark matter; · Baryons outside of galaxies; and · Supermassive black holes and quasar power sources. DEVELOPMENT OF PRESENT-DAY STRUCTURES How did present-day structures develop from a universe that was once almost perfectly smooth and uniform? This question has a fundamental significance because the early development of structure was crucial to the concentration of baryons (i.e., familiar matter composed of protons, neutrons, and electrons) that in turn eventually led to the existence of life. Observations of the cosmic microwave background indicate that the universe in the distant past was hot and extremely uniform. There are only weak fluctuations in intensity which reflect slight non-uniformities in the density of the universe at early epochs; the over- and under-densities amount to only 1 part in lOO,OOO. These perturbations were the seeds from which stars, galaxies, quasars, and large clusters of galaxies later formed. But exactly how and when the variety of structures seen today formed is still a matter of speculation. In this picture, galaxies condensed at an early time from gas in the expanding universe. At the present epoch, approximately 15 billion years later, the gas in intergalactic space can be expected to be composed of primordial material that was never incorporated into galaxies, plus any gas that was processed through stars and returned to intergalactic space. A full understanding of the evolution of galaxies thus also requires an understanding of the nature and evolution of the gas between the galaxies. Key Questions About the Development of Present-Day Structures Key questions to be addressed in studies of the development of present-day structures include the following: · How do the distribution of galaxy properties such as size and mass, and the variety of their morphological forms, relate to the physical conditions of the universe in its first billion years? · What triggered the birth of the first generation of stars? · What regulated or influenced the rates of formation of subsequent generations of stars that drove the chemical evolution of the Milky Way? To answer questions like these, researchers need to turn the clock back, stepping through the evolution of galaxies over cosmic time, to the epoch of birth. It is difficult to reconstruct the evolutionary history of a galaxy from the "fossil record," but astronomers have the unique opportunity of viewing the history of the universe directly as it happened, by exploiting the cosmological "look-back" time. Events at great distances are observed after a period of time equal to the light-travel time: the farther away the source of light, the farther back in time it is seen. Astronomers now have the capability to make routine measurements of ordinary galaxies at distances of at least 7 billion light-years (that is, a look-back time of at least 7 billion years), to be compared to the age of the universe of around 15 billion years. These time intervals are large enough that evolutionary changes are expected to be apparent (Figure 4.1~. The distance to a galaxy is measured by the redshift seen in its spectral lines- the higher the redshift, the greater the distance, and the earlier in cosmological history astronomers see the galaxy. As just mentioned, normal
GALAXIES AND STELLAR SYSTEMS 37 galaxies half the age of the universe are now studied fairly easily, but special cases are known of galaxies much more distant and much younger than this some galaxies are seen when the universe was oniv 15~ of its nr~.~.nt ~. .. . . . . ~. . . . ~ _ , ~r-~~ age. By extending search techniques to fainter sensitivity levels, observers can expect to see normal galaxies at comparable distances, and special cases to distances that are limited not by instrumentation, but by the sources themselves. Recent Progress in Understanding the Development of Structures Images of distant galaxies from the Hubble Space Telescope (MST) have provided direct evidence for the evolution in galaxy populations and morphology with look-back time. This information, combined with ground- based spectroscopic observations with the 10-meter Keck telescope, have led to the first identifications of star- forming galaxies 2 billion to 3 billion years after the big bang. Many of the objects studied appear surprisingly advanced in their formation even at this early epoch. These early systems contain enormous amounts of gas-enough, perhaps, to be the forerunners of all the galaxies now seen in the universe. This notion has motivated ongoing attempts to detect starlight from galaxies associated with a class of absorption lines seen in the spectra of quasars, the so-called damped Lyman-alpha systems. Similarly, visible galaxies have been associated with metal absorption lines in the spectra of quasars; these correspondences provide powerful probes of the clouds containing these atoms. X-ray absorption has been detected in sources close to, and perhaps surrounding, bright, high-redshift quasars. In addition, observations of CO at millimeter wavelengths reveal that large masses of molecular gas have been found around some of the highest-redshift quasars, implying enormously energetic star-formation episodes at this early time. Recent numerical simulations of the evolution of dissipationless particles (including dark matter) and gas are producing a more accurate picture of the universe in its gas-rich, pregalactic phase. It is becoming possible to simulate the early universe with sufficient fidelity to model the formation of a galaxy and to simulate the clumping of hot and cold gas that gives rise to quasar absorption lines. Refined simulations now allow exploration of the mechanisms triggering bursts of star formation and activity in the nuclei of galaxies. Future Directions for Understanding the Development of Structures - -r Further progress in the area of galaxy formation and evolution suggests the following three different ap- proaches, in no particular order: 1. Conducting a census of galaxies to the earliest possible epoch to observe their formation directly. To see the gaseous precursors of galaxies and the later evolution of interstellar gas clouds and dust will require instruments capable of reaching far-infrared, submillimeter, and millimeter wavelengths. The proposed ground- based Millimeter Array is a good example of the kind of tool needed to make these observations. Further gains can be achieved with a large, cooled space telescope ontimizer1 for nhotom~.trv Anti ~n~.rtrn~ron~ in the ~nnrr`Yim~tP wavelength range from 20 to 800 am. Facilities like the Near-Infrared Camera and Multi-Object Spectrometer (NICMOS), recently installed on HST, and the Space Infrared Telescope Facility (SIRTF) will identify the most luminous galaxies at very early times and will study what happens when star formation occurs at explosive rates in dust-laden galaxies. SIRTF will detect ordinary galaxies to redshifts (z) of ~3,* and more luminous galaxies at higher redshifts. The direct observation of starlight in typical galaxies at very early times will require a large-aperture space telescope optimized for the 1- to 5-~m band in the near infrared. With an aperture large enough to achieve 0.1-arc see images or better, and cooled sufficiently to be limited by the natural background, such a telescope would have the sensitivity to detect galaxies to redshifts of at least z = 5, and might be able to search for galaxies at z = 10. - r - - r ~~ rip ^A~ ~r~~ i- he-Pro ^ ~^-Bran ~ See redshift in the glossary for the definition of z.
38 A NEW SCIENCE STRATEGY FOR SPACE ASTRONOMY AND ASTROPHYSICS 2. Studying the structural and kinematic properties of very young galaxies and following the evolution of these properties to the present epoch. Achieving this goal requires high angular resolution at optical and near-infrared wavelengths, and moderate-resolut~on spectroscopy over a wide range of wavelengths (to accommo- date the range of redshifts). Large ground-based telescopes will contribute mainly to observations at visible wavelengths, over wide fields of view, and where angular resolution better than 0.5 arc see is not required. Measuring spectral features in the ultraviolet requires space-based instrumentation. Sensitivity in the near infrared is greatly enhanced with space-based instrumentation, partly because of the lower background, and partly because of the higher attainable angular resolution. 3. Understanding the interactions among stars, interstellar gas, and dust that drive the formation of stars in a galaxy and chemical evolution. Individual supernovae trace the first generation of stars and can be detected with a low-background, high-resolution, near-infrared telescope. Since these same supernovae enrich the surrounding gas with heavy elements, their detection allows a direct measure of the chemical evolution of the universe. Some of the light emitted by stars is absorbed by gas or dust and subsequently re-radiated at longer wave- lengths before it can escape the galaxy. Information on the dust grains is provided by the continuum re-radiation occurring at far-infrared wavelengths. Information on the gas comes from line re-radiation, in particular, lines in the submillimeter spectral band. CHEMICAL COMPOSITION OF THE UNIVERSE The chemical composition of the universe and its variation over time are of great importance. Three minutes after the big bang, the universe consisted of hydrogen, helium, and trace amounts of other light elements. Synthe- sis of the heavier elements came several hundred million years later, following the formation of the first generation of stars. Key Questions About Chemical Composition Thanks to the increasing amount of high-quality spectroscopic data on elemental abundances in stars and in the gas of high-redshift systems, unprecedented progress in the study of the chemical evolution of the universe is possible in the near future. Answers to a number of key questions will guide future understanding of the history of star formation and nucleosynthesis in galaxies. These questions include the following: · What are the precise values of the primordial abundances that emerged from the big bang? · How and when were the heavier elements first synthesized and dispersed after the big bang? · What was the nature of the very first generations of massive stars and associated supernovae explosions? Recent Progress in Understanding the Chemical Composition In addition to the hydrogen lines in the Lyman-alpha forest seen in the spectra of distant quasars, other absorption lines have been observed from atoms like carbon, magnesium, and silicon. The presence of these heavy elements indicates that some material has indeed been processed through stars at very early times; indeed, the spectra of most clouds at the earliest epochs show heavy-element lines. The abundance ratio of silicon to carbon can be used, in principle, to measure the star-formation time scale, which is a measure of the time scale for nucleosynthesis in stars. The study of the connection between the hydrogen clouds and those clouds in which lines from heavier elements have also been detected is an active field of research, particularly now that HST has greatly expanded access to ultraviolet wavelengths. Also important is the ability of ground-based telescopes to obtain high-resolution spectra at high signal-to-noise ratios. X-ray telescopes have discovered that hot diffuse gas is a major component of the mass of rich clusters of galaxies. Moreover, both continuum radiation and line emission are detected at x-ray wavelengths, and the gas is
GALAXIES AND STELLAR SYSTEMS 39 deduced to have a heavy-element fraction as large as that seen in the solar system. These developments have raised a host of new questions. How, in detail, is metal-enriched gas ejected from galaxies, and how does the efficiency of related processes depend on the dynamical state or age of the cluster? How is the intracluster gas heated to such high temperatures? What are the specific contributions to the mix of metals from Type Ia and Type II supernovae, and how have the relative contributions tracked the evolution of the clusters? Visible-light spectro- scopic measurements of chemical abundances in the stars in the galaxies, and abundances deduced from quasar absorption lines, have yielded complementary constraints on such issues. Spectroscopic studies of the stars in the Milky Way's halo have made important contributions to this discus- sion. These stars are notable because they have extremely low heavy-element concentrations, presumably because they formed very early in the history of the Milky Way. Observations with both HST and ground-based tele- scopes, have, however, detected signatures of supernova enrichment and thus, perhaps, an epoch of heavy-element synthesis very early in the Milky Way's history. Studies of stars in nearby, dwarf galaxies give important information about the history of galactic star formation; rather than being simple, as previously thought, the dwarfs display complex episodes of star formation occurring irregularly over much of their lifetimes. Campaigns to detect distant supernovae (z > 0.5) have been successful thanks to large-format charge-coupled devices (CCDs), sophisticated image-processing software, and the availability of time on large telescopes; analyses of light curves have yielded key parameters such as the supernova type and the maximum flux. Variations of elemental abundance with time for stars in the Milky Way's disk have been determined, indicating that the mass fraction of heavy elements has increased by only a factor of 3 to 5 over the last 10 billion to 12 billion years. Measurements of the abundance ratio of deuterium to hydrogen in intergalactic gas clouds have provided important constraints on the history of star-formation activity as a function of redshift. These observations have, as a consequence, further tested the predictions of the synthesis of the light elements in the big bang. Future Directions for Understanding Chemical Composition The future directions identified below are intended to elucidate the history of star formation, the major factor in compositional evolution. Critical issues to be examined with the next generation of instrumentation include the following, in no particular order: Measuring the composition of the gas, dust, and stars of which the universe is composed, as a function of redshift. This requires identification of galaxies (and protogalaxies) to large redshift, and spectros- copy of these galaxies across a wide range of wavelengths. Surveys with SIRTF will provide target lists. 2. Conducting a census of supernovae at high redshifts to measure their early rate of occurrence. The rate of occurrence of Type II supernovae is an index of the formation of massive stars. Thus, the rate for Type II relative to that for Type Ia provides a diagnostic of the first stellar generations. This diagnostic, in turn, allows the elemental abundances at high redshift to be interpreted. The detection of supernovae at the greatest distances requires high angular resolution, sky coverage adequate to find these rare events, and a low background in the near-infrared spectral range. 3. Determining the amount of the hot gas in clusters of galaxies as a function of redshift. Such a program requires sensitive x-ray spectroscopy. The Advanced X-Ray Astrophysics Facility (AXAF) will begin this pro- gram, but higher-sensitivity spectroscopy is needed. Moreover, wide-field capability in the x-ray band is required to identify unbiased samples of distant clusters. 4. Using intergalactic gas clouds at high redshift to probe the early history of elemental abundances. High sensitivity combined with high spectral resolution at ultraviolet wavelengths is necessary both to measure important atoms (e.g., neutral and ionized helium) and to compare absorption lines at low redshift with those at high redshift.
40 There A NEW SCIENCE STRATEGY FOR SPACE ASTRONOMY AND ASTROPHYSICS DARK MATTER Is convincing evidence that matter in the familiar form of stars, gas, and dust constitutes only a small fraction of the matter in the universe. The missing mass is 10 to 100 times larger than that which is observed. Some of this missing, or dark, matter must be in the halos of galaxies; more is probably smoothly distributed outside galaxies. Strong theoretical arguments suggest that the density of the universe is close to the critical value required to eventually halt the expansion resulting from the big bang. If so, there must be even more dark matter than astronomers have already inferred from available measurements. Key Questions About Dark Matter Our understanding of galaxies, clusters of galaxies, and the universe itself is seriously incomplete until the nature of dark matter is determined. Three key questions are as follows: · What is dark matter? · Where is it located? · How does it influence the evolution of structures in the universe? Recent Progress in Understanding Dark Matter Analysis of the gravitational potential of groups of galaxies, clusters, and individual galaxies based on x-ray observations by the Roentgensatellit (ROSAT) and the Advanced Satellite for Cosmology and Astrophysics (ASCA), has shown that much dark matter exists in these systems. Similarly, gravitational tensing of background sources has been observed in images taken with HST and ground-based optical and radio telescopes (Figure 4.2~. Lensing by galaxies and clusters of galaxies cannot be understood unless there is dark matter. The discovery of strong tensing by rich clusters shows that the distribution of dark matter probably has a large central cusp. The detection of gravitational microlensing by objects in the Milky Way's halo demonstrates an important new technique that will allow astronomers to probe dark matter in the form of small collapsed objects, such as burnt-out stars. Comparison of the mass distribution inferred from large-scale galaxy motions with the mass distribution observed as luminous galaxies has been made possible thanks to complete surveys of nearby galaxies. Such studies have given evidence for very large mass concentrations on supercluster scales, and they point to a universe with a density that may approach or equal the critical density. Future Directions for Understanding Dark Matter Future directions for the study of dark matter include the following, in no particular priority: 1. Detecting dark matter through its gravitational influences on the light from distant objects. For gravitational lens studies, angular resolution and surface-brightness sensitivity are critical. Large-aperture, high- resolution telescopes working at visible and near-infrared wavelengths will greatly increase the number of objects in which gravitational lens effects can be observed. 2. Searching for fine splittings in the images of point-source-like quasars. Ground-based telescopes with adaptive optics can achieve very high angular resolution and are well matched to this task. However, adaptive- optics technology is still under development, and the first applications will focus on the near-infrared spectral range. At shorter wavelengths the field of view is very small, which limits the use of ground-based adaptive-optics observations for untargeted surveys for tensed objects. 3. Studying tensed radio sources. Much work is still to be done with the Very Large Array (VLA) and now with the recently completed Very Long Baseline Array (VLBA). However, space-based application of very long baseline interferometry (VLBI) will provide the highest angular resolution at radio wavelengths. The VLBI Space
GALAXIES AND STELLAR SYSTEMS 4 1 Observatory Program (VSOP) will be limited in its surface-brightness sensitivity, but if successful it will point the way to more sensitive missions. 4. Measuring mass distributions on supercluster scales (1 degree of arc) by the technique of weak tensing. Such studies require the development of ground-based, wide-field-of-view cameras with excellent image sharpness and careful control of the uniformity of the image sharpness across large areas of sky. While MST's current cameras have the requisite angular resolution and uniformity, they are limited by their small fields of view. Future surveys of weakly tensed objects will extend current capabilities by obtaining more solid-angle coverage, fainter surface-brightness limits than those achieved in the Hubble Deep Field, and a wavelength range extending into the near infrared. 5. Tracing the gravitational potential of galaxies and clusters by mapping their x-ray emissions. Such observations require more sensitive telescopes with large fields of view. While AXAF will represent a giant step forward in resolution and sensitivity, limitations in the size of its mirrors will prevent the application of these techniques to the most distant objects. To extend the work begun by AXAF, large-scale x-ray optics must be developed. 6. Improving understanding of the large-scale motions of galaxies. More accurate maps of the distribu- tion of galaxies and more accurate measurements of distance are essential to this task. The Sloan Digital Sky Survey and the 2-Micron Sky Survey (2MASS) will provide some of the information on the distribution. Studies of Type Ia supernovae and surface-brightness fluctuations are particularly promising techniques to provide infor- mation on distances. High angular resolution is critical, and MST's potential has not yet been fully realized for such studies. A space telescope with a larger aperture than HST would allow these techniques to be applied at greater distances. Ground-based telescopes equipped with adaptive optics will probably be of great value as well. BARYONS OUTSIDE OF GALAXIES If current ideas about the production of elements in the early universe are correct, then a large fraction of all baryonic matter is not contained in observed galaxies. In the standard cosmological picture, approximately 5% of the mass (if the universe has the critical density) is expected to be in the form of baryons, but less than 1% can be accounted for within the visible boundaries of galaxies. Key Questions About Baryons Outside of Galaxies Key questions to guide astronomers' studies of the baryons not tied up in galaxies include the following: · Are these "missing" baryons in large halos around galaxies, or are they distributed uniformly throughout intergalactic space? · How much primordial gas was left over in intergalactic space after the initial enoch of ~alaxv formation? · How can observers identify the gaseous predecessors of galaxies? Recent Progress in Understanding Baryons Outside of Galaxies r G ~ For years, astronomers have had two sets of observations demonstrating that gas exists between the galaxies: as just mentioned, numerous absorption lines seen in the spectra of high-redshift quasars, and x-ray observations of the hot intracluster gas in nearby clusters of galaxies. New insights into the important role played by intergalac- tic gas in the evolution of galaxies and structure in the universe have come from a variety of sources. These include space missions (e.g., HST, the Hopkins Ultraviolet Telescope (HUT), ROSAT, and ASCA), large ground- based optical telescopes (e.g., Keck), and extensive numerical calculations. The spectra of distant quasars show numerous weak absorption features attributable to the Lyman-alpha transition of neutral hydrogen in intergalactic gas clouds that lie along the line of sight. Absorption lines associated with this Lyman-alpha forest are exceedingly numerous: several hundred lines per quasar, far too numerous to be attributable to normal galaxies. Also, unlike galaxies, the clouds are more uniformly distributed
42 A NEW SCIENCE STRATEGY FOR SPACE ASTRONOMY AND ASTROPHYSICS along the line of sight; they are not clustered as strongly as galaxies. These clouds seem to dissipate as the universe evolves; the forest is densest in the spectra of the most distant quasars and systematically thins out with time. These properties led to the speculation that the Lyman-alpha forest clouds might be pregalactic gas fragments, which later collapsed to form galaxies. However, observations of lines of ionized carbon, oxygen, and silicon now indicate that at the epoch of the forest clouds, stellar processing had already occurred. In addition, ground- and space-based observations of normal quasars and ones gravitationally tensed show that the moderate- to high- redshift structures producing absorption are very much larger than galaxies. Some idea of what these structures are has come from theoretical studies. Numerical simulations of the evolution of the intergalactic medium indicate that the features causing the absorption are typically sheets and filaments, aligned with the sites of future galaxies and clusters. These calculations also suggest that such features are far from dynamical equilibrium. In an . . important breakthrough, ionized intergalactic helium was detected by both HST and HUT at very high redshift, providing a direct probe of the state of diffuse gas at early epochs. The Cosmic Background Explorer's (COBE's) observations of the uniformity of the cosmic microwave background rule out a hot intergalactic medium as the origin of the diffuse x-ray background at energies greater than 1 keV. However, theoretical studies indicate that a substantial fraction of the volume of the local universe is filled with a low-density hot gas, which may account for the x-ray background at lower energies. The major baryonic component seen in local clusters of galaxies is x-ray-emitting gas. If, as most theoretical studies indicate, rich clusters are fair samples of the universe, this result suggests that the density of the universe is only one-third of the critical density. Many poor groups of galaxies have large dark-matter halos as well, and so it may well turn out that a significant fraction of the baryonic dark matter in the universe resides there. Future Directions for Understanding Baryons Outside of Galaxies order: Future directions in the study of the baryons found outside of galaxies include the following, in no particular 1. Measuring physically important quantities such as density, temperature, and velocity shear in inter- galactic clouds. Future use of quasars as probes of the intergalactic gas requires greater sensitivity at high spectral resolution. This is especially true in the ultraviolet spectral region, which has the richest set of spectral diagnos- tics. High spectral resolution is required to fully resolve the lines and measure important physical quantities. The Space Telescope Imaging Spectrograph (STIS), installed in HST in February 1997, and the Far-Ultraviolet Spec- trographic Explorer (FUSE) will provide useful data but will be able to study only the brightest and nearest sources. 2; Determining the sizes and geometry of intergalactic structures. Correlating the hydrogen absorption lines seen in quasars along neighboring lines of sight is the most appropriate technique. However, doing that requires observing faint quasars to increase the likelihood of finding close pairs. A modest gain in the sensitivity of space-based ultraviolet spectrometers should yield a spectacular gain in the number of lines of sight observable. Increased sensitivity will also allow good searches for trace metals and primordial deuterium in the gas. Progress is most likely to come from a combination of ground- and space-based efforts. 3. Studying the physical properties, evolution, and interactions between intergalactic gas and galaxies. Such studies are best performed by observations of the far-ultraviolet absorption of neutral and ionized helium in intergalactic gas. Currently, HST and HUT can observe helium transitions in the far ultraviolet along just three lines of sight that are unusually clear of near-ultraviolet absorption by hydrogen. In the places where helium has been seen, only very low resolution spectra are currently attainable. Thus, whether the gas is uniformly distributed or clumped in clouds or sheets is unknown. Observing helium along many lines of sight will require a larger telescope than HST, with better ultraviolet spectroscopic sensitivity. Such an instrument, however, need not be as costly as HST since it can sacrifice image quality to gain increased light-gathering power. Until they are replaced, MST's unique near-ultraviolet imaging and spectroscopic capabilities should be maintained and, if possible, upgraded.
GALAXIES AND STELLAR SYSTEMS 43 4. Determining the masses of clusters and the chemical composition of the intracluster gas. While these quantities are calculable from spatially resolved, x-ray spectroscopy, no x-ray instrument currently available combines the requisite high spectral and spatial resolution. Thus, this technique has been limited by the need to make simplifying assumptions such as spherical symmetry and hydrostatic equilibrium, and the neglect of turbu- lence or clumping. Observations to date are biased toward low-redshift systems because of the low angular resolution of satellites such as ASCA. Future missions such as the X-Ray Multi-Mirror (XMM) mission and AXAF will have improved sensitivity but will allow clusters to be studied only to moderate distances. A large x- ray telescope, with good angular resolution, could push studies of the chemical and dynamical history of the hot intracluster gas to cosmologically significant distances. Such an instrument will also permit studies of the absorption spectra of C, N. O. Ne, Si, and S ions, redshifted to the soft x-ray band, and will provide unique information about the ionization and abundance of the intergalactic gas as a function of time. SUPERMASSIVE BLACK HOLES AND QUASAR POWER SOURCES Quasars are characterized by their enormous power output, exceeding that of entire galaxies; their (astronomi- cally speaking) tiny sizes, comparable to that of the solar system; their spectra, which show no characteristic temperature; and, sometimes, their ejection of matter at nearly the speed of light. The high luminosities of quasars allows them to be seen to larger distances, and therefore to earlier times, than any other object in the universe. Surveys show that quasar activity peaked about 2 billion years after the big bang. What is the nature of the power sources responsible for such huge luminosities, and what is the nature of the strong evolution that is observed? Motivated by these puzzles, researchers have studied quasars over the last 20 years across the electromagnetic spectrum with space-based telescopes. These observations suggest that the energy source is ultimately derived from the gravitational energy of falling matter, in particular, gas that is falling into a black hole that has 1 million to 1 billion times the mass of the Sun. Inevitably the gas acquires a swirling motion as it falls, creating an accretion disk in orbit about the black hole. This gaseous disk suffers frictional (and possibly magnetic) heating to very high temperatures, perhaps creating a corona above the disk. As the resulting light is radiated away (the light observers see as the quasar), the material sinks closer to the black hole, eventually to fall inside the black hole's event horizon, thus increasing the mass of the black hole. In the meantime, new material has fallen onto the accretion disk to replenish the fuel supply. This picture has a number of testable features. It is predicted, for example, that the inner edge of the accretion disk is hotter than the outer edge, and, to first order, the emergent spectrum should reflect this difference in temperature. The inner edge will be orbiting more rapidly; in fact, the force of gravity this close to the event horizon is so large that effects of general relativity must be included in the calculations. The speed of rotation can be estimated from the Doppler-broadening of x-ray spectral lines. The disk-like geometry has bipolar symmetry, which is in accordance with the double-lobed structure frequently seen. Jets of material that are observed to be moving close to the speed of light are thought to be collimated by the disk and its magnetic field. A good candidate for a source of infalling gas is a surrounding host galaxy, and indeed quasars do seem to lie at the centers of galaxies (which is why they are commonly called active galactic nuclei, or AGN). While this model has provided a fruitful basis for the interpretation of new data, there are still a number of outstanding questions. Where did the original seed black holes come from? How does material in a galaxy- typically tens of thousands of light-years across get funneled into a volume that is light-hours across? How is it that the large masses of quasars and their chemical composition both indicate that they are highly evolved objects, yet the most distant quasars are seen so soon after the big bang? Questions like these need to be resolved before astronomers can construct a comprehensive picture of these important objects. Key Questions About Supermassive Black Holes and Quasar Power Sources The details of the physical processes governing the formation and growth of supermassive black holes, the prodigious luminosities of quasars, and the evolution of quasars (both collectively and individually) are still far from being resolved. Key questions in this area include the following:
44 A NEW SCIENCE STRATEGY FOR SPACE ASTRONOMY AND ASTROPHYSICS · What is the specific connection between the observed early peak in quasar activity and the formation history of massive black holes? · How big are the "seeds" of massive black holes and how fast do they grow? Do all galaxies harbor a large black hole? Do galaxy mergers lead to black hole mergers? Do quasars store up all their fuel early on, or do they acquire it gradually from outside? (If the fuel supply is stored, what turns the supply on and off? If the fuel supply is acquired from outside, what brings the fuel in?) How does matter interact with supermassive black holes close to the event horizon? How are energetic jets formed and collimated by the black hole? Why do some active galactic nuclei apparently never form jets? . . . Recent Progress in Understanding Supermassive Black Holes and Quasar Power Sources If quasars are considered as marking the locations of supermassive black holes, an empirical statistical picture of the origin and growth of supermassive black holes can be made by conducting a census of quasars up to high redshifts and down to low luminosities. On the other hand, to study the power sources of quasars requires a complementary approach, namely observing the detailed processes at work in the central regions of individual objects. Quasars are now known to a redshift of almost 5, corresponding to a time when the universe was less than 10% of its present age. However, at this early time they appear in reduced numbers, suggesting that observers are close to seeing the epoch of initial quasar formation. A potentially large population of quasars not evident from visible- light surveys was uncovered among x-ray and far-infrared sources. This finding suggests that many quasars may be obscured at visible wavelengths by surrounding dust. Careful studies of the host galaxies of quasars by HST and ground-based infrared imaging have revealed much new information. Moreover, similar observations have revealed low-power activity in the nuclei of many nearby galaxies (Figure 4.31. A number of significant advances in the study of quasar power sources have been made in recent years. Some important examples include the following: · The VLBA and HST were used to measure the masses of black holes in quasars. Having reliable masses sharpens the problem enormously and makes more detailed tests possible. · ASCA discovered extremely broad x-ray lines, showing the imprints of strong-field general relativity. This is a first in astrophysics and a clear signature of black holes. The lines come from matter close to the event horizon and tell us the spin of the black hole; potentially they could reveal much more. · The Energetic Gamma-Ray Experiment Telescope (EGRET) on CGRO revealed that blazers, the most active of galactic nuclei, emit most of their power as previously unobserved high-energy gamma rays. This observation confirmed the hypothesis that blazers contain energetic jets of matter moving toward us at close to the speed of light. · Features such as shocks moving along quasar jets at 5 to 10% the speed of light and surprisingly small and complex inner structures of active nuclei were mapped by applying the powerful technique of reverberation mapping. The technique relies on a combination of observations from multiple satellites covering visible to gamma-ray wavelengths and takes advantage of the intrinsic variability of the AGN to map structure surrounding the nucleus at scales too small to be resolved directly. The time delay between an increase in the continuum strength and a corresponding increase in line-emission strength, given plausible assumptions, reveals the location of the line-emitting gas relative to the photoionizing source. Similarly, in a blazer undergoing a flare, the delay in the appearance of the flare at different wavelengths reflects the time of propagation of a shock down the jet itself. · Short-lived analogs of quasar jets have been associated with binary star systems near the galactic center, according to observations made by the CGRO and ground-based radio and millimeter-wave telescopes.
GALAXIES AND STELLAR SYSTEMS 45 Future Directions for Understanding Supermassive Black Holes and Quasar Power Sources Progress on the origin and growth of supermassive black holes requires a more complete census of quasars than is currently available. Progress on quasar power sources requires measurements that reveal processes at still smaller physical scales. To answer the key questions relating to supermassive black holes requires the following necessary efforts: . 1. Searching for quasars out to the highest redshifts, including dust-obscured quasars, in the x-ray and far-infrared bands. AXAF and SIRTF will make deep surveys for obscured quasars; both will also search for the most luminous obscured quasars at the highest redshifts. Surveys for dust-enshrouded early quasars can also be undertaken with large, near-infrared and far-infrared telescopes. An x-ray telescope with 100 times the sensitivity of AXAF and X~M is needed to search for less luminous, obscured quasars, some of which may be in the process of formation. Good angular resolution is needed for reliable source identification at these faint levels. A wide field of view (10 to 30 arc min) is needed to obtain adequate numbers of quasars for statistical studies. 2. Examining the nuclei of nearby galaxies for weak activity and dormant black holes. Ultraviolet imaging and spectra and x-ray imaging can be used to look for faint active nuclei in nearby galaxies. Good angular resolution (subarc second in the ultraviolet and optical bands, arc second in the x-ray) is needed to resolve the galaxy in whose nucleus the quasar resides. The current instruments on the HST are well suited to detect weak active nuclei in nearby galaxies. A large collecting area in the x-ray band (100 times that of AXAF) is essential to gather enough photons from nearby, low-luminosity nuclei to study their spectra and variability. Radio sources with no recent injection of relativistic electrons radiate most strongly at low frequencies because the high-energy electrons responsible for high-frequency radio emission lose energy relatively quickly via synchrotron radiation. A low-frequency antenna could search for such relic radio sources. A location far from Earth, e.g., the far side of the Moon, would shield such a telescope from the serious limitations of telecommunications interference (see the section titled "New Astrophysical Windows and Cosmic Mysteries" in Chapter 5~. 3. Using gravitational tensing to search for hypothetical black holes outside galaxies. Deep subarc- second imaging of high-redshift objects to search for gravitational tensing may be the only way to find supermassive black holes outside the cores of galaxies. If the high-redshift objects are targeted, this search could be undertaken with an adaptive-optics system on a ground-based telescope. The wider field of view accessible from space, however, provides a special opportunity for serendipitous discovery. To answer key questions relating to quasar power sources, it will be necessary to accomplish the following: 4. Exploring the regions immediately surrounding black hole event horizons. X-ray lines distorted by general relativity provide astronomers' only access to the black hole event horizon. To exploit this information, an x-ray telescope with 100 times the collecting area of AXAF and good angular resolution (preferably approaching that of AXAF) is needed to remove spectral contamination from other sources when the target quasars are moderately faint. New x-ray optics technologies that promise to achieve large collecting area and good angular resolution at reasonable cost need to be pursued. 5. Understanding accretion and outflow processes associated with disks and jets. Observers can indi- rectly make images of the interior regions of quasars with reverberation-mapping techniques using data obtained simultaneously at multiple wavelengths. X-ray information is especially critical and will be available with limited collecting area from AXAF, Astro-E, and XMM. HST and SIRTF will extend the spectrum of the peak power- emitting regions in quasar jets and in the powerful central continuum source. Simultaneous ultraviolet/x-ray and infrared/ultraviolet reverberation maps the larger-scale accretion disk regions. Quasar jets can be "imaged" with simultaneous gamma-ray/x-ray/ultraviolet/optical monitoring, as could be done by a successor to EGRET with multiwavelength capability, specifically, on-board x-ray and optical/ultraviolet monitors. Ground-based TeV Cerenkov telescopes would be a highly cost-effective adjunct to this experiment. Optical/ultraviolet variations caused by gravitational microlensing offer another route to mapping out the accretion disk.
46 A NEW SCIENCE STRATEGY FOR SPACE ASTRONOMY AND ASTROPHYSICS Direct mapping of accretion disks the size of the solar system (10 AU) at the distance of the nearest giant galaxies (10 Mpc) requires an angular resolution of 1 microarc see; this sets the long-term goal. In the radio band, interferometric imaging of bright quasar cores at the 10-microarc see level can be demonstrated with current technology (e.g., VSOP). With further development, such interferometric imaging can also be accomplished in the infrared, visible, ultraviolet, and possibly even x-ray bands. In addition, next-generation space-based VLBI experiments, with enhanced sensitivity and frequency coverage, will explore jet structure and maser lines at resolutions of some 10 microarc sec. 6. Placing the active nucleus in the context of the host galaxy. Spectroscopy at moderate spectral and high angular resolution can map the kinematic field of the surrounding gas and stars. For nearby (redshift <0.04) active or dormant nuclei, near-infrared ground-based spectroscopy of the CO lines at 2.3 ,um, coupled with adaptive optics, will be an extremely powerful tool. Spectroscopy from space is necessary to apply this technique at higher redshifts. A separate approach would use thousands of channels of radio spectroscopy with the Green Bank Telescope (GET), VLA, and VLBA. which will allow sensitive searches for lines diagnostic of cool cast A small. multi wavelength (ultraviolet/x-ray) mission could map hotter matter in the cores of active galactic nuclei. CONCLUSIONS As detailed above, astronomers have an emerging picture for the origins of structures on scales ranging from those of clusters of galaxies to the compact, active nuclei found in individual galaxies. This evolutionary picture needs a firm empirical framework, for which there are exciting and realizable prospects in future space missions. TGSAA's priorities for the study of galaxies and stellar systems are, in rank order (but with items 2 and 3 of equal priority), as follows: 1. Detail the processes at work in the high-redshift universe by conducting a census of the universe as it was 1 billion years after the big bang. Such a census would identify the ancestors of the familiar structures observers see relatively nearby the gaseous precursors of galaxies, the earliest star-forming regions in galaxies, and the quasars at the highest redshift. Searches for these widely different astrophysical objects naturally require widely different observational techniques and capabilities. For example, the near-infrared spectral range is critically important because the energy from starlight in galaxies at high redshift peaks there, and the noise background is low in space. Angular resolution of 0.1 arc see or better is required to detect faint point sources, avoid source confusion, and study structure in distant galaxies. Likewise, x rays are an integral part of the census, for they can tell us of the earliest massive condensations and the formation epoch of quasars. Much of the starlight may be reprocessed by gas and dust in forming galaxies. Detecting such systems and studying their constituents at any redshift require high sensitivity at submillimeter wavelengths for both line and continuum radiation. At all wavelengths significant increases in collecting area can yield significant increases in knowledge, both through increased spatial resolution and increased spectroscopic sensitivity. 2. Link the high-redshift objects to their descendants by following the evolution to lower redshift, and undertake a detailed study of the underlying physical processes. In particular, the dynamical, structural, and chemical evolution of the various structures are all fundamental and need to be pursued. Opportunities include mapping the temperature distribution, kinematics, and chemical composition of x-ray-emitting gas in clusters, and detecting weak absorption lines at high spectral resolution at ultraviolet wavelengths. 3. Understand the formation and evolution of supermassive black holes in the nuclei of galaxies, and elucidate the processes at work there. Of particular interest is the physics of matter in extremely high gravita- tional fields near the black hole event horizon. The most promising approaches include measuring changes in x- ray line profiles, since these lines may be formed close to the horizon and may allow mapping of the space-time metric. A greatly increased aperture is needed to collect enough x rays for fully time-resolved reverberation mapping, which also requires simultaneous multiwavelength measurements over a wide spectral range.