Chapter 2

Chapter 2, The Science Case for a GSMT

NOAO Logo    Gemini Logo

The 20th century witnessed a revolution in our knowledge of the universe. We now know that the Sun is a star, composed primarily of hydrogen, and only one among more than 10 billion stars in our Galaxy, the Milky Way. Our Galactic home itself is but one among billions of galaxies spread out over the vast expanse of space and time in a distribution whose complexity we are only beginning to appreciate. We have also learned that these galaxies are expanding away from each other as the result of the explosive birth of the universe more than 10 billion years ago. Closer to home, we have found compelling evidence for planets around nearby stars, which raises the possibility of life and intelligence elsewhere in the Galaxy.

As a result of these discoveries, our questions about the universe have become progressively more sophisticated. What is the origin of structure in the universe? From the cosmic soup of the Big Bang, how did clusters, galaxies, stars, and planets arise? How did the chemical evolutionary history of the universe result in the genesis of life on Earth? And how common is life in the universe? The keys to answering these questions lie in deciphering the past: ancient events observable in the distant universe, the historical record preserved in the stellar populations of the Milky Way and nearby galaxies, and present-day analogues of the birth of the Sun and solar system.

Although current 8-m and 10-m telescopes will greatly advance our understanding of these questions, we already know that many answers lie beyond their reach. As we describe below, answering them will require a novel combination of greater sensitivity, higher angular resolution, and larger fields of view (FOVs) than are currently available. Moreover, the discoveries that will be made by planned space-based or multi-wavelength ground-based facilities will only be fully realized with a new large-aperture optical-infrared telescope. For example, the stars that first illuminated the universe may be detected by NASA's James Webb Space Telescope (JWST), but investigating their astrophysical properties (e.g., age, metallicity) is well beyond the capabilities of either JWST or any extant ground-based telescope. Similarly, the Atacama Large Millimeter Array (ALMA) will probe the current birthplaces of stars in the Galaxy, but understanding the formation of planets in our solar system will require a joint effort, relying heavily on a next generation large optical-infrared telescope. Indeed, without investment in such a telescope with sensitivity and angular resolution matched to the power of ALMA, JWST, and SKA (Square Kilometer Array), the US community will lack access to capabilities critical to exploring the astrophysical frontiers.


Advances in astronomy in the next decade will be enabled by a diverse set of observing tools covering a broad range of wavelengths, on the ground and in space. By 2015, JWST will have been launched, and perhaps will have already completed its nominal five-year mission to explore the early universe and the creation of the first stars and galaxies. It will undoubtedly have revolutionized our view of the near- to mid-infrared universe. ALMA will be operational (target completion date 2011) and examining an equally broad range of astronomical problems, from the study of dusty galaxies at high redshift to the initial conditions and physical processes that produce stars and planets. For the first time, a facility operating at mm wavelengths will provide sensitive observations of the cold, molecular universe at angular resolutions ~ 10 mas—nearly 100x that achievable today and 10x that of the Keck telescope. Before the end of the next decade, SKA will bring analogous sensitivity and angular resolution to bear on the problem of mapping the evolution of primordial hydrogen gas from the epoch of recombination to the formation of the first gaseous structures in the universe. Chandra will have long completed its inventory of the X-ray background, while Constellation-X will be on the verge of probing the formation of galaxy clusters and the distribution of hot, ionized gas in the universe.

By 2015, 8-10-m class ground-based OIR (optical-infrared) telescopes will have been operating for nearly two decades. Experience with adaptive optics (AO) will have produced huge gains in sensitivity and perhaps brought us to the point of replicating at near- and mid-IR wavelengths the excellent imaging quality available in space. We will by then have glimpsed the high-redshift universe, perhaps to z > 8, as well as the large scale structure of galaxies at z ~ 4 based on studies of the brightest (> L*) galaxies. However, the nature of the first collapsed objects (z > 8) and the clustering properties of sub-L* (i.e., typical) galaxies, which provide the critical test of hierarchical formation models, will lie beyond the sensitivity and angular resolution provided by these facilities. Closer to home, we will have studied planet forming systems within ~ 200 pc using 8-10-m ground-based telescopes and thereby obtained, from samples of a few tens of systems, tantalizing hints to the questions of when and how planets form. However, definitive answers to these questions and the central question of how frequently planets form, as well as an understanding of the role of dynamical evolution and its impact on the habitability of planetary systems, will await a more powerful facility capable of providing a census of planetary architectures for thousands of forming stars and their planet forming accretion disks.

In this climate of intense activity, in which numerous discoveries will be made and fundamental astrophysical problems solved, the availability of forefront ground-based optical and infrared telescopes and instrumentation will remain central to expanding the frontiers of astrophysical knowledge. Thesefacilities will not only drive unique, ground-breaking science on their own, but will also be essential to the scientific success of capabilities at other wavelengths and in space.

The reasons for this are several. First, most of the abundant atomic species have important transitions at UV-optical wavelengths. These include the characteristic suites of transitions that act as a ``fingerprint'' for the identification of individual atomic species (in particular, the strong resonance transitions, which appear in the optical in the spectra of redshifted objects, and important forbidden line transitions). In addition, many abundant molecular species have their rovibrational transitions in the near- to mid-IR. Thus, the observational diagnostics on which our understanding of stars and galaxies is based (chemical composition, gravity, stellar mass, age, etc.) lie in the optical and infrared. Our ability to observe these well-understood and modeled diagnostics provides the foundation for both future progress and for the interpretation of observations made at other wavelengths.

The availability of ground-based telescopes with sensitivity matched to next generation space- based and radio facilities is essential to astrophysical progress. For example, while the SKA will probe the formation of the first gaseous structures, and JWST will determine the morphologies of the first stars and galaxies, spectroscopy with a 30-m class optical-IR telescope will be needed to determine the physical properties (age, stellar content, mass, etc.) that are needed to understand the formation and evolution of these objects. A 30-m ground-based telescope is the natural spectroscopic complement to JWST in much the same way that the Keck telescopes are for the Hubble Space Telescope (see Figure 1).

Figure 1 A comparison of the signal/noise gain with a 30-m GSMT and a 6.5-m NGST for
selected spectral resolving powers. As Keck provided the sensitivity critical to spectroscopic
follow-up to HST discoveries, GSMT will play a similar role in the NGST era, owing to its power to
enable moderate- to high- resolution spectroscopy in the near-IR.

The scientific legacy of the Hubble Deep Field is a compelling example of how space-based imaging plus ground-based spectroscopy is a powerful combination for astrophysical discovery. Similarly, while space missions such as MAP will probe the primordial density fluctuation spectrum, and Constellation-X will probe the structure of hot gas in the universe, ground-based optical-infrared spectroscopy will be needed to obtain the critical complementary information on the large scale structure of galaxies. As another example, while ALMA will determine the range of initial conditions and physical processes that give rise to stars, optical-infrared spectroscopy will be needed to determine the properties of the ultimate products of star formation process: the stars themselves.

Secondly, compared with space-based missions, ground-based facilities can deploy much more complex and renewable instrumentation that is not subject to the mass and energy limitations of space-based platforms. For example, large FOVs can be used to maximize scientific gain in situations where observations of large samples of objects are critical. The instruments needed to accept large FOVs are typically prohibitively large for space-based facilities. As a result, JWST will have a relatively small FOV (5'), which largely precludes studies of, for example, the large scale structure of galaxies, the structure of the Galactic halo, and the dynamical structure and merger history of nearby clusters. As another example, high spectral resolution also typically requires larger instruments. As a result, JWST will not explore spectral resolutions R > 10,000, which consequently precludes, for example, the possibility of detailed studies of planet formation environments, the identification of merger remnants in nearby galaxies, and the chemical enrichment histories of those galaxies.


Although it is clear that the next generation ground-based OIR telescope will be central to our ability to expand the frontiers of astrophysical knowledge, an important unanswered question is what kind of OIR telescope would represent the optimal combination of scientific productivity, cost effectiveness, and technological readiness for deployment by 2015. To help address this question, we have identified several potential "discovery spaces" for a next generation telescope i.e., potential opportunities for significant scientific discovery that would be accessible if the telescope were designed to allow certain combinations of sensitivity, FOV, wavelength coverage, and spatial and spectral resolution. Sensitivity and FOV

A large aperture, ground-based telescope can provide the critical combination of sensitivity and FOV that is needed (e.g., for definitive spectroscopic studies of the evolution of large scale structure). The same combination of sensitivity and FOV would be critical to other studies, such as stellar population studies of the Milky Way halo that would address the formation and evolution of our own Galaxy. This capability would enable the next generation telescope to fully exploit those periods (perhaps 30% of the time at the very best sites) when atmospheric conditions (e.g., light cirrus) preclude full adaptive correction to carry out observations essential to linking observed fluctuations in the cosmic background to the initial appearance of large scale structure in the gaseous and stellar components of the universe.

Such a high throughput, large aperture, ground-based telescope will have the sensitivity to categorize the physical properties (e.g., star formation rates, metallicities, interstellar media, and internal dynamics) of galaxies over a range of redshifts and, potentially, of the first luminous objects in the universe from observations of their integrated spectra. Sensitivity and Angular Resolution

A large aperture, ground-based telescope, when equipped with a moderate Strehl (~ 30%) AO system over moderate FOVs (~ 1'), will have the critical combination of sensitivity and angular resolution needed to resolve and analyze the components of forming galaxies (HII regions, nascent bulges and disks), as well as individual stars in crowded stellar fields. With this capability, we will be able to trace the properties of just-formed galaxies and putative pre-galactic fragments at redshifts 3 and greater, and at intermediate redshifts, follow the merger and star formation history of nearby galaxies. Similarly, the same combination of sensitivity and angular resolution can be used to measure the masses of galaxies from the morphological (i.e., gravitationally lensed) distortions that they induce in the appearance of background galaxies (see Galaxy Formation and Resolved Stellar Populations). Sensitivity for High Resolution IR Spectroscopy

A large aperture, ground-based telescope designed to minimize thermal emissivity will have the potential for high resolution (R ~ 100,000) spectroscopy at unprecedented sensitivity at IR wavelengths. With this capability, we will be able to carry out detailed studies of planet formation environments in samples large enough to address fundamental questions such as when, where, and how frequently planets form. This capability will provide an essential complement to ALMA, whose sensitivity and angular resolution are best matched to probing cooler environments inaccessible to a next generation OIR telescope. High Dynamic Range Capability

A large aperture, ground-based telescope, when equipped with high-Strehl AO and a coronagraph, can also enable detailed studies of high contrast (> 106 between a bright central star and objects or regions located within 0.1") situations at high angular resolution. A 30-m next generation OIR telescope should have the angular resolution and sensitivity needed to, for example, directly detect the light from extra-solar planets and thereby characterize their physical characteristics (masses, radii) and atmospheres for comparison with planets in our own solar system.

In Galaxy Formation, we illustrate how these discovery spaces might translate into specific science opportunities, i.e., we illustrate the kind of science that would be possible if a next generation telescope were designed to allow those specific combinations of sensitivity, FOV, wavelength coverage, and spatial and spectral resolution. We wish to stress at the outset that the science opportunities discussed are not meant to be comprehensive. Rather, our goal is to describe in some detail the magnitude of the scientific gain that is possible for a selected subset of science in several areas if a next generation telescope were designed to enable a given discovery space.1

1A greater diversity of scientific programs that could be carried out with a next generation OIR telescope has been described in our workshop panel reports and in reports for other large telescope projects such as the Canadian XLT and the OWL.

November 2002