Stargazing as done by computers

To entertain his dinner guests at his home in Birr Castle, it was the custom of the third Earl of Rosse to demonstrate his 72-inch telescope, the largest in the world in 1845. The massive wooden tube of the telescope was housed between two large masonry walls just outside the castle and was moved only by an ingenious arrangement of pulleys and weights which were manipulated by workers from the earl's estate. If the cloudy skies of an Irish winter permitted, the earl would direct his telescope to the heavens, then to the accompaniment of his guests' social chatter, the shouts of his workmen, and the creaking and groaning of his wooden telescope, he would laboriously sketch the nebula, many of them seen through his telescope for the first time.

One hundred and thirty-five years later, I study these same nebula (now known to be galaxies) using a computer controlled telescope (of similar size to Lord Rosse's) on an Arizona mountaintop. While my television camera, at the focus of the telescope, integrates the light from the galaxy, stores it in a computer, and displays it on a screen so that I can do my observing in a heated room, I wonder what Lord Rosse would think of modern astronomical techniques. Mechanical genius that he was, Lord Rosse would quickly have understood and appreciated the gadgetry that characterizes modern astronomy. Perhaps the most startling advance he would see would be the fast compact digital computer, which enables the astronomer to control the telescope from a terminal in another room or another building.

The human eye (Lord Rosse's detector) has long been replaced as an astronomical detector by the photographic plate; this, in turn, is now being replaced by the CCD (charge- coupled device) camera. This camera, which has only recently been adapted for astronomical studies, consists of 10,000 tiny detecting elements packed into an area less than one tenth that of my thumbnail; each of the elements detects photons of light with high efficiency and gives a digital output that can be immediately processed by computer. The device can be used either to take a direct image of the sky or to register the output of a spectrograph. In either case, the strength of the CCD lies in its high efficiency, wide dynamic range, and computer compatible output.

Although the final data processing must use the large facilities at central laboratories, it is essential that the observer get some instant feedback from his data so that he can plot the future course of his observations. Hence, most observing programs now require the availability of computers on site. These have the function not only of analyzing the data, but of controlling the instrument, the telescope mount, and perhaps the telescope mirrors. The astronomer thus becomes one step removed from the actual observing and his major function is the operation of a computer terminal.

One of the unique features of modern astronomical research is that it requires the use of very sophisticated equipment in some of the most remote and inaccessible corners of the globe. Driven by limitations of atmospheric "seeing" (essentially, the requirement that the atmosphere above the telescope be stable) optical observatories are almost invariably located on high mountains. To escape the effect of back-scattered city lights, these mountaintops must be as far as possible from civilization. Some of the best sites are on high mountains in the Andes in Chile, on the Baja California peninsula in Mexico, in Hawaii, in Arizona, or in the Canary Islands. None of these remote sites at first sight would seem likely places for modern research.

Since computers have taken over so many of the classical astronomer's observing skills, it has not escaped notice that the most archaic part of this whole process is having the astronomer at the telescope. Since telescopes in space are operated from control rooms at major city laboratories, why not operate ground-based telescopes in the same way? Although, for convenience, the ground-based telescope need not be fully automated; i.e., a technician may still be present at the observatory, the observing program can be completely dictated by an astronomer at his office some thousands of miles away. Given a telephone data link, the astronomer can almost immediately get the data back from the telescope so that he can make real time decisions about his program.

When the Smithsonian Astrophysical Observatory started its Mt. Hopkins Observatory in southern Arizona 13 years ago, it was realized that if there was not some kind of remote observing system, travel costs would be a major factor in research programs. Using Arizona-based technicians under the supervision of a few resident astronomers, this kind of program was first put into effect in satellite tracking and ground-based gamma ray astronomy programs. Later it was expanded to include optical astronomy involving routine measurements of standard stars on a small telescope.

By the end of the 1970s approximately half of the observations at the observatory's 24- inch and 60-inch telescopes were being taken by remote observing. When the Multiple Mirror Telescope (equivalent aperture 176 inches) goes into operation this year, it is expected that it will be operated in the same way.

Although this kind of program really only works well for relatively routine kinds of observations, computer technology has expanded the number of observations that can be classified as routine. It is most suited for programs that involve a large number of similar observations such as surveys to classify a particular kind of object.

An astronomer sitting at a computer terminal in his office in Boston and directing an observing program on an Arizona mountaintop is a far cry from Lord Rosse's after-dinner observing in his "front yard." I cannot help feeling that Lord Rosse would not have been very enthralled by this new method of doing astronomy. This feeling is shared by many of the astronomers who were attracted to the discipline by the romance of observing the skies on clear cold nights from remote mountaintops. A computer terminal in an astronomer's office may be comfortable and efficient, but for many it lacks the thrill of actually making the observations, of finding the object among the myriad of stars, of making minor adjustments to the instrument, of watching the signal grow out of noise in real time, and of feeling the personal responsibility for making the observation. If the astronomer becomes too far removed from the instrument, there is a real danger that he may lose touch with its idiosyncrasies.

However, to dwell on the instruments and methods of the modern astronomer is to miss the real wonder of astronomy in the past decade. Because of technical advances, there has been an explosion in our understanding of the cosmos, an understanding that would have seemed impossible in Lord Rosse's day. Every scientific field has felt this explosion but none more so than astrophysics. One has only to list the new astronomical vocabulary -- quasars, pulsars, neutron stars, bursters, and black holes -- to realize that we live in a far richer and more diverse (but perhaps ultimately more understandable) universe than we thought only a decade ago. Lord Rosse might think that our methods are strange and not very attractive, but he would surely have been excited by the almost weekly catalog of new and unexpected discoveries.