MEMBERS OF THE GSMT SCIENCE WORKING GROUP

Transcription

1

2 MEMBERS OF THE GSMT SCIENCE WORKING GROUP ELIZABETH BARTON-GILLESPIE, Steward Observatory, University of Arizona JILL BECHTOLD, Steward Observatory, University of Arizona MICHAEL BOLTE, University of California, Santa Cruz RAY CARLBERG, University of Toronto MATTHEW COLLESS, Australian National University IRENE CRUZ-GONZALEZ, UNAM, Instituto de Astronomía, Mexico ALAN DRESSLER, Observatories of the Carnegie Institution of Washington TERRY HERTER, Cornell University PAUL HO, Harvard-Smithsonian Center for Astrophysics ROLF-PETER KUDRITZKI, Institute for Astronomy, University of Hawaii, Chair JONATHAN LUNINE, University of Arizona Lunar and Planetary Lab. CLAIRE MAX, LLNL and University of California, Santa Cruz CHRISTOPHER MCKEE, Physics Department, University of California, Berkeley FRANÇOIS RIGAUT, Gemini Observatory DOUG SIMONS, Gemini Observatory CHUCK STEIDEL, California Institute of Technology STEPHEN E. STROM, NOAO, Vice-Chair NOAO-NIO SUPPORT STAFF SAM BARDEN ROBERT BLUM ARJUN DEY JOAN NAJITA KNUT OLSEN STEPHEN RIDGWAY LARRY STEPP EXTERNAL OBSERVER TETSUO NISHIMURA, NAOJ COVER A montage of multi-color images obtained with the Hubble Space Telescope of the Antennae, a pair of colliding galaxies. The bright blue regions are newly-formed clusters of stars produced as a consequence of the collision. Systems similar to the Antennae were common when our expanding Universe was much younger, the average distance between galaxies much smaller, and the likelihood of collision much higher. Galaxy collisions can lead to mergers of the gas and stars initially bound to the two systems. Systems like the Milky Way appear to be product of multiple mergers. GSMT will be able to image colliding systems during the earliest evolutionary phases of the Universe, providing direct observation of the processes that give rise to the Milky Way. Courtesy: B. Whitmore, STScI

3 FRONTIER SCIENCE ENABLED BY A GIANT SEGMENTED MIRROR TELESCOPE (GSMT) TABLE OF CONTENTS REPORT OF THE GSMT SCIENCE WORKING GROUP EXECUTIVE SUMMARY...1 FUTURE ACTIVITIES OF THE GSMT SCIENCE WORKING GROUP...3 Members of the GSMT Science Working Group, 4 INTRODUCTION...5 THE ORIGIN OF LARGE-SCALE STRUCTURE IN THE UNIVERSE...6 Scientific Context and Questions, 6 The Role of GSMT, 7 BUILDING THE MILKY WAY AND OTHER GALAXIES...9 Scientific Context and Questions, 9 The Role of GSMT, 10 EXPLORING OTHER SOLAR SYSTEMS...12 Scientific Context and Questions, 12 The Role of GSMT, 14 SUPPORTING SCIENTIFIC REPORTS 1 THE POWER OF GSMT A THREE-DIMENSIONAL MAP OF GALAXIES AND GAS IN THE EARLY UNIVERSE STAR FORMATION IN THE VERY EARLY UNIVERSE BUILDING GALAXIES: THE PHYSICS OF GALAXY EVOLUTION BUILDING GALAXIES: THE HISTORIES OF MATURE GALAXIES THE ORIGIN OF THE STELLAR INITIAL MASS FUNCTION STUDY OF PLANET FORMATION ENVIRONMENTS CHARACTERIZATION OF EXTRA-SOLAR PLANETS Frontier Science Enabled by a Giant Segmented Mirror Telescope i

4 TABLE OF CONTENTS APPENDIX A TECHNOLOGY NEEDED TO ENABLE EXTREMELY LARGE TELESCOPES (ELTS)...A-1 Telescope Systems, A-1 Adaptive Optics, A-2 Site Evaluation, A-4 Instrumentation, A-4 The Need to Invest Now, A-5 ii Frontier Science Enabled by a Giant Segmented Mirror Telescope

5 REPORT OF THE GSMT SCIENCE WORKING GROUP BACKGROUND EXECUTIVE SUMMARY In spring 2002, the National Optical Astronomy Observatory (NOAO) was asked by the National Science Foundation to organize a Science Working Group (SWG) to advise the NSF Division of Astronomical Sciences on a strategy for guiding federal investment in a Giant Segmented Mirror Telescope (GSMT). Following a broad community solicitation (via the June 2002 NOAO Newsletter) seeking interested potential members, and close consultation with the Foundation, sixteen scientists active in ground-based astronomy research, and/or representing national and international groups that expect to play a role in developing next-generation telescopes, accepted invitations to join the SWG. Rolf-Peter Kudritzki, Director of the Institute for Astronomy at the University of Hawaii, agreed to serve as the SWG Chair. Over the period July 2002 through May 2003, the SWG held four meetings. Its primary focus has been on developing an understanding of: The forefront astrophysical problems likely to emerge over the next decade The science potentially enabled by-next generation telescopes Design options that can achieve that potential Technologies that must be advanced or developed in order to realize viable telescopes at acceptable cost The SWG analyzed these major GSMT issues in detail, collecting information through reports and presentations of the four principal private telescope design groups: University of Hawaii; the Magellan 20 consortium (Carnegie, Harvard, Arizona, Michigan, MIT); the Large Atacama Telescope (LAT) consortium (Cornell, Illinois, Chicago, Northwestern); and the CELT consortium (California Institute of Technology and the University of California). The SWG also heard reports from adaptive optics experts drawn from throughout the U.S. community, and received vital information from individual scientists who carried out simulations or calculations critical to assessing the potential performance of next generation telescopes. Presentations to the SWG and meeting summaries are available on the SWG web site ( Frontier Science Enabled by a Giant Segmented Mirror Telescope, the first report of the GSMT Science Working Group, summarizes and provides supporting material for the SWG s conclusions and recommendations. CONCLUSIONS The unique challenge of astronomy in the 21st century is to study the evolution of the universe in order to relate causally the physical conditions during the Big Bang to the development of RNA and DNA (Riccardo Giacconi, 2002 Nobel Prize in Physics). A 20-m to 30-m telescope will provide capability to meet this challenge. It will, for the first time, permit direct observations of hundreds of extra-solar giant planets, the disks from which planetary systems take form, the building blocks of galaxies and the process of galaxy assembly; the early evolution of chemical elements heavier than helium, and the emergence of large scale structure as mapped by galaxies and intergalactic gas during the first billion years following the Big Bang. Frontier Science Enabled by a Giant Segmented Mirror Telescope 1

6 EXECUTIVE SUMMARY This telescope will have the light gathering power and angular resolution to open up discovery spaces that virtually assure uncovering of unanticipated phenomena. From extensive analysis carried out by several groups, a 20-m to 30-m telescope can be built for costs within the envelope estimated by the most recent NAS/NRC decadal survey (Astronomy and Astrophysics in the New Millennium, 2001), i.e., approximately $700M. While there are significant technical challenges to building telescopes of this size, there appear to be no show stoppers. In order to reap the enormous potential synergy between the James Webb Space Telescope (JWST) and a 20-m to 30-m telescope, it is essential to initiate major design and technology development efforts now to ensure that facility operations coincide with the early JWST era. Private consortia are open to public-private partnerships to design, build, and operate a nextgeneration telescope. Federal investment now in a major technology development program targeted at key areas can advance multiple design programs, and will ensure a strong public voice at all stages in the development of next-generation telescopes. RECOMMENDATION The U.S. community appears poised to embrace a new paradigm: public-private partnerships to advance flagship research facilities. The SWG urges NSF to seize the moment and provide funding for advancing key technologies. 2 Frontier Science Enabled by a Giant Segmented Mirror Telescope

7 FUTURE ACTIVITIES OF THE GSMT SCIENCE WORKING GROUP This report, the product of one year s study, discussion, and analysis, represents an initial response to the NSF charge to the GSMT Science Group, which was to examine the science case and justification for any federal investment by NSF or other agencies in GSMT. As efforts proceed on telescope designs both in the U.S. and abroad, the SWG envisions a continuing, active role in ensuring a strong community voice in setting science performance goals, and in maintaining a productive dialogue with federal agencies. Over the next few years, the SWG specifically plans to: Seek further feedback from the community regarding the science cases we have identified, the key observations we foresee, and the methods by which they could be made Use the science cases to develop a community-based view of the performance goals and requirements for GSMT Review scientific instrument concepts Monitor the progress of key technology development Establish mechanisms to provide scientific feedback to groups involved in GSMT design efforts Establish working relationships with international groups in Europe, Canada, Australia, Mexico and Asia to serve as a forum where mutually beneficial collaborations could evolve Work closely with the JWST Science Working Group to prepare a plan for an optimized complementary scientific use of GSMT and JWST Report periodically to NSF Division of Astronomical Sciences, National Astronomy and Astrophysics Advisory Committee (NAAAC), and the Committee on Astronomy and Astrophysics (CAA) Evolve its membership to ensure input from a broad cross-section of the U.S. research community SWG intends to continue to post minutes of all meetings as well as all presentations on its web site: Comments regarding this report and other SWG activities are welcome and may be addressed to the Chair, R. Kudritzki and Vice-Chair, S. Strom. Frontier Science Enabled by a Giant Segmented Mirror Telescope 3

8 FUTURE ACTIVITIES OF THE GSMT SCIENCE WORKING GROUP MEMBERS OF THE GSMT SCIENCE WORKING GROUP ELIZABETH BARTON-GILLESPIE, Steward Observatory, University of Arizona JILL BECHTOLD, Steward Observatory, University of Arizona MICHAEL BOLTE, University of California, Santa Cruz RAY CARLBERG, University of Toronto MATTHEW COLLESS, Australian National University IRENE CRUZ-GONZALEZ, UNAM, Instituto de Astronomía, Mexico ALAN DRESSLER, Observatories of the Carnegie Institution of Washington TERRY HERTER, Cornell University PAUL HO, Harvard-Smithsonian Center for Astrophysics ROLF-PETER KUDRITZKI, Institute for Astronomy, University of Hawaii, Chair JONATHAN LUNINE, University of Arizona Lunar and Planetary Lab. CLAIRE MAX, LLNL and University of California, Santa Cruz CHRISTOPHER MCKEE, Physics Department, University of California, Berkeley FRANÇOIS RIGAUT, Gemini Observatory DOUG SIMONS, Gemini Observatory CHUCK STEIDEL, California Institute of Technology STEPHEN E. STROM, NOAO, Vice-Chair NOAO SUPPORT STAFF SAM BARDEN ROBERT BLUM ARJUN DEY JOAN NAJITA KNUT OLSEN STEPHEN RIDGWAY LARRY STEPP EXTERNAL OBSERVER TETSUO NISHIMURA, NAOJ 4 Frontier Science Enabled by a Giant Segmented Mirror Telescope

9 INTRODUCTION According to 2002 Physics Nobel prize winner Riccardo Giacconi, 21st century astronomy is uniquely positioned to study the evolution of the universe in order to relate causally the physical conditions during the Big Bang to the development of RNA and DNA. Taking the next major step toward achieving this grand synthesis will require a ground-based telescope of unprecedented power: the Giant Segmented Mirror Telescope (GSMT) The evolution of the universe from the dawn of known physics to the formation of atoms has been made plain by the Wilkinson Microwave Anisotropy Probe (W-MAP) cosmic microwave background mission. As seen in the diagram below, the GSMT will pick up the story from the appearance of the structural seeds manifest in gas in the most distant observable reaches of the universe, and follow it through the appearance of the first stars and pre-galactic forms, through 12 billion years of often violent evolution toward becoming the grand spiral and elliptical galaxies that populate the nearby universe, to forming solar systems surrounding newborn stars, and finally to mature planets surrounding neighboring suns. FIGURE 1 A schematic overview depicting the main chapters of the saga linking the origin of our Universe in the cosmic explosion that signaled the birth of time, to the origin of life on Earth. The enormous increase in light-gathering power and angular resolution provided by a Giant Segmented Mirror Telescope will allow us to decipher the text in five of the most consequential chapters of the saga: from the appearance of the first stars and galaxies through the formation of solar systems. Frontier Science Enabled by a Giant Segmented Mirror Telescope 5

10 THE ORIGIN OF LARGE SCALE STRUCTURE IN THE UNIVERSE SCIENTIFIC CONTEXT AND QUESTIONS It took a century. Fifty years after Einstein s remarkable theory of general relativity, cosmology was not considered a serious subject of scientific inquiry. How could fanciful speculations of a Big Bang or the steady state possibly be tested? Legitimacy came suddenly in the mid-1960 s with discovery of the cosmic microwave background the relic light of the Big Bang itself. A quarter-century later, a rocket-borne thermometer of exquisite precision removed all doubt about the Hot Big Bang model: the agreement of measurements with the predicted black body spectrum converted the last skeptics. Within a few years, observations confirmed astronomers suspicion that the universe is filled with a pervasive dark matter, mysteriously impervious to nature s dominant force of electromagnetism. The gravity of this abundant dark matter is in fact responsible for the growth of structures that made our own existence possible. By the end of the 1990s, astronomical measurements of supernovae explosions in distant galaxies, coupled with even-more precise cosmic background measurements from space, revealed that the universe is accelerating, a possibility raised in Einstein s theory, but dismissed by everyone, even Einstein. This acceleration is strong evidence of an unknown force in the universe that had not been discovered indeed could not have been discovered with a century s worth of particle accelerators. Cosmology the expression of the laws of physics in the ultimate experiment: the creation of the universe itself had gone from dismissed speculation to the alluring frontier of fundamental physics. Exquisitely sensitive all-sky maps have now been made at millimeter wavelengths by NASA s COBE (Cosmic Background Explorer) and W-MAP (Wilkinson Microwave Anisotropy Probe) satellites. Such maps record intricate patterns imprinted in the relic, fading light of the cosmic explosion that gave birth to our universe. These patterns miniscule fluctuations in the temperature of the ubiquitous 3 K background radiation produced by the Big Bang encode the basic physics of the universe, including the origin of visible structures in the universe: the star-filled galaxies and the gas between them. FIGURE 2 A color-coded representation of minute temperature fluctuations (typically one part in a million) in the otherwise uniform 3K radiation emanating from the cosmic explosion the Big Bang that gave rise to the universe and its constituents. This all-sky picture was made recently with W- MAP. Courtesy: NASA Cosmologists are busy decoding these maps; already they have made enormous progress in inferring the geometry, age and expansion rate of the universe, and inferring its likely fate. Physicists have developed new theories to explain the origin and 6 Frontier Science Enabled by a Giant Segmented Mirror Telescope

11 THE ORIGIN OF LARGE SCALE STRUCTURE IN THE UNIVERSE features of the W-MAP temperature fluctuation patterns, and the mysterious combination of dark matter and dark energy that they imply. Over the next decade, even more precise maps sensitive to an intensity variation of one part in a million will be made with ground-based millimeter-wave radio telescopes. Combining these maps with observations of the large scale distribution of galaxies, and precise distances to thousands of high redshift distant galaxies provided by supernovae, will give us the means to refine both our understanding of the laws that govern the evolution of our universe in space and time, and constrain theories that seek to describe the basic nature of dark matter and dark energy. We have good reason to believe that this course will lead to a revolution in our understanding of basic physics rivaling the landmark discoveries of the early 20 th century. Astronomers are equally excited by the challenge of understanding how, in the context of a rapidly expanding, post Big- Bang universe, these microscopic variations in the uniformity of the universe, engraved at the moment of origin, ultimately manifest themselves in the structures we can observe: the million light-year long webs traced by galaxies and intergalactic FIGURE 3 Results of a numerical simulation starting from a model universe whose basic parameters derive from analysis of COBE and WMAP data. This simulation illustrates the web of intergalactic gas threading the early universe. The brightest areas indicate regions where dark matter has concentrated gas into higher density clumps. GSMT has enough lightgathering power to use faint, distant galaxies as probes to detect intervening intergalactic gas, and to develop a tomographic map of the how it is distributed. The goal is to understand the link between the 3-dimensional distribution of this gas and the fluctuations observed in the microwave background, and to discover the interplay in dynamics and chemical composition between this gas and the first generations of stars in nascent galaxies. Courtesy: L. Hernquist gas. How do the largest structures in the universe evolve? What is the interplay between ordinary matter (gas, dust, stars) and dark matter? How does large-scale structure influence the formation and early evolution of galaxies? How do galaxies and their constituent stars take form from intergalactic gas, and how do galaxies through their stellar cycles enrich the gas? The GSMT will play a critical, unique role in answering these questions. THE ROLE OF GSMT The power of GSMT lies in its unique ability to obtain spectra of large numbers of galaxies at sufficiently high resolution to reveal via their sharp shadows against galactic light a forest of absorption features arising in intergalactic gas and diagnostic both of its distribution (via its Doppler-shifted Frontier Science Enabled by a Giant Segmented Mirror Telescope 7

12 THE ORIGIN OF LARGE SCALE STRUCTURE IN THE UNIVERSE velocity ) and chemical composition (via the strength of a heavy-element feature compared to the strength of a corresponding hydrogen absorption feature at the same Doppler-shifted velocity). The crucial importance in using faint galaxies as probes is to achieve a density of background sources that is high enough to map the clustering structure at its natural scale of tens of millions of light years. Only GSMT will have this ability. By using a large number of galaxies (on order a million probing a volume of 300 million light years on a side) and searching for features of common Doppler velocity along multiple lines of sight, it will be possible to produce a tomographic map of the intergalactic medium as it looked during the first 10% of the lifetime of the universe. Combined with the distribution of galaxies, this map will provide the basis for confronting model universes designed to predict large-scale structure from the fluctuations encoded in the cosmic background radiation. FIGURE 4 A simulation of the forest of hydrogen and metal absorption lines as they might appear observed against the spectrum of a faint quasar (top). The middle spectrum reveals the best result we could expect for today s 8 to 10-m diameter telescopes even with an all-night exposure. The bottom spectrum illustrates the potential of GSMT to deliver spectra capable of analyzing chemical composition and motions of intergalactic gas. To gather the million galaxy sample capable of providing a tomographic map of the intergalactic medium will require two full years of observation with a 30-m GSMT. Courtesy: J. Bechtold These same observations will provide a direct probe not only of luminous matter (contained in galaxies and stars) and gas, but of dark matter as well traced indirectly through its gravitational effect on the motions of both galaxies and gas. Hence we can learn how dark and luminous matter are distributed during the earliest evolutionary phases of the universe, and how dark matter influences the formation and evolution of galaxies. Furthermore, the distribution, composition and motions of the intergalactic gas can be compared with the distribution of galaxies in order to learn how elements produced by exploding supernovae in young galaxies enter the intergalactic medium and enrich it. 8 Frontier Science Enabled by a Giant Segmented Mirror Telescope

13 BUILDING THE MILKY WAY AND OTHER GALAXIES SCIENTIFIC CONTEXT AND QUESTIONS Observations with the Hubble Space Telescope and the largest ground-based telescopes point to the building of giant galaxies like our own Milky Way through multiple mergers and accretions of smaller systems. A very successful theory of the growth of structure in the universe, developed and advanced through numerical simulations with ever-more-powerful computers, paints much the same picture: Galaxy-sized concentrations of gas and stars emerged less than a billion years after the Big Bang. Pulled together by the gravitational lasso of invisible dark matter, these first galactic nurseries were soon ablaze with the intense energy of newborn stars. Galaxies as majestic as our Milky Way came later, assembled from tens of these embryonic galaxies that were initially spread out over a volume 1000 times larger than the Milky Way is today. Over the next several billion years these growing, infant galaxies were themselves united by mutual gravitation and reshaped, often by violent dynamical interactions, into today s familiar galactic forms. Each major merger was accompanied by spectacular bursts of newlyformed stars and stellar clusters. These fireworks were not only a celebration of the galactic birth process but also the source of the chemical elements heavier than hydrogen and helium, the first step in the long chain of complex chemistry that would one day lead to Earth-like planets, life, and sentient beings. As violent supernovae exploded these first heavy elements into the surrounding galactic and intergalactic gas, a complex, but eminently decipherable record of their origins was left for us to discover. Already we have seen hints of these remote events: observations of the most distant systems within reach of the Hubble and the Keck telescopes reveal adolescent galaxies which are young compared to our 13 billion-years-later perspective, but already mature in comparison to the building blocks that preceded mature galaxies. Closer to home, we have begun to examine in detail the death-spiral embrace of those few nearby galaxies that are only now merging an encore of a largely finished performance. And in the Milky Way and its close neighbors, we have begun to dig out evidence of long-completed galactic annexations through the fossil-like record of populations of stars left behind. FIGURE 5 A numerical simulation of galaxies in formation, showing vast clouds of gas (green) drawn together by the gravitational pull of dark matter. The higher density yellow and red regions are nascent galaxies. Courtesy: R. Dave, N. Katz, D.H. Weinberg Now we are ready to ask more specific questions. What are the masses, structures, and dynamics of pre-galactic systems? What varieties of stars were first formed and what mix of the elements heavier than helium did they produce? How did these proto-galaxies grow Frontier Science Enabled by a Giant Segmented Mirror Telescope 9

14 BUILDING THE MILKY WAY AND OTHER GALAXIES and what was the detailed history of their chemical enrichment? How did they come together to form the Milky Way and the other galaxies we see today? The answer will come by fitting together our observations of the earliest pre-galactic forms with our studies of the resulting galaxies of the modern universe that can endeavor to answer these fundamental questions. THE ROLE OF GSMT Observing the light of the first stars in the universe and imaging the earliest phases of the galaxy building process awaits the extraordinary sensitivity of the James Webb Space Telescope. A thorough analysis of those stars, including the physical and chemical properties of the pre-galactic building blocks in which they reside, will further require the enormous light gathering power of GSMT. Its spectrographs will be able to resolve light emitted from dense knots of gas and stars, enabling us to assay the evolving chemical mix. From precise measurements of the Doppler shift of this light, GSMT spectrographs will also measure the dynamical motions of individual clumps to determine the all-important total mass of the each system. GSMT s superb angular resolution will provide images of distant mergers and their spectacular star-forming regions. Together, JWST and GSMT will pull back the curtain on the last great mystery of galaxy genesis. FIGURE 6 Simulation (left) depicting a time sequence of a merger of two galaxies, drawn together by their mutual gravitational attraction to form a single system. On the right is a nearby pair of merging galaxies, the Antennae. In the color composite HST image of the central part of the Antennae are spectacular blue clusters of new stars induced to form in the violent collision that will eventually produce a single, merged galaxy. Courtesy: C. Mihos and L. Hernquist (simulation); B. Whitmore (HST image) GSMT will also have the power to carry out surveys of millions of galaxies spanning a wide range of ages from the earliest pre-galactic condensations to fully mature galaxies. From the rich information encoded in their spectra, astronomers will be able to chart the courses different galaxies followed as their total masses and abundances of heavy chemical elements increased. This is the key to developing a complete picture of how galaxies passed from infancy, through adolescence, to adulthood. The final piece of the puzzle will come from analysis of the archaeological record preserved in the chemical composition and orbital motions of stars in our own Milky Way and its neighbor galaxies. Distinct cohorts of stars with unique chemical compositions and orbital motions, identified through spectroscopy and through multi-color imaging in the crowded halos and disks of more distant neighbor galaxies, will allow us to decode the 10 Frontier Science Enabled by a Giant Segmented Mirror Telescope

15 BUILDING THE MILKY WAY AND OTHER GALAXIES number of major merger events and when they took place. This detailed record of what happened here in our own home galaxy will put in perspective how our own origins are connected to the vast array of possibilities we have observed for the panoply of galaxies. FIGURE 7 A simulation depicting streams of dynamically and chemically distinct stars colorcoded remnants of multiple past merger events comprising the halo of a Milky Way-like galaxy. Analysis of these cohorts via spectroscopy with GSMT will provide a complete genealogical record for the Milky Way and other nearby galaxies. Courtesy: P. Harding. Frontier Science Enabled by a Giant Segmented Mirror Telescope 11

16 EXPLORING OTHER SOLAR SYSTEMS SCIENTIFIC CONTEXT AND QUESTIONS In his 16 th century manifesto for the Copernican Revolution, the Italian philosopher Giordano Bruno asserted that: There are countless suns and countless earths all rotating around their suns in exactly the same way as the seven planets of our system. We see only the suns because they are the largest bodies and are luminous, but their planets remain invisible to us because they are smaller and non-luminous. The countless worlds in the universe are no worse and no less inhabited than our Earth. Up to the 1990 s, even the basic notion that planets existed around stars other than the Sun remained conjecture. We now know that the processes that give birth to stars are also likely to give birth to other planetary systems. Observations carried out with the Infrared Astronomy Satellite, the Hubble Space Telescope, and ground-based telescopes, reveal disks of orbiting gas and dust having size similar to the solar system and mass comparable with that of the planets, asteroids and comets now orbiting the sun. That some of these disks indeed form planets is now established with certainty, thanks to the pioneering observations of several groups in the U.S. and Europe. Today, over 100 planetary mass bodies have been discovered around stars near the Sun, and it appears that perhaps as many as one out of five stars like the Sun possesses at least one giant planet. FIGURE 8 Disks of solar system size and mass observed by the Hubble Space Telescope against the bright background of the Orion Nebula. Courtesy: M.J. McCaughrean & C.R. O Dell 12 Frontier Science Enabled by a Giant Segmented Mirror Telescope

17 EXPLORING OTHER SOLAR SYSTEMS Most surprising is the diversity of these systems few resemble our solar system. What accounts for the diversity of planetary architectures that we observe? How and when do giant planets like Jupiter form?. Did such gas giant planets form in a multi-step process of solids sticking together followed by accretion of gas a process likely to lead as well to rocky planets like the Earth or did they form like tiny stars, disrupting their nascent disks and hence mitigating against the existence of habitable worlds? How frequently can terrestrial planets similar to our own Earth and its near neighbors Venus and Mars form? How many are located in favorable locations for life? How many enjoy the protection offered by our own Jupiter which both shields our planet from an untoward number of collisions with asteroids, while redirecting icy bodies earthward to transport the water which gave rise to life-bearing oceans? Answers to these questions are within grasp. Over the next decade, continued ground-based observations will increase the number of planetary systems by several fold, both extending the sample of exo-solar systems and exploring domains of planetary separations heretofore beyond the reach of the past decade s studies. In space, both the Kepler satellite (which will detect planets as they eclipse their parent stars), and the Space Interferometry Mission (which will have the sensitivity to discover orbiting earths around nearby stars via their effects on the apparent motions of their parent stars) are expected to advance our understanding of planetary demographics. Both JWST and the largest ground-based telescopes may image a few nearby gas giants, provided they are located far enough from their parent stars. FIGURE 9 Artist s conception of the planetary system orbiting the nearby star 55 Cancri as seen from near the location of a Jupiter-like planet located at a distance from this star close to that of Jupiter from our own Sun. GSMT will be able to analyze the light from this planet, determine its chemical composition and infer the mechanism by which it formed. Courtesy: L. Cook Frontier Science Enabled by a Giant Segmented Mirror Telescope 13

18 EXPLORING OTHER SOLAR SYSTEMS THE ROLE OF GSMT GSMT is uniquely poised with its huge aperture (hence sensitivity) and advanced adaptive optics (hence sharpness) plus coronagraphy to pluck the faint light of large numbers of Jupiters out from under the glare of the parent stars. Young Jupiter-like planets can be imaged around stars out to the nearest star-forming regions, more than 200 light years away. More mature planets can be imaged around near-neighbors of the sun, out to distances of light years. GSMT s light-gathering power will enable spectroscopic analysis of the constituents of giant planet atmospheres. Their relative abundances will provide direct insight into how gas giant planets typically form and thus whether terrestrial planets are likely outcomes of the planet-formation process. FIGURE 10 A numerical simulation of a planet-forming disk. Note the gaps produced by the gravitational effects of forming planets on the distribution of orbiting disk gas The location of these gaps can be inferred by exploiting the light gathering power of GSMT to feed a sensitive infrared spectrograph capable of deconstructing the structure of the disk from high resolution measurements of emission arising from disk gas. Courtesy: University of Washington High Performance Computing Center. GSMT will also have the light-gathering power to peer into the disks surrounding just-born stars, to learn whether planetary systems begin to take shape within the first few million years of a sun s life. While light from the planets themselves will be too weak to see against the bright emission arising from the disk, their presence can be revealed through observations gaps created by the effects of a forming planet s gravity on orbiting gas and dust. From these observations, we can determine when and where giant planets form, and whether their location is benign or hostile to the development of life-bearing terrestrial planets. 14 Frontier Science Enabled by a Giant Segmented Mirror Telescope

19 SUPPORTING SCIENTIFIC REPORTS 1 THE POWER OF GSMT During its first two meetings, SWG members discussed the potential of next-generation telescopes to open new discovery spaces and to advance understanding of problems at the frontiers of current research. Members easily generated a long list of exciting research programs that would benefit from a telescope that provides an order of magnitude gain in sensitivity compared with today s largest facilities. However, marshaling the resources public and private needed to build and operate such a facility requires meeting a higher standard: namely, that in addition to the facilitation of important research programs, the GSMT must enable fundamental advances in understanding. After considerable debate, the SWG selected a representative set of fundamental problems for which key measurements lie well beyond the reach of current telescopes but which could be made with a GSMT. TABLE 1 Problem Key Scientific Problems and GSMT Performance Characteristics Key GSMT Performance Characteristics 3-D Map of Galaxies and Gas Galaxy Assembly History of Mature Galaxies Origin of the IMF Planet-forming Environments Direct Observations of Planets Early Star Formation 10 gain in sensitivity needed to obtain high signal-to-noise spectra of absorption features in the IGM 10x gain in sensitivity needed to obtain spectral diagnostics of mass, composition and star-formation rate 3x gain in angular resolution needed to resolve structure in forming galaxies and enable spectroscopic analysis 3x gain in angular resolution needed to resolve individual stars in nearby galaxies and obtain accurate photometry diagnostic of age and chemical composition 3x gain in angular resolution needed to resolve individual stars in dense, newly-formed star clusters and to obtain accurate photometry diagnostic of the distribution of stellar masses 10x gain in sensitivity needed to carry out ultra-high resolution spectroscopy capable of diagnosing the orbital motions of orbiting molecular gas 3x gain in angular resolution needed to separate planets from the glare of their parent stars 10x gain in sensitivity needed to carry out spectroscopy diagnostic of atmospheric chemical composition and structure 10x gain in sensitivity needed to enable spectroscopy diagnostic of the dominant masses of newly-formed star clusters and the chemical composition of the gas from which they formed Frontier Science Enabled by a Giant Segmented Mirror Telescope 1-1

20 1 THE POWER OF GSMT SWG members examined quantitatively the gains afforded by a GSMT of nominal size comparable to the 30-m diameter identified in the AASC decadal survey with the performance of the current generation of 8-m to-10-m class telescopes; the role of a GSMT in complementing JWST and ALMA, the two major facilities that will be completed at the beginning of the next decade, was also explored. A key component of these studies involved simulations of the potential performance of a GSMT equipped with realistic and achievable adaptive optics systems. Table 1 above lists the problems examined by the SWG over the past year and indicates the key characteristics of GSMT that will enable fundamental advances. 1-2 Frontier Science Enabled by a Giant Segmented Mirror Telescope

21 SUPPORTING SCIENTIFIC REPORTS 2 A THREE-DIMENSIONAL MAP OF GALAXIES AND GAS IN THE EARLY UNIVERSE Problem From our current understanding of how galaxies form, it is clear that there is an important feedback cycle between the galaxies and the intergalactic gas. The major questions that need to be addressed are: How does the primordial mix of hydrogen and helium gas collapse to form the first generation of stars and galaxies? How are the heavier elements returned to the intergalactic gas after they are produced by nuclear fusion in stars? How does this cycling of material between gas and stars affect the evolution of subsequent generations of stars and galaxies? And how does the large-scale structure of the universe the cosmic web affect galaxy formation? Approach These questions can be most directly tackled by observations at early epochs (using a sample of high redshift galaxies located at large distances and correspondingly large look-back times) when the bulk of stars in galaxies were being formed. A high-resolution three-dimensional map of the distribution of the galaxies and gas at large distances with spectra of high redshift, together with information on the enrichment of the gas with heavy elements, is an essential requirement for understanding the complex interactions between star-forming galaxies and the intergalactic medium. The overall approach is to map the galaxies by means of a redshift survey and to map the gas by locating it in absorption as a shadow against background sources (quasars and galaxies). With sufficient background sources the full threedimensional distribution of the intergalactic gas (and its enrichment) can be reconstructed from the many sight-lines. These maps of both the stellar and gaseous components will reveal the mechanisms by which gas is converted into stars and galaxies, and then recycled by supernovae, galactic outflows and other feedback processes. The maps will also reveal how larger-scale structures affect the processes of galaxy formation and produce the variety in morphological type, stellar content, age and metallicity that is observed in different environments. Key Observations The essential requirements for this program are the following: 1. The survey volume must be large enough to contain a sufficiently large number of galaxies so as to be immune to cosmic variance. A 10 8 Mpc volume co-moving with the cosmic expansion, comparable to the 2dF and SDSS surveys at low redshift, is an appropriate choice. 2. The survey must reach high enough redshift (or distance) so that it can probe the universe prior to the major epoch of star formation (i.e., redshift z>2 redshift, z, is the wavelength shift of the observed spectral lines redward from their laboratory wavelengths divided by their laboratory wavelengths) 3. The survey must sample the universe at that epoch sufficiently densely that the details of galaxy formation can be well constrained for typical galaxies, including all objects that are progenitors (or Frontier Science Enabled by a Giant Segmented Mirror Telescope 2-1

22 2 A THREE DIMENSIONAL MAP OF GALAXIES AND GAS IN THE EARLY UNIVERSE at least dominant components of the merger trees) of typical bright galaxies today, and ideally for significantly fainter objects. 4. The intergalactic medium (IGM) must be probed with at least the same level of resolution as the galaxy distribution (i.e., at least one sight-line passing close to every galaxy, and ideally more). Galaxy Survey A galaxy survey that fulfils these requirements would cover a huge 5 x5 region of the sky over the redshift range 2.5<z<3.5 corresponding to a volume 600Mpc x 600Mpc x 900 Mpc, or 3x10 8 Mpc 3. Galaxies can be efficiently selected in broad redshift ranges over 2<z<5 using multi-color optical photometry to identify the Lyman break, which is caused by the bound-free absorption of hydrogen in its ground state. These Lyman break galaxies (LBGs) are sufficiently numerous to be the likely progenitors of typical bright galaxies today. A 30-m telescope can obtain redshift identifications for LBGs down to magnitude R 26.5 in two-hour exposures. At this faint limit the surface density of LBGs is about 10 arcmin -2, which corresponds to a co-moving density slightly higher than the comoving density of bright galaxies today. A survey to this limit would include almost one million galaxies, comparable to the Sloan survey at low redshift. With a suitable wide-field (diameter ~ 15 arcmin), multi-object (multiplex ~ 2000) spectrograph, this survey could be carried out in about 100 nights and would provide an extensive inventory of the properties of the galaxy population and a precise determination of the large-scale structure at z~3. IGM Tomography To achieve a comparable density of probes of the IGM throughout this volume requires that the mean projected separation of background sources be comparable to the mean actual separation of galaxies i.e., we require of order (10 6 ) 2/ background sources over the 25 deg 2 of the survey, corresponding to a surface density of 0.1 arcmin -2. This is an order of magnitude more than the density of background quasars, but corresponds to the density of background LBGs at around magnitude R 24. Thus, a full three-dimensional reconstruction of the IGM requires using LBGs down to magnitude R 24 as the background sources (although quasars will also be used, since despite being relatively rare, they provide the smoothest and brightest continuum against which to measure Lyman absorption by the IGM). Suitable high-resolution, high-s/n spectra of magnitude R 24 LBGs for identifying and characterizing the absorption-line systems in the IGM can be obtained on a 30-m telescope with whole-night integrations. A high-efficiency, wide-field (diameter ~ 15 arcmin), multi-object (multiplex ~ 20) spectrograph could then carry out a complete tomographic survey of the full volume in 400 nights, yielding a complete map of the IGM and its enrichment on all scales down to individual galaxies. Need for GSMT A 20-m to 30-m telescope is required for its light-gathering power. For the galaxy survey, this allows redshifts to be measured for intrinsically less luminous galaxies and, for the progenitors of today's typical bright galaxies, metallicities, dust content, galactic wind energetics and stellar population distributions. For the IGM tomography survey, the gain is more than just aperture, since only a very large telescope can reach a sufficiently large number of background sources to provide enough sight lines for re-constructing 2-2 Frontier Science Enabled by a Giant Segmented Mirror Telescope

23 2 A THREE DIMENSIONAL MAP OF GALAXIES AND GAS IN THE EARLY UNIVERSE the full three-dimensional distribution of the IGM on the scales relevant to galaxy formation; with an 8- m to 10-m telescope, the limiting magnitude for a background source is less than magnitude R 23, and the surface density of potential background sources (quasars and LBGs) is more than 10x lower than is required for true tomography of the IGM. Synergy with JWST and ALMA The survey envisaged here is a necessary complement to the deep galaxy survey that is the centerpiece of the JWST science program. JWST will image a smaller area of sky in the near-infrared and midinfrared part of the electromagnetic spectrum, which corresponds to the rest-frame optical and restframe near-infrared light of galaxies at redshift z=2-5, and will obtain low-dispersion spectroscopy of the brightest galaxies in this range. ALMA will be capable of mapping the molecular content of these galaxies, an important part of understanding the process by which stars form. Just as the Keck 10-m was required to obtain spectra of galaxies in the HST Deep Field, spectroscopy with a 20-m to 30-m telescope will be necessary to realize the full scientific potential of the deep imaging surveys of JWST and ALMA. Only a very large telescope has the light-gathering power to measure redshifts for the faintest galaxies at low spectral resolution, and, at higher resolution, to determine masses, ages, and metallicities for the brighter galaxies, and map the locations and enrichment of the IGM gas clouds, in order to give a complete picture of the lifecycles of gas and galaxies at high redshift. Frontier Science Enabled by a Giant Segmented Mirror Telescope 2-3

24 SUPPORTING SCIENTIFIC REPORTS 3 STAR FORMATION IN THE VERY EARLY UNIVERSE Problem The aging stellar populations observed in well-formed distant objects at redshifts of z=4 indicate that the collapse of gas to form galaxies in the universe must have happened shortly after the Big Bang. However, with the exception of the microwave background, no astronomical sources with significantly larger distances corresponding to established redshifts > 7 have ever been detected. This leaves fundamental questions unanswered. When did the early stars and galaxies in the universe form? What were the properties of their stars? When did the intergalactic medium shift from neutral to ionized hydrogen and what objects were responsible for this shift? Approach Because most distant objects with extremely high redshifts emit little or no light at visual wavelengths, observations in the range of the I-band filter (z < 6) and the near-infrared (z > 6) are the best wavelengths to detect and study them. The Lyα line is a promising feature for detecting high-z star formation; at least one object at z > 6 has already been discovered. Other spectral features that may be used for confirmation are the Lyman break (caused by the continuous absorption of hydrogen in its ground-state) and the HeII line (1640 Å). Thus, an ideal approach to studying these objects would be discovery through either the Lyman break technique or through narrow-band imaging of Lyα, followed by spectroscopy for confirmation and more detailed study. Key Requirements The search for emission lines emanating from extremely high-redshift objects (z ~ 7-10) must be carried out at resolutions sufficient to isolate galactic emission from emission arising from terrestrial night sky emission. This requires relatively high spectral resolution and, correspondingly, the sensitivity provided by a GSMT. The survey must cover enough cosmic volume to sample the large-scale structure well and eliminate cosmic variance. In addition, it must resolve star-forming lumps well enough to determine their spatial distribution. Complications arise from the potentially high opacity of the interstellar medium (ISM) within the galaxy and the intergalactic medium (IGM) outside of it. However, recent analyses indicate that at least 10% of high redshift galaxies (z ~ 3) have low ISM opacities to Lyα. In addition, if these objects emit enough high-energy photons to create large bubbles of surrounding ionized gas, as much as 20% -30% of the Lyα photons will escape the IGM, even before the epoch of reionization. Thus, with the predicted high star formation rates of these early objects, their Lyα emission is readily detectable. Studying the Earliest Stars At present, we have no empirical knowledge of the properties of objects at z > 7. However, we can estimate the requirements for studying them using hydrodynamical simulations of galaxy formation (see Figure 11 below). A 2 x2 field of view covers a volume of ~ 5 Mpc co-moving with the cosmic expansion at a redshift of z=10, or an area slightly larger than the Local Group. Within this field, simulations predict several tens of objects that are detectable in Lyα with a ~30-m GSMT. With a 2 to 5 field of view (and multi-conjugate or ground-layer adaptive optics), a fair sampling of the universe can be achieved in 60- Frontier Science Enabled by a Giant Segmented Mirror Telescope 3-1

25 3 STAR FORMATION IN THE VERY EARLY UNIVERSE 400 telescope pointings (~ nights) using a narrow-band filter. After the initial discovery, follow-up spectroscopy requires a similar number of nights with a multi-object spectrograph in the near-infrared (multiplex ~ ). This follow-up will reveal the strength of the HeII (1640 Å) line, a probe of the initial mass function of these early stars. FIGURE 11 Early star formation from a hydrodynamical simulation by Dave, Katz, & Weinberg. Panel (a) shows a simulation of a collapsing galaxy in the newly forming universe at z=10, as seen in Ly-alpha radiation from cooling clouds of gas (green) and from newly forming stars (yellow, red). The remaining panels show simulated 8-hour observations of a forming galaxy with a 30-m telescope; the region is 1 x 1 comoving Mpc/h across and 1000 km/s deep. The observations are assumed to be between the OH lines at high resolution (R=3000) using ground-layer adaptive optics. The image represents 10 stacked exposures taken with a tunable filter centered at several selected wavelengths or, equivalently, a "collapsed" image of 10 resolution elements near Lyman alpha with an IFU. The panels show (b) observations of the Lyman-alpha line at z=10 assuming a Salpeter IMF with 1/5 Solar metallicity and a total escape fraction of 20\%, (c) observations of the HeII (1640) line at z=10 for a "top-heavy" IMF with a Salpeter slope and stars from solar masses with zero metallicity, and (d) observations of the HeII (1640) line at z=10 for a top-heavy IMF with zero metallicity. A GSMT will detect extremely high redshift galaxies in formation and diagnose the properties of initial mass function of these early stars. Courtesy: E. Barton-Gillespie Need for GSMT The sensitivity of a 20- to 30-m telescope is indispensable for the study of objects at extreme distances corresponding to redshifts of z ~ 10 objects. Although JWST may discover these objects, only GSMT will be able to take spectra with high enough resolution and sensitivity to probe the profile of the Lyα line and, potentially, the absorption in the nearby IGM. These measurements have the potential to probe the epoch of reionization of hydrogen in the early universe directly. The sensitivity of GSMT at high spectral resolution (~3000, between the contaminating OH lines of the Earth s atmosphere) substantially exceeds that of the tunable narrow-band filters designed for JWST (R=100). Thus, if objects at z > 6-10 prove fainter than expected but have strong emission lines, GSMT may be the only means of detecting them. In addition, the factor of ~5 potential improvement in diffraction-limited image quality over JWST will also enable morphological studies of these objects on scales < 100 pc using adaptive optics. 3-2 Frontier Science Enabled by a Giant Segmented Mirror Telescope

26 3 STAR FORMATION IN THE VERY EARLY UNIVERSE Synergy with JWST The broad and medium-band sensitivity of JWST will be highly complementary to GSMT. With broadband filters in the near-infrared, JWST will detect the ultraviolet continua of luminous high-redshift objects, revealing any systems in which Lyα is absent. If the JWST broad or medium-band surveys are efficient, they may replace the need for discovery with GSMT, directly enabling pointed follow-up spectroscopy with a multi-object spectrograph on GSMT. In any case, it is GSMT, through spectroscopic observation, that will enable a physical understanding of the stars that formed in the very early universe Frontier Science Enabled by a Giant Segmented Mirror Telescope 3-3

27 SUPPORTING SCIENTIFIC REPORTS 4 BUILDING GALAXIES: THE PHYSICS OF GALAXY EVOLUTION Problem Motivated by models of the collapse of matter in the universe, the hierarchical assembly theory provides an indispensable framework for studying the formation and evolution of galaxies. In this picture, galaxies are built through the smooth accretion of gas and the merging of smaller units. The central, remaining questions about this process can be broadly summarized as follows: When did the stars in the universe form? What are the intrinsic properties of the primeval galaxies we see at high redshift and how did they come to resemble the galaxies we see today? Approach The most direct method of answering these questions is through a census of the detailed properties of galaxies at large distances (and correspondingly large redshifts and look-back times), and the most efficient and comprehensive way to achieve this goal is an integrated program of spectroscopic surveys with three tiers. On the first tier, a broad optical spectroscopic survey over the full range of the distribution function of galaxy luminosities down to very faint objects will measure the star formation rates of galaxies as a function of redshift. The second tier will involve observing a subset of galaxies selected from the first-tier survey and obtaining deeper, higher-quality optical spectra that reveal the metal contents of the galaxies and the properties of their stellar population. The third tier of the survey will explore the spatially resolved properties of a subset of the galaxies in the near-infrared, revealing the dynamical masses and dynamical status of these galaxies as well as internal variations in their star formation and chemical enrichment histories. Together, these observations will reveal not only the early star formation history of the universe, but also the physical properties of the systems in which these stars formed. Key Requirements The typical galaxy in the early universe is not a very luminous system. Lower-luminosity galaxies are important sites of star formation; these systems may have merged to form many of the more luminous systems we see today. Thus, an understanding of the majority of the galaxies in the universe requires optical spectroscopy of fainter galaxies In addition, the first and second tier surveys of these objects must be large enough to avoid the problems associated with cosmic variance and to sample the galaxy population well as a function of redshift, luminosity, metallicity, and morphological type. The third tier study of the spatially resolved internal kinematics, abundances, and stellar populations of high-redshift galaxies will necessarily involve the most vigorously star-forming objects; it requires sensitive, spatially resolved spectroscopy of emission lines in galaxies on small scales (< 100 pc -1 kpc). These measurements must be made in the near-infrared because only rest-frame optical lines avoid extinction problems well enough to serve as kinematic indicators. The rest-frame optical emission lines, available only in the near infrared at high redshift, provide the best-understood measures of the gas-phase metallicities of star-forming regions. Mid-infrared capabilities should also be considered, because they would enable a complete census of the star formation histories of survey galaxies, including stars that are obscured by dust. Frontier Science Enabled by a Giant Segmented Mirror Telescope 4-1

28 4 BUILDING GALAXIES: THE PHYSICS OF GALAXY EVOLUTION The Building of Galaxies The first tier survey requirements can be met by the galaxy survey of ~10 6 galaxies sampling an area of 5 x5 to a depth of magnitude R ~ 26.5, as described in section 2 above, by designing the survey to sample the galaxy population well in redshift, luminosity, and morphology. In the two-hour exposures required for the survey, the GSMT can take optical spectra of sufficient quality for both redshift and star formation rate measurements from the red-shifted ultraviolet continuum. For the second tier, four-hour exposures for objects brighter than magnitude R 25.5 will result in high-signal-to-noise (S/N ~ 20) spectra ideal for measurements of the metallicities and initial stellar mass functions at redshift z = Within a arcminute field of view there are > 1000 potential targets. Thus, with a multiobject spectrograph (resolution 2000, multiplex ~ 1000), a high S/N survey of 10 5 galaxies will require 50 nights. The IGM tomography survey, also described above, will provide extremely high-quality spectra of 10 4 additional targets. Internal Properties of Forming Galaxies The third-tier survey of the internal properties of high redshift galaxies in the near-infrared will map the luminous galaxies. A 30-m GSMT will detect regions of high star formation rate to redshifts as high as z~5 (Figure 12). Targets with magnitude R < 25 appear at a rate of ~3 per square arcminute at 2.5 < z < 3.5 and ~1 per square arcminute at z > 3.5. With 8 hours of exposure time and multiplexed integral field spectroscopy of all 40 objects in the full ground-layer adaptive optics field of view (~10 10 ), measurements of the high-resolution internal kinematics of 1,000 galaxies with a spatial resolution < 1 kpc can be completed in 25 nights. For a detailed study of the distribution of star formation within the galaxies using multi-conjugate adaptive optics, a subset of ~240 objects (16 per 2 2 field) can be observed with ~100 pc resolution and greater sensitivity with 24-hour exposures in another 45 nights. Although high-redshift galaxies have very irregular structures, the measurements of internal motions within these galaxies will provide important constraints on the masses of their dark matter halos. These measurements will also enable us to distinguish between collapsing galaxies, dynamically hot collapsed systems, and the rotational motions of forming disks. Emission-line fluxes and flux ratios in the rest- FIGURE 12 Observing emission lines at high redshift. Panel (a) shows a narrow-band H-alpha image of the "Antennae", a nearby pair of interacting, star-forming galaxies. Panels (b)-(d) show simulated 8-hour observations of the Antennae at z=4.74 in the [OII](3727) emission line (K band) observed at high resolution (R=3000), between the OH lines. The panels show observations with (b) an 8-m telescope and MCAO (0.05" pixels), (c) a 30-m telescope and MCAO (0.01" pixels), and (d) a 30-m telescope and ground-layer adaptive optics (0.05" pixels). A GSMT equipped with adaptive optics will enable studies of the internal kinematics, starformation rates, and metallicities of extremely high-redshift, star-forming galaxies. Courtesy: E. Barton-Gillespie 4-2 Frontier Science Enabled by a Giant Segmented Mirror Telescope

29 4 BUILDING GALAXIES: THE PHYSICS OF GALAXY EVOLUTION frame optical also allow measurements of the star formation rate and the metal abundances as a function of position within these assembling or assembled galaxies. Need for GSMT Thus, a study of the physics of galaxy formation requires both the sensitivity and the spectral resolution of a GSMT. Existing facilities allow redshift measurements of the most luminous star-forming galaxies in the distant (z > 3) universe. A 30-meter GSMT would extend the sensitivity of an optical redshift survey to objects fainter at least 1.5 magnitudes down the luminosity function of galaxies. It would also allow us to take high-quality spectra at z=3; currently, such spectra are available only for high-redshift galaxies whose light has been amplified by gravitational lensing. In addition, 8- to 10 -m class telescope studies of internal kinematics are limited to just beyond z = 1. Since JWST will carry no instruments with sufficient spectroscopic resolution to measure the internal kinematics of high-redshift galaxies, a 20 to 30-m GSMT is an absolute requirement for the study of the internal kinematics of high-redshift galaxies. With adaptive optics, the spatial resolution of GSMT will also exceed that of JWST by a factor of ~5 for highsurface brightness objects, allowing the study of distant galaxies on sub-kiloparsec scales. Synergy with JWST and ALMA The spectroscopic capabilities of the GSMT will provide the perfect complement to JWST and its remarkable broad-band sensitivity in the near and mid-infrared. In particular, the imaging surveys with JWST can be used to select targets and reveal morphologies for the redshift surveys with GSMT. In the near-infrared, integrated spectroscopy with JWST will provide the best measure of the overall (spatially unresolved) metallicities of high-redshift galaxies in the rest-frame optical. The capabilities of ALMA will supplement the GSMT surveys by revealing obscured star formation regions and provide complementary measurements of the molecular gas distribution and kinematics in distant galaxies. GSMT will serve as the primary tool of choice for developing a deep understanding of the detailed physical processes involved in galaxy evolution. Frontier Science Enabled by a Giant Segmented Mirror Telescope 4-3

30 BUILDING GALAXIES: THE PHYSICS OF GALAXY EVOLUTION FIGURE 13 This figure depicts schematically the evolution of the universe from the fluctuations manifest in the microwave background, to the web-like structure of intergalactic gas and pre-galactic clumps drawn together by dark matter, to merging pre-galactic structures, to the mature galaxies that populate the universe today. At the earliest epochs, the universe is dominated by gas largely composed of hydrogen and helium. As time progresses, the first generation of stars and clusters form in the precursors of galaxies; the most massive of these stars become supernovae whose ejecta enrich the gas with chemical elements heavier than helium. Drawn together by gravity, galactic precursors merge, and over time develop the morphologies -- spirals and ellipticals -- that we see in the nearby universe. As time goes on, more and more gas is turned into stars, which in turn enrich the gas with metals produced and ejected during the course of stellar evolution. GSMT will have the power to map the 3-dimensional structure of gas and link the observed structure to the fluctuations in the microwave background, follow the evolution of the chemical evolution of the early universe through spectroscopic observations of intergalactic gas, and to determine the masses, star-forming rates, and chemical composition of pre-galactic clumps. 4-4 Frontier Science Enabled by a Giant Segmented Mirror Telescope

31 SUPPORTING SCIENTIFIC REPORTS Problem 5 BUILDING GALAXIES: THE HISTORIES OF MATURE GALAXIES Determining the formation epoch, star formation history, and chemical evolution history of galaxies and understanding the physical processes that govern these properties is one of the major challenges in astronomy and astrophysics. Outstanding questions are: (1) How are galaxies of different morphological types assembled and how do they evolve? (2) What are the star-formation and element production histories of typical spiral, elliptical and dwarf galaxies? And (3) How is the evolution of galaxies affected by their environment? Approach In the nearby universe, we can use the resolved stars of mature galaxies to reconstruct their entire star formation and chemical evolution histories. Comparing this knowledge with that obtained from in situ measurements of distant galaxies seen at large look-back times is critical for developing a consistent picture of galaxy formation and evolution. The diagnostic tools are optical and near-infrared stellar photometry with high enough precision and large enough sample sizes to allow the separation and quantitative characterization of stellar populations of different ages and overall metal/hydrogen ratios; medium-resolution spectroscopy of individual stars for determining the kinematics of major population subgroups; and high-resolution spectroscopy for measuring the detailed patterns of stellar element abundances that illuminate the complex relationship between star formation and chemical enrichment. Key Requirements Resolved-star studies suitable for recovering the evolutionary history of a galaxy have so far been carried out only for the Milky Way Galaxy and a few of its nearest neighbors. The power of current generation telescopes is not adequate to the task of resolving and analyzing stellar populations comprising the full range of galaxy morphologies found outside the Local Group which contains just two luminous spiral galaxies nor to observe the effect of cosmic variance on galaxy evolutionary histories. The distribution of galactic morphological types in the nearby universe is such that we require the capability of resolved-star studies in galaxies as distant as 10 Mpc; within a volume of this size, we find galaxies spanning the full range of morphological types (see Table 2 below). The requirements for imaging studies are precise (5% or better) photometry over modest fields of view (< 1 arcmin) at the diffraction limit of a 30-m telescope. These requirements demand an adaptive optics system with excellent image improvement (moderate to high Strehl ratio) in the near-infrared and a wellcharacterized point-spread-function; Optical and near-infrared spectroscopic studies for chemical element abundances drive the need for spectral resolution of ~40000, with kinematic studies possible with resolutions of Need for GSMT For crowded-field, point-source photometry, the gains provided by diffraction-limited resolution with large apertures are spectacular (see Figure 13 below). While in uncrowded fields longer exposure times provide greater depth of vision, in the high surface brightness main bodies of spiral and elliptical Frontier Science Enabled by a Giant Segmented Mirror Telescope 5-1

32 5 BUILDING GALAXIES: THE HISTORIES OF MATURE GALAXIES galaxies, crowding imposes hard limits on the achievable photometric depth within exposure times of only a few minutes! The crowding limit can only be lowered by dramatically increasing spatial resolution; lengthening the exposure time alone is futile. With the current generation of large telescopes, studies of the evolutionary histories of galaxies as described in this section are limited by crowding to the Galactic complement of low-surface brightness dwarf galaxies and to the outskirts of M31. The JWST will not provide any improvement over existing facilities, as its near-infrared resolution is no greater than the largest AO-corrected ground-based telescopes. As demonstrated in Figure 13, photometry reaching below the level of the horizontal branch (M J ~ 1) greatly improves the ability to separate distinct populations. To reach these limits in a significant sample of galaxies requires the high-angular resolution of a GSMT. Color-magnitude diagrams are useful diagnostic tools of a galaxy s star-forming past. Each point in a diagram represents the temperature and apparent luminosity of a single star, with fainter stars falling lower on the diagram; stars with different ages and chemical compositions fall along distinct, predictable sequences. The colored lines in the middle and right columns represent these sequences for the mix of stars input to the JWST and GSMT simulations. The number of faint stars visible with each telescope is limited entirely by the telescopes resolutions, not by the integration times (below) On the left is a diagram showing the relative star formation rates (SFR), ages (in Gyr), and bulk metallicities ([Fe/H]) of the stars used to simulate the JWST and GSMT M32 images shown above. A maximum-likelihood technique is then used to extract the mix of stars represented by the simulated color-magnitude diagrams, for both the JWST (middle) and GSMT (right). The simulation shows that for JWST and 8 to 10-m ground-based telescopes, crowding precludes diagnosing the mix of stellar populations in M32. By contrast, with diffraction limited near-ir images, GSMT can obtain photometry with the precision needed to determine the mix of ages and compositions in M32. FIGURE 14 A 2µm image of the inner 30 of the Local Group galaxy M32 taken with Gemini N. and the Hokupa a AO system (left). The middle and right images simulate the appearance of the same region as observed with JWST and GSMT, respectively. Courtesy: Olsen, NOAO, F. Rigaut, & B. Ellerblok, Gemini Obs. 5-2 Frontier Science Enabled by a Giant Segmented Mirror Telescope

33 5 BUILDING GALAXIES: THE HISTORIES OF MATURE GALAXIES FIGURE 15 The top figure depicts a color-magnitude diagram in which we plot the brightness (ordinate) and color (abscissa) measured for artificial star fields emulating the surface density of stars in the inner 30 of M32, and comprising a mix of stellar populations of different ages and metal abundance. The left diagram illustrates the accuracy of brightness and colors recovered using an image quality comparable to that delivered by the Hokupa a adaptive optics system on Gemini; the middle and right hand diagrams are for JWST and GSMT, respectively. The superior angular resolution of GSMT enables much higher precision photometry. High precision is critical to deducing the percentage of stars of differing ages and metal abundances. The bottom figure depicts the maximum likelihood recovered population mix for a threecomponent input population: [10%, 1 Gyr, solar metal abundance]; [45%, 5Gyr, solar metal abundance], [45% 10Gyr 1/2 solar metal abundance]. GSMT provides by far the most accurate recovery of input populations. Courtesy: K. Olsen, NOAO, and F. Rigaut, Gemini Obs. Thus, as shown in Table 2 below, a 30-m or larger GSMT is required to measure the evolutionary histories of a statistically significant sample of galaxies spanning the entire range of morphological types. Frontier Science Enabled by a Giant Segmented Mirror Telescope 5-3

34 5 BUILDING GALAXIES: THE HISTORIES OF MATURE GALAXIES For medium- and high-dispersion spectroscopy, the large collecting areas of GSMT-class telescopes are crucial for deriving kinematics of population subgroups and detailed abundances of chemical elements of intrinsically luminous but observationally faint red giant stars at distances of 3 10 Mpc. (Table 2 includes exposure times for obtaining near-infrared spectroscopy of red giants with resolutions of 1500 and at each of the listed distances.) TABLE 2 Power of Next-Generation Telescopes to Diagnose Stellar Populations in Crowded Fields Distances 1 Mpc 3 Mpc 6 Mpc 10 Mpc Luminous spirals Number of galaxies of type: (Tully 1988) Giant ellipticals LMC/M33 - like galaxies Dwarf galaxies ~70 ~170 ~300 Photometric depth in magnitudes 8-m (M J ) S/N=10 at 20-m J mag arcsec m Exposure time to reach imaging depth (seconds) 8-m m m Resolution 1500, 8-m 7500 X X X Wavelength 2.2µ Spectroscopic exposure time 20-m X X for red giants S/N=20 30-m X Resolution 40000, 8-m X X X X Wavelength 2.2µ Spectroscopic exposure time 20-m X X X X for red giants S/N=20 30-m X X X 5-4 Frontier Science Enabled by a Giant Segmented Mirror Telescope

35 SUPPORTING SCIENTIFIC REPORTS 6 THE ORIGIN OF THE STELLAR INITIAL MASS FUNCTION Problem The large majority of stars composing galaxies are believed to form in rich clusters containing between 10 4 and 10 6 stars confined within volumes of 3 to 30 pc 3. Such clusters are expected to be the most prominent features in pre-galactic clumps and newly forming galaxies. Understanding the kinds of stars that form in these regions and what processes control their formation represents an essential first step in understanding the star formation history of galaxies, including such basic properties at the overall star formation rate which relies on the extrapolation to lower masses of an initial mass function (IMF) determined locally. The shape of the IMF, i.e., the number of stars of a given mass, N(M), for stars more massive than 5 M, will control the total quantity and relative abundances of heavy elements enriching the galactic ISM and intergalactic IGM. Of equal interest to studying the emergent IMF in massive clusters is gaining an understanding of the star formation process itself in such dense, rich, and varied environments. Approach A combination of deep imaging and spectroscopy will be used to investigate massive star forming clusters in a range of environments from the Milky Way to M33. The approach will follow the recent work done in the nearby star-forming region of Orion, the Trapezium cluster. The basic idea is to sample the mass range as deeply as possible using spectroscopy to identify spectral types as a function of position in the observed near infrared color magnitude diagram (K magnitude vs. H K magnitude, CMD ). These data, when coupled with near infrared (JHK) colors, will allow distributions of age and excess color (due to accretion disks) to be determined as a function of position in the CMD. The mass function follows from a ``best fit'' to model isochrones for massive stars (main sequence) and low mass stars (main sequence and/or pre-main sequence) accounting for the interstellar extinction toward each star (or group of stars in the CMD). Key Requirements The observations require superb image quality (~20 milliarcseconds) in order to reach the faintest magnitudes in these crowded clusters. A GSMT is required in order to provide the accurate photometry that will yield accurate mass functions. A wide field of view is needed (1-2 arcminutes) to efficiently cover the nearby targets and thus allow a determination of the mass function with position within a given cluster. Such studies will be a primary tool to study how the emergent mass spectrum depends on local stellar density. The spectroscopic observations require modest spectral resolution (R = 1000 to 3000), and again very good angular resolution; the crowded fields and target densities require integral field spectroscopy and multi-object capability. Need for a GSMT The overarching requirement for this science case is angular resolution. The large diameter of the GSMT coupled to adaptive optics will enable diffraction limited imaging observations of individual stars with masses well below 10 solar masses in M33 (three M at two effective radii in the densest clusters)! Frontier Science Enabled by a Giant Segmented Mirror Telescope 6-1

36 6 THE ORIGIN OF THE STELLAR INITIAL MASS FUNCTION Observations in the Galaxy and LMC will reach well below the hydrogen burning limit to sample brown dwarfs in the densest clusters known (~0.01 M in a Galactic center-like-arches cluster in the LMC). Current 8m telescopes do not have the resolving power to reach even the hydrogen burning limit in the denser regions at a distance of the LMC. The large aperture of the GSMT will also allow for integral field spectroscopic observations reaching below a solar mass in dense clusters in the LMC and to nearly 10 solar masses in M33. Synergy with JWST This program is complementary to the primary science case for JWST which will explore the earliest epochs of star and galaxy formation in the high redshift universe. Those studies will depend critically on inputs derived from this study. The most basic properties describing the early universe and the epoch of galaxy formation depend on the IMF, its characteristic shape and mass limits. 6-2 Frontier Science Enabled by a Giant Segmented Mirror Telescope

37 SUPPORTING SCIENTIFIC REPORTS 7 STUDY OF PLANET FORMATION ENVIRONMENTS Problem The study of planet formation environments is a central part of our quest to understand the origin of the Earth and solar system. Indeed, our motivation to study planet formation environments is all the more intense today given the discovery of planets outside the solar system. The unexpected diversity in the properties of the extra-solar planet population has challenged traditional theories of planet formation and highlights fundamental questions: What are the protoplanetary disk conditions that lead to the formation of planetary systems? How common is the formation of solar systems similar to our own? Approach The likely complexity of the planet formation process emphasizes the need for direct observational study of young planet-forming systems (at ages 1 Myr or less) in order to identify the basic physical processes that are responsible for the diversity of planetary architectures. This framework is needed to assess whether and why our solar system is a common or rare outcome of the planet formation process. Since the angular scale subtended by the planet formation region of disks (~ < 10 AU: 1 AU is the distance between the Earth and the Sun) is beyond our ability to resolve spatially at the distance of the nearest star forming regions (~150 pc), high resolution spectroscopy is needed to probe disk structure: measuring the motions of orbiting gas can provide a map of disk temperature, density and composition as a function of radius, and locate the signatures of forming giant planets. Key Observations We need to measure the environmental conditions at planet formation distances in order to provide observational constraints on the efficiency of a variety physical processes (e.g., grain growth, orbital migration of forming planets) during the planet formation epoch. With velocity resolved profiles, we can determine the region of the disk responsible for the emission. Traditional emission line spectroscopic techniques can then be employed to measure the physical properties of disks. For example, from the measurement of multiple resolved line profiles, physical properties such as temperatures, densities, and column densities can be determined as a function of disk radius. We also need to carry out a census of a large sample of young planetary systems, measuring masses, orbital radii and eccentricities of planetary companions. These results can be compared with the properties of older (ages of several Gyr) systems (identified from ongoing radial velocity searches, and later in the decade, from astrometric measurements with the Space Interferometric Mission in order to begin to chart the evolution of planetary systems. We anticipate using GSMT to attempt detection of young giant proto-planets through the gaps that they induce in the disks in which they form. Since the width of the gap depends on planetary mass, measurements of the location and width of gaps in protoplanetary disks, deduced through spectroscopy, provides a potentially powerful tool for inferring both the masses and orbital radii of planets at their epoch of formation. Observations in the thermal infrared (4-30 µm) are ideal for the study of planet formation environments at radial distances < 5-10 AU, since the Planck function for disk material at these radii peaks in the mid- Frontier Science Enabled by a Giant Segmented Mirror Telescope 7-1

38 7 STUDY OF PLANET FORMATION ENVIRONMENTS infrared. At the warm temperatures ( K) and high densities of disks at < 5-10 AU, molecules are expected to be abundant in the gas phase, and sufficiently excited to produce a rich ro-vibrational and rotational spectrum. Despite the tremendous potential of thermal infrared spectroscopy to probe planet formation environments, this spectral region has remained largely unexplored due to the severe sensitivity limitations imposed by large thermal backgrounds, strong telluric absorption and inadequate sensitivity. Since JWST will not have high spectral resolution capability in the mid-infrared, the GSMT with its high sensitivity, has the opportunity to make the first detailed studies of the dynamics, chemistry, and physical structure of planet formation environments within 5-10 AU. Jy ) Fl ux ( Wavelength( µ m) FIGURE 16 A simulation (left) depicting the dynamical effects of a newly-formed gas giant planet (the red dot embedded within a dark elliptical ring) on a disk of circumstellar gas and dust surrounding a young star The gravitational effect of the planet on surrounding disk material opens up a gap, or ring, within which the amount of residual gas and dust is miniscule compared to the regions inward and outward of the ring. The residual gas produces a spectral signature (right panel, a simulated profile produced by a Jupiter mass planet orbiting a solar mass star at a distance of 1 AU), in this case a double-horned profile manifest in emission from carbon monoxide. The wavelength separation of the horns diagnoses the distance of the planet from its parent sun, while the width of each horn measures the width of the gap, which in turn diagnoses the mass of the planet. The simulated spectrum is representative of the expected performance of an R =100,000 mid-ir spectrograph on a 30-m GSMT for an 8-hour exposure. Courtesy G. Bryden (left) and J. Najita (right). Potential diagnostics include the CO fundamental lines at 4.7µm wavelength, mid-infrared rotational lines of water, and the pure rotational lines of molecular hydrogen, the most important of which (as far as ground-based observations are concerned) is located at 17µm wavelength. In order to measure resolved line profiles for gas at distances ~5-10 AU around solar mass stars, we require a velocity resolution of ~3 km/s or R=100, Frontier Science Enabled by a Giant Segmented Mirror Telescope

39 SUPPORTING SCIENTIFIC REPORTS Need for GSMT The order of magnitude gain in light gathering provided by a 30-m class telescope is critical to achieving the science goals outlined above: high signal/noise (s/n > 100) spectra with spectroscopic resolving powers R ~ 10 5 are required for a statistically significant sample of targets spanning a range of starforming environments. A large sample is essential for determining the frequency with which particular planetary architectures are found, and determining what fraction of systems are similar to our solar system. A typical survey target will be located in the Orion association (d ~ 480pc). With a 30-m GSMT, observing an Orion target will require a four- hour exposure to achieve s/n = 100 for molecular tracers spanning a range of disk temperatures. To study a sample of 1000 stars (approximately 10% of which may harbor giant planets) will require a year of on-target time with GSMT. Synergy with JWST and ALMA The pure rotational lines of molecular hydrogen and other gas phase diagnostics will be studied at moderate spectral resolution with SIRTF (R=600) and JWST (R= ). What will remain unclear even after these measurements are made is where in the disk the gas resides, and whether it resides in the region in which planets are believed to form (~ 5 AU) or at much larger distances. Thus, the role of the GSMT would be to measure the orbital radii from which the emission originates using high resolution, high s/n spectroscopy. GSMT will complement ALMA, which will probe the outer, cooler regions of planet-forming disks. Frontier Science Enabled by a Giant Segmented Mirror Telescope 7-3

40 SUPPORTING SCIENTIFIC REPORTS 8 CHARACTERIZATION OF EXTRA-SOLAR PLANETS Problem Precision radial velocity studies have revealed the existence of planets outside the solar system and have provided a hint of the diversity in the properties of extra-solar planetary systems based on their demographics (i.e., their orbital distances, eccentricities, and projected mass distributions). Even so, we know relatively little about the fundamental properties (such as mass, radius, and composition) of any individual extrasolar planet. Determining quantities such as planetary mass and radius can provide fundamental constraints on the evolutionary history of the planet. (An example is the star HD209458b, which is eclipsed an orbiting giant planet allowing for a determination of the planetary radius.) Measurements of the chemical composition of extra-solar planets can also provide a new way to address the question of how planets form. Approach With high contrast imaging and spectroscopy, it will be possible to directly image and disperse the light from extra-solar giant planets. With this capability, we will be able to measure the system inclination from the visual binary orbit, which will establish the mass of the planet. Further multi-wavelength imaging or spectroscopy of the planet will characterize the planetary atmosphere in terms of its cloud composition and planetary albedo. Given a measured albedo, the planetary radius can be deduced from the luminosity of the planet. Monitoring will reveal large-scale circulation patterns. GSMT will have the power to expand planetary science to encompass planets outside the solar system and usher in a new era of comparative planetology. By measuring the relative metallicity of the planet compared to its parent star, it should be possible to constrain the mechanism by which the planet formed. Two competitive formation theories are: (1) the classical accumulation scenario (agglomeration of solids followed by accretion of gas) which predicts metal-rich planets, and (2) the disk instability scenario, which predicts no compositional difference between the star and planet. Both of these theories appear to be able to explain the general properties of extra-solar giant planets, but they make different predictions for the formation of terrestrial planets: terrestrial planets appear to be much more likely to form and survive in the classical picture. Thus, by using measurements of chemical compositions for extra-solar planets, a statistically significant sample of planets, we can determine the dominant planet formation pathway, and address thereby the likelihood that systems similar to our own are rare or commonplace. Key Observations Precision radial velocity surveys have now been in operation for long enough that they are beginning to detect Jupiter-mass objects with orbital separations similar to that of Jupiter in our own solar system, i.e., at orbital separations of ~5 AU. (An example is the star 55 Cancri, see Figure 17 below.) With GSMT, we will be able to detect such planets directly (with low resolution [ 10] spectroscopy at 1.2 µm, s/n=25 in 10 5 sec) if they are in orbit around nearby stars (distance < ~ 50 pc). We will be able to physically Frontier Science Enabled by a Giant Segmented Mirror Telescope 8-1

41 8 CHARACTERIZATION OF EXTRA-SOLAR PLANETS characterize these planets in detail (with medium resolution of 1000) spectroscopy at 1.2 µm, s/n=25 in 10 5 sec) if they are in orbit around the nearest stars (distance < ~ 10 pc). Similarly, old (~5 Gyr) Jupiter mass planets at smaller orbital radii (~ 0.5 AU) can also be studied for distances > ~ 10 pc. FIGURE 17 Top: A plot of the predicted spectra from (a) the nearby solar-like star 55 Cnc; (b) the warm, gas-giant planet (55Cnc b) located within 0.11 AU of its parent star, and (c) a Jupiter-like planet (55 Cnc d) located at nearly the same distance from 55 Cnc as Jupiter is from the Sun. (Courtesy: Sudarsky, Burrows, and Lunine, private communication.) Spectral features arising from methane, ammonia, and water are indicated in the upper and lower panels. Lower panel: The ratio of planetary flux to parent star flux for the two companions to 55Cnc. The red line indicates the level to which the light of 55 Cnc could be suppressed by the occulting disk of a well-designed coronagraph fed by an adaptivelycorrected GSMT image of the Cnc system. GSMT will have the power to detect both 55 Cnc b and d, and to determine the chemical composition of the atmosphere of 55 Cnc b from analysis of its spectrum. 8-2 Frontier Science Enabled by a Giant Segmented Mirror Telescope

42 SUPPORTING SCIENTIFIC REPORTS We can potentially expand the sample of planets accessible for spectroscopic study by searching among younger, and therefore brighter, extra-solar giant planet population. Surveys of the solar neighborhood (d < 100 pc) have identified a population of young (~100 Myr) stars that is a target subsample of a major SIRTF Legacy Program. Some of these systems have recently been found to have spectral energy distributions indicating the presence of debris disks, possibly with central holes that are carved by young, inner giant planets. The GSMT will be able to characterize these planets at 1.25 µm (Resolution = 1000, s/n=25, 10 5 sec) out to distances of ~70 pc. FIGURE 18 A plot depicting the mass of known exo-planets detected from radial velocity studies against their inferred distance (astronomical units) from their parent stars; the locations of Venus (V), Earth (E), Jupiter (J), Saturn (S), Uranus (U) and Neptune (N) are indicated for reference. The scales located at the interior top and bottom of the plot indicate the distance (parsecs) within which a planet of a given distance from its parent star can be detected by a 30m diameter GSMT at 1.2 µm (top) and 5 µm (bottom); detectability at intermediate wavelengths can be judged by the blue and green lines. As an example, a 1 Jupiter mass planet located 0.6 AU from its parent star can be imaged and characterized by GSMT at 1.2 µm provided its parent star is located within 10 pc; a 1 Jupiter mass planet located at 1.8AU can be seen out to 30 pc. In both cases, planets more distant than 0.6 AU and 1.8 AU, respectively, can be resolved easily. The red line indicates the mass range of planets that will be accessible to next generation radial velocity searches with 5m/s precision; planets with masses and separations that place them above the line can be detected. Courtesy: T. Herter, Cornell University Frontier Science Enabled by a Giant Segmented Mirror Telescope 8-3

43 8 CHARACTERIZATION OF EXTRA-SOLAR PLANETS Need for GSMT With its high sensitivity and angular resolution, the GSMT has the potential to make a significant contribution to the study of extra-solar planets. High angular resolution is needed to separate the planet from its central star, and high sensitivity is needed to acquire spectra of sufficient spectral resolution to measure the metallicities of planets. GSMT will be able to resolve planets located at angular distances of 40 mas and 160 mas at 1.2 microns and 4.6 microns respectively, corresponding to 0.4 AU and 1.6 AU at a distance of 10 pc. Successful realization of the full potential of GSMT will require development of high performance adaptive optics systems coupled to innovative coronagraphs, which together can suppress the light from the central star by 10 6 early in GSMT s operational lifetime and between 10 8 and 10 9 later in the GSMT era. Synergy with JWST The ability to detect spectral features across a wide range of wavelengths will provide the best constraints on abundances, albedoes, and surface gravities. While GSMT will excel in the region shortward of 5µm, JWST will be the facility of choice at longer wavelengths. The much smaller diffraction limit of GSMT compared to JWST, and the possibility of sophisticated coronography with GSMT, will allow it to probe planets at much smaller angular separations than is accessible to JWST. 8-4 Frontier Science Enabled by a Giant Segmented Mirror Telescope

44 APPENDIX A TECHNOLOGY NEEDED TO ENABLE ELTS Key to realizing the science goals for the ELT will be investment aimed at advancing technologies central to achieving its performance potential. The areas of most urgent need follow from the science goals and the derived telescope/instrument requirements. These are summarized in the following table. TABLE 3 Capabilities Required to Enable Five Primary ELT Science Objectives SCIENCE OBJECTIVE CAPABILITY REQUIRED 1 The formation of galaxies as derived from their archeological record Near-diffraction limited performance over a ~2 arcmin field of view 2 Characterization of extra-solar planets High dynamic range imaging 3 The birth of planetary systems High sensitivity mid-ir spectroscopy 4 The formation of galaxies witnessing the process directly Enhanced seeing observations over ~5 arcmin field 5 Formation of large scale structure in the universe Seeing-limited multi-object spectroscopy over a wide field (~ 20 arcmin) In turn, instrument requirements map into needed component performance. In its analysis of these requirements, the SWG identified the need for near-term technology investments in four areas: (1) telescope systems, (2) site evaluation, (3) adaptive optics, and (4) instrumentation. Telescope Systems Telescope technology development will be needed in several areas, foremost among them optics, coatings, and advanced control systems. Figure 19 below illustrates three different ELT concepts currently under development by various groups, and highlights areas of technology development in common to these concepts. For example, adaptive secondary mirrors are needed in the Magellan 20 concept, as well as the GSMT concept, where an emphasis has been placed on low system emissivity to support thermal infrared science (Science Objective 3 in table above). An adaptive secondary will also be used for correcting windshake-induced tip/tilt errors in these telescopes. It is also likely that such secondary mirror technology can be used in Ground Layer Adaptive Optics (GLAO) applications, which offer the potential of wide-field adaptive optics correction to the telescope s focal plane (Science Objective 4 in table above) albeit with reduced resolution compared to more conventional, narrowerfield high order adaptive systems. A-1 Technology Needed To Enable Extremely Large Telescopes

45 A TECHNOLOGY NEEDED TO ENABLE ELTS Figure 19 further indicates the need for high reflectivity coatings on the primary mirror that can be easily cleaned in place, and do not require recoating for a decade or more. The driver for this technology is twofold. First, all science programs will benefit from high-reflectivity broadband (e.g., µm) coatings because integration times to yield the same sensitivity will be reduced. Second, it is crucial that the large number of segments required by the segmented-mirror ELT concepts not require frequent swapping with freshly-coated segments; the operational costs and risk associated with segment swapping, integrated over the lifetime of the observatory, will be prohibitively high. The viability of light-weight substrates for mirrors also needs to be assessed early in the development of segmented mirror ELT concepts, given the long lead time required to manufacture and polish on order 1000 individual segments. The use of advanced substrates like SiC will significantly reduce the overall weight of segmented mirror support systems, which in turn will lead to lower fabrication costs for the telescope, potentially higher bandwidth control systems, and improved image quality. CELT GSMT Magellan 20 Adaptive Secondary Mirrors High Performance Durable Coatings Advanced Primary Mirror Materials Active Optics Control Systems FIGURE 19 Three proposed ELT concepts, with key technologies that will be needed to support these different telescope designs. Courtesy: CELT, AURA-NIO; Magellan 20 Consortium Adaptive Optics The second general area of ELT technology investment needed is adaptive optics (AO). Unlike many current telescopes, advanced AO systems will be designed to serve as an integral part of the ELT and its controls system from the outset. Several types of AO systems are required, as shown in Table 3 above. For example, a multi-conjugate adaptive optic (MCAO) system, working in conjunction with a sodium laser system, will be needed to isolate distinct stellar sub-populations which are in turn traceable back to ancestral galactic components (Science Objective 1). To produce MCAO systems on the scale required for a 30-m ELT, detailed system design studies will be needed to optimize performance. Moreover, investment is needed to ensure the availability of together low cost and highly reliable sodium lasers and deformable mirrors with upwards of 5000 actuators. For the latter, a completely new manufacturing technology may be needed, as current technologies for producing stacked actuator mirrors are not easily A-2 Technology Needed To Enable Extremely Large Telescopes

46 APPENDIX scalable to the level of many thousands of actuators given the costs, high voltage drive electronics, and beam sizes required. Instead, other approaches to making deformable mirrors should be explored, including the use of Micro-Electronic Mechanical Systems (MEMS) technology, which stands to yield deformable mirrors on vastly more compact and lightweight scales than has been possible to date. GLAO is another adaptive optics mode that an ELT will need to enable observations at modest strehls but over wider fields than MCAO can provide. A GLAO system will drive deformable mirrors to even larger formats than MCAO systems, particularly if it must work at optical wavelengths. Figure 20 shows predicted ELTdelivered image quality with and without GLAO corrections. For fields of view ~5 arcmin diameter or less, the level of correction offered by GLAO will yield significantly enhanced throughput for integral field spectroscopy, and should be well matched to faint extended targets over large fields. While GLAO holds promise, it is predicated on the assumption that the bulk of the line-of-sight turbulence lies within a few hundred meters of the ground. For this technique to succeed, the ELT must be located on a site that has modest ground-layer turbulence, and relatively steady small contributions to seeing from high altitude turbulence. Investment in detailed site evaluation is therefore an important factor in implementing a GLAO system. FIGURE 20 GLAO model delivered image performance is plotted assuming a Cerro Pachón turbulence profile. Note how for fields of view smaller than ~5 arcmin, this approach to adaptive optics correction provides significant improvement at optical and near-infrared wavelengths. The natural seeing image sizes are indicated by the lines Laser systems represent another above for V (0.5 µ) [solid line], J (1.2.µ) [dotted line], and K key technology needed for ELTclass adaptive optics. ELTs will (2.2 µ)[dashed line]. The GLAO-corrected image sizes are indicated by the curves below for the same wavelengths. require low cost ~50 W sodium Courtesy: F. Rigaut, Gemini Obs. lasers with good beam quality, high power efficiency, and reliability that is as high as any other subsystem on the telescope. Such lasers will vastly increase sky coverage compared to natural guide star systems (which typically enable adaptive correction over 10% of A-3 Technology Needed To Enable Extremely Large Telescopes

47 A TECHNOLOGY NEEDED TO ENABLE ELTS the sky or less), as well as enhance photometric accuracy across extended fields by enabling accurate wavefront sensing and correction. New large format, high speed, low noise detectors will also be critical to accurate wavefront sensing. Current wavefront sensing detectors must be advanced to formats with ~2 e - read noise and khz read rates. Even larger format detectors will be needed for LGS AO configurations. These represent significant capability enhancements when compared to current typical ~ format detectors used for wave front sensing. In summary, ELT AO systems will require investment in deformable mirror technology, sodium lasers, and wavefront sensors to effectively define the optical bridge between the telescope and its instruments Site Evaluation and Selection Site evaluation and selection lies on the critical path for most, if not all ELT programs. A variety of characteristics must be assessed for a large number of potential sites, including fractional cloud cover, precipitable water vapor, prevailing winds and wind flow across local topographic features, seeing (turbulence profiles including the ground layer), geologic activity, etc. Investment in site characterization has already begun via analyses of large global satellite databases, as well as in situ measurements using remote weather stations and seeing monitor facilities. More investment along these lines will be needed to build a comprehensive database to enable assessment of potential ELT sites. Finally, new devices such as the Multi Aperture Scintillation Sensor (MASS 1 ) need to be procured and stationed at several sites to provide real-time atmospheric turbulence as a function of height. All these activities are key to making a timely decision as to where an ELT should be located. Though fragmentary progress has been made in evaluating a few selected sites, many more need to be assessed within the next few years if an ELT is to begin operation early in the JWST era. FIGURE 21 Part of the site evaluation work needed for an ELT involves remote sensing of weather patterns through existing satellite-based databases. Courtesy: M. Sarazin, ESO 1 Tokovinin A.A. A new method to measure atmospheric seeing. Pis'ma v AZh., 1998, V. 24, P (Astronomy Letters, P ). A-4 Technology Needed To Enable Extremely Large Telescopes