Below is the uncorrected machine-read text of this chapter, intended to provide our own search engines and external engines with highly rich, chapter-representative searchable text of each book. Because it is UNCORRECTED material, please consider the following text as a useful but insufficient proxy for the authoritative book pages.
Data Processing and Computational Facilities I. INTRODUCTION The Astronomy Survey Committee charged the Panel on Data Processing and Computational Facilities with an assess- ment of the computational requirements of the astronomical community in three areas: theoretical calculations, observational data processing and analysis (especially image processing), and data archiving and access. The Panel was additionally asked to consider, within each of these areas, the astronomical and astrophysical problems to be addressed; the computational resources (hardware, software, documentation, personnel) required for the solu- tion of these problems; the present and future availabil- ity of computational resources; choices among local, regional, and national distribution of computational resources; possible leadership roles for the National Astronomy Centers; the existence or establishment of coordinating or advisory groups to guide the orderly and effective operation and growth of facilities; and, above all, the means to ensure a high level of input from, and responsiveness to, the astronomical community. The Panel members heartily support the recognition of the importance of computational facilities to astronomy represented by the establishment of this Panel. Although computers first began to be used by astronomers in the 1950's, and although the Greenstein committee (see Astronomy and Astrophysics for the 1970's, National 302
303 Academy of Sciences, Washington, D.C., 1972) established a panel to investigate astronomical computing, the present document is the first formal report dealing with computa- tional facilities to be included in the deliberations of an Astronomy Survey Committee. Indeed, computers are now so widely entrenched in astronomy that one can scarcely conceive of returning to the manner in which astronomy was performed just 10 years ago, much less before the 1950's. Future technological developments are sure to make the role of computers in astronomy even more important. Figure 5.1 is a schematic representation of the manner in which progress is made in astronomy (indeed, in any I~NSTRUME ~ OBSERVAT S I MULAT I ON I _ CONTROL-FEEDBACK / DES I ON / DATA COLLECT I ON / \ TEST / - \ QUICK-LOOK / INTERPRETAT I ON \ / PROCESS I NG S I MULAT I ON \ / CORRECT I ON MODEL I NG \ / CAL I BRAT I ON W<\\ //>~ \ \ ~ / / i\ ~ / / \~ \ / an/ \ / a\ \ / /~ \ | CATALOGS \ \ / / ARCHIVES STORAGE \ \ / / STORAGE CROSS REFERENCE \ ~ / I ND I CES \ ACCESS /` ye ye by\ ACCESS / ANALYS I S I NFO. EXTRACT I ON | MODEL F I TT I NG / FIGURE 5.1 Schematic representation of progress in astronomy.
304 science). As one traverses the diagram in the direction of the arrows, new knowledge of the Universe is obtained. At the same time, additional questions are raised that require further traversals of the diagram before they can be answered. Observations are performed, processed, analyzed, and interpreted, resulting in suggestions for new instruments and new observations, which in turn lead to another round of processing, analysis interpretaton, and so on. Because astronomy is an observational science and because it deals with large numbers of diverse objects, archives and catalogs have been especially important for astronomy (perhaps more so than for many other sciences) and will continue to be important in the future. Included with each major task in Figure 5.1 are thos e functions in which computers play a prominent role. With- out computers it would be impossible to design, build, test, or operate the high-technology instruments and detectors that are in use today and contemplated for the future. Digital data from high-quantum-efficiency detectors could not be collected, processed, analyzed, archived, or reduced to catalogs. Extensive numerical simulations required to evaluate simple physical laws in complex astrophysical situations could not be performed. In short, computational facilities are an integral part of astronomy as we know it today. Progress in astronomy depends not only on the development of more powerful instruments and detectors but also on the development of the capability to digest and interpret the large quantity of high-quality data generated by these modern instruments and detectors. This Panel was primarily concerned with the computa- tional facilities required for the processing, analysis, interpretation, archiving, and cataloging functions shown in Figure 5.1. Although we recognized the importance of computers in the design, construction, test, and operation of instruments and detectors, computers used for such pur- poses are properly considered part of the instrument rather than as separate computational facilities. The costs of computational facilities are modest compared with the costs of a major ground-based or earth- orbiting telescope or a planetary mission. This fact is both a curse and a blessing. A curse, because the re- quirements for computational facilities to support a tele- scope or space mission are often ignored or put on the "back burners with the idea that "if we get the data, we'll figure out some way to process and analyze them."