Desktop Astronomy

There is a growing realization in the astronomical community that the combination of constant or decreasing federal support for astronomy, a continuing increase in the number of young and talented astronomers, and an increase in the expense of many of the tools needed to tackle the most difficult and interesting problems, portends an increasingly destabilizing economic situation, especially for more junior astronomers. An oft-repeated solution for this dilemma is the possibility that a growing number of astronomers will perform research in existing data banks, rather than designing experiments and/or performing observations to acquire new data ab initio. Although this reasoning seems plausible, it is also the case that it is often not eagerly embraced by those faced with the prospect, as "data from the archive" are commonly viewed by potential users as rigidly pre-defined, difficult or impossible to recalibrate to personal preferences, and, worst of all, as having had the cream already skimmed.

We wish to point out that SDSS will almost certainly embody all the best advantages, and suffer among the fewest defects, of this new style of "Desktop Astronomy". Some of the few hardware tools for astronomical research that continually grow cheaper yet more capable with time are fast workstations, large mass storage devices, and high-throughput internet access, precisely the three resources required for use of the SDSS data. We can thus be reasonably assured that the SDSS will remain accessible for use to the entire community, regardless of the uncertain future of federal, philanthropic, or institutional support. The amount of data to be produced by the SDSS is so enormous that conventional data analysis methods will not be possible, and an integral part of our effort is the development of data bases and access tools for the collaboration (see Appendices C and D) which will also be provided to the community. The architecture of the SDSS photometric (imaging) archive is such that virtually no assumptions have been made on the style of future use, not just on which scientific problems will be tackled, but even on such basic issues as algorithms for flux calibration. Essentially all the information read from the CCD at the instant that the exposure is completed is available for all later users, at all later times.

It is of course the case that one basic and unchangeable advance assumption has been made in the spectroscopic portion of the SDSS, namely the selection of which objects to observe. We believe, however, that the international community of users will be more than amply recompensed for this particular lack of flexibility in the form of an entirely new capability never before available in astronomical spectroscopy: huge spectroscopic samples of faint objects, all observed by the identical instrument with high sensitivity, decent spectral resolution and extremely thorough flux calibration. A perfectly homogeneous data bank of 10⁶ galaxy spectra, 10⁵ QSO spectra, 10⁴ unusual stellar spectra, and spectra of 10⁴ ROSAT optical counterparts, all accompanied by fully-calibrated 5-color imaging and astrometric data, is simply an unprecedented capability.

Finally, we make a presumably obvious remark about any data base of 8 Terabytes volume: fears of a small group of workers quickly "skimming the cream" of the most important intellectual content are simply unrealistic. The group who obtained the original Palomar Observatory Sky Survey plates quite correctly realized they had nothing to lose through rapid release of that material, and indeed almost 50 years later, not only are discoveries routinely still made from the POSS, but no one group has become identified with its most successful use.

Whether or not one personally looks forward to the prospect of a larger fraction of the community working in large digital data bases, and regardless of one's assessment of how rapidly these resources might come to predominate, it is probably inevitable that such tools will indeed become more and more common in many or even most astronomers' work. The SDSS data base will almost certainly remain unique for decades, and provide a rich, permanent, and virtually inexhaustible source of research data for any astronomer with access to a few thousand dollars worth of hardware. We thus believe that the positive contribution of SDSS to this new (and some will say inevitable) era of "Desktop Astronomy" -- that of front-line, highest quality, yet universally accessible data, is a most significant one.