The First Workshop on the O/IR Ground-based System

October 27-October 28, 2000

Phoenix, Arizona

What is the O/IR system and why are we talking about its evolution?

The report of the O/IR Panel of the Decadal Survey has yet to be released due to delays in the lengthly editing process --- this is the first time Panel reports have also been refereed. As a result, details about the O/IR Panel's principal recommendation, that the O/IR ground-based facilities, both public and private, be considered as one system, have been limited to the terse description in the Executive Summary. We are concerned that this has led to a substantial misconception of the tone and intent of the O/IR Panel report on this crucial issue.

Previous Decadal Surveys took the position that, with regard to O/IR ground-based facilities, only national facilities such as NOAO or Gemini were subject to recommendations of the Committee. This meant that the majority of U.S. facilities were de facto excluded from the discussion. The O/IR panel in this Decadal Survey believed that this was a serious roadblock in our attempts to fully exploit the resources that U.S. astronomers have at their disposal, especially since the formulation of NSF policies with regard to these independent facilities received no guidance from the critical decadal review.

As described in the excerpts below from the Panel Report, we think that the suite of ground-based O/IR telescopes, without any master plan, already functions as an effective system. The diversity expressed in this often incoherent system is one of its principal strengths, and the O/IR Panel is not advocating a top-down structure that would remove important decisions from individual observatories and universities to a central planning body. Rather, the idea was to recognize this suite of facilities to be a virtual system and, as elements of its positive evolution, encourage the twin goals of greater access for the entire community and greater NSF support for all components, particularly in the area of instrumentation. Broad access to the new generation of large-aperture telescopes was considered a scientific priority as well as a key step towards securing new NSF funding for instrumenting these telescopes. All the initiatives proposed by the Panel, GSMT, LSST, and TSIP, were seen as tools with which to connect elements of the system, providing powerful new capabilities via consideration of public and private resources together. The Panel believed that, simply by appealing to self-interest, the various components of this system could be encouraged to work together towards improving all the facilities, and that the NSF could develop policies that recognize and encourage this complex but powerful combination which characterizes O/IR ground-based astronomy in the United States.


Excerpts from the O/IR Panel Report


Exploiting the Diverse, Unique Facilities of U.S. Ground-based O/IR Astronomy

Astronomers now probe the physics of exotic and extreme environments using observations across the electromagnetic spectrum, ranging from radio waves to gamma rays and everything in between. Ground-based O/IR research remains at the heart of this endeavor. Ground-based spectroscopic observations have been crucial to research with the HST for such diverse topics as high- redshift galaxies, the extragalactic distance scale and cosmological parameters, and stellar populations in the Milky Way and other nearby galaxies. The study of radio galaxies relies on a combination of radio, optical, and infrared data, as does the goal of linking the evolution of the interstellar medium to stellar evolution. The x-ray halos of galaxies and the hot gas in rich clusters are being understood thanks to both space-based observations and ground-based optical data, and the recent discovery that gamma-ray bursts take place in distant galaxies, making them the most energetic known phenomenon, is the result of hard-won spectroscopic data from the Keck telescopes and rapid-response imaging from a variety of ground-based instruments.

The ambitious scientific program outlined above requires a broad suite of telescopes with a range of aperture sizes and powerful state-of-the-art instrumentation employing the latest array detectors. Ground-based O/IR facilities available to U.S. astronomers include both national and independent installations in a unique combination that has kept U.S. astronomy strong. Remarkably, most of the "glass" resides at independent observatories, but the National Science Foundation's (NSF's) role is nevertheless vital to the health of the entire enterprise: NSF not only supports NOAO and Gemini but also plays a crucial role at the independent observatories by virtue of its grant support for research and instrumentation.

Strong competition for leadership in astronomy is now coming from the European Southern Observatory (ESO), which has made huge investments at two Chilean observatories. The most recent of the investments was for the Very Large Telescope (VLT), four 8-m telescopes at Cerro Paranal costing approximately $700 million. Eight first-generation instruments are currently under construction at a cost of approximately $90 million. An ongoing instrumentation program funded at $10 million per year is anticipated to provide state-of-the art facilities throughout the decade. A world-class interferometer is also under construction for the VLT. Early results from the VLT are most impressive. In addition, ESO has announced its intention to build a truly enormous telescope, currently planned to have a 100-m aperture, a billion-dollar-plus initiative that is unprecedented in the history of ground- based astronomy.

To meet this challenge, the United States must use its unique combination of federal, state, and private resources to best advantage. The spirit of independence and individual initiative that has characterized U.S. ground- based O/IR astronomy should continue, for it has had highly productive and creative results. But federal resources are sufficiently scarce, and the need at both national and independent observatories is so acute, that a greater degree of cooperation is urgently needed. Put simply, the suite of U.S. observatories should function as a coherent system to ensure that U.S. astronomers will have the means to participate fully in pursuing the fundamental goals outlined above. Although the recommended new facilities and programs described below are completely justified by scientific arguments alone, the panel sees as equally important the additional overarching goal of strengthening the system and creating a process for its development. To reach this goal, NSF and other agencies funding O/IR astronomy should appreciate all the elements of the U.S. O/IR system and implement policies that guide the system's development so that federal funding may achieve maximum science and maximum opportunity.


Moderate Initiative, Priority 1: Telescope System Instrumentation Program: Leveraging Nonfederal Investment and Increasing Public Access


The U.S. ground-based astronomy inventory consists of nine 6.5- to 10-m-class telescopes (operating or under construction), nine 2.5- to 5-m telescopes, and numerous smaller instruments. Considered together, this collection of both national and independent observatory facilities represents the world's most powerful ground-based telescope arsenal, with unequaled opportunities for maintaining leadership in research. However, these facilities have traditionally not worked together as a coherent system, in contrast with astronomy facilities abroad, which are dominated by national or international observatories. The panel believes that better coordination and cooperation are essential to realizing the full potential of this system and that NSF should work to achieve such coordination and to ensure that facilities and data are made widely available to the entire astronomy community.



The panel proposes a new program, the Telescope System Instrumentation Program (TSIP), modeled on the McCray report's Facilities Instrumentation Program of support for instrumentation and instrumentalists at the independent observatories. The TSIP would guide the evolution of the telescope system so that it becomes more powerful and more diverse, for example, by favoring instruments with unique capabilities and those that would be particularly effective in reaching the scientific goals described in this report. The panel supports the twin goals of achieving greater public access to these facilities and encouraging and leveraging the contribution of institutions that contribute nonfederal funds to the U.S. astronomy enterprise, to be accomplished by an equal weighting of the funding for both goals.


Science Drivers for 8-m Telescopes with Advanced Instrumentation

With the new generation of 8-m-class, ground-based O/IR telescopes and state-of-the-art instrumentation, many of astronomy's primary science goals for the next two decades can be achieved. Some examples follow:


  • Assemble very large samples of galaxy photometry and spectroscopy over the redshift range 0 < z < 2 to measure the chemical and, to the extent possible, structural evolution of galaxies in the context of the growth of large- scale structure. This effort will require wide-field imaging and highly multiplexed moderate-resolution (R ~ 5000) spectroscopy. Instrumentation required: moderate- resolution, low-background spectrographs with multimillion-pixel, low-noise, near-IR array detectors.
  • Identify, through low-resolution spectroscopy, distant supernovae found through large-area surveys. Use type Ia supernovae in the redshift range 0 < z < 1.5 to measure cosmological parameters, particularly to confirm recent measurements of a nonzero cosmological constant. Compile observations of all supernovae to study the history of star formation rate over a range of galaxy types and luminosities. Instrumentation required: high-sensitivity optical and near-IR, low- to moderate-resolution spectrographs fed by AO systems.
  • Probe the stellar content of the Galactic halo to characterize its assembly and chemical enrichment history. Find thousands of blue horizontal branch stars and RR Lyraes in the Galactic halo and obtain medium-resolution spectroscopy to map the dark halo to 200 kpc. Search for extremely metal-poor stars and other subpopulations and measure their kinematics (three-dimensional motions) to reveal how the halo was assembled through early agglomeration and later accretion. Study with high- dispersion spectroscopy the r- and s-process element distributions in the most metal-poor stars to probe early chemical enrichment of the halo. Extend these studies and the studies of Galactic globular clusters to Local Group members and, to the extent possible, to more distant Local Supercluster galaxies. Instrumentation required: wide- field, multicolor imaging to identify candidates; follow-up with moderate- to high-resolution spectrographs.
  • Perform spectroscopic follow-up of samples of brown dwarfs in the Galaxy to study the frequency of binary systems and the evolution of their atmospheres; find evidence of chromospheres, flares, winds, and x-ray coronae. Instrumentation required: moderate-resolution, low-background, near-IR spectrographs.
  • Study the infalling envelopes, accretion disks, and post-accretion disks around young stars. Use adaptive optics and ground-based interferometry in the IR to measure the structure and temperature distribution in the 10 to 100 AU region, and use high-resolution spectroscopy to probe the 1 to 10 AU region via observations of velocity- resolved molecular transitions such as those of CO and H2O. Characterize the outflow regions of young stars and understand the transport of angular momentum and the evolution of magnetic fields. Look for structure in post- accretion disks indicative of planet formation and use coronagraphic imaging to search for high-mass planets and brown dwarf companions around neighboring stars. Instrumentation required: IR interferometers; AO-fed, near- to mid-IR, moderate- to high-resolution spectrographs; coronographs.
  • Use adaptive optics imaging to monitor weather on Mars and the Jovian planets, climatic variations on Titan, volcanic eruptions on Io, and make the first high-resolution maps of surface features of Mercury. Adaptive optics with spectroscopy will provide spatially resolved spectra of the atmospheres of Jovian planets. Measure the binary frequency of KBOs and obtain spectra of brighter objects to study composition. Instrumentation required: optical to mid-IR AO imaging; AO-fed, moderate-resolution, near- to mid-IR spectrographs.
  • Provide rapid-imaging follow-up of gamma-ray bursts to identify optical counterparts, and use sensitive spectroscopy to obtain host galaxy redshifts. Instrumentation required: instant-access O/IR cameras and low-resolution spectrographs.
  • Use AO and IR interferometry to map the structure of AGNs at very small scales in order to study the kinematics, temperature, and density structure of material close to the black hole/accretion disk. Instrumentation required: AO-fed, O/IR spectrographs; IR interferometers.


The Telescope System Instrumentation Program

Because public and private resources have a history of uncoordinated development, U.S. capabilities in ground- based O/IR astronomy represent a strong but--at best-- loosely organized approach to astronomical research, with agendas set by a wide range of institutions pursuing a variety of goals. Although this diversity is one of the strengths of U.S. astronomy, it is imperative that the new generation of 8-m telescopes be used as a total system in order for the nation to compete effectively. It has also become clear with the building of the 8-m telescopes that the resources necessary to instrument these telescopes properly, as well as to process, analyze, and distribute the data, are woefully inadequate. Traditionally, NSF grants have provided critical support for the processes of data gathering and reduction, as well as for supporting theory and laboratory astrophysics programs essential for progress in O/IR research. However, the panel believes that NSF's already important role can become even more important in the coming decade as it enables national and independent observatories to work together, as a system, to accomplish the scientific goals described in this report. Through a process of peer review, NSF can use its grants programs to focus limited federal resources in a way that will maximize the scientific return on these huge investments, by supporting the development of instrumentation that provides special, as yet unavailable observing opportunities.

A key component of U.S. leadership has been the excellence of U.S. instrument builders, who have continually provided innovative, powerful, and cost- effective instrumentation for ground-based telescopes. Of particular concern is the training and support of these future instrumentalists. The building of high-quality instrumentation, particularly in university environments, where students can be trained, is an essential component of a vigorous, diverse, ground-based telescope system. This instrumentation includes both traditional smaller (principal- investigator-scale) instruments as well as new, state-of-the- art facility instruments for 8-m-class telescopes. Compared with the former, the latter present special challenges: they are more expensive and more difficult to build, and they require larger groups management structures, as well as longer production times. These challenges, and a disturbing lack of recognition in some university environments for this essential contribution to the scientific process, mean that many instrument builders see their opportunities for research hard pressed. Because instrument building at the universities and observatories in the United States is crucial for the nation's continued success, the panel believes that it is vital for NSF to focus its efforts on ensuring a healthy mix of smaller-scale instrumentation programs as well as large-scale facility instruments.

The facilities instruments for the new generation of 8-m telescopes are far more capable than their predecessors, but because of their scale and complexity, they are an order of magnitude more expensive, typically $5 million to $10 million apiece (the VLT average is $11 million). Furthermore, data storage, analysis, and dissemination costs will be substantial. Fortunately, the investment required is incremental to the funds already expended at both private and public observatories for construction and continued operation of the new facilities. The panel proposes a new investment beginning at $5 million per year in instrumentation for the independent observatories, concentrated on the new 8-m-class telescopes, which would leverage a scientific yield several times over. Although it believes that the program originally proposed in the NRC's McCray report, recently reviewed and endorsed by the Committee on Astronomy and Astrophysics, provides the framework for the administration of this program, the panel emphasizes elements of the McCray report that have not been applied so far.

The McCray report recognized that maximizing the quality and quantity of astronomical research in the United States depends on a vigorous investment from NSF; the report also acknowledged NSF's long-standing commitment to provide wide access to astronomical facilities, so that studies outlined in excellent research proposals identified through peer review could be carried out at premier facilities, public and private.

It is important to recognize that private facilities now support a large fraction of the U.S. astronomy community (~50 percent according to the recent NRC report Federal Funding of Astronomical Research, National Academy Press, 2000), a situation very different from that existing at the founding of the Kitt Peak National Observatory in the 1950s. Although there are now many additional new opportunities for astronomical research, including space-based facilities, radio observatories, and data archives, it is abundantly clear that some measure of public access is vital to the health of U.S. astronomy.

The panel reaffirms the critical importance of maximizing scientific return and ensuring greater public access, and it emphasizes the crucial importance of regarding all of the U.S. O/IR facilities as a system. The NSF should administer the TSIP so as to achieve all of these objectives. It is vitally important to encourage independent observatories to participate, an approach that depends on recognizing their important contribution of nonfederal funds for astronomy. To encourage their participation, the TSIP should be broader than the program first implemented, when the NSF's goal of acquiring telescope time on private facilities dominated the process. Borrowing from the McCray report's recommendations, the panel strongly advocates the following guidelines:


  • The TSIP should apply to facility instruments for independent observatories only, for which NSF grants at least $1 million in support of the proposed instrument. It should not replace NSF's existing Advanced Technologies and Instrumentation (ATI) or Major Research Instrumentation (MRI) programs.
  • Successful proposals that include an offer of observing time will provide nights on the telescope whose value (based on amortized investment and operations) amounts to 50 percent of the granted funds. This 50/50 split properly recognizes the initiative of the independent observatory researchers in bringing nonfederal funds to astronomical research and supports their science, while still attending to the important goal of providing observing time for the best peer-reviewed proposals, regardless of institutional affiliation. (The panel also notes a finding from Federal Funding of Astronomical Research that 50 percent of the users of ground-based O/IR facilities have access to independent observatories.) The 50/50 split should not be negotiable: to negotiate it would undermine the cooperative spirit needed to ensure the success of the overall O/IR system. The proposing institution may specify additional guidelines, for example, whether the time is available only on the proposed instrumentation or on all instrumentation, and it may include requests that specify operating modes, for example, minimum observing run duration. Such conditions are to be evaluated along with other aspects of the proposal.
  • In lieu of some or all of the telescope time, proposals may be accepted that offer other comparable benefits to the astronomy community, for example, the production and dissemination of surveys and the archiving of data from this or other instruments on the telescope for which the instrument is proposed.

The panel considers the 50/50 split to be an essential part of its recommended guidelines. It represents a fair division in which both the independent observatory and public access communities benefit. It is clear that a "dollar of telescope time for a dollar of instrumentation funding" does not recognize or encourage the contribution of universities and private institutions in raising funds and does not recompense them for the talent and time of their scientists and engineers in building facilities. On the other hand, an unrestricted NSF grant of funds with no benefit provided to the broader astronomy community would frustrate the aspirations of the scientists who would like to use the unique facilities outside the national observatories. Something for both groups is the only appropriate solution; the 50/50 split has the added benefit of conveying the traditional notion of fairness. Negotiating the split, as was tried previously, promotes competition that can only undermine the goal of cooperation within the system.

Effective development of the entire suite of community facilities as a system depends on a common perception of how the parts of that system interact; a common vision of their strengths, deficiencies, and potential evolution; and an implementation plan with some level of feedback and accountability. The panel looks to NOAO for leadership in establishing a common vision by involving all segments of the ground-based O/IR community in discussions aimed at evaluating elements of the system and devising plans for the future. Based on the results of these discussions, NOAO should develop a strategic plan for the ground-based O/IR system that includes an analysis of the benefits, costs, and risks for various prioritized implementation alternatives. Such an analysis would help the NSF decide how to invest its resources, either proactively, through solicitations for particular capabilities or negotiations for telescope time, or through TSIP, in response to instrument proposals.

Given the wide variety of instrument capabilities and performance, scientific potential, public benefit, terms of use of telescope time, and importance to the system of O/IR facilities, the choice of successful proposals should be made annually by an NSF-constituted peer review committee. Any recommendations from the NOAO-led strategic planning effort should be provided to this NSF committee, but the peer review process should allow a full consideration of all factors. Every year, the NOAO-led, community-based strategic planning group should provide structured feedback to the NSF regarding the perceived efficacy of its investments in meeting the strategic goals of the community.

The panel also considered the effectiveness of direct purchase of telescope time by the NSF and concluded that in some circumstances, this could be the most efficient way for the entire community to gain access to a unique capability. However, the panel prefers the TSIP approach because it--as opposed to the time purchase option--ensures the production of critically needed instrumentation and provides for community involvement (through peer review) in the types and capabilities of instruments that will be built and made available. For this reason, the panel advises that the direct purchase of telescope time should be a second option, used sparingly so as not to significantly decrease the resources needed for TSIP, which would be sized according to the instrumentation needs of the independent observatories. It also urges that decisions by the NSF to buy telescope time should be guided by an understanding of the broad needs of the community, as is intended in TSIP.

The goal of building a more cohesive ground-based O/IR community--with increased incentives for private fund- raising, continued commitment to public access to premier facilities, and maximized scientific creativity and output-- should guide NSF policy throughout the decade. Without these investments and increased funding for research support, both provided by the NSF grants program, U.S. astronomers will not be able to continue in leadership roles in the most basic astronomical research and will not be able to take full advantage of the new, powerful space telescopes such as Chandra, SIRTF, and NGST.

Technology Issues

The NSF could enhance the ground-based O/IR system through continued investment in the development of technologies that will ultimately enable new capabilities. Development of detectors, especially large-format, near- and mid-IR arrays, is a key area. AO systems to feed near- IR spectrographs are extremely important to reduce background noise and to work in crowded regions of the sky. Large-scale surveys require new, more efficient data- handling techniques. IR interferometry will open a new discovery space for the study of high-surface-brightness objects and may provide competition for the evolution of ground- and space-based filled-aperture telescopes.

Cost Issues

The investment for the first complement of VLT instruments is $91 million, with an expected continuing investment of at least $10 million per year. These figures reflect a realistic assessment of the ongoing investment needed to take full advantage of the half-billion-dollar investment in new telescopes (excluding the investment for infrastructure at Paranal). The United States has a comparable capital investment and an even greater number of telescopes, so a continuing U.S. investment for major instrumentation of at least this magnitude is required. NSF has separate commitments for instrumentation at NOAO and Gemini, and some nonfederal support for instrumentation is available, leading the panel to conclude that an additional $5 million per year for independent observatories is critically needed.

Context Issues

The importance of ground-based facilities for space-based telescopes and ground-based radio astronomy is clear. Most astronomy projects, regardless of wavelength domain, have an O/IR component that is effectively addressed with ground-based O/IR facilities. The data collected should be archived and made available to the astronomy community, following the lead of the National Virtual Observatory (see the report of the Panel on Theory, Computation, and Data Exploration, Chapter 6). Support for instrumentalists is a particularly vital part of this program; the NSF can and should initiate programs that encourage and support instrument builders throughout the community. The greater participation of theoretical astrophysicists in the planning of large programs also needs encouragement and support.