Scientific American, Volume 247, Number 1, July 1982, pp. 40-51

The Space Telescope

The largest astronomical telescope designed to operate beyond the interfering effects of the earth's atmosphere is scheduled to be transported into orbit by the U.S. space shuttle in 1985

by John N. Bahcall and Lyman Spitzer, Jr.

The earth's atmosphere is an imperfect window on the universe. Electromagnetic waves in the optical part of the spectrum (that is, waves longer than X rays and shorter than radio waves) penetrate to the surface of the earth only in a few narrow spectral bands. The widest of the transmitted bands corresponds roughly to the colors of visible light; waves in the flanking ultraviolet and infrared regions of the optical spectrum are almost totally absorbed by the atmosphere. In addition, atmospheric turbulence blurs the images of celestial objects, even when they are viewed through the most powerful ground-based telescopes.

Accordingly the advantages of making astronomical observations from outside the atmosphere have long been recognized. In the past few decades considerable experience has been gained in the remote operation of telescopes that have been carried above most or all of the atmosphere by suborbital rockets, high-altitude balloons and artificial earth satellites. Significant findings have come from these efforts, altering theories of the structure and evolution of the universe.

The next stage in this program of exploration is the Space Telescope, which is scheduled to be put into orbit around the earth by the U.S. space shuttle in 1985. The Space Telescope will be a conventional reflecting telescope with unconventional capabilities. It will be the largest astronomical telescope ever orbited. It will also be the first long-term international scientific facility in space.

The Space Telescope, which is now under construction, is designed as a multi-purpose astronomical observatory. It will have a 2.4-meter (94-inch) primary mirror capable of concentrating electromagnetic radiation in the entire optical part of the spectrum. It will be equipped initially with an assortment of scientific instruments for recording extraordinarily high-resolution astronomical images, for detecting extremely faint objects, for collecting various kinds of spectrographic data and for making very precise measurements of the position of radiant sources in the sky. The observations will be made from an altitude of some 500 kilometers (300 miles), well above the obscuring layers of the atmosphere.

The plans for the Space Telescope have been developed by a large number of scientists and engineers, working for almost a decade under the supervision of the National Aeronautics and Space Administration (NASA). The prime contractors charged with the actual construction are the Perkin-Elmer Corporation (responsible for the telescope itself) and the Lockheed Missiles and Space Company, Inc. (responsible for the supporting spacecraft system and for the integration of the components into a working whole). The total cost of the project is currently estimated at S750 million.

The projected lifetime of the Space Telescope is 15 years, although in principle there is no reason it could not be operated for many decades. An essential element in ensuring such a long lifetime (and in keeping costs within reasonable limits) is the availability of the space shuttle, which not only will deploy the telescope but also will service it on a regular basis. Astronauts from the shuttle will visit the Space Telescope whenever the instruments on board the observatory need maintenance, repair or replacement. At longer intervals (perhaps every five years) the entire Space Telescope will be returned to the earth by the shuttle for refurbishment of the mirror and other major components. The telescope will then be returned to orbit.

With suitable instrumentation the Space Telescope should be able to respond to electromagnetic waves ranging in length from about 115 nanometers (billionths of a meter) in the far-ultraviolet region of the spectrum to about a million nanometers (or one millimeter) in the far-infrared. Thus the spectral band accessible to the telescope could extend over a range of wavelengths that differ by a factor of 10,000. In contrast, ground-based telescopes have a clear view of colors that range in wavelength from about 300 to 1,000 nanometers, a span of less than a factor of 10.

Because the Space Telescope will be immune to the blurring effects of atmospheric turbulence it will be able to obtain much sharper images of celestial objects than ground-based telescopes can, even at the same wavelengths that are observable from the ground. The maximum spatial resolution attainable with the Space Telescope will be on the order of a tenth of an arc-second, most astronomical images made with ground-based instruments have a resolution not much better than an arc-second. The tenfold improvement in resolution will make possible more detailed observations of extended objects. It is also expected to enable astronomers to see stars some seven times farther from the solar system than is now possible.

The observing program for the Space Telescope will be administered for NASA by the Association of Universities for Research in Astronomy (AURA), a consortium of 17 universities organized originally to operate several facilities for the National Science Foundation, including the national astronomical observatories at Kitt Peak in Arizona and at Cerro Tololo in Chile. The center for the initial processing and analysis of data from the telescope will be the Space Telescope Science Institute, a new facility that is being established by AURA on the campus of Johns Hopkins University. The first director of the institute is Riccardo Giacconi, who led the scientific teams for the highly successful Uhuru and Einstein X-ray satellites. The operation of the Space Telescope will be the joint responsibility of the institute and of NASA's Goddard Space Flight Center in Greenbelt, Md. The Goddard center will have direct control of the satellite and will serve as the collection point for the data transmitted back to the earth.

The European Space Agency (ESA) is covering approximately 15 percent of the cost of the Space Telescope and will have an independent data-analysis center at the headquarters of the European Southern Observatory in Munich. The ESA is supplying the solar panels for powering the observatory, a high-resolution faint-object camera for the instrument section and a number of scientists and technicians for the staff of the Space Telescope Science Institute. In return European astronomers will be entitled to 15 percent of the observing time. Astronomers from other parts of the world will also work with the telescope, making it a truly international observatory.

PRIMARY MIRROR for the Space Telescope was photographed at the Wilton, Conn., plant of the Perkin-Elmer Corporation just after its front surface had been coated with a reflective film of aluminum three millionths of an inch thick, followed by a protective layer of magnesium fluoride a millionth of an inch thick. The mirror, which is made of fused silica glass with an extremely low coefficient of thermal expansion, is 2.4 meters (94 inches) in diameter and weighs about 1,800 pounds. It consists of a lightweight cellular core approximately 10 inches thick sandwiched between two endplates, each about an inch thick. Some 200 pounds of material were removed from the front plate in the course of the 28 months of grinding and polishing required to give the surface its proper figure, which is that of a concave hyperboloid. The masked man seen enlarged in reflection is standing next to the photographer some 60 feet from the mirror. Another technician, also wearing a mask and a special suit to maintain the cleanliness of the mirror's surface, is at the right. A metal plate temporarily covers the hole in the center of the mirror through which light from the telescope's secondary mirror will pass.

The first astronomical observations from space were made in the late 1940's with captured German V-2 rockets. Some of these early liquid-fuel rockets were brought to the U.S. after World War II and were used to send various scientific instruments far above the atmosphere for several minutes of observation. Smaller solid-fuel rockets were later developed specifically for scientific research; they typically lifted a payload of about 100 pounds to a maximum altitude of 100 miles, giving an observation time of a few minutes above the obscuring layers of the atmosphere. The subsequent development of lightweight, solid-state electronic devices has made it possible to build increasingly complex and capable scientific instruments for such missions without prohibitive increases in the power needed to lift them.

The first application of the high-altitude technology was to the study of the sun. In 1946 a rocket-launched spectrometer developed by workers at the U.S. Naval Observatory obtained the first ultraviolet solar spectrogram, revealing absorption features not previously detected in the radiation from any celestial object. It was not until 1957 that ultraviolet radiation from a star was recorded. The spectrographic resolution of this early measurement was quite coarse, with a measuring bandwidth of several tens of nanometers. The early rockets could not be aimed accurately; they rotated freely in space and so could not give the long exposure needed for a precise measurement of the faint radiation from a distant star. In the 1960's techniques were developed for pointing rocket-borne instruments at a star, utilizing small gyroscopes to provide an inertial reference system. As a result stellar spectrograms were recorded with a measuring bandwidth of about a tenth of a nanometer. This achievement marked the beginning of active research on many aspects of stellar atmospheres and interstellar matter.

Meanwhile another group of astronomers employed balloons to lift optical telescopes to altitudes of about 20 miles, above the densest part of the atmosphere. In the late 1950's a 12-inch telescope of this type, named Stratoscope I, obtained extraordinarily sharp pictures of the sun. In the following decade its successor, the 36-inch Stratoscope II, made several photographs of planets and star systems with a resolution close to a tenth of an arc-second.

An artificial earth satellite, which can operate in orbit for years, offers a much better platform for mounting an optical telescope than either a suborbital rocket or a balloon. As aerospace technology has progressed satellites have become the primary vehicles for extraterrestrial astronomy.

With satellites as with the earlier rockets and balloons the first observations made were of the sun. The process of finding an object in the sky and pointing a telescope at it is much easier with the sun than with a more distant star. Beginning in the 1960's NASA built and operated a series of Orbiting Solar Observatories, equipped with various instruments for studying the solar atmosphere.

The first NASA satellites designed for stellar observations were named the Orbiting Astronomical Observatories. Two satellites of this type were operated successfully, one from 1968 to 1973 and the other from 1972 to 1981. Both of them were used mainly for analyzing ultraviolet radiation from stars. The first one had a fairly low spectrographic resolution: its measuring bandwidth was 1.2 nanometers. The second, named Copernicus, was far superior in this respect: its measuring bandwidth was .005 nanometer. The development of precise guidance systems for such satellites was a major technological achievement. The Copernicus telescope, which had a mirror 32 inches in diameter, could stay pointed toward a star for several minutes with a maximum deviation of about .02 arc-second.

The two Orbiting Astronomical Observatories yielded a wealth of data. For example, observations made with the Copernicus satellite showed that much of the hydrogen in interstellar clouds is in the form of molecules rather than individual atoms.

COLOR-CODED TOPOGRAPHIC MAPS of the surface of the primary mirror were made on the screen of a computer-graphics terminal as an aid in determining the corrective action needed for each of the 24 cycles in the final, eight-month polishing process. The maps were based on precise interferometric measurements of the shape of the surface. The two maps shown were made at the start and the finish of the computer-controlled polishing process. The white areas represent the average surface of the mirror; the dark blue and dark red areas correspond respectively to highs and lows. At the start deviations from the prescribed shape were as great as 100 millionths of an inch; at the finish the maximum deviation was less than a millionth of an inch over most of the surface. The finished primary is the finest large astronomical mirror ever made. According to Perkin-Elmer, ``so nearly perfect is the surface that if the mirror were scaled up to the width of the continental United States, no hill or valley would depart from the mean surface by more than about 2 1/2 inches.''

Moreover, many oxygen atoms in the regions between the clouds were found to be highly ionized, indicating that the gas between the clouds is very hot: on the order of a million degrees Kelvin. The satellite data also showed that the cosmic ratio of atoms of deuterium, or heavy hydrogen, to atoms of ordinary hydrogen is about one to 100,000. According to certain cosmological theories, this measurement supports the view that the universe will continue to expand forever.

The most recent optical telescope in space is the International Ultraviolet Explorer, a satellite developed jointly by NASA, the ESA and the British Science Research Council; it has been measuring the ultraviolet spectrum of comparatively faint objects since 1978. Although the performance of this instrument is limited by the size of its mirror (which is 18 inches in diameter), it has been particularly effective in obtaining ultraviolet spectrograms of galactic nuclei and in analyzing the interstellar gas in remote parts of our galaxy.

The concept of a much larger space telescope has evolved slowly over the past two decades. The first official notice of such a project appeared in 1962 in the report of a group of scientists organized for NASA by the National Academy of Sciences to study the future of space science. The group recommended the development of a large space telescope as a logical long-range goal of the U.S. space-science program. The recommendation was repeated by a similar study group in 1965. Soon afterward the National Academy established a committee chaired by one of us (Spitzer) to define the scientific objectives of a proposed space telescope with an aperture of approximately three meters. The report of this group was issued in 1969. In spite of the many advantages cited for such a large space telescope, most astronomers were simply too busy at the time to take an active part in promoting its development. Ground-based astronomy had entered an exciting ``olden era'' with the discovery of phenomena such as quasars, the cosmic microwave background radiation and pulsating neutron stars, and few people were prepared to devote the many years of effort needed to develop a facility as complex and costly as a large space telescope.

In 1972 another committee of the National Academy of Sciences, chaired by Jesse L. Greenstein of the California Institute of Technology, reviewed the needs and priorities of astronomy in the 1970's and again drew attention to the capabilities of a large space telescope. Although the nature and cost of such a device were then only partially defined, it was viewed as a realistic and desirable long-range goal.

Meanwhile NASA had assembled a small group of astronomers under the direction of Nancy G. Roman to provide scientific guidance for the space-telescope feasibility studies then being done at Goddard and at the George C. Marshall Space Flight Center in Huntsville, Ala. Representatives of academic institutions, NASA research centers and industrial contractors assisted in the initial effort.

In 1973 NASA selected a group of scientists from several academic institutions to help establish the basic design of the telescope and its instruments. The group worked with NASA scientists and engineers to determine what objectives for the telescope were feasible and which of them should be given priority. The main scientific guidance was provided by a 12-member working group (on which both of us served) chaired by C. R. O'Dell of the University of Chicago. In order to head the scientific effort for the still unfunded Space Telescope project O'Dell left his positions as professor and chairman of the astronomy department at Chicago and as director of the Yerkes Observatory.

In 1977 NASA selected a new group of 60 scientists from 38 institutions to participate in the design and development of the proposed observatory. The scientific direction of this effort is again guided by a science working group headed by O'Dell; the current membership of the working group includes key NASA employees, the principal investigators responsible for the initial scientific instruments, several interdisciplinary scientists (including Bahcall) and specialists in data handling, spacecraft operations and telescope optics.

ATMOSPHERIC ABSORPTION of electromagnetic radiation limits ground-based optical astronomy primarily to the narrow spectral band corresponding to visible light. Radiation in the flanking ultraviolet and infrared regions is almost totally blocked. The upper edge of the gray areas indicates the boundary where the intensity of the radiation at each wavelength is reduced to half its original value. A nanometer is a billionth of a meter, or 10 angstrom units.

The Space Telescope program almost didn't happen. Between 1974 and 1978 the project was repeatedly in danger of being canceled or postponed indefinitely as a result of congressional and executive budgetary reviews. After an intensive lobbying effort, joined not only by hundreds of astronomers but also by many interested scientists in other fields, construction was finally authorized in 1977. The program survived its first appropriations test in Congress in 1978, and since then it has consistently met with a sympathetic and informed response on Capitol Hill.

SPACE SHUTTLE will carry the Space Telescope to an altitude of approximately 500 kilometers (300 miles) and then release it into orbit with the aid of a mechanical arm The solar-power panels, communications antennas and aperture door, which will be stowed while the satellite is being carried in the shuttle's cargo bay, will be deployed by the satellite after its release. The telescope will be visited by the shuttle for maintenance, repair and replacement of parts. Every five years or so the entire satellite will be returned to the earth for refurbishment.

By the time the Space Telescope was formally approved detailed NASA studies had led to a comprehensive design, which is being followed for the most part in the actual construction of the observatory. The telescope itself consists of two hyperboloidal reflecting surfaces: the 94-inch concave primary mirror and a much smaller convex secondary mirror mounted about 16 feet in front of the primary. Light striking the primary mirror is reflected to the secondary, where it is directed through a hole in the center of the primary; the image comes to a focus several feet be hind the primary. The telescope is described as a Ritchey-Chretien type of Cassegrain optical system.

The scientific instruments that detect and measure the radiation concentrated in the focal plane are installed in an array of boxes mounted behind the primary mirror. Four of the boxes are aligned parallel to the optical axis of the telescope and four are arranged radially around the axis. Of the four radial boxes three house the telescope's fine-guidance system. The tube of the telescope extends more than 10 feet in front of the secondary in order to shield the optical system from stray light, most of which is direct light from the sun and scattered sunlight from the earth and the moon. A system of internal baffles provides additional shielding. Electronic equipment and other devices are housed in a toroidal section surrounding the telescope tube at its base. Two panels of solar cells for powering the equipment and two dish-shaped radio antennas for communicating with the earth extend from the midsection. The cylindrical body of the satellite is about 42 feet long and 14 feet in diameter.

The most remarkable feature of the Space Telescope will be the unprecedented quality of the images formed at its focal plane. The optical surfaces will be as nearly perfect as modern technology can make them: the average deviation of the two reflecting surfaces from their ideal contour will not exceed 10 nanometers. To avoid thermal distortions the mirrors are made of fused silica glass with an extremely low coefficient of thermal expansion. In addition they will be maintained thermostatically at a nearly constant temperature while they are in space. The position of the two mirrors with respect to each other and to the focal surface will be adjustable by remote control to yield the sharpest images possible. The fine-guidance system, which will take a fix on stellar images in the outer part of the telescope's field of view, is expected to be able to hold the optical axis steady to within .01 arc-second for as long as 10 hours. (Internal reaction wheels will serve to aim the telescope and hold it steady; commanding such a wheel to rotate faster in one direction will cause the entire telescope to turn in the opposite direction.)

Six major scientific instruments are scheduled to be included in the Space Telescope's instrument section from the time it is launched through its first few years of operation. The first five are called the wide-field/planetary camera, the faint-object camera, the faint-object spectrograph, the high-resolution spectrograph and the high-speed photometer. In addition the fine-guidance system will give the telescope an astrometric capability, that is, an ability to measure the exact position of stars. Although the two mirrors will have a high reflection efficiency for radiation at all wavelengths in the optical region of the spectrum, no infrared-sensitive instrument will be included in the initial stage. Nevertheless, all aspects of the observatory are planned to be consistent with the possible future inclusion of an instrument sensitive to radiation with wavelengths as long as a millimeter.

The entrance apertures of the four axially mounted instruments are at the focal plane of the telescope. There the total field of view, which measures 28 arc-minutes in angular units, is almost half a meter in linear diameter; the resulting scale of the image at the focal plane is 3.58 arc-seconds per millimeter. With suitable pointing commands the image of any object in the field of view can be directed toward any one of the four axial instruments or toward the fifth, radially mounted one. Each instrument is designed so that it can be removed in orbit and a new instrument installed in its place by a space-suited astronaut operating from the space shuttle.

An on-board computer, external to the scientific instruments, will control the operation of the observatory and handle the flow of data. The computer will be reprogrammable, making it possible to modify the procedures as experience is gained with the instruments. Astronomers and spacecraft controllers will communicate with the Space Telescope by means of the NASA Tracking and Data Relay Satellite System. All data will be relayed back to the earth through this system also, for delivery ultimately to the Space Telescope Science Institute.

The principal investigators responsible for developing the initial set of instruments were chosen after intense competition. By the time the satellite is launched each of these investigators and his colleagues will have spent more than eight years building a general-purpose instrument for the potential use of all astronomers. In recognition of this effort each principal investigator and his team will be awarded more than a month of observing time.

INTERNAL COMPONENTS are drawn in black and external components in color in this overall perspective view of the Space Telescope in its deployed configuration. The cylindrical body of the satellite is approximately 42 feet long and 14 feet in diameter. The scientific instruments are designed so that they can be replaced in orbit by a space-suited astronaut operating from the space shuttle.

The principal investigator for the wide-field/planetary camera is James A. Westphal of Cal Tech. This instrument, as its name suggests, can be operated in either of two modes: as a wide-field camera or as a higher-resolution camera suitable for, among other things, planetary observations. In each mode the detection system consists of four charge-coupled devices (CCD's): microelectronic silicon ``chips'' that convert a pattern of incident light into a sequence of electrical signals. Each chip is a square measuring almost half an inch on a side and is subdivided into an array of pixels, or individual picture elements, with 800 pixels on a side. A single chip therefore has a total of 640,000 pixels, and the four part mosaic image formed by a set of four CCD's has more than 2.5 million pixels. Each pixel yields an electrical signal proportional to the number of photons, or quanta of electromagnetic radiation, reaching it during an exposure.

The wide field/planetary camera is mounted on the side of the telescope that will generally be kept away from the sun. Incoming light passing along the optical axis of the telescope is directed outward at a right angle by means of a flat ``pick-off'' mirror held by a rigid arm at a 45-degree angle to the optical axis. The diagonal mirror diverts only the central part of the incoming beam; the rest of the light passes around the mirror to the other instruments.

In the wide field mode the camera has a square field of view 2.67 arc-minutes on a side, the largest field of any of the instruments. Each pixel in this mode subtends an angle of .1 arc-second. In a sense the wide field camera compromises the angular resolution of the telescope in order to provide a field of view large enough for the study of extended sources such as planetary nebulas, galaxies and clusters of galaxies. Even so, the field of view is much smaller than the field that can be recorded on a photographic plate by a ground-based telescope. In the Space Telescope the field is limited by the size of the microelectronic detectors available for remotely acquiring, storing and digitizing pictures. The CCD's for the wide field/planetary camera, which are being supplied by Texas Instruments, Inc., have more pixels than any other CCD's used for astronomical purposes.

In the planetary mode the square field of view of the camera covers an area of the sky about a fifth as large as it does in the wide field mode; the field in the planetary mode measures 68.7 arc-seconds on a side, and an individual pixel subtends an angle of .043 arc-second. The planetary camera takes advantage of almost the full resolution of the optical system while providing a field of view that is more than adequate for full disk images of the planets. The high sensitivity of the CCD detection system makes possible the short exposure time required for certain planetary observations. The planetary mode will also be employed by many observers for high resolution studies of extended galactic and extragalactic objects.

The wide field/planetary camera is unique among the Space Telescope's instruments in several respects. It will gather by far the greatest number of bits of information: more than 30 million bits per picture. The spectral response of the detector will also be the widest available with any of the telescope's instruments: the camera will be sensitive to wavelengths ranging from 115 nanometers in the far-ultraviolet region to 1,100 nanometers in the near infrared. The wide spectral coverage is made possible by coating the CCD's with an organic phosphor, called Coronene, that converts photons of ultraviolet radiation into photons of visible light, which the silicon sensors can detect. The excellent response at the red end of the visible band is attributable to the natural sensitivity of the CCD's.

The CCD's used in both the wide field mode and the planetary mode have a low level of background electrical ``noise'' and hence are well suited for making pictures of faint sources. Part of the noise in such a device is thermal, and it will be reduced by cooling the detector elements thermoelectrically to about -95 degrees Celsius. The heat generated by the cooling system will be dissipated by a radiator that will form part of the outside surface of the satellite.

The incoming light to the instrument can be directed onto either the four CCD's of the wide field camera or the four CCD's of the planetary camera by means of a pyramidal mirror that can be rotated by 45 degrees about its axis, thereby allowing two essentially independent optical systems to be housed in one instrument compartment. Any of 48 filters can be inserted into the optical path. Thus the wide field/planetary camera is an extremely versatile instrument that will serve a broad range of astronomical purposes. We shall mention here just two of the many investigations that will be undertaken with this instrument.

The camera will be employed in both modes to make a series of images of certain nearby stars to see if they have planetary companions. The 10 or so stars selected for the study have been chosen because they all have a large proper motion (that is, motion across the sky). If any of the stars does have a planetary system, it may be possible, given the extraordinary resolution and accurate guidance of the Space Telescope, to detect periodic ``wobbles'' in the path of the star caused by the gravitational attraction of an unseen companion. The measurements are difficult ones, but the Space Telescope may finally resolve the long standing question of whether there are planetary systems similar to the solar system among the nearby stars.

OPTICAL PATH in the Space Telescope is said to be folded: light from the concave primary mirror is reflected from the convex secondary mirror and passes through a hole in the center of the primary before coming to a focus at the image plane in the instrument section several feet behind the primary. Technically the telescope is described as a Ritchey-Chrétien type of Cassegrain optical system.

Quasars are the most distant and the most energetic objects known in the universe. Each of these compact sources emits on the order of 100 times as much energy as a bright galaxy made up of 10 billion stars. Several competing theories have been put forward to explain how a quasar produces such an enormous amount of energy in such a small space, but some crucial observational tests required to settle the matter are not feasible with ground-based instruments. Some of the theories are based on the idea that quasars are ``sick'' galaxies; in other words, the quasars are supposed to represent a transient, disease-like stage in the evolution of an otherwise normal galaxy. To test these theories high-resolution images of quasars will be obtained with the wide-field/planetary camera to determine whether the bright objects that appear as point sources from the earth are surrounded by the fainter, more diffuse light of a galaxy. It should even be possible to tell whether the quasar stage is a disease of young galaxies or of old ones. This fundamental question is currently unanswerable because of the fuzziness of the images obtained with ground-based instruments.

The faint-object camera that will be supplied by the ESA is one of the four axially mounted instruments. The primary purpose of this second camera is to exploit the full optical power of the Space Telescope. It will detect the faintest objects visible with the telescope and will record images having the highest angular resolution attainable with the optical system. The project scientist for the faint-object camera is F. Macchetto of the ESA.

The faint-object camera is complementary in several ways to the wide- field/planetary camera. The faint-object camera will have a higher spatial resolution, whereas the wide-field/planetary camera will have a larger field of view. In the spectral region between 120 and 400 nanometers the faint-object camera will acquire an image more rapidly than the wide-field/planetary camera will. In the longer-wavelength, redward regions of the spectrum, however, the wide-field/planetary camera will be faster. In addition to forming images the faint-object camera will be able to determine the polarization of the detected radiation and to make spectroscopic measurements of both point objects and extended objects. The two cameras are not redundant, but they are designed to be sufficiently similar in function to ensure that an operable camera of some kind will be among the initial instruments even if a camera were to fail in orbit.

In the faint-object camera two similar but independent optical systems are provided to form an image of a point source. One system has a very small, square field of view, measuring 11 arc-seconds on a side; it has a pixel size of only .022 arc-second. The other system has a square field of view 22 arc-seconds on a side and a pixel size of .044 arc-second. In each system the detector consists of an image-intensifying device similar to the light-sensitive cathode-ray tube in a television camera. Unlike the CCD's in the wide-field/planetary camera, a detector of this kind counts individual photons.

INCOMING LIGHT is routed in different directions by an array of small ``pick-off'' mirrors positioned near the center of the Space Telescope's scientific-instrument section behind the primary mirror. The diamond-shaped flat mirror mounted diagonally on the optical axis directs light outward to the radially mounted wide-field/planetary camera. The three arc-shaped flat mirrors arranged around the outside of the incoming beam send light to the three fine-guidance sensors, which are also radially mounted. The light that bypasses these four mirrors comes to a focus at an image plane at the entrance apertures near at the front of the four axially mounted instrument boxes. The projections of the pick-off mirrors on this focal plane are shown in dark gray in the plan view at the bottom. Because the incoming beam is interrupted by the pick-off mirrors well in advance of the focal plane the areas blocked by the mirrors are slightly enlarged; the additional vignetted zones are represented by the light gray bands outlining the projected mirror zones. At the focal plane the field of view is 28 arc-minutes in angular diameter. The wide-field/planetary camera views a square region about three arc-minutes on a side in the center of the field. The remainder of the field out to a radius of about nine arc-minutes is divided into quadrants, each of which is viewed by one of the four axially mounted instruments. The outermost part of the field, roughly between nine and 14 arcminutes from the optical axis, is sampled by the fine-guidance system, which is designed not only to point the telescope but also to make precise measurements of the position of stars.

The faint-object camera is designed so that each point-source image produced by the telescope is sampled by several pixels. Hence it will be the instrument of choice when the highest possible resolution and the maximum contrast against the background sky are required. The camera will also be able to carry out spectroscopic and polarimetric studies of comparatively faint objects. In addition the camera will be able to view extremely narrow fields with an even smaller pixel size (approximately .007 arc-second).

The scientific tasks of the wide-field/ planetary camera and the faint-object camera are expected to overlap. Depending on the specific resolution, field of view and spectral region required, an observer may choose to work with one camera or the other. We shall mention here only one type of observation for which the faint-object camera should be particularly suited.

Globular clusters are spherical collections of millions of stars that can be seen from the ground on a clear night with a small telescope or even with binoculars. Because all the stars in a cluster are at approximately the same distance from the solar system one can test theoretical models of stellar evolution simply by counting the stars of various types in a cluster. The standard theory predicts that each globular cluster should include between about 10,000 and 100,000 of the stars called white dwarfs. These compact objects represent the last stage in the evolution of stars that have exhausted their nuclear fuels, cooled and collapsed. Because white dwarfs are very faint they cannot be seen at the great distances of the globular clusters with ground-based instruments. The Space Telescope's faint-object camera, however, should be able to detect many white dwarfs in globular clusters. By studying their properties it will be possible to learn much more about the evolution of stars.

The Space Telescope will have two spectrographs: optical devices that divide the incoming light from an astronomical source into separate beams according to wavelength. In spectroscopy resolution is usually defined as the ratio of the wavelength of the incoming light to the smallest separation that can be measured between two wavelengths. One of the two spectrographs on board the observatory, the faint-object spectrograph, will be able to observe faint stellar objects with a spectrographic resolution of 1,000 (equivalent to a measuring bandwidth of 1/1,000th of the wavelength). The principal investigator for this instrument is Richard J. Harms of the University of California at San Diego.

The faint-object spectrograph will be equipped with two systems of detectors. Both detectors are devices called Digicons; one is sensitive to red light and the other to blue light and ultraviolet radiation. A Digicon sensor operates on the basis of the photoelectric effect. The incoming light is spread out according to wavelength by a diffraction grating and strikes the surface of a thin photocathode layer deposited on a transparent plate. Light of a particular wavelength reaches a particular position along the photocathode, producing a spray of free electrons known as photoelectrons. A magnetic field focuses the photoelectrons at a point whose position depends on where they emerge from the photocathode and hence on the wavelength of the incident light. The photoelectrons are collected by a linear array of 512 diodes, each of which records the intensity of the incident light at a particular wavelength.

The faint-object spectrograph will be sensitive to radiation ranging in wavelength from about 115 to 800 nanometers. In addition the instrument will have two special features: it will be able to measure the polarization of the incoming light and to detect extremely fast variations (perhaps as brief as a few milliseconds) in the spectrum of radiation emitted by bright sources. Because the investigation of many astronomical problems depends on the spectral analysis of the radiation from extremely faint objects, this instrument is expected to be one of the busiest on the Space Telescope. By measuring the spectra of very distant quasars, for example, it should be possible to study the properties of the universe more than 10 billion years ago, perhaps 85 percent of the way back to the beginning of time (if, as the standard big-bang model of cosmology assumes, time actually had a beginning). Spectrograms of the most distant quasars are expected to indicate the chemical constitution of matter at that early stage in the evolution of the universe.

WlDE-FlELD/PLANETARY CAMERA is one of the instruments scheduled to be included in the Space Telescope during its first few years of operation The camera is designed to operate in either of two modes. In each mode the detection system consists of a rectangular array of four light-sensitive silicon ``chips'' called charge-coupled devices (CCD's). The incoming light reflected into the radially mounted instrument compartment by the diagonal pick-off mirror can be directed onto either the four CCD's of the wide-field camera or the four CCD's of the higher-resolution planetary camera by means of a pyramidal mirror that can be rotated by 45 degrees about its axis. Any of 48 filters can be inserted into the optical path. The external radiator serves to dissipate the heat generated by the cooling system associated with the detectors.

The investigation of some astronomical questions requires a higher spectrographic resolution than can be attained with the faint-object spectrograph, because the width of many emission and absorption features is narrower than the measuring bandwidth of the instrument. The high-resolution spectrograph will meet this need. Under normal operating conditions it will have a spectrographic resolution of 20,000. Narrow spectral features that might not even be detected with the lower-resolution faint-object spectrograph will be accurately measured, yielding information about the physical conditions under which the radiation was emitted. The high-resolution spectrograph will also have an ultrahigh-resolution mode of operation in which the spectrographic resolution will be improved by an additional factor of five to about 100,000. The principal investigator for the high-resolution spectrograph is John C. Brandt of Goddard.

Of course, there is a price to be paid for the higher resolution of this second spectrograph. Dividing the spectrum into a much larger number of bands in order to measure the flux of photons separately in each band has the effect of decreasing the number of photons detected per band. Thus higher resolution results in lower sensitivity, and the larger quantity of information provided by the high-resolution spectrograph can be obtained only for stars that are some 60 times brighter than those that can be studied with the faint-object spectrograph. This difference amounts to about 4.5 stellar magnitudes. For the ultrahigh-resolution mode the difference in brightness is a factor of more than 300, or the equivalent of about six stellar magnitudes.

The high-resolution spectrograph has six interchangeable diffraction gratings, each of which disperses light of different wavelengths in different directions. A camera mirror or grating then forms an image of the spectrum on the photoelectron-emitting surface of a Digicon sensor. By rotating the carousel on which the gratings are mounted, any one of them can be brought into the optical path of the instrument, making it possible to obtain a spectrographic reading at any wavelength between 110 and 320 nanometers.

This spectrograph with its normal resolution should be able to observe stars as faint as the 13th magnitude, or about six stellar magnitudes fainter than those observed by the Copernicus telescope. The gain in sensitivity over the spectrograph on the International Ultraviolet Explorer is not as great-about four magnitudes-but the spectrographic resolution and the photometric accuracy will be significantly better for the instrument on the Space Telescope.

The power of this instrument should open up a number of interesting new lines of inquiry. For example, the high-resolution spectrograph will make possible the study of interstellar gas at places in our galaxy and other galaxies where it cannot now be observed. Preliminary measurements by the International Ultraviolet Explorer have shown that the gas in the galactic ``halo'' between the earth and the nearest neighboring galaxy (one of the two Magellanic clouds) includes carbon atoms that have been stripped of three electrons indicating that the temperature in this region is about 100,000 degrees Kelvin. With the high-resolution spectrograph much more accurate data will be obtainable, perhaps revealing the relation between this gas and the even hotter material detected by Copernicus. Measurements of the way in which the properties of our galaxy vary from place to place will provide much-needed clues to the evolution of the system as a whole.

The high-resolution spectrograph will also be applied to the study of interstellar clouds. Ground-based observations of such clouds are able to detect only a few dark lines in the spectrum, created when the gas of the cloud absorbs radiation from background stars. In many instances each absorption line is split into multiple subfeatures, which can be attributed to separate clouds along the same line of sight. The clouds are moving with somewhat different speeds toward the solar system or away from it, altering the characteristic wavelengths at which they absorb radiation. The splitting of the absorption lines makes it possible to study each cloud separately, provided the spectrographic resolution is high enough. With the high-resolution spectrograph it will be possible to analyze a wide range of ultraviolet absorption features from various atoms and molecules and to determine the physical conditions in each cloud. Our understanding of how such interstellar clouds come together and contract to form stars may depend critically on the results of such studies.

The high-speed photometer, which is being developed by Robert C. Bless and his colleagues at the University of Wisconsin at Madison, is designed to make highly accurate measurements, with an extraordinary temporal resolution, of the intensity of the light from astronomical sources over a wide range of wavelengths. The photometer will be capable of distinguishing events separated in time by only 10 microseconds. Observations of sources that vary over time scales this short are difficult or impossible with ground-based instruments because of fluctuations in the atmosphere.

RANGE OF WAVELENGTHS potentially accessible to the Space Telescope extends from the far-ultraviolet part of the spectrum (left) to the far-infrared (right). For comparison the spectral bands that can be observed with the unaided human eye and with a large ground-based telescope (in this case the 200-inch Hale telescope on Palomar Mountain) under normal observing conditions are also indicated. The vertical scale gives the relative brightness (in terms of stellar magnitude) of the faintest celestial object that can be imaged; an increase of one unit in stellar magnitude corresponds to a decrease in apparent brightness by a factor of about 2.5.

The high-speed photometer is the simplest of the instruments in the initial group on board the Space Telescope. It has no moving parts and relies entirely on the fine pointing of the spacecraft to direct the light from an astronomical target onto one of its 100 or so combinations of spectral filters and entrance apertures. The photometer has four independent, magnetically focused detectors, called image dissectors; they resemble photomultiplier tubes in operation, except that they can be made to respond only to photoelectrons coming from the small region of the photocathode on which the light is focused. Each image dissector is mounted behind a plate that holds an assortment of filters and entrance apertures.

The overall spectral response of the image dissectors extends from about 115 to 650 nanometers. The instrument is also equipped with a red-sensitive photomultiplier tube and a system for measuring the polarization of ultraviolet radiation with the aid of one of the image dissectors.

The high-speed photometer will be capable in principle of detecting the smallest objects observable with any of the instruments on the Space Telescope. The ability to distinguish events that are separated in time by only 10 micro-seconds implies (according to the special theory of relativity) that variations in the light output of a star as small as three kilometers across could be detected. This is an extraordinarily small linear dimension for a star; indeed, it is very close to the diameter the sun would have if it were compressed to such a high density that it formed a black hole. Accordingly, one program scheduled for the high-speed photometer is to search for extremely fast variations in astronomical systems that are suspected of harboring a black hole, in the hope of finding further evidence of these elusive entities. The high-speed photometer will also be used for less exotic observations, including an attempt to identify optically faint objects observed mainly at radio or X-ray wavelengths.

Under the best observing conditions ground-based measurements of the position of any star are limited by the size of the star's blurry ``seeing disk,'' which is generally at least one arc-second in diameter. In determining the angular distance between two stars an uncertainty of about .1 arc-second, or a tenth of the diameter of the stellar image, is typical for a single observation. By averaging many exposures the uncertainty can be reduced to about .01 arc-second. Random errors result in corresponding uncertainties in the determination of a star's parallax. (Parallax is the average angular change in the apparent position of a star resulting from the revolution of the earth about the sun.) The determination of distance beyond the solar system is based largely on measurements of the parallax of comparatively nearby stars. Since the measurement of a stellar image with the Space Telescope will be accurate to within about .002 arc-second, the determination of stellar position, and hence of stellar parallax, should be about five times better than it is with ground-based telescopes. The fivefold improvement in the accuracy of stellar- parallax measurements is of fundamental importance to all of stellar astronomy. For example, knowing the precise distance of certain comparatively young star clusters in our galaxy will enable astronomers to determine the absolute brightness of the stars in the clusters. This knowledge in turn will make it possible to extend the calibrated distance scale, which is based on the comparison of apparent brightness and absolute brightness, to stars that are much farther away.

The Space Telescope has not been equipped with a separate instrument for astrometry because the fine-guidance system will be accurate enough to make the necessary measurements of the angular distance between stars. The leader of the team for astrometry is William H. Jefferys of the University of Texas at Austin.

Observing time on the Space Telescope will be allocated to astronomers from all parts of the world by the Space Telescope Science Institute, which will be responsible for facilitating the most effective scientific use of the powerful new observatory. To provide visiting astronomers with the most efficient operating systems, to assist and advise observers on the optimum use of the various instruments and to help create a stimulating atmosphere for research with the Space Telescope outstanding astronomers from the U.S. and abroad are being recruited to serve on the institute's staff. It is expected that half of their time will be devoted to the diverse tasks of the institute, with the other half available for their own research programs. The new institute will also make recommendations to NASA on broad policy matters pertaining to the Space Telescope. The involvement of outside astronomers in determining the policies of the institute is being ensured through a number of external committees.

The institute will solicit outside proposals for specific observing programs for the Space Telescope. With the aid of peer-review groups the institute will evaluate the proposals and select the most promising programs for inclusion in the telescope's schedule. In many cases the programs selected will be combined with those submitted by the original scientific-instrument teams, by other members of the Space Telescope working group and by the European groups. The final scheduling and the preparation of a complete list of commands for the operating computer will be done by NASA, which will retain responsibility for the day-to-day operation of the observatory.

Astronomers on the staff of the institute will advise outside astronomers on the formulation of observing plans. Other staff astronomers will be responsible for maintaining the calibration of the instruments and for the initial processing of data. Computer specialists will help to develop suitable programs for use by the astronomers in analyzing the data. Finally, the Space Telescope Science Institute will assist astronomers in communicating the results of their studies to other scientists, to NASA, to Congress and to the public.

The Space Telescope will help to solve many outstanding astronomical puzzles. The greatest excitement, however, will come when the pictures returned from the satellite reveal things no one in this generation of astronomers has dreamed of, phenomena that only the next generation will be privileged to understand.

TENFOLD IMPROVEMENT in spatial resolution expected with the Space Telescope will enable astronomers to make more detailed observations of extended objects. In this simulation the picture at the top represents the image of a distant spiral galaxy obtained with the 200-inch Hale telescope and the picture at the bottom represents the corresponding image obtained with the Space Telescope. Actually the picture at the bottom is a digitized version of a photograph of a nearby galaxy made with the 200-inch telescope and the picture at the top is a blurred version of the same image made by defocusing the original by an amount proportional to the difference in the effective resolution obtainable with the two instruments. The simulation was prepared by John L. Tonry of the Institute for Advanced Study.

Back to John Bahcall's Home Page
Address questions and comments about this server to