Plate 5

• Computers and information technology. Advances in these areas enable astronomers to analyze efficiently and develop meaningful models for the vast flood of data produced by the new instruments on OIR telescopes. They also permit greater versatility and accuracy in telescope control and the ability to assess and analyze data in real time. Thanks to high-speed telecommunications networks, it is now becoming possible for astronomers to operate telescopes located thousands of miles away with computer terminals in their home offices.

• Adaptive optics. Technology to correct for rapidly changing image distortion due to atmospheric turbulence, pioneered by the Department of Defense, is now becoming available to astronomers. A factor-of-two reduction in the full width at half maximum (FWHM) of the resolution of the image implies a factor-of-four increase in peak flux and a factor-of-eight improvement in accuracy of moment analysis, image distortion analysis, and morphological classification. Already, the Canada-France-Hawaii Telescope (CFHT) has demonstrated the ability of an adaptive tip-tilt system to reduce the seeing FWHM from 0.8'' to 0.4''. More sophisticated technologies to correct wavefront distortion more completely are currently under development (Plate 6). They have the potential of providing image quality that can now be obtained only by far more costly telescopes in space.

Plate 6

• Interferometry. The technology to combine in phase the light from separated telescopes, a standard technique for radio astronomy, opens the possibility of observing sources with angular resolution hundreds or thousands of times sharper than currently feasible from telescopes on the ground or in space. As discussed in the AASC report, such technology would enable astronomers to address exciting problems currently beyond reach. Great challenges remain to bring the technology to fruition.

The AASC report identified the outstanding scientific opportunities in astronomy and astrophysics for the 1990s and laid out a prioritized strategy for realizing those opportunities. The AASC strategy for OIR astronomy is part of a larger strategy for research in astronomy and astrophysics that includes facilities on the ground and in space. The current revolution in our understanding of the cosmos comes largely from our new-found ability to observe the sky at every wavelength of the electromagnetic spectrum, ranging from radio to gamma rays. In this strategy, OIR astronomy plays a central role. Almost every new astronomical source, whether discovered by radio telescopes on the ground or by infrared, ultraviolet, X-ray, or gamma-ray telescopes in space, must be observed by ground-based OIR telescopes to understand its physical nature and significance.

Conversely, observations with ground-based OIR telescopes are essential for the efficient use of far more costly telescopes in space. For example, the Hubble Space Telescope (HST) has a very narrow field of view and can observe only a tiny fraction of the sky. We can realize the full benefits of the HST's superior image quality and unique ultraviolet spectroscopic capability only if we identify its targets on the basis of extensive studies with ground-based OIR telescopes. Moreover, the HST will image distant sources so faint that their spectra can be measured only by ground-based OIR telescopes of far greater aperture. The same considerations apply to other NASA programs under development, such as the NICMOS infrared instrument on HST, the Space Infrared Telescope Facility (SIRTF) and the Stratospheric Observatory for Infrared Astronomy (SOFIA) infrared telescopes, and the Advanced X-ray Astrophysics Facility (AXAF) X-ray telescope. Even ignoring the scientific discoveries enabled by OIR telescopes alone, NSF's $40 M annual expenditure to support ground-based OIR astronomy can be justified easily on the basis of the enhanced scientific yield from NASA's $800 M annual funding of space astrophysics.

The AASC report pointed out that major opportunities to address fundamental cosmic questions will be enabled by new technologies and instrumentation for ground-based OIR telescopes. For example:

• How do stars form? Telescopes equipped with modern infrared instruments will be able to observe newly forming stars that are enshrouded in dust clouds from which optical light cannot emerge. The images will reveal the morphology of the disks and jets around these stars, and the spectra will tell us about the gas temperatures, velocities, and magnetism that control the star formation dynamics.

• What is the origin of the heavy elements in the universe? Astronomers believe that the heavy elements are formed as a result of nuclear reactions in stars, particularly in their final convulsions as novae and supernovae. Surveys with 2- to 4-meter telescopes will find many more of these events, and large telescopes will obtain detailed spectra, particularly at infrared wavelengths where newly formed elements are most apparent, to confirm and enrich this theory. With powerful new spectrometers, astronomers will be able to understand better how the products of supernova nucleosynthesis are dispersed and built up in stars, galaxies, and interstellar and intergalactic gas.

• How many stars have planetary systems? With infrared telescopes, astronomers will be able to detect and image disks of dust particles around stars from which planetary systems are believed to form.

• How do galaxies form and evolve? With large optical and infrared telescopes, astronomers will be able to find newly forming galaxies at high redshifts and learn about their dominant physical processes.

• What powers the central engines of active galaxies and quasars? Are they supermassive black holes? Do many other galaxies, including the Milky Way, also contain quiescent black holes? If so, what are the environmental conditions that determine the rich variety of phenomena associated with quasars and galactic nuclei? To answer these questions, astronomers need to observe many galactic nuclei with OIR telescopes having high angular resolution, broad spectral range, and polarimetric capability. The coordination of such observations with observations with radio, ultraviolet, X-ray, and gamma-ray telescopes is also necessary.

• How did the matter in the universe coalesce into clusters and superclusters of galaxies, separated by huge voids? With new-technology 2- to 8-meter wide-field telescopes instrumented to measure spectra of hundreds of galaxies at a time, astronomers will be able to map the distribution and velocities of many thousands of galaxies at moderate and high redshifts, and to understand the forces and motions caused by the unseen "dark matter" that appears to dominate the mass of the universe. With infrared telescopes, they will be able to search deeply for faint red stars that may contribute to the dark matter in the halos of galaxies.

The AASC report recognized that the most dramatic advances in these and other areas would probably come from observations in the infrared band, where many of these phenomena are most easily observed. Great advances in our ability to obtain infrared images and spectra are now being achieved with new large-scale array detectors with high quantum efficiency and very low noise and dark current. Moreover, the opportunity to obtain much sharper images from ground-based telescopes will be realized first at infrared bands, for which the effects of atmospheric distortion are most easily compensated. For these reasons, as well as the scientific promise of proposed NASA observatories such as SOFIA and SIRTF, the AASC report called the 1990s "the decade of the infrared."

Dramatic confirmation of the prescience of that remark comes from the impact of comet Shoemaker-Levy 9 with Jupiter. The effects are most clearly evident in infrared images taken with telescopes equipped with adaptive optics correctors and wide-format infrared array detectors that were not available five years ago. Even as this report is written, these images are appearing on the front pages of the world's newspapers, magazine covers, and television news broadcasts. This remarkable event, the likes of which may not recur for millennia, will tell us much about the nature of comets, the atmosphere of Jupiter, and the mechanisms for mass extinctions that occur on Earth on time scales of tens of millions of years.

The sky contains literally billions of sources visible to OIR telescopes, representing an amazing variety of phenomena. A partial list includes:

• Planets, moons, comets, and asteroids;
• Violent magnetic storms on nearby stars;
• Giant stars that are blowing their outer layers into interstellar space;
• Violent stellar explosions in novae and supernovae;
• Interacting binaries containing the collapsed remnants of dead stars;
• Vast clouds of magnetized interstellar gas violently disturbed by stellar outflows and explosions;
• Newly forming stars surrounded by whirling disks and shooting out jets of gas;
• Galaxies with a vast variety of sizes, shapes, content, and dynamical behavior, which are observed to evolve as we look further back in distance and time;
• Active galaxies and quasars containing compact sources of enormous power at their centers;
• Clouds of diffuse gas between the galaxies observable by their absorption of ultraviolet radiation from distant quasars; and
• An expanding universe in which the galaxies are distributed on filaments separating great voids and move under the influence of a far greater mass of invisible matter.

Even using all the telescopes available, only a tiny fraction of these sources can be observed in a human lifetime. A strategy to optimize progress in understanding such a sky will not be highly focused--it will require great diversity of facilities, observing strategies, and ideas.

The commissioning of the two powerful Gemini telescopes in 2000 and 2003 will open new opportunities for research by the U.S. astronomical community. The 8-meter Gemini North telescope on Mauna Kea was the AASC report's highest-priority recommendation for a ground-based facility. It will be optimized for diffraction-limited operation at infrared wavelengths and will be a unique facility using revolutionary infrared array detectors to make the high-spatial- and high-time-resolution observations needed to study phenomena ranging from protoplanetary disks around young nearby stars to the most distant galaxies in the early universe. The 8-meter Gemini South telescope, located in Chile (see back cover), will provide U.S. astronomers with a vital window to the Magellanic clouds, the center of the Milky Way, and other southern sky objects.

The estimated annual cost for the IGP to operate the two Gemini telescopes will be about $11 M, of which the United States is obliged to provide half. In addition, NOAO estimates an annual cost of $2.5 M to support access by the U.S. community to the Gemini telescopes, including partial support for continued instrument development, observer support, and analysis and archiving of data. The net cost, $8 M, is well within the normal guidelines for the operation of any major astronomical facility, which is about 10% per year of the construction costs.

Modern OIR astronomy involves a mix of telescope sizes and types. The largest and most expensive telescopes, such as the 8-meter Gemini telescopes (see Plate 1) and the 10- meter Keck telescopes (see Plate 2), will have unique power to record images and spectra of the faintest and most distant sources in the sky. But it would be extremely wasteful to use these great telescopes to observe systems that can be observed equally well, and often far more efficiently, by smaller telescopes. For example, these great telescopes have relatively narrow fields of view, whereas modern 2- to 4- meter-class telescopes can observe a far greater number of sources at once because they have larger fields of view. Thus, a strategy for efficient use of the large telescopes requires smaller telescopes to select the most promising targets from the myriad of sources. Moreover, there are many projects of great scientific merit, such as redshift surveys and mapping of extended sources, that can be done more efficiently with smaller telescopes.

In addition to large and moderate general-purpose telescopes, an efficient infrastructure for OIR astronomy will include telescopes designed for special purposes. Some important programs can be accomplished at great savings in telescope construction and operation by sacrificing versatility. For example, the Hobby-Eberly Telescope (HET; see Table 2) has a 10- meter fixed spherical primary mirror and a movable secondary mirror--the optical equivalent of the Arecibo radio telescope. By sacrificing pointing and steering capability, the HET can measure spectra of faint objects at a small fraction of the cost of doing so with a general-purpose telescope of comparable effective aperture (6 to 8 meters). Other very important projects, such as the Two Micron All Sky Survey (2MASS) of millions of infrared sources and the Sloan Digital Sky Survey (SDSS) to measure the colors and spectra of millions of galaxies and quasars, can be carried out only with dedicated special-purpose telescopes.

An efficient strategy for OIR astronomy will also accommodate a diversity of observing modes. Programs to develop new instrument technology will require substantial amounts of dedicated telescope time. Some observations, which push the performance limits of telescopes and instruments, can be carried out successfully only by astronomers intimately familiar with the facilities. Uniform surveys of large numbers of sources may require tens or hundreds of nights of telescope time but can be carried out according to an established routine. Some such programs may now be accomplished most efficiently by remote observing. At the other extreme, a new discovery at radio or X-ray wavelengths may require a snapshot taking only a few minutes of telescope time or, as is frequently the case, it may lead to an extensive campaign for coordinated ground- and space-based OIR observations. Much important science can be achieved most efficiently by creating a large uniform data set and analyzing the results later, as was the case with the Infrared Astronomy Satellite and will likely be so for the SDSS. These various observing modes will complement, but not replace, the traditional observing run of a few nights, which will still be needed for experienced astronomers to carry out many kinds of programs and to provide hands-on training of new astronomers.

Some observations might be done best if scheduled in a queue and executed by staff astronomers instead of the investigator, much as most observations with space observatories are carried out. Queue scheduling can be efficient because it permits (1) observations that require rare conditions such as exceptional seeing; (2) greater efficiency in executing short observations; (3) greater flexibility in ensuring that observations of highest scientific priority are executed; (4) ease of scheduling time-critical observations such as targets of opportunity and synoptic studies; and (5) optimal scheduling of observations to ensure observations at minimum air mass and correct lunar phase.

Most of the major OIR telescopes in the United States are located at independent observatories, owned and managed by state and private institutions (see Section II). This situation, in which the majority of the capital assets were provided by private and state sources, is a unique and enormous asset to U.S. physical science. Because these independent observatories operate more than two-thirds of the major U.S. telescopes and are used primarily by about half of the OIR astronomers in the United States (Section II), they can support scientific programs of great merit that are beyond the resources of the NOAO. In particular, the independent observatories can devote greater fractions of their telescope time to testing of innovative instrumentation and to extensive observing projects requiring tens or hundreds of nights of telescope time. The independent observatories also make a major contribution to the research of astronomers not affiliated with their own institutions, through informal collaborations, guest observer programs, and the data that they disseminate to the community.

NOAO adds a vital dimension to OIR astronomy (and solar astronomy) in the United States. Since KPNO began operations in 1960, NOAO has provided world-class telescopes, particularly the 4-meter telescopes at KPNO and CTIO. The CTIO has been especially important to U.S. astronomers because its facilities have provided vital access to the southern sky (the only other major U.S.-owned Southern Hemisphere telescope is the 2.5-meter Dupont telescope of the Las Campanas Observatory). NOAO enables many astronomers at universities without major telescopes to carry out frontier research on the basis of open peer-reviewed competition. NOAO also provides crucial observing options not otherwise available to astronomers at independent observatories. Likewise, NOAO provides vital access to OIR telescopes for radio and space astronomers.

The NOAO includes the National Solar Observatories (NSO), which provide the U.S. solar physics community with access to observing capabilities not available elsewhere in the United States. These include the infrared capabilities of NSO facilities on Kitt Peak and the high-angular-resolution facilities in the optical at Sacramento Peak.

Recently, NOAO has entered into a number of successful partnerships with university instrument groups and independent observatories, such as the deployment at CTIO of the OSIRIS infrared spectrometer that was developed by Ohio State University, and the construction and joint operation at KPNO of the WIYN telescope, a partnership of the University of Wisconsin, Indiana University, Yale University, and NOAO. NOAO has exerted leadership in some areas of instrumentation development. Outstanding recent examples are the Hydra multifiber spectrograph (see Plate 5) and the deployment of large-format optical and infrared detector arrays. NOAO has acted as a national resource for instrumentalists by providing advice and technical information freely. NOAO has also developed and supported standards for data archiving and analysis, including the IRAF data-reduction software that is used by astronomers worldwide.

OIR astronomy in the United States gains strength not only from the infrastructure of the independent observatories and the NOAO, but also from a growing variety of international collaborations. The State of Hawaii has the good luck to have, on Mauna Kea, the best site in the world for many kinds of OIR astronomy (see front cover); as a result, the University of Hawaii is a partner in the operations of several international telescopes, notably the 3.6-meter CFHT, the 3.8-meter United Kingdom Infrared Telescope, and the 8-meter Japanese Subaru Telescope, currently under construction. International cooperation will become a much greater part of the U.S. OIR astronomy infrastructure with the completion of the two 8- meter telescopes of the international Gemini project, a collaboration between the United States (50% share), the United Kingdom (25%), Canada (15%), Chile (5%), Brazil (2.5%), and Argentina (2.5%). In addition, a number of independent observatories have undertaken to build major OIR telescopes in partnership with other countries, notably the Large Binocular Telescope (LBT), the HET, and the SDSS (see "The Independent Observatories" in Section II).

Given the diversity of scientific challenges for OIR astronomy, it is no easy task to suggest mechanisms to optimize the productivity of the complex infrastructure that is required to meet them. Indeed, this panel cannot dictate how this infrastructure will develop or foresee the scientific and technical problems and opportunities that will arise. The best that it can do is to identify some general principles to increase the scientific yield of the enterprise, and suggest mechanisms for making ongoing decisions that are likely to lead to a more optimum infrastructure. The principles are as follows:

• It is wasteful to maintain a full complement of instruments on every telescope. Losses accrue from leaving valuable instruments on the shelf most of the time and from the necessity to change instruments. Significant savings can be realized by supporting fewer instruments on each telescope.

• If telescopes become more specialized, the diversity of observing options required for OIR astronomy can be maintained by arrangements facilitating access by astronomers to a variety of specialized telescopes at independent and national observatories. Various successful examples already exist of such arrangements, which are often informal. They include bartering of telescope time, exchange of telescope time for instruments, and service observing. Rapidly developing technology for remote observing will make it easier to provide such access.

• A broad distribution function of the length of observing runs will probably result in the greatest scientific yield. Long-term projects, by experienced observers with one or two instruments, can be of great scientific merit and can be carried out at the lowest cost per night. Many significant observations, particularly those on the largest telescopes, will require less than a night and might be accomplished most efficiently by queue scheduling and remote or service observing.

• Cooperation at every level should be encouraged. Already evident are excellent examples of cooperation between NOAO and various universities in building and operating telescopes (e.g., the WIYN telescope), in the deployment of instruments (e.g., the Ohio State University OSIRIS infrared spectrometer and the Rutgers University Fabry-Perot camera at CTIO), in the development of optical and infrared detector arrays, and in software development for instrument and telescope control as well as data analysis. Ongoing efforts to establish and maintain standards for user-telescope-instrument interfaces will encourage and facilitate such cooperation.

• Excellent opportunities will likely arise for international cooperation beyond the various agreements already mentioned to build new telescopes. For example, the Anglo-Australian Telescope and the CTIO are already discussing arrangements to barter telescope time. In the future, the Keck, Gemini, and European Southern Observatories may find that barter arrangements may reduce the need to build similar instruments at each observatory.

• Mechanisms for such cooperation will be most effective if the terms can be arranged by the working scientists and can evolve with changing circumstances.